T h e Ox f o r d H a n d b o o k o f

T H E P OL I T IC A L
E C ON OM Y OF
I N T E R NAT IONA L
T R A DE
 The Oxford Handbook of

THE POLITICAL
ECONOMY OF
INTERNATIONAL
TRADE
Edited by
LISA L. MARTIN

1
 3
Oxford University Press is a department of the University of
Oxford. It furthers the University’s objective of excellence in research,
scholarship, and education by publishing worldwide.

Oxford New York
Auckland  Cape Town  Dar es Salaam  Hong Kong  Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto

With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam

Oxford is a registered trademark of Oxford University Press

in the UK and certain other countries.

Published in the United States of America by

Oxford University Press
198 Madison Avenue, New York, NY 10016

© Oxford University Press 2015

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the prior
permission in writing of Oxford University Press, or as expressly permitted by law,
by license, or under terms agreed with the appropriate reproduction rights organization.
Inquiries concerning reproduction outside the scope of the above should be sent to the
Rights Department, Oxford University Press, at the address above.

You must not circulate this work in any other form

and you must impose this same condition on any acquirer.

Library of Congress Cataloging-in-Publication Data

The Oxford handbook of the political economy of international
trade / edited by Lisa L. Martin.
pages cm
Includes bibliographical references.
ISBN  978–0–19–998175–5 (hardback : alk. paper)  1.  International trade.
2.  Commercial policy.  I.  Martin, Lisa L., 1961–
HF1379.O9965 2015
382—dc23
2014039333

1 3 5 7 8 9 6 4 2
Printed in the United States of America
on acid-free paper
 Contents

About the Contributors  ix

1. Introduction  1
Lisa L. Martin

PA RT I   H I STOR IC A L , T H E OR E T IC A L ,
A N D  M E T HOD OL O G IC A L
DE V E L OP M E N T S
2. Explaining the GATT/WTO: Origins and Effects  19
Joanne Gowa
3. The Free Trade Idea  37
Gordon Bannerman
4. Trade Policy Instruments over Time  57
Chad P. Bown
5. Methodological Issues  77
Raymond Hicks

PA RT I I   D OM E ST IC S O C I E T Y
6. Individual Attitudes  99
Jason Kuo and Megumi Naoi
7. Labor and Protectionist Sentiment  119
Erica Owen
8. Domestic Politics and International Disputes  138
B. Peter Rosendorff
vi   Contents

PA RT I I I    I N D U ST RY- L E V E L P ROT E C T ION

9. Industry-Level Protection  159
Lucy M. Goodhart
10. Intra-Industry Trade and Policy Outcomes  177
Timothy M. Peterson and Cameron G. Thies
11. Heterogeneous Firms and Policy Preferences  196
Michael Plouffe
12. The Politics of Market Competition: Trade and Antitrust
in a Global Economy  213
Tim Büthe
13. Connected Channels: MNCs and Production Networks
in Global Trade  233
Walter Hatch, Jennifer Bair, and Günter Heiduk

PA RT I V   D OM E ST IC I N S T I T U T ION S
14. New Democracies  259
Bumba Mukherjee
15. Electoral Systems and Trade  280
Stephanie J. Rickard
16. Authoritarian Regimes  298
Daniel Yuichi Kono
17. Domestic Geography and Policy Pressures  316
Kerry A. Chase

PA RT V   I N T E R NAT IONA L
N E G OT IAT ION S  A N D  I N S T I T U T ION S
18. The Design of Trade Agreements  337
Leslie Johns and Lauren Peritz
19. Deep Integration and Regional Trade Agreements  360
Soo Yeon Kim
 Contents   vii

20. WTO Membership  380

Christina L. Davis and Meredith Wilf
21. Dispute Settlement in the WTO  400
Marc L. Busch and Krzysztof J. Pelc

PA RT V I   I S SU E L I N KAG E S
22. Trade and War  419
Erik Gartzke and Jiakun Jack Zhang
23. Trade and Environment  439
J. Samuel Barkin
24. Bridging the Silos: Trade and Exchange Rates in International
Political Economy  457
Mark S. Copelovitch and Jon C. W. Pevehouse
25. Trade and Development  475
Mark S. Manger and Kenneth C. Shadlen
26. A Match Made in Heaven? The Wedding of Trade
and Human Rights  493
Susan Ariel Aaronson
27. Trade and Migration  518
Margaret E. Peters

Index   535
 About the Contributors

Susan Ariel Aaronson is a research professor in the Elliott School of International

Affairs at George Washington University.
Jennifer Bair is an associate professor of sociology at the University of Colorado,
Boulder.
J. Samuel Barkin is a professor in the Department of Conflict Resolution, Human
Security, and Global Governance at the McCormack Graduate School of Policy and
Global Studies at the University of Massachusetts Boston.
Gordon Bannerman is a tutor at International Correspondence Schools.
Chad P. Bown is lead economist at the World Bank in Washington, DC, and a research
fellow at CEPR in London.
Marc L. Busch is Karl F. Landegger Professor of International Business Diplomacy at the
School of Foreign Service, Georgetown University.
Tim Büthe is an associate professor of political science and public policy, as well as a
senior fellow for the Rethinking Regulation Project at the Kenan Institute for Ethics at
Duke University.
Kerry A. Chase is an associate professor of politics at Brandeis University.
Mark S. Copelovitch is an associate professor of political science and public affairs and
Trice Faculty Scholar at the University of Wisconsin-Madison.
Christina L.  Davis is a professor of politics and international affairs at Princeton
University.
Erik Gartzke is a professor of political science at the University of California, San Diego.
Lucy M. Goodhart is a lecturer at Brandeis University.
Joanne Gowa is William D. Boswell Professor of World Politics of Peace and War at
Princeton University.
Walter Hatch is an associate professor of government and the director of the Oak
Institute for Human Rights at Colby College.
Günter Heiduk is a professor in the East Asian Center at the Warsaw School of
Economics.
x   About the Contributors

Raymond Hicks is a statistical programmer at the Woodrow Wilson School of

International Affairs, Princeton University.
Leslie Johns is an assistant professor of political science at the University of California,
Los Angeles.
Soo Yeon Kim is an associate professor of Political Science at the National University of
Singapore.
Daniel Yuichi Kono is an associate professor of political science at the University of
California, Davis.
Jason Kuo is a PhD candidate in the Department of Political Science at the University of
California, San Diego.
Mark S. Manger is an associate professor of political economy and global affairs in the
Munk School at the University of Toronto.
Lisa L. Martin is a professor of political science at the University of Wisconsin–Madison.
Bumba Mukherjee is an associate professor of political science at the Pennsylvania State
University.
Megumi Naoi is an associate professor of political science at the University of California,
San Diego.
Erica Owen is an assistant professor of political science at Texas A&M University.
Krzysztof J.  Pelc is an associate professor of political science and William Dawson
Scholar at McGill University.
Lauren Peritz is a PhD candidate in the Department of Political Science at the University
of California, Los Angeles.
Margaret E. Peters is an assistant professor of political science at Yale University.
Timothy M. Peterson is an assistant professor of political science at the University of
South Carolina.
Jon C. W. Pevehouse is a professor of political science at the University of Wisconsin–
Madison.
Michael Plouffe is an assistant professor of international political economy in the
Department of Political Science at University College London’s School of Public Policy.
Stephanie J. Rickard is an associate professor in the Department of Government at the
London School of Economics.
B. Peter Rosendorff is a professor of politics at New York University.
Kenneth C.  Shadlen is a professor in development studies at the London School of
Economics.
 About the Contributors    xi

Cameron G. Thies is a professor and the director of the School of Politics and Global
Studies at Arizona State University.
Meredith Wilf is an assistant professor in the Graduate School of Public and
International Affairs at the University of Pittsburgh.
Jiakun Jack Zhang is a PhD candidate in the Department of Political Science at the
University of California, San Diego.
 Chapter 1

Introdu c t i on

Li sa L. Martin

Introduction

International trade offers opportunities and risks to countries, groups, and individu-
als. Opening an economy up to trade with other economies provides potential advan-
tages for all, as it allows for specialization that leads to higher levels of productivity and
efficiency, enlarges the markets available to firms, and increases the scope for consumer
choice. However, the process of specialization entails redeployment of assets that is
costly for some actors, and exposure to the global economy can create new sources of
risk. For these reasons, trade is highly politicized. No government in the modern state
system has ever pursued an entirely hands-off approach to trade policy.
This volume features chapters that view the politics of international trade from a wide
variety of angles. It considers the concepts that have driven trade policy, the interests
that compete over it, and the institutions that channel these interests. The actors that
care about trade policy range from individuals and firms to interest groups, government
agencies, and international institutions. Economists have made great strides in under-
standing the factors that create a demand for international trade and the economic con-
sequences of trade for various actors. Many of the approaches studied in this volume
build on such economic models. However, the focus here is very much on politics: who
wins and who loses from different policies, who is able to organize and to influence gov-
ernments, and thus how politics influences actual patterns of trade.
In this introductory chapter I provide some historical and conceptual background
to the more advanced theoretical perspectives and sophisticated empirical work show-
cased in the rest of the volume. The next section looks at the progression of eras of trade
policy, from the mercantilism that emerged along with the modern state system to
today’s landscape of global, regional, and bilateral trade institutions. The third section
turns to theories, providing an overview of economic models of trade that help us under-
stand the interests that go into making trade policy and the domestic and international
2   Introduction

institutions that aggregate these interests. Finally, I briefly discuss the organization of
this volume.

Historical Overview

In a very broad-brush sense, the history of the politics of international trade is about
movement from extensive government intervention and control of trade, to an appre-
ciation of the potential value of freer trade, to a search for mechanisms that liberalize
trade while providing some basic protections against the pressures of international
competition. However, within this big picture of long-term trends there is a tremen-
dous amount of variation. Pressures for protection ebb and flow with business cycles,
political trends, and new ideas. Some countries now have only modest impediments to
trade, while others continue with more protectionist policies. Some industries lobby for
more openness to trade, while others demand protection. This volume is about explain-
ing such variations. This section highlights some of the major shifts in general trends in
trade policy over time.
Prior to 1500, trade between economic and political units was rare, costly, and thus
concentrated in luxury goods. However, as European states began conquering and set-
tling the rest of the world at the turn of the sixteenth century, regular patterns of trade
emerged within Europe and between European countries and their colonies. European
governments engaged in colonization largely because of their desire to assure access to
natural resources and markets in the rest of the world. The economic and political doc-
trine that informed this period of expansion was mercantilism. Under mercantilism,
military power and state wealth were understood as opposite sides of the same coin.
As Thomas Hobbes famously proclaimed during this era, “Wealth is power, and power
is wealth” (Viner 1948, 15). During the mercantilist era, which roughly covered the six-
teenth through the eighteenth centuries, imperial governments established monopolies
that controlled trade. Governments regulated and intervened extensively in these new
trading networks, manipulating the terms of trade (the ratio of the cost of imports to
exports) to enrich themselves and their domestic supporters.
Under mercantilism, European monarchies put in place laws and policies that
ensured they gained access to resources from their colonies at artificially low prices.
Trade between colonies and states other than the colonial power was prohibited or
extremely limited. In addition, imperial powers required that their colonies import
manufactured goods only from them, thus constraining supply and driving up the
prices of their exports. This meant that colonists paid high prices for imported manufac-
tured goods and received low prices for the commodities that they exported to Europe;
in exchange, they received military protection from the empire. For example, Thomas
(1965) estimated that English restrictions on trade with the thirteen North American
colonies cost the colonists approximately $2,255,000 per year, primarily through the
artificially low prices the colonies received for exports of tobacco and rice to Britain.
 Lisa L. Martin    3

During the three centuries that mercantilism dominated the pursuit of international
politics, underlying economic and political interests, as well as the ideas structuring
trade, underwent fundamental changes.1 The next major era of international trade can
be dated from 1815, when the Napoleonic wars ended with the defeat of France by Britain
and its allies, to 1914 and the outbreak of World War I. Known as the Pax Britannica,
this period was characterized by growing economic exchange among European
powers—the first era of what we now recognize as globalization. This period saw, in gen-
eral, more cooperation and less warfare among European states, as well as the spreading
Industrial Revolution. Peace and more integrated economies via increased capital flows
led to rapidly expanding international trade and substantial economic growth in much
of the world. The end of mercantilist doctrines took hold first in Britain, the vanguard of
the Industrial Revolution. Industrialists opposed restrictions on the import of food and
other resources, as this raised the costs of their workforce and other industrial opera-
tions. They also strongly desired the end of protectionism, which blocked them from
exporting to other markets. Urban areas of Britain thus became opponents of restrictive
devices such as the British Corn Laws (which taxed imported grain) and proponents of
free trade. While remaining protectionist interests in Britain, especially farmers, lob-
bied in favor of maintaining government restrictions on trade, by the 1840s Britain had
overturned most of its mercantilist policies.
Other governments, both in Europe and the Americas, followed Britain’s lead and
began dismantling their systems of trade protection. They built on this unilateral lib-
eralization with bilateral, and then multilateral, agreements to reduce trade barriers. If
we focus on the rich countries of the late nineteenth century, we see that their trade over
the course of the century grew at a rate that was nearly three times that of the growth
of those rich-countries’ economies. By 1900 trade represented a much larger share of
the global economy than it had in 1800, up to eight times as large (Maddison 1995, 38).
Beyond changes in doctrine and spreading industrialization, the growth of trade was
encouraged by new technologies, such as telephones and steamships, and the concomi-
tant rapidly declining costs of long-distance trade. In addition, the establishment of the
classical gold standard throughout most of the industrialized world and beyond facili-
tated trade as well as global movements of capital and people.2
As the twentieth century began, the relative decline of Britain and rise of Germany,
as well as the ongoing collapse of remaining empires within Europe (Ottoman, Austro-
Hungarian, and Russian), led to rising tension in Europe and the formation of rigid
alliances. These alliances contributed to the outbreak of World War I in 1914, and the
growing importance of the United States and Japan meant that this war spread far
beyond European boundaries. The war devastated European economies and the trade
networks that had grown over the previous century, and neither recovered rapidly when
the war ended in 1917. The United States emerged as the major economic power in the
world during the war and increased its international trade and investment to a remark-
able degree during this time. However, after the war isolationist forces won political
battles back home, and the United States pulled back from its participation in European
economies.
4   Introduction

Lack of leadership and very weak economies during the interwar period meant
that trade did not return to prewar levels. This situation was exacerbated by the Great
Depression beginning in 1929, which was characterized by competitive devaluations
and renewed protectionism that devastated international trade as well as national
economies. World War II brought an initially reluctant United States fully back into
European politics and economics, and the economic destruction caused by that war left
it the undisputed leader in the global economy. With the advent of the Cold War, the
Eastern bloc broke nearly all of its economic ties with the West, and the United States
set out to create renewed economic integration within the West. One novel aspect of the
postwar economic order was its highly institutionalized nature, as the existing network
of bilateral and multilateral trade agreements was overlaid with more formal global
and regional trade organizations. Attempts to create a far-reaching International Trade
Organization (ITO) failed, largely because they were too ambitious and ran into resis-
tance in the United States. However, the failed ITO negotiations left in their wake the
General Agreement on Tariffs and Trade (GATT), an informal set of deals to reduce bar-
riers to trade. Under the GATT, impediments to trade fell rapidly over the next decades,
and trade flows increased.3
While the GATT was intended as simply an interim accord, by default it became the
institutional basis for global trade. Rounds of multilateral negotiations greatly expanded
its reach, both to new types of trade and to more of the world. The GATT allowed for
regional trade agreements (RTAs) that even further liberalized trade, and countries
took advantage by creating and strengthening RTAs, especially beginning in the 1980s.4
Many developing countries, especially in Latin American, initially resisted the trend
toward liberalization, following instead policies of import-substituting industrializa-
tion (ISI). While the early years of ISI saw some notable successes, such as the establish-
ment of automobile industries in Brazil and Mexico, the debt crises of the 1980s forced
most developing-country governments to drop ISI and follow the export-led develop-
ment strategies that were pioneered in East Asia. Liberalization thus spread beyond the
rich world to the developing world and new democracies. By the 1990s, measures of
global economic integration including trade flows were back to pre-World War I levels.
In 1995 the GATT was replaced by the more formal and institutionalized World Trade
Organization (WTO).
The existing trade regime faces a number of challenges. As the WTO has expanded
to include economies in which the state plays a major role, such as China, other mem-
bers worry that these states are breaking the rules and undermining the system. Many
worry that the WTO’s proliberalization agenda undermines local economies and cul-
ture. Ambitious negotiations within the WTO to further expand its scope have not
resulted in major new trade deals, although the ongoing dispute resolution mechanisms
of the WTO are an essential element of the global trade regime.5 In the absence of break-
throughs in the WTO, further liberalization has occurred primarily through the prolif-
eration of RTAs and other preferential trade agreements (PTAs). In spite of protectionist
pressures, global trade flows have remained surprisingly robust. For example, in the ini-
tial stages of the Great Recession in 2008, many expressed concern that the economic
 Lisa L. Martin    5

downturn would empower protectionist forces and lead to decreased trade. While trade
did show a small dip as economies contracted, fears of a surge of protection and collapse
of trade proved unfounded.
Global forces, particularly the rise of China, could in the future lead to another fun-
damental transformation in the pattern and practice of international trade. However,
for the medium term the current system seems stable. A network of organizations and
agreements promote liberalized trade while offering loopholes that governments use to
offer temporary protection to threatened industries.6 Economic integration, on both
global and regional scales, rules the day. The countries that have resisted integration into
the modern trade system, such as North Korea, are remarkably rare. Economic integra-
tion, in turn, has created vested interests in the current system that are a major source of
its stability.

Approaches to the Study

of International Trade

Studies of the politics of international trade aim to explain variation in the demand for
and supply of protection over time, across countries, and across sectors. To do so, they
build on understandings of economic interests, domestic institutions, and international
institutions. The chapters in this volume explore those factors, and their interaction, in
great detail. In this section I provide some basic background on the major economic
models of interests and types of institutions that go into analyses of the politics of trade.
To understand the interests that go into the making of trade policy, we need to begin
with theories of what kinds of goods countries are likely to export or import. This analy-
sis is rooted in the concept of comparative advantage, as explained in 1776 by Adam
Smith, the founder of classical economics (Smith 1937). When countries trade with one
another, they no longer need to be self-sufficient and can specialize in producing goods
that they make more efficiently than other goods; that is, in the goods in which they have
a comparative advantage. The process of specialization leads to gains in productivity and
improvements in aggregate welfare for all countries. However, to build on this insight to
explain the specifics of trade policy, we require a theory of comparative advantage. How
do we know what goods a particular country will be able to produce the most efficiently?
The dominant approach to this question was developed in the 1920s by Swedish econ-
omists Eli Heckscher and Bertil Ohlin. The Heckscher-Ohlin model focuses on coun-
tries’ factor endowments: the distribution of resources that go into production of goods
and services. Typically we divide these factors into three groups: land, labor, and capi-
tal. Many modern analyses further divide capital into capital for investment and human
capital. Countries differ in their capital endowments. Rich countries are abundant in
capital and relatively scarce in labor compared to poor countries. Heckscher-Ohlin
models argue that a country will produce most efficiently the goods that use the
6   Introduction

endowments that it has in abundance. Thus, rich countries will specialize in produc-
ing capital-intensive goods such as complex machinery or high-technology products.
Countries abundant in land will tend to be agricultural exporters, while countries abun-
dant in unskilled labor will focus on production and export of consumer goods such as
clothing and simple manufactured goods.
In 1941 Wolfgang Stolper and Paul Samuelson built on the Heckscher-Ohlin
approach to trade to develop a model of the losers and winners within each country
from increased exposure to trade. The Stolper-Samuelson theorem argues that abun-
dant factors within a country will benefit from increased trade, as resources will flow
into sectors that use abundant factors as trade opens and new export markets become
available. For example, China is abundant in unskilled labor. As China opened to trade,
resources poured into the basic manufacturing sector there, benefiting workers in that
sector and leading to the massive income gains that China has experienced over the last
decades. In contrast, scarce factors will tend to be hurt by increased exposure to trade, as
resources leave their sector and returns to the scarce factor decrease. In the United States
unskilled labor is a scarce factor, and the wages of unskilled labor have fallen as trade has
increased, as resources have left sectors such as textiles and apparel.
The Stolper-Samuelson (or HOSS) model predicts that abundant factors within
a country will tend to support freer trade, while scarce factors will prefer protection-
ist policies. In the aggregate, the HOSS model is a good place to start developing an
understanding of domestic interests regarding trade. Labor in rich countries such as
the United States does lean toward protectionism, while those who are well educated
(endowed with human capital) and rich tend to support free trade. Where land is scarce,
for example in much of Western Europe or crowded developing countries, the agri-
cultural sector is highly protectionist. Many models of trade politics thus build on the
HOSS approach.
However, empirically we can also identify plenty of situations that don’t fit the HOSS
predictions, and economists have developed alternative models of interests regarding
trade policy. Analysts frequently observe that the predicted split between labor and cap-
ital over trade policy does not obtain. For example, HOSS would predict that in debates
over protection for the automobile industry in the United States, capital within the
industry would lobby for reduction in levels of protection, while labor in the automo-
bile industry would lobby for increased protection. However, most debates over protec-
tion for General Motors or Ford have not played out in that manner. Instead, we usually
find  that both capital and labor within the industry strongly support protection.7 How
do we account for these anomalies?
The HOSS model makes a strong assumption about the mobility of factors of pro-
duction (their ability to move from one productive activity to another). It assumes that
factors are fully mobile within a country (but not at all mobile across countries). So, for
example, it assumes that if the US automobile industry is under pressure from imports,
capital within that industry can easily deploy to a new use, such as production of phar-
maceuticals. So capital’s interests are not determined by the specific industry in which
it is invested. Obviously factors are not always fully mobile; sometimes they are barely
 Lisa L. Martin    7

mobile at all, being sunk in a particular industry and very costly to redeploy (Hiscox
2002). An alternative to the HOSS model is thus known as the specific factors approach,
or the Ricardo-Viner model. Ricardo-Viner assumes that factors are tied to particular
industries. Thus, when that industry is under threat from imports, all factors tied to it
will support protection. Likewise, when an industry is highly productive and gaining
export markets, all factors within it will favor freer trade. While HOSS predicts cleav-
ages in trade policy based on broad categories of factors (land, labor, capital) (Rogowski
1990), Ricardo-Viner predicts cleavages based on the specific sector in which a factor
is invested or employed. In periods, countries, or industries in which factor mobility is
limited, Ricardo-Viner may provide a better starting point for understanding economic
interests than HOSS.
Modern work in economics has challenged the assumptions of both HOSS and
Ricardo-Viner and provides theoretical frameworks that may better fit a globalized
world in which capital and other factors freely flow across international borders. New
trade theory, which Paul Krugman and others developed in the late 1970s, challenged
the assumption of constant returns to scale in production. If instead production exhibits
increasing returns to scale—that is, production becomes more efficient when carried
out on a large scale—then industries and countries that initially begin production can
outcompete new, smaller entrants into the industry (Krugman 1979). This model can
provide justification for government protection of new industries, so that they can get
a head start on potential competitors and grow in scale to become efficient exporters.8
In addition, new trade theory assumes that consumers prefer to have choice among a
number of brands within an industry; thus, for example, the variety of luxury car brands
that we observe. These assumptions lead to models that help to explain why we observe
so much trade among countries with similar factor endowments, while HOSS pre-
dicts trade among dissimilar countries on the basis of comparative advantage. Under
new trade theory, countries specialize in producing a few brands of a given product
and will trade these brands with one another. Germany exports BMWs to Sweden, and
Sweden exports Volvos to Germany.
While new trade theory drew attention to the industry as the major unit of analysis,
the so-called new new trade theory delves even deeper, focusing on variation among
firms within an industry (Melitz 2003). Within any industry in a country, firms will
vary in their level of productivity. When trade opens up, the most productive firms
will become exporters, and the least productive firms will be driven out of business.
Thus, even within an industry we expect to observe variation in preferences for pro-
tection, with globally competitive, productive firms preferring free trade and less pro-
ductive firms demanding protection. With international competition, the population of
firms changes as the least productive are driven out of business. Within this new, smaller
population of remaining firms, the threshold at which firms are productive enough to
export rises, so firms that were formerly marginally in support of freer trade can now
become protectionist instead. Empirical explorations of new new trade theory are
emerging, providing the foundation for a wave of work in political science asking about
individual-level preferences for protection.9
8   Introduction

One additional foundational economic model of trade, this one focused on the politi-
cal economy of trade, has become essential to the analysis of the politics of trade. Gene
Grossman and Elhanan Helpman, in what has become known as the protection-for-sale
(PFS) or Grossman-Helpman model, examine the interplay between governments and
sectors that demand protection.10 They assume that the government values both aggre-
gate welfare in the country, which is harmed by trade protection, and contributions
from economic actors who demand protection in exchange. Sectors of the economy can
be either organized or not. Organized sectors can offer the government a specified con-
tribution in exchange for a particular level of protection. The government chooses to
provide protection to some industries, until it reaches a point where further protection
would reduce aggregate welfare enough to make the government worse off.
This model offers some clear predictions about the level of protection given to vari-
ous industries and has received substantial empirical evaluation. Some of the main
insights of the PFS model are discussed here. First, governments that more highly
value aggregate welfare will provide less trade protection. Political scientists have thus
linked the PFS model to variation in domestic institutions that determine the extent to
which governments must be responsive to general welfare concerns. Second, organized
industries will receive protection, while industries that are not organized will receive
no protection or be subject to taxes on exports.11 Third, government will provide more
protection to those industries for which trade distortions will have the smallest impact
on aggregate welfare. In this model, that means that protection will be higher for goods
for which demand is relatively inelastic. If demand is inelastic—not highly responsive
to price—then providing protection will generate lower distortions to the overall econ-
omy and have less impact on overall welfare. Empirical tests of the PFS model have gen-
erally found support, in that industries where demand is inelastic do tend to receive
higher levels of protection (Goldberg and Maggi 1999). However, empirical tests have
also led to some surprising results, such as the finding that the US government puts a
much greater weight on aggregate welfare than on contributions, to an extent that seems
implausible. When connecting economic models of trade interests to the actual provi-
sion of protection, many political scientists begin with some version of the PFS model.
Models of economic interests in trade are an essential starting point in studying the
politics of trade. They provide insight into the demand for protection. The PFS model
provides one way of connecting that demand for protection to its supply by govern-
ments, but its model of the government is very simple: just an actor that maximizes
some weighted combination of aggregate welfare and contributions from sectors.
Where these weights come from is not examined, nor is any other aspect of govern-
ment. Political scientists probe much more deeply into the institutions, both domes-
tic and international, that channel demand for protection or for freer trade into actual
policy outcomes.
As mentioned in discussion of the PFS model, preferences regarding trade policy
interact with the organization of economic interests. Interests that are diffuse and so
difficult to organize, or that face other obstacles to mobilization such as unfortunate
geographical distribution,12 will find it more difficult to have their preferences reflected
 Lisa L. Martin    9

in policy outcomes. Protection tends to help more particular, well-organized interests,

while harming diffuse interests such as general consumer welfare. Trade policy battles
thus often come down to concentrated, particularistic interests demanding protection
from international competition against more diffuse, disorganized interests in free trade
that reduces distortions to the national economy. Domestic institutions become essen-
tial in explaining policy at this point, as they help to determine the degree to which gov-
ernments are responsive to particularistic versus general interests.
One crucial set of institutions that play into this process are social institutions that
organize broad coalitions and effectively present their demands to government.
If unskilled labor, for example, is organized through encompassing and effective
unions—as in the United States in the mid-twentieth century—labor’s interests are
more likely to influence trade policy than in the absence of such unions. When interests
are narrowly organized, along sectoral or regional lines, they are likely to put less weight
on aggregate welfare and be more protectionist. The more narrow the organization of
interests, the more protectionist demands are likely to be. Thus it is not surprising that
where labor is organized on a country-wide basis, as in much of Western Europe, labor’s
demands are less protectionist (McGillivray 2004).
Political institutions become linked to the social institutions that organize interests.
Some political institutions encourage tight links between narrowly organized inter-
ests and government; others create governments that are necessarily more responsive
to broad interests. In a very broad-brush sense, democracy is an important factor in
this equation. To varying degrees, democratic institutions force politicians to respond to
broad constituencies rather than the narrow base of support that we might find in a per-
sonalistic dictatorship, for example. This logic suggests that in general, moving toward
more democratic institutions should empower broad interests and in many cases sup-
port movement toward freer trade. We do observe, again in a very general sense, that
recent waves of democratization have gone hand in hand with trade liberalization,
although attributing causal mechanisms in this process is difficult. In a more systematic
examination, Milner and Kubota (2005) find that developing countries that are more
democratic are also more likely to engage in trade liberalization, corroborating the gen-
eral insight.13
Scholars also tie more specific elements of political institutions to patterns of protec-
tion. When politicians are elected on a narrow regional basis, they are likely to be highly
responsive to the economic conditions of their regional constituency and less respon-
sive to national welfare concerns. We would therefore expect that politicians elected
on a broad national basis—such as the president in the United States—would tend to
support free trade, while members of the legislature who are elected by a small district
would be more protectionist. Lohman and O’Halloran (1994), for example, provide sup-
port for this hypothesis about the relationship between the size of the constituency and
the politician’s stance on trade policy. Similarly, political parties that are organized along
broad lines of class are more likely to support free trade than are those that are tied to
specific industries or regions. Rogowski (1987) has argued that proportional representa-
tion systems, in which voters choose a party rather than an individual candidate, will
10   Introduction

result in less protectionist outcomes than those in which elections are for a specific indi-
vidual candidate, as in the United States.14 Beyond electoral systems, other aspects of
domestic institutions will determine trade policy. For example, in the United States a
shift in authority to provide protection away from Congress to the president coincided
with the move away from very high levels of protection in the interwar period (Bailey,
Goldstein, and Weingast 1997; Lohmann and O’Halloran 1994). In many developed
countries today, much of the action on trade policy takes place within agencies rather
than legislatures, as the agencies are empowered to provide temporary relief from pro-
tection through safeguard provisions in international agreements. These agencies can
operate without much public visibility and can be captured by narrow interests, perhaps
leading to an increased tendency to protect threatened industries.
Studies of domestic institutions tend to examine individual countries in isolation.
However, as the previous section of this chapter discussed, over time more and more
trade policy is made not by individual governments, but in the process of international
negotiation and through the work of international institutions. A complex web of bilat-
eral and multilateral trade agreements, regional and other preferential trade organi-
zations, and the global WTO regime now constrain individual states’ trade policies.
Governments enter these agreements voluntarily, but once in it is very costly to exit such
commitments, and as trade competitors enter agreements, choosing to remain outside
of them can become prohibitively expensive. Thus, international bargaining and inter-
national institutions are another crucial channel through which interests are aggregated
into trade policy outcomes.
Governments that enter into international negotiations on trade or that request
admission to a trade organization are typically driven by the desire to gain access to
additional markets (driven, in turn, by the interests of exporters and potential export-
ers back home). While economists focused on comparative advantage decry such
language, negotiations therefore take the form of offering reciprocal “concessions” in
terms of access to the home market. Reciprocity is central to international trade politics.
During initial bargaining, governments offer to lower barriers to specific types of trade
in exchange for lower barriers to their exports in other countries. Reciprocity contin-
ues to dominate the functioning of agreements and trade institutions after governments
reach deals. Countries are often tempted to cheat on trade agreements, providing pro-
tection to home industries during hard times (Downs and Rocke 1997). If they do and
are caught, enforcement takes the form of retaliating by imposing punitive tariffs on
that country’s exports in the affected market. Complementing reciprocity, the concept
of most-favored-nation (MFN) status extends deals that countries reach bilaterally, or
among a small group, to all other members of an organization. Reciprocity and MFN
status characterize the agreements and institutions that structure modern international
trade, although we increasingly see regional deals offering market access that goes
beyond what countries receive through MFN status.
As discussed in the previous section, the modern global trade regime dates to the cre-
ation of the GATT in 1947, replaced by the more formalized WTO in 1995. Throughout
the history of the GATT/WTO, rounds of multilateral negotiations have led to
 Lisa L. Martin    11

commitments to reduce barriers to trade. Small groups of states reach deals on the basis
of reciprocity and then use MFN to extend these deals to all members of the regime.
The WTO also has rules for providing temporary relief to threatened industries and a
dispute resolution mechanism for members. While technically all members of the WTO
have the same voting power, in practice the process of negotiation among the major
trading powers—the United States, European Union, and Japan—has meant that deals
have favored the interests of these rich, developed countries, and the concerns of devel-
oping countries have not been as high on the agenda (Gowa and Kim 2005).
The WTO launched the latest round of negotiations, known as the Doha Round, in
2001, and over a decade later the round has not yet reached a final deal. Doha Round
negotiations have been characterized by confrontation between developing countries
who want to see liberalization of agricultural trade in the rich world and developed
countries concerned about trade in services and protection of intellectual property.
A draft deal in December 2013 renewed hopes that the Doha Round would eventu-
ally be completed. But even in the absence of major new trade deals, the WTO dis-
pute resolution mechanism and existing framework of rules provide an institutional
structure that sets the foundation for members’ trade policies. The WTO also engages
in monitoring of members’ trade policies and in a variety of ways provides extensive
information about trading activities. The provision of information and mechanism for
resolving disputes are crucial to the maintenance of multilateral cooperation on trade
issues.
As negotiations in the WTO have stalled or failed to address individual states’ trade
concerns, governments have increasingly turned to RTAs and PTAs. As of 2013, the
WTO estimates that there were 379 RTAs in force. Some of the most important RTAs
are the European Union (EU); the North American Free Trade Agreement (NAFTA);
and Mercosur, an agreement among South American countries in which Brazil
plays a leading role. Some RTAs are bilateral, and others are very large—the EU has
twenty-seven members and continues to expand. Like the WTO, RTAs operate on the
basis of reciprocity. They vary greatly in the extent to which they are formalized and
provide institutional mechanisms for dispute resolution, monitoring of trade policies,
and so forth. One ongoing debate among economists and political scientists revolves
around the relationship between the WTO and RTAs. Does increasing reliance on RTAs
as the mechanism of liberalization undermine the multilateral trade regime and intro-
duce distortions into global trade? Or do RTAs allow for more rapid, deeper integration
among small groups of countries that will eventually translate into deep multilateral
liberalization?15 Regardless of the answer to this question, in the absence of successful
negotiations in the WTO, states will pursue their interests through other international
trade institutions.
The chapters in this volume build on these historical and conceptual foundations.
Economic models of trade interests provide a starting point. Political scientists focus on
aggregation of interests through organization, bargaining, and institutions, on both the
domestic and international levels. In the next section I explain how this general frame-
work is reflected in the organization of this volume.
12   Introduction

Organization of This Volume

The chapters in this volume are organized into six sections, which loosely follow the
conceptual and theoretical framework just discussed. The contributions survey both
classic work in the field and new developments. They outline new research agendas and
standing puzzles and present new, cutting-edge research on the politics of international
trade.
The first section provides an overview of historical, theoretical, and methodologi-
cal developments in the study of trade. Joanne Gowa explains the GATT/WTO system,
both its origins in international negotiations and its effects. She argues that the benefits
of the system have largely been concentrated among its large founding members and ties
this result to major powers’ security concerns. Gordon Bannerman provides a concep-
tual overview of the idea of free trade. What began as a unilateral, voluntary commit-
ment to removing all restrictions on international exchange has evolved into a much
more nuanced understanding of the benefits and costs of trade liberalization. Chad
P. Bown focuses on the specific instruments that governments have used in the modern
era to limit and channel trade flows. The GATT/WTO has drastically reduced what was
previously the major impediment to trade, tariffs (taxes on imports at the border). But
that process has revealed, or perhaps even encouraged, the use of administrative means
of protection known collectively as nontariff barriers (NTBs). Finally, Raymond Hicks
delves into the methodological challenges of testing theories of trade politics by concen-
trating on the debate over the effects of the GATT/WTO. The availability of new data
and the rapid development of new statistical methods have challenged scholars attempt-
ing to stay at the frontiers of research. Hicks offers some central lessons of methodologi-
cal developments for ongoing empirical research.
Parts  2 through 5 consider a particular set of actors in international trade. Part  2
begins by looking at domestic society. Jason Kuo and Megumi Naoi examine what is
probably the most micro-level of analysis, the individual. Advances in survey research
techniques have opened up the ability to test theories of trade by asking whether the pre-
dicted effects obtain at the individual level. Kuo and Naoi survey this new research on
individual attitudes and present some new results. Within domestic society—especially
in developed countries—the role of labor in debates over trade policy has often been
central. Erica Owen looks at the work on whether labor is an important source of pro-
tectionist sentiment and how the role and attitudes of labor have changed over time.
Finally, B. Peter Rosendorff links the domestic to the international level through the lens
of trade disputes. Because domestic actors are subject to economic shocks and put pres-
sure on governments to shield them from these shocks, international trade agreements
allow for some flexibility in the enforcement of trade deals. However, governments can
be tempted to use this flexibility in unintended ways, creating the need for dispute settle-
ment procedures in international agreements. Rosendorff argues that these procedures
are best seen as information-providing devices.
 Lisa L. Martin    13

Part 3 turns to firms as the unit of analysis. The five chapters in this section consider
the role of the firm in trade policy from different angles. Lucy M. Goodhart begins by
looking at efforts to explain variations in the level of protection that different indus-
tries receive. She presents a detailed discussion of the PFS model, empirical studies of
PFS, and where it falls short of explaining observed variation in levels of protection. She
discusses new work that goes beyond PFS to present new models of the demand and
supply of protection by industries. Timothy M. Peterson and Cameron G. Thies focus
on the role of international trade within industries. A number of analysts have studied
the implications of intraindustry trade for domestic outcomes such as lobbying activity.
Peterson and Thies expand this discussion to international-level implications, such as
political affinity and militarized conflict. Michael Plouffe analyzes the new new trade
theory summarized above and its implications for variation in protection. As discussed,
the key insight of this body of work is that, even within a given industry, firms are het-
erogeneous in their levels of productivity. This heterogeneity gives rise to variation in
demands for protection or liberalized trade. Tim Büthe examines the interaction of
antitrust policies and trade policies. Economists have typically understood these policy
areas as substitutes; there is little need for antitrust regulation to keep the domestic mar-
ket competitive if the market is open to international competition. However, empirical
evidence does not support this interpretation. A resolution to this conflict may lie in the
recognition that firms lobby regulators not just in their own country, but across national
borders. Finally, Hatch, Heiduk, and Bair consider the implications of the growth of
multinational corporations (MNCs) for trade policy. Increasingly, MNCs use a model
of production that relies on networks—different elements of the productive process are
scattered around the globe, building on comparative advantage and variation in legal
environments. Focusing on production networks in different regions, Hatch, Heiduk,
and Bair show how global production networks fundamentally change the context of
international trade.
Part 4 concentrates on domestic political institutions. As mentioned above, studies
of newly democratizing states have generally found that democratization and trade lib-
eralization go hand in hand. However, Bumba Mukherjee argues that this generaliza-
tion misses a great deal of nuance in the pattern of protection in developing countries
that undergo democratization. While these governments typically do liberalize trade
in low-skill goods, they often maintain or increase protection of skill-intensive indus-
tries. In addition, Mukherjee examines the question of whether trade liberalization
itself is a contributing factor to democratization. Stephanie J. Rickard turns to politi-
cal institutions in established democracies, where a major source of variation is in the
details of electoral systems. While there is a great deal of work on the effects of electoral
systems, there is surprisingly little consensus on those effects. Rickard argues that the
lack of robust empirical findings in this literature so far is not a reason to declare the
research program a failure and identifies new directions for research. Daniel Yuichi
Kono takes on the relatively new and challenging topic of trade policy in authoritar-
ian regimes. As nondemocracies, especially China, play a larger role in the global
trade regime, it is crucial that we understand trade politics within these states. Yet we
14   Introduction

have few generalizations about how trade policy is made in authoritarian states. Kono
identifies four elements of authoritarian regimes (economic structure, coalition size,
time horizons, and leader’s mode of entry into power) as promising bases for devel-
oping theories of authoritarian regimes. Finally, Kerry A. Chase looks at the interac-
tion between domestic geography and institutions. Different industries have different
geographical organization within countries, which in turn affects their ability to influ-
ence government policy. Chase identifies the central questions in this research area,
such as whether geographic concentration is helpful or harmful for the pursuit of trade
interests.
Part 5 jumps to the international level of analysis. As discussed above, international
trade politics is highly institutionalized, and the four chapters in this section look at
these international institutions from different perspectives. Leslie Johns and Lauren
Peritz focus on the design of trade agreements. They provide insight into why PTAs
are designed in particular ways, but also suggest that this literature needs to move
beyond examining individual PTAs to take a more systemic view that includes the
interaction of PTAs with one another. Soo Yeon Kim studies RTAs, in particular their
role in promoting “deep integration” that goes well beyond removing impediments at
the border to trade flows. As global trade negotiations stall, this regional promotion
of deep integration may represent the future of international trade politics. Christina
L.  Davis and Meredith Wilf consider the expansion of the WTO over time. What
determines the pattern of WTO expansion, and what are its implications? Davis and
Wilf argue that noneconomic interests are important in explaining decisions to join
the WTO. Finally, Marc L. Busch and Krzysztof J. Pelc focus on the dispute settle-
ment elements of the WTO, which provide ongoing support to liberalization efforts
even in the face of stalled negotiations. Busch and Pelc argue that existing political
science literature on dispute settlement does not pay sufficient attention to its legal
dimensions.
Trade policy does not exist in a vacuum. It interacts with other policy areas, such
as national security, human rights, and capital flows. Part 6 examines these issue link-
ages. Erik Gartzke and Jiakun Jack Zhang take on the perennial topic of the relationship
between trade and war, arguing that this literature needs to incorporate better models
of domestic politics to make sense of conflicting empirical results. Mark S. Copelovitch
and Jon C. W. Pevehouse note the surprising deficit of work linking exchange rate and
trade policies, in spite of widespread understanding that these policies are codeter-
mined. They present initial work that bridges these “silos” in the literature. Does inter-
national trade harm the natural environment, and how should environmental concerns
enter into trade agreements? J. Samuel Barkin argues that these questions are more
complex and multilayered than has been appreciated in the literature. Trade policy
has always been a central part of governments’ development policies, as the discussion
above of ISI indicated. Mark S. Manger and Kenneth C. Shadlen provide an in-depth
examination of the relationship between trade and development, arguing that the inter-
ests of exporters in developing countries weigh heavily in the policy process and force
a shift in theories of trade policy. Susan Ariel Aaronson takes on the fraught question
 Lisa L. Martin    15

of the relationship between trade and human rights. While much of the literature has
focused on the introduction of human rights provisions in trade agreements, Aaronson
finds that this literature misses the big picture in its focus on personal integrity rights to
the exclusion of other, perhaps more relevant, aspects of human rights for trade. Finally,
Margaret E. Peters studies the interaction of migration policies and trade policies. She
argues that flows of people and goods are to a large extent substitutes for one another,
with predictable implications for the policy preferences of firms and individuals regard-
ing immigration and trade protection.
Understanding the politics of international trade requires a grasp of the econom-
ics of trade, how economics translates into variation in preferences over trade policy,
and how these preferences are aggregated through negotiations and institutions into
policy outcomes. The chapters in this volume present foundational and path-breaking
research on each stage of this process. Both theories and empirical studies of trade are
a thriving area of research in international political economy, and they draw on a wide
variety of analytical perspectives. These perspectives are represented in this volume,
which we hope will serve as both an introduction to the subject and the impetus for
further research.

Notes
1. See Bannerman chapter in this volume.
2. See Copelovitch and Pevehouse in this volume for analysis of the relationship between
trade and exchange rate policies; and Peters for the relationship between trade and
migration.
3. However, a lively debate rages about the extent to which the GATT itself actually contrib-
uted to increased trade flows. See chapters by Hicks and by Gowa in this volume.
4. On PTAs, see the Kim chapter in this volume.
5. See chapters by Rosendorff and by Busch and Pelc in this volume.
6. See the Bown chapter in this volume.
7. As the industry has become dominated by multinational corporations with global pro-
duction networks, the pattern of lobbying has become more complex. See the chapter by
Hatch, Heiduk, and Bair in this volume.
8. Krugman and others acknowledge that the difficulty of guessing which particular indus-
tries might succeed often outweighs the potential benefits of government support.
9. See the Kuo and Naoi chapter in this volume.
10. See the Goodhart chapter in this volume for further discussion of the PFS model.
11. Taxes on exports are highly unusual in the modern economy, so this implication of the
PFS model has not received empirical validation.
12. See the Chase chapter in this volume.
13. See also the chapters by Mukherjee in this volume on new democracies and by Kono on
dictatorships.
14. But see Rogowski and Kayser (2002). The Rickard volume in this chapter studies the
impact of electoral systems on trade policy.
15. See Lawrence (1991).
16   Introduction

References
Bailey, Michael, Judith Goldstein, and Barry Weingast. 1997. The Institutional Roots of
American Trade Policy. World Politics 49 (3): 309–338.
Downs, George W., and David M. Rocke. 1997. Optimal Imperfection? Domestic Uncertainty
and Institutions in International Relations. Princeton, NJ: Princeton University Press.
Goldberg, Pinelopi Koujianou, and Giovanni Maggi. 1999. Protection for Sale: An Empirical
Investigation. American Economic Review 89 (5): 1135–1155.
Gowa, Joanne, and Soo Yeon Kim. 2005. An Exclusive Country Club: The Effects of the GATT
on Trade, 1950–94. World Politics 57 (4): 453–478.
Grossman, Gene M., and Elhanan Helpman. 1994. Protection for Sale. American Economic
Review 84 (4): 833–850.
Hiscox, Michael. 2002. Commerce, Coalitions, and Factor Mobility: Evidence from Congres­
sional Votes on Trade Legislation. American Political Science Review 96 (3): 591–608.
Krugman, Paul. 1979. Increasing Returns, Monopolistic Competition, and International Trade.
Journal of International Economics 9 (4): 469–479.
Lawrence, Robert A. 1991. Emerging Regional Arrangements: Building Blocs or Stumbling
Blocks? In Finance and the International Economy 5: The AMEX Bank Review Prize Essays,
edited by Richard O’Brien, 25–35. New York: Oxford University Press.
Lohmann, Susanne, and Sharyn O’Halloran. 1994. Divided Government and U.S. Trade Policy:
Theory and Evidence. International Organization 48 (4): 595–632.
Maddison, Angus. 1995. Monitoring the World Economy, 1820-1992. Paris: Development Centre
of the Organisation for Economic Co-operation and Development.
McGillivray, Fiona. 2004. Privileging Industry: The Comparative Politics of Trade and Industrial
Policy. Princeton, NJ: Princeton University Press.
Melitz, Marc J. 2003. The Impact of Trade on Intra-Industry Reallocations and Aggregate
Industry Productivity. Econometrica 71 (6): 1695–1725.
Milner, Helen V., and Keiko Kubota. 2005. Why the Move to Free Trade? Democracy and Trade
Policy in the Developing Countries. International Organization 59 (1): 107–143.
Rogowski, Ronald. 1990. Commerce and Coalitions:  How Trade Affects Domestic Political
Alignments. Princeton, NJ: Princeton University Press.
Rogowski, Ronald. 1987. Trade and the Variety of Democratic Institutions. International
Organization 41 (2): 203–223.
Rogowski, Ronald, and Mark A. Kayser. 2002. Majoritarian Electoral Systems and Consumer
Power:  Price-Level Evidence from the OECD Countries. American Journal of Political
Science 46 (3): 526–539.
Smith, Adam. 1937 [1776]. An Inquiry into the Nature and Causes of the Wealth of Nations.
New York: Modern Library.
Thomas, Robert Paul. 1965. A Quantitative Approach to the Study of the Effects of British
Imperial Policy on Colonial Welfare. Journal of Economic History 25 (4): 615–638.
Viner, Jacob. 1948. Power versus Plenty as Objectives of Foreign Policy in the Seventeenth and
Eighteenth Centuries. World Politics 1 (1): 1–29.
 Pa rt  I

H I STOR IC A L ,
T H E OR E T IC A L , A N D
M E T HOD OL O G IC A L
DE V E L OP M E N T S
 Chapter 2

E x pl aining th e G AT T / W TO
 origins and effects

Joanne  Gowa

It has long been conventional wisdom in international relations that institutions can
help states cooperate with each other when their decentralized efforts to do so fail. The
canonical problem that creates a role for a trade institution arises when more than one
state can exert power over its terms of trade—that is, the price of its exports in terms
of its imports. In this case, each state has an incentive to impose a tariff to raise its real
income. If all do so, however, they are worse off than if they had all adhered to free trade.
This creates the familiar “prisoners’ dilemma” (PD). An institution has value added in
this case because it can help states realize the Pareto-improving equilibrium outcome of
open markets.
The institutions that have governed international trade have varied widely over time.
During the first golden age of globalization in the late nineteenth century, a network
of bilateral trade treaties regulated commerce. Between the world wars of the twenti-
eth century, a now infamous set of trade blocs did so, the legacy of repeated diplomatic
failures to establish a liberal regime and of the Great Depression. The interwar system
instantiated discrimination and is widely regarded as having wreaked havoc on global
trade while exacerbating the political tensions that erupted into war in 1939. In con-
trast, the second golden age of globalization, which began after World War II, witnessed
the emergence of an overarching multilateral organization, the General Agreement on
Tariffs and Trade (GATT), and its successor, the World Trade Organization (WTO).
The variation in international economic institutions over time is due partly to shifts
in the distribution of power at home and abroad that make the institutions’ construc-
tion more or less viable. As several detailed histories of trade make clear (e.g., Findlay
and O’Rourke 2007; O’Rourke and Williamson 1999), it is also the product of apoliti-
cal changes. During the first golden age of globalization, for example, sharply falling
transport costs led to a secular increase in trade. Shifts in exchange-rate regimes can
also induce variations in the demand for institutions. The same is true of changes in the
20    Explaining the GATT/WTO: Origins and Effects

composition of trade: trade in differentiated products, for example, increases the gains

from trade and the appeal of open international markets.
This chapter examines the role of international politics in the creation and operation
of the GATT/WTO. Other chapters in this volume address the role that domestic poli-
tics plays in world trade. The question I address here is whether political-military alli-
ances influenced trade expansion after World War II. Allies have an especially strong
interest in trading with each other and relatively weak concerns about cheating. This
biases them toward increasing their trade with each other despite the explicit prohibi-
tion against discrimination in the GATT/WTO charter. I show that alliances do exert a
positive impact on the trade of contracting parties, effectively embedding discrimina-
tion in the postwar regime.
I begin by briefly explaining the value added of trade institutions and the impact on
trade of security interests. Then I explain the intended role of the GATT and the extent
to which it fulfilled the role that theory assigns to institutions. Finally, I examine the
data, showing that alliances do exert a positive influence on trade between GATT/WTO
members.

Why Institutions? Economics

and Politics

International institutions arise to solve market-failure problems (Keohane 1984). Their

raison d’être in the case of trade inheres in their ability to solve the public-good prob-
lem that arises because some states have incentives to use tariffs and other trade barri-
ers. According to standard trade theory, a large state—that is, one capable of exercising
market power—can use a tariff to increase its real income at the expense of other states
(e.g., Bagwell and Staiger 2002). A tariff increases a state’s welfare because it reduces its
demand for imports and lowers their price. Although a tariff also imposes costs in terms
of foregone trade, a state can always set its tariff to ensure that its net gain is positive. Its
trading partner bears the cost of the price drop and the deadweight loss associated with
the inefficient income transfer a tariff effects.
In contrast, standard trade theory does not offer a small state—that is, a state that can-
not exercise either monopoly or monopsony power over its terms of trade—any incen-
tive to use a tariff. While leaving the state’s terms of trade unchanged, its tariff will reduce
its demand for imports. The associated loss of consumer surplus exceeds the gains that
its producers and the government reap. This means that open markets maximize a small
state’s welfare regardless of the trade policies of other states. Even if it is the target of a
large state’s tariff, for example, the small state is better off absorbing the associated cost
than retaliating: a tariff will only decrease its real income without reducing the large
state’s welfare. As such, a small state’s deviation will not induce punishment by a large
state; the small state only shoots itself in the foot when it uses a trade barrier.1
 Joanne Gowa   21

This implies that there is no role for a trade institution when only small states exist or
only one large state exists.2 The same applies when only two large states exist. In this case,
the incentive structure of each corresponds to a PD game in which the dominant strategy is
defection. Mutual cooperation can emerge in an infinitely repeated PD game. Although an
infinite number of equilibria exist, including perpetual defection, the use of a grim-trigger
strategy can, for example, enable two large states to realize the Pareto-improving equilib-
rium of free trade. The strategy mandates that both states defect forever in the event that
either deviates.3 In this situation, each state has an incentive to monitor the other’s com-
pliance, because it bears the full cost of the latter’s deviation. As long as the discounted
sum of payoffs for cooperation exceeds the sum of a one-shot gain from defection and the
discounted sum of payoffs from mutual defection, neither state will wander off the equi-
librium path of play. Although implicit, this mechanism presumably enabled a network of
trade treaties to emerge sua sponte in the late nineteenth century.4
When more than two states capable of exercising market power exist, however, a
public-good problem arises. Each state has an incentive to free ride on the efforts of
others to monitor compliance. While information about defection is excludable in the-
ory, in practice states have incentives to make it common knowledge, because punish-
ment is incentive compatible for each large state when any defection occurs. This means
that making private information public leads to harsher punishment for the defecting
state and a more effective deterrent. In practice, therefore, the supply of information is a
public good. An institution can help in this situation to the extent that it centralizes the
monitoring of compliance.5
Existing theory is silent about many other aspects of the “rational design” of a trade
institution (Koremenos, Lipson, and Snidal 2001). Among these is the set of rules that
should govern negotiations under its auspices. The most basic, perhaps, is whether tariff
cuts should be extended to all members—that is, whether to adopt most-favored nation
(MFN) treatment. While small states have little reason to join an institution absent
MFN treatment and nondiscrimination reduces the costs associated with distinguish-
ing imports by country of origin, it may not be welfare enhancing relative to a regime in
which tariff cuts apply more narrowly (Horn and Mavroidis 2001). An MFN clause can
induce free riding as states reap the benefits of tariff cuts that other members make with-
out lowering their own trade barriers. This can stall the process of trade liberalization as
each state waits for others to cut their tariffs.
The literature about the value added of institutions also has relatively little to say
about the role that international politics plays in their creation and operation. As Martin
and Simmons (1998, 737) note, “distributive consequences” fell rapidly “from the cen-
ter of consideration.” Attempts to integrate international politics into the study of
regimes have proved controversial. Some argue that national interests and institutions
are isomorphic (e.g., Mearsheimer 1994/95). Others maintain that institutions affect the
interests and behavior of even large states, socializing them to respect the norms of the
organizations they join (e.g., Johnston 2001; Finnemore 1993). Still others maintain
that institutions can be “simultaneously causes and effects”—that is, they are “both the
objects of state choice and consequential” (Martin and Simmons 1998, 743).
22    Explaining the GATT/WTO: Origins and Effects

Perhaps the most basic issue that the politics of regimes engenders is whether states
can cooperate with each other in a world in which each is the guarantor of its security.
When war is always an option, some argue, states decide whether to cooperate not only
on the basis of the gains they stand to realize but also on the basis of the welfare increases
that accrue to other states:  an asymmetrical distribution of the gains from coopera-
tion, they contend, can place the security of states at risk (e.g., Waltz 1979; Grieco 1988;
Mearsheimer 1994/95, Powell 1991). Krasner (1991) claims that the equilibrium out-
comes that institutions enforce are themselves endogenous to international politics: the
distribution of power among its members determines the choice among equilibria on
the Pareto frontier.6
That a link exists between trade and security is the premise of the literature about allies
and trade. Trade generates both private and social welfare gains (Gowa and Mansfield
1993; Gowa 1995). Trade with an ally generates a positive security externality, raising the
real income of both the home state and its ally and increasing the aggregate power of
their alliance. Trade with an adversary, on the other hand, produces private gains and
social costs: the income that accrues to the adversary, because it increases its potential
political-military power, creates a negative security externality. This gives allies stronger
incentives to trade with each other than with other states. Although the original claim
applied exclusively to large states, because only they can affect their terms of trade, sub-
sequent work shows that it applies to allies of all sizes (e.g., Gowa and Mansfield 1993;
Mansfield and Bronson 1997; Long 2003).
Nonetheless, the extent, if any, to which alliances affect the trade expansion that insti-
tutions generate has rarely been systematically examined.7 This is so even though the
existence of security externalities generates stronger incentives to lower trade barriers
between allies than between other states. Security externalities also reduce the demand
for monitoring, because the payoff that accrues to any member of an alliance from
defecting also increases aggregate alliance strength. As long as the sum of the payoffs
from the temptation to defect and unilateral cooperation is lower than the aggregate
gains from cooperation, however, allies retain an interest in monitoring each other,
albeit one that is weaker than in the case of other states.
Before asking about the impact international politics exerted on the postwar trade
regime, I briefly review the principles and purposes that motivated the creation of the
original GATT to make clear the rules that its architects intended to govern its operation.

The Creation of the GATT

Secretary of State Cordell Hull’s belief that trade discrimination generates hostile polit-
ical relations has become legendary among students of international politics. He and
other US officials set out to create a multilateral regime based on nondiscrimination.
Intent on eliminating any vestiges of the interwar blocs, they began to work on a blue-
print for the postwar era even before the United States entered World War II. Using
 Joanne Gowa   23

Lend Lease aid as leverage, the United States pressed Britain to sign the 1941 Atlantic
Charter, which pledged its signatories to “endeavor, with due respect for their existing
obligations, to further the enjoyment by all States, great or small, victor or vanquished,
of access, on equal terms, to the trade and to the raw materials of the world … needed
for their economic prosperity.”8
American officials assumed that a reversion to the interwar system was all too likely
absent an overarching trade institution. Before 1914 the gold standard had corrected
payment imbalances that otherwise might have impeded global trade; privileging
exchange-rate stability over domestic output and employment was designed to ensure
the automatic reversal of deficits and surpluses. Tariffs had begun to rise before 1914,
but it took the Great War to write “an abrupt and fiery coda to the preceding century’s
unremitting expansion of trade” (Rogowski 1989, 61). The war per se did not, however,
produce the trade and currency blocs that would later preoccupy US officials. Rather,
they were endogenous to the onset of the Great Depression and the efforts states made
to keep intact what Eichengreen (1992) describes as their “golden fetters.”9 If the post-
war exchange-rate regime enabled states to sustain growth, it seemed very unlikely that
the years after World War II would witness a replay of the interwar experience. But the
specter of the interwar blocs nonetheless motivated the construction of a system of rules
designed to facilitate the opening of international markets.
The architects of the postwar trade regime envisaged the GATT as “simply a trade
agreement” under the aegis of an International Trade Organization (ITO) that would
govern not only trade but also policies related to competition, employment, investment,
and economic development (Irwin, Mavroidis, and Sykes 2008, 100). In early 1946 the
United States prompted the Economic and Social Council (ECOSOC) of the United
Nations to appeal for the creation of a regime to establish the “rules and norms by which
trade would occur and would provide oversight, through a large administrative struc-
ture, of the adherences to these rules” (Barton et al. 2006, 35). Given its domain, the
ITO required member states to delegate “far more authority to an international institu-
tion than did the GATT” (Barton et al. 2006, 36). The US Congress, however, refused to
sanction the creation of the ITO even though the United States “had taken the principal
initiative” to develop its charter (Jackson 1989, 34).
Thus, the GATT evolved as a somewhat accidental organization. The bilateral trade
agreements that the United States had concluded under the 1934 Reciprocal Trade
Agreements Act (RTAA) became its template. Its founding members accepted several
provisions included in the RTAA agreements, such as an unconditional MFN clause,
regulations about government purchases, and procedures to compensate changes in
agreed-upon tariffs. The GATT, as Irwin, Mavroidis, and Sykes (2008, 12) note, “was not
created from scratch, but represented a continuation and expansion of US efforts during
the 1930s.”
Article I  of its charter formally established nondiscrimination, expressed as MFN
treatment, as the basis of the postwar regime. The charter also, however, allowed states
to retain the preferences that they already had in force, including those that Britain had
extended to Imperial Preference System (IPS) members. It also condoned the formation
24    Explaining the GATT/WTO: Origins and Effects

of preferential trade agreements (PTAs) that lowered trade barriers among a subset of
GATT members as long as their ex post barriers to trade were no higher on average than
they had been ex ante. The PTA provision facilitated European integration, a process
that had begun with the proposal for a customs union between France and Italy in July
1947 (Odell and Eichengreen 1998, 193).10
Despite the lengthy negotiations that produced it, the GATT was ill-equipped to per-
form the critical role that market-failure theory assigns to institutions. As noted previ-
ously, the principal value added of a multilateral trade organization is its ability to solve
the public-good problem created by the need to monitor compliance. The contracting
parties, however, refused to endow the GATT with the capacity to perform this role.
They agreed neither on a budget nor on “a fully fledged secretariat” (Kim 2010, 7). They
were unwilling to “create, or pay for, any kind of standing committees” (Barton et al.
2006, 43). The UN Conference on Trade and Employment created the secretariat, for-
mally known as the Interim Commission for the International Trade Organization
(Hoekman and Kostecki 2001, 38). Yet it too had neither “direct oversight nor judicial
powers” (Barton et al. 2006, 48). The contracting parties’ lack of interest in an adminis-
trative apparatus, including a mechanism to monitor cheating, reflected a culture that
discouraged “the Secretariat form taking a hard line against its own members in any-
thing other than the most diplomatic language” (Daunton, Narlikar, and Stern 2012).
Although several “ad hoc surveillance schemes” existed (Daunton, Narlikar, and
Stern 2012), it was only in 1989 that the contracting parties agreed to establish a provi-
sional Trade Policy Review Mechanism (TPRM), tasked with reviewing their policies.
The inception of the WTO in 1995 made the TPRM permanent. Although this mecha-
nism might seem to enable the monitoring of compliance, its reports typically review
broad economic policies while taking note of “the consistency of certain trade measures
with multilateral principles” (Daunton, Narlikar, and Stern 2012). Members repeatedly
refused to make the reports an integral element of the dispute-resolution mechanism
(DRM). Even now, states subject to a review have the right to approve or revise it prior to
its publication.
Rather than being the centralized provider of information about cheating that
market-failure theory calls for, the GATT/WTO remains a “fire-alarm” system
(McCubbins and Schwartz 1984), one in which enforcement depends on member com-
plaints (Martin 2008). To help its members penetrate the camouflage that states often
use to cloak their deviations, the GATT authorizes them to request informal consulta-
tions with the alleged offender. Plaintiffs can also opt to request a panel of experts to
adjudicate a dispute. The ability of the GATT to do this requires the acquiescence of the
prospective defendant. Even if the latter agrees, it retains the option to block the imple-
mentation of a panel decision.
The creation of a new DRM was an integral element of the effort to establish the WTO
that began with the 1990 suggestion by Canada to replace the GATT (Hoekman and
Kostecki 2001, 49). Perhaps the major shift associated with its inception was to make
permanent the changes to the DRM that had occurred during the Uruguay Round.
Under the new rules, a defendant cannot veto panel formation. It is still the case,
 Joanne Gowa   25

however, that the enforcement of panel rulings depends on the willingness of the defen-
dant to change its behavior and of other states to punish it in the event that it refuses to
do so.11 Because the WTO DRM remains a fire-alarm system, it has a public-good prob-
lem of its own: each prospective plaintiff has an incentive to let another bear the costs of
litigation.
This very brief discussion highlights the role played in the creation of the GATT/
WTO of the lessons of the past that its architects had absorbed. In particular, its genesis
makes clear their determination to eliminate all vestiges of the discrimination that had
sullied trade between the wars. Yet from the outset it sanctioned the trade preferences
that some of its members had previously adopted. Moreover, as it evolved, discrimina-
tion, albeit of a different type, would reemerge. In the next section I review the existing
empirical evidence about the effects of the GATT/WTO on trade between its members.

The Effects of the GATT/WTO

The failure of the GATT/WTO to respond to the market-failure problem that an institu-
tion is designed to solve implies that it would exert a distinctly limited impact on trade
flows. Yet students of international relations long believed that the secular increase
in global trade that occurred after World War II was attributable at least partly to its
existence. While some earlier studies of other trade-related issues included a GATT
membership variable (e.g., Mansfield and Bronson 1997),12 it was not until 2004 that a
systematic empirical test of the organization’s impact on trade appeared.
In a seminal analysis, Rose (2004a) added to the covariates standard in the gravity
model two dummy variables to capture the effects of the GATT/WTO. The first assigns
a one to dyads that include two of its members; it is zero otherwise. The second takes on
a value of unity for country pairs in which one state is a GATT/WTO member and the
other is not. Rose examines trade between 1948 and 1999 using an ordinary least squares
analysis that includes year fixed effects.13 He finds that the postwar regime does not exert
any significant impact on trade between its members. Although the addition of country
or dyadic fixed effects produces a positive and significant coefficient on the GATT/WTO
variable, Rose argues that the impact is nonetheless “small compared to other effects
(e.g., regional trade associations), the long-term growth of trade, intuition, and the hype
surrounding the GATT/WTO” (2004, 204). In another publication Rose (2004b) sug-
gests that the GATT/WTO’s lack of impact is due to the fact that trade-policy liberaliza-
tion does not accompany state entry into it.
Dissent followed quickly. Gowa and Kim (2005) took issue with Rose’s assumption
that the postwar trade regime exerted uniform effects on its members’ trade. They
argue  that the principal-supplier rule enabled its largest members to dominate the
organization. A  legacy of US interwar treaty making, the “so-called chief-source, or
‘Principal Supplier, rule” arose in response to the US effort “to respect the MFN clause,
expand exports, and protect domestic industry” (Hawkins 1951, 81). Harry C. Hawkins,
26    Explaining the GATT/WTO: Origins and Effects

then head of the Trade Agreements Division of the Department of State, advised its
adoption because it would allow the United States to lower its tariffs on “particular prod-
ucts [only] to the country that … supplied the greatest proportion” of its imports of
them (Hawkins 1951, 81 n. 16). This would protect the administration against complaints
that “other countries were getting something for nothing.”14
The rule meant that each country would offer to reduce its tariff on a particular good
only if the state to which it extended the offer was the principal supplier of the item
to its market. Canada, for example, supplied 96 percent of the agricultural goods and
99 percent of the fish and lumber products on which the United States cut tariffs in the
1935 US-Canadian trade accord (Butler 1998, 135 n. 14). Tariff specialization—that is, the
“separating or ‘ex-ing out’ of portions of a tariff item for negotiating purposes”—served
the same end (Pomfret 1988, 6). Because Congress insisted on a principal-supplier
rule, postwar tariff cuts would take place on an item-by-item basis rather than across
the board as the British preferred. Successive trade rounds therefore opened with the
exchange of lists of products on which a state was willing to offer concessions and the
exports on which it hoped to lower foreign levies.
Although participants in the Kennedy Round formally adopted a linear approach—
that is, they agreed to cut their tariffs by a specified percentage across the board—bilateral
negotiations continued to play an integral role. As Hoda observes, the “really meaning-
ful discussions” at the Kennedy Round occurred bilaterally between “the major trading
partners.” Rather than “obviate the need for bilateral negotiation,” the linear approach
only gave participants “an additional tool” to use in their negotiations (Hoda 2001, 47).
Although a linear-tariff cutting process in principle increases the efficiency of negotiat-
ing, it was frequently held hostage to the reciprocity that “will always remain the key to
any trade bargain, whatever the formula for tariff reduction agreed upon” (Curzon and
Curzon 1976, 161).
Gowa and Kim argue that the principal-supplier rule effectively privileged the goods
the largest members of the GATT produced and exchanged. They test their idea by com-
paring trade flows in the years between the wars to trade after the GATT came into exis-
tence. They find that the only contracting parties that realized a significant expansion of
trade with each other between 1950 and 1994 were what they label core states: the United
States, Britain, France, Germany, and Canada. With the exception of some states that
had previously been members of the interwar trade blocs, the GATT exerted no other
significant effect on trade.
Subramanian and Wei (2007) also dispute Rose’s results. Distinguishing between
industrial countries and other states,15 they estimate a gravity model that includes
importer and exporter year fixed effects. They find that trade between industrial coun-
tries increases strongly and significantly as a consequence of their membership in
the GATT/WTO. In contrast, exports of developing countries to industrial countries
increase, but their imports from them do not. This is consistent with the steady drop in
tariffs on manufactured goods over time and the adoption of import-substituting indus-
trialization in many developing countries.
 Joanne Gowa   27

It is also consistent with the de facto restraints on trade that the core states in the
trade regime imposed. These included voluntary export restraints on cars and steel and
quantitative restrictions on agriculture and textiles, precisely the sectors in which devel-
oping countries were likely to enjoy a comparative advantage (Subramanian and Wei
2007, 153). GATT/WTO rules about agriculture in general were “shaped around” US
domestic politics (Davis 2003, 7). The organization also endorsed reverse discrimina-
tion: the special and differential treatment it adopted in 1965 released developing coun-
tries from the obligation to reciprocate the tariff cuts of other members. Designed to
block a Soviet effort to construct a trade regime under the auspices of the United Nations
Conference on Trade and Development (UNCTAD), the exemptions ruffled few feath-
ers: because most developing countries could not affect their terms of trade, other con-
tracting parties could afford to be indifferent to the conditions of their participation.
Tomz, Goldstein, and Rivers (2007) take a different tack in their analysis of the effects
of the regime. They argue that Rose underestimates the number of GATT members,
because he codes states as such only if they had completed the formal accession process.
States, they point out, also could and did become members at the discretion of their colo-
nial “parents.” Colonies that became independent could remain members on the same
terms they had previously enjoyed. “Provisional” members of the organization—that
is, states that had begun but not yet completed the formal accession process—were also
entitled to the benefits of membership that existing members agreed to provide them.
Including dyadic, country, and year fixed effects in their analysis, Tomz, Goldstein,
and Rivers find that the GATT/WTO increased trade between its members by about
72 percent.
The existing empirical literature about the postwar trade regime is virtually devoid
of efforts to examine systematically whether international politics influenced them.16
As Irwin, Mavroidis, and Sykes (2008, 197) point out, however, it was the onset of the
Cold War that induced the United States to press ahead with the GATT despite the ada-
mant refusal of the British to eliminate imperial preferences. The US State Department
opposed a delay over the IPS, fearing that it would empower the Soviets to “exploit fully
any such differences between [the] US and UK just as they are now trying to capitalize
on British weakness by increasing pressure throughout Eastern Europe and Near East”
(cited in Irwin, Mavroidis, and Sykes 2008, 197).
That the problems market-failure theory identifies are insufficient to explain the
practices of the postwar trade regime seems clear. From the perspective of the the-
ory, for example, US support for the creation of the European Economic Community
(EEC) and the extension of Marshall Plan aid made little sense. The EEC increased the
market power that its constituent states wielded, although it would take them several
years to converge on a joint bargaining position that would enable them to exploit it.
American support of the EEC reflected its willingness to endow other states with the
market power that transformed the opening of postwar markets into a PD game. Absent
its national-security interests in European integration, US support for the creation of a
rival trade bloc is inexplicable.
28    Explaining the GATT/WTO: Origins and Effects

It seems unlikely, then, that the postwar trade regime operated independently of the
political context in which it was embedded. To the extent that it did so, this implies that
states were treated differently as a function of the security ties between them, contraven-
ing the de jure nondiscrimination embedded in the GATT charter. In the next section
I examine the extent to which these interests, expressed as alliance ties, modify existing
empirical findings about the impact of the regime.

Effects of Political-Military
Alliances on GATT/WTO Trade

To measure the impact of alliance ties on trade, I use the specification of the gravity
model that has become the industry standard in empirical studies of trade in the eco-
nomics literature. The dependent variable is the annual imports of each state in a dyad
from the other. This creates two observations in each year for every country pair: one
records the imports of A from B, and the other records the imports of B from A. I exam-
ine the years between 1946 and 2003, because data about one key variable end in 2003.
I include directed-dyad fixed effects to take into account the unobserved
time-invariant heterogeneity that can occur at the country-pair level. I cluster the stan-
dard errors at the directed-dyad level to correct for heteroskedascity. I  also include
importer-year and exporter-year dummy variables to control for time-varying
state-level factors that affect trade (Anderson 2010, 24). These “multilateral-resistance”
terms proxy for yearly changes that affect a nation’s trade but that cannot be easily mea-
sured (Mathy and Meissner 2011, 18). They eliminate the omitted-variable bias that
might otherwise result, because bilateral trade depends not only on the costs two states
incur when they trade with each other, but also on the costs they incur when they trade
with other countries. As Felbermayr and Kohler (2010, 1446) note, the country-year
dummies are “perfect controls for country-specific policies that apply to all trading
partners.” Because they control for annual changes in some standard gravity-model
variables, it is not necessary to add separately other covariates that the gravity model
typically specifies, such as population and gross domestic product (GDP).
The estimates that this model produces measure the change in dyadic trade that
occurs between, for example, two states in a “treatment” group relative to the trade of
each state with states in the base group.17 The coefficient on the GATT/WTO variable,
then, estimates the effect of joining the regime on trade between two states relative to
the trade of each with nonmember states. Existing studies that include only dyad- and
year-specific effects instead estimate the shift in trade between states after they join the
regime.18 This risks conflating an increase in trade induced by the GATT/WTO with an
overall increase in world trade that could occur, for example, if the cost of transporting
goods from some subset of other states drops at the same time or if the volume of trade
in differentiated products increases.19
 Joanne Gowa   29

As in Gowa and Kim (2005), I distinguish three subsets of GATT/WTO members.

In the “core” group are pairs of major trading nations (i.e., Britain, Canada, France, the
United States, and West Germany). In the “other-industrial country” group are pairs
of industrial states other than those in the core group. Following the designation of the
International Monetary Fund (IMF), this group includes Australia, Austria, Belgium,
Denmark, Finland, Greece, Iceland, Ireland, Italy, Japan, the Netherlands, New Zealand,
Norway, Portugal, South Africa, Spain, Sweden, Switzerland, and Turkey.20 Finally, in
the “developing-country” group are all other member dyads that include at least one
developing country.
To construct the variable representing alliance ties, I use the information about alli-
ances that the project on Alliance Treaties and Obligations (ATOP) makes available
(Leeds et al. 2002). It includes data on coalitions between 1815 and 2003. I use these data
to create a dummy variable that assumes a value of one if when dyads include allied
states in a given year. I create three other variables that interact each of the GATT/WTO
group variables with the alliance term. These terms indicate pairs of states that are both
regime members and each other’s allies. For example, the variable labeled “core-state
allies” includes the pairs of core states that are each other’s allies.
About 16 percent of GATT/WTO member dyads include two allies. The aggregate
statistic conceals sizeable differences in the alliance composition of the three groups.
Almost 97 percent of country pairs in the core group include two allies. The correspond-
ing statistic is about 52 percent for other industrial country pairs and about 11 percent
for member dyads that include at least one developing country. If security concerns
do influence trade expansion within the postwar regime, the dispersion in the alliance
composition of the different groups suggests that the GATT/WTO may affect different
subsets of its members differently, not only on the basis of whether they are industrial-
ized but also on the basis of whether they are allies.
For data on trade and GATT membership, I use the information in the data set that
Tomz, Goldstein, and Rivers assembled. It includes information about bilateral trade
flows between states. It also codes PTA member states and the signatories of unilateral
trade treaties authorized under the GATT. These agreements, collectively labeled the
Generalized System of Preference (GSP), provide one-way preferential market access
for developing countries to the markets of an advanced industrialized country or coun-
tries. The developing countries that receive GSP preferences are chosen at the discretion
of the advanced country that extends them. I also include a dummy variable that takes
on a value of one if states are members of a currency union.
As a baseline, I first analyze the effects of the GATT/WTO on bilateral trade using
the variables that indicate the subsets of core states, other industrial states, and
developing-country members. I control for alliance membership separately. Table 2.1,
column 1, displays these results. It shows that neither pairs of core-group states nor
pairs of states that include one developing country realize a significant increase in their
trade relative to their trade with base-group states. Pairs of states in the other-industrial
group witness a marginally significant increase of about 10  percent in their trade
relative to trade between these states and those in the base group (p-value ≤ 0.08).
30    Explaining the GATT/WTO: Origins and Effects

Alliances raise trade between their members by about 16  percent, irrespective of
regime membership.
Next I examine whether alliances affect trade expansion in the postwar regime. The
results of this analysis are presented in Table 2.1, column 2. They show that core-state
allies trade about 15 percent more with each other than do other states in the core group,
a statistically significant difference.21 In the case of states that include at least one devel-
oping country, dyads that include allies trade about 42 percent more with each other
than do other states in the same group, a large and significant difference. Relative to
their trade with base-group states, allied pairs of states in the other industrial-country
group trade about 18 percent more with each other, a significant increase.
In other respects, the results in Table 2.1 conform to existing studies. States that are
GSP members actually experience a small but significant fall in their trade of about 8 per-
cent. This is not surprising given the constraints that the advanced countries impose on

Table 2.1  The Effects of the GATT/WTO and Alliances on

Trade, 1946–2003

Core-state pairs 0.01 –0.27

(0.17) (0.15)
Core-state allies 0.46***
(0.13)
Other industrial-country pairs 0.10 0.10**
(0.06) (0.06)
Other industrial-country allies 0.12**
(0.06)
County pairs with at least one 0.05 –0.00
developing country (0.03) (0.16)
Country pairs with at least one 0.39***
developing country: allies (0.04)
Alliances 0.15*** –0.06
(0.03) (0.04)
GSP –0.09*** –0.08***
(0.02) (0.02)
PTAs 0.24*** 0.23***
(0.02) (0.02)
Currency union 0.41*** 0.43***
(0.09) (0.09)
N 379476 379476

Note: dyadic and import-year and exporter-year fixed effects included;

standard errors clustered on directed dyads in parentheses; ** p-value
≤ 0.05, *** p-value ≤ 0.01; t-tests are two-sided.
 Joanne Gowa   31

the countries to which they grant GSP status. Often the rules of origin they specify are so
onerous that countries decline to claim preferential access altogether. Other trade agree-
ments have a positive and significant impact on trade: PTAs raise trade between their
members by about 26 percent, while members of a currency union realize an increase of
about 54 percent in their trade with each other.
The results in the second column of Table 2.1 show that the impact of the GATT/WTO
on trade does vary as a function of the alliance ties of its members, embedding discrimi-
nation in the postwar regime for large subsets of its contracting parties. Although the
discrimination at work in the GATT/WTO differs from that which motivated the for-
mation of the interwar trade blocs, the trade it created nonetheless differed across its
members in systematic ways prohibited by Article I of its 1947 charter.

Conclusion

Historical accounts of the origin of the postwar economic regime emphasize the intense
and prolonged diplomacy it involved. They amply document extended Anglo-American
debates about the rules and procedures that would govern trade and finance. In contrast,
the network of trade treaties that existed during the first golden age of globalization
emerged as a series of decentralized responses to the signing of the Cobden-Chevalier
Treaty in 1860. The interwar trade blocs also arose as a consequence of a decentralized
process of response to the advent of the Great Depression and the slow deterioration of
the gold standard.
Despite the long negotiations that produced the ITO charter and that of the GATT,
gaps quickly emerged between the de jure rules embedded in the charter of the
GATT and its operation. As Hoekman and Kostecki observe (1996, 38–39), the “fairly
complex and carefully crafted basic legal text was extended or modified by numerous
supplementary provisions, special arrangements, interpretations, waivers, reports by
dispute settlement panels, and Council decisions.” Some of these modifications, as well
as de facto practices, torpedoed the most basic provisions that had been so laboriously
embedded in the original GATT charter.
Among them was the importance states attached to increasing trade with their allies.
Because international politics prompted de facto discrimination among its members,
the GATT/WTO represented less of a clean break from the interwar system than is often
assumed. In each case, politics motivated discrimination. The “hub-and-spoke” trade
that developed between members of the IPS and the Reichsmark bloc bear the imprint
of the security interests that drove Britain and Germany, respectively, to create them
(Gowa and Hicks 2013). In the postwar era, security interests also drove a substantial
amount of trade within the GATT/WTO. Indeed, absent political interests, it is diffi-
cult to know what would have become of the GATT, given its lack of ability to monitor
the compliance of its members. As noted previously, alliances mitigate the monitoring
problem because they reduce the cost of cheating: while it is still true that the deviator
32    Explaining the GATT/WTO: Origins and Effects

gains at the expense of other states, some of its gains nonetheless accrue to the benefit
of the alliance as a whole, encouraging more postwar trade expansion than might have
occurred otherwise.

Notes
1. For a discussion of how a system of multilateral enforcement can trigger punishment in
this situation, see Maggi (1999).
2. Domestic politics can, of course, induce small states to impose trade barriers even if they
reduce overall welfare. In addition, the time-inconsistency problem free trade creates for
small states can lead them to join an institution to tie the government’s hands if that insti-
tution can punish its deviations (e.g., Levy 1999). As this engages domestic politics, I do
not discuss it further here.
3. A grim-trigger strategy requires each actor to defect in perpetuity whenever any defects.
A less severe punishment regime can also support cooperation (e.g., Gilligan and Johns
2012).
4. The actual effect of these treaties on trade is disputed. See Accominotti and Flandreau
(2008) for an argument that they had little impact on trade, and Lampe (2009) for a
different view.
5. Punishing all defections is not necessarily optimal, however. See, for example, Downs and
Rocke (1997); Koremenos (2001); and Milner and Rosendorff (2001).
6. Some claim that the distribution of gains does not matter (e.g., Gilligan and Johns 2012, 227),
based on work showing that war is not profitable under some conditions (Powell 1999). But
Powell does not rule out war and makes clear that the security effects of cooperation are irrel-
evant if and only if neither state allocates all its gains to its military sector. Otherwise, the “dis-
tribution of power would shift in favor of the state that obtained a larger gain.” Anticipating
this, its prospective partner “would refuse to cooperate in the first place” (1999, 76).
7. For an analysis of the impact of alliances on developing-country GATT members, see
Gowa (2010).
8. http://www.ssa.gov/history/acharter2.html (accessed June 17, 2013).
9. Although the fall in trade that followed the Great Depression has long been attributed
partly to the interwar blocs, recent work shows that they did not on the whole result in
trade diversion or contribute to the depression of trade more generally (Gowa and
Hicks 2013).
10. All references to the provisions of the GATT charter are to the language of the charter itself,
which can be found at wto.org/english/docs_e/legal_e/gatt47_02_e.htm#articleXXVI.
11. For an incisive analysis of the political processes involved in trade disputes, see Davis
(2012).
12. Mansfield and Bronson include a GATT variable in their analyses of PTA effects between
1960 and 1990. They find that it did not increase trade among its members (1997, 99).
13. Neither Rose nor his successors control for selection into the GATT/WTO, because no
viable instrument exists:  any variable associated with selection is also associated with
bilateral trade. Baier and Bergstrand (2007) advocate using dyad fixed effects, discussed
below, to take into account endogeneity. They assume that the source of the endogeneity is
time-invariant unobserved heterogeneity at the dyad level. See Hicks (this volume) for a
more extensive discussion of the gravity model.
 Joanne Gowa   33

14. John Leddy, oral history, http://trumanlibrary.org/oralhist/leddyj, 8–9.

15. They assign states to industrial status based on the classification that the IMF uses—that
is, those countries, listed below, that have International Financial Statistics (IFS) country
codes less than 200.
16. For an exception, see Kim (2010).
17. Some studies argue that estimating the effects of the GATT/WTO on both the exten-
sive and intensive margins of trade requires taking into account dyads with zero trade.
Felbermayr and Kohler (2010, 1451) note, however, that estimating effects at the exten-
sive margin rules out the use of a panel-data approach, because doubts exist about “the
reliability of time series variation on the extensive margin of trade.” Using time-series
cross-sectional analyses, they find that accounting for the zeroes does not change the esti-
mate of the trade regime’s effects. Here, as in earlier analyses to which we compare our
results, we rely on panel-data estimation.
18. Subramanian and Wei (2007) is an exception.
19. Estimating the model in the Tomz, Goldstein, and Rivers article (2007) but adding
importer and exporter year fixed effects produces a much smaller coefficient on the
GATT/WTO than they estimate. Formal members trading with each other realize about
an 8 percent increase in their trade. The corresponding statistic for nonmember partici-
pant dyads is about 17 percent. No significant increase in trade occurs between country
pairs that include one formal and one nonmember participant. Complete results are avail-
able from the author.
20. The IMF also designates as industrial states Luxembourg, San Marino, and Malta. But no
trade data exist for them. Data for Belgium exist as of 1997.
21. This is the sum of the coefficients on the core group, core-group allies, and allies (calcu-
lated as eb – 1).

References
Accominotti, Olivier, and Marc Flandreau. 2008. “Bilateral Treaties and the Most-Favored-
Nation Clause: The Myth of Trade Liberalization in the Mid-Nineteenth Century.” World
Politics 60 (2): 147–88.
Anderson, James E. 2010. The Gravity Model. Working Paper 16576. Cambridge,
Mass.:  National Bureau of Economic Research. Available at http://www.nber.org/papers/
w16576. Accessed 20 June 2012
Bagwell, K., and R. Staiger. 2002. The Economics of the World Trading System. Cambridge,
MA: MIT Press.
Baier, S., and J. H. Bergstrand. 2007. Do Free Trade Agreements Actually Increase Members’
International Trade? Journal of International Economics 71 (1): 72–95.
Barton, J., et al. 2006. The Evaluation of the Trade Regime: Politics, Law and Economics of the
GATT and the WTO. Princeton, NJ: Princeton University Press.
Butler, M. 1998. Visionary: Cordell Hull and Trade Reform, 1933–1937. Kent, OH: Kent State
University Press.
Curzon, G., and Curzon, V. 1976. The Management of Trade Relations in the GATT in
International Economic Relations of the Western World 1959-1971, Vol. 1. Politics and Trade,
edited by H. Oliver, 141-283. London: Royal Institution for International Affairs and Oxford
Press.
34    Explaining the GATT/WTO: Origins and Effects

Daunton, M., A. Narlikar, and R. Stern. 2012. The Oxford Handbook on the World Trade
Organization. New York: Oxford University Press.
Davis, C. 2003. Food Fights over Free Trade: How International Institutions Promote Agricultural
Trade liberalization. Princeton, NJ: Princeton University Press.
Davis, C. 2012. Why Adjudicate? Enforcing Trade Rules in the WTO. Princeton, NJ: Princeton
University Press.
Downs, G., and D. Rocke. 1997. Optimal Imperfection: Domestic Uncertainty and Institutions in
International Relations. Princeton, NJ: Princeton University Press.
Eichengreen, B. 1992. The Gold Standard and the Great Depression. New  York:  Oxford
University Press.
Felbermayr, G., and W. Kohler. 2010. Modeling the Extensive Margin of World Trade: New
Evidence on GATT and WTO Membership. The World Economy 33 (11): 1430–1469.
Findlay, R., and K. O’Rourke. 2007. Power and Plenty: Trade, War, and the World Economy in
the Second Millennium. Princeton, NJ: Princeton University Press.
Finnemore, M. 1993. International Organizations as Teachers of Norms: UNESCO and Science
Policy. International Organization 47 (4): 565–598.
Gilligan, M., and L. Johns. 2012. Formal Models of International Institutions. Annual Review of
Political Science 15: 221–243.
Gowa, J. 1995. Allies, Adversaries, and International Trade. Princeton, NJ: Princeton University
Press.
Gowa, J. 2010. Alliances, Market Power, and Postwar Trade:  Explaining the GATT/WTO.
World Trade Review 9 (3): 487–504.
Gowa, J., and R. Hicks. 2013. Politics, Institutions, and Trade: Lessons of the Interwar Era.
International Organization 67 (Summer): 439–467.
Gowa, J., and S. Kim. 2005. An Exclusive Country Club: The Effects of the GATT on Trade,
1950–94. World Politics 57 (4): 453–478.
Gowa, J., and E. Mansfield. 1993. Power Politics and International Trade. American Political
Science Review 87 (2): 408–420.
Grieco, J. M. 1988. Anarchy and the Limits of Cooperation—a Realist Critique of the Newest
Liberal Institutionalism. International Organization 42 (3): 485–507.
Hawkins, H. C., 1951. Commercial Treaties and Agreement: Principles and Practice. New York:
Rinehart.
Hoekman, B., and M. Kostecki. 2001. The Political Economy of the World Trading System. 2nd
ed. New York: Oxford University Press.
Hoda, Anwarul. 2001. Tariff Negotiations and Renegotiations Under the GATT and the WTO.
NY: Cambridge University Press.
Horn, H., and P. Mavroidis. 2001. Economic and Legal Aspects of the Most Favored Nation
Clause. European Journal of Political Economy 17 (2): 233–279.
Irwin, D., P. Mavroidis, and A. Sykes. 2008. The Genesis of the GATT. New York: Cambridge
University Press.
Jackson, John H. 1989. The World Trading System: Law and Policy of International Relations.
Cambridge: MIT Press.
Johnston, A. 2001. Treating International Institutions as Social Environments. International
Studies Quarterly 45 (3): 487–515.
Keohane, R. 1984. After Hegemony: Cooperation and Discord in the World Political Economy.
Princeton, NJ: Princeton University Press.
 Joanne Gowa   35

Kim, S. 2010. Power and the Governance of Global Trade: From the GATT to the WTO. Ithaca,
NY: Cornell University Press.
Koremenos, B. 2001. Loosening the Ties That Bind: A Learning Model of Agreement Flexibility.
International Organization 55: 289–325.
Koremenos, B., C. Lipson, et al. (2001). The Rational Design of International Institutions.
International Organization 55 (4): 761-80.
Krasner, S. 1991. Global Communication and National Power. World Politics 43: 338–366.
Lampe, Marcus. 2009. Effects of Bilateralism and the MFN Clause on International Trade:
Evidence for the Cobden-Chevalier Network, 1860-1875. Journal of Economic History 69 (4):
1012–1040.
Leeds, B., et  al. 2002. Alliance Treaty Obligations and Provisions, 1815–1944. International
Interactions: Empirical and Theoretical Research in International Relations 28 (3): 237–260.
Levy, P. 1999. Lobbying and International Cooperation in Tariff Setting. Journal of International
Economics 47 (2): 345–370.
Maggi, G. 1999. The Role of Multilateral Institutions in International Trade Cooperation.
American Economic Review 89 (1): 190–214.
Mansfield, E., and R. Bronson. 1997. Alliances, Preferential Trading Arrangements, and
International Trade. American Political Science Review 91 (1): 94–107.
Martin, L. 2008. Introduction. In Global Governance, edited by Lisa Martin, xi-xxi. Aldershot,
England: Ashgate Publishing.
Martin, L., and B. Simmons. 1998. Theories and Empirical Studies of International Institutions.
International Organization 52 (4): 729–757.
Mathy, G., and C. Meissner. 2011. Trade, Exchange Rate Regimes and Output Co-Movement:
Evidence from the Great Depression. NBER Working Paper No. 16925.
McCubbins, M., and T. Schwartz. 1984. Congressional Oversight Overlooked: Police Patrols
versus Fire Alarms. American Journal of Political Science 28 (1): 165–179.
Mearsheimer, J. 1994/95. The False Promise of International Institutions. International Security
19 (3): 5–49.
Milner, H., and P. Rosendorff. 2001. The Rational Design of International Institutions.
International Organization 55 (4): 829–857.
O’Rourke, K., and J. Williamson. 1999. Globalization and History:  The Evolution of a
Nineteenth-Century Atlantic Economy. Cambridge, MA: MIT Press.
Odell, J., and J. Eichengreen. 1998. The United States, the ITO, and the WTO: Exit Options,
Agent Slack, and Presidential Leadership. In The WTO as an International Organization,
edited by A. Kruger, 181-209. Chicago: University of Chicago Press.
Pomfret, R. 1988. Unequal Trade: The Economics of Discriminatory Trade Policy. New York: Basil
Blackwell.
Powell, Robert. 1991. “Absolute and Relative Gains in International Relations Theory. The
American Political Science Review 85 (4): 1303–1320.
Powell, R. 1999. Shadow of Power: States and Strategies in International Politics. Princeton,
NJ: Princeton University Press.
Rogowski, R. 1989. Commerce and Coalitions: How Trade Affects Domestic Political Alignments.
Princeton, NJ: Princeton University Press.
Rose, A. 2004a. Do We Really Know That the WTO Increases Trade? American Economic
Review 94 (1): 98–114.
Rose, A. 2004b. Do WTO Members Have More Liberal Trade Policy? Journal of International
Economics 63 (2): 209–235.
36    Explaining the GATT/WTO: Origins and Effects

Rosendorff, B., and H. Milner. 2001. The Optimal Design of International Trade Institutions:
Uncertainty and Escape. International Organization 55 (4): 829–857.
Subramanian, A., and S. Wei. 2007. The WTO Promotes Trade, Strongly But Unevenly. Journal
of International Economics 72: 151–175.
Tomz, M., Goldstein, J. and D. Rivers. 2007. “Do We Really Know that the WTO Increases
Trade? Comment. American Economic Review 97 (5): 2005-2018.
Waltz, K. 1979. Theory of International Politics. Reading, MA: Addison Wesley.
 Chapter 3

The Free Tra de  I de a

Gord on Ba nn e r m an

The success of the free trade idea in attaining scholarly credibility has led one distin-
guished writer to claim for it an “intellectual status unrivalled by any other doctrine
in the field of economics” (Irwin 1996, 217). The powerful ideological resonance of
free trade in nineteenth-century Britain led many historians and economists to fol-
low Rostow in adopting a Whiggish historical approach to describing the progressive
advance of nations toward commercial liberalism (1959, 7). Underpinning this inter-
pretation was a scholarly tendency, largely emanating from a preoccupation with the
scientific basis of economic theory, to describe free trade and open markets as optimal
policy for economic growth and wealth creation (Irwin 1996, 217). In fact, free trade
never attained unchallenged ideological hegemony, and recently scholars have more
readily accepted a growing number of exceptions to its applicability and the limitations
of the policy itself (Irwin 1996, 217–230). The development of a more critical, nuanced
approach, partly a consequence of the emergence of new academic disciplines, examines
commercial policy armed with a methodology concerned with the multiple contexts
of policy formation, processes, and outcomes (Krugman 1987, 131–132). This approach
considers the evolution of free trade more precisely in terms of historical context, while
assessing alternative models of political economy and economic development more
closely. Despite a moralistic assumption of natural justice and an economic assumption
of equality being made about it, free trade was, and often is, characterized as far from
fair and equal, and historically as a policy has seldom been uncontested.
This chapter seeks to trace the evolution of free trade ideas from earlier centuries by
reference to the development of economic theory in tandem with the economic organi-
zation and activities of nation-states within the context of prevailing political configura-
tions and imperatives. The organization of the chapter is primarily chronological and
seeks to chart long-term developments such as ideological influence and policy forma-
tion in time and space. This approach is accompanied by the thematic consideration of
how the free trade idea was received, assessed, and modulated by states with diverging
economic portfolios and sectors, within different stages of economic development. The
chapter takes a historicist and quasi-deterministic approach in concluding that the free
38   The Free Trade Idea

trade idea was highly mutable, as well as a contested and controversial policy option,
because of divergent socioeconomic contexts and the global imbalance of international
economic forces. Translated into policy terms, the ideas became entangled within the
international state system, and while commerce had always been an integral component
of foreign affairs and diplomacy, the cosmopolitanism of free trade ideas, although pow-
erful, could not supplant the vital diplomatic and foreign policy interests of nation-states
in the international policy space. Rather, commercial policy became annexed, and in
many cases subordinated, to these wider interests.

Defining the “Free Trade Idea”

What is meant by free trade? In theoretical terms precision is possible, because a body
of work exists widely acknowledged as containing the core theoretical principles of
the policy (Smith 2012; Ricardo 1821; J. Mill 1824; J. S. Mill 1848). Broadly speaking, the
absence of artificial legislative enactments imposing taxes on exports and imports, as
described by Adam Smith and the classical economists, as opposed to natural restric-
tions or technological limitations on commercial activity, accurately defines free trade.
What actually constitutes “free trade” in practice is more difficult to assess. Trade is
never really free so long as tariffs are levied for revenue purposes, and no nation has yet
been able to dispense with tariffs completely. Yet the incidence, extent, and nature of tar-
iffs have traditionally determined the nature of commercial policy, and tariff portfolios
have always been dependent on wider political and economic conditions, relationships,
imperatives, and traditions.
The term free trade emerged at the end of the sixteenth century, referring to com-
merce unhindered by exclusionary guild regulations or monopoly privileges. That was
“freedom to trade,” concerned with access to markets, as opposed to the application of
wider philosophical principles (Irwin 1996, 46). The first debate about free trade in the
modern sense concerned calico imports shipped to England by the East India Company,
which damaged the fledgling domestic cotton industry and raised calls for protection.
Import restrictions were imposed by a series of Calico Acts early in the eighteenth cen-
tury, allowing the English cotton industry to develop as an “infant industry” free from
foreign competition (Irwin 1996, 53). This series of events included many of the ele-
ments and arguments that have since informed the wide-ranging and intricate debate
on the purposes, nature, and application of commercial policy both within and between
sovereign states.
The idea of unrestricted trade developed from an early period, but before the sev-
enteenth century ethical, religious, and moral beliefs mainly determined economic
thought. The intellectual lineage from Aristotle and Aquinas to the moral philoso-
phers of the Scottish Enlightenment and Adam Smith was clearly, if unevenly, delin-
eated (Irwin 1996, 25). By the seventeenth century global trade expansion and the rise
of nation-states and empires led to a more favorable view of commerce and a desire to
 Gordon Bannerman   39

understand the economic distribution of goods and resource allocation facilities. The
natural law of Aquinas proved influential in facilitating invocation of the Creation as
a means of explaining global resource allocation. More widely, the role of religious
thought proved to be enduring, especially in Britain (Viner 1960, 48).
Religious thought underpinned the doctrine of “Universal Economy,” which
described the providential plan of unequal resource allocation and distribution
throughout the world “to promote commerce between different regions” and contained
implications of “design” that were influential from the seventeenth century onward
(Irwin 1996, 15). Dean Tucker viewed providential design in the distribution and allo-
cation of goods and resources throughout the world as conducive to “an Intercourse
mutually beneficial, and universally benevolent” (1763, 32). Into the nineteenth century,
“the unerring and clearly revealed will of a beneficent Creator”1 was transformed into
an antiprotectionist and especially anti–Corn Law argument against market interven-
tion. In 1841 Richard Cobden condemned the protectionist Corn Laws on the basis
that they were “an interference with the disposition of divine Providence as proved by
the revealed word of God and the obvious laws of nature” (Noble 2005, 105; Matthew
1997, 56). Yet the obvious metaphor for the mechanism driving resource distribution
and allocation, Smith’s “Invisible Hand,” was not presented by Smith as “a proof of either
providential order or design” (Milgate and Stimson 2009, 94). This omission repre-
sented a significant break with the past, as well as providing an indication of the future
trajectory of economic theory toward international trade.
Nevertheless, despite theoretical developments relating to the contours of inter-
national trade, policy choices remained primarily determined by revenue, employ-
ment, and developmental considerations, most notably in mercantilist doctrine (Irwin
2002, 46). Mercantilism was the predominant theory and practical operation of political
economy prevailing in Europe after the decline of feudalism. It was based on national
policies of accumulating bullion, establishing colonies and strong maritime forces, and
developing national industries, all as a means of attaining a favorable balance of trade
in commerce; that is, a country would produce and export more products than it would
import. The belief that a finite amount of global commerce meant it was vital to secure
a favorable trade balance was based on a static view of wealth creation and economic
growth, for the concept of increasing global wealth was “wholly alien” to mercantil-
ists (Heckscher 1969, 25–26). Moreover, in the seventeenth century most mercantilists
held the conventional contemporary view that self-interest governed human behavior
(Child 1681, 29). Mercantilists like Colbert and Child identified commerce as an aspect
of national power, with commercial rivalry an element of “perpetual combat in peace
and war among the nations of Europe as to who will gain the upper hand” (Economides
and Wilson 2002, 36).
Mercantilist doctrine was perhaps the first manifestation of “globalization,” and
despite a later tendency in Anglo-American historiography, it was not simply crude
protectionism. Most mercantilists were critical of commercial restrictions, and many
acknowledged that international commerce promoted domestic manufacturing, thus
displaying an understanding of reciprocal trading relationships. Some even disparaged
40   The Free Trade Idea

excessive domestic consumption as less important and less profitable than foreign trade
(Magnusson 1994, 125–128; Irwin 1996, 29–33). Even historians hostile to mercantilism
have acknowledged that it was more efficient in practice than in theory, and that mercan-
tilists did not pursue self-sufficiency at the expense of foreign trade (Fay 1928, 24). While
it was unclear whether import tariffs could reverse an unfavorable trade balance, by
reducing import penetration tariffs almost always increased domestic production and
employment in specific sectors. The formula for achieving this outcome—low import
duties on inputs and raw materials and high import duties on processed goods—was a
crucial point of convergence between mercantilism and modern industrial protection-
ism (Irwin 1996, 37–41; Viner 1937, 73).

Classical Political Economy

and Free Trade

Mercantilism was never an entirely coherent doctrine, and the work of David Hume
arguably represented the first substantial erosion of its central tenets, for Hume
demonstrated the fallacy of balance-of-trade arguments by describing how the
“price-specie flow” mechanism ensured regulation of distribution and exchange of
gold, facilitating balance-of-payments equilibrium (1987, 311–312). In Of the Balance
of Trade, Hume demonstrated how “real world” factors of labor and fixed assets were
prime determinants of commercial expansion and wealth creation, and in Of the
Jealousy of Trade he illustrated how commercial transactions promoted mutual pros-
perity and enhanced the prospects for long-term economic growth (1987, 316–317, 334).
Anticipating Smith, Hume criticized trade restrictions bred by national antagonisms
and argued for mutual expansionary gains for foreign trade and domestic manufac-
turing accruing from free trade (Irwin 1996, 70–72). Despite the rather unsystematic
dispersal of economic ideas throughout his works, Hume clearly exercised consider-
able influence on Smith.
The search for antecedents for Smith’s work has produced many ingenious, if not
always convincing, connections (Viner 1960, 54). Before The Wealth of Nations, there
had been considerable exaggeration of free trade arguments, with many proposals
advanced for particular political or economic objectives (Viner 1937, 92). The influence
of the physiocrats was limited, and though Smith met, through Hume, many leading
physiocrats, familiarity with the ideas of Quesnay, Mirabeau, and Turgot did not lead
to the doctrine substantially informing his work (Irwin 1996, 67). In fact, Smith criti-
cized the physiocrats for placing too much emphasis on domestic agricultural output,
and by thus imposing restraints on manufacturing and foreign trade, indirectly and
inadvertently undermining the prospects for agricultural prosperity. In effect, Smith
argued that by undervaluing the industry of the towns, the physiocrats made the oppo-
site mistake to that Colbert had made (Smith 2012, 662, 685).
 Gordon Bannerman   41

Other writers provided further grist to Smith’s mill. The Dutch merchant Pieter de
la Court, opposing the trading monopoly of the Dutch East India Company, lauded
free competition in Interest van Holland (1662) as central to national wealth and pros-
perity. Despite interested motives, this type of analysis linked previous definitions
of “freedom to trade” with broader conceptions of “free trade,” a linkage informing
Smith’s critique of mercantilism (Smith 2012, 629–630). A cross-current of influences,
beginning fundamentally from the Hobbesian relationship between self-interest
and society, influenced and informed Smithian political economy. Mercantilists
had understood the notion of “economic man” responding to the rational stimulus
of pecuniary gain, but Smith argued that they subordinated welfare gains from trade
to power considerations (Viner 1960, 56–57; Heckscher 1969, 93). Tucker described
the harmony of private and public interests in the economic sphere but considered
self-love to outweigh benevolence; Frances Hutcheson bridged the gap between moral
philosophy and economic behavior by arguing that empathy provided a moral sense
to temper self-interest.
Tucker had also convinced Hume that Britain could safely adopt freer trade without
risking its prosperity (Crowley 1990, 339–360). It was not a view anticipating hegemonic
power, for other nations could exploit their own product advantage, as well as using pro-
tective duties if necessary (Fieser 1999, 312–314). Assailing monopolies as early as 1749,
Tucker believed government should support commerce by a regulatory framework but
attacked the idea that a country could become prosperous by bankrupting its neighbors,
thus promoting a reciprocal, not adversarial, view of commerce (1763, 30–32; Williams
1972, 394). All of these policy positions can be found in Smith’s work.
Commercial policy debates generally revolve around how policies influence the most
effective use of primary factors of production to produce the greatest possible national
income. For Smith, mercantilist policies were primarily a misallocation of resources,
which endorsed and fortified restrictive practices (Gomes 2003, 4). Nevertheless, Smith
shared with mercantilist writers the view that defense was fundamental to wealth cre-
ation and “of much more importance than opulence,” and that there were other points
of convergence (Smith 2012, 454). Smith accepted a case for protective duties for “infant
industries” and defense industries, acknowledged the potency of retaliatory duties, and
considered countervailing duties appropriate in particular cases. He qualified this posi-
tion by questioning the efficiency and productivity of protected industries, but this did
not prevent protectionists later citing his claims in relation to industrializing economies
(Smith 2012, 442, 447, 452–462).
Despite linking mercantilism with international “discord and animosity,” Smith
rejected the notion of freer commerce promoting international harmony (Smith 2012,
484; Walter 1996, 5–28). Later political economists were more optimistic, with Ricardo
describing the moral exemplar of free trade uniting “by one common interest and inter-
course, the universal society of nations throughout the civilized world” (1821, 139). The
cosmopolitanism of free trade received greater currency in the nineteenth century from
Cobden, but earlier, although Hume and Smith had proclaimed the economic effi-
ciency of free commerce, they were essentially “economic liberals but political realists,”
42   The Free Trade Idea

convinced of the enduring qualities and patriotic attachment of citizens to nation-states

(Sally 2008, 40–41).
Smith promoted free trade in terms of economic efficiency rather than international
harmony, by demonstrating how international trade was an extension of the interna-
tional division of labor, with domestic surplus exchanged for foreign surplus promot-
ing mutual prosperity. The natural human instinct to exchange would be facilitated by
removing constraints and promoting commercial freedom (Smith 2012, 15, 22–26). With
market activity incentivized, individual self-interest would be “led by an invisible hand”
toward socially beneficial rather than socially unproductive activities (Smith 2012, 445).
Domestic altruism was thus extended, and Britain would gain by exchanging expen-
sive exports for cheap ones (Smith 2012, 453–454). Smith did not pronounce on factor
endowments, commodity composition, or stages of economic development, because he
did not consider these factors as in any way determinants of policy outcomes. For Smith,
the benefits of free commerce were universally applicable, and possessing ethical and
moral value in maintaining individual rights and liberties, went beyond commercial
transactions (Irwin 1996, 84–86; Viner 1960, 60).

Political Economy and

Commercial Policy

The coherence of Smith’s worldview led Hugh Blair to suggest that The Wealth of Nations
would become the “Commercial Code of Nations” (Mossner and Ross 1987, 187–190).
As policy issues of the 1770s and 1780s indicate, Smith’s ideas were slow to be adopted.
In Ireland, “free trade” agitation, an attempt to participate in the imperial trading sys-
tem, was a recrudescence of seventeenth-century debates over monopolies and unequal
treatment (Cunningham 1886, 285; Davis 1962, 290–291). Proposals to equalize customs
duties raised protests from British manufacturers concerned about Irish competition,
fears dismissed by Smith in 1779 on account of the underdeveloped Irish economy
(Bowden 1924, 655–674; Browning 1886, 309). Nevertheless, the manufacturers received
political support, indicating the limits to which even progressive politicians were pre-
pared to go when links between political control and economic dependency were
threatened (Kelly 1975, 562).
The 1786 Anglo-French commercial treaty presented a similar picture of commer-
cial policy subordinate to, or supplementary to, other policy issues. The commitment
of manufacturers to trade liberalization was questionable, for the predominant empha-
sis was on particular sectors rather than wider conceptions of commercial policy
(Williams 1972, 186). The manufacturers of 1786 were not forerunners of the Anti-Corn
Law League, for “it was no consistent economic theory or enlightened social ideal that
dictated the attitude of Wedgwood, Garbett, the Ironmasters, and the Manchester
manufacturers,” but rather self-interest, that fueled their support for a more liberal
 Gordon Bannerman   43

commercial relationship (Ehrman 1962, 48). Philosophical support for the treaty was
best expressed by Pitt and especially Eden, who, like Montesquieu and Kant, referred
to the mutual dependence established by trading relationships and anticipated Cobden
in linking commerce and peace.2 Reaction to the treaty was largely dictated by location
and sector, welcomed by southern wine-growers and by manufacturers in central and
western France, whose products did not compete directly with Britain, but unpopular
with northern industrialists facing competition from cheaper British imports (Sloane
1898, 214; See 1930, 308). As a model of reciprocal tariff reductions the treaty was sig-
nificant, but in the late eighteenth century commercial policy remained largely confined
within a mercantilist framework of protective duties and prohibitions.3
The Napoleonic and Revolutionary Wars further undermined commercial prog-
ress, with the commercial policy of combatants, the Continental System and Orders
in Council, used as strategic instruments of war (Williams 1972, 230). Wartime com-
mercial restrictions provoked middle-class radicalism in Britain and calls for commer-
cial freedom, but revolution and war hindered “the practice of an economic doctrine
which linked commerce and peace” throughout Europe (Howe 2002, 195). After 1815
quasi-autarkic policies were adopted by most states, and revenue concerns and notions
of economic independence, often based on the ideas of Hamilton, Fichte, and List, were
pervasive. Even progressive politicians were concerned that “habitual dependence on
foreign supply” was dangerous (Huskisson 1827, 8). The passing of the Corn Laws in
Britain crushed any hope that protectionism would be easily reformed (Howe 2002,
195–196). Wartime disruption also encouraged industrial protectionism. Most nota-
bly, building on the Hamiltonian thesis that technological advances made by industrial
nations meant greater efficiency and cheaper manufacturing than in emerging nations,
the United States adopted protective policies on the basis of industrial development
and revenue accumulation. Similarly, a succession of postwar tariff increases in France
imposed a highly protectionist tariff regime (Williams 1972, 223; Levasseur 1892, 23–24).
In an increasingly protectionist world, free trade entered the policy arena more force-
fully as a highly contested policy, but the translation of the idea from its scientific eco-
nomic base to the popular arena took time. During the war Smith’s work was widely
hailed as inaugurating a new era in political economy, with ideas that were “radical
solvents of the old consensus” (Howe 2002, 194). Dugald Stewart’s Life and Writings of
Adam Smith (1793) robustly promoted Smith’s theories and exerted a powerful influ-
ence on the next generation of classical economists (Milgate and Stimson 2009, 99–101).
Smithian political economy was now at least accompanied, if not superseded, by
Evangelical, Malthusian, and Ricardian variants (Howe 2002, 194). Most pertinently,
Smith’s theories were refined and augmented by Ricardo’s theory of comparative advan-
tage. It was a major theoretical advance, vital to understanding international trade pat-
terns. Ricardo illustrated that a country was best served by directing labor and allocating
resources to the sector where its comparative advantage was greater (Sally 2008, 36).
Those sectors with greatest relative efficiency advantage would export with the great-
est success, and the resultant trade would be mutually beneficial. Comparative advan-
tage meant every party to a transaction gained by specialization, and no country was
44   The Free Trade Idea

excluded, for purchasing goods from abroad provided foreign countries with necessary
purchasing power (Gourevitch 1986, 39; Irwin 2002, 28).
Ricardo was also concerned with domestic economic sectors, and his theory of
“diminishing returns” in agriculture was based on a critique of agricultural protection.
Influenced by Malthus but drawing different conclusions, Ricardo argued that since
higher corn prices meant higher wages for workers and lower profitability for capital-
ists, only landlords gained by differential rent; that is, the financial return they secured
from their possession of land, a return that varied according to the fertility of the land
and the productivity and yield. Ricardo thus concluded that “the interest of the land-
lord is always opposed to the interest of every other class in the community” (1821, 399).
This stance placed him in radical ranks in the Corn Laws controversy, and Marx, cit-
ing Ricardo’s Principles, argued that wage-cutting, consistent with Ricardo’s “Iron Law
of Wages,” meant prospective corn imports accounted for manufacturing support for
repeal (Hollander 2011, 232–233). Marx thus identified the class basis of support for free
trade, a relationship that Cobden later conceded was accurate.
Initially, support for free trade was sluggish. On presentation of the London mer-
chants’ petition in 1820 for removal of “injurious restrictions,” Ricardo was “astonished”
that free trade principles had taken so long to obtain mercantile support.4 Intellectual
support was forthcoming, notably from Benthamites and philosophic radicals incor-
porating free trade within a broader program of political and social reform, and the-
oretical advances were made by classical economists, notably McCulloch and Mill, in
claiming discovery and formulation of the “laws” of political economy. Mill compla-
cently described these “laws” as tenets of a progressive science, whose application would
destroy the remnants of feudalism and monopoly from an aristocratic political system.
The “people’s edition” of Mill’s Principles of Political Economy, published for working
men, sold more than ten thousand copies, lending credence to the popularity of free
trade in mid-Victorian Britain (Mill 1874, 278–279). By the mid-nineteenth century,
the theoretical case was well-stated and increasingly popularized as part of the liberal
reform agenda. With industrial, financial, and commercial strength facilitating the
potential for penetrating global markets, British commercial policy objectives pointed
toward an open trading regime.

Economic Development
and Commercial Policy

Why were free trade ideas more popular in some countries than others, and why were
those principles enacted as governmental policy in only a partial and limited way, albeit
in a reasonably large number of countries? While classical economists rigidly argued
that the principles were applicable to countries at all stages of economic development, in
practice fiscal pressures fueled protectionism in “new” countries and in older economies
 Gordon Bannerman   45

where tariffs were an important source of revenue. The influence, impact, and popular-
ity of free trade ideas varied according to the political and economic dynamics within
each nation, and broadly speaking it is possible to agree with a recent writer that while
free trade ideas “may have fostered liberalization in the long run … they explain little of
its timing” (Federico 2012, 181). We might also add “or its location.”
Fears raised during the Revolutionary Wars of greater manufacturing capability
among Britain’s rivals were not assuaged by British industrial preeminence after 1815,
for recurrent depressions and rapid American and European industrialization pointed
to the need for new markets. Government responded by cautiously ending monopo-
lies, relaxing restraints, and eliminating differential shipping duties through reciprocity
treaties (Williams 1972, 446). Gradual tariff reduction was enacted after 1822 with lower
import duties on raw materials, to facilitate a better competitive position for British
manufacturers. Britain also discarded prohibitions on artisans leaving the country and
exporting machinery, mercantilist-type measures that had aimed at defending indus-
trial advantages (Gomes 2003, 210–211).
After 1815 a dichotomous discourse of “free trade” and “protection” emerged, illus-
trating the contested nature of commercial policy (Matthew 1997, 135; Howe 2013). The
influence of political economists in political circles, with McCulloch in particular deliv-
ering lectures attended by ministers and officials, extended the popular appeal of politi-
cal economy, but it was a controversial policy area, and opposition, deeply felt if not
necessarily broadly based, attacked “the dismal science and its practitioners” and the
“cash nexus” that appeared to be enveloping Britain (Sack 1993, 182). Yet by the 1830s,
despite retaining protective duties and imperial preference, Britain was the leading
promoter of commercial liberalism. Peel’s 1842 budget reduced import duties on raw
materials, partly manufactured goods, and export duties, in totality representing a tri-
umph for export-led industries. An additional motive for enacting liberal measures was
the wish to “prevent prohibitive enactments in foreign countries” (Williams 1972, 452).
However, France, Prussia, Spain, Russia, and the United States resisted moves toward
freer trade, and critics increasingly questioned the value of reciprocity treaties, since
European nations competed with Britain in other markets (Williams 1972, 454–456).
Gladstone was exasperated by prevalent suspicion of British motives:
When we maintain the restraints we find in existence, they use our conduct as their
apology for inventing new ones. When we remove such restraints, they perceive only
a deeper plan for bringing about their ruin by cheap production, which requires of
them still more imperiously the multiplication of their repressive and prohibitive
enactments.
(Gladstone 1845, 60)
This intransigent stance was especially true of France and Prussia, which justified their
own protectionism by reference to British wine, timber, and corn duties, but a myriad of
cultural, political, and economic differences lay behind the broader policy divergence.
International market position to some degree determined policy options. The
“production profile” of nations, encompassing the timing and extent of industrial
46   The Free Trade Idea

development, the nature of international economic relationships, and institutional

and political configurations within states, convincingly describes the broad range
of policy influences (Gourevitch 1986, 55–60, 63–64; Wallerstein 1979, 66–73). It was
against this backdrop that the struggle took place between higher costs incurred by
consumers under domestic protection and the job losses suffered by producers under
a free trade regime of cheaper imports. Pareto’s theory of the role of concentrated
interests in maintaining protective duties demonstrates the importance of endog-
enous elements in policy making (Kindleberger 1975, 22). In developmental terms,
maintaining protective duties for most European states was greatly overdetermined.
Conversely, developmental theory identified free trade as the policy preference of
the most-advanced industrial nation in an open trading system (Irwin 2002, 144).
That nation was Britain, but as Corn Law repeal demonstrates, ideology as well as
economic interests, and the political institutions through which ideas and interests
were expressed, provide greater clarity for understanding the necessary conditions
for adopting free trade (Schonhardt-Bailey 2006). Most pertinently, the 1832 Reform
Act facilitated middle-class agitation for economic reforms, and parliamentary and
pressure-group activity was more intensively focused in Britain than elsewhere.
Corn Law repeal, the ending of imperial preference, and repeal of the Navigation
Acts reconfigured the British state from its mercantilist base toward unilateral free
trade. By 1860 import duties were levied only on noncompeting products, and protec-
tive duties were effectively removed from Britain’s fiscal system. Tariffs remained on
high-yielding, noncompeting imports for revenue purposes. In light of this taxation
structure, the “optimistic” view of free trade finance promoted by Matthew and Howe
in terms of its contribution toward a moralistic minimal “knaveproof ” state is per-
haps overstated (Matthew, 1997, 171; Howe, 1997, 113).
As the adoption of free trade was dependent on economic development, human
agency, ideology, and political culture, it was certainly not easily transferable. The
expectation that repeal would stimulate other nations to reduce tariffs and exploit their
comparative advantage in noncompeting products to British manufactures stemmed
from a belief that the Corn Laws had inhibited commerce and forced European nations
to industrialize (Kindleberger 1975, 33–34). Even Cobden regretted European industrial
development, fueled by “the fostering bounties which the high-priced food of the British
artisan has offered to the cheaper-fed manufacturer of those countries” (1878, 65).
Did such attitudes amount to pursuit of a hegemonic strategy? In Europe, free trade
was often viewed inseparably from perceptions of British hegemony, and limited accep-
tance of free trade presents a convincing refutation of hegemonic power (Nye 2007, 98).
The less-obvious “second face” of hegemony, using international market power to influ-
ence prices as a means of influencing economic behavior and attitudes, was largely
absent after 1846 (James and Lake 1989, 8). Altering investment patterns and factor and
sector returns abroad to “ensure complementary production and the free exchange of
primary goods for British manufactured goods” may have motivated classical econo-
mists and Board of Trade officials, but the idea of a “conscious motivation” for hege-
monic leadership appears misplaced (James and Lake 1989, 27). Promoting free trade
 Gordon Bannerman   47

ideas should not be conflated with the conceptually distinct notion of hegemony. This
is especially true given that despite a shift toward lower tariffs between 1846 and 1873,
primarily to advance industrial expansion, most European nations remained generally
opposed to radical reductions (Irwin 1996, 98).
As Cobden discovered on his European tour of 1846–1847, the constellation of eco-
nomic interests and political forms meant that ideological commitment to free trade
was limited and free trade movements largely ineffectual, at least partly owing to the
strength of protectionist financial and industrial interests within monarchical and
authoritarian regimes. In France, even after the 1848 Revolution, the political system
remained highly conservative, with the Chamber of Deputies dominated by protec-
tionist interests. Despite the efforts of Bastiat and Chevalier and calls from industri-
alists for lower input costs, French politicians, from Proudhon to Thiers, denounced
free trade as a means by which “perfidious Albion” ensured European subservience
to Britain (Cain 1979, 243; McKeown 1983, 73–91). Similarly, in the 1830s Palmerston
doubted the commitment of the Zollverein to freer commerce, and by 1847, having wit-
nessed the preferential treatment given to Belgian iron, he was convinced it existed
“for the purpose of excluding by high duties the importation of British manufactures
into Germany” (Howe 1997, 81). Austria resisted further tariff reductions because it was
not satisfied with the subordinate role of supplying raw materials (Williams 1972, 211).
Successive parliamentary inquiries into foreign commercial policy revealed a similar
pattern of organized protectionist strength, widespread perceptions of protection as a
“national” policy, and resistance to free trade allied to deep-seated hostility to British
economic dominance.
The necessity Britain was under in making the 1860 Anglo-French commercial treaty
indicates that “the hegemonic story is virtually turned on its head,” for the treaty con-
tradicted any notion of hegemonic power (Howe 1997, 94–95; Nye 2007, 109). The inter-
locking network of European treaties that followed, underscored by tariff reductions
and most-favored nation (MFN) clauses, was not of Britain’s making and represented
a return to bargaining and negotiation. For some historians, the treaty system was a
paradigm shift, appearing to represent the “cornerstone of a new international trading
system” (O’Brien and Pigman 1992, 100). However, attempts to provide a comprehen-
sive explanation for this movement to the point of “near free trade” have fallen back on
emphasizing specific and unique, and crucially temporary, national factors (Federico
2012, 166).
While undoubtedly a facet of the reality of international commercial relations, British
hegemony was more often not coercively exercised or applied. The “soft power” of per-
suasion and influence was used, with variable success, by British government depart-
ments connected with commerce, such as the Board of Trade and the Foreign Office.
Wallerstein’s broad claim that “non-hegemonic” powers, fearful of larger economic
entities, adopted protection, while apparently historically accurate and contemporarily
relevant, unnecessarily inserted hegemonic language into the discussion (Wallerstein
1979,  19). Much of the association made between free trade and British hegemony
emanated from those, like Thiers and Bismarck, who skilfully reworked traditional
48   The Free Trade Idea

prejudices for their own ends. Hegemonic power was a useful concept for the enemies of
free trade, and propagation of the idea often acted as a justificatory ideological weapon
for protectionism.

Alternatives to Free Trade

The most influential alternative to free trade came from Friedrich List, whose vehe-
ment opposition to “British” free trade was driven by British economic dominance (List
1856, 437). The “infant industry” argument promoted by List, of tariffs fostering indus-
trial development, was based on historical judgments rather than economic analysis
(Irwin 1996, 124–126). As the relative absence of the theory in Ireland demonstrated, it
was a theory predicated on a preexisting but nascent industrial sector (Collison Black
1960, 140–143). The theory was not new. In the eighteenth century Tucker identified the
relationship between protection and economic development (Semmel 1965, 762). More
famously, Smith and Mill both cautiously stated that the theory was not inconsistent
with the main tenets of classical political economy (Irwin 1996, 118–119). Such quali-
fied acceptance, easily interpreted as approval, combined with scholarly expositions by
Torrens and Mill demonstrating how tariffs improved a country’s terms of trade, pro-
vided intellectual Delete this respectability for protectionist political economy (Irwin
1991, 202).
While Listian analysis provided ideological impetus, the intellectual challenge to freer
trade was fundamentally driven by the “great depression” of 1873–1896. In identifying
causes for the damage inflicted by cheaper imports, falling prices, and unemployment,
freer trade was an obvious target. Across Europe, powerful coalitions were constructed
on the basis of economic grievances, incorporating reconciliation of hitherto opposing
economic sectors, particularly in Germany (Rosenberg 1943, 65–66; Schonhardt-Bailey
1998, 293–296). Confrontational in form, with incisive arguments, economic nation-
alism represented for its leading ideologues not just protectionism but a redefinition
of the duties and responsibilities of the state (Viner 1969, 64; Heckscher 1969, 93–94).
Historical economists developed historical and inductive methods in attacking ortho-
dox political economy for failing to address wider national interests, aside from wealth
and individual self-interest (Cunningham 1925, 738). The influence of Listian histori-
cism was transparent in allusions made to British protectionism of earlier centuries and
the subsequent inclination of foreign countries “to imitate the steps by which England
attained to greatness,” with a corresponding reluctance to embrace policies they consid-
ered detrimental to their national interests (Cunningham 1925, 869).
Economic nationalism was also aided by a flaw in Ricardian political economy, for
there were never any illusions harbored that comparative advantage would ensure even
distribution.
The disproportionate gains secured by developed economies provided economic
nationalists with a powerful rallying cry. Nor was protectionist revival geographically
 Gordon Bannerman   49

confined. Australian protectionism drew on populist elements of colonial nation-

alism, harnessing resources, providing employment, and raising revenue, to pro-
mote a “national” economic policy free from imperial control (Gourevitch 1986, 110).
Similarly, after a period of lower American tariffs after 1846, protectionist resurgence,
propelled by Carey’s virulent anti-British rhetoric, enlarged on Hamiltonian themes
of self-sufficiency and independence. Developmental theory has been used to explain
why Britain, as export leader in an open trading system, retained free trade in the 1870s
(Gourevitch 1986, 77). Similarly, Wallerstein described adoption of free trade as “the
definitive mark of the self-confident leader in a capitalist market system” (1979, 32).
While rational choice interpretations describe the economic imperatives, ideology was
also crucial in converging with and articulating national economic interests. In Britain,
free trade was a Weltanschauung, a worldview revolving around consumer interests
and free exchange (Trentmann 1997, 73). Economic theory and the intellectual domi-
nance of free trade ideas in scholarly and political circles made a significant contribu-
tion to this process. Significantly, List’s work was not translated into English until 1885,
more than forty years after its publication in Europe (Gourevitch 1986, 83).
Improvements in international transport and communications and technological
developments in industry and agriculture meant Britain faced significant import com-
petition in the domestic market and foreign tariffs overseas. The practical operations
and realities of the international economy were increasingly counterposed to abstract
theory, and “unequal” or “one-sided” free trade emerged as a critique of a set of dys-
functional commercial relationships.5 However, reform was problematic because any
reimposition of food taxes, even for fiscal or imperial reasons, was politically difficult
(Brown 1943, 68). However, in concentrated domestic industries like silk and sugar, the
applicability of comparative advantage and the existence of a “self-regulating mecha-
nism” modulating displacement of labor were increasingly questioned (Trentmann
1997, 90; Green 1995, 109–110; Marrison 1996, 110–112).
At the beginning of the twentieth century the fiscal landscape displayed a familiar
pattern, with the trenchant protectionism of Russia and the United States accompanied
by less severe protectionist regimes in France, Germany, and Austria (Bastable 1923, 106;
Matthew 1997, 569–570). The collapse of the treaty system as a mode of extending freer
commerce meant policy outcomes were increasingly driven by “economic circum-
stance” and international position. Internationally, colonial, economic, and national
rivalries undermined freer trade; domestically, the emergence of collectivism and
socialism had a similar effect by raising fiscal pressures (Gourevitch 1986, 123).
The ideological contours were changing, with the relationship among free trade,
wealth, and welfare assuming greater importance, and support for free trade (and oppo-
sition to protectionism) increasingly based on different policy assumptions. Socialist
thought, giving precedence to the collective standard of living over individual freedom,
denied the primacy of consumerism, claiming that free trade was too narrowly mate-
rialistic in excluding important societal considerations (Trentmann 1996, 221–223).
Similarly, the labor movement promoted a political economy model of regulation and
rights, influenced by atavistic notions of “moral economy” and modern economic
50   The Free Trade Idea

planning. Politically, accelerating democratization shifted the policy debate from

wealth to welfare, and intellectually, from individualism to collectivism (Trentmann
1997, 82, 97–98). While there was clear erosion of the economic case for free trade,
Cobdenite internationalism still had advocates, with Hobson fiercely denouncing pro-
tectionism and imperialism as “enemies of international morality” destroying “the free
self-expression and intercourse of nations” (1903, 374). Cobden’s vision of a produc-
tive, peaceful, and prosperous “industrial” society, based on Smithian and Spencerian
notions, in which “the aggressive instincts of men were sublimated in work,” remained
influential (Cain 1979, 230). Nowhere was this so more than in Angell’s The Great Illusion
(1911), in which the shift from militant to industrial societies and forging of economic
interdependence by commerce was taken as meaning war was virtually impossible, a
view completely undermined by the outbreak of war in 1914 (Howe 1997, 296).

Decline and Reformulation

of the Free Trade Idea

The classical case for free trade rested on the assumption of full employment and
elastic labor markets, but once undermined, the economic case was weakened.
Correspondingly, from the 1930s onward greater state intervention inevitably meant
closer political control of economic policy, with economic planning and collectivist
social policies the antithesis of free trade ideology (Howe 1997, 273). Wartime develop-
ments changed the intellectual and economic landscape, and the absence of European
manufacturing products in third markets motivated “infant industry” protectionism in
Japan, Australia, and India. After 1918 tariffs were raised by new European states strug-
gling to achieve economic stability, and Western European economies combined inter-
vention, regulation, and fiscal readjustment as a means of restoring equilibrium.
Restoring pre-1914 economic structures was the fundamental aim of many govern-
ments, but comparative responses to reconstruction were largely determined by pre-
vailing financial and monetary constraints (Gourevitch 1986, 127–128). Financial
instability was exacerbated by the restored gold standard proving unable to operate a
balance-of-payments correction mechanism on account of higher tariffs and wage and
price rigidity. In pursuit of a better balance between sectors, import tariffs were raised
to offset diminishing domestic demand, while seeking protection from external volatil-
ity (Landes 1969, 392). The 1930s witnessed a loss of faith in classical political economy,
fatally undermined by economic depression and political turmoil. The international
financial crisis fundamentally altered financial and economic structures and practices.
Extensive state intervention in the form of neo-orthodox policies of economic plan-
ning and regulation protecting domestic producers and satisfying domestic demand
represented a departure from orthodox deflation (Landes 1969, 399). Symptomatic of
these new realities was the highly protectionist American tariff, which exacerbated the
 Gordon Bannerman   51

decline in domestic incomes and production and prompted a debilitating retaliatory tar-
iff cycle. Despite the implementation of New Deal ideas and the 1934 Reciprocal Trade
Agreements Act, the United States showed little inclination to lower tariffs (Gourevitch
1986, 160).
Keynesian demand stimulus accelerated after 1945, but in the 1930s Germany,
Sweden, and America all practiced it to some degree as part of a more pronounced pro-
tectionist, insular, and autarkic set of policy preferences (Gourevitch 1986, 142). Others
remained more cautious. Britain retained orthodox deflation until 1931, when financial
crisis forced devaluation and a general tariff. The reimposition of protective duties and
imperial preference in 1932 destroyed the notion of a self-regulating, free-market econ-
omy (Howe 1997, 296). The tenets of classical trade theory, factor mobility, and price
flexibility were held inoperative under high unemployment, and even Keynes, with the
zeal of the convert, enthusiastically endorsed protection (Irwin 1996, 200). France fol-
lowed the same path, via a demand stimulus interlude in 1936–1937, but financial crisis
converged with historical policy predilections based on economic composition, for the
French economy was predominantly small scale, labor intensive, and protectionist, and
opposed to an open, internationalist strategy (Gourevitch 1986, 156).
Tariffs continued on an upward trajectory throughout the 1930s, along with import
quotas and controls, potentially more damaging than tariffs in being less influenced
by price mechanisms, within economic recovery programs. Wartime intervention
only exacerbated existing trends, and after 1945, with the prospect of a liberal interna-
tional order far removed from contemporary political realities, a formal international
and institutional commitment to freer commerce was made with the 1947 General
Agreement on Tariffs and Trade (GATT), and in the following decade with formation of
the European Common Market (Irwin 2002, 161–165). In containing protectionist pres-
sures and promoting reciprocal tariff reductions, the GATT has been a practical way of
advancing freer trade, for wider policy influences continued to be variable in incidence
and impact and have necessarily required constant attention, modification, and negotia-
tion on an international basis.6
In broad terms, the pace of ideas and pressure of events outstripped the principles
and practices of the mid-Victorian world. Historians may accurately refer to a “free
trade era” when classical political economy met with less opposition rather than enthu-
siastic acceptance, but changing patterns of global political authority and economic
power meant continual reconfiguration and renegotiation. Nevertheless, the concept
of “globalization,” fueled by technological and communications revolutions rather
than ideological consensus, has revealed some familiar debates, particularly the rela-
tionship between policy formation and outcomes and economic development. In place
of “hegemonic” Britain stands “the West,” and in place of emergent industrial econo-
mies stand less-developed countries, where cheap labor supply and rudimentary facili-
ties and resources reflect a wider disparity in wage rates and living standards. Closer
identification of commercial policy with national and regional development, diversity,
and sustainability, underpinned by ethical and moral imperatives, has been a significant
development, resulting in the traditional protectionism of import-competing domestic
52   The Free Trade Idea

industries being supplemented by nongovernmental organizations (NGOs) and anti-

capitalist protestors in broad-based attacks against open markets (Irwin 2002, 225–228).
The pursuit of free trade remains controversial, for the policy continues to attract crit-
icism on the basis that it is a tool of the major powers to ensure global market domina-
tion. Even avid proponents of laissez-faire economics, while considering the free market
a great impersonal force in resource allocation, price determination, and product dis-
tribution, concede that it was never universally applicable (Viner 1960, 64). Based on
short-term static and long-term dynamic gains of comparative advantage, economies of
scale, and technology transfer, a strong case for economic efficiency underpinned free
trade. However, economically efficient outcomes, while constituting the main consid-
eration for previous generations, are now rarely considered alone. Ethical policy begins
for many antiglobalization thinkers and activists by identifying the imperfections of the
free-market Smithian Invisible Hand and the inefficient economic and unjust political
outcomes it produces (Stiglitz 2002, 4–10).

Conclusion

Even at the height of its popularity, free trade was a contested policy, and by the twen-
tieth century it no longer appeared sufficiently robust to deal with the demands of col-
lectivism and social democracy. Untrammeled, free-market operations had long been
deprecated, and after 1945 open markets coexisted with “distributive justice” in mixed
economies (Viner 1960, 68). Despite the exalted intellectual status accorded to classical
political economy, alternative models always existed. The idea of the universal appli-
cability of free trade was born out of the political culture of an expanding capitalist
economy in the late eighteenth and early nineteenth centuries, but ultimately the policy
failed to fully encapsulate the economic experience of less-developed nations. While it
would be reductionist to attribute too much influence to import/export orientation, it
clearly contributed to the range of exogenous and endogenous factors of economic com-
position, political culture, and perceptions of place within the international economic
order influencing commercial policy.
Ricardian political economy became increasingly divorced not only from Smith’s
original vision but also from practical economic and political experience, for complex
welfare and employment issues in an age of political democracy posed unprecedented
fiscal and political demands. Even if the economic efficiency argument retained validity,
it is now not the only or even primary factor in the policy space. Not only has free trade
“irretrievably lost its innocence,” but it “can never again be asserted as the policy that
economic theory tells us is always right” (Krugman 1987, 131–132). In fact, free trade had
never achieved this status. There were always dissenters from the policy and the philo-
sophical view of the world from which it emerged.
Commensurate with the internationalization of governance and regulatory authority,
free trade has now cast off any utopian, unilateral, and voluntarist associations, but the
 Gordon Bannerman   53

complexity of the international economy makes the modern case for free trade appear
“too narrow and mechanical” and even “a little unreal” (Sally 2008, 47–48). Smith’s
observation that it would be “absurd” to expect the restoration of free trade has great
contemporary resonance (Smith 2012, 460). As in 1776, free trade remains an aspiration
rather than a reality, and the variable success of the free trade idea reveals the disparity
between theory and practice in the domestic and international commercial policy envi-
ronment (Irwin 2002, 161–164).
Free trade remains a powerful idea, with strong associations with “progressive”
ideas of freedom, liberty, and democracy, but despite elements of ideological continu-
ity, the greater mobility of capital and labor, fueled by technological innovation and
combined with supranational political authority and multinational business organi-
zation, has complicated global trading patterns. The free trade/protection dichotomy
between nations has been supplanted by a complex configuration of national, regional,
and sectorial models, with the result that notions of reciprocity and perceived vital
national interests largely determine commercial policy. The cosmopolitanism of the
free trade idea, while not totally discarded in its conception of the universal benefits
bestowed by commercial liberalization, has been replaced by a more nuanced, practi-
cal, and hard-nosed acceptance of trade liberalization based on and subject to negotia-
tion and bargaining.

Notes
1. The Anti-Corn Law Almanack for 1842 (Manchester: J. Gadsby), 27.
2. William Eden to Robert Liston, September 27, 1786, Liston Papers, National Library of
Scotland, NLS MS 5545 fols. 84–85.
3. Ralph Woodford to Robert Liston, December 29, 1786, Liston Papers, National Library of
Scotland, NLS MS 5545 fols. 151–152.
4. Commercial Restrictions—Petition of the Merchants of London, HC Debates, Hansard,
May 8, 1820, vol. 1. c. 191.
5. “If ever the battle of Free Trade has to be fought over again, it will be decided, not upon its
abstract merits, but by the encounter of sundry conflicting interests.” The Graphic, no. 939,
November 26, 1887.
6. For the origins of GATT, see the chapter in this volume by Joanne Gowa and for the
contours and configuration of trading relations instigated by GATT, see the chapter by
Raymond Hicks.

References
Bastable, C. F. 1923. The Commerce of Nations. 9th ed. London: Methuen.
Bowden, W. 1924. The Influence of the Manufacturers on Some of the Early Policies of William
Pitt. American Historical Review 29 (4): 655–674.
Brown, B. H. 1943. The Tariff Reform Movement in Great Britain, 1881–1895. New York: Columbia
University Press.
54   The Free Trade Idea

Browning, O. 1886. Adam Smith and Free Trade for Ireland. English Historical Review 1
(2): 308–311.
Cain, P. J. 1979. Capitalism, War and Internationalism in the Thought of Richard Cobden.
British Journal of International Studies 5 (3): 229–247.
Child, Josiah. 1681. A Treatise Concerning the East India Trade. London: T.J.
Cobden, Richard. 1878. Speeches on Questions of Public Policy. Edited by J. Bright and J. E. T.
Rogers. London: Macmillan.
Collison Black, R. D. 1960. Economic Thought and the Irish Question, 1817–1870. London:
Cambridge University Press.
Crowley, J. E. 1990. Neo-Mercantilism and the Wealth of Nations: British Commercial Policy
after the American Revolution. Historical Journal 33 (2): 339–360.
Cunningham, W. 1925. The Growth of English Industry and Commerce in Modern Times.
Cambridge, UK: Cambridge University Press.
Cunningham, W. 1886. The Repression of the Woollen Manufacture in Ireland. English
Historical Review 1 (2): 277–294.
Davis, R. 1962. English Foreign Trade, 1700–1774. Economic History Review 15 (2): 285–303.
Economides, Spyros, and Peter Wilson. 2002. The Economic Factor in International
Relations: A Brief Introduction. London: I.B. Tauris.
Ehrman, John. 1962. The British Government and Commercial Negotiations with Europe,
1783–1793. Cambridge, UK: Cambridge University Press.
Fay, C. R. 1928. Great Britain from Adam Smith to the Present Day: An Economic and Social
Survey. London: Longmans, Green.
Federico, Giovanni. 2012. The Corn Laws in Continental Perspective. European Review of
Economic History 16 (2): 166–187.
Fieser, James, ed. 1999. Early Responses to Hume’s Moral, Literary and Political Writings II.
Bristol, UK: Thoemmes Press.
Gladstone, W. E. 1845. Remarks upon Recent Commercial Legislation. London: John Murray.
Gomes, Leonard. 2003. The Economics and Ideology of Free Trade:  A  Historical Review.
Cheltenham, UK: Edward Elgar.
Gourevitch, Peter. 1986. Politics in Hard Times:  Comparative Responses to International
Economic Crises. Ithaca, NY & London: Cornell University Press.
Green, E. H. H. 1995. The Crisis of Conservatism: The Politics, Economics, and Ideology of the
British Conservative Party, 1880–1914. London: Routledge.
Heckscher, Eli F. 1969. Mercantilism. In Revisions in Mercantilism, edited by D. C. Coleman,
19–34. London: Methuen.
Hobson, J. A. 1903. The Inner Meaning of Protectionism. Contemporary Review 84: 365–374.
Hollander, Samuel. 2011. Friedrich Engels and Marxian Political Economy. Cambridge,
UK: Cambridge University Press.
Howe, Anthony. 1997. Free Trade and Liberal England, 1846–1946. Oxford, UK: Clarendon Press.
Howe, Anthony. 2013. Popular Political Economy. In Languages of Politics in Nineteenth-
Century Britain, edited by. David Craig and James Thompson, 118–141. London: Palgrave
Macmillan.
Howe, A. 2002. Restoring Free Trade:  The British Experience, 1776–1873. In The Political
Economy of British Historical Experience, 1688–1914, edited by D. Winch and P. K. O’Brien,
192–213. Oxford: Oxford University Press.
Hume, David. 1987. Essays Moral, Political and Literary. Rev. ed. Edited by Eugene F. Miller.
Indianapolis, IN: Liberty Fund.
 Gordon Bannerman   55

Huskisson, W. 1827. A Letter on the Corn Laws, by the Right Hon. W. Huskisson, to One of His
Constituents, in 1814. London: James Ridgway.
Irwin, D. A. 1991. Retrospectives: Challenges to Free Trade. Journal of Economic Perspectives 5
(2): 201–208.
Irwin, Douglas A. 1996. Against the Tide:  An Intellectual History of Free Trade. Princeton,
NJ: Princeton University Press.
Irwin, Douglas A. 2002. Free Trade under Fire. Princeton, NJ: Princeton University Press.
James, S. C., and D. A. Lake. 1989. The Second Face of Hegemony: Britain’s Repeal of the Corn
Laws and the American Walker Tariff of 1846. International Organization 43 (1): 1–30.
Kelly, P. 1975. British and Irish Politics in 1785. English Historical Review 90 (356): 536–563.
Kindleberger, C. P. 1975. The Rise of Free Trade in Western Europe, 1820–1875. Journal of
Economic History 35 (1): 20–55.
Krugman, Paul. 1987. Is Free Trade Passé? Journal of Economic Perspectives 1 (2): 131–141.
Landes, David. 1969. The Unbound Prometheus. Cambridge, UK: Cambridge University Press.
Levasseur, E. 1892. The Recent Commercial Policy of France. Journal of Political Economy 1
(1): 20–49.
List, Friedrich. 1856. National System of Political Economy. Philadelphia: J. P. Lippincott.
Magnusson, Lars. 1994. Mercantilism:  The Shaping of an Economic Language. London:
Routledge.
Marrison, Andrew. 1996. British Business and Protection, 1903–1932. Oxford:  Oxford
University Press.
Matthew, H. C. G. 1997. Gladstone, 1809–1898. Oxford: Clarendon Press.
McKeown, T. J. 1983. Hegemonic Stability Theory and 19th Century Tariff Levels in Europe.
International Organization 37 (1): 73–91.
Milgate, Murray, and Shannon C. Stimson. 2009. After Adam Smith: A Century of Transfor­
mation in Politics and Political Economy. Princeton, NJ: Princeton University Press.
Mill, James. 1824. Elements of Political Economy. 2nd ed. London: Baldwin, Cradock, and Joy.
Mill, John Stuart. 1874. Autobiography. 4th ed. London: Longmans, Green, Reader, and Dyer.
Mill, John Stuart. 1848. Principles of Political Economy with Some of Their Applications to Social
Philosophy. London: John W. Parker.
Mossner, E. C., and I. S. Ross. 1987. The Correspondence of Adam Smith. 2nd ed. Oxford:
Clarendon Press.
Noble, David F. 2005. Beyond the Promised Land:  The Movement and the Myth. Toronto:
Between the Lines.
Nye, J. V.  C. 2007. War, Wine, and Taxes:  The Political Economy of Anglo-French Trade,
1689–1900. Princeton, NJ, and Oxford: Princeton University Press.
O’Brien, Patrick K., and Geoffrey Allen Pigman. 1992. Free Trade, British Hegemony and the
International Economic Order in the Nineteenth Century. Review of International Studies 18
(2): 89–113.
Ricardo, David. 1821. On the Principles of Political Economy and Taxation. London:  John
Murray.
Rosenberg, Hans. 1943. Political and Social Consequences of the Great Depression of 1873–1896
in Central Europe. Economic History Review 13 (1–2): 58–73.
Rostow, W. W. 1959. The Stages of Economic Growth. Economic History Review 12 (1): 1–16.
Sack, James J. 1993. From Jacobite to Conservative. Cambridge, UK: Cambridge University Press.
Sally, Razeen. 2008. Trade Policy, New Century: The WTO, FTAs, and Asia Rising. London:
Institute of Economic Affairs.
56   The Free Trade Idea

Schonhardt-Bailey, Cheryl. 2006. From the Corn Laws to Free Trade:  Interests, Ideas, and
Institutions in Historical Perspective. Cambridge, MA: MIT.
Schonhardt-Bailey, Cheryl. 1998. Parties and Interests in the “Marriage of Iron and Rye.” British
Journal of Political Science 28 (2): 291–330.
See, H. 1930. The Normandy Chamber of Commerce and the Commercial Treaty of 1786.
Economic History Review 2 (2): 308–313.
Semmel, B. 1965. The Hume-Tucker Debate and Pitt’s Trade Proposals. Economic Journal 75
(300): 759–770.
Sloane, W. M. 1898. The Continental System of Napoleon. Political Science Quarterly 13
(2): 213–231.
Smith, Adam. [1776] 2012. An Inquiry into the Nature and Causes of the Wealth of Nations.
London: Wordsworth.
Stiglitz, Joseph. 2002. Globalization and Its Discontents. London: Penguin.
Trentmann, Frank. 1996. The Strange Death of Free Trade:  The erosion of the Liberal
Consensus in Great Britain, c. 1903–1932. In Citizenship and Community: Liberals, Radicals
and Collective Identities in the British Isles, 1865–1931, edited by E. F. Biagini, 219–250.
Cambridge, UK: Cambridge University Press.
Trentmann, Frank. 1997. Wealth versus Welfare:  The British Left between Free Trade and
National Political Economy before the First World War. Historical Research 70 (171): 70–98.
Tucker, Josiah. 1763. The Case of Going to War, for the Sake of Procuring, Enlarging, or Securing
of Trade, Considered in a New Light. London: R and J. Dodsley.
Viner, Jacob. 1960. The Intellectual History of Laissez-Faire. Journal of Law & Economics
3: 45–69.
Viner, Jacob. 1969. Power versus Plenty as Objectives of Foreign Policy in the Seventeenth
and Eighteenth Centuries. In Revisions in Mercantilism, edited by D. C. Coleman, 61–91.
London: Methuen.
Viner, Jacob. 1937. Studies in the Theory of International Trade. New York: Harper.
Wallerstein, Immanuel. 1979. The Modern World System: Capitalist Agriculture and the Origins
of the European World-Economy in the Sixteenth Century. New York: Academic.
Walter, Andrew. 1996. Adam Smith and the Liberal Tradition in International Relations.
Review of International Studies 22 (1): 5–28.
Williams, J. B. 1972. British Commercial Policy and Trade Expansion, 1750–1850. Oxford:
Clarendon Press.
 Chapter 4

Tr ade P olicy Inst rume nts

over T i me

Cha d P. B ow n

Introduction

Throughout history governments have applied import protection through many dif-
ferent policy instruments. One extreme is an instrument that is widespread, simple, and
relatively transparent: most customs authorities currently apply a nondiscriminatory,
ad valorem import tariff. Toward the other extreme are dozens if not hundreds of eclec-
tic instruments that are more complex and less transparent and frequently entail addi-
tional political-economic distortions. Consider the experience of only one industry in
only one country—US steel—even since the 1960s. While being protected by applied,
nondiscriminatory, ad valorem import tariffs, US steel has also experienced quotas,
minimum price arrangements, voluntary export restraints, antidumping, countervail-
ing duties, suspension agreements, preferential tariffs to NAFTA and other partners,
import licenses and monitoring, safeguards, expanded trade adjustment assistance, and
local content requirements.
That import protection arises through such a variety of instruments raises funda-
mental questions for political-economic research. Are there apparent trends in the use
and nonuse of particular import protection instruments? Furthermore, what are the
political-economic and institutional determinants of, interrelationships between, and
implications for government use of these various instruments?
The next section begins by examining historical trends for a number of import pro-
tection instruments. Perhaps the most pervasive trend throughout the international
trading system since 1947 has been the reduction in applied tariffs and opening of econ-
omies to international commerce. Nevertheless, it is worth recalling the analogy from
a New York Times article that Robert E. Baldwin made famous at the conclusion of the
Kennedy Round, by including it in his comprehensive economic analysis of nontariff
protection: “The lowering of tariffs has, in effect, been like draining a swamp. The lower
58    Trade Policy Instruments over Time

water level has revealed all the snags and stumps of non-tariff barriers that still have to
be cleared away” (1970, 2).
While individual nations may have drained their swamps by lowering applied tariffs,
the more complete story of import protection demonstrates substantial heterogeneity.
Sometimes these low applied tariffs have not been accompanied by the legal commit-
ments that “bind” them under the World Trade Organization (WTO). Sometimes the
low applied tariffs are not offered to all trading partners but are instead offered only on
a discriminatory basis through regional, bilateral, or preferential regimes. Sometimes a
low applied tariff on a particular product is then more than offset by application of a new
and higher tax through antidumping, safeguards, or countervailing duties. Sometimes
even more innovative, nontransparent, and complex nontariff barriers have come
into use.
These trends suggest complexities between tariff and nontariff protection that push
beyond even Baldwin’s substantial foresight. Draining the swamp of tariff protection
may have done more than just reveal the existence of previously unobserved nontariff
protection. It may have stimulated growth in levels of old and new forms of nontar-
iff protection. It also may have resulted in an altered political-economic ecosystem in
which governments conduct trade policy.
The third section reviews some of the theories that guide the literature on import
protection as well as the implications for the relationship between tariff and nontariff
protection. These theories draw from both economic efficiency and redistributive
motives for trade policy; many are also informed by the evolution of the multilat-
eral trading system, from the start-up of the General Agreement on Tariffs and Trade
(GATT) through the WTO period that began in 1995.
The fourth section turns to major areas of empirical study in light of this theory.
Econometric research has begun to provide new insight into the determinants of gov-
ernment application of import tariffs, as well as the forces that influence their negotiated
reduction under international trade agreements. An emerging literature also explores
empirically how trade agreement commitments, as well as political-economic shocks
and incentives, jointly affect governments’ resorting to other instruments of import
protection.
The fifth section touches on recent and promising research innovations, and the last
section concludes by considering open questions and future research.

Institutions and Trends

in Policy Instruments

This section describes key institutional developments and trends in the use of tariff and
nontariff instruments of import protection since the 1940s.
 Chad P. Bown    59

Institutional Influences under the GATT/WTO System

Before the advent of the multilateral trading system in its current form, tariff and non-
tariff barriers were very high. The US Smoot-Hawley tariffs of 1930 and the subsequent
retaliatory response by the international community led to an accumulation of various
instruments of protection that impeded the worldwide resumption of trade in the after-
math of the Great Depression. Irwin, Mavroidis, and Sykes (2008) provide a fascinating
account of the political-economic negotiations that led to the GATT’s creation in 1947,
signed by twenty-three contracting parties, which includes the following in its charter’s
preamble:
Recognizing that their relations in the field of trade and economic endeavour should
be conducted with a view to raising standards of living, ensuring full employment
and a large and steadily growing volume of real income and effective demand, devel-
oping the full use of the resources of the world and expanding the production and
exchange of goods,
Being desirous of contributing to these objectives by entering into reciprocal and
mutually advantageous arrangements directed to the substantial reduction of tariffs
and other barriers to trade and to the elimination of discriminatory treatment in inter-
national commerce. . . .
(GATT 1947; emphasis added)
The broad strategy for those working within the GATT system at its inception was two-
fold.1 First get countries to convert their quantitative restrictions and other nontariff
barriers into nondiscriminatory, or most-favored-nation (MFN), tariff form. Then
get the GATT contracting parties together in periodic negotiating rounds to recipro-
cally exchange concessions to lower these applied tariffs and legally “bind” the tariffs
to prevent them from increasing. For the more industrialized economies, the result
was a general downward trend in applied import tariff rates and binding levels over
the next fifty years. For developing countries—when they ultimately became GATT
contracting parties—the import tariff story evolved differently. Most self-selected
out of the reciprocal, multilateral negotiations framework for reducing MFN tariffs.
Instead, developing-country reductions to applied tariffs typically occurred much later
and frequently resulted from a unilateral liberalization that took place independently
or in conjunction with preferential trade liberalization. Furthermore, most develop-
ing countries have not immediately “locked in” those tariffs by making legally binding
commitments at the similarly low rates at which their MFN tariffs have actually been
applied.
While the GATT framework attempted to cajole import protection toward the
instrument of applied MFN tariffs, the 1947 agreement also contained a variety of
exceptions spelling out conditions under which governments could subsequently
resort to other instruments of protection. For example, while the GATT’s Article XI
contained a general prohibition on the use of quantitative restrictions, Article XII
60    Trade Policy Instruments over Time

permitted governments to use quotas to address balance of payments problems. While

the GATT did not allow governments to simply raise their applied MFN import tariffs
unilaterally without notice, Article VI (antidumping and countervailing duties) and
Article XIX (safeguards) permitted governments to impose higher duties in response
to certain economic shocks, provided they went through particular procedural steps.2
Finally, while the GATT’s Article I  demanded participants offer nondiscriminatory
MFN tariffs to all other contracting parties, Article XXIV provided a fundamental
exception by allowing formation of (preferential) free trade agreements and customs
unions.3

Important Trends in Use of Different Policy Instruments

Beginning from near autarky during the Great Depression, applied MFN import tar-
iff rates have fallen considerably across countries over time. And while they may have
fallen during different periods and for different reasons, the resulting simple average of
applied MFN tariffs by 2010, for example, was extraordinarily low for a number of major
economies: 3.5 percent for the United States, 4.4 percent for Japan, 5.1 percent for the
European Union, 9.6 percent for China, 13.0 percent for India, and 13.7 percent for Brazil
(WTO 2011).
The remainder of this subsection describes three other important trade policy instru-
ments of the GATT/WTO era:  temporary trade barrier policies, voluntary export
restraints (VERs), and preferential trade agreements (PTAs).
Policies such as antidumping, safeguards, and countervailing (antisubsidy) duties
constitute one major class of nontariff protection that took on greater significance in an
environment characterized by low and legally bound applied MFN tariffs. These poli-
cies are often jointly referred to as trade remedies, contingent or administered protec-
tion, trade defense instruments, or temporary trade barriers—the latter stemming from
the fact that WTO rules identify maximum specific periods over which these policies
are permitted to remain in effect. Antidumping has been the most frequently used of
these instruments; the first antidumping law predates the GATT and was enacted by
Canada in 1904. Australia, Canada, the European Union, and the United States domi-
nated global use of antidumping until the early 1990s.
Many major emerging economies became important users of antidumping and other
temporary trade barriers beginning in the 1990s (Bown 2011). As Table 4.1 reveals, while
the United States and European Union continued to have periods in which a signifi-
cant share of their imports were covered by such import restrictions, comparable and
sometimes higher levels of import coverage under these nontariff instruments arose,
seemingly out of nowhere, for countries like Argentina, Brazil, China, India, Mexico,
and Turkey. Most emerging economies had virtually no experience using these particu-
lar instruments before each underwent its own major episode of tariff liberalization in
the late 1980s or 1990s.
 Chad P. Bown    61

Table 4.1  Imports Covered by Imposed Temporary Trade Barriers, 1995–2011

Estimated share of non-oil manufacturing imports
covered by TTBs in

Year of first
antidumping Peak year, (Peak
Economy law/initiation 1995 2000 2005 2010 1995–2011 year)

Argentina 1972/NA 0.7 2.4 1.8 2.6 3.5 (2002)

Australia 1906/NA 0.8 0.4 0.5 0.5 0.8 (1996)
Brazil 1987/1988 0.5 1.0 1.6 1.4 2.1 (2008)
Canada 1904/NA 1.2 1.9 1.0 0.7 1.9 (2001)
China 1997/1997 0.0 0.5 2.9 2.8 4.5 (2003)
European Union* 1968/1968–69 2.7 3.4 2.9 1.8 4.2 (2002)
Indonesia 1995/1996 0.0 1.1 0.2 0.8 1.3 (2001)
India 1985/1992 0.1 2.6 2.4 4.3 6.3 (2011)
Japan 1920/1982 0.1 0.1 0.0 0.0 1.2 (2009)
Mexico 1986/1987 1.7 1.3 0.7 0.3 1.8 (1998)
South Africa 1914/1921 0.3 0.7 0.9 0.4 1.0 (2002)
South Korea 1963/1986 0.2 0.5 0.5 0.5 0.5 (2009)
Turkey 1989/1989 2.4 1.8 2.6 4.2 4.9 (2008)
United States 1916/1922 1.9 5.2 4.3 3.6 5.6 (2001)

Source: Data taken from Bown (2013, Figure 1a) and Bown (2012).
*EU at the time was the European Economic Community; some EU member states have AD laws that
predate 1968. Temporary trade barriers (TTBs) include antidumping, countervailing duties, global
safeguards, and China-specific transitional safeguards.
NA = not available.

Voluntary export restraints are a second nontariff instrument that experienced a

period of proliferation; most of this occurred prior to the more recent turn to tempo-
rary trade barriers. They work like an import quota but with one important differ-
ence: because the exporters voluntarily agree to restrict their quantities of foreign sales,
the exporting country gets the “rents” associated with the consumer price increase above
its free trade level. While the economic outcome for exporters is typically not as good as
free trade, it is frequently better than under a comparably trade-restricting import tariff
or import quota.
At least two related factors took hold in the 1960s that contributed to the prolifer-
ation of VERs. The first was Japan’s entry into the GATT in 1955 and the integration
of its export-led growth model into the trading system. Increased Japanese exports of
textiles and related products put adjustment pressure on import-competing industries
62    Trade Policy Instruments over Time

across many countries, including a number of industrialized economies. Between

Japan’s accession in 1955 and 1970, roughly fifty GATT contracting parties invoked
their Article XXXV rights not to apply MFN tariffs to Japan’s exports; many imposed
country-specific quotas. The United States, on the other hand, sought to negotiate VERs
with Japan. The second factor was pragmatism. In an environment in which govern-
ments were committed to imposing some new import protection, VERs were more pal-
atable to exporters than import quotas or tariffs, for with a VER the exporter at least
received the rents associated with the trade restriction.4
The proliferation of VERs was ultimately headlined by the Multi-Fiber Arrangement
(MFA)—a system of VERs and quantitative restrictions that governed global trade in
textile and apparel products for thirty years before it was eventually phased-out in 2005.
A number of other US-Japanese VERs continued to arise through the 1980s, including
in footwear, semiconductors, automobiles, steel, photo paper, and chemicals. In fact, an
explicit outcome of the Uruguay Round of GATT negotiations was attempts to rein in
VERs; the more recent increase in temporary trade barrier use may be partially related
to VERs falling out of favor by the late 1980s.5
A third important trend is that applied import tariffs for a number of countries
have also fallen preferentially, through two main instruments, for a number of coun-
tries since 1947.6 The first are preference regimes offered to developing countries under
programs such as the Generalized System of Preferences, Everything But Arms, and
African Growth and Opportunity Act. The second are free trade areas (FTAs) and cus-
toms unions; important examples include the European Union, the North American
Free Trade Agreement (NAFTA), and MERCOSUR, which involves Argentina, Brazil,
Paraguay, and Uruguay. Under both arrangements, not only did applied rates fall to dif-
ferent levels for preferential versus MFN tariffs, but the tariff reductions frequently took
place during different time periods.
To summarize, the major intertemporal movements in the rates of import protection
under these different policy instruments—for example, applied MFN tariffs, preferen-
tial tariffs, VERs, antidumping, safeguards, and countervailing duties—suggest com-
mon political-economic determinants. The remainder of this chapter highlights formal
research examining these questions.

Theoretical Studies

Before turning to the determinants of import protection instruments, an examination

of two more fundamental questions is in order. First, why do governments impose
import-restricting policies at all? Second, why do governments voluntarily commit
to limit their access to certain import-restricting policy instruments by signing trade
agreements? Once informed by frameworks that address these questions, this section
then turns to the implications of these trade agreement commitments (covering tariffs)
for government use of other instruments of import protection.
 Chad P. Bown    63

Theories of Import Restrictions: Tariffs and Everything Else

Research has coalesced around two main government motives to explain policies that
interfere with free trade. The first is economic efficiency. If the country is “large”—so
that an alteration in its consumption or supply of a product leads to changes in world
(or the foreign exporter’s received) prices—the government can use trade policy to shift
the terms of trade in its country’s favor and improve national well-being relative to free
trade (Johnson 1953–1954).
The second motive is redistributive. Even if the country is “small”—so that changes to
its trade policy do not affect world prices, thus ruling out the terms-of-trade motive by
assumption—governments may want to impose a “politically optimal” import tariff if
its benefits are greater than the costs. The economic costs of protection are well known;
they include the deadweight losses associated with too little domestic consumption and
too much induced production by relatively inefficient domestic industries. However, in
political-economy models of import protection, political benefits can overcome such
costs. For example, in the Grossman and Helpman (1994) workhorse, political-economy
model of lobbying for protection, the benefits arising from tariffs can outweigh their
efficiency costs if governments value campaign contributions in addition to the eco-
nomic well-being of producers and consumers.
A long-standing literature has evolved that compares the political-economic impli-
cations of various instruments of import protection. Import tariffs are generally inef-
ficient for small countries relative to free trade; furthermore, they are inefficient relative
to outcomes whereby governments use domestic policy instruments (taxes, subsidies)
to address particular market failures in ways that do not create by-product distortions.7
Nevertheless, conditional on an outcome wherein a government is committed to impos-
ing some form of import protection, a nondiscriminatory ad valorem tariff is typically
the least inefficient instrument.8 To provide one illustrative example, consider a nontar-
iff alternative such as a quantitative restriction, an instrument that theoretically can be
structured so as to be exactly as trade distorting as an import tax. Because quotas also
require the government to undertake a separate and discretionary decision about how
to allocate the licenses, additional opportunities for wasteful rent-seeking (Krueger,
1974) can arise in addition to the potential for discrimination between foreign sources.

Large Countries, Import Protection, and Trade Agreements

Why do large countries sign trade agreements, and what are the implications? To begin,
consider a world without trade agreements, in which rational governments thus set
optimal (Nash) import tariffs so as to maximize their countries’ economic well-being.
One strain of the literature suggests that governments of large countries find trade
agreements efficiency-improving if the agreement helps coordinate theirs and another
large country’s trade policies so as to avoid a “prisoner’s dilemma” outcome (Bagwell
and Staiger 1999, 2002). For import tariffs, the trade agreement is needed because
64    Trade Policy Instruments over Time

neither country has an incentive to lower its tariff unilaterally; a unilateral reduction
for a large country starting from its optimal tariff would lead to an increase in the world
price of its imported good, which would make it worse off. Interestingly, the GATT/
WTO principle of reciprocity can be interpreted as facilitating a negotiated outcome
whereby one country trades off a reduction to its import tariffs (affecting a foreign trad-
ing partner’s exports) against a reduction to the foreign country’s import tariffs (affect-
ing the home country’s exports). The trade agreement’s coordinated reduction of import
tariffs improves each country’s economic well-being by expanding the total volume of
trade while neutralizing what would otherwise be an adverse impact on each country’s
terms of trade, defined as each country’s price of exports relative to its price of imports.
One acknowledged limitation of this approach is its abstraction from the complexities
of enforcement—a particularly acute issue in the context of trade agreements between
sovereign states. Thus a related literature examines the self-enforcing nature of trade
agreements between large countries. Bagwell and Staiger (1990), for example, examine a
repeated game played between governments in which each stage is a prisoner’s dilemma,
and there is uncertainty over trade volumes. A trade agreement is modeled as the most
liberal or “cooperative” trade policy that can be supported without either government
having an incentive to “defect” by imposing its unilaterally optimal (noncooperative) tar-
iff. This approach reveals that governments may need to increase their equilibrium levels
of import protection to maintain cooperation in response to import surges, because such
surges create a new and strong incentive for a government to defect by raising its tariff
because it would improve its nation’s economic well-being. However, an implication for
countries that have taken on commitments to lower and bind their applied MFN tariffs at
the WTO is that the government may need to switch to some other policy instrument to
increase protection if it hopes to remain consistent with its basic WTO obligations.
To summarize, large countries have a unilateral incentive to impose beggar-
thy-neighbor import protection policies that improve their economic well-being by
shifting some of the protection’s costs onto trading partners. Trade agreements can thus
be used to help two or more such large countries coordinate the import tariff reduc-
tions that each would not undertake unilaterally. Nevertheless, trade agreements that
constrain national use of tariffs come with their own caveats. In particular, a cooperative
trade agreement on tariff commitments may be unsustainable in an environment char-
acterized by trade flow volatility if the agreement does not provide governments with
the flexibility to sometimes access additional import protection.9

Small Countries, Import Protection,

and Trade Agreements
A second strain of the literature examines “small” countries that do not have (national)
economic efficiency motives to impose tariffs, yet may do so for redistributive purposes.
This literature posits that trade agreements may play a “commitment” role for govern-
ments to tie their own hands with respect to their private sectors.
 Chad P. Bown    65

Take the classic Grossman and Helpman (1994) political-economy model of a small
country. Here the government would not want to sign a trade agreement that would tie
its hands and constrain its tariffs, because this would foreclose its ability to extract rents
from domestic producers through their lobbying activities and campaign contributions.
However, Maggi and Rodríguez-Clare (1998) relax the model’s assumption on capital
immobility and allow factors to be mobile across sectors in the long run to show that the
expectation of import protection can lead unproductive industries to make excessive
investment that even a politically motivated government would prefer to avoid. A trade
agreement that commits the government to impose lower tariffs in the long run can
therefore help prevent excessive lobbying for import protection.10
There are, however, potential costs for politically motivated governments that seek
to sign trade agreements to constrain their use of tariffs, especially when other instru-
ments of protection are available. Copeland (1990) develops a two-staged game in which
governments can first negotiate cooperatively over tariffs and then implement nontariff
protection noncooperatively in the second stage. If governments are politically moti-
vated and place a high value on the interests of producers, trade agreements that elimi-
nate tariffs can lead governments to substitute policy toward more economically costly
forms of nontariff protection in the second stage so that consumers are made worse off.
Overall, this literature identifies a number of trade-offs associated with trade agree-
ments, even for small countries. Politically motivated governments may use trade
agreements as a commitment device to help prevent their own excessive application
of import tariffs. However, trade agreements with loopholes that make it too costless
to replace (once high) applied tariffs with protection through another instrument can
undermine the agreement’s commitment value. Furthermore, agreements that con-
strain tariffs too much may risk pushing import protection into nontariff instruments
that may be even more distorting.

Empirical Studies

This section turns to the major empirical studies of import protection. It follows a simi-
lar organizational structure, first identifying empirical support for the determinants
of import tariffs and the role of trade agreements. The analysis then builds from these
results to assess their implications for alternative instruments of import protection.

Evidence on Import Tariff Formation Before

and After the Signing of Trade Agreements
The first foundational piece from the empirical literature examines why coun-
tries impose tariffs in the absence of international trade agreements. Broda, Limão,
and Weinstein (2008) provide evidence from disaggregated data and export supply
66    Trade Policy Instruments over Time

elasticities in line with the terms-of-trade theory: governments impose higher import

tariffs where they have greater market power. Evidence on the determinants of these
optimal tariffs is supported via four different empirical exercises. First is evidence for
the applied import tariffs for fifteen nonmembers of the WTO. The second and third are
evidence from the United States for its “column 2” tariffs (which apply to countries not
granted MFN status) and for measures of its nontariff protection. Fourth is the lack of
statistical evidence for the US-applied MFN tariff rates. This meets the expectation that
US-applied tariffs would be statistically unrelated to the theoretically predicted optimal
tariff, because these tariffs have been continually reduced after being subjected to suc-
cessive GATT/WTO negotiating rounds.
A second set of studies provides evidence that the terms-of-trade theory can help one
understand the tariff cuts that countries take on under trade agreements like the WTO.
Bagwell and Staiger (2011), for example, examine determinants of the one-time tariff
cuts for sixteen countries that acceded to the WTO between 1995 and 2005. The evidence
indicates that negotiated tariff levels are related to prenegotiation levels of tariffs, import
volumes, import prices, and trade elasticities. Furthermore, Ludema and Mayda (2013)
examine the Uruguay Round negotiations and the resulting MFN tariffs for thirty-six
developing and high income economies, including Australia, Canada, the European
Union, Japan, South Korea, and the United States. Evidence suggests that the level of
the negotiated import tariff is also negatively related to the importer’s market power and
the Herfindahl-Hirschman index of exporter concentration. The free-rider problem of
dispersed export interests can help explain why some products—such as agriculture,
prepared food, textiles, and footwear—have experienced much smaller reductions to
their applied MFN tariffs under multiple rounds of multilateral negotiations than did
more concentrated export interests.
There are two main insights from this empirical literature. First, in the absence of
trade agreements, governments impose higher tariffs where they have market power.
Second, these and related economic forces affect the negotiation outcomes for WTO
member countries and thus implicitly also affect the size of their remaining (“politically
optimal”) applied MFN import tariffs.

The Decline in Applied Tariffs and the Rise in Other

Instruments of Nontariff Protection
Now armed with a better understanding of the political-economic forces behind the dif-
ferent import tariffs that countries apply in the absence of trade agreements and when
they are parties to the GATT/WTO, this subsection turns to research on the implica-
tions of trade agreements for other instruments of import protection.
The fundamental fact that applied import tariff reductions resulting from GATT/
WTO negotiations are at least partially responsible for the levels and patterns of non-
tariff protection is implicit, if not explicit, in much of the rest of the literature.11 In the
 Chad P. Bown    67

influential empirical studies of the Grossman and Helpman (1994) theory of the politi-
cal economy of import protection, both Goldberg and Maggi (1999) as well as Gawande
and Bandyopadhyay (2000) do not use data on applied US import tariff rates. Instead,
they estimate structural determinants of measures of US nontariff protection.
Other approaches attempt to more directly examine implications of the terms-of-
trade theory for countries that have taken on MFN tariff commitments under the WTO
and must then use alternative policy instruments to increase their levels of import
protection. For example, Bown and Crowley (2013b) construct measures of nontar-
iff protection from US application of antidumping and safeguards between 1997 and
2006 to investigate the repeated game model of Bagwell and Staiger (1990). Evidence
is consistent with the theory that countries in self-enforcing trade agreements increase
their levels of import protection in response to positive trade volume shocks, with varia-
tion arising according to industry-level trade elasticities. Market power may thus still
affect the levels of import protection that governments impose, even when applied MFN
import tariffs are constrained by trade agreement commitments, provided that mea-
sures of import protection reflect the influence of the appropriate additional nontariff
instruments.
Studies on different countries and time periods examine other aspects of the relation-
ship between tariff and nontariff protection. Limão and Tovar (2011) examine Turkey’s
experience during the 1990s, when Turkey signed a customs union arrangement with
the European Union and legally bound some of its tariffs under the WTO. Evidence sug-
gests that Turkey’s tariff commitments increased the likelihood and restrictiveness of its
government’s subsequent use of nontariff instruments of import protection.
Another interesting case study is India, which made massive cuts to its applied
MFN import tariffs in the 1990s and subsequently became the trading system’s heavi-
est user of antidumping and safeguards. Bown and Tovar (2011) estimate structural
determinants of the Grossman and Helpman (1994) political economy model of pro-
tection on repeated cross-sections of Indian data—before and after India’s unilateral
applied MFN import tariff rate cuts associated with its 1991–1992 standby arrange-
ment with the IMF. Changes to India’s applied MFN tariffs during this period appear
to have resulted in an exogenous, trade-liberalizing shock to India’s trade policy.
Nevertheless, by 2002 India had unwound much of the applied MFN tariff reductions
of the 1990s by accumulating a substantial stock of imposed antidumping and safe-
guards nontariff restrictions.
Finally, a number of studies provide cross-country evidence that an important deter-
minant of import protection through “new” policies such as temporary trade barriers is
the commitments that governments take on that limit their access to “old” instruments
of protection. For example, combined evidence from Bown and Crowley (2013a, 2014)
for eighteen high-income and emerging economies between 1989 and 2010 finds that as
more products’ applied MFN tariffs push up against the trade agreement constraints of
WTO tariff-binding commitments, governments implement new import protection by
turning to instruments like temporary trade barriers.
68    Trade Policy Instruments over Time

Preferential Tariff Reductions and

Multilateral Tariff Reductions
A separate stream of research examines potential interrelationships between a govern-
ment’s preferential tariffs and its MFN tariffs offered under the WTO. These empirical
studies have begun to inform understanding of whether and when preferential trade
agreements can be “building blocks” versus “stumbling blocks” for multilateral coopera-
tion in trade policy (Bhagwati 1991). This has been an open empirical question given the
conflicting results that arise from the theoretical literature.12
The first major empirical study to examine this question used product-level data
from the United States (Limão 2006). It considered the impact of US PTAs on the mul-
tilaterally negotiated MFN tariff cuts that the United States subsequently made under
the Uruguay Round. Evidence suggests a stumbling block effect: US multilateral tar-
iff reductions were smaller for products imported under its PTAs relative to similar
products imported only from PTA nonmembers. In a follow-up study of the European
Union, Karacaovali and Limão (2008) provide related evidence indicating that this
stumbling block effect is not limited to just the United States and its PTAs. Their evi-
dence indicates that the EU reduced its multilateral tariffs on goods not imported under
FTAs by almost twice as much as the tariffs imported under its PTAs.
Because the spread of PTAs is so pervasive, there are many opportunities to inves-
tigate the extent to which the stumbling block phenomenon extends to other settings.
In one important study, Estevadeordal, Freund, and Ornelas (2008) reach the opposite
conclusion after examining the experience of ten Latin American countries in the 1990s.
For these Latin American countries, the preferential reductions in applied tariffs were,
on average, subsequently followed by governments making applied MFN tariff reduc-
tions. In this context, Latin American PTAs were found to be a building block to future
multilateral tariff liberalization.
Can theory help explain the difference in these results across empirical settings, and
whether any particular PTA is likely to hinder or promote subsequent multilateral trade
liberalization?13 Part of the explanation for the stumbling block evidence may be the
importance of nontrade objectives particular to the US and EU PTAs. Limão (2007), for
example, suggests a theoretical motivation that the United States and EU need to main-
tain preferences with certain partners to compensate for their commitments to higher
labor and environmental standards and intellectual property rights protection.
Furthermore, Estevadeordal, Freund, and Ornelas (2008) note an important differ-
ence arising in the context of the Latin American countries’ liberalization episodes.
Unlike the US and EU cases, the Latin American PTA negotiations would have resulted
in a given product’s “preference margin”—defined as the difference between its applied
MFN tariff and its PTA tariff—being quite large. Large preference margins present the
opportunity for substantial economic efficiency costs to arise through trade diversion
(Viner 1950). One explanation is that Latin American governments were cognizant of
this concern and thus minimized the potential negative PTA impact and deliberately
reduced preference margins by also cutting their MFN tariffs toward PTA nonmembers.
 Chad P. Bown    69

Additional Influences on Import Protection: Retaliation

Capacity and WTO Dispute Jurisprudence
Because trade policy is a repeated game played between sovereign states, the expected
trading partner reaction to a new import restriction is also likely to endogenously affect
how countries implement import protection in the first place. Historical evidence from
the Great Depression has long established expectations that retaliation can have impor-
tant effects; for example, Irwin (2011) argues that trading partner retaliation against the
US Smoot-Hawley tariffs in 1930 severely curtailed US exports.
Research on more recently implemented policies shows how the latent threat of
meaningful foreign trade retaliation is also likely to affect the channels through which
import protection arises ex ante. Blonigen and Bown (2003) use the setting of industry
applications for US antidumping to illustrate how the presence of this tailor-made—that
is, product-specific and trading partner–specific—instrument presents opportunities
for import protection to be funneled toward certain countries and/or industries that
lack the capacity to retaliate. Similarly, Bown (2004) provides cross-country evidence
that the incentives inherent in the multilateral system’s formal dispute settlement pro-
cedures affected government choices of whether to implement protection through
GATT-consistent instruments during 1973–1994.
Finally, the WTO’s dispute settlement system is also likely to influence the instruments
of import protection that member governments apply through means other than its abil-
ity to authorize retaliation.14 Sykes (2003), for example, identifies problems stemming
from Panel and Appellate Body decisions resulting from legal challenges to national use
of import restrictions under the WTO’s Agreement on Safeguards. Not only did these
WTO rulings strike down virtually all challenged instances in which governments had
applied this nontariff instrument, but the resulting jurisprudence failed to provide policy
makers with useful guidance on how to actually apply a safeguard in a WTO-consistent
manner.15 In a trading system in which governments have access to relatively substitut-
able instruments of import protection, discouraging the use of safeguards in isolation
may not lead to less protection overall. While such rulings may help explain the decline in
safeguard use, they may also help explain the steady increase in use of antidumping.

New Approaches

Notwithstanding the additional questions raised by the burgeoning empirical literature

on the determinants of use of different trade policy instruments over time, a number of
new theoretical and methodological approaches are also worth introducing.
Ossa (2011) has provided an innovative theoretical approach that identifies an inter-
national cost-shifting effect of trade policy that is separate from the terms-of-trade
externality found in Bagwell and Staiger (1999, 2002). The approach starts from a
monopolistically competitive market of the “new trade theory” models, which feature
70    Trade Policy Instruments over Time

shipping costs and two-way trade in similar (but differentiated) products. One impor-
tant result is that a beggar-thy-neighbor motive for trade policy intervention can arise
from a “firm-delocation” (or “profit-shifting”) effect, as governments have incentives to
attract more of the world’s firms to locate locally so as to save on transport costs.
Antràs and Staiger (2012) have introduced a new approach that recognizes the global
fragmentation of production, offshoring, and the increasing economic importance of
trade in intermediate inputs.16 They conclude that one change to a simple assumption
maintained in most prior economic theory on trade agreements—prices being deter-
mined not from market-clearing conditions but due to bilateral bargaining between
international buyers and sellers—can have profound implications. Intuitively, an
international “hold-up” problem can arise when relationship-specific investments
are required if contracts between buyers and sellers are incomplete. In such instances,
because one party (e.g., the buyer or importer) may be able to hold up the other (e.g.,
the exporter) and renegotiate the terms of their deal after the exporter has made a sunk
investment, the seller will not make the jointly efficient level of investment in the first
place. The hold-up problem can thus result in volumes of input trade across countries
under free trade that are inefficiently low, suggesting an additional, efficiency-enhancing
motive for trade policy intervention.
Such approaches have the potential to affect understanding of not only why govern-
ments impose tariff protection, but also why they voluntarily sign agreements to con-
strain tariff protection, and thus the implications of these agreements for alternative
instruments of protection.
Before concluding this section, it is also worth highlighting that one of the fundamen-
tal questions arising from the disparate empirical research on different instruments of
import protection involves comparability. Any attempt to make comparisons—across
countries, industries, and time—is complicated by the variety and complexity of the
instruments in use. For example, how does one compare the restrictiveness of the “pro-
tection” inherent in one country, which may apply low MFN tariffs but is also a frequent
user of nontariff protection, with that of a different country, which has zero nontariff
protection but much higher applied MFN tariffs? One important sign of progress in this
area is the effort of Kee, Nicita, and Olarreaga (2009) to more accurately aggregate and
measure the many different forms of tariff and nontariff protection through application
of the theory of trade restrictiveness indices (Anderson and Neary 2005).

Conclusion

This chapter has examined the evolving nature of import protection and research on
the political-economic and institutional determinants of various instruments in use
over time.
Nevertheless, one chapter on this topic cannot be comprehensive. Omitted areas
include the relationship between import protection and pressures stemming from
 Chad P. Bown    71

choice of exchange rate regime; Irwin (2012), for example, describes important link-
ages between the constraints imposed on monetary policy under the gold standard and
the outbreak of protectionism in the early 1930s.17 A second omission is subsidy instru-
ments. Bagwell and Staiger (2006), for example, provide a theory to help understand
WTO rules on subsidies, identifying some of the trade-offs that arise and suggesting
somewhat provocatively that the existing rules on subsidy use may be too stringent.
Furthermore, this chapter has deliberately avoided many of the instruments that will
arise in attempts to address the “twenty-first-century” trade agreement issues, such as
state-owned enterprises, foreign direct investment, intellectual property rights protec-
tion, labor and environmental standards, technical and health standards, and achieving
other forms of “regulatory coherence.”
To conclude, it is worth reconsidering Baldwin’s original analogy—that applied tar-
iff reductions are like draining a swamp—in light of this chapter’s highlighted body of
research. Draining the swamp has more than simply revealed the existence of the “snags
and stumps” of nontariff protection. Some of the major instruments of nontariff pro-
tection to arise were not even present at the initial draining of some national swamps.
Perhaps the GATT/WTO institutional framework helped the seedlings for such instru-
ments blow in from somewhere and take root in new locations. Perhaps the newfound
exposure to sunshine, alongside the inevitable rain that followed, also triggered the
growth of the new, and still-adapting, forms of nontariff protection. However, it is also
possible that draining the tariff swamp may have begun to reveal fundamental limits
to feasible international cooperation over import protection. A still open question is
whether a multilateral system with fully enforceable, time-invariant, free trade would be
possible or even desirable in the long run.
For what outsiders may deride as a swamp, earth scientists refer to more fondly
as a wetland. And what the process of swamp draining has also revealed, once its
broader contributions to areas such as flood control and biological diversity have
come to be accounted for and appreciated, is that such wetlands may play unex-
pectedly critical roles in sustaining a larger ecosystem than had been previously
understood.

Data Sources

Only relatively recently have panel data for product-level trade policy instruments for
many countries become publicly and freely available and widespread enough for a wide
group of political-economic researchers to access.
Important sources for product-level applied and bound tariff data across countries
include the WTO’s Integrated Database and Consolidated Tariff Schedule; these are
available from both the WTO Web site and through the World Bank’s free, online World
Integrated Trade Solutions (WITS) software platform. Preferential tariff data compiled
by UNCTAD (TRAINS) have also been made available through WITS.
72    Trade Policy Instruments over Time

Antidumping, countervailing duty, and safeguards policy use data across countries
have been freely and publicly available in electronic format since 2005 through the
World Bank’s Temporary Trade Barriers Database (Bown 2012) and Global Antidumping
Database.
With respect to other, more difficult to measure, nontariff instruments of import pro-
tection, other useful sources for data construction include the information collected
by the Global Trade Alert, WTO’s Trade Policy Reviews, and the WTO Committees on
Technical Barriers to Trade (TBT), as well as Sanitary and Phytosanitary (SPS) measures.

Acknowledgments

Thanks to Petros Mavroidis, Baybars Karacaovali, and Patricia Tovar for helpful com-
ments on an earlier draft. Aksel Erbahar and Carys Golesworthy provided outstanding
research assistance. Any opinions expressed in this paper are the author’s and should
not be attributed to the World Bank. All errors are my own.

Notes
1. Important treatments of the evolution of the trading system under the GATT and WTO
include Hoekman and Kostecki (2009) and Barton, Goldstein, Josling, and Steinberg
(2006). Furthermore, Dam (1970), Hudec (1990), and Jackson (1997) are seminal and
accessible studies from the perspective of international law. Gowa (1994) provides a com-
parison of postwar trade developments under the GATT to political-military trends under
NATO. See also Gowa (this volume).
2. The GATT’s Article XXVIII also allows contracting parties to renegotiate their MFN
tariff commitments so long as they compensate adversely affected trading partners.
Antidumping requires evidence of injury to domestic competitors of a like product caused
by dumped (low-priced) imports. Countervailing duties require evidence of injury caused
by subsidized imports. Safeguards require evidence of injury caused by an unexpected
surge in imports. Mavroidis, Messerlin, and Wauters (2008) provide a legal-economic
introduction to these import protection instruments.
3. There are other important GATT exceptions to Article I  that allow for discriminatory
treatment. A  GATT waiver initially permitted contracting parties to offer tariff pref-
erences to developing countries before this was formalized in 1979 by the “Enabling
Clause.” Governments can also apply antidumping and countervailing duties on a
discriminatory basis.
4. Baldwin (1970, 30–46) also highlights the frequency with which quotas were used during
this period in areas unrelated to Japan’s GATT entry. While many of the quotas that had
been imposed as Article XII (balance of payments) exceptions after World War II were
phased out by the mid-1960s, industrial economy quotas continued to affect coal, petro-
leum, and agricultural products. Furthermore, Baldwin finds that forty industries were
covered by quota bills introduced in the US Congress in the fall of 1968 alone.
 Chad P. Bown    73

5. The 1995 WTO Agreement on Safeguards prohibits VERs as the outcome of investigations.
Paradoxically, the WTO’s Agreement on Antidumping encourages investigations being
resolved by exporters voluntarily agreeing to “price undertakings.” VERs have thus not
completely disappeared: in 2005 the EU, the United States, and China negotiated VERs
to address China’s export surge around the timing of the expiration of the MFA, and in
2013 VERs were considered as a potential solution to address China’s large exports of solar
panels.
6. Mansfield and Milner (1999) provide a more comprehensive review of the regionalism
phenomenon; Hoekman and Ozden (2007) survey the literature on foreign preference
regimes facing exporters in developing countries.
7. There are numerous “second-best” motives for import protection to improve economic
well-being when more appropriate domestic policy instruments are unavailable. See the
seminal work of Bhagwati and Ramaswami (1963).
8. Ad valorem tariffs are frequently preferred to specific tariffs because the restrictive-
ness of the latter depends also on price levels. For the specific duties found in the US
Smoot-Hawley tariffs, for example, Irwin (1998b) shows how the deflation of the early
1930s increased their trade restrictiveness, while Irwin (1998a) attributes more of the
1940s trade liberalization to the period’s import price inflation than to trade policy
negotiations.
9. Rosendorff and Milner (2001) provide a related approach that extrapolates from a number
of the economic market issues and predictions highlighted in Bagwell and Staiger (1990).
10. Maggi and Rodríguez-Clare (2007) construct a “large” country version of the model to
illustrate how the key commitment insights are affected by governments that also have
terms-of-trade motivations. Staiger and Tabellini (1987) present a related modeling
approach for a small country that is concerned with the time inconsistency problem of
trade policy announcements in the presence of discretionary policy.
11. Much of the research described here was also informed by Trefler’s (1993) critique of the
existing state of the empirical literature on the political-economic determinants of import
protection. Trefler identified the importance of this endogeneity for empirical estimates of
the formation of trade policy; his application to US data in 1983 argued that treating trade
policy as exogenous underestimated its actual impact on trade flows by a factor of ten.
12. One fundamental economic efficiency concern arising from preferential tariffs is the trade
diversion identified initially by Viner (1950). For excellent surveys of the theoretical litera-
ture on the incentives that can arise under preferential versus multilateral liberalization,
see Freund and Ornelas (2010) and Panagariya (2000).
13. A  related question is the extent to which multilateral liberalization affects subsequent
efforts at preferential liberalization. On the one hand, when focusing on tariffs alone, full
multilateral liberalization over tariffs would obviate the need for additional (redundant)
preferential tariff reductions. Mansfield and Reinhardt (2003) argue that impediments to
further multilateral liberalization may push members to pursue PTAs instead.
14. Surveys of the empirical literature on GATT/WTO dispute settlement include Busch
and Reinhardt (2002), Busch and Pelc (this volume), and Bown (2009, ch. 4); the latter
includes a discussion of the potential for GATT/WTO disputes to affect the endogenous
formation of trade policy and thus choice of instruments of import protection.
15. Goldstein and Martin (2000) offer other examples of ways through which the increased
legalization of the WTO system may have unintended consequences for domestic political
economy forces and trade liberalization.
74    Trade Policy Instruments over Time

16. In their extensive work studying value-added trade, Johnson and Noguera (2012),
for example, find that intermediate inputs may account for as much as two-thirds of
international trade.
17. See also Copelovitch and Pevehouse (this volume).

References
Anderson, J., and P. Neary. 2005. Measuring the Restrictiveness of Trade Policy. Cambridge,
MA: MIT Press.
Antràs, P., and R. W. Staiger. 2012. Offshoring and the Role of Trade Agreements. American
Economic Review 102: 3140–3183.
Bagwell, K., and R. W. Staiger. 1990. A Theory of Managed Trade. American Economic Review
80: 779–795.
Bagwell, K., and R. W. Staiger. 1999. An Economic Theory of GATT. American Economic
Review 89: 215–248.
Bagwell, K., and R. W. Staiger. 2002. The Economics of the World Trading System. Cambridge,
MA: MIT Press.
Bagwell, K., and R. W. Staiger. 2006. Will International Rules on Subsidies Disrupt the World
Trading System? American Economic Review 96: 877–895.
Bagwell, K., and R. W. Staiger. 2011. What Do Trade Negotiators Negotiate About? Empirical
Evidence from the World Trade Organization. American Economic Review 101: 1238–1273.
Baldwin, R. E. 1970. Nontariff Distortions of International Trade. Washington, DC: Brookings
Institution Press.
Barton, J. H., J. L. Goldstein, T. E. Josling, and R. H. Steinberg. 2006. The Evolution of the Trade
Regime: Politics, Law, and Economics of the GATT and the WTO. Princeton, NJ: Princeton
University Press.
Bhagwati, J. 1991. The World Trading System at Risk. Princeton, NJ:  Princeton University
Press.
Bhagwati, J., and V. K. Ramaswami. 1963. Domestic Distortions, Tariffs, and the Theory of
Optimum Subsidy. Journal of Political Economy 71: 44–50.
Blonigen, B. A., and C. P. Bown. 2003. Antidumping and Retaliation Threats. Journal of
International Economics 60: 249–273.
Bown, C. P. 2004. Trade Disputes and the Implementation of Protection under the GATT: An
Empirical Assessment. Journal of International Economics 62: 263–294.
Bown, C. P. 2009. Self-Enforcing Trade: Developing Countries and WTO Dispute Settlement.
Washington, DC: Brookings Institution Press.
Bown, C. P. 2011. Taking Stock of Antidumping, Safeguards and Countervailing Duties,
1990-2009. The World Economy 34: 1955–1998.
Bown, C. P. 2012. Temporary Trade Barriers Database. The World Bank. May. http://econ.
worldbank.org/ttbd/ (accessed June 15, 2013).
Bown, Chad P. 2013. Emerging Economies and the Emergence of South-South Protectionism.
Journal of World Trade 47: 1–44.
Bown, C. P., and M. A. Crowley. 2013a. Import Protection, Business Cycles, and Exchange
Rates: Evidence from the Great Recession. Journal of International Economics 90: 50–64.
Bown, C. P., and M. A. Crowley. 2014. Emerging Economies, Trade Policy, and Macroeconomic
Shocks. Journal of Development Economics 111: 261–273.
 Chad P. Bown    75

Bown, C. P., and M. A. Crowley. 2013b. Self-Enforcing Trade Agreements:  Evidence from
Time-Varying Trade Policy. American Economic Review 103: 1071–1090.
Bown, C. P., and P. Tovar. 2011. Trade Liberalization, Antidumping, and Safeguards: Evidence
from India’s Tariff Reform. Journal of Development Economics 96: 115–125.
Broda, C., N. Limão, and D. E. Weinstein. 2008. Optimal Tariffs and Market Power:  The
Evidence. American Economic Review 98: 2032–2065.
Busch, M. L., and E. Reinhardt. 2002. Testing International Trade Law:  Empirical Studies
of GATT/WTO Dispute Settlement. In The Political Economy of International Trade
Law:  Essays in Honor of Robert Hudec, edited by D. M. Kennedy and J. D. Southwick,
457–481. Cambridge, UK: Cambridge University Press.
Copeland, B. R. 1990. Strategic Interaction Among Nations: Negotiable and Non-Negotiable
Trade Barriers. Canadian Journal of Economics 23: 84–108.
Dam, K. W. 1970. The GATT:  Law and International Organization. Chicago:  University of
Chicago Press.
Estevadeordal, A., C. Freund, and E. Ornelas. 2008. Does Regionalism Affect Trade Liber­
alization toward Non-Members? Quarterly Journal of Economics 123: 1531–1575.
Freund, C., and E. Ornelas. 2010. Regional Trade Agreements. Annual Review of Economics
2: 139–166.
GATT. 1947. The Text of the General Agreement on Tariffs and Trade.
Gawande, K., and U. Bandyopadhyay. 2000. Is Protection for Sale? Evidence on the Grossman-
Helpman Theory of Endogenous Protection. Review of Economics and Statistics 82: 139–152.
Goldberg, P. K., and G. Maggi. 1999. Protection for Sale: An Empirical Investigation. American
Economic Review 89: 1135–1155.
Goldstein, J., and L. L. Martin. 2000. Legalization, Trade Liberalization, and Domestic Politics:
A Cautionary Note. International Organization 54: 603–632.
Gowa, J. 1994. Allies, Adversaries, and International Trade. Princeton, NJ: Princeton University
Press.
Grossman, G. M., and E. Helpman. 1994. Protection for Sale. American Economic Review
84: 833–850.
Hoekman, B. M., and M. M. Kostecki. 2009. The Political Economy of the World Trading
System: The WTO and Beyond. 3rd ed. New York: Oxford University Press.
Hoekman, B. M., and C. Ozden. 2007. Introduction to Trade Preferences and Differential
Treatment of Developing Countries, edited by B. M. Hoekman and C. Ozden, xi–xlii.
Cheltenham, UK: Edward Elgar.
Hudec, R. E. 1990. The GATT Legal System and World Trade Diplomacy. Salem, NH:
Butterworth Legal Publishers.
Irwin, D. A. 1998a. Changes in U.S. Tariffs: The Role of Import Prices and Commercial Policies.
American Economic Review 88: 1015–1026.
Irwin, D. A. 1998b. The Smoot-Hawley Tariff:  A  Quantitative Assessment. The Review of
Economics and Statistics 80: 326–334.
Irwin, D. A. 2011. Peddling Protectionism: Smoot-Hawley and the Great Depression. Princeton,
NJ: Princeton University Press.
Irwin, D. A. 2012. Trade Policy Disaster: Lessons from the 1930s. Cambridge, MA: MIT Press.
Irwin, D. A., P. C. Mavroidis, and A. O. Sykes. 2008. The Genesis of the GATT. Cambridge,
MA: Cambridge University Press.
Jackson, J. H. 1997. The World Trading System:  Law and Policy of International Economic
Relations. 2nd ed. Cambridge, MA: MIT Press.
76    Trade Policy Instruments over Time

Johnson, H. G. 1953–1954. Optimum Tariffs and Retaliation. Review of Economic Studies

21: 142–153.
Johnson, R. C., and G. Noguera. 2012. Accounting for Intermediates: Production Sharing and
Trade in Value Added. Journal of International Economics 86: 224–236.
Karacaovali, B., and N. Limão. 2008. The Clash of Liberalizations: Preferential vs. Multilateral
Trade Liberalization in the European Union. Journal of International Economics 74: 299–327.
Kee, H. L., A. Nicita, and M. Olarreaga. 2009. Estimating Trade Restrictiveness Indices.
Economic Journal 119: 172–199.
Krueger, A. 1974. The Political Economy of the Rent-Seeking Society. American Economic
Review 64: 291–303.
Limão, N. 2006. Preferential Trade Agreements as Stumbling Blocks for Multilateral Trade
Liberalization: Evidence for the U.S. American Economic Review 96: 896–914.
Limão, N. 2007. Are Preferential Trade Agreements with Non-trade Objectives a Stumbling
Block for Multilateral Liberalization? Review of Economic Studies 74: 821–855.
Limão, N., and P. Tovar. 2011. Policy Choice:  Theory and Evidence from Commitment via
International Trade Agreements. Journal of International Economics 85: 186–205.
Ludema, R., and A. M. Mayda. 2013. Do Terms-of-Trade Effects Matter for Trade Agreements?
Theory and Evidence from WTO Countries. Quarterly Journal of Economics 128: 1837–1893.
Maggi, G., and A. Rodríguez-Clare. 1998. The Value of Trade Agreements in the Presence of
Political Pressures. Journal of Political Economy 106: 574–601.
Maggi, G., and A. Rodríguez-Clare. 2007. A Political-Economy Theory of Trade Agreements.
American Economic Review 97: 1374–1406.
Mansfield, E. D., and H. V. Milner. 1999. The New Wave of Regionalism. International Orga­
nization 53: 589–627.
Mansfield, E. D., and E. Reinhardt. 2003. Multilateral Determinants of Regionalism: The Effects
of GATT/WTO on the Formation of Preferential Trading Arrangements. International
Organization 57: 829–862.
Mavroidis, P. C., P. A. Messerlin, and J. M. Wauters. 2008. The Law and Economics of Contingent
Protection in the WTO. Cheltenham, UK: Edward Elgar.
Ossa, R. 2011. A “New Trade” Theory of GATT/WTO Negotiations. Journal of Political
Economy 119: 122–152.
Panagariya, A. 2000. Preferential Trade Liberalization:  The Traditional Theory and New
Developments. Journal of Economic Literature 38: 287–331.
Rosendorff, B. P., and H. V. Milner. 2001. The Optimal Design of International Trade
Institutions: Uncertainty and Escape. International Organization 55: 829–857.
Staiger, R. W., and G. Tabellini. 1987. Discretionary Trade Policy and Excessive Protection.
American Economic Review 77: 823–837.
Sykes, A. O. 2003. The Safeguards Mess:  A  Critique of WTO Jurisprudence. World Trade
Review 2: 261–295.
Trefler, D. 1993. Trade Liberalization and the Theory of Endogenous Protection:  An
Econometric Study of U.S. Import Policy. Journal of Political Economy 101: 138–160.
Viner, J. 1950. The Customs Union Issue. New York: Carnegie Endowment for International
Peace.
World Trade Organization (WTO). 2011. World Tariff Profiles 2011. Geneva: WTO, ITC, and
UNCTAD.
 Chapter 5

Met hod ol o gi c a l  I s su e s

Raymond H ic k s

Political science has long lagged behind economics in the methods used to analyze
trade, usually following innovations or advances made a few years earlier. The focus of
research also differs between the two disciplines. While economists examine why coun-
tries trade, the welfare effects of trade, and the effect of institutions on trade, political
scientists generally focus on the political aspects, especially the effects of institutions.
By their nature, trade institutions such as the General Agreement on Tariffs and Trade
(GATT)/World Trade Organization (WTO) or preferential trade agreements (PTAs)
are political. Political factors such as democracy and veto players influence selection
into and ratification of PTAs as well as the choice of members (Gray 2013; Mansfield,
Milner, and Rosendorff 2002; Mansfield, Milner, and Pevehouse 2007, 2008; Mansfield
and Milner 2012).
Whether these institutions have an independent effect on trade arouses heated
debates among observers. In part, the lack of consensus results from the use of differ-
ent methods to analyze the questions. As the methods used to examine trade advance,
the estimated effects of institutions also change. From its beginning as a cross-sectional
method, the gravity model has grown from a panel method that includes dyadic and
year fixed effects to a panel method that includes dyadic fixed effects and controls for
multilateral resistance. As the gravity model has evolved, I argue, the questions that we
can ask about the effects of institutions have also changed.
This chapter focuses on the gravity model to address several issues. First, how do
the methods affect the questions we can ask? We can consider two separate questions
about the GATT/WTO’s effects on trade: Does the GATT/WTO increase trade among
its members compared to their pre-GATT/WTO trade? Second, do countries in the
GATT/WTO trade more with each other than they do with non-GATT/WTO coun-
tries? In other words, what comparison do we want to make to determine whether the
GATT/WTO increases trade? Interpretation of coefficients with dyadic and year fixed
effects is different than when including dyadic fixed effects and country-year fixed
effects, which influences the questions we can ask with each method. Also, how do
the different methods used to control for multilateral trade resistance compare to one
78   Methodological Issues

another? A few methods have been advocated to incorporate multilateral trade resis-
tance, but their findings have not been compared with real-world data. Finally, using the
most recent methods from the economics literature, what effects has the GATT/WTO
had on trade?
A brief history of the gravity model is presented in the next section. The relationship
of “the bilateral barrier between [two regions] relative to average trade barriers that both
regions face with all their trading partners” (Anderson and Van Wincoop 2003, 176),
or the various methods economists have incorporated to control for multilateral trade
resistance, are reviewed in the second section. I then review the literature on the trade
effects of the GATT/WTO in the third section, focusing on differences in the methodol-
ogy used. In the fourth section I estimate the effects of the GATT, controlling for multi-
lateral trade resistance. I also discuss reasons for the differences in results. Finally, I offer
some suggestions for “best practice guidelines.”

Gravity Model

It is axiomatic that studies refer to the gravity model as the “workhorse” model of trade.
Although it has been around for decades,1 economists have only recently provided its
microfoundations.2 Political scientists are less concerned about its foundations and
more interested in using the model to determine the effects of institutions on trade.
Thus, we as a field rely on the specification that is au courant in economics.
The basic gravity model in trade is similar to the gravity model in physics: trade is
a function of the size of countries, their wealth, and the distance between them. That
is, just as gravity has a stronger effect on larger and closer objects, so should larger and
more proximate countries trade more with one another.3 Because the marginal propen-
sity to import is positive, trade should increase with income. An inverse relationship
exists with distance because trade costs increase across space. Mathematically,

Tij = (Yi * Yj ) / Dij (1)

where Tij is trade between countries i and j; Yi is the gross domestic product (GDP)
of country i; Yj is the GDP of country j; and Dij is the distance between countries i and j.
To estimate the gravity model with ordinary least squares (OLS), we log both sides of
(1) to derive the log-linear equation:

( ) ( ) ( )
ln Tij = a + b1 * ln (Yi ) + b2 * ln Yj − b3 ln Dij + eij (2)

where eij is an error term.

Equation (2) is the cross-sectional form of the model and lacks a time dimension.
Cross-sectional models assume that trade is relatively constant from year to year—that
is, the coefficients on the different variables should have similar values in each year—but
 Raymond Hicks   79

this is probably not the case. Rose’s cross-section results, for example, show that the
GATT/WTO effect varies across time (2004, 105). Given the ease of computing capacity
now, most studies use panel data that utilize changes over time and changes within the
dyad. Equation (2) can be extended to panel data by adding a time subscript and includ-
ing a term for year-specific factors (Gt).

( ) ( ) ( )
ln Tij = a + b1 * ln ( Yit ) + b2 * ln Yjt − b3 ln Dijt + b4 * G t + eijt (3)

Equation (3) is the specification typically used now. If trade were dependent only on
GDP and distance, then we would not have to worry about other variables. But there are
a host of other dyadic and monadic factors that may influence trade, so we can think of
Dij as the set of dyadic factors and Yi and Yj as the set of monadic factors that determine
trade. The dyadic factors include, for example, contiguity, distance, common language,
combined area, and number of islands, variables that generally do not vary within a
dyad, as well as colonial status and membership in trade institutions such as the GATT
or trade agreements, which may vary over time. The monadic components include
country-level factors such as GDP, population, exchange rates, and tariff and nontariff
barriers that influence trade and trade costs.
The use of logged trade values has consequences for interpretation of the coefficients.
Because GDP, population, and distance are also logged, their coefficients measure the
percentage change in trade when they increase by 1 percent (see Goldstein, Rivers, and
Tomz 2007, 47). The effect of a dummy variable is a little different. We use the formula
eβ – 1 to estimate the percentage change in trade when a dichotomous variable changes
from 0 to 1.
A second consequence of using the logged value of trade is that trade values of zero
are dropped from the analysis, because the log of 0 is undefined. Some studies attempt
to accommodate missing trade values by adding a small value to trade before logging
it (e.g., ln(Tij +1)), which ensures that zero values are included in the analysis while not
affecting actual trade values. Santos Silva and Tenreyro suggest, however, that both this
procedure and ignoring zero values are likely to “lead to inconsistent estimators of the
parameters of interest” (2006, 643). Instead, they advocate the use of Poisson models
with nonlogged trade values, showing that it works as well as log-linear OLS.
The choice of trade flows is also important. We could examine total bilateral trade
flows (imports plus exports), exports alone or imports alone, or average trade. Because
imports from country A to country B should equal exports from country B to country
A, using bilateral trade flows or average trade as the dependent variable involves non-
directed dyads, or only one observation per dyad-pair. On the other hand, focusing
on either imports or exports entails a directed-dyad approach, or two observations for
each dyad-pair, Tij and Tji. Imports are typically viewed as more reliable than exports,
because countries have an incentive to track them more closely to maximize tariff rev-
enues; Baldwin and Taglioni, however, suggest that because exporters receive a rebate
for correctly announcing exports in the European Union (EU), exports are now equally
reliable (2006, 13).
80   Methodological Issues

In the standard gravity model, researchers attempt to directly estimate the monadic
and dyadic components by including observed variables. The effect of institutions such
as the GATT/WTO or PTAs is estimated with a dichotomous variable indicating the
presence of the institution in the dyad-year. As Baier and Bergstrand (2007) point out,
though, endogeneity is possible when estimating the effects of these institutions with a
standard gravity model. The same factors that lead to greater dyadic trade may also lead
them to form a trade agreement. Failing to account for this will bias the trade agree-
ment coefficient upward. To control for selection into institutions, Baier and Bergstrand
advocate the use of dyadic fixed effects, assuming that “the source of the endogeneity
bias in the gravity equation is unobserved time-invariant heterogeneity” (2007, 84). That
is, specific unobserved attributes both cause dyads to trade and enter into a PTA and
are not accounted for by the standard dyadic variables. Including dyadic fixed effects
will control for any unobservable dyadic effect as well as the observable dyadic effects
represented by contiguity, distance, or colonial status. Any dyadic variable that does
not change over the sample will be dropped from the model. This also means that for
the trade institution variables, we need at least one year in the sample when a country
was not in an institution. If both countries in a dyad are in an institution for the entire
sample, their data will not be used to estimate the institution’s effect. Specifically for the
GATT, we need at least one year of data before 1948, or trade between founding mem-
bers does not contribute to the both-in coefficient.
The number of dyads in most data sets is too large to allow the use of dummy vari-
ables for the dyadic fixed effects. For example, there are almost 17,500 dyads in the
Goldstein, Rivers, and Tomz (2007) analyses that cover the years 1946 to 2003. Rather
than dyadic dummy variables, we condition the fixed effects out of the data and then
adjust the standard errors appropriately. The coefficient for each dyadic fixed effect is
the average trade value for the dyad. We can partition out this effect by subtracting the
mean dyadic value of every variable from each observation. For variables that do not
change over time within the dyad, such as distance or contiguity, the mean will equal
the value for each observation. The time-invariant dyadic variables will drop from
the model, making it impossible to estimate how contiguity or distance affects trade.
Fortunately, these variables are typically a second-order concern, especially in political
science.
The coefficient on the trade institution variable in a model with dyadic fixed effects
shows the change in dyadic trade when both are members compared to their trade
when at least one is not.4 This will be the trade-creating effect of the institution. This
trade-creating effect does not rely on the use of dyadic fixed effects. Instead, it is the
coefficient on the variable indicating that both dyad members are in a trade institution.
Without dyadic fixed effects, the comparison is not to preinstitution trade, but to trade
among the base, or excluded, group, which includes all dyads where at least one country
is not a member.
At the same time, the institution may have a trade-diverting effect; that is, the insti-
tution may shift a member’s trade away from nonmembers (Viner 1950; Ghosh and
 Raymond Hicks   81

Yamarik 2004). To measure this effect, we include a variable that takes a value of 1 if only
one of the dyad members belongs to the institution. (We will call the trade-creation vari-
able “both-in” and the trade-diversion one “one-in” for short.) With dyadic fixed effects,
the one-in variable measures the change in dyadic trade when one of the countries is
in an institution.5 Trade institutions are not necessarily trade-diverting. It is possible
that they have no effect on one-in trade or actually exert a positive effect on trade.6 We
want to compare the both-in and one-in coefficients to determine the relative effects of
trade-creation and trade-diversion.
Since the year-specific factors (Gt) in panel data are common to all countries, most
studies include year fixed effects. One benefit of the year fixed effects is that we do not
need to worry about converting trade flows into constant dollars. The coefficients on the
right-hand side variables will be the same if we use current or constant values for trade.
Since the conversion is usually done using the US GDP deflator or US consumer price
index, which is constant for all countries in a given year, it will be collinear with the year
fixed effects.7

Multilateral Trade Resistance

For a long time, aside from GDP, population, and maybe exchange rates, researchers
ignored the monadic factors affecting trade. Despite the recognition that trading
costs play an important role in trade, they were largely consigned to the error term.
Trading costs include not just the cost of shipping goods, but also factors such as
tariffs and nontariff barriers, which affect the prices at which goods trade. Largely,
these costs were ignored because time-series data on tariff and nontariff barriers
are sparse. In the mid-1990s, however, McCallum identified a “border puzzle” in his
comparison of trade within Canadian provinces to trade between Canadian prov-
inces and US states in 1988. He finds that “trade between two provinces is more than
20 times larger than trade between a province and a state” (1995, 616). Anderson
and Van Wincoop (2003) argue that McCallum’s large border effect results from
the omission of what they label “multilateral trade resistance” (MTR) or the fact
that “trade between two regions depends on the bilateral barrier between them
relative to average trade barriers that both regions face with all their trading part-
ners” (2003, 176). Multilateral trade resistance includes any policies at the country
(or monadic) level that potentially affect trade with all partners and may “include
trade policy and tariffs, exchange controls, the effective multilateral exchange rate
regime, fiscal policies, financial crises and so forth” (Mathy and Meissner 2011, 13).
Anderson and Van Wincoop derive MTR separately for each region in the analysis,
and separate terms are estimated for both the importing region and the exporting
region.8 Because they include ten Canadian provinces, thirty US states, and the rest
of the United States, they solve forty-one equations to calculate the MTR for each
82   Methodological Issues

region. If more countries are added, the number of equations to be solved would
also increase.
As Baldwin and Taglioni (2006) show, the Anderson and Van Wincoop method is
suitable only for cross-sectional data. For panel data, Baldwin and Taglioni suggest
including country-year fixed effects, or dummy variables for each importer-year and
for each exporter-year, as an alternative solution. Just as the dyadic fixed effects con-
trol for any unobserved dyadic factors, the importer-year and exporter-year fixed effects
account for any unobserved monadic factors that may influence trade in a year.
As a formula, we incorporate MTR as

( ) ( ) ( )
ln Tij = a + b1 * ln ( Yit ) + b2 * ln Yjt − b3 ln Dijt + b 4 * ln (G t )

( )
+ b5 * ln Ωjt + b6 * ln (Pit ) + eijt
(4)

where Ωj is the “market potential” of the exporter (Baldwin and Taglioni 2006, 4) and Pi
is the price of trading with the exporter.
The alternative methods discussed below are easier to understand if we regroup the
MTR elements thus:

( ) ( )
ln Dijt − ln (Pit ) − ln Ω jt + ln (G t ) (5)

Because both Pit and Ωjt are constants within a country-year, we can estimate them
with country-year fixed effects, as Baldwin and Taglioni (2006) argue, but this causes Gt
to drop.9
There are benefits and limitations to including country-year dummies. Because
we are controlling for unobserved monadic factors that influence trade, we get bet-
ter estimates of the effects of trade institutions.10 Because many models limit the
monadic variables to GDP, the effects of other monadic variables, such as tariff and
nontariff barriers and shipping costs, are consigned to the error term. If countries
that join the GATT or PTAs have lower trading costs, then the institution’s effects
will be overestimated, even relative to one-in trade. Because the country-year fixed
effects control for all these unobservable monadic factors, we can derive more accu-
rate estimates of the effects of institutions, obviating the need to collect tariff and
nontariff barrier data.
A second benefit is that the sample of countries is less restricted than is the case with
only dyadic and year fixed effects; specifically, we can use observations that lack GDP
data. Because tariffs were an important source of government revenue for many coun-
tries before World War II, detailed sources of bilateral trade are available for the inter-
war period and earlier. Gross domestic product, on the other hand, was not calculated
before World War II, and pre-1939 estimates of GDP are based on estimated growth
rates, limiting country coverage. Country-year fixed effects allow us to bypass these and
other limitations and use all of the available trade data.
 Raymond Hicks   83

On the other hand, the questions we can ask about trade differ because the vari-
ables we can include are limited and the coefficients are interpreted differently. When
country-year fixed effects are included, we cannot include control variables such as GDP,
population, or exchange rate variability (Magee 2008). The model controls for these
variables and other unobserved monadic level factors. If a dummy variable approach
is used, the coefficients for these variables may be estimated, but the coefficient will
be for an excluded country-year dummy variable rather than the GDP or population
effect. This point cannot be stressed enough. Depending on which time-varying coun-
try effects are dropped from the model, the coefficient on GDP may fluctuate widely and
should not be interpreted.
Similarly, we cannot simultaneously include variables for both-in and one-in for an
institution, as together they will be collinear with the country-year fixed effects. The
coefficients in a model with country-year and dyadic fixed effects show the change in
trade when both countries are in the institution compared to the change in trade when
only one country in the dyad is in the institution. (If both countries are in an institution,
then it cannot be the case that neither is an institution for the year.) The one-in effect is
picked up by the country-year coefficients; the coefficients on the both-in institutions
are picking up the change in trade relative to this excluded group.
We cannot tell if an institution is trade-creating or trade-diverting, and we cannot
determine whether the institution led to an increase in trade compared to preinstitu-
tion trade; because the trade institution coefficient shows the change in trade relative
to one-in trade, it could be positive for two reasons. First, one-in and both-in trade
could both increase, but both-in trade by more than one-in trade. Alternatively, both
could decrease, but both-in trade by less than one-in trade. In either case, both-in trade
increases relative to one-in trade, and its coefficient will be positive with country-year
and dyadic fixed effects. This shows the importance of knowing which questions we are
asking. It would be incorrect to interpret the coefficient on trade institutions as showing
that it increased trade compared to preinstitution trade.
The biggest limitation with the dummy variable approach to country-year fixed
effects is that the number of variables needed to account for the country-year
fixed  effects grows quickly as the number of trading partners and years in the
data increases and may exceed the computing power of many statistical packages.
Stata, for example, allows a maximum of 11,000 variables on the right-hand side.
With 60 years of trade data and an average of about 150 importers and 150 export-
ers per year, the variable limit is quickly reached. While we can condition out or
absorb the dyadic fixed effects, we may have to account for the importer-year and
exporter-year fixed effects with dummy variables. With large data sets, this just may
not be possible.
The study of interwar trade by Gowa and Hicks (2013) is one of the few political sci-
ence papers that empirically controls for country-year effects using dummy variables.
Because their sample covers only about sixty countries over twenty years, the number
84   Methodological Issues

of importer-year and exporter-year fixed effects is manageable. They find that the cur-
rency and trade blocs had little effect on trade. Insofar as any of the blocs did affect trade,
they decreased trade relative to trade with nonbloc members. The only exceptions were
the trade blocs centered around the United Kingdom and Germany, which did see an
increase in hub-spoke trade relative to their trade with other countries.

Alternative Methods
For other studies, especially those examining trade after World War II, the number of
dummy variables needed to account for importer-year and exporter-year fixed effects
is too large to include. In order to resolve the computing issues, economists have advo-
cated a number of alternative methods to approximate multilateral trade resistance.
I discuss a few of these methods here.
Tetrad approach Head, Mayer, and Ries (2010) utilize the “tetrad” approach, which
involves eliminating the monadic effects through the ratio of each variable with a com-
mon importer (k) and exporter (l).11 Essentially, we take the value of a variable between
country 1 and country 2 (Dijt) relative to country 1’s value with the common importer
(Pikt) and then take the ratio of that quantity with country 2’s value with a common
exporter (Ωjlt) relative to the value between the common importer and exporter
(Gklt).12 The values of Pikt and Ωjlt are constant within importer-year and exporter-years
and Gklt constant within a year. After transforming the variables, the dyadic values are
demeaned to remove the dyadic fixed effects. Finally, the standard errors are corrected
to account for the clustered nature.
The coefficients should then be similar as when including country-year and dyadic
fixed effects. So rather than including dummy variables for the time-varying country
fixed effects, we remove their effect. As with the inclusion of country-year dummy vari-
ables, variables such as GDP and GDP per capita should not be included in the model.
The major shortcoming with the tetrad approach, as Head, Mayer, and Ries admit, is
that the results will depend on the reference importer and exporter chosen (2010, 3).
Specifically, if a country does not trade with either reference country in a year, the ratio
of ratios cannot be constructed, and the observations for that country-year will be
dropped.
Bonus  vetus Unlike the tetrad method, Baier and Bergstrand (2009, 2010) esti-
mate MTR terms for each variable. As discussed above, MTR is based on the bilateral
costs of trade (Ωjt—Pit) compared to the average costs that each faces with all trading
partners (Gt; Anderson and Van Wincoop 2003). Baier and Bergstrand discuss both
an unweighted method (2010) and a weighted method (2009) to estimate MTR. The
unweighted method adds in MTR as the average of the importer costs and the average
of the exporter costs minus the world costs for each variable.13 The weighted method
weights the value by the partner country’s share of world income and then sums the
weighted values by country. Again, the average for the world, this time weighted, is
subtracted from each observation.
 Raymond Hicks   85

In both the weighted and unweighted versions, the logged trade value is normalized
by subtracting the logged GDP for both the importer and the exporter. After these trans-
formations, standard OLS can be used on the new variables. Thus, Baier and Bergstrand
are able to control for MTR rather than adding in dummy variables for each country or
country-year. In their two articles, Baier and Bergstrand examine cross-sectional data.
In the comparison below, I add a year component to the transformations to account for
the panel structure of the data.14
Reg3hdfe The most promising innovation is a user-written Stata command called
reg3hdfe, which stands for regression with three high dimension fixed effects (Guimaraes
2010; Carneiro, Guimaraes, and Portugal 2012).15 Similar to the tetrad approach, the
command conditions out three levels of fixed effects instead of including dummy vari-
ables. Rather than calculating a coefficient for each of the separate fixed effects, we cal-
culate a new variable for each set of fixed effects.
The estimation process involves iterating over regressions, changing the values of the
fixed effects observations after each iteration, until no further improvement in model fit can
be made. The first step involves running a normal OLS with the values of the fixed effects
set to 0. The residuals for the model are computed. Each fixed effect is then calculated as
the mean of the residual plus the fixed effect coefficient multiplied by the value of the fixed
effect. These values are then included in the next iteration of the regression. Iterations con-
tinue until the difference between the residual sum of squares in the new and previous
model is below a certain threshold or tolerance. At that point, the coefficient on the fixed
effects variable will be equal to 1. The value of each fixed effects observation will be equal to
the coefficient on the fixed effects if each had been entered as a dummy variable.16
With two fixed effects, we would solve a system of three equations in a similar man-
ner. Because the system of equations becomes more computationally intensive as
the number of fixed effects increases, we can condition out one of the fixed effects by
demeaning the data first. We again iterate over estimations until the desired tolerance
is reached.
A comparison of the results of Gowa and Hicks (2013) that include dyadic fixed effects
and dummy variables for the county-year fixed effects to their results using the reg3hdfe
method shows the benefit of the method. As shown in Table 5.1, the coefficients are
the same to the fourth or fifth decimal place. More important, there is a vast improve-
ment in the computation time, as shown in the last row of the table. The dummy vari-
able approach took almost seven minutes to run (409 seconds), while reg3hdfe took just
under a minute and a half (84 seconds) with the tolerance set to .001. The coefficients
for a couple of the variables differ very slightly at the fourth decimal place. As shown in
model 3, the similarity of the coefficients can be improved by decreasing the tolerance
level, though at the cost of increased computational time.17
Moreover, because the reg3hdfe command does not fall victim to the variable limita-
tion that has plagued the application of multilateral trade resistance to the post–World
War II analysis of trade, we can use the method as a baseline with which to compare
dyadic and year fixed effects models and the alternative methods.18 First, though,
I briefly discuss the data used for the comparison of methods.
86   Methodological Issues

Table 5.1  Comparison of Methods: Interwar Period

(2) Reg3hdfe (3) Reg3hdfe
(1) Dummy Variable Tolerance = .001 Tolerance = .0001
b/se b/se b/se
Gold bloc 0.03979 0.03988 0.03976
(0.09191) (0.09193) (0.09191)
IPS –0.03758 –0.03759 –0.03760
(0.24971) (0.24973) (0.24969)
Reichsmark bloc –0.31964 –0.31969 –0.31963
(0.19908) (0.19900) (0.19909)
Exchange controls –0.72973*** –0.72967*** –0.72972***
(0.12707) (0.12705) (0.12706)
Sterling –0.10207 –0.10211 –0.10208
(0.09925) (0.09923) (0.09924)
Alliance –0.17840** –0.17841** –0.17840**
(0.06968) (0.06966) (0.06967)
Joint democracy 0.15609* 0.15609* 0.15609*
(0.08108) (0.08108) (0.08108)
N 35199 35199 35199
Time to run 411.437 83.820 122.211

Note: Model 1 uses Stata’s areg with dyadic fixed effects absorbed and dummy variables
for importer-year and exporter-years. Models 2 and 3 use reg3hdfe with dyadic,
importer-year, and exporter-year fixed effects.

Trade and the GATT

Only relatively recently have scholars examined whether the GATT/WTO increases
trade among its members, finding different effects of the GATT.19 These discrepan-
cies are due partly to the different methods employed. Most of Rose’s specifications
include only year fixed effects with standard gravity model controls for both dyadic
and monadic factors. Other authors use both year and dyadic fixed effects. All of them
ignore the unobservable monadic components of trade, assuming that GDP and popu-
lation will be sufficient to capture them.
Another limitation of the existing studies is that they focus only on whether both-in
trade increased compared to its pre-GATT trade. That is, they do not examine whether
a GATT member’s trade with other GATT members increased more than its trade with
non-GATT members. While Rose (2004) and Goldstein, Rivers, and Tomz (2007) both
 Raymond Hicks   87

include a one-in GATT variable in their articles, neither compares its effect to the effect
when both dyad members participate in the GATT. There is no attempt to measure the
relative trade-creation and trade-diversion effects. Since the both-in and one-in GATT
coefficients are both positive, this will reduce the GATT’s overall effect.
Data To compare the estimation techniques, I use data from Goldstein, Rivers, and
Tomz (2007). They have the most comprehensive list of GATT/WTO members, having
gone back to the GATT archives to code formal members and nonmember participants.
Their models include dichotomous variables indicating whether both countries are
GATT members and whether only one participates, as well as dichotomous variables for
the presence of a reciprocal PTA, a nonreciprocal PTA, and a currency union. There are
also variables indicating whether the generalized system of preferences (GSP) exists in
the dyad and whether the countries are in the same colonial orbit—that is, whether both
countries are existing colonies of the same metropole or the dyad includes the metro-
pole and a colony. Finally, they include the logged product of GDP of the countries in
the dyad. The dependent variable is the log of imports in 1967 US dollars.
With dyadic fixed effects, we need at least one year in each dyad where one of the
dichotomous variables, such as GATT participation, is equal to 0.  To accommodate
this, Goldstein, Rivers, and Tomz collected trade for 1946 and 1947 to supplement the
IMF data, which begin in 1948, the same year the GATT began. While the extra data
are sufficient to estimate a GATT effect, it is questionable how robust the estimate is, as
I discuss below.

Comparison of Methods

Table 5.2 shows the results of the different analyses. Model 1 replicates the model using
dyadic and year fixed effects from Table 2 of Goldstein, Rivers, and Tomz (2007). The
coefficient of 0.35 for GATT participation represents an increase in dyadic trade of
about 42 percent. This probably overestimates the impact of the GATT. Trade between
one GATT country and one non-GATT country also increased by 24 percent during
this period. The difference between the two coefficients is still highly significant, so the
GATT still had a significant impact on trade, but the effect is smaller than as reported by
Goldstein, Rivers, and Tomz (~18 percent).
Model 2 shows the results of including importer-year and exporter-year fixed effects
with dyadic fixed effects (using reg3hdfe). Because of collinearity issues, the GATT
one-in variable cannot be included in the regression; neither can the GDP variable.
Any year-to-year factors that impact all partner countries equally will be picked up by
the country-year fixed effects. The coefficient on the GATT both-in variable picks up
change in trade in the year relative to the base group of trade; that is, relative to its change
in trade with a non-GATT country. There is a much smaller, though still significant,
effect of GATT membership. Compared to a GATT member’s trade with non-GATT
Table 5.2  Comparison of Methods: Post-World War II
(2) Country-year (4) Tetrad: (5) Country + (6) Bonus Vetus (7) Bonus Vetus
(1) Dyad + Year + Dyad (3) Tetrad: US–UK US–Japan Year (unwt) (wt)
b/se b/se b/se b/se b/se b/se b/se
GATT: Both-in 0.354*** 0.064** 0.177*** 0.064 0.383*** 0.150*** 0.424***
(0.034) (0.033) (0.055) (0.062) (0.042) (0.055) (0.051)
GATT: One-in 0.200***
(0.029)
Reciprocal PTA 0.343*** 0.290*** 0.500*** 0.370*** 0.301*** 0.238*** 0.250***
(0.022) (0.022) (0.029) (0.034) (0.027) (0.035) (0.031)
Nonreciprocal PTA –0.053* 0.113*** 0.146*** 0.045 0.087** 0.111** 0.275***
(0.032) (0.032) (0.035) (0.042) (0.037) (0.046) (0.039)
GSP –0.099*** –0.122*** –0.072*** –0.096*** –0.104*** –0.184*** –0.402***
(0.019) (0.020) (0.023) (0.021) (0.023) (0.029) (0.022)
Currency union 0.492*** 0.287*** 0.537*** 0.146 1.079*** 1.146*** 1.326***
(0.088) (0.078) (0.065) (0.098) (0.089) (0.099) (0.103)
Same colonial orbit 0.836*** 0.631*** 0.381*** 1.103*** 1.611*** 1.612*** 1.610***
(0.081) (0.087) (0.092) (0.125) (0.127) (0.154) (0.166)
GDP log product 0.661***
(0.011)
Log of distance –1.016*** –1.042*** –0.955***
(0.014) (0.020) (0.015)
N 381656 381656 346366 338073 381656 381656 381656
 Raymond Hicks   89

members, trade with other GATT members increases by about 7 percent. As Baldwin
and Taglioni (2006) argue, if the unobserved monadic factors are positively correlated
with greater trade, then the coefficients will overestimate the effect. Since countries that
join the GATT are more likely to trade more, the unobserved monadic factors should be
positively correlated with trade.
Interestingly, compared to model 1, the coefficient for nonreciprocal PTAs in model 2
switches signs. Nonreciprocal PTAs have a positive impact on trade when dyad and
country-year fixed effects are included. This difference probably reflects the fact that at
least one member of each nonreciprocal PTA dyad is a GATT member. There is thus
complete overlap between both-in GATT, one-in GATT, and nonreciprocal PTAs. In
model 1, the nonreciprocal variable is similar to an interaction, showing trade in nonre-
ciprocal PTAs decreased compared to both-in or one-in GATT trade. But the coefficient
is showing the average difference from both-in and from one-in. Goldstein, Rivers, and
Tomz examine the relationship between nonreciprocal PTAs and GATT membership
more closely, finding that nonreciprocal PTAs do have a positive effect in the one-in
GATT dyads (2007, 60–61).
This is confirmed by the dyadic and country-year fixed effects results. Since these
compare the nonreciprocal PTA coefficient to the excluded group, they show that
nonreciprocal PTAs increase trade, relative to non-GATT trade. All the countries that
give nonreciprocal benefits are in the GATT, but not all recipients are. For the provid-
ers of nonreciprocal benefits, the nonreciprocal PTA variable removes the beneficiaries
from the base group. For the non-GATT beneficiaries, the nonreciprocal PTA coeffi-
cient shows the change in trade relative to the base group. All of this suggests that nonre-
ciprocal PTAs do have some trade benefits, though the benefits are not as large as when
both countries belong to the GATT.
There is not much change in the coefficients for the other variables—the effects
of PTAs and the GSP are very similar. The coefficients on GSP suggest that not only
does trade decrease when the GSP is introduced, but trade with other countries is not
affected. The similarity of the coefficients on reciprocal PTAs suggests that they do
increase trade within a dyad, and that there may not be that much trade-diversion going
on.20 Currency unions and colonial orbit both continue to have very strong effects on
trade, though the effect weakens somewhat with the country-year fixed effects.
In models 3 and 4 I present two models using the tetrad approach. As Head, Mayer,
and Ries caution, the results can change depending on the countries used as the refer-
ence categories. In model 3, the United States is used as the reference importer and the
United Kingdom as the reference exporter; in model 4, the United States is again used as
the reference importer and Japan as the reference exporter.
The first thing to notice is that the sample size changes, both compared to models 1
and 2 and also across tetrad models. Data cannot be used in years where a country did
not trade with the reference importer or the reference exporter. Thus, it is important
when using the tetrad command to choose reference countries with many trading
partners.
90   Methodological Issues

Second, changing the reference countries has a very large impact on the estimated
coefficients.21 (As with the country-year fixed effects models, we cannot include vari-
ables for both-in and one-in the GATT in the same model.) The sizes of all of the coef-
ficients except GSP change drastically, and the coefficients in neither tetrad model are
particularly close to those estimated using country-year fixed effects. In model 3 the
coefficients on both-in GATT, PTAs, nonreciprocal PTAs, and currency unions are all
bigger than those in model 2.
The fifth model shows results with importer-year and exporter-year fixed effects,
including the log of distance as a time-invariant dyadic variable. With cross-section
data, including importer and exporter fixed effects is equivalent to controlling for mul-
tilateral resistance. Adding the time element accommodates for the panel aspect of the
data. There are no dyadic fixed effects in this model; instead dyad factors are accounted
for with the distance variable. Again, the effect of reciprocal PTAs and the GSP is very
similar to that in the other models. Nonreciprocal PTAs again have a positive coeffi-
cient. The effects of GATT both-in, currency unions, and colonial orbit are much bigger
than in the other models, probably reflecting unobserved dyadic factors.
Model 6 shows the results using the unweighted bonus vetus method. With the excep-
tion of the GATT both-in and the GSP coefficients, the coefficients are very similar to
those in model 5. While the coefficient on both-in GATT is more similar to the other
models, the GSP coefficient is smaller. The effects of currency unions and colonial orbit
are again larger than in the other models.
Finally, the results using Baier and Bergstrand’s weighted version of MTR are differ-
ent than the other results. Both-in GATT and nonreciprocal PTAs have much larger
positive effects and GSPs a much larger negative effect on trade.
4.1  Discussion While there is some overlap among the different methods, the results
are not as similar as in Monte Carlo simulations done by the methods’ advocates or by
Head and Mayer (2013), all of which show a remarkable similarity in results. One differ-
ence is that most of the simulations are done with a cross-sectional structure. Probably
the more important reason is that real-world data are messy. Countries break apart and
unify, so dyads will enter and leave the data set. Political systems change and alter trad-
ing patterns. Both will prevent data from having a nice rectangular structure. Indeed,
Head and Mayer (2013) show that as the number of missing observations increases in
simulations, the alternatives to the dummy variable approach perform more poorly. We
should not be surprised, then, that they do not do so well with real-world data.
Nevertheless, the comparison of the methods is enlightening both because of the dif-
ferences in coefficients and their similarities. The effects of reciprocal PTAs and the GSP
do not vary much—across the methods, the coefficient on reciprocal PTAs is around 0.3,
while the coefficient on GSP is around –0.10. Both of these variables have lots of prein-
stitution and institution observations, and neither has much of an effect on one-in trade.
The coefficients on the other variables, especially currency unions, colonial orbit, and
GATT participation, vary tremendously across methods. The largest effects for cur-
rency unions and colonial orbits are in the models that do not explicitly incorporate
 Raymond Hicks   91

dyadic factors, except for distance. This suggests that, as with Baier and Bergstrand’s
argument about PTAs, both are endogenous to trade.
Finally, why does the GATT effect vary so much? While it is significant in most of
the models, only sometimes does it exert a large impact on trade. The most important
reason may simply be the lack of pre-GATT data. The effect of a dichotomous variable
will be more consistent when there are a larger number of control and treatment obser-
vations. For the founding members of the GATT, there are only two years of pre-GATT
data available, 1946 and 1947, and the world was experiencing a lot of changes then.
One solution may be to collect trade data before 1946. While such data do exist, the
years before 1946 saw, first, the Great Depression, which lasted from 1930 to about 1938,
and then World War II from 1939 to 1945, both of which had their own effects on trading
patterns. This will present its own set of issues. If either the Depression or World War II
changed trading patterns, then adding in the extra years without accounting for
the change will not improve our estimates of the GATT effect. Imagine, for example,
that countries that were allies during World War II were more likely to trade with one
another during the war and were more likely to be founding GATT members. Including
the World War II years may attenuate the GATT effect, because average dyadic trade
before the GATT is larger.
A second limitation is, as discussed above, the fact that some countries such as the
United States extended their GATT tariff cuts to all countries with which they had MFN
status. Non-GATT trading partners of the United States benefited from the GATT even
as nonmembers. That is, in the absence of the GATT, it is doubtful that the United States
would have reduced trade barriers as much as it did under the GATT. The coefficient
on one-in GATT, then, will pick up some of the GATT reforms, perhaps causing us to
underestimate its effects. Thus, it may be difficult to obtain a true estimate of the GATT’s
effects given the historical peculiarities of its formation.

Conclusion

Recent advances in trade theory advocate controlling for multilateral resistance.

Including country-year fixed effects accomplishes this task. They also control for any
unobservable monadic factor that might affect trade. There is still a dyadic component
to trade, which should be accommodated for with the inclusion of dyadic fixed effects.
Ignoring multilateral trade resistance and including only dyad and year fixed effects
will lead to biased estimates, because any monadic unobservable variables are omitted,
but the bias will depend on the relationship between the unobserved factors and trade.
If there is a positive correlation, the effects of institutions such as the GATT will be over-
estimated. It seems likely that countries that trade more are also more likely to join the
GATT/WTO, so we would expect dyadic and year fixed effect models to overestimate
the GATT effect.
92   Methodological Issues

Economists have devised a few methods to control for MTR. Overall, the methods
do a good job when the data are cross-sectional or a rectangular panel. The more miss-
ing trade values in the data and the more that new dyads enter the data and old ones
leave, the worse these methods perform in comparison to the country-year and dyadic
fixed effects approach (see also Head and Mayer 2013). Because the reg3hdfe command
does produce the same results as including dyadic and country-year fixed effects, there
is no excuse for not running models with time-varying country factors. The drawback of
country-year fixed effects is that they cannot tell us whether the GATT increased trade;
they only tell us whether GATT membership increased trade relative to trade with
non-GATT members.
To conclude, I offer some recommendations for using the gravity model.

Recommendations
1.  When using fixed effects, be aware of the potential for collinearity.

This applies not just to gravity models, but to any analysis using fixed effects, espe-
cially ones that use a dummy variable approach. For example, when GDP is included in
a model with dummy variables for time-varying country effects, many statistical pack-
ages will drop one of the time-varying country dummies in order to calculate a coef-
ficient for GDP. The GDP coefficient should not be interpreted as the effect of GDP on
trade. If the fixed effects are conditioned out of the data, the collinear coefficients will
simply drop.

2.  Estimating the effects of institutions when using fixed effects requires a sizeable num-
ber of observations when countries are in and not in an institution.

As discussed above, one of the reasons that the GATT coefficient varies across meth-
ods is that there are only two years of pre-GATT trade data. The coefficients on recip-
rocal PTAs and GSPs, for which there is much more data both before and after the
institutions form, are much more consistent across methods.

3.  Use the method that most closely fits the question asked.

While this may be obvious, the different ways to estimate gravity models do affect
the questions that can be answered. Models that include dyadic and year fixed effects
are most flexible. They can answer whether institutions were trade-creating and
trade-diverting as well as the relative effects of trade-creation and trade-diversion. But
the effects will probably be overestimated.
In contrast, models that include dyadic and country-year dummies can only
answer whether trade between GATT members increased more than their trade with
non-GATT members. The coefficients show the change in trade when both countries
 Raymond Hicks   93

are in the institution compared to the change in trade when only one country in the
dyad is in the institution. The country-year fixed effects are picking up the change in
trade of the excluded category. This also means that if we were to swap the both-in vari-
able with a one-in GATT variable, the coefficient on the one-in variable would be half
the size of the both-in coefficient and of the opposite sign. If trade between two GATT
members increases relative to the trade of each with a non-GATT member, then trade of
the latter has to decrease relative to the former.
The best strategy, then, is to report results using dyadic and year fixed effects as well
as country-year and dyadic fixed effects. We can use the dyadic and year fixed effects
to determine the direction and approximate size of trade-creation and trade-diversion,
and the country-year fixed effects results will give us more precise estimates of the rela-
tive effects of the two. While we may never conclusively determine the true effects of
the GATT, using models with dyadic and year fixed effects and country-year and
dyadic fixed effects does give us a more precise picture of trade and the effects of trade
institutions.

Notes
1. Tinbergen was one of the first to apply the gravity model to trade, in 1962.
2. In the economics literature, Anderson (2011) provides a useful overview of the founda-
tions of the gravity model; Head and Mayer (2013) discuss the microfoundations of the
gravity model, but are also interested in different methods to estimate multilateral resis-
tance. Finally, Shepherd (2012) provides a more general discussion.
3. Distance is not only geographic distance, but also includes cultural closeness, such as colo-
nial linkages or a common language.
4. With no other right-hand side variables in the model, the coefficient would equal the dif-
ference between the average trade under the institution and the average trade before the
countries joined the institution.
5. Without dyadic fixed effects, the interpretation of the both-in coefficient changes when
the one-in variable is added. It is no longer in comparison to all excluded dyads, but to
dyads where neither member is in the institution.
6. This may be the case if countries in an institution apply the institution’s rules to both mem-
bers and nonmembers. The United States, for example, extended tariff cuts made under
the GATT to all countries with which it had most-favored nation (MFN) status. Similarly,
European Economic Community countries extended tariff cuts made in their move
toward a customs union to the countries with which they had MFN status. As discussed
more below, this complicates estimates of the GATT/WTO effect.
7. The coefficients on the year variables will obviously differ in the two models.
8. The value for MTR will be a constant for each region. Thus, we can include fixed effects for
each region as importer and for each region as exporter to approximate MTR, as Anderson
and Van Wincoop admit (2003, 180). Indeed, many cross-section analyses of trade have
done this.
9. Note that if we include country-year and dyadic fixed effects, Yit, Yjt, Dijt, and Gt will all
drop from the model. We can still estimate the effect of variables that are not constant
within a country-year, such as trade institutions (or democracy and alliances).
94   Methodological Issues

10. Because trade costs vary over time, it is not sufficient to include time-invariant dummy
variables for importers and exporters. They will also be collinear with the dyadic fixed
effects and will drop out of the model.
11. There is a Stata version of the tetrad command available on Keith Head’s web page: http://
strategy.sauder.ubc.ca/head/sup/.
12. Head, Mayer, and Ries (2010) use exports as the dependent variable. Country 1 would be
the exporter and country 2 the importer in their setup.
13. Recall that ln(A/B) is equal to ln(A)-ln(B). Thus, ln((Ωj*Pi)/ G) is equal to ln(Ωj)+ln(Pi)–
ln(G). Since we are removing this effect from each variable, we get (–ln(Ωj)–ln(Pi)+ln(G)).
14. Baier and Bergstrand suggest that “[i]ncluding GDP-share-weighted multilateral trade
costs could create an endogeneity bias” and prefer the unweighted trade costs (2010, 104).
15. The command is available at http://www.aeaweb.org/articles.php?doi=10.1257/mac.4.2.133.
16. Technically, the value is equal to the difference between the dummy variable coefficient
and the constant term.
17. The default tolerance level is much smaller (1.192e-07), and it takes more iterations for
the transformation of each variable to converge. The model in Table 5.1 took about five
minutes to run (297 seconds) at the default tolerance, still faster than the dummy variable
approach.
18. Head and Mayer (2013) do a similar comparison of techniques, with two notable excep-
tions. First, they run a Monte Carlo simulation, based on a value of 1 for distance and 0.5
for reciprocal PTAs. Second, they use a cross-section approach. In contrast, this chapter
uses real-world panel data (from Goldstein, Rivers, and Tomz 2007).
19. See Gowa’s chapter in this volume for a review of the studies. There are also large litera-
tures that focus on the effects of preferential trade agreements on trade (Rose 2004; Baier
and Bergstrand 2007; Goldstein, Rivers, and Tomz 2007; Magee 2008) and the forma-
tion of trade agreements (Baccini and Dür 2010; Baier and Bergstrand 2004; Baldwin and
Jaimovich 2010; Egger and Larch 2008; Magee 2003; Mansfield, Milner, and Rosendorff
2002; Mansfield, Milner, and Pevehouse 2007; Mansfield and Milner 2012).
20. Indeed, if we construct a one-in PTA variable that takes on the value of 1 if either the
importer or the exporter has a reciprocal PTA in the year, but not with one another, and
include it in the model from Table 5.2, model 1, the coefficient is insignificant, with a value
of about 0.003.
21. I ran the results for a variety of combinations and found similar variation in coefficients.

References
Anderson, John. 2011. The Gravity Model. Annual Review of Economics 3: 133–160.
Anderson, James, and Eric van Wincoop. 2003. Gravity with Gravitas: A Solution to the Border
Puzzle. American Economic Review 93(1): 170–192.
Baccini, Leonardo, and Andreas Dür. 2010. Investment Discrimination and the Proliferation of
Preferential Trade Agreements. Paper Presented at the Annual Meeting of the International
Political Economy Society, Cambridge, MA, November 12–13.
Baier, Scott, and Jeffrey Bergstrand. 2010. Approximating General Equilibrium Impacts of
Trade Liberalizations Using the Gravity Equation. In The Gravity Model in International
Trade, edited by Peter van Bergeijk and Steven Brakman, 88–134. Cambridge, UK:
Cambridge University Press.
 Raymond Hicks   95

Baier, Scott, and Jeffrey Bergstrand. 2009. Bonus Vetus OLS:  A  Simple Method for
Approximating International Trade-Cost Effects Using the Gravity Equation. Journal of
International Economics 77 (1): 77–85.
Baier, Scott, and Jeffrey Bergstrand. 2007. Do Free Trade Agreements Actually Increase
Members’ International Trade? Journal of International Economics 71, 1: 72–95.
Baier, Scott, and Jeffrey Bergstrand. 2004. Economic Determinants of Free Trade Agreements.
Journal of International Economics 64: 29–63.
Baldwin, Richard, and Dany Jaimovich 2010. Are Free Trade Agreements Contagious? NBER
Working Paper No. 16084.
Baldwin, Richard, and Daria Taglioni. 2006. Gravity for Dummies and Dummies for Gravity
Equations. NBER Working Paper No. 12516.
Carneiro, Anabela, Paulo Guimaraes, and Pedro Portugal. 2012. Real Wages and the Business
Cycle:  Accounting for Worker, Firm, and Job Title Heterogeneity. American Economic
Journal: Macroeconomics 4 (2): 133–152.
Egger, Peter, and Mario Larch. 2008. Interdependent Preferential Trade Agreement
Memberships: An Empirical Analysis. Journal of International Economics 76: 384–399.
Ghosh, Sucharita and Steven Yamarik. 2004. “Are Regional Trading Arrangements Trade
Creating?: An application of extreme bounds analysis.” Journal of International Economics
63(2): 369–395.
Goldstein, Judith, Douglas Rivers, and Michael Tomz. 2007. Institutions in International
Relations:  Understanding the Effects of the GATT and the WTO on World Trade.
International Organization 61 (1): 37–67.
Gowa, Joanne, and Raymond Hicks. 2013. Politics, Institutions, and Trade:  Lessons of the
Interwar Era. International Organization 67 (3): 439–468.
Gray, Julia. 2013. The Company States Keep:  International Economic Organizations and
Investor Perceptions. New York: Cambridge University Press.
Guimaraes, Paulo. 2010. Estimation of High-Dimensional Models. Presentation to the
Portuguese Stata Users Group, September 17, Braga, University of Minho.
Head, Keith, and Thierry Mayer. 2013. Gravity Equations: Workhorse, Toolkit, and Cookbook.
Sciences Po Economics Discussion Papers No. 2013–02.
Head, Keith, Thierry Mayer, and John Ries. 2010. The Erosion of Colonial Trade Linkages after
Independence. Journal of International Economics 81 (1): 1–14.
Magee, Christopher. 2003. Endogenous Preferential Trade Agreements. Contributions to
Economic Analysis & Policy 2 (1): 1–17.
Magee, Christopher. 2008. New Measures of Trade Creation and Trade Diversion. Journal of
International Economics 75: 349–362.
Mansfield, Edward, Helen Milner, and B. Peter Rosendorff. 2002. Why Democracies Cooperate
More: Electoral Control and International Trade Agreements. International Organization 56
(3): 477–513.
Mansfield, Edward, and Helen Milner. 2012. International Trade and Cooperation: The Domestic
and International Politics of Preferential Trade Agreements. Princeton, NJ:  Princeton
University Press.
Mansfield, Edward, Helen Milner, and Jon Pevehouse. 2007. Vetoing Co-operation: The Impact
of Veto Players on Preferential Trade Agreements. British Journal of Political Science 37
(3): 403–432.
Mansfield, Edward, Helen Milner, and Jon Pevehouse 2008. “Democracy, Veto Players, and the
Depth of Regional Integration.” The World Economy 31(1): 67–96.
96   Methodological Issues

Mathy, Gabriel, and Christopher Meissner. 2011. Trade, Exchange Rate Regimes and Output
Co-movement. NBER Working Paper No. 16925.
McCallum, John. 1995. National Borders Matter:  Canada-U.S. Regional Trade Patterns.
American Economic Review 85 (3): 615–623.
Rose, Andrew. 2004. Do We Really Know That the WTO Increases Trade? American Economic
Review 94 (1): 98–114.
Santos Silva, J. M. C., and Silvana Tenreyro. 2006. The Log of Gravity. Review of Economics and
Statistics 88 (4): 641–658.
Shepherd, Ben. 2012. The Gravity Model of International Trade. ARTNeT Gravity Modeling
Initiative. Bangkok: United Nations.
Tinbergen, Jan. 1962. Shaping the World Economy. New York: Twentieth Century Fund.
Viner, Jacob. 1950. The Customs Union Issue. New York: Carnegie Endowment for International
Peace.
 Pa rt  I I

D OM E ST IC S O C I E T Y
 Chapter 6

In dividual At t i t u de s

Jason Kuo a nd Me gum i Naoi

Introduction

Studies on individual attitudes toward trade have flourished over the past two decades.
These studies sought to understand how ordinary citizens form positions on trade pol-
icy and how voter attitudes can influence trade-policy outcomes. This research question
differs from the dominant approaches to trade policy, which have focused on the roles of
political and business elites (e.g., Schattscheider 1935; Bauer, de Sola Poole, and Dexter
1963; Milner 1989; Grossman and Helpman 1994; Goodhart, this volume; Plouffe, this
volume; Hatch, Bair and Heiduk this volume) and domestic political institutions in
making trade policy (e.g., Lohmann and O’Halloran 1994; Gilligan 1997; Mansfield,
Milner, and Rosendorff 2000; Mukherjee, this volume; Rickard, this volume; Kono, this
volume; Chase, this volume).
Three factors have contributed to the emergence of individual-level studies on trade
policy. First, technological advancements in survey research (such as use of the Internet,
randomized experiments, and computer-assisted enumeration) have vastly reduced the
costs and time required for public opinion research. Second, formal, political economy
models of trade-policy making, such as endogenous trade-policy theory developed
since the 1980s (Magee, Block, and Young 1989; Mayer 1994; Grossman and Helpman
1994), explicitly address how legislators weigh votes and interest group contributions.
Yet the development of this research program has been lopsided; that is, formal mod-
els with little empirical understanding of how voters view and influence trade policy
have accumulated. The literature on individual attitudes has sought to fill this gulf.
Finally, the previous elite-level analyses have failed to address emerging empirical puz-
zles. Examples of such puzzles are why we see more free-trade policies when elite-level
theories have predicted a prevalence of protectionism (Grossman and Helpman 1994;
Gawande, Krishna, and Olarreaga 2009) and why consumers, who are assumed to be
free-trading in economic studies, are often protectionist (Coughlin 2002; Naoi and
Kume 2011).
100   Individual Attitudes

This chapter argues that although these motivations have substantially advanced
this research program, the research has suffered from two major issues. The first is the
missing link between individual attitudes and behaviors. Because the majority of work
has analyzed citizens’ stated preferences via public opinion surveys, the validity of the
results comes into question, since it is unclear that costlessly expressed preferences can
actually inform more costly behaviors, such as citizens’ voting behaviors in elections or
consumption decisions.
The second is the missing link between individual attitudes and group behaviors,
through collective action and aggregation via social, political, and economic institutions.
While the elite-centered approach has addressed how organized groups influence trade
policy through collective actions (e.g., Olson 1965), the individual attitudes literature
has largely ignored the question of interest aggregation and collective action, thus leav-
ing much to be explained. Relatedly, research linking elite behaviors and attitudes with
those of citizens has rarely gone beyond correlational studies using a one-shot public
opinion survey. This begs the question of whether elites mold citizens’ minds or respond
to and represent citizens’ trade-policy preferences. The remainder of this chapter criti-
cally discusses the development of this research program, the remaining questions to be
solved, and some promising lines for future research.

Economic Origins of Trade Preferences

Earlier empirical research on individual attitudes toward trade was largely aimed at
solving the theoretical disagreement over whether class (i.e., factor of production à la
the Stolper-Samuelson theorem) or industry of employment (i.e., exporting or import-
ing sectors à la the Ricardo-Viner model) determines individual positions on trade.1
Kenneth Scheve and Mathew Slaughter (2001b) offered the first empirical test of the
two competing claims using nationally representative public opinion data from the
American National Electoral Study (ANES). They showed that in the United States,
respondents’ factor profile, which was measured by years of education, better explained
the trade preferences of citizens than their industry of employment. Consistent with the
Stopler-Sameulson theorem, respondents with higher levels of education were more
likely to support free trade than were those with less education.
The factor-based trade theorem finds empirical support beyond the US case.
Using public opinion data obtained from twenty-four countries surveyed in the 1995
International Social Survey Program (ISSP), O’Rourke and Sinnott (2001) provided one
of the first proper cross-national tests of the Stopler-Sameulson theorem, by controlling
for country-level factor endowment, measured by GDP per capita. They demonstrate
that in advanced industrialized countries, high-skilled respondents (measured by years
of education) are more likely to support free trade, but this pattern is reversed in devel-
oping economies. Mayda and Rodrik (2005) obtained similar results, using the same
ISSP 1995 data as well as the 1995–1997 World Value Survey (WVS) covering forty-seven
 Jason Kuo and Megumi Naoi    101

countries. These findings were consistent with the factor-based trade theorem rather
than the industry-based models of trade.
Despite the fact that the original Stolper-Samuelson theorem formulates both pro-
ducer and consumer benefits of free trade (Baker 2005; Helpman 2011), empirical tests
of these models using public opinion surveys have been limited to a focus on job-related
attributes, such as respondents’ skill levels, income and sector of employment (Naoi and
Kume 2015). Andy Baker (2005) shifted the focus from production profiles of respon-
dents to consumer profiles; that is, what bundle of goods citizens consume. He relaxed
the assumption that consumers consume the same bundle of goods and hypothesized
that consumers of exported goods would be more likely to prefer protection to free
trade, as the expansion of exports raises the price of those particular goods. On the
flip side, consumers of imported goods would be more likely to support free trade, as
the expansion of imports lowers the prices on these goods. The 1995–1997 WVS data
covering both developed and developing countries lends support to this hypothesis,
although the survey does not directly measure the consumption bundle. The question
thus remains whether we can use respondents’ income level as a proxy for differences in
their consumption bundles, as done by Baker (2005).
These early studies also found several deviations from what trade models have
predicted. For instance, Scheve and Slaughter (2001b) showed that homeowners in
depressed regions were more protectionist than non-homeowners, controlling for their
factor profiles and sectors of employment. The authors interpreted from the results that
homeowners in regions that are adversely affected by trade are more likely to worry
about declining asset prices in a region. Hence, the housing market ties the income of
homeowners to that of declining industries, regardless of the homeowners’ sectors of
employment.
These deviations, however, do not necessarily mean that we can no longer improve
our understanding of the economic origins of individual trade-policy preferences.
We see several promising lines for future research. First, these early studies took fac-
tor mobility in the national economy and individual skill specificity (to the extent that
individual work skills are useable in other industries) as given when testing the rival
theories of international trade, despite the fact that they are two of the key conditions
under which class or sectoral conflicts arise in trade (Magee, Block, and Young 1989;
Hiscox 2002). National-level measures of factor mobility over time, such as wage dif-
ferentials across industries discussed in Hiscox (2002), would be a useful addition to
future cross-national analysis of individual attitudes. Likewise, new measures of the
individual-level skill specificity, like the one calculated by Torben Iversen for several
waves of ISSP data (see Cusack, Iversen, and Rehm 2006; Iversen and Soskice 2001),
could be a useful addition.
Second, scholars have pitched two competing trade models as rival theories by con-
trolling for possible covariates. This mode of causal inference, however, is based on a
strong statistical assumption that one theory trumps the other for the entire set of obser-
vations analyzed. The assumption does not allow scholars to specify an empirical scope
under which one theory does better than the others. To address this problem, Imai and
102   Individual Attitudes

Tingley (2012) developed the finite mixture modeling approach, which allowed some
observations to be better explained by one theory and the others by its rival. Adopting
this statistical method could help improve our understanding of the empirical scope of
cases in which one theory does better than the other.
Third, economic self-interest hypotheses have not been fully tested, for two reasons.
The first reason is that this research program needs a better measure of respondents’
economic interests in trade. The early studies, for instance, used years of respondents’
education as a proxy for their skill levels (i.e., factor profile), yet education could highly
correlate with other things that determine trade attitudes, such as ideology, gender, and
openness to foreign culture and new products (i.e., “cosmopolitanism”). Likewise, as
Melitz’s (2003) model of heterogeneous firms suggests, even individuals working for the
same industry can have diverging economic interests in trade. Workers in highly pro-
ductive firms can support free trade, and workers in low productive firms can oppose it
(see Naoi and Urata 2013).
The second reason is that while the menu of trade policy options that governments
can adopt is complex in reality, the majority of survey instruments used have asked for
respondents’ opinions about “placing new limits on imports” or “increasing trade” (see
Table 6.1). While these questions might allow scholars to solicit citizens’ gut-based, gen-
eral reactions to trade, these survey instruments often diverge from the day-to-day con-
text in which citizens think about trade policy.
We note three points in particular. First, political debates surrounding trade are most
often about the increase in exports from a particular country (e.g., China or the United
States) or the signing of free trade agreements with particular countries. Yet only a hand-
ful of studies have examined individual attitudes toward trade with particular countries
(Baldwin and Magee 2000 on North America Free Trade Agreement, NAFTA; Hicks,
Miller, and Tingley 2014 on Central America Free Trade Agreement, CAFTA; Chiang,
Liu, and Wen 2013 on Economic Cooperation Framework Agreement, ECFA; Naoi and
Urata 2013 on Trans-Pacific Partnership Agreements, TPP).2 Moreover, survey instru-
ments on trade policy rarely incorporate institutional designs of international trade
agreements negotiated between or among states (e.g., Koremenos, Lipson, and Snidal
2001; Johns and Peritz, this volume; Kim, this volume; Davis and Wilf, this volume;
Busch and Pelc, this volume). Survey conducted by the Program on International Policy
Attitudes (PIPA) showed that public support for lowering trade barriers seemed contin-
gent upon the principle of reciprocity (see Table 6.1 or Scheve and Slaughter 2001a, 16).
Exploring effects of “rational design” of international trade institutions on trade prefer-
ences is a promising line of future research.
The second point is that the majority of the remaining trade barriers today are non-
tariff barriers, comprising regulatory and nonregulatory differences among states.
Although existing survey instrument phrasing such as “placing new limits on imports”
can include nontariff barriers, asking citizens more pointed questions about raising or
lowering the standards of regulation, such as the safety and quality of products, would
offer much needed insight. Similarly, because liberalization of service trade is the key
 Jason Kuo and Megumi Naoi    103

Table 6.1  A Variety of Survey Instruments for Trade Policy Preferences

Data sources Question wording Examples of works

ANES 1992 & 1996 Do you favor or oppose placing Scheve & Slaughter (2001b);
new limits on imports, or haven’t Hainmueller & Hiscox (2006)
you thought much about this?’
ISSP & GSS 1995 How much do you agree or O’Rourke & Sinnott (2001); Mayda
disagree with the following & Rodrik (2005); Daniels & von der
statement? (respondent’s country) Ruhr (2005); Hainmuller & Hiscox
should limit the import of foreign (2006)
products in order to protect its
national economy.
WVS 1995-1997 Do you think it is better if Mayda & Rodrik (2005); Baker
(1) Goods made in other countries (2005); Kaltenthaler et al. (2004)
can be imported and sold here if
people want to buy them; or that
(2) There should be stricter limits
on selling foreign goods here, to
protect the jobs of people in this
country.
TESS 2004 Do you favor or oppose increasing Hiscox (2006)
trade with other nations?
PIPA 1999 Tell me which one you most Scheve & Slaughter (2001a, 16)
agree with:
A: The U.S. should lower barriers
even if other countries do not,
because …
B: The U.S. should only lower its
barriers if other countries do,
because …
Original (Japan) & CCES 2012 Do you support or oppose the Naoi & Urata (2013)
Japanese (or U.S.) government’s
participation in the Trans-Pacific
Partnership agreement?

agenda of trade negotiations today, soliciting citizens’ attitudes toward service trade lib-
eralization would be an interesting line of inquiry.
Finally, citizens’ opinions about trade could be industry specific. Scheve and Slaughter
(2001a, 24–25) reviewed the battery of surveys done in the United States on public sup-
port for protecting specific industries, such as steel, textiles, and footwear. Naoi and
Kume (2011) found that Japanese citizens’ support for protectionism is low for general
imports, yet their support for protectionism more than doubles for agriculture, espe-
cially among workers with low job security. Lu, Scheve, and Slaughter (2012) also found
that both American and Chinese citizens show higher support for protecting low-skill
104   Individual Attitudes

industries than high-skill industries; they attribute this pattern to citizens’ inequity
aversion and redistributive concerns. Asking citizens’ opinions about placing limits
on particular industries, as done by Rho and Tomz (2013), is a promising line of future
research.

Beyond Economic Self-Interests

The second wave of research has challenged these economic, self-interested explana-
tions by identifying alternative sources of individual trade-policy attitudes, such as ide-
ology and social identity.

Ideational Factors
Scholars have documented the effect of ideational factors beyond economic self-
interests, such as economic ideology, openness to outsiders, and religious beliefs, on
individual attitudes toward trade.
To the extent that trade is viewed as causing an increase in inequality, citizens’ eco-
nomic ideology, especially their normative views on the role of government in income
redistribution, should affect their attitudes toward trade. Those who hold left-leaning
ideology (i.e., supporting a larger role of government in income redistribution and
regulation) should oppose free trade, while those who hold right-leaning ideology (i.e.,
supporting small government and an environment in which the market determines
winners and losers) should support it. Indeed, studies on US legislators found that
the ideology of legislators affected their positions on trade (Milner and Tingley 2011).
Kaltenthaler, Gelleny, and Ceccoli (2004), however, found no evidence of the effect of
ideology among the masses, using the 1995–1997 WVS data.
Another series of studies has shown that citizens might see international trade as cul-
tural exchanges (i.e., benefits or threats) rather than economic exchanges. As a result,
citizens’ open-mindedness toward “outsiders” shapes their support for free trade.
Kaltenthaler, Gelleny, and Ceccoli (2004), for instance, using the same WVS data, show
that citizens’ degree of “cosmopolitanism,” a belief system viewing international trade as
“an opportunity to try new products and interact with the broader world,” impacts their
relative support for free trade.
Mansfield and Mutz (2009) have also shown that citizens who have anxieties about
being involved with “outer-groups” (immigrants, foreign culture, etc.) are more likely
to oppose free trade, and that low educational attainment is a good proxy for this
“outer-group” anxiety among American citizens. Consistent with these findings,
Margalit (2012) has shown cross-nationally and experimentally in the United States that
when individual citizens view international trade as a cultural threat rather than an eco-
nomic opportunity, citizens’ opposition to free trade increases.
 Jason Kuo and Megumi Naoi    105

Finally, citizens’ religious beliefs might shape their attitudes toward trade through
their predisposition against “outsiders.” Daniels and von der Ruhr (2005) demonstrated
that American voters who were pre–Vatican II Catholic or members of a fundamen-
talist Protestant denomination were more likely to prefer protection than those who
were post-Vatican II Catholic or members of liberal or moderate Protestant denomina-
tions. Evidence drawn from the 1995 ISSP and General Social Survey (GSS) supported
their claim.
These ideational findings are also consistent with earlier findings that protectionist
trade attitudes were associated with nationalist sentiments or psychological attach-
ments to regions, communities or nations, (O’Rourke and Sinnott 2001; Mayda and
Rodrik 2005).

Socialization Through Group Membership

Individuals might form their trade attitudes by socializing and interacting with others
through group membership. Goldstein, Margalit, and Rivers (2008) showed that mar-
ried individuals are more protectionist than single or divorced people, controlling for
other factors. They interpret this result as a consequence of married individuals having
concern for the labor market prospects of the spouse, where even individuals who are
themselves beneficiaries of free trade take more protectionist positions in concordance
with their spouse’s position. Whether the mechanism of marital convergence is through
self-selection into the legal arrangement, socialization, or income maximization (or the
minimization of income losses) for the household unit still remains a question.
Similarly, Ahlquist, Clayton, and Levi (2014) showed that workers who belong to anti-
trade unions are more protectionist than workers in the same sector who do not belong
to the union. The original survey of dockworkers on the US West Coast from 2007 to
2011 lends support to this argument.
Finally, Mansfield and Mutz (2009) have shown that citizens are more likely to form
their attitudes toward trade based on its perceived effects on the national economy,
rather than on their own jobs and income. They call this mechanism “socio-tropic” for-
mation of attitudes. Investigating more directly what contributes to this “socio-tropic”
formation—for example, whether it is driven by media coverage of economic news or
strong national identities—is a promising line of future research.

Gender
The most robust empirical finding that emerged from the first two waves of research is
that women were more likely to support trade protection than men, controlling for the
other factors. Disagreements exist over why. Burgoon and Hiscox (2006) showed that
neither the distributional effects of trade nor gender differences in skills and job char-
acteristics account for female protectionism in the United States. They alluded to the
106   Individual Attitudes

role of the gender gap in what people learned during college—that is, women take fewer
neoclassical economics courses than men—building on Hainmueller and Hiscox’s find-
ings (2006).
Analyzing cross-national survey data from the 1995 and 2003 ISSP, Beaulieu
and Napier (2008) showed that the gender gap in support of free trade could not be
explained by measurable characteristics such as education, labor market participation,
and country location. They thus concluded that the robust gender gap in trade-policy
stances could only be explained by differences in unobservable attributes of men and
women, such as biology.
Using a survey experiment in Japan, Naoi and Kume (2015) identified two distinct
sources of female protectionism, one on the labor market condition and another on
consumption. They found that women are more protectionist when a rigid and gen-
der-segregated labor market makes their job security lower than men’s (Estevez-Abe
2006), and they are also more protectionist consumers because they care more about
the safety and quality of food than men do. Indeed, marketing studies on organic food
consumption have shown that women are more likely than men to purchase organic
and natural food, which could lead to their opposition to cheap, imported food.
Further disentangling the sources of female protectionism is thus a promising line of
future research.

Risk Orientation
Finally, building on findings from behavioral economics, studies have found that
individual dispositions toward risk can account for individual attitudes toward trade.
Mayda, O’Rourke, and Sinnott (2007), using the Asia-Europe Survey (ASES) covering
eighteen countries in Europe and Asia, reported that risk aversion was associated with
anti-trade attitudes. Ehrlich and Maestas (2010) further showed that the college edu-
cated were more likely to support free trade if they were more risk-accepting to begin
with; by contrast, the non-college educated were less likely to support free trade if they
were risk-averse.
Although these studies challenge the economic self-interests approach in a funda-
mental way, there are several issues that need to be addressed. First, there is a chicken
and egg problem regarding the relationship between ideational factors and trade atti-
tudes. With single-shot survey data, we cannot trace which precedes the other (i.e.,
whether social identity shapes trade attitudes or trade attitudes lead to particular social
identities). Thus, these studies cannot rule out well-known problems of simultaneity
bias and reverse causality. Fordham and Klienberg (2012) therefore called for a more
careful look at the process of socio-tropic formation of individual attitudes.
Similarly, most of the time citizens’ decisions to join any particular group are by no
means random. This leads to a self-selection issue between group membership and
policy attitudes. It is thus important to be more mindful about identification strate-
gies, such as establishing in the first stage that group membership is orthogonal to trade
 Jason Kuo and Megumi Naoi    107

attitudes or identifying group membership that is highly path dependent. Alternatively,

propensity score matching or lab and survey experiments might help, though none of
these techniques fully addresses the problem of self-selection into groups that occurs in
reality.
Beyond identification strategies, a promising line of future research is to examine
the process and medium of information transmission from groups to individuals and
among individuals and groups. For example, asking citizens directly about how they
learn about trade policy and its distributional effects in surveys, as done by Bauer, de
Sola Poole, and Dexter (1963), would be an easy step forward.

Political Origins of Trade Preferences

The studies discussed above have examined how the economic and social attributes of
individuals are associated with their attitudes toward trade. By contrast, a series of stud-
ies have shown how institutions and the national political economy that individuals are
embedded in, such as electoral systems, partisan competition, and the variety of capital-
ism defining their economic environment, can shape citizens’ attitudes toward trade. In
particular, studies have shown that these political economy institutions influence citi-
zens’ attitudes through two distinct channels: income and information.
First, emerging empirical studies have verified the micro-foundation of what Ruggie
(1982) has termed “embedded liberalism,” that government policies to shield citizens
from income shocks of trade liberalization, through welfare and job-training programs,
can mobilize support for free trade among citizens. These studies show that as long as
citizens’ assessments of trade depend on its effects on income, their attitudes toward
trade are mutable with such government programs.
Using cross-national data on government spending and individual-level opinion data
on trade attitudes among OECD countries, Hays, Ehrlich, and Peinhardt (2005) have
shown that individual support for trade is higher in countries where governments pro-
vide generous welfare programs for labor and retirees. Ehrlich and Hearn (2013) further
support this finding using a survey experiment in the United States. Respondent groups
that were provided with facts about the Trade Adjustment Assistance (TAA) program
were more likely to support free trade than the control group without the information
about the TAA, and this increase was especially large among low-income respondents.
Walter (2010) further disentangles the causal chains with the use of public opinion data
from Switzerland and demonstrates that welfare expansion occurs through increased
support for the Left among trade losers.
Political and economic institutions can also shape citizens’ trade attitudes through
controlling consumer price. Through cross-national observational studies, Rogowski
and Kayser (2002) have shown that majoritarian electoral systems have lower consumer
prices because they raise legislators’ responsiveness to consumer interests over pro-
ducer interests. Building on this logic, Naoi and Kume (2015) have shown, using ISSP
108   Individual Attitudes

2003 data, that citizens’ support for free trade is actually higher in majoritarian systems
than in proportional representation systems when they are asked about trade’s benefits
for their consumer welfare. This finding is contrary to the conventional wisdom about
majoritarian electoral systems leading to lower support for free trade due to citizens’
concerns about lower job security (Hays, Ehrlich, and Peinhardt 2005).
An informational effect of institutions on citizens’ trade attitudes comes through
political persuasion and issue linkage in partisan competition. While the empirical evi-
dence is mixed at best on whether trade continues to be a partisan issue (Milner and
Judkins 2004; Naoi and Urata 2013), recent studies have shown that issue framing and
political persuasion during elections and referendums can powerfully shape individ-
ual attitudes toward trade. Using district-level referendum results on CAFTA in Costa
Rica, Hicks, Milner, and Tingley (2014) have shown that the left-leaning party was suc-
cessful at persuading voters to oppose signing a free trade agreement with the United
States. Likewise, Naoi and Urata (2013) show that anti-trade campaigns organized by
agricultural cooperatives and the opposition Liberal Democratic Party mobilized
much broader mass opposition to Japan’s participation in the Trans-Pacific Partnership
Agreement than could have been predicted from trade theorems.
Partisan orientations can also shape how consumers make consumption choices
regarding domestically produced versus imported goods. Ehrlich (2010a) has shown
that fair trade concerns (i.e., concerns about the labor conditions in developing econo-
mies) can mobilize protectionist sentiments among highly educated and wealthy liber-
als in the United States. Exploring how trade issues are linked to partisan debates and
ideologies in citizens’ minds is thus a promising line of future research.
The third informational effect of institutions on citizens’ trade attitudes is through the
structure of the media market in a given country. Because citizens learn the distribu-
tional effects of forthcoming trade policy mostly through media, who owns and finances
media can have profound effects on citizen attitudes. No study exists, to our knowledge,
of how citizens’ views on trade can be influenced by what the media report in authori-
tarian systems with government-controlled or censored media. In democratic contexts,
media ownership and marketing can also influence what trade-related news is reported.
Using the case of the TPP, a comprehensive trade agreement currently under negotia-
tion among twelve Asia-Pacific countries, Kuno and Naoi (2013) found that local news
outlets, such as prefectural-level newspapers and geographically targeted television pro-
grams, are more likely to report negative estimates on the effect of the TPP on the local
economy than are the national media in Japan. A promising line of future research is to
study how citizens learn about the distributional effects of trade policies through media
and interactions with elites.

International Politics: Trade as Foreign Policy

As much as domestic politics and concerns for redistribution influence how citizens
view trade, international politics can also affect citizens’ attitudes toward trade. Indeed,
 Jason Kuo and Megumi Naoi    109

cross-national studies have shown that military allies are more likely than nonallies to
sign free trade agreements, as trade generates security externalities to any third party
in the international anarchy (Gowa and Mansfield 1993; Gowa 1994; Gowa, this vol-
ume). Also, democratic pairs of countries are four times as likely to sign preferential
trade agreements as autocratic pairs, because to secure reelection, leaders in democra-
cies need to signal voters that they are working hard to grow the economy and lower the
price of goods (Mansfield, Milner and Rosendorff 2000).
Although studies linking trade and security alliances abound, most of these studies
have focused on the role of elites in maximizing national welfare or seeking reelection.
Few studies directly examine how voters view the effects of forming trade agreements
on their economic and security welfare. Public opinion studies, however, have shown
that anti-American or anti-Chinese sentiments among citizens, generated by the histor-
ical experience of conflicts, occupation, or the existence of military bases, can increase
citizens’ opposition to signing trade agreements with the United States or China (Kim
2011 on South Korea; Chiang, Liu, and Wen 2013 on Taiwan; Kume and Kohno 2011 on
Japan). These findings direct us to study how voters perceive the benefits and costs of
forming trade agreements with particular countries, such as security allies versus non-
allies and democracies versus autocracies.
In sum, in spite of substantial progress in understanding the economic origins of
individual trade attitudes, more research is called for on the political origins of trade
attitudes, rooted in domestic politics and international politics.

Experimental Approaches

The third wave of research on individual attitudes toward trade has used experimental
approaches, such as randomized survey experiments. Two considerations have spurred
this approach. The first is a lack of faith in the efficacy of using public opinion surveys as
proxies for citizens’ attitudes toward trade. Experimental studies have demonstrated an
instability and volatility in respondents’ positions on trade depending on how questions
are framed. Another concern is more empirical and aims to disentangle various sources
of endogeneity that caused stagnation in the progress in research on individual attitudes
toward trade. We review both of these literatures below.

Challenging the Validity of Survey

Responses: Framing Experiments
While their study was not experimental, Scheve and Slaughter (2001a) were among
the first to systematically document that the American public’s support for free trade
differs dramatically depending on the framing of the survey questions (see espe-
cially Chapter 2). For instance, American respondents were more likely to opt for
110   Individual Attitudes

protectionism if the wording of the question justified protectionism as “protecting

American industries and jobs” as opposed to when the wording justified free trade
as “permitting the widest choice and lowest prices for American consumers” (23).
Moreover, they show that public support for free trade is much stronger when trade
is described in broad and general terms (e.g., “how positive or negative you think the
growth of international trade is”), rather than with specific reference to the US econ-
omy (e.g., “trade with other countries … good or bad for the U.S. economy?”; 33–34).
American citizens’ support for forming “trade agreements” in general seems high,
if there is no specific reference to the United States or other countries (34–35). They
interpret this gap as showing that the reference to the US economy provokes respon-
dents’ mercantilist sentiments.
To what extent is this instability in respondents’ attitudes due to different issue fram-
ing and the wording of the instruments (e.g., telling respondents trade is good or bad)
or to different trade policies having diverging distributional effects? This is an impor-
tant distinction. For instance, “increasing trade” or “trade agreement” questions might
make respondents consider both imports and exports (both losing and winning sides
in a Ricardo-Viner sense), while a “limits on imports” question might make respon-
dents only think about imports (i.e., the losing side in the R-V model). These two ques-
tions suggest different distributional implications of trade, and that this difference goes
beyond the wording and framing of issues.
Hiscox (2006) provides one of the first survey experiments on the trade policy pref-
erences of individuals with a nationally representative sample in the United States. He
found that American citizens were very sensitive to job-threat framing—that is, trade
being a threat to American jobs—but that they were not sensitive to consumer-benefits
framing. He further found that less educated respondents were generally more sensitive
to issue framing, and in particular, to the job-threat framing.
The experiment spurred interest in understanding which citizens are more likely to
be framed by what type of information. In particular, there remains a question regard-
ing whether the large effect of “job threat” framing for less educated respondents is due
to their predisposition that trade is harmful to their job security (due to their low skill
levels) or to their sensitivity to the framing in general (due to their lack of sophistication
with political discourse).
Ardanaz, Murillo, and Pinto (2013) addressed this question directly by replicat-
ing the Hiscox experiment in Argentina, where there is a reversal of factor endow-
ment from that of the United States (scarce high-skilled labor, and an abundance
of low-skilled labor). This allowed them to separate out the effects of educational
attainment and job security on respondents’ sensitivity to framing. They found that
respondents’ material interests did shape their sensitivity to framing. The framing
was effective among those whose economic interests were ambiguous, such as service
sector workers, regardless of their levels of educational attainment or skills.3 When
respondents had clearly defined economic interests in trade, such as employment in
import-competing manufacturing sectors, their trade positions were stable and resis-
tant to various types of framing.
 Jason Kuo and Megumi Naoi    111

Disentangling Endogeneity and Collinearity:

Priming Experiments
The second wave of experimental studies has sought to address common endoge-
neity and collinearity issues that have hampered the progress on individual attitudes
toward trade.
For instance, Margalit (2012) documents with cross-national survey data the finding
that citizens who see trade as a cultural threat are more likely to oppose free trade. Yet
respondents who see the expansion of trade as a cultural threat tend to be those with low
educational attainment, making it harder to identify the causal role of “cultural threat
perception” that is orthogonal to respondents’ standing in the labor market. To address
this collinearity issue, Margalit (2012) also conducted a survey experiment in the United
States, which primed respondents to think about cultural threats versus economic
opportunity and proceeded to ask about their support for free trade. The results sug-
gest that, among respondents with low educational attainment, those who were primed
by “cultural threat” treatment showed higher opposition to free trade than those who
received no priming. How much of this priming effect is due to less educated respon-
dents’ strong prior attitudes about trade’s negative economic effects, and how much can
be attributed to their lack of sophistication with political discourse, remains a task for
future research.
Another group of studies has shown that citizens’ concern for others in the country
or community (“other-regarding preferences”), rather than self-interests, could dictate
their trade attitudes. This line of inquiry sees trade policy as a government’s decision to
redistribute income across society.
Naoi and Kume (2011) sought to explain why there is high public support for agri-
cultural protectionism in advanced industrialized nations, despite the fact that it
financially burdens consumers. The difficulty in getting at this question, they argue,
is that citizens can oppose agricultural liberalization as income-earners (e.g., they feel
sympathetic to low-wage/low-skill farmers), or, as consumers (e.g., they care about
the quality and safety of food). To solve this “duality of interests” problem, they con-
ducted a survey experiment in Japan, which randomly assigned visual images that
prime respondents to think about their jobs or about consumption and proceeded
to solicit their support for agricultural protectionism. The results suggest that the
main source of opposition to agricultural imports is labor market concerns, espe-
cially among the respondents with low job security, raising the support for protection
out of their own concerns for job security (Naoi and Kume dub this “the projection”
mechanism).
Lu, Scheve, and Slaughter (2012) have sought to solve the puzzle of why both
developed and developing countries protect low-skilled workers more heavily than
high-skilled workers. To do so, they conducted coordinated, nationally representative
survey experiments in the United States (where low-skilled workers should be trade los-
ers) and China (where low-skilled workers should be trade winners). They found that
112   Individual Attitudes

redistributive concerns, not self-interests, dictate citizens’ attitudes toward protecting

the low-skilled (proxied by their wage levels) workers regardless of low-skill workers’
position in the international economy, which they call “altruism,” and citizens oppose
protecting high-skilled (high-wage) workers in developing economies due to “envy”
toward the rich.
A promising line of future research is to ask whether these inequality aversions are
innate, psychological traits of individuals or byproduct of politics and policy programs
(e.g., profarmer discourse in Japan; small welfare programs in the United States and
China).

Conclusion: Research Frontiers

Research on individual attitudes toward trade has progressed substantially thanks to

newly available data, computational advancement, and the introduction of experimen-
tal methods to the field of international political economy. Yet the research program has
large room for improvement, such as measuring respondents’ economic self-interests
more carefully and paying more attention to how individual preferences are rooted in
the national political economy and international politics.
To conclude, we discuss two lines of the future research frontier. Both address the “so
what” criticisms that are commonly levied on the literature covering individual trade
attitudes discussed thus far. The first big critique is that opinions expressed through sur-
veys are a relatively costless form of stated preferences and have little bearing on the
actual behavior of individuals—such as voting and lobbying. In the survey, moreover,
respondents are often forced to take positions on trade issues among multiple choice
answers, despite their general low political salience.4 Indeed, Guisinger (2009) found
that trade is a low salience issue in the United States and that American voters rarely cast
their votes based on party or legislators’ positions on trade policy.5

Frontiers (1): Linking Attitudes to Behaviors

There have been three lines of inquiry to rebut this concern. One is to abandon public
opinion survey methods altogether and instead use behavioral data such as voting on
referendums on trade agreements or on an incumbent party. For instance, Irwin (1994)
has shown that in the 1906 British general election, where trade was the defining issue
for voters, voters in import-competing sectors opposed free trade, and those employed
in exporting or nontraded goods sectors supported it. This is one of the first empirical
studies that used district-level voting data to study constituents’ interests in trade. The
results lend strong support to the sector-based, Ricardo-Viner model.
Margalit (2011) has shown that trade-related job losses, measured by the number of
applications for the TAA program in US counties, led to vote losses for the incumbent
 Jason Kuo and Megumi Naoi    113

Republican Party in the 2004 presidential election. Likewise, Hicks, Milner and Tingley
(2014) examined district-level referendum data on the CAFTA trade agreement in Costa
Rica and found that districts with a higher proportion of low-skilled labor opposed the
agreement, contrary to the Stolper-Samuelson prediction. Using the same case of the
CAFTA referendum in Costa Rica, Urbatsch (2013) showed that districts with a high
proportion of pensioners were more likely to support the agreement, while districts
with more Left party supporters opposed it.
While these studies explicitly address the “so what” question with voting data, there
is an ecological inference problem (King 1997). The theory to be tested is at the individ-
ual level, yet empirical analyses use aggregate data from electoral districts or localities.
Another common issue in using the voting data on general elections (not referendums
on specific trade policies) is a potential omitted variable bias that trade is only one
among multiple issues determining citizens’ votes.
Second, to address these ecological inference problems and omitted variable biases,
an emerging group of studies involves survey respondents’ making optional, costly
behaviors that can be easily monitored by scholars, such as sending e-mails to legisla-
tors regarding a policy issue (Betchel, Hainmueller, and Margalit 2014). Finally, scholars
have devised field experiments in retail stores to estimate consumers’ price premiums
for ethical products compared to nonethical alternatives, such as goods produced with
high labor standards in developing countries (Hiscox and Smyth 2011) and fair trade
coffee (Hainmueller, Hiscox, and Sequeira 2011).

Frontiers (2): Linking Attitudes to Policy Outcomes

The second line of criticisms of the survey-based research concerns how individual
opinions are aggregated into trade-policy outcomes. Indeed, although economists
have formally shown how voting rules in democracies affect trade-policy outcomes
(Baldwin 1976; Mayer 1994), this line of inquiry has been lopsided, with more for-
mal models than empirical tests. When empirical tests are conducted, scholars tend
to bypass the individual-level attitudinal data and estimate the effect of national-level
institutions on trade-policy outcomes (e.g., Rogowski and Kayser 2002; one exception
is Kono 2008).
Aggregation of individual attitudes can occur through group membership, such as
labor union or partisanship. It can also occur through more elite-driven processes that
are not rooted in socioeconomic organizations, such as media framing and political
campaigns. A promising line of future research is examining how citizens form their
attitudes through group membership (and nonmembership), with careful identifica-
tion strategies for observational data (see Ahlquist, Clayton, and Levi 2014) or subgroup
analyses of experimental data (Naoi and Kume 2013a).
Another solution to the missing link between mass attitudes and policy is to study
the concordance between public attitudes and elites’ position taking on trade. Although
district-level studies on how constituents influence legislators’ position taking are
114   Individual Attitudes

abundant, they have used district-level industry profiles and factor endowments as
proxy measures of voters’ interests. As a result, individual-level data are rarely incorpo-
rated into these studies. It would be worthwhile to examine direct legislative responsive-
ness to policy-specific public opinion, such as done by Lax and Phillips (2009) in the
United States.
In sum, an exciting direction for future research is to link individual trade attitudes
with costly behaviors of individuals, collective actions of groups, or policy responsive-
ness of legislators to address “so what” criticisms.

Notes
1. Rodrik (1995) first called for more research on individual preferences over trade policy
to better understand voters’ positions on trade. Milner (2002) similarly advocated for
more studies on trade policy preferences of individual voters to better explain the societal
demand for trade policy.
2. There is an abundance of survey data, conducted by the media and public opinion research
organizations, on citizen’s opinions about trade agreements (general and specific), but
scholarly works using these has been scant. See Scheve and Slaughter (2001a, Chap 2) for
the detailed reviews.
3. They used a separate measure for skill-levels called “occupational scores”, which is an
index of how many subordinates employees respondents have. The higher this score, more
“white-collar” management jobs respondents have.
4. Many existing surveys have given respondents options to choose “don’t know” or neutral
responses, such as “can’t say one way or the other”. Yet, the empirical analyses using these
data often treat them as missing (in the case of “don’t know”) or as a baseline (in the case of
neutral responses), rather than the outcome to be examined.
5. One could argue, though, that low salience of trade is expected from the low tariff rates
and low ratio of export and import in the U.S. economy—the U.S. is thus a deviant case
among OECD countries. Pelc (2013) provides the evidence to the contrary by using
google’s search volume data to examine ordinary people’s “information-seeking” activi-
ties. He demonstrated that U.S. citizens were concerned with WTO dispute-related infor-
mation involving the U.S. case.

References
Ahlquist, John S., Amanda B. Clayton, and Margaret Levi. 2014. Provoking Preferences:
Unionization, Trade Policy, and the ILWU Puzzle. International Organization 68 (1): 33–75.
Ardanaz, Martin, M. Victoria Murillo, and Pablo Pinto. 2013. Sensitivity to Issue Framing on
Trade Policy Preferences: Evidence from a Survey Experiment. International Organization
67 (2): 411–437.
Baker, Andy. 2005. Who Wants to Globalize? Consumer Tastes and Labor Markets in a Theory
of Trade Policy Belief. American Journal of Political Science 49 (4): 924–938.
Baldwin, Robert. 1976. The Political Economy of Postwar U.S. Trade Policy. The Bulletin.
New York University Graduate School of Business Administration.
 Jason Kuo and Megumi Naoi    115

Baldwin, Robert, and Christopher S. Magee. 2000. Is Trade Policy for Sale? Congressional
Voting on Recent Trade Bills. Public Choice, 105 (1-2): 79–101.
Bauer, Raymond, Ithiel de Sola Poole, and Lewis Dexter. 1963. American Business and Public
Policy. New York: Atherton Press.
Beaulieu, Eugene, and Michael Napier. 2008. Why Are Women More Protectionist Than Men?
Working Paper. Department of Economics, The University of Calgary.
Betchel, Michael, Jens Hainmueller, and Yotam Margalit. 2014. “Preferences for International
Redistribution: The Divide Over the Eurozone Bailouts. American Journal of Political Science
58 (4): 835-856.
Burgoon, Brian, and Michael J. Hiscox. 2006. The Mysterious Case of Female Protectionism:
Gender Bias in Attitudes Toward International Trade. Working Paper. Harvard University
Weatherhead Center for International Affairs.
Chiang, Chun-Fang, Jin-Tan Liu, and Tsai-Wei Wen. 2013. Individual Preferences for Trade
Partners in Taiwan. Economists and Politics 25 (1): 91–109.
Coughlin, Cletus C. 2002. The Controversy Over Free Trade: The Gap between Economists
and the General Public. Review, Federal Reserve Bank of St. Louis 84 (1): 1–22.
Cusack, Thomas, Torben Iversen, and Philip Rehm. 2006. Risks at Work: The Demand and
Supply Sides of Government Redistribution. Oxford Review of Economic Policy 22 (3):
365–389.
Daniels, Joseph P., and Marc von der Ruhr. 2005. Gold and the Global Economy: Religion and
Attitudes toward Trade and Immigration in the United States. Socio-Economic Review 3 (3):
467–487.
Ehrlich, Sean. 2010a. The Fair Trade Challenge to Embedded Liberalism. International Studies
Quarterly 54 (4): 1013–1033.
Ehrlich, Sean. 2010b. Who Supports Compensation? Individual Preferences for Trade-Related
Unemployment Insurance. Business and Politics 12 (1): Article 3.
Ehrlich, Sean, and Eddie Hearn. 2013. Does Compensating Those Harmed Increase Support
for Trade? An Experimental Test of the Embedded Liberalism Thesis. Foreign Policy Analysis
9 (1): 1–16.
Ehrlich, Sean D., and Cherie Maestas. 2010. Risk Orientation, Risk Exposure, and Policy
Opinions: The Case of Free Trade.” Political Psychology 31 (5): 657–684.
Estevez-Abe, Margarita. 2006. Gendering the Varieties of Capitalism: A Study of Occupational
Segregation by Sex in Advanced Industrial Societies. World Politics 59 (1): 142–175.
Fordham, Benjamin, and Katja Kleinberg. 2012. How Can Economic Interests Support for Free
Trade? International Organization 66 (2): 311–328.
Gawande, Kishore, Pravin Krishna, and Marcelo Olarreaga. 2009. What Governments
Maximize and Why: The View from Trade. International Organization 63 (3): 491–532.
Gilligan, Michael. 1997. Empowering Exporters: Reciprocity, Delegation, and Collective Action in
American Trade Policy. Ann Arbor: University of Michigan Press.
Goldstein, Judith, Yotam Margalit, and Dougalas Rivers. 2008. Producer, Consumer, Family
Member: The Relationship between Trade Attitudes and Family Status. Paper presented at
the Conference on Domestic Preferences and Foreign Economic Policy, Working Paper,
Stanford University.
Gowa, Joanne. 1994. Allies, Adversaries, and International Trade. Princeton, NJ:  Princeton
University Press.
Gowa, Joanne, and Edward D. Mansfield. 1993. Power Politics and International Trade.
American Political Science Review 87 (2): 408–420.
116   Individual Attitudes

Grossman, Gene M., and Elhanan Helpman. 1994. Protection for Sale. American Economic
Review 84 (4): 833–850.
Guisinger, Alexandra. 2009. Determining Trade Policy: Do Voters Hold Politicians
Accountable? International Organization 63 (3): 533–557.
Hainmueller, Jens, and Michael J. Hiscox. 2006. Learning to Love Globalization: Education and
Individual Attitudes toward International Trade. International Organization 60 (2): 469–498.
Hainmuller, Jens, Michael J. Hiscox, and Sandra Sequeira. 2011. Consumer Demand for the
Fair Trade Label: Evidence from a Field Experiment. Review of Economics and Statistics.
(Forthcoming 2015).
Hays, Jude C., Sean Ehrlich, and Clint Peinhardt. 2005. Government Spending and Public
Support for Trade in the OECD: An Empirical Test of the Embedded Liberalism Thesis.
International Organization 59 (2): 473–494.
Helpman, Elhanan. 2011. Understanding Global Trade. Cambridge, MA:  Harvard
University Press.
Hicks, Raymond, Helen Milner, and Dustin Tingley. 2014. Trade Policy, Economic Interests
and Party Politics in a Developing Country:  The Political Economy of CAFTA-DR.
International Studies Quarterly 58 (1): 106–117.
Hiscox, Michael J. 2002. International Trade and Political Conflict: Commerce, Coalitions, and
Mobility. Princeton, NJ: Princeton University Press.
Hiscox, Michael J. 2006. Through a Glass and Darkly: Attitudes Toward International Trade
and the Curious Effects of Issue Framing. International Organization 60 (4): 755–780.
Hiscox, Michael J., and Nicholas Smyth. 2011. Is There Consumer Demand for Fair Labor
Standards? Evidence from a Field Experiment. SSRN. http://papers.ssrn.com/sol3/papers.
cfm?abstract_id=1820642 (accessed November 15, 2013).
Imai, Kosuke, and Dustin Tingley. 2012. A Statistical Model for Empirical Testing of Competing
Theories. American Journal of Political Science 56 (1): 218–236.
Irwin, Douglas. 1994. The Political Economy of Free Trade:  Voting in the British General
Election of 1906. Journal of Law and Economics 37 (1): 75–108.
Iversen, Torben, and David Soskice. 2001. An Asset Theory of Social Policy Preferences.
American Political Science Review 95 (4): 875–893.
Kaltenthaler, Karl, Ronald D. Gelleny, and Stephen J. Ceccoli. 2004. Explaining Citizen Support
for Trade Liberalization. International Studies Quarterly 48 (4): 829–851.
King, Gary. 1997. A Solution to Ecological Inference Problem: Reconstructing Individual
Behaviors from Aggregate Data. Princeton, NJ: Princeton University Press.
Kim, Mi-kyung. 2011. Foreign Economic Policy and Social Conflicts in South Korea:
Preferences, Policy Paradigms, and the Formation of Political Coalitions. Korean Political
Science Review 45 (5): 147–173.
Koremenos, Barbara, Charles Lipson, and Duncan Snidal. 2001. The Rational Design of
International Institutions. International Organization 55 (4): 761–799.
Kono, Daniel Y. 2008. Does Public Opinion Affect Trade Policy? Business and Politics 10
(2): Article 2.
Kume, Ikuo, and Masaru Kohno. 2011. Political Conflicts over TPP Reflect the Public Divide
over Dipolomacy. Nikkei Shimbun, Keizai Kyositsu Series, December 22.
Kuno, Arata, and Megumi Naoi. 2013. Framing Business Interests: How Campaigns Affect Firms’
Positions on Preferential Trade Agreements. Working Paper, Kyorin University and UCSD.
Lax, Jeffery R., and Justin H. Phillips. 2009. Gay Rights in the States: Public Opinion and Policy
Responsiveness. American Political Science Review 103 (3): 367–386.
 Jason Kuo and Megumi Naoi    117

Lohmann, Susanne, and Sharyn O’Halloran. 1994. Divided Government and U.S. Trade
Policy: Theory and Evidence. International Organization 48 (4): 595–632.
Lu, Xiaobo, Kenneth Scheve, and Matthew J. Slaughter. 2012. Inequity Aversion and the
International Distribution of Trade Protection. American Journal of Political Science
56(3): 638-654.
Magee, Stephen P., William A. Block, and Leslie Young. 1989. Black Hole Tariffs and Endogenous
Policy Theory. Cambridge, UK: Cambridge University Press.
Mansfield, Edward D., Helen V. Milner, and B. Peter Rosendorff. 2000. Free to
Trade:  Democracies, Autocracies, and International Trade. American Political Science
Review 94 (2): 305–321.
Mansfield, Edward, and Diana Mutz. 2009. Support for Free Trade: Self-Interest, Sociotropic
Politics, and Out-group Anxiety. International Organization 63 (3): 425–457.
Margalit, Yotam. 2011. Costly Jobs:  Trade-related Layoffs, Government Compensation, and
Voting in U.S. Elections. American Political Science Review 105 (1): 169–188.
Margalit, Yotam. 2012. Lost in Globalization:  International Economic Integration and the
Sources of Popular Discontent. International Studies Quarterly 56 (3): 484–500.
Mayda, Anna Maria, and Dani Rodrik. 2005. Why Are Some People (and Countries) More
Protectionist Than Others. European Economic Review 49 (6): 1393–1430.
Mayda, Anna Maria, Kevin H. O’Rourke, and Richard Sinnott. 2007. Risk, Government and
Globalization: International Survey Evidence. Working Paper No. 13037. National Bureau of
Economic Research.
Mayer, Wolfgang. 1994. Endogenous Tariff Formation. American Economic Review 74: 970–985.
Melitz, Marc J. 2003. The Impact of Trade on Intra-Industry Reallocations and Aggregate
Industry Productivity. Econmetrica 71 (6): 1695–1725.
Milner, Helen, and Dustin Tingley. 2011. Who Supports Global Economic Engagement? The
Sources of Preferences in American Foreign Economic Policy. International Organization 65
(1): 37–68.
Milner, Helen V. 2002. International Trade. In Handbook of International Relations, edited by
Walter Carlsnaes, Thomas Risse, and Beth Simmons, 448–461. New York: Sage.
Milner, Helen V. 1989. Resisting Protectionism: Global Industries and the Politics of International
Trade. Princeton, NJ: Princeton University Press.

Naoi, Megumi, and Ikuo Kume. 2013a. Coalition of Losers: Why Agricultural Protectionism
Has Survived During the Great Recession. In Politics in the New Hard Times, edited by Miles
Kahler and David Lake, 190–211. Ithaca, NY: Cornell University Press.
Naoi, Megumi, and Ikuo Kume. 2011. Explaining Mass Support for Agricultural Protectionism:
Evidence from a Survey Experiment during the Global Recession. International Organi­
zation 65 (4): 771–795.
Naoi, Megumi, and Ikuo Kume. Forthcoming. Workers or Consumers? A Survey
Experiment on the Duality of Citizens’ Interests in the Politics of Trade. Comparative
Political Studies.
Naoi, Megumi, and Shujiro Urata. 2013. Free Trade Agreements and Domestic Politics: The
Case of TPP. Asian Economic Policy Review 2 (2): 326–349.
Olson, Mancur. 1965. The Logic of Collective Action. Cambridge, MA: Harvard University Press.
O’Rourke, Kevin H., and Richard Sinnott. 2001. The Determinants of Individual Trade-Policy
Preferences: International Survey Evidence. Brookings Trade Forum: 157–196.
118   Individual Attitudes

Pelc, Krzysztof J. 2013. Googling the WTO:  What Search-Engine Data Tell US About the
Political Economy of Institutions. International Organization 67 (3): 629–655.
Rho, Sungmin, and Michael Tomz. 2013. Industry, Self-Interest, and Individual Preferences
over Trade Policy. Working Paper prepared for the 2013 International Political Economy
Colloquium, Goa, India.
Rodrik, Dani. 1995. Political Economy of Trade Policy. In Handbook of International
Economics, edited by Gene M. Grossman and Kenneth Rogoff, 3: 1457–1494. Amsterdam,
Netherlands: Elsevier Science Press.
Rogowski, Ronald. 1987. Trade and the Variety of Democratic Institutions. International
Organization 41 (2): 203–223.
Rogowski, Ronald, and Mark Kayser. 2002. Majoritarian Electoral Systems and Consumer
Power: Price-level Evidence from the OECD Countries. American Journal of Political Science
46 (3): 526–539.
Ruggie, John G. 1982. International Regimes, Transactions, and Change: Embedded Liberalism
in the Postwar Economic System. International Organization 36 (2): 379–415.
Schattschneider, Elmer E. 1935. Politics, Pressures, and the Tariff:  A  Study of Free Private
Enterprise in Pressure Politics, as Shown in the 1929–1930 Revision of the Tariff. Prentice-Hall.
Scheve, Kenneth F., and Matthew J. Slaughter. 2001a. Globalization and the Perceptions of
American Workers. Washington, DC: Institute for International Economics.
Scheve, Kenneth F., and Matthew J. Slaughter. 2001b. What Determines Individual Trade
Policy Preferences. Journal of International Economics 54 (3): 267–292.
Urbatsch, Robert. 2013. A Referendum on Trade Policy: Voting on Free Trade in Costa Rica.
International Organization 67 (1): 197–214.
Walter, Stefanie. 2010. Globalization and the Welfare State: Testing the Micro-foundation of
the Compensation Hypothesis. International Studies Quarterly 54 (2): 403–426.
 Chapter 7

L ab or and Prot e c t i oni st

Sentim e nt

E ri ca Owe n

The distributional consequences of trade in labor markets are one of the primary chan-
nels through which globalization affects the economic well-being of the public. This
chapter focuses on the demand for protection by labor interests, which may correspond
to producer demands more generally or put the interests of labor and capital at odds.
Two questions are of particular importance: (1) What are the underlying cleavages or
latent shared interests of labor? (2) When are protectionist interests likely to be reflected
in policy outcomes? Failure by governments to include labor interests in their policies
today will likely set the stage for unrest and protectionist backlash in the future.
Individual-level studies find mixed evidence regarding the role of economic theory in
shaping preferences about trade. However, Fordham and Kleinberg (2012) suggest that
it would be premature to dismiss self-interest as a determinant of trade preferences, and
argue that groups (organized or not) play an important role in shaping attitudes toward
trade. They suggest that people need not know clearly how trade will affect them for eco-
nomic factors to shape trade preferences. Studies of mass public opinion show evidence
of systematic preferences, eliminating noise in individual survey responses and reiterat-
ing the importance of latent shared interests.
Trade protection is the subject of large literatures in both political science and eco-
nomics. This chapter reviews primarily the literature in political science, while draw-
ing attention to work in economics that can shed light on lingering questions in
political science. In particular, our thinking about protectionist demands and ways to
satisfy these interests must evolve to keep pace with the changing nature of globaliza-
tion. Recent advances in economic trade theory, in particular the heterogeneous firms
theory, have important implications for the politics of protection (see Plouffe, this vol-
ume). In the first section I briefly review the various cleavages among labor predicted
by trade theories prominent in political science. In the second section I discuss how
the influence of labor interests on policy has been examined empirically. The third sec-
tion covers policy alternatives to tariffs and nontariff barriers that may be used to satisfy
120    Labor and Protectionist Sentiment

protectionist demands, specifically temporary trade barriers and compensation. In the

fourth section, I examine the implications of heterogeneous firms theory for protection-
ist demands. The fifth section presents directions for future research, followed by a sec-
tion of conclusions.

Distributional Consequences
and Demands for Protection

This section only briefly reviews the canonical models of trade theory that shape prefer-
ences regarding trade, as they have been extensively reviewed elsewhere (e.g., Rogowski
2006), before discussing other sources of protectionist sentiment. Our understanding
of how trade affects domestic factors of production, and subsequent preferences about
trade policy, are primarily drawn from two underlying trade models.
The first, and arguably most widely used, is the factor endowments model. The
Heckscher-Ohlin theory suggests that countries will have a comparative advantage in
those factors that they have in abundance and a comparative disadvantage in those that
are relatively scarce. The Stolper-Samuelson (HOS) theorem suggests that trade will
increase returns to abundant factors and decrease returns to scarce factors through spe-
cialization and exchange (Stolper and Samuelson 1941). This insight is the foundation
of a large body of work in political science because of the implied cleavages that will
form. In his seminal work Commerce and Coalitions, Rogowski (1989) argues that pro-
tection (liberalization) will benefit owners of scarce (abundant) factors; thus cleavages
will divide interests with respect to trade along factor lines. Owners of a given factor
of production will have shared interests. This leads to the well-known hypothesis that
labor will be protectionist in developed countries, because unskilled labor constitutes
a majority of the workforce yet is relatively scarce; conversely, labor will be pro-trade in
developing countries. In one test of this hypothesis, Baker (2005) calculates a country-
level measure of support for free trade and finds that countries with greater skill abun-
dance have lower aggregate support for free trade.
The Ricardo-Viner, also known as specific factors, model is the second primary the-
ory of the distributional consequences of trade. Industries of comparative advantage use
the abundant factor relatively intensively, while industries of comparative disadvantage
use the scarce factor intensively. Unlike in the HOS model, factors are not fully mobile
domestically. This means that owners of the same factor may have heterogeneous pref-
erences; thus coalitions tend to form around industry rather than along class lines.
The key difference between the factor endowments and specific factors models is the
level of domestic factor mobility. Hiscox (2002) suggests that cleavages will form along
industry or factor lines depending on the level of inter-industry factor mobility. When
asset mobility is high, cleavages are likely to form along factor lines; when mobility is low,
industry cleavages form, because it is more difficult to move from one industry to another.
 Erica Owen   121

The assumption about the degree of factor mobility has implications for the types of coali-
tions that form, as well as the ease of collective action (Alt and Gilligan 1994). Conventional
wisdom suggests that collective action is likely to be more successful when factors orga-
nize along industry lines, due to the lower costs of organizing a smaller number of actors.
Along with factor mobility, designation of the factors of production shapes the nature
of protectionist demands. Traditionally, labor is treated as a homogenous factor of
production. However, Chase suggests that scholars have emphasized the “antagonism
between labor and capital in the United States” at the expense of examining cleavages
within labor (2008, 658). The distinction between skilled and unskilled labor is par-
ticularly important for the distributional consequences of trade in developed countries
because trade, in combination with other aspects of economic integration and techno-
logical change, has affected skilled and unskilled labor differently.
Empirical inconsistencies in the neoclassical (HOS and specific factors) models
led to the development of new trade theory. In short, trade on the basis of compara-
tive advantage predicts inter-industry trade between countries with dissimilar factor
endowments. However, much international trade is between countries with similar fac-
tor endowments and occurs within the same industry. New trade theory accounts for
this by introducing imperfect competition and differentiated goods. Standard distribu-
tional pressures are mitigated in this case because goods are differentiated and thus are
not perfect substitutes. Therefore, intra-industry trade between countries with similar
endowments should spur fewer demands for protection (Krugman 1981).
Macroeconomic factors like unemployment and the real exchange rate are also
important sources of protectionist sentiment. Mansfield and Busch (1995) review the
intuition guiding this theory. Recessions, and high unemployment in particular, are
likely to lead to more demands for protection (Corden 1993). Workers who lose their
jobs are likely to have a harder time finding new employment and may be forced to
accept a lower wage. Cassing, McKeown, and Ochs (1986) suggest that politically effec-
tive demands for protection by import-competing sectors will be higher during peri-
ods of unemployment. Similarly, rising current account imbalances, or an appreciated
exchange rate more generally, may lead to increasing protectionist sentiment, due in
part to competitive pressures and perhaps perceptions of unfair practices. In the United
States, anti-Japanese sentiment in the 1980 and anti-Chinese sentiment in the 2000 are
excellent examples of this. Such conditions are likely to give rise to economic national-
ism. Indeed, Baker (2005) finds that as aggregate nationalist sentiment increases, overall
mean support for free trade decreases.

Influence of Labor on Trade Policy

Rogowski (1989) and Hiscox (2002) are exemplary accounts of the coalitions that form
around trade, but given a set of preferences, the next step is to examine how and under
what conditions labor interests influence trade policy. Therefore, we must examine
122    Labor and Protectionist Sentiment

how the preferences generated by distributional consequences are translated into effec-
tive political demands and the incentives of politicians to place weight on competing
interests. This literature largely follows the open economy politics (OEP) framework.1
Theories of preference aggregation can generally be divided into two camps: those that
emphasize the incentives created by institutions for politicians to cater to different con-
stituencies and those that emphasize the collective action problem and political activity
by organized interest groups.
Because it is difficult to observe lobbying behavior by groups with latent shared
interests, we frequently rely on the incentives of politicians to cater to the interests of
one group over another. Certain domestic institutional arrangements are thought to
privilege narrow or broad coalitions and often correspond to factoral or sectoral cleav-
ages. Institutional accounts thus emphasize the incentives of politicians to supply pro-
tection, given a set of preferences held by labor, rather than political activity by labor
interests demanding protection. For example, democracies tend to place more weight
on the preferences of citizens, who are primarily owners of labor (rather than capital).2
Therefore, democracies are more likely to implement policies that benefit labor than are
nondemocracies (e.g., Milner and Kubota 2005), although the effect of democracy on
openness will depend on relative factor endowments (e.g., Tavares 2008; Milner and
Mukherjee n.d.).
Government partisanship also affects the political influence of labor. Like regime
type, arguments about partisanship tend to assume HOS cleavages. Parties have broad
constituencies, and the Left, as a representative of labor, is more likely to promote the
interests of labor with respect to trade, whereas the Right is more likely to represent
the interests of capital. Thus we should see more protection under Left governments in
developed countries and more openness under Left governments in developing coun-
tries. Indeed, in a study of twenty-five developed countries, Milner and Judkins (2004)
find that Right governments are more likely to take pro-trade positions than Left gov-
ernments, while Dutt and Mitra (2005) find that the reverse is true in countries with
abundant labor. However, as labor interests become more heterogeneous, the constitu-
ency of Left parties is not obvious. For example, in a study of welfare spending, Rueda
(2007) finds that the Left represents the interests of labor market insiders at the expense
of outsiders. Similar pressures are likely to occur with trade as well.
Industry cleavages typically appear in the endogenous protection literature,
which emphasizes the lobbying process and the ease of collective action. The
protection-for-sale (PFS) model is the most prominent theory of endogenous trade
protection. Grossman and Helpman’s (1994) seminal PFS model emphasizes that in
the policy-making process, governments face a trade-off between satisfying special
interests that provide political contributions and increasing aggregate social welfare to
appeal to voters. Traditionally, PFS models do not attempt to model collective action
and assume protectionist interests are better able to overcome the collective action
problem than are the winners from trade, which includes diffuse groups like consum-
ers, and also exporters (as an exception, see Milner 1988). Busch and Reinhardt (2000)
introduce geographic concentration as one factor that captures the ease of collective
 Erica Owen   123

action and likelihood of mobilization, while Gawande and Magee (2012) introduce free
riding into the model to address the empirical finding that governments appear to care
almost solely about aggregate welfare. Labor does not often play an explicit role in PFS
models, which largely lump protectionist interests together and in which societal wel-
fare is based on consumer interests.
The norm has been to posit either factor—or industry—cleavages. However, in addi-
tion to Hiscox’s work documenting changes in the type of cleavages, other scholars have
sought to explicitly measure the level of factor mobility (especially of labor) and incor-
porate it into theories of protection. The level of factor mobility is an important deter-
minant of the adjustment costs faced by labor and thus of demands for protection and,
as discussed below, compensation. Mukherjee, Li, and Smith (2009) incorporate fac-
tor mobility, measured as a continuous variable, into a test of the PFS model, thereby
allowing for either HOS or Ricardo-Viner cleavages to emerge. In a panel of majoritar-
ian countries, they find that governments raise tariffs when factor mobility decreases,
consistent with the notion that collective action becomes easier as the size of coalitions
decreases.
As an alternative to trade on the basis of endowment-driven comparative advantage,
new trade theory, which includes imperfect competition and intra-industry trade, has
also been used to explain the politics of trade protection, though labor does not play
a prominent role in these accounts. Milner (1988) suggests that internationally ori-
ented firms tend to be less protectionist than domestically oriented ones. She further
argues that these pro-trade preferences can act as a counterweight to protectionist
ones, thus explaining why trade barriers do not increase significantly during periods
of high import competition. Similarly, Hathaway (1998) suggests that openness damp-
ens protectionist sentiment in industries hurt by imports, because larger firms are able
to increase capital intensity or move labor-intensive activities abroad, while others are
more likely to exit the industry. See also Chase (2003).
Protectionist sentiment is also associated with larger macroeconomic trends.3 During
bad economic times, governments often must respond to increased demands for protec-
tion. Seminal work by Goldstein (1986) and Milner (1988) suggests that the level of pro-
tection in the United States was lower than expected during the economic decline of the
1970s. Milner suggests that this was so because anti-protectionist interests limit the sup-
ply of protection, while Goldstein suggests that institutions and ideas biased toward free
trade persist in the United States. On the other hand, Takacs (1981) finds that bad eco-
nomic conditions increased petitions for safeguards in the United States between 1949
and 1979. In an analysis of US industry-level tariffs during the 1980s, Bradford (2006)
finds that an industry’s level of trade protection increases with the number of unem-
ployed workers and decreases with the job turnover rate.
Protectionist pressures in the face of bad macroeconomic outcomes are one com-
ponent of a society’s pluralist interests, and thus institutions play an important role in
shaping when those interests are reflected in policy. Mansfield and Busch (1995) exam-
ine the effect of the real exchange rate and unemployment on nontariff barriers (NTBs),
conditional on variation in democratic institutions. Similarly, Henisz and Mansfield
124    Labor and Protectionist Sentiment

(2006) examine the effects of unemployment on trade openness, conditional on the

number of institutional constraints. They find that trade openness decreases as unem-
ployment rises in democracies, though this effect is attenuated as the number of veto
players increases. Together this research suggests that democratic governments are
responsive to domestic protectionist pressures, though the degree of responsiveness
varies in systematic ways.
Political activity and lobbying by labor is the subject of a much smaller literature.
Due to the common assumption that protectionists are more likely to overcome col-
lective action, labor interests are typically incorporated into the political economy of
trade through institutional mechanisms or are assumed to align with producer inter-
ests. However, organized labor in particular shapes politicians’ incentives to weigh the
interests of labor interests more heavily. Matschke and Sherlund (2006) propose a labor-
augmented Grossman-Helpman model that allows for lobbying by trade unions. They
find that active lobbying and other labor market variables are important determinants
of trade policy in a study of NTBs in US industries. Bradford (2006) also finds a posi-
tive effect of the level of unionization on the level of protection. Similarly, scholars have
found that campaign contributions by labor groups affect legislative votes. For example,
Steagall and Jennings (1996) demonstrate that House representatives who depended on
labor political action committee (PAC) contributions were more likely to vote against
the North American Free Trade Agreement (NAFTA). Similarly, Baldwin and Magee
(2000) find that campaign contributions by labor unions increase the likelihood that US
members of Congress will vote against trade liberalization bills.
The preceding discussion emphasizes protectionist demands by labor in developed
countries because neoclassical models of trade suggest that capital (and skilled labor)
is the source of protectionist sentiment in labor-abundant developing countries. In the
section “Heterogeneous Firms and Labor Market Outcomes,” I discuss new research in
economics, which implies that skilled labor in developing countries also benefits from
globalization relatively more than does unskilled labor.

State of the Literature: Protection

and Redistribution

New research in international political economy (IPE) builds on earlier work by intro-
ducing additional distributional consequences and extending the analysis of protec-
tion to include other tools politicians can use to satisfy protectionist sentiments (aside
from tariffs and NTBs). The development of global supply chains and the trade in
intermediate goods and services has changed the shape of international trade. Other
facets of openness, including the activity of multinational firms and immigration, may
also fuel protectionist sentiments (e.g., Rogowski 2006). Politicians must find a way to
respond to this new protectionism and do so in a world trading system that is much
more institutionalized. “The WTO mediates the relationship between the domestic
 Erica Owen   125

political processes that generate government policy preferences and the tariff rates
that one observes” (Oatley 2011, 320). Because politicians are more constrained in their
ability to use traditional protection measures (even if they should want to for domestic
political reasons), we must consider alternative trade barriers like safeguards, as well as
other tools like compensation.

Alternative Forms of Protection

Governments face a menu of options when responding to protectionist demands.
For example, Naoi (2009) suggests that when governments choose to provide protec-
tion, they can choose unilateral instruments like subsidies, bilateral tools like volun-
tary export restraints, or measures compliant with the General Agreement on Tariffs
and Trade (GATT)/World Trade Organization (WTO), such as escape clauses and
antidumping. This is a fruitful area of further research, especially empirical research.
Increasing legalization has limited the ability of governments to use tariffs and NTBs.
However, flexibility built into international institutions, including countervailing duties
(CVD) and antidumping (AD) and other escape clauses, provides governments with the
means of responding to domestic protectionist pressures.4
Though AD and CVD are not new inventions, they are increasingly important as pol-
icy tools used to supply protection.5 The United States established variants of these laws
in the 1930s to deal with unfair trade practices or cheating by trade partners (Goldstein
1986). Though the largest users of AD laws traditionally were the United States, the
European Union, Canada, and Australia, developing countries have adopted and are
using AD laws at a higher rate in the WTO period (Bown 2011). These administered pro-
tection measures are sanctioned by the WTO, and of these, AD tools are the most widely
used. Although the Great Recession of the late 2000s did not lead to large tariff increases,
it did lead to an increase in the use of these temporary trade barriers (Bown 2011).
Hansen (1990) finds that political factors determine the supply of protection by the
International Trade Commission (ITC) in the United States, while Drope and Hansen
(2004) find that PAC spending in favor of protection has a positive effect on the likeli-
hood of a favorable decision by the ITC. Temporary trade barriers can also be used to
smooth over the costs of adjustment for workers in industries that experience shocks to
terms of trade, an important component of the welfare consequences of trade (Davidson
and Matusz 2004).
Finally, labor and environmental standards in free trade agreements (FTAs) may level
the playing field and constitute another form of protection from the adjustment costs
of trade. However, these standards are often criticized as being unenforceable. After
NAFTA came into effect, US unions felt that labor standards were not enforced; thus
there has been increasing opposition to more recent FTAs. On the other hand, Ehrlich
(2010) argues that fair trade differs from general protectionism, suggesting that such side
agreements are not protectionism in disguise. His findings suggest that labor and envi-
ronmental side agreements, when enforced, may be a way to mitigate some protectionist
demands that cannot be met by increasing barriers to trade or providing compensation.
126    Labor and Protectionist Sentiment

Compensation in Lieu of Protection

Politicians can use compensation, as an alternative to trade protection, to address labor
demands for protection, while simultaneously maintaining the current level of open-
ness. Although labor as an actor in its own right does not play a central role in much
of the IPE literature, it is central in studies of welfare outcomes. Compensation is a key
component of the embedded liberalism hypothesis, which states that governments can
“compensate” citizens for rising insecurity associated with international economic inte-
gration, in exchange for continued support for openness (Ruggie 1982). Boix (2006)
highlights trade policy and redistribution as alternative policy choices that can be used
to achieve a given domestic political goal.6 Thus, protection and compensation can both
be used to satisfy the demands of protectionists.
Compensation generates or maintains support for openness and may take the form
of job retraining programs, unemployment benefits, or other general welfare programs.
Survey data from the United States and elsewhere suggest that individuals have reserva-
tions about the effects of trade on jobs and job security and are more likely to support
free trade when it is accompanied by policies to help those workers hurt by trade. Walter
(2010) finds that globalization generates increased feelings of job insecurity, which in
turn leads to increased preferences for welfare compensation among those hurt by glo-
balization. Similarly, Rehm (2011) argues that preferences about redistribution are a
function of income and income prospects. In a study of aggregate public opinion, Owen
and Quinn (forthcoming) demonstrate that rising imports increase demands for larger
government, and that the effect of aggregate public opinion on spending is conditional
on the level of imports. In summary, compensation is likely to be demanded by those
who experience or anticipate an increase in job insecurity or decline in income.
Whether the compensation mechanism successfully reduces anti-trade sentiment is
the subject of a growing literature. Hays, Ehlrich, and Peinhardt (2005) find that com-
pensation is an effective means to garner support for free trade. At the individual level,
they find that higher levels of active labor market spending lead to higher levels of sup-
port for free trade. See also Hays (2009). In a different approach, Margalit (2011) exam-
ines the effect of trade adjustment assistance (TAA) on President George W. Bush’s vote
share in the 2004 election and finds that higher levels of TAA decreased the loss in vote
share associated with negative trade outcomes. Thus constraints on the ability of gov-
ernments to provide compensation increase the risk of protectionist backlash.7
This raises the question of when governments will choose different tools to appease
protectionist interests. Hwang and Lee (2013) suggest that the form of redistribution
depends on factor mobility and find that welfare politics are more likely when mobil-
ity is high, while targeted subsidies are more likely when mobility is low, conditional
on government partisanship. Fewer studies have explored choices about protection and
compensation simultaneously. In a theoretical paper, Dixit and Londregan suggest that
protection is more politically manageable, though less efficient economically, than com-
pensation, which encourages resource reallocation to more productive sectors. Ehlrich
 Erica Owen   127

(2010) suggests that true fair trade attitudes may limit the effectiveness of compensation
as a means of addressing protectionist concerns, and that governments may need to use
multiple policy tools to address anti-trade sentiments, including through the design of
international agreements.

Heterogeneous Firms and

Labor Market Outcomes

Recent advances in economics, namely heterogeneous firms theory, introduce new dis-
tributional consequences for labor that are likely to increase protectionist sentiment,
even in developing countries. In new new trade theory (NNTT), firms are heteroge-
neous in terms of size, export status, and productivity. In a seminal article, Melitz (2003)
argues that firms have different levels of productivity; below a certain cutoff, firms are
no longer active in the market. There is a second cutoff, and only firms with productiv-
ity levels above this threshold are able to engage in export behavior. Exporting firms
expand through trade, while import-competing firms are likely to fail or change indus-
tries, and their plants are more likely to close and have lower employment and wage
growth (Bernard, Jensen, and Schott 2006). Exceptions to this are firms that have reallo-
cated production abroad and are engaged in importing back into their home countries.
Trade therefore changes the composition of firms in the economy, increasing the pro-
ductivity level of an industry. This has important labor market implications, because
firms with different productivity levels have systematically different demand schedules
for labor, including demand for skilled relative to unskilled labor, wages paid, and the
elasticity of demand. As suggested by Scheve and Slaughter, the distributional conse-
quences (labor market outcomes) of globalization include “the level of real and relative
wages, and insecurity in terms of the volatility of wages and employment” (2006, 225).
The consideration of foreign direct investment (FDI) and trade together is crucial
and flows naturally from NNTT.8 Protectionism leads not only to demands for com-
pensation or trade protection, but also, for example, to new restrictions on cross-border
mergers and acquisitions. Heterogeneous firms theory offers new, in some cases con-
flicting, distributional consequences relative to those generated by neoclassical theories,
and which have important implications for political economy of protection.

Trade and Unemployment

Traditional trade theory assumes full employment; as a result, standard political
economy stories tend to emphasize the wage effects of trade and do not account for
the possibility of equilibrium unemployment, even in the presence of rising corpo-
rate profits. Economists have proposed several mechanisms, including fair wages and
128    Labor and Protectionist Sentiment

labor market frictions, to explain how trade increases unemployment. For example,
Egger and Kreickemeier introduce firm-specific wages where “the wage considered to
be fair depends inter alia on the productivity level (and thus the economic success) of
the employer” (2009, 188). In this model, wages for identical workers are higher at more
productive firms. Owners of highly productive firms are unwilling to reduce wages to
clear the labor market, as is the case in traditional trade theory, leading to unemploy-
ment in equilibrium. Changes in overall productivity also affect the level of employ-
ment. Although output increases from trade and raises demand for labor, increasing
overall industry productivity decreases labor demand, which may make some of the
workforce redundant. Alternatively, Helpman and Itskhoki (2009) suggest that greater
labor market frictions lead to higher unemployment, which can account for the differ-
ent labor market outcomes in the United States (higher inequality) and Western Europe
(persistent unemployment).
Moreover, those who lose their jobs because of import competition are unlikely to be
hired by exporting firms, because these workers are likely to have lower skill levels than
those demanded by exporters. Trade changes the composition of firms in the economy,
thus inducing job creation, destruction, and dislocation. Workers displaced by import
competition are likely to face reduced wages and perhaps jobs in the service sector
(Kletzer 2001).
The number of jobs is also affected by offshoring and foreign direct investment. In a
model proposed by Mitra and Ranjan (2010), offshoring can decrease unemployment if
there is perfect factor mobility, but in the presence of labor market frictions, offshoring
increases unemployment in the offshoring sector and decreases it in the non-offshoring
sector. Empirical testing of these implications of NNTT is in the early stages. However,
this finding that trade increases unemployment calls into question the conclusion that
trade increases aggregate welfare.

Trade and Inequality

The introduction of heterogeneous firms also provides two insights into the relation-
ship between trade and inequality.9 First, the channels through which trade leads to
higher inequality are more numerous than suggested by traditional trade theory. In par-
ticular, trade increases inequality even within industries of comparative advantage (e.g.,
Lawrence and Slaughter 1993). Moreover, this effect is not limited to inter-industry trade
as predicted by the Stolper-Samuelson theorem. Intra-industry trade between countries
with similar factor endowments (i.e., two developed countries) also increases inequality,
because trade raises the skill premium as more productive firms increase their market
share and less productive firms exit. While traditional trade theory predicts inequality
will rise in labor scarce countries, and the empirical evidence bears this out, the findings
here suggest that it may increase more than expected.
Second, NNTT predicts that trade will increase inequality in developing countries, in
stark contrast to Stolper-Samuelson. Exporters in developing countries are again more
 Erica Owen   129

skill intensive, and as they increase market share through trade, the skill premium, and
by extension inequality, increase as well. Even without reference to the skill premium,
Egger and Kreickemeier (2009) demonstrate that trade increases inequality among
identical workers as more productive firms, which pay higher wages, increase their
employment share. Intuitively, the positive effect of trade on inequality only strengthens
in the presence of unemployment.
These effects are magnified by the offshoring of production activities and services,
which contributes to rising inequality in both developed and developing countries.
Feenstra and Hanson (1996) propose a model of offshoring in which firms can reallocate
tasks to workers in another country based on skill intensity of the task and factor endow-
ments, increasing inequality in both home and host. In political science, Chase (2008)
emphasizes the importance of outsourcing as a source of cleavages between skilled and
unskilled workers. He finds that inequality is correlated with offshoring, and that collec-
tive action in favor of countervailing duty petitions splits along skill level in the motion
picture industry. The heterogeneous firms model of offshoring predicts a similar effect
on inequality (Harrison et al. 2011).

Directions for the Future

This section addresses three directions for future research. First, I argue that offshoring
in services introduces new cleavages over trade and has the potential to increase the
salience of trade in developed countries. Second, I discuss the implications of global
production for the political influence of labor. Finally, I discuss the implications of het-
erogeneous firms theory for the politics of trade in developing countries.

Services Offshoring and New Trade Cleavages

Offshoring is the sourcing of inputs (goods and services) from foreign countries,
through the migration of jobs from one country to another, regardless of whether activ-
ity remains wholly within the firm. Offshoring in manufacturing is not a new phenom-
enon and is the subject of a large literature (see, e.g., Feenstra and Hanson 1996). The
development of global supply chains has increased the kinds of manufacturing activi-
ties that can be offshored, including the production of intermediate goods. Offshoring
in manufacturing follows traditional patterns of comparative advantage; therefore, the
production activities exported abroad from developed countries tend to be those that
utilize unskilled labor.
However, declining transportation and telecommunications costs have generated
not only trade in intermediate goods, but also trade in services. In the 1990s offshoring
included more skill-intensive tasks, from a broader range of sectors, including services.
Services, which were previously considered nontradable, are a large part of the economy
130    Labor and Protectionist Sentiment

in developed countries, and thus a sizable segment of the labor force may be exposed to
international competition for the first time. Though there is some debate about how to
measure the number of jobs that may be offshored, in general the literature agrees that
the main characteristic of such jobs is that they may be provided from a distance; that
is, over a wire. Blinder (2007) has generated an index of offshoreability, which empha-
sizes whether services are personal (requiring face-to-face interaction; e.g., child-care
workers or surgeons), or impersonal (e.g., call center operators and software develop-
ers). Only the latter are potentially offshoreable, or vulnerable to offshoring. Though
not all offshoreable jobs will be offshored, the scope for trade in services is increasing as
telecommunications costs continue to fall (Blinder 2007).
In addition to exposing a large segment of the public to international competition for
the first time, competitive pressures from services offshoring differ in important ways
from those of manufacturing. This is likely to generate new cleavages over trade, not
along factor or industry lines as in traditional trade theory, but rather along occupa-
tional lines. This suggests that within skill levels, industries, and even firms, workers
may have heterogeneous preferences about trade. Empirical findings suggest that work-
ers displaced due to offshoring sustained longer and more persistent earnings losses
than those displaced for other reasons, and the effect is more pronounced for unskilled
workers. Blinder (2007), for instance, finds that workers in offshoreable jobs experi-
ence lower wages, and that vulnerability to offshoring is not correlated with education.
Across skill levels, some occupations will be exposed to offshoring, while others, espe-
cially high-skill jobs requiring creativity, will benefit from increasing exports of ser-
vices. These fractionalized labor interests may make collective action by labor more
difficult.
Offshoring is particularly politically salient. For example, Mankiw and Swagel (2006)
discuss and document the political climate surrounding the offshore/outsourcing
debate during the 2004 US election. They point to more than a thousand references to
the issue of outsourcing in four major newspapers during that year (compared to fewer
than three hundred in the preceding years). Margalit (2011) finds that voters are sen-
sitive to offshoring-related job loss. Owen (n.d.-a) examines floor speeches on House
debate over four FTAs between 2003 and 2004 and finds both that the issue of offshor-
ing is raised frequently and that greater vulnerability to offshoring among constituents
reduces the likelihood that House members will vote in favor of FTAs. As technological
innovation and the economic integration of countries like China and India continue,
offshoring is likely to become a salient political issue in developed countries, especially
given that attitudes toward offshoring appear to be shaped by noneconomic factors such
as nationalism and out-group attitudes, as suggested by Mansfield and Mutz (2013).

Elasticity of Demand and the Influence of Labor

Economic outcomes influence political ones, per Rogowski (1989), and thus the labor
market outcomes of international trade have important implications for the ability of
 Erica Owen   131

labor to influence trade and redistributive policy. Not only do trade and offshoring
affect the composition and number of jobs in the economy, they also increase the elas-
ticity of demand for labor and heterogeneity of labor interests. Together these affect
the political power of protectionist labor interests, because they limit the bargain-
ing power of unions, as well as the ease of collective action and nature of cleavages
regarding trade.
In addition to generating feelings of job insecurity, the elasticity of demand is an
important determinant of the bargaining power of labor. Fabbri, Haskel, and Slaughter
(2003) find that outward FDI from the United States increased elasticity of demand for
production and nonproduction workers. Multiple studies suggest that multinationals
are more likely to close plants in response to wage shocks (e.g., Doms and Jensen 1998),
suggesting the threat to exit is credible. The well-known article by Scheve and Slaughter
(2004) suggests that workers are aware of this, because inward FDI is positively associ-
ated with feelings of job insecurity.
This weakens the economic position of labor, particularly in developed countries.
Scholars have begun to examine the ways in which globalization hurts the bargaining
power of unions over wages (e.g., Dumont, Rayp and Willeme 2006). Even the threat
of relocation of production is enough to force unions to accept wage cuts. For exam-
ple, trade unions in Belgium agreed to an increase in the workweek from thirty-five to
thirty-eight hours, without additional pay, when Volkswagen threatened to close a plant
in Brussels that was home to three thousand jobs.10
Globalization also is likely to impose constraints on the political influence of labor.
Rogowski (1989) proposes a simple political model in which an increase (decrease)
in wealth is associated with an increase (decrease) in political power and likely policy
influence. For organized labor, it is a challenge to represent heterogeneous interests;
thus the organization of unions becomes an important determinant of the political
influence of labor. In other words, union density (rate of unionization) alone is not a
sufficient measure of the strength of organized labor. It is an open question as to whether
encompassing unions (i.e., those that are concentrated) are better able to navigate these
internal divisions (Garrett 1998), or whether internal cleavages limit their effectiveness,
as suggested by Chase (2008) in his concluding remarks. Unions in the past have been
effective at influencing what comes into a country (e.g., imports and inward FDI) under
certain conditions. For example, Owen (n.d.-b) finds that greater unionization leads to
more barriers to inward FDI only when unions are highly concentrated. However, they
have been less effective in shaping the behavior of domestically based firms that engage
in outward FDI or offshoring.
Further research should examine how globalization interacts with labor market char-
acteristics to shape the influence of labor interests on other policy outcomes, including
compensation and redistribution. These trends are not limited solely to organized labor
interests nor exclusively to developed countries. For example, Rudra (2008) argues that
the bargaining power of labor is a function of the amount of surplus labor. She finds that
certain aspects of globalization increase surplus labor, such that labor does not benefit
much from openness even in labor-abundant countries.
132    Labor and Protectionist Sentiment

New Challenges of Globalization

in Developing Countries
Traditionally, the politics of trade were thought to be very different in developed
and developing countries. However, the above NNTT suggests that openness tends
to benefit skilled labor relatively more than unskilled labor, even in labor-abun-
dant countries. This is consistent with survey findings from developing countries,
in which more-educated respondents are more likely to favor openness than those
who are less educated (Ardanaz, Murillo, and Pinto 2013). Thus, globalization is
likely to generate political and social tension. Relative gains, even in the presence
of absolute gains, have the potential to generate dissatisfaction, and inequality may
be the source of demands for trade protection (e.g., Lu, Scheve, and Slaughter 2012).
Similarly, Robertson and Teitelbaum (2011) argue that inward FDI is likely to increase
social dislocations and lead to rising expectations and find that inward FDI leads
to a higher incidence of strikes in low- and middle-income countries (especially in
nondemocracies).
NNTT makes clear predictions that inequality in developing countries will increase.
Though the empirical research on the relationship between inequality and globalization
is vast, using a variety of measures and types of data (cross-national, subnational, etc.),
recent work suggests that globalization does contribute to rising inequality. For exam-
ple, Goldberg and Pavcnik (2007) find that the skill premium is rising in countries like
Brazil, China, and Mexico. Regional differences in the level of development are exacer-
bated by differing levels of engagement with global economic forces, which contributes
to rising within-country inequality.
As large developing countries like India and China become more integrated into
the world economy, problems regarding the distribution of income will have to be
addressed to mitigate social tension. Income inequality is a potential source of social
instability, as suggested by Robertson and Teitelbaum, and something governments
in developing countries appear to be taking seriously through a variety of policy mea-
sures. For example, the Chinese government has implemented a series of prorural poli-
cies, including the abolition of school fees in rural areas, though research suggests the
inequality of opportunity may be just as important as, or even more important than,
real income inequality (Lu 2013). Similarly, the Brazilian government has implemented
incentives for foreign investors to locate in less-developed regions in the Northeast
and Amazon regions. Brazil and other Latin American countries have also introduced
policies like cash transfer programs targeted to the poor to address inequality.
The recent experiences of developing countries suggest that governments cannot
continue to privilege economic growth without consideration of the distributional con-
sequences. Moreover, they further suggest that globalization can lead to many of the
same political issues for both developed and developed countries. Indeed, the poten-
tial for backlash may be greater given tighter constraints on countercyclical spending in
developing and emerging markets.
 Erica Owen   133

Conclusions

As international trading relations evolve and become more open, the costs of adjust-
ment facing labor will become increasingly important to navigate, particularly in light
of the effects of trade on unemployment and inequality suggested by NNTT. In democ-
racies in particular, failure to do so may lead to a protectionist backlash. Going for-
ward, IPE scholars should reexamine how policies regarding trade and redistribution
serve as substitutes or complements, and the influence of labor interests on both policy
areas.
Continued openness has implications for the political influence of labor as well. Not
only does the presence of mobile capital limit the ability to demand trade protection, but
governments must be aware of a growing gap between labor and corporate income share
as openness limits the ability to demand wage increases.

Notes
1. See Lake (2009) for a summary of the OEP literature and Oatley (2011) for a critique.
2. Voters are consumers as well as workers. Baker (2005) concludes that even after consider-
ing the consumer interest, a HOS model seems to do well in explaining variation in indi-
vidual level preferences.
3. See Milner (2002) for a summary of the debate over whether bad economic times allow
politicians to push through necessary reforms like trade liberalization or makes them
more likely to supply protection, as in the Great Depression.
4. On increasing legalization, see Goldstein and Martin (2000). On flexibility, see Rosendorff
and Milner (2001) for an example emphasizing the role of institutional design, including
flexibility and escape clauses, in the ability to promote cooperation.
5. See Bown (this volume) for more detailed discussion.
6. Industry subsidies are a different mechanism of redistribution. I focus on compensation
because it is targeted toward workers.
7. The ability of governments to provide compensation is the subject of a large literature,
which I do not address here. See Hays (2009) and Rickard (2012) as recent examples and
reviews of existing literature.
8. Though they are not synonymous, I use FDI and offshoring interchangeably here.
9. See Harrison et  al (2011) for a more extensive discussion and review of the empirical
literature.
10. European Foundation for the Improvement of Living and Working Conditions (2008, 20).

References
Alt, James, and Michael Gilligan. 1994. The Political Economy of Trading States:  Factor
Specificity, Collective Action Problems and Domestic Political Institutions. Journal of
Political Philosophy 1 (2): 165–192.
134    Labor and Protectionist Sentiment

Ardanaz, Martin, M. Victoria Murillo, and Pablo M. Pinto. 2013. Sensitivity to Issue Framing on
Trade Policy Preferences: Evidence from a Survey Experiment. International Organization
67 (2): 411–437.
Baker, Andy. 2005. Who Wants to Globalize? Consumer Tastes and Labor Markets in a Theory
of Trade Policy Beliefs. American Journal of Political Science 49 (4): 924–938.
Baldwin, Robert, and Christopher Magee. 2000. Is Free Trade for Sale? Public Choice:  105
(1/2): 79–101.
Bernard, Andrew B., J. Bradford Jensen, and Peter K. Schott (2006). Survival of the Best
Fit:  Exposure to Low-wage Countries and the (Uneven) Growth of U.S. Manufacturing
Plants. Journal of International Economics 68: 219–237.
Blinder, Alan. 2007. How Many U.S. Jobs Might be Offshorable? CEPS Working Paper No. 142
(March).
Boix, Carles. 2006. Between Redistribution and Trade: The Political Economy of Protectionism
and Domestic Compensation. In, Globalization and Egalitarian Distribution, edited by
Samuel Bowles, Pranab Bardan, and Michael Wallerstein, 192–216, Princeton, NJ: Princeton
University Press.
Bown, Chad. 2011. Taking Stock of Anti-dumping, Safeguards and Countervailing Duties,
1990–2009. World Economy 34(12): 1955–1998.
Bradford, Scott. 2006. Protection and Unemployment. Journal of International Economics 69
(2): 257–271.
Busch, Marc, and Eric Reinhardt. 2000. Geography, International Trade, and Political
Mobilization in U.S. Industries. American Journal of Political Science 44 (4): 703–719.
Cassing, James, Timothy McKeown, and Jack Ochs. 1986. The Political Economy of the Tariff
Cycle. American Political Science Review 80 (3): 843–862.
Chase, Kerry. 2003. Economic Interests and Regional Trading Agreements:  The Case of
NAFTA. International Organization 57 (1): 137–174.
Chase, Kerry. 2008. Moving Hollywood Abroad: Divided Labor Markets and the New Politics
of Trade in Services. International Organization 62 (4): 653–687.
Corden, Max W. 1993. The Revival of Protectionism in Developed Countries. In Protectionism
and World Welfare, edited by Dominick Salvatore, 54–79, New  York:  Cambridge
University Press.
Davidson, Carl, and Steven Matusz. 2004. An Overlapping-generations Model of Escape
Clause Protection. Review of International Economics 12 (5): 749–768.
Dixit, Avinash, and John Londregan. 1995. Redistributive Politics and Economic Efficiency.
American Political Science Review 89 (4): 856–866.
Doms, Mark, and J. Bradford Jensen. 1998. Comparing Wages, Skills, and Productivity Between
Domestically and Foreign-owned Manufacturing Establishments in the United States. In
Geography and Ownership as Bases for Economic Accounting, edited by Robert Baldwin,
Robert Lipsey, and J. David Richardson, 235–255, Chicago: University of Chicago Press.
Drope, Jeffrey, and Wendy Hansen. 2004. Purchasing Protection? The Effect of Political
Spending on U.S. Trade Policy. Political Research Quarterly 57 (1): 27–37.
Dumont, Michael, Glenn Rayp, and Peter Willeme. 2006. Does Internationalization Affect
Union Bargaining Power? An Empirical Study for Five EU Countries. Oxford Economic
Papers 58: 77–102.
Dutt, Pusha, and Devashish Mitra. 2005. Political Ideology and Endogenous Trade Policy: An
Empirical Investigation. Review of Economics and Statistics 87 (1): 59–72.
 Erica Owen   135

Egger, Hartmut, and Udo Kreickemeier. 2009. Firm Heterogeneity and the Labor Market
Effects of Trade Liberalization. International Economic Review 50 (1): 187–216.
Ehrlich, Sean. 2010. The Fair Trade Challenge to Embedded Liberalism. International Studies
Quarterly 54: 1013–1033.
European Foundation for the Improvement of Living and Working Conditions. 2008.
Perceptions of Globalisation: Attitudes and Responses in the EU. http://www.eurofound.
europa.eu/pubdocs/2008/07/en/1/ef0807en.pdf.
Fabbri, Francesca, Jonathan Haskel, and Matthew Slaughter. 2003. Does Nationality of
Ownership Matter for Labor Demands? Journal of the European Economic Association 1
(2–3): 698–707.
Feenstra, Robert, and Gordon Hanson. 1996. Foreign Investment, Outsourcing, and Relative
Wages. In Political Economy of Trade Policy: Essays in Honor of Jagdish Bhagwati, edited by
Robert Feenstra and Gene Grossman, 89–128. Cambridge, MA: MIT Press.
Fordham, Benjamin, and Katja Kleinberg. 2012. How Can Economic Interests Influence
Support for Free Trade? International Organization 66 (2): 311–328.
Garrett, Geoffrey. 1998. Partisan Politics in the Global Economy. New  York:  Cambridge
University Press.
Gawande, Kishore, and Christopher Magee. 2012. Free Riding and Protection for Sale.
International Studies Quarterly 56 (4): 1–13.
Goldberg, Pinelopi Koujianou, and Nina Pavcnik. 2007. Distributional Effects of Globalization
in Developing Countries. Journal of Economic Literature 45 (1): 39–82.
Goldstein, Judith. 1986. The Political Economy of Trade: Institutions of Protection. American
Political Science Review 80 (1): 161–184.
Goldstein, Judith, and Lisa Martin. 2000. Legalization, Trade Liberalization, and Domestic
Politics: A Cautionary Note. International Organization 54 (3): 603–632.
Grossman, Gene, and Elhanan Helpman. 1994. Protection for Sale. American Economic Review
84 (4): 833–850.
Hansen, Wendy. 1990. The International Trade Commission and the Politics of Protectionism.
American Political Science Review 84 (1): 21–46.
Harrison, Ann, John McLaren, and Margaret McMillan. 2011. Recent Perspectives on Trade
and Inequality. World Bank Policy Research Working Paper 5754.
Hathaway, Oona. 1998. Positive Feedback:  The Impact of Trade Liberalization on Industry
Demands for Protection. International Organization 52 (30): 575–612.
Hays, Jude. 2009. Globalization and the New Politics of Embedded Liberalism. New York: Oxford
University Press.
Hays, Jude, Sean Ehlrich, and Clint Peinhardt. 2005. Government Spending and Public
Support for Trade in the OECD: An Empirical Test of the Embedded Liberalism Thesis.
International Organization 59 (2): 473–494.
Helpman, Elhanan, and Oleg Itskhoki. 2009. Labor Market Rigidities, Trade and
Unemployment. Centre for Economic Policy Research Working Paper 7502.
Henisz, Witold, and Edward Mansfied. 2006. Votes and Vetoes: The Political Determinants of
Commercial Openness. International Studies Quarterly 50 (1): 189–211.
Hiscox, Michael. 2002. International Trade and Political Conflict: Commerce, Coalitions and
Mobility. Princeton, NJ: Princeton University Press.
Hwang, Wonjae, and Hoon Lee. 2013. Globalization, Factor Mobility, Partisanship, and
Compensation Policies. International Studies Quarterly 58(1): 1–14.
136    Labor and Protectionist Sentiment

Kletzer, Lori. 2001. Job Loss from Imports: Measuring the Costs. Washington, DC: Institute for
International Economics.
Krugman, Paul. 1981. International Trade and Income Distribution: A Reconsideration. NBER
Working Paper No. 356.
Lake, David. 2009. Open Economy Politics:  A  Critical Review. Review of International
Organization 4: 219–244.
Lawrence, Robert, and Matthew Slaughter. 1993. Trade and U.S. Wages: Giant Sucking Sound
or Small Hiccup? BPEA, Microeconomics 2: 161–210.
Lu, Xiaobo. 2013. Equality of Education Opportunity and Attitudes Toward Income
Inequality: Evidence from China. Quarterly Journal of Political Science 8 (3): 271–303.
Lu, Xiaobo, Kenneth Scheve, and Matthew Slaughter. 2012. Inequity Aversion and the
International Distribution of Trade Protection. American Journal of Political Science 56
(3): 638–654.
Mankiw, N. Gregory, and Phillip Swagel. 2006. The Politics and Economics of Offshore
Outsourcing. Journal of Monetary Economics 53 (5): 1027–1056.
Mansfield, Edward, and Marc Busch. 1995. The Political Economy of Non-tariff Barriers:
A Cross-National Analysis. International Organization 49 (4): 723–749.
Mansfield, Edward, and Diana Mutz. 2013. Us vs. Them:  Mass Attitudes toward Offshore
Outsourcing. World Politics 65 (4): 571–608.
Margalit, Yotam. 2011. Costly Jobs:  Trade-related Layoffs, Government Compensation, and
Voting in U.S. Elections. American Political Science Review 105 (1): 166–188.
Matschke, Xenia, and Shane Sherlund. 2006. Do Labor Issues Matter in the Determination of
U.S. Trade Policy. American Economic Review 96 (1): 405–421.
Melitz, Marc. 2003. The Impact of Intra-industry Reallocations and Aggregate Industry
Productivity. Econometrica 71 (6): 1695–1725.
Milner, Helen. 2002. International Trade. In The Handbook of International Relations, edited by
Walter Carlsnaes, Thomas Risse, and Beth Simmons, 448–461. London: Sage Publications.
Milner, Helen. 1988. Resisting Protectionism. Princeton, NJ: Princeton University Press.
Milner, Helen, and Bumba Mukherjeee. N.d. Democracy and the Skill-Bias in Trade Policy in
Developing Countries. Unpublished manuscript.
Milner, Helen, and Benjamin Judkins. 2004. Partisanship, Trade Policy, and Globalization: Is
There a Left-Right Divide on Trade Policy? International Studies Quarterly 48: 95–119.
Milner, Helen, and Keiko Kubota. 2005. Why the Move to Free Trade? Democracy and Trade
Policy in Developing Countries. International Organization 59 (1): 107–143.
Mitra, Devashish, and Priya Ranjan. 2010. Offshoring and Unemployment: The Role of Search
Frictions Labor Mobility. Journal of International Economics 81 (2): 219–229.
Mukherjee, Bumba, Dale Smith, and Quan Li. 2009. Labor (Im)mobility and the Politics of
Trade Protection in Majoritarian Democracies. Journal of Politics 71 (1): 291–308.
Naoi, Megumi. 2009. Shopping for Protection: The Politics of Choosing Trade Instruments in a
Partially Legalized World. International Studies Quarterly 53: 421–444.
Oatley, Thomas. 2011. The Reductionist Gamble:  Open Economy Politics in the Global
Economy. International Organization 65: 311–341.
Owen, Erica. N.d.-a. Labor, Offshoring, and the Political Economy of Trade Liberalization.
Paper presented at the meeting of the International Political Economy Society, Pomona, CA.
Owen, Erica. N.d.-b. The Strength of Organized Labor and the Politics of Foreign Direct
Investment: The Impact of Union Membership and Concentration on Barriers to Inward
 Erica Owen   137

FDI in Developed Democracies. Paper presented at the meeting of the International Studies
Association, San Francisco, CA.
Owen, Erica, and Dennis Quinn. Forthcoming. Does Economic Globalization Influence the
U.S. Policy Mood? A Study of U.S. Public Sentiment, 1956–2011. British Journal of Political
Science.
Rehm, Philipp. 2011. Risk Inequality and the Polarized American Electorate. British Journal of
Political Science 41: 363–387.
Rickard, Stephanie. 2012. Welfare versus Subsidies: Governmental Spending Decisions in an
Era of Globalization. Journal of Politics 74 (4): 1171–1183.
Robertson, Graeme, and Emmanuel Teitelbaum. 2011. Foreign Direct Investment, Regime
Type, and Labor Protest in Developing Countries. American Journal of Political Science 55
(3): 665–667.
Rogowski, Ronald. 1989. Commerce and Coalitions. Princeton, NJ: Princeton University Press.
Rogowski, Ronald. 2006. Trade, Immigration, and Cross-Border Investment. In The Oxford
Handbook of Political Economy, edited by. Barry Weingast and Donald Wittman, 814–830,
Oxford: Oxford University Press.
Rosendorff, Peter, and Helen Milner. 2001. The Optimal Design of International Trade
Institutions: Uncertainty and Escape. International Organization 55 (4): 829–857.
Rudra, Nita. 2008. Globalization and the Race to the Bottom in Developing Countries:  Who
Really Gets Hurt? Cambridge, UK: Cambridge University Press.
Rueda, David. 2007. Social Democracy Inside Out:  Partisanship and Labor Market Policy in
Industrialized Democracies. Oxford: Oxford University Press.
Ruggie, John. 1982. International Regimes, Transactions, and Change: Embedded Liberalism
in the Postwar Economic Order. International Organization 36 (2): 203–223.
Scheve, Kenneth, and Matthew J. Slaughter. 2004. Economic Insecurity and the Globalization
of Production. American Journal of Political Science 48 (4): 662–674.
Scheve, Kenneth, and Matthew Slaughter. 2006. Public Opinion, International Economic
Integration, and the Welfare State. In Globalization and Egalitarian Distribution, edited by
Samuel Bowles, Pranab Bardan, and Michael Wallerstein, 217–260, Princeton, NJ: Princeton
University Press.
Stolper, Wolfgang, and Paul Samuelson. 1941. Protection and Real Wages. Review of Economic
Studies 9 (1): 58–73.
Steagall, Jeffrey, and Ken Jennings. 1996. Unions, PAC Contributions, and the NAFTA Vote.
Journal of Labor Research 17 (3): 515–521.
Takacs, Wendy. 1981. Pressures for Protectionism: An Empirical Analysis. Economic Inquiry
19: 687–693.
Tavares, Jose. 2008. Trade, Factor Proportions, and Political Rights. Review of Economics and
Statistics 90 (1): 163–168.
Walter, Stephanie. 2010. Globalization and the Welfare State: Testing the Microfoundations of
the Compensation Hypothesis. International Studies Quarterly 54 (2): 403–426.
 Chapter 8

D om estic P ol i t i c s a nd
Internationa l Di sp u t e s

B. Pete r Ro se n d orf f

International trade in goods and services occurs within a complex international legal
framework, constituted by both negotiated treaties and customary international law.*
The General Agreement on Tariffs and Trade (GATT), which is still operative under the
structure of the World Trade Organization (WTO), subject to the modifications of 1994,
forms the backbone of international trade law. Together with regional or preferential
trade agreements (PTAs), it is designed to lower trade barriers and increase the flows
and benefits of trade among member states.
States that accede to these international regulatory environments in general intend to
comply with their obligations (Downs, Rocke, and Barsoom, 1996; Chayes and Chayes,
1993). Occasionally political pressures to protect certain industries, sectors, or inter-
ests may be difficult to avoid, even if the international agreements restrict such action
(Rosendorff and Milner, 2001; Rosendorff, 2005). States therefore must balance the need
to comply with their international obligations with the domestic political need to protect
political supporters from international competition. Domestic politics matters for under-
standing the pattern of compliance with international obligations. Policy makers deviate
from WTO and PTA obligations in search of enhanced electoral returns (Simmons 1998;
Naoi 2009).
Responding to the needs of politically influential groups or voters at large, policy
makers enter the gray area between compliance with and abrogation of their interna-
tional commitments. Subsidies (usually implicit) in favor of export interests, for exam-
ple, are rationalized as a legal response to unfair protection abroad; it is often argued
that tariffs at home are reasonable and legal responses to foreign dumping or to provide
temporary protection while an industry retools, or necessary to protect the health and
safety of domestic consumers. Such trade policy choices may or may not be legal under
the WTO or the PTAs to which a state is a signatory. Aware of this ambiguity, states have
developed and refined dispute settlement mechanisms to more effectively adjudicate
 B. Peter Rosendorff    139

these disputes, to clarify obligations, to offer avenues for escape, and to enhance flexibil-
ity within the world trading system.
Domestic politics (actors and institutions) affects the decision to include a dispute
settlement procedure (DSP) in an international trade agreement in the first place.
Politics also affects the design of, and the rules that govern the practice of, the DSP and
the choice to use the mechanism once in place. This chapter explores how domestic
actors and institutions influence the presence or absence of dispute settlement mech-
anisms and how they influence the use and outcome of the adjudicatory procedure.
The chapter closes with a recognition of a feedback effect: that the dispute settlement
process and the disputes themselves in turn have an effect on the domestic politics
(and economics) of the signatory states, an effect not unanticipated by the authors
and designers of these international agreements and the drafters of international law.

Domestic Politics and Design

The procedures specified in the Dispute Settlement Understanding (DSU) adopted dur-
ing the Uruguay Round of negotiations at the WTO permit a contracting party to file
a complaint with the WTO regarding a perceived violation of the treaty on the part of
another member. If formal, bilateral consultations are unproductive (an attempt at a
negotiated resolution), the complainant may request that a panel of independent experts
investigate the matter and make a recommendation. If the panel finds that the offending
action is “inconsistent,” the offending party is obliged to terminate the violating mea-
sure and bring its practice back into conformity with its obligations or compensate the
complainant for damages.1 Ultimately, if a measure is not rescinded and compensation
is not paid, retaliation can be authorized. The finding by the panel can also be appealed
to the Standing Appellate Body.
Across the realm of PTAs, there is significant variation in dispute settlement pro-
cedures, ranging from “weaker,” nonbinding, informal opportunities for consultation
and negotiation (a more diplomatic route); to “stronger,” formal, binding, standing,
third-party courts, the decisions of which have the force of international law, with
appeals processes; to even stronger measures, including the possibility of sanctions for
violation (more “legalistic” design). Scholars have developed a variety of measures of
the degree of legalism, including Hofmann and Kim (2013a) and Haftel (2013).
The DESTA (Dür, Baccini, and Elsig 2013) data set is probably the most comprehen-
sive and compelling compilation of facts regarding PTAs. Dür, Baccini, and Elsig stud-
ied 587 agreements for more than 100 measures, including details on embodied dispute
settlement mechanisms. They find that 87 percent of agreements include some type of
dispute settlement provision, and 6 percent create a standing legal body to adjudicate
cases. Figure 8.1 offers a summary of the variation in the characteristics of DSPs identi-
fied in the DESTA data set.
140    Domestic Politics and International Disputes

Monetary sanctions 3
Creation of standing body 6
Sanctions cross-sector 14
Sanctions within sector 14
Mediation provision 17
Defendant chooses amount 17
Restriction to one forum 17
Provision on forum shopping 18
Delegation to external body 19
Third party chooses amount 21
Arbitration provision 36
Complainant chooses amount 47
Provision on sanctions 51
Consultation provision 79
Some type of dispute settlement 83
0 10 20 30 40 50 60 70 80 90
Percentage of PTAs

Fig.  8.1  Variation in DSP Design across  PTAs.

Source:  Dür, Baccini, and Elsig (2013). Percentages of the 587 PTAs coded that have the feature specified.

Political Uncertainty, Dispute Settlement, and Flexibility

Once an international agreement is in place, leaders experience uncertainty about
the costs of compliance. Industries occasionally experience unanticipated surges of
imports or technological shifts abroad that lower competitiveness of local sectors
or particular skill sets held by local professions. These groups, if they are organized,
will influence trade policy makers to provide protection. This has been described as a
“time-consistency” problem, in which a treaty design that is optimal in expectation ex
ante may not be so ex post.
Of course at the time these international trade agreements are negotiated, the mem-
ber states are aware of these potential contingencies and are unlikely to build institu-
tions and agreements they know they will have difficulty complying with. Leaders build
flexibility into the agreements to manage this time-consistency problem, of which DSPs
are a central flexibility-enhancing mechanism.
Some argue that the signatories use an instrument of international law to prevent
them from succumbing to the particularistic demands of special interests to the ben-
efit of social welfare. This “tying-of-the-hands” interpretation of trade agreements (e.g.,
Maggi and Rodriguez-Clare 1998) suggests that the stronger the agreement is in pre-
venting violations, the more leaders can withstand domestic political pressure and the
greater is the degree of international cooperation.
States interested in maximizing the gains from trade might then be expected to
design very rigid agreements, with deep concessions and few (if any) opportunities
 B. Peter Rosendorff    141

for abrogation. This would especially be true of democratic states. If democracies are
more responsive to political pressure to protect affected industries (at least on average),
we might be inclined to predict that democracies will build agreements that bind very
tightly.
In fact the agreements regulating international trade (and for that matter interna-
tional investment, environmental, and other issue area agreements) to varying degrees
embody a variety of flexibility-enhancing devices designed to reduce the rigidity of the
agreements and to manage these periods of intense, unexpected political pressure to
protect a sector, interest, or industry.2 A “flexibility provision” is “any provision of an
international agreement that allows a country to suspend the concessions it previously
negotiated without violating or abrogating the terms of the agreement” (Rosendorff and
Milner 2001).3
The WTO articles, for instance, explicitly allow signatories to renege on their com-
mitments under certain circumstances. Article XIX allows a country, for a limited time,
to suspend its obligations when increased imports “cause or threaten serious injury to
domestic producers” of import-competing goods.4 Article VI of the GATT, the anti-
dumping (AD) and countervailing duties (CVD) codes, allow member states to apply
duties when imports are “dumped” or when the foreign competitor is being subsidized.
Balance of payments exceptions (Articles XVII and XII), infant industry protection
(XVIII), and tariff renegotiation (XXVII) are also opportunities for WTO-legal applica-
tion of trade barriers. Kim (2010) offers a historical account of how this need for flexibil-
ity resulted in these provisions in the original GATT texts. She finds, for example, that
“uncertainty regarding domestic political support for free trade and trade liberalization
in the United States … was a significant factor leading to flexibility in the design of the
General Agreement” (2010, 41).
A state experiencing a set of circumstances that results in clamoring by a group or
sector for a policy of protection may, in an exercise of policy discretion, choose to oblige
the special interest. This runs the risk, of course, of opening up that state to complaints
from its trading partners. A provision that provides protection for a domestic industry
by definition reduces access to the market by the member’s trading partners. The trading
partners’ complaints may eventually take the form of formal trade disputes under either
the auspices of the WTO or the mechanisms of the applicable PTAs.5
If there are opportunities for signatories to escape their obligations (at least temporar-
ily, until the unexpected political pressure passes and a more “normal” state of affairs
prevails), an affected country may subject itself to the discipline of a DSP under the
WTO or a PTA. The use of a DSP therefore allows a contracting partner to violate the
(spirit, if not the letter of, the) agreement, yet remain within the community of coop-
erating nations and not be immediately punished or banished for its actions. Trading
partners acquiesce, for they may be in need of similar tolerance sometime in the not
too distant future. The dispute settlement processes built into the international trading
regime add flexibility to an otherwise rigid system.
These opportunities for “tolerated defection” (Rosendorff 2005) (or “ ‘forbear-
ance” [Bowen  2010], “efficient breach” [Sykes 1991; Dunoff and Trachtman  1999], or
142    Domestic Politics and International Disputes

“involuntary defection” [Putnam 1988]) all suggest that states that are more suscepti-
ble to political pressure are less likely to sign onto agreements that lack these flexibility
provisions. Before being willing to sign agreements, political leaders need to know that
when their political survival is on the line, those agreements will permit them to take
ameliorative action. Compliance is a political decision; so is the decision to accede to a
compliance regime.
The evidence suggests that democratic polities are more likely to join PTAs
(Mansfield, Milner, and Pevehouse 2007), to trade more freely (Mansfield, Milner, and
Rosendorff 2000; Rosendorff 2006), and to join agreements that embody stronger dis-
pute settlement procedures (Pevehouse, Hafner-Burton, and Zierler 2002; Pevehouse
and Buhr 2013).
Further evidence that flexibility provisions make accession to trade agreements
more likely comes from Kucik and Reinhardt (2008), who show that states that
are able to take advantage of a domestic antidumping mechanism (a key flexibility
device) are significantly more likely to join the WTO, agree to more tightly binding
tariff commitments, and implement lower applied tariffs. That is, flexibility makes
accession at deeper levels more likely. Kucik and Reinhardt take seriously the problem
of endogeneity when understanding the effect of a flexibility provision on the willing-
ness to join an agreement. They correctly remark that “several different endogenous
variables are co-determined: specifically, 1) the existence of flexibility provisions in an
agreement; 2) a state’s decision to enter into the agreement; and 3) the level of conces-
sions” of the depth of the agreement (2008, 479). Johns (2013) establishes that once we
control for the benefits of trade liberalization, deep agreements will be more flexible,
while shallow agreements will be more rigid. Only if a leader knows that she or he has
the flexibility to violate a trade agreement when times are tough will she or he commit
to deeper tariff concessions, initially at the time of trade agreement negotiations.

Variation in Dispute Settlement Design across PTAs

Hofmann and Kim (2013a) identify a set of hypotheses in the literature that explain the
degree of delegation and obligation (terms defined by Goldstein et al. 2000), proxies
for the depth of legalized agreements. Pevehouse and Buhr (2013) and Smith (2000)
suggest that the depth of commercial integration in terms of trade volumes predicts
DSP legalism; depth in terms of specific assets susceptible to capture and holdup such
as foreign direct investment also predict a stronger DSP (Büthe and Milner 2008).
Haftel (2013) shows that strength of the negotiated DSP is correlated with the depth of
implementation of the agreement. Colonial history among the members makes stron-
ger DSPs less likely, but democracy makes stronger DSPs more likely (Hofmann and
Kim 2013a).
There is significant cross-country spillover when it comes to PTA and DSP design.
More legalistic approaches to PTA formation appear to diffuse across borders. Within
 B. Peter Rosendorff    143

the set of fifty-seven PTAs in the Asia-Pacific, Hofmann and Kim (2013a) find that
despite the “ASEAN way” of Asian countries not engaging in open conflict with each
other, more legalistic dispute settlement mechanisms are diffusing across agree-
ments: the more legalistic the agreements are that the members have already joined,
the more likely it is that a new PTA between them will be more legalistic. Like much
of the literature on diffusion (Simmons and Elkins 2004), little explanation is offered
for what is driving this diffusion: Is it learning, competition, emulation, or some other
process?
These studies exploring the domestic and international sources of variation in the
degree of “legalism” generally suffer from the problem of endogeneity. The strength of
the dispute settlement procedure is codetermined with other dimensions of the treaty
design: the depth, scope, and the scale of its membership (Johns 2013). Hence to iso-
late one dimension and look for independent causal variables, perhaps taking the other
dimensions as given and exogenous, generates substantial misspecification bias.6

Functioning and Outcomes of DSPs

Once a mechanism is in place for dealing with the uncertainty associated with domes-
tic politics and the pressure for protection, it follows that domestic politics will influ-
ence the incidence of disputes, or the relative frequency with which the various dispute
settlement mechanisms are appealed to. Similarly, if political uncertainties are driving
the design of DSPs and the frequency with which they are appealed to, they will nec-
essarily influence the outcome of the dispute settlement process—in particular, how
often cases are settled before trial and the circumstances under which cases actually
make it to trial.

On the Incidence of Disputes: Who Files?

Democracy and Development
Within the context of the WTO, substantial modifications were made in the DSP in the
shift from GATT to WTO in 1994. The diplomatic nature of the GATT made the hearing
of a dispute rare and somewhat weak and ineffective. The WTO now entrenches a com-
plainant’s right to a panel, providing for the automatic adoption of panel reports (except
in the case of “negative consensus,” in which all member states, including the complain-
ant and defendant, agree to void the report) and making appellate review available.
The DSP in addition has jurisdiction over all disputes arising under the various GATT
agreements that were rolled into the WTO.
Busch (2000) argues that the democratic dyads are more likely to escalate their dis-
putes to the panel stage than are other dyads. Reinhardt (2000), using data on dispute
144    Domestic Politics and International Disputes

initiation within all GATT/WTO directed dyads from 1948 through 1998, finds that
democracies participate in more, not fewer, GATT/WTO disputes, and they resolve
those disputes less cooperatively as well. Rosendorff and Smith (2013) find that the
most democratic nations are between 6.5 and 51 times more likely to initiate a WTO
dispute than the most autocratic nations. Dyads that have had recent disputes (defined
as a dispute within the dyad in either of the two previous years) are far more likely to
have future disputes than dyads without a recent history of disputes.7 The rate of WTO
dispute onset is less than 6 per 10,000 dyad years without a recent history of disputes.
When there is a prior history of disputes, then the rate of dispute onset jumps to about 39
per 10,000 dyad years.
WTO disputes occur between large economically powerful nations. Rosendorff and
Smith (2013) estimate that the predicted probability for a country at the 75th percentile
of GDP and population of initiating a dispute is a mere 0.00007, less than one dispute per
10,000 dyad years. In larger nations (GDP and population at the 95th percentile) the rate
of dispute onset is substantially higher: approximately 20 disputes per 10,000 dyad years.8

“Who Files?” Is Endogenous to “Who Violates?”

This literature ignores to a substantial degree the selection effect—not all violations
end up as formal disputes. Some are settled via a negotiation process and are not even
reported to the WTO. Davis reports that the United States National Trade Estimate
Report (which monitors the trade barriers of US trade partners) listed 126 trade barriers
by Japan between 1995 and 2004, of which only 6 were addressed in WTO dispute settle-
ment (2012, 9). There is therefore an outstanding question: What determines the set of
cases that are in fact adjudicated (from the set of actual violations)?
First, it could be the tough cases, in which there is genuinely an issue of interpretation
or clarification of obligations required and not simply a matter of opportunistic pro-
tectionism taking place. Second, cases that are more likely to result in compliance once
decided are perhaps more likely to be brought. Rulings are more likely to be limited to
those that can be implemented cheaply rather than those that are (politically) expensive
for the defendant. Hence bringing a case for which the expected ruling is unlikely to be
complied with may potentially bring the court’s legitimacy into question, an outcome a
potential plaintiff state may not prefer to initiate.9 If expected rulings are a function of
the member state’s and the court’s concern for the legitimacy of the court, this will affect
which cases are brought from the set of actual violations.10
Davis (2012) suggests that the checks and balances that characterize democracies
bias dispute settlement toward public lawsuits (filings) and away from informal set-
tlements (negotiations). Davis argues that leaders, to enhance their domestic politi-
cal prospects, must be seen by their legislatures and politically influential sectors and
industries to be enforcing international trade obligations abroad. Filing a lawsuit is a
public signal of their commitment to redistributing in favor of that industry; nego-
tiation as an alternative strategy may solve the problem, but it is done behind closed
doors and prevents a credible signal of the leader’s commitment to that industry from
being sent. There is therefore a political demand for what Davis calls “adjudication,”
 B. Peter Rosendorff    145

and this demand is heightened in polities in which leaders are more accountable, nota-
bly democracies.
The decision to file therefore is indeed a political one; we have argued that the deci-
sion to violate (or at least to appeal to the flexibility-enhancing devices as a basis for
the violation) is also a political one. Most of the literature, however, treats the sample
of filings as given. Davis and Shirato (2007) and Davis (2012) do consider the political
determinants of the decision to file, given a set of observed violations. But the decision
to violate depends crucially on whether the potential violator expects to be filed against.
Hence taking the set of violations as somehow exogenous from domestic politics in both
the domestic and foreign countries is probably a flawed approach.11 The answer to the
question “Who files?” is determined simultaneously with the answer to the question
“Who violates?” Current work doesn’t manage to address this endogeneity.

Leader Turnover
Leaders rely on a coalition of supporters to remain in office. Government changes,
especially those associated with changes in the underlying coalition of supporters, gen-
erate a shift in the profile of trade policy. Leadership changes, especially those asso-
ciated with shifts in the underlying support coalition, lead therefore to changes in a
nation’s tariff and subsidy profile. When a trade policy profile—a set of policies across
industries or sectors—changes, it is likely that some outstanding disputes will now
be settled, and other sectors will see new disputes initiated. Sectors that see their pro-
tection decline are associated with settlement of preexisting disputes; sectors receiv-
ing enhanced protection may be associated with new disputes initiated against their
government.
Leader change, therefore, and especially those changes in leadership associated
with changes in the underlying support coalition, are associated with changes in
the pattern of dispute settlements and filings at the WTO. Bobick and Smith (2013)
find that leader change leads to more new disputes and more settlements of disputes.
Rosendorff and Smith (2013) show that this effect is accentuated when the leader turn-
over is associated with a change in the underlying support coalitions, compared to
situations of either no leader change or leader change without a change in the under-
lying support coalition.
These findings are conditioned by regime type. The effect of leader change on dispute
initiation is much larger in nondemocracies than in democracies. Elections in democra-
cies that bring new leaders to office have a relatively small effect on new dispute initia-
tion; leader changes in autocracies, especially those associated with changes in support
coalition, have a much larger effect on dispute initiation.12 Simulation of the size of the
substantive effect suggests leader change in an autocracy increases the risk of dispute
onset about eightfold when compared with a democracy (Rosendorff and Smith 2013).
A democracy requires a larger supporting coalition; protection is likely to cover a
wider set of industries. However, protection for more sectors comes at a greater cost
to the individual consumers and voters by way of higher prices for goods. In contrast,
autocracies have smaller winning coalitions and little regard for consumers in general;
146    Domestic Politics and International Disputes

protection is likely to be deeper for a smaller set of industries. Since protection benefits
industries (and workers in those industries), but harms consumers in general, protec-
tion will be wider but shallower in democracies than in autocracies that provide deep
but narrow protections.
Since protection for any specific sector is shallower in democracies, the likelihood
that the state will become a defendant in a WTO dispute filing in that sector is lower. The
deep protection offered by an autocracy makes the likelihood of WTO dispute initiation
much greater. Hence leader change in an autocracy is likely to be associated with more
new dispute initiation than leader change in a democracy.

Politics and Outcomes: Adjudication or Settlement?

Disputes (once filed) can be concluded in one of three ways. The case can simply be
“dropped”—35 percent of all cases that are filed fail to report any formal settlement with
the WTO, and no panel is established, although many of these may actually have been
settled but the settlement was simply not reported (Chaudoin, Kucik, and Pelc 2013).
Settled cases (those that reach a mutually agreed solution) and those that go to rulings
via litigation and a panel report (about 50 percent of all cases) make up the rest.
Busch (2000) establishes that cases are more likely to be paneled when the dispu-
tants are democracies. He also finds that those cases paneled by democratic dyads are
more likely to end with concessions by the defendant states. Democracies are also more
likely to be named in petitions filed by the United States in antidumping cases (Busch,
Raciborski, and Reinhardt 2008). Democracies are more willing to participate in the
adjudicatory mechanisms of the WTO and comply with its findings.
“Early settlement” offers the greatest chance of obtaining greater concessions from a
defendant at the GATT/WTO. Developing countries, however, have been more likely to
have disputes proceed to the panel stage against developed countries (Busch 2000). As
a result, they have fared less well in extracting concessions from defendants. The newer,
stronger DSP system reinforces this outcome, given both the incentives to litigate as
well as developing countries’ lack of capacity to push for early settlement (Busch and
Reinhardt 2003).
The designers of the WTO’s DSP clearly intended for member states to solve their
differences via negotiation and settlement. Even when dispute reaches the panel stage, a
ruling often takes the form of an interpretation of the text in the context of the dispute,
rather than a finding of explicit violation. The WTO’s various provisions require a (re-)
negotiated solution whenever possible. Johns and her coauthors (Johns and Pelc 2014;
Johns 2013; Gilligan, Johns, and Rosendorff 2010; Gilligan and Johns 2012) describe this
as “post-adjudicative bargaining.” Stronger courts—with higher levels of enforcement
and jurisdiction—may come with hidden costs: if a plaintiff has private information
about the value of a legal claim, as international courts grow stronger, legal claims mat-
ter more, and the asymmetrical information becomes more important. This lessens the
 B. Peter Rosendorff    147

likelihood of an early settlement to the dispute and increases the likelihood of costly
litigation, exacerbating conflicts between states.

Closing the Loop: The Effects

of DSPs on Trade, Trade Policy,
and Domestic Politics

Domestic politics has been shown to matter for the design of these mechanisms for set-
tling international disputes. It influences the frequency with which violations occur,
the decision to file a dispute, and the outcomes of the DSP. Domestic politics affects the
design, form, and function of international institutions.
There is also a feedback effect. A dispute and the dispute settlement institution can
influence domestic politics. It is of course a central question, of both international polit-
ical economy scholars and those in international law, to ascertain the conditions under
which international economic law can affect state behavior (Martin and Simmons 1998).
State behavior, in turn, is the product of domestic political bargaining and conflict.
Disputes are “fire alarms” that sound when states violate their obligations (McCubbins
and Schwartz 1987). They can activate domestic actors to punish their government politi-
cally for offering distortionary protection to industries (McGillivray and Smith 2000,
2004). If the public cannot observe the policy choices of their executives directly, these
fire alarms may improve the quality of the information that the public can use to infer
their government’s behavior. An election in adverse conditions could result in the “unfair
eviction” of the executive: eviction because the voters suspect the executive has been
extractive; unfair because bad aggregate economic conditions rather than extractive
behavior could have generated the adverse economic environment. To insure themselves
against this unfair eviction, executives sign PTAs, especially those with DSPs, so that
their nonextractive behavior can be more easily observed by the public. Democratically
accountable executives are particularly sensitive to this logic and hence are more likely
to trade away extractive opportunities in return for holding onto office during economic
downturns by signing PTAs. Autocratic executives, less sensitive to the will of the vot-
ers, sign PTAs less frequently or are less concerned about embodying an effective DSP.
As a result democracies are more likely to sign PTAs, especially those with DSPs, than
are autocracies (Mansfield, Milner, and Rosendorff 2000, 2002; Rosendorff 2006).
Democracies are also likely to cooperate more on trade issues due to other mechanisms.
Mansfield, Milner, and Rosendorff (2000) argue, for instance, that the separation of pow-
ers characteristic of democracies leads to more cooperative behavior than in unitary
states.
This feedback implies that there are both economic and political effects of trade
disputes.
148    Domestic Politics and International Disputes

Economic: DSPs and Trade Policy

Domestic politics appears to condition the effect of DSPs on trade volumes. A recent flurry
of papers (Chaudoin, Kucik, and Pelc 2013; Hofmann and Kim 2013b; Bechtel and Sattler
2011; Bown 2001) explore whether disputes increase trade. After all, a dispute is initiated
if a measure adopted by a signatory state has restricted market access by another mem-
ber state. Once the dispute is finalized—via either settlement or ruling—and the offending
measure is removed or an equivalent concession is offered, one might expect trade vol-
umes to resume to levels that existed prior to the implementation of the offending measure.
Bown (2001) and Chaudoin, Kucik, and Pelc (2013) find almost no effect of disputes
on trade volumes; Bechtel and Sattler (2011) find a positive effect of dispute settlement on
trade volumes—exports from a complainant country in a WTO dispute rise significantly
in the three years after a panel ruling, and similar effects are evident in the exports of pro-
complainant third parties. Furthermore, procomplainant third parties experience more
exports than neutral or prodefendant third parties. Hofmann and Kim (2013b) find that
settled cases end with more trade. However, among cases that end and then return to the
WTO’s DSP for compliance hearings—indicating that compliance is in dispute—trade
falls. Furthermore, consistent with what Bhagwati (1988) has called the “law of constant
protection,” new barriers appear to rise when compliance is in dispute.
This confusing picture is probably the result of a misperception of the role of the DSP
in the international trading system. The question that should be put is not, Does trade
increase after a dispute? but rather, Is there more trade in a world trading system with a
DSP compared to a world trading system without one?
There is a temptation to think of panel findings as ordering the removal of a policy that
is ruled as a violation. There is no question that panels clarify obligations, demand that
the parties return to the bargaining table, and renegotiate given the clarifications of the
obligations that the panel has made. The panels, of course, have no enforcement power.
The effect, as we have argued above, is to permit tolerated defection or forbearance and
to add flexibility and enhance stability of the system. Panel rulings encourage further
(“postadjudicative”) negotiation and may actually facilitate collusion. Consequently, the
DSP is not a mechanism to ensure that trade increases on average; it is instead a mecha-
nism to manage tolerated defection.
A system with a DSP therefore better manages defections than a system without one.
Defections are tolerated without the system collapsing into retaliatory trade wars. This
enhanced stability of the system comes at some cost, however. There is less per period
cooperation relative to a system without a DSP, but such a system is more prone to col-
lapse when severe political shocks hit one or more of the member states. A DSP permits
the member states to reap the long-term benefits of a more stable trading regime, even at
the cost of some short-term defection.
The extra flexibility a DSP offers within a trading system also makes countries who are
otherwise less likely to want to sign such an agreement more willing to do so. Leaders,
concerned that the caps on the levels of permitted protection will prevent them from
offering policies in favor of an industry in their supporting coalition when adverse
 B. Peter Rosendorff    149

shocks hit, will find that a DSP permits a government the flexibility to do what it needs
to in times of enhanced political pressure. Such a system therefore is more inclusive and
permits a greater number and variety of states to join it (Rosendorff 2005).
The benefits of a DSP therefore are not associated with more trade after a dispute is
concluded; they are increased stability of the system and a larger membership within a
more liberalized world trading system.

Political: DSPs, Information, and Leader Survival

The previous section suggested that the extra flexibility a trading system with a DSP
offers reduces the likelihood of breakdown and increases membership. Presumably
these features improve the political prospects of the leaders who join these agreements.
Clearly the ability to protect an influential sector in times of political crisis has benefits
for leaders concerned with their political survival. But the leaders are committing them-
selves to systems in which the levels of protection are capped, the procedures by which
trade policy can be changed are restricted and limited, and heightened transparency
over trade policy is required. They make these commitments because their (re-)electoral
prospects are heightened by doing so.
Mansfield, Milner, and Rosendorff (2002) argue that the information generated by
the DSP plays a crucial role in enhancing the electoral survival of leaders in democra-
cies. The DSP, in its role as a clearinghouse for information regarding trade conflicts,
clarifies whether a violation has occurred and when it has been resolved to the satisfac-
tion of both sides. Protection for an industry acts like a tax on consumers and voters
at large, but if voters are poorly informed about the trade policy choices of their lead-
ers, it may be difficult for them to identify how much their current economic distress
is the result of poor policy choices by their leaders and how much is due to unfavorable
economic conditions. A trade agreement with a well-functioning DSP helps the voters
disentangle these two effects. The information provided by the DSP helps the voters dis-
cern whether the bad current conditions are indeed the result of reallocation of income
via trade policy or a consequence of circumstances beyond the leader’s control.
Where the will of the voters matters for leader survival—say in democracies—leaders
will be more likely to sign agreements with DSPs to insure themselves from being
evicted from office in bad times. Leaders in autocracies, who are insulated from popular
concerns, will find the information generated by a DSP less helpful and perhaps even a
nuisance. Information plays a key role in the explanation here, and DSPs are providers
of information voters use to discipline their leaders and keep them from being exces-
sively extractive. The DSP, in addition to adding flexibility to the world trading system,
adds transparency to the policy-making process domestically, which facilitates moni-
toring by voters, at least in democracies.
Mansfield, Milner, and Rosendorff (2002) don’t actually test this informational
mechanism. They offer a reduced-form empirical specification, linking democracies to
trade. While consistent with the theoretical story, it is not quite a complete test. In an
150    Domestic Politics and International Disputes

even more direct test of the mechanism—that voters use information generated by the
international organization to make political decisions—Pelc (2012) investigates Google
searches as a measure of the information flows from the WTO to voters. Using this
innovative measure, he finds that US citizens are concerned about their country being
branded a violator of international law, even when they have no direct material stake in
the case at hand. He finds little evidence, however, that material interests magnify the
reaction to US filings against trade partners.13
Hollyer and Rosendorff (2012) revisit this question and test two aspects of this prob-
lem more closely. First, presumably leaders who sign PTAs survive in office longer—on
the basis of the political gains from signing discussed above. Second, this is more true of
democratic leaders than of autocratic leaders.
PTAs codify and clarify the permitted policy choices by member states, and reduce the
volatility of trade policy, and limit the trade-policy uncertainty experienced by domes-
tic actors. When policy making becomes more predictable, domestic economic agents
make better and more efficient investment and resource allocation decisions. Reduced
uncertainty improves economic performance, strengthening the sitting government’s
hold on office, especially for governments responding to the will of the broader elector-
ate. The effect of signing a PTA is accentuated by the degree to which the leadership is
accountable to the electorate—the degree of democracy.
Hollyer and Rosendorff (2012) build a formal model in which PTAs reduce the vola-
tility of noisy signals about the economy. A risk-averse voter hears these signals and
chooses an action that most closely matches the true state of the economy. After learning
the true state of the economy, the voter decides whether to reelect the government. PTAs
provide an electoral benefit: less volatility increases the voter’s expected utility, which in
turn increases the likelihood that the voter will reelect the government. However, PTAs
are costly for the government because they limit discretion over trade policy. When the
government decides how many PTAs to sign, it balances the electoral benefits against
the policy discretion costs. Hollyer and Rosendorff also provide convincing evidence
that PTAs increase the likelihood that a government will survive in office, and this effect
is heightened in democracies.
PTA accession increases leader survival, especially in democracies by virtue of reduc-
ing uncertainty. DSPs are a mechanism by which information is generated within
a PTA:  filings, panels, and rulings all produce information relevant to inferring the
actions of policy makers and leaders. Less uncertainty and more transparency leads
to improved resource allocation decisions and higher expected welfare of the voters at
large—hence the enhanced effect within democratic polities.

Conclusion

The WTO’s Dispute Settlement Understanding (DSU) plays an increasingly central

role in the “legalized” realm of world trade (Goldstein and Martin 2000), especially in a
 B. Peter Rosendorff    151

period when the diplomatic or negotiating track of the multilateral system seems to be
stalled and the gains from the Uruguay Round incomplete.
The WTO and the “spaghetti bowl”14 of PTAs all attempt to manage a fundamental
trade-off: deeper concessions in the form of lower tariff bindings increase the gains from
freer trade, but countries are reluctant to join an institution that binds too tightly. If they
do join, tighter bindings may lead to increased violations. The rules that govern interna-
tional trade can’t be too onerous, or states won’t cooperate; treaties can’t be too lax either,
or they won’t change behavior in any significant way and will have little effect in bolster-
ing a freer trading regime.
Member states search for the line between permitted policy responses to domestic
political pressure and violations of their international obligations. Policy choices, espe-
cially those that tack closely to the fine line between legal and not, are motivated by the
political benefits that accrue to the government and the policy makers. When the legal-
ity of those policies is disputed by a trading partner, the dispute itself becomes subject to
and explained by domestic political concerns.
Domestic political institutions—regime type, electoral rule, sectoral political
influence—affect the presence and the form of DSPs in the agreements that regulate
international trade. These political determinants are influential in understanding both
the set of violations, as well as the set of disputed cases. Moreover, once a case is filed,
whether it proceeds to adjudication or is settled via negotiation is a political question.
Domestic politics affects the design and functioning of the international trade dispute
settlement process.
It is essential to remember, however, that these agreements and the broader inter-
national trade law regime did not emerge independently of the leaders (and the policy
choices they make) who find themselves subject to that law. That is, they choose a legal
architecture via a negotiated process with an eye to its effects on their political survival.
Hence there is a need to balance the benefits of freer trade with the flexibility associ-
ated with opportunities to protect politically influential industries in times of distress.
Leaders design these instruments fully anticipating the effects of these institutional
choices on the domestic politics they currently face (and might face in the future).
The inclusion of a DSP and its use are designed therefore to enhance political survival
of the leaders who negotiate them in the first place. The presence and/or functioning of a
DSP feeds back and affects domestic politics.
Consequently, this simultaneity of international design and domestic politics makes
the identification of causal arguments difficult. This is especially true when one of the
causal processes is via “information”—where information generated by the DSP affects
a domestic political conflict—that is difficult to measure and observe.
This simultaneity makes one of the central questions of international relations schol-
arship more difficult to answer. What is the effect of international organizations on states’
policies and behavior? Identifying the causal effects is all the more difficult because the
international organizations are themselves the product of a political process. An emer-
gent research agenda will take this simultaneity more seriously and develop theory and
evidence consistent with this feedback process.
152    Domestic Politics and International Disputes

Notes
* Many thanks to Brett Allen Casper, Emine Deniz, and Pedro Silva for excellent research
assistance on this chapter.
1. Article 3.5 of the DSU requires that “all solutions … be consistent with [the WTO] agree-
ments.” Even bilateral deals reached in private consultations must, under the terms of the
treaty, extend any trade concessions to all WTO members. Also, third parties may join a
dispute’s proceedings and observe and contribute to proceedings at every stage.
2. Just a few non-trade examples: Hafner-Burton, Helfer, and Fariss (2011) view the opportu-
nity for states to “derogate” from human rights treaties as a form of tolerated escape/flex-
ibility necessary during emergencies; flexibility provisions have been studied in climate
change treaties (von Stein 2008; Thompson 2010) and regional trade agreements (Baccini
2010), among others.
3. Pelc (2009) suggests that “escape clauses, or ‘pressure valves,’ allow members of an agree-
ment to temporarily suspend their obligations under that agreement following an exog-
enous shock, while assuring other state members a return to compliance in the following
period.”
4. Sykes (1991) suggests that the purpose of Article XIX safeguards is to permit policy makers
to respond to the political pressures of materially injured sectors.
5. See Busch (2007) on forum shopping.
6. Johns and Peritz (this volume) comprehensively catalog the important elements along
which international trade institutions vary. In addition to depth, scope, and membership,
they identify rigidity and institutionalization as being interdependent design elements
that in combination affect state behavior.
7. Davis and Bermeo (2009) make a similar finding: past experience in trade adjudication,
as either a complainant or a defendant, increases the likelihood that a developing country
will initiate disputes. States that frequently file GATT/WTO complaints are, however, less
likely to be targeted in US antidumping decisions (Blonigen and Bown 2003; Bown 2001).
8. Simmons and Guzman (2005) find evidence that poorer countries file fewer cases because
they lack the resources to do so, but not that they fear retaliation from more powerful
trading partners. In contrast, Sattler and Bermauer (2011) suggest that power does play a
role in preventing cases from making it to the WTO, which are perhaps instead dealt with
outside the WTO.
9. Courts are usually treated as nonstrategic actors in this literature. Courts, however, con-
cerned about their own legitimacy and anticipating increased compliance, might be will-
ing to rule more severely (drawing from the same set of cases). Or if the court is concerned
about having its rulings ignored—and undermining its legitimacy—it may rule less
severely.
10. Busch and Pelc (this volume) offer an innovative suggestion for dealing with this
long-standing problem of selection. They suggest that there are now new data on specific
trade concerns (STCs), reported to the committees on technical barriers to trade and sani-
tary and phytosanitary measures at the WTO. The key here is that not all of these STCs
escalate into dispute settlement cases, giving the scholar a sense of the universe from
which disputes may be drawn and insight into the dogs that didn’t bark, the cases that
didn’t get filed.
11. Rickard (2010) goes some way to explore the decision to violate—governments elected
via majoritarian electoral rules and/or single-member districts are more likely to violate
 B. Peter Rosendorff    153

GATT/WTO agreements than those elected via proportional electoral rules and/or mul-
timember districts—but does so independent of the effect of the violation on the potential
dispute that might follow.
12. This work makes use of two new data sets, one collected by Bobick and Smith (2013), which
is an extension of the data collected by Busch and Reinhardt (2003) on the list of cases filed
at the WTO, and the second (the CHISOLS data set) on leader and coalition change, col-
lected by Leeds and Mattes (2013).
13. In contrast to Davis (2012), who argues that countries file disputes to credibly convey to
industries that they are serving their interests. Filing, in Davis’ view (partly), is pandering
to exporters.
14. A term credited to Bhagwati (1988).

References
Baccini, Leonardo. 2010. Explaining Formation and Design of EU Trade Agreements: The Role
of Transparency and Flexibility. European Union Politics 11 (2): 195–217.
Bechtel, M., and Thomas Sattler. 2013. What Is Litigation in the World Trade Organization
Worth? International Organization, forthcoming.
Bhagwati, Jagdish. 1988. Protectionism. Cambridge, MA: MIT Press.
Blonigen, Bruce, and Chad P. Bown. 2003. Antidumping and Retaliation Threats. Journal of
International Economics 60: 249–273.
Bobick, Talya, and Alastair Smith. 2013. The Impact of Leader Turnover on the Onset and the
Resolution of WTO Disputes. Review of International Organizations 8(4): 423–445.
Bowen, T. Renee. 2013. Forbearance in Optimal Multilateral Trade Agreements Stanford
University Working Paper.
Bown, Chad P. 2001. On the Economic Success of GATT/WTO Dispute Settlement. Review of
Economics and Statistics 86(3)811-823
Busch, Marc L. 2000. Democracy, Consultation, and the Paneling of Disputes under GATT.
Journal of Conflict Resolution 44 (4): 425–446.
Busch, Marc L. 2007. Overlapping Institutions, Forum Shopping, and Dispute Settlement in
International Trade. International Organization 61 (04): 735–761.
Busch, Marc L, Rafal Raciborski, and Eric Reinhardt. 2008. Does the Rule of Law Matter? The
WTO and US Antidumping Investigations. Working Paper, Georgetown University.
Busch, Marc L., and Eric Reinhardt. 2003. Developing Countries and GATT/WTO Dispute
Settlement. Journal of World Trade 37 (4): 719–735.
Büthe, Tim, and Helen V. Milner. 2008. The Politics of Foreign Direct Investment into
Developing Countries: Increasing FDI through International Trade Agreements? American
Journal of Political Science 52 (4): 741–762.
Chaudoin, Stephen, Jeffrey Kucik, and K Pelc. 2013. Do WTO Disputes Actually Increase
Trade? SSRN 2299651.
Chayes, Abram, and Antonia Handler Chayes. 1993. On Compliance. International Organi­
zation 47 (2): 175–205.
Davis, Christina. 2012. Why Adjudicate? Enforcing Trade Rules in the WTO. Princeton, NJ:
Princeton University Press.
Davis, Christina L., and Sarah Blodgett Bermeo. 2009. Who Files? Developing Country
Participation in GATT/WTO Adjudication. Journal of Politics 71 (03): 1033–1049.
154    Domestic Politics and International Disputes

Davis, Christina L, and Yuki Shirato. 2007. Firms, Governments, and WTO Adjudication:
Japan’s Selection of WTO Disputes. World Politics 59 (2): 274–313.
Downs, George W., D. M. Rocke, and P. B. Barsoom. 1996. Is the Good News About Compliance
Good News About Cooperation? International Organization 50(3): 379–406.
Dunoff, Jeffrey L., and Joel P. Trachtman. 1999. Economic Analysis of International Law. Yale
Journal of International Law 24 (1): 1–59.
Dür, Andreas, Leonardo Baccini, and Manfred Elsig. 2014. The Design of International
Trade Agreements: Introducing a New Dataset. Review of International Organizations.
9(3):353–375.
Gilligan, Michael J., and Leslie Johns. 2012. Formal Models of International Institutions.
Annual Review of Political Science 15: 221–243.
Gilligan, Michael J., Leslie Johns, and B. Peter Rosendorff. 2010. Strengthening International
Courts and the Early Settlement of Disputes. Journal of Conflict Resolution 54 (1): 5–38.
Goldstein, Judith, Miles Kahler, Robert Keohane, and Anne-Marie Slaughter. 2000.
Introduction: Legalization and World Politics. International Organization 54 (3): 385–400.
Goldstein, Judith, and Lisa L. Martin. 2000. Legalization, Trade Liberalization, and Domestic
Politics: A Cautionary Note. International Organization 54 (3): 603–632.
Hafner-Burton, Emilie Marie, Laurence R. Helfer, and Christopher J. Fariss. 2011. Emergency
and Escape: Explaining Derogation from Human Rights Treaties. International Organization
65 (4):673–707.
Haftel, Yoram Z. 2013. Commerce and Institutions: Trade, Scope, and the Design of Regional
Economic Organizations. Review of International Organizations 8 (3): 389–414.
Hofmann, Tobias, and Soo Yeon Kim. 2013a. Designing Credible Commitment: The Political
Economy of Dispute Settlement Design in PTAs. Unpublished manuscript, 1–40.
Hofmann, Tobias, and Soo Yeon Kim. 2013b. Does Trade Comply? The Economic
Effect(iveness) of WTO Dispute Settlement. Unpublished manuscript.
Hollyer, James R., and B. Peter Rosendorff. 2012. Leadership Survival, Regime Type, Policy
Uncertainty and PTA Accession. International Studies Quarterly 56 (4): 748–764.
Johns, Leslie. 2014. Depth versus Rigidity in the Design of International Trade Agreements.
Journal of Theoretical Politics, 26 (3): 468–495.
Johns, Leslie, and Krzysztof J. Pelc. 2014. Who Gets to Be in the Room? Manipulating
Participation in WTO Disputes. International Organization 68 (3): 663–699.
Kim, Soo Yeon. 2010. Power and the Governance of Global Trade: From the GATT to the WTO.
Ithaca, NY: Cornell University Press.
Kucik, Jeffrey, and Eric Reinhardt. 2008. Does Flexibility Promote Cooperation? An
Application to the Global Trade Regime. International Organization 62 (3): 477–505.
Leeds, Brett Ashley, and Michaela Mattes. 2013. Change in Source of Leader Support
(CHISOLS) Dataset: Coding Rules. Unpublished manuscript.
Maggi, Giovanni, and Andres Rodriguez-Clare. 1998. The Value of Trade Agreements in the
Presence of Political Pressures. Journal of Political Economy 106: 574–601.
Mansfield, Edward D., Helen V. Milner, and Jon C. Pevehouse. 2007. Vetoing Cooperation: The
Impact of Veto Players on Preferential Trading Arrangements. British Journal of Political
Science 37 (03): 403.
Mansfield, Edward D., Helen V. Milner, and B. Peter Rosendorff. 2000. Free to Trade:
Democracies, Autocracies and International Trade. American Political Science Review 94
(2): 305–322.
 B. Peter Rosendorff    155

Mansfield, Edward D., Helen V. Milner, and B. Peter Rosendorff. 2002. Why Democracies
Cooperate More:  Electoral Control and International Trade Agreements. International
Organization 56 (3): 477–514.
Martin, Lisa L., and Beth Simmons. 1998. Theories and Empirical Studies of International
Institutions. International Organization 52 (4): 729–757.
McCubbins, Matthew D., and T. Schwartz. 1987. Congressional Oversight Overlooked: Police
Patrols verus Fire Alarms. In Congress:  Structure and Policy, edited by Matthew D.
McCubbins and Terry Sullivan, 426–440. New York: Cambridge University Press.
McGillivray, Fiona, and Alastair Smith. 2004. The Impact of Leadership Turnover on Trading
Relations Between States. International Organization 58: 567–600.
McGillivray, Fiona, and Alastair Smith. 2000. Trust and Cooperation Through Agent-specific
Punishments. International Organization 54 (4): 809–824.
Naoi, M. 2009. Shopping for Protection:  The Politics of Choosing Trade Instruments in a
Partially Legalized World. International Studies Quarterly 53(2): 421–444.
Pelc, Krzysztof J. 2012. Googling the WTO:  What Search Engine Data Tell Us About the
Political Economy of Institutions. International Organization 67 (3): 629–655.
Pelc, Krzysztof J. 2009. Seeking Escape:  The Use of Escape Clauses in International Trade
Agreements. International Studies Quarterly 752: 349–368.
Pevehouse, Jon C., and Renee Buhr. 2013. Democracy, Legalism, and the Design of Preferential
Trade Agreements. Unpublished manuscript, University of Wisconsin.
Pevehouse, Jon C., Emilie Hafner-Burton, and Matthew Zierler. 2002. Regional Trade and
Institutional Design: Long After Hegemony? Paper presented at the 2002 Midwest Political
Science Association Meetings.
Putnam, Robert D. 1988. Diplomacy and Domestic Politics: The Logic of Two-Level Games.
International Organization 42 (3): 427–460.
Reinhardt, Eric. 2000. Aggressive Multilateralism: The Determinants of GATT / WTO Dispute
Initiation, 1948–1998. Working Paper, Emory University.
Rickard, S. J. 2010. Democratic Differences: Electoral Institutions and Compliance with GATT/
WTO Agreements. European Journal of International Relations 16 (4): 711–729.
Rosendorff, B. Peter. 2006. Do Democracies Trade More Freely? In Democratic Foreign Policy
Making: Problems of Divided Government and International Cooperation, 83–106, edited by
Robert Pahre. London: Palgrave.
Rosendorff, B. Peter. 2005. Stability and Rigidity: Politics and the Design of the WTO’s Dispute
Resolution Procedure. American Political Science Review 99 (3): 389–400.
Rosendorff, B. Peter, and Helen V. Milner. 2001. The Optimal Design of International Trade
Institutions: Uncertainty and Escape. International Organization 55 (4): 829–857.
Rosendorff, B. Peter, and Alastair Smith. 2013. Domestic Political Determinants of the Onset of
WTO Disputes. Working Paper, New York University.
Sattler, Thomas, and Thomas Bermauer. 2011. Gravitation or Discrimination? Determinants
of Litigation in the World Trade Organisation. European Journal of Political Research 50 (2):
143–167.
Simmons, B. A. 1998. Compliance with International Agreements. Annual Review of Political
Science 1: 75–93.
Simmons, Beth A., and Zachary Elkins. 2004. The Globalization of LIberalization:  Policy
Diffusions in the International Political Economy. American Political Science Review 98 (1):
171–189.
156    Domestic Politics and International Disputes

Simmons, Beth A., and Andrew Guzman. 2005. Power Plays & Capacity Constraints:  the
Selection of Defendants in WTO Disputes. Journal of Legal Studies 34: 557–598.
Smith, James McCall. 2000. The Politics of Dispute Settlement Design: Explaining Legalism in
Regional Trade Pacts. International Organization 54 (1): 137–180.
Sykes, Alan. 1991. Protectionism as a Safeguard. University of Chicago Law Review 58: 255–305.
Thompson, Alexander. 2010. Rational Design in Motion: Uncertainty and Flexibility in the
Global Climate Regime. European Journal of International Relations 16 (2): 269–296.
von Stein, Jana. 2008. The International Law and Politics of Climate Change: Ratification of
the United Nations Framework Convention and the Kyoto Protocol. Journal of Conflict
Resolution 52: 243–268.
 Pa rt  I I I

I N DU ST RY- L E V E L
P ROT E C T ION
 Chapter 9

Indu stry - L ev e l
Protecti on

Lucy M. Go odhart

Empirical studies abound with descriptions of the industries that are the classic targets
of protective measures, highlighting the loyalty that they inspire within regional strong-
holds (McGillivray 2004). Similarly, followers of American football will recognize that
an industry often becomes a local emblem and source of identity, with the Steelers and
Packers both named for industries that were once the mainstay of local prosperity. It
would not be odd if we wondered, as scholars, whether their many connections to local
welfare and consciousness also assist those industries in gaining protection.
While we wish to avoid explanations that amount to “just so stories,” tailored to indi-
vidual industries, the distribution of trade barriers across industries is a key starting
point for theory-building. One test of theory is that it must be able to explain the cross-
sectional variation that we actually observe. Moreover, our capacity to both build and
test theory is growing as disaggregated data on protection by industry become more
available.1
Yet the study of industry-level protection contains a central conundrum. The cross-
sectional, industry-level patterns of protection are well-established from existing
empirical analysis (with some recent updating for developing countries). Further, these
patterns appear stable. Many industries, once established as the recipients of trade pro-
tection, see consistent protection over time, even given changes in institutional regime
that imply dramatic shifts in the mechanisms used to protect them (Goldstein and
Gulotty 2014; Bown and Tovar 2011). The stylized facts of industry-level protection, in
other words, are clear and persistent. We would be blind not to recognize that steel and
sugar in the United States have been protected by governments of the left and right, that
the auto industry has counted on government support in many European countries, and
that this has continued despite the leadership of these countries in negotiating multilat-
eral agreements for free trade.
And yet, while these empirical patterns are well understood, we can find little evi-
dence that being employed in a given industry affects an individual’s attitudes to trade.2
160   Industry-Level Protection

The striking evidence of industry effects in protection, therefore, cannot be linked

quickly and easily to a causal mechanism that involves individuals in either lobbying
or voting. Indeed, we face clear difficulties in connecting the empirical record of pro-
tection across industries to the available conceptual frameworks. Even the most widely
respected and frequently invoked model of trade protection, the “protection for sale”
model (discussed later in this chapter and elsewhere in this volume), offers little in the
way of explanation for some of the key findings on industry-level protection.
In this chapter, and as part of efforts to achieve greater resolution of theory and data,
I do four things. First I summarize the key findings on the industry-level pattern of pro-
tection, advancing beyond existing reviews where possible, to include accounts of pro-
tection in the developing world and the agricultural sector. Next I turn to the related
work on individual attitudes showing the disjuncture between trade-policy outcomes
and preferences. Third, I relate industry-level protection to models of endogenous pro-
tection based on lobbying and indicate where the empirical findings are at odds with
the “protection for sale” model. Finally, I introduce and discuss more contemporary
work that uses the conundrum as a launching pad for research that extends or departs
from that last model in key ways. This work acknowledges particular features of indus-
tries that enable them to overcome the collective action issues of lobbying but also con-
siders alternative explanations that neither depend on lobbying nor assume an a priori
interest in compensating workers whose jobs are threatened by international trade. This
new analysis goes far beyond the simple treatment of industry as a fixed effect and enters
into the political dynamics that help to generate trade protection for particular catego-
ries of industry. As such, this exploration of industry-level protection plays a key role in
extending our understanding of the political economy of protection more generally.

Empirical Findings

Canonical reviews of the literature on the political economy of trade policy are available
from Rodrik (1995) and Alt and colleagues (1996). Gawande and Krishna (2003) update
the earlier reviews and offer a more synthesized discussion of the relationship between
theory and empirical analysis.
The broad outlines of the findings on industry-level protection, however, have not
altered since these reviews and indeed are generally consistent with earlier findings for
the United States and other industrialized nations (Ray and Marvel 1984; Marvel and
Ray 1983; Ray 1991). That earlier work has examined trade protection measured by nom-
inal and effective tariffs, the coverage of nontariff barriers (NTBs), and exemptions from
multilateral agreements on trade liberalization. The key findings can be summarized as
follows. An industry is more likely to receive protection if it

• is labor intensive and/or employs low-skill or low-wage workers,

• has high or increasing import penetration or has been in decline,
 Lucy M. Goodhart    161

• produces consumer goods rather than intermediate products, and

• is not mainly engaged in intra-industry trade.3

It is worth considering in greater detail the results of two relatively recent studies,
by Trefler (1993) and Lee and Swagel (1997), both of which look at the coverage ratio
of NTBs by industry. Trefler (1993) analyzes NTB coverage ratios by industry for the
United States in 1983, while Lee and Swagel (1997) look at NTB coverage ratios by indus-
try for forty-one countries in 1988, incorporating both industry and country fixed
effects. These two studies are highlighted because both take into account the endogene-
ity of regressors to the dependent variable, so that import penetration (for example) is
responsive to the level of protection. Thus, and in response, both papers estimate the
determinants of NTBs in a simultaneous equation model in which NTBs and trade
flows are jointly estimated.
Both of these papers find a large and significant effect of import penetration on pro-
tection (although Trefler finds that it is the change in import penetration that is sig-
nificant and substantively important). The result for import penetration was significant
across nearly all specifications estimated by Lee and Swagel (1997) and was substan-
tively large in the one equation in which it could not be precisely estimated. The finding
that industries are more likely to be protected when they see more competition from
imports, then, bears up when the sample set is expanded beyond the United States, and
the specification takes account of endogeneity.
Results on the responsiveness of trade protection to low wages (or the intensive use of
low-skilled workers) are less clear-cut. Lee and Swagel (1997) show that NTBs are higher
in low-wage industries (a finding that echoes that of Baldwin 1985 for the United States)
but that this result does not hold once country fixed effects are introduced. For his part,
Trefler (1993) finds that protection is actually higher in industries in which higher-skill
occupations (including scientists and engineers) are well-represented. Thus, there are
indications that protection might go to low-skill and low-wage industries, but this is not
a consistent finding.4
To summarize, and in two studies that set a high standard for the methodological
treatment of data on trade protection, we see the following results. First, and using the
terminology employed by Gawande and Krishna (2003), there is strong support for a
“status quo” tendency in trade policy.5 Protection seems to be responsive to deviations
from status quo industry conditions, as if trade policy protects workers and capital own-
ers from adjustment costs, and with protection advancing as import penetration and
industry unemployment rise (see Corden 1974 for a classic statement of this approach).6
In addition, we see some support, albeit mixed, for a “social justice” tendency in which
policy makers use trade policy to address issues of inequality, directing the benefits of
trade policy to workers with lower pay.
The studies of protection by industry cited above are mostly based on analysis of the
industrialized economies and some emerging economies. Our potential to expand
upon these analyses has been expanded with the availability of data on trade restric-
tiveness for a much greater number of countries (see Nicita and Olarreaga 2007) and
162   Industry-Level Protection

the compilation of data on temporary trade barriers such as antidumping and counter-
vailing duties for developed and developing countries (see Bown 2005). Initial analysis
of those data suggests that while developing countries have seen significant liberaliza-
tion over time, trade barriers are still higher in countries where gross domestic product
(GDP) per capita is lower, but that developing countries also encounter higher barriers
to their exports (Milner and Kubota 2005; Looi Kee, Nicita, and Olarreaga 2009). While
there has been less attention directed to the inter-industry patterns of protection in
developing countries, Milner and Mukherjee (2009) find that developing countries that
are also democratic are marked by a skill-bias whereby high-skill sectors have seen more
protection than low-skill sectors.7 They connect this finding to the additional respon-
siveness of democratic policy makers to both electoral competition and campaign con-
tributions. Where labor is the abundant factor, a reduction in trade barriers to low-skill
goods is a boon for the low-skill median voter if it can be paired with access to foreign
markets through reciprocal liberalization. Trade liberalization may then reap electoral
gains. By contrast, democratic leaders in developing countries might offer selective pro-
tection to the high-skill, relatively scarce factor in return for campaign contributions.
This pattern suggests that the “social justice” tendency seen by Baldwin (1985) might be
reversed in developing countries, and that the use of trade policy to benefit particular
skill groups could be related to electoral strength.
The findings above come, obviously, from the study of manufacturing. This does
not exhaust the potential for protection, however, with trade distortions pervasive in
the market for agricultural commodities, and with trade policy for services now being
incorporated into the World Trade Organization (WTO) structure (see Barton et  al
2006). Francois and Hoekman (2010) provide a review of the emerging work on ser-
vices. They conclude that the General Agreement on Trade in Services (GATS) of 1994
has been ineffectual in promoting broad liberalization. When it has been accomplished,
services liberalization has been part of the negotiation of bilateral investment treaties or
the investment chapters of preferential trade agreements (see Manger 2008).
The literature on agricultural protection is far more extensive (see Swinnen 2010).8
As canonical work by both Lindert (1991) and Anderson (1986) highlights, agricultural
protection is higher in wealthier economies; “anti-trade” in that protection is higher (or
taxation lower) in economies in which agricultural commodities are mainly imported,
whereas export taxes are often applied where agricultural commodities are exported;
and “anticomparative advantage” in that protection rises as a country’s comparative
advantage in the production of a given commodity falls. As such, agricultural protection
in the developed world may be thought of as exhibiting the same “status quo” tendency
as trade policy in manufactured goods, which is biased against trade and toward declin-
ing industries. The key difference is that agricultural price distortions, and protection to
domestic producers, have been consistently higher than those seen for manufacturing
goods.9
To help explain the observed features of agricultural protection, Thies and Porche
(2007) test what they describe as “mid-range” theories of agricultural protection, noting
the greater ability of farmers (vis-à-vis consumers) to organize as their numbers decrease
 Lucy M. Goodhart    163

and the lower electoral salience of agricultural protection as the percent of expenditure
devoted to food falls with greater wealth. Their conclusion is that agricultural pro-
tection, although an “inefficient curiosity” for economists, can be readily explained
by acknowledging and modeling the role of interest groups organized as lobbies (see
also Anderson, Rausser, and Swinnen 2013). The political consequence of higher food
imports and a shrinking domestic agricultural sector is that a smaller nucleus of remain-
ing farmers is easier to organize. However, this statement of the literature serves to high-
light the central issue in studying the political economy of industry-level protection.
We lack a unifying framework that would explain why it is that certain interest groups
become politicized around trade and effective as lobbies, while others do not. I turn next
to the role of individual preferences regarding trade policy as possible drivers of political
action, including lobbying.

Attitudes to Trade and

Industry-Level Protection

As the preceding discussion implies, political economy theories relate trade policy
outcomes to preferences, with preferences regarding trade policy generally related to
the distributive consequences for the factors (i.e., different types of labor and capital)
involved in the production of a particular type of good.10 Thus, and in explaining why
it is that protection, for industrial goods, so often goes to industries that use low-skilled
labor intensively (at least in advanced, industrialized economies) and that have high or
increasing import penetration, we might first look to attitudes toward trade among indi-
viduals who are employed in low-skill or import-competing industries. Owners of capi-
tal might also lobby, and might be very effective, but many political economy accounts
focus explicitly on the role of labor in lobbying.11
Thus, and before proceeding to the formal models of trade protection and their impli-
cations for the distribution of trade barriers across industries, it is worth mentioning
the empirical findings on attitudes toward trade. In addition, it is important to connect
those attitudes to our understanding of labor as a mobile or specific factor. If labor is
a mobile factor, as in the standard Heskscher-Ohlin model of trade, then employees
move easily among different industries, wage rates equalize across sector, and trade
interests are broad. In other words, the attitude of labor (and most individuals) toward
trade policy will depend on whether greater protection helps or hurts the interests of
labor (or different skill groups within labor), and this in turn will depend on whether the
imported good uses labor intensively.
On the other hand, and if labor is a fixed or specific factor (at least in the short term),
as implied by the Ricardo-Viner model, then individuals cannot move seamlessly across
industries and will face a drop in wages if they do so. In this case, trade interests will be
narrow, in that the welfare of workers will depend on how trade policy helps or hurts
164   Industry-Level Protection

their specific industry. As an implication, and once again if attitudes influence policy,
trade-policy outcomes will then be very differentiated by industry.
While most accounts of trade policy assume that labor and other factors have limited
mobility, we do not have to accept this constraint on modeling. Mayer (1984) develops a
model in which labor is mobile, the median voter is everywhere reliant on labor income
only, and trade policy follows from the interest of the median voter given the country’s
endowments (so that we should see import tariffs in richer countries in which labor is
relatively scarce and import subsidies elsewhere). As Gawande and Krishna (2003) note,
however, empirical evidence is stacked against this model in that trade policy is protec-
tionist everywhere, even where labor is the relatively abundant factor.12
More important is that the distribution of protection across goods is also inconsistent
with an account of trade policy that assumes mobile factors. Rather than legislating and
maintaining a broad set of trade barriers that act similarly for all goods produced by
the relatively scarce factor or factors, most countries administer a relatively small set of
tariffs or NTBs, with large variance in protection across goods, even within industries,
and noticeable tariff “peaks” amid striking dispersion in protection levels (Hoekman
2001). In many countries, tariff and nontariff measures are distinct to the eight- or even
ten-digit level of commodity codes, using the harmonized system of codes operated by
the World Customs Organization. Trade policy, in other words, is granular.
One of the challenges, then, for political economy analysis of protection is to con-
nect policy outcomes to preferences. To start on that challenge, scholars have first
examined how trade policy preferences are manifested in individual attitudes. Recent
work has examined this issue in detail, looking directly at stated preferences on trade
policy. Scheve and Slaughter (2001), for example, examine responses of individuals in
the United States, while Mayda and Rodrik (2005) analyze data from individuals in
twenty-three countries included in the World Values Survey (WVS) or the International
Social Survey Programme (ISSP).
The two studies consider the broad array of factors that might affect preferences, but
both also take care to include controls for the economic implications of trade policy
for individual welfare. If factors are relatively mobile, under the “factor endowments”
model that assumes a Stolper-Samuelson result and follows from Heckscher-Ohlin,
then interests should be associated with an individual’s membership in broad catego-
ries of labor defined by skill level and will vary with the country’s labor endowments.
If factors are relatively immobile, as in the “specific factors” or Ricardo-Viner model,
then preferences will be far more closely tied to individual industries or sectors, since it
is the fate of that industry that will affect returns to labor. The different authors investi-
gate this by including controls for skill level (proxied by years of education or the aver-
age wage for the respondent’s occupation as a measure of skill) and for the exposure of
the respondent’s industry or sector to international trade (generally measured from net
exports from or imports into the industry). Both studies offer strong support for the
factor endowments model and at best lukewarm support for a model of the world in
which individual preferences are linked to their specific industry of employment and its
competitiveness.13
 Lucy M. Goodhart    165

If these results do not give us pause, then perhaps they should. Most political economy
models of trade policy have assumed a specific-factors model, with small, cohesive inter-
est groups formed from capital and labor, and this has been so standard that Beaulieu
(2002b) calls it the “maintained assumption.” By contrast, analysis of survey data indi-
cates that individual preferences are more consistent with interests linked to broad factor
type. Indeed, Scheve and Slaughter (2001, 288) suggest that canonical models of trade
policy be reworked to follow a Hechscher-Ohlin, mobile factors account of economic
interests over trade policy. Yet the variation in the ultimate dependent variable, trade
protection, is hardly consistent with a model in which the preferences of broad factor
coalitions are enacted after an electoral and legislative struggle. While accounts of major
historical shifts in trade regime have been couched in these terms, the current pattern of
trade protection is quite different (see Rogowski 1989 for a discussion of broad coalitions
motivated by trade in the nineteenth century). In the advanced, industrialized countries,
trade protection is industry-specific, highly differentiated, often implemented through
the imposition of temporary trade barriers, and (in between the negotiation of major
trade liberalization agreements) carried out through technocratic and administrative
politics (Kono 2006; Caddel 2014). It does not, in short, conform to a view of politics in
which the distributive effects of trade mobilize popular coalitions for systematic liberal-
ization or protection across a range of goods associated with a given factor.14
So why might popular preferences regarding trade policy indicate a strong role for
broad interests related to skill endowments when actual outcomes seem more consistent
with highly differentiated interests? One answer might be that public preferences are
relatively disconnected from outcomes because trade policy has not, generally, been a
salient issue in electoral politics and mass mobilization.15 When it has been so, as in the
1988 Canadian federal election, votes have followed economic interests based on fac-
tor endowments (Beaulieu 2002a). Further, when a particular industry becomes more
important to residents, industry interests also become more meaningful. Scheve and
Slaughter (2001), for instance, find that respondents who own a home in a county in
which import-competing industries are concentrated, and who are therefore exposed
to the broader effects of industry decline, are more likely to oppose trade. Relatedly,
Busch and Reinhardt (2000) show that there is far greater political action around
trade-exposed industries that are also geographically concentrated (see also Busch and
Reinhardt 2005).
Individual preferences, in other words, are mutable and can be activated as trade
becomes more salient. In general, however, voters’ inferred concerns about the distribu-
tive effects of trade do not appear to be directly instantiated into political outcomes.
Where such effects do not operate, we might expect that trade politics reflect the pref-
erences of a narrower group of actors, who are not well-represented in mass public
surveys. If this is so, however, we need to understand how an interest group account of
trade policy could help us explain the industry-level patterns of protection described in
the first section of this chapter. What are the main precepts of the core, specific-factors,
political economy models of protection, those that are centered on lobbying behavior,
and what are their implications?
166   Industry-Level Protection

Protection for Sale and

Industry-Level Protection

In this section I offer a condensed description of the core political economy models of
protection based on lobbying by the owners of specific factors.16 I then describe the pre-
dictions of these models for industry-level protection and assess how well they explain
the stylized facts presented earlier.
The most recent contribution, the “protection for sale” model developed by Grossman
and Helpman (1994), is the preeminent existing theoretical account, but built off precur-
sors that tapped some of the same dynamics. The first of those earlier works was the
tariff formation function of Findlay and Wellisz (1982), in which tariffs are the result of
competition between two lobbies that represent individuals as consumers and produc-
ers. The rewards to lobbying come from trade policy that increases the domestic price of
the protected good and raises the returns to the owners of the specific factor involved in
its production, with losses to consumers as prices rise. The government is represented
only as a “tariff formation function,” which assigns a particular tariff to each combina-
tion of lobbying dollars from the two groups and assumes a marginal rate of substitu-
tion between the dollars spent by each group (so that government favors one lobby over
another). This model captured the dynamics of competition between groups but does
not assign any objectives to government, beyond the maximization of dollars received
from lobbies.
In turn, the political support function, developed by Hillman (1982), assumes that the
government has a political interest in maintaining the welfare of a particular specific
factor and thus protects the good it produces as world prices fall, but also cares about
the effects on consumer welfare.17 In determining the trade policy for each industry,
therefore, the government trades off benefits to the producer group and the welfare costs
to consumers from protection. As Rodrik (1995) notes, this model is far more explicit
about government objectives, but is in turn opaque on the strategic behavior of potential
interest groups and how much they will lobby.
In addition, and as Gawande and Krishna (2003) highlight, these models are untest-
able because they incorporate unknown quantities.18 The protection for sale ((PFS)
model from Grossman and Helpman (1994) overcomes these issues by assuming that
the government trades off welfare and contributions from the different lobbies in a linear
fashion (and this linear rate of substitution can be estimated from the model). Further,
lobbies act to maximize the joint welfare of members given the expected contributions
from other lobbies and the anticipated decision rule of government. This decision rule
is modeled as a menu auction (due to Bernheim and Whinston 1986). Each lobby (for
any industry that is organized as a lobby) presents a “menu” of offers to the govern-
ment of the different contributions that it will deliver in return for a given schedule of
trade policy measures that raise the domestic price of the good produced by lobby mem-
bers. The technical assuredness of the model, its comprehensive depiction of groups and
 Lucy M. Goodhart    167

government, and its relative tractability have put the PFS model front and center in dis-
cussions of the political economy of protection. Moreover, a signal virtue of the PFS
model is that it serves as a focal point for empirical analysis, identifying the factors that
are assumed to affect outcomes if the model holds true and providing a framework for
more precise testing of assumptions (Gawande and Krishna 2003).
While the derivation of the model is more complicated, its central prediction, indicat-
ing the factors that will affect the industry-level pattern of protection, is relatively simple
and is shown below:

ti I − α L  zi 
= i , i = 1, ..., (1)
n.
1 + ti a + α L  ei 

Here, ti = ( pi − p ) / p is the ad valorem tariff or subsidy on good I where pi is the

domestic price and p is the world price of the good, Ii is an indicator variable that
takes the value 1 if the industry is organized and 0 otherwise, and αL is the proportion
of industries that are organized into lobbies (not all industries are organized so that
αL < 1).19 Further, a is the weight that the government places on consumer welfare rela-
tive to lobbying dollars, zi is the inverse import penetration ratio (or more formally, it
is domestic output of good i over imports), and ei is the price elasticity of demand for
the good.
To clarify the implications of the model, I indicate what Equation 1 would predict for
industries that produce an imported good. In this case, and if the industry is organized,
then the expected tariff is positive, because (Ii − α L ) > 0, and tariffs are higher if less of
the productive sector is organized (because sectors do not waste money bidding against
each other). If, however, the sector is not organized, then the expected outcome is that
the government offers a subsidy on the imported good, reducing its price. Why does it
do so? The government can offer benefits to organized groups at a lower welfare cost if
it recompenses consumers for part of the consumer surplus they lose with subsidies on
the goods produced by groups that are not organized.
In addition, and whatever the fraction of industries that are organized, αL, tariffs (and
other trade barriers) are expected to be lower for industries with higher import penetra-
tion (and thus lower zi).20 Why is this so? If imports into a sector are very high, then the
benefits to producers of a tariff are relatively small in relation to welfare costs borne by
consumers, simply because there are relatively few of those producers. Tariffs also fall as
the price elasticity of demand rises, for the reasons associated with the Ramsey rule; any
price distortion will have larger welfare costs when demand is elastic.
The PFS model has been empirically tested with results from Gawande and
Bandyopadhyay (2000) and Goldberg and Maggi (1999). That analysis essentially tests
the marginal effects that flow from PFS; that protection should rise as zi rises (and
import penetration falls) for organized industries with the reverse effect for industries
that are not organized and for which a higher zi (and lower import penetration) is asso-
ciated with lower subsidies to imports.21 These marginal effects are confirmed by the
empirical analysis. Yet in other ways the PFS model is markedly inconsistent with what
168   Industry-Level Protection

we know about protection. For example, and while the model predicts that sectors that
are not organized will see import subsidies, these trade measures are almost unknown
in practice.22 Perhaps more disturbing is that one of the main and consistent empiri-
cal findings on trade protection is that industries facing high or rising import penetra-
tion are more likely to receive protection and to exhibit higher trade barriers. Yet, and
within PFS, it is precisely the industries that are most exposed to trade that should see
the smallest trade distortions and those that are the least involved in trade that are pre-
dicted to have the largest. How then can we explain the pattern of protection that we
actually observe in the world?

Bridging Theory and Empirical Results

on Industry-Level Protection

The research that has developed since the publication of the PFS model, and which has
sought to resolve the disjuncture between theory and broad findings, has proceeded in
two directions. The first has been to extend the PFS framework, especially by discarding
the assumption of a given αL, or proportion of industries that are organized, and devel-
oping a more explicit framework to explain which industries will be able to lobby and
how the selection into lobbying will affect outcomes. The second direction has been to
articulate a political theory of protection that neither depends on campaign contribu-
tions nor assumes an a priori objective of assisting declining industries.
Under the first heading, and endogenizing the formation of lobbies, authors have
built upon intuitions first modeled by Olson (1965). Lobbying is a form of collective
action; those firms that do not lobby cannot be excluded from policy benefits; and so
individual firms may shirk from joining lobbies, reducing the overall amount of lob-
bying activity we see. If lobbying is comparatively easy for declining industries, then
we will see more effort from these groups and more protection provided to them. This
is precisely the logic that Baldwin and Robert-Nicoud (2007) set out in a theoretical
contribution. Lobbies create rents for any firm producing the protected good. In their
model, Baldwin and Robert-Nicoud (2007) show that firm entry will tend to erode
the rents that go to existing industry “incumbents.” Thus, it is only in declining indus-
tries, in which the future prospects are sufficiently dim that new firms do not enter, that
firms have an incentive to lobby. Baldwin and Robert-Nicoud therefore have a simple
explanation for the “status quo” tendency within protection that supports declining
firms; “It is not that government policy picks losers, it is that losers pick public policy”
(2007, 1064).
The empirical work on lobbying has not directly tested the predictions of the Baldwin
and Robert-Nicoud (2007) model to determine whether “losing” firms and industries
lobby more. Rather, Baldwin and Robert-Nicoud have built in additional assump-
tions on lobbying cohesion and effectiveness in order to see whether this improves the
 Lucy M. Goodhart    169

predictive power of the PFS model and, particularly, to check whether this can account
for anomalous findings in the original tests. For example, the relatively low incidence of
actual protection implies that policy makers place an extremely large weight on welfare,
relative to contributions. Equally, contributions appear very low when compared to the
gains from existing protection. More contemporary research has thus sought to explain
which industries can lobby and whether modeling the lobbying decision more closely
can improve model fit.
Bombardini (2008), for example, notes that larger firms (relative to the size of the
industry) internalize more of the benefits from lobbying, so that industries with a
small number of relatively large firms are likely to be able to lobby more effectively.
She then uses the dispersion of firm size within an industry to proxy for the sectors
that should be able to overcome the collective action issues of lobbying. Estimating
a model that includes the control for size dispersion adds predictive power over
the original PFS model in explaining the distribution of protection across sectors.
Similarly, Gawande and Magee (2012) add a category of “partially organized” indus-
tries to the existing dichotomy of organized or not to account for free-riding behav-
ior and find that this extended modeling of lobbying types helps to explain variation
across sectors and yields more realistic estimates of the weight placed on welfare.
Finally, and in the same vein, Gawande, Krishna, and Olarreaga (2012) allow for com-
petitive lobbying between upstream and downstream producers and find similar
results.
The research referenced above makes advances on the standard PFS model. It brings
greater detail and realism into the depiction of lobbying behavior. It is not clear, how-
ever, whether this research brings us closer to understanding the mechanism that might
be contributing to the “status quo” tendency in the observed pattern of industry-level
protection and the benefits extended to declining industries. Is it that only losers lobby?
If not, what can explain the anti-trade bias seen in trade policy and the relationship
between protection and import competition?
In attempting to confront this conceptual issue more directly, some authors have
re-emphasized risk or loss aversion on the part of both the public and policy makers.
These motivations help to explain why governments are more likely to protect indus-
tries that are losing market share to imports. Tovar (2009), for instance, shows that if
loss aversion among the public is sufficiently strong, protection will go to industries that
are declining due to trade pressures and there will be an anti-trade bias. Freund and
Ozden (2008) develop an augmented PFS model with loss aversion and show that if loss
aversion is sufficiently deep, there will be compensating protection in the face of price
declines.
This work returns to a classic theme of the political economy literature on protec-
tion (Corden 1974; Hillman 1982) and makes important contributions to our under-
standing of how trade policy is often depicted and perceived (see also Lü, Scheve, and
Slaughter 2012). It too, however, confronts questions that derive from the stylized facts.
First, and if public preferences on loss aversion have an impact on implemented policy,
why is that policy more consistent with a specific factor’s account than with the broad
170   Industry-Level Protection

interests related to factor endowments that are evident in mass preferences? Second,
and if the public cares about loss, why haven’t we seen more broad protective measures
in response to dramatic declines in manufacturing output across the developed world?
Last, and as argued forcefully by Rodrik (1995), given the tremendous inefficiencies and
social costs inherent in compensating for risk via trade policy, why is it that policy mak-
ers use this mechanism to do so? As Mayda and Rodrik (2005) point out, trade policy is
one of the few policy areas in which all expert advisors agree. Trade distortions are bad
for welfare. And yet, with other options available, policy makers still use this one. Can
we explain this outcome without simple reliance on exogenous loss aversion preferences
among policy makers?
Goodhart (2014) develops a model that contains a separate, political economy mech-
anism by which policy makers choose trade protection. In her account, policy makers
wish to target benefits to particular districts within a multi-district electoral system
(consistent with a majoritarian electoral system or limited proportionality). However,
and in the presence of legal prohibitions against direct cash transfers to potential voters,
politicians use trade policy to target benefits to districts. The reason for protecting the
declining industries that have lost ground to imports is similar to the rationale set out
by Baldwin and Robert-Nicoud (2007). In expanding industries, the increase in domes-
tic prices that is generated by protection would likely trigger investment in new firms,
meaning that the additional output and jobs associated with protection will take place at
facilities whose location is uncertain when the government sets policy. As such, protect-
ing expanding industries is not an efficient mechanism for targeting transfers. Declining
industries, however, solve the problem of targeting benefits to particular places. The
spare capacity created by decline enables politicians to generate additional rents in spe-
cific locations using trade barriers that divert production to underutilized facilities at
existing firms.23
Further, the industries with the highest start-up costs of capital are also those in which
firms are slowest to leave, and these industries offer the greatest potential for targeting
benefits, because they will have high levels of excess capacity during a downturn. Using
data on temporary trade barriers in the United States, Goodhart (2014) finds that the
predictions of the model are borne out. Decline is associated with protection, but only
in industries in which the start-up capital costs of production are large, and this result
holds even when controlling for campaign contributions.
This work could be further formalized, but it offers a meaningful alternative to the
PFS model and allows testing for related predictions that focus on electoral laws and
the rewards to targeting swing districts. Moreover, it provides a theory that is consis-
tent with the “status quo” tendency apparent in industry-level protection and the highly
differentiated pattern of protection across sectors. Industries like steel, chemicals, and
autos have been frequent recipients of trade barriers, while others, like footwear and
telecommunications apparatus, have undergone wrenching contractions with little sup-
port via trade policy. Last, but not least, it can explain why, as others have noted, we
see “political” effects on trade policy even after we control explicitly for the campaign
 Lucy M. Goodhart    171

contributions that are the central explanatory variable in the PFS model (Ludema,
Mayda, and Mishra 2010; Conconi, Facchini, and Zanardi 2011).

Conclusion

It is a sobering undertaking to re-read Rodrik (1995) and Alt and colleagues (1996)
as summaries of the available literature on the political economy of protection that
were available nearly twenty years ago. It is sobering because so many of the concep-
tual challenges laid down by Rodrik (1995), in particular, are still relevant, and the dis-
juncture between theory and findings on the inter-industry pattern of protection still
holds. A major question for the field, then, is what we have gained in understanding
industry-level protection during that time.
There has been some progress. First, a number of new data sets have been developed
that enable much more direct testing of new theory. Second, the development of the PFS
model has focused empirical analysis either on this model as the maintained assump-
tion or on alternative mechanisms and has put paid to generalized empirical analysis
that is consistent with a number of different theories. Third, we have seen an emerging
interest in patterns of protection in developing countries and emerging economics. This
can only be to the good, in part because of the difference in factor endowments from
the advanced, industrialized world that allow us to test predictions related to the fac-
tor endowments model. Last, but not least, we have seen progress even on the thorny
issue of connecting theory to findings more closely in the developed economy cases that
are best known. Moreover, that work offers tantalizing opportunities for new research
to both economists and political scientists, both because of its focus on the industrial
organization of lobbying and because of the development of models that allow for polit-
ical, and electoral, influences on protection independent of campaign contributions.

Notes
1. This chapter focuses on barriers to trade, including tariff and nontariff barriers, at the
industry level. As Rodrik (1995) highlights, our focus on observed trade measures often
reproduces the anti-trade bias that is endemic to trade policy, with little use seen of import
subsidies, export subsidies, or export taxes. A full and richer account of trade policy would
explain both the observed trade restrictions and the absence of subsidies to trade.
2. See the chapter by Kuo and Naoi in this volume for a comprehensive discussion of indi-
vidual attitudes to trade.
3. See the chapter by Thies in this volume for a discussion of intra-industry trade and that by
Plouffe on heterogeneous firms. The emergence of “new new” trade theory from Melitz
(2003) has highlighted the role of firm productivity in determining participation in trade
and preferences regarding trade liberalization. Work inspired by this literature has stressed
172   Industry-Level Protection

the potential for variation in preferences across firms, within a given sector or industry,
similarly to the work on intra-industry trade. Both point to the role of heterogeneity in
countering the lobbying efforts of protectionist forces. For recent work on this topic see
Kim (2013) and Osgood (2013).
4. See Lü, Scheve, and Slaughter (2012), who report a negative correlation between tariffs
and wages by industry for a large cross-section of countries, and Milner and Mukherjee
(2009) for an apparent reversal of this pattern among developing countries that are also
democracies.
5. This terminology is originally from Baldwin (1985).
6. This is related to the finding that trade policy is “counter-cyclical,” with protection rising as
the macro-economy declines (Bown and Crowley 2013).
7. See also the chapter in this volume by Mukherjee on trade protection in new democracies.
8. Empirical work on agricultural protection has been enriched in recent years by the com-
pilation of new data sets, notably the World Bank’s database of Distortions to Agricultural
Incentives; see Anderson and Valenzuela (2008).
9. Anderson, Rausser, and Swinnen (2013, 2) offer a particularly good statement of the higher
relative costs of agricultural protection: “In 2004, existing agricultural and trade policies
[on agricultural and food products] accounted for an estimated 70 percent of the global
welfare cost of all merchandise trade distortions, even though the agricultural sector con-
tributed only six percent of global trade and three percent of global GDP.”
10. This does not, however, exhaust the potential explanations for individual attitudes toward
trade. For an analysis of the role of consumer interests, see Baker (2003, 2006), and for a
consideration of cosmopolitan interests over globalization, see Hainmueller and Hiscox
(2006). Last, Mansfield and Mutz (2009) explore the role that sociotropic concerns over
the impact of trade on the economy as a whole play in determining attitudes.
11. Magee, Brock, and Young (1989), for instance, argue that lobbying from declining industries
will increase as the industry wage falls, reducing the opportunity cost of lobbying for labor.
12. However, Dutt and Mitra (2002) test a related proposition, given the evidence of consis-
tent protectionism, that the degree of protectionism will depend on the degree of inequal-
ity, with higher inequality conditioning the level of protection because it further separates
the interest of the median voter from that of capital. Thus, higher inequality should be
associated with greater protection in capital rich countries and lower protection in poor,
capital-scarce countries, a finding that holds in the data.
13. O’Rourke et  al. (2001) similarly find evidence for preferences that are consistent with
mobile factors and the Heckscher-Ohlin model of trade effects on welfare, while Beaulieu
(2002b) finds an important role for skill in determining preferences around the 1988 fed-
eral election in Canada (seen as a referendum on the Canada-US Free Trade Agreement),
with some role for industry of employment.
14. The literature on whether factor or industry effects operate in campaign contributions
and congressional votes is also lively (see Ladewig 2006; Beaulieu 2002b). Milner and
Mukherjee (2009) find that it is when labor is relatively immobile (so that industry con-
cerns might be operative) that majoritarian democracies raise trade barriers.
15. This also, of course, raises the issue of why trade is not, in the current era, a salient electoral
issue (Hiscox 2002).
16. This review is primarily based on Rodrik (1995) and Gawande and Krishna (2003).
 Lucy M. Goodhart    173

17. Hillman (1982) assumes that the government is motivated to offset losses to workers in the
import-competing industry and spreads the costs of support to that industry across con-
sumers via trade policy.
18. In Findlay and Wellisz (1982) the unknown parameter is the marginal rate of substitu-
tion between the dollars from different lobbying groups, and in Hillman (1982) it is the
trade-off between the income of the affected interest group and consumer welfare.
19. If all sectors are organized, their lobbying efforts cancel each other out, and there are no
deviations from the free trade equilibrium.
20. Sectors with a lower zi that do not organize will have lower import subsidies imposed
on them.
21. Industries are categorized as organized if they contribute more or if their contributions
vary more with trade flows.
22. To be fair, few if any of the existing political economy models can explain why we do not
observe the full range of potential trade-policy instruments, including import subsidies.
23. Margalit (2011) highlights the local, electoral gains from compensating workers for job
losses associated with foreign trade if trade policy is not used to maintain employment.

References
Alt, James E., Jeffry Frieden, Michael J. Gilligan, Dani Rodrik, and Ronald Rogowski. 1996.
The Political Economy of International Trade: Enduring Puzzles and an Agenda for Inquiry.
Comparative Political Studies 29 (6): 689–717.
Anderson, Kym, Yujiro Hayami, and Aurelia George. 1986. The Political Economy of Agricultural
Protection: East Asia in Comparative Perspective. Sydney and Boston: Allen and Unwin.
Anderson, Kym, Gordon Rausser, and Johan Swinnen. 2013. Political Economy of Public
Policies: Insights from Distortions to Agricultural and Food Markets. Journal of Economic
Literature 51 (2): 423–477.
Anderson, Kym, and Ernesto Valenzuela. 2008. Estimates of Global Distortions to Agricultural
Incentives, 1955–2007. Washington, DC: World Bank.
Baker, Andy. 2006. Who Wants to Globalize? Consumer Tastes and Labor Markets in a Theory
of Trade Policy Beliefs. American Journal of Political Science 49 (4): 925–939.
Baker, Andy. 2003. Why Is Trade Reform So Popular in Latin America? A Consumption-Based
Theory of Trade Policy Preferences. World Politics 55: 423–455.
Baldwin, Robert E. 1985. The Political Economy of US Import Policy. Cambridge, MA: MIT
Press.
Baldwin, Richard E., and Frédéric Robert-Nicoud. 2007. “Entry and Asymmetric Lobbying:
Why Governments Pick Losers.” Journal of the European Economic Association 5 (5):
1064–1093.
Barton, John H., Judith L. Goldstein, Timothy E. Josling, and Richard H. Steinberg. 2006. The
Evoluation of the Trade Regime: The Politics, Law and Economics of the GATT and WTO.
Princeton, NJ: Princeton University Press.
Beaulieu, Eugene. 2002a. Factor or Industry Cleavages in Trade Policy? An Empirical Analysis
of the Stolper–Samuelson Theorem. Economics & Politics 14 (2): 99–131.
Beaulieu, Eugene. 2002b. The Stolper–Samuelson Theorem Faces Congress. Review of
International Economics 10 (2): 343–360.
174   Industry-Level Protection

Bernheim, B. Douglas, and Michael D Whinston. 1986. Menu Auctions, Resource Allocation,
and Economic Influence. Quarterly Journal of Economics 101 (1): 1–31.
Bombardini, Matilde. 2008. Firm Heterogeneity and Lobby Participation. Journal of
International Economics 75 (2): 329–348.
Bown, Chad P. 2005. Global Antidumping Database Version 1.0. World Bank Policy Research
Working Paper.
Bown, Chad P., and Meredith A. Crowley. 2013. Import Protection, Business Cycles, and
Exchange Rates:  Evidence from the Great Recession. Journal of International Economics
90(1): 50–64.
Bown, Chad P., and Patricia Tovar. 2011. Trade Liberalization, Antidumping, and
Safeguards: Evidence from India’s Tariff Reform. Journal of Development Economics 96 (1):
115–125.
Busch, Marc L., and Eric Reinhardt. 2000. Geography, International Trade, and Political
Mobilization in US Industries. American Journal of Political Science 44(4): 703–719.
Busch, Marc L., and Eric Reinhardt. 2005. Industrial Location and Voter Participation in
Europe. British Journal of Political Science 35 (4): 713–730.
Caddel, Jeremy. 2014. “Domestic Competititon over Trade Barriers in the International Trade
Commission.” International Studies Quarterly 58 (2): 260-268.
Conconi, Paola, Giovanni Facchini, and Maurizio Zanardi. 2011. Policymakers’ Horizon
and Trade Reforms. Discussion Paper Series, Center for Economic Policy Research,
London, UK.
Corden, W. Max. 1974. Trade Policy and Economic Welfare. Oxford: Clarendon Press.
Dutt, Pushan, and Devashish Mitra. 2002. Endogenous Trade Policy Through Majority
Voting: An Empirical Investigation. Journal of International Economics 58 (1): 107–133.
Findlay, Ronald, and Stanislaw Wellisz. 1982. Endogenous Tariffs, the Political Economy of
Trade Restrictions, and Welfare.” In Import Competition and Response, edited by Jagdish
Bhagwati, 223–244. Chicago: University of Chicago Press.
Francois, Joseph, and Bernard Hoekman. 2010. Services Trade and Policy. Journal of Economic
Literature 48 (3): 642–692.
Freund, Caroline, and Caglar Ozden. 2008. Trade Policy and Loss Aversion. American
Economic Review 98 (4): 1675–1691.
Gawande, Kishore, and Christopher Magee. 2012. Free Riding and Protection for Sale
International Studies Quarterly 56 (4): 735–747.
Gawande, Kishore, Pravin Krishna, and Marcelo Olarreaga. 2012. Lobying Competition
over Trade Policy. International Economic Review 53 (1): 115–132.
Gawande, Kishore, and Usree Bandyopadhyay. 2000. Is Protection for Sale? Evidence on the
Grossman-Helpman Theory of Endogenous Protection. Review of Economics and Statistics
82 (1): 139–152.
Gawande, Kishore, and Pravin Krishna. 2003. The Political Economy of Trade Policy:
Empirical Approaches In Handbook of International Trade, edited by E. Kwan Choi and
James Harrigan, 139–152. Oxford: Blackwell.
Goldberg, Pinelopi Koujianou, and Giovanni Maggi. 1999. Protection for Sale: An Empirical
Investigation. American Economic Review 89 (5): 1135–1155.
Goldstein, Judith, and Robert Gulotty. 2014. America and Trade Liberalization: The Limits of
Institutional Reform. International Organization 68(2): 263–295.
 Lucy M. Goodhart    175

Goodhart, Lucy M. 2014. “Protection as Targeting: Why Governments Protect Declining

Industries.” Paper presented at the Annual Meetings of the Midwest Political Science
Association, Chicago IL.
Grossman, Gene, and Elhanan Helpman. 1994. Protection for Sale. American Economic Review
84 (4): 833–850.
Hainmueller, Jens, and Michael J. Hiscox. 2006. Learning to Love Globalization: Education
and Individual Attitudes Toward International Trade. International Organization 60 (02):
469–498.
Hillman, Arye. 1982. Declining Industries and Political Support Protectionist Motives.
American Economic Review 72 (5): 1180–1187.
Hiscox, Michael J. 2002. International Trade and Political Conflict: Commerce, Coalitions, and
Mobility. Princeton, NJ: Princeton University Press.
Hoekman, Bernard, Francis Ng, and Marcelo Olarreaga. 2001. Tariff Peaks in the Quad and
Least Developed Country Exports. Washington, DC: World Bank.
Kim, In Song. 2013. “Political Cleavages within Industry:  Firm-level Lobbying for Trade
Liberalization.” Paper presented at the Annual Meeting of the American Political Science
Association, Chicago, IL.
Kono, Daniel Y. 2006. Optimal Obfuscation: Democracy and Trade Policy Transparency.
American Political Science Review 100 (3): 369–384.
Ladewig, Jeffrey W. 2006. Domestic Influences on International Trade Policy: Factor Mobility
in the United States, 1963 to 1992. International Organization 60 (1): 69–103.
Lee, Jong-Wha, and Phillip Swagel. 1997. Trade Barriers and Trade Flows across Countries and
Industries. Review of Economics and Statistics 79 (3): 372–382.
Lindert, Peter H. 1991. Historical Patterns of Agricultural Policy. In Agriculture and the
State: Growth, Employment, and Poverty in Developing Countries, edited by C. Peter Timmer,
29–83. Ithaca, NY: Cornell University Press.
Looi Kee, Hiau, Alessandro Nicita, and Marcelo Olarreaga. 2009. Estimating Trade
Restrictiveness Indices. Economic Journal 119 (534): 172–199.
Lü, Xiaobo, Kenneth Scheve, and Matthew J. Slaughter. 2012. Inequity Aversion and the
International Distribution of Trade Protection. American Journal of Political Science 56 (3):
638–654.
Ludema, Rodney D., Anna Maria Mayda, and Prachi Mishra. 2010. Protection for Free? The
Political Economy of US Tariff Suspensions. Discussion Paper Series, Centre for Economic
Policy Research, London, UK.
Magee, Stephen P., William A. Brock, and Leslie Young. 1989. Black Hole Tariffs and
Endogenous Policy Theory: Political Economy in General Equilibrium. New York: Cambridge
University Press.
Manger, Mark. 2008. International Investment Agreements and Services Markets: Locking
in Market Failure? World Development 36 (11): 2456–2469.
Mansfield, Edward D., and Diana C. Mutz. 2009. Support for Free Trade: Self-Interest,
Sociotropic Politics, and Out-Group Anxiety. International Organization 63(3): 425–457.
Margalit, Yotam. 2011. Costly Jobs: Trade-related Layoffs, Government Compensation, and
Voting in U.S. Elections. American Political Science Review 105 (1): 166–188.
Marvel, Howard P., and Edward J. Ray. 1983. The Kennedy Round: Evidence on the Regulation
of International Trade in the United States. American Economic Review 73 (1): 190–197.
Mayda, Anna Maria, and Dani Rodrik. 2005. Why Are Some People (and Countries) More
Protectionist Than Others? European Economic Review 49 (6): 1393–1430.
176   Industry-Level Protection

Mayer, Wolfgang. 1984. Endogenous Tariff Formation. American Economic Review 74 (5):
970–985.
McGillivray, Fiona. 2004. Privileging Industry: The Comparative Politics of Trade and Industrial
Policy. Princeton, NJ: Princeton University Press.
Melitz, Marc J. 2003. The Impact of Trade on Intra-industry Reallocations and Aggregate
Industry Productivity. Econometrica 71 (6): 1695–1725.
Milner, Helen V., and Keiko Kubota. 2005. Why the Move to Free Trade? Democracy and
Trade Policy in the Developing Countries. International Organization 59 (01): 107–143.
Milner, Helen V., and Bumba Mukherjee. 2009. “Democracy and the Skill-Bias in Trade Policy
in Developing Countries.” Unpublished manuscript, Princeton University.
Nicita, Alessandro, and Marcelo Olarreaga. 2007. Trade, Production, and Protection Database,
1976–2004. World Bank Economic Review 21 (1): 165–171.
O’Rourke, Kevin H., Richard Sinnott, J. David Richardson, and Dani Rodrik. 2001. The
Determinants of Individual Trade Policy Preferences: International Survey Evidence
[with Comments and Discussion]. Brookings Trade Forum (January): 157–206.
Olson, Mancur. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups.
Cambridge, MA: Harvard University Press.
Osgood, Iain. 2013. “Resistance at Home and Allies Abroad: Firms, Industry Associations and
Support for Trade.” Paper presented at the Annual Meeting of the American Political Science
Association, Chicago, IL.
Ray, Edward J. 1991. Protection of Manufactures in the United States. In Global Protectionism: Is
the U.S. Playing on a Level Field?, edited by David Greenaway, 12–36. London: Macmillan.
Ray, Edward J., and Howard P. Marvel. 1984. The Pattern of Protection in the Industralized
World. Review of Economics and Statistics 66 (3): 452–458.
Rodrik, Dani. 1995. Political Economy of Trade Policy. Handbook of International Economics 3:
1457–1494.
Rogowski, Ronald. 1989. Commerce and Coalitions. Princeton, NJ: Princeton University Press.
Scheve, Kenneth F., and Matthew J. Slaughter. 2001. What Determines Individual Trade-Policy
Preferences? Journal of International Economics 54 (2): 267–292.
Swinnen, Johan F. M. 2010. Political Economy of Agricultural Distortions: The Literature to
Date. In The Political Economy of Agricultural Price Distortions, edited by Kym Anderson,
81–104. Cambridge, UK: Cambridge University Press.
Thies, Cameron G., and Schuyler Porche. 2007. The Political Economy of Agricultural
Protection. Journal of Politics 69 (1): 116–127.
Tovar, Patricia. 2009. The Effects of Loss Aversion on Trade Policy: Theory and Evidence.
Journal of International Economics 78 (1): 154–167.
Trefler, Daniel. 1993. Trade Liberalization and the Theory of Endogenous Protection: An
Econometric Study of U.S. Import Policy. Journal of Political Economy 101 (1): 138–160.
 Chapter 10

In tra-Industry T ra de
an d P olicy Ou tc ome s

Timothy M. Pete r son

and Ca me ron G. Th ie s

Introduction

There is little dispute that international trade influences domestic politics and has con-
sequences for a wide variety of international interactions. How the composition of trade
conditions these relationships is less understood. Trade has evolved over time such that
traditional patterns in distinct commodities—aka, inter-industry trade—has dimin-
ished, while the two-way exchange of similar, yet differentiated, commodities—aka,
intra-industry trade—has grown. Given that classical theories of trade explain only the
existence of inter-industry trade, economists have devoted considerable attention to the
economic determinants of intra-industry trade. Scholars of international political econ-
omy (IPE) have also examined the consequences of intra-industry trade for lobbying,
trade policy, and economic integration. Less research has explored other international
political outcomes of intra-industry trade.
The central argument of this chapter is that there is considerable opportunity to syn-
thesize theories of intra-industry trade with those linking trade more broadly to a wide
variety of international interactions. These include topics of perennial interest to IPE
scholars, such as the formation and subsequent performance of preferential trade agree-
ments (PTAs), the design of international economic institutions, and the relative use
of the dispute resolution mechanism within the World Trade Organization (WTO).
If we broaden traditional IPE to include conflict studies, then we might also consider
how intra-industry trade conditions militarized conflict between states and the for-
mation of alliances or rivalries. Implicit in contributions to these topics would be an
increased understanding of the domestic distributional consequences associated with
increases in intra-industry trade and how such changes affect the supply and demand
178    Intra-Industry Trade and Policy Outcomes

of protectionism. In essence, we suggest that carefully considering the composition of

trade may cause us to rethink many of our well-worn arguments in the IPE literature.
In the next section we review the evolution of the literature on intra-industry trade,
noting that its consequences for politics have been explored primarily in studies
of lobbying and trade policy. In the subsequent section we summarize three ques-
tions on the conceptualization and operationalization that political scientists might
consider when designing studies linking intra-industry trade to international poli-
tics. Then we explore some ways in which bilateral intra-industry trade could influ-
ence political affinity, as well as the occurrence of militarized conflict. We conclude
with a summary of the opportunities to explore broader political consequences of
intra-industry trade.

Gains from Comparative Advantage

versus Gains from Scale

David Ricardo (1817) laid the foundations for an understanding of international trade
that has endured hundreds of years. His theory holds that by specializing in produc-
tion of commodities wherein a state has a comparative advantage and trading for other
commodities, the state can reap efficiency gains that would be impossible under autarky.
Ricardo provides a simple illustration of his theory using two countries (England and
Portugal) and two commodities (textiles and wine). England can produce textiles more
efficiently, while Portugal can produce wine more efficiently.1 Accordingly, Ricardo
argues that English production of wine and Portuguese production of textiles would
be (relatively) wasteful. England and Portugal could increase their welfare if England
exports textiles to Portugal, while Portugal exports wine to England, each state fore-
going production of the other good. The Heckscher-Ohlin (HO) model (e.g., Ohlin
1933), which has become the dominant model of inter-industry trade, extends Ricardo’s
basic argument to incorporate three factors of production. Until the twentieth century,
essentially all international trade could be explained in terms of trading states’ relative
endowments of land, labor, and capital.
Although the theory of comparative advantage and the HO model remain relevant and
useful today, scholars have noted that an increasing proportion of international trade no
longer fits the expectations of these models. Since the end of World War II there has
been an increasing proportion of trade within, rather than between, industries—often
among states with very similar factor endowments. A large body of research in eco-
nomics has followed from this observation that the composition of trade did not mir-
ror theoretical expectations. To explain the two-way exchange of similar commodities,
scholars have noted that states might seek to capture gains from scale rather than gains
from specialization, and from the satisfaction of varied consumer tastes (e.g., Krugman
1979). Accordingly, whereas inter-industry trade consists of relatively homogenous
 Timothy M. Peterson and Cameron G. Thies    179

commodities, intra-industry trade is characterized by the exchange of varied products

(e.g., Grubel and Lloyd 1975).

Intra-industry Trade, Domestic

Politics, and Trade Policy

Early economists did not ignore the potential political consequences of trade; however,
research in this vein focused on its consequences for lobbying and trade policy. Following
from classical models of inter-industry trade, Stolper and Samuelson (1941) extended
the Heckscher-Ohlin model to predict how returns to holders of scarce and abundant
factors would respond to liberalization of trade. Expanding on this research, others
(e.g., Rogowski 1987) demonstrate that support or opposition to trade openness can be
explained by the factor of production that a given group utilizes. Using similar logic, Viner
(1950) predicts that increasing economic integration in Europe following World War II
could lead to heightened political resistance by groups harmed by foreign competition.
Yet protectionist backlash against European integration was minimal. Scholars
noted that resistance to liberalization in Europe might be lessened because of the
intra-industry nature of European trade in the post-World War II period (e.g., Balassa
1961). Observing the two-way trade of similar commodities, scholars reasoned that
adjustment costs could be lower when trade did not reflect comparative advantage
stemming from distinct factor endowments. This follows because intra-industry trade
does not lead to the elimination of relatively unproductive industries; although a given
industry in a given state could be less productive, the unique variety of commodities
produced therein remains marketable for export. Accordingly, given the presence of
internal economies of scale, the industry is less likely to mobilize for protectionism
when it is exposed to imports of other varieties of the same commodity. As a result,
resistance to liberalization should also be lower in the presence of two-way trade.
However, a study by Gilligan (1997) suggests that intra-industry trade could increase
lobbying by firms; it is associated with greater ease of overcoming collective action
problems, because the presence of differentiated commodities suggests that the bene-
fits of lobbying will not be distributed across the entire industry. Kono (2009) extends
this work, noting that the nature of domestic political institutions will condition the
influence of intra-industry trade on lobbying—whether for or against protectionism.
Specifically, when institutions reward narrow interests, lobbying will increase when
firms engage in intra-industry trade. Conversely, in the presence of intra-industry
trade, lobbying will decrease when political institutions foster wider competition,
such as along geographic or party lines. Recent work by Madeira (2013) suggests that
intra-industry trade will cause shifts in the nature of political coalitions, reducing incen-
tives to form industry-level associations and increasing firm-level lobbying. While
an unproductive industry engaging in intra-industry trade is not likely to exit due to
180    Intra-Industry Trade and Policy Outcomes

foreign competition, less productive firms within an industry will face pressure to exit.
Taken together, these studies suggest that we could see a qualitative shift in the makeup
of lobbying entities rather than a mere decrease in overall lobbying.
While political scientists have begun to explore the relationship between intra-
industry trade and political competition, work has emphasized preferences regarding lib-
eralization, ability to organize for collective action, and trade-policy outcomes. Wider,
international political determinants and consequences of intra-industry trade have
received far less consideration. Yet the studies conducted thus far suggest that there could
be important links between the structure of trade and international politics. For example,
Gowa and Mansfield (2004) argue that alliances should have a greater trade-facilitating
impact in the presence of scale economies and similar factor endowments, such that
increasing trade is more likely to flow within industries. One study has examined the
connection between intra-industry trade and conflict-propensity among trade part-
ners, noting that intra-industry trade should be robustly pacifying, while inter-industry
trade has more ambiguous effects (Peterson and Thies 2012a). This follows because, at the
state level, inter-industry trade can provoke asymmetrical dependence more easily than
intra-industry trade. Specifically, inter-industry trade could lead one state to rely on its
trade partner for its supply of vital commodities like fuel, metals, or food. While the trade
partner also benefits from such exchange, it could perceive potential to make a demand of
the dependent state, using its advantageous position as leverage (e.g., Hirschman [1945]
1980). Conversely, with intra-industry trade, neither state imports commodities that it
does not also produce domestically. Furthermore, at the subnational level, a higher pro-
portion of intra-industry trade suggests the presence of fewer interests opposed to trade
exposure, who might otherwise lobby for protectionism and who could prefer militarized
conflict to protect their bottom line. Finally, there could be consequences for the Kantian
mechanism of increased cultural understanding for trade in branded commodities,
which lead to greater consumer awareness and deeper understanding of trade partners.

Controversies and Refinements

in Our Understanding
of Intra-industry Trade

While theoretically there should be little controversy surrounding the notion that intra-
and inter-industry trade may have different effects on policy outcomes, empirically
demonstrating that this is the case poses some particular challenges. In the following
discussion we review some of the basic controversies surrounding the measurement
of intra-industry trade, including our own views on what seems most appropriate for
the kind of research typically carried out by political scientists. We argue that there are
three critical concepts to consider when designing studies linking intra-industry trade
to international politics. First, what level of product-level aggregation is appropriate?
Second, is it appropriate to separate horizontal and vertical intra-industry trade? Third,
 Timothy M. Peterson and Cameron G. Thies    181

is it better to examine intra-industry trade at the state or dyad level? As we show below,
each of these decisions carries implications for empirical analysis. We explore how a
researcher’s theoretical orientation suggests answers to each of these three questions.

Product-level Aggregation (What Is an Industry?)

Measures of intra-industry trade can vary when using different levels of aggrega-
tion (Grimwade 1989, 101–106). Early research into intra-industry trade often disag-
gregated trade to the Standard International Trade Classification three-digit (SITC-3)
level. However, some argued that intra-industry trade was a statistical artifact of this
aggregation; the resulting “industries” were arguably broad enough to obscure whether
exchange consisted truly of similar commodities or internationalization of production
led to the exchange of related but distinct commodities (e.g., Finger 1975). We argue that
it is vital for the measure of intra-industry trade to be consistent with the theoretical
expectations of the study. For example, in Peterson and Thies (2012a) we are most con-
cerned with a measure that captures consumer substitutability of commodities. This
leads us to use the following conceptual and empirical approach.
Specifically, given our theoretical focus on the benefits of intra-industry trade in part
following from satisfying consumer demand for variety, we define an “industry” as a
group of trade goods that can generally be considered substitutes by consumers (a group
in which we include firms as well as individuals). This distinction is useful because some
trade that appears to flow within industries could in fact follow from the internation-
alization of production. Specifically, a case in which a commodity is imported, experi-
ences further production, and then is re-exported while retaining the same commodity
code would likely be recorded as vertical intra-industry trade.2 However, this scenario
suggests that the value added to the traded commodity by each trade partner does not
follow from intra-industry specialization, but rather from differing factor endowments.
Indeed, this form of trade is most similar to classic HO trade.3 Accordingly, consumers
will not benefit from a variety of similar goods. To operationalize intra-industry trade
that reflects consumer tastes, disaggregated enough that substitute commodities are
coded the same while internationally produced commodities are distinct, we use data
recording commodity flows from exporter to importer at the Standard International
Trade Classification four-digit level (Peterson and Thies 2012a).
Table 10.1, created using data available from the US Census Bureau, presents a few
examples of aggregation at the SITC three-, four-, and five-digit levels. It demonstrates
that the three-digit level aggregates commodities that consumers would not view as
substitutes. For example, all motorcars intended to carry individuals (other than public
transportation), including passenger vehicles and racing vehicles, are aggregated into
one commodity code. Similarly, all motorcycles and nonmotorized bicycles are aggre-
gated to the same commodity code. Conversely, the SITC five-digit level in many cases
either offers no additional disaggregation from the four-digit level or disaggregates
commodities too much with respect to the goal of capturing consumer substitutes. For
example, cars intended for passenger transportation are already disaggregated fully at
182    Intra-Industry Trade and Policy Outcomes

Table 10.1  Aggregation by SITC Commodity Codes

SITC 3-, 4-,
and 5-Digit Description

781 MOTOR CARS AND OTHER MOTOR VEHICLES PRINCIPALLY DESIGNED FOR THE
TRANSPORT OF PERSONS (NOT PUBLIC TRANSPORT), INCLUDING STATION
WAGONS AND RACING CARS
7811 VEHICLES SPECIALLY DESIGNED FOR TRAVEL ON SNOW; GOLF CARTS AND SIMILAR
VEHICLES
78110 VEHICLES SPECIALLY DESIGNED FOR TRAVEL ON SNOW; GOLF CARTS AND SIMILAR
VEHICLES
7812 MOTOR VEHICLES FOR THE TRANSPORT OF PERSONS (OTHER THAN PUBLIC TRANSPORT),
NOT ELSEWHERE SPECIFIED (N.E.S.)
78120 MOTOR VEHICLES FOR THE TRANSPORT OF PERSONS (OTHER THAN PUBLIC TRANSPORT),
N.E.S.
782 MOTOR VEHICLES FOR THE TRANSPORT OF GOODS AND SPECIAL PURPOSE
MOTOR VEHICLES
7821 MOTOR VEHICLES FOR THE TRANSPORT OF GOODS
78211 DUMPERS DESIGNED FOR OFF-HIGHWAY USE
78219 MOTOR VEHICLES FOR THE TRANSPORT OF GOODS, N.E.S.
785 MOTORCYCLES (INCLUDING MOPEDS) AND CYCLES, MOTORIZED AND NOT
MOTORIZED; INVALID CARRIAGES
7851 MOTORCYCLES (INCLUDING MOPEDS) AND CYCLES FITTED WITH AN AUXILIARY MOTOR,
WITH OR WITHOUT SIDE-CARS; SIDE-CARS
78511 MOTORCYCLES WITH RECIPROCATING INTERNAL COMBUSTION PISTON ENGINE OF
A CYLINDER CAPACITY NOT EXCEEDING 50 CC
78513 MOTORCYCLES WITH RECIPROCATING INTERNAL COMBUSTION PISTON ENGINE OF
A CYLINDER CAPACITY EXCEEDING 50 CC BUT NOT 250 CC
78515 MOTORCYCLES WITH RECIPROCATING INTERNAL COMBUSTION PISTON ENGINE OF
A CYLINDER CAPACITY EXCEEDING 250 CC BUT NOT 500 CC
78516 MOTORCYCLES WITH RECIPROCATING INTERNAL COMBUSTION PISTON ENGINE OF
A CYLINDER CAPACITY EXCEEDING 500 CC BUT NOT 800 CC
78517 MOTORCYCLES WITH RECIPROCATING INTERNAL COMBUSTION PISTON ENGINE OF
A CYLINDER CAPACITY EXCEEDING 800 CC
78519 MOTORCYCLES (INCLUDING MOPEDS) AND CYCLES FITTED WITH AN AUXILIARY MOTOR,
WITH OR WITHOUT SIDE-CARS, N.E.S.

Note: Bold text denotes 3-digit level. Italics denote 4-digit level. Plain text denotes 5-digit level.

the four-digit level. On the other hand, motorcycles are grouped into a single four-digit
product code, whereas at the five-digit level they are disaggregated into six different cat-
egories based on relatively minor variations in engine size. This latter disaggregation
arguably fails to group together motorcycles of varying engine power that consumers
might consider purchasing.
 Timothy M. Peterson and Cameron G. Thies    183

Practically, this level of product aggregation is useful because the UN Comtrade sys-
tem records data at the SITC-4 level (revision 1) for the period spanning 1962 to the
present. In addition, these Comtrade data include records for unit value, which allows
for the disaggregation of vertical and horizontal intra-industry trade using price thresh-
olds. However, the SITC four-digit level is not used universally. For example, Kono
(2009) advocates the Harmonized System (HS) six-digit level of aggregation. Notably,
however, the SITC four-digit level is quite similar to the HS six-digit level, with close to a
one-to-one correspondence. Yet dyadic SITC-4 level data are more widely available on a
yearly basis,4 while dyad-year HS-6 level data are more difficult to obtain.

Horizontal and Vertical Intra-industry Trade

Disaggregation of trade to the commodity rather than industry level has facilitated
improvement in the measurement of intra-industry trade. For example, an examina-
tion of commodities has led to the distinction between horizontal and vertical vari-
ants of trade within industries (e.g., Falvey 1981). Horizontal intra-industry trade is
the exchange of commodities that perform essentially the same function but are dif-
ferentiated by variety. It occurs primarily among states with similar factor endow-
ments and in the presence of monopolistic competition and is determined primarily
by consumer tastes for variety and the presence of increasing returns to scale for firms
in each trade partner (Krugman 1979). Horizontal intra-industry trade as a share of
total trade appears larger between states that have higher income per capita and have
similar income levels (Fontagné, Freudenberg, and Péridy 1998). Conversely, vertical
intra-industry trade is an exchange of commodities that fulfill the same function but
are distinguished by quality. In some ways, vertical intra-industry trade is more like
inter-industry trade than to horizontal intra-industry trade (Blanes and Martin 2000).
Specifically, vertical intra-industry trade can arise due to comparative advantage among
states with differing factor endowments under the condition of perfect competition
(Falvey 1981). According to the findings of Fontagné, Freudenberg, and Péridy (1998),
vertical intra-industry trade composes a larger share of dyadic trade when trade part-
ners have differing incomes, as does inter-industry trade.
The distinction between horizontal and vertical intra-industry trade could have
implications for international politics. Horizontal intra-industry trade suggests the pres-
ence of similar factor endowments (and most likely similar development levels as well).
Accordingly, coalitions within each state are more likely to have complementary inter-
ests. Critically, participating firms in neither state have unilateral incentives to lobby for
protectionism. With vertical intra-industry trade, distributional considerations could
spark resistance in one or both participating states. However, this consequence is less
certain than when trade follows primarily from inter-industry specialization.
Whereas horizontal and vertical intra-industry trade could vary somewhat in their
influence on producers, we contend that the influences of both on consumer prefer-
ences, as well as on the structure of trade dependence, are similar. Accordingly, from
our consumer-oriented, product substitutability perspective, horizontal and vertical
184    Intra-Industry Trade and Policy Outcomes

intra-industry trade are nearly identical in their influence on a number of international

political relationships.5 Notably, if researchers wish to focus on the role of producers
in lobbying for protection, then isolating vertical intra-industry trade could be impor-
tant, because exposure to it might involve relatively greater distributional consequences.
However, from the Kantian perspective, as well as from a focus on dependence and
vulnerability, quality distinctions are probably no more meaningful than horizontal
variations.

The (Sub)state and Dyad Levels of Analysis

An understanding of how intra-industry trade affects international politics depends
in part on whether we examine the phenomenon at the state or the dyad level. At the
state level, intra-industry trade exists when a given state imports and exports simi-
lar commodities in a given industry. As discussed above, such patterns of trade have
implications for the productivity of firms and therefore influence incentives to lobby
for trade policy, as well as resulting levels of protectionism. However, there is no guar-
antee that country-level indicators of imports and exports within a given industry
imply the existence of bilateral intra-industry trade. When two countries engage in
bilateral intra-industry trade, there is potential for additional political consequences.
All else being equal, it is more likely that coalitions favoring the continuance of ami-
cable bilateral trade relations will exist in both states. An examination of bilateral
intra-industry trade is also important to consider given that economic and political
agreements between states could be intended to create an advantage relative to third
parties. We discuss this more fully below. For example, a firm might lobby for PTAs
with trade partners with which it engages in intra-industry trade, while seeking pro-
tection against other states that could also export similar commodities, but with which
existing intra-industry trade levels are low. Bilateral intra-industry trade suggests
a complementarity of interests in the two participating states. As Peterson and Thies
(2012a) note, bilateral intra-industry trade also suggests that trade gains are high and
symmetrical. Dyads engaging in intra-industry trade should therefore be less likely
to experience asymmetrical vulnerability, which could lead to coercion attempts and
conflict.

Operationalizing Intra-industry Trade

for Use in Models of International Politics
Taking the three considerations above into account, we suggest that a bilateral ver-
sion of Grubel and Lloyd’s intra-industry trade measure, using data disaggregated to
the SITC-4 (or HS-6) level, incorporating horizontal and vertical variants, is useful for
many studies of international politics. Specifically, at the commodity level,
 Timothy M. Peterson and Cameron G. Thies    185

 Xijk − Xijk 
Gijk = 1 −  k k 

 Xij + Xij 
k
where Xij is the value of exports from country i to country j (or, conversely, imports of
k
j from i) of commodity k, and Xij is the value of exports from country j to country i (or
imports of i from j) of commodity k. To create a single measure for a given dyad-year, we
take the weighted average of each commodity-level measure with respect to the propor-
tion of dyadic trade comprised by the given commodity, as follows:
N Xijk + Xijk
Gij = ∑ Gijk –
j =1 Xijk + Xijk

where Xij is the value of exports from i to j across all commodities and Xji is the value of
exports from j to i across all commodities. Our final measure varies from 0 to 1, where
0 represents no intra-industry trade and 1 signifies that all trade within the dyad, in a
given year, flows within industries.6
Notably, this measure might be less useful if one’s theoretical aim is to explain firm
lobbying for a state’s multilateral trade policy. To explain firm- and industry-level pref-
erences within a single state, researchers might prefer a measure of intra-industry trade
that uses the combined total of imports from all other states to state i, rather than the
dyadic measure we propose above. Looking within one state, researchers can use the
industry or commodity as the unit of analysis. Doing so reduces possible aggrega-
tion bias inherent in the dyadic measure. Specifically, because the dyadic measure
is a weighted average of intra-industry trade across all industries, an indicator of 0.5
obscures whether 50 percent of trade is two-way in every industry or some industries
are characterized entirely by two-way trade while others are characterized entirely by
one-way trade. Lobbying for protection at the industry level likely depends primarily on
the proportion of two-way trade specifically in that industry.7

Potential Advances in the Study

of International Political Economy

Given that intra-industry trade follows from economies of scale, participating firms in
both countries face similar incentives to promote openness—although this favorability
could be limited to specific trade partners benefiting from bilateral intra-industry trade.
Conventional wisdom has long asserted a causal connection between the simultaneous
increase in intra-industry trade and reduced protectionism in the post–World War II
period. As noted above, lessened distributional consequences associated with reducing
trade barriers render liberalization—whether multilateral or preferential—associated
186    Intra-Industry Trade and Policy Outcomes

with intra-industry trade politically more feasible (e.g., Balassa 1961, 1966; Aquino
1978). However, recent research challenges this conventional wisdom (e.g., Gilligan
1997; Kono 2009), leaving open a number of new research questions. We explore a few
of these below.

Bilateral Intra-industry Trade and PTA Formation

Most of the research in IPE on PTA formation focuses on the domestic institutional
constraints on moving from a state of protectionism to one of preferentially liberalized
trade. A variety of approaches to understanding how to change this domestic status
quo have focused on the role of regime type (e.g., Mansfield, Milner, and Rosendorff
2002; Baccini 2012), bureaucratic interests (Elsig and Dupont 2012), electoral con-
cerns (Hollyer and Rosendorff 2011), the use of trade institutions as a means of lock-
ing in domestic commitment (Maggi and Rodriquez-Clare 2007), interest groups
(e.g., Grossman and Helpman, 2002), and veto players (Henisz and Mansfield 2006;
Mansfield, Milner, and Pevehouse 2007, 2008). Most large-N statistical research has
tended to use veto players as a useful surrogate for domestic-level political activity, since
this concept represents an easily comparable measure of such activity across nations and
time. While having higher numbers of veto players leads to decreased likelihood of PTA
formation, the determinants of domestic support for PTAs are largely left unexamined
in this stream of research.
In most cases there is likely to be considerable variation in domestic support for
PTAs. All else being equal, this variation should affect the likelihood that states enter
into trade agreements. When there is considerable support for PTA formation (e.g.,
when firms want export markets or foreign products), we should expect greater lobby-
ing on behalf of the agreements. Conversely, we expect more lobbying against PTAs as
detractors (primarily import competitors) proliferate. Our own research suggests that
the aforementioned intra-industry trade index is a useful way to gauge this balance of
supporters and opponents of PTA formation and preferential liberalization. The higher
the proportion of a dyad’s trade that is intra-industry, the more likely is the support for
PTA formation.
We contend that higher intra-industry trade within a dyad will encourage firms
in each dyad member to lobby for preferential trade agreements in order to facili-
tate gains from increased trade without risking the potential loss that might accrue
if trade barriers were reduced for all states. PTAs result in expanded markets, which
allow firms engaging in intra-industry trade to benefit further from economies of
scale (e.g., Chase 2003, 2005).8 However, whereas Chase examines industry-level
determinants of lobbying in favor of a PTA, we consider the dyad-level likelihood
of PTA formation. At the firm level, the existence of economies of scale alone is suf-
ficient to encourage lobbying in favor of a PTA. Yet if firms in one state enjoy produc-
tivity advantages over their counterparts in a potential PTA partner, resistance to the
agreement by their potential competitors could thwart their own support for PTA
 Timothy M. Peterson and Cameron G. Thies    187

formation, regardless of how many resources they invest in lobbying in favor of the
agreement. Yet when intra-industry trade already exists, it suggests the presence of
economies of scale and a mutual benefit thereof for firms in each state. Accordingly,
there is mutual willingness to form a PTA across state borders. One might argue that,
logically, the increased competition from firms in partner states would cancel out the
benefit associated with a larger market. While ambiguous gain is perhaps true regard-
ing multilateral liberalization, the net gain associated with PTA formation becomes
evident once one considers the third-party effects of these trade agreements as dis-
cussed by Baldwin (1995). Our argument in this regard rests in part on the role of
third parties.

Bilateral Intra-industry Trade and Third Parties

Previous research shows that intra-industry trade is associated with industry-level
productivity gains because exposure to international competition will drive the least
productive firms to exit, while the more productive firms benefit from increased sales
(e.g., Melitz 2003). While it has been noted that this added efficiency adds to the gains
provided by intra-industry trade, scholars have paid less attention to the fact that
these more productive industries in trading states A and B could represent a threat
to competing industries in third-party state C. An increase in productivity of a given
firm in state A or B would have an effect similar to a lowered tariff in that industry in
state C. Accordingly, import competitors in C could lobby to increase protectionism in
response to this shock. If intra-industry trade leads states A and B to form a preferential
trade agreement, then state C might seek some other state D with which to form a com-
peting agreement.
The indirect implications of intra-industry trade between two states on each of
these states’ political relationship with third parties is worth further examination. For
example, PTA formation would result in productivity gains for member-state firms
that engaged in intra-industry trade due to the enlargement of markets (Baldwin 1995;
Melitz 2003; Melitz and Ottaviano 2008); as such, competing firms in nonmember
states are rendered relatively less efficient and therefore less competitive. Alternatively,
if third-party firms are considerably more efficient than those in states contemplating
a PTA, its formation is attractive because it would lead to expansion of markets with
similarly unproductive firms, while trade barriers could be maintained or even raised
against nonmembers (see Levy 1997). This potential for PTAs to be “trade diverting” has
been suggested repeatedly in extant literature (e.g., Viner 1950; Bhagwati 1991). While
trade diversion is typically viewed in negative terms, rational firms should pursue it
when it protects them from more efficient competitors. As such, there should be con-
siderable domestic lobbying for entrance into PTAs on behalf of sectors and industries
engaged in intra-industry trade.
Importantly, because intra-industry trade signifies that trade partners do not spe-
cialize (as they do under conditions of inter-industry trade), it is less likely that firms
188    Intra-Industry Trade and Policy Outcomes

(or entire industries) will be driven out of business because a trade partner has the
comparative advantage in producing a given traded good. As such, there will be fewer
losers due to expanded trade, and therefore there will be fewer actors lobbying against
entrance into PTAs, relative to cases in which there is a high degree of inter-industry
trade. It is important to note that, counter to conventional wisdom, this aspect of
intra-industry trade would not be sufficient for PTA formation if not for the third-party
competitive element discussed above. Yet the combination of substantial gains for
exporters in both potential PTA members (relative to third parties) and relatively little
loss for importers meets Grossman and Helpman’s (1995) necessary conditions for the
formation of a PTA.
In our own work, we empirically test these arguments about the importance of
intra-industry trade to PTA formation (Peterson and Thies 2011; Thies and Peterson
forthcoming). We find that intra-industry trade is associated with a higher likelihood
of PTA formation, regardless of how PTAs are conceived. Substantively, we find that
moving from one standard deviation below to one standard deviation above mean
intra-industry trade produces an average 25  percent increase in the probability of
entrance into a PTA. We find that PTA formation is also more likely when there are
third-party PTAs in force. In addition to these central findings, we also reconfirm the
negative effects of veto players on PTA formation found in the existing literature as well
as the effects of the standard control variables in this literature.

WTO Dispute Resolution

The WTO dispute settlement procedure (DSP) represents a legalization of international
trade disputes that serves to increase transparency and foster credible commitments of
WTO members to abide by WTO rules (e.g., Goldstein and Martin 2000). By “bind-
ing” states to the cause of liberalization, the DSP provides insulation for national gov-
ernments against potentially powerful domestic interests that favor protectionism (e.g.,
Goldstein and Martin 2000; Rosendorff 2005). However, previous research has shown
that the DSP does not necessarily foster decreased trade competition between states,
and that democracies, because they are more responsive to domestic interests, are more
likely to initiate trade disputes (Reinhardt 1999).
We suggest that bilateral intra-industry trade could serve as an indicator of decreased
willingness to initiate a trade dispute against a given trade partner via the DSP. This logic
follows because a higher proportion of intra-industry trade suggests that fewer indus-
tries face harm due to trade exposure. Furthermore, given the tendency of states to ini-
tiate countersuits (Reinhardt 1999), firms engaging in intra-industry trade should be
especially resistant to the use of the DSP, because they are exporters as well as import-
ers. Initial results from Thies and Peterson (forthcoming) confirm the expectation that
higher proportions of intra-industry trade within a trading relationship reduce the like-
lihood of either party making use of the DSP.
 Timothy M. Peterson and Cameron G. Thies    189

Intra-industry Trade and the

Trade-Conflict Literature

A large literature examines the link between trade policy and international politi-
cal relations, with a focus on the potentially pacifying impact of liberalized trade (e.g.,
McDonald 2004; McDonald and Sweeney 2007; Gartzke and Zhang, this volume). A syn-
thesis of these theories with those linking intra-industry trade to liberalization suggests
at least an indirect link between a higher proportion of intra-industry trade and a lower
propensity for conflict. Specifically, to the extent that resistance to liberalization is lower
in the presence of more intra-industry trade, exposure to this two-way trade could lead to
the reduction—or at least the maintenance—of preexisting trade barriers. Accordingly,
all else being equal, a given value of trade could imply lower trade barriers, and therefore
have a stronger pacifying impact, when the trade is intra-industry in nature.
Although theories of trade and conflict advanced in recent years tend to focus on the
role of domestic politics, the period from the end of World War II through the end of the
Cold War saw numerous state-level theories focusing with realist origins. These theories
emphasized vulnerability from trade, noting that gains from trade were also potential
losses to be leveraged by trade partners. From the perspective of interdependence the-
ory, a higher proportion of intra-industry trade could imply a lower likelihood of asym-
metrical dependence. We have argued in previous work that the resulting symmetry of
trade gains (and therefore opportunity costs) maximizes the pacifying impact of trade
(Peterson and Thies 2012a). We find that higher levels of intra-industry trade are associ-
ated with lower levels of dyadic militarized conflict. Overall trade interaction typically
has no effect on the likelihood of dyadic conflict when controlling for intra-industry
trade. This finding stands in contradiction to much of the peace-through-trade lit-
erature. Finally, development has no effect on the likelihood of dyadic conflict in the
absence of intra-industry trade. These findings have helped to define some of the
boundary conditions for the commercial peace thesis.

Bilateral Intra-industry Trade, Preference

Similarity, and Allied/Rival Relationships
Most of the existing literature has focused on the effects of alliances and political simi-
larity on trade, with little explicit work on how rivalry might affect trade and no explicit
consideration of the effects that the composition of trade might have on alliances, politi-
cal similarity, or rivalry. For example, Gowa (1989) argues that free trade agreements
are more likely to occur within alliances than outside them and more likely to occur in a
bipolar than a multipolar system. These arguments are replicated in Gowa and Mansfield
(1993), who provide statistical analysis demonstrating that alliances significantly
190    Intra-Industry Trade and Policy Outcomes

increase trade, in particular during periods of bipolarity. Morrow, Siverson, and Tabares
(1998) provide analyses that support the idea that similarity of political relations and
democratic dyads increase trade flows. Alliances increase trade under multipolarity, but
decrease it under bipolarity (or in many cases have no significant effect in the models).
Subsequent work also tends to confirm a positive relationship between alliances and
trade under varying conditions (e.g., Long 2003; Long and Leeds 2006).
Long (2008) is one of the few empirical pieces that includes explicit consideration of
rivalry, which is conceived of as a measure of the “shadow of conflict.” The argument is
that firms engaged in trade are risk-averse and look for indications that armed conflict
may disrupt their normal trading relationship (Li and Sacko 2002). If such indications
look likely, as in the case of a rivalry, then trade should be lower. Long’s empirical analy-
ses demonstrate a negative relationship between rivalry and trade.
Gowa and Mansfield (2004) is the only article in the literature on alliances and trade
that began to investigate the distinction between intra-industry and inter-industry trade,
supporting our belief that this is an important avenue of investigation. They note that
increasingly in the post–World War II era, trade between developed states has taken on
a different form than in the preceding century. As they note, “Ex ante, a firm has a wide
array of alternative investment and production opportunities available to it. Ex post, how-
ever, the firm locks itself into a bilateral monopoly whenever its investment is to some
degree ‘specific to an export destination’ ” (2004, 780). The sunk costs associated with
intra-industry trade make exporting firms especially vulnerable with regard to ex post
attempts to renegotiate, thus weakening their incentive to export. While firms them-
selves may create mechanisms to deter opportunism on the part of other states, alliances
between governments are argued to be especially desirable in this regard. Alliances reduce
the risks associated with sunk costs in export production, since the income of both the
exporting and importing states depends on the relationship. Government and firms have
an interest in maintaining stability and peace under conditions of intra-industry trade.
Gowa and Mansfield (2004) thus argue that alliances will promote trade overall, but that
due to this time inconsistency problem, they will foster intra-industry trade to an even
greater extent than inter-industry trade. They find statistical support for their argument.
Rather than focus exclusively on the effects of alliances, rivalry, and preference similar-
ity on trade, we suggest that the composition of trade may condition the likelihood of alli-
ance and rivalry formation as well as preference similarity between states. In essence, the
causal arrow may have been pointed in the wrong direction by scholarship heavily shaped
by the overlay of the Cold War. We draw inspiration from Rogowski (1987), who focuses
on the effects that trade may have on shaping democratic institutions within the state, to
consider how different forms of trade may shape political relationships between states. If
trade can generate changes within domestic political structures through the reshaping of
societal preferences, and if those preferences are also related to decisions about restric-
tion or expansion of trade, then we believe it stands to reason that decisions about the
characterization of states’ larger political relationships with each other are also at stake.
Preference similarity, and even formal alliances and rivalries, may be born of many expe-
dient factors, but underlying societal views about other states must also figure into this
 Timothy M. Peterson and Cameron G. Thies    191

decision. Those views are shaped by the cultural information conveyed by different forms
of trade. Trading relationships should therefore be an important structural consideration
for policy makers when forging larger political relationships between states. Those deci-
sions could then influence future trade between states, as an allied relationship casts a
very different light on intersocietal relations than a rival relationship does.
We have begun to test some of these relationships in our own work (Peterson and
Thies 2012b; Thies and Peterson forthcoming). Accordingly, we argue that intra-industry
trade promotes the emergence of similar foreign policy preferences among trade part-
ners. However, we contend that the reverse case does not follow: more similar states
do not necessarily engage in more intra-industry trade because restrictions on trade in
strategic commodities are lower against friendly states. Specifically, security-conscious
states might not wish to rely on foreign imports of fuels, minerals such as iron, or
food, which could then be used as leverage for coercion (e.g., Hirschman [1945] 1980).
However, when a state is confident that its trade partner will remain friendly, perhaps
due to naturally harmonious preferences or the presence of a common enemy, the
incentive to prevent such dependence could give way to the incentive to maximize gains
from trade. Simultaneous equations models and error correction models spanning 1962
to 2000 confirm that intra-industry trade promotes political similarity among trade
partners. We also find some evidence that intra-industry trade is associated with a lower
likelihood that states will become rivals while, simultaneously, rivalry is associated with
relatively more intra-industry trade.

Conclusion

If our arguments and evidence are correct, then future work on intra-industry trade and
political outcomes holds the possibility of overturning the predominant views held by
scholars of IPE about the role that trade plays in shaping global institutions and the bal-
ance of peace and conflict between states. The importance of trade and war, as well as the
relationship between these two processes, is well known to scholars of systemic leader-
ship, those in the interdependence and conflict literature, and those studying the forma-
tion of institutional arrangements. Yet we believe that often these scholars are relying
on outdated theoretical arguments about trade, as well as inappropriate measures of the
kind of trade that matters most for political outcomes. Our work suggests that the time
has come to completely rethink the relationship between trade and politics in our global
political economy, because while the composition of trade has changed dramatically
in the twenty-first century, our thinking remains rooted in nineteenth-century under-
standings of trade and its political effects.
The study of intra-industry trade is not without its conceptual and operational dif-
ficulties, as we have noted above. We expect that further refinements at the conceptual
level may be possible as considerations of trade composition are synthesized into theo-
ries of cooperation and conflict. Conceptual refinements will then lead to more accurate
192    Intra-Industry Trade and Policy Outcomes

operational measures dependent on the context of the study in question. We have sug-
gested a number of future directions for research in various substantive literatures,
including PTA formation, WTO dispute initiation, conflict outcomes, and preference
similarity. These literatures are enriched by the incorporation of intra-industry trade
concepts and measures. We believe that future work will build on this agenda to gener-
ate a more nuanced understanding of the relationship between trade and politics for the
twenty-first century.

Notes
1. In this case, each state holds an absolute as well as comparative advantage. However, the
same logic would hold if, for example, England were more efficient producing both textiles
and wine. It would still hold a comparative advantage in textiles if its relative efficiency
with respect to Portugal were greater for textiles than for wine.
2. Its classification as vertical is probable because price differentials would likely exceed
the 15  percent or 25  percent threshold used to distinguish vertical from horizontal
intra-industry trade.
3. However, the fact that production is internationalized could imply the existence of
intra-firm trade across states. This form of trade could also have implications for interna-
tional politics, although we leave examination of these effects to future researchers.
4. In addition to those data hosted by the UN, Feenstra et al. (2005) have made available
SITC4-level data.
5. Again, we except trade following from what appears to be vertical intra-industry special-
ization, but which actually follows from production sharing, from our conceptualization
and operationalization of intra-industry trade.
6. In Peterson and Thies (2012a), we used a version of this measure that adjusts for multilat-
eral trade imbalances as suggested by Aquino (1978) in order to avoid under-representing
intra-industry trade. For example, if, for example, a country has a severe trade deficit, then
even if all of its imports and exports were within a single industry (and hence the measure
should equal 1), the simple measure would report a smaller proportion of intra-industry
trade because exports minus imports cannot equal zero. However, all results are essen-
tially identical using the simpler Grubel and Lloyd (1975) measure (at the dyad-year level).
Furthermore, given criticism of the trade-balance weights (e.g., Greenaway and Milner
1981), we now prefer using the simpler measure.
Because total imports and exports are at the country-year level, we generally avoid
problems of divi3sion by zero, as there are no examples of years in which countries do not
import or export at all. However, to ensure that we do not lose observations at the com-
modity level, we add.01 to all commodity-level trade flows.
7. In theory, researchers could use the dyad-industry level of analysis. However, this would
require a large number of observations, potentially straining computer resources, particu-
larly if the analysis spans multiple years.
8. Economies of scale refer to advantages obtained due to increased size or scale of operation.
The cost per unit generally declines with increased scale as the fixed costs associated with
production are spread over more units. These increasing returns from economies of scale
are one of the mechanisms that generates the gains from intra-industry trade.
 Timothy M. Peterson and Cameron G. Thies    193

References
Aquino, Antonio. 1978. Intra-industry Trade and Inter-industry Specialization as Concurrent
Sources of International Trade in Manufactures. Review of World Economics 114 (2): 275–296.
Baccini, Leonardo. 2012. Democratization and Trade Policy:  An Empirical Analysis of
Developing Countries. European Journal of International Relations 18 (3):  455–479.
doi: 10.1177/1354066110391307
Balassa, Bela. 1961. Towards a Theory of Economic Integration. Kyklos 14 (1): 1-17.
Balassa, Bela. 1966. Tariff reductions and trade in manufactures among industrial countries.
American Economic Review 56: 466-73.
Baldwin, Richard. 1993. A Domino Theory of Regionalism. NBER working paper #4465.
Bhagwati, Jagdish. 1991. The World Trading System at Risk. Princeton, NJ:  Princeton
University Press.
Blanes, José V., and Carmela Martín. 2000. The Nature and Causes of Intra-industry
Trade: Back to the Comparative Advantage Explanation? The Case of Spain. Review of World
Economics 136: 423–441.
Chase, Kerry A. 2003. Economic Interests and Regional Trading Arrangements: The Case of
NAFTA. International Organization 57: 137–174.
Chase, Kerry A. 2005. Trading Blocs: States, Firms, and Regions in the World Economy. Studies
in International Political Economy, University of Michigan Press.
Elsig, Manfred, and Cedric Dupont. 2012. European Union Meets South Korea: Bureaucratic
Interests, Exporter Discrimination and the Negotiations of Trade Agreements. Journal of
Common Market Studies 50 (3): 492-507.
Falvey, R. E. 1981. Commercial Policy and Intra-industry Trade. Journal of International
Economics 11 (4): 495–511.
Feenstra, Robert, Robert E. Lipseym, Haiyan Deng, Alyson C. Ma, and Hengyong Mo. 2005.
World Trade Flows: 1962–2000. NBER Working Paper 11040. http://www.nber.org/papers/
w11040.
Finger, J. M. 1975. Trade overlap and intra-industry trade. Economic Inquiry 13: 581-589.
Fontagné, L., Freudenberg M. and N. Péridy. 1998. Intra-industry Trade and the Single
Market: Quality Matters. CEPR Discussion Paper, #1959.
Gilligan, Michael J. 1997. Lobbying as a Private Good with Intra-Industry Trade. International
Studies Quarterly 41 (3): 455–474.
Goldstein, Judith, and Lisa L. Martin. 2000. Legalization, Trade Liberalization, and Domestic
Politics: A Cautionary Note. International Organization 54: 603–632.
Gowa, Joanne. 1989. Bipolarity, Multipolarity, and Free Trade. American Political Science
Review 83 (4): 1245–1256.
Gowa, Joanne, and Edward Mansfield. 2004. Alliances, Imperfect Markets, and Major Power
Trade. International Organization 58 (4): 775–805.
Gowa, Joanne, and Edward Mansfield. 1993. Power Politics and International Trade. American
Political Science Review 87 (2): 408–420.
Greenaway, David, and Chris Milner. 1981. Trade Imbalance Effects in the Measurement of
Intra-Industry Trade. Weltwirtschaftliches Archiv 117 (4): 756–762.
Grimwade, Nigel. 1989. International Trade: New Patterns of Trade, Production, and Investment.
London and New York: Routledge.
Grossman, Gene M., and Elhanan Helpman. 2002. Interest Groups and Trade Policy. Princeton,
NJ: Princeton University Press.
194    Intra-Industry Trade and Policy Outcomes

Grossman, Gene M., and Elhanan Helpman. 1995. The Politics of Free-Trade Agreements.
American Economic Review 85: 667–690.
Grubel, H. G., and D. J. Lloyd. 1975. Intra-industry Trade, The Theory and Measurement of
International Trade in Differentiated Products. London: Macmillan.
Henisz, Witold J., and Edward D. Mansfield. 2006. Votes and Vetoes:  The Political Deter­
minants of Commercial Openness. International Studies Quarterly 50 (1): 189–212.
Hirschman, Albert O. [1945] 1980. National Power and the Structure of Foreign Trade.
Berkeley: University of California Press.
Hollyer, James, and Peter Rosendorff. 2011. Leadership Survival, Regime Type, Policy Uncer­
tainty and PTA Accession. Unpublished manuscript.
Intra-industry Trade: Cooperation and Conflict in the Global Political Economy
Kono, Daniel Y. 2009. Market Structure, Electoral Institutions, and Trade Policy. International
Studies Quarterly 53: 885–906.
Krugman, Paul R. 1979. Increasing Returns, Monopolistic Competition, and International
Trade. Journal of International Economics 9: 469–479.
Levy, Philip I. 1997. A Political-Economic Analysis of Free-Trade Agreements. American
Economic Review 87 (4): 506–519.
Long, Andrew G. 2008. Bilateral Trade in the Shadow of Armed Conflict. International Studies
Quarterly 52 (1): 81–101.
Long, Andrew G. 2003. Defense Pacts and International Trade. Journal of Peace Research 40
(5): 537–552.
Long, Andrew G., and Brett Ashley Leeds. 2006. Trading for Security: Military Alliances and
Economic Agreements. Journal of Peace Research 43 (4): 433–451.
Madeira, Mary Anne. 2013. The New Politics of the New Trade: The Political Economy of
Intra-Industry Trade. PhD diss., University of Washington.
Maggi, Giovanni, and Andres Rodriquez-Clare. 2007. A Political Economy Theory of Trade
Agreements. American Economic Review 97 (4): 1374–1406.
Mansfield, Edward D., Helen V. Milner, and Jon C. Pevehouse. 2008. Democracy, Veto Players
and the Depth of Regional Integration. World Economy 31 (1): 67–96.
Mansfield, Edward D., Helen V. Milner, and Jon C. Pevehouse. 2007. Vetoing Co-operation: The
Impact of Veto Players on Preferential Trading Arrangements. British Journal of Political
Science 37: 403–432.
Mansfield, Edward D., Helen V. Milner, and Peter Rosendorff. 2002. Why Democracies
Cooperate More:  Electoral Control and International Trade Agreements. International
Organization 56: 477–514.
McDonald, Patrick J. 2004. Peace through Trade or Free Trade? Journal of Conflict Resolution
48: 547–572.
Melitz, Marc J. 2003. The Impact of Trade on Intra-Industry Reallocations and Aggregate
Industry Productivity. Econometrica 71: 1695–1725.
Melitz, Marc J., and Giancarlo I. Ottaviano. 2008. Market Size, Trade, and Productivity. Review
of Economic Studies 75: 295–316.
Morrow, James D., Randolph M. Siverson, and Tressa E. Tabares. 1998. The Political
Determinants of International Trade:  The Major Powers, 1907–90. American Political
Science Review 92 (3): 649–661.
Ohlin, Bertil. 1933. Interregional and International Trade (Cambridge: Harvard University.
 Timothy M. Peterson and Cameron G. Thies    195

Peterson, Timothy M., and Cameron G. Thies. 2012a. Beyond Ricardo: The Link between Intra-
Industry Trade and Peace. British Journal of Political Science 42 (04): 747–767. http://dx.doi.
org/10.1017/S0007123412000129.
Peterson, Timothy M., and Cameron G. Thies. 2011. Intra-industry Trade, Veto Players, and the
Formation of Preferential Trade Agreements. Paper presented at the International Political
Economy Society, University of Wisconsin, Madison, WI, November 11-12.
Peterson, Timothy M., and Cameron G. Thies. 2012b. Trade and the Variety of International
Relationships:  How Intra- and Inter-Industry Trade Conditions Alliances and Rivalries.
Paper presented at the American Political Science Association, New Orleans, LA, August
30–September 2.
Reinhardt, Eric. 1999. Aggressive Multilateralism: The Determinants of GATT/WTO Dispute
Initiation, 1948–1998. Paper presented at International Studies Association Meetings,
Washington, DC.
Ricardo, David. [1817] 1981. The Principles of Political Economy and Taxation. Cambridge:
Cambridge University Press.
Rogowski, Ronald. 1987. Trade and the Variety of Democratic Institutions. International
Organization 41 (2): 203–223.
Rosendorff, B. Peter. 2005. Stability and Rigidity: Politics and Design of the WTO’s Dispute
Settlement Procedure. American Political Science Review 99 (3): 389–400.
Stolper, Wolfgang, and Paul Samuelson. 1941. Protection and Real Wages. Review of Economic
Studies 9 (1): 58-73.
Thies, Cameron G., and Timothy M. Peterson. Forthcoming. Intra-industry Trade: Cooperation
and Conflict in the Global Political Economy. Palo Alto, CA: Stanford University Press.
Viner, Jacob. 1950. The Customs Union Issue. New York: Carnegie Endowment for International
Peace.
 Chapter 11

Het ero geneou s Fi rms a nd

P olicy Prefe re nc e s

Mi chae l Pl ouf f e

The redistributional effects of international trade create clear winners and losers,
and a large body of research in international political economy has been devoted to
identifying them and the ways in which they might impact potential trade-policy out-
comes. Scholarship on the sources of demand for trade policy has primarily fallen into
two broad categories. One body of work concentrates on aggregated groups of actors,
whether industries or owners of various factors of production. The second research
program examines characteristics of voters at the individual level to ascertain the
sources of preferences regarding trade policy.1 However, a third strand of research, rely-
ing on recent advances in modern trade theory, has brought about an increased focus
on firms as key actors in the formation of trade policy. Like factor owners, industries,
and individuals, firms may be positively or negatively impacted by a change in trade
policy, and these potentially significant shifts of fortune lead to divergences in policy
preferences.
Trade engagement within industries is exceedingly dependent on firm-level charac-
teristics. Highly productive firms can afford to overcome the significant costs associ-
ated with lucrative foreign trading, while low-productivity firms cannot. In the face of
liberalization, high-productivity producers favor reducing these costs or gain access to
foreign markets, while low-productivity firms seek protection from relatively cheap for-
eign imports. Helen Milner’s work on the efforts of multinational corporations (MNCs)
to further trade liberalization was very prescient in this regard.2 However, MNCs are not
alone among producers seeking to reduce trade barriers. In the global economy, a small
but significant portion of producers engages foreign markets through trade or invest-
ment, while many more firms benefit from these links indirectly.
In this chapter I briefly overview some of the influential international political econ-
omy research on firms and trade policy that predates the firm heterogeneity revolu-
tion. I then outline the impact of firm heterogeneity on the economics of trade and the
 Michael Plouffe   197

supporting empirical findings. I extend the model to its political implications and dis-
cuss the importance of firms’ policy preferences.

Producers and Trade Politics

The study of the role of producers in trade politics has a rich history that long predates
the firm heterogeneity revolution in economics. Bauer, Pool, and Dexter3 provide an
early effort to open up the black box of decision-making processes within the firm,
focusing on trade-policy engagement. A number of their findings have become com-
mon starting points for subsequent research on trade-policy preferences. While the
individual decision-makers at any firm may hold a number of distinct sources of pref-
erences over trade policy stemming from sociological or economic concerns,4 when
they need to request policy-based support for their firms, these people frequently make
requests in response to specific problems that go against their own personal trade-policy
preferences. Different sources of preferences may be activated through different sorts of
stimuli (framing), but the stability of firm-based preferences has formed a starting point
for a significant and growing body of research.
Perhaps the seminal approach to the firm in trade politics is found in Helen Milner’s5
work, which disaggregates the trade-policy interests of MNCs and exporters from those
of domestic producers. Exporting firms and MNCs depend on access to markets abroad,
leading to policy positions that diverge from those of other producers. While domestic
producers would seek protection from foreign-produced goods, MNCs seek access to
these markets. This leads to a clear delineation of producers based on their activities.
While the predictions of this model match nicely with some of those made by current
firm heterogeneity models, there are some key differences, detailed later in this chapter.
Woll (2008) extends Milner’s argument to address MNC-government relations in ser-
vice industries.
Whereas Milner’s work anticipated the preference heterogeneity among producers
that is a consequence of firm heterogeneity, the dominant approach linking produc-
ers to trade politics, the “protection for sale” paradigm,6 ignores this, largely relying
on a Ricardo-Viner industry-based framework with homogenous industry-level pref-
erences. In the usual formulation of models in this vein, industries organize to lobby
politicians through campaign contributions for import protection. Subsequent
advancements have introduced producer heterogeneity, but retained industry-based
preferences.7 While the incorporation of heterogeneous firms has contributed to a
model that better explains observed patterns of political engagement, the directionality
of political activities has largely remained constant: firms are assumed to pursue protec-
tionist policies. Recent extensions to these models have incorporated aspects of prefer-
ence heterogeneity, but they have remained largely overlooked, perhaps because of the
increased demands they require of data.8
198    Heterogeneous Firms and Policy Preferences

Revolution in the Economics of Trade

The trade models currently utilized by international political economists are largely based
on variants of the Hecksher-Ohlin model, which focused on an economy’s endowments
in factors of production as a way to explain patterns in cross-national trade flows. These
models were replaced in the economics profession in the 1980s by new trade theory,9
which took advantage of the relatively late formalization of monopolistic competi-
tion by Avinash Dixit and Joseph Stiglitz (1977) to create more parsimonious models,
incorporating increasing returns to scale and product differentiation. The new trade
theory allowed for improved understanding of trade flows, providing an accounting of
intra-industry trade as well as inter-industry trade; however, its shortcomings became
apparent with the advent of new firm-level data sets in the early 1990s. Despite its ability
to describe and predict macro-level patterns of trade, new trade theory could not explain
patterns in the internationalization of production (such as intra-firm trade), nor could it
link firm-level characteristics to patterns of international economic engagement.
In response to these shortcomings, a growing literature has focused on firm hetero-
geneity in what has been inelegantly referred to as “new new trade theory.” Theoretical
frameworks to explain these behaviors arose in models by Melitz10 and Bernard, Eaton,
Jensen, and Kortum,11 with the Melitz model becoming a workhorse for many recent
innovations in trade theory.12 This new generation of trade models accounts for the
incorporation of intra-firm trade and also reflects the reality of the rarity of trade engage-
ment among firms. In addition, because trade varies significantly across industries,
theories have been developed to integrate the Melitz model into the Heckscher-Ohlin
framework,13 to account for variations in product differentiation across industries.14
These theoretical innovations have provided a solid framework for elucidating patterns
revealed in the growing body of empirical work on trade.

The Empirics of Heterogeneous Firms

As more sources of firm-level data have become available, a number of empirical regu-
larities have emerged and become accepted as stylized facts. Exporters differ from firms
that only serve the domestic market. Compared to nonexporters, they are more pro-
ductive, larger in terms of both shipments and employment, are more skill and capital
intensive, and pay higher wages.15 Even prior to entry into the export market, prospec-
tive exporters are larger and more productive than other producers. This indicates that
high-productivity firms select into exporting; less-productive firms cannot afford to pay
the high fixed (and sunk) costs of export-market entry.16
Given that exporters are both large and highly productive, one might expect
the productivity between them and nonexporters to be merely evidence of exploi-
tation of economies of scale. Controlling for firm size, Bernard et al. (2007, 5) show
 Michael Plouffe   199

that American exporters remain significantly more productive than nonexporters,

so exporters possess some intrinsic characteristics that enhance their productivity, in
addition to their economies of scale. This combination of inputs is described as total
factor productivity (TFP). Although not all firms with high levels of TFP engage in
trade, all traders have relatively high levels of TFP. This relationship is characteristic of
firms worldwide.17
The relative skill- and capital-intensive nature of exporters holds across economies at
all levels of development, in contrast with expectations based on a standard factor-based
trade model. In a developed (capital or skilled-labor abundant) economy, this is to be
expected, but in a developing economy (land or unskilled-labor abundant), comparative
advantage would seem to dictate that exporters would utilize production processes that
intensively use the abundant factors of production (land or unskilled labor). However,
this is not the case; exporters remain relatively skilled-labor or capital intensive, even
within comparative advantage industries.18 Thus, both sector- and factor-based models
face significant difficulties in explaining these behaviors.

The Melitz Model and Other Approaches

to Heterogeneous Firms
The Melitz model (first presented in Melitz 2003) is the starting point for most theo-
retical approaches to trade that incorporate heterogeneous firms. Like Krugman
(1980), who introduced product differentiation to a single industry, Melitz models
intra-industry dynamics in a monopolistic competition framework. Unlike previous
models, Melitz’s model accounts for empirically observed asymmetries in firm pro-
ductivity and size, rather than dismissing them by assumption. These dissimilarities in
firms’ characteristics drive variations in their market behaviors.
The Melitz model introduced “new new trade theory,” which deals explicitly with
these intra-industry features by incorporating heterogeneous firms into a trade model
based on monopolistic competition. According to Melitz’s model and others like it, firm
heterogeneity occurs because, prior to entering the domestic production market, firms
face uncertainty over both their immediate and future levels of productivity. Immediate
productivity is only revealed if the firm chooses to pay the fixed (then sunk) cost of entry
to enter the market.
At this point, if productivity is revealed to be at the low end of the distribution, a firm
may never realize any profits and will choose to exit without producing anything, in
order to avoid incurring further costs (this point is often referred to as the zero-profit
productivity cutoff). Where production is profitable, firms do not exit the market, and
they incur variable costs of production. In the face of increased exposure to trade, an
industry becomes more productive on average. This happens as inter-firm reallocations
lead to the most productive firms gaining in market share and factors of production,
while the least productive firms exit.
200    Heterogeneous Firms and Policy Preferences

Firm heterogeneity also provides an explanation for variations in firms’ export behav-
iors. Engaging a foreign market requires payment of a high fixed cost of entry, much
like the initial domestic market entry decision, and the variable costs associated with
producing goods sold abroad are higher (and in iceberg form—more than one unit must
be exported for one unit to be sold abroad) than for those sold domestically. Similarly,
export sales are lucrative relative to domestic sales. Consequently, only an industry’s
most productive firms can afford to enter export markets, leaving a large portion of pro-
ducers engaging the domestic market alone.

Firm Heterogeneity Beyond Exporting

While much of the firm heterogeneity literature has focused on exporting, produc-
tivity also segments other forms of internationalization by firms. For example, many
exporting firms also engage in importing, sourcing intermediate goods from abroad.19
Importers share many characteristics with exporters. They are rare, they are more pro-
ductive, and they are larger; in addition, they spend more on intermediate goods than
do nonimporters.20 They also employ more workers and pay them higher wages and are
more skill- and capital-intensive than nontraders.21 Firms that only import, while more
productive than nonimporters, are slower growing and less innovative than two-way
traders and firms that only export.22 This could be due to the fact that importers, when
grouped by behaviors, are a relatively diverse lot:  some firms import intermediate
inputs, while others import final products for domestic distribution. Even the economic
implications of these activities are understudied, possibly due to a lack of availability of
sufficiently disaggregated and detailed data.
Similarly, firms that engage directly in trade can be differentiated from those that
trade through intermediaries, employing the distribution networks of other firms to
either source inputs or final products abroad or sell portions of their own production in
foreign markets. This rapidly growing body of research focuses on explaining variation
in firms’ export behaviors,23 and much of it deals with characteristics of the intermediar-
ies themselves, rather than the firms employing them. The use of an intermediary firm is
driven primarily by country-specific fixed costs, meaning a firm that directly exports to
one country may use an intermediary to reach another market.24 Intermediaries them-
selves tend to be smaller than direct traders in terms of employment, but this differential
decreases when total sales are compared.
An early extension of the Melitz model incorporated the firm-level decision to engage
in foreign direct investment (FDI). A very small subset of the most productive firms
may choose to invest directly in a foreign market to avoid transportation costs asso-
ciated with exporting finished goods (horizontal FDI); similar to exporting, FDI is
lucrative compared to domestic sales. However, compared to export sales, FDI entails
higher sunk costs, but potentially higher profits over time.25 Relative to exporters, firms
investing in plants abroad are more productive. Similarly, firms that outsource portions
of their production processes overseas are more productive than their counterparts.26
 Michael Plouffe   201

While theoretical approaches to these behaviors have explored the boundaries of the
firm, they have not yet incorporated trading activities. Ultimately, firms have a large
menu for accessing international markets; heterogeneity in their characteristics, espe-
cially productivity, leads to significant diversity in their actual engagement activities.
These patterns of engagement can provide valuable insight for political economists
attempting to understand and explain firms’ political positions and behaviors in a global
economy.
In addition, a number of researchers have extended the Melitz model to study the
implications of heterogeneous firms for other economic phenomena. For example,
Bernard, Redding, and Schott (2007) follow Helpman and Krugman (1985) in work-
ing new theoretical innovations into the existing Heckscher-Ohlin framework.
Nesting the Melitz model into a two-sector integrated equilibrium structure allows
for the examination of cross-industry adjustments to trade, in addition to the intra-
industry adjustments revealed in the Melitz model. While postliberalization asset and
market-share reallocations benefit high-productivity firms in both industries, they
also benefit the comparative advantage industry over the comparative disadvantage
industry. In short, the comparative-advantage industry grows, while its counterpart
shrinks.

The New Political Economy

of Trade: Incorporating
Firm Heterogeneity

Plouffe (2011b)27 was the first to extend a model of heterogeneous firms to include their
political interests and behaviors. As in other heterogeneous firm models, producers
within an industry are differentiated by their levels of TFP, which determines firms’ abil-
ity to produce, as well as their ability to engage foreign markets through trade, invest-
ment, or other means. Policy preferences then follow from these economic abilities.
The production market in this model follows that of Melitz. The behaviors firms
engage in can be stylized as occurring sequentially within a period of time. These peri-
ods can then be treated as repeating infinitely, with industry structure evolving dynami-
cally over time. Industry dynamics reflect adjustments to trade and trade-policy reform;
business cycles are omitted from the model.
Firms are differentiated by TFP, but a firm’s TFP draw is unobserved prior to domestic
market entry. Payment of the fixed (and then sunk) cost of domestic market entry allows
new firms to observe their initial productivity draws. If a firm’s TFP draw is sufficiently
high for it to produce profitably, meaning it will have relatively low variable costs of pro-
duction, it will begin to produce for the domestic market; if the TFP draw would lead to
unprofitable production, the firm will exit the market without producing. For firms that
entered the market in a previous period, the analogous decision that must be made is
202    Heterogeneous Firms and Policy Preferences

whether to continue operating (which does not require a repayment of the sunk costs of
entry) or to exit the market. Firms that incurred losses in the previous period will exit,
whereas those that either broke even or realized a profit will choose to continue.
After firms make the market entry/continuation decision, they choose whether to
export a portion of their production or pursue political action (these actions are not
mutually exclusive). Decisions about exporting tend to be linked with certain invest-
ment and research decisions28 and to be implemented in conjunction with each other.
The same can be expected for political decisions, which are discussed in the following
section.
Firms pursue export sales, as they are more profitable than domestic-market sales;
however, export-market entry is limited to firms with sufficiently high productivity to
overcome the significant fixed costs of entry and higher variable costs—this point is
referred to as the exporting cutoff. Likewise, other forms of internationalization—FDI,
offshore outsourcing, and importing—can be lucrative activities, provided firms can
overcome the significant fixed costs of entry for these activities.
In general, the claim can be made that firms in the upper portion of the productivity
distribution—those with TFP such that engaging foreign markets is a potentially profit-
able prospect—can expect to gain from trade liberalization. It follows that these firms,
as well as the factors of production they employ, both prefer and seek trade liberaliza-
tion, so they will have incentives to enter the political market to gain access to foreign
markets and goods. However, less-productive firms cannot surmount the high costs of
export-market entry. Assuming reciprocity, they face reduced margins from liberalized
trade, with the least productive firms forced out of the market. While these unproduc-
tive firms may favor protection, they lack the necessary resources to enter the political
market.
Finally, firms produce, incurring variable costs and realizing profits or losses based on
their TFP draw and production decision. At this point, the model transitions to the next
period, repeating these steps with a new mass of firms.

Policy Preferences with Heterogeneous Firms

Perhaps unsurprisingly, models of the political economy of trade with heterogeneous
firms point unequivocally to intra-industry cleavages over trade policy. In most cases,
highly productive firms favor increased access to foreign markets through liberaliza-
tion, while unproductive producers favor protection from competition that may be
intensified by heightened exposure to imports.
Likewise, the configuration of preferences matters for the incorporation of trade-
policy interests into policy. High-productivity firms are likely to pursue their inter-
ests through lobbying activities, while their unproductive counterparts do not. Plouffe
accounts for this disparity in activities by focusing on the high costs of political engage-
ment. As in other markets, highly productive firms benefit from participation, while
unproductive firms cannot afford entry.
 Michael Plouffe   203

In this model, the political market resembles a miniature version of the produc-
tion market. Evidence suggests that the market for lobbying is structured in the same
manner as a production market: firms face a significant fixed (and sunk) cost of entry
and incur relatively low variable costs of continuation, should they choose to remain
politically active.29 Regardless of the manner of political engagement, entry costs may
be associated with gaining an understanding with relevant political institutions, actors,
policies, and regulations, or may be linked directly to searching for a good lobbyist and
gaining access to key players. Costs of continuation may not be insignificant, but famil-
iarity with the components of the political system should make them relatively low rela-
tive to the fixed costs.
Firms that choose to engage in political activities—or reveal their preferences—are
assumed to make truthful contributions, meaning they seek the trade-policy outcome
that enhances their business activities. Thus, highly productive firms that lobby are
likely to pursue liberalization, while unproductive firms, if they can afford to enter the
political market, will seek protection.30
Both Kim (2013) and Osgood (2013) frame these observed patterns as the result of
the level of product differentiation within an industry. In Kim’s model, the protection
for sale logic is flipped on its head: in highly differentiated industries, exporting inter-
ests are concentrated among a small number of firms that seek profit-improving export
sales, while potentially protectionist interests are diffuse. Those pursuing liberaliza-
tion can more easily overcome any collective-action problem over trade policy, while
import-competing firms are relatively unthreatened by imported varieties.
For within-industry preferences, Osgood largely ignores collective action, instead
focusing on the theoretical differences between a shift from autarky to liberalization
and from postliberalization to increased liberalization. In the first instance, firms
are cleanly divided: exporters favor liberalization, while domestic-focused produc-
ers prefer autarky. In the second, a shift from a postliberalization environment to
one with even freer trade, the producers that would favor further liberalization are
exporters of middling productivity, as these firms stand to gain the most in terms
of the percentage gain in their exporting profits. Consequently, the most productive
firms, like those solely competing with imports, oppose liberalization, as the poten-
tial reductions to their costs would be countered by increased competition, both at
home and abroad.
Plouffe (2012) similarly omits collective action, arguing that highly productive
firms—both exporters and nonexporters—are more likely to favor liberalization than
unproductive firms. However, as productivity increases above the exporting cutoff,
trade liberalization is less likely to be a salient issue, as the marginal gains for a firm with
extremely low variable production costs are smaller than for one with relatively high
variable costs. Thus, the firms that are most likely to hold strong liberalizing preferences
are those with TFP draws nearest to, but greater than, the exporting cutoff.
Finally, these preference expectations are shaped by status quo barriers to trade. High
tariffs and nontariff barriers should lead to a relatively large mass of high-productivity
firms seeking liberalization; consequently, low-productivity firms will be less likely to
204    Heterogeneous Firms and Policy Preferences

seek increased protection. Likewise, relatively low tariffs and nontariff barriers should
be associated with fewer high-productivity firms pursuing further liberalization and a
greater portion of low-productivity firms seeking protection.

Evidence of Heterogeneous Preferences

The evaluation of firms’ heterogeneous policy preferences is often a data-intensive task.
While the firm-level financial data that provide the inputs for TFP estimation are gen-
erally readily available, international trade and investment data are less rigorously col-
lected and distributed at the firm level. Cross-nationally, the availability of information
on firms’ political activities varies widely. Under the Lobbying Disclosure Act of 1995,
the United States requires lobbyists to report data on clients, expenditures, and issues.
Other countries do not have similar transparency requirements.
Following a number of broad-based studies of business lobbying,31 some efforts have
been made to assess the variations in trade-policy lobbying activities among American
firms.32 Both Plouffe (2012) and Kim (2013) find a positive relationship between produc-
tivity and the likelihood of lobbying. In the same manner as market activities, lobbying
is a costly endeavor. The high fixed costs of lobby-market entry and variable continu-
ation costs preclude participation by less-productive producers, while highly produc-
tive producers may not seek to incur lobbying’s additional costs. Lobbying among the
most productive firms is more likely when products are highly differentiated or highly
substitutable, while lobbying is less likely among firms in industries with middling lev-
els of product differentiation.33 These highly productive firms, across all industries, are
very likely to be lobbying for liberalization, relative to less-productive firms, and this
productivity-lobbying relationship persists across both exporting firms and domestic
producers.34
While lobbying-based studies capture producers’ political activities or revealed pref-
erences, they do not capture the variation of latent preferences along the productivity
distribution of firms. Plouffe (2011b) uses a survey-based design to capture latent pref-
erences among Japanese manufacturers, finding that highly productive manufacturers
are more likely to favor liberalization than are less-productive firms. Furthermore, the
relationship between productivity and pro-trade preferences persists when controlling
for actual trading behaviors.
The data-intensive nature of firm-level empirical studies makes it difficult to explore
the historical nature of these theories using the same approach that can be applied to
contemporary examinations, but a surprisingly rich wealth of data exists for use in his-
torical case studies. For example, Plouffe (2013) examines the impact of productivity
heterogeneity on producer cleavages over the Smoot-Hawley Tariff of 1930, relying on
firm size (a common proxy for TFP among economists), geographically induced differ-
ences in TFP distributions, and mechanization rates to find evidence of intra-industry
cleavages over the Tariff. Similar studies could utilize these techniques to reinvestigate
 Michael Plouffe   205

other watershed events, such as the formation of the Coalition of Iron and Rye or the
coalitions behind the passage of the Reciprocal Trade Agreements Act of 1934.

Comparative Advantage and Heterogeneous Firms

In addition to predicting variations in trade-policy preferences within industries, het-
erogeneous firm theories can explain differences in how preferences are distributed
across industries. In keeping with a rich literature emphasizing the impact of factor
endowments on trade flows,35 Plouffe (2011a, 2012) identifies variations in political
interests between comparative-advantage and comparative-disadvantage industries.
Because production in comparative-advantage industries is relatively cheap when com-
pared to same-industry production abroad, a relatively large mass of firms can expect
to benefit from trade liberalization. Likewise, because production is relatively dear in
comparative-disadvantage industries, a fairly small portion of highly productive pro-
ducers gains from liberalization.
Following a liberalizing reform to trade policy, both output and the employ-
ment of the factors of production shift from comparative-disadvantage industries to
comparative-advantage industries. At the same time, both of these shares also shift from
unproductive firms, especially those that are forced out of the production market, to
highly productive producers. The shifts in the employment of factors of production lead
to an adjustment in the composition of their demand. Recall that across all industries
(and at all levels of economic development), highly productive firms are more skill- and
capital-intensive than their counterparts. Thus, following liberalization, the prices of
these factors increase as market share and means of production are reallocated to these
high-productivity firms.
However, while the aggregate postliberalization trends follow the Ricardian pre-
dictions of production specialization according to comparative advantage, reliance
on a classical trade model generates the overprediction of specialization, as well as
painting an inaccurate portrait of the micro-level implications of liberalization.
Incorporating heterogeneous firms into a comparative-advantage framework gener-
ates predictions for developing economies that markedly contradict the conventional
wisdom. Factor owners employed by highly productive firms gain from trade liberal-
ization, while those employed by unproductive firms lose. Similarly, across different
levels of development, skill and capital owners face better postliberalization prospects
than labor owners. As market share is reallocated to the firms that benefit from trade,
demand for capital and skills increases, drawing from the pool of these factors freed
from the unproductive firms forced to exit the market. Because the producers that
gain from trade do not require labor as intensely as their counterparts who ceased
production, labor loses from liberalization. The reduction in demand may lead to
decreases in nominal returns, even as the growth in trade brings about welfare growth
in real terms.
206    Heterogeneous Firms and Policy Preferences

Ultimately, while factor abundance in classical models leads to variation in the distri-
butional consequences from liberalization at different levels of economic development
in classical models, this is not the case in models with heterogeneous firms. Rather,
across all levels of development, owners of skills and capital are predicted to benefit the
most from trade, while owners of unskilled labor face reduced demand for employment.

Product Differentiation Across Industries

Kim (2013) and Osgood (2013) focus on intra-industry product differentiation as a
means of distinguishing industry-level variations.36 High levels of product differen-
tiation reduce the impact of import liberalization on the margins of existing products
within an industry relative to the impact in an industry with low levels of product differ-
entiation, as individual products are less readily substitutable. Thus, in industries with
low levels of product differentiation, where even a small shift in competitors’ prices can
have large consequences for competitiveness, even highly productive firms may favor
protection, seeking to maintain their domestic margins.37
While neither model explicitly incorporates the dynamics of adjustment to a liber-
alizing reform, the adjustment dynamics would likely resemble those discussed in the
previous section. In a skill- or capital-abundant economy, output and factor employ-
ment shift to both highly productive firms and industries with high levels of product
differentiation, and away from unproductive firms and industries with relatively undif-
ferentiated products. In an economy abundant in unskilled labor, employment and out-
put would shift to highly productive firms in industries with low levels of differentiated
products, with these producers competing in foreign markets based primarily on price
advantages. As an economy develops, it experiences increasing product differentiation
in skill- and capital-intensive sectors.
Regardless of whether industries are divided based on product differentiation of
comparative-advantage lines, models of trade policy with heterogeneous firms gener-
ate political cleavages that differ from those of models based on the Ricardo-Viner and
Stolper-Samuelson frameworks. Rather than political cleavages forming along factor or
sector lines, they are dictated by firm-level productivity.

The Policy Implications of Firm Heterogeneity

Just as the distributional and political predictions of heterogeneous firm models dif-
fer from those of classical models, so do the policy implications deviate from the con-
ventional wisdom. Where the workhorse theories in international political economy
predict political victories as accrued by those seeking protection, heterogeneous firm
models primarily predict the opposite outcome: liberalization results from the political
activities of pro-trade firms. This outcome is the consequence of two features of political
 Michael Plouffe   207

engagement by producers. First, only highly productive firms possess the capacity to
engage in political behavior, as such activities divert resources from production. These
businesses are more likely than others to favor liberalization, biasing policy outcomes
toward liberalization rather than protection. Second, because the productivity distribu-
tion of firms within a given industry is typically shaped such that the majority firms are
relatively unproductive, with a small highly productive portion, collective action can be
expected to be less problematic for these politically active firms than for the many firms
that do not stand to benefit from liberalization. Thus, in contrast with the ubiquitous
protection for sale paradigm, we should generally expect to see political activities that
incentivize policy makers to implement liberalizing reforms.
However, there are important theoretical and empirical exceptions, arising especially
from Osgood’s prediction of highly productive firms opposing further liberalization in a
postliberalization environment. Here it is instructive to differentiate between tariffs and
nontariff barriers (NTBs). The highest productivity firms never favor increased tariffs.
However, they do prefer increases in NTBs, as long as trade can be maintained. A uni-
lateral increase in NTBs shifts the distribution of export profits away from the least pro-
ductive exporters to the most productive exporters.38 This effect is discussed in greater
detail in the following section.

Firms and Nontariff Measures

Nontariff measures/barriers (NTM/Bs) include both behind-the-border and at-the-
border restrictions on trade and form an important, if difficult to study, component of
trade policy. The extent to which these regulations supplement or complement tariffs is
beyond the scope of this chapter, but characteristics of the demand for them runs some-
what counter to the conventional wisdom that MNCs promote liberalization because
of their heavy investment in foreign markets.39 Under certain conditions in Osgood’s
model, the most productive firms in an industry seek the implementation of protection
through NTBs.
Where tariffs impact a trading firm’s variable costs on top of its productivity draw,
NTBs increase the fixed cost of market entry. This serves to restrict market access to
potential entrants, preventing the less productive ones from engaging the market,
regardless of the intent behind the NTB. Thus, multinationals might support the imple-
mentation of NTBs to restrict the foreign market expansion of their competitors, reduc-
ing competition in those markets. Gulotty (2013) illustrates this using the case of a 2007
reform of European Union chemical legislation. The reform, which required the regis-
tration and testing of chemical products, was initially strongly opposed by American
industry members, but both Dow Chemical and BASF vocally supported its implemen-
tation. Further investigation provides evidence that highly productive MNCs support
the implementation of NTBs, providing an explanation for the persistence and progres-
sion of NTBs while tariffs have steadily decreased.
208    Heterogeneous Firms and Policy Preferences

Looking Forward

The body of research incorporating heterogeneous firms into the study of the interna-
tional political economy of trade is very much in its infancy, but these theories have the
potential to induce a sea change in the study of the subject. The focus of much of the
literature on factors and industries lacks both solid micro-foundations and mechanisms
for translating preferences into policy. Heterogeneous firm approaches provide both.
The implications of these theories range beyond the topic of trade politics. Similar
productivity-based breakdowns can be identified for other foreign economic behaviors,
like FDI, importing, behaving as an intermediating trade, or engaging in intermediated
trade.40 For each of these activities, despite differing entry cutoffs, highly productive
firms can expect to benefit from participation, while unproductive firms cannot afford
to participate. These activities are not mutually exclusive, so a single producer may
engage in some combination of any or all of them (in addition to exporting) to reach dif-
ferent target markets.
Similarly, productivity impacts firm-level preferences over tasks trade. The most pro-
ductive firms engaging in tasks trade offshore various functions, while less productive
task traders will onshore, outsourcing production to another domestic firm.41 Like other
markets, participation entails a fixed entry cost, precluding the participation of the least
productive. Firms that can obtain benefits from reducing the costs to offshoring will
favor this, as it resembles a factor-specific increase in productivity.42 Firms that can-
not afford to offshore face two effects: the factor that is offshored becomes cheaper as a
result of its increased productivity, while competition from offshoring firms potentially
becomes more intense as the unit price of their output is adjusted for their new variable
costs. Firms that cannot offshore are likely to oppose the liberalization of offshoring,
as the availability of the cheaper factor is likely to lag behind the competition effects,
although this depends on the extent of product differentiation within the industry.
Although the effects of offshoring break down in a similar manner to those of trade,
the effect of increasing the dispersion of an industry’s TFP distribution leads to politi-
cal battles that are often highly pitched relative to those over trade policy, as shifts in
employment across international borders exacerbate domestic tensions between pro-
ducers based on offshoring capabilities as well as among factor owners fearing their own
future employment prospects.
In addition to improving political economy understandings of economic flows, het-
erogeneous firm studies can enhance research in other related fields. For example, while
much progress has been made in determining the factors that lead firms to lobby on
trade policy in the United States, little work has been done to expand these findings
to other political activities and institutional arrangements. Similarly, there is much
room for research exploring the configurations of producer interests within domestic
and international institutions. Producer associations vary significantly in membership,
organizational structure, and issue scope. Firm-focused studies of the nature of those
 Michael Plouffe   209

discussed in this chapter should provide a stepping stone for building an improved
producer-association literature through an increased appreciation of the variations in
members’ political interests.
While research on the implications of heterogeneous firms for the international polit-
ical economy of trade is still in its youth, it challenges the conventional wisdom of the
structure of demand for trade policy. Furthermore, firms are thrust into a central role in
the policy-making process, refocusing the emphasis on policy mechanisms from voting
to lobbying. These models provide explanations for the general postwar reductions in
barriers to trade, as well as the more recent turn to NTBs. By further assessing the varied
nature of producer interests, based on firm-level heterogeneity, political economists will
be able to build on existing understanding of the interactions between political interests
and policy makers to improve descriptions, explanations, and predictions of policy out-
comes and their consequences.

Notes
1. In economics, “preference” is typically used to refer to individual-level attitudes. However,
the convention in international political economy, followed here, aggregates the use of
“preference” to also refer to the collective position taken on an issue by an interest group.
2. Milner (1988); Milner and Yoffie (1989).
3. Bauer, Pool, and Dexter (1972).
4. Megumi Naoi discusses the sources of individual trade-policy preferences in Chapter Six.
5. Milner (1988); Milner and Yoffie (1989).
6. Grossman and Helpman (1994). Lucy Goodhart discusses this body of research in much
greater detail in Chapter 9.
7. Bombardini (2008).
8. Chang and Willmann (2006); Abel-Koch (2010).
9. Krugman (1980); Helpman and Krugman (1985).
10. Melitz (2003).
11. Bernard et al. (2003).
12. Redding (2011) provides an overview of the developments of these and related models.
13. Bernard, Redding, and Schott (2007).
14. Melitz and Ottaviano (2008).
15. Bernard et al. (2011).
16. Roberts and Tybout (1997); Bernard and Jensen (1999); De Loecker (2007).
17. Bernard et al. (2007); Helpman (2006); and Bernard et al. (2011) provide more in-depth
surveys.
18. Alvarez and Lopez (2005); Bernard, Redding, and Schott (2007).
19. Bernard et al. (2007); Helpman (2006).
20. Kasahara and Lapham (2013); Gibson and Graciano (2011).
21. Bernard et al. 2007.
22. Seker 2011; Plouffe 2011a.
23. Ahn et al. 2011; Antràs and Costinot 2011; Bernard et al. 2010.
24. Bernard et al. 2010.
25. Helpman, Melitz, and Yeaple (2004); Head and Ries (2003).
210    Heterogeneous Firms and Policy Preferences

26. Antràs 2003; Antràs and Helpman 2004.

27. Kim (2013) and Osgood (2013) provide alternative formal specifications.
28. See Costantini and Melitz (2008) for a look at linkages between investment and export
decisions.
29. Kerr, Lincoln, and Mishra (2011).
30. Osgood (2013) and Gulotty (2013) provide an important counterpoint to this logic, dis-
cussed later in this chapter.
31. See, for example, Hansen and Mitchell (2000); Hansen, Mitchell, and Drope (2005);
Richter, Samphantharak, and Timmons (2009).
32. Plouffe (2012); Kim (2013).
33. Kim (2013, 26).
34. Plouffe (2012).
35. Helpman and Krugman (1985); Bernard, Redding, and Schott (2007).
36. Kim (2013); Osgood (2013).
37. It is interesting to note that in developed economies, comparative-advantage sectors tend
to have high levels of product differentiation, whereas comparative-disadvantage sectors
tend to have low levels of product differentiation.
38. Osgood (2013, 25).
39. Milner (1988).
40. For more on intermediated trade, see Bernard, Grazzo, and Tomasi (2010) and Plouffe (2011a).
41. Antràs and Helpman (2004).
42. Grossman and Rossi-Hansberg (2008).

References
Abel-Koch, Jennifer. 2010. “Endogenous Trade Policy with Heterogeneous Firms.” Unpub­
lished manuscript.
Ahn, JaeBin, Amit K. Khandelwal, and Shang-Jin Wei. 2011. “The Role of Intermediaries in
Facilitating Trade.” Journal of International Economics 84 (1): 73–85.
Alvarez, Roberto and Ricardo A Lopez. 2005. “Exporting and Performance: Evidence from
Chilean Plants.” Canadian Journal of Economics 38 (4): 1384–1400.
Antràs, Pol. 2003. “Firms, Contracts, and Trade Structure.” National Bureau of Economic
Research, Working Paper No. w9740.
Antràs, Pol, and Elhanan Helpman. 2004. Global Sourcing. Journal of Political Economy
112: 552–580.
Antràs, Pol, and Arnaud Costinot. 2011 “Intermediated Trade.” The Quarterly Journal of
Economics 126 (3): 1319–1374.
Bauer, Raymond A., Ithiel de Sola Pool, and Lewis Anthony Dexter. 1972. American Business
and Public Policy: The Politics of Foreign Trade. Chicago: Aldine-Atherton, Inc.
Bernard, Andrew B., and J. Bradford Jensen. 1999. “Exceptional Exporter Performance: Cause,
Effect, or Both?” Journal of International Economics 47 (1): 1–25.
Bernard, Andrew B., Jonathan Eaton, J. Bradford Jensen, and Samuel S. Kortum. 2003. Plants
and Productivity in International Trade. American Economic Review 93 (4): 1268–1290.
Bernard, Andrew B., Marco Grazzi, and Chiara Tomasi. 2010. Intermediaries in International
Trade: Direct vs. Indirect Modes of Export. National Bank of Belgium Working Paper Series
no. 199.
 Michael Plouffe   211

Bernard, Andrew B., J. Bradford Jensen, Stephen J. Redding, and Peter K. Schott. 2011. The
Empirics of Firm Heterogeneity and International Trade. NBER Working Paper No.
w17627.
Bernard, Andrew B., J. Bradford Jensen, Stephen J. Redding, and Peter K. Schott. 2007. Firms in
International Trade. Journal of Economic Perspectives 21 (3): 105–130.
Bernard, Andrew B., Stephen J. Redding, and Peter K. Schott. 2007. Comparative Advantage
and Heterogeneous Firms. Review of Economic Studies 74: 31–66.
Bombardini, Matilde. 2008. “Firm Heterogeneity and Lobby Participation.” Journal of
International Economics 75 (2): 329–348.
Chang, Pao-Li and Gerald Willmann. 2006. “Protection for Sale with Heterogeneous Interests
within Industries.” Unpublished manuscript.
Costantini, James and Marc Melitz. 2008. “The Dynamics of Firm-Level Adjustment to Trade
Liberalization.” In The Organization of Firms in a Global Economy, edited by Elhanan
Helpman, Dalia Marin and Thierry Verdier. Cambridge: Harvard University Press. 107–141.
De Loecker, Jan. 2007. “Do Exports Generate Higher Productivity? Evidence from Slovenia.”
Journal of International Economics 73 (1): 69–98.
Dixit, Avinash K., and Joseph E. Stiglitz. 1977. “Monopolistic Competition and Optimum
Product Diversity.” The American Economic Review: 297–308.
Gibson, Mark J., and Tim A. Graciano. 2011. Trade Models with Heterogeneous Firms: What
about Importing? Unpublished manuscript.
Grossman, Gene M., and Elhanan Helpman. 1994. “Protection for Sale.” American Economic
Review 84 (4): 833–850.
Grossman, Gene M., and Esteban Rossi-Hansberg. 2008. Trading Tasks: A Simple Theory of
Offshoring. American Economic Review 98 (5): 1978–1997.
Gulotty, Robert. 2013. Regulatory Protectionism, MNC Profits, and the Limits of International
Cooperation. Unpublished manuscript.
Hansen, Wendy L., and Neil J. Mitchell. 2000. Disaggregating and Explaining Corporate
Political Activity:  Domestic and Foreign Corporations in National Politics. American
Political Science Review 94 (4): 891–903.
Hansen, Wendy L., Neil J. Mitchell, and Jeffrey M. Drope. 2005. The Logic of Private and
Collective Action. American Journal of Political Science 49 (1): 150–167.
Head, Keith and John Ries. 2003. “Heterogeneity and the FDI versus Export Decision of
Japanese Manufacturers.” Journal of Japanese and International Economies 17 (4): 448–467.
Helpman, Elhanan. 2006. Trade, FDI, and the Organization of Firms. Journal of Economic
Literature XLIV: 589–630.
Helpman, Elhanan, and Paul Krugman. 1985. Market Structure and Foreign Trade. Cambridge,
MA: The MIT Press.
Helpman, Elhanan, Marc Melitz, and Stephen Yeaple. 2004. Export Versus FDI with
Heterogeneous Firms. American Economic Review 94 (1): 300–316.
Kasahara, Hiroyuki, and Beverly Lapham. 2013. “Productivity and the Decision to Import and
Export: Theory and Evidence.” Journal of International Economics 89 (2): 297–316.
Kerr, William R., William F. Lincoln, and Prachi Mishra. 2011. The Dynamics of Firm Lobbying.
NBER Working Paper No. 17577.
Kim, In Song. 2013. Political Cleavages within Industry:  Firm-level Lobbying for Trade
Liberalization. Unpublished manuscript, Princeton University.
Krugman, Paul. 1980. Scale Economies, Product Differentiation, and the Pattern of Trade.
American Economic Review 70: 950–959.
212    Heterogeneous Firms and Policy Preferences

Melitz, Marc. 2003. The Impact of Trade on Intra-Industry Reallocations and Aggregate
Industry Productivity. Econometrica 71 (6): 1695–1725.
Melitz, M. J., and G. I. P. Ottaviano. 2008. Market Size, Trade, and Productivity. Review of
Economic Studies 75 (1): 295–316.
Milner, Helen. 1988. Resisting Protectionism: Global Industries and the Politics of International
Trade. Princeton, NJ: Princeton University Press.
Milner, Helen V., and David B Yoffie. 1989. Between Free Trade and Protectionism: Strategic
Trade Policy and a Theory of Corporate Trade Demands. International Organization 43
(2): 239–272.
Osgood, Iain. 2013. Differentiated Products, Divided Industries:  Firms and the Politics of
Intra-Industry Trade. PhD diss., Harvard University. ProQuest/UMI 10982.
Plouffe, Michael. 2011a. Firms and Foreign Market Engagement: Internationalization Activities
among Japanese Firms. Paper presented at the meeting of the Japan Society of International
Economics, Tokyo, 23 October 2011.
Plouffe, Michael. 2011b. The New Political Economy of Trade:  Heterogeneous Firms and
Trade-Policy Positions. Paper presented at the meeting of the International Political
Economy Society, Madison, WI, 11 November 2011.
Plouffe, Michael. 2012. Liberalization for Sale: Heterogeneous Firms and Lobbying over FTAs.
Paper prepared for the meeting of the American Political Science Association, New Orleans,
Louisiana, 30 August 2012.
Plouffe, Michael. 2013. The New Political Economy of Trade: Heterogeneous Firms and Trade
Policy. PhD diss., University of California, San Diego.
Redding, Stephen. 2011. Theories of Heterogeneous Firms and Trade. Annual Review of
Economics 3: 77–105.
Richter, Brian Kelleher, Krislert Samphantharak, and Jeffrey F. Timmons. 2009. Lobbying and
Taxes. American Journal of Political Science 53, (4): 893–909.
Roberts, Mark J., and James R. Tybout. 1997. “The Decision to Export in Colombia:  An
Empirical Model of Entry with Sunk Costs.” The American Economic Review: 545–564.
Seker, Murat. 2011. “Importing, Exporting, and Innovation in Developing Countries.”
Unpublished manuscript.
Woll, Cornelia. 2008. Firm Interests:  How Governments Shape business Lobbying on Global
Trade. Ithaca, NY: Cornell University Press.
 Chapter 12

T he P olitics of Ma rk et
C om peti t i on
 Trade and Antitrust in a Global Economy

Ti m Bü t h e

The Trade-Antitrust Linkage

Antitrust law and enforcement—known as “competition policy” outside the United

States—is one of the most powerful tools that states have at their disposal to shape
the structure and operation of markets (e.g., Baron 2010, 265–310; Fligstein 1990).
Competition law authorizes and regulates government intervention against anticom-
petitive behavior, such as price-fixing or bid-rigging, and the concentration of economic
power, for instance through mergers and acquisitions. When it succeeds in safeguard-
ing or increasing market competition such that both buyers and sellers are generally
price-takers, it brings widely recognized economic benefits, boosting economic effi-
ciency, growth, and innovation, increasing both consumer and aggregate welfare (e.g.,
Gellhorn, Kovacic, and Calkins 2004; Rey 1997; Weiss 1989). In various countries and
at various times, competition policy has also had a number of other legitimate objec-
tives—many of them quite familiar to scholars of trade politics and policy—ranging
from industrial policy and economic development goals to economic freedom.1 But
even when it only seeks to enhance economic welfare, effective competition policy is
inherently deeply political, since it entails the use of political power to constrain or even
redistribute economic power.
Ever since the adoption of the first modern antitrust statutes in Canada and the
United States at the end of the nineteenth century, competition policy has been
closely intertwined with trade policy.2 To be sure, a diverse coalition of interests came
together in passing these first national competition laws (e.g., Baggaley 1991; Letwin
1965, esp. 53–99), but a central motivation was clearly an understanding of antitrust as
a substitute for trade openness. Senator John Sherman (for whom the first US federal
214    The Politics of Market Competition

antitrust statute, the Sherman Act of 1890, is named) sought antitrust enforcement in
large part to counterbalance the protectionist tariff for then-nascent northern indus-
tries. As a Republican senator from Ohio, he was a proponent of the controversial pro-
tectionist measures that subsequently came to be known as the McKinley tariff, but
he also recognized that such protection would threaten the effective operation of still
emerging and often oligopolistic markets for many of those industries’ products. Like
many of his contemporaries, Sherman thought that vigorously enforced laws against
anticompetitive practices, especially the abuse of market power, would make a protec-
tionist foreign economic policy compatible with an efficient market economy domesti-
cally (DiLorenzo 1985, esp. 82f.; Hazlett 1992).
The contemporary belief that antitrust could substitute for free trade is even more
important for understanding the politics of the original Canadian competition legisla-
tion, the Act for the Prevention and Suppression of Combinations Formed in Restraint
of Trade of 1889. “Combines” (known as “trusts” in the United States)—groups of
companies cooperating to “manage” production and prices and keep competitors at
bay—were still rather rare and commercially unimportant in the Canadian economy.
Consequently, populist hostility to the power of ever larger corporations, an important
source of political support for the US antitrust legislation, was much weaker in Canada
(e.g., Cohen 1938, 453f.; Bliss 1974, 33–54; though cf. Benidickson 1993). Why then did
the Canadian legislature pass the first modern, national competition law fourteen
months before the US Congress? Archival research and a close reading of the Canadian
legislative record suggest that the parliamentary majority’s enthusiasm for the law was
mostly about trade policy—“a calculated manoeuvre” by N. Clarke Wallace of the gov-
erning Conservative Party, the champion of the 1889 Act, “to deflect criticism from the
combine-creating effects of the protective tariff ” that his Conservative majority had put
in place (Bliss 1973, 182f., 185, 188). And the opposition Liberals in the Canadian House
of Commons accordingly portrayed the Act as a cynical distraction from the ill effects
of tariff protection (Bliss 1973, 182). Some went so far as to call it a legislative or political
“fraud” (Halladay 2012, 158; Trebilcock 1991).
The view of trade openness and antitrust enforcement as substitutes, which informed
Sherman and his contemporaries, can ultimately be traced back to the work of the classi-
cal political economists. Adam Smith, who warned against anticompetitive behavior in
numerous passages of the Wealth of Nations and other works, saw such private restric-
tions on market competition as equally as damaging as governmental measures such
as royal grants of monopolies. David Ricardo, best known to trade scholars for having
shown that the principles of comparative advantage can (under classical assumptions)
ensure that every country is better off under free trade than autarky, formalized some
of Smith’s key insights in his theory of rents (a key element of Ricardo’s Principles of
Political Economy and Taxation).3 From these sources, the substitutability of trade open-
ness and competition policy became the traditional view in trade economics, discussed
more fully in the next section.
Starting in the 1990s, however, this traditional approach faced a major empiri-
cal anomaly when the rapid, institutionalized spread of trade openness (evidenced in
 Tim Büthe   215

particular by the shift from the General Agreement on Tariffs and Trade to the World
Trade Organization and the explosive growth in the number of preferential trade agree-
ments) coincided with a similarly explosive diffusion of antitrust legal norms. During
the first century after the US Congress passed the Sherman Act, between 1890 and 1990,
the number of countries with competition laws on the books increased from 2 to 37.
During the next twenty years, between 1990 and 2010, the number of countries with
competition laws more than tripled, to about 120, even though the sustained, institu-
tionalized increase in international market integration during the same period of time
should have made the adoption of such laws ever more superfluous if trade openness
were a substitute for competition policy.
This observation prompted scholars of the law and economics of competition policy
to rethink the relationship between trade and competition policies. Newer theoretical
models tend to emphasize the possibility that competition law might—through selec-
tive, discriminatory enforcement—be abused as a trade barrier. These models differ
regarding the level of aggregation at which they operate, the attribution of causality, and
their implicit or explicit causal mechanisms. Yet, they have in common that they view
competition policy as a substitute for protectionism rather than as a substitute for free
trade. These new theoretical perspectives have been very influential in policy circles and
have informed sometimes heated debates over the question of whether antitrust law and
enforcement need to be brought under the umbrella of the WTO.4
The remainder of this chapter examines these different theoretical approaches to
thinking about the relationship between trade openness and competition policy, focus-
ing in the second section on how compelling they are deductively and in the third
section on how useful they are for understanding key aspects of antitrust in open econo-
mies. I argue that both perspectives suffer from truncated or missing models of politics,
which undermines their explanatory leverage. In the fourth section, I provide a sketch
of ongoing research that seeks to address these weaknesses of the existing theoretical
approaches and aims to provide a sounder empirical basis for understanding the inter-
national dimension of antitrust enforcement.

Theoretical Approaches: The
Law, Economics, and Politics
of Market Competition

Free Trade as a Substitute for Competition

Policy—and Vice Versa
Trade economists have traditionally viewed trade openness as a substitute for competi-
tion policy, because they see barriers to entry as the key economic rationale for competi-
tion policy. In any standard neoclassical model with profit-maximizing, self-interested
216    The Politics of Market Competition

economic agents, barriers to entry are a prerequisite for maintaining cartels and other
forms of anticompetitive behavior through which some economic actors benefit at
everyone else’s expense. Competition policy seeks to constrain anticompetitive behav-
ior directly by prohibiting such behavior and identifying and punishing those who
engage in it. Such a policy lowers barriers to market entry insofar as competition regu-
lators target anticompetitive behavior that creates or raises such barriers or industries
in which barriers to entry are high for other reasons, e.g., industries with high start-up
costs. Free trade promises to constrain anticompetitive behavior indirectly by open-
ing national markets to foreign producers. Free trade thus allows foreign producers to
enter and seize market share whenever domestic anticompetitive behavior keeps prices
at supra-competitive levels. And sustaining the expectation that foreign producers will
indeed do so requires us only to assume that competitive foreign producers for the prod-
uct in question exist and are materially self-interested, which should lead them to offer
their goods at lower prices, up to the point where those prices approximate the foreign
producers’ marginal costs.
In sum, trade liberalization checks market power, reducing the need for regula-
tory intervention to safeguard competition. As Blackhurst puts it: “Trade Policy is
Competition Policy” (1991). In fact, from the point of view of economic efficiency under
the conventional assumption of trade economics, free trade is superior to competition
policy because it achieves the same ends without the need to create and maintain a siz-
able bureaucracy and without the potential for abuse. Trebilcock (1991) hence refers to
competition policy as a “second best” approach to safeguarding market competition.
The logic of the argument that economic openness increases competition is impec-
cable, given its assumptions, which are common and conventional in the trade litera-
ture. In more recent decades, it has found support not just in formal models but also
some support from empirical analyses. As summarized by Irwin, “there is … overwhelm-
ing [empirical] evidence that free trade improves economic performance by increasing
competition in the domestic market” (2009, 43).5 Note, however, that it does not neces-
sarily follow that trade liberalization goes sufficiently far to constitute a real substitute
for regulatory enforcement of competition law—nor that economic openness is equally
effective in checking market power and safeguarding against anticompetitive behavior
in internationally integrated industries.
Three concerns motivate my skepticism about viewing free trade as a substitute for
competition policy. First, as Dixit (1984) pointed out thirty years ago, much interna-
tional trade occurs in industries whose structure more closely approximates oligopoly
than perfect competition. Second, competition policy may have important objectives
beyond maximizing consumer welfare or generally enhancing economic efficiency.
Ordo-liberals, for instance, attribute to competition policy an essential role in safe-
guarding economic (and therefore inherently also political) freedom, because high con-
centrations of economic power can be readily leveraged into political influence.6 The
debate about “too-big-to-fail” industrial and financial firms in the last few years suggests
that concerns about economic power remain timely. It is unclear, however, how trade
openness would address them. Third, and maybe most importantly, the theoretical logic
 Tim Büthe   217

underpinning the argument that free trade is a substitute for competition policy rests
on the assumption that firms are simply reactive to the constraints that public policy
might place on them, so that those constraints can be theoretically treated as exogenous.
But a large literature on the domestic politics of trade policy shows that, even though
genuine liberalization is possible (Gawande, Krishna, and Olarreaga 2009; Milner 1987;
Rogowski 1989), it is difficult and politically costly to use trade policy to force foreign
competition on a previously well-protected industry. Such long-protected industries are
likely—not least thanks to the rents afforded by being protected from foreign competi-
tors—to have material resources, strong incentives, and concentrated interests (includ-
ing often an organizational structure for overcoming collective action problems) to
simply replace one kind of protection with another (e.g., Grieco 1990; Limão and Tovar
2009; Kono 2006; Naoi 2009; WTO 2012). In sum, once the politics of trade policy are
taken seriously, the theoretical case for viewing a generally liberal foreign economic
policy as a substitute for domestic competition policy becomes deductively much less
compelling.
When we look at the theoretical argument the other way (competition policy as a sub-
stitute for free trade), it is similarly built on shaky foundations. As a theory of antitrust
enforcement, this approach assumes that competition regulators—and the politicians
who are in a position to influence them through statutory changes, resource allocation,
or other means—are disinterested guardians of market efficiency.7 Such heroic assump-
tions about policy makers’ objectives might not be surprising given that this approach is
devoid of politics. But competition law enforcement is inherently deeply political in that
it entails the use of public authority to constrain private actors’ exercise of market power.
Once we take politics seriously, the theoretical case for viewing competition policy as a
substitute for free trade becomes much weaker.

Competition Policy as a Substitute for Protectionism

A second perspective on the relationship between trade openness and competition
policy exhibits a greater (though still truncated) awareness of the politics of both trade
and competition policy. It posits competition policy as a substitute for trade restrictions
rather than trade openness, based on the assumption that competition law can be selec-
tively enforced to the benefit of domestic stakeholders and the detriment of their foreign
counterparts, including the foreign competitors of domestic firms. It is theoretically
useful to distinguish two variants of this antitrust-as-protectionism approach.
What may be called the aggregate national welfare variant, with strong affinities to
statist theories of International Relations, treats governments as unitary actors and
assumes that each government seeks to maximize the country’s aggregate welfare.
Under this assumption, (net) imports create an incentive for overly stringent enforce-
ment, because such an “oversupply” of antitrust enforcement creates benefits for domes-
tic consumers, whereas the costs are borne disproportionately by foreign producers. By
contrast, (net) exports create an incentive for overly lax enforcement, because the gains
218    The Politics of Market Competition

from uncorrected anticompetitive behavior are disproportionately enjoyed by domestic

producers, whereas the costs are disproportionately borne by foreign consumers (e.g.,
Horn and Levinsohn 2001; Williams and Rodriguez 1995). Such selective enforcement
is clearly attractive (and hence expected) for economically large countries, because it
should yield a gain in aggregate economic welfare for them: for countries with suffi-
ciently large markets, changes in their production or consumption induced by such
selective antitrust enforcement affect the world price and hence improve the country’s
terms of trade (Guzman 1998, 2004).
The other variant, which may be called the domestic political economy variant of the
“antitrust-as-protectionism” perspective, has strong affinities to the public choice cri-
tique of regulation, with its assumption that the state and any specific manifestations of
public authority tend to get captured by special, usually private interests. This approach
takes firms more seriously as (potential) political actors than trade economists tradi-
tionally have done. Specifically, scholars in this tradition assume that domestic firms,
when faced with trade liberalization that threatens to expose them to increased foreign
competition, will turn to the government or state agencies to achieve a similar level
of protection through other (ab)uses of public authority. And firms’ “actions aimed at
effectively locking competing imports or foreign investors out of their domestic mar-
ket” (Trebilcock and Howse 2005, 591) can include the “use of antitrust to subvert com-
petition” (Baumol and Ordover 1985, 247). The argument still lacks an explicit theory
of politics or policy making, but typically assumes a pluralistic responsiveness of pol-
icy makers—including competition regulators—to political lobbying (e.g., Shughart,
Silverman, and Tollison 1995). Consequently, it yields observable implications similar to
those noted for the aggregate national welfare variant but for all countries (rather than
just economically large countries), because there is no assumption that policy makers
seek to maximize aggregate welfare and therefore will only engage in selective enforce-
ment that is “efficient” for the national economy.
While both variants of this theoretical approach address certain weaknesses of the
traditional trade economics approach, they also both suffer from theoretical weak-
nesses of their own. The logical structure of the argument underpinning the aggregate
national welfare variant, for instance, is essentially the same as for optimal tariff theory
(e.g., Krugman 1986), so that the well-known critiques of optimal tariff theory should
equally apply here. For example, critics of optimal tariff theory have long pointed out
that a country’s attempts to achieve welfare gains at the expense of its trading partners
(here through selective enforcement) invite retaliation by some of those trading part-
ners, since no country is economically large in all industries. Such retaliation turns the
hypothetically possible tactical welfare gain for the one side into a welfare loss for both.
And since the loss is predictable, we have no reason to expect that we will empirically
observe the selective enforcement that was the original cause of concern.
Moreover, both variants, while more attentive to the politics of antitrust than the tra-
ditional trade economics perspective, still rely on an overly truncated model of poli-
tics. Even in the domestic political economy variant of the antitrust-as-protectionism
perspective, firms are recognized as political actors only domestically, where they lobby
 Tim Büthe   219

regulators and the politicians who oversee them. They are denied transnational politi-
cal agency, even though substantial research shows firms to be increasingly important
transnational political actors (e.g., Baron 2001; Büthe 2010; Büthe and Mattli 2011;
Cutler, Haufler, and Porter 1999). A similar problem affects the conceptualization of
governments: Although they are recognized as pertinent political actors across borders,
they are conceptualized only as unitary actors. Much research about the politics of inter-
national economic relations and global governance, however, emphasizes the increasing
importance of transnational politics and transgovernmental networks (e.g., Keohane
and Nye 1989; Newman and Posner 2011; Slaughter 2004).
In sum, if we take both firms and regulators seriously as transnational political actors,
then there is no reason to expect firms to have a preference for public over private
protection, which motivates the exclusive focus on the [ab]use of public power in the
domestic political economy model. Nor need we assume that regulatory practices and
actions are shaped only by the domestic political environment in which the regulators
operate.8 Once we relax those assumptions, there is no compelling reason to expect that
the increasing international integration of product markets would result in the (ab)use
of competition policy for protectionist purposes.

Empirical Analyses

Putting aside, for the moment, the above critiques of the deductive logic of the existing
theoretical approaches, I now turn to a brief assessment of their explanatory lever-
age for understanding key aspects of competition policy in open economies. I under-
take this assessment with the caveat that few of the existing empirical studies were
set up to test the theoretical approaches presented in the preceding section vis-à-vis
each other. Moreover, many gaps remain, even in the basic descriptive information
available about antitrust law and enforcement across countries and over time. Yet, the
theoretical approaches discussed above yield a number of clearly distinct observable
implications concerning various aspects of competition policy (King, Keohane, and
Verba 1994). Specifically, we should want these approaches to provide insights into
whether and, if so how, international economic integration and in particular trade
openness affect

• countries’ decisions about whether and when to adopt competition legislation,

allowing for the possibility of network effects;
• variation in the substantive focus or stringency of antitrust laws across countries
and over time;
• variation in antitrust enforcement across countries and over time;
• the broader pattern of international diffusion of antitrust law and enforcement;
• patterns of competition and anticompetitive behavior; and
• international conflict and cooperation over antitrust law and enforcement.
220    The Politics of Market Competition

Overall, with regard to these six sets of explananda, many and maybe most of the
implications of the different theoretical approaches remain untested. A few clear find-
ings nonetheless emerge, and identifying the gaps in our knowledge also helps specify
avenues for future research.
Adoption of national antitrust law and its content or stringency: Here the trade eco-
nomics approach suggests that there should be a negative relationship between a coun-
try’s trade openness and the likelihood that the country adopts a competition law or
increases its stringency. By contrast, the antitrust-as-protectionism approach implies
that there should be a positive relationship between the likelihood that a country adopts
or strengthen its competition law and the country’s level of net imports.
Strikingly, there is, as noted by Gutmann and Voigt, “no established dataset based on
one consistent definition of what constitutes a competition law” (2014, 6) and no com-
prehensive database of countries’ competition laws, which would allow easy access to
even such basic information as the year when a country adopted a competition law for
the first time.9 There is consequently very little large-N, comparative empirical work
on the adoption of national competition legislation. In joint work with Anu Bradford,
I am currently building a comprehensive panel data set of national antitrust laws, based
on a detailed coding of their substantive provisions and secondary legislation/imple-
menting regulations. That database, however, is still a work in progress. To provide a
preliminary sense of the explosive spread of competition laws over the 125 years since
1889, I have included in Figure 12.1 information from the gap- and error-prone antitrust-
worldwiki (see also Hylton and Deng 2007), complemented by data from Kronthaler
(2010), Petersen (2013), Voigt (2009), and Waked (2010), each of whom has collected

150
No. of Jurisdictions w| Competition Law

100

50

0
1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

Year

Fig.  12.1  Increase in the Number of Jurisdictions with Competition Laws, 1945–2010.
 Tim Büthe   221

more specific information for different subsets of countries. The figure shows the striking
pattern of the number of countries with a competition law that at least prohibits cartels,
from 1945, when only Canada and the United States had such a law on the books, to 2010.
Given that trade openness has generally increased over the course of the sixty-five
years captured in the graph and especially during the two decades after the WTO
replaced GATT, the overall pattern in Figure 12.1 seems clearly inconsistent with a view
of trade openness and competition policy as substitutes. The graph does not, however,
provide us with insights into what motivated the adoption of antitrust laws across these
countries, especially whether an increase in net imports might have driven their adop-
tion, as suggested by the antitrust-as-protectionism perspective. In the absence of a
well-established empirical model, Table 12.1 takes a first cut at this question by reporting
the trade patterns in the years prior to a country’s first adoption of an antitrust law for
the ninety-three countries that enacted their first competition law between 1980 and
2010 (subject to data availability).
To be sure, bivariate correlations with no consideration of possible confounding
factors cannot yield conclusive results. That said, the empirical findings in Table 12.1
offer no support for the trade economics approach, which suggests that the probability
of adopting antitrust legislation increases during periods of unusual low trade open-
ness. Table 12.1 instead show slightly higher average levels of trade openness in the 1–3
years immediately preceding the adoption of a first antitrust law, thought the difference
is not statistically significant. Moreover, net imports are on average lower in the years
preceding the adoption of a first antitrust law, and significantly so for the year t-1. This
preliminary empirical finding directly runs counter to the expectation of the antitrust-
as-protectionism approach.

Table 12.1  Trade Patterns Before the Enactment of a Country’s First Antitrust Law,
1980–2010
1 year before All other years and 3 years before All other years and
first AT Law countries first AT law avg. (t-1, t-2, t-3) countries

trade openness 84.2% 83.9% 84.6% 83.9%

(4.70) (0.731) (3.06) (0.739)
net imports 3.69% 6.90% 5.68% 6.88%
(1.32) (0.252) (0.940) (0.253)
N 84 4839 222 4797
Countries 84 187 74 187

Note:  Mean levels of trade (X + M) and net imports, as a percentage of GDP; standard errors in
parentheses. Trade data from World Development Indicators (Nov. 2011). Analysis of trade patterns
1 year before the adoption of the country’s first antitrust law omits, due to missing trade data, 9
jurisdictions with an antitrust law (first enacted since 1980): Argentina, Channel Islands, Faroe Islands,
Greenland, Kazakhstan, Liechtenstein, Montenegro, Poland, and Taiwan. Missing data cause the loss
of another ten countries with AT laws in the analysis of trade flows during the 3 years prior to the
adoption of the first antitrust law: Belarus, the Czech Republic, Estonia, Lithuania, Latvia, Moldova,
Russia, Slovenia, Ukraine, and Uzbekistan.
222    The Politics of Market Competition

Enforcement: Analyses of legal provisions must be supplemented by analyses that

consider their enforcement in order to avoid falling into the trap of the old institutional-
ist literature of mistaking formal rules for practice.10 Information about enforcement
also requires careful interpretation: A highly effective antitrust enforcement agency, in
equilibrium, needs to undertake little or no enforcement actions, because it successfully
deters all violations.
Unfortunately, internationally comparable enforcement data are even scarcer than data
about national laws, and as Petersen notes, “there is no [agreed] perfect way to measure
the existence of an effective antitrust regime” (2013, 606, emphasis added). Consequently,
very few comparative analyses have been conducted—with mixed results.11 In an unpub-
lished paper based on field research in Africa and Asia, Waked (2010) finds a negative
empirical relationship between trade openness and input measures of antitrust enforce-
ment (budget and staff), which might be interpreted as support for the view that trade is
a substitute for antitrust law and enforcement. The estimated coefficients for her mea-
sures of enforcement, however, are statistically significant in only seven of her twelve
models, and her analysis is limited to twenty-eight to thirty developing countries, with
five observations over time on average. Moreover, some other empirical analyses reach
opposite findings. Consistent with earlier anecdotal evidence that “the one action taken”
under the 1910 Canadian Combines Investigation Act “was against a foreign corporation”
(McFall 1922, 182), Shughart, Silverman, and Tollison (1995) find significant support for
the antitrust-as-protectionism approach in their analysis of the antitrust enforcement
budgets for the US Department of Justice’s Antitrust Division and for the Federal Trade
Commission. However, their work is limited to a single country, ends in 1981, and does
not address concerns about multicollinearity and spurious correlation, which inevitably
arise when regressing a trending time series on several other time series (see Büthe and
Morgan with Bradford 2014 for a critique). More robust analyses of the effect of economic
openness on antitrust enforcement would include a more diverse set of countries, more
comprehensive measures of the stringency of antitrust enforcement, and/or longer time
series.
Patterns of competition and anticompetitive behavior: Patterns of competition
are merely an indirect but substantively important implication of the theoretical
approaches discussed above. Analyses of patterns of competition and anticompetitive
behavior moreover provide information about the suitability of key assumptions of the
different models (see Coase 1994). A number of studies have found that increases in
trade openness increase the level of competition within a country, confirming one of
the key assumptions of the trade economics approach. Other studies find that the strin-
gency and/or independence of antitrust enforcement increases market competition in
a country (e.g., Voigt 2009). Without assessing the effect of trade openness and com-
petition policy simultaneously, however, these studies do not speak to the question of
whether free trade and antitrust work as substitutes or in some other relationship with
each other. Including both factors in multicountry, industry-level analyses is one of the
key contributions of Kee and Hoekman (2007). And while they confirm that free trade
 Tim Büthe   223

increases competition, they also find—contrary to what we should expect based on the
trade economics model—that competition policy significantly further increases the
level of competition, at least indirectly.

An Alternative Approach: Free
Trade and Competition Policy
as Genuine Complements

To address the theoretical weaknesses of the analytical approaches discussed in the sec-
ond section and provide a better understanding of the law and politics of antitrust in a
global economy I have, initially in joint work with Anu Bradford, developed an alterna-
tive approach, which views trade openness and competition policy as genuine comple-
ments (see, e.g., Büthe and Bradford 2012). I briefly sketch this approach here. It departs
from the previous approaches in two main ways.
First, I take firms seriously as (potential) political as well as economic actors—and
not just vis-à-vis their domestic governments but also transnationally. Conceptualized
in this way, firms can not only turn to their government when they seek protection but
also pursue “private protection,”12 for instance by colluding with other firms. And trade
liberalization, by putting firms that previously did not meaningfully compete into com-
petition with each other, creates new opportunities to gain from transnational collusion,
at the same time that it lowers the risk of detection because monitoring global markets
is more difficult and costly than monitoring purely domestic markets, and because evi-
dence of transnational collusion can more easily be kept out of reach of enforcement
agencies that operate at the national level.13
Second, this approach takes governments and regulatory agencies seriously as politi-
cal actors at both the national and inter/transnational levels. Specifically, it assumes that
governments (and competition regulators) understand that free trade creates opportuni-
ties and incentives for private protection. While recognizing that regulatory agencies and
elected officials can be captured by special interests, this approach does not elevate the
possibility to a general assumption.14 To the contrary, it assumes that competition regula-
tors see safeguarding market competition as their primary objective unless institutional
features (such as the lack of agency independence) provide specific reasons to expect
otherwise.
What are the empirical implications of this new approach to thinking about the rela-
tionship between economic openness and competition policy? The first point above
implies that we should see an increase in transnational anticompetitive behavior as a
function of the institutionalization of product market integration—that is, as govern-
ments’ options for providing public protection decline and firms’ opportunities and
incentives for transnational collusion increase. Transnational anticompetitive behavior
224    The Politics of Market Competition

is therefore expected to increase even while we might see a decrease in purely domes-
tic anticompetitive behavior in industries that have been opened to foreign competition.
It is naturally difficult to establish comprehensive patterns of anticompetitive business
behavior, because what is in most jurisdictions now illegal behavior will surely not be
fully detected. But a striking number of multinational cartels have been discovered in
recent years (e.g., Bond 2005; Connor 2007; Kovacic et al. 2007; Levenstein and Suslow
2008). And the often near-global reach, complexity, and persistence of many of these
cartels show that anticompetitive behavior does not stop at borders, as the traditional
approaches assume, and that cartels are not nearly as unstable as most economics text-
books would have us believe.
Interviews with competition regulators suggest that the increase in the detection
of such cartels has been at least in part a function of increased enforcement efforts,
including increased monitoring of international rather than just domestic markets and
increased transgovernmental enforcement cooperation (discussed below). Importantly,
the transnational collusion appears to have begun, for most of the major transnational
cartels that have been detected, only after the relevant markets experienced a substantial
increase in international openness.
The second point above implies that competition regulators will recognize the
increased probability of transnational anticompetitive behavior, and that they will
seek to counteract it.15 Empirically, this implies a positive relationship between trade
openness and the probability of adopting a competition law, as in fact shown in Table
12.1, which yielded anomalous findings for the other theoretical approaches. It implies
further that we should expect to see greater trade openness (across countries and
over time) result in more resources being devoted to monitoring international rather
than just domestic markets. There is strong anecdotal evidence that this has occurred,
at least in the United States, several European countries, and the European Union,
though research completed to date does not allow a full assessment of this observable
implication.
Relatedly, we should expect to see trade openness result in increased efforts to estab-
lish and institutionalize transgovernmental antitrust enforcement collaboration.
Preliminary empirical findings provide substantial support for this hypothesis. The
long-existing but loose and entirely informal transgovernmental network of competi-
tion regulators, now known as the International Competition Network, has over the
last twenty years become increasingly institutionalized; it also has grown tremendously
(Aydin 2010; Djelic and Kleiner 2006; Svetiev 2010). It has been complemented by
efforts to foster international and transgovernmental collaboration on antitrust enforce-
ment through the Organisation for Economic Co-operation and Development and the
UN Conference on Trade and Development.
There also is growing evidence of increased bilateral (and occasionally minilateral)
enforcement cooperation among competition regulators. In a detailed analysis of
antitrust provisions in a random sample of 182 preferential trade agreements (PTAs),
Bradford and Büthe (2015) find that competition provisions generally were rare and
 Tim Büthe   225

usually minimalist through the early 1990s, but since the mid-1990s have become
a common and prominent feature of PTAs. This increase in the frequency and detail
of antitrust provision in PTAs has coincided with the qualitative shift toward a much
more institutionalized multilateral trade regime under the WTO (and with a general
increase in specificity and scope of PTAs, i.e., the increase in the institutionalization
of minilateral trade agreements). Importantly, Bradford and Büthe find that provisions
for information exchange and mutual assistance in antitrust enforcement are strik-
ingly common, suggesting a real interest in facilitating transgovernmental cooperation.
By contrast, provisions that seek to exempt a country’s firms from the other country’s
domestic competition regime or in other ways signal concern about the other side’s
abuse of competition policy for protectionist purposes are relatively rare (Bradford and
Büthe 2015).
This institutionalization of enforcement cooperation in trade agreements is supple-
mented by a nearly simultaneous growth in dedicated antitrust enforcement coopera-
tion agreements over the same time period, shown in Figure 12.2, as well as a similar
increase in antitrust provisions in international agreements from broad cooperation
and friendship agreements and general (non-PTA) economic agreements to bilateral
law enforcement agreements. These findings provide further support for the hypothesis
that governments do indeed see effective competition policy as a complement to trade
openness (Büthe and Bradford 2012).

35

30
Number of Agreements in Force

25

20

15

10

0
1980 1985 1990 1995 2000 2005 2010
Year

Fig.  12.2  Increase in Dedicated International Competition Agreements, 1980–2010.

226    The Politics of Market Competition

Conclusion

Political economists since Adam Smith and David Ricardo have recognized increased
market competition as an important benefit of a free trade policy. Antitrust law and
its enforcement is supposed to achieve the same end by intervening directly against
price-fixing, bid-rigging, and other forms of anticompetitive behavior and increases
or abuses of market power. And indeed, trade policy and competition policy have
been deeply intertwined since the very beginning of modern competition law in late
nineteenth-century North America.
Yet, the major schools of thought that have shaped antitrust law and its enforce-
ment—structuralism, ordo-liberalism, the Chicago school, and even the new
industrial organization approach—have largely failed to recognize and address this trade-
competition policy nexus: They generally assume, at least implicitly, that the boundaries
of the market coincide with the boundaries of the antitrust jurisdiction. The question of
how economic openness affects competition policy has therefore become the focus in a
distinct set of theoretical approaches.
This chapter has provided an overview of different ways of thinking theoreti-
cally about the relationship between the international integration of product markets
on the one hand and competition law and enforcement on the other. Based on both
a theoretical and an empirical assessment, I argued that the dominant approaches in
economics and law (the disciplines that have dominated the existing literature on the
trade-competition policy nexus) suffer from either being devoid of politics or relying on
a model of politics that deprives both firms and competition regulators of transnational
and transgovernmental agency, respectively.
I have sketched a more overtly political approach, based on recent and forthcoming
joint work (e.g., Büthe and Bradford 2012; Bradford and Büthe 2015). It takes firms seri-
ously as sophisticated transnational actors, not just able to lobby their domestic gov-
ernments for public protection but alternatively capable of setting up systems of private
protection in an increasingly global economy. At the same time, it takes government
regulators seriously as political actors with often considerable independence and hence
a capacity to pursue their own, distinct interests, including in a transgovernmental net-
work of competition regulators, in which they are embedded.
The empirical analysis of the international and comparative dimension of competi-
tion policy is still in its infancy. Preliminary results suggest, however, that a theoretical
model that allows for transnational as well as transgovernmental politics can explain
the simultaneous spread of competition law and trade openness in the last two decades
and more generally yields an understanding of trade and antitrust in the global econ-
omy that is at least equally as plausible as the traditional model of trade economics and
models that posit antitrust as a substitute for protectionism. This finding should not
come as a surprise: antitrust enforcement inherently entails the use of public authority
to constrain private actors’ concentration and exercise of market power. Hence, if we
 Tim Büthe   227

want to understand competition policy in a global economy, we need not just legal and
economic but also political analysis.

Author’s Note

For helpful comments, I thank Anu Bradford, Cindy Cheng, Lisa Martin, and members
of the audience at presentations at the World Trade Institute, Bern, and at Yale University
(Leitner Seminar). Research for this chapter has been supported in part by grants from the
Law and Social Sciences program of the National Science Foundation (grant no. 1228483)
and from the Josiah Charles Trent Memorial Foundation (grant 12-06).

Notes
1. See, e.g., Böhm (1961); Crane and Hovenkamp (2013); Eucken (2001); Fox (2011); Gerber
(2010); Lande (1982). Even within economics, multiple distinct notions of competition are
common (Stigler 1957; Martin 2012, esp. 5–11) and in fact have persisted for a long time; see
De Roover (1951) and McNulty (1968).
2. Federal antitrust legislation in the United States was preceded by state-level antitrust
statutes in more than twenty states, though in most cases only by a few years or even just
months. Previously, English common law had been seen as providing a sufficient safe-
guard through its prohibitions against anticompetitive behavior since at least the 1700s,
but its competition law elements had successively weakened since the 1840s, eventually
prompting the adoption of statutory competition law in Canada and the United States
(see, e.g., Bliss 1973, 178f.; Letwin 1965, esp. 19–52).
3. For excerpts of key passages from Smith and Ricardo’s works, see Crane and Hovenkamp
(2013, 5-40).
4. As Iacobucci (1997, 5f.) points out, the ultimately abandoned 1947/1948 Havana Charter
for an International Trade Organization already included rules concerning anticompeti-
tive business practices. Proposals for a supranational competition regime in the context
of the global trade regime in fact have often sought to increase the aggregate welfare gains
from competition policy rather than “just” to constrain national competition regulators
in order to forestall abuse (e.g., Anderson and Holmes 2002; Fox 2003; Guzman 2004;
Marsden 2003; Zäch and Correa 1999). For key critiques of a WTO-based competition
regime, see Bradford (2007) and McGinnis (2004). From a trade policy perspective, argu-
ably the greatest benefit of a formal international competition policy regime linked to the
WTO and the international trade regime might be that it would provide an opportunity to
reign in the abuses of antidumping, since allegations of dumping would be antitrust viola-
tions if it were not for the foreign nationality of the allegedly dumping producer.
5. Moreover, anecdotal evidence suggests some substitutability insofar as (detected) anti-
trust violations in Canada during the early decades appeared to be disproportionately
concentrated in industries “protected from foreign competition by substantial tariffs or
other barriers to trade” (Trebilcock et al. 2002, 642).
228    The Politics of Market Competition

6. Ordo-liberalism is a philosophical school of thought that is liberal in rejecting government

intervention when it seeks to direct economic activity but sees the state as having a neces-
sary “ordering” function in the economy to safeguard individuals against any concentra-
tion of political and economic power that would threaten their freedom and equality of
opportunity. From the 1950s through at least the 1990s, it shaped the European (EU-level)
approach to competition policy.
7. These assumptions are usually entirely implicit, but I submit that they are necessary and in
effect drive the models’ conclusions.
8. “Private protection” refers to economic actors protecting themselves against the conse-
quences of increased exposure to competitors through collusion or other anticompetitive
measures. It is discussed in greater detail below.
9. The challenge is illustrated by the history of Austria’s competition laws. The country
adopted its first “cartel legislation” in 1951. That legislation, however, primarily required
cartel agreements to be notified to a state agency in order to be enforceable as contracts.
It subjected cartel agreements to certain disciplines and created a quasi-judicial proce-
dure allowing the Austrian federal government to prohibit a cartel agreement if it had
detrimental consequences for the Austrian economy, but it made clear that agreements
on production quotas, prices, etc., would not per se be considered as having detrimen-
tal consequences. Amendments in 1956, 1957, 1958, 1962, and 1968 and a new version of
the law in 1972 provided for more elaborate administrative and judicial review processes
and increasingly restricted the permissiveness of cartels that sought to raise prices or keep
them from falling, but a general prohibition of cartels was not included in the law until
1988. Which year should count as the first year for Austria’s competition legislation?
10. Some of the most comprehensive, stringent antitrust laws are found in countries with gen-
erally very limited state/administrative capacity and a weak rule-of-law tradition.
11. An alternative approach is to conduct time series analysis within a single country with a
long history of antitrust enforcement, such as the United States. Existing longitudinal data
sets of US enforcement, however, lack information on the nationality of the parties against
which enforcement actions have been directed, as well as detailed information about the
industries involved (for disaggregated analyses of the relationship between trade patterns
and enforcement). Büthe and Bradford are developing a new data set for US Department
of Justice enforcement actions, 1960–2010, including such information.
12. See Ludema (2001); Trebilcock (1996); and Williams and Rodriguez (1995).
13. Furthermore, insofar as institutionalized trade liberalization succeeds in constraining gov-
ernments’ urge to protect domestic firms, it creates further incentives for firms to turn to
private protection.
14. This assumption is consistent with the finding that capture is much less common than
often assumed (Carpenter and Moss 2014).
15. Put another way, governments (or at least those within the “state” who see it as their objec-
tive to guard competition) should see trade openness and the need for vigorous competi-
tion policy as complements.

References
Anderson, Robert D., and Peter Holmes. 2002. Competition Policy and the Future of the
Multilateral Trading System. Journal of International Economic Law 5 (2): 531–563.
 Tim Büthe   229

Aydin, Umut. 2010. The International Competition Network: Cooperation and Convergence

in Competition Laws. Rekabet Derisi [Competition Journal] 11 (3): 51–78.
Baggaley, Carman. 1991. Tariffs, Combines and Politics:  The Beginning of Canadian
Competition Policy, 1880–1900. In Historical Perspectives on Canadian Competition Policy,
edited by R. S. Khemani and W. T. Standbury, 1–51. Halifax, NS: Institute for Research on
Public Policy.
Baron, David P. 2001. Private Politics, Corporate Social Responsibility, and Integrated Strategy.
Journal of Economics & Management Strategy 10 (1): 7–45.
Baron, David P. 2010. Business and Its Environment. 6th ed. Upper Saddle River, NJ: Prentice
Hall.
Baumol, William J., and Janusz A. Ordover. 1985. Use of Antitrust to Subvert Competition.
Journal of Law and Economics 28 (2): 247–265.
Benidickson, Jamie. 1993. The Combines Problem in Canadian Legal Thought, 1867–1920.
University of Toronto Law Journal 43 (4): 799–850.
Blackhurst, Richard. 1991. Trade Policy Is Competition Policy. In Competition and Economic
Development, 253–258. Paris: OECD.
Bliss, Michael. 1973. Another Anti-Trust Tradition:  Canadian Anti-Combines Policy,
1889–1910. Business History Review 47 (2): 177–188.
Bliss, Michael. 1974. A Living Profit: Studies in the Social History of Canadian Business, 1883–1911.
Toronto: McClelland and Stewart Ltd.
Böhm, Franz. 1961. Democracy and Economic Power. In Cartel and Monopoly in Modern
Law: Reports on Supranational and National European and American Law, Presented to
the International Conference on Restraints of Competition at Frankfurt am Main, June
1960, edited by the Institut für ausländisches und internationales Wirtschaftsrecht and
der Johann-Wolfgang-Goethe-Universität, Frankfurt am Main, in cooperation with the
Institute for International and Foreign Trade Law of the Georgetown University Law Center,
Washington D.C. 25–45. Karlsruhe: Verlag C. F. Müller.
Bond, Eric W. 2005. Antitrust Policy in Open Economies: Price Fixing and International Cartels. In
Handbook of International Trade: Economic and Legal Analyses of Trade Policy and Institutions,
edited by E. Kwan Choi and James C. Hartigan, 333–357. Malden, MA: Blackwell Publishing.
Bradford, Anu. 2007. International Antitrust Negotiations and the False Hope of the WTO.
Harvard International Law Journal 48 (2): 383–440.
Bradford, Anu, and Tim Büthe. 2015. “The Politics of Competition Policy and Free Trade:
Antitrust Provisions in PTAs.” In Trade Cooperation: The Purpose, Design and Effects
of Preferential Trade Agreements, edited by Andreas Dür and Manfred Elsig, 246–274.
Cambridge, UK: Cambridge University Press.
Büthe, Tim. 2010. Private Regulation in the Global Economy:  A  (P)Review. Business and
Politics 12 (3): 1–38.
Büthe, Tim, and Anu Bradford. 2012. Wettbewerbspolitik und internationaler Handel—Macht
und Politik in Offenen Märkten: A Research Agenda and Some Preliminary Results. Paper
presented at the 25th Biennial Convention of the German Political Science Association
(DVPW), Tübingen, September 25.
Büthe, Tim and Stephen Morgan with Anu Bradford. 2014. Antitrust Enforcement and Foreign
Competition: Special Interest Theory Reconsidered. Paper presented at the Conference on
Empirical Legal Studies, University of California, Berkeley, November 8.
Büthe, Tim, and Walter Mattli. 2011. The New Global Rulers: The Privatization of Regulation in
the World Economy. Princeton, NJ: Princeton University Press.
230    The Politics of Market Competition

Carpenter, Daniel P., and David A. Moss, eds. 2014. Preventing Regulatory Capture: Special
Interest Influence and How to Limit It. New York: Cambridge University Press.
Coase, Ronald H. 1994. How Should Economists Choose? First delivered as the 1981 G. Warren
Nutter Lecture in Political Economy. In Essays on Economics and Economists, 15–33.
Chicago: University of Chicago Press.
Cohen, Maxwell. 1938. The Canadian Anti-Trust Laws: Doctrinal and Legislative Beginnings.
Canadian Bar Review 16 (6): 438–465.
Connor, John M. 2007. Global Price Fixing. 2nd ed. Berlin/New York: Springer.
Crane, Daniel A., and Herbert Hovenkamp, eds. 2013. The Making of Competition Policy: Legal
and Economic Sources. Oxford: Oxford University Press.
Cutler, A. Claire, Virginia Haufler, and Tony Porter, eds. 1999. Private Authority and
International Affairs. Albany: State University of New York Press.
De Roover, Raymond. 1951. Monopoly Theory Prior to Adam Smith: A Revision. Quarterly
Journal of Economics 65 (4): 492–524.
DiLorenzo, Thomas J. 1985. The Origins of Antitrust:  An Interest-Group Perspective.
International Review of Law and Economics 5 (1): 73–90.
Dixit, Avinash. 1984. International Trade Policy for Oligopolistic Industries. Economic Journal
94 (supp.): 1–16.
Djelic, Marie-Laure, and Thibaut Kleiner. 2006. The International Competition Network:
Moving Towards Transnational Governance. In Transnational Governance: Institutional
Dynamics of Regulation, edited by Marie-Laure Djelic and Kerstin Sahlin-Andersson,
287–307. New York: Cambridge University Press.
Eucken, Walter. 2001. Wirtschaftsmacht und Wirtschaftsordnung:  Londoner Vorträge zur
Wirtschaftspolitik und zwei Beiträge zur Antimonopolpolitik. Münster: LIT Verlag für das
Walter-Eucken-Archiv.
Fligstein, Neil. 1990. The Transformation of Corporate Control. Cambridge, MA: Harvard
University Press.
Fox, Eleanor M. 2003. International Antitrust and the Doha Dome. Virginia Journal of
International Law 43: 911–932.
Fox, Eleanor M. 2011. Competition, Development, and Regional Integration:  In Search of a
Competition Law Fit for Developing Countries. In Competition Policy and Regional Integration
in Developing Countries, edited by Josef Drexl et al., 273–290. Cheltenham, UK: Edward Elgar.
Gawande, Kishore, Pravin Krishna, and Marcelo Olarreaga. 2009. What Governments
Maximize and Why: The View from Trade. International Organization 63 (3): 491–532.
Gellhorn, Ernest, William E. Kovacic, and Stephen Calkins. 2004. Antitrust Law and Economics
in a Nutshell. 5th ed. St. Paul, MN: West Publishing.
Gerber, David J. 2010. Global Competition: Law, Markets, and Globalization. New York: Oxford
University Press.
Grieco, Joseph M. 1990. Cooperation Among Nations: Europe, America, and Non-Tariff Barriers
to Trade. Ithaca, NY: Cornell University Press.
Gutmann, Jerg, and Stefan Voigt. 2014. Lending a Hand to the Invisible Hand? Assessing
the Effects of Newly Enacted Competition Laws. Unpublished manuscript, University of
Hamburg, February.
Guzman, Andrew T. 1998. Is International Antitrust Possible? New York University Law Review
73 (5): 1501–1548.
Guzman, Andrew T. 2004. The Case for International Antitrust. Berkeley Journal of
International Law 22 (3): 355–374.
 Tim Büthe   231

Halladay, Casey W. 2012. The Origins of Canada’s Cartel Laws. Canadian Competition Law
Review 25 (1): 157–163.
Hazlett, Thomas W. 1992. The Legislative History of the Sherman Act Re-Examined. Economic
Inquiry 30 (2): 263–276.
Horn, Henrik, and James Levinsohn. 2001. Merger Policies and Trade Liberalisation. Economic
Journal 111 (470): 244–276.
Hylton, Keith N., and Fei Deng. 2007. Antitrust Around the World: An Empirical Analysis of
the Scope of Competition Laws and Their Effects. Antitrust Law Journal 74 (2): 271–341.
Iacobucci, Edward M. 1997. The Interdependence of Trade and Competition Policies. World
Competition 21 (2): 5–33.
Irwin, Douglas A. 2009. Free Trade Under Fire. 3rd ed. Princeton, NJ: Princeton University
Press. First published 2002.
Kee, Hiau Looi, and Bernard Hoekman. 2007. Imports, Entry, and Competition Law as Market
Disciplines. European Economic Review 51 (4): 831–858.
Keohane, Robert O., and Joseph S. Nye. 1989. Power and Interdependence. 2nd ed. New York:
HarperCollins. First published 1977.
King, Gary, Robert O. Keohane, and Sidney Verba. 1994. Designing Social Inquiry: Scientific
Inference in Qualitative Research. Princeton, NJ: Princeton University Press.
Kono, Daniel Y. 2006. Optimal Obfuscation:  Democracy and Trade Policy Transparency.
American Political Science Review 100 (3): 369–384.
Kovacic, William E., et al. 2007. Lessons for Competition Policy from the Vitamins Cartel. In
The Political Economy of Antitrust, edited by Vivek Ghoshal and Johan Stennek, 149–176.
Amsterdam/Boston: Elsevier.
Kronthaler, Franz. 2010. Factors Influencing the Implementation of Recently Enacted
Competition Laws: An Empirical Analysis. International Research Journal of Finance and
Economics, no. 51, 71–87.
Krugman, Paul R., ed. 1986. Strategic Trade Policy and the New International Economics.
Cambridge, MA: MIT Press.
Lande, Robert H. 1982. Wealth Transfer as the Original and Primary Concern of Antitrust: The
Efficiency Interpretation Challenged. Hastings Law Journal 34 (1): 65–151.
Letwin, William. 1965. Law and Economic Policy in America: The Evolution of the Sherman
Antitrust Act. New York: Random House.
Levenstein, Margaret C., and Valerie Y. Suslow. 2008. International Cartels. Issues in Compe­
tition Law and Policy 2: 1107–1126.
Limão, Nuno, and Patricia Tovar. 2009. Policy Choice: Theory and Evidence. NBER Working
Paper No. 14655.
Ludema, Rodney D. 2001. Market Collusion and the Politics of Protection. European Journal of
Political Economy 17 (4): 817–833.
Marsden, Philip. 2003. Competition Policy for the WTO. London: Cameron May.
Martin, Stephen. 2012. Globalization and the Natural Limits of Competition. In The
International Handbook of Competition, edited by Manfred Neumann and Jürgen Veigand,
4–56. Cheltenham, UK/Northampton, MA: Edward Elgar.
McFall, Robert James. 1922. Regulations of Business in Canada. Political Science Quarterly 37
(2): 177–210.
McGinnis, John O. 2004. The Political Economy of International Antitrust Harmonization. In
Competition Laws in Conflict: Antitrust Jurisdiction in the Global Economy, edited by Richard
A. Epstein and Michael S. Greve, 136–151. Washington, DC: American Enterprise Institute.
232    The Politics of Market Competition

McNulty, Paul J. 1968. Economic Theory and the Meaning of Competition. Quarterly Journal of
Economics 82 (4): 639–656.
Milner, Helen V. 1987. Resisting the Protectionist Temptation:  Industry and the Making of
Trade Policy in France and the United States During the 1970s. International Organization
41 (4): 639–666.
Naoi, Megumi. 2009. Shopping for Protection: The Politics of Choosing Trade Instruments in a
Partially Legalized World. International Studies Quarterly 53 (2): 421–444.
Newman, Abraham L., and Elliot Posner. 2011. International Interdependence and Regulatory
Power: Authority, Mobility, and Markets. European Journal of International Relations 17 (4):
589–610.
Petersen, Niels. 2013. Antitrust Law and the Promotion of Democracy and Economic Growth.
Journal of Competition Law & Economics 9 (3): 593–636.
Rey, Patrick. 1997. Competition Policy and Economic Growth. Mimeograph, Unversité de
Toulouse (IDEI), September.
Rogowski, Ronald. 1989. Commerce and Coalitions:  How Trade Affects Domestic Political
Alignments. Princeton, NJ: Princeton University Press.
Shughart, William F., Jon D. Silverman, and Robert D. Tollison. 1995. Antitrust Enforcement
and Foreign Competition. In The Causes and Consequences of Antitrust:  The Public
Choice Perspective, edited by Fred S. McChesney and Willam F. Shughart II, 179–187.
Chicago: University of Chicago Press.
Slaughter, Anne-Marie. 2004. A New World Order. Princeton, NJ: Princeton University Press.
Stigler, George J. 1957. Perfect Competition, Historically Contemplated. Journal of Political
Economy 65 (1): 1–17.
Svetiev, Yane. 2010. Partial Formalization of the Regulatory Network. Unpublished manu-
script, Brooklyn Law School. Online at http://ssrn.com/abstract=1564890
Trebilcock, Michael J. 1996. Competition Policy and Trade Policy: Mediating the Interface.
Journal of World Trade 30 (4): 71–106.
Trebilcock, Michael J. 1991. Competition Policy, Trade Policy, and the Problem of Second Best.
In Canadian Competition Law and Policy at the Centenary, edited by R. S. Khemani and W. T.
Stanbury, 29–44. Halifax, NS: Institute of Research on Public Policy.
Trebilcock, Michael J., and Robert Howse. 2005. The Regulation of International Trade. 3rd ed.
New York: Routledge.
Trebilcock, Michael J., et al. 2002. The Law and Economics of Canadian Competition Policy.
Toronto: University of Toronto Press.
Voigt, Stefan. 2009. The Effects of Competition Policy on Development:  Cross-Country
Evidence Using Four New Indicators. Journal of Development Studies 45 (8): 1225–1248.
Waked, Dina I. 2010. Antitrust Enforcement in Developing Countries:  Reasons for
Enforcement & Non-Enforcement Using Resource-Based Evidence. Mimeograph, Harvard
Law School, July, http://ssrn.com/abstract=1638874).
Weiss, Leonhard W., ed. 1989. Concentration and Price. Cambridge, MA: MIT Press.
Williams, Mark D., and A. E. Rodriguez. 1995. Antitrust and Liberalization in Developing
Countries. International Trade Journal 9 (4): 495–518.
World Trade Organization (WTO). 2012. World Trade Report 2012: Trade and Public Policies—A
Closer Look at Non-Tariff Measures in the 21st Century. Geneva: World Trade Organization.
Zäch, Roger, and Carlos M. Correa, eds. 1999. Towards WTO Competition Rules: Key Issues and
Comments on the WTO Report (1998) on Trade and Competition. Berne: Stämpfli and The
Hagues: Kluwer Law International.
 Chapter 13

C onnected C ha nne l s
MNCs and production networks in global trade

Walte r Hatch, Je n n if e r Bair,

a nd G ünte r H e iduk

Introduction

Formal and informal business groups are now associated with a large and perhaps
increasing share of international trade. Lanz and Miroudot (2011, 12) estimate that
intrafirm shipments account for one-third of the world’s commodity exports, while
Borga and Zeile (2004, 2)  estimate that “vertical trade”—by which intermediate
goods produced in one country move through other countries, where value is added
via fabrication, processing, or assembly—accounts for roughly one-quarter of global
exports.
We face two major questions. The first has to do with the fact that, although there
is widespread recognition of the scale of this activity, there is only limited agreement
on how to explain it. Some, whom we might call liberals, tend to view it as the sum of
economic flows moving through largely disembodied or neutral entities. They use equi-
librium models to measure relative factor intensities of processes and relative factor
endowments of locations, giving only scant attention to the organizational context in
which transnational production is embedded. Others, whom we might call structur-
alists, prefer to view it as a coordinated activity. Instead of corporate vessels seeking
efficiencies, the business groups engaged in this transnational activity are political and
social agents striving to control assets and exercise influence.
A related question is this: How should we characterize these groups? The traditional
approach focuses narrowly on multinational corporations (MNCs), which are held
together by equity or ownership ties. A new approach, however, insists on the increas-
ingly important role of global production networks (GPNs), which are more loosely
affiliated. The latter is a broad category that includes, but is not limited to, the former.
234   Connected Channels

Let us address the second question first. Scholars have long been fascinated with
MNCs, far-flung conglomerates such as GE, Procter & Gamble, Nestlé, Monsanto, and BP
that engage in foreign direct investment (FDI), operating factories, stores, and hotels or
providing services such as legal advice and logistics through affiliated companies or sub-
sidiaries in another country. Dunning (1977), who developed the complex model upon
which much analysis of FDI is now based, suggests that MNCs seek to exploit ownership,
location, and internalization (OLI) advantages as they move overseas. Ownership has to
do with a firm’s legal title to particular assets, such as technical and managerial know-how.
Location has to do with the distinctive endowment of resources and factors in a particular
place that determines its comparative advantage for specialized production and export-
ing. And internalization has to do with a conglomerated firm’s ability to reduce transac-
tion costs by establishing intrafirm channels rather than relying on the external market.
More recently, however, scholars have looked beyond MNCs to analyze GPNs, which
have been described variously as “international fragmentation” (Jones and Kierzkowski
1990), “the slicing up of the value chain” (Krugman 1995), “the international disinte-
gration of production” (Feenstra 1998), “global production sharing” (Yeats 2001), and
“international outsourcing” (Milberg 2004). Unlike MNCs, they do not necessarily con-
sist of equity ties; instead, these networks tend to be held together informally by close
and durable business relationships between, for example, a machine assembler and dif-
ferent parts suppliers. Like some MNCs in manufacturing, however, they rely on a geo-
graphic dispersion of production activities that yields “vertical trade” and is guided by
both economic and political calculations. The former includes the technological capac-
ity of, and wage levels in, different business environments, while the latter includes tax
rates, export incentives, and many other policies in the home and host countries.
Global production networks may be more obscure simply because they tend to resist
measurement by our standard empirical tools for studying transnational activity. For
example, FDI data can tell us about the countries and industries in which parent firms
invest (or disinvest) each year, but they are not able to capture transactions such as off-
shore outsourcing that take place outside both established hierarchies and atomized mar-
kets (Williamson 1975). However, some sociologists have suggested that, to measure the
level of fragmentation or networking in a particular sector, we can use the ratio of global
trade to global value-added in that sector. Mahutga (2012, 12), building on an insight from
Feenstra (1998, 34), describes the ratio this way: “As the number of countries involved in
the production of a good increases, the extent to which value added in manufacturing is
double counted in trade will also increase for the industry in which the good is classified.”
This empirical discussion brings us back to the first, and perhaps meatier, theoretical
question in the literature on MNCs and GPNs: How do they operate? Do they function
according to liberal logic as neutral entities for mostly market-driven economic flows,
or in line with structuralist logic, as political and social organizations that seek control
and influence?
There is division even among economists. Mundell (1957), perhaps the pioneer in this
field and clearly a liberal, uses neoclassical analysis to portray FDI as the global move-
ment of physical capital in search of higher returns. But Hymer (1976), inspired by
 Walter Hatch, Jennifer Bair, and Günter Heiduk    235

Marxist theory, counters that MNC behavior is strategic and conforms better to models
of industrial organization than to traditional trade theories such as comparative advan-
tage. He argues that these transnational entities seek to monopolistically control their
assets as they invest abroad.
It should come as no surprise that sociologists have tended to view these transnational
business groups as social and strategic actors, not as disembodied vessels. A valuable con-
tribution has come from Gereffi (1994, 1995) and others (see Bair 2009), who focus on
global commodity chains (GCCs) and global value chains (GVCs).1 These scholars distin-
guish between producer- and buyer-driven chains. The former tend to be characterized
by higher levels of vertical integration and are orchestrated by major manufacturers, such
as Matsushita or Siemens, that use a far-flung division of labor, sourcing raw materials
and intermediate goods from certain countries, assembling final goods in another coun-
try, and then distributing those products to markets around the world. The latter, by con-
trast, are led by firms that organize both supplier and consumer markets for brand-name
goods produced by independent contractors to the precise specifications provided by
those organizers. Initially, buyer-driven commodity or value chains were associated with
labor-intensive consumer goods such as apparel, footwear, and toys, and were dominated
by retailers such as Walmart or merchandisers such as Nike. More recently, however,
higher-technology goods such as computers, video game consoles, and cell phones are
being produced via similar arrangements. Apple is perhaps the best-known “manufac-
turer without a factory,” contracting with original design manufacturers (ODMs) such as
Foxconn Technology Group of Taiwan to produce its popular iPads and iPhones.2
As Sturgeon (2009, 128–130) notes, the GCC/GVC approach draws our attention
to the governance of transnational production. Lead firms are able to coordinate the
chain’s activities, choosing and replacing suppliers, but also directing them to reduce
costs, adopt specific business strategies, and even invest in particular locations. Other
scholars have focused on different aspects of power, especially the technological capa-
bility of a firm or coalition of firms to move an entire industry in its favored direction.
Gawer and Cusumano (2002) have referred to this as “platform leadership,” while Kim
and Hart (2002) coined a more descriptive term, “Wintelism,” to describe the way in
which the coalition of Microsoft (producer of Windows) and Intel set the standard for
many years in the personal computer industry. Some contend that these forms of power
depend on a relatively high level of asset specificity, based on brands or technology.
Political scientists used to write extensively about the influence of transnational busi-
ness organizations, especially as they set up shop in “Third World” countries. For exam-
ple, Moran (1974) analyzed the relationship between MNCs and host states, outlining a
model of bargaining power that evolves over time. Gilpin (1973), by contrast, presented
MNCs as tools of powerful states in the global political economy. But with notable
exceptions such as Mosley (2011), political scientists today adopt a liberal line, tending
to treat MNCs and GPNs as disembodied vessels for capital flows rather than as orga-
nized actors. Li and Resnick (2003) document the different ways in which democratic
institutions in the host country can promote or retard FDI, while Jensen (2006) shows
how federal systems empower local elites and thereby may attract flows of investment.
236   Connected Channels

For some geographers and culturally minded political economists, this treatment of
MNCs and GPNs is troubling. Because they view transnational business groups as spa-
tially differentiated actors that shape, and are shaped by, the locations in which they oper-
ate, these scholars are aligned squarely with the structuralists. Henderson and colleagues
(2002) note that these transnational business groups are more or less embedded in spe-
cific places, whether they be nations or regions, and thus reflect and influence the social
dynamics (norms and institutions) of those places. A US MNC (e.g., General Motors)
operates differently from a Japanese MNC in the same industry (e.g., Toyota), even when
they invest in the same foreign market (e.g., Mexico). Likewise, transnational business
organizations may have different developmental consequences, depending on how they
are embedded in host environments. For example, a highly flexible network of footloose
manufacturers may not allow local actors—economic or political—to capture many of the
gains generated by that cross-border activity. Among other effects, wages may remain rel-
atively flat in this sector of the host economy. Katzenstein (2005) has extended this analy-
sis to regions, suggesting that government and business actors in East Asia, for example,
are motivated by distinctive values, compared to their counterparts in Europe. Both, how-
ever, are subject to a common, globalizing force he calls “the American Imperium.”
Liberals, meanwhile, suggest that mobile capital always poses an opportunity and
challenge for states and citizens, regardless of its organizational shape. States, they
argue, may attract FDI by creating export-processing zones and conversely by adopting
trade policies that are protectionist, which allow foreign manufacturers to jump trade
barriers and enjoy rents on domestic-oriented production. But the structure of invest-
ment should not matter. Likewise, citizens will prosper or suffer from investment flows,
but not based on the organization of transnational investors. For example, Rudra (2008)
offers an analysis of the impact of FDI on the economic standing of different social
classes in host countries, ignoring the structure of investing business groups. Freed
from this taxonomical condition, capital flow becomes a more easily measured variable.
This chapter looks at three regions of the world—East Asia, Latin (especially North)
America, and Europe—and summarizes the research on MNCs and GPNs in each.
For the sake of comparison, the discussion is limited to transnational automobile and
electronics operations in these three regions. The chapter concludes with some general
observations based on this comparative analysis.

MNCs and Production

Networks in East Asia

East Asia today enjoys perhaps the world’s thickest web of interconnected transnational
business organizations.3 Although scholars writing about this region have highlighted
the activities of both MNCs and GPNs, they increasingly tend to emphasize the latter.
Liberals note that economies in East Asia differ dramatically in their factor endowments
 Walter Hatch, Jennifer Bair, and Günter Heiduk    237

and add that states there tend to pursue relatively open trade policies—a set of condi-
tions that conspires to attract heavy flows of export-oriented FDI. Structuralists focus
instead on the strategies used by dominant firms to stitch those economies together,
turning this region into the “factory to the world” that manufactures a rich blend of
electronics, automobiles, machine tools, and other goods. Those products tend to be
assembled in Asian locations filled with cheap labor, using designs and sophisticated
inputs from Asian countries with high-tech capabilities and less sophisticated inputs
from Asian countries with more modest capabilities. Over time, the vertical division of
labor established and led by Japanese MNCs in the 1980s has turned into a more hor-
izontal and open model. Developing economies in the region, such as Indonesia and
Vietnam, have benefited as they have become more deeply incorporated into the thick
web of business ties.

History
Until the 1980s, Japanese industry had been reluctant to invest heavily overseas.
Managers of manufacturing firms, in particular, preferred to exploit social networks
in the domestic political economy, networks that included close personal relationships
with government officials; bankers, parts suppliers, and distributors; and long-term
employees. That reluctance crumbled after 1985, when the governments of Japan, the
United States, France, West Germany, and the United Kingdom agreed to a signifi-
cant appreciation of nondollar currencies, especially the yen. Export-oriented firms
in Japan, often supported by low-interest loans from their own government, increas-
ingly shifted production to South Korea, Taiwan, Thailand, Malaysia, Indonesia, and the
Philippines. Over time, and especially after Deng Xiaoping’s more aggressive economic
reforms began to take effect, manufacturing FDI flowed even more heavily from Japan
to coastal China.
Observers described an emerging “yen bloc” in East Asia as Japanese machine
assemblers, followed by their long-standing parts suppliers, general trading com-
panies, and main banks, set up operations in the region. Even government agencies,
especially JETRO (Japan External Trade Organization) and JICA (Japan International
Cooperation Agency), jumped into the networking business, advising both Japanese
manufacturers investing in East Asia and host country officials striving to turn their
economies into nodes in the evolving web of networks. In the automobile industry
these production networks manufactured parts for all of Southeast Asia and finished
vehicles for the domestic market in this subregion. Toyota, for example, set up large
assembly plants in Thailand, Malaysia, Indonesia, and the Philippines. But it also helped
set up specialized supply operations in different locations: Thailand became the base
for pressed parts, diesel engines, and electronics; Malaysia for steering gears and elec-
tronics; Indonesia for pressed parts and gasoline engines; and the Philippines for trans-
missions. Toyota sourced different parts from these specialized bases throughout the
region, as well as from suppliers in Japan, then sold finished vehicles in the same market
238   Connected Channels

in which they were assembled.4 By contrast, in the electronics industry Japanese pro-
duction networks often relied on a triangular pattern of trade, importing components
from Japan or Japanese suppliers in the region, assembling them at affiliates in East Asia,
and exporting finished goods to consumers in the United States and Europe. As of 1992,
for example, Sony’s Penang affiliate purchased half of its computer drive inputs from the
parent firm in Japan and another 40 percent from suppliers in Malaysia, Singapore, and
Thailand. But 95 percent of these “Southeast Asian” suppliers were Japanese affiliates
operating in the region. Sony exported all of the drives it assembled in Penang, and most
shipments went to the United States or Europe.5
In the 1990s Hatch and Yamamura (1996), as well as scholars at the Berkeley
Roundtable on the International Economy (BRIE), conducted empirical research on
East Asian production networks in the electronics industry. Ernst (1994, 1997) noted,
for example, that large parent firms based in Japan tended to control manufacturing
affiliates in the region, requiring them to use Japanese expatriates as managers and tech-
nicians, and to acquire inputs from long-standing domestic suppliers or their Asian
affiliates. This, he argued, made Japanese production networks less flexible and less open
than networks directed by MNCs based in other countries. Borrus (1997) extended this
finding, arguing that a primary reason for the revival of the US electronics industry in
the 1990s was the ability of American MNCs to tap into the relatively flexible and open
supply base of what he called the “China circle,” consisting of ethnic Chinese manu-
facturers on the mainland, in Taiwan, and throughout Southeast Asia. Thanks to their
access to this dynamic supply base, he contended, US producers gained a competitive
edge over their stodgier rivals from Japan.
At the same time, scholars noted that despite efforts by Japan to maintain its tech-
nological edge and avoid “industrial hollowing,” less advanced economies in East Asia
were steadily catching up. The region’s network-based macro-economy was developing
in a “flying geese pattern,” with industries in Japan, the lead goose, adopting successively
more sophisticated technologies and shedding others, which second-tier economies
then acquired, adapted, and eventually shed to third-tier economies. It quickly became
apparent that “Little Tigers” or “New Dragons” were chasing Japan and building impres-
sive, high-tech industries of their own.6
Scholars using the GCC or GVC approach (see, e.g., Feenstra and Hamilton 2006)
heralded the rise of what they called “demand-responsive economies” in Asia, especially
South Korea and Taiwan. These economies boasted a large number of manufacturers
that initially gained technical expertise by capitalizing on backward linkages, produc-
ing household appliances, footwear, and other simple goods for big retailers and mer-
chandisers in the United States, Europe, and Japan. Eventually, however, many of these
suppliers and OEM producers became MNCs, with their own production networks
churning out relatively sophisticated products. In South Korea, chaebol owners met the
demands of buyer-driven value chains primarily via internalization. They used their
well-greased connections to secure internal financing for expanded production in exist-
ing firms or new members of the corporate group. In Taiwan, by contrast, the smaller
 Walter Hatch, Jennifer Bair, and Günter Heiduk    239

suppliers satisfied demand by subcontracting with other, even nimbler, firms, many in
Southeast Asia.
With the dawn of a new millennium, some of these Korean and Taiwanese manufac-
turers emerged as technological rivals to the Japanese MNCs that had spun the very first
web of East Asian production networks. In semiconductors, for example, Samsung of
Korea acquired its own brand reputation, becoming nearly as well-regarded as Hitachi.
And in computers, for example, Acer of Taiwan closed some of the gap with Toshiba.
But many scholars note that smaller companies in other, less developed parts of Asia
also have tapped into the region’s sprawling web of production networks, helping their
economies grow without large “national champions.” Baldwin (2012) notes that heavily
populated developing economies like China and India may enjoy success in squeezing
technology from MNCs, but less populous developing economies might do better to
emulate Thailand, where dozens of suppliers now earn profits producing and exporting
automobile parts for Japanese, Korean, American, and European assemblers. Thailand,
he writes, is benefiting from an economic policy that encourages its manufacturers to
join a GVC rather than build one of their own: its “success in becoming the Detroit of
Southeast Asia shows supply-chain industrialization can work even without muscular
technology transfer policies” (34).
Gone, then, is the traditional model of vertically organized networks dominated by
Japanese MNCs. Scholars tend to agree that Asia today is home to a complex, criss-
crossing pattern of production networks that are at once more global and more regional
than before. They are more global in the sense that they increasingly include produc-
tion of intermediate goods for European and American MNCs, not just for Japanese
and Korean MNCs. But they also are more regional in the sense that the finished goods
produced by these networks increasingly are consumed in Asia. As a result of successful
supply-chain industrialization, consumers in the region are now wealthy enough to pur-
chase a larger and larger share of the fruits of network labor. Urban Chinese families, in
particular, have developed a strong appetite for high-technology goods. This means that
triangular trade is giving way to greater intraregional (or intra-Asian) trade.

The Impact of MNCs and Production Networks on East Asia

By spinning the first of what has become a dynamic web of regional production net-
works, Japanese MNCs and Japanese state agencies, collectively and individually,
enjoyed tremendous political clout in Asia, especially in Southeast Asia, from the
mid-1980s until roughly the end of the twentieth century. On the public side, agen-
cies such as MITI (Ministry of International Trade and Industry) dispatched officials
to Bangkok, Jakarta, Kuala Lumpur, and Manila, advising their counterparts on such
things as how to build supporting industries for automobile and electronics manufac-
turing. Concessionary yen loans helped finance “hard infrastructure” such as roads,
power lines, and new industrial estates, while Japanese government grants helped
240   Connected Channels

finance “soft infrastructure” such as new Japanese-style organizations reflecting coop-

eration between the state and industry.7 On the private side, Japanese MNCs became
leading suppliers of jobs, exports, and (to a more limited extent) new technology in
countries throughout Southeast Asia. It is safe to say that those developing coun-
tries came to depend increasingly on Japan, and that dependence manifested itself
in government policy and industry practice. In 1985, for example, Malaysia turned
to Mitsubishi for help in developing its first “national automobile,” the Proton Saga.
And in 1991, when Chrysler was trying to introduce the Jeep Cherokee into Thailand,
the Thai government treated the four-by-four SUV as a luxury car, which called for a
hefty excise tax, even though it had tagged Mitsubishi’s rival four-by-four as a pickup
truck, which qualified it for a far lower excise tax. American and European automo-
bile manufacturers also complained routinely that they could not persuade Japanese
parts producers, who dominated the supply base in Southeast Asia, to sign contracts
with them; they chose instead to deal exclusively with Japanese assemblers like Toyota
or Nissan.8
Before long, however, host governments cooled on Japan and its MNCs in the region.
They were especially frustrated by exclusionary behavior (hiring only Japanese man-
agers, using predominantly Japanese suppliers) that seemed to limit technology trans-
fer. The Thai Board of Investment, in a brochure designed to encourage non-Japanese
MNCs to invest in Thailand’s automotive and auto parts industries, expressed it this
way: “The Japanese have never been keen on transferring design and engineering exper-
tise to their Thai counterparts.”9
But the auto industry was unique. In other industries, alternative production net-
works emerged, attracting the interest of host governments and suppliers in Asia.
Nearly every economy in the region is now part of networks created by not only
Japanese but also Korean, Taiwanese, European, and American MNCs, capitalizing
on complementary labor skills and technology levels and creating a remarkable num-
ber of backward and forward linkages within and across industries. In the production
of electronic machinery, Hayakawa, Ji, and Obashi (2009, 9) find that the scale of an
industry in one country is positively correlated with that industry’s scale in neighbor-
ing countries. This suggests that Asian countries, by participating in regional networks
that break the manufacturing process into different pieces, are enjoying simultane-
ous expansion of output. Wang, Powers, and Wei (2009) demonstrate that developing
economies in Asia, led by China, have become far more integrated into the region’s web
of networks, especially in electronics. “In contrast, automobile production still mainly
involved Japan and Korea in 2000, with developing Asia just starting to show up in the
value chain” (32–33).
It is rather obvious today that the web of production networks in East Asia has dra-
matically reshaped host economies, from Dalian to Jakarta, while also influencing the
regional and global economies. These networks have dramatically expanded trade,
raising both levels of employment and income in areas most integrated into the web
through their different advantages in labor costs or technological capacity. A joint study
carried out by IDE-JETRO (Japan) and the WTO (2011, 68) suggests that cross-border
 Walter Hatch, Jennifer Bair, and Günter Heiduk    241

trade generated new employment opportunities for all of the ten Asian countries it stud-
ied, but that China, which has emerged as the new hub of the region’s networks, ben-
efited the most—by far. In 2000 intraregional trade generated fifty-six million new job
opportunities for China; in 2005 that number was eighty-nine million. What studies like
this often fail to note, however, is that the effects of trade generated by FDI and produc-
tion networks vary wildly within individual countries. For example, cities and towns
along China’s Pacific coast, compared to more rural and inland areas, have benefited
disproportionately from this paradoxical process of manufacturing fragmentation and
trade integration. The process has contributed to the widening income gap in various
Asian countries, but especially China.

The Impact of Public Policy on MNCs and

Production Networks in East Asia
Governments have influenced the behavior and organization of MNCs and produc-
tion networks in Asia. As Hatch (2004) notes, this is perhaps most obvious through
a quick comparison of the region’s automobile and electronics industries. Most host
governments have maintained relatively high tariff walls around auto assemblers. This
allowed Japanese automotive MNCs to hang onto market share in the region, but also
caused them to sell heavily in domestic rather than foreign markets. Baldwin (2012,
17–19) notes that Malaysia also used an import substitution policy to build its own auto-
mobile network, but fell short because its car manufacturers proved to be inefficient.
By contrast, host governments eliminated tariffs on electronics, making that industry
highly competitive. Japanese computer and semiconductor MNCs, for example, lost
their dominant position in the region’s markets, which became relatively efficient and
export-oriented.
Kimura and Obashi (2011, 13–14) argue that although countries in the region, espe-
cially members of the Association of Southeast Asian Nations (ASEAN), have tried to
liberalize trade both unilaterally and through bilateral and regional agreements, much
more could be done. MNCs and their production networks have prospered despite
limited progress in reducing tariffs. They note that some countries in the region have
emerged as major nodes in production networks by investing in shipping terminals and
other transportation infrastructure. Singapore is the region’s leading entrepot.

Summary
Although scholars offer competing explanations, they tend to agree that East Asia
has emerged as the world’s leading center of transnational business networks for
export-oriented manufacturing. Some attribute this outcome to local conditions, such
as an abundant supply of cheap labor and different levels of technological capacity; oth-
ers emphasize the organizational strategies used by firms enjoying market power.
242   Connected Channels

MNCs and Production Networks

in Latin America

The Americas have long been at the center of debates regarding the costs and benefits
of economic globalization, including the role of MNCs and production networks in
the development process. North America in particular is unique because it is home to
neighboring countries with starkly different levels of development and diverse resource
endowments. Today the region’s developing economies are scrambling not just to attract
foreign investment, but also to cultivate the manufacturing capabilities necessary to
upgrade beyond their traditional comparative advantage in lower-wage labor. This sec-
tion recounts the history of FDI in Latin America and examines the developmental impli-
cations of the region’s role in transnational production coordinated by foreign lead firms.

History
During the nineteenth and early twentieth centuries, European investors played an
important role in Latin America (Grosse 1989). Key sectors included railroads and
utilities, but foreign ownership was also significant in mining and oil in Mexico and
South America, and in agricultural commodities such as bananas and sugar in Central
America and the Caribbean (Chapman 2007; Ayala 1999). The political clout of foreign
capital was amply demonstrated during the opening decades of the twentieth century,
with multiple interventions by the US military in what came to be known as “banana
republics.”
Through the middle decades of the twentieth century, international investors repo-
sitioned themselves within changing political economies. Mexico is a case in point. In
1938 President Lazaro Cardenas nationalized the country’s oil industry in what was one
of the region’s earliest and most significant expropriations. By 1946 FDI from the United
States had plunged 50  percent (Gereffi and Evans 1981). When investment resumed,
much of it flowed into the manufacturing sector—a pattern that was repeated elsewhere,
as the adoption of import-substituting industrialization (ISI) strategies stimulated
market-seeking FDI into the region’s largest economies. Latin American governments
encouraged foreign investment in manufacturing because they believed that multina-
tionals could play an important role in developing the domestic industrial base prized
by the ISI model. Companies in strategic sectors such as autos were invited to establish
subsidiaries from which they could serve a growing (and protected) national market.
In return, they were expected to comply with policies, such as local content rules and
joint ownership requirements, designed to ensure technology transfer and other posi-
tive spillovers. By 1967 two-thirds of US FDI in Mexico and Brazil was in manufactur-
ing, with foreign firms generating between 30 and 40 percent of total industrial output
in these economies (Bulmer-Thomas 1994).
 Walter Hatch, Jennifer Bair, and Günter Heiduk    243

The debt crisis and the subsequent “lost decade” witnessed a sharp contraction of
FDI in Latin America, which fell from $7.5 billion in 1981—the year before Mexico’s
default—to $2.8 billion in 1986. Within a few years the region’s position as the leading
destination for foreign investment had eroded. In 1980 almost 70 percent of all FDI
flows to developing countries went to Latin America; by 1986, this percentage fell to less
than 20 percent (Paus 1989).
In the aftermath of the debt crisis, countries throughout the region implemented
reforms to liberalize trade and restore macroeconomic stability. Governments elimi-
nated restrictions that prohibited foreign investment in certain sectors or required joint
ventures with national firms. Privatization of state-owned assets created new ownership
opportunities for multinational corporations. Tariff barriers were reduced, bringing to
an end the import protection that foreign and domestic enterprises had enjoyed. Many
of the policies designed to promote linkages between multinationals and local firms
were eliminated. The far-reaching process of economic restructuring and liberalization
carried out during the 1980s and early 1990s replaced the earlier emphasis on domestic
industrialization with an outward-looking strategy of global economic integration and
export promotion. This shift in development strategy, from an import-substitution to
an export-led model, corresponded to a shift in the nature of FDI, from “horizontal”/
domestic market to “vertical”/foreign market, as subsidiaries that had been established
to serve domestic markets were reconfigured as links in international value chains
(Dunning and Narula 2004).

The Impact of Reform and Regional Integration

on MNCs and GPNs in North America
Following the economic reforms of the late 1980s, governments in North America
turned their attention to the promotion of preferential trade agreements to stimulate
investment and intraregional trade flows. The initial impetus for the North American
Free Trade Agreement (NAFTA) originated with Mexican President Carlos Salinas
de Gortari, who hoped that a free trade agreement with its northern neighbors would
assure foreign investors of Mexico’s commitment to reform. Signed by the leaders of
Canada, the United States, and Mexico in 1993, NAFTA went into effect on January 1
of the following year. The agreement proved successful in stimulating increased invest-
ment in Mexico; between 1994 and 2002, an annual average of $13 billion in FDI flowed
into Mexico, as compared with average yearly inflows of $4.5 billion in the eight years
prior to the agreement.
Fueled by increased FDI, Mexico’s manufactured exports soared in the immediate
post-NAFTA period. Growth in clothing exports, in particular, caused significant con-
cern among the countries of Central America and the Caribbean, whose less-diversified
industrial sectors depend heavily on the garment trade (Heron 2003). These nations
lobbied the US government for an agreement giving them “NAFTA parity”; their efforts
resulted in the passage of the Caribbean Basin Trade Partnership Act in 2000 and in
244   Connected Channels

2004 the signing of the Central American Free Trade Agreement (CAFTA) among
the United States, the Dominican Republic, El Salvador, Guatemala, Honduras, and
Nicaragua (Bair and Peters 2006).

Implications of MNCs and GPNs in North America

The passage of NAFTA and then CAFTA triggered a new round of debate about the
degree to which the increased investment and trade stimulated by these agreements
would yield more enduring benefits than those associated with the export-processing
model typified by Mexico’s maquiladoras. Unlike in East Asia, where numerous econo-
mies managed to parlay “triangular trade” into sustained growth and industrial upgrad-
ing during the post–World War II period, cross-border production networks between
US firms and assembly plants in Mexico and the Caribbean had produced mostly disap-
pointing results through the 1980s (Sklair 1993). However, there is some evidence that
this is changing since the establishment of NAFTA, at least in Mexico’s two largest man-
ufacturing sectors: autos and electronics.
As the global auto industry has become organized into regional production blocs,
Mexico has emerged as a central manufacturing hub for the North American market. In
1994 the country hosted eight auto assembly plants and produced fewer than one million
vehicles per year. By 2010 the number of plants had more than doubled, to twenty-two,
and output reached 2.2 million units. More important, domestic suppliers are playing
a growing, if still relatively small, role in these production networks (Contreras and
Carrillo 2012). The evolution of the electronics industry has been more uneven. After
a sharp contraction in the early 2000s, Mexico’s electronics sector upgraded to more
sophisticated, higher value-added manufacturing, including larger products (such
as flat-screen TVs) for which proximity to the US confers particular advantage. Yet in
Mexico’s most dynamic electronics cluster, located near Guadalajara, upgrading has not
been accompanied by expanded employment, and national companies have a minimal
presence (Sturgeon and Kawakami 2010).
The effect of FDI on productivity in Mexico has been found to be positive (Waldkirch
2010), though at least one study showed that including reverse FDI flows (i.e., repa-
triated profits and dividends) in the net FDI capital variable attenuated the effect
(Ramirez 2006). Aitken, Harrison, and Lipsey (1996) concluded that FDI had a posi-
tive effect on wages in Mexico, but Waldkirch (2010), using data from the post-NAFTA
period, found no positive effect on overall wages and a marginal negative effect on
the wages of unskilled workers. Beyond research on FDI, scholars are asking how the
incorporation of subcontractors and suppliers into nonequity production networks is
affecting the region’s developing economies. Some have expressed concern that inso-
far as countries are hosting narrower “slices” of a disaggregated value chain, they may
find opportunities for positive spillover and learning effects minimized, especially
if these “slivers of specialized activity” are of the low value-added sort (Buckley and
Ghauri 2004).
 Walter Hatch, Jennifer Bair, and Günter Heiduk    245

Although MNCs based in the global North continue to play the pivotal role in most
of the GPNs crisscrossing Latin America, intraregional investment is growing as a
percentage of all FDI flows. In 2012, 14 percent of all FDI in Latin America originated
from another Latin American country. This trend reflects the emergence of indigenous
multinational firms within the region, referred to as “multilatinas” (Santiso 2008)
or “trans-Latinas” (ECLAC 2012). Mexico is home to the largest number of domestic
MNCs, while Brazil has the largest FDI stock abroad. Latin America’s largest multina-
tionals have emerged from a diverse set of sectors, including construction, mining, tele-
communications, and food and beverages.

Summary
The manufacturing sector absorbs about half of the FDI flowing into the economies
of Mexico, Central America, and the Caribbean. In contrast to South America, which
receives significant investment from European countries in primary sectors, the devel-
oping economies of the northern subregion continue to depend heavily on the United
States, both as a source of manufacturing FDI and as a market for their exports. Given
the role that these countries play in production networks coordinated by foreign lead
firms, competition with Asian exporters in global industries is increasingly seen as a
prerequisite for successful development.

MNCs and Production

Networks in Europe

The fall of the Iron Curtain in late 1989 has resurrected Europe’s prewar economic land-
scape, consisting of western, central, and eastern European countries. The fast-emerging
east-west integration built on centuries-old cultural proximity gained momentum from
the economic underdevelopment of countries in Central and Eastern Europe (CEE)
due to their isolation from growing markets in the west. Unlike in East Asia and Latin
America, the debate over MNCs and GPNs in this region appears to revolve around
their role in deepening the European Union (EU). Scholars have carefully analyzed the
transformational and integrative impact of production networks in Central and Eastern
Europe. Before 1989, the industrialized CEE differed from Asia and Latin America in
that it did not offer a large low-skilled, cheap labor force that could be used to absorb
low-end production processes Today, however, this area not only offers opportuni-
ties for MNCs to establish production networks, but also motivates small and medium
enterprises (SMEs) to cooperate with suppliers in CEE. The common cultural tradition
and the intensive and deep prewar ties create a welcoming environment that clearly dif-
fers from the tensions that US and Japanese MNCs often face in their regions.
246   Connected Channels

From a theoretical point of view, the focus of the literature is on trade and FDI as
integrative vehicles. The major concerns about Asia and Latin America—the difference
between horizontal and vertical intra-industry trade (IIT), efficiency-seeking or stra-
tegic FDI, and the political clout of MNCs—do not appear to dominate the debate over
Europe.

History
Enlargement of the EU in 2004 and 2007 propelled western European manufacturing
companies to allocate production of parts and components to the new member states
in CEE because they hoped to exploit comparative advantages. Although MNCs did not
prune production networks in peripheral economies along the Mediterranean Sea, FDI
flowed even more heavily to CEE countries. Only Greece failed to attract higher levels
of FDI. According to Kokkinou and Psycharis (2004), this was a result of disincentives
associated with the Greek government, especially its complicated tax system and its rel-
atively corrupt bureaucracy.
The 1989 miracle offered western European companies new opportunities to invest
and produce in a large and diversified region. In 1997, well ahead of eastern enlarge-
ment, the EU signaled clearly in its “Agenda 2000” that it would follow through on the
pre-accession strategy by negotiating and implementing accession agreements with the
applicant countries in CEE.
From the middle of the 1990s, progress in transition (“transition factor”), along with
the advantages of pre-accession agreements with the EU (“EU factor”), combined to
induce western companies to explore local capabilities and/or to use CEE as a regional
supply base. There is evidence that MNCs predominantly create intrafirm production
networks via FDI and extend them by subcontracting with local companies. SMEs
on the German-Polish and German-Czech border have established contract-based
regional production networks that are often embedded in the institutional concept
of Euroregions. Grix and Knowles (2002) stress the importance of the “networks of
reciprocal trust” as a basis for cross-border businesses. Due to the strong prewar ties
Germany had with countries on its eastern border, it is unsurprising that German com-
panies are the largest investors in CEE.
A number of studies (e.g., Damijan et al. 2013) confirm that FDI promoted the indus-
trial restructuring and export growth of CEE economies. At the end of the 1990s around
80 percent of Hungary’s exports and 40 percent of Poland’s exports to the EU originated
from foreign firms (Kaminski 2001, 33). Evidence suggests that foreign firms increas-
ingly integrated local CEE producers into their regional production networks.
In addition to the “transition factor” and the “EU factor,” scholars have identified
gravity determinants, especially proximity to Germany, as the major forces driving FDI
in and manufacturing-related trade with CEE in the 1990s (see, e.g., Kaminski 2001).
Mode of privatization, changes in the institutional environment, macroeconomic stabil-
ity, success in reducing sovereign debt, and speed of opening-up seem to be the major
 Walter Hatch, Jennifer Bair, and Günter Heiduk    247

factors explaining the differences in FDI inflows to the new members. Multinationals
engaged in manufacturing were attracted to CEE by the large pool of skilled workers,
low unit labor costs, and existing foundation of capital-intensive production.
Resmini’s (2000) analysis of FDI flows to this subregion shows a pattern that var-
ies according to industry. Market-driven FDI dominates in traditional sectors; tran-
sition success, as well as proximity to Western Europe, attracts FDI in science-based
and capital-intensive sectors; and wage differentials affect FDI in scale-intensive sec-
tors. Overall, Güngör and Binatli (2010, 26) conclude that efficiency-seeking motives
prevail across the region rather than market-seeking and resource-seeking motives.
Approaches that go beyond the advantage- and gravity-based explanation of production
networks improve the explanatory value of western European companies’ engagement
in CEE. According to Altomonte (1998), institutional climate and economic uncertainty
are factors that significantly influence the decision to reach out to transition countries as
well as the location, mode of entry, scope, and type of relocated activities.

The Impact of New West-East Production

Networks on CEE Countries
Studies by Mateev and Tsekov (2013), as well as others, reveal the following about MNCs’
FDI in CEE. Germany is the leading source of FDI, Hungary the leading destination,
and the automotive industry the leading investor. FDI stock, as of 2010, had reached
more than 50 percent of GDP (Pojar, 2012, 23). Last but not least, FDI positively contrib-
uted to export restructuring via MNCs’ production networks (see, e.g., Damijan et al.
2013). Spillover effects to domestic firms are not generally confirmed.
Despite limits on measuring cross-border production networks, the considerable
west-east FDI flows, accompanied by a significant increase in IIT in intermediate and
equipment goods, especially in machinery, electrical, and automotive industries, sug-
gest a vertical specialization pattern (Damijan et al. 2013). Production networks allow
MNCs to improve their market competitiveness by exploiting CEE countries’ produc-
tion competitiveness.
A simplified view of intra-EU west-east trade generates the following conclusions. First,
in the early 1990s low-tech industries shifted labor-intensive production to CEE coun-
tries, mostly for assembling. Second, at the end of the 1990s those CEE countries became
net exporters of parts in capital- and skill-intensive industries. Third, the quality of their
exports has increased markedly, reaching the level of western EU industries in recent
years. Fourth, since the mid-1990s FDI has been an important driver of the trade special-
ization pattern in CEE. Ambroziak’s (2012, 20) empirical test (which excludes Bulgaria
and Romania) confirms that the estimated impact of FDI on all types of IIT was positive
and highly significant. Likewise, Kutan and Vuksic (2007) argue that FDI has increased
production capacity in CEE countries, resulting in increasing export supply capacity.
Kravtsova and Radosevic (2012) suggest that FDI in CEE countries has generated
positive spillovers for domestic firms, helping them increase their competitiveness.
248   Connected Channels

Negative effects would occur if foreign firms “steal the market” of domestic firms.
Kosová (2010), using data from the Czech Republic, shows evidence of both possible
outcomes.
The effect of FDI on employment in CEE countries seems ambiguous. Due to the rapid
evolution from low-tech to medium- and high-tech FDI, the originally large demand for
low-skilled workers was replaced by a relatively small demand for high-skilled workers.
Empirical studies suggest that the effects of FDI on employment in CEE generally have
been modest. FDI inflows in manufacturing may have a negative effect, indirectly, on
local suppliers, but this may be outweighed by the positive effect of direct employment
by MNCs (Hunya and Geishecker 2005). Studies on the “feedback” effect of this FDI
on labor markets in western Europe do not show a general pattern; in fact, results dif-
fer among countries, industries, and the skill level of workers (see, e.g., Castellani et al.
2008). Empirical studies on the effects of FDI and trade on wages in CEE countries show
different results depending on the country, industry, type of FDI, and trade, as well as
on the relation between imports and exports (Egger and Stehrer 2001; Stehrer and Wörz
2006).

EU’s Multilevel Governance and Its

Impact on European MNCs
It is difficult to evaluate the impact of European MNCs on the policy-making process in
the European Union, just as it is difficult to evaluate the impact of EU policy making on
MNC practice. In both cases one has access to only limited evidence and must resort to
speculation.
European production networks are affected most critically by two sets of poli-
cies: external trade policy, which is laid out and enforced entirely at the level of the
EU, and single market policies, which are largely determined by the Council of the EU
and the European Commission but implemented by member states. The power of the
Commission is enhanced by its administrative expertise, while the power of the Council
is weakened by the constraints of the complex system of qualified majority voting and
the resulting need to secure compromise. In highly specialized cases (e.g., antidump-
ing policy or technical regulations), the Commission turns to industry for advice. It is
well known that the Commission would like to adopt industrial policies designed to
increase the competitiveness of core industries. This likely would strengthen the power
of national and European trade associations. In addition, fifty of Europe’s top corpo-
rate executives pool their individual knowledge and company-backed power in the
European Round Table of Industrialists (ERT), lobbying policy makers on everything
from antitrust to intellectual property rights. The ERT may represent a distinctively
EU style of lobbying, which Woll (2012, 194)  calls informational, constructive, and
consensus-oriented, compared to the aggressive advocacy approach used in the United
States. MNCs with headquarters in large member states and production networks
that are spread over many smaller/medium-sized member states may benefit from
 Walter Hatch, Jennifer Bair, and Günter Heiduk    249

the Council’s voting system when lobbying for their interests (Coen 2007). The higher
the number of involved national governments, the higher the potential power of these
MNCs. But even in the case of common interests, large member states are often forced to
logroll and package deals. The existing literature does not clarify whether trade associa-
tions and business groups drive the Commission into a more protectionist or more open
direction, which enhances competition (Vassalos 2012).

Summary
The thick west-east production networks in Europe emerged from the political change
in 1989 and the following integration of CEE countries into the EU. This allows western
European MNCs to make full use of CEE’s cost and proximity advantages. The deep ver-
tical specialization is documented by high shares of intermediate and equipment goods
in the EU’s west-east IIT. Due to the EU’s multilevel governance system and its complex
decision-making processes, MNCs’ power to shape external trade policy as well as com-
mon market policies is difficult to evaluate and needs further industry-specific research.

Conclusion

Although they agree that a large and perhaps increasing share of global trade occurs
inside corporate channels, scholars are divided on how to explain this phenomenon.
Some attribute it to economic flows that are pushed and pulled by market forces; others
credit corporate strategies. At the same time, scholars differ over how to characterize the
business groups engaged in this cross-border activity. For manufacturing investment,
the traditional approach has described them as MNCs that either circumvent trade bar-
riers by establishing offshore but domestic market-oriented operations or capitalize on
host country incentives by assembling parts in foreign enclaves and shipping finished
goods back home. More recently, students of FDI have highlighted a new model: the
GPN that slices up the manufacturing process into separate pieces located in different
countries based on the presumed technological capacity of host economies.
This chapter has shown that relatively complex networks are emerging in the three lead-
ing regions of the global economy, reshaping trade patterns and development prospects in
Asia, Latin (especially North) America, and Europe. These networks tend to be dominated
by Japanese or Korean, US, and German firms, either manufacturers or retailers, which
have established close and long-standing ties with suppliers and contractors in neighbor-
ing countries. Lead firms promote and exploit government policies, especially preferen-
tial trade regimes at the regional level, while also using novel business practices that allow
them to capture the gains of specialization via segmentation. For example, automobile
assemblers—whether they are headquartered in Nagoya, Detroit, or Stuttgart—are out-
sourcing more and more production to nominally independent but often foreign affiliates
250   Connected Channels

in regional hubs like Poland, Mexico, and Thailand. But our case studies also revealed dif-
ferences in the nature and depth of networking in these three regions.
Asia is home to the most extensive production networks in the world today. Japanese
multinationals initiated this trend in the 1980s, but Korean and Taiwanese producers
soon got into the act. China then emerged as a low-cost production and export base
for MNCs from around the world. Other economies in the region—including India,
Bangladesh, and Vietnam—are now becoming alternative sites for low-cost manufac-
turing. Even though formal-legal institutions in Asia, especially Northeast Asia, are
relatively weak, GPNs are well-established here and are characterized by massive spa-
tial and technological differentiation. This suggests that transnational firms, rather than
states, are driving the process of integration in Asia.
In Latin America, US producers have established low-cost bases in Mexico and dif-
ferent parts of Central America, often manufacturing “reverse imports” for the US mar-
ket. NAFTA and CAFTA have contributed to the US-led process of production sharing,
though the advantages these regimes provide may not be sufficient to offset pressures
from Asian competitors. Further south and east, in places like Brazil and Argentina, US
producers are less dominant; European and Latin American investors are more active
here, especially in this subregion’s primary industries.
The eastward expansion of the European Union has, by contrast, created a far-flung
single market that capitalizes on the lower costs, especially lower wages, in the new
members’ economies. Intermediate and equipment goods account for a significant share
of the EU’s west-east IIT, indicating the progress of vertical specialization.

Notes
1. In the 1990s, “value” upstaged “commodity” as the unit of analysis for many scholars in
this school. They argued that it covered a broader range of transnational business activity.
2. ODMs used to be referred to as OEMs (original equipment manufacturers).
3. See Aminian, Fung, and Ng (2009) for a comparison of East Asia and Latin America. See
Kimura, Takahashi and Hayakawa (2007) for a comparison of East Asia and Europe.
4. See Hatch and Yamamura (1996, 26).
5. See Hatch and Yamamura (1996, 160).
6. This discussion relies heavily on Hatch (2010).
7. See Hatch and Yamamura (1996, 130–145).
8. See Yamamura and Hatch (1997).
9. Thai Board of Investment, Investment Opportunities Study: Automotive and Autoparts
Industries in Thailand, 1995.

References
Aitken, B., A. Harrison, and R. Lipsey. 1996. Wages and Foreign Ownership: A Comparative
Study of Mexico, Venezuela, and the United States. Journal of International Economics 40
(3–4): 345–371.
 Walter Hatch, Jennifer Bair, and Günter Heiduk    251

Altomonte, Carlo. 1998. FDI in the CEECs and the Theory of Real Options:  An Empirical
Assessment. LICOS Discussion Paper, 76/1998, KU Leuwen.
Ambroziak, Lukasz. 2012. FDI and Intra-Industry Trade:  Theory and Empirical Evidence
from the Visegrad Countries. International Journal of Economics and Business Research 4
(1–2): 180–198.
Aminian, Nathalie, K. Fung, and Francis Ng. 2009. A Comparative Analysis of Trade and
Economic Integration in East Asia and Latin America. Economic Change and Restructuring
42 (1): 105–137.
Ayala, C. 1999. American Sugar Kingdom: The Plantation Economy of the Spanish Caribbean,
1898–1934. Chapel Hill: University of North Carolina Press.
Bair, J., and E. D. Peters. 2006. Commodity Chains and Endogenous Growth:  Export
Dynamism and Development in Mexico and Honduras. World Development 34 (2): 203–221.
Bair, Jennifer, ed. 2009. Frontiers of Commodity Chain Research. Stanford, CA:  Stanford
University Press.
Baldwin, Robert. 2012. Trade and Industrialization after Globalization’s 2nd Unbundling: How
Building and Joining a Supply Chain Are Different and Why It Matters. NBER Working Paper
No. 17716. http://siteresources.worldbank.org/INTRANETTRADE/Resources/Baldwin_
NBER_Working_Paper_17716.pdf.
Borga, Maria, and William J. Zeile. 2004. International Fragmentation of Production and the
Intrafirm Trade of U.S. Multinational Companies. Bureau of Economic Analysis, Commerce
Department, Working Paper No. 2004-02.
Borrus, Michael. 1997. Left for Dead:  Asian Production Networks and the Revival of U.S.
Electronics. BRIE Working Paper No. 100. http://brie.berkeley.edu/publications/WP100.
pdf.
Buckley, P., and P. Ghauri. 2004. Globalisation, Economic Geography and the Strategy of
Multinational Enterprises. Journal of International Business Studies 35 (2): 81–98.
Bulmer-Thomas, V. 1994. The Economic History of Latin America since Independence.
New York: Cambridge University Press.
Castellani, Davide, Ilaria Mariotti, and Lucia Piscitello. 2008. The Impact of Outward
Investment on Parent Company’s Employment and Skill Composition. Evidence from the
Italian Case. Structural Change and Economic Dynamics 19: 81–94.
Chapman, P. 2007. Bananas:  How the United Fruit Company Shaped the World. New  York:
Cannongate.
Coen, David. 2007. Lobbying in the European Union. Paper requested by the European
Parliament’s Committee on Constitutional Affairs, European Parliament, PE 393.266,
Brussels, Belgium.
Contreras, O., and J. Carrillo. 2012. Local Entrepreneurship within Global Value Chains: A Case
Study in the Mexican Automotive Industry. World Development 40 (5): 1013–1023.
Damijan, Joze, Crt Kostevc, and Matija Rojek. 2013. Global Supply Chains at Work in Central
and Eastern European Countries: Impact of FDI on Export Restructuring and Productivity
Growth. Vives Discussion Paper No. 37, KU Leuwen.
Dunning, J., and R. Narula. 2004. Multinationals and Competitiveness: A New Agenda.
Cheltenham, UK: Edward Elgar.
Dunning, John H. 1977. Trade, Location of Economic Activity and the MNE:  A  Search
for an Eclectic Approach. In The International Allocation of Economic Activity, edited
by Bertil Ohlin, Per-Ove Hesselborn, and Per Magnus Wijkman, pp. 395-418. London:
Macmillan.
252   Connected Channels

Economic Commission for Latin America and the Caribbean (ECLAC). 2012. Foreign Direct
Investment in Latin America and the Caribbean. Santiago, Chile: Economic Commission for
Latin America and the Caribbean.
Egger, Peter, and Robert Stehrer. 2001. International Outsourcing and the Skill-specific Wage
Bill in Easter Europe. The Vienna Institute for International Economic Studies, Research
Reports No. 17.
Ernst, Dieter. 1994. Carriers of Regionalization:  The East Asian Production Networks of
Japanese Electronics Firms. BRIE Working Paper No. 73. http://brie.berkeley.edu/publica-
tions/wp%2073.pdf.
Ernst, Dieter. 1997. Partners for the China Circle? The Asian Production Networks of Japanese
Electronics Firms. BRIE Working Paper No. 91. http://brie.berkeley.edu/publications/
wp%2091.pdf.
Feenstra, Robert C. 1998. Integration of Trade and Disintegration of Production in the Global
Economy. Journal of Economic Perspectives 12 (4): 31–50.
Feenstra, Robert C., and Gary Hamilton. 2006. Emergent Economies, Divergent Paths: Economic
Organization and International Trade in South Korea and Taiwan. Cambridge, UK:
Cambridge University Press.
Gawer, Annabell, and Michael A. Cusumano. 2002. Platform Leadership: How Intel, Microsoft
and Cisco Drive Industry Innovation. Cambridge, MA: Harvard Business School Press.
Gereffi, Gary. 1995. Global Production Systems and Third World Development. In Global
Change, Regional Response:  The New International Context of Development, edited by
Barbara Stallings, 100–142. Cambridge, UK: Cambridge University Press.
Gereffi, Gary. 1994. The Organization of Buyer-Driven Global Commodity Chains:  How
US. Retailers Shape Overseas Production Networks. In Commodity Chains and Global
Capitalism, by Gary Gereffi and Miguel Korzeniewicz, 95–122. Westport, CT: Praeger.
Gereffi, G., and P. Evans. 1981. Transnational Corporations, Dependent Development, and
State Policy in the Semiperiphery: A Comparison of Brazil and Mexico. Latin American
Research Review 16 (3): 31–64.
Gilpin, Robert. 1973. US Power and the Multinational Corporation: The Political Economy of
Foreign Direct Investment. New York: Basic Books.
Grix, Jonathan, and Vanda Knowles. 2002. The Euroregion as a Social Capital Maximizer: the
German-Polish Euroregion Pro-Europa Viandrina. Regional and Federal Studies 12 (4):
154–176.
Grosse, R. 1989. Multinationals in Latin America. London: Routledge.
Güngör, Hakan, and Ayla O. Binatli. 2010. The Effect of European Accession Prospects on
Foreign Direct Investment Flows. Izmir University of Economics, Izmir, Working Papers on
Economics No. 10/06.
Hatch, Walter F. 2010. Asia’s Flying Geese:  How Regionalization Shapes Japan. Ithaca,
NY: Cornell University Press.
Hatch, Walter. 2004. When Strong Ties Fail:  US-Japan Manufacturing Rivalry in Asia. In
Beyond Bilateralism: US-Japan Relations in the New Asia-Pacific, edited by Ellis Krauss and
T. J. Pempel, pp. 154-175, Stanford, CA: Stanford University Press.
Hatch, Walter, and Kozo Yamamura. 1996. Asia in Japan’s Embrace:  Building a Regional
Production Alliance. Cambridge, UK: Cambridge University Press.
Hayakawa, Kazunobu, Zheng Ji, and Ayako Obashi. 2009. Agglomeration versus
Fragmentation: A Comparison of East Asia and Europe. IDE Discussion Paper No. 212.
 Walter Hatch, Jennifer Bair, and Günter Heiduk    253

Henderson, Jeffrey, Peter Dicken, Martin Hess, Neil Coe, and Henry Wai-Chung Yeung.
2002. Global Production Networks and the Analysis of Economic Development. Review of
International Political Economy 9 (3): 436–464.
Heron, T. 2003. The U.S.-Caribbean Apparel Connection and the Politics of NAFTA Parity.
Third World Quarterly 23 (4): 753–767.
Hunya, Gabor, and Ingo Geishecker. 2005. Employment Effects of Foreign Direct Investment
in Central and Eastern Europe. The Vienna Institute for International Economic Studies,
Research Reports No. 321.
Hymer, Stephen H. 1976. The International Operations of National Firms: A Study of Direct
Foreign Investment. Cambridge, MA: MIT Press. (1960 PhD diss. published posthumously.)
IDE-JETRO and WTO. 2011. Trade Patterns and Global Value Chains in East Asia: From Trade
in Goods to Trade in Tasks. Geneva: World Trade Organization Centre.
Katzenstein, Peter J. 2005. A World of Regions: Asia and Europe in the American Imperium.
Ithaca, NY: Cornell University Press.
Kim, Sangbae, and Jeffrey A. Hart. 2002. The Global Political Economy of Wintelism: A New
Mode of Power and Governance in the Global Computer Industry. In Information
Technologies and Global Politics: The Changing Scope of Power and Governance. edited by
James N. Rosenau and J. P. Singh, 143–168. Albany: State University of New York Press.
Kimura, Fukunari, and Ayako Obashi. 2011. Production Networks in East Asia:  What We
Know So Far. ADBI Working Paper 320. Tokyo: Asian Development Bank Institute. http://
www.adbi.org/working-paper/2011/11/11/4792.production.networks.east.asia/.
Kimura, Fukunari, Yuya Takahashi, and Kazunobu Hayakawa. 2007. Fragmentation and Parts
and Components Trade:  Comparison between East Asia and Europe. North American
Journal of Economics and Finance 18 (1): 23–40.
Krugman, Paul R. 1995. Growing World Trade. Brookings Papers on Economic Activity No. 1.
Jensen, Nathan M. 2006. Nation-States and the Multinational Corporation:  The Political
Economy of Foreign Direct Investment. Princeton, NJ: Princeton University Press.
Jones, Ronald W., and Henryk Kierzkowski. 1990. The Role of Services in Production and
International Trade: A Theoretical Framework. In The Political Economy of International
Trade: Essays in Honor of Robert E. Baldwin, edited by Ronald Jones and Anne Krueger,
pp. 31–48. Oxford: Basil Blackwell.
Kaminski, Bartlomiej. 2001. How Accession to the European Union Has Affected External
Trade and Foreign Direct Investment in Central European Economies. World Bank
Working Papers 2578. Washington, DC.
Kokkinou, Aikaterini, and Psycharis, Ioannis. 2004. Foreign Direct Investment, Regional Incentives
and Regional Attractiveness in Greece. Department of Planning and Regional Development,
School of Engineering, University of Thessaly, Volos, Greece, Discussion Paper Series No. 10(11).
http://www.prd.uth.gr/uploads/discussion_papers/2004/uth-prd-dp-2004-11_en.pdf.
Kosová, Renáta. 2010. Do Foreign Firms Crowd Out Domestic Firms? Evidence from the
Czech Republic. Review of Economics and Statistics 92 (4): 861–881.
Kravtsova, Victoria, and Slavo Radosevic. 2012. Are Systems of Innovation in Eastern Europe
Efficient? Economic Systems 36 (1): 109–126.
Kutan, Ali M., and Goran Vuksic. 2007. Foreign Direct Investment and Export
Performance: Empirical Evidence. Comparative Economic Studies 49: 430–445.
Lanz, R., and S. Miroudot. 2011. Intra-Firm Trade:  Patterns, Determinants and Policy
Implications. OECD Trade Policy Papers No. 114. OECD Publishing.
254   Connected Channels

Li, Quan, and Adam Resnick. 2003. Reversal of Fortunes: Democratic Institutions and Foreign
Direct Investment to Developing Countries. International Organization 57 (1): 175–211.
Mahutga, Matthew C. 2012. When Do Value Chains Go Global? A Theory of the Spatialization
of Global Value Chains. Global Networks 12 (1): 1–21.
Mateev, Miroslav, and Ilya Tsekov. 2013. Do Central and Eastern European Countries Possess
FDI Advantages to More Developed Western Countries? Paper presented at the Annual
Meeting of the Midwest Finance Association, Chicago, March 13–16. http://bmabg.s801.
sureserver.com/assets/var/docs/papers/do-cee-FDI.pdf.
Milberg, William S. 2004. The Changing Structure of Trade Linked to Global Production
Systems: What are the Policy Implications? International Labour Review 143 (1–2): 45–90.
Moran, Theodore H. 1974. Multinational Corporations and the Politics of Dependence: Copper in
Chile. Princeton, NJ: Princeton University Press.
Mosley, Layna. 2011. Labor Rights and Multinational Production. New  York:  Cambridge
University Press.
Mundell, Robert. 1957. International Trade and Factor Mobility. American Economic Review
47 (3): 321–335.
Paus, E. 1989. Direct Foreign Investment and Economic Development in Latin
America: Perspectives for the Future. Journal of Latin American Studies 21 (2): 221–239.
Pojar, Simona-Valeria. 2012. The Influence of Foreign Direct Investment on Firms’ Performance
in Central and Easter in European Countries. Master’s thesis, Aarhus University. http://pure.
au.dk/portal-asb-student/files/44523221/Master_Thesis.pdf.
Ramirez, M. 2006. Is Foreign Direct Investment Beneficial for Mexico? World Development
34 (5): 802–817.
Resmini, Laura. 2000. The Determinants of Foreign Direct Investment in the CEECs—New
Evidence from Sectoral Patterns. Economics of Transition 8 (3): 665–689.
Rudra, Nita. 2008. Globalization and the Race to the Bottom in Developing Countries:  Who
Really Gets Hurt? New York: Cambridge University Press.
Santiso, J. 2008. The Emergence of Latin Multinationals. CEPAL Review 95: 7–30.
Sklair, L. 1993. Assembling for Development: The Maquila Industry in Mexico and the United
States. San Diego, CA: Center for U.S.-Mexican Studies.
Stehrer, Robert, and Julia Wörtz. 2006. Attract FDI!—A Universal Global Rule? Empirical
Evidence for Europe and Asia. Paper presented at the 3rd Conference on RTN “Trade,
Industrialization and Development,” Universita degli Studi di Milano. http://dev3.cepr.org/
meets/wkcn/2/2360/papers/Woerz.pdf.
Sturgeon, T., and M. Kawakami. 2010. Global Value Chains in the Electronics Industry: Was
the Crisis a Window of Opportunity for Developing Countries? The World Bank Policy
Research Working Paper No. 5417.
Sturgeon, Timothy J. 2009. From Commodity Chains to Value Chains:  Interdisciplinary
Theory Building in an Age of Globalization. In Frontiers of Commodity Chain Research,
edited by Jennifer Bair, 110–135. Stanford, CA: Stanford University Press.
Vassalos. 2012. Who’s Driving the Agenda at DG Enterprise and Industry? Alliance for
Lobbying Transparency and Ethics Regulation—ALTER-EU, Brussels. http://www.alter-eu.
org/sites/default/files/documents/DGENTR-driving_0.pdf.
Waldkirch, A. 2010. The Effects of Foreign Direct Investment in Mexico since NAFTA. World
Economy 35 (5): 710–745.
 Walter Hatch, Jennifer Bair, and Günter Heiduk    255

Wang, Zhi, William Powers, and Shang-jin Wei. 2009. Value Chains in East Asian Production
Networks: An International Input-Output Model Based Analysis. U.S. International Trade
Commission, Office of Economics Working Paper No. 2009-10-C.
Williamson, Oliver. 1975. Markets and Hierarchies. New York: Free Press.
Woll, Cornelia. 2012. The Brash and the Soft-Spoken:  Lobbying Styles in a Transatlantic
Comparison. Interest Groups & Advocacy 1 (2): 193–214.
Yamamura, Kozo, and Walter Hatch. 1997. A Looming Entry Barrier? Japan’s Production
Networks in Asia. NBR Analysis 8 (1). http://nbar.org/publications/analysis/pdf/vol8no1.pdf.
Yeats, Alexander J. 2001. Just How Big Is Global Production Sharing. In Fragmentation: New
Production Patterns in the World Economy, edited by Sven Arndt and Henryk Kierzkowski,
108–143. Oxford: Oxford University Press.
 Pa rt  I V

D OM E ST IC
I N ST I T U T ION S
 Chapter 14

New Demo c rac i e s

Bumba Mukhe r je e

What is the impact of democratization on trade protection in developing states? Are

governments in newly democratized developing countries more likely to adopt trade
reforms? Political scientists and economists have invested substantial effort in address-
ing these two questions (e.g., Frye and Mansfield 2004; Giavazzi and Tabellini 2005;
Milner and Kubota 2005; Rudra 2005; Eichengreen and Leblang 2008; López-Córdova
and Meissner 2008; Tavares 2008; Milner and Mukherjee 2009a). While debates persist
over how democratization increases the likelihood of trade liberalization, scholars gen-
erally converge on the view that governments in newly democratized developing states
have political incentives to reduce trade barriers (Giavazzi and Tabellini 2005; Milner
and Kubota 2005; Eichengreen and Leblang 2008; Milner and Mukherjee 2009a).
Mutual agreement among students of political economy about the impact of domestic
political institutions (in this case, the transition to democracy) on economic outcomes
such as trade policy is rare.
Indeed, the widely accepted view that democratization fosters trade reforms stands in
sharp contrast to debates about the impact of established democracies on trade protec-
tion. Some scholars suggest, for instance, that democracy is mutually compatible with
free trade (Mansfield, Milner, and Rosendorff 2000; Guisinger 2001; Eichengreen and
Leblang 2008). Others claim that democracy promotes free trade only under certain
electoral systems1 or economic conditions (Tavares 2008). Furthermore, some research-
ers find that democratic governments employ nontraditional means such as nontariff
barriers (NTBs) to protect their economies (Kono 2006). Although claims about the
effects of established democratic institutions on trade policy are far from settled, it is
clear from the preceding summary of the relevant literature that students of interna-
tional political economy (IPE) are far more confident that the transition to democracy
leads to trade liberalization in developing states. Is this confidence justified?
A close look at some cases suggests that the link between democratization and trade
reforms in developing countries is more complex than suggested in extant studies. For
example, during the initial post-transition years between(i) 1986 and 1989, the Sarney
administration in Brazil’s newly democratized state and (ii) 1987 and 1990, the Corazon
260   Democracies

Aquino–led government in Philippines’ new democracy sharply reduced trade restric-

tions (Weyland 2002; UNCTAD 2005). After Bangladesh and Ghana experienced a
transition to democracy in 1991 and 1996, respectively, the first post-transition govern-
ment in each of these two countries increased tariffs to protect domestic industries from
import competition (UNCTAD 2005). These examples suggest that extant research on
trade politics in developing states has arguably painted the link between democratiza-
tion and trade reforms with a broad brush.
This should not be taken to mean that insights provided by current studies on
democratic transitions and trade policy are inadequate. These studies provide an
excellent foundation for conducting further research on when—rather than just
why—democratization is likely to engender trade liberalization in developing coun-
tries. Conducting further research on how the emergence of democracy influences
trade protection is substantively important, as it may help us to address the variation in
trade policy in the post-transition years across “new democracies.” Developing nuanced
insights about the conditions under which leaders in newly democratized states adopt
trade reforms also allows us to understand whether (and when) democratic transitions
promote economic growth, if we believe that more trade openness fosters growth.
Given the importance of understanding the link between democratic transitions and
trade policy, the objective of this chapter is thus twofold. The first objective is to consider
the state of the literature on democracy (including democratization) and trade policy.
One cannot provide an exhaustive analysis of all the relevant literature in this issue area.
That said, key insights from this literature are highlighted in this review. The second
objective is to address the following three questions:

• What do the data tell us about trade reforms during the immediate post-transition
years in newly democratized regimes in the developing world? Are some post-
transition governments more likely to liberalize trade policies than others?
• Is there variation in industry-level tariffs in new democracies during the initial
years after the transition to democracy? If so, what is the nature of this variation,
and what does it tell us about democratization and trade politics?
• Does trade liberalization positively influence the likelihood of democratization? Is
there statistical evidence for an endogenous relationship between democracy and
trade reforms?

The next section of this chapter begins with a discussion of the existing literature on
the impact of democratic transitions as well as established democracies on trade policy.
In the third section I analyze some data that reveal both variation in import duties and
trade protection of skill-intensive versus low-skilled industries in newly democratized
states. This is followed by a concise discussion of various arguments that may account
for the variation in trade barriers mentioned above. The fourth section explores both
the data and the literature that evaluate the impact of trade openness on the likelihood
of democratization. I conclude with a discussion of key research issues that need to be
addressed in the study of democratization and trade policy.
 Bumba Mukherjee   261

Studies on Democracy and Trade

Economists generally agree that trade protection generates deadweight losses.2 Yet it is
common knowledge that governments in both developed and developing countries pro-
tect domestic industries from import competition through the use of tariffs and NTBs.
Why do governments protect their domestic industries from import competition? The
literature that addresses this question is vast, and there is no space to discuss this litera-
ture in detail in this chapter. Stated briefly, however, academic research on the political
economy of trade protection can be broadly divided into two categories: “demand-side”
and “supply-side” explanations for trade policy. Demand-side explanations largely
employ the Stolper-Samuelson theorem and the Ricardo-Viner model to account for the
trade-policy preferences of individuals. Demand-side studies of trade policy primarily
explain how trade-policy preferences of individuals or firms may influence the pros-
pects for trade liberalization (Rogowski 1989; Busch and Reinhardt 1999, 2000; Scheve
and Slaughter 2001; Hiscox 2002; Baker 2005; Mayda and Rodrik 2005; Mansfield and
Mutz 2009).
According to these “demand-side” studies, the trade policy preferences of individu-
als, including voters, are determined by their degree of occupational specificity, educa-
tion, or income; their beliefs about whether or not trade openness is beneficial for them;
and their sociotropic perceptions of trade’s influence on the national economy. Theories
about trade-policy preferences held by individuals have been extensively tested by using
survey-response data sets.3 In contrast to the focus on individual preferences, other
demand-side explanations examine how campaign contributions and lobbying pressure
from “special interests” affect trade politics (Baldwin and Magee 2000; Grossman and
Helpman 2006). Empirical tests of the effects of industry-based campaign contributions
provided by industries on trade policy outcomes are rare, however, given the paucity of
publicly available data about contributions provided by industries.
Unlike the demand-side approach, supply-side explanations about trade policy ana-
lyze the impact of political institutions on international trade. Some of these studies
have explored the effects of international institutions such as the General Agreement on
Tariffs and Trade/World Trade Organization (GATT/WTO) on international trade flows
or the settlement of inter-state trade disputes (Gowa and Kim 2005; Tomz, Goldstein,
and Rivers 2007; Davis and Bermeo 2009). Other supply-side studies theorize and find
empirically that democratic regimes—including newly democratized states and estab-
lished democracies—are more likely to liberalize their trade policies (Frye and Mansfield
2004; Milner and Kubota 2005; Eichengreen and Leblang 2008; Milner and Mukherjee
2009a). Several studies also statistically evaluate and find robust support for the influence
of specific democratic institutions on trade policy (Nielsen 2003; Hankla 2006; Henisz
and Mansfield 2006). These institutions include electoral systems, the number of veto
players, electoral particularism, political partisanship, and the numbers of effective par-
ties in democracies. Notwithstanding these demand-side and supply-side explanations
for the determinants of trade protection, there is little ambiguity in the fact that trade
262   Democracies

barriers have decreased sharply across developing states during the past three decades.
This is corroborated by a substantial body of statistical evidence showing that trade bar-
riers have declined significantly across several developing countries since 1980 (Harrison
and Hanson 1999; Goldberg and Pavcnik 2004; Milner and Kubota 2005).
Preceding and concurrent with the move to free trade in the developing world, there
has been a global movement toward democracy. For instance, in their careful and
detailed empirical study of democratic transitions and economic growth, Papaionnou
and Siourounis (2004) identify as many as fifty-nine episodes of “full” democratiza-
tion and twenty-four episodes of “partial” democratization between 1970 and 2002.4
This is further confirmed by Milner and Kubota (2005, 158–159), who point out that
the number of democracies in the developing world grew at least threefold between
1975 and 2002. It is precisely this “dual transition” phenomenon, characterized by
simultaneous democratization and trade liberalization in the developing world
(since the early 1980s), that has encouraged researchers to claim that the transition
to democracy leads to trade liberalization. Scholars have also employed a variety of
statistical models to evaluate whether the emergence of democracy leads to a decline
in trade protection. These studies show that democratization indeed exerts a posi-
tive and statistically significant effect on trade liberalization or, in other words, leads
to a significant decline in aggregate country-year trade protection measures such as
import duties (Giavazzi and Tabellini 2005; Milner and Kubota 2005; Eichengreen
and Leblang 2008).
In addition to the empirical research alluded to above, a fairly rich literature exam-
ines why democratization engenders trade liberalization (Stokes 2001; Weyland 2002;
Milner and Kubota 2005). The most comprehensive theoretical analysis of democ-
ratization on trade liberalization to date is by Milner and Kubota (2005). They argue
that since democratization expands the “selectorate” in a polity, it empowers whole
groups in society that were formally excluded from the political process. In the con-
text of developing countries, this implies that democratization empowers and enfran-
chises particularly labor in the rural and urban informal sectors, who, given the
Stolper-Samuelson theorem, favor trade openness, as they are the abundant factors
that gain from trade in developing countries. As a result, governments in new democ-
racies have political incentives to respond positively to the preferences of labor by low-
ering trade barriers.
Unlike Milner and Kubota (2005), Weyland (2002, 60)  suggests in his work that
democratization weakened the political power of vested interests that favored protec-
tionism in Latin American states. This facilitated the adoption of trade reforms by polit-
ical leaders in various newly democratized Latin American states such as Argentina and
Brazil (Weyland 2002, 60–61). An important difference between Weyland (2002) and
Milner and Kubota’s (2005) theoretical analysis is that the latter set of scholars identify
which social classes are more likely to benefit from trade reforms. Milner and Kubota
(2005) also develop a logically consistent causal mechanism that links democratization
specifically to trade liberalization, while Weyland’s (2002) book focuses more broadly
on a larger set of economic reform (including trade reform) policies.
 Bumba Mukherjee   263

Other scholars (mainly comparativists) have also suggested either implicitly or

explicitly that democratization generally promotes economic reform and more spe-
cifically trade liberalization. For instance, early research on the topic of political
regimes and economic reform questioned the presumption that authoritarian gov-
ernments have an inherent advantage with respect to implementing trade reform
(Geddes 1995; Remmer 1998). Geddes (1995) and Remmer (1998) instead suggest that
democratizing states are quite capable of implementing reform policies, including
trade liberalization. Stokes (2001) argues in her book that if trade reforms provide
benefits to large sections of society in developing countries, then democratization
will foster trade liberalization in these states. This is because leaders in new democra-
cies will be subject to electoral punishment if they resist trade liberalization. In addi-
tion, since governments in newly democratized states anticipate political rewards
from liberalization ex post, they have political incentives ex ante to liberalize trade
policies (Stokes 2001).
The discussion in the preceding paragraphs indicates that extant studies on demo-
cratic transitions and trade policy primarily focus on why democratization leads to trade
reforms. While numerous studies emphasize that both democratization and higher
levels of democracy are negatively correlated with trade protection, not all researchers
agree with the claim that democracy helps to reduce trade barriers. For example, Kono
(2006) finds that democracies have a negative effect on tariffs but a positive impact on
NTBs. Garrett (2000, 973) posits that even though “on the one hand, democracy” helps
with respect to “promoting trade liberalization. … On the other hand, democracy also
empowers distributional coalitions with intense interests, making higher levels of pro-
tectionism more likely.”
The possibility that democracy may not always promote free trade, as suggested by
Garrett (2000), has motivated scholars to focus on variation in trade barriers across
democracies. Some studies find, for instance, that left-leaning parties espouse more
protectionist policies than their right-leaning counterparts in developed countries
but are more pro free-trade in developing countries (Milner and Judkins 2004; Dutt
and Mitra 2006). Nielson (2003) argues that trade protection is lower in presiden-
tial systems across developing countries, while Henisz and Mansfield (2006) show
that the conditional effect of veto players matters for trade policy. Rogowski (1989)
argued that closed-list proportional representation (PR) systems are more likely to
have lower trade barriers. Building on this, Mansfield and Busch (1995) suggested
more broadly that PR systems tend to have lower levels of NTBs. Finally, Grossman
and Helpman (2005) claim that in majoritarian systems, parties cannot prevent indi-
vidual legislators from pursuing protectionist policies for the interests of their own
constituencies.
It is important to note here that extant research on variation in trade barriers across
democratic states focuses on trade policies in established and “consolidated” democ-
racies across the developed and developing world.5 Less attention has, however, been
paid to the possibility that the degree of trade protection (and thus the prospects for
trade reform) may also vary substantially during the initial post-transition years in
264   Democracies

newly democratized developing states. Indeed, as discussed earlier, anecdotal evidence

reveals that although post-transition governments in Brazil and the Philippines sharply
reduced trade barriers, other newly democratized states like Ghana and Bangladesh
increased trade restrictions during the initial post-transition years. Moreover, as shown
below, variation in trade barriers during the immediate post-transition period is also
prevalent in a larger pooled sample of “new democracies.”

Democratization and Trade

Barriers: The Data

The possibility that trade protection varies substantially across newly democratized
states during the immediate postdemocratic transition years is supported in a data set
on trade barriers for the years following a democratic transition across several develop-
ing countries. Consider figure 14.1, which plots the moving-average of the import duty
coverage ratio (a widely accepted measure of tariff barriers)6 during the first five years
after a democratic transition across a comprehensive list of fifty-six developing coun-
tries that have experienced a transition to democracy in the previous three decades.
These countries and the year in which each made a transition to democracy (as identi-
fied in the Polity V and the Przeworski, Alvarez, Cheibub, and Limongi 2000 data sets)
are listed in table 14.1.7
Figure 14.1 shows that although import duties decreased in the first five post-transition
years for 62 percent of these “new democracies,” it increased or maintained the status
quo in the remaining 38 percent of the developing countries that also successfully made
a transition to democracy. This figure therefore reveals that trade liberalization (i.e., a
decrease in import duties) in the initial post-transition years occurred in a certain share
of but not all newly democratized developing states.
It is worth emphasizing here that it is not merely an aggregate country-year mea-
sure of trade barriers across several industries—for example the import duties mea-
sure employed to extract figure 14.1—that varies across newly democratized states
in the post-transition period. If, for instance, one makes a distinction between trade
protection on high skilled goods produced by skill-intensive industries (e.g., the pet-
rochemical industry) and low skilled goods (e.g., agriculture) in developing states,
one finds that the relationship between democratization and trade protection in the
developing world is more complex.8 Consider figure 14.2, which illustrates the mean
of output-weighted (average) tariffs for several high skilled goods and low skilled
goods at the three-digit International Standard of Industrial Classification9 (ISIC)
level (i)  for a sample of established developing country democracies10 between
1980 and 2006 and (ii) during the first five postdemocratic transition years for the
fifty-six developing countries listed in table 14.1.11 Before discussing the pattern of
industry-level tariffs in this figure, it is important to explain the three-step procedure
 Bumba Mukherjee   265

35
Decline (62%)
Increase (38%)
30

25
Import duties (% )

20

15

10
Years prior to Democratic Years in which transition
democratic transition transition at t countries are observed as
5 new democracies

0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5 t+6
Time

Fig.  14.1  New Democratic Regimes and Import  Duties.

Note:  Data sources used to compute the moving average of import duties are listed in the  text.

Column A Column B

18 High skilled goods

Output weighted tariff (%)

15

12
High skilled goods
9

6
Low skilled goods Low skilled goods

3
Established democracies New democracies

Mean output-weighted tariff level

Fig. 14.2  Tariffs for High Skilled and Low Skilled Goods in New and Established Democracies.

that I  employed to operationalize the output-weighted average tariff level for low
skilled goods (labeled low skilled tariff) and high skilled goods (skilled tariff) illus-
trated in figure 14.2. First I collected ad valorem tariff country-year data for thirty
industries—including the agricultural sector—that are disaggregated and classified
at the three-digit ISIC level (Revision 2)  for each established developing democ-
racy and newly democratized state. Tariff data at the disaggregated three-digit ISIC
level for developing countries are only available for these thirty industries.12 These
Table 14.1  New Democracies in the Developing World
First 5 Post- First 5 Post-
Democratic Transition Years Democratic Transition Years
Transition (includes Transition (includes
Country Year transition year) Country Year transition year)

Albania 1992 1992–1996 Lithuania 1991 1991–1995

Argentina 1983 1983–1987 Madagascar 1993 1993–1997
Bangladesh 1991 1991–1995 Malawi 1994 1994–1998
Benin 1991 1991–1995 Mali 1992 1992–1996
Bolivia 1982 1982–1986 Mexico 1997 1997–2001
Brazil 1985 1985–1989 Mongolia 1992 1992–1996
Bulgaria 1990 1990–1994 Mozambique 1994 1994–1998
Cape Verde 1991 1995–1999 Nepal 1991 1991–1995
Central African Republic 1993 1993–1997 Nicaragua 1990 1990–1994
Chile 1990 1990–1994 Nigeria 1999 1999–2003
Comoros 1990 1990–1994 Pakistan 1988 1988–1992
Croatia 2000 2000–2004 Panama 1994 1994–1998
Czech Republic 1993 1993–1997 Peru 1980 1980–1984
Dominican Republic 1978 1978–1982 Philippines 1987 1987–1991
Ecuador 1979 1979–1983 Poland 1990 1990–1994
El Salvador 1994 1994–1998 Slovak Republic 1993 1993–1997
Estonia 1991 1991–1995 Slovenia 1992 1992–1997
Ethiopia 1995 1995–1999 South Africa 1994 1994–1998
Ghana 1996 1996–2000 Thailand 1992 1992–1996
Guatemala 1996 1996–2000 Uruguay 1985 1985–1989
Guyana 1992 1992–1996 Paraguay 1993 1993–1997
Haiti 1994 1994–1998 Russia 1993 1993–1997
Honduras 1982 1982–1986 Senegal 2000 2000–2004
Hungary 1990 1990–1994 Suriname 1991 1991–1995
Indonesia 1999 1999–2003 Tanzania 1995 1995–1999
Korea, Republic of 1988 1988–1992 Turkey 1983 1983–1987
Latvia 1991 1991–1995 Ukraine 1991 1991–1995
Lesotho 1993 1993–1997 Zambia 1991 1991–1995
 Bumba Mukherjee   267

Table 14.2  Three-Digit ISIC Industries

ISIC # Description S/L Ratio ISIC # Description S/L Ratio

331 Wood products except 0.109 342 Printing and publishing 0.397
furniture
323 Leather products 0.091 354 Miscellaneous petroleum & 0.414
coal products
311 Food products 0.164 355 Rubber products 0.396
111 Agriculture & agric. raw 0.047 356 Plastic products 0.358
materials
313 Beverages 0.126 361 Pottery, china & earthenware 0.140
314 Tobacco 0.075 362 Glass and glass products 0.216
322 Wearing apparel except 0.193 372 Nonferrous metal basic 0.392
footwear industries
331 Wood products except furniture 0.138 381 Fabricated metal products 0.404
332 Furniture except metal 0.125 382 Machinery except electrical 0.426
324 Footwear except rubber or 0.216 383 Electrical machinery 0.611
plastic
312 Textiles 0.112 390 Other manufactured products0.587
371 Iron and steel 0.283 353 Petroleum refineries 0.630
369 Other nonmetallic mineral 0.205 351 Industrial chemicals 0.592
products
372 Nonferrous metals 0.176 384 Transport equipment 0.465
341 Paper products 0.391 385 Professional goods & 0.726
scientific equipment

industries and the goods that they produce are listed in table 14.2. The goods listed
there range from low skilled goods such as wood products and goods produced in the
agricultural sector to relatively high skilled goods that include, for example, scientific
equipment.
Second, after collecting country-year tariff data for the thirty industries, I classified
them and therefore the goods that they produce into two broad categories: low and high
skilled goods, as follows. First, following existing studies, I calculated the S/L ratio, that
is, the ratio of skilled workers S (workers with greater than or equal to twelve years of
schooling) to low-skilled workers L (workers with fewer than twelve years of schooling)
employed in each industry to rank-order the goods produced in the industry according
to their skill-intensity of production (Nunn and Trefler 2006).13 As suggested by Nunn
and Trefler (2006), if the ratio of skilled to low skilled workers employed in an industry
listed in table 14.2 is S/L ≥ 0.39, then the good produced by the industry is classified as
268   Democracies

high skilled; if S/L < 0.39 for an industry, then the good produced by that industry is
classified as low skilled.
Using the S/L ratio of 0.39 as a threshold to classify the thirty industries and the
goods that they produce into low or high skilled, I obtained seventeen low skilled goods
that are produced in low skilled industries and thirteen high skilled goods produced
in skill-intensive industries. The mean low skilled and skilled tariff levels illustrated in
figure 14.2 do not change when I use any value in the S/L ∈ [0.36, 0.42] range as the
threshold for the S/L ratio to categorize the goods produced in the thirty industries as
low or high skilled. Third, following extant studies on trade protection at the three-digit
ISIC industry level,14 I used the tariff data for each low and high skilled good to compute
the output-weighted average tariff for the seventeen low skilled goods (low skilled tariff)
and thirteen high skilled goods (skilled tariff) for each country in the group of (i) estab-
lished developing democracies and (ii) newly democratized states.15 I then calculated
the within-group mean of low skilled tariff and skilled tariff for each of the two groups
mentioned above to extract figure 14.2.
Figure 14.2 shows the group mean of these two measures between established
developing democracies and newly democratized states. Column B and a simple
difference-of-means test show that the mean of skilled tariff is substantially and sig-
nificantly (in the statistical sense) higher than the mean of low skilled tariff in the
immediate post-transition years for newly democratized states. In contrast, column
A reveals that there is little or no difference between the mean of skilled tariff and low
skilled tariff in the set of established developing country democracies. Figure 14.2 also
shows that the mean of skilled tariff in newly democratized developing states is much
higher than the mean of this measure for established democracies across the devel-
oping world. But the difference in the mean level of low skilled tariff between newly
democratized states and established developing democracies is marginal. Hence, we
learn from this figure that there is a “skill bias” in trade protection during the initial
post-transition years in new democracies, in that trade barriers on high skilled goods
are much higher than those for low skilled goods in the post-transition years for these
states.
Figures 14.1 and 14.2 therefore unambiguously reveal that there is no simple inverse
relationship between the emergence of new democracies and trade restrictions in the
developing world. Instead, these figures show that both aggregate measures of trade
barriers (e.g., import duties) as well as industry-level tariffs—conceptualized here as
the distinction in tariffs on goods produced by low skilled industries versus those pro-
duced by skill-intensive industries—varies significantly across new democracies in the
developing world. They also suggest that some but not all newly democratized states
enthusiastically embraced trade reforms during the initial post-transition years. These
two figures also raise the following questions: Why do import duties decrease in the
immediate years following a democratic transition in some new democratic regimes
across the developing world but not others? Why are tariffs on goods produced by
skill-intensive industries significantly higher than tariffs placed on low skilled goods
in the initial post-transition years in new democracies? Answering these questions is
 Bumba Mukherjee   269

critical to understanding the link between new democratic regimes and trade protec-
tion. This is substantively vital, given that the variation illustrated in figures 14.1 and 14.2
has not to the best of my knowledge been carefully studied by scholars. Addressing these
questions is also important given that trade policy has critical economic consequences
in developing countries. The next section thus provides some brief answers to the two
questions posited above.

Why Trade Protection Varies

Across New Democracies

Various arguments can be put forth to address the first question, which focuses on why
some but not all newly democratized states adopt trade reforms. One approach that
can be used to answer this question is to explore how the trade-policy preferences of
the empowered electoral majority in new democracies, namely workers, influence the
trade-policy responses of political parties in these states. To see this in more detail, con-
sider a situation in which political parties “strategically interact” in a Downsian setting
with workers (i.e., labor)—a key productive group in society (who, as mentioned above,
are also an electoral majority)—during the immediate post-transition period in a new
democracy. Suppose that the trade-policy preferences of workers in the new democ-
racy are determined by their degree of inter-industry occupational mobility. As shown
in some “demand-side” theories of trade policy, occupationally mobile workers tend to
support trade liberalization, as they gain monetarily from higher levels of trade open-
ness (Wacziarg and Wallack 2004; Mukherjee, Smith and Li 2009). Conversely, occu-
pationally specific workers typically favor higher levels of trade protection, because
they are more likely to suffer from unemployment when import competition increases
(Mukherjee, Smith, and Li, 2009).
Building on these “demand-side” claims in the literature, suppose further that the
inter-industry occupational mobility of workers in the newly democratized state’s
economy is sufficiently high. If this is the case, then it follows that the distribution of
the workers’ occupational mobility in the economy will be positively skewed. In this
positively skewed distribution, the median voter (who is also a worker) is more occu-
pationally mobile than the mean voter. In other words, the median voter in this case
will be a “mobile-type” individual characterized by a relatively high degree of occupa-
tional mobility across industries. This is important, because when the median voter is
indeed occupationally mobile between industries in the new democratic state, he or she
is more likely to gain from more trade openness and will thus be receptive toward trade
liberalization.
Note that in the Downsian equilibrium, the political parties in the newly democ-
ratized state will observe that the median voter’s level of inter-industry occupational
mobility is sufficiently high and will thus recognize that the median voter favors trade
270   Democracies

reforms. Since the parties in the new democracy are interested in winning the elec-
tion, they have political incentives to appeal to the “occupationally mobile” median
voter by reducing tariff barriers in equilibrium. The main hypothesis that emerges
from this simple story is that policy makers in new democracies will reduce tariffs
(i.e., adopt trade reforms) during the initial post-transition years if the inter-industry
occupational mobility of workers in the economy is sufficiently high. Is this hypoth-
esis empirically valid? I briefly assess this hypothesis by broadly evaluating the impact
of a recently developed measure of inter-industry occupational mobility of labor (in
short, inter-industry labor mobility) on import duties for the first five post-transition
years across the fifty-six newly democratized developing states listed in table 14.1. This
measure of inter-industry occupational mobility of labor (labeled here labor mobility)
has been developed by Wacziarg and Wallack (2004) and Hiscox and Rickard (2005).
The labor mobility measure is computed for each country-year in my sample by isolat-
ing the fraction of jobs that move from one industry to another independent of overall
employment gains or losses. Let E tj ,i denote employment in industry j in country i at
time t. Hence

Σ Nj =1 E tj ,i − E tj ,−i δ − Σ Nj =1 E tj ,i − Σ Nj =1 E tj ,−i δ
labor mobility =
1 N t
Σ (E + E tj ,−i δ ) (1)
2 j =1 j , i

where the summation ( Σ Nj =1 ) is over all N = 30 industries at the three-digit ISIC level
listed in table 14.2.
The difference between the expression E tj ,i − E tj ,−δ i on the left and the expression
Σ Nj =1 E tj ,i − Σ Nj =1 Eon
t −δ
j ,i
the right in the numerator in equation 1 gives the employment changes
that result from the pure shifts of jobs across different industries.16 The denominator
in labor mobility computes the average of total employment for the industries in con-
sideration between t and t – δ. I set δ = 1 year given that I am interested in examining
the impact of the labor mobility measure on import duties on for each of the five initial
post-transition years across the fifty-six new democracies. The labor mobility measure is
a continuous variable that ranges from low to high inter-industry occupational mobility.
The scatterplot in figure 14.3 shows that higher levels of labor mobility indeed have a
strong negative effect on import duties during the initial five post-democratic transition
years. This indicates that higher levels of inter-industry labor mobility encourage policy
makers in newly democratized developing states to embrace trade liberalization during
the immediate post-transition period. Figure 14.3 also shows that import duties tend
to be much higher in new democracies when labor mobility is low or, in other words,
workers are characterized by a high level of occupational specificity. This is perhaps not
surprising, as a high level of occupational specificity in the labor force will ensure that
the median voter is occupationally specific and will hence favor higher trade protection.
This in turn will drive Downsian parties to raise trade barriers in the post-transition
period. The simple scatterplot shown in figure 14.3 therefore provides an intuitive
 Bumba Mukherjee   271

account of why there is variation in trade barriers across new democracies. That said, it
needs to be studied in far more depth.
The results discussed above are interesting, but they merely tell us something about
the correlation between labor mobility (or lack thereof) and trade protection in new
developing country democracies. They do not explain causality per se between demo-
cratic transitions and trade liberalization in developing countries. Furthermore, it may
be possible that it is not the level of inter-industry labor mobility in new democracies
but other alternative arguments that may account for why policy makers in some newly
democratized states embrace trade liberalization immediately after the transition to
democracy. For instance, there are sound reasons to suspect that in some new democra-
cies “geographically concentrated” export-oriented industries—which are less suscep-
tible to collective action problems owing to their geographic concentration—may have
extensively lobbied their respective governments to reduce trade barriers. Examples
from the post-transition years in newly democratized states such as Brazil and
Indonesia suggest that geographically concentrated export-oriented industries indeed
put pressure on policy makers in these countries to slash tariffs.17 Others have suggested
that new democracies in especially Central and Eastern Europe acquiesced to pressure
from international institutions like the International Monetary Fund and the World
Bank to reduce trade restrictions (Shafaeddin 2006). Such pressure may have induced
policy makers in these states to adopt trade reforms in the initial post-transition years.
In short, the claims posited above, as well as these alternative causal claims, suggest
that studies on trade politics need to develop parsimonious yet lucid causal theories to

20

15
Import duties (%)

10

1 2 3 4 5
Labor mobility

Fig.  14.3  Labor Mobility and Import Duties in New Democracies.

Notes: Scatterplot of import duties against the index offender-industry labor mobility for the first five post-transition
years for the 56 new democracies in table 14.1. Each point represents one postdemocratic transition country-year for
which both variables are observed. The scatterplot is overlaid with a dashed line, which is the pooled OLS best fit
line that accounts for the within-country correlation of import duties across  time.
272   Democracies

account for the variation in trade protection across “new democracies” that is illustrated
in figure 14.1.
Let us next briefly explore the second question posited above:  Why are tariffs on
goods produced by skill-intensive industries significantly higher than tariffs placed on
low skilled goods in the initial post-transition years in new democracies? A variety of
explanations can arguably answer this question. Recent studies by some economists that
examine industry-level tariffs at the six-digit ISIC level in developed and developing
democracies suggest that leaders in majoritarian democracies adopt higher trade bar-
riers on goods produced by skill-intensive industries but reduce trade restrictions on
low skilled goods such as agricultural products (Ardelean and Evans 2013). This claim is
intuitive. Yet a cursory examination of my data on skilled tariff and low skilled tariff for
new democracies shows that this claim does not hold for newly democratized states.
Specifically, the vast majority (73 percent) of the fifty-six new democracies in table 14.1
adopted the PR electoral system or variants of it immediately after the transition to
democracy. The mean of skilled tariff during the initial post-transition years in these
new democracies that adopted the PR system is higher than the mean of low skilled tariff
in these countries. It is also higher than the mean of both skilled and low skilled tariff
in new democracies that adopted the majoritarian system. This suggests that focusing
on the distinction between electoral systems may not be fruitful in accounting for why
trade restrictions on goods produced by skill-intensive industries are high during the
post-transition period in new democracies. Instead, it might be worthwhile to pay more
attention to how certain domestic industry-level characteristics within new democ-
racies may shape the economic and political relationship between governments and
industries in these states. This relationship in turn may potentially account for different
patterns of trade protection that we observe at the industry level in new democracies.

Does Trade Openness Influence

Democratization?

Understanding the impact of democratization on trade policy is undoubtedly a cen-

tral area of research for students of IPE. Yet scholars have suggested that there is the-
oretically a potential for reverse causality between democratic transitions and trade
reforms; that is, higher levels of trade openness that result from trade liberalization may
increase the likelihood of democratization in developing countries (Lipset 1959; Boix
2003; Acemoglu and Robinson 2006). The theoretical literature on the impact of trade
openness on democratization is indeed quite rich. For instance, Lipset (1959) suggested
almost five decades ago that trade can spark development and create a larger middle
class, which in turn might foster the emergence of democracy.
Acemoglu and Robinson (2006) build on their well-known models of the link between
inequality and democratization to explore how trade openness influences the prospects
 Bumba Mukherjee   273

of democratic transition in developing states. They incorporate a Stolper-Samuelson

model of international trade, which assumes perfect labor mobility, and essentially
argue that higher levels of trade openness decreases income inequality in developing
countries, which consequently increases the probability of democratization in these
states. This is because if one assumes that developing countries are abundant in labor,
then one would expect increased trade openness to increase real returns for labor (the
relatively abundant factor) and reduce real returns for owners of capital (the relatively
scarce factor) under the Stolper-Samuelson theorem. This in turn reduces income
inequality in these countries. Lower inequality then reduces the costs of tolerating
democracy for elites in authoritarian developing countries, as tax redistributions are less
severe for elites when income inequality is low. In short, because greater trade openness
reduces income inequality in labor-abundant developing authoritarian states and lower
inequality reduces the costs that elites may incur from a democratic transition, political
elites in these states have incentives to adopt democracy.
In contrast to Acemoglu and Robinson (2006), Boix (2003) suggests that the effect
of trade openness on democracy is contingent on the distribution of factors within a
particular economy. When skilled workers are the abundant factor, trade openness
increases income inequality within society by driving up the wages of skilled workers
and deflating the wages of already poorer low skilled or unskilled workers. Growing
income inequality in turn discourages democratization. Adsera and Boix (2002) argue
that greater trade openness actually endangers democratic institutions in emerging
democracies. They emphasize that in countries in which democratic institutions are less
well established, interest groups that benefit from more trade openness might try to void
democratic institutions and impose openness through a dictatorship.
A careful examination of the relevant empirical literature indicates that statistical
support for the theoretical claim that trade openness positively influences democrati-
zation is either weak or at best conditioned by other domestic factors. Some studies,
for example, find that trade openness does not have a significant effect on the likeli-
hood that democracy may occur (Bussmann 2002). Li and Reuveny (2003) find that
trade openness is negatively associated with democratic institutions. Rudra (2005) sug-
gests that the effect of trade openness on democratization is conditional on the bargain-
ing power of labor in developing countries. The one exception is Lopez-Cordova and
Meissner (2008), who find that trade openness positively influences democratization.
The studies cited in the previous paragraph employ samples that include developed
and developing countries for their statistical tests, as well as (in some cases) using a long
temporal domain that typically includes observations from the late nineteenth century
(Lopez-Cordova and Meissner 2008; Eichengreen and Leblang 2008). While including
developed and developing countries in the sample and using a longer temporal sample range
increases the generalizability of the reported results, it also raises the following question: Is
there statistical support for the theoretical prediction that trade openness has increased the
probability of democratization in developing countries in the previous three to four decades?
To address this question, I briefly conducted a simple empirical exercise to check
whether trade openness (the sum of exports and imports as a share of gross domestic
274   Democracies

product) has a statistically positive effect on the likelihood of democratization in a sam-

ple of 128 developing countries from 1972 to 2008. To this end, I estimated a dynamic
probit model with random effects to assess the statistical effect of trade openness on
the dummy dependent variable democracy. This dummy variable is set equal to 1 for
democratic countries that satisfy Przeworski and colleagues’ (2000) criteria for a
democracy.18 Standard control variables such as log GDP per capita, ethno-linguistic
fractionalization, log population, and the number of past transitions to authoritarian-
ism are also included in the dynamic probit specification, as these variables are known
to influence the probability of democratization (for this see Przeworski et al. 2000).
Estimates from the dynamic probit model, which are not reported here to save space,
reveal that the effect of trade openness on the dummy dependent variable democracy
is positive but statistically insignificant. This result requires more research. But it does
confirm extant findings that cast doubt on the claim that trade openness positively
influences democratic transitions.19
Researchers should not conclude from the consistent lack of robust statistical sup-
port for the effect of trade openness on the likelihood of democratization that this
issue area is not worth assessing in more depth. I  believe that future research on
democratic transitions and trade policy may potentially benefit more by rigorously
evaluating when—rather than whether—trade openness may positively influence the
probability of democratization in developing states. Doing so may help to empiri-
cally clarify the potential link between trade openness and democratization. This
is substantively necessary considering that proponents of the view that trade open-
ness leads to democratization arguably accept this causal relationship far too easily.
Alternatively, skeptics tend to be excessively dismissive a priori about the potential
effects that trade liberalization may have on the prospects for democratic transition in
developing states.

Conclusion: Future Research

The third wave of democratization across the developing world, combined with the
sharp decline in trade barriers in these states, has produced a rich literature that explores
the impact of democracy on trade policy. While much progress has been made with
respect to unpacking the link between democracy and trade policy, less attention has
been paid to understanding the political dynamics of trade reforms during the initial
post-transition years in newly democratized developing states. That said, some stud-
ies have theorized extensively about why the emergence of democracy fosters trade
reforms. However, the most significant advance made in the study of democracy and
trade (as discussed in the previous sections) is empirical.
Despite the empirical progress made so far, it would be far-fetched to claim that
scholars have developed a thorough and comprehensive understanding of the design
of trade policy in both newly democratized developing states. Instead, in this section
 Bumba Mukherjee   275

I briefly discuss three main areas for future research that may help to move the research
agenda on democracy and trade forward. First and foremost, this chapter has analyzed
variation in tariff barriers such as import duties and output-weighted industry-level
tariffs during the immediate postdemocratic transition period. Yet an important area
of research on trade politics has shown that established democracies tend to use NTBs
rather than tariffs for protectionist purposes (Mansfield and Busch 1995; Kono 2006).
This suggests that it may not be sufficient to simply explore easily observable tariff bar-
riers in newly democratized states, given that developing country democracies may
use NTBs as a nontransparent tool for protectionist purposes. Instead, it may be worth
exploring whether democratization in developing countries discourages or encourages
the use of nontariff instruments for trade protection. Doing so will be challenging given
the paucity of reliable data on NTBs in the developing world. But given the importance
of NTBs, it makes sense to at least broadly understand the relationship between democ-
ratization and the use of NTBs as an instrument for trade protection.
Second, I emphasized earlier that political scientists often use aggregate country-year
measures such as import duties to assess the political economy of trade protection in
developed as well as developing countries (e.g. Milner and Kubota 2005; Henisz and
Mansfield 2006; Eichengreen and Leblang 2008). These measures are limited, as they
are not designed to capture substantial variation in industry-level tariff barriers across
both new and established democracies in the developing world. This is unfortunate,
because as I showed in a previous section of this chapter, even a cursory examination
of industry-level tariffs reveals that governments in newly democratized states tend to
protect domestic skill-intensive industries from import competition. In contrast, policy
makers in new democracies tend to reduce trade restrictions on low skilled goods.
I presented some brief arguments that may explain why we observe the variation in
industry-level tariffs posited above. Yet these arguments are at best conjecture. It is dif-
ficult to provide a thorough answer here to the issue of variation in industry-level tar-
iffs across new democracies, given space constraints. But the illustration in figure 14.2
certainly needs to be explored in greater empirical and theoretical detail. The intrigu-
ing relationship between the emergence of democracy and variation in industry-based
tariffs also suggests that researchers should continue to develop trade-policy measures
that account for both cross-sectional and temporal variation in industry tariffs across
developing countries rather than relying on standard aggregate tariff measures.
Third, some scholars debate whether or not capital account openness and trade
liberalization are substitutes or complements for developing country governments
(McKinnon 1991; Giavazzi and Tabellini 2005). A key lesson that one learns from this
debate is that governments in newly democratized developing states may account for
the extent to which their country’s capital account is open when deciding to adopt trade
reforms. It is beyond the scope of this essay to explore the link between capital account
openness and trade liberalization. But it might be worthwhile for scholars to explore
(i) if capital account policies matter for trade reforms in newly democratized states and
(ii) how domestic politics may potentially influence the simultaneous choice of capital
account and trade reforms during the initial post-transition period.
276   Democracies

Notes
1. Grossman and Helpman (2005) argue that majoritarian democracies are associated with
higher trade barriers. Rogowski (1989) and McGillivray (2004) claim that trade restric-
tions are likely to be lower in democracies that employ the closed-list PR electoral system.
2. For the claim that trade protection engenders deadweight losses in developed and devel-
oping countries, see, for example, Harrison and Hanson (1999).
3. For studies that employ survey-response data sets or conduct survey experiments to
assess trade-policy preferences, see, for example, Scheve and Slaughter (2001); Mayda and
Rodrik (2005); Baker (2005); and Mansfield and Mutz (2009).
4. Papaioannou and Siourounis (2004, 8)  define full democratization as “where both the
Polity and the Freedom House indicators have reached an almost perfect score like Spain,
Portugal, or Argentina).” They define “partial democratization” as including countries
that have abandoned autocratic rule, but in which civil rights protection has not reached
Western world levels (like Nigeria, Guatemala, or Albania).
5. The term “consolidated” democracies, as used here, refers to new democratic regimes that
did not relapse back into autocratic rule within three years after these countries made a
formal transition to a full-fledged democracy.
6. The import duty coverage ratio, labeled import duties, is defined as the total value of
a country’s import duties divided by the total value of its imports in a given year and is
expressed as a percentage. Data for this variable are drawn from World Bank (2011).
7. The democratic-transition year for each country in table 14.1 is identified using the Polity
indicator of the year of democratic transition in the Polity V database. The Polity V data-
base identifies a total of sixty-five democratic transition episodes across all developing
states from 1976 to 2005. Since data on import duties are available for fifty-six develop-
ing countries out of the total of sixty-five (sixty-one) developing states that experienced a
democratic transition, Figure 18.1 is derived based on the available import duties data for
these fifty-six “new democracies” listed in table 14.1.
8. Scholars are increasingly focusing on industry-level tariffs, including tariffs on goods pro-
duced by skill-intensive industries versus those produced by low skilled industries (see,
e.g., Milner and Mukherjee 2009b; Lu, Scheve, and Slaughter 2012).
9. The low and high skilled goods for which 1 collected output-weighted industry-level tariff
data at the three-digit ISIC level and which are included in Figure 14.2 are listed in table 14.2.
10. This sample includes a total of fifty-one established developing-county democracies such as
Brazil, India, Costa Rica, and other developing democracies (i) where democracy has lasted
for more than fifteen years and (ii) that have not relapsed back into an authoritarian regime.
11. The criterion that I employ to classify produced goods into either low or high skilled is
drawn from economists such as Nunn and Trefler (2006); it is described below.
12. The data for skilled and low skilled tariff are drawn from Milner and Mukherjee (2009b). The
primary and secondary sources employed to operationalize skilled and low skilled tariff are
listed in detail in Milner and Mukherjee (2009b). These sources are not listed here to save space.
13. The data to calculate the S/L ratio are taken from Milner and Mukherjee (2009b);
the sources employed to compute the S/L ratio are also listed in detail in Milner and
Mukherjee (2009b).
14. See, e.g., Nunn and Trefler (2006). Output-weighted tariffs are used because they accu-
rately capture the relative weight (and thus importance) in the economy of each industry
that produces tradable goods.
 Bumba Mukherjee   277

15. Data on output to calculate output-weights for goods produced by each industry at
the three-digit ISIC level listed in table 14.2 are drawn from Milner and Mukherjee
(2009b).
,i on the left of labor mobility refers to the number of job changes
16. The expression E tj ,i − E tj −δ
between t and t − δ while the expression Σ Nj=1E tj ,i − Σ Nj=1E tj ,−i δ on its left refers to the number of
job losses or gains not offset by a gain or loss in other sectors.
17. For Brazil and Indonesia, see UNCTAD (2005).
18. Their criteria for a democracy are that (i)  the chief executive and legislature must be
directly elected, (ii) there must be more than one party in the legislature, and (iii) incum-
bents must allow a lawful alternation of office if defeated in elections.
19. See, e.g., Milner and Mukherjee (2009a).

References
Acemoglu, D., and J. A. Robinson. 2006. Economic Origins of Dictatorship and Democracy.
New York: Cambridge University Press.
Adsera, A., and C. Boix. 2002. Trade, Democracy, and the Size of the Public Sector: The Political
Underpinnings of Openness. International Organization 56 (2): 229–262.
Ardelean, A., and C. L. Evans. 2013. Electoral Systems and protectionism: An Industry-Level
Analysis. Canadian Journal of Economics 46 (2): 725–764.
Baker, A. 2005. Who Wants to Globalize: Consumer Tastes and Labor Markets in a Theory of
Trade Policy Beliefs. American Journal of Political Science 49 (4): 924–938.
Baldwin R., and C. Magee. 2000. Is Trade Policy for Sale? Congressional Voting on Recent
Trade Bills, Public Choice 105 (1-2): 79–101.
Boix, Carles. 2003. Democracy and Redistribution. New York: Cambridge University Press.
Busch, M. L., and E. Reinhardt. 2000. Geography, International Trade, and Political
Mobilization in US Industries. American Journal of Political Science 44: 703–719.
Busch, M. L., and E. Reinhardt. 1999. Industrial Location and Protection: The Political and
Economic Geography of U.S. Nontariff Barriers. American Journal of Political Science 43 (4):
1028–1050.
Bussmann, Margit. 2002. Examining Causality Among Conflict, Democracy, Openness, and
Economic Growth. University of Alabama Working Paper.
Davis, C. L., and S. B. Bermeo. 2009. Who Files? Developing Country Participation in GATT/
WTO Adjudication. Journal of Politics 71 (3):1033–1049.
Dutt, P., and D. Mitra. 2006. Labor versus Capital in Trade Policy: The Role of Ideology and
Inequality Journal of International Economics 69 (2): 310–320.
Eichengreen, B., and D. Leblang. 2008. Democracy and Globalization. Economics and Politics
20 (3): 289–334.
Frye, T., and E. D. Mansfield. 2004. Timing Is Everything: Elections and Trade Liberalization in
the Post-Communist World, Comparative Political Studies 37 (May): 371–398.
Garrett, G. 2000. The Causes of Globalization. Comparative Political Studies 33: 941–991.
Geddes, B. 1995. The Politics of Economic Liberalization. Latin American Research Review
30 (2): 195–214.
Giavazzi, F., and G. Tabellini. 2005. Economic and Political Liberalization. NBER Working
Paper No. 10657.
278   Democracies

Goldberg, P., and N. Pavcnik. 2004. Trade, Inequality, and Poverty:  What Do We Know?
Evidence from Recent Trade Liberalization Episodes in Developing Countries. NBER
Working Paper No. 10593.
Gowa, J., and S. Y. Kim. 2005. An Exclusive Country Club: The Effects of the GATT on Trade,
1950–1994. World Politics 57 (July): 453–478.
Grossman, G. M., and E. Helpman. 2006. Interest Groups and Trade Policy. Princeton,
NJ: Princeton University Press.
Grossman, G. M., and E. Helpman. 2005. Protectionist Bias in Majoritarian Politics. Quarterly
Journal of Economics 120 (4): 1239–1282.
Guisinger, A. 2001. The Determinants of Trade Tax Liberalization. Unpublished manuscript,
Yale University.
Hankla, C. 2006. Party Strength and International Trade:  A  Cross National Analysis.
Comparative Political Studies 39 (9): 1133–1156.
Harrison, A., and G. Hanson. 1999. Who Gains from Trade Reform? Some Remaining Puzzles.
Journal of Development Economics 59 (June): 125–154.
Henisz, W. J., and E. D. Mansfield. 2006. Votes and Vetoes:  The Political. Determinants of
Commercial Openness. International Studies Quarterly 50 (1): 189–211.
Hiscox, M. J. 2002. International Trade and Political Conflict:  Commerce, Coalitions, and
Mobility. Princeton, NJ: Princeton University Press.
Hiscox, M. J., and S. Rickard. 2005. Birds of a Different Feather? Varieties of Capitalism, Factor
Specificity, and Inter-industry Labor Movements. Unpublished manuscript, London School
of Economics.
Kono, D. Y. 2006. Optimal Obfuscation: Democracy and Trade Policy Transparency. American
Political Science Review 100: 369–384.
Li, Q., and R. Reuveny. 2003. Economic Globalization and Democracy: An Empirical Analysis.
British Journal of Political Science 33 (1): 29–54.
Lipset, S. M. 1959. Some Social Requisites of Democracy: Economic Development and Political
Development. American Political Science Review 53: 69–105.
Lu, X., K. Scheve, and M. J. Slaughter. 2012. Inequity Aversion and the International
Distribution of Trade Protection. American Journal of Political Science 56 (3): 638–654.
Lopez-Cordova, E., and C. Meissner. 2008. The Globalization of Trade and Democracy:
A Long-run Perspective. World Politics 60 (4): 539–575.
Mansfield, E. D., and M. Busch. 1995. The Political Economy of Nontariff Barriers: A
Cross-National Analysis International Organization 49: 723–749.
Mansfield, E. D., H. V. Milner, and B. P. Rosendorff. 2000. Free to Trade:  Democracies,
Autocracies, and International Trade. American Political Science Review 94 (2): 305–321.
Mansfield, E. D., and D. Mutz. 2009. Support for Free Trade: Self-Interest, Sociotropic Politics,
and Out Group Anxiety. International Organization 63 (3):425–457.
Mayda A. M., and D. Rodrik. 2005. Why Are Some People (and Countries) More Protectionist
Than Others? European Economic Review 49 (6):1393–1430.
McGillivray, F. 2004. Privileging Industry:  The Comparative Politics of Trade and Industrial
Policy. Princeton, NJ: Princeton University Press.
McKinnon, R. 1991. The Order of Economic Liberalization: Financial Control in the Transition to
a Market Economy. Baltimore, MD: John Hopkins University Press.
Milner, H. V., and B. Judkins. 2004. Partisanship, Trade Policy, and Globalization: Is there a
Left–Right Divide on Trade Policy? International Studies Quarterly 48 (1): 95–119.
 Bumba Mukherjee   279

Milner, H. V., and K. Kubota. 2005. Why the Move to Free Trade? Democracy and Trade Policy
in the Developing Countries. International Organization 59: 107–143.
Milner, H. V., and B. Mukherjee. 2009a. Democratization and Economic Globalization.
Annual Review of Political Science 12: 163–181.
Milner, H. V., and B. Mukherjee. 2009b. Democracy and Skill-Bias in Trade Protection in
Developing Countries. Princeton University, Department of Politics Working Paper.
Mukherjee, B., D. L. Smith, and Q. Li. 2009. Labor (Im)mobility and the Politics of Trade
Protection in Majoritarian Democracies. Journal of Politics 71 (1): 1–18.
Nielson, D. 2003. Supplying Trade Reform:  Political Institutions and Trade Policy in
Middle-Income Democracies. American Journal of Political Science 47: 470–491.
Nunn, N., and D. Trefler. 2006. Putting the Lid on Lobbying: Tariff Structure and Long-Term
Growth When Protection Is for Sale. NBER Working Paper No. 12164.
Papaioannou, E., and G. Siourounis. 2004. Democratization and Growth. Unpublished manu-
script, London Business School.
Przeworski, A., M. E. Alvarez, J. A. Cheibub, and F. Limongi. 2000. Democracy and Develop­
ment: Political Institutions and Well-being in the World, 1950–90. New  York: Cambridge
University Press.
Remmer, K. 1998. The Politics of Neoliberal Economic Reform in South America, 1980–1994.
Studies in Comparative International Development 33 (2): 3–29.
Rogowski, R. 1989. Commerce and Coalitions: How Trade Affects Domestic Political Alignments.
Princeton, NJ: Princeton University Press.
Rudra, N. 2005. Globalization and the Strengthening of Democracy in the Developing World.
American Journal of Political Science 49: 704–730.
Scheve, K., and M. Slaughter. 2001. What Determines Trade Policy Preferences? Journal Of
International Economics 54 (2): 267–292.
Shafaeddin, M. 2006. Does Trade Openness Favor or Hinder Industrialization and
Development? Working Paper., Kuala Lumpur, Malaysia: TWN Trade and Development
Series.
Stokes, S. 2001. Mandates and Democracy: Neoliberalism by Surprise in Latin America. New York:
Cambridge University Press.
Tavares, J. 2008. Trade, Factor Proportions, and Political Rights. Review of Economics and
Statistics 90 (1): 163–168.
Tomz, M., J. L. Goldstein, and D. Rivers. 2007. Do We Really Know That the WTO Increases
Trade? Comment. American Economic Review 97 (5): 2005–2018.
UNCTAD. 2005. Trade and Development Report. New York and Geneva: United Nations.
Wacziarg, R. and J. Wallack. 2004. Trade liberalization and intersectoral labor movements,
Journal of International Economics 64(2): 411–439.
Weyland, Kurt. 2002. The Politics of Market Reform in Fragile Democracies: Argentina, Brazil,
Peru, and Venezuela. Princeton, NJ: Princeton University Press.
World Bank. 2011. World Development Indicators. Washington, DC: World Bank.
 Chapter 15

Electoral Syst e ms
and Tr a de

Ste phanie J . Ric kard

In democracies, politicians compete to win votes, and subsequently office, in free

and fair elections. Elected representatives should therefore be responsive to citizens’
demands. Yet leaders in some democracies are more reactive to pressures for trade pro-
tection than are leaders in other countries. The varied responsiveness of elected officials
to protectionist demands may result from the dissimilar rules that govern democratic
elections around the world.
A growing body of scholarship examines the relationship between electoral rules
and trade policy. Despite increased attention to electoral institutions, no consensus
exists about which system makes politicians most responsive to protectionist demands.
Some argue that majoritarian systems provide legislators with the greatest incentives
to cater to demands for trade barriers (e.g., Rogowski 1987; Grossman and Helpman
2005). Others contend that proportional electoral systems lead politicians to be most
responsive to protectionists (e.g. Mansfield and Busch 1995; Rogowski and Kayser 2002;
Hays 2009). The debate continues unresolved as both arguments find credible empirical
support.
Conflicting conclusions about the effects of electoral systems are surprising given
that these institutions lie at the heart of democratic politics. Elections, and accordingly
electoral rules, are a defining feature of democracy. Electoral systems connect voters to
policy makers; thus, it is reasonable to expect different rules to produce different policy
outcomes. Why then have scholars reached inconsistent inferences about the effects
of electoral systems on trade policy? This question is the central focus of this chapter.
Understanding why scholars have reached this stalemate is important if future research
is to progress beyond it. Failure to resolve the impasse may result in the premature
demise of this research agenda.
The following section outlines the contours of the debate. Potential reasons for the
disagreement about the relationship between electoral systems and trade policy are dis-
cussed. The concluding section suggests possible ways forward.
 Stephanie J. Rickard    281

Defining Electoral Systems

An electoral system is a set of laws and regulations that govern electoral competition
between candidates or parties or both (Cox 1997, 38). These procedures comprise a mul-
titude of items, such as the electoral formula, the ballot structure, and district magni-
tude. Of these items, the electoral formula has received the most attention from trade
scholars. Electoral formulas refer to the method by which vote totals translate into
claims upon legislative seats (Cox 1990). Two main categories of electoral formulas
exist: majoritarian and proportional. Together they govern 80 percent of elections held
across the world (Clark, Golder, and Golder 2012; Inter-Parliamentary Union 2013).
Majoritarian rules stipulate that the candidates or parties who receive the most
votes win. Votes are cast for individual candidates, and the top k vote-recipients win
seats, where k is the magnitude of the district (Cox 1990, 906). The most commonly
used majoritarian formula is the single-member district plurality system (SMDP). In an
SMDP system, citizens cast one vote for a candidate in a single-member district, and the
candidate with the most votes is elected. This system is used, for example, in the United
States. Forty-five percent of countries use some type of majoritarian electoral formula
(Clark, Golder, and Golder 2012, 590).
Majoritarian electoral formulas typically help the largest party obtain a legislative
majority. For example, in the post-World War II period, governments in Great Britain
received 45 percent of the popular vote on average, but 54 percent of the legislative seats
(Norris 1997). In contrast, proportional electoral rules (PR) are intended to produce
proportional outcomes (Cox 1997). In a fully proportional system, a party that won
45 percent of the vote would win 45 percent of the legislative seats. In order to achieve
proportional outcomes, PR systems employ multimember districts and a quota- or
divisor-based electoral formula that determines the number of votes that a candidate or
party needs to win a seat (Clark, Golder, and Golder 2012, 564). Thirty-seven percent of
countries use some type of proportional electoral formula.
While most countries employ either a majoritarian or a proportional electoral formula,
nearly one-fifth of countries employ a mixed system. Mixed systems combine majoritar-
ian and proportional formulas in the same election. In such elections, voters select repre-
sentatives through two different electoral formulas that run alongside each other.

The Protectionist Bias

in Majoritarian Systems

The idea that trade openness is systematically related to countries’ electoral systems
has a long intellectual history. Early studies by Ronald Rogowski (1987) and Peter
Katzenstein (1985) suggested a natural affinity exists between trade openness and
282   Electoral Systems and Trade

proportional electoral rules. More recently, Grossman and Helpman (2005) developed
a theoretical model that envisages a protectionist bias in majoritarian systems. In this
seminal model, two parties compete in legislative elections, and each party has an equal
chance of winning a given seat in a given district. There are three electoral districts;
each contains one-third of the population and elects one legislator. Legislators repre-
sent districts with interests tied to district-specific industries. Upon forming the gov-
ernment, the delegation from the majority party seeks to maximize the welfare of its
constituents. If the party in power represents all three districts, then the legislature will
work to maximize the welfare of the entire country by setting tariffs at zero. In contrast,
if the governing party represents only two of the three districts, it will set a positive tar-
iff rate. A nonzero tariff emerges because trade policy is chosen by the majority delega-
tion, and legislators in the minority have limited means to influence policy decisions.
Legislators in the majority use tariffs to redistribute income to industries in their own
districts, rather than maximizing national welfare by setting an optimal tariff of zero.
In majoritarian systems, the governing party is unlikely to represent all three districts,
whereas in proportional systems it is more likely to represent all three. Consequently,
the prediction emerges that tariffs will be lower in PR systems than in majoritarian
systems.
Several studies provide empirical support for the predicted “protectionist bias” in
majoritarian systems. In a sample of 147 countries from 1981 to 2004, Evans (2009) finds
that majoritarian systems have higher average tariffs than countries with proportional
electoral systems, all else being equal. Similarly, Ardelean and Evans (2013) demonstrate
that tariffs are higher on average in majoritarian systems than in proportional systems
using product-level tariff rates for a cross-section of developed and developing countries
between 1988 and 2007. These results suggest that a protectionist bias exists in majoritar-
ian systems with regard to tariffs, but leave unanswered the question of whether or not
this bias extends to other forms of trade protection.
Governments increasingly use nontariff barriers (NTBs), rather than tariffs, to pro-
tect domestic producers. Studies that examine only tariffs, such as Evans (2009), leave
open the possibility that electoral rules have an ambiguous effect on total trade protec-
tion. Governments in majoritarian systems may provide no more (or less) protection
than those in proportional systems; instead, majoritarian governments may simply use
tariffs more often than other forms of trade protection.
Different types of governments may prefer different forms of trade barriers because
they entail different cost-benefit analyses. Tariffs are a relatively blunt policy tool; they
typically cover a specific product and thereby affect all firms producing that product.
Although tariffs are blunt and thus potentially inefficient, they generate revenue for gov-
ernments. In contrast, subsidies cost governments money.1 While subsidies are expen-
sive, they can be targeted narrowly. For example, a subsidy can be provided to a single
firm rather than to an entire industry. Different governments may prefer different types
of trade barriers given their varied policy goals and budget constraints. An important
but unanswered question therefore exists; does the protectionist bias in majoritarian
politics extend beyond tariffs to other forms of trade protection?
 Stephanie J. Rickard    283

A recent study suggests that the protectionist bias in majoritarian politics extends to
subsidies, which are an increasingly common NTB (Ford and Suyker 1990; OECD 1998;
Rickard 2012c). Rickard (2012b) reports that the share of total government expenditures
allocated to subsidies is higher in countries with majoritarian electoral rules than in
those with proportional electoral rules, holding all else equal. On average, governments
in majoritarian systems spend 2.5 percentage points more on subsidies than govern-
ments in PR systems (Rickard 2012b). This evidence suggests that the protectionist bias
in majoritarian politics does, in fact, extend beyond tariffs to subsidies.
The majoritarian protectionist bias appears to extend even further, to “illegal” forms
of trade protection. Democracies with majoritarian electoral rules are named as defen-
dants in disputes litigated via the World Trade Organization (WTO) more often than
those with proportional electoral rules (Rickard 2010; Davis 2012). One interpretation
of this finding is that majoritarian democracies simply have more barriers to trade than
do PR democracies. Another possibility is that majoritarian democracies have relatively
more non-WTO-compliant trade barriers. The electoral incentives to provide protec-
tion may be so compelling in majoritarian systems that legislators are willing to supply
trade barriers even when doing so violates international rules. In contrast, legislators
may choose to provide only WTO-compliant protection in countries where electoral
systems are proportional (Naoi 2009).
In sum, convincing empirical evidence suggests a protectionist bias exists in majori-
tarian systems. This bias appears to extend beyond tariffs to subsidies and other trade
barriers proscribed by WTO rules.

Trade Protection in
Proportional Rule Systems

Although convincing evidence exists of a protectionist bias in majoritarian systems,

some scholars argue that proportional electoral rules generate higher barriers to trade.
A model developed by Rogowski and Kayser (2002), although not exclusively a model
of trade protection, implies that trade barriers will be higher in PR democracies than in
majoritarian systems. In the Rogowski and Kayser model, the key distinction between
electoral systems is the seat-vote elasticity. Majoritarian systems have greater seat-vote
elasticities than PR systems; as a result, a loss of votes translates into a greater loss of
legislative seats for parties competing in majoritarian systems. In proportional systems,
politicians are able to cater to narrow interests without having to be overly concerned
with any election losses they might incur for doing so. In contrast, politicians in plural-
ity systems cannot stray far from the preferences of the median voter, because a small
change in vote share can produce a large change in seat share. Rogowski and Kayser
posited that politicians in proportional rule systems will therefore be relatively more
responsive to narrow interests, such as industry-specific demands for trade protection.
284   Electoral Systems and Trade

A theoretical model developed by De Mesquita and Smith (2005) also implies that
trade barriers will be higher in PR systems than in majoritarian systems. Their model
examines the political consequences of a winning coalition’s size. A winning coalition
is a subset of the selectorate large enough to allow it to endow leadership with politi-
cal power to negate the influence of the remainder of the selectorate and the disenfran-
chised members of the society (De Mesquita and Smith 2005, 51). The winning coalition
is larger in majoritarian systems than in PR systems, according to De Mesquita and
Smith (2005).2 As the size of the winning coalition grows, the cost of private goods, such
as trade barriers, increases. According to their logic, trade protection should be lower in
majoritarian systems than in PR systems.
Some evidence exists to suggest that proportional electoral rules incentivize relatively
higher trade barriers. Nontariff barriers are higher, on average, in PR democracies than
in majoritarian systems (Mansfield and Busch 1995). Proportional rule systems are also
associated with higher consumer prices (Rogowski and Kayser 2002; Chang, Kayser,
and Rogowski 2008; Chang et al. 2010). Higher consumer prices arguably reflect gov-
ernmental policies that privilege producer groups at the expense of consumers. One
such policy is trade protection. Legislatively imposed barriers to trade raise the prices of
consumer goods. The presence of larger trade barriers in PR countries may explain why
consumer prices are higher in PR systems than in majoritarian systems.
In sum, it remains unclear which electoral formula generates higher levels of
trade protection. Possible reasons for this impasse are explored in the following sec-
tion. Understanding why scholars have reached this stalemate is important for future
research to progress beyond it and advance understanding of how a fundamental demo-
cratic institution impacts trade policy.

Resolving the Debate

One explanation for the mixed empirical results may be that electoral rules have var-
ied effects on different types of trade barriers. Existing studies employ diverse mea-
sures of trade protection; Evans (2009) uses tariffs; Rickard (2012b) uses subsidies;
and Mansfield and Busch (1995) use nonsubsidy, nontariff barriers, such as quotas and
restrictive import licensing requirements. The varied conclusions reached by these
investigations may simply reflect the dissimilar measures of trade protection examined.
Electoral rules might have different effects on different types of trade barriers because
they shape the incentives of parties and politicians to target benefits more or less nar-
rowly (Milesi-Ferretti, Perotti, and Rostagno 2002; Persson and Tabellini 2003; Rickard
2009). In majoritarian systems, electoral incentives exist to narrowly target economic
benefits. With majoritarian rules and single-member districts, a party need only receive
a plurality of votes in half the districts plus one to win an election and form a government.
Politicians need to win only a plurality of votes in their geographically defined electoral
districts. Therefore, politicians’ optimal reelection strategy in majoritarian rule systems
 Stephanie J. Rickard    285

is to supply benefits narrowly to only those voters in their districts (Milesi-Ferretti,

Perotti, and Rostagno 2002; Persson and Tabellini 2003).
In PR systems, politicians and parties have fewer incentives to target benefits nar-
rowly (Lizzeri and Persico 2001; Persson and Tabellini 2003; Rogowski 1987). Parties
competing under PR electoral rules do not win elections district by district. In fact, no
single district is critical to the electoral success of a party (McGillivray 2004). Instead,
parties work to maximize their aggregate vote share, because this determines the num-
ber of legislative seats a party will control. By targeting policies to a broad segment
of the electorate, parties are able to buy the support of a wide range of voters. Thus,
leaders in proportional systems have greater incentives to supply broadly beneficial
policies.
If some forms of trade protection are inherently “narrower” than others (i.e., easier to
target to select constituencies), then electoral rules may have varied effects on different
types of trade protection. Narrow trade barriers should be more prevalent in majoritar-
ian systems than in proportional systems. In contrast, broad trade barriers should be
more widespread in proportional systems. Two key questions arise from this discus-
sion: Are some forms of trade protection inherently “narrower” than others (i.e., eas-
ier to target to select constituencies). If so, which ones? These questions remain largely
unexplored.3 Nonetheless, it is easy to imagine that some forms of trade protection may
be more targetable than others. Subsidies, for example, can be provided to a single firm.
In contrast, restrictive import licenses often apply to products that typically benefit
more than just one firm. In this example, subsidies should be higher in majoritarian sys-
tems, while import licenses should be more restrictive in PR systems. Findings reported
by Rickard (2012b) and Mansfield and Busch (1995) provide preliminary support for
these expectations.
The idea that electoral rules may have varied effects on different types of trade barriers
has several implications. First, it suggests a straightforward explanation of why existing
studies have come to conflicting conclusions about the empirical relationship between
electoral rules and trade protection:  it depends on which type of trade barrier they
examine. Second, this idea offers a possible explanation of why governments choose a
particular policy instrument from their vast arsenal of policy tools. Why, for example,
do some governments choose to use discriminatory procurement practices rather than
tariffs to privilege domestic producers?4 Choice of policy instrument remains an impor-
tant research agenda, because even today no form of trade protection is trivial: tariffs
still account for approximately 30 percent of global protection, while NTBs and subsi-
dies account for the remaining 70 percent (Kono 2009).
The idea that electoral rules may have varied effects on different types of trade protec-
tion implies that the challenge of relating electoral rules to policy outcomes might be
unique to trade because of the myriad tools governments can use to protect domestic
producers from competition with low-cost foreign imports. However, research in other
areas is plagued by similar ambiguity. For example, it remains unclear which electoral
formula generates better outcomes in terms of inflation, economic growth, or budget
deficits (Taagepera and Qvortrup 2011, Table  1). The mixed effects of electoral rules
286   Electoral Systems and Trade

across a multitude of policy areas suggest that the problem may reside on the right-hand
side of the equation rather than the left.

Variation within Electoral Systems

The simple distinction between majoritarian and proportional electoral rules may be
too blunt to explain the cross-national variation in trade protection. Most investigations
of trade policy use a rudimentary measure of electoral rules; that is, a dichotomous vari-
able that takes PR (and majoritarian) electoral systems to be a monolithic phenomenon.
Yet what trade scholars typically characterize as being one uniform phenomenon in fact
contains multiple structures each with different dynamics and political implications.
For example, among systems classified as being majoritarian, eight different electoral
formulas are used to translate votes into seats (Clark, Golder, and Golder 2012, 543).
Similar diversity exists among PR systems-even though most can be characterized as
list systems.5 In a list PR system, each party presents a list of candidates for a multimem-
ber district, and parties receive seats in proportion to their overall share of the votes.
Despite sharing this common feature, list PR systems vary significantly. Seven distinct
electoral formulas are used to allocate seats to parties in list PR systems. List PR systems
also differ in their district magnitude, the use of higher electoral tiers, the use of elec-
toral thresholds, and the type of party list employed (Gallagher, Laver, and Mair 2006,
354). These dissimilar proportional systems are often grouped together in a single cat-
egory labeled “PR”.
Mixed electoral systems introduce even more variance. Mixed electoral systems
combine majoritarian and proportional electoral formulas in the same election. Many
scholars simply include mixed systems together with “pure” systems for convenience.6
Germany, for example, is classified as being a mixed-proportional system because the
total number of legislative seats received by a party is proportional to its list-tier results
(Shugart and Wattenberg 2001; Thames and Edwards 2006). Germany is therefore
grouped together with pure PR systems in many empirical studies (e.g., Rickard 2012a).
Because of this common practice, the dichotomous “PR versus majoritarian” variable
includes even greater diversity than suggested by the already significant variation in
proportional (or majoritarian) systems alone.
This diversity raises questions about the usefulness of the blunt distinction between
PR and majoritarian systems. Taagepera and Qvortrup (2011) warn that only non-
specialists can persuade themselves that all electoral systems can be characterized by
a single dichotomous indicator. They call the majoritarian/proportional distinction a
“procrustean bed” (Taagepera and Qvortrup 2011, 255) and warn researchers to “forget
about such coarse dichotomy” (Taagepera and Qvortrup 2011, 253). Scholars of trade
politics should heed this warning and take seriously the variation that exists within
majoritarian and proportional electoral systems. Doing so may help to clarify the pre-
cise relationship between electoral systems and trade policy. For example, some types of
 Stephanie J. Rickard    287

majoritarian systems may generate higher levels of trade protection than some types of
proportional systems. Such within-system variance could explain the mixed empirical
results reported to date and help to clarify the precise mechanism through which elec-
toral institutions influence trade policy. If some majoritarian systems generate higher
levels of trade protection than others, a common feature may characterize these particu-
lar majoritarian systems, and it may be this shared attribute that influences trade policy,
rather than the electoral formula itself. Several such possible traits are explored in the
following section.

Identifying Causal Mechanisms

The causal mechanism linking electoral formulas to trade policy remains unclear. Few
theories articulate how the translation of votes into seats affects trade policy.7 Instead,
scholars typically refer to other features of a country’s electoral system that tend to
covary with electoral formula, such as district size. Rogowski (1987), for example, argues
that the large electoral districts that typify proportional systems insulate politicians
from protectionist demands. The implication is that the electoral formula itself may not
matter much for trade policy; instead, other features of countries’ electoral systems that
tend to go together with electoral formulas, such as district size, are important.

District Size
District size may help to explain the apparent protectionist bias in majoritarian coun-
tries. A  district’s size is the number of people living in an average electoral district.
District size is assumed to increase with district magnitude (i.e., number of represen-
tatives elected in a district). The single-member districts that characterize majoritar-
ian systems are presumed to be smaller than the multimember districts used in PR
systems.8 Smaller districts are understood to produce higher levels of trade protection,
because they give protectionists greater influence over elected representatives (e.g., Alt
and Gilligan 1994; Mansfield and Busch 1995; McGillivray 2004). McGillivray (2004,
28) provides the following illustrative example. An industry with 100 employees rep-
resents 10 percent of the electorate in a district with 1,000 voters. The same industry
represents only 0.1 percent of the electorate in a district of 100,000 voters. In the larger
district, refusing to protect the industry is unlikely to affect the politician’s reelection
chances, because the industry is only 0.1 percent of the representative’s electorate. Given
this, politicians elected from smaller districts are more likely to supply trade protection
than politicians elected from larger districts.
Few empirical studies test this claim. One of the only cross-national measures of dis-
trict size is provided by Hankla (2006). He divides the total number of seats in a coun-
try’s lower legislative chamber by mean district magnitude and divides that number into
288   Electoral Systems and Trade

the country’s total population to estimate the number of people living in each electoral
district. This variable most closely measures the concept of district size articulated by
McGillivray (2004) and Rogowski (1987). Hankla reports a negative correlation between
this measure and import duty coverage ratios: “District size can explain decreases in
import duty coverage ratios of more than 3.5 percent” (2006, 1149). In other words,
larger districts correspond with lower levels of trade protection, as expected. Similarly,
Rogowski (1987) reports a negative correlation between the number of parliamentary
constituencies and trade openness. He asserts that the number of electoral districts is
“an inverse measure of average constituency size.” Rogowski’s evidence also suggests
that larger districts are associated with lower trade barriers.
A series of studies makes use of the differences in constituency size that occur within
the United States. Using tariff votes from the US Senate in the late nineteenth and early
twentieth centuries, Hauk (2011) finds that industries concentrated in smaller constit-
uencies receive more trade protection than those located in larger constituencies. In
contrast, Karol (2007) finds no correlation between constituency size and protection-
ism in the United States in recent decades. He concludes that constituency size does not
account for the differences in preferences among the House, Senate, and presidency on
trade issues.
Although most arguments relate district size to trade protection, it is possible that
district magnitude, that is, the number of representatives elected per district, influences
trade policy. Single-member districts allow voters to assign credit (or blame) for trade
barriers. In multimember districts, however, voters observe the total amount of protec-
tion provided to the district but not the amount produced by individual legislators. As
a result, voters do not know which of their representatives to credit for providing trade
protection (Ashworth and Bueno de Mesquita 2006). Thus, the electoral benefits of pro-
viding protection are relatively lower in multimember districts. If many legislators can
claim responsibility for a trade policy with local ramifications, each individual’s incen-
tives to provide such policies decrease (Lancaster 1986). Politicians in multimember dis-
tricts may thus provide less trade protection than politicians in single-member districts
(Magee, Brock, and Young 1989).
As anticipated by this logic, democracies with single-member districts appear to have
more “illegal” trade barriers than those with multimember districts, holding all else
constant (Rickard 2010). The number of WTO disputes filed against democracies with
single-member districts is 186  percent higher, on average, than against democracies
with multimember districts, all else being equal (Rickard 2010). Moving from a multi-
member district system with seven seats on average to a single-member district system
increases the probability of being named as a defendant in a GATT/WTO dispute by
more than 6 percentage points in a given year (Rickard 2010). This evidence suggests
that district magnitude has a direct effect on trade policy.
District magnitude may also have an indirect effect on trade policy by mediating
the influence of electoral formulas. The effects of a proportional electoral formula may
depend on the district magnitude (Carey and Shugart 1995; Carey and Hix 2011). When
district magnitude is high, electoral systems are relatively more proportional, because
 Stephanie J. Rickard    289

smaller parties are more likely to win seats (Cox 1997; Rae 1967; Taagepera and Shugart
1989). A party would need to win more than 25 percent of the vote to guarantee a seat
in a three-seat district, but it would need to win only a little more than 10 percent of the
vote to guarantee winning a seat in a nine-seat district. The electoral system is likely
to be disproportional whenever the district magnitude is small, irrespective of the par-
ticular formula used to translate votes into seats. For example, when PR is used in very
small districts, as in Australia or Ireland, its effects become similar to those of plurality
elections. For this reason, many political scientists, such as Duverger (1964), argue that
district magnitude is the single most important dimension by which electoral systems
differ.

The Nature of Electoral Competition

District magnitude cannot, however, explain the nature of electoral competition, which
has been shown to influence trade policy (Nielson 2003). Electoral competition is char-
acterized as being either candidate centered or party centered (Carey and Shugart 1995).
Party-centered competition encourages voters to emphasize their party preference over
that for specific candidates. In contrast, candidate-centered competition encourages
the voter to see the basic unit of representation as the candidate rather than the party
(Shugart 1999, 70).
Candidate-centered electoral competition can emerge from either single-member or
multimember districts. In multimember districts with open party lists, voters are able
to indicate their preferred party and also their favored candidate within that party. As a
result, candidates have incentives to appeal directly to voters. Candidates may go against
the interests of their party to curry favor with constituents in their districts. Similar
incentives exist in single-member districts, like those in the United States, where can-
didates are rewarded by voters for bringing pork-barrel projects home to their districts.
In candidate-centered systems, the optimal reelection strategy is to cultivate a per-
sonal vote (Shugart 1995). A personal vote occurs when an individual votes based on the
characteristics of a particular candidate rather than the characteristics of the party to
which the candidate belongs (Carey and Shugart 1995). To develop a personal vote, rep-
resentatives can provide private or local public goods and services to their geographi-
cally defined constituents. Subsidies and other trade barriers are expedient means by
which to achieve this goal. As such, trade protection may rise as politicians’ incentives to
cultivate personal votes increase (Nielson 2003).
Few incentives exist for politicians to cultivate a personal vote in party-centered
systems, because voters emphasize their party preferences over those for specific can-
didates. Closed-list PR systems, for example, engender party-centered competition
because voters are not able to express a preference for a particular candidate. Ballot
papers in closed-list systems often do not even contain the names of individual candi-
dates. Voters select a party and parties then receive seats in proportion to the number
of votes that they obtain. These seats are filled by the party using a predetermined list of
290   Electoral Systems and Trade

candidates. On this list, candidates are rank-ordered by party leaders. Candidates closer
to the top of the list are more likely to get a seat in the legislature. Therefore, candidates’
best electoral strategy is to work to promote the party’s national popularity. Doing so
maximizes the party’s vote share and candidates’ own positions on the party list. Because
politicians have few incentives to cultivate personal votes, tariffs will tend to be lower in
party-centered systems than in candidate-centered systems (Nielson 2003).
In a sample of eighteen middle-income presidential democracies, Nielson (2003)
finds that both collected tariffs and official tariff rates are higher in candidate-centered
systems than in party-centered systems. This evidence suggests that the nature of elec-
toral competition influences trade policy. The distinction between candidate- and
party-centered systems may prove to be more useful for understanding trade policy out-
comes than the PR/majoritarian dichotomy.

Party Strength
The nature of electoral competition is often conflated with party strength.9 However,
these two concepts are distinct. Party strength typically refers to party discipline, which
is defined as the ability of a political party to get its legislators to support the policies
of the party’s leadership. Party discipline is a legislative phenomenon; party-centered
competition is an electoral arena phenomenon. Cox (1987) argues in support of the
coevolution of these two concepts, but other scholarship shows only a weak relationship
between electoral competition and the party-centered nature of the legislature (Martin
2014). Candidate-centered electoral competition exists, for example, in systems with
high party discipline, as is the case in Ireland. It is also possible to have low levels of party
discipline in party-centered electoral systems (e.g., in South Korea). Scholars must be
careful not to conflate party-centered electoral competition with party strength when
theorizing about their potential effects on trade policy. Conflating these two concepts
adds further confusion to an already muddled field of inquiry.
Party discipline has been proposed as an explanation for the relatively low levels of
trade protection in proportional electoral systems (e.g., Rogowski 1987). Such argu-
ments typically assume that PR systems produce disciplined parties that insulate politi-
cians from protectionist interests. However, the origins of party discipline are unclear.
While electoral systems may have some effect on party discipline, many other factors
also matter, including, for example, regime type, party organization structures, and the
allocation of power inside legislatures. In short, proportional electoral rules are neither
necessary nor sufficient to engender high levels of party discipline.10
Arguments that identify party discipline as a key causal mechanism therefore can-
not be tested using measures of proportionality. Instead, it is necessary to develop
cross-country indicators of party discipline. Such measures can then be included in
empirical models along with measures of countries’ electoral formula, subjecting
them to various statistical horse races. Empirical tests have been conducted on limited
Organisation for Economic Co-operation and Development (OECD) data. Ehrlich
 Stephanie J. Rickard    291

(2007), for example, initially finds that proportional electoral rules have a robust nega-
tive effect on tariffs.11 Specifically, he reports that PR is associated with 7.3 percent lower
tariffs in the long run in a sample of twenty-one OECD countries, from 1948 to 1994
(Ehrlich 2007, 595). However, after controlling for party strength (and the number of
electoral districts), Ehrlich finds that electoral rules no longer have a robust long-run
effect on tariffs. Failing to control for party strength may lead to inaccurate conclusions
about the role of electoral rules in trade policy making.

Interests and Institutions

A concluding explanation for the mixed results found to date is that electoral institutions
alone cannot explain why some countries are more open to trade than others. Instead,
the impact of electoral rules on trade policy may depend on the nature of the electorate,
or what Rae called “the surrounding envelope of societal forces” (1971, 167) (Rogowski
1987, 210). Electoral institutions aggregate citizens’ interests. Therefore, to fully under-
stand the effects of electoral rules on trade policy, it may be necessary to understand
voters’ economic interests. Although a growing body of literature examines individuals’
preferences about trade,12 such individual-level analyses have not yet been integrated
into research on electoral institutions. However, two recent studies link economic inter-
ests, electoral rules, and trade policy.
Rickard (2009) argues that the electoral benefits of providing trade protection are
jointly determined by a country’s electoral rules and voters’ preferences regarding trade
policy. Voters’ preferences about trade are shaped by their mobility. Voters who find it
prohibitively costly to move to a new job prefer policies that target protection only to
their current industry of employment. In contrast, voters able to move easily between
industries are less interested in narrowly targeted protection. Politicians’ responsiveness
to these demands is a function of a country’s electoral system.
Examining the prevalence of countervailing duties and narrow trade barriers alleged
to violate WTO rules, Rickard (2009) finds that policy makers in PR systems are more
responsive to increases in demand for narrow trade barriers. The implication is that
under certain conditions, trade protection will be higher in PR countries than in majori-
tarian countries. Specifically, when voter demand for narrow trade barriers is high, pro-
tection will be greater in PR countries than in majoritarian countries. When demand for
trade protection is low, majoritarian systems will tend to have greater trade protection
than PR systems. This study demonstrates how voters’ economic interests can help to
resolve the debate over which electoral system is most prone to protectionism.
In a subsequent study, Rickard (2012a) argues that the geographic concentration of
protectionist interests mediates the impact of electoral rules on trade policy. When pro-
tectionists are geographically diffuse, politicians in PR systems will be more responsive
to their demands than politicians in majoritarian countries. Governmental spend-
ing patterns in fourteen countries over a twenty-year period provide support for this
292   Electoral Systems and Trade

conditional argument. When voters with a shared economic interest in industrial subsi-
dies are geographically diffuse, spending on subsidies constitutes a larger share of gov-
ernment expenditures in PR systems than in majoritarian systems. In contrast, spending
on subsidies constitutes a larger share of government expenditures in majoritarian sys-
tems when voters with an interest in subsidies are geographically concentrated.13 In
short, the geographic dispersion of protectionist interests provides a potential bridge
between two prominent, rival arguments about the effects of electoral systems on trade
policy and specifies the conditions under which one is more appropriate than the other.
Together, these studies point to the value of considering voters’ economic interests
and geographic location along with electoral institutions to understand trade policy
outcomes.

Going Forward

No consensus exists about which electoral system makes politicians most responsive to
protectionist demands. Explanations for the conflicting conclusions reached to date are
put forward in this chapter. By means of this discussion, several suggestions emerge for
future research. First, trade scholars must take seriously the variation that exists within
proportional and majoritarian electoral systems. To this end, an expedient first step
would be to move away from using the dichotomous PR/majoritarian variable to “mea-
sure” countries’ electoral institutions. Credible alternative measures exist (e.g. Gallagher
1991; Johnson and Wallack 2012). However, before adopting an off-the-shelf-measure,
researchers should think carefully about precisely what aspect of a country’s electoral
system matters for trade policy outcomes. The widespread use of the dichotomous
PR/majoritarian variable has allowed scholars to obfuscate the exact mechanism link-
ing electoral institutions and trade policy. To advance this research agenda and move
beyond the current impasse, scholars need to think carefully about precisely how elec-
toral institutions influence trade policy.
Rather than trying to connect trade policy outputs to broad labels such as “propor-
tional representation,” scholars should investigate the connection of policy outputs to
theoretically motivated and potentially intervening variables, such as district magni-
tude. One way to achieve greater clarity empirically would be to specify unique impli-
cations of each causal mechanism beyond a simple correlation between electoral rules
and trade barriers—for example, by explaining which type of barrier politicians will
prefer under different electoral institutions. Alternatively, it would be useful to find
cross-country indicators that capture one causal claim or another and use them as inde-
pendent variables in equations estimating trade protection levels, subjecting them to
various statistical horse races.14
One way to shed new empirical light on potential causal mechanisms is to isolate the
effects of specific institutions, as suggested above. Investigating institutions individu-
ally helps identify which dimensions of institutions matter for which outcomes. This
 Stephanie J. Rickard    293

type of knowledge will enable better theories of institutions to be developed and practi-
cal policy recommendations to emerge (Taagepera and Qvortrup 2011, 654–655). The
strategy of “unbundling institutions”—that is, understanding the role of specific com-
ponents of the broad bundle of laws and regulations that make up a country’s electoral
system—may be critically important.
An alternative approach is to consider the combined effect of “institutional bundles”
in an effort to build a general theory that can incorporate multiple institutional differ-
ences within a single dimension. Several such general theories exist, such as veto players
(Tsebelis 2011) and access points (Ehrlich 2007). These theories incorporate the effects
not only of electoral institutions but of other institutional features as well, such as the
executive-legislative relationship and the policy-making bureaucracy. For scholars
interested specifically in the effect of electoral institutions on policy outcomes, addi-
tional work is needed. Future research may make progress by specifying how various
features of a country’s electoral system work together to influence legislators’ electoral
incentives and subsequent policy decisions.
Finally, studies of electoral institutions must not neglect the importance of interests.
Electoral institutions aggregate interests; therefore, to fully understand the effects of
institutions, such as electoral rules, one must understand citizens’ economic interests.
Important advances have been made in understanding voters’ preferences about trade
policy.15 Building on these developments may help engender an improved understand-
ing of the interactive effects of interests and institutions on trade policy.

Conclusion

Research on trade and electoral systems has come full circle. Initial arguments linking
electoral systems and foreign trade proposed a causal connection that ran from coun-
tries’ dependence on international trade to leaders’ choice of electoral institutions. More
recent work assumes that the choice of electoral systems is exogenous to trade. Once
instituted, electoral systems are believed to generate incentives for leaders to be more or
less responsive to protectionist demands. While these two arguments are not mutually
exclusive, they highlight the causal complexities that plague research on this topic.
A requisite question is whether a robust correlation exists between electoral systems
and trade policy. No consensus has yet emerged on this issue. However, a second, even
more challenging, question exists. Is it possible to assess whether electoral rules “cause”
distinct trade policy outcomes, given the myriad unobservable factors that drive the
selection of electoral systems and trade policies? Ultimately, to know this we would need
to answer the counterfactual question: If we picked a country at random and went back
in history to change its electoral rules, how would this alter its current trade policies?
The problem, of course, is that we cannot observe the relevant counterfactual.
These difficulties should not lead scholars to abandon research on electoral systems
and trade. Understanding how a fundamental democratic institution influences policy
294   Electoral Systems and Trade

outcomes is an important and valuable research agenda. Besides, there is still much to be
learned from observational data, as demonstrated by the suggestions for future research
offered in this chapter. Although observational data do not typically identify causal
effects, they can uncover interesting patterns and robust relationships, which can then
be used as valuable inputs into theory building. Armed with better theories about pre-
cisely how and why electoral systems might affect trade policy, scholars can then work to
empirically identify causal effects, using, for example, within-country changes in elec-
toral institutions.

Notes
1. See Rickard (2012c) for a discussion of how these fiscal effects matter for governments fac-
ing tight budget constraints.
2. However, Persson and Tabellini (2003) make the opposite claim.
3. However, trade barriers in GATT/WTO disputes are coded as being either narrow or
broad based on a set of criteria in Rickard (2010).
4. On this point, see Kono and Rickard (2014).
5. Some proportional systems do not employ any type of party list. In single transferable vote
systems, candidates’ names appear on the ballot, often in alphabetical order, and voters
rank at least one candidate in order of their own preferences. Candidates who surpass a
specified quota of first-preference votes are immediately elected. In successive counts,
votes from eliminated candidates and surplus votes from elected candidates are reallo-
cated to the remaining candidates until all the seats are filled (Clark, Golder, and Golder
2012, 578).
6. See, for example, Rickard (2012b). However, Davis (2012) cautions against such as a strat-
egy, as some of her findings are sensitive to the way in which mixed systems are coded.
7. The notable exception is Rogowski and Kayser (2002).
8. Although all PR systems employ multimember districts, the magnitude of these dis-
tricts varies significantly from one country to another. In the Netherlands, for exam-
ple, all 150 of the legislators in the lower chamber are elected from a single, national
district. In contrast, Chile elects its legislators in 60 relatively small, two-seat districts.
Even allowing for the cross-national variation in the size of multimember districts, sin-
gle-member districts are on average smaller than multimember districts (Powell and
Vanberg, 2000).
9. For example, party discipline is often estimated using variables that capture the distinction
between candidate-centered and party-centered competition (e.g., Ehrlich 2007).
10. The strength of parties also varies within majoritarian systems (McGillivray 1997, 2004).
For example, national parties in Canada and Great Britain are generally stronger than par-
ties in the United States.
11. McGillivray (1997) finds evidence that strong parties in majoritarian systems provide
trade policies favorable to voters in marginal districts.
12. See Kuo and Naoi (current volume) for an excellent review of this literature.
13. In a study of two majoritarian countries, McGillivray (1997) finds that electorally concen-
trated industries receive higher levels of protection than electorally decentralized indus-
tries (602–603).
 Stephanie J. Rickard    295

14. See, for example, Ehrlich (2007). His results suggest that the impact of electoral rules on
trade protection is attenuated when party strength and district magnitude are added to the
statistical model.
15. As illustrated clearly by Kuo and Naoi’s discussion in this volume.

References
Alt, James E., and Michael Gilligan. 1994. The Political Economy of Trading States:  Factor
Specificity, Collective Action Problems and Domestic Political Institutions. Journal of
Political Philosophy 2 (2): 165–192.
Ardelean, Adina, and Carolyn L. Evans. 2013. Electoral Systems and Protectionism:  An
Industry‐Level Analysis. Canadian Journal of Economics 46 (2): 725–764.
Ashworth, Scott, and Ethan Bueno de Mesquita. 2006. Delivering the Goods:  Legislative
Particularism in Different Electoral and Institutional Settings. Journal of Politics 68
(1): 168–179.
Carey, John M., and Simon Hix. 2011. The Electoral Sweet Spot: Low-Magnitude Proportional
Electoral Systems. American Journal of Political Science 55 (2): 383–397.
Carey, John M., and Matthew Soberg Shugart. 1995. Incentives to Cultivate a Personal
Vote: A Rank Ordering of Electoral Formulas. Electoral Studies 14 (4): 417–439.
Chang, Eric C. C., Mark Andreas Kayser, and Ronald Rogowski. 2008. Electoral Systems and Real
Prices: Panel Evidence for the OECD Countries. British Journal of Political Science 38: 739–751.
Chang, Eric C.  C., Mark Andreas Kayser, Drew A. Linzer, and Ronald Rogowski. 2010.
Electoral Systems and the Balance of Consumer-Producer Power. Cambridge, UK: Cambridge
University Press.
Clark, William Roberts, Matt Golder, and Sona N. Golder. 2012. Principles of Comparative
Politics. Washington DC: Sage CQ Press.
Cox, Gary W. 1990. Centripetal and Centrifugal Incentives in Electoral Systems. American
Journal of Political Science 34: 903–935.
Cox, Gary W. 1987. Electoral Equilibrium under Alternative Voting Institutions. American
Journal of Political Science 31(1): 82–108.
Cox, Gary W. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems.
Cambridge, UK: Cambridge University Press.
Davis, Christina. 2012. Why Adjudicate? Enforcing Trade Rules. Princeton, NJ:  Princeton
University Press.
De Mesquita, Bruce Bueno, and Alastair Smith. 2005. The Logic of Political Survival. Cambridge,
MA: MIT Press.
Duverger, Maurice. 1964. Political Parties: Their Organization and Activity in the Modern State.
Taylor & Francis.
Ehrlich, Sean D. 2007. Access to Protection:  Domestic Institutions and Trade Policy in
Democracies. International Organization 61(3): 571–605.
Evans, Carolyn L. 2009. A Protectionist Bias in Majoritarian Politics. Economics & Politics
21: 278–307.
Ford, Robert, and Win Suyker. 1990. Industrial Subsidies in the OECD Economies. OECD
Department of Economics and Statistics Working Papers No. 74. Paris: OECD Publications.
Gallagher, Michael. 1991. “Proportionality, Disproportionality and Electoral Systems.” Electoral
Studies 10(1): 33–51.
296   Electoral Systems and Trade

Gallagher, Michael, Michael Laver, and Peter Mair. 2006. Representative Government in
Modern Europe. Berkshire: McGraw-Hill.
Grossman, Gene M., and Elhanan Helpman. 2005. A Protectionist Bias in Majoritarian Politics.
Quarterly Journal of Economics 120: 1239–1282.
Hankla, Charles R. 2006. Party Strength and International Trade A Cross-National Analysis.
Comparative Political Studies 39 (9): 1133–1156.
Hauk, William R. 2011. Protection with Many Sellers:  An Application to Legislatures with
Malapportionment. Economics & Politics 23 (3): 313–344.
Hays, Jude C. 2009. Globalization and the New Politics of Embedded Liberalism. Oxford: Oxford
University Press.
Inter-Parliamentary Union. 2013. PARLINE Database. http://www.ipu.org/parline/parline-
search.asp
Johnson, Joel W. and Jessica S. Wallack. 2012. “Electoral Systems and the Personal Vote”, http://
hdl.handle.net/1902.1/17901 V1 [Version]
Karol, David. 2007. Does Constituency Size Affect Elected Officials’ Trade Policy Preferences?
Journal of Politics 69 (2): 483–494.
Katzenstein, Peter J. 1985. Small States in World Markets: Industrial policy in Europe. Ithaca,
NY: Cornell University Press.
Kono, Daniel Yuichi. 2009. Market Structure, Electoral Institutions, and Trade Policy.
International Studies Quarterly 53: 885–906.
Kono, Daniel Yuichi, and Stephanie J. Rickard. 2014. Buying National: Democracy and Public
Procurement. International Interactions 40 (5): 657-682.
Lancaster, Thomas D. 1986. Electoral Structures and Pork Barrel Politics. International Political
Science Review 7 (1): 67–81.
Lizzeri, Alessandro, and Nicola Persico. 2001. The Provision of Public Goods under Alternative
Electoral Incentives. American Economic Review 91: 225–239.
Magee, Stephen P., William A. Brock, and Leslie Young. 1989. Black Hole Tariffs and Endogenous
Policy Theory:  Political Economy in General Equilibrium. Cambridge, UK:  Cambridge
University Press.
Mansfield, Edward, and Marc Busch. 1995. The Political Economy of Nontariff Barriers: A Cross
National Analysis. International Organization 49: 723–749.
Martin, Shane. 2014. Why Electoral Systems Don’t Always Matter: The Impact of “Mega-seats”
on Legislative Behavior in Ireland. Party Politics 20(3): 467–479.
McGillivray, Fiona. 1997. Party Discipline as a Determinant of the Endogenous Formation of
Tariffs. American Journal of Political Science 41(2): 584–607.
McGillivray, Fiona. 2004. Privileging Industry: The Comparative Politics of Trade and Industrial
Policy. Princeton, NJ: Princeton University Press.
Milesi-Ferretti, Gian Maria, Roberto Perotti, and Massimo Rostagno. 2002. Electoral Systems
and Public Spending. Quarterly Journal of Economics 117: 609–665.
Naoi, Megumi. 2009. Shopping for Protection: The Politics of Choosing Trade Instruments in a
Partially Legalized World. International Studies Quarterly 53 (2): 421–444.
Nielson, Daniel L. 2003. Supplying Trade Reform:  Political Institutions and Liberalization
in Middle‐Income Presidential Democracies. American Journal of Political Science 47
(3): 470–491.
Norris, Pippa. 1997. Choosing Electoral Systems:  Proportional, Majoritarian and Mixed
Systems. International Political Science Review 18 (3): 297–312.
OECD. 1998. Improving the Environment through Reducing Subsidies. Paris: OECD.
 Stephanie J. Rickard    297

Persson, Torsten, and Guido Tabellini. 2003. The Economic Effects of Constitutions. Cambridge,
MA: MIT Press.
Powell, G. Bingham, and Georg S. Vanberg. 2000. Election Laws, Disproportionality and
Median Correspondence:  Implications for Two Visions of Democracy. British Journal of
Political Science 30 (3): 383–412.
Rae, Douglas. 1967. The Political Consequences of Electoral Laws. New Haven, CT:  Yale
University Press.
Rickard, Stephanie J. 2010. Democratic Differences:  Electoral Institutions and Compliance
with GATT/WTO Agreements. European Journal of International Relations 16 (4): 711–729.
Rickard, Stephanie J. 2012a. Electoral Systems, Voters’ Interests and Geographic Dispersion.
British Journal of Political Science 42 (4): 855–877.
Rickard, Stephanie J. 2012b. A Non-Tariff Protectionist Bias in Majoritarian
Politics: Government Subsidies and Electoral Institutions. International Studies Quarterly
56 (4): 777–785.
Rickard, Stephanie J. 2012c. Welfare versus Subsidies: Governmental Spending Decisions in an
Era of Globalization. Journal of Politics 74 (4): 1171–1183.
Rickard, Stephanie J. 2009. Strategic Targeting the Effect of Institutions and Interests on
Distributive Transfers. Comparative Political Studies 42 (5): 670–695.
Rogowski, Ronald. 1987. Trade and the Variety of Democratic Institutions. International
Organization 41: 203–223.
Rogowski, Ronald, and Mark A. Kayser. 2002. Majoritarian Electoral Systems and Consumer
Power: Price-level Evidence from the OECD Countries. American Journal of Political Science
46: 526–539.
Shugart, Matthew Soberg. 1999. Presidentialism, Parliamentarism, and the Provision of
Collective Goods in Less-Developed Countries. Constitutional Political Economy 10
(1): 53–88.
Shugart, Matthew, and Martin Wattenberg. 2001. Mixed-Member Electoral Systems:
A Definition and Topology. In Mixed-Member Electoral Systems: The Best of Both Worlds?,
edited by Matthew Shugart and Martin Wattenberg, 9–24. Oxford: Oxford University Press.
Taagepera, Rein, and Matt Qvortrup. 2011. Who Gets What, When, How—Through Which
Electoral System. European Political Science 11 (2): 244–258.
Taagepera, Rein, and Matthew Soberg Shugart. 1989. Seats and Votes:  The Effects and
Determinants of Electoral Systems. New Haven, CT: Yale University Press.
Thames, Frank C., and Martin S. Edwards. 2006. Differentiating Mixed-Member Electoral
Systems Mixed-Member Majoritarian and Mixed-Member Proportional Systems and
Government Expenditures. Comparative Political Studies 39 (7): 905–927.
Tsebelis, George. 2011. Veto Players: How Political Institutions Work. Princeton, NJ: Princeton
University Press.
 Chapter 16

Au thoritaria n Re g i me s

Danie l Y ui c h i Kon o

Although democracy is on the rise, autocracies remain a major part of the global
political economy. According to one popular measure, 39 percent of countries were
autocratic as of 2008,1 and major trading powers such as China remain under auto-
cratic rule. Moreover, when it comes to trade policy, these autocracies are remark-
ably diverse. Some, such as Botswana and Namibia, have overall tariff equivalents of
less than 1 percent.2 Others, such as Sudan and Tanzania, have overall rates of pro-
tectionism in excess of 50 percent. Trade policies in the rest of the world’s autocra-
cies vary continuously across this range. As these figures make clear, autocratic trade
policies vary tremendously. Acquiring a full understanding of the global economy
thus requires us to ask why autocracies choose to participate (or not) in international
trade.
Extant research provides surprisingly few answers to this question. The voluminous
literature on regime type and trade shows that autocracies have different trade policies
than democracies (e.g., Frye and Mansfield 2004; Kono 2006, 2008; Milner and Kubota
2005; Mansfield, Milner, and Rosendorff 2000, 2002; Tavares 2008; Verdier 1998).3
Research on electoral institutions and trade shows that democracies themselves differ
in important ways (Hankla 2006; Kono 2009; Mansfield and Busch 1995; Nielson 2003;
Rogowski 1987).4 However, neither body of research asks whether or why some autocra-
cies are more open to trade than others. Instead, to the extent that autocracies are con-
sidered at all, they are treated as an implicitly homogeneous reference category against
which democracies are compared.
This inattention to autocratic trade policy is perhaps less surprising than it seems, as
it reflects a more general neglect of politics in authoritarian regimes. As Geddes (1999)
notes, the study of autocracies has lagged behind that of democracies, perhaps because
the former are less institutionalized and transparent. Whatever the reason, “this focus
has left [us]… with few shoulders of giants on which to stand as we try to understand
political change in less democratic and less institutionalized settings” (Geddes 1999, 1).
Likewise, those studying trade policy in autocracies have few well-developed models on
which to build.
 Daniel Yuichi Kono   299

This chapter surveys the small but growing literature on trade policy in autocracies.
Following Alvarez and colleagues (1996), I define autocracies as regimes that do not
choose their chief executives and legislatures through regular, contested elections. This
is admittedly a residual definition—autocracies are not democracies—but it does high-
light what is common to all autocracies while allowing for their diversity. My central
question is: Why are some nonelected governments open to international trade, while
others are not?
Because I seek to explain variation among autocracies, I largely avoid work that is
generic with respect to regime type (e.g., Grossman and Helpman 1994), as well as the
extensive literature on regime type and trade (e.g., Frye and Mansfield 2004; Kono 2006,
2008; Milner and Kubota 2005; Mansfield, Milner, and Rosendorff 2000, 2002; Tavares
2008; Verdier 1998), which emphasizes variation across regime types. I  say “largely”
because this work provides a natural starting point for thinking about autocratic trade
policy. I thus begin by considering extensions of this work to autocracies, before moving
on to research that focuses on autocracies alone.
Specifically, I consider four sets of factors that may influence autocratic trade pol-
icy: economic structure, coalition size, time horizons, and the leader’s mode of entry
into power. My central conclusion is that while all four factors help explain autocratic
trade policy, research on this topic remains largely an extension of earlier work on
democracies and regime type. This is not meant as a criticism: research on this topic has
barely begun,5 and scholars have understandably built on well-established and familiar
models. Nonetheless, scholars have yet to develop trade policy models that are uniquely
tailored to autocracies. Doing so will require trade policy scholars to build on more basic
research on authoritarian politics (e.g., Gandhi and Przeworski 2006; Geddes 1999;
Magaloni 2006; Svolik 2009; Weeks 2008; Wright 2008a), just as research on democratic
trade policy built on earlier work on comparative electoral institutions. This process has
only begun, but seems likely to accelerate as trade policy scholars become increasingly
familiar with comparative work on authoritarian regimes.

Economic Structure

When trying to explain policy outcomes, political scientists are naturally inclined to
look to political institutions. If such institutions do not matter, then this chapter’s focus
on autocracies—an institutional distinction—makes little sense. For precisely this rea-
son, however, it is worth asking whether autocracies really are different. Do autocratic
institutions really matter, or are trade policies driven by more fundamental, noninstitu-
tional forces? Much work on the political economy of trade points to the importance of
economic structure: factor endowments (e.g., Rogowski 1989), sectoral characteristics
(e.g., Grossman and Helpman 1994), and so forth. If trade policy is largely determined
by such factors, then there is little need to refer to political institutions or to study autoc-
racies as a distinct group.
300   Authoritarian Regimes

In support of this “structural” approach, I note that all political leaders face the same
fundamental problem of gaining and maintaining power (Bueno de Mesquita et  al.
2003). This requires all leaders to respond to societal demands of various kinds. Political
institutions arguably matter because different institutions may subject leaders to differ-
ent societal pressures. Yet scholars from Marx onward have argued that some pressures
are too fundamental for any politicians to ignore. As Alesina and Rodrik (1994, 466) note,
“Even a dictator cannot completely ignore social demands, for fear of being overthrown.
Thus, even in a dictatorship, distributional issues affecting the majority of the population
will influence policy decisions.” For example, the Chinese leadership allegedly sets high
growth targets to deter popular unrest. More generally, societal pressures may have simi-
lar effects in all political systems, if they are sufficiently powerful. If so, then models based
on these pressures may explain trade policy in autocracies as well as democracies.
The dominant model in the trade policy literature, Grossman and Helpman’s
(1994) “protection for sale” model, implies that trade policy variation—specifically,
variation across sectors—can be explained by a small number of sectoral character-
istics. Protectionism should fall with the elasticity of import demand, rise with the
output-import ratio in organized sectors, and fall with the output-import ratio in unor-
ganized sectors. The cross-sector structure of protectionism can thus be explained
with very few variables—demand elasticities, output-import ratios, and sectoral
organization—all of which are arguably exogenous to political institutions.6 Tests of the
model’s predictions find strong empirical support (Gawande and Bandyopadhyay 2000;
Goldberg and Maggi 1999).
At first glance the Grossman-Helpman (1994) model appears to describe trade pol-
itics in democracies: the frequent references to lobbying and campaign contributions
suggest a democratic political process. In fact, however, the model is quite general. The
government’s objective function is very simple: it maximizes a weighted sum of finan-
cial contributions and aggregate welfare. It seems likely that all governments care about
these things, and hence that the Grossman-Helpman model is broadly applicable.
Empirical research supports this point. Although initial tests of the model’s predic-
tions focused on US trade policy (Gawande and Bandyopadhyay 2000; Goldberg and
Maggi 1999), subsequent tests have examined autocracies as well. For example, Mitra,
Thomakos, and Ulubasoglu (2002) have tested the model’s predictions using Turkish
data, both before and after Turkey became a democracy. Their tests support these
predictions in both democratic and nondemocratic periods. This suggests that the
Grossman-Helpman (1994) model is useful for explaining trade policy in autocracies as
well as in democracies.
Another canonical trade policy model, that of Mayer (1984), also points to the impor-
tance of economic structure. In this model trade policy is determined by the median
voter, whose preferences are in turn determined by her factor ownership and national
factor endowments. If the median voter’s capital-labor ratio is lower than the national
mean, the median voter will seek protectionism for labor-intensive imports but an
import subsidy for capital-intensive imports. If the median voter’s capital-labor ratio is
above the national mean, she will have the opposite preferences.
 Daniel Yuichi Kono   301

It may seem odd to refer to Mayer’s model (1984) as “structural,” since its reliance
on the median voter suggests an important role for political institutions. Indeed, Mayer
(1984) provides the foundation for several arguments about regime type and trade (Kono
2008; O’Rourke and Taylor 2007; Tavares 2008), a point I return to below. Nonetheless,
Mayer specifies precisely how economic variables should affect trade policy, and some
research shows that these variables have similar effects across regime types. For exam-
ple, Dutt and Mitra (2002) extend Mayer to predict that inequality affects trade policy.
Since higher inequality increases the distance between the (invariably capital-poor)
median voter and the national mean, higher inequality also intensifies the median
voter’s trade policy preferences. Higher inequality thus leads to higher protectionism
in capital-abundant countries but to lower protectionism in capital-scarce countries.
Dutt and Mitra (2002) find strong empirical support for this hypothesis. Interestingly,
they also find that the inequality-trade policy relationship is much the same in democ-
racies and autocracies. This implies that regime type does not alter the identity of the
median “voter”—that is, the person whose political support is crucial to maintaining
power7—and hence that cross-country variation in trade policy is largely driven by eco-
nomic factors.
In a study focused on autocracies, Wu (2014) explores this relationship in greater
depth. Building on Acemoglu and Robinson (2005) and Boix (2003), Wu notes that
rising inequality generates rising demands for redistribution—and possibly for
democratization—in dictatorships. Dictators seeking to prolong their rule must
respond to these demands in some way, through either increased repression or income
redistribution. Assuming Stolper-Samuelson (1941) effects, trade liberalization pro-
vides one means for autocrats in labor-abundant countries to redistribute income
toward unskilled workers. Empirically, Wu (2014) finds that increases in inequality
lead to lower trade openness in labor-scarce autocracies but to higher trade openness in
labor-abundant autocracies. He also finds that higher trade openness reduces inequality
in autocracies, even after controlling for the former’s endogeneity to the latter. This sug-
gests that autocrats indeed liberalize trade in response to popular pressures for income
redistribution.
These studies suggest that at least some economic-structural variables affect trade
policy similarly in democracies and autocracies.8 Scholars may thus be able to employ
a common set of models to understand trade policy under both regime types. This is
worth remembering, as there is much to be said for parsimonious explanations that do
not require us to reinvent the wheel. That said, it also seems unlikely that we can fully
understand autocratic trade policies without delving into the institutional structure of
authoritarian regimes.
First, the large literature on regime type and trade (e.g., Frye and Mansfield 2004;
Kono 2006, 2008; Milner and Kubota 2005; Mansfield, Milner, and Rosendorff 2000,
2002; O’Rourke and Taylor 2007; Tavares 2008; Verdier 1998) demonstrates that autoc-
racies have different trade policies than democracies. This alone points to the impor-
tance of autocratic political institutions. Second, even the abovementioned work on
economic structure indicates important differences across regimes. For example, Mitra,
302   Authoritarian Regimes

Thomakos, and Ulubasoglu (2002) find that the Turkish government attached a higher
weight to general welfare—that is, adopted lower trade barriers—after Turkey democ-
ratized. Hence, while the Grossman-Helpman (1994) sectoral variables explained the
cross-industry pattern of protectionism in both autocratic and democratic periods,
average protectionism tended to be lower during the latter. This suggests that the weight
attached to general welfare—exogenous in the Grossman-Helpman model—is endog-
enous to political institutions.
Finally, there is evidence that the impact of economic structure itself varies across
regime types. Mayer (1984) noted that the identity of the median voter depends on
political institutions: she is the median member of society given universal suffrage and
complete voter turnout, but the median member of a small elite when the franchise is
restricted. Given the nature of real-world income distributions, scholars have taken
this to mean that the median voter is capital-poor in democracies but capital-rich in
autocracies (Kono 2008; Milner and Kubota 2005; O’Rourke and Taylor 2007; Tavares
2008). If so, then factor endowments should affect trade policy differently under the two
regime types: an increase in an economy’s capital-labor ratio should lead to higher trade
barriers in democracies but to lower trade barriers in autocracies.
As this example suggests, the impact of economic-structural variables may depend
on political institutions. This limits our ability to explain autocratic trade policy with
economic variables alone. A more complete explanation must take autocracies’ institu-
tional structures into account.

Coalition Size

Political institutions vary in many ways. Perhaps the most central concept in the vast
literature on such institutions is constituency size. Put simply, every politician must sat-
isfy some constituency to stay in power. The US president needs the support of half the
voting public. Members of Congress require majorities within their electoral districts.
Legislators in closed-list proportional representation systems are beholden to party
leaders, who in turn must capture the largest possible share of the electorate. Although
constituencies are most clearly defined in democracies, autocratic leaders must satisfy
constituencies as well, whether the latter are military elites, party members, nobles, or
some other group. A key insight from the literature on political institutions is that the
size of a leader’s constituency affects her policy choices.
One prominent model based on constituency size is Bueno de Mesquita and col-
leagues (2003) “selectorate model.” In this model, leaders seek the support of a “win-
ning coalition,” defined as the set of individuals whose support is needed to gain and
maintain power. The winning coalition can be small—for example, nobles in a monar-
chy, military elites in a military dictatorship, party officials in a single-party regime—or
large, for example, half the electorate in a democracy. As these examples suggest, democ-
racies tend to have larger winning coalitions than autocracies. According to Bueno de
 Daniel Yuichi Kono   303

Mesquita and colleagues, larger winning coalitions induce leaders to provide more pub-
lic goods, which benefit large groups of people at minimal cost, and fewer private goods,
which benefit only a small elite. Democracies should thus provide more public goods
than autocracies. Bueno de Mesquita and colleagues employ this logic to explain, inter
alia, why democracies are more open to international trade.
Milner and Kubota (2005) apply the concepts—though not the strict logic—of the
selectorate model to explain the relationship between regime type and trade policy in
developing countries. Following Bueno de Mesquita and colleagues (2003), they note
that democracies have larger winning coalitions than autocracies. However, Milner
and Kubota’s (2005) argument rests more on the income effects of trade policy than
on the distinction between public and private goods. As the winning coalition grows
larger, it is more and more likely to include unskilled workers, who, in labor-abundant
developing countries, should benefit from freer trade (Stolper and Samuelson 1941).
Democratization thus induces politicians to adopt the liberal trade policies that benefit
unskilled workers.
As the Milner and Kubota (2005) example suggests, constituency-size arguments
are often used to explain differences between democracies and autocracies. However,
research on electoral institutions demonstrates that constituency size varies within as
well as across regime types. Some electoral institutions cause politicians to represent
larger constituencies than others: for example, some legislators may represent larger and
more heterogeneous electoral districts than others, and popularly elected presidents
represent larger districts than individual legislators. Research shows that such variation
in constituency size matters for trade policy. As Rogowski (1987) notes, small electoral
districts create incentives to protect industries within those districts, since the costs of
such protectionism largely fall elsewhere. In contrast, large electoral districts internalize
more of the costs of protectionism, thus creating countervailing pressures for liberal-
ization. Lohmann and O’Halloran (1994) formalize this logic, showing that presidents
should be more liberal than legislatures because the former represent larger constituen-
cies that bear the costs as well as the benefits of protectionism. Similarly, Nielson (2003)
and Hankla (2006)—building on Carey and Shugart (1995)—show that strong party dis-
cipline, which induces legislators to cater to larger (party-wide) constituencies, leads to
more liberal trade policies.
The works discussed above show, from a variety of perspectives, that constituency
size affects trade policy: larger constituencies generally encourage freer trade. They also
show that constituency size can vary within as well as across regime types. Therefore it
is not surprising that initial attempts to understand autocratic trade policies have rested
heavily on the concept of constituency size. In this sense, the early work on trade policy
in autocracies is a natural extension of that on democracies.
Such an extension requires an understanding of autocratic institutions:  how they
vary, and how such variation affects constituency size. Although various scholars have
conceptualized autocratic institutions in different ways, studies of autocratic trade pol-
icy have relied heavily on Geddes’s (1999, 2003) typology of such institutions.9 I thus
preface my discussion of these studies by briefly reviewing this typology.
304   Authoritarian Regimes

Geddes (1999, 2003) divides autocracies into three broad subtypes: military, single-
party, and personalist. She summarizes these regime subtypes as follows (2003, 51):
In military regimes, a group of officers decides who will rule and exercises some
influence on policy. In single-party regimes, one party dominates access to politi-
cal office and control over policy, though other parties may exist and compete as
minor players in elections.[10] Personalist regimes differ from both military and
single-party in that access to office and the fruits of office depend much more on the
discretion of an individual leader. The leader may be an officer and may have created
a party to support himself, but neither the military nor the party exercises indepen-
dent decision-making power insulated from the whims of the ruler.
As the last point makes clear, the crucial distinction between personalist regimes and
the other two subtypes is one of institutional autonomy. In institutionalized military
regimes—for example, Brazil from 1964 to 1985 or Argentina from 1976 to 1983—“senior
officers have agreed upon some formula for sharing or rotating power, and consultation
is somewhat routinized (Geddes 2003, 52). Likewise, in institutionalized single-party
regimes—for example, the Institutional Revolutionary Party in Mexico or Communist
parties in Eastern Europe—the party organization “exercises some power over the
leader at least part of the time, controls the selection of officials, organizes the distribu-
tion of benefits to supporters, and mobilizes citizens to vote and show support for party
leaders” (Geddes 2003, 52). In contrast, personalist military rulers (e.g., Rafael Trujillo
in the Dominican Republic; Idi Amin in Uganda) have concentrated power in their own
hands and are relatively unconstrained by their militaries, just as personalist party lead-
ers (e.g., Manuel Odria in Peru; Etienne Eyadema in Togo) have effectively monopolized
control over their parties. Militaries and parties in personalist regimes are thus subser-
vient to their leaders rather than checking the latter’s power.
A fourth regime subtype, not considered by Geddes (1999, 2003), is monarchies,
defined by Hadenius and Teorell (2007, 146)  as “regimes in which a person of royal
descent has inherited the position of head of state in accordance with accepted practice
or the constitution.” Examples include Saudi Arabia and other “dynastic” Gulf monar-
chies, “where successors are chosen by a consensus of the royal family” (Hadenius and
Teorell 2007, 146). To be a monarchy, dynastic succession must be institutionalized in
norms or the constitution: hence, for example, Hadenius and Teorell do not consider
North Korea a monarchy, as the father-to-son transitions are not institutionalized but
ad hoc. Institutionally, monarchies share aspects of all of Geddes’s (1999, 2003) regime
subtypes. Like personalist regimes, they often concentrate power in the hands of a single
ruler. However, like military and single-party regimes, they transfer power in an orderly
manner that requires the support of a well-defined group.
This brief discussion suggests that autocratic regime subtype may influence win-
ning coalition size. Military and single-party leaders are accountable to a wider group
of officers or party members, respectively. They should thus have larger winning coali-
tions than personalist leaders, who are accountable only to themselves. Military and
single-party leaders probably also have larger coalitions than monarchs, since the
 Daniel Yuichi Kono   305

latter require support from only a small group of nobles. Milner and Kubota (2005)
and Hankla and Kuthy (2013) build on this insight in their research on autocratic trade
policy.
Although Milner and Kubota (2005) focus mainly on the democracy-autocracy
distinction, they note that their argument also predicts trade policy differences
among autocratic subtypes. Building on Geddes (1999), they argue that single-party
dictatorships have the largest winning coalitions, followed by military dictatorships
and then by personalist regimes. Given their general argument about coalition size,
this implies that single-party regimes should be most responsive to unskilled work-
ers and hence the most liberal. Empirically, they find that single-party autocracies
have significantly lower tariffs than personalist regimes, suggesting that coalition
size affects trade policy within autocracies as well as within democracies and across
regime types.
Hankla and Kuthy (2013) build on this insight, focusing exclusively and in greater
depth on variation among autocracies. They argue that more institutionalized autoc-
racies, such as single-party and multiparty regimes, should be more liberal than less
institutionalized ones such as monarchies and personalist regimes. Like Milner and
Kubota (2005), Hankla and Kuthy (2013) stress the importance of constituency size,
arguing that more institutionalized autocracies have larger winning coalitions.
Unlike Milner and Kubota (2005), Hankla and Kuthy (2013) distance themselves from
Stolper-Samuelson (1941) or any particular model of trade; instead, they argue that pro-
tectionism generally provides private goods to concentrated interests, while free trade
provides a public good to the economy as a whole, and hence to a majority of people. In
this sense, Hankla and Kuthy’s argument is both more faithful to the selectorate model
and potentially more generalizable to capital-abundant as well as labor-abundant
autocracies. Hankla and Kuthy perform a wide range of empirical tests and find that
single-party and multiparty regimes are generally more liberal than other autocratic
regime types.
Frye and Mansfield (2003) advance a similar argument. Focusing on postcommu-
nist countries, they argue that fragmentation of power in autocracies encourages trade
liberalization. Fragmentation of power is “the extent to which a national government
includes competing partisan and institutional actors whose agreement is necessary to
make policy. These actors include rival branches of government, as well as legislative
and executive coalitions involving different political parties” (2003, 639). As Frye and
Mansfield note, the degree of fragmentation varies greatly even among nondemocratic
regimes.
Fragmentation matters for trade policy because highly concentrated power facili-
tates protectionist rent seeking by a small elite. In contrast, the fragmentation of
power increases the likelihood that policy will also be influenced by groups that gain
from liberalization. Although Frye and Mansfield (2003) do not employ the language
of constituency size, their argument is essentially the same as that of Rogowski (1987)
and Lohmann and O’Halloran (1994): as the government’s support coalition grows, it
is more and more likely to include groups that are hurt by trade protectionism. Frye
306   Authoritarian Regimes

and Mansfield (2003) find strong support for this argument, showing that autocra-
cies are more likely to liberalize trade when their power is fragmented than when it is
concentrated.
Wu (2013) provides a different perspective on the effects of constituency size in autoc-
racies. Like Milner and Kubota (2005) and Hankla and Kuthy (2013), Wu (2013) starts
from the premise that single-party autocracies have larger winning coalitions than other
autocratic regime subtypes. However, in contrast to the former scholars, Wu emphasizes
the impact of coalition size on the structure rather than the level of trade protectionism.
Building on Ehrlich (2007, 2011), Wu (2013) argues that larger coalitions provide more
“access points” for interest groups that benefit from protectionism. Large-coalition dic-
tators may thus face pressures for trade liberalization, but they also have incentives to
tailor trade policies specifically to those groups that have access. The result is a wedge
between tariffs for coalition insiders and outsiders: the former receive high tariff protec-
tion, while the latter do not. Hence, while large-coalition dictatorships may have lower
average tariff levels, they should also have higher tariff dispersions. Empirically, Wu
finds that single-party autocracies in fact have more dispersed tariffs. This suggests that
Ehrlich’s (2007, 2011) access point theory can help us understand trade policy in autocra-
cies as well as democracies.
These studies point to the potential importance of coalition size for autocratic trade
policies. This is encouraging, as it suggests that the central concept in the study of democ-
racies and regime type also helps explain trade policy in autocracies. However, these
studies also raise questions about the causal mechanisms at work. First, because autocra-
cies’ winning coalitions are often not explicitly defined, regime subtypes are proxies for
coalition size—and, like most proxies, probably exhibit measurement error. Perhaps for
this reason, the studies that use these proxies generate empirical anomalies. For exam-
ple, Milner and Kubota (2005) find that military regimes are (insignificantly) more pro-
tectionist than personalist regimes, while Hankla and Kuthy (2013) find that monarchies
are as liberal as multiparty regimes and that military and single-party regimes do not
differ significantly from each other. In all of these cases, the authors’ coalition-size argu-
ment predicted a different result. While these anomalies do not undermine the authors’
central results, they do raise questions about the mapping between regime subtypes and
coalition size, and perhaps about the importance of coalition size itself.
Second, regime subtypes possess other characteristics that covary with coalition size.
As discussed below, Hankla and Kuthy (2013) argue that large-coalition regimes also
have longer time horizons, which may also encourage trade liberalization. Although
this correlation between coalition size and time horizons strengthens the theoretical
case for Hankla and Kuthy’s predictions, it also raises questions about which features of
autocratic subtypes actually drive differences in trade policy.
Despite these caveats, the studies discussed above constitute an important first step
toward understanding trade policy in autocracies. The evidence strongly suggests that
coalition size affects such policies. More conclusive tests, however, will require scholars
to disentangle the effects of coalition size from those of other factors that vary across
regime subtypes, such as the time horizons of autocratic leaders.
 Daniel Yuichi Kono   307

Time Horizons

Another dimension along which autocratic leaders vary is their time horizons: that is,
how long they expect to stay in power. Some autocratic regimes are highly stable, while
others exhibit frequent leadership turnover. Scholars have long argued that time hori-
zons affect policy choice in democracies; for example, the proximity of elections may
influence macroeconomic policies (Franzese 2002). We might thus expect time hori-
zons to affect autocratic policies as well.
Much work on autocratic time horizons builds on Clague and colleagues (1996) argu-
ment about time horizons and property rights. They argue that leaders have short-term
incentives to violate property rights by seizing property they can enjoy now. However,
they also have long-term incentives to respect property rights, since secure property
rights encourage long-run economic growth and higher tax revenues. It follows that
leaders’ respect for property rights should depend on their time horizons. Short-sighted
leaders, who expect to lose power soon, should violate property rights because they will
not be around to enjoy the long-term revenue gains from respecting them. In contrast,
far-sighted leaders who expect to stay in power should respect property rights to maxi-
mize these long-term gains.
Although Clague and colleagues (1996) focus on property rights, their argument
should apply more generally to any policies that have different short-run and long-run
effects. For example, Wright (2008b) builds on this argument to explore the effects of
foreign aid on economic growth. He argues that leaders with short time horizons are
likely to spend resources—including foreign aid—in unproductive ways. Not only
might they keep this money for themselves, but they also have strong incentives to use
it to prolong their rule by repressing or buying off threats to the regime. In contrast,
leaders with long time horizons have more freedom and incentive to spend resources
in growth-enhancing ways. Wright (2008b) thus hypothesizes, and finds empirically,
that foreign aid is more likely to promote growth when given to leaders with long time
horizons.
These arguments have implications for trade policy, because the costs and benefits of
trade policy are also unevenly distributed over time: the costs are immediate, as trade
liberalization causes politically unpopular adjustment costs (Porto and Hoekman 2010),
but the benefits are deferred, in the form of longer-term growth and higher potential
tax revenues (Edwards 1998; Frankel and Romer 1999; Sachs and Warner 1995). The
incentives to liberalize trade should thus depend on leaders’ time horizons. Politically
insecure leaders should fear the short-term costs of liberalization while discounting the
long-term benefits; politically secure leaders should be willing to bear the short-term
costs for the sake of long-term gains. Longer time horizons should thus encourage trade
liberalization.
Hankla and Kuthy (2013) employ this logic in their study of autocratic subtypes and
trade policy. They argue that more institutionalized autocracies, such as single-party
308   Authoritarian Regimes

and multiparty regimes, are more stable and have longer time horizons than less institu-
tionalized autocracies such as monarchies and personalist regimes. More institutional-
ized autocracies should thus have more liberal trade policies. Hankla and Kuthy find
empirical support for this hypothesis, although their test is indirect: regime subtypes
serve as proxies for time horizons. This seems reasonable, as Wright (2008b) finds that
single-party regimes are indeed less likely to fail than military, personalist, or monarchi-
cal regimes. However, because regime subtypes also have different winning coalitions, it
is difficult to say whether Hankla and Kuthy’s (2013) results point to the importance of
time horizons, coalition size, or both.
Kono and Montinola (forthcoming) test the time-horizons hypothesis more directly,
using Wright’s (2008b) measure of time horizons. In brief, this measure is based on
an analysis of authoritarian regime failure: higher failure probabilities correspond to
shorter time horizons. Using various measures of trade policy, Kono and Montinola
(forthcoming) find that longer time horizons lead to more liberal trade policies. They
also find, however, that foreign aid attenuates this relationship. Because foreign aid
helps unstable governments survive, it effectively lengthens their time horizons, thus
weakening the relationship between Wright’s (2008b) time horizons measure and trade
policy. This nonetheless tends to support the posited link between time horizons and
trade policy, as foreign aid can be seen as operating primarily through a time-horizons
channel.
Like the work on coalition size, research on autocratic time horizons has its roots
in earlier work on democracies. The literature on political business cycles has long
maintained that short time horizons—induced by the proximity of elections—induce
short-sighted macroeconomic policies in democracies (Franzese 2002). Although the
factors that influence time horizons in autocracies are different—as are the time hori-
zons themselves—the central intuition is the same.
Where the democratic and autocratic literatures differ is in their degree of theoreti-
cal and empirical development. In the literature on democracies, there is widespread
agreement that elections are the main determinant of time horizons. This theoretical
consensus has allowed scholars to narrow their focus and specify more precisely both
the conditions under which elections should matter and how to measure their effects
empirically. The political business cycle literature has consequently generated ever
more nuanced and context-contingent results (Franzese 2002). In contrast, there is still
little consensus about what determines autocratic time horizons. And, while measures
like Wright’s (2008b) have the advantage of being multidimensional, they also leave us
wondering exactly what political circumstances induce short-sighted or far-sighted
behavior. It would be useful, in future research, to explore these circumstances in
greater depth.
It would be particularly useful to know what causes autocratic time horizons to vary
over time. Perhaps the main contribution of the political business cycle literature is
that it sheds light on why democratic policies fluctuate. To date, there is no analog in
the literature on autocratic trade policy. Hankla and Kuthy (2013) identify regime sub-
type as a main determinant of time horizons. However, since these subtypes are quite
 Daniel Yuichi Kono   309

stable, they cannot explain why autocratic trade policies vary over time. This is impor-
tant, since autocratic trade policies are in fact quite volatile: for example, the Democratic
Republic of Congo’s average tariff fell from 25 percent in 1990 to only 4 percent in 1992,
then jumped back to 34 percent in 1994.11 Explaining such variation will require more
attention to less stable features of autocratic regimes.

Mode of Entry

Economic structure, coalition size, and time horizons are features of all polities and may
have similar effects across regime types. This conceptual continuity between work on
democracies and autocracies is useful, but it also begs the question of whether and how
autocracies are unique. Svolik (2009, 479) notes at least one way in which autocracies
differ qualitatively from democracies: “[P]‌olitical power in dictatorships, even when
nominally exercised through political institutions, must be ultimately backed by a cred-
ible threat of violence.” This is not the case in democracies, where the central political
actors generally obey constitutional rules without being implicitly or explicitly coerced.
Moreover, the threat of violence is not a constant, but varies considerably across space
and time (Svolik 2009). Given this, it is worth asking whether threats of violence affect
trade politics in authoritarian regimes.
Chow and Kono (2014) explore this possibility, developing a model of trade politics in
which coup threats play a central role. They begin by noting that autocratic leaders enter
power in different ways: some enter in “a regular manner, according to the prevailing
rules, provisions, conventions and norms of the country,” while others enter in “an irreg-
ular manner, such as a coup” (Goemans, Gleditsch, and Chiozza 2009, 272–273). Chow
and Kono (2014) refer to the first mode of entry as “legal” and to the second as “extrale-
gal.” Both modes of entry are common in autocracies, with extralegal entries accounting
for about 40 percent of autocratic power transitions (Goemans, Gleditsch, and Chiozza
2009). This clearly distinguishes autocracies from democracies, in which essentially all
power transitions are legal.12
Why might the mode of entry matter for trade policy? Chow and Kono (2014) argue
that it affects the new leader’s vulnerability to coups. All autocrats must fear coup
attempts by other ruling elites. However, the magnitude of this threat depends on the
ruler’s ability to identify and neutralize coup plotters through co-optation or repres-
sion. This can be difficult to do because, as Svolik (2009, 480) notes, autocratic regimes
are particularly prone to “secrecy and back-channel politics.” In other words, autocratic
leaders often do not know who is up to what. In such an environment, knowledge is the
key to survival. Chow and Kono (2014) argue that legal leaders are more knowledgeable
than extralegal ones and hence less vulnerable to coups.
New legal leaders are generally regime “insiders”: they emerge from the officer corps
in military regimes, the party leadership in single-party regimes, the royal family in
monarchies, and so forth. They are thus familiar with other ruling elites: who is allied
310   Authoritarian Regimes

with whom, who wants what, who can be bought, who must be incarcerated, and so
forth. This knowledge helps new legal leaders identify and neutralize coup plot-
ters. In contrast, new extralegal leaders are typically regime “outsiders”:  they enter
power extralegally precisely because they cannot do so under the established regime.
Compared with new legal leaders, new extralegal leaders are thus less familiar with
the political landscape, less able to neutralize coup plotters, and more vulnerable to
attempted coups.
Vulnerability to coups heightens the incentives to placate mass publics, since public
unrest increases both the likelihood of coup attempts and the chances that they will
succeed (Galetovic and Sanhueza 2000; Powell 2012; Thyne 2010). Extralegal lead-
ers who feel threatened by coups thus have strong incentives to adopt policies—such
as trade liberalization—that benefit mass publics. Politicians have understood since
Roman times that high prices can lead to protests and riots, and that providing cheap
“bread” is one solution to such problems. When high commodity prices in 2007–2008
led to food riots around the globe,13 many leaders responded to this unrest by using
trade policy to reduce prices (World Bank 2008). Chow and Kono (2014) argue,
more generally, that leaders worried about coups should discourage public unrest by
liberalizing trade.
For these reasons, a leader’s mode of entry affects trade policy. Legal leaders who are
relatively unthreatened by coups adopt high tariffs to maximize rents. In contrast, extra-
legal leaders who are vulnerable to coups reduce tariffs, forgoing rents to reduce pub-
lic unrest and hence to improve their chances of survival. This difference between legal
and extralegal leaders is short-lived, however: over time, extralegal leaders accumulate
knowledge about other ruling elites, eventually becoming regime insiders. As this hap-
pens, they become similar to legal leaders and adopt similar trade policies. Chow and
Kono (2014) thus hypothesize that extralegal entries lead to short-term tariff reductions
that are reversed as new leaders consolidate their rule. They support this hypothesis
with data on average tariffs, the Sachs-Warner (1995) openness measure, and the leader’s
mode of entry. Their results point to the mode of entry as an important determinant of
autocratic trade policy.
Chow and Kono’s (2014) main contribution is to incorporate into the study of trade
policy the threat of violence that is both central and largely unique to authoritar-
ian regimes. This represents a clear departure from previous work on trade policy in
democracies. In addition, because the mode of entry varies over time within regimes,
Chow and Kono help explain longitudinal variation in autocratic trade policies. This
is also a departure from previous work on autocratic trade policy (Hankla and Kuthy
2013; Milner and Kubota 2005), which focuses on institutions that rarely change over
time. That said, Chow and Kono’s (2014) model implies that the mode of entry can-
not explain persistent cross-national differences among autocratic regimes. Because
the effects of extralegal entries diminish rather quickly, countries governed by extra-
legal leaders are, in the long run, no less protectionist than countries governed by legal
leaders.
 Daniel Yuichi Kono   311

The Road Ahead

This chapter has highlighted four sets of variables that may affect trade policy in autoc-
racies:  economic structure, coalition size, time horizons, and the autocrat’s mode of
entry. These factors collectively help explain the level, structure, and changes in protec-
tionism in authoritarian regimes. In this sense, research on trade policy in autocracies
is off to a good start. It is also clear, however, that theories of autocratic trade policy
have barely begun to move beyond previous work on democracies and regime type.
Economic-structural research on autocracies (Dutt and Mitra 2002; Mitra, Thomakos,
and Ulubasoglu 2002) is a direct application of Grossman and Helpman (1994) and
Mayer (1984). Similarly, research on autocratic coalition size (Hankla and Kuthy 2013;
Milner and Kubota 2005; Wu 2013) builds closely on Bueno de Mesquita and colleagues
(2003) and Ehrlich (2007, 2011), while work on autocratic time horizons (Hankla and
Kuthy 2013; Kono and Montinola, forthcoming) has roots in the democratic political
business cycle literature. Although theories about the mode of entry (Chow and Kono
2014) are sui generis, they are also at an early stage of development. It thus seems fair
to say that scholars have barely begun to develop trade policy models that are uniquely
tailored to autocracies.
This is not a criticism, but an acknowledgment of the current state of the field. The first
article to focus exclusively on autocratic trade policy was Hankla and Kuthy (2013). This
is thus a very new field of study, and one that, understandably, has begun by building on
well-established and familiar models. At this point, however, it is possible to do more.
Trade policy scholars can now draw on a growing body of “basic” research on autocratic
politics. This research illuminates, for example, the factors that promote regime stabil-
ity (Geddes 1999, 2003), elite power sharing (Svolik 2009), policy concessions and rent
sharing between the leader and the opposition (Gandhi and Przeworski 2006), and
audience costs (Weeks 2008). The implications of this research for trade policy are not
yet clear. However, this was also initially the case for research on democratic institutions.
For example, Downs’s (1957) median voter theorem did not find its way into trade pol-
icy research until the work of Mayer (1984), just as Carey and Shugart’s (1995) work on
electoral particularism was not applied to trade policy until Nielson (2003) and Hankla
(2006). There is always a lag between basic research on political institutions and its appli-
cation to a specific topic such as trade policy. The task for trade policy scholars is now to
give autocracies the same attention that they have historically given to democracies.

Notes
1. Based on the dichotomous regime type measure developed by Alvarez et al. (1996) and
extended by Cheibub, Gandhi and Vreeland (2010).
2. Based on Kee, Nicita, and Olarreaga’s (2009) index of overall protection, measured as the
ad valorem tariff equivalent of all trade barriers in 2009. Data are available at http://econ.
312   Authoritarian Regimes

worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/0,,contentMDK:2257
4446~pagePK:64214825~piPK:64214943~theSitePK:469382,00.html.
3. See also Mukherjee, this volume.
4. See also Rickard, this volume.
5. The first article to focus exclusively on autocratic trade policy is Hankla and Kuthy (2013).
6. I say “arguably” because one could also argue that these variables are influenced by politi-
cal institutions: for example, perhaps a country’s institutional structure affects the pattern
of sectoral organization. Although such arguments are plausible, the Grossman-Helpman
(1994) model itself treats these variables as exogenous to political institutions. For a more
detailed discussion of this model, see Goodhart, this volume.
7. Dutt and Mitra (2002) emphasize that the median-voter language is not to be taken liter-
ally, but is a metaphor for any political process in which mass pressures are brought to bear.
8. However, in contrast to Dutt and Mitra (2002), Wu (2014) obtains different results for
democracies and autocracies.
9. For extensions and refinements of this typology, see, for example, Hadenius and Teorell
(2007) and Wright (2008a).
10. As the reference to minor parties suggests, Geddes (1999, 2003) classifies most multiparty
autocracies as single-party regimes.
11. Based on World Bank tariff data, available at http://econ.worldbank.org/WBSITE/
EXTERNAL/EXTDEC/EXTRESEARCH/0,,contentMDK:21051044~pagePK:64214825~
piPK:64214943~theSitePK:469382,00.html (accessed August 28, 2012).
12. Over 98 percent, according to Goemans, Gleditsch, and Chiozza (2009).
13. “The New Face of Hunger,” The Economist, April 17, 2008.

References
Acemoglu, Daron, and James A. Robinson. 2005. Economic Origins of Dictatorship and
Democracy. New York: Cambridge University Press.
Alesina, Alberto, and Dani Rodrik. 1994. Distributive Politics and Economic Growth.
Quarterly Journal of Economics 109 (2): 465–490.
Alvarez, Mike, José Antonio Cheibub, Fernando Limongi, and Adam Przeworski. 1996.
Classifying Political Regimes. Studies in Comparative International Development 31 (2): 3–36.
Boix, Carles. 2003. Democracy and Redistribution. Cambridge, UK:  Cambridge University
Press.
Bueno de Mesquita Bruce, Alastair Smith, Randolph M. Siverson, and James D. Morrow. 2003.
The Logic of Political Survival. Cambridge, MA: MIT Press.
Carey, John M., and Matthew S. Shugart. 1995. Incentives To Cultivate A Personal Vote: A Rank
Ordering of Electoral Formulas. Electoral Studies 14 (4): 417–439.
Cheibub, José Antonio, Jennifer Gandhi, and James Raymond Vreeland. 2010. “Democracy
and Dictatorship Revisited.” Public Choice 143 (1–2): 67–101.
Chow, Wilfred M., and Daniel Yuichi Kono. 2014. Entry, Vulnerability, and Trade Policy: Why
Some Autocrats Like International Trade. Working paper.
Clague, Christopher, Philip Keefer, Stephen Knack, and Mancur Olson. 1996. Property and
Contract Rights in Autocracies and Democracies. Journal of Economic Growth 1: 243–276.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper & Row.
 Daniel Yuichi Kono   313

Dutt, Pushan, and Devashish Mitra. 2002. Endogenous Trade Policy Through Majority
Voting: An Empirical Investigation. Journal of International Economics 58: 107–133.
Edwards, Sebastian. 1998. Openness, Productivity and Growth: What Do We Really Know?
Economic Journal 108: 383–398.
Ehrlich, Sean D. 2011. Access Points: An Institutional Theory of Policy Bias and Policy Complexity.
New York: Oxford University Press.
Ehrlich, Sean D. 2007. Access to Protection:  Domestic Institutions and Trade Policy in
Democracies. International Organization 61 (3): 571–605.
Frankel, Jeffrey A., and David Romer. 1999. Does Trade Cause Growth? American Economic
Review 89 (3): 379–399.
Franzese Robert J. 2002. Electoral and Partisan Cycles in Economic Policies and Outcomes.
Annual Review of Political Science 5: 369–421.
Frye, Timothy, and Edward D. Mansfield. 2003. Fragmenting Protection:  The Political
Economy of Trade Policy in the Post-Communist World. British Journal of Political Science
33 (4): 635–657.
Frye, Timothy, and Edward D. Mansfield. 2004. Timing Is Everything: Elections and Trade
Liberalization in the Postcommunist World. Comparative Political Studies 37 (4): 371–398.
Galetovic, Alexander, and Ricardo Sanhueza. 2000. Citizens, Autocrats, and Plotters: A Model
and New Evidence on Coups d’Etat. Economics & Politics 12 (2): 183–204.
Gandhi, Jennifer and Adam Przeworski. 2006. “Cooperation, Cooptation, and Rebellion
Under Dictatorships.” Economics & Politics 18 (1): 1–26.
Gawande, Kishore, and Usree Bandyopadhyay. 2000. Is Protection for Sale? Evidence on the
Grossman-Helpman Theory of Endogenous Protection. Review of Economics and Statistics
82 (1): 139–152.
Geddes, Barbara. 1999. Authoritarian Breakdown: Empirical Test of a Game Theoretic Argu­
ment. Paper presented at the annual meeting of the American Political Science Association,
Atlanta, GA, September.
Geddes, Barbara. 2003. Paradigms and Sand Castles: Theory Building and Research Design in
Comparative Politics. Ann Arbor: University of Michigan Press.
Goemans, H. E., Kristian Skrede Gleditsch, and Giacomo Chiozza. 2009. Introducing
Archigos: A Data Set of Political Leaders. Journal of Peace Research 46 (2): 269–283.
Goldberg, Penelopi, and Giovanni Maggi. 1999. Protection for Sale: An Empirical Investigation.
American Economic Review 89: 1135–1155.
Grossman, Gene M., and Elhanan Helpman. 1994. Protection for Sale. American Economic
Review 84 (September): 833–850.
Hadenius, Axel, and Jan Teorell. 2007. Pathways from Authoritarianism. Journal of Democracy
18 (1): 143–157.
Hankla, Charles R. 2006. Party Strength and International Trade: A Cross-National Analysis.
Comparative Political Studies 39 (9): 1133–1156.
Hankla, Charles R., and Daniel Kuthy. 2013. Economic Liberalism in Illiberal Regimes:
Authoritarian Variation and the Political Economy of Trade. International Studies Quarterly
57 (3): 492–504.
Kee, Hiau Looi, Alessandro Nicita, and Marcelo Olarreaga. 2009. Estimating Trade
Restrictiveness Indices. Economic Journal 119 (534): 172–199.
Kono, Daniel Yuichi. 2008. Democracy and Trade Discrimination. Journal of Politics 70 (4):
942–955.
314   Authoritarian Regimes

Kono, Daniel Yuichi. 2009. Market Structure, Electoral Institutions, and Trade Policy. Inter­
national Studies Quarterly 53 (4): 885–906.
Kono, Daniel Yuichi. 2006. Optimal Obfuscation: Democracy and Trade Policy Transparency.
American Political Science Review 100 (3): 369–384.
Kono, Daniel Yuichi, and Gabriella R. Montinola. Forthcoming. Foreign Aid, Time Horizons,
and Trade Policy. Forthcoming in Comparative Political Studies.
Lohmann, Susanne, and Sharyn O’Halloran. 1994. Divided Government and U.S. Trade
Policy: Theory and Evidence. International Organization 48 (4): 595–632.
Magaloni, Beatriz. 2006. Voting for Autocracy: Hegemonic Party Survival and Its Demise in
Mexico. New York: Cambridge University Press.
Mansfield, Edward D., and Marc L. Busch. 1995. The Political Economy of Nontariff Barriers:
A Cross-National Analysis. International Organization 49 (4): 723–749.
Mansfield, Edward D., Helen V. Milner, and Peter Rosendorff. 2000. Free to Trade:
Democracies, Autocracies, and International Trade. American Political Science Review 94
(2): 305–322.
Mansfield, Edward D., Helen V. Milner, and Peter Rosendorff. 2002. Why Democracies
Cooperate More:  Electoral Control and International Trade Agreements. International
Organization 56 (3): 477–514.
Mayer, Wolfgang. 1984. Endogenous Tariff Formation. American Economic Review 74
(5): 970–985.
Milner, Helen V., and Keiko Kubota. 2005. Why the Move to Free Trade? Democracy and Trade
Policy in the Developing Countries. International Organization 59 (1): 107–144.
Mitra, Devashish, Dimitrios D. Thomakos, and Mehmet A. Ulubasoglu. 2002. “Protection
for Sale” in a Developing Country: Democracy vs. Dictatorship. Review of Economics and
Statistics 84 (3): 497–508.
Nielson, Daniel L. 2003. Supplying Trade Reform:  Political Institutions and Liberalization
in Middle-Income Presidential Democracies. American Journal of Political Science 47
(3): 470–491.
O’Rourke, Kevin H., and Alan M. Taylor. 2007. Democracy and Protectionism. In The New
Comparative Economic History: Essays in Honor of Jeffrey G. Williamson, edited by Timothy
J. Hatton, Kevin H. O’Rourke, and Alan M. Taylor, 193–216. Cambridge, MA: MIT Press.
Porto, Guido, and Bernard M. Hoekman, eds. 2010. Trade Adjustment Costs in Developing
Countries: Impacts, Determinants and Policy Responses. Washington, DC: World Bank.
Powell, Jonathan. 2012. Determinants of the Attempting and Outcome of Coups d’état. Journal
of Conflict Resolution 56 (6): 1017-1040.
Rogowski, Ronald. 1989. Commerce and Coalitions:  How Trade Affects Domestic Political
Alignments. Princeton, NJ: Princeton University Press.
Rogowski, Ronald. 1987. “Trade and the Variety of Democratic Institutions.” International
Organization 41 (2): 203–223.
Sachs, Jeffrey D., and Andrew Warner. 1995. Economic Reform and the Process of Global
Integration. Brookings Papers on Economic Activity 1: 1–118.
Stolper, Wolfgang F., and Paul A. Samuelson. 1941. Protection and Real Wages. Review of
Economic Studies 9 (November): 58–73.
Svolik, Milan W. 2009. Power Sharing and Leadership Dynamics in Authoritarian Regimes.
American Journal of Political Science 53 (2): 477–494.
Tavares, José. 2008. Trade, Factor Proportions and Politics. Review of Economics and Statistics
90 (1): 163–168.
 Daniel Yuichi Kono   315

Thyne, Clayton L. 2010. Supporter of Stability or Agent of Agitation? The Effect of US Foreign
Policy on Coups in Latin America, 1960–99. Journal of Peace Research 47 (4): 449–461.
Verdier, Daniel. 1998. Democratic Convergence and Free Trade. International Studies Quarterly
42 (1): 1–24.
Weeks, Jessica L. 2008. Autocratic Audience Costs:  Regime Type and Signaling Resolve.
International Organization 62 (1): 35–64.
World Bank. 2008. Rising Food Prices: Policy Options and World Bank Response. Background
Note for the Development Committee 43225 (http://www-wds.worldbank.org/exter-
nal/default/WDSContentServer/WDSP/IB/2008/04/09/000334955_20080409041811/
Rendered/PDF/432250BR0SecM211081020080Box327353B.pdf).
Wright, Joseph. 2008a. Do Authoritarian Institutions Constrain? How Legislatures Affect
Economic Growth and Investment. American Journal of Political Science 52 (2): 322–343.
Wright, Joseph. 2008b. To Invest or Insure? How Authoritarian Time Horizons Impact Foreign
Aid Effectiveness. Comparative Political Studies 41 (7): 971–1000.
Wu, Wen-Chin. 2014. When Do Dictators Decide to Open Trade Regimes? Inequality and
Trade Openness in Authoritarian Countries. International Studies Quarterly. doi: 10.1111/
isqu.12149
Wu, Wen-Chin. 2013. Disguised Protectionism under Authoritarianism:  Why Do Some
Dictatorships Have Lower But More Dispersed Tariffs? Working paper.
 Chapter 17

D omestic G e o g ra ph y
and P olicy Pre s su re s

Ke rry A. C hase

In the study of political economy, policy pressures figure prominently. Policy pressures
arise because self-interested economic agents have preferences about policies (and cor-
respondingly, policy-making institutions). Stakeholders act on their preferences by pol-
iticking on their own or in groups. Whenever government action has the potential to
redistribute income, policy making tends to be politicized.
Analytical approaches to policy pressures work from three building blocks. The first
is preferences, defined in terms of individual interests; the second is strategies for pursu-
ing these preferences by way of lobbying and collective action; and the third is the aggre-
gation of preferences into policy through the workings of political institutions (Frieden
1999; Lake 2009). Domestic geography can factor in at every step. If the material endow-
ments that shape underlying interests are location specific in some way, geography may
influence preference formation. Geography also affects collective action capabilities,
helping or hindering efforts to pressure government. If elected office is organized terri-
torially, geography can structure the representation of interests inside political systems.
Political economy research has analyzed the channels through which domestic geog-
raphy conditions policy pressures and the ultimate effects on policy outputs. Many of
these studies touch on geography without engaging it squarely. Only recently have place
and space entered the literature more systematically. One author remarks that “the use
of geography as an explanatory variable has been underutilized in theories of inter-
est group politics” (Schiller 1999, 770). Fifteen years later progress is evident, yet this
research agenda remains young.
Geography shows up in distinctive ways in different lines of research—thus it takes
on different meanings. In one perspective, geography is the distribution of economic
endowments, or of other locational advantages and disadvantages, across place. This
literature carves national landscapes into zones to explain how geographically special-
ized production shapes policy preferences, forms cleavages, and creates distributional
conflict. In a second perspective, geography is about concentration in space. The focus
 Kerry A. Chase    317

of this research is whether spatial proximity helps common interests exert pressure
through collective action. In a third perspective, geography is defined by the electoral
map: it is the location of policy pressures in relation to institutional sources of lever-
age in democratic political systems. These studies assess how the concentration or dis-
persion of organized interests across electoral units affects political responsiveness to
policy pressures.
Because domestic geography is still new to political economy—and rigorous analysis
of it is a large-scale data enterprise that demands good measures of underlying concepts,
challenges that have held back scholarship—big puzzles persist, and conflicting answers
abound. When and why is geography a primary line of cleavage, independent of indus-
try? Is it better for organized interests to be concentrated or dispersed? If concentration
is better, then where: in space or in high-leverage electoral districts?
While political economy research generally takes patterns of specialization as given,
the study of economic geography provides theoretical and methodological underpin-
nings. Before turning to policy pressures, this survey begins with a brief synopsis of eco-
nomic geography.

Economic Geography, Old and New

Geography matters to political economy because economic activities are not evenly
distributed in space. Inside countries, as between them, areas differ in endowments, so
they specialize in different industries. Natural advantages rooted in local endowments
parcel out crop production according to land and climate suitability; mining clusters
around resource deposits; logging flourishes in timberlands. To reduce transport costs
and prevent spoilage, resource-using enterprises ordinarily bunch near sources of basic
inputs—sugar processing near cane and beet fields; wineries alongside vineyards—while
producers of many processed foods and beverages disperse for much the same reasons.
Industries such as shipbuilding and petroleum refining are often hitched to coastal
topography. These forms of natural advantage, dubbed “first-nature geography” because
they reflect attributes of the physical environment independent of human activity, are
easily understood in the traditional Heckscher-Ohlin trade model.
Stretching the limits of this “old” economic geography, the “new” economic geogra-
phy, in spotlighting efficiency gains that flow from spatial relationships between eco-
nomic agents, introduces “second-nature geography.” Disciples of the new economic
geography point out that economic activity is more concentrated than the neoclassi-
cal approach expects; this tendency to localize, they argue, must reflect advantages of
proximity between or among firms, workers, and consumers. Without delving into
the nuances, one set of approaches, pioneered by Krugman (1991), finds incentives for
industries with increasing returns to scale to congregate near consumers to econo-
mize on transport costs; over time, these clusters may be self-reinforcing as inward
labor migration further enlarges the local market, enhancing the temptation for firms
318    Domestic Geography and Policy Pressures

to set up shop close by.1 Another set of approaches formalizes the insights of Alfred
Marshall (1920), who attributed agglomeration to externalities and spillovers from
close economic interactions. These gains from proximity generally fall into three cat-
egories: advantages of specialized skilled labor pools, which reduce labor costs and
deepen labor markets; synergies between downstream producers and upstream sup-
pliers, which increase supply and cut prices for specialized inputs; and knowledge
spillovers between firms that use similar technologies or that share common labor
requirements.2
Empirical appraisals of formal models inspired by the new economic geography
have lagged.3 Studies have documented the unevenness of industrial location across
the United States, Europe, and Japan (e.g., Ellison and Glaeser 1997; Brülhart 2001;
Kalemli-Ozcan, Sørensen, and Yosha 2003). Others assess the relative importance of
natural advantage, increasing returns to scale, and spillovers in variation across indus-
tries and time (e.g., Ellison and Glaeser 1999; Kim 1995, 1999). Most pertinent for politi-
cal economy research is the difficulty of measuring location patterns. Concentration,
Krugman (1991, 5) writes, “is the most striking feature of the geography of economic
activity,” but calculating it is no simple task. Following Krugman’s example, many stud-
ies use a “location Gini,” a Herfindahl index of an industry’s regional and national
employment shares. Standard concentration indices such as this—though straightfor-
ward to compute and relatively manageable in their data requirements—are problem-
atic. These measures do not take account of industrial concentration; thus they cannot
distinguish geographic agglomeration from production that is centralized in a few large
firms at specific places. More significant for scholars interested in the political effects of
industrial location, not its underlying economic causes, geographic areas (however they
are defined territorially) differ in economic size, population, and other spatial features
that introduce measurement bias.
To overcome these drawbacks, scholars have devised theoretically informed and
highly sophisticated geographic concentration measures. The most widely cited, the
“dartboard approach” developed by Ellison and Glaeser (1997), controls for variation
in the size distribution of plants in an industry and in the size of geographic areas. This
index finds that industrial location in US manufacturing looks more centralized than
darts thrown aimlessly at a board: well over half of industries are at least somewhat con-
centrated, and more than a quarter are highly concentrated. Yet the measure presents
problems of its own; like the simpler location Gini, it misses neighborhood effects in
industry agglomerations that cross spatial aggregates (in the US context, states, coun-
ties, and zip codes). This “checkerboard problem” afflicts any measure that neglects
proximity in the distribution of economic activities across areal units of analysis—on
the figurative checkerboard, there is no way to distinguish congregation in adjacent
spaces from opposite ends.
An alternative approach by Busch and Reinhardt (1999) solves the checkerboard
problem, calculating concentration around a geographic central point as a negative
exponential function of distance. (Their findings are reviewed later in the chapter.)
Despite its advantages over other methods, the measure has not found its way into
 Kerry A. Chase    319

empirical work on economic geography. It has, however, established a new standard for
political economy research on domestic geography and policy pressures.4
This brief review of economic geography has three important implications for the
study of policy pressures. First, economic activity indeed tends to congregate in space;
at the same time, the degree of concentration varies across different activities. If domes-
tic geography affects the policy pressures that governments feel, the patterns should be
visible in industry and country cross-sections. Second, though more speculatively, geo-
graphic concentration—in any theoretical framework—signals some sort of advantage,
natural or created. Particularly where clustering reflects increasing returns to scale or
location-specific externalities, economic agents are likely to be involved in foreign trade.
Economic geography models in fact suggest that the concentration-trade relationship
may be self-reinforcing: the more firms and workers congregate, the easier it is to trade;
concentrated industries in turn possess unique advantages, which should facilitate
exports. If the same economic processes drive both location decisions and trade expo-
sure, then it is reasonable to expect concentrated and nonconcentrated industries to
behave differently politically. Third, testing hypotheses about domestic geography and
policy pressures requires good measures of the spatial distribution of economic activ-
ity. Core concepts such as concentration and dispersion, however, lack clear operational
definition. Not only does effective measurement involve large data requirements, but
the consideration of space also raises methodological obstacles that continue to compli-
cate empirical studies of location patterns.
If geographic concentration poses methodological challenges for economic geog-
raphy, it raises even more complications for political economy research, to which this
chapter now turns.

Geography and Preference Formation

The geographic sources of policy preferences are a recurrent theme of pressure group
analyses. Gerschenkron (1966) famously spotlights the agricultural lands of Imperial
German East Elbia, where the topography favored grain cultivation on big estates, as
a champion of higher tariffs. Kindleberger’s (1975) classic study of nineteenth-century
European trade repeatedly identifies organized interests—not only Manchester tex-
tile merchants—by city as well as industry. Another landmark study, by Bauer, Pool,
and Dexter (1963), dubbed the seat of the US automotive industry, Detroit, a “hotbed
of free traders” for its civic activism on trade issues in the 1950s. Both Manchester and
Detroit, like Germany’s Junkers, eventually swung over to protectionism amid competi-
tive decline. Yet the linkage of policy preferences with place has a long and rich his-
tory. Behind this association are the conjectures that geographic area is synonymous
with a dominant industry, a connection drawn from factor endowments or other
location-specific traits, and that spatial proximity nurtures common interests, even
shared identity, on economic policy questions. In most accounts, the imputation of
320    Domestic Geography and Policy Pressures

preferences from place functions as an interpretive scheme not intended for rigorous
testing. Only lately have more careful and systematic investigations been undertaken of
domestic geography as a source of policy preferences.
The study of sectionalism in the United States provides an example of place as a proxy
for preferences. “Historically, preferences regarding the tariff have been geographically
determined,” O’Halloran (1994, 45) writes of nineteenth-century tariff debates. The notion
that trade policy orientations sorted into “politically relevant regional groupings” (Lake
and James 1989, 12) is characteristic of analyses of this period. A considerable literature
interprets tariff swings in terms of the shifting positions of three geographic sections, the
Northeast, South, and Midwest. In the early nineteenth century, northeastern shipping
and mercantile interests joined southern planters in opposing tariffs, while midwestern
areas favored protectionism. In midcentury these alignments reshuffled, the industrializ-
ing North turning to protectionism as the agrarian West allied with the pro-trade South,
a union that Lake and James (1989) attribute to the boost in grain exports after Britain’s
repeal of the Corn Laws. From the Civil War to the 1920s, not only production struc-
tures but the party system also organized along sectional lines, creating partisan divisions
on tariff issues. According to Hiscox (1999, 682), the interweaving of party and section
reflected in protariff capitalists of the Republican Northeast and antitariff agrarians in
the Democrats’ Solid South “emerges as a consequence of the factoral-regional distinc-
tiveness of party constituencies.” Partisan and regional cleavages are likewise apparent
in the making of the Smoot-Hawley tariff, though by this time both industry and agri-
culture were fracturing over tariffs (Eichengreen 1989). Only in the 1930s did partisan
conflict—and by implication sectional division—begin to subside as production grew
less regionally concentrated, causing regions to converge in their trade profiles, and as
party constituencies became more diversified (O’Halloran 1994; Hiscox 1999, 2002).
Geography also features prominently in accounts of US foreign policy. Earlier works
by Smuckler (1953) and Rieselbach (1966) consider the geographic insularity of the
Great Plains and mountain West a source of political attachment to isolationism after
the world wars. Recent research continues this line of thought. Politicians from the
Northeast were more inclined to support, and those from the South and West to oppose,
military action against Germany in World War I, a division that Fordham (2007) links
to wartime changes in state-level exports. During the early Cold War, senators from the
Midwest, mountain West, and South were less supportive than northerners of military
buildup and overseas commitments, primarily because of regional differences in trade
exposure and overseas investment (Fordham 1998). Over a longer time span, Trubowitz
(1998, 4) accentuates “America’s regional diversity as the most important source of ten-
sion and conflict over foreign policy.” In his view foreign policy divisions in Congress
reflect regional cleavages over the costs and benefits of international economic exchange
and national defense programs. After 1890 these rifts split the Republican Northeast and
Midwest from the Democratic South; in the 1930s, the pivot of conflict separated coastal
areas from the interior; and in the 1960s, new fault lines formed between the traditional
manufacturing belt and the Sunbelt. At each turn, sectional and often partisan division
in foreign policy attitudes mounted (Trubowitz 1998).
 Kerry A. Chase    321

While the historical US context is the focus of many sectional interpretations, this
mode of analysis travels more widely. In a study of imperial German tariffs, for example,
Schonhardt-Bailey (1998) posits that a regional split between Ruhr heavy industry and
Saxon light industry fractured the party system and eventually the historic “marriage of
iron and rye” with eastern plantation owners. Her empirical analysis confirms that the
industrial regions voted for different parties, while their Reichstag deputies broke over
agricultural tariffs after 1879.
All in all, the works referenced in the preceding paragraphs are strongly suggestive of
connections between domestic geography and preference formation, even if geography
is often incidental to the story. These studies also point to directions for future research.
For one thing, policy pressures are rarely observed directly: it is difficult to systematically
document collective behavior on the part of regional economic actors, so preferences are
generally seen by proxy in political behavior at higher levels, such as legislative voting.
More centrally for the relationship between place and preferences, a region—in many
of these examples, groupings of ten or more US states, defined less by economic struc-
tures than by conventional classifications such as date of statehood or status in the Civil
War—is too blunt an indicator for industry location. Not only are the exact boundaries
of the three canonical US regions of the pre-1930 period arbitrary, but each was more
economically diverse than simplified labels like industrial North or agrarian South can
apprehend. Getting at the causal links between geography and policy preferences calls
for more exact spatial measurement of economic activity—a demanding task for sure,
but a potentially surmountable one with available census data. The cross-referencing of
industry-level indicators and electoral districts in Schonhardt-Bailey (1998) is an exem-
plar in this regard.
One way around the problem of directly observing policy preferences in spatial con-
text is to exploit election referendums and survey-generated data. For example, Irwin
(1994) analyzes the British general elections of 1906, in effect a referendum on free trade.
His models reveal that areas where export industries and traded goods occupations con-
gregated went strongly for the antitariff party, the Liberals, who carried the election,
perpetuating the free-trade policy. Urbatsch (2013) examines Costa Rica’s 2007 referen-
dum on the Central American Free Trade Agreement; one result was that districts with
more manufacturing workers were more supportive of ratification, a surprise given the
country’s trade deficit in manufactured goods. Ardanaz, Murillo, and Pinto (2013) use
survey responses to evaluate urban-rural cleavages over trade in postcrisis Argentina.
They find that respondents from import-competing industrial areas, predominantly
the major cities, are less favorable to trade than are respondents from the agricultural
hinterlands, where import exposure is lower. This suggests that trade policy attitudes
are specific to place as much as to industry, at least in contemporary Argentina, where
urban workers in both tradable manufactures and nontradable services feel the effects
in their food bills.
Other research indicates that geography factors into political backlash against job loss.
In the United States, regional disparities are present in the incidence of trade-related
layoffs in manufacturing, which have concentrated in the Northeast, Rust Belt, South,
322    Domestic Geography and Policy Pressures

and Midwest but have largely bypassed the Great Plains. Higher layoffs, Margalit
(2011) finds, reduced votes for the incumbent, George W. Bush, in the 2004 presiden-
tial race, but federal Trade Adjustment Assistance for hard-hit areas cushioned these
losses at the ballot box. Compared to manufacturing, agglomeration is less common
in services; however, geographically concentrated services are more tradable interna-
tionally, placing workers in these sectors at greater risk of offshoring. Using a locational
Gini index, Jensen and Kletzer (2006) identify the sectors and occupations in which
offshoring is easiest. An implication of their study is that job loss should be centralized
in the urban areas where service agglomeration is most prevalent. In that case, political
reactions—lobbying activity, petitions for administered remedies, or shifts in individual
attitudes and voting behavior—may replicate these spatial lines of cleavage. Focusing on
a single industry, motion pictures, one of the most heavily traded and densely localized
services, Chase (2008) explains that an outmigration of filming from Hollywood set the
area’s labor groups at odds: low-skilled workers organized to demand policy measures
against offshoring, but high-skilled occupations were less inclined to join this campaign.
Close analysis of specific sectors allows more nuanced empirical tests as a comple-
ment to the wider national and cross-national studies that dominate political economy
scholarship. Sector-level approaches may be especially useful for deepening under-
standing of trade in services, an area of growing importance. While most services are
scattered geographically to be near end markets, a few, such as motion pictures and
investment banking, tend to cluster (Kolko 2010). Not only in Hollywood does motion
picture production crowd in a central area; in other places too it is highly centralized, for
example in France, where the industry bunches in and near the eighth arrondissement
of Paris, and in England west of London (Scott 2000). Financial services likewise con-
solidate in financial centers, and some have persisted as “capitals of capital” for centuries
(Kindleberger 1974). In both instances, geographic concentration signals the presence of
location-specific externalities of the sort accentuated in the new economic geography.
Both sectors are highly exposed to trade, as agglomeration models would expect.
These attributes point to some testable political implications. One is a close identity
of interests, hence common preferences, among firms and workers in agglomerations
of these sorts. Scott (2000, 4) notes of motion pictures that “producers’ individual com-
petitive destinies are all intertwined with one another, which implies … that they are also
bound together in a sort of de facto political coalition.” These coalitions may begin to
fray, however, if agglomeration gives way to dispersion, as Chase (2008) intimates about
labor-market divisions over offshoring. Another implication is that core-periphery
cleavages are likely to be cut between producers housed inside and outside these indus-
trial clusters, suggesting that geography can function as a primary line of cleavage
within an industry. In this context, Verdier (1998) posits that the internationalization,
amalgamation, and spatial centralization in banking under the gold standard divided
money-center banks, which benefited from the concentration of capital, and financial
institutions in the hinterlands, which lost out as deposits flowed to the center.5 Though
this hypothesis is not tested directly, it offers a compelling picture of the political rifts
that are bound to arise when location differentiates how globalization affects a sector.
 Kerry A. Chase    323

Future research could find theoretical footing in the new economic geography or
in other theoretical models that give preference formation a spatial component. Most
existing studies of place and policy preferences build off factor-based or sector-based
trade models. Geography only enters these approaches ad hoc when certain advan-
tages and disadvantages happen (or are believed) to be unevenly distributed in space.
Rarely does theoretical work connect geography and preference formation directly. If
economic assets are specific both to industry and to place, however, then there should
be “a coincidence of spatial economic interests among a region’s residents” (Cassing,
McKeown, and Ochs 1986, 848). This raises the possibility of within-industry cleavages
between producers in different regions, for example at points in the business cycle at
which tariff rents are not evenly shared. Suggestive examples exist of regional realign-
ment within US industry, such as the migration of textiles from New England to the
Carolinas or of steel mini-mills to the Sunbelt as integrated steelworks suffered Rust Belt
decline. Yet the trade preferences of “old” and “new” industrial regions have not been
systematically analyzed.
Further theoretical groundwork is needed to add rigor to the study of geography
and preference formation. Geography has loosely guided sectional accounts of broad
national trends, especially in the history of US foreign economic policy, and impres-
sionistic examples of place as a factor in preferences abound. But a fuller understanding
of the locational determinants of policy preferences demands more spatially informed
research designs.

Geography and Collective Action

Geography has an impact on collective action because the location of economic agents
in space influences the capacity to organize in pursuit of common policy goals. The
causal linkage between political mobilization and geography originates in Olson’s (1971)
Logic of Collective Action, though geography is nowhere in the study. Olson’s insight
was that shared interests do not guarantee successful organization; to exert political
pressure as a group, individuals with common concerns must first overcome free rid-
ing. A favorable policy, in this setup—a tariff, in the canonical example—benefits all
producers independent of how much each one contributes to obtaining it. Action on
behalf of collective interests therefore resembles a public good, which tempts free rid-
ers to contribute less than their fair share in the hopes of partaking of the fruits of oth-
ers’ labor. Interests that cannot effectively monitor contributions and police shirking,
Olson (1971, 165) concluded, are fated to be “forgotten groups,” consigned to “suffer in
silence.”
While group size was Olson’s focus, concentration dominates the empirical analysis
of collective action. Concentrated interests, the thinking goes, have fewer numbers to
manage than dispersed ones, aiding the assignment of costs, and each contributor cap-
tures a larger share of the gains, making the benefits of the effort easier for all to pocket.
324    Domestic Geography and Policy Pressures

In these ways, concentration aids collective action—a proposition widely accepted

before Olson, perhaps most of all through Schattschneider’s (1935) demonstration that
concentrated producer interests dominated diffuse consumer interests in the making of
the historic Smoot-Hawley tariff.
A persistent puzzle is that while concentration helps to mobilize group pressure, it
ought to be detrimental to forming majoritarian coalitions in a democracy. Pincus’s
(1975) analysis of the 1824 Tariff Act in the United States was one of the first to bring
geography into the study of collective action. On the one hand, Pincus reasoned, the
communications of the time were poor, and interests spread over a wide expanse could
not easily find one another; thus potential lobby partners needed to be in close proxim-
ity to exchange information, monitor effort, and punish slacking. On the other hand,
the interests most favored organizationally would be too narrow to sway many votes.
Leaving the apparent contradiction unresolved, geographically concentrated indus-
tries received higher tariff rates in the 1824 bill, but so did industries present in many
states.
It is conceivable that concentration and dispersion could interact in supportive ways,
with concentration helping to mobilize pressure and dispersion increasing the number
of legislators who will plead a pressure group’s case. An illustrative example is Britain’s
repeal of the Corn Laws (Schonhardt-Bailey 1991). The core export interest, textiles, was
centered in Lancashire County, its bull’s-eye the city of Manchester. This localization
of textile production concentrated the gains from free trade and the resources to pur-
sue it, spawning a free-trade lobby in the anti–Corn Law League, likewise centralized
in Lancashire. But pressure from this one region was not enough to overturn agricul-
tural duties, which required a parliamentary majority.6 The key to the anti–Corn Law
League’s ultimate success, according to Schonhardt-Bailey, was the growing industrial
diversification of British exports after 1830, hence the wider spread of export interests
across parliamentary constituencies. Because of growing textile concentration, which
provided financial and organizational backing, and the increased dispersion of other
exporters, which turned out the votes on a national basis, “the league had the best of
both worlds” (Schonhardt-Bailey 1991, 48).
The coincidence of concentration and dispersion helps to clarify Britain’s move to
free trade in the 1840s, raising the question of whether coalitions of concentrated and
dispersed interests have formed in other settings to pool their political advantages. For
any one industry, however, concentration comes at the expense of dispersion and vice
versa; it is not possible to have both. At the industry level, the competing implications of
concentration versus dispersion have been tested in statistical models of trade protec-
tion, but with inconclusive results. Studies focused on the United States find support for
one or the other effect, or neither (e.g., Ray 1981; Lavergne 1983; Godek 1985; Gardner
1987; Trefler 1993). Analyses of other countries are equally mixed: geographically con-
centrated industries receive more trade protection in Australia (Conybeare 1978) and
the United Kingdom (Greenaway and Milner 1994), but not in Canada (Caves 1976). In
a study of thirty-two democracies, geographic concentration is a statistically significant
correlate of trade protection in developed countries; it is not significant, however, for
 Kerry A. Chase    325

developing countries or in the full sample (Mukherjee, Smith, and Li 2009). In sum,
results are sensitive to measurement and model specification, and conflicting findings
abound.
Articles by Busch and Reinhardt (1999, 2000, 2005) significantly clarify the terms of
the concentration-dispersion debate. The first of these contributions points out three
major flaws in earlier research (Busch and Reinhardt 1999). First, authors tend to con-
flate geographic and political concentration without precisely identifying which effect
is being measured and assessed. Geographic concentration in physical space is part of
a demand-side explanation of group pressure exerted via collective action, while politi-
cal concentration across electoral constituencies is a supply-side hypothesis about how
political systems react to these pressures in policy outputs.7 Second, commonly used
measures of geographic concentration, by omitting distance, fall prey to the “checker-
board problem” described in the first section; thus they cannot discern whether concen-
trated interests live side by side. Third, earlier studies neglect conditional relationships
between concentration and group size, which raises the possibility that too much politi-
cal concentration harms a large group less than a small one, even if size, as Olson (1971)
suggests, handicaps large groups in collective action.
Unraveling the expected effects of economic concentration and political concen-
tration yields two alternative hypotheses about geography. One proposition, the
“close-group” hypothesis, is about organizing costs: it expects close physical proxim-
ity among the members of a group to allow greater contact, aid information exchange,
facilitate knowledge sharing, and increase the density of social networks, lowering orga-
nizational and monitoring costs for the lobbying effort. Geographically concentrated
interests, in the close-group hypothesis, are more apt to effectively mobilize pressure,
increasing the chances that policy demands translate into outcomes. The other proposi-
tion, the “dispersed-group” hypothesis, is about representation: it expects a presence in
many constituencies to broaden representation, enhancing the prospects for legislative
coalition-building, hence influence, in majoritarian systems. To test the independent
effects of each, Busch and Reinhardt (1999) introduce measures to capture concentra-
tion on two maps, a spatial one in which distance matters, and an electoral one in which
it does not.
The results of their analysis of nontariff barriers in the United States strongly favor
the close-group hypothesis:  geographically concentrated industries are more likely
to obtain import relief than industries that are not concentrated in space. This effect
trumps the benefits of political dispersion—as one might expect, given that Congress
does not legislate trade remedies, instead delegating this authority to administrative
judgment of industry petitions—though it is notable that a few very large, politically
concentrated industries receive nontariff barriers as well (Busch and Reinhardt 1999).
A  follow-up study considers how geographic concentration affects individual-level
behavior (Busch and Reinhardt 2000). Using survey data, this inquiry finds that workers
in trade-exposed industries are more inclined to have an opinion about trade, contrib-
ute to political campaigns, and cast ballots at election time if their industry is geographi-
cally concentrated than if it is not. Moving beyond the United States to Europe, a third
326    Domestic Geography and Policy Pressures

study shows that higher turnout rates at the polls by workers in geographically con-
centrated traded goods industries hold up across countries—even in the Netherlands,
where the electorate votes as a national constituency rather than for local office (Busch
and Reinhardt 2005). These results indicate that the advantages of geographic concen-
tration in amplifying “voice” do not depend on how political systems represent interests;
they translate to any electoral setting.
Taken together, these are the strongest conclusions the literature on geographic con-
centration has produced to date. Still, questions for future research persist. Above all,
more direct observation would help to flesh out the causal pathways from proximity to
pressure to policy. Convincing evidence tying spatial proximity to effective mobiliza-
tion would document and explain some kind of behavior taking place at the group level.
Existing studies of geography and collective action instead generally rely on policy out-
puts, such as tariffs or nontariff barriers, to proxy for the pressure that organized groups
exert in politics. While this is a sensible way to pit clashing arguments about concentra-
tion and dispersion against one another, it assumes that all industries want government
assistance, other things being equal, and that they want it in the form that the researcher
chooses to study. A standard measure such as nontariff barriers, it is true, may well bias
statistical tests against (not in favor of) finding an effect, since industries that do not
ask for or obtain import relief count against one hypothesis or the other. Yet it is still
worth knowing whether geographically concentrated interests are more inclined to file
petitions for import relief, testify at hearings, initiate contacts with legislators, expend
funds on lobbying efforts, or vote in blocks. While these variables may harbor their
own biases, precise measures of pressure are lacking in this line of research. In the few
instances where they have been tried (e.g., Bombardini and Trebbi 2012), the evidence
is mixed.
Without direct observation, it is also hard to know whether technological progress
affects the geography of collective action. Physical proximity, it stands to reason, was
pivotal in the nineteenth century, when common interests had to collaborate to send
petitions to the capital or refashion voter rolls nationwide, operations made difficult by
poor communications (Pincus 1975; Schonhardt-Bailey 1991). But why should distance
continue to inhibit collective action in today’s information age? Busch and Reinhardt
(2000, 2005) suggest that there remains no good substitute for direct social contact.
But their outcomes of interest—opinion formation, campaign contributions, and voter
turnout—are individual behaviors with tenuous connections to collective action. Other
research concludes that geographically concentrated US industries do not in fact con-
tribute more money to political campaigns, precisely because they have better ways to
flex their political muscle, such as lobbying or direct appeals to voters (Grier, Munger,
and Roberts 1994; Damania, Fredriksson, and Osang 2004). This raises doubts that indi-
vidual behavior maps directly into group pressure, just as one might expect in light of
Olson’s famous theory of conflict between individual and group. Until research takes a
closer look at collective action, the close-group hypothesis, though compelling, remains
unproven.
 Kerry A. Chase    327

Geography and Political

Representation

Geography has the potential to affect the representation of interests in politics because
interests are not evenly scattered across electoral divisions. Whether geography mat-
ters for representation and exactly how it matters depends, however, on electoral rules
(see Rickard, this volume). In political systems in which elected office is geographically
assigned, the location of interests in electoral space is pivotal for office seekers, who
must compete for power on a territorial basis. In situations where electoral rules allocate
representation according to territory irrespective of population, such as the US Senate
or Japan’s erstwhile single-nontransferable vote, minority groups—namely farmers in
lightly settled rural areas—are reputed to hold sway. On these grounds, Rogowski (2002,
204) adds to the criterion of a territorially defined electorate that “[o]‌nly in small-district
electoral systems does the geographical concentration or dispersion of interests affect
their institutional power.” Under proportional representation, in contrast, politics are
less responsive to geography: candidates for office compete nationally, not locally, and
constituencies tend to be larger in size. It is therefore little surprise that research often
focuses on the United States, where geographic representation in single-member dis-
tricts forces members of Congress to pay close attention to the location of interests on
the electoral map.
Arguments about geography and representation, like those emphasizing collective
action by organized interests, place the “close-group” and “dispersed-group” hypotheses
in opposition. In its simplest form, the dispersed-group hypothesis posits that the larger
the number of electoral constituencies in which an interest resides, the wider its repre-
sentation in politics. But clearly dispersion does not translate so easily into influence in
politics—bakeries and milk distributors can be present in every district without com-
manding any one legislator’s attention. Thus there is a tension between interests being
a wide enough presence nationally for their voices to be heard by more than a few rep-
resentatives and being a large enough presence in local electorates for representatives
to do their bidding (Pincus 1975; Caves 1976; Salamon and Siegfried 1977). On the one
hand, larger employment and production shares in a district (or state) give an indus-
try its representative’s ear, so Hollywood holds the attention of California’s congressio-
nal delegation, and the stature of aircraft production in Seattle earned Senator Henry
Jackson (D-WA) the memorable moniker “Senator from Boeing.” On the other hand,
centralization in a few places limits the number of legislators pleading an industry’s case,
necessitating coalition-building to secure patronage. Viewed in this way, the most suc-
cessful interests must be large industries whose voting strength is geographically dis-
persed enough to command extensive concern in the legislature—turning on its head
Olson’s pronouncement about the advantages of small groups and concentration. An
excellent example of this sort is the US steel industry.8
328    Domestic Geography and Policy Pressures

One resolution to the concentration versus dispersion trade-off is to split the differ-
ence. Moderately concentrated industries are dispersed enough to be present in many
electoral districts and also concentrated enough to have clout with their elected repre-
sentatives. Thus it pays to be concentrated, but not too concentrated. Rogowski (2002)
measures employment concentration across congressional districts and finds that
industries widely thought to be influential in politics, such as automobiles, sugar, and
steel, are in fact moderately concentrated. He concludes that the increased geographic
dispersion of US industry since 1960 may explain why trade has become more contested
in Congress—protectionist pressures are gaining strength as they grow less centralized.
Future research might test whether the nonlinear effect of concentration on protection
stands up to scrutiny in a larger sample.
Other studies come down more strongly in favor of political dispersion over geo-
graphic concentration. In this line of thought, the central factor in political system
responsiveness is not how much pressure organized groups can bring to bear; it is
how much electoral gain politicians can expect from channeling rewards to certain
interests rather than others. McGillivray (2004, 8–9) takes this tack, stating that
“politicians do not automatically respond to the loudest demands. They listen to the
industries that affect their chances of reelection.” From this standpoint, concentrated
interests may be able to mobilize pressure, as studies of geography and collective
action suggest, but they will lack sufficient representation when electoral rules favor
dispersed interests. If the influence effect of political dispersion in fact outweighs the
mobilization effect of geographic concentration, then redistributive rewards by gov-
ernment should strongly correlate with the distribution of interests across electoral
boundaries, irrespective of their centralization in physical space. McGillivray (2004)
finds evidence for this: large, decentralized industries gain higher tariffs in the United
States, as do electorally concentrated industries, which hold greater salience with
their local representatives.9
Where industry is precisely located on the electoral map may matter even more than
how dispersed it is. On this count, McGillivray (2004) posits that the effects of industrial
location depend on the level of party discipline in the political system. In a strong-party
system such as Britain’s, parties have powerful incentives to grant assistance to groups
concentrated in marginal electorates that are up for grabs in the next election. In a
weak-party system such as the United States, alternatively, seniority and committee
positions channel redistributive rents to safe seats because these politicians can more
credibly promise future vote trades. The evidence confirms that in high-party-discipline
Canada, marginal districts receive higher tariffs, whereas in the United States the oppo-
site occurs: districts whose representatives hold greater seniority in office and powerful
committee assignments instead receive more protection.10 Having firms in the district
of powerful committee members also increases the chances that an industry will win
administered trade relief in the United States (Hansen 1990). Outside the US context,
Hankla (2008) finds that increased party centralization in India shifted party patronage
to battleground areas for upcoming elections, a pattern that did not emerge when the
parties were decentralized. These analyses lend support to the expectation that strong
 Kerry A. Chase    329

party systems reward electoral marginality, while weak party systems favor proximity to
institutional sources of power.
Partisan effects have received less attention. In the aggregate, one study finds, indus-
tries concentrated in districts held by the majority party in Congress secure higher
tariffs (Fredriksson, Matschke, and Minier 2011). On its face this seems to be further
evidence of agenda control, yet attribution is not so easily assigned. Rarely are industrial
location and party constituencies both organized along the same geographic lines—at
least not today; industry-party overlap may have been more extensive in the past, as
the sectional approaches to US foreign economic policy reviewed earlier suppose—so
most industries have incentives to cultivate and maintain support across the party spec-
trum. Yet a handful of US industries do show strong party associations, a byproduct of
their centralization in areas with little party competition; examples include cotton and
tobacco growers in the South and corn and dairy producers in the Midwest (Fordham
and McKeown 2003, 535). In these instances, industry-party alliances may form. The
consequences of these alliances, however, remain to be fleshed out more fully.
If geography shapes an industry’s overall representation in politics and its access
to well-placed politicians, then industries will have incentives to build coalitions that
accentuate their political strengths and mitigate their shortcomings. How the geography
of representation cycles back into lobbying strategies has not been extensively studied.
An exception is Schiller (1999), who proposes that industrial location across US states
shapes coalition-building. Industries concentrated in populous states, she posits, will
be strong in the House through their presence in a number of electoral districts, but
weak in the Senate; conversely, industries concentrated in less populous states will be
strong in the Senate, where representation is population independent, but weak in the
House. To improve the prospects for passing favorable legislation in both chambers,
industries with different strengths ought to seek each other out as coalition partners.
Case study evidence lends cautious support to this assertion. In the 1980s the US tex-
tile industry, with strong champions in the Senate from the North and South Carolina
delegations, joined forces with apparel producers, whose primary backers were north-
eastern House Democrats. This alliance pushed multiple import quota bills through
Congress. However, a counter-coalition of apparel retail and farm exporters—one
influential in the House, the other in the Senate—mobilized against quotas, assembling
enough votes to block the override of successive presidential vetoes. Lobbying over tex-
tile quotas contrasts with the case of steel, whose “more balanced political geography”
(Schiller 1999, 379) gave the industry leverage with both parties and in both chambers of
Congress, enabling protectionists to “go solo,” mobilizing on their own without having
to barter compromises with coalition allies.
Even the redistributive policy instruments that governments choose may be an out-
come of interactions between industrial geography and political representation. For
example, electoral rules and party systems may predispose the use of one set of policy
tools over another. In weak-party majoritarian systems, distributionally blunt policies
such as tariffs and nontariff barriers make it possible to scatter industry rents nation-
wide, facilitating the construction of legislative coalitions. High party discipline,
330    Domestic Geography and Policy Pressures

however, allows the parties to target constituencies in specific areas where the poten-
tial for electoral gain is greatest, for example by allocating subsidies on a regional basis
(McGillivray 2004). In that case geographic cleavages are likely to form, not in response
to differential demand for redistribution from producers in different areas (e.g., as in
Cassing, McKeown, and Ochs 1986), but rather because areas where industry locates
differ in their electoral value.

Conclusion

Domestic geography surely matters to the policy pressures that organized interests bring
to bear. It matters because interests are unevenly distributed, both in physical space and
on the electoral map, and this unevenness filters through multiple channels: preference
formation, collective action, and political representation. When it comes to redistribu-
tion, we might say, all politics is spatial. Yet research on geography and policy pressures
is still in its infancy, and long-standing puzzles persist. Researchers often define geogra-
phy in different ways, hampering comparison of findings, and empirical work remains
largely confined to the US context, where data are easier to obtain and geography is
reputed to have its greatest effects.
These shortcomings aside, research on the geographic sources of policy pressures
is gaining theoretical and empirical sophistication as better measures are formulated
and spatially organized data are more fully exploited. The proliferation of information
sources, such as geocoded census results and other geographically referenced indica-
tors, and the enhancement of desktop applications to organize, integrate, and analyze
these data augur future advancement in this research agenda. To date, political econ-
omy scholarship has not taken maximum advantage of geographic software systems,
which hold enormous promise for overcoming challenges that have held back earlier
work.11 Better data and measurement allow for more rigorous empirical tests, resulting
in more refined spatial hypotheses and stronger theorizing. With these advances, politi-
cal economy is poised to develop substantial new knowledge about domestic geography
and policy pressures in the years ahead.

Notes
1. For Krugman (1991), this helps to illuminate the historical persistence of the US “manu-
facturing belt” stretching from the mid-Atlantic coast to the Great Lakes region.
2. These three causes of agglomeration also appear in Krugman (1991). Surveys of the new eco-
nomic geography include Krugman (1998); Ottaviano and Puga (1998); and Neary (2001).
3. There is nevertheless a large and growing empirical literature on industrial localization
and its causes; this section singles out just a few of the highlights. For a survey, see Redding
(2010).
 Kerry A. Chase    331

4. Another issue is that extant measures of geographic concentration are sensitive to the scale
of aggregation, known as the “modifiable areal unit problem,” which originates in the arbi-
trary boundaries that spatial units of analysis impose. As a result, any one industry’s geo-
graphic concentration relative to others is dependent on the level of aggregation. Some US
industries, for example, are more concentrated at the state or county level than at smaller
units such as zip codes or metropolitan statistical areas; for others, it is the reverse. See
Kolko (2010). For a good portion (though not all) of political economy research, concen-
tration matters only in relation to predefined political aggregates, such as electoral dis-
tricts, making the modifiable areal unit problem a lesser concern.
5. This core-periphery division appears to characterize lobbying for a central bank in the
United States, as the New York City financial elite pushed for the creation of the Federal
Reserve while consumer banks in the interior showed little interest in financial reform
(Broz 1997).
6. Employment in cotton textiles surpassed 1 percent of the voting population in only 5 per-
cent of electoral districts, so the industry’s clout was too limited in Parliament. McKeown
(1989, 358–359).
7. This distinction is clear in Pincus (1975), but it was clouded in later empirical studies.
8. Steelworks historically were centralized in a few states with significant numbers of elec-
toral votes (Pennsylvania, Ohio, Illinois, Michigan, and Indiana). This industrial geogra-
phy earned the industry the backing of a bipartisan caucus in both chambers of Congress
dedicated to its interests. Moore (1996) cites recent changes in this industrial geography as
the central factor behind “the waning influence of big steel.”
9. As noted in the last section, Busch and Reinhardt (1999) qualify this conclusion after test-
ing both economic and political geography in the same model: only a few very large indus-
tries reap benefits from electoral dispersion in the setting of nontariff barriers.
10. These results are surprising in light of the heavy time dependence in US tariff rates (see
Lavergne 1983). If the interindustry tariff structure persists over decades even as seat hold-
ers come and go, then district-level protection may be less responsive to legislator senior-
ity and agenda-setting power than the evidence appears to indicate.
11. Cognate areas of research, particularly studies of elections, redistricting, and ethnic
conflict, have used geographic software systems more extensively. These packages offer
numerous possibilities, including the ability to geocode spatial data and to match infor-
mation compiled at different spatial levels; enhanced visualization tools, such as maps
overlain with symbols, shading, or coloring to identify spatial patterns; and statistical rou-
tines for multilevel modeling and geographically weighted regression analysis. See Tam
Cho and Gimpel (2012).

References
Ardanaz, M., M. V. Murillo, and P. M. Pinto. 2013. Sensitivity to Issue Framing on Trade Policy
Preferences: Evidence from a Survey Experiment. International Organization 67: 411–437.
Bauer, R. A., I. De Sola Pool, and L. A. Dexter. 1963. American Business and Public Policy: The
Politics of Foreign Trade . New York: Atherton Press.
Bombardini, M., and F. Trebbi. 2012. Competition and Political Organization:  Together or
Alone in Lobbying for Trade Policy? Journal of International Economics 87: 18–26.
332    Domestic Geography and Policy Pressures

Broz, J. L. 1997. The International Origins of the Federal Reserve System . Ithaca, NY: Cornell
University Press.
Brülhart, M. 2001. Evolving Geographical Concentration of European Manufacturing
Industries. Weltwirtschaftliches Archiv 137: 215–243.
Busch, M. L., and E. Reinhardt. 2000. Geography, International Trade, and Political
Mobilization in US Industries. American Journal of Political Science 44: 703–719.
Busch, M. L., and E. Reinhardt. 1999. Industrial Location and Protection: The Political and
Economic Geography of US Nontariff Barriers. American Journal of Political Science
43: 1028–1050.
Busch, M. L., and E. Reinhardt. 2005. Industrial Location and Voter Participation in Europe.
British Journal of Political Science 35: 713–730.
Cassing, J., T. J. McKeown, and J. Ochs. 1986. The Political Economy of the Tariff Cycle.
American Political Science Review 80: 843–862.
Caves, R. E. 1976. Economic Models of Political Choice: Canada’s Tariff Structure. Canadian
Journal of Economics 9: 278–300.
Chase, K. A. 2008. Moving Hollywood Abroad: Divided Labor Markets and the New Politics of
Trade in Services. International Organization 62: 653–687.
Conybeare, J. 1978. Public Policy and the Australian Tariff Structure. Australian Journal of
Management 3: 49–63.
Damania, R., P. G. Fredriksson, and T. Osang. 2004. Collusion, Collective Action and
Protection: Theory and Evidence. Public Choice 121: 279–308.
Eichengreen, B. 1989. The Political Economy of the Smoot-Hawley Tariff. Research in Economic
History 12: 1–43.
Ellison, G., and E. L. Glaeser. 1997. Geographic Concentration in US Manufacturing
Industries: A Dartboard Approach. Journal of Political Economy 105: 889–927.
Ellison, G., and E. L. Glaeser. 1999. The Geographic Concentration of Industry: Does Natural
Advantage Explain Agglomeration? American Economic Review 89: 311–316.
Fordham, B. O. 1998. Building the Cold War Consensus: The Political Economy of US National
Security Policy, 1949–51. Ann Arbor: University of Michigan Press.
Fordham, B. O. 2007. Revisionism Reconsidered: Exports and American Intervention in World
War I. International Organization 61: 277–310.
Fordham, B. O., and T. J. McKeown. 2003. Selection and Influence:  Interest Groups and
Congressional Voting on Trade Policy. International Organization 57: 519–549.
Fredriksson, P. G., X. Matschke, and J. Minier. 2011. Trade Policy in Majoritarian Systems: The
Case of the US. Canadian Journal of Economics 44: 607–626.
Frieden, J. A. 1999. Actors and Preferences in International Relations. In Strategic Choice and
International Relations, edited by D. A. Lake and R. Powell, 39–76. Princeton, NJ: Princeton
University Press.
Gardner, B. L. 1987. Causes of US Farm Commodity Programs. Journal of Political Economy
95: 290–310.
Gerschenkron, A. 1966. Bread and Democracy in Germany. Berkeley: University of California
Press.
Godek, P. E. 1985. Industry Structure and Redistribution through Trade Restrictions. Journal of
Law and Economics 28: 687–703.
Greenaway, D., and C. Milner. 1994. Determinants of the Inter-industry Structure of Protection
in the UK. Oxford Bulletin of Economics and Statistics 56: 399–419.
 Kerry A. Chase    333

Grier, K. B., M. C. Munger, and B. E. Roberts. 1994. The Determinants of Industry Political
Activity, 1978–1986. American Political Science Review 88: 911–926.
Hankla, C. R. 2008. Parties and Patronage: An Analysis of Trade and Industrial Policy in India.
Comparative Politics 41: 41–60.
Hansen, W. L. 1990. The International Trade Commission and the Politics of Protectionism.
American Political Science Review 84: 21–46.
Hiscox, M. J. 2002. International Trade and Political Conflict:  Commerce, Coalitions, and
Mobility . Princeton, NJ: Princeton University Press.
Hiscox, M. J. 1999. The Magic Bullet? The RTAA, Institutional Reform, and Trade
Liberalization. International Organization 53: 669–698.
Irwin, D. A. 1994. The Political Economy of Free Trade: Voting in the British General Election
of 1906. Journal of Law and Economics 37: 75–108.
Jensen, J. B., and L. G. Kletzer. 2006. Tradable Services: Understanding the Scope and Impact of
Services Offshoring. In Brookings Trade Forum 2005: Offshoring White-Collar Work, edited
by S. M. Collins and L. Brainard, 75–133. Washington, DC: Brookings Institution Press.
Kalemli-Ozcan, S., B. E. Sørensen, and O. Yosha. 2003. Risk Sharing and Industrial
Specialization: Regional and International Evidence. American Economic Review 93: 903–918.
Kim, S. 1995. Expansion of Markets and the Geographic Distribution of Economic
Activities:  The Trends in U.S. Regional Manufacturing Structure, 1860–1987. Quarterly
Journal of Economics 110: 881–908.
Kim, S. 1999. Regions, Resources and Economic Geography:  The Sources of US Regional
Comparative Advantage, 1880–1987. Regional Science and Urban Economics 29: 1–32.
Kindleberger, C. P. 1974. The Formation of Financial Centers: A Study in Comparative Economic
History. Princeton Studies in International Finance no.  36. Princeton, NJ:  International
Finance Section, Department of Economics, Princeton University.
Kindleberger, C. P. 1975. The Rise of Free Trade in Western Europe, 1820–1875. Journal of
Economic History 35: 20–55.
Kolko, J. 2010. Urbanization, Agglomeration, and Coagglomeration of Service Industries.
In Agglomeration Economics, edited by E. L. Glaeser, 151–180. Chicago:  University of
Chicago Press.
Krugman, P. 1998. What’s New about the New Economic Geography? Oxford Review of
Economic Policy 14: 7–17.
Krugman, P. R. 1991. Geography and Trade . Cambridge, MA: MIT Press.
Lake, D., and S. James. 1989. The Second Face of Hegemony: Britain’s Repeal of the Corn Laws
and the American Walker Tariff of 1846. International Organization 43: 1–29.
Lake, D. A. 2009. Open Economy Politics:  A  Critical Review. Review of International
Organizations 4: 219–244.
Lavergne, R. P. 1983. The Political Economy of US Tariffs:  An Empirical Analysis. Toronto:
Academic Press.
Margalit, Y. 2011. Costly Jobs: Trade-related Layoffs, Government Compensation, and Voting
in US Elections. American Political Science Review 105: 166–188.
Marshall, A. 1920. Principles of Economics . London: Macmillan.
McGillivray, F. 2004. Privileging Industry:  The Comparative Politics of Trade and Industrial
Policy . Princeton, NJ: Princeton University Press.
McKeown, T. J. 1989. The Politics of Corn Law Repeal and Theories of Commercial Policy.
British Journal of Political Science 19: 353–380.
334    Domestic Geography and Policy Pressures

Moore, M. O. 1996. Steel Protection in the 1980s: The Waning Influence of Big Steel? In The
Political Economy of American Trade Policy, edited by A. O. Krueger, 73–132. Chicago:
University of Chicago Press.
Mukherjee, B., D. L. Smith, and Q. Li. 2009. Labor (Im)mobility and the Politics of Trade
Protection in Majoritarian Democracies. Journal of Politics 71: 291–308.
Neary, J. P. 2001. Of Hype and Hyperbolas: Introducing the New Economic Geography. Journal
of Economic Literature 39: 535–561.
O’Halloran, S. 1994. Politics, Process and American Trade Policy . Ann Arbor: University of
Michigan Press.
Olson, M. 1971. The Logic of Collective Action:  Public Goods and the Theory of Groups .
Cambridge, MA: Harvard University Press.
Ottaviano, G. I. P., and D. Puga. 1998. Agglomeration in the Global Economy: A Survey of the
“New Economic Geography”. World Economy 21: 707–731.
Pincus, J. J. 1975. Pressure Groups and the Pattern of Tariffs. Journal of Political Economy
83: 757–778.
Ray, E. J. 1981. The Determinants of Tariff and Nontariff Trade Restrictions in the United States.
Journal of Political Economy 89: 105–121.
Redding, S. J. 2010. The Empirics of New Economic Geography. Journal of Regional Science
50: 297–311.
Rieselbach, L. N. 1966. The Roots of Isolationism:  Congressional Voting and Presidential
Leadership in Foreign Policy . Indianapolis, IN: Bobbs-Merrill.
Rogowski, R. 2002. Trade and Representation: How Diminishing Geographic Concentration
Augments Protectionist Pressures in the US House of Representatives. In Shaped by War and
Trade: International Influences on American Political Development, edited by I. Katznelson
and M. Shefter, 181–210. Princeton, NJ: Princeton University Press.
Salamon, L. M., and J. J. Siegfried. 1977. Economic Power and Political Influence: The Impact of
Industry Structure on Public Policy. American Political Science Review 71: 1026–1043.
Schattschneider, E. E. 1935. Politics, Pressure, and the Tariff . New York: Prentice-Hall.
Schiller, W. J. 1999. Trade Politics in the American Congress: A Study of the Interaction of
Political Geography and Interest Group Behavior. Political Geography 18: 769–789.
Schonhardt-Bailey, C. 1991. Lessons in Lobbying for Free Trade in 19th-century Britain: To
Concentrate or Not. American Political Science Review 85: 37–58.
Schonhardt-Bailey, C. 1998. Parties and Interests in the “Marriage of Iron and Rye”. British
Journal of Political Science 28: 291–332.
Scott, A. J. 2000. French Cinema: Economy, Policy and Place in the Making of a Cultural-
products Industry. Theory, Culture and Society 17: 1–38.
Smuckler, R. H. 1953. The Region of Isolationism. American Political Science Review
47: 386–401.
Tam Cho, W. K., and J. G. Gimpel. 2012. Geographic Information Systems and the Spatial
Dimensions of American Politics. Annual Review of Political Science 15: 443–460.
Trefler, D. 1993. Trade Liberalization and the Theory of Endogenous Protection:  An
Econometric Study of US Import Policy. Journal of Political Economy 101: 138–160.
Trubowitz, P. 1998. Defining the National Interest: Conflict and Change in American Foreign
Policy . Chicago: University of Chicago Press.
Urbatsch, R. 2013. A  Referendum on Trade Theory:  Voting on Free Trade in Costa Rica.
International Organization 67: 197–214.
Verdier, D. 1998. Domestic Responses to Capital Market Internationalization under the Gold
Standard, 1870–1914. International Organization 52: 1–34.
 Pa rt  V

I N T E R NAT IONA L
N E G OT IAT ION S A N D
I N ST I T U T ION S
 Chapter 18

The Design of T ra de
Agreem e nts

L eslie Johns and L aure n  Pe ritz

Introduction

In recent decades there has been an explosion in the number and importance of prefer-
ential trade agreements. As of 2012, more than 250 of these agreements are in effect, and
dozens more are under negotiation. Every country in the world has signed at least one,
while some have signed more than thirty-five. This increase has prompted many econo-
mists and political scientists to analyze the design of trade agreements.
These agreements have a common objective—to promote international trade—and
all operate independently of the multilateral trade regime. Yet they vary dramatically
in their design. Some preferential trade agreements (PTAs) are bilateral, while oth-
ers are regional. Some impose shallow trade obligations, whereas others impose deep
obligations. Some exist only on paper; others have complex institutions. This variation
suggests two questions. First, how does the design of a PTA affect the behavior of its
members? Second, what designs are optimal for a given political-economic context?
Our primary objective in this chapter is to distill the existing research on PTA design.
Most of this research focuses on the treaty level. That is, scholars have examined how the
design of individual treaties shapes state behavior. Our secondary objective is to suggest
that future research should consider the design of PTAs at the system level. In addition
to studying agreements as isolated “observations” or data points, we should also ask how
these agreements interact with one another and with the multilateral trade regime, pre-
viously manifested in the General Agreement on Tariffs and Trade (GATT) and now in
the World Trade Organization (WTO).
A preferential trade agreement creates legal restrictions on its members’ trade poli-
cies. Such agreements aim to promote economic integration among member countries
by improving market access. For example, the North American Free Trade Agreement
(NAFTA) creates a free trade area in which trade barriers are reduced or eliminated on
338    The Design of Trade Agreements

products exchanged among Canada, Mexico, and the United States. In contrast, customs
unions like the Andean Community eliminate internal trade barriers and set common
tariffs for nonmember countries. Empirical scholars use the term “PTA” in slightly dif-
ferent ways. We use this term to denote any international agreement with limited mem-
bership that restricts the trade policies of its members. This inclusive definition covers
bilateral and regional agreements, free trade areas, customs unions, common mar-
kets, and economic unions (Bhagwati 1993; Krueger 1999; Mansfield and Milner 2012).
While these kinds of agreements differ in certain respects, all constrain trade policy in
some way.
We adopt a strategic, game-theoretic perspective to examine how a PTA changes the
incentives, and hence the behavior, of its members. This perspective allows us to con-
sider the relationship between an institution’s design and its effect. We focus on prefer-
ential trade agreements because these instruments collectively illustrate broad variation
in institutional design. Different design features constrain members in particular ways;
we leverage PTA variation to explain their potential effects on countries’ behavior.
This chapter builds on insights from the literature on the multilateral trade regime.
Over time, new design features have been incorporated into the GATT and now the
WTO. Scholars have shown how changes in these design elements—flexibility, dispute
settlement procedures, scope, and so forth—can influence cooperation among WTO
members (Rosendorff and Milner 2001; Rosendorff 2005; Barton et al. 2008; Kucik and
Reinhardt 2008). We draw on these insights throughout this chapter, integrating our
study of institutional design with a discussion of institutional effects.
Throughout this chapter we use data from several sources, drawing primarily on the
WTO’s Regional Trade Agreements Database, which includes all preferential trade agree-
ments notified to the WTO between 1951 and 2012. These agreements are reciprocal,
meaning that all members commit to granting market access to one another. Because
the WTO records the date a PTA enters into force and whether it is active, we are able
to approximate the number of agreements in effect at a given time.1 We supplement this
information with data from Mansfield and Milner (2012). In addition, the Design of
Preferential Trade Agreements data set from Kucik (2012) provides information on many
design features—flexibility, dispute settlement procedures, and so forth—for approxi-
mately 330 PTAs signed between 1960 and 2008. For a measure of the depth and scope
of trade agreements, we refer to two additional data sets: Reciprocal Trade Agreements
in Asia from Hicks and Kim (2012), which records the depth and pace of trade commit-
ments in sixty-seven Asian trade agreements, and the World Trade Report 2011, which
is an updated version of the data in Horn, Mavroidis, and Sapier (2010) that documents
the scope of a sample of approximately one hundred PTAs.2
We begin by focusing on the treaty level. We describe the immense variation in the
design of PTAs and examine how this variation shapes the behavior of members. We
then step back and examine PTAs at the system level. We show that the proliferation of
PTAs has created complex networks of overlapping treaties and ask how these networks
might promote or hinder international trade cooperation.
 Leslie Johns and Lauren Peritz    339

Design Elements

The growing plethora of trade agreements has created immense variation in the design
of PTAs. In this section we discuss five major PTA design elements: depth, scope, mem-
bership, rigidity, and institutionalization.

Depth
Preferential trade agreements vary in depth, the extent to which a PTA constrains state
behavior. Deeper agreements place more significant limits on state behavior. Depth
often refers to tariff bindings, which are the highest tariffs that a state can impose while
still complying with the agreement. Lower tariff bindings require deeper cooperation
because they grant a state less policy discretion.
Depth can also refer to nontariff barriers, including quantitative restrictions. When
a state imposes a quantitative restriction, such as an import quota, it limits the quan-
tity of an imported good to protect domestic producers. For example, the United States
restricts sugar imports—especially from Mexico and Brazil—to protect US farm-
ers. Many PTAs prohibit quantitative restrictions or limit the circumstances under
which they can be used. Deeper agreements impose stricter limits on the use of these
restrictions.
Preferential trade agreements also often address regulations that create barriers to
trade, such as customs procedures, licensing rules, product standards, and government
procurement rules. For example, states often require government entities to purchase
goods and services from domestic firms. The American Recovery and Reinvestment Act
of 2009 contained a “Buy American” clause that required fund recipients to purchase
goods and services from US companies (Uchitelle 2009). Preferential trade agreements
often limit government procurement rules like the “Buy American” clause to ensure that
foreign firms can compete for government contracts. Nearly 40 percent of trade agree-
ments address government procurement (World Trade Report 2011).
In short, deeper agreements have lower tariff bindings and make it more difficult for
treaty members to impose nontariff barriers, such as quantitative restrictions and regu-
lations that restrict trade.

Scope
Preferential trade agreements also vary in scope, the number of issue areas covered by
the agreement. All PTAs regulate the trade of goods, but not all cover agricultural goods
(Davis 2003). Nearly 73 percent of the PTAs in our sample address nontariff barriers on
agricultural goods, but these provisions tend to differ significantly from one agreement
340    The Design of Trade Agreements

to the next, reflecting members’ domestic political concerns. In addition, roughly half of
PTAs regulate the trade of services, such as accounting and telecommunications (World
Trade Report 2011).
These agreements can also address nontrade issues. Approximately half of PTAs have
intellectual property provisions (World Trade Report 2011), which typically require
members to recognize and protect copyrights, patents, and trademarks. Similarly, many
PTAs regulate competition policy (53 percent) by prohibiting monopolies and cartels
and protect foreign investment (42 percent) by, for example, requiring expropriating
countries to provide compensation to foreign investors. Trade agreements also increas-
ingly address noneconomic issues, such as human rights and environmental protection
(Hafner-Burton 2005; Steinberg 2002).3
On average, the scope of PTAs has increased over time. Whereas the average PTA in
the 1970s regulated nine trade and nontrade issues, the average PTA in the 2000s cov-
ers fifteen issues (World Trade Report 2011). There is some ambiguity in the empirical
literature about the difference (if any) between the depth and scope of a PTA (Horn,
Mavroidis, and Sapier 2010; Baccini, Dür and Elsig 2012). From a theoretical perspec-
tive, agreements with a broader scope place more constraints on state behavior and
therefore require deeper cooperation. We thus consider scope to be one specific form
of depth.

Membership
Trade agreements vary in the size of their membership. The majority of PTAs are bilat-
eral, but there are also many important multilateral trade agreements (Mansfield and
Milner 2012). These vary dramatically in size. While NAFTA has only three members,
the Global System of Trade Preferences among Developing Countries has forty-five.
Most multilateral PTAs are regional—they include members from the same geographi-
cal area (Baccini and Dür 2012). This is not particularly surprising, because countries
tend to trade most with their neighbors.
Some scholars argue that there is a fundamental trade-off between the depth of a
treaty and the size of its membership. When an agreement has more members, they
argue, it will demand less because the agreement must be acceptable to the state least
willing to cooperate. Agreements with more members may thus be shallower than
agreements with fewer members (Downs, Rocke, and Barsoom 1998; Gilligan and
Johns 2012). Agreements like the Association of South-East Asian Nations (ASEAN)
PTA—which has many members and is very shallow—support these arguments.
However, many other trading agreements, like the European Union’s single mar-
ket program—which also has many members and is relatively deep—do not. There is
thus mixed evidence about whether agreements with fewer members lead to deeper
commitments.
The relationship between membership and depth is likely to be influenced by the
design of treaty commitments. When a PTA requires all members to commit to a
 Leslie Johns and Lauren Peritz    341

common policy, then more members may lead to shallower commitments following the
logic stated above. However, when a PTA allows different members to commit to differ-
ent policies, then treaties with many members can generate deep cooperation, because
the states least willing to cooperate can accept shallow commitments without hindering
those states that want to make deeper commitments (Gilligan 2004).

Rigidity
There is also tremendous variation in the rigidity of trade agreements. Flexible agree-
ments sometimes allow states to violate their trading obligations without abrogating
the treaty, while more rigid agreements allow few (if any) opportunities for temporary
escape. For example, most PTAs contain safeguard rules that allow states to restrict trade
if an import surge harms a domestic industry. Preferential trade agreements also often
contain antidumping rules, which allow members to impose an additional tax—an anti-
dumping duty—if a domestic industry is harmed by an import that is sold below its nor-
mal value. Similarly, roughly 28 percent of PTAs allow states to impose a countervailing
duty if a subsidized import harms a domestic industry (Kucik 2012). All three of these
mechanisms reduce the rigidity of a PTA.
Even among those treaties that do contain flexibility mechanisms, treaty rules vary
dramatically in the criteria and procedures that states must use to restrict trade and
in the limits on the duration and magnitude of these restrictions. In the case of safe-
guards, some treaties require consultations between the importer and exporter before
safeguards can be used, whereas others allow states to take unilateral action. In addi-
tion, many (but not all) PTAs place strict constraints on the duration and magnitude
of safeguard measures (29  percent and 66  percent, respectively). For example, the
US-Australian PTA only allows safeguards for two years and limits the size of the safe-
guard on certain products. By contrast, the Pacific Island Countries Trade Agreement
allows four-year renewable safeguard measures and does not limit the size of these
safeguards.
Some PTAs—often called “fair trade” agreements—also give states the discretion to
restrict trade to accommodate competing values. For example, many European PTAs
allow trade restrictions that are “necessary to protect human, animal or plant life or
health.”4 Similarly, some PTAs promote labor protection through rules on minimum
wages, child labor, and occupational safety (Kim 2012). These provisions all give coun-
tries flexibility to balance trade liberalization with other social objectives.

Institutionalization
Finally, PTAs vary widely in their institutionalization. Some treaties simply articulate
goals for cooperation, while others create complex bureaucracies and dispute settlement
procedures (DSPs).
342    The Design of Trade Agreements

When PTAs have bureaucracies, members can delegate the authority to modify and
create rules (Johns and Pelc 2014). Bureaucracies can also monitor compliance with
the treaty. For example, the European Union (EU) is a highly institutionalized customs
union with a bureaucratic entity—the European Commission—that monitors policy
implementation. The commission publishes annual implementation reports and can file
lawsuits against a state that violates its treaty obligations.
Most PTAs also contain DSPs, which are rules for trade disputes among members.
These DSPs play a vital role in providing information about state behavior and coor-
dinating informal enforcement of treaty rules (Johns and Rosendorff 2009; Johns
2012).5 Few scholars have examined DSPs in a comparative context, so there is no stan-
dard method for measuring their institutionalization. Some DSPs are relatively infor-
mal and only require states to conduct good faith negotiations to resolve their trade
disputes. Others allow members to refer disputes to the International Chamber of
Commerce—which oversees international arbitration—or the International Court of
Justice.
Approximately 70 percent of PTAs currently in force contain formal dispute settle-
ment procedures (Kucik 2012). They usually allow an individual or panel to hear argu-
ments from affected parties and then issue some form of opinion. The DSPs used vary
in the selection of individuals or panels, as well as the legal status of the opinion. Some
DSPs only allow the individual or panel to make nonbinding recommendations; others
allow legally binding rulings. The latter sometimes have implementation procedures to
ensure the disputants adopt the decision in a reasonable period of time.
Most scholars believe that more formal DSPs make a treaty more rigid, but they dis-
agree about how the existence of DSPs affects rigidity. We believe that this confusion
stems from differences in how scholars conceptualize dispute settlement in the absence
of formal procedures. Some scholars believe that treaties without DSPs are sustained
using grim-trigger punishment, which leads to the total collapse of a treaty if any state
commits a single violation. For these scholars, the existence of DSPs reduces rigidity
because it provides opportunities for states to violate their substantive treaty obliga-
tions without abrogating the agreement (Rosendorff 2005; Rosendorff and Milner
2001). Dispute settlement procedures are therefore more “forgiving” than anarchy.
Other scholars assume (usually implicitly) that a treaty without DSPs continues to be in
force even if a state violates it. For these scholars, the existence of DSPs increases rigid-
ity because it allows for the punishment of treaty violations that would otherwise be
ignored (Goldstein et al. 2000; Smith 2000; Guzman 2002). They believe that DSPs are
therefore more “punishing” than anarchy.

Impact of Design on State Behavior

Because the primary objective of trade agreements is to promote international trade,

most trade scholars care inherently about the causal effect of PTAs: the extent to which
 Leslie Johns and Lauren Peritz    343

trade agreements actually increase international trade. As Martin (2013, 605) notes, “[a]‌
causal effect is stated as a counterfactual: how does state behavior in the presence of
an institution differ from the behavior that would have occurred in the absence of the
institution?” Because it is very difficult, if not impossible, to measure something that has
not happened—for example, how much a PTA member would have traded if it were not
a PTA member—we must rely on theoretical tools like game theory, which allows us to
compare state behavior both with and without a PTA to understand the impact of PTA
design on state behavior.
The design of a PTA determines the extent to which a treaty requires a state to change
its behavior. Ceteris paribus, a PTA will be more effective when it imposes deep obliga-
tions for a broad scope of issues on many members. However, rules can only be effective
if member states actually change their behavior. In the short term, PTAs will be more
effective when members comply with treaty rules. In the long term, PTAs will  be
more effective when they create stable regimes that endure over time. In this section we
provide a summary of previous research that examines the impact of treaty design on
compliance and stability.6

Domestic Political Pressure

The study of trade agreements cannot be divorced from domestic politics (Goldstein
and Martin 2000). While trade agreements can generate aggregate economic benefits
for a society, they have a distributional impact:  some domestic groups benefit from
trade liberalization while others suffer. This distributional impact means that trade
agreements have political implications. When governments negotiate treaties and
choose trade policies, they must balance the competing interests of three domestic
groups:  import-competing industries, exporters, and consumers. Import-competing
industries are harmed by trade liberalization because free trade increases market com-
petition. They therefore want their government to impose high tariffs on imported
goods. Exporters are not directly affected by their own government’s tariffs, but they
support trade liberalization if it leads to lower tariffs in other countries, because export-
ers want to compete in foreign markets. Finally, consumers care about the price of goods
and services in their country. Consumers are not directly affected by the tariffs of other
countries, but they benefit from the market competition that is created by free trade. Of
course the real world is considerably more complex than this simple account, but this
basic framework provides insight into the interests competing over trade policy.
The relative political power of these three groups will fluctuate over time. As dis-
cussed elsewhere in this volume by Rosendorff, economic factors, such as import
surges and changes in foreign production technology, can increase pressure to protect
an import-competing industry from foreign competition at the expense of consum-
ers and exporters. Similarly, political factors, such as a closely contested election, can
affect the willingness of a government to meet the demands of an import-competing
industry. Finally, random and unexpected events—such as floods, droughts, and
344    The Design of Trade Agreements

earthquakes—can also lead governments to temporarily protect domestic industries by

imposing trade barriers.
These changes in domestic political pressure from import-competing industries, rela-
tive to consumers and exporters, ensure that a leader’s incentive to cooperate fluctuates
over time. When states negotiate trade agreements, they know that domestic politi-
cal pressure from import-competing industries will fluctuate in the future. But states
cannot perfectly anticipate the amount of future domestic political pressure; they will
always be at least somewhat uncertain about the future difficulty of trade cooperation.

PTA Components
Trade agreements must therefore contain at least two elements. First, a treaty must cre-
ate primary rules that specify appropriate behavior. Some examples of primary rules are
tariff bindings, the maximum tariffs that are permissible under the treaty; the prohibi-
tion of quantitative restrictions; and government procurement rules. In the following
discussion we focus on tariff bindings, but our arguments apply to any primary rule that
constrains a government’s trade policy. The deeper the treaty, the more it constrains its
members. When treaty members follow these primary rules, there is first-order compli-
ance with the treaty.
Second, a treaty must create secondary rules about how members will respond when
a member violates its primary obligations. These secondary rules are usually created by
dispute settlement procedures.7 When secondary rules are relatively lax, the treaty is
flexible. As secondary rules grow stricter, the treaty becomes more rigid. In the interna-
tional system, states cannot be compelled to follow a treaty’s secondary rules. For exam-
ple, a state that violates a tariff binding can either abide by a regime’s secondary rules or
leave the treaty altogether. When treaty members follow these secondary rules, there is
second-order compliance with the treaty.
In the following discussion we conceptualize DSPs as imposing a penalty for treaty vio-
lations. Penalties can come in many forms, including trade retaliation, equivalent con-
cessions, technical assistance, reputational costs, and even the cost of negotiations and
litigation. The specific design of a DSP will affect a state’s expected cost of second-order
compliance. Rather than modeling specific DSPs, scholars have focused on the underly-
ing concept of rigidity; namely, more rigid treaties make it more costly for a state to vio-
late first-order rules and remain a member of the regime. We therefore model the rigidity
of a PTA by the size of the penalty that it imposes for first-order violations.

Equilibrium Behavior with and without a PTA

Absent a cooperative agreement, a leader’s optimal action in each period is to choose the
tariff that maximizes his or her own one-period utility. This “defection tariff ” increases
as the domestic political pressure to protect grows larger. Leaders have little incentive
 Leslie Johns and Lauren Peritz    345

to choose low tariffs, because these will harm their import-competing industries with-
out benefiting their exporters. Consumers prefer low tariffs that reduce the domestic
price of goods and services, but they usually exert less influence than import-competing
industries, which can lobby the government for protection.
When states write trade agreements, they are implicitly “trading” benefits for export-
ers. When state A lowers its trade barriers, it helps exporters in another state, B, and
harms its own import-competing industries. State A is only willing to do this if, in
exchange, state B lowers its trade barriers, thus benefiting A’s exporters. Cooperation
is therefore conditional: one country is only willing to make trade concessions if other
members do so as well. Members may want to allow treaty violations when a state faces
high pressure to protect, but a treaty must impose a penalty for a rule violation, or
members will dissemble about their true political pressure and the treaty will collapse
(Rosendorff and Milner 2001).8 So a treaty member must make two decisions: (1) Will
it comply with the treaty? and (2) If it does not comply, will it accept the penalty to settle
the dispute or leave the treaty regime?
Many theoretical models generate the equilibrium behavior shown in Figure 18.1.9
The horizontal axis represents domestic political pressure to protect, and the verti-
cal axis represents the tariff. The dotted horizontal line shows the tariff binding that is

Defection
tariff

Settlement
tariff
Tariff level

Tariff
binding

C S D
Domestic political pressure

Fig.  18.1  Equilibrium Tariffs under a Trade Agreement.

346    The Design of Trade Agreements

mandated by the trade agreement. The “defection tariff ” line shows the state’s optimal
tariffs when the PTA is not in effect. The “settlement tariff ” line shows the optimal
tariffs if a state violates a PTA but then accepts the penalty to settle the resulting dis-
pute and remain a member of the PTA in future periods. The “settlement tariff ” line
is therefore the “defection tariff ” line shifted downward by the penalty amount. The
bold line denotes the tariffs that a leader will choose in equilibrium under the trade
agreement.
Suppose that domestic political pressure is low. If a leader were to choose the optimal
tariff without regard to the country’s international obligations, then the tariff he or she
would choose (the “defection tariff ”) is lower than the treaty binding. In such situations,
a leader’s trade policy is unconstrained by the treaty, and he or she complies. However,
as domestic political pressure increases, the defection tariff breaks the binding. To com-
ply with the treaty, the leader will need to choose the tariff binding, as shown by the flat
portion of the bold line. The full compliance region is shown in the bottom portion of
the figure by region C.
As domestic political pressure grows even larger, leaders are better off violating the
binding and then settling under the treaty’s dispute settlement procedures. The opti-
mal tariff in these circumstances, the “settlement tariff,” rises with the level of domestic
political pressure. However, the settlement tariff is always less than the defection tariff,
because the treaty’s violation penalty tempers the magnitude of treaty violations. The
more rigid a treaty is, the larger the violation penalty. The settlement interval is indi-
cated by region S of Figure 18.1.
Finally, if domestic political pressure grows very large, leaders will no longer be will-
ing to settle trade disputes to remain within the treaty regime. It is instead optimal to
defect by reverting to the defection tariff and leaving the trade regime altogether. This
occurs in region D of the figure.
To understand how depth and rigidity affect the likelihood of full compliance and
stability, we must investigate how the design of the treaty affects regions C and D. As
region C expands, the likelihood of full compliance increases, because leaders are more
likely to choose tariffs at or below their binding. As region D expands, stability decreases
because a treaty member is more likely to defect by violating the treaty terms and exiting
the regime.

Impact of Depth on State Behavior

When a trade agreement grows deeper, it imposes stricter constraints on its mem-
bers. For example, a deep tariff binding requires lower tariffs than a shallow binding.
As a treaty grows deeper, it thus becomes more difficult for members to comply and,
not surprisingly, the likelihood of compliance decreases. This is the primary reason
we should not confound the concepts of compliance and effectiveness:  if a treaty
places few constraints on states, it is unlikely to change behavior, even if we observe
 Leslie Johns and Lauren Peritz    347

high compliance rates (Downs, Rocke, and Barsoom 1996). In addition, by requir-
ing more from its members a deep treaty makes it more difficult for a state to tem-
porarily violate its obligations and then settle the resulting dispute. If a state breaks
its binding, the magnitude of any violation—the difference between the chosen tar-
iff and the binding—increases as a treaty grows deeper. This in turn makes it more
costly for a state to settle its dispute within the DSP. Accordingly, deeper treaties are
inherently less stable than shallower treaties, because states are more likely to exit
deep treaties.
Figure 18.2 shows the effect of lowering the tariff binding from a shallow to a deep
binding. This makes the agreement more demanding on the member states, because the
treaty requires deeper cooperation. The solid line denotes equilibrium tariffs under the
shallow binding, and the dashed line shows equilibrium tariffs under the deep binding.
As the bottom portion of Figure 18.2 indicates, when the tariff binding grows deeper, the
region of full compliance shrinks from C to C and the region of instability (defection)
grows from D to D. A treaty regime is stable if there is either full compliance or settle-
ment. Trade agreements that demand deeper cooperation thus lead to less compliance
(region C) and less stability (regions C and S combined).

Defection
tariff

Settlement
tariff
Tariff level

Shallow tariff
binding

Deep tariff
binding

Shallow
C S D
Deep
C S D
Domestic political pressure

Fig.  18.2  Deep Agreements Reduce Compliance and Stability.

348    The Design of Trade Agreements

Impact of Rigidity on State Behavior

When a treaty is flexible, it is relatively permissive of occasional defections and imposes
only small penalties on treaty violators. As a treaty grows more rigid, it becomes less per-
missive of these violations and imposes larger penalties on treaty violators. Increasing
the rigidity of a treaty thus makes it more costly for a state to violate its binding and
then settle the resulting dispute. When a leader faces relatively low domestic political
pressure to protect, increasing rigidity makes full compliance more desirable relative
to settlement. However, when a leader faces high domestic political pressure, rigidity
makes defection more desirable relative to settlement. Rigidity—which makes it more
difficult for a state to temporarily violate and then settle—thus increases the probability
of full compliance and decreases the stability of the regime.
Figure 18.3 shows the impact of moving from a flexible to a rigid design. The solid line
denotes equilibrium tariffs under a flexible agreement, and the dashed line shows equi-
librium tariffs under a rigid agreement. The first thing to note is that states choose lower
settlement tariffs under the rigid regime: if a state is going to be more severely penalized
for a violation, then it will violate less by choosing a lower tariff. For the settlement (S)
region of the figure, states do not fully comply with either treaty, but the magnitude of

Defection
tariff

Flexible
settlement tariff

Rigid
settlement tariff
Tariff level

Tariff
binding

Flexible
C S D

Rigid
C S D
Domestic political pressure

Fig.  18.3  Rigid Agreements Increase Compliance but Decrease Stability.

 Leslie Johns and Lauren Peritz    349

violations—how much the chosen tariff varies from the tariff binding—is larger under
the more flexible treaty. In addition, as rigidity increases, the zones for both full compli-
ance and instability increase (from C to C and D to D, respectively), which demonstrates
that rigidity increases the likelihood of full compliance (region C), but decreases stabil-
ity (regions C and S combined).
The specific design of a trade agreement changes the way its members behave. The
overall effectiveness of a treaty in reducing trade barriers is thus determined by the trea-
ty’s rules and how states respond to changes in domestic political pressure. This frame-
work does not provide clear predictions about which specific designs will be optimal in
promoting trade liberalization. However, it does suggest that the design of a trade agree-
ment should be conditioned, at least in part, on the cost of leaving the treaty regime
(Johns 2015).
When a state leaves a treaty, it can no longer expect to receive the benefits of coopera-
tion. Yet some trade agreements are nested in political environments that increase the
cost of exit even further. If treaty membership is linked to a multilateral organization,
such as the WTO or the EU, exit should be more costly. In such situations treaty design-
ers can worry less about the impact of the treaty design on stability and write trade
agreements that are both deeper and more rigid. However, when trade agreements are
not nested in such a political context, stability is a greater concern, and treaty designers
should thus be more likely to write shallow and flexible treaties.

Networks and Complexity

In this section we move to a system-level analysis of trade agreements. As the number of

PTAs has increased, an overlapping and complex network has emerged. We examine the
interaction among PTAs and ask how this complex network of agreements affects trade
liberalization.

PTA Proliferation and Expansion

Preferential trade agreements have proliferated over time. As shown in Figure 18.4, the
first PTAs began to appear in the 1950s, and the number of PTAs in force grew at a steady
pace until around 1990. After 1990 the number of PTAs exploded. Figure 18.4 also shows
that the number of countries participating in at least one PTA has increased over time.10
European countries have contributed significantly to the proliferation of PTAs, but as of
this writing, almost every country is a member of at least one PTA.
Several factors drive the global proliferation of PTAs. These agreements reduce
trade barriers and thereby lower the price of imports from member countries, which
benefits both exporters and consumers at the expense of import-competing indus-
tries. However, PTAs can generate negative externalities for exporters in nonmember
350    The Design of Trade Agreements

250

Number of PTAs in force

Number of countries in (at least one) PTA
200 200

Countries participating
160
150
PTAs in force

120
100

80

50
40

0 0
1960 1970 1980 1990 2000
Year

Fig.  18.4  Proliferation of PTAs over  Time.

Source: Data from Kucik (2012) and the World Trade Organization
Regional Trade Agreements Database.

states (Chang and Winters 2002). All else being equal, a PTA lowers the profit of
nonmember exporters because they face higher tariffs than their competitors from
member states. Preferential trade agreements with deeper commitments or broader
scope magnify this effect. Economists have shown that such trade diversion—to
members and away from nonmembers—encourages PTA membership (Grossman
and Helpman 1995; Krishna 1996). Moreover, the tariffs applied by PTA members
against nonmembers may be even higher than they were in the absence of the PTA
(Panagariya and Findlay 1996). By granting benefits to members and generating
losses for nonmembers, countries have an incentive to either create new PTAs or join
existing ones.
Some scholars argue that when the membership of a PTA expands, incentives to
join intensify because larger PTAs magnify the trade diversion effect (Hoekman and
Kostecki 2009, 499). Remaining outside the regime becomes more costly. So each
time a country joins a PTA, the pressure on other countries to join increases, creat-
ing a “domino effect” (Baldwin 1995). The more countries are involved in a PTA and
the more effective that PTA is at generating in-group advantages, the more appeal-
ing it is for nonmembers to accede (Baldwin 2006). The EU exemplifies this “dom-
ino effect” of PTA expansion. Originally a trade agreement among six countries, the
European Economic Community went through successive enlargements. By 2007
there were twenty-seven members and several other countries undergoing the acces-
sion process.
 Leslie Johns and Lauren Peritz    351

Number of PTAs each country has signed

1 38

Fig.  18.5  Density of PTA Membership.

Source: Data from Kucik (2012); image created with R package “rworldmaps.”
We only show PTAs in force and notified to the  WTO.

As PTAs proliferate and expand, nonmembers also encounter stronger incentives

to form their own trade blocs. Some scholars argue that PTA members might limit
accession of new members, prompting outsiders to create separate trade agreements
(Panagariya 2000). By cooperating with one another, these other countries strengthen
their multilateral bargaining position. For example, a primary objective of Mercado
Común del Sur (MERCOSUR) was to improve the bargaining power of its members
vis-à-vis NAFTA and the EU (Whalley 1998, 72; Bevilaqua, Catena and Talvi 2001, 153).
Its members enjoy more bargaining leverage as a group than they would as individuals.
Some empirical studies show that MERCOSUR is not a unique case; countries’ decisions
to form PTAs are highly interdependent (Egger and Larch 2008; Baccini and Dür 2012).
Thus membership limits, an important design choice, can affect PTA proliferation.

Network Complexity
As PTAs have proliferated and expanded, a complex network has emerged. While some
countries enter only one or two trade agreements, others join many. Figure 18.5 shows
global PTA density by country. Darker colors indicate that a country has signed more
352    The Design of Trade Agreements

PTAs. The most active users of PTAs to date are the European countries, Singapore,
India, and Chile.
Much of the recent growth in PTAs is through “hub and spoke” agreements, in which
an established trade regime (the “hub”) creates separate agreements with other coun-
tries outside the regime (the “spokes”).11 These agreements usually vary in the benefits
and obligations created for the “spoke” members. For example, the EU functions as a
“hub” of members that have committed to a common market. Over time the EU has
signed many preferential trading agreements with non-EU countries, including Chile,
Korea, and Mexico. These “spoke” arrangements vary widely in their depth, scope, and
rigidity, but none impose the same terms as the “hub” agreement. In addition, PTAs
sometimes bind together two existing “hub” regimes, like the 2008 agreement between
the EU and CARIFORUM, a group of Caribbean countries that have committed to
regional integration.
Because each country can be a member of multiple agreements, and some of these
agreements are linked to other trade regimes, countries usually have complex and over-
lapping trade commitments. Figure 18.6 shows the PTA connections for four major
(“focal”) PTA signatories:  the European Union, India, Japan, and the United States.
Each of these four members has numerous trade agreements with other countries. For
example, the United States has PTAs with Australia, Chile, Israel, and many others. In
addition, many nonfocal countries have trade agreements with multiple focal countries.
For example, Chile has PTAs with all four of the focal members shown in Figure 18.6.
Individual countries often have overlapping treaties with multiple trade partners, and
PTAs as a whole create a complex network of trading obligations.

European
Union

India

USA
Japan

Fig.  18.6  Overlapping PTA Networks.

Note: Each panel shows the PTAs between the focal countries (labeled)
and their trade partners (black  dots).
Source: World Trade Organization Regional Trade Agreements Database.
 Leslie Johns and Lauren Peritz    353

Perils of Complex Networks

While PTAs individually should increase trade cooperation, a network of overlapping
PTAs can hinder trade cooperation by creating conflicting legal obligations. These con-
flicts can be most easily seen in the WTO, which has heard several cases over conflict-
ing WTO and PTA rules.12 For example, all WTO members are required to adhere to
the most-favored nation (MFN) principle, which requires that if a member extends a
benefit to another WTO member, it must extend the same benefit to all other WTO
members. Yet Article XXIV of the GATT allows members to sign PTAs, which implies
that benefits provided to PTA members do not have to be extended to all WTO mem-
bers. There is thus a tension between a core tenet of the multilateral trade regime—the
MFN principle—and the proliferation of PTAs. One WTO panel report noted that this
relationship between the MFN principle and Article XXIV “has not always been harmo-
nious” and has led to numerous trade disputes.13
For example, Turkey has signed multiple trade agreements with the European
Communities (EC) since 1963 that gradually developed into a Turkey-EC customs
union.14 As part of this process, Turkey negotiated new agreements with its trading part-
ners in the early 1990s so that its textile rules would match those of the EC. India, a major
textile exporter, refused to participate in these negotiations, and Turkey imposed unilat-
eral restrictions on textile imports from India in 1996. India quickly filed a WTO dispute
against Turkey, arguing that Turkey had violated multiple WTO rules.15 Turkey argued
that it should be exempt from these rules under Article XXIV because it was changing
its laws to form a Turkey-EC customs union. Both the WTO panel and Appellate Body
ruled against Turkey, stating that “Article XXIV does not allow Turkey to adopt, upon
the formation of a customs union with the European Communities, quantitative restric-
tions … which were found to be inconsistent with” WTO rules.16 This created a legal
quandary: Turkey violated WTO rules by changing its policies to match those of the
EC, but Turkey would have violated its agreement with the EC if it had not changed its
policies.
This example illustrates one of the perils of using PTAs to promote international
trade: they create complex and often contradictory networks of legal obligations. The
multilateral trade regime—first under the GATT, then under the WTO—has cre-
ated detailed legal obligations for its members, which now include almost every state.
The growth of PTAs has created complex sets of rules that overlap with the multilat-
eral regime and even with other PTAs. Overlapping rules can create uncertainty about
legal commitments and prompt trade disputes, which hinder international coopera-
tion (Gilligan, Johns, and Rosendorff 2010). This complexity and uncertainty may be
an inevitable, but acceptable, cost if PTAs promote aggregate international coopera-
tion. However, if PTAs merely divert trade, rather than increase it, the cost of PTAs may
outweigh their benefits. A  network of overlapping bilateral and regional agreements
may be less beneficial than shallower multilateral cooperation or even no agreements
whatsoever.
354    The Design of Trade Agreements

Complexity may be an inevitable result of the growing PTA network, but the ill effects
of conflicting and uncertain legal obligations can be mitigated through the design of
PTA DSPs. While DSPs can clarify ambiguous or conflicting legal commitments, the
proliferation of PTA-specific DSPs could be problematic if they adopt conflicting juris-
prudence. Some recent PTAs allow members to choose which DSP they will use when a
conflict arises, effectively “outsourcing” dispute settlement to another institution, most
commonly the WTO.17 We believe that this trend is promising because it increases the
likelihood that trade disputes will be adjudicated under common legal principles and
jurisprudence. If more PTAs were designed to share dispute settlement capacity, then
conflicts over mismatched commitments might be more easily resolved. It may even be
optimal for states to create a single dispute settlement mechanism with jurisdiction to
hear all trade disputes involving the WTO and PTAs.
In addition, we suspect that more specialized PTAs could avert problems that arise
from overlapping commitments by restricting their obligations to a narrow domain. For
example, new PTAs could be designed with narrower scopes that focus more precisely
on particular trade issues like agriculture, intellectual property, and technical barriers.
Moreover, specialized PTAs could be designed to account for issue-specific cooperation
problems, such as lenient safeguard measures for agriculture and greater institutional-
ization for technical issues like intellectual property. We find little evidence, however, of
any such specialization; most recently signed PTAs have tended to be broader in scope
than their earlier counterparts.

Steppingstones or Stumbling Blocks to Multilateralism?

As the network of PTAs becomes increasingly tangled, many scholars question whether
these agreements promote or hinder multilateral trade cooperation.18 We unfortunately
lack a clear answer. Preferential trade agreements may foster multilateralism if their ini-
tial reciprocal tariff cuts allow export sectors to expand and thereby change the balance
of political power between import-competing industries and exporters (Baldwin and
Freund 2011). In addition, many policy experts believe that successful PTAs may per-
suade nonmembers to support deeper multilateral cooperation by demonstrating the
benefits of international trade.19
On the other hand, PTAs may hinder multilateralism. Members can leverage their
market power and increase their tariffs against the rest of the world, which makes mul-
tilateral cooperation less attractive (Saggi and Yildiz 2010). Some scholars argue that
the economic context of a PTA determines whether it becomes a steppingstone or
stumbling block for multilateral cooperation.20 However, this scholarship reaches no
clear conclusions, because its results are highly contingent on assumptions about the
structure of the member economies. The impact of PTAs on multilateral cooperation is
important but understudied. We hope that future scholarship, particularly in political
science, will address this topic.
 Leslie Johns and Lauren Peritz    355

Conclusion

Existing research has provided many insights into PTA design variation and its impact
on state behavior. This treaty-level analysis shows that PTAs must carefully balance their
impact on compliance and stability. Agreements that are nested in political-economic
contexts that make exit infeasible can impose deep and rigid constraints and create
institutions with the authority to monitor behavior and enforce obligations. However,
agreements that are nested in contexts that do not constrain exit must be more care-
fully calibrated to the trade-off between compliance and stability. Deeper concessions
must be paired with less rigidity in order for these agreements to survive over time, and
weaker institutions are best suited for promoting cooperation.
We know relatively less about the effects of PTAs at the system level. The prolifera-
tion and expansion of PTAs has created an increasingly complex network of overlap-
ping legal obligations. These agreements must also operate in the shadow of the WTO.
While it is reasonable to expect that more PTAs will lead to more cooperation, com-
plexity comes with costs. Overlapping PTAs can create ambiguous and even contradic-
tory legal obligations that provoke trade disputes. These disputes reduce the efficiency of
the system as a whole and reduce stability. This suggests that PTAs with more members
can have a multiplier effect on cooperation because they create more consistent—and
hence more precise—rules than can be created by a network of overlapping bilateral and
regional agreements. Multilateralism is more than just the sum of its parts.

Notes
1. This is the most comprehensive current list of PTAs available, since almost all countries
are members or observers of the WTO.
2. Most PTAs in the World Trade Report 2011 data set involve the United States or European
Union as a member and have a large volume of trade.
3. See also the chapters in this volume by Aaronson and Barkin.
4. Some of these PTAs include: EC-Mexico 2000, EC-CARIFORUM 2008, and EC-Eastern
and Southern Africa 2012 Interim Agreement.
5. See also the chapters in this volume by Rosendorff (­chapter  8) and Busch and Pelc
(­chapter 5).
6. These arguments come from Johns (2014), which builds on Rosendorff (2005) and
Rosendorff and Milner (2001).
7. Like any institution, a trade agreement must create common beliefs about how members
should respond to first-order noncompliance, even if the treaty does not contain formal
DSPs (North 1990). The law of treaties, as articulated in the Vienna Convention on the
Law of Treaties, creates secondary rules for those trade agreements that lack formal DSPs.
8. Pelc (2009) provides an alternative account in which states can violate a treaty without any
penalty whatsoever. However, this account relies on the assumption that a treaty creates an
adjudicative body that can perfectly observe each state’s domestic pressure to protect and
356    The Design of Trade Agreements

impose punishments on states that cheat. This account is therefore more compelling for
understanding a well-developed institution, like the WTO, than PTAs generally.
9. See Johns (2014, 2015). Rosendorff (2005) and Rosendorff and Milner (2001) present simi-
lar models, but do not allow settlement tariffs.
10. The occasional decrease in PTAs is due to EU enlargements in 2004 and 2007.
11. This term refers to any arrangement in which one partner has “a network of radial bilateral
PTAs with some of these trading partners but these trading partners do not have PTAs
with each other” (Baldwin and Freund 2011, 129).
12. For example, Argentina—Poultry (DS241) and Mexico—Taxes on Soft Drinks (DS308)
both addressed jurisdictional conflicts for dispute settlement. In other cases, countries
invoking safeguards excluded their PTA partners, contrary to WTO obligations: for exam-
ple, Argentina—Footwear (DS121), United States—Wheat Gluten (DS166), and United
States—Line Pipe (DS202).
13. See the panel report for Turkey—Restrictions on Imports of Textile and Clothing Products,
DS34, para. 2.3. WTO Document WT/DS34/R.
14. This case summary is based on WTO records.
15. India invoked claims from both the GATT and the WTO Agreement on Textiles and
Clothing.
16. See the Appellate Body report for Turkey—Restrictions on Imports of Textile and Clothing
Products, DS34, para. 64. WTO Document WT/DS34/AB/R.
17. Those PTAs that allow members to use the WTO’s dispute settlement body to resolve
conflicts include NAFTA, the Australia-Chile PTA, the Israel-Mexico PTA, and the
CARICOM-Costa Rica PTA. See also Busch (2007) on forum shopping in dispute
settlement.
18. For example, see Bagwell and Staiger (1998); Baldwin and Freund (2011); Bhagwati (1991);
Bhagwati and Panagariya (1999); Krueger (1999); Kono (2002); Limão (2006); and Limão
(2007).
19. Remarks by Robert Zoellick, former US Trade Representative, at Princeton University on
November 29, 2012.
20. For example, see Bagwell and Staiger (2001); Freund (2000); Goyal and Joshi (2006); and
Saggi (2006).

References
Baccini, Leonardo, and Andreas Dür. 2012. The New Regionalism and Policy Interdependence.
British Journal of Political Science 42: 57–79.
Baccini, Leonardo, Andreas Dür, and Mafred Elsig. 2012. The Politics of Trade Agreement
Design: Depth, Scope and Flexibility. Paper presented at the Annual Conference on the
Political Economy of International Organizations, Philadelphia, PA, January 26–28.
Bagwell, Kyle, and Robert W. Staiger. 2001. Reciprocity, Non-discrimination and Preferential
Agreements in the Multilateral Trading System. European Journal of Political Economy
17: 281–325.
Bagwell, Kyle, and Robert W. Staiger. 1998. Will Preferential Agreements Undermine the
Multilateral Trading System? Economic Journal 108: 1162–1182.
 Leslie Johns and Lauren Peritz    357

Baldwin, Richard. 1995. A Domino Theory of Regionalism. In Expanding Membership of the

European Union, edited by R. Baldwin, P. Haaparanta, and J. Kiander, 25–53. Cambridge,
UK: Cambridge University Press.
Baldwin, Richard. 2006. Multilateralising Regionalism: Spaghetti Bowls as Building Blocs on
the Path to Global Free Trade. World Economy 29: 1451–1581.
Baldwin, Richard, and Caroline Freund. 2011. Preferential Trade Agreements and Multilateral
Liberalization. In Preferential Trade Agreement Policies for Development: A Handbook, edited by
Jean-Pierre Chauffour and Jean-Christophe Mau, 121–141. Washington, DC: The World Bank.
Barton, John H., Judith L. Goldstein, Timothy E. Josling, and Richard H. Steinberg. 2008.
The Evolution of the Trade Regime: Politics, Law, and Economics of the GATT and the WTO.
Princeton, NJ: Princeton University Press.
Bevilaqua, Alfonso S., Marcelo Catena, and Ernesto Talvi. 2001. Integration, Interdependence
and Regional Goods: An Application to Mercosur. In Economia, edited by Andres Velasco,
153–207. Washington, DC: Brookings Institution Press.
Bhagwati, Jagdish. 1993. Regionalism and Multilateralism: An Overview. In New Dimensions
in Regional Integration, edited by Jaime de Melo and Arvind Panagariya, 22–51. Cambridge,
UK: Centre for Economic Policy Research.
Bhagwati, Jagdish. 1991. The World Trading System at Risk. Princeton, NJ:  Princeton
University Press.
Bhagwati, Jagdish, and Arvind Panagariya. 1999. Preferential Trading Areas and
Multilateralism—Strangers, Friends, or Foes? In Trading Blocs: Alternative Approaches to
Analyzing Preferential Trade Agreements, edited by Jagdish Bhagwati, Arvind Panagariya,
and Pravin Krishna, 33–100. Cambridge, MA: MIT Press.
Busch, Marc L. 2007. Overlapping Institutions, Forum Shopping, and Dispute Settlement in
International Trade. International Organization 61: 735–761.
Chang, W., and L. A. Winters. 2002. How Regional Blocs Affect Excluded Countries: The Price
Effects of MERCOSUR. American Economic Review 92: 889–904.
Davis, Christina L. 2003. Food Fights Over Free Trade: How International Institutions Promote
Agricultural Trade Liberalization. Princeton, NJ: Princeton University Press.
Downs, George W., David M. Rocke, and Peter N. Barsoom. 1996. Is the Good News about
Compliance Good News about Cooperation? International Organization 50: 379–406.
Downs, George W., David M. Rocke, and Peter N. Barsoom. 1998. Managing the Evolution of
Multilateralism. International Organization 52: 397–419.
Egger, Peter, and Mario Larch. 2008. Interdependent Preferential Trade Agreement
Memberships: An Empirical Analysis. Journal of International Economics 76: 384–399.
Freund, Caroline. 2000. Different Paths to Free Trade: The Gains from Regionalism. Quarterly
Journal of Economics 115: 1317–1341.
Gilligan, Michael J. 2004. Is There a Broader-Deeper Trade-off in International Multilateral
Agreements? International Organization 58: 459–484.
Gilligan, Michael J., and Leslie Johns. 2012. Formal Models of International Institutions.
Annual Review of Political Science 15: 221–243.
Gilligan, Michael J., Leslie Johns, and B. Peter Rosendorff. 2010. “Strengthening International
Courts and the Early Settlement of Disputes.” Journal of Conflict Resolution 54: 5–38.
Goldstein, Judith, Miles Kahler, Robert Keohane, and Anne-Marie Slaughter. 2000.
Introduction: Legalization and World Politics. International Organization 54: 385–399.
358    The Design of Trade Agreements

Goldstein, Judith, and Lisa L. Martin. 2000. Legalization, Trade Liberalization, and Domestic
Politics: A Cautionary Note. International Organization 54: 603–632.
Goyal, Sanjeev, and Sumit Joshi. 2006. Bilateralism and Free Trade. International Economic
Review 47: 749–778.
Grossman, Gene, and Elhanan Helpman. 1995. The Politics of Free Trade Agreements.
American Economic Review 85: 667–690.
Guzman, Andrew T. 2002. The Cost of Credibility: Explaining Resistance to Interstate Dispute
Resolution Mechanisms. Journal of Legal Studies 31: 303–326.
Hafner-Burton, Emilie M. 2005. Trading Human Rights: How Preferential Trade Agreements
Influence Government Repression. International Organization 59: 593–629.
Hicks, Raymond, and Soo Yeon Kim. 2012. Reciprocal Trade Agreements in Asia: Credible
Commitment to Trade Liberalization or Paper Tigers? Journal of East Asian Studies 12: 1–29.
Hoekman, Bernard M., and Michel M. Kostecki. 2009. The Political Economy of the World
Trading System. New York: Oxford University Press.
Horn, Henrik, Petros C. Mavroidis, and André Sapir. 2010. Beyond the WTO? An Anatomy of
EU and US Preferential Trade Agreements. The World Economy 33: 1565–1588.
Johns, Leslie. 2012. Courts as Coordinators:  Endogenous Enforcement and Jurisdiction in
International Adjudication. Journal of Conflict Resolution 56: 257–289.
Johns, Leslie. 2014. Depth versus Rigidity in International Trade Agreements. Journal of
Theoretical Politics 26: 468–295.
Johns, Leslie. 2015. Strengthening International Courts: The Hidden Costs of Legal­ization. Ann
Arbor: University of Michigan Press.
Johns, Leslie, and Krzysztof J. Pelc. 2014. Who Gets to Be in the Room? Manipulating
Participation in WTO Disputes. International Organization 68: 663–699.
Johns, Leslie, and B. Peter Rosendorff. 2009. Dispute Settlement, Compliance, and Domestic
Politics. In Trade Disputes and the Dispute Settlement Understanding of the WTO:  An
Interdisciplinary Assessment, edited by James C. Hartigan, 139–163. San Diego, CA: Emerald
Group Publishing.
Kim, Moonhawk. 2012. Ex Ante Due Diligence: Formation of PTAs and Protection of Labor
Rights. International Studies Quarterly 56: 704–719.
Kono, Daniel Yuichi. 2002. Are Free Trade Areas Good for Multilateralism? Evidence from the
European Free Trade Association. International Studies Quarterly 46: 507–527.
Krishna, Pravin. 1996. A Political Economy Analysis of Preferential Trading and Multilater­
alism. Eastern Economic Journal 22: 477–483.
Krueger, Anne O. 1999. Are Preferential Trading Arrangements Trade-Liberalizing or
Protectionist? Journal of Economic Perspectives 13: 105–124.
Kucik, Jeffrey. 2012. The Domestic Politics of Trade Agreement Design: Producer Preferences
over Trade Agreement Rules. Economics & Politics 24: 95–118.
Kucik, Jeffrey, and Eric Reinhardt. 2008. Does Flexibility Promote Cooperation? An
Application to the Global Trade Regime. International Organization 62: 477–505.
Limão, Nuno. 2006. “Preferential Trade Agreements as Stumbling Blocks for Multilateral
Trade Liberalization: Evidence for the U.S.” American Economic Review 96: 896–914.
Limão, Nuno. 2007. Are Preferential Trade Agreements with Non-trade Objectives a Stumbling
Block for Multilateral Liberalization? Review of Economic Studies 74: 821–855.
Mansfield, Edward D., and Helen V. Milner. 2012. Votes, Vetoes, and the Political Economy of
International Trade Agreements. Princeton, NJ: Princeton University Press.
 Leslie Johns and Lauren Peritz    359

Martin, Lisa L. 2013. Against Compliance.” In Interdisciplinary Perspectives on International

Law and International Relations: The State of the Art, edited by Jeffrey L. Dunoff and Mark A.
Pollack, 591–610. New York: Cambridge University Press.
North, Douglass C. 1990. Institutions, Institutional Change and Economic Performance.
Cambridge, UK: Cambridge University Press.
Panagariya, Arvind. 2000. Preferential Trade Liberalization: The Traditional Theory and New
Developments. Journal of Economic Literature 38: 287–331.
Panagariya, Arvind, and Ronald Findlay. 1996. “A Political Economy Analysis of Free Trade
Areas and Customs Unions.” In The Political Economy of Trade Policy: Papers in Honor of
Jagdish Bhagwati, edited by Robert Feenstra, Gene Grossman, and Douglas Irwin, 265–288.
Cambridge, MA: MIT Press.
Pelc, Krzysztof. 2009. Seeking Escape:  Escape Clauses in International Trade Agreements.
International Studies Quarterly 53: 349–368.
Regional Trade Agreements Database. Geneva: The World Trade Organization. http://rtais.wto.
org/UI/PublicSearchByCr.aspx.
Rosendorff, B. Peter. 2005. Stability and Rigidity: Politics and the Design of the WTO’s Dispute
Resolution Procedure. American Political Science Review 99: 389–400.
Rosendorff, B. Peter, and Helen V. Milner. 2001. The Optimal Design of International
Institutions: Uncertainty and Escape. International Organization 55: 829–857.
Saggi, Kamal. 2006. Preferential Trade Agreements and Multilateral Tariff Cooperation.
International Economic Review 47: 29–57.
Saggi, Kamal, and Halis Murat Yildiz. 2010. Bilateralism, Multilateralism and the Quest for
Global Free Trade. Journal of International Economics 81: 26–37.
Smith, James McCall. 2000. The Politics of Dispute Settlement Design: Explaining Legalism in
Regional Trade Pacts. International Organization 54: 137–180.
Steinberg, Richard H. 2002. Understanding Trade and the Environment:  A  Conceptual
Framework. In The Greening of Trade Law: International Trade Organizations and Environ­
mental Issues, edited by Richard H. Steinberg, 1–20. New York: Rowman & Littlefield.
Uchitelle, Louis. 2009. “Buy America” in Stimulus (but Good Luck with That). New York Times,
February 20. http://www.nytimes.com/2009/02/21/business/21buy.html (accessed May
31, 2013).
Whalley, John. 1998. Why Do Countries Seek Regional Trade Agreements? In The
Regionalization of the World Economy, edited by Jeffrey A. Frankel, 63–90. Chicago:
University of Chicago Press.
World Trade Report 2011—The WTO and Preferential Trade Agreements: From Coexistence to
Coherence. 2011. Geneva: World Trade Organization.
 Chapter 19

Deep Integ rat i on

and Regiona l T ra de
Agreem e nts

S o o Y e on  K im

Deep integration provisions in regional trade agreements (RTAs) seek to strengthen

the contestability of markets for firms in partner economies. Such provisions have
three main functions: protection of foreign firms and their interests; liberalization of
“beyond-the-border” barriers to trade (i.e., domestic trade-related laws and regula-
tions that go beyond traditional “barriers-at-the-border,” such as tariffs and quotas);
and harmonization of domestic trade rules to enhance the efficiency of international
production. The road to deep integration has been paved by the demands of multina-
tional firms, their production networks, and the increasing complexity of the interna-
tional supply chain (Hatch, Bair, and Heiduk, this volume). Trade governance, which
first focused under the General Agreement on Tariffs and Trade (GATT) on reducing
border measures such as tariffs, quotas, and other quantitative restrictions, has moved
toward constraining domestic policies such as technical regulations. The role of RTAs
in fostering deep integration is also shaped by the failure of the Doha Round to reach
a successful conclusion and the emergence of RTAs as the main institutional venue for
forging new trade rules.
Deep integration is the hallmark of the modern RTA. Deep integration RTAs
emerged in the late twentieth century as international production and trade became
more complex, driven by extensive foreign direct investment and the internationaliza-
tion of production. Prominent examples of deep integration RTAs are the US-Mexico
part of the North American Free Trade Agreement (NAFTA); the European Union’s
RTAs; and more recently, Japan’s push for economic partnership agreements (EPAs)
that seek to incorporate an extensive array of issues, as well as the US-led Trans-Pacific
Partnership (TPP) agreement currently under negotiation. Deep integration RTAs are
not exclusive to the developed North or only between North and South countries. Many
 Soo Yeon Kim    361

South-South RTAs have also begun to feature deep integration commitments. The most
advanced deep integration RTA is the customs union, in which member countries not
only remove trade barriers but also adopt a common external tariff (CET) and pool
their economic sovereignty in economic policy. The exemplar is the European Union
(EU), which created a single market that not only provided for the free flow of the
factors of production but also established piecemeal the institutional convergence of
domestic policies.
This chapter examines the relationship between deep integration and RTAs and
how commitments for achieving greater compatibilities in national regulatory sys-
tems are manifested in the provisions of trade agreements. The objective is to delin-
eate the main lines of research that has been conducted in this area, emphasizing the
questions that have been raised in the literature and the extensive data collection
efforts that have attended scholarly interest in deep integration and in RTAs. The
chapter is divided into four main parts. The first section discusses the properties, both
conceptual and practical, of deep integration and their legal counterparts in RTAs.
The second and third sections examine, respectively, why states make deep integra-
tion commitments in RTAs and the effects of such deep integration commitments,
focusing on the prospects for convergence and competition of deep integration mod-
els at the systemic level. The chapter concludes with a discussion of avenues for fur-
ther research.

What is Deep Integration, and

How is it Related to Regional
Trade Agreements?

Deep integration refers to a process of economic integration that erodes differences

in national economic policies and regulations and renders them more compat-
ible for economic exchange. As a concept, deep integration is widely associated with
Robert Z. Lawrence (1996), who defined it essentially as “behind-the-border” inte-
gration in a study that examined the tensions between RTAs and national autonomy.
Deep integration is also synonymous with “positive integration,” which is associated
with the liberalization approach of the World Trade Organization (WTO) era. Positive
integration involves the active establishment of domestic rules and regulations that
render a country’s trade regime more consistent with its obligations under the “sin-
gle undertaking” that is the WTO. Positive integration, or deep integration, can also
be contrasted with the “negative integration” or “shallow integration” approach of
the GATT era. Under shallow integration, member states had the option to enroll in
GATT agreements in an à la carte fashion. For example, only some countries chose
to sign onto plurilateral agreements such as the Tokyo Round’s Standards Code
362    Deep Integration and Regional Trade Agreements

governing technical regulations, which precluded the use of the GATT’s dispute set-
tlement process by signatories against nonsignatories. More broadly, states’ commit-
ments during the GATT era extended only to the agreement not to raise “barriers at
the border,” such as raising their tariffs above the bound level, imposing quantitative
restrictions, or implementing indirect taxes to protect domestic industries and dis-
criminate against foreign firms.

The Role of Regional Trade Agreements

Deep integration and RTAs are not naturally related. Rather, they are currently con-
nected due to the weakened legislative function of the WTO, which has been unable to
conclude its first round of multilateral trade negotiations. With the Doha Round well
past its first decade of negotiations and without an end in sight, RTAs have emerged
as the most active institutional setting for governments to negotiate the lowering of
“behind- the-border” barriers to trade, thus either superseding or continuing the work
of the multilateral trade regime that began essentially with the Kennedy Round under
the GATT (Hoekman and Kostecki 2009, 582).
In the context of regional trade agreements, deep integration refers to provisions
that go to the heart of domestic policies that concern the contestability of the market.
Deep integration commitments promote harmonization or at least mutual recognition
of trade rules (Birdsall and Lawrence 1999), an idea that can be traced back as far as
Tinbergen’s (1954) work on economic integration. Deep integration facilitates the glo-
balization of production and trade along the international supply chain. RTA provi-
sions for deep integration have two defining features. First, the commitments focus on
“behind-the-border” trade rules; that is, domestic laws, regulations, and administrative
mechanisms that comprise a country’s trade regime. Second, in addition to their focus
on “behind-the- border barriers,” deep integration provisions in RTAs are also wide in
scope (Koremenos, Lipson, and Snidal 2001). They cover an extensive range of issues
that are mostly trade related but in some cases fall outside the governance purview of
existing WTO agreements.
Figure 19.1 is an example of a deep integration agreement that is, at the time of
writing, under active negotiation. The TPP agreement, involving twelve coun-
tries including the United States, is a twenty-first-century deep integration agree-
ment that features “regulatory coherence” as a key objective.1 Prospective members
advance the TPP as a high-standard agreement among “like-minded” countries. It
covers the key areas of deep integration, such as investment, services, competition,
labor, and environment, as well as the traditional areas of market access in goods
and trade remedies. Most relevant to this discussion of deep integration is that one
important goal of the agreement is to make the regulatory systems of the members
more compatible so that firms may operate more seamlessly and efficiently across the
partner economies.
 Soo Yeon Kim    363

Areas of Negotiation
Market Access for Goods
Trade Remedies
Legal Issues/Dispute Settlement

Cross-Border Services
Financial Services
Telecommunications

Competition Policy
Government Procurement
Intellectual Property
Investment
Sanitary and Phytosanitary Standards (SPS)
Technical Barriers to Trade (TBT)
Temporary Entry
Rules of Origin
Textiles and Apparel/ROOs

Cooperation and Capacity Building

Customs
E-Commerce
Environment
Labor

Fig. 19.1  The Trans-Pacific Partnership (TPP)

Agreement: A  Deep Integration RTA for the
21st Century.

Properties of Deep Integration Provisions in RTAs:

Protection, Liberalization, and Harmonization
Deep integration provisions in RTAs have three common properties: protection, lib-
eralization, and harmonization. They concern adjustments in domestic laws, regula-
tions, and administrative mechanisms that allow greater market contestation by foreign
firms. Provisions for deep integration allow for the leveling of the playing field in
national economies between domestic and foreign firms, overall decreasing the levels
of discrimination against the latter. Protection involves the representation of the eco-
nomic interests of foreign firms by enhancing their legal rights in the domestic arena
364    Deep Integration and Regional Trade Agreements

and providing for mechanisms of redress in cases where their interests are harmed.
Liberalization refers to the enhancement of market access for foreign firms by allowing
greater participation in all stages of economic activities. Last but not least, harmoniza-
tion entails consistency across countries of rules and policies that govern production
for the purposes of trade.
In the existing scholarship, studies of deep integration commitments in RTAs vary
in their relative emphasis on protection, liberalization, and harmonization depending
on the particular issue area. The relative emphasis on these properties is determined
by the cooperation problems inherent in the particular issue area under consideration.
For RTA provisions on technical barriers to trade (TBTs), for example, the cooperation
problems emanate from cross-national differences in standards and technical regu-
lations. The motivation to reduce production costs by making standards and techni-
cal regulations more compatible across countries calls for commitments to mutual
recognition of partner countries’ standards, technical regulations, and conformity of
assessment procedures, at a minimum.2 The strongest provisions in TBTs expressly
stipulate harmonization though the promotion of specific regional or international
standards. Piermartini and Budetta’s (2009) study of TBT provisions in RTAs reflects
this emphasis on harmonization:  beyond its reference to the TBT agreement of the
WTO, their classification focuses on the approach to integration—mutual recognition
or harmonization—in the areas of standards, technical regulations, and conformity
assessment.
In another issue area, Kotschwar (2009) examines investment provisions in RTAs
that emphasize the properties of protection and liberalization. The presence or absence
of an investor-state dispute resolution mechanism, as well as provisions for expropri-
ation, minimum standard of treatment, and treatment in case of strife, are common
features in RTAs to protect investor firm interests and assets. In areas relevant to lib-
eralization, Kotschwar’s classification includes the approach to the scheduling of
bindings under the most-favored nation (MFN) and national treatment principles.
Scheduling refers to the specific listing of liberalization commitments, and it may take a
positive-list or negative-list approach. A positive-list approach liberalizes only the sec-
tors that are listed, while a negative-list approaches liberalizes all sectors for investment
except those for which parties claim exemptions. Kotschwar’s study also specifies the
scope of coverage across the stages of investment: establishment, acquisition, postest-
ablishment, and resale. In the area of competition, another key area of deep integration
provisions, Teh (2009) classifies commitments that emphasize liberalization to enhance
the contestability of national markets. Teh’s study covers national regulations govern-
ing competition, including provisions by sector and improving upon earlier work by
the Organisation for Economic Co-operation and Development (OECD) (Solano and
Sennekamp 2006). The study examines nine areas of competition-related regulations,
including commitments to cooperation; promotion of national competition and reg-
ulation of anticompetitive behavior; dispute settlement; institutional arrangements;
and general and horizontal principles of nondiscrimination, procedural fairness, and
transparency.
 Soo Yeon Kim    365

The above studies reflect the broader direction of the existing literature, which
exhibits an overall trend toward mapping of RTA provisions, but are also relevant for
the study of deep integration. Current RTA-mapping projects include the Design of
Trade Agreements (DESTA) project (Dür et al. 2012), perhaps the largest-scale proj-
ect, comprising more than 500 coded agreements; Estevadeordal, Suominen, and
Teh’s (2009) focused study of a sample of seventy-four agreements chosen for diver-
sity in economic development, trade flows, and geography in their participants; and
Chauffour and Maur’s (2011) study of developing country strategies in negotiating
RTAs, building on seminal works by De Melo and Panagariya (1996) and Schiff and
Winters (2003) and the World Bank’s Trade Blocs (2000). While they stand as inde-
pendent RTA-mapping projects, each also demonstrates how the properties of deep
integration—protection, liberalization, and harmonization—are written into specific
commitments in RTAs.

Measuring Deep Integration

Though a voluminous literature exists on the impact of RTAs on trade flows, the lit-
erature on the effects specifically of deep integration commitments in RTAs is less
well-developed. One reason has been the difficulty of measuring what exactly consti-
tutes deep integration provisions in RTAs. The literature so far features several types
of measures of deep integration, from qualitatively driven indexes based on the agree-
ment’s scope to statistically driven scores.
The widely known metric devised by Horn, Mavroidis, and Sapir (2009) is a qualita-
tive measure to distinguish the RTAs of the United States and the EU. Though the study
does not focus explicitly on deep integration commitments, the analysis does enumer-
ate the kinds of issue areas that are relevant. Provisions in US and EU RTAs are classified
as WTO-plus and WTO-X. WTO-plus provisions cover issue areas under the purview
of WTO agreements but whose commitments go beyond current disciplines. WTO-X
provisions cover policy areas that are not currently covered by WTO agreements.
They include but are not limited to “competition policy, environmental laws, invest-
ment, IPRs, capital movements, consumer and data protection, cultural cooperation,
education, energy, health, human rights, illegal immigration, illicit drugs, money laun-
dering, R&D, SMEs, social matters, statistics, taxation, and visa and asylum policies”
(pp. 12–15).
The WTO’s annual World Trade Report (2011) extended the Horn, Mavroidis, and
Sapir study and their assessment of deep integration provisions to ninety-seven more
RTAs, in addition to the original thirty-three for the EU and eleven for the United States.
The report found that RTAs increasingly include deep integration provisions that are
legally enforceable. RTA provisions include not only WTO-plus commitments, but also
more and more WTO-X areas. In particular, the report finds a striking pattern in par-
ticular issue areas that are included in RTAs: competition policy, movement of capital,
intellectual property rights beyond TRIPS, and investment.
366    Deep Integration and Regional Trade Agreements

Existing scholarship has also proposed alternative methodologies to measure

deep integration commitments in terms of the general “depth” of a trade agreement.
Perhaps the most straightforward approach has been to construct an additive and
usually unweighted index of all the provisions in a particular RTA that are geared
toward liberalization, including not only border barriers such as tariffs but also
behind-the-border deep integration provisions and enforcement mechanisms. Hicks
and Kim (2012) and Baccini, Dür, and Elsig (2013) construct such indexes, though
the latter explicitly label the measure as “depth.” Kim (2012) also develops an addi-
tive index of deep integration commitments in RTAs based on TBTs and competition
provisions and analyzes the impact of production network trade on the strength of
commitments in these two issue areas. Orefice and Rocha (2011) also employ an addi-
tive index, but the index is constructed as an unweighted measure of the number of
legally enforceable provisions in an agreement defined according to Horn, Mavroidis,
and Sapir (2010). A more data-driven approach utilizes exploratory factor analysis
or principal components analysis (Orefice and Rocha 2011; Baccini, Dür, and Elsig
2013) to generate an index that captures the overall (and latent) depth of an RTA.
This approach yields a weighted index of deep integration in an RTA, and it is com-
puted using all the provisions included in the agreement. The index constructed as
a comprehensive measure of deep integration is essentially the factor score for each
agreement, in which the index is weighted according to the factor loadings for each
provision of the RTA.
There is no widely employed measure of “deep integration” commitments in RTAs.
Rather, studies have operationalized deep integration using both substantively driven
and data-driven measures. These measures were uniformly constructed to capture the
“depth” of RTA commitments, whether in a comparison of EU and US trade agreements
or a broader set of RTAs. The literature has so far not produced discrepant findings that
highlight the strengths and drawbacks of different measurements of deep integration;
however, the few studies that exist do form a critical mass of scholarship to draw major
conclusions, and the measurement issue would benefit from additional and method-
ologically systematic investigation.

Why Do States Commit to Deep

Integration Agreements?

The most important factor pushing governments toward deep integration commit-
ments in RTAs is the increasing fragmentation of international production and trade
along the international supply chain. Declining transport costs, as well as advances in
information, communications, and technology, have allowed for greater geographical
fragmentation of the production process. In this connection, firms that rely on inter-
mediate inputs, whether multinational or national, are especially likely to push for deep
 Soo Yeon Kim    367

integration commitments in investment, services, and the protection of intellectual

property rights in their countries’ agreements with trade partners.

Production Networks and Deep Integration

An emerging empirical literature in economics has examined the links between pro-
duction networks and deep integration RTAs. The studies consistently find that high
levels of trade in parts and components—the favored proxy for measuring production
network trade (Yeats 1998; Hummels, Ishii, and Yi 2001; Athukorala 2011)—is associ-
ated with strong deep integration commitments in RTAs.3 The success of production
networks in generating profits for firms relies on low(er) trade costs that derive from
domestic-trade-related regulations and, equally important, stable expectations regard-
ing these regulations. As barriers at the border such as tariffs and other quantitative
restrictions have become lower through the successful rounds of GATT negotiations,
trade and production costs generated by incompatibilities in trade-related domestic
regulations across countries now figure more prominently in the strategies of multi-
national firms that rely heavily on the movement of intermediate goods. Their opera-
tions are affected by inadequate infrastructure and services, such as transportation and
communications.
Baldwin (2011) calls this intensification and increasing complexity of the interna-
tional supply chain the “second unbundling.” It has created a trade-investment-services
nexus in which the conventional links between factories and offices have been inter-
nationalized and dispersed across different countries. The expansion of production
networks thus requires sufficient convergence in legal and regulatory systems, as well
as economic institutions, in relevant areas such as product and production standards,
intellectual property rights, and protection and liberalization of investment. Such ini-
tiatives toward harmonization are thus encoded into provisions in RTAs, and commit-
ments in these areas are supported by a strong and highly legalized dispute settlement
system. The literature also suggests that a two-way link exists between production
networks and deep integration RTAs (Orefice and Rocha 2011; WTO 2011). Countries
strongly embedded in the international supply chain and engaged in the fragmentation
of production are more likely to sign on to deep integration RTAs so as to secure their
position as providers of intermediate goods and services. Kim (2012) has found that
high levels of production network trade are associated with stronger commitments in
TBT provisions, but only weak effects in the area of competition policy. Moreover, there
is a distinct pattern in US and EU RTAs, in which the US RTAs produce stronger com-
mitments in TBTs, while the EU RTAs demonstrate stronger commitments to liberal-
ization of competition policy.
It should be noted that deep integration commitments are practical only after states
achieve a certain level of “shallow integration,” as only when the more traditional trade
policy measures such as tariff and quantitative restrictions have been lowered does the
need arise to attend to differences in national regulatory systems. Thus deep integration
368    Deep Integration and Regional Trade Agreements

RTAs with strong commitments are most likely to be found for countries that have
already achieved low levels of border protection, such as developed countries. For
North-South RTAs and South-South RTAs, “shallow integration” or “negative integra-
tion” can still be expected to figure prominently in RTA provisions (see also Manger and
Shadlen, this volume).

Multinational Firms and the Political

Economy of Deep Integration
In the political economy of deep integration, the role of multinational firms comprises
an important area of research. In earlier eras multinational firms supported trade lib-
eralization (Milner 1988), but more recently the operational needs of international
business and multinational firms have been instrumental in promoting regional trade
arrangements. The EU initiative toward a single market was buttressed by European
firms that sought economies of scale to boost international competitiveness. NAFTA
was actively supported by US firms as well as Mexico’s large industrial groups that
formed regional production networks (Fishlow and Haggard 1992). In Asia, where lev-
els of production network trade are among the highest in the world (Athukorala 2005,
Athukorala, 2011), regional integration has been largely informal and led largely by pri-
vate foreign and regional multinational firms.
The existing literature also identifies variability across sectors. Trade in goods, for
example, is different from trade in services. Where production networks and deep
integration are concerned, the political economy of trade in services is strongly related
to foreign direct investment (FDI). Indeed, in the area of services, where export inter-
ests are weak relative to the manufacturing or agricultural sectors, it is the inves-
tors—the multinational firms—that function as “export” interests that drive trade
negotiations. Kerry Chase has produced studies on the political economy of integra-
tion in different sectors, including investment (2004), rules of origin (2008a), and
services (2008b). Though the studies largely focus on the United States, they are as a
collection illustrative of the interests at play in each sector and also indicative of the
kinds of cross-linkages between sectors that would complicate negotiations in a deep
integration RTA.
Governments sign on to deep integration RTAs in an effort to facilitate the operations
of production networks. Multinational firms are the key promoters of making national
regulatory systems compatible, as this helps to reduce the cost of doing business abroad
and generally improves the operation of the international supply chain. In the politics
of production networks in the design of RTAs, firms that rely on trade along the inter-
national supply chain, whether they are multinational or national, are likely to figure as
important political actors that have incentives to lobby governments for deep integra-
tion commitments.
 Soo Yeon Kim    369

The Effects of Deep Integration RTAs

The effects of deep integration RTAs may be assessed in terms of costs in two areas. On the
one hand, the mix of developmental stages of agreement partners results in uneven costs
and benefits for agreement partners. North-North, North-South, and South-South RTAs
are affected by the level of asymmetry in market size. North-South agreements increase the
prospects for the importation of the more developed country’s regulatory regime by the less
developed country, as the latter is likely to have an underdeveloped domestic trade regime
in the first place. Disparities in levels of economic development will be reflected in insti-
tutional preferences as well, as the less-developed economy is likely to focus on industrial
market access, while the more developed economy is likely to seek to export its regulatory
regime. Second, the cost of deep integration will also depend on how non-members are
likely to be affected by deeper integration between RTA members. Competing and largely
incompatible models of trade governance written into these RTAs will increase discrimi-
nation against nonmembers and also lower the possibility that deep integration norms in
RTAs can serve as a focal point for cooperation in the WTO or other multilateral forum.
Current scholarship on the effects of deep integration RTAs is limited (WTO 2011,
145), due largely to the problem of measuring deep integration, as noted above. Existing
studies can be divided into those that examine economic effects, specifically the impact
on production network trade, and those that examine the consequences for multilateral
trade liberalization, largely along the lines of RTAs as “stumbling blocks” or “building
blocks.” Empirical studies of the trade effects of deep integration RTAs include Orefice
and Rocha (2011), who found strong positive effects of deep integration RTAs on pro-
duction network trade defined in terms of trade in parts and components. This effect
is pronounced in trade in automobile parts and in intermediate products in informa-
tion and technology. The WTO’s (2011) report on trade agreements corroborates these
results. Deep integration RTAs have a positive and statistically significant impact on
trade in parts and components, and every additional deep integration provision adds
to this positive effect. The WTO study finds further evidence that individually, commit-
ments to harmonize standards (under TBT provisions) and to adopt competition laws
also enhance production network trade among members.

Convergence or Competition?
Another area of the literature has extended the debate on RTAs as “stumbling blocks”
or “building blocks” to multilateral trade liberalization (see also Johns and Peritz, this
volume). This debate has focused on the degree to which the deep integration com-
mitments in these agreements can be multilateralized to provide an alternative venue
for the negotiation of multilateral trade rules—the convergence route for managing
370    Deep Integration and Regional Trade Agreements

international trade. The other side of this debate privileges the role of power politics in
producing “hubs of governance,” or competing models of trade governance in which
leading economies pursue particular models of deep integration that are essentially dis-
criminatory to non-RTA members.
Multilateralization of deep integration commitments in RTAs, which would extend
commitments beyond the agreement partners, is a significant prospect. Given that deep
integration commitments reach into domestic laws, regulations, and administrative
mechanisms, multilateralizaton may be easier to achieve, as reforming these laws may be
equivalent to unilateral liberalization. The issue of convergence is not exclusive to RTAs,
but rather is a broader issue of concern when it comes to the governance of globaliza-
tion (Kahler and Lake 2003). In arguing for a convergence route to multilateral trade lib-
eralization, Hoekman and Kostecki (2009, 509) point to regulatory convergence in the
process of deep integration under the single market program of the EU. Across trade in
goods and services, the European Commission and the European Court of Justice actively
intervened to remove not only barriers-at-the border such as tariffs and quotas, but also
regulatory limitations on the movement of goods and services. In the area of standards,
a key component of deep integration, the European Commission and European Court
of Justice provided for mutual recognition of standards across member countries, thus
achieving regulatory convergence to facilitate the movement of goods and services.

Hubs of Governance
Another scenario in this debate privileges the role of power politics in which the web
of RTAs that populate the global trading system diverge in a hub-and-spokes pattern,
with each “hub of governance” centered on a leading economy. This form of “hegemonic
liberalization” (Hoekman and Winters 2009) is one form of cooperation through RTAs
on market regulation and domestic economic policies. It may result in the multilateral-
ization of trade rules, but could just as likely lead to competition between models of gov-
ernance. That is, leading economies may utilize their market power to negotiate deep
integration provisions in RTAs. To the extent that partner economies accept and adopt
these deep integration provisions, some degree of multilateralization of trade rules
results. Moreover, where partner countries extend these same deep integration commit-
ments in their subsequent RTAs, a higher-order effect is also achieved for the particu-
lar deep integration model of the RTA, such as the spread of RTAs between the United
States and its regional neighbors across the Americas (Shadlen 2005, Shadlen, 2008).
However, the multilateralization scenario assumes that there is little divergence across
the deep integration models pursued by the leading economies. This assumption is not
tenable, as studies have shown that leading economies such as the United States and EU
pursue different priorities in their deep integration RTAs. The United States, for exam-
ple, utilizes a standard template for all partners in bilateral investment treaties (BITs)
and also follows a consistent template for the protection of intellectual property rights in
RTAs (World Bank 2005). The EU prioritizes reform and harmonization of competition
 Soo Yeon Kim    371

policy and also seeks to maintain its system of geographical indications (GI), which
denote the geographical origins of certain products (such as French wines and Swiss
cheeses), in its RTAs (Schiff and Winters 2003). Horn, Mavroidis, and Sapir (2009), in
their comparative study of US and EU FTAs, find that the United States pursues a “hard”
legalization approach (Abbott et al. 2001; Abbott and Snidal 2000), while the EU pur-
sues a “soft” legalization approach in its RTAs. The US RTAs contain more WTO-plus
provisions that are highly legalized and enforceable. While US RTAs do not include as
many WTO-X provisions, they do include them consistently in the areas of anticor-
ruption (not found in EU RTAs), competition policy, environmental laws, investment,
IPRs, labor market regulation, and capital movement. The EU is the “market leader” in
WTO-X provisions; however, the majority of provisions lack any meaningful enforce-
ment mechanisms. Even superficially, the “hard” legalization approach of the United
States and the “soft” legalization approach of the EU reflect a broader divergence in the
deep integration strategies and governance models of these two leading economies.
The different priorities that leading economies place on the wide range of issue areas
that fall under deep integration produce different governance models for trade, and to
the extent that they are supported by the accumulation of second-order RTAs created
by RTA partners, are likely to obstruct efforts to harmonize or to adopt mutual recogni-
tion practices at the multilateral level. The problem of competition in governance mod-
els becomes more acute when we take into account the role of emerging markets—the
BRICS countries including Brazil, Russia, India, China, and South Africa, and beyond
(Mattoo and Subramanian 2009). In Asia, for example, where the importance of China as
an economic power is most evident, Kim and Manger (2013) find striking differences in
the services liberalization provisions in RTAs signed by the United States and by China.
More broadly, the idea of “hubs of governance” formed by RTAs with varying degrees
of deep integration commitments is especially compelling given the recent trend toward
forming “mega-FTAs.” Agreements under negotiation—such as the Transatlantic Trade
and Investment Partnership (TTIP) between the United States and EU, the TPP between
the United States and Asian trade partners, and the Regional Comprehensive Economic
Partnership Agreement (RCEP) between China and its trade partners—bring under
one large-scale umbrella agreement countries that already have existing RTAs between
them. These agreements not only offer competing models of trade liberalization and
deep integration, but they also reflect how the world’s leading economies are competing
for influence in global trade relations.

Avenues for Future Research

The study of deep integration commitments in RTAs remains a highly promising area
for fruitful research in trade agreements and trade governance more broadly. In an
effort to promote more scholarship in this frontier in the study of trade governance, this
section poses questions that would benefit from further study.
372    Deep Integration and Regional Trade Agreements

What is the Baseline?

Existing studies of deep integration in RTAs would benefit greatly from an assessment
of “depth” with respect to some baseline in the regulatory status quo of signatories.
Studies have emphasized the degree to which states make legally binding commitments
in a range of governance areas that purportedly increase the contestability of markets in
partner economies. However, this raises the question of the baseline that constitutes the
basis for comparison: Do RTA commitments really promote reform to protect, to lib-
eralize, and to harmonize, or are states simply formalizing and internationalizing their
existing regulatory regimes? Downs, Rocke, and Barsoom (1996) advanced the widely
acknowledged argument that the contributions of international institutions are greatly
exaggerated, as they mostly provide for only modest commitments that do not depart
significantly from what signatories would have undertaken in the absence of such insti-
tutional obligations. Hoekman and Winters (2009) similarly find little evidence that
deep integration provisions actually promote regulatory reform. Rather, the study finds
that in most cases, states pursue regulatory reform unilaterally.
The absence of a baseline that reflects the regulatory sophistication of a country also
makes it difficult to determine causal direction: one cannot infer that the templates for
reform provided by RTA provisions are promoting regulatory changes and improve-
ments beyond what governments have already decided autonomously to pursue. Put
differently, the lack of a baseline as the basis of comparison may also lead to selection
effects (Von Stein 2005; Simmons and Hopkins 2005), in which an international institu-
tion simply reflects signatories’ preferences rather than a commitment to change. An
implication of this argument is that the net benefits of such RTAs would be marginal.
Many of the mapping projects in existence today rely on the coding and interpreta-
tion of legal text and do not base assessments of deep integration on existing multilateral
or international commitments or on current levels of national legislation in the relevant
issue areas. This baseline problem is exacerbated by the fact that RTA provisions them-
selves do not imply or explicate the extent to which signatory governments must change
national laws or regulations. One approach for a comparative assessment would be to
take commitments in existing multilateral and bilateral treaties as the baseline for com-
parison. One baseline that researchers have employed is the relevant WTO agreement
itself, in an effort to assess the extent to which an RTA advances commitments beyond
the status quo defined under the multilateral trade regime.
Teh (2009) and Teh, Prusa, and Budetta (2009) evaluate the strictness of trade rem-
edy provisions of RTAs vis-à-vis current commitments on antidumping duties, coun-
tervailing duties, and bilateral and global safeguards under WTO rules. The study finds
that RTAs on average do not exceed the strictness of current WTO rules. However,
deep integration RTAs do stand out as agreements that place prohibitions on at least
one type of trade remedies, such as exempting the partner country from any invoca-
tion of global safeguards. A study on services undertaken in Roy (2011) and expand-
ing on Roy, Marchetti, and Lim (2009) and Marchetti and Roy (2008), assesses service
 Soo Yeon Kim    373

commitments in RTAs for fifty-three countries across sixty-seven agreements.4 For

every RTA to which a country is a signatory, the RTA’s commitments were evaluated
against commitments in the General Agreement on Trade in Services (GATS) and
against current Doha Round offers. This study explicitly avoids a “legal interpretation”
of the agreements and provides only an assessment of the extent to which the selected
countries have gone beyond current levels of GATS commitments and Doha Round
offers. It nonetheless constitutes one of the few studies that provide a basis of compari-
son for RTA commitments. The data also cover only mode 1 (cross-border trade) and
mode 3 (commercial presence), but these are the sectors in which most services trade is
conducted.
While these studies take the WTO agreements as the baseline, and scholarship would
certainly benefit from more detailed study of individual issue areas, this approach has
two weaknesses. First, it cannot accommodate newer issues that are not governed by the
WTO. In this connection, future work may look toward, for example, existing obliga-
tions in other multilateral treaties, such as multilateral environmental agreements for
environmental standards, International Labour Organization (ILO) criteria for labor
standards, and bilateral investment treaties for investment provisions.5 Second and
more important, this approach does not capture the domestic regulatory landscape of
the members and therefore cannot gauge the extent to which an RTA actually represents
a commitment beyond the regulatory status quo of its signatories. Assessment vis-à-vis
WTO rules or any other international agreement is essentially a comparison between
legal texts, and the current literature provides little insight into how these rules map onto
the domestic regulatory landscape. Indeed, assessing not only the actual level of deep
integration commitments but also the benefits that can be accrued from them hinges on
whether the agreement requires actual regulatory reform on the part of signatories.

Regulatory Disarray in Trade Governance

The issue of whether the increasingly complex and growing network of RTAs is likely
to lead to convergence or discrimination in international trade is closely related to the
question of how states’ memberships in numerous RTAs matters. At a time when the
rule-making function of the multilateral trade regime is greatly impaired, the dramatic
expansion of RTAs in the global economy advances the much-needed development of
national trade regimes in areas either underdeveloped in the WTO or currently outside
its governance purview. There is, however, a crucial drawback to this phenomenon: the
possible emergence of a “spaghetti bowl” of rules from overlapping agreements, which
can contribute to regulatory disarray.
The multitude of RTAs and the overlapping memberships of countries pose a chal-
lenge to the management of RTA obligations. According to Chaffour and Maur (2011, 1),
countries in Africa belong, on average, to at least four RTAs. In Latin America, the
average country has signed at least seven agreements. Modern deep integration RTAs
are legally complex, and each agreement has its own tariff schedule, list of exclusions
374    Deep Integration and Regional Trade Agreements

and implementation times, rule of origin, and customs procedures. Thus, in the aggre-
gate, obligations across overlapping agreements are likely to diminish the efficiency of
managing trade and its rules, as well as the transparency that a single multilateral trade
regime would make possible. In the absence of efforts to multilateralize RTA obligations
or toward open regionalism in which all trading partners may benefit, legal diversity
and conflicts between rules are likely to have a negative impact on progress toward a
more developed multilateral trading system (Bhagwati 2008).
Scholarship needs to move beyond the long-standing debate on trade-creation ver-
sus trade-diversion from RTAs (Viner 1950) to take account of the impact of deep inte-
gration commitments. In thus shifting the debate, Mansfield and Reinhardt (2008),
for example, have found that membership in RTAs leads to less trade volatility among
their members, as RTAs reduce uncertainty. The regulatory disorder that can result
from overlapping RTAs may undermine not only WTO agreements but also other exist-
ing international institutions. In the case of environmental provisions, where there are
treaties—multilateral environmental agreements (MEAs)—in force, RTAs can either
conflict with the legal obligations in them or go beyond them to push the current bound-
aries of environmental standards in trade (Kernohan and De Cian 2007). The mapping
projects discussed above are equipped with the basic materials to examine the broader
issue of regulatory conflict.
Finally, the question of regulatory disorder beyond the contracting parties of the
RTA also calls for more extensive research into nonviolation cases in WTO disputes.
Nonviolation complaints fall under Article 26 of the WTO’s Dispute Settlement
Understanding (DSU). The DSU allows for the filing of cases involving a domestic policy
measure that does not necessarily violate WTO obligations but has the effect of “nullifi-
cation and impairment” of the benefits of an existing concession (Trebilcock and Howse
2005, 143).6 Nonviolation complaints are especially suited for examining the impact of
regulatory reform brought about by deep integration RTAs, given that the core provi-
sions of deep integration RTAs concern protection, liberalization, and harmonization
of trade-related domestic regulatory rules. Such rules, to the extent that they produce
regulatory conflicts with the concessions already in effect under the WTO, are likely
to bring an increase in nonviolation complaints under the WTO system and would be
strongly indicative of regulatory disorder.

Conclusion

Deep integration is not a new phenomenon in the politics of international trade.

Negotiations for the liberalization of domestic trade rules began with the Kennedy
Round of the GATT as the negotiating rounds achieved success in reducing tariffs
and quantitative restrictions. As tariff levels receded, national differences in domestic
laws and regulations governing trade became more visible and thus a focal point for
negotiating trade rules. GATT members proposed coordination of national policies
 Soo Yeon Kim    375

and outright harmonization of regulatory systems. This trend continues in the Doha
Round era through deep integration RTAs, incorporating not only competition pol-
icy and investment but also long-standing initiatives on labor and environmental
standards.
Deep integration as encoded in the legal texts of RTAs, however, is indeed a recent
development in the regulation and management of trade. Robert Lawrence’s early study
of deep integration (1996) focused on the tension between national autonomy in eco-
nomic policy and the demands on domestic reform of regional trade agreements. Since
this path-breaking study, the depth in deep integration RTAs has only intensified, as
RTAs have increasingly widened the scope of issue areas they cover. This chapter has
provided a survey of the existing scholarship on the defining features and functions of
deep integration, the determinants of states’ choices to enter into deep integration RTAs,
and the consequences for international trade of these agreements. The discussion has
also raised questions and avenues for future research.
Deep integration in RTAs emphasizes the reduction or dismantling of behind-
the-border barriers, thus largely supporting the “positive integration” approach of the
WTO in actively reforming national regulatory systems to be more WTO-consistent.
Provisions for deep integration are wide in scope and increasing over time to include
issue areas not under the purview of the WTO agreements. Deep integration commit-
ments in various issue areas protect the interest of foreign firms, liberalize domestic
trade rules to level the playing field for foreign firms, and harmonize regulatory systems
to make them more compatible between RTA partners. Current scholarship provides
measures of deep integration in individual issue areas as well as aggregate measures to
gauge the overall depth of an agreement.
Governments enter into deep integration RTAs as a response to the international-
ization and increasing geographic fragmentation of international production. The
trade-investment-services nexus (Baldwin 2011) figures importantly in the strength of
commitments in deep integration RTAs, as they facilitate the operation of production
networks and reduce the cost of doing business. In the political economy of deep inte-
gration, multinational firms are key political actors that drive the negotiation of RTAs.
Existing studies show that deep integration RTAs lead to higher levels of production
network trade and vice versa. At the systemic level, the network of RTAs may lead to
convergence in deep integration provisions and subsequently national regulatory sys-
tems within “hubs of governance,” as well as to competition between these hubs and
models of deep integration espoused by the leading economies.
Scholarship would benefit from research that provides a baseline for comparison
of the actual regulatory change that would be required for RTA members. Whereas
for less developed countries regulatory reform under a deep integration RTA may be
far-reaching, developed countries with established and well-functioning domes-
tic regulations may simply be formalizing their institutional status quo. In addition,
there is a need for the debate on regionalism to move beyond trade-creation versus
trade-diversion, to examine the possibility of regulatory disarray in international trade
and inconsistencies between deep integration commitments in RTAs and obligations in
376    Deep Integration and Regional Trade Agreements

relevant multilateral treaties, such as between provisions on environmental standards

and legal obligations in multilateral environmental agreements. Finally, given that deep
integration RTAs seek greater regulatory coherence among partners, research on non-
violation complaints in WTO dispute cases would lend insights into the impact of these
agreements on existing commitments under the multilateral trade regime.

Notes
1. The current members of the agreement are Australia, Brunei Darussalam, Canada, Chile,
Japan, Malaysia, Mexico, New Zealand, Peru, Singapore, the United States, and Vietnam.
(http://www.ustr.gov).
2. Following the language of the WTO’s Agreement on Technical Barriers to Trade (TBT
Agreement), standards refer to voluntary standards for products, technical regulations are
mandatory standards, and conformity assessment refers to the formal process by which
products are certified as having met the standards and technical regulations of a particular
country. The WTO’s TBT Agreement covers the above three governance areas for prod-
ucts but not for production standards.
3. Athukorala’s (2011) definition of parts and components trade is the most expansive of the
three studies. In measuring parts and components, the study identifies six product catego-
ries in which network trade is heavily concentrated: SITC 75-SITC 78, including 1) office
machines and automatic data processing machines, 2)  telecommunications and sound
recording equipment, 3) electrical machinery, 4) road vehicles; and 5) SITC 87-SITC 88,
which include professional and scientific equipment and photographic apparatus. This
measure differs from the OECD approach and from Yeats (1998) and Hummels, Ishii, and
Yi (2001). Athukorala’s study also provides a very useful survey and critique of existing
measures of production network trade (67–69).
4. The coding scheme is based on Hoekman (1996), which evaluated GATS commitments of
GATT members during the Uruguay Round.
5. For the case of investment provisions, Kotschwar (2009) has pointed out that RTAs go
beyond many bilateral investment treaties in terms of protection, liberalization, and har-
monization in investment provisions.
6. A well-known case of a nonviolation dispute is the Kodak-Fuji dispute between the United
States and Japan (DS44), in which the former claimed that competition policies of the
Japanese government regarding the distribution and domestic sale of imported film and
paper constituted nonviolation nullification and impairment of benefits.

References
Abbott, Kenneth W., and Duncan Snidal. 2000. Hard and Soft Law in International Governance.
International Organization 54: 421–456.
Abbott, Kenneth W., Robert O. Keohane, Andrew Moravcsik, Ann-Marie Slaughter, and
Duncan Snidal. 2001. The Concept of Legalization. International Organization 54: 401–419.
Athukorala, Prema-chandra. 2005. Product Fragmentation and Trade Patterns in East Asia.
Asian Economic Papers 4 (3): 1–27.
 Soo Yeon Kim    377

Athukorala, Prema-chandra. 2011. Production Networks and Trade Patterns in East

Asia: Regionalization or Globalization? Asian Economic Papers 10 (1): 65–95.
Baccini, Leonardo, Andreas Dür, and Manfred Elsig. 2013. Depth, Flexibility, and International
Cooperation: The Politics of Trade Agreement Design. Unpublished manuscript.
Baldwin, Richard. 2011. 21st Century Regionalism: Filling the Gap between 21st Century Trade
and 20th Century Trade Rules. Centre for Economic Policy Research Policy Insight No. 56.
Bhagwati, Jagdish. 2008. Termites in the Trading System: How Preferential Trade Agreements
Undermine Free Trade. Oxford: Oxford University Press.
Birdsall, Nancy, and Robert Z. Lawrence. 1999. Deep Integration and Trade Agreements. In
Global Public Goods:  International Cooperation in the 21st Century, edited by Inge Kaul,
Isabelle Grunberg, and Marc A. Stern, 128–151. Oxford: Oxford University Press.
Chase, Kerry. 2004. From Protectionism to Regionalism:  Multinational Firms and
Trade-Related Investment Measures. Business and Politics 6 (2): 1–36.
Chase, Kerry. 2008a. Moving Hollywood Abroad: Divided Labor Markets and the New Politics
of Trade in Services. International Organization 62 (4): 653–687.
Chase, Kerry. 2008b. Protecting Free Trade:  The Political Economy of Rules of Origin,
International Organization 62 (3): 507–530.
Chauffour, Jean-Pierre, and Jean Christophe Maur, eds. 2011. Preferential Trade Agreement
Policies for Development:  A  Handbook. Washington, DC:  International Bank for
Reconstruction and Development/World Bank.
De Melo, J., and A. Panagariya, eds. 1996. New Dimensions in Regional Integration. Cambridge,
UK: Cambridge University Press.
Downs, George W., David M. Rock, and Peter N. Barsoom. 1996. Is the Good News about
Compliance Good News about Cooperation? International Organization 50 (3): 379–406.
Dür, Andreas, Leonardo Baccini, and Manfred Elsig, 2013. The Design of International Trade
Agreements: Introducing a New Database. Review of International Organizations. DOI 10.1007/
s11558-013-9179-8.
Estevadeordal, Antoni, Kati Suominen, and Robert Teh, eds. 2009. Regional Rules in the Global
Trading System. Cambridge, UK: Cambridge University Press.
Fishlow, Albert, and Stephan Haggard. 1992. The United States and the Regionalization of the
World Economy. Paris: OECD Development Centre.
Hicks, Raymond, and Soo Yeon Kim. 2012. Reciprocal Trade Agreements in Asia: Credible
Commitment to Trade Liberalization or Paper Tigers? Journal of East Asian Studies 12
(1): 1–30.
Hoekman, Bernard. 1996. Assessing the General Agreement on Trade in Services. In The
Uruguay Round and Developing Countries, edited by Will Martin and L. Alan Winters,
88–124. Cambridge, UK: Cambridge University Press.
Hoekman, Bernard, and Michel M. Kostecki. 2009. The Political Economy of the World Trading
System. 3rd ed. Oxford: Oxford University Press.
Hoekman, Bernard, and L. Alan Winters. 2009. Multilateralizing Preferential Trade Agree­
ments: A Developing Country Perspective. In Multilateralizing Regionalism: Challenges for
the Global Trading System, edited by Richard Baldwin and Patrick Low, 636–680. Cambridge,
UK: Cambridge University Press.
Horn, H., P. C. Mavroidis, and A. Sapir. 2010. Beyond the WTO? An Anatomy of EU and US
Preferential Trade Agreements. World Economy 33 (1): 1565–1588.
Hummels, David, Jun Ishii, and Kei-Mu Yi. 2001. The Nature and Growth of Vertical
Specialization in World Trade. Journal of International Economics 54: 75–96.
378    Deep Integration and Regional Trade Agreements

Kahler, Miles, and David Lake, eds. 2003. Governance in a Global Economy: Political Authority
in Transition. Princeton, NJ: Princeton University Press.
Kernohan, David, and Enrica De Cian. 2007. Trade, the Environment and Climate
Change: Multilateral versus Regional Agreements. In Climate and Trade Policy: Bottom-up
Approaches Towards Global Agreement, edited by Carlo Carraro and Christian Egenhofer,
70–93. Cheltenham, UK: Edward Elgar.
Kim, Soo Yeon. 2012. Negotiating the Nexus: Production Networks and Behind-the-Border
Commitments in Regional Trade Agreements. Unpublished manuscript.
Kim, Soo Yeon, and Mark Manger. 2013. Hubs of Governance:  Path-Dependence and
Higher-Order Effects of PTA Formation. Unpublished manuscript.
Koremenos, Barbara, Charles Lipson, and Duncan Snidal. 2001. The Rational Design of
International Institutions. International Organization 55 (4): 761–799.
Kotschwar, Barbara. 2009. Mapping Investment Provisions in Regional Trade Agreements:
Towards and International Investment Regime? In Regional Rules in the Global Trading
System, edited by Antoni Estevadeordal, Kati Suominen, and Robert Teh, 365–417.
Cambridge, UK: Cambridge University Press.
Lawrence, Robert Z. 1996. Regionalism, Multilateralism, and Deeper Integration. Washington,
DC: The Brookings Institution.
Mansfield, Edward D., and Eric Reinhardt. 2008. International Institutions and the Volatility of
International Trade. International Organization 62 (4): 621–652.
Marchetti, Juan A., and Martin Roy. 2008. Services Liberalization in the WTO and in
RTAs. In Opening Markets for Trade in Services; Countries and Sectors in Bilateral and
WTO Negotiations, edited by Juan A Marchetti and Martin Roy, 61–112. Cambridge,
UK: Cambridge University Press and WTO.
Mattoo, Aaditya, and Arvind Subramanian. 2009. From Doha to the Next Bretton Woods:
A New Multilateral Trade Agenda. Foreign Affairs 88 (1): 15.
Orefice, Gianluca, and Nadia Rocha. 2011. Deep Integration and Production Networks: An
Empirical Analysis. World Trade Organization Staff Working Paper No. ERSD-2011-11.
Piermartini, Roberta, and Michele Budetta. 2009. A Mapping of Regional Rules on Technical
Barriers to Trade. In Regional Rules in the Global Trading System, edited by Antoni
Estevadordal, Kati Suominen, and Robert Teh, 250–315. Cambridge, UK:  Cambridge
University Press.
Roy, Martin. 2011. Services Commitments in Preferential Trade Agreements: An Expanded
Dataset. World Trade Organization Staff Working Paper No. ERSD-2011-18.
Roy, Martin, Juan Marchetti, and Hoe Lim. 2009. Services Liberalization in the New Generation
of Preferential Trade Agreements: How Much Further Than the GATS? In Regional Rules in
the Global Trading System, edited by Antoni Estevadeordal, Kati Suominen, and Robert Teh,
316–364. Cambridge, UK: Cambridge University Press.
Schiff. M., and L. A. Winters 2003. Regional Integration and Development. Washington,
DC: World Bank.
Shadlen, Kenneth C. 2005. Exchanging Development for Market Access? Deep Integration and
Industrial Policy under Multilateral and Regional-Bilateral Trade Agreements. Review of
International Political Economy 12 (5): 750–775.
Shadlen, Kenneth C. 2008. Globalisation, Power and Integration: The Political Economy of
Regional and Bilateral Trade Agreements in the Americas. Journal of Development Studies
44 (1): 1–20.
 Soo Yeon Kim    379

Simmons, Beth A., and Daniel J. Hopkins. 2005. The Constraining Power of international
Treaties: Theory and Methods. American Political Science Review 99 (4): 623–631.
Solano, Oliver, and Andreas Sennekamp. 2006. Competition Provisions in Regional Trade
Agreements. OECD Trade Policy Working Papers No. 31. Paris: OECD.
Teh, Robert. 2009. Competition Provisions in Regional Trade Agreements. In Regional Rules in
the Global Trading System. edited by Antoni Estevadordal, Kati Suominen, and Robert Teh,
418–491. Cambridge, UK: Cambridge University Press.
Teh, Robert, Thomas J. Prusa, and Michele Budetta. 2009. Trade Remedy Provisions in
Regional Trade Agreements. In Regional Rules in the Global Trading System, edited by Antoni
Estevadeordal, Kati Suominen, and Robert Teh, 166–249. Cambridge, UK:  Cambridge
University Press.
Tinbergen, J. 1954. International Economic Integration. Amsterdam: Elsevier.
Trebilcock, Michael J., and Robert Howse. 2005. The Regulation of International Trade. 3rd ed.
London: Routledge.
Viner, J. 1950. The Customs Union Issue. New York: Carnegie Endowment for International
Peace.
Von Stein, Jana. 2005. Do Treaties Constrain or Screen? Selection Bias and Treaty Compliance.
American Political Science Review 99 (4): 611–622.
World Bank. 2005. Global Economic Prospects:  Trade, Regionalism, and Development.
Washington, DC: World Bank.
World Bank. 2000. Trade Blocs. Oxford: Oxford University Press.
World Trade Organization. 2011. The WTO and Preferential Trade Agreements:  From
Co-existence to Coherence (World Trade Report 2011). Geneva: WTO.
Yeats, A. J. 1998. Just How Big Is Global Production Sharing? World Bank Policy Research
Working Paper No. 1871. Washington, DC: World Bank.
 Chapter 20

W TO M em be rsh i p

C h risti na L. Davi s an d M e re dith  W ilf

Introduction

Those who created the post–World War II trade regime sought to establish rules
that would prevent the kind of breakdown of the economic order that occurred in
the 1930s. Secretary of State Cordell Hull of the United States led in the commitment
to establish free trade as the basis for peaceful international relations. The General
Agreement on Tariffs and Trade (GATT), adopted in 1947, emerged as the compromise
outcome to reduce trade barriers and promote growth of world trade.1 The refusal
of the United Kingdom to end the imperial preference system and the reluctance of
the US Congress to accept ambitious institutional commitments for the negotiated
International Trade Organization resulted in the GATT serving as the core institution
that would govern trade for the next forty-eight years. Despite its origin as a tempo-
rary arrangement, the GATT oversaw an era of rapidly expanding trade. It evolved
as an institution by means of members negotiating more tariff reductions and new
rules in trade rounds while using panels of diplomats in a form of dispute settlement
process that helped avoid trade wars when disagreements arose among members.
The creation of the World Trade Organization (WTO) in 1995 further broadened the
scope of agreement and increased the strength of the existing, legalized enforcement
process.
Whether the trade regime is responsible for the phenomenal postwar growth of trade
is hotly debated, as world trade volume and trade as a percent of country’s gross domes-
tic product (GDP), a standard measure of a country’s trade openness, has grown steadily
among both members and nonmembers (see Figure 20.1). Nevertheless, the GATT/WTO
is widely credited with helping states to expand trade. Members choose to resolve major
trade disputes through the regime’s adjudication process and have largely complied with
its rulings. In line with the motivation of leaders at the founding of the institution, major
world economic downturns have not led to rampant protectionism or to a breakdown of
the rules.
 Christina L. Davis and Meredith Wilf    381

1.0
Members
Nonmembers

0.8

0.6

0.4

0.2

0.0

1950 1960 1970 1980 1990 2000 2010

Fig.  20.1  Trade Openness and Membership.*

*Mean value of trade openness (imports plus exports as percent of GDP) for GATT/WTO members and all other
countries. In any given year, members and nonmembers enjoy similar levels of trade openness.

As states manage incentives to pursue mercantilist policies against broader gains

from free trade, power and interest underlie their support of the regime. But members’
uncertainty about future preferences and economic circumstances has led them to adopt
an institutional design with flexibility for escape from commitments, discretionary
terms over new member accession, and decentralized enforcement. Far from high
delegation to a supranational organization, the GATT/WTO remains a largely member-
driven organization. Therefore it is critical to understand the conditions that explain
membership. What are the gains from membership that lead countries to join?
Membership in the regime has grown from 23 to 159, from 23 percent of all countries in
1948 to 80 percent of all countries in 2012 (see Figure 20.2). The Ministerial Conference
in December 2013 approved the accession of Yemen to become the 160th member, and
twenty-three other countries are currently in negotiations to join. Clearly, states view
advantages from membership such that most seek entry to the organization. At the
same time, this member expansion presents problems as the increasingly heterogeneous
preferences of the broad membership prevent agreement, as seen in the latest trade
negotiating round. Why were major powers willing to grant the expanded membership
that brought forth this scenario?
This chapter begins with a brief overview of the literature on why states benefit
from the trade regime. The theories about regime formation and empirical research
382   WTO Membership

100

GATT/WTO members as percent of all countries 80

60

GATT WTO
40

20

1948 1955 1965 1975 1985 1995 2005 2012

Year

Fig. 20.2  GATT/WTO Membership Growth: Expanding Membership of the Trade Regime.*
*Growing share of countries that have become members. “Member” is defined as any state that has completed
formal accession to GATT/WTO.

on the effects of membership provide insights into the motivation of states to join
the trade regime. Next we examine the process of joining the regime and explain
selection patterns that result from both the decisions of countries to apply and deci-
sions of the membership to accept new members. The final section summarizes the
discussion.

The Benefits of Regime Membership

Several explanations—each placing a different emphasis on the role of power, prefer-
ences, institutions, and social purpose—have been given for the establishment and
evolution of the multilateral regime to support free trade. These theories also highlight
competing reasons for states to join the regime. One view is that states simply join as
an act that follows the lead of the dominant state. Alternatively, states join because they
need the regime as a tool for coordinating behavior or share principled beliefs with
other members.
 Christina L. Davis and Meredith Wilf    383

Why Form a Multilateral Trade Regime?

When the state system is characterized by a highly asymmetrical distribution of power,
the dominant hegemon will support free trade that will enhance its own income and
power over other states (Krasner 1976). Hegemonic states build a framework of rules to
maintain stable order. Kindleberger (1986) blamed the US failure to ascend to this role in
the 1930s as the source of breakdown in the economic order. When the United States did
rise to the leadership position and shape the postwar trading regime, it closely adapted
the rules to its interests. For industries in which the United States sought exceptions to
free trade—such as in agriculture—liberalization demands were lenient. Further, excep-
tions to free trade were crafted around US domestic laws. As the US economic structure
shifted to become more service and technology based in the 1970s, it pushed to expand
the regime’s scope to include service trade and intellectual property rights, beginning in
the Tokyo Round. Some see the WTO as an “American creation” (Gilpin 2001, 222). In
his book on hierarchy in international relations, Lake (2009) emphasizes the contrac-
tual relationship of the United States with subordinate states and offers membership in
the GATT/WTO as an example of how the United States exercises authority over other
states without the costs of coercive enforcement of its interests. The mechanism by which
the United States exerts this power is identified by Kim (2010) as the product-by-product
bargaining protocol that gives advantage to market size. The informal trade round nego-
tiation process has allowed the United States to shape rule making on an ongoing basis
through ad hoc threats of exclusion or exit that would leave other states worse off if they
lost access to the large US market (Steinberg 2002; Stone 2011). To the extent that the
rules reflect US interests, delegation to a legal enforcement mechanism allows the United
States to avoid the hostility provoked by unilateral market access demands.
While none deny the influential role of the United States as first among equals in
the formation and evolution of the trade regime, others contend that US power alone
is insufficient to drive the multilateral trade regime. Even as US economic domi-
nance declined in the 1970s, the trade regime was sustained. This outcome arose from
the “stickiness” of the trade rules within an institution—once created, the institution
persisted—and also from the functional demand for the regime. From the perspective of
institutional theories, the GATT acts as a central bargaining forum that promotes more
efficient communication among states with repeated interactions. It supports linkage
and punishment strategies to uphold commitments to free trade (Keohane 1982; Martin
1992; Morrow 1994). One prominent explanation of the trade regime contends that large
states would be tempted to set unilateral, high tariffs in an effort to manipulate terms
of trade to their own advantage. The costs of such protective measures are forced onto
foreign exporters and consumers. Based on the core principles of reciprocity and non-
discrimination, multilateral agreement allows states to escape this “prisoner’s dilemma”
and to achieve greater welfare benefits through a low tariff equilibrium (Johnson
1953–1954; Bagwell and Staiger 1999). This logic extends beyond the largest states to also
explain the behavior of a wide range of countries (Broda, Limao, and Weinstein 2008).
384   WTO Membership

These theories present the WTO as a self-enforcing set of rules based on state interest.
The implication is that all states will gain from regime membership.
Solving the prisoner’s dilemma at the international level, however, does not elimi-
nate domestic pressures for protection. Even as the supposed leader of the move to free
trade, the United States resisted liberalization for many of the tariff lines that were politi-
cally too sensitive to cut (Goldstein and Gulotty 2014). Tariff reductions have gener-
ally occurred through slow reduction of rates over repeated rounds of negotiations, with
exceptions leaving high barriers to protect sensitive sectors such as agriculture and tex-
tiles. Herein lies a second rationale for joining the regime. Leaders who see gains from
free trade can use the trade rules to overcome domestic resistance. Governments face
a time inconsistency problem whereby short-term lobbying and electoral pressures
may induce protection that would harm long-term prospects for economic growth and
for free trade that benefits diffuse consumers (Staiger and Tabellini 1999; Maggi and
Rodriquez-Clare 1998). Issue linkage in the form of reciprocity, and package deals for
broad trade reform, mobilize exporters who benefit from reciprocity and help to build
domestic coalitions in support of liberalization (Gilligan 1997; Davis 2003). Through
design of flexible rules, it is possible to lower tariffs without going beyond what is sus-
tainable at home. Members address their uncertainties about preferences and future
trade shocks through use of escape clauses and tolerance of limited noncompliance as
defined by system rules (Rosendorff and Milner 2001; Rosendorff 2005).
Legal obligations also create constraints against protection. Jackson (1997) portrays
the trade regime as changing the nature of the interaction among states from power
politics toward the rule of law. The relatively high levels of precision in agreements
and use of third-party dispute settlement make the trade regime stand out as having
greater legalization than other institutions (Goldstein et al. 2000). This legitimacy has
constrained even the most powerful states—including the United States—by making
unilateral trade threats less effective (Schoppa 1997; Pelc 2010). Rule of law is upheld
not only by the empowerment of specific configurations of domestic actors, but also by
shared norms. The normative basis to the regime has allowed it to withstand many chal-
lenges and to evolve so that even as rules carved out exceptions to free trade—due to the
kinds of political exigencies noted above—members’ shared purpose to pursue liberal
trade supported continuity of the regime (Ruggie 1982). The establishment of the WTO
in 1995 was the next step of delegation by members toward a centralized trade regime
with a more robust legal enforcement mechanism and a new trade-policy review pro-
cess for peer monitoring.

Do Members Enjoy Higher Trade Levels?

Andrew Rose (2004) was the first to empirically test the proposition that the GATT
facilitated postwar trade growth. Analyzing trade at the dyadic level and using a grav-
ity model to control for factors that affect states’ natural affinities to trade, he surpris-
ingly found that pairs of states that are members of the GATT/WTO did not seem to
 Christina L. Davis and Meredith Wilf    385

enjoy higher trade than pairs of nonmember states.2 He concluded that member states
do not enjoy significant trade increases, with the implication that the trade regime may
not actually fulfill its basic goal to increase trade for members.
Judith Goldstein, Douglas Rivers, and Michael Tomz (2007) highlight the importance
of the GATT’s de facto practice that granted most-favored nation (MFN) benefits to
entire categories of nonmember states—specifically, colonies and former colonies—as
if they were regime members. In contrast to Rose’s focus on formal members, Goldstein,
Rivers and Tomz argue that it is more appropriate to compare regime “participants”—
both formal members and states that received MFN multilateral trade benefits—to
“non-participants,” states that were nonmembers without MFN benefits. Re-running
Rose’s analysis using participation (instead of formal membership) and focusing on a
country’s exports within a directed dyad (instead of nondirected, average dyadic trade),
they found that regime participants do enjoy higher levels of trade than do nonpartici-
pants. They attribute a causal mechanism—a state’s access to export markets at MFN
tariff rates—to tie membership to trade levels, moving beyond Rose’s focus on the asso-
ciation between formal membership and trade levels.
Allee and Scalera (2012) posit an alternative causal mechanism—a state’s own level of
tariff liberalization—as the channel through which the GATT affects trade levels. States
that significantly liberalize tariffs upon joining the regime—determined by their type of
accession process—enjoy increased levels of trade compared to states that join the insti-
tution without significant trade-policy liberalization.
The many studies debating trade regime effects on trade use slightly different meth-
ods and units of analysis. Rose uses nondirected dyads (where the dependent variable
is the average of both states’ bilateral trade), but subsequent research has largely used
directed dyads to model each state’s individual bilateral trade level. Similarly, Rose uses
year fixed effects, while other studies use year and dyad fixed effects. Subramanian and
Wei (2007, 158) argue that dyadic fixed effects capture the multilateral resistance of each
country—a key assumption of the gravity model—and therefore including dyadic fixed
effects provides more accurate estimates. Imai and Kim (2013), who examine the debate
with specific attention to fixed effects methodology, confirm that empirical results are
highly model dependent. Their analysis finds that, as a result of greater across-country
than across-year variation in membership, dyadic fixed effect specifications yield more
volatile results of WTO membership effect than do year fixed effect specifications. That
results are highly model dependent means that substantive researchers must use theory
to justify the assumptions underlying their chosen specification.
There is some consensus that the effects of membership and participation on trade are
greatest for developed countries and for trade in manufactured products. Subramanian
and Wei (2007) decompose the effects on imports versus exports, level of country devel-
opment, and country trade profile; they find that significant trade benefits accrue to
industrial countries. Dutt, Mihov, and Van-Zandt (2011) find that the WTO influences
extensive product margin of trade rather than intensive margin; members increase
the goods that they trade with each other rather than trading higher volumes of the
same goods. Helpman, Melitz, and Rubinstein (2008) show that nearly 50 percent of
386   WTO Membership

possible dyadic ties are lost if ties between states that do not trade with each other at all
are excluded from analysis. However, studies that incorporate these new insights into a
reexamination of the debate continue to arrive at varying conclusions. For example, Liu
(2009) and Chang and Lee (2011) find WTO membership associated with large boosts
to bilateral trade, while Eicher and Henn (2011) and Barigozzi, Fagiolo, and Mangioni
(2011) find little effect of membership on trade. Using a network model of trade, De
Benedictis and Tajoli (2011) find that WTO members are closer trading partners than
are non-WTO members.
Beyond direct linkages between membership and trade, authors focus on how trade
benefits accrue to developed and developing country trade regime members. Gowa and
Kim (2005) show that trade benefits accrued to the five largest member states at the time
of GATT establishment: the United States, Great Britain, Germany, France, and Canada.
While many studies emphasize benefits of the regime accruing to developed countries,
others emphasize benefits for developing countries. Developing country members who
appear to benefit most from the regime are those who have alliances such that the trade
rules reinforce concessions granted to them on the basis of a strong political relationship
(Gowa 2010). In contrast, Carnegie (2014) emphasizes the value of the regime to those
states that have weak political ties, because these states need the regime as protection
against having trade interests held up in exchange for foreign policy concessions.
Other scholars analyze whether the trade regime provides stability of trade as dis-
tinct from the volume of trade. This line of reasoning follows closely the logic within
the GATT that even a high fixed tariff is to be preferred over discretionary adminis-
trative measures. Mansfield and Reinhardt (2008) emphasize sustainable trade flows.
They propose that members may seek lower trade volatility rather than higher trade lev-
els and show that membership in the GATT/WTO lowers trade-policy volatility and
trade-export volatility. Their finding counters Rose (2005), who finds no significant dif-
ference in trade volatility of members and nonmembers. As there is no standard method
to calculate trade volatility, Mansfield and Reinhardt attribute the conflicting findings
to the differences in how their study operationalizes the dependent variable. They apply
four separate measures of annual trade volatility, in contrast to Rose’s use of one measure
based on a country’s standard deviation of trade over twenty-five-year panels. Dreher
and Voigt (2011) examine the impact of membership in international organizations on
country risk ratings and find that the GATT/WTO has a statistically significant impact
on lowering perceived risk by investors. In summary, membership—both formal and
informal—is often used to identify the regime’s effects on trade flows, but the literature
has not reached a consensus. There is need for further research to reach more conclusive
answers to this basic question.

Do Members Benefit from Multilateral Enforcement?

A strong legal framework that underlies third-party dispute settlement is a central ben-
efit of membership. All members of the GATT/WTO have standing to file complaints
 Christina L. Davis and Meredith Wilf    387

under the Dispute Settlement Understanding when they allege that another member
has a policy in violation of the agreement. The process begins with the filing of a formal
complaint, which initiates a consultation period. If settlement is not achieved through
consultation, the complainant may request a panel of third parties to rule on the legal
status of the challenged policy. This process has remained largely consistent across both
the GATT and WTO, although reforms of the Uruguay Round Agreements codified a
more robust process. Key reforms included the automatic right to panel and the intro-
duction of an appellate body of standing judges to hear appeals as an additional legal
layer. The dispute process offers enforcement authority that is important to resolve the
specific disputes and to deter future violations.
Within the set of cases filed under the dispute settlement process, compliance with
consultative agreements and panel rulings has been generally good. Hudec (1993), in
the first attempt to tally dispute outcomes, found that the large majority of the early
GATT panels resolved the dispute. Busch and Reinhardt (2003, 725) find that GATT/
WTO disputes produce substantial concessions in 50 percent of cases and partial con-
cessions for another 20 percent of cases. Interest in upholding the overall credibility of
the system of rules leads countries to comply with most rulings (Kovenock and Thursby
1992; Jackson 1997; Hudec 2002).3 The director of the WTO legal affairs division, Bruce
Wilson, acclaims members specifically for high compliance with rulings (Wilson 2007).
Further, the direct costs of authorized retaliation and the potential reputational costs
from noncompliance increase the incentives to cooperate (Bagwell and Staiger 1999,
2002; Maggi 1999).
Critics argue that developing countries benefit less from multilateral enforcement
than other members, because they lack legal capacity and retaliatory power. Evidence
from filed complaints suggests that developing countries gain fewer trade benefits from
winning cases (Bown 2004a, 2004b), suffer from weak legal capacity (Busch, Reinhardt,
and Shaffer 2009), and may be less willing to impose countermeasures if the defendant
refuses implementation (Bagwell, Mavroidis, and Staiger 2007). However, these stud-
ies are motivated by the question of whether a developing country WTO member that
files a legal complaint would get more from the system if it had more resources, which
is different from asking whether the developing country would get a better outcome in
the absence of the WTO dispute settlement process. The latter is the relevant question
when considering the benefits of membership, which provides access to WTO dispute
settlement. Davis (2006) compares the negotiation process for Peru and Vietnam in two
“all else being equal” disputes over fish labeling policies based on nearly identical legal
terms on technical barriers to trade. Peru used its rights as a WTO member to file a legal
complaint at the WTO against Europe, while Vietnam pursued bilateral negotiations
with the United States because it had not yet completed the WTO accession process. The
divergent outcomes are quite stark—Peru won concessions, while Vietnam got nothing.
Officials in Vietnam saw this case as further motivation to continue its negotiations to
join the WTO. Once Vietnam joined the WTO, it used the dispute settlement process to
bring two cases against the United States. Small countries would do even better if they
were larger and if institutions were revised to auction countermeasures or to implement
388   WTO Membership

collective retaliation. Yet this does not detract from the reality that WTO dispute settle-
ment under the multilateral trade regime offers small states greater leverage against large
states than they would otherwise have outside the regime in asymmetrical negotiations.
To assess the effectiveness of dispute settlement, it is important to include counter-
factual outcomes of trade disputes that were not addressed through adjudication. Davis
(2012) compares outcomes of similar trade barrier disputes raised in negotiations to
those adjudicated through a WTO complaint. On average, barriers that were also chal-
lenged through a WTO complaint were more likely to show progress toward removal of
the trade barrier than those that were not subject to a formal WTO complaint.
Others challenge whether “victory” in a dispute matters in the sense of bringing
observable changes to trade flows. The findings of Goldstein, Rivers, and Tomz (2007),
who show that nonmember participants experience the same or even larger trade ben-
efits from the regime than formal members, imply that the dispute system does not play
a large role in protecting trade flows; nonmember participants shared all rights with
formal members, with the exception of the right to use the dispute settlement process. In
recent work, Chaudoin, Kucik, and Pelc (2013) and Hofmann and Kim (2013) find little
evidence that a WTO ruling produces increased exports to the market that has been
found in violation. If true, this raises the question of why states use the dispute process. It
may be that they are more interested in political signals sent by dispute actions to actors
both at home and abroad. Other work, however, does find a positive effect between dis-
pute settlement and increased trade flows (Bechtel and Sattler forthcoming).
Even though all members have a right to use the dispute system, the structure of gov-
ernment institutions influences how frequently countries use it (Rickard 2010; Davis
2012). Governments push harder in dispute actions on behalf of key interest groups
(Fattore 2012; Rosendorff, this volume), and complainants select the timing of disputes
against the United States during election years, which could bring more pressure for
compliance (Chaudoin forthcoming). Looking at trade partners and products beyond
those directly at stake in the specific dispute, governments may seek a deterrent effect,
as winning one case could prevent future disputes. Deterrence of cheating counts as a
benefit that accrues to members from their access to a legal dispute system.
There are many quantities of interest for assessing the effectiveness of the trade
regime. Any single study highlights one aspect of how the regime may benefit members,
such as increasing bilateral trade with other members or resolving trade disputes. Each
of these outcomes may have varying importance to different members. Where one state
seeks more trade, another wants to avoid negative trade linkages with foreign policy or
excessive volatility. The average effect of membership may be less relevant than the effect
conditional on characteristics of states that influence what they need from membership.
A fuller appreciation of regime benefits would need to take into account all dimensions
across which the regime serves its members.
Different levels of analysis—beyond dyads and individual states—suggest broader
trade regime effects. For example, Oatley (2011) emphasizes that an institution such as
the trade regime is a systemic factor that affects outcomes, because political decisions
on a range of issues are made conditional upon the context provided by the multilateral
 Christina L. Davis and Meredith Wilf    389

regime. Another approach is to compare how states respond to crises. Davis and Pelc
(forthcoming) suggest that the ability of the regime to restrain protection must be
understood within the context of different types of economic crises that give rise to new
demands for protectionist policies. In contrast to local crises in which only one or a few
states have incentives to increase trade protection, during pervasive crisis—when many
members simultaneously face hard economic times—the trade regime establishes a focal
point and coordinates a collaborative response to limit the increase in protection. Cao
(2012) analyzes trade flows within a network analysis and establishes how trade simi-
larity, not trade flows, is associated with domestic regulatory convergence. Economic
researchers have also begun to analyze trade at the network level (e.g., De Benedictis
and Tajoli 2011). These new perspectives that move beyond dyadic analysis emphasize
the importance for scholars of specifying clear counterfactuals.

The Problem of Using Membership

to Identify Regime Effectiveness
GATT/WTO membership status is central to empirical analyses about the effects of the
trade regime. In the research discussed in the previous section, membership is used as
an identification strategy whereby the effect of membership is inferred from analysis
that compares policy outcomes for members and nonmembers. The average effect of
membership is taken to reflect the regime’s effect as a whole. It is necessary to flag the
limitations that arise from using membership to identify institutional effects.4
This approach assumes that membership is exogenously given and has no spillover
effects. These two assumptions may be problematic. First, selection bias is a concern,
given that the process generating membership in the trade regime is not random. Thus,
these studies capture the different performance between members and nonmembers
but neglect the prior question of what leads countries to join or stay outside of the
trade regime in the first place. Second, spillover effects are likely because the policy
changes related to being a member of the trade regime are often comprehensive. For
example, domestic laws that have been reformed to enhance protection of intellectual
property rights will offer this protection equally to members and nonmembers. Even
tariff changes are often implemented through MFN policies without exclusively privi-
leging members. Other functions of the regime may induce stability at the systemic level
through reducing the volatility of trade flows and supporting open markets during eco-
nomic crises.
Theories about regime formation and the effects of membership must take into
account the endogenous nature of institutions, which are created to serve the interests
of members (Martin and Simmons 2001). Downs, Rocke, and Barsoom (1996) high-
light that if states accept rules that already reflect their preferences, compliance is not
a meaningful concept. From this perspective, liberalization is driven by an exogenous
process that leads to preference changes. Hence states should pursue gradual expansion
of membership to include new members as they become more liberal (Downs, Rocke,
390   WTO Membership

and Barsoom 1998). Theories about why states join the institution need to be connected
to the empirical debate about the extent to which the trade regime has been responsible
for liberalization of trade flows and resolution of trade disputes.

Decisions to Join
Membership in any regime is voluntary. States must decide to join an institution, and
the current members must be willing to accept the applicant. The organizational rules of
the GATT and WTO support an inclusive mandate to achieve broad membership. Any
country or territory with autonomous control over its trade policy is eligible to join. This
section reviews the process of becoming a member and theories about the political and
economic variables that shape when countries join. A brief case study of Mexico’s acces-
sion to the GATT illustrates the factors that can influence decisions to join the regime.

The Accession Process

Under the GATT, two accession processes existed. The standard accession process out-
lined in GATT Article 33 was decentralized, with most details of accession delegated to
the working party, a group of existing contracting parties that opted-in to oversee the
accession process for a given applicant. The working party and applicant country held
consultations and bargained over an initial tariff schedule, which was then presented to
the GATT General Council. GATT Article 33 formally required a two-thirds majority
vote to adopt the arrangement, although in practice few votes occurred. Nonmember
countries meeting certain criteria, mostly former colonies, were eligible to join the
GATT under a simpler accession process outlined in GATT Article 26, which granted
membership on the basis of sponsorship (i.e., the written support) of its former colo-
nial power and a note of intention to join the regime. Upon receipt of these notices,
the GATT granted immediate membership, bypassing the negotiations required
under GATT Article 33. This lowered the entry cost for countries eligible to accede under
GATT Article 26 as compared to most countries that acceded under GATT Article 33.5
Since the establishment of the WTO in 1995, a single process has governed accession.
The WTO Article 12 process is formally the same as the GATT Article 33 accession pro-
cess, but nonmember countries acceding through WTO Article 12 have a longer list of
required agreements to negotiate (including GATS, TRIMS, and TRIPS) before gaining
WTO membership. Further, trade rounds under the WTO are no longer associated with
negotiating entry of applicants, which means that there are few opportunities for reci-
procity during WTO Article 12 negotiations. Pelc (2011) highlights how asymmetrical
bargaining shapes WTO accession negotiations, as applicants are forced to make con-
cessions while members do not concurrently change their commitments. All applicant
countries that were not contracting parties upon transition from the GATT to the WTO
in 1995 must complete the WTO Article 12 accession process. Observers have noted the
 Christina L. Davis and Meredith Wilf    391

prevalence of WTO-plus commitments, whereby the accession protocol of an applicant

country goes beyond the obligations of WTO member countries. Although the rules for
accession have not changed greatly between the GATT and WTO, there is a widespread
perception that the WTO accession process is more demanding than the GATT acces-
sion process.
The accession process is a key moment of leverage when regime members can dictate
entry terms for nonmembers. Jones (2009) establishes that the length of WTO accession
negotiations has been increasing over time. As the scope of rules has become broader,
new members must accept more obligations. The higher cost of being outside the grow-
ing regime also allows members to demand more of entrants. Kim (2010) argues that
those who join the trade regime under WTO rules enjoy higher increases in trade than
earlier entrants who joined under the GATT rules. Allee and Scalera (2012) demon-
strate that states given a free ride to enter the regime without any negotiation, an acces-
sion option under the GATT available to the former colonies of members, receive the
smallest trade gains. The more rigorous negotiation demands placed on other states
bring more liberalization concessions that also lead to more trade gains. The demands
placed on entrants also vary as a function of their economic size. Pelc (2011) identifies
middle-income states—those not protected by norms of special and differential treat-
ment and those whose export markets are not large enough to have strong bargaining
power—as those that offer the most liberalization concessions during accession nego-
tiations. Neumayer (2013) finds that trade interests determine which members opt into
the GATT/WTO working party that negotiates accession terms. Members recognize the
risks of competition from expanding membership and use the accession window to
strategically delay accession and negotiate terms that will protect their key interests.
This flexibility surrounding accession becomes a source of informal institutional power
(Stone 2011).

The Economic and Political Determinants of Membership

New research directly models the economic, political, and legal factors that create dif-
ferential costs and benefits for a country to join the trade regime. The interests of the
members ultimately determine entry; these interests reflect not only their economic
gains from an expanding membership in the trade regime, but also their broader foreign
policy objectives. Applicants weigh a mix of economic risks and benefits from expand-
ing market access and competition and may see the rules as a guarantor against arbi-
trary sanctions. Thus on both the supply and demand sides, one must examine multiple
dimensions.
The largest draw of membership is market access. For states dependent on trade
with GATT/WTO members, entry will lower their trade costs. In an innovative study,
Copelovitch and Ohls (2012) examine the former colonies of members who were
granted “free pass” entry under Article 26 provisions of GATT accession. Since these
countries did not face any bargaining or policy reform conditions that could act as a
392   WTO Membership

barrier to entry, the study highlights the demand side of membership. They find that
high trade dependence with members was an important factor in bringing some of
these countries to join early after independence. In contrast, a country such as Vietnam
waited until later—after expansion of the agreement terms under the WTO made acces-
sion more difficult—before it decided to join.
Ironically, the ability to protect sensitive industries under the trade regime is one
important condition for accession into the institution. Kucik and Reinhardt (2008)
provide evidence that states with stronger antidumping laws are more likely to join
the trade regime and to liberalize more when they join. As described by Rosendorff
(this volume), flexibility measures have been essential to allow free trade commitments
by countries that want to be responsive to domestic political actors. A heterogeneous
group of countries with mixed preferences on free trade has joined the institution
because it does not demand rigid commitments to fully open markets. This in turn
presents challenges for cooperation. India was among the founding members of the
GATT and has also been among the most reluctant to open its markets, to the point
where its demand for a special safeguard measure to protect its agricultural sector was
given as one reason for the collapse of the Doha Round ministerial meeting in 2008.
Former US Trade Representative Susan Schwab commented on the long deadlock in
WTO negotiations: “No future multilateral negotiation will succeed, however, without
addressing the very real differences in economic strength, prospects, and capabilities
within the so-called developing world. It is worth recalling that one of the WTO’s most
important characteristics is the inclusion of these developing economies in governance
and decision-making from its origins as the General Agreement on Tariffs and Trade in
1948” (2011, 117).
In addition to trade levels, geopolitical alignment and democracy appear to be
equally important for determining membership. The GATT, negotiated as part of the
ITO at the Bretton Woods conference, brought together twenty-three countries under
the leadership of the United States and United Kingdom. Countries that subsequently
joined were colonies and friends of these countries. The Cold War made it less likely
that the Soviet Union or its allies would join (although the Soviet Union was invited to
join the GATT, and Czechoslovakia was a founding member). Davis and Wilf (2014)
argue that geopolitical ties between applicants and members provide a basis of common
interest that draws states rapidly into the regime. The discretionary accession process
has allowed members to use the GATT/WTO as a tool of statecraft, raising and lower-
ing the bar for nonmember states to enter the organization. By modeling the timing
of nonmember application—instead of formal accession year—Davis and Wilf avoid
the endogeneity problem associated with many states’ long accession processes. Their
statistical analysis of applicants shows that the allies of members and democratic states
are significantly more likely to join the regime. In some cases, members encourage the
application of an ally or dependency, such as US sponsorship of Japan’s application
in 1952 or the US role in encouraging Iraq and Afghanistan to apply in 2004. When
East European countries were being courted as part of foreign policy strategy, Poland,
Hungary, and Romania were granted lenient accession terms so that they could join
 Christina L. Davis and Meredith Wilf    393

in the 1960s without substantial tariff liberalization or reforms of their state-planned

economies.

Mexico’s Debate over Membership

The case of Mexico illustrates a number of factors that can impact membership deci-
sions. Economic preferences shape whether governments are willing to engage in trade
liberalization, and both domestic political interests and the structure of the economy
impact views about economic benefits from joining the regime. In addition, broader
political concerns related to nationalism and relations with other countries shape the
calculus of leaders.
Reluctance to embrace free trade delayed application and accession by Mexico.
The first rejection of membership occurred in 1947, when the Mexican finance minis-
ter announced that the country would not ratify the Havana Charter to establish the
International Trade Organization or participate in the GATT because the organiza-
tion would ruin national industries (Ortiz Mena 2005, 217). Import substitution poli-
cies led to declining shares of trade as percent of GDP throughout the 1950s and 1960s
and at first delivered strong economic growth. With economic downturn in the 1970s,
however, the government began to experiment with different economic policies. Under
pressure from the International Monetary Fund after receiving funds to help address
the 1976 balance of payments crisis, the Mexican government launched a series of
reforms, including steps to liberalize trade policies (Ortiz Mena 2005, 219). It appeared
that Mexico was turning from import substitution to export promotion, and this shift
in policy orientation explained the move to apply for GATT membership in 1979. The
working group on accession met five times to produce an accession protocol that gave
Mexico wide latitude to pursue its development strategies (Story 1982, 772–773).
At this point, Mexican President Lopez Portillo engaged in a remarkably open
domestic debate on the merits of GATT accession. The deliberation process, with input
from economic analysts and consultation with industry and labor groups, was covered
in front-page media commentary. Ortiz Mena (2005, 221–222) notes the irony that this
debate went forward entirely separate from the accession offer, which was negotiated
with extremely favorable terms for Mexico. Rather than specific industries lobbying
against import competition, attention focused on economic sovereignty and the devel-
opment model writ large. The final vote of the cabinet in March 1980 opposed accession,
and the government informed the GATT that it would postpone its application. Mexico
had decided to remain committed to state intervention and oil exports. The push for
economic reform was set aside as rising oil prices raised the prospect for a return to high
growth.
The United States had actively lobbied Mexico on the need for it to join the GATT.
Story (1982) contends that the desire to assert independence from the United States per-
versely pushed Portillo to decline joining. Portillo’s relations with the Carter adminis-
tration were poor at this time. Opponents of the GATT in Mexico portrayed accession
394   WTO Membership

as a move toward dependence on the United States. Portillo announced the postpone-
ment of GATT accession on the 42nd anniversary of Mexico’s expropriation of US
oil companies, framing rejection of the GATT in terms of Mexican nationalism and
anti-American policy (Story 1982, 775).
Caught by surprise when Mexico postponed its application in 1980, the United
States began to impose higher countervailing duties against imports from Mexico
(Lara-Fernandez 1987, 20). Export subsidies used as part of Mexican sectoral promo-
tion strategies were seen as inimical to US interests. Friction between the countries was
resolved under desperate circumstances. When the collapse of oil prices and spiral-
ing cost of servicing debt obligations threw the Mexican economy into turmoil in the
early 1980s, the United States used the Baker plan as leverage to encourage economic
reforms in Mexico, including trade liberalization. The 1982 National Development Plan
launched structural reforms to restore stable growth and improve competitiveness.
Mexico renewed its application to the GATT in 1985, stating its desire to improve
market access and reduce its reliance on petroleum exports.6 As domestic reforms had
begun a process to liberalize its own markets, many urged the government to bargain for
access rather than simply adopt unilateral liberalization (Ortiz Mena 2005, 225). Yet the
government had not made a complete reversal in course. It only agreed to the accession
protocol when assured that it would be allowed to continue sectoral promotion policies,
exclude sensitive agricultural products, and retain its pricing system.
Mexico applied after its economic and geopolitical alignment became more similar to
that of trade regime members. Its negotiations also displayed the willingness of mem-
bers to let Mexico in with low conditionality. The accommodating stance of members,
in combination with more interest in trade policy reform by Mexican leaders, led the
country to finally accept entry into the organization in 1986 after having twice declined
to join.

Conclusion

We have highlighted the role of membership in access to markets and dispute settle-
ment. The gains to members, however, remain contested within the literature. The role
of economic interests, geopolitical ties, and democracy in shaping member selection
means that one needs to be cautious about using membership to identify the effective-
ness of the institution. Furthermore, the commonly used dyadic framework neglects
broader systemic impact from the regime. More attention needs to be given to discov-
ering how membership benefits some countries in different ways from others or has
changed over time.
As the trade regime reaches nearly universal membership, the comparison of mem-
bers with nonmembers will become increasingly irrelevant. Indeed, the expansion of
the regime had already become so inclusive at the time of establishing the WTO that
 Christina L. Davis and Meredith Wilf    395

it is very difficult to compare the effect of membership during the WTO period and
during the earlier GATT period. By 1995, the year the GATT transitioned to the WTO,
128 countries (two-thirds of total countries in the world) had formally joined the
regime, and outsiders were newly independent former Soviet Union countries and an
unusual group of laggards. Only two nonmembers—China and Russia—individually
held greater than 1 percent of world trade. Since then 32 countries—including China in
2001 and Russia in 2012—completed the accession process under the WTO, 23 countries
are in the process of negotiating accession, and only a handful of countries are neither
members nor formal applicants in the process of negotiating membership (as of 2013).
The remaining outsiders are such a biased sample of countries that one could not gener-
alize from analysis of this group of states.
Instead, attention is turning to the WTO-plus agreements being concluded among
subsets of members. The proliferation of bilateral and regional agreements raises the
question of how these overlapping memberships impact the benefits of membership in
the WTO (Mansfield and Reinhardt 2003). Diversion of trade flows and fragmentation
of legal obligations could erode benefits, but a competitive liberalization process could
push broader and deeper commitments at both levels. A serious challenge lies in the
imbalance of power across negotiation and enforcement of rules. The large membership
of the WTO contributes to the stalled negotiations over new rules in the Doha Round,
while more limited membership allows countries to forge ahead in the rapid conclu-
sion of bilateral and regional trade agreements. Yet the asymmetrical nature of many
bilateral and regional trade agreements weakens their enforcement power, while the
multilateral system draws broader enforcement credibility from the wider membership.
Increasingly, progress to negotiate new rules occurs outside of the WTO, while enforce-
ment issues continue to be taken up largely within the WTO.

Notes
1. See Irwin, Mavroidis, and Sykes (2008) on the founding of the GATT.
2. Rose’s (2004) analysis includes multiple specifications of the model, and in his conclusion
he identifies a number of factors that could drive his finding. In cross-section analysis, he
found a positive effect of early trade rounds and a negative effect of later rounds. See Hicks
and Gowa chapters in this volume.
3. See Busch and Pelc, in this volume, for a full review.
4. See also Hicks, in this volume.
5. At the beginning of a wave of decolonization in 1960, the GATT determined that each
territory eligible to become a contracting party under the GATT Article 26 accession pro-
cess would gain “de facto’ status within the GATT while the territory deliberated about
whether or not to accede to the regime. Goldstein, Rivers, and Tomz (2007, 42) treat these
nonmember states as participants because de facto participants received MFN treatment
and could observe GATT proceedings. These states, however, could not participate in
decision making or dispute settlement and lacked status as formal members.
6. GATT, November 27, 1985, L/5919.
396   WTO Membership

References
Allee, Todd L., and Jamie E. Scalera. 2012. The Divergent Effects of Joining International
Organizations: Trade Gains and the Rigors of WTO Accession. International Organization
66 (2): 243–276.
Bagwell, Kyle, Petros Mavroidis, and Robert Staiger. 2007. Auctioning Countermeasures in the
WTO. Journal of International Economics 73: 309–332.
Bagwell, Kyle, and Robert Staiger. 1999. An Economic Theory of GATT. American Economic
Review 89 (1): 215–248.
Bagwell, Kyle and Robert Staiger. 2002. The Economics of the World Trading System. Cambridge,
MA: MIT Press.
Barigozzi, Matteo, Giorgio Fagiolo, and Giuseppe Mangioni. 2011. Identifying the Community
Structure of the International-Trade Multi-Network. Physica A 390 (11): 2051–2066.
Bechtel, Michael M., and Thomas Sattler. Forthcoming. What Is Litigation in the World Trade
Organization Worth? International Organization.
Bown, Chad. 2004a. Developing Countries as Plaintiffs and Defendants in GATT/WTO Trade
Disputes. World Economy 27 (1): 59–80.
Bown, Chad. 2004b. On the Economic Success of GATT/WTO Dispute Settlement. Review of
Economics and Statistics 86 (3): 811–823.
Broda, Christian, Nuno Limao, and David E. Weinstein. 2008. Optimal Tariffs and Market
Power: The Evidence. American Economic Review 98 (5): 2032–2065.
Busch, Marc, and Eric Reinhardt. 2003. Developing Countries and GATT/WTO Dispute
Settlement. Journal of World Trade 37 (4): 719—735.
Busch, Marc, Eric Reinhardt, and Greg Shaffer. 2009. Does Legal Capacity Matter? A Survey of
WTO Members. World Trade Review 8 (4): 559–577.
Cao, Xun. 2012. Global Networks and Domestic Policy Convergence: A Network Explanation
of Policy Changes. World Politics 64 (3): 375–425.
Carnegie, Allison. 2014. States Held Hostage: Political Hold-Up Problems and the Effects of
International Institutions. American Political Science Review. 108 (1): 54–70.
Chang, Pao-Li, and Myoung-Jae Lee. 2011. The WTO Trade Effect. Journal of International
Economics 85 (1): 53–71.
Chaudoin, Stephen. Forthcoming. Information Transmission and the Strategic Timing of
Trade Disputes. International Organization.
Chaudoin, Stephen, Jeff Kucik, and Krzysztof Pelc. 2013. Do WTO Disputes Actually Increase
Trade? Unpublished manuscript: University of Pittsburgh, University College London and
McGill University.
Copelovitch, Mark S., and David Ohls. 2012. Trade, Institutions, and the Timing of GATT/
WTO Accession in Post-Colonial States. Review of International Organizations 7 (1): 81–107.
Davis, Christina. 2003. Food Fights Over Free Trade: How International Institutions Promote
Agricultural Trade Liberalization. Princeton, NJ: Princeton University Press.
Davis, Christina. 2006. Do WTO Rules Create a Level Playing Field for Developing Countries?
Lessons from Peru and Vietnam. In Negotiating Trade: Developing Countries in the WTO
and NAFTA, edited by John Odell, 219–256. Cambridge, UK: Cambridge University Press.
Davis, Christina L. 2012. Why Adjudicate? Enforcing Trade Rules in the WTO. Princeton,
NJ: Princeton University Press.
Davis, Christina, and Meredith Wilf. 2014. Joining the Club? Accession to the GATT/WTO.
Unpublished manuscript, Princeton University.
 Christina L. Davis and Meredith Wilf    397

Davis, Christina L., and Krzysztof J. Pelc. Forthcoming. Cooperation in Hard Times: Self
Restraint of Trade Protection. Journal of Conflict Resolution.
De Benedictis, Luca, and Lucia Tajoli. 2011. The World Trade Network. World Economy 34
(8): 1417–1454.
Downs, George, David Rocke, and Peter Barsoom. 1996. Is the Good News about Compliance
Good News about Cooperation? International Organization 50 (3): 379–406.
Downs, George, David Rocke, and Peter Barsoom. 1998. Managing the Evolution of
Multilateralism. International Organization 52 (2): 397–419.
Dreher, Axel, and Stefan Voigt. 2011. Does Membership in International Organizations
Increase Governments’ Credibility? Testing the Effects of Delegating Powers. Journal of
Comparative Economics 39: 326–348.
Dutt, Pushan, Ilian Mihov, and Timothy Van-Zandt. 2011. Does WTO Matter for the Extensive
and the Intensive Margins of Trade? Unpublished manuscript, INSEAD.
Eicher, Theo S., and Christian Henn. 2011. In Search of WTO Trade Effects: Preferential Trade
Agreements Promote Trade Strongly, But Unevenly. Journal of International Economics
83: 137–153.
Fattore, Christina. 2012. Interest Group Influence on WTO Dispute Behaviour: A Test of State
Commitment. Journal of World Trade 46 (6): 1261–1280.
Gilligan, Michael. 1997. Empowering Exporters: Reciprocity, Delegation, and Collective Action in
American Trade Policy. Ann Arbor: University of Michigan Press.
Gilpin, Robert. 2001. Global Political Economy:  Understanding the International Economic
Order. Princeton, NJ: Princeton University Press.
Goldstein, Judith, and Robert Gulotty. 2014. America and Trade Liberalization: The Limits of
Institutional Reform. International Organization 68 (2): 263–295.
Goldstein, Judith, Miles Kahler, Robert O. Keohane, and Anne-Marie Slaughter.
2000. Introduction:  Legalization and World Politics. International Organization 54
(3): 385–399.
Goldstein, Judith, Douglas Rivers, and Michael Tomz. 2007. Institutions in International
Relations:  Understanding the Effects of the GATT and the WTO on World Trade.
International Organization 61 (1): 37–67.
Gowa, Joanne. 2010. Alliances, Market Power, and Postwar Trade:  Explaining the GATT/
WTO. World Trade Review 9 (3): 487–504.
Gowa, Joanne, and Soo Yeon Kim. 2005. An Exlusive Country Club: The Effects of the GATT
on Trade, 1950–94. World Politics 57 (4): 453–478.
Helpman, Elhanan, Marc Melitz, and Yona Rubinstein. 2008. Estimating Trade Flows: Trading
Partners and Trading Volumes. Quarterly Journal of Economics 123 (2): 441–487.
Hofmann, Tobias, and Soo Yeon Kim. 2013. Does Trade Comply? The Economic Effect(ivenes)
s of WTO Dispute Settlement. Unpublished manuscript, National University of Singapore.
Hudec, Robert. 2002. The Adequacy of WTO Dispute Settlement Remedies: A Developing
Country Perspective. In Development, Trade, and the WTO, edited by Bernard Hoekman,
Aaditya Mattoo, and Philip English, 81–91. Washington, DC: The World Bank.
Hudec, Robert. 1993. Enforcing International Trade Law: The Evolution of the Modern GATT
Legal System. Salem, NH: Butterworth.
Imai, Kosuke, and In Song Kim. 2013. On the Use of Linear Fixed Effects Regression Estimators
for Causal Inference. Unpublished manuscript, Princeton University.
Irwin, Douglas, Petros Mavroidis, and Alan Sykes. 2008. The Genesis of the GATT. Cambridge,
UK: Cambridge University Press.
398   WTO Membership

Jackson, John H. 1997. The World Trading System: Law and Policy of International Economic
Relations. 2nd ed. Cambridge, MA: The MIT Press.
Johnson, Harry. 1953–1954. Optimum Tariffs and Retaliation. Review of Economic Studies
21: 142–153.
Jones, Kent. 2009. The Political Economy of WTO Accession. World Trade Review 8 (2):
279–314.
Keohane, Robert. 1982. The Demand for International Regimes. International Organization 36
(2): 325–355.
Kim, Soo Yeon. 2010. Power and the Governance of Global Trade: From the GATT to the WTO.
Ithaca, NY: Cornell University Press.
Kindleberger, Charles P. 1986. The World in Depression. Berkeley: University of California Press.
Kovenock, Dan, and Marie Thursby. 1992. GATT, Dispute Settlement and Cooperation.
Economics and Politics 4 (2): 151–170.
Krasner, Stephen D. 1976. State Power and the Structure of International Trade. World Politics
28 (3): 317–347.
Kucik, Jeffrey, and Eric Reinhardt. 2008. Does Flexibility Promote Cooperation? An
Application to the Global Trade Regime. International Organization 62 (3): 477–505.
Lake, David. 2009. Hierarchy in International Relations. Ithaca, NY: Cornell University Press.
Lara-Fernandez, Martha. 1987. Mexico’s Accession to the GATT: A Turning Point in Mexican
Foreign Trade Policy and Its Possible Implications. Geneva: Institut universitaire de hautes
etudes internationales.
Liu, Xuepeng. 2009. GATT/WTO Promotes Trade Strongly:  Sample Selection and Model
Specification. Review of International Economics 17 (3): 428–446.
Maggi, Giovanni. 1999. The Role of Multilateral Institutions in International Trade
Cooperation. American Economic Review 89 (1): 190–214.
Maggi, Giovanni, and Andres Rodriquez-Clare. 1998. The Value of Trade Agreements in the
Presence of Political Pressures. Journal of Political Economy 106 (3): 574–601.
Mansfield, Edward D., and Eric Reinhardt, 2003. “Multilateral Determinants of Regionalism:
The Effects of GATT/WTO on the Formation of Preferential Trading Arrangements.”
International Organization 57(4): 829--862.
Mansfield, Edward D., and Eric Reinhardt. 2008. International Institutions and the Volatility of
International Trade. International Organization 62: 621–652.
Martin, Lisa. 1992. Interests, Power, Multilateralism. International Organization 46 (3):
765–792.
Martin, Lisa, and Beth Simmons. 2001. Theories and Empirical Studies of International
Institutions. In International Institutions: An International Organization Reader, edited by
Lisa Martin and Beth Simmons, 437–465. Cambridge, MA: The MIT Press.
Morrow, James. 1994. The Forms of International Cooperation. International Organization 48
(3): 387–423.
Neumayer, Eric. 2013. Strategic Delaying and Concessions Extraction in Accession
Negotiations to the World Trade Organization: An Analysis of Working Party Membership.
World Trade Review 12 (4): 669–692.
Oatley, Thomas. 2011. The Reductionist Gamble:  Open Economy Politics in the Global
Economy. International Organization 65: 311–341.
Ortiz Mena, Antonio. 2005. Mexico. In The World Trade Organization: Legal, Economic and
Political Analysis, Volume 3, edited by Patrick F. J. Macrory, Arthur E. Appleton, and Michael
G. Plummer, 217–247. New York: Springer.
 Christina L. Davis and Meredith Wilf    399

Pelc, Krzystof. 2010. Constraining Coercion? Legitimacy and its Role in U.S. Trade Policy,
1975–2000. International Organization 64: 65–96.
Pelc, Krzysztof J. 2011. Why Do Some Countries Get Better WTO Accession Terms Than
Others? International Organization 65 (4): 639–672.
Rickard, Stephanie. 2010. Democratic Differences: Electoral Institutions and Compliance with
GATT/WTO Agreements. European Journal of International Relations 16 (4): 711–729.
Rose, Andrew K. 2004. Do We Really Know That the WTO Increases Trade? American
Economic Review 94 (1): 98–114.
Rose, Andrew K. 2005. Does the WTO Make Trade More Stable? Open Economies Review 16
(1): 7–22.
Rosendorff, Peter. 2005. Stability and Rigidity:  Politics and Design of the WTO’s Dispute
Settlement Procedure. American Political Science Review 99 (3): 389–400.
Rosendorff, Peter, and Helen Milner. 2001. The Optimal Design of International Trade
Institutions: Uncertainty and Escape. International Organization 55 (4): 829–857.
Ruggie, John Gerard. 1982. International Regimes, Transactions, and Change:  Embedded
Liberalism in the Postwar Economic Order. International Organization 36 (2): 379–415.
Schoppa, Leonard. 1997. Bargaining With Japan: What American Pressure Can and Cannot Do.
New York: Columbia University Press.
Schwab, Susan. 2011. After Doha: Why the Negotiations are Doomed and What We Should Do
About It. Foreign Affairs 90 (3): 104–117.
Staiger, Robert, and Guido Tabellini. 1999. Do GATT Rules Help Governments Make Domestic
Commitments? Economics and Politics 2 (2): 109–144.
Steinberg, Richard. 2002. In the Shadow of Law or Power? Consensus-Based Bargaining and
Outcomes in the GATT/WTO. International Organization 56 (2): 339–374.
Stone, Randall. 2011. Controlling Institutions:  International Organizations and the Global
Economy. Cambridge, UK: Cambridge University Press.
Story, Dale. 1982. Trade Politics in the Third World:  A  Case Study of the Mexican GATT
Decision. International Organization 36 (4): 767–794.
Subramanian, Arvind, and Shang-Jin Wei. 2007. The WTO Promotes Trade, Strongly but
Unevenly. Journal of International Economics 72 (1): 151–175.
Wilson, Bruce. 2007. Compliance by WTO Members with Adverse WTO Dispute Settlement
Rulings: The Record to Date. Journal of International Economic Law 10 (2): 397–403.
 Chapter 21

Dispu te Set t l e me nt
in the  W TO

M arc L. Busch and K r z ysz tof J. Pe l c

Introduction

Not long ago, scholars had to justify studying litigation under the General Agreement
on Tariffs and Trade (GATT). After all, the institution was derided as a “court with
no bailiff.”1 Today, studying dispute settlement under the World Trade Organization
(WTO) is very much in vogue, for good reason. There is no other instance of as many
countries delegating as much decision making to an international legal body over the
enforcement of binding commitments. Since its creation in 1995, the WTO’s Dispute
Settlement Understanding (DSU) has developed a sophisticated body of jurisprudence,
leading many to agree with Bhala (1998–1999) that the institution features de facto bind-
ing precedent, meaning past rulings constrain current ones. Add to this that compliance
with rulings is high, and it is little wonder that the WTO gets so much attention. But
how far has the literature come since the old GATT days? And where should it go in the
future?
Recent scholarship has delivered some impressive insights. For example, while the lit-
erature once took it as a matter of faith that developing countries were at a disadvantage
in litigation because of their lack of power, it turns out that weak legal capacity—political,
legal, and economic resources—matters more (Busch, Reinhardt, and Shaffer 2009;
Davis and Bermeo 2009). Not only are the two variables only loosely correlated, the
solutions to addressing this disparity differ dramatically. The record also reveals no
disadvantage for developing countries in terms of winning verdicts or compliance
from defendants. Rather, they trail rich complainants in rates of settlement: develop-
ing countries not only lack legal capacity, but tend to file disputes against large wealthy
markets, which draw in other parties, making them harder to resolve in pretrial negotia-
tions (Busch and Reinhardt 2003). These insights are not what the GATT-era literature
expected.
 Marc L. Busch and Krzysztof J. Pelc    401

We submit, however, that for all of its advances, the literature needs retooling. Three
areas stand out. First, scholars have always struggled with the question of selection: Is
there something different about the cases that get filed, in relation to the universe of all
trade disputes? In the 1990s critics of GATT insisted that cases brought to Geneva were
likely less contentious, or “easy,” than those taken up as matters of high politics. This
view fit with the prevailing debate over institutions more generally, forcing students of
GATT to tease out implications of selection bias to show that these disputes mattered. In
recent years more involved statistical methods have been used in making the point, and
with some success, as relatively few pages are now devoted to the issue. Yet more can be
done. Most important, there are data on so-called specific trade concerns (STCs) raised
before the WTO’s committees on technical barriers to trade (TBT) and sanitary and
phytosanitary (SPS) measures. For example, on technical regulations, standards, and
conformity assessment, which include some of the most contentious regulatory politics,
the WTO’s TBT committee has handled 379 STCs, versus 47 dispute settlement cases
referencing this agreement. Similarly, the ratio of STCs to litigation on health and safety
standards is almost 10:1. These STCs can shed new light on selection issues, especially
because the data are coded according to the type of infraction alleged and whether there
was a resolution reported to the WTO. Some scholars are already pursuing this lead
(see, e.g., Horn, Mavroidis, and Wijkstrom 2013). GATT-era scholarship could not have
imagined this treasure trove.
Second, much of the criticism of the institution’s dispute settlement procedures
(DSPs) has consisted of normative claims about a lack of transparency and unequal
participation. In this telling, small countries are excluded not only from negotiations
during multilateral rounds, but also from participation in disputes that may bear on
their interests. These critics bemoan the lack of access to the room during consultations
and litigation. They call for open hearings, such as take place in the European Court of
Justice (ECJ), and greater access for nongovernmental organizations (NGOs). Yet these
calls often ignore the means that members do have to participate, the full costs and ben-
efits entailed, and the drivers of the decision to participate or not. The bottom line is that
normative claims, such as the ones made about participation and transparency, carry
empirical implications, and these implications can be tested. In doing so, scholars are
likely to find the normative case more ambiguous than it appears at first blush. This mat-
ters, since it leads to divergent policy solutions.
Third, scholars give insufficient attention to the institution’s words. For all the interest
in the acquis of WTO case law, few studies unpack the content of the decisions rendered.
It is as if the words matter only because members think they do, not because of the words
themselves. But the words do matter. WTO verdicts are longer, and more complicated,
than the GATT-era documents, and say things that the field should take seriously. There
is, for starters, an inventory of legal precedents, called the WTO Analytical Index, which
can guide scholars in making more use of these rulings. Indeed, if the WTO increases
the predictability and stability of the global economy, it does so through precedent. The
field of American politics has started to take the utterances of courts seriously, as should
WTO scholars. It may be the case, for example, that precedent-heavy rulings bear on
402    Dispute Settlement in the WTO

rates of compliance or settlement. Apart from these outcomes, moreover, WTO panel
reports detail procedural aspects of cases that can be important, like the decision not
to rule on certain legal claims, which can limit the scope of the resulting precedent.
This practice, called judicial economy, is explicitly explained in panel reports and offers
a crucial link between the interests of the wider membership and the dispute at hand
(Busch and Pelc 2010). Given all of the new methodologies for doing content analysis,
the words reasoned by the WTO merit a closer look. GATT-era scholarship could only
have dreamed of allowing the institution to speak for itself.
This chapter first surveys some of the main themes in the literature on WTO dis-
pute settlement and provides an apercu of the stages of a typical dispute. It then pro-
poses tackling trade specific concerns, participation, and the WTO’s words. Finally, the
conclusion.

The Literature on WTO

Dispute Settlement

The study of WTO dispute settlement is a multidisciplinary venture. Lawyers, econo-

mists, and political scientists have drawn on each other’s research more in this than in
most other areas of international affairs. Like many, we attribute this to Robert Hudec, a
legal scholar by training but a political economist at heart. Hudec (2000) set a high stan-
dard, demanding that research be grounded in the institution’s mechanisms, but sensi-
tive to the “theater” he saw domestically and internationally, reflecting governments’
efforts to curry favor at the ballot box. This same high standard is no less relevant today.
The defining feature of dispute settlement is that it is decentralized. This distinguishes
the WTO from domestic courts or supranational enforcement, the latter of which, in
the European context, boasts at least some centralized prosecutorial functions. In con-
trast, WTO disputes begin when a member considers that its rights have been impaired.
Disputes are typically responses to domestic interest groups that lobby for better market
access abroad (Davis 2012), although some are pursued for “systemic” interests, includ-
ing cases over measures that have never been used.
In keeping with its decentralized character, the WTO has no enforcement power to
speak of. It merely produces legal rulings, which are referred to as “recommendations”
but considered binding on states. The WTO cannot force compliance; it cannot punish
violators. The WTO can only authorize a complainant to “suspend concessions” to a
defendant, by raising tariffs so that foreign exporters lobby for compliance.
In truth, retaliation never happens. Less than 1 percent of cases result in the suspen-
sion of concessions. In fact, only 40 percent of cases ever go to a panel. This is because
most cases are solved or withdrawn in consultation, a mandatory sixty-day period pre-
ceding the request for a panel. Consultations are a holdover from the GATT days, when
the system was (purportedly) more about diplomacy than law. Consultations are held
 Marc L. Busch and Krzysztof J. Pelc    403

behind closed doors, and this is essential to their success. Indeed, countries are more
likely to reach an agreement if they do not need to pander to powerful domestic inter-
est groups. The design of the WTO’s DSU, more generally, reflects the fact that govern-
ments negotiate with domestic interest groups as much as they do with one another.
Consultations, in particular, shield negotiators from these demands at home.
Consultations are not completely closed off to all comers, however. Other WTO
members can enter as “third parties” if they have a substantial trade interest in the dis-
pute at hand, such as when they are affected by the barrier at issue (Busch and Reinhardt
2006). Yet this only happens if both the complainant and the defendant agree to have
those other countries present. Complainants welcome such third parties in about half
of all cases, and defendants rarely block them, for the simple reason that a blocked third
party can always bring a suit of its own, which leads to greater trouble for the defendant
(Johns and Pelc 2014). The one most telling finding about the effect of privacy on nego-
tiations is that the greater the number of third parties present, the less likely is a negoti-
ated solution (Busch and Reinhardt 2006; Davey and Porges 1998). Indeed, third parties
add voices and issues to the disputes, and they may also create an incentive on the part
of litigants to posture, to “act tough” and demonstrate resolve, with an eye toward subse-
quent disputes (Stasavage 2004).
The high rate of settlement is good news. Litigation is an inefficient outcome (Gilligan,
Johns, and Rosendorff 2010): it entails considerable legal and bureaucratic costs, and
from the point of view of litigants, it throws a spotlight on the underlying issue. Litigants
no longer enjoy insulation from domestic interests, since the outcome of litigation—the
panel report—is made public, along with both litigants’ full arguments. Governments
can no longer claim that they wrought a “hard won deal” and misrepresent any losses as
wins. The panel report lays it out for all to see. For these reasons, concessions on the part
of the defendant are most likely to take place prior to the ruling. This is a sign of a func-
tioning legal system: most of the action takes place in the shadow of the law.
These beliefs are empirically supported. On average, complainants are more likely to
obtain concessions if they can reach a settlement than if the case moves on to litigation
and results in a ruling, even if the complainant wins on all counts (Busch and Reinhardt
2001). Indeed, what pushes respondents to concede is the benefit of doing so privately,
which allows governments to shape the perception of these concessions domestically.
That, together with governments’ desire to avoid the normative condemnation that
comes with a ruling by the WTO, explains why the odds of agreement are greatest before
the panel hands down a ruling. This, in short, is the institution’s single greatest forte.
The trade regime effectively relies on self-enforcement. One reason this is achievable
in international politics is that countries have a wealth of information about each other’s
behavior. When there is reason for ambiguity about the legality of a given country’s poli-
cies, litigation is a means of resolving it. In this light, the main function of WTO verdicts
is to clarify the meaning of the institution’s rules, lending greater predictability and sta-
bility to the global economy.
Another reason self-enforcement is achievable is that governments continually inter-
act with one another; the trade regime, after all, is nearly seventy years old. And while
404    Dispute Settlement in the WTO

it remains the prerogative of a sovereign country to ignore a WTO ruling if it chooses,

it does so at the risk that a subsequent ruling it wins may be spurned; that its demands
will be less likely to be met; and that the concessions it offers during trade rounds will be
devalued. In short, countries join the institution knowing that in some cases the rules
will work against them, but on average, membership is a net benefit (Pelc 2010). This is
also the reason why, despite the ability to block unfavorable rulings during the GATT,
surprisingly few countries availed themselves of this option over the half century prior
to the WTO’s inception. Just as countries can spurn rulings now, they could block rul-
ings then. In both cases, they avoid doing so for analogous reasons.
To avoid normative condemnation, defendants will dig in their heels once an unfavor-
able ruling is handed down. Indeed, there is little left to lose at that point. As Schelling
(1960) noted, a threat that needs to be exercised has lost all potency; it is an ineffec-
tive threat. As a result, concessions are most likely to occur at the prelitigation stage. Of
course, whether cases proceed to litigation and result in a panel report is not a random
occurrence. Some cases stand little chance of settling early. Disputes like US—Cotton
or EC—Hormones touch on such sensitive issues that the domestic political costs of
conceding, even behind closed doors, would have been prohibitive (Heinisch 2006;
Rountree 1998–1999; Shumaker 2007; Sien 2006–2007). Some disputes, in other words,
are launched even though it is known that they will not settle. Moreover, it is likely that
countries sometimes file cases not principally for the sake of the aggrieved domestic
industry, but rather because they are seeking a favorable interpretation of the rules (Pelc
2013). This is why a country that “wins” at the panel stage may still appeal the verdict,
looking to secure more favorable legal logic from a body new to the WTO: the Appellate
Body (AB).
The AB is comprised of seven full-time jurists, three of whom hear any given case.
Insofar as the WTO establishes precedent, this is the body that does it. The AB weighs
the validity of the appealed portion of a contested panel report. It has no fact-finding
power, in that it cannot collect additional information or ask questions, and it cannot
send a case back to a panel to complete unfinished analysis. Some 70 percent of panel
verdicts are appealed.
If the defendant loses its appeal, and the complainant perceives that the policy at fault
has not been remedied, the next step is a compliance panel, where the same panelists
from the panel report now address a single question: Has the defendant followed the
panel’s recommendations? If the defendant is found not to have complied, then the
complainant can ask for the authorization to retaliate.
This final step, the DSU’s last resort, has preoccupied much of the literature, despite its
rarity. One thing is readily agreed on: retaliation serves no balancing function, because
it does not remedy the harm done (Hudec 2000, 22). In economic terms, it amounts
to shooting oneself in the foot. Indeed, it is worth reflecting on the apparent irony of a
system built on the principle of free trade that allows aggrieved countries to respond to
injury by raising tariffs in turn.
Adam Smith foresaw retaliation for what it was:  economic folly in the service of
potentially rational politics, belonging “not … so much to the science of a legislator, …
 Marc L. Busch and Krzysztof J. Pelc    405

as to the skill of that insidious and crafty animal, vulgarly called a statesman or politi-
cian” (1937, 435). Although nearly 250 years old, this view is not out of date. As the rep-
resentative from Ecuador put it recently, retaliation at the WTO “does not restore the
balance lost, … but rather tends to inflict greater injury on the complaining party, as
occurred in the banana dispute.”2 Indeed, retaliatory tariffs end up hurting end-users
in the complaining country, who, as Ecuador further explains, “have nothing to do with
the dispute. [N]‌ever mind consumers.”3 This is not the position of developing countries
alone. The European Union (EU) has declared that “the use of suspension of trade con-
cessions involves a cost not only for the defending party, but also for the economy of the
complaining Member.”4
Most important, however, the threat most likely to drive behavior is that of an unfa-
vorable ruling, rather than the suspension of concessions, no matter how credible the
latter may be. The GATT period, wherein the trade system successfully resolved trade
conflicts, serves as good evidence on this count. Despite the lack of any credible threat
of retaliation—since defendants could always block a case—the system worked. It did so
because it led countries to settle their disagreements, to avoid being branded as violators.

Selection and Specific

Trade Concerns

The methodological challenge that has gnawed at WTO scholars most systematically is
the issue of selection (see Rosendorff, this volume). While the problem of endogeneity
plagues much of social science inquiry, selection is a particular obstacle for the study
of courts, since disputes only become observable once they are officially launched, and
observers rarely have any sense of what potential disputes were not acted on.
This has considerable consequences for empirical analysis. Imagine a research
design that seeks to investigate whether panel recommendations that require legislative
approval in the defendant country are less likely to be complied with. If we believe that
WTO members are rational, we would then also expect them not to file such cases as
complainants, or to file them with much less frequency. Those filed might have higher
legal merit, making compliance more likely, all things being equal. What WTO scholars
lack, in sum, is a sense of the disputes that are not filed: the proverbial nonbarking dogs.
A range of new statistical approaches has sought to deal with such selection effects,
chief among them being the search for valid instrumental variables, which is of limited
use in getting a sense of nonfiled disputes in the study of courts. Others have attempted
to delve into the process of selection itself. Davis (2012), for example, looks at the step
prior to dispute initiation in the United States by examining the petitions of industries
for enforcement of trade rules in Congress.
In recent years another promising avenue for exploring the issue of selection has
become available. Specifically, the WTO collects data on STCs reported to the TBT
406    Dispute Settlement in the WTO

and SPS committees. The key is that these are not (yet) formal dispute settlement cases.
Rather, they are complaints brought to the attention of the committee with the aim of
seeking a resolution. The data are organized just as in a dispute settlement case: the
member(s) raising the concern is identified; the member(s) implementing the mea-
sure at issue, and its description, are recorded; and the case’s “status” is listed, including
whether it was partially or fully resolved or an outcome was not reported. These data are
instructive in thinking about selection issues.
The first thing that stands out about STCs is the numbers. With respect to TBT mea-
sures, there have been 379 STCs versus 47 dispute settlement cases. In terms of SPS,
STCs outnumber instances of formal litigation by a wide margin: 350 to 40. So what
exactly are STCs? Horn, Mavroidis, and Wijkstrom (2013, 1) describe them as an “infor­
mal form of resolution of trade conflicts” over TBT and SPS measures. These measures
are complex because they involve regulatory politics. The stakes are thus often high, and
the WTO invests in making these measures transparent. Since TBT and SPS measures
can act as nontariff barriers, members are required to “notify” others of any changes in
policy. Many STCs arise as a result of these notifications, but STCs can also be brought
in reaction to measures that are not notified. As the WTO explains it, STCs “relate nor-
mally to the proposed draft measures … or to the implementation of existing measures”
(WTO, 2011 Document G/TBT/1/Rev.10).
The point is that STCs are informal and cheap to bring, and there is little incentive for
governments not to let them proceed. The bureaucratic processes that select disputes
to initiate formally are not yet at play. Indeed, because the WTO equates notifications
with transparency, governments are likely to err on the side of more, not fewer, STCs.
Data on STCs can thus be expected to be rather inclusive, permitting a look at easy and
hardest-test cases, and everything in between.
This brings us to the second aspect that stands out about STCs: the subject matter.
Unsurprisingly, some are partly, or wholly, probative. Most basically, a member has a
question about a measure(s) that has an impact on its trade. For example, in 261 STCs
concerning TBT, one or more parties asked for “further information” on, or “clarifica-
tion” of, the measure. But many STCs invoke the sort of language that one expects to
read in a dispute settlement case. Staying with TBT, fully 100 STCs cite a measure’s dis-
criminatory effect; 151 question the rationale and/or legitimacy of a given measure; and
141 ask how the measure relates to international standards, the sine qua non of a case
litigated under TBT Article 2.
Turning to SPS, Australia, Canada, and the United States brought the first STC on
January 6, 1995, over Korea’s shelf-life requirements. A day later the WTO recorded a
partial resolution. A  year later the EU brought an STC concerning Brazil’s import
requirements for wine. No resolution has been reported. These STCs are a window on
the dogs that did not bark. True, STCs still involve a decision to go to Geneva. There is
still a selection issue in that something motivated the EU to bring its worry to the SPS
committee, even if it opted not to litigate this case right away. The point is that STCs help
us understand why certain contentious trade conflicts, once brought to the WTO, go to
litigation, but not others. It is in this sense that STCs flesh out the cases that weren’t.
 Marc L. Busch and Krzysztof J. Pelc    407

Then again, STCs sometimes precede cases that were. Returning to TBT, STC 42,
raised on February 25, 2000, by Brazil, Mexico, and Egypt, concerns US dolphin-safe
labeling. In 2008 Mexico requested consultations with the United States, commencing
for formal dispute settlement. The case, US—Tuna II, is one of the most important TBT
cases to date, involving core aspects of the agreement. There is little doubt that the STC
was focused on these issues; in addition to requesting information and clarification,
Brazil, Mexico, and Egypt focused on the non-product-related process and production
methods that trigger the label “dolphin-safe.” These measures are increasingly wide-
spread and the subject of much debate. That this STC preceded formal litigation is not
surprising. What is surprising is that Brazil and Egypt did not join as co-complainants,
and only Brazil reserved its third-party rights. Moreover, all of this took place in full
knowledge of what happened in the GATT-era case US—Tuna I. This suggests that
STCs can offer insights into the politics and legal strategy leading up to the request for
consultations.
STCs can teach us still more. Horn, Mavroidis, and Wijkstrom (2013), for example,
find that the relationship between the number of STCs and dispute settlement cases is
not obvious. Our instinct would be to match the content of STCs to the specific legal
claims raised in dispute settlement cases, an effort that, to our knowledge, has yet to be
undertaken. Not that this is easy. In the case where an STC precedes dispute settlement,
a litigant is likely to learn from the responses to its initial inquiries and adjust its legal
claims accordingly. This may make it difficult, at times, to directly map dispute settle-
ment cases onto STCs. On the other hand, in cases like US—Tuna II, this mapping is
fairly straightforward.
The payoff for doing the mapping promises to be significant. Not only will it give us
a better sense for selection issues, but it will tell us new things about the protagonists.
Horn, Mavroidis, and Wijkstrom (2013) observe that developing countries are dispro-
portionately involved in STCs, relative to their trade volumes. Do STCs teach develop-
ing countries how to prosecute trade conflicts? Busch and Reinhardt (2006) suggest that
the high rates of developing-country participation as third parties might be part of a
“learning by watching” strategy. Perhaps the same is true of STCs. Or perhaps develop-
ing countries derive something quite different from bringing STCs.
To be sure, it would be a mistake to view STCs as just the prelude to litigation. The fact
that they get the TBT and SPS committees involved in interpreting rules is tremendously
valuable. This may be especially helpful for developing countries. One intriguing possi-
bility is that STCs get the TBT and SPS committees to reflect on the implications of ongo-
ing litigation. If so, lessons may be incorporated into a resolution faster than if members
had to wait out the verdicts issued by the WTO’s judicial bodies. Consider, for example,
Cuba’s involvement in the dispute over Australia’s tobacco measures. Cuba filed this case
in May 2013, joining the Dominican Republic, Honduras, and Ukraine, all of which filed
in 2012. Interestingly, Cuba, the Dominican Republic, and Honduras, together with nine
other developing countries, raised TBT STC 377 in March 2013 against EU measures on
tobacco, including packaging restrictions. We submit that the TBT committee has some
latitude to informally discuss the litigation, anticipate the politics of an outcome, and
408    Dispute Settlement in the WTO

suggest the contours of a negotiated settlement in the shadow of the law. This is exactly
the logic behind DSU Article 5, “Good Offices, Conciliation and Mediation,” another
informal provision to be used alongside formal dispute settlement, but one that is almost
never used. Given the technical subject of TBT and SPS, and since STCs are not a part of
formal dispute settlement per se, perhaps they fill in where DSU 5 has fallen short.
Studying STCs will not end concerns about selection bias, but it will go a long way in
getting at the sources of potential bias. The data speak explicitly to those cases that look
more like information gathering than allegations of discrimination. This in itself is a win
that GATT-era scholars found elusive.
Finally, there are alternative means of addressing our inability to observe the cases
that weren’t. Imagine a typical empirical claim, drawn from the political economy of
trade literature, that politically powerful industries get more enforcement of their griev-
ances. The difficulty of testing such a claim rests in the same problem of selection. If we
look at the cases filed and find that they disproportionately concern politically powerful
industries, what are we to conclude? This could be a sign of domestic institutions favor-
ing such industries, or it could be that, for example, the type of industry that is politically
powerful is also more likely to be faced with WTO-inconsistent import barriers abroad.
How are we to know if we cannot inspect the pool of potential cases from which dis-
putes are drawn? One way of getting at such a story is to invert conditional hypotheses
(Johns and Pelc 2014). If domestic institutions favor politically powerful cases, then con-
ditional on such a case being filed, it should have on average lower legal merit than a case
not favored for this reason. This institutional bias should then have observable effects at
the settlement and ruling stages—all things being equal, cases representing politically
powerful industries should result in less favorable panel rulings. If, instead, those cases
are filed because the industries genuinely face more trade barriers abroad, then there
should be no difference in panel outcomes.
In sum, the literature has reached the stage where it needs to take methodological
concerns such as selection bias seriously. Although no one solution is a panacea, there
are different ways of addressing these concerns. Selection effects can sometimes be iden-
tified by examining subsequent legal outcomes. Or scholars can examine the data that
have emerged about that prior step, by looking at domestic petitions or STCs within the
WTO itself.

Participation

A significant portion of the literature on the WTO is made up of normative claims. This
is not surprising, given how the institution brings together developed and developing
countries under an agreement with common objectives that addresses increasingly
sensitive political issues. One of these normative assessments has grown particularly
salient: observers have claimed that developing countries are effectively excluded from
the system.5
 Marc L. Busch and Krzysztof J. Pelc    409

As with multilateral negotiations during trade rounds, most countries are kept out-
side of the room during WTO legal proceedings, even if their interests are at stake
(Smythe and Smith 2006; Steinberg 2002; Heisenberg 2007). And neither domestic
interest groups nor NGOs have any access to DSU hearings.6 As the literature has begun
addressing these claims empirically, some unexpected conclusions have emerged.
It is worth noting that in many respects the WTO’s level of transparency compares
favorably with that of other international organizations. Public access to adopted panel
reports, by itself, constitutes a unique window on decision making in the institution, the
likes of which does not exist in the sister Bretton Woods institutions. The comparison is
less favorable with other courts, such as the European Court of Justice, where hearings
are open to the public. And there is far from a harmony of interests among participa-
tion advocates. The often-overlooked truth is that while developing countries demand
greater representation in negotiations and greater access to legal proceedings, they also
seek to prevent opening up the process to NGOs, such as through the acceptance of
amicus briefs, submissions from “friends of the court.” Now that they are engaged in the
process of governance, developing countries resist the entrance of another set of actors
that would dilute their influence (Kahler 2004). In other words, every increase in par-
ticipation also holds distributional effects.
The institution offers opportunities for empirical assessment of the claims underlying
the participation debate. Indeed, the WTO has championed the option of third-party
participation during disputes. Third-party status, whereby nonlitigant countries
can join a dispute from its very beginning, is one means of allowing developing and
least-developed countries to be in the room when an issue in which they may have a
stake is being negotiated. It is also a means for countries to learn the intricacies of trade
law and build their legal capacity (Busch, Reinhardt, and Shaffer 2009).
There is also evidence that the presence of third parties has a positive effect on the con­
tent of settlements. Whereas complainants are found to capture a disproportionate por-
tion of gains from settlement when consultations are conducted entirely in private, this
is not the case when third parties are present (Kucik and Pelc 2013). In those instances,
the complainant, on average, gains no more than the remainder of the membership.
Third parties, in other words, perform an unwitting enforcement function, by ensuring
that the gains reached in a settlement are extended to the remainder of the member-
ship, per DSU rules.7 Naturally countries do not join as third parties out of altruism;
above and beyond the desire to have their views recorded and building legal capacity,
countries join as third parties to ensure that they are not left out of an otherwise private
settlement.
Indeed, countries that choose not to sign on as third parties run the risk of being
left out of potential concessions if the parties reach early settlement (Nakagawa 2007;
Alschner 2014; Bown 2009). When parties settle before proceeding to the panel stage,
the terms of the settlement remain private. Defendants wary of increasing market
access across the board can carefully design concessions to appease the complainant,
while limiting benefits to the broader membership. Such discriminatory settlements are
prohibited by the rules of the DSU. Nevertheless, the secrecy surrounding settlements
410    Dispute Settlement in the WTO

makes this difficult to enforce (Nakagawa 2006). A country that has an interest in the
dispute but elects not to sign on as a third party may find itself out in the cold unless it
follows up by filing a new dispute. In sum, third parties join to ensure that they get their
share of a settlement, yet in doing so, they also have a positive effect on the equity of the
deal for everyone else.
There is a catch. These beneficial effects of third parties pertain to those disputes in
which a settlement is reached. Yet it turns out that the presence of third parties also
affects the odds of such a settlement being reached in the first place. Despite all the ben-
efits of third-party participation, this step toward greater inclusivity thus comes at a
cost. The presence of an audience makes settlement significantly less likely (Busch and
Reinhardt 2006; Davey and Porges 1998; Stasavage 2004). There is enough incentive to
posture for the sake of these other member states that the odds of reaching agreement
between the litigants decline precipitously as the number of other countries in the room
rises. In fact, beyond five third parties, the odds of agreement are effectively nil (Busch
and Reinhardt 2006). In other words, if settlement is the goal of the system, then greater
inclusiveness through third-party participation exerts a cost. And this is the case even
when the audience in the room is made up of other countries, rather than NGOs and
domestic interest groups.
The effect of third parties on the odds of prelitigation settlement, it turns out, is a
special concern for developing countries. Herein lies one of the true cleavages between
rich and poor countries in dispute settlement. It is not found in the odds of winning a
case, which are statistically indistinguishable for rich versus poor countries. The true
difference is that, on average, developed countries settle, and developing countries liti-
gate. This happens for two reasons. First, developing countries tend to file cases against
large, rich countries with whom countless others trade, meaning that third parties are
more likely to join, with their participation undermining the prospects for early settle-
ment (Busch and Reinhardt 2006). Second, developing countries generally lack the legal
capacity to make the most of consultations (Busch, Reinhardt, and Shaffer, 2009), fur-
ther reducing these odds.
Wider participation, as in other fora, also has the effect of politicizing outcomes (see
Rosendorff, this volume). It takes away from the unambiguous legal condemnation that
is intended to drive country behavior. Courts are intended to shift disputes and deci-
sion making from the political to a legal arena. These bodies can promote cooperation
by creating a “functional domain to circumvent the direct clash of political interests”
(Burley and Mattli 1993, 44). Third- party participation achieves the opposite. And such
repoliticization has distributional effects: the ambiguity it generates over legal rulings
helps the losing party in a dispute and hurts the party that prevails on the legal merits
(Johns and Pelc 2014).
Yet the most striking aspect of third-party participation is that for all its
benefits—voicing one’s interests, building legal capacity, ensuring one is not left out
of a settlement—relatively few countries avail themselves of the opportunity (Horlick
1998, 690). Despite its low cost, the average dispute counts about four third parties,
while estimates suggest that based on countries’ trade stakes, the number should be
 Marc L. Busch and Krzysztof J. Pelc    411

closer to fourteen (Johns and Pelc 2014). This outcome would seem to complicate the
simple prescription of allowing for greater access:  the option to participate exists,
yet relatively few countries exercise it. The reason seems to be that countries take the
aforementioned consequences of participation in settlement into account. Given that
more countries in the room means lower odds of settlement, and litigation is an inef-
ficient outcome for all involved, countries appear to exercise restraint. Rather than try
to capture a larger slice of a potentially smaller pie, states strategically choose to settle
for a potentially smaller slice of a larger pie. The literature until now has focused on
how power considerations may deter participation: Elsig and Stucki (2012) describe
in detail the reluctance of African countries to join the cotton dispute filed by Brazil
against the United States, for fear of jeopardizing US aid. What these recent findings
suggest, however, is that controlling for such power considerations, a greater driver
of nonparticipation, on average, appears to be every country’s interest in avoiding the
negative consequences of overcrowding. The result is that paradoxically, every addi-
tional third party makes everyone else less likely to participate. In short, low rates of
participation may be the result not of power plays and exclusionary rules, but of strate-
gic country behavior.
In sum, cleavages between rich and poor countries exist. Yet they do not lie where we
most often look for them. Success in dispute settlement hinges most of all on the abil-
ity to assess the legal merit of potential cases (Busch, Reinhardt, and Shaffer 2009) and
to settle disputes in the shadow of the law, before a ruling is handed down. Developing
countries are at a disadvantage in this respect, as they tend to settle less and litigate
more. Yet scholars have shown that countries can rapidly learn to use dispute settle-
ment (Davis and Bermeo 2009). Gaining legal capacity is more easily achievable than
growing one’s market power or retaliatory capacity. Finally, there are institutional solu-
tions to the participation deficit. The option of joining as a third party is beneficial, as it
allows poor countries to be present and participate in proceedings, to voice their views
and potentially affect legal outcomes, and to protect their trade interests in so doing.
Yet third-party participation also comes at a cost to the institution: as diplomats know,
audiences make agreement less likely (Keohane and Nye 2001). There exists an observ-
able trade-off between inclusiveness and the odds of settlement, meaning that there is
no easy solution to the participation deficit. The challenge of the institution is to balance
these competing objectives.

Words

With a wink and a nod at the role of precedent, economists and political scientists
reduce WTO rulings to which country won or lost a case (e.g., Busch and Reinhardt
2001; Petersmann 1994). There is a sense that more predictable legal decisions, through
precedent, make the system work, but there is seldom much attention paid to what the
rulings actually say or the structure of the legal arguments rendered. Lawyers generally
412    Dispute Settlement in the WTO

do a better job mulling over the references to past cases, but tend to focus on one or two
disputes, offering little perspective on broader trends, which after all are at the heart of
the matter. To capture these trends, we need to combine the wider accounting of econo-
mists and political scientists with the attention to detail of the lawyers, something the
literature has yet to do. Put simply, we need both the forest and the trees if we are to mea-
sure the extent to which past rulings hold sway over present ones.
Law is a fundamentally rhetorical exercise. And while the rules have remained con-
stant since the WTO’s inception, their agreed-upon meaning has changed. This has
taken place through rulings spanning the GATT and WTO periods.
The traditional search for sources of law falls short in trying to delimit between
what binds countries and what does not. Despite DSU Article 3.2 explicitly stating
that “recommendations and rulings of the DSB cannot add to or diminish the rights
and obligations provided in the covered agreements,” echoing Article 59 of the ICJ
Statute, panel rulings plainly shape members’ understanding of the meaning of the
rules. Even unadopted GATT panel reports, such as the ones for the Tuna—Dolphin
decision against the United States, have shaped the interpretation of subsequent WTO
law concerning nontrade issues, taken up in some of the WTO’s most important dis-
putes (Venzke 2012). And this is in spite of having no actual legal standing whatsoever.
Indeed, WTO rulings have had most to say about issues that countries find hard to
arrive at a consensus on during negotiations. In this way, it is legal rulings that have cre-
ated expectations about how policies addressing nontrade concerns must be “primarily
aimed at” addressing these concerns (whereas the text of Article XX[g]‌merely states
that they must be “related to” these policies) and that such measures must be “least trade
restrictive” (whereas the text at issue in Article XX[d] states that the policy must be
“necessary”) (Venzke 2012). Where the legislative function of an institution falls short,
judges often step in. The way they do so cannot be captured by looking at who wins and
who loses.
Scholars lag behind policy makers in the realization that words affect the obligations
that bind them. There is growing evidence that countries recognize the importance of
words in their behavior. Countries sometimes appeal favorable rulings, because the
particular language employed by the panelists does not suit their long-term interests.
They have strong preferences about judges’ use of judicial economy, even as it is meant
to leave the final verdict unaffected (Busch and Pelc 2010). And they file some cases
not with the typical goal of satisfying some export-oriented industry, but rather with a
longer-term goal of modifying the meaning of the rules in a way that favors their inter-
ests (Pelc 2013).
The EU’s pursuit of safeguards disputes successfully achieved such a change.
Safeguards are the institution’s quintessential escape clause. The EU, an infrequent safe-
guards user, filed a number of low-profile cases soon after the WTO’s inception, tar-
geting Korean skim powdered milk and Argentinean footwear—insignificant markets
by any measure. It successfully argued for an interpretation of the new Agreement on
Safeguards through its predecessor, GATT Article XIX. In doing so, it significantly raised
the threshold for the use of safeguards, reenlisting the GATT’s mention of “unforeseen
 Marc L. Busch and Krzysztof J. Pelc    413

developments” as a prerequisite for any exercise of a safeguard (Pelc 2009). As soon as

this legal strategy bore fruit, the EU turned to bigger fish: it took on the United States in
a series of safeguards cases that rank among the WTO’s most important in terms of their
commercial value. The United States lost these cases and immediately cut its reliance on
safeguards. The rules did not change, but their meaning did, as the EU convinced the
AB to adopt its interpretation of the text. Scholars need to recognize the importance of
words: their ability to affect state behavior.
While this discussion emphasizes how words figure in countries’ strategic behavior,
the same can naturally be said of judges themselves. Judges have preferences: panelists
would rather not have their rulings overturned on appeal; both panelists and AB judges
would rather see compliance with their recommendations. More subtly, WTO judges
may have some preferences about the evolution of the institution that differ from those
of states. Words are the means by which judges act on these preferences. While the inter-
play between legislative politics and judicial decisions has been studied most closely in
the European context (Burley and Mattli 1993), similar conclusions have obtained in
other contexts. American legal scholars have recently shown that US Supreme Court
judges write less “readable” opinions when they anticipate a hostile reaction in the
House (Owens, Wedeking, and Wohlfarth 2013). All of this variation gets lost when one
reduces rulings to wins or losses, as the literature has largely done up to now. The time is
ripe for taking words seriously.

Conclusion

The literature on WTO dispute settlement has come a long way. It is richer and deeper
in so many ways than its earliest contributors could have imagined. It is also more
multidisciplinary than many would have predicted. But along the way, as issues like
selection and participation have preoccupied scholars, some of the institution’s most
axiomatic features have been overlooked. As a matter of course, the literature con-
trols for things like trade dependence or dummy SPS measures as being especially
politically contentious. But it does not always ask about what the litigants argued or
what the panel’s verdict did or did not say. It has been a long time since the literature
moved beyond the question of whether institutions matter and turned to the question
of why they matter. If we are to take dispute settlement seriously, part of the answer to
why they matter must be the words uttered in a panel or AB report, the calibration of
legal arguments to key precedents, and the acquis of case law that extends the longest
shadow on the decision to file in the first place. None of this will be as easy to code as
variables like trade dependence. But then again, nothing was ever easy about coding
GATT disputes back in the day. The methods and technology that we have today make
it possible to be as creative as we are rigorous in studying WTO dispute settlement. It is
an institution of laws that renders legal judgments. It is time to let the institution speak
for itself.
414    Dispute Settlement in the WTO

Acknowledgements

Our thanks to Lisa Martin for comments on this chapter, and Amanda Kennard for her
research assistance.

Notes
1. Rossmiller (1994, 232).
2. Contribution of Ecuador to the Improvement of the Dispute Settlement System of the
WTO, 2002, TN/DS/W/9. The dispute being referred to is EC—Bananas, WTO/DS27.
3. WTO, TN/DS/W/9.
4. WTO, TN/DS/W/1.
5. For instance, at a WTO meeting in 2011, the representatives of Bolivia, Cuba, Ecuador, and
Nicaragua complained that “[i]‌n practice, the WTO has become an organization that is
not led by its Members, in which decision-making based on facts is not governed by con-
sensus, and negotiation meetings are not open to participation by all Members.”
6. Oxfam International and WWF International, together with eight other NGOs, released a
memo in 2003 claiming that the WTO features “proliferation of ‘informal’, undocumented
and exclusive meetings, amounting to lack of transparency and inability of many countries
to participate.”
7. See Article 3.2 of the DSU.

References
Alschner, W. 2014. Amicable Settlements of WTO Disputes: Recent Trends, New Concerns
and Old Problems. World Trade Review13: 65–102.
Bhala, Raj. 1998–1999. The Myth about Stare Decisis and International Trade Law (Part One of
a Trilogy). American University International Law Review 14: 845–956.
Bown, Chad P. 2009. Self-Enforcing Trade: Developing Countries and WTO Dispute Settlement.
Washington, DC: Brookings Institution Press.
Burley, Anne-Marie, and Walter Mattli. 1993. Europe before the Court: A Political Theory of
Legal Integration. International Organization 47 (1): 41–76.
Busch, Marc L., and Eric Reinhardt. 2001. Bargaining in the Shadow of the Law:  Early
Settlement in GATT/WTO Disputes. Fordham International Law Journal 24 (1): 158–172.
Busch, Marc L., and Eric Reinhardt. 2003. Developing Countries and General Agreement on
Tariffs and Trade/World Trade Organization Dispute Settlement. Journal of World Trade
37: 719–735.
Busch, Marc L., and Eric Reinhardt. 2006. Three’s a Crowd: Third Parties and WTO Dispute
Settlement. World Politics 58: 446–477.
Busch, Marc L., Eric Reinhardt, and Gregory Shaffer. 2009. Does Legal Capacity Matter?
Explaining Dispute Initiation and Antidumping Actions in the WTO. World Trade Review
8: 559–577.
Busch, Marc L., and Krzysztof J. Pelc. 2010. The Politics of Judicial Economy at the World Trade
Organization. International Organization 64: 257–279.
 Marc L. Busch and Krzysztof J. Pelc    415

Davey, William J., and Amelia Porges. 1998. Comments on Performance of the System I:
Consultations and Deterrence. International Lawyer 32: 695–707.
Davis, Christina. 2012. Why Adjudicate? Enforcing Trade Rules. Princeton, NJ:  Princeton
University Press.
Davis, Christina L., and Sarah Blodgett Bermeo. 2009. Who Files? Developing Country
Participation in GATT/WTO Adjudication. Journal of Politics 71: 1033–1049.
Elsig, Manfred, and Philipp Stucki. 2012. “Low-income Developing Countries and WTO
Litigation: Why Wake Up the Sleeping Dog?” Review of International Political Economy 19
(2): 292–316.
Gilligan, Michael, Leslie Johns, and B. Peter Rosendorff. 2010. Strengthening International
Courts and the Early Settlement of Disputes. Journal of Conflict Resolution 54: 5–38.
Heinisch, E. L. 2006. West Africa Versus the United States on Cotton Subsidies: How, Why and
What Next? Journal of Modern African Studies 44 (2): 251–274.
Heisenberg, Dorothee. 2007. Informal Decision-Making in the Council: The Secret of the EUs
Success? In Making History: European Integration and Institutional Change at Fifty, edited by
Sophie Meunier and Kathleen R. McNamara, 67–88. Oxford: Oxford University Press.
Horlick, Gary N. 1998. The Consultation Phase of WTO Dispute Resolution:  A  Private
Practitioner’s View. International Lawyer 32: 685–693.
Horn, H., P. C. Mavroidis, and E. N. Wijkstrom. 2013. In the Shadow of the DSU: Addressing
Specific Trade Concerns in the WTO SPS and TBT Committees. Journal of World Trade 47
(4): 729–759.
Hudec, Robert E. 2000. Broadening the Scope of Remedies in the WTO Dispute Settlement:
Issues and Lessons from the Practice of Other International Court and Tribunals. London:
Cameron May.
Johns, Leslie, and Krzysztof J. Pelc. 2014. Who Gets to Be in the Room? Manipulating
Participation in WTO Disputes. International Organization 68(3): 663-699.
Kahler, Miles. 2004. Defining Accountability Up:  the Global Economic Multilaterals.
Government and Opposition 39(2): 132–158.
Keohane, Robert, and Joseph S. Nye. 2001. The Club Model of Multilateral Cooperation
and Problems of Democratic Legitimacy. In Efficiency, Equity, and Legitimacy:  The
Multilateral Trading System at the Millennium, edited by Roger B. Porter, Pierre Sauv, Arvind
Subramanian, and Americo Beviglia Zampetti, Washington, DC:  Brookings Institution
Press, 264-294.
Kucik, Jeff, and Krzysztof J. Pelc. 2013. Measuring the Cost of Privacy:  A  Look at the
Distributional Effects of Private Bargaining. McGill University and University College
London Working Paper.
Nakagawa, J. 2007. “No More Negotiated Deals?: Settlement of Trade and Investment Disputes
in East Asia.” Journal of International Economic Law 10(4):837–867.
Owens, Ryan J., Justin Wedeking, and Patrick C. Wohlfarth. 2013. How the Supreme Court
Alters Opinion Language to Evade Congressional Review. Journal of Law and Courts
1(1): 35–59.
Pelc, Krzysztof J. 2010. Constraining Coercion? Legitimacy and Its Role in U.S. Trade Policy,
1975–2000. International Organization 64: 65–96.
Pelc, Krzysztof J. 2009. Seeking Escape:  The Use of Escape Clauses in International Trade
Agreements. International Studies Quarterly 53 (2): 349–368.
Pelc, Krzysztof J. 2013. The Politics of Precedent in International Law:  A  Social Network
Application. American Political Science Review 108(3): 1-18.
416    Dispute Settlement in the WTO

Petersmann, Ernst-Ulrich. 1994. The Dispute Settlement System of the World Trade
Organization and the Evolution of the GATT Dispute Settlement System Since 1948.
Common Market Law Review 31: 1157–1244.
Rossmiller, George E. 1994. Discussion. In Agricultural Trade Conflicts and GATT:  New
Dimensions in US-European Agricultural Trade Relations, edited by Giovanni Anania et al.,
Boulder, CO: Westview Press, 262–263.
Rountree, G. H. 1998–1999. Raging Hormones: A Discussion of the World Trade Organization’s
Decision in the European Union-United States Beef Dispute. Georgia Journal of International
and Comparative Law 27: 607–634.
Schelling, Thomas C. 1960. The Strategy of Conflict. Cambridge, MA: Harvard University Press.
Shumaker, M. J. 2007. Tearing the Fabric of the World Trade Organization:  United
States-Subsidies on Upland Cotton. North Carolina Journal of International Law and
Commercial Regulation 32: 548–604.
Sien, I. A. R. 2006–2007. Beefing up the Hormones Dispute: Problems in Compliance and
Viable Compromise Alternatives. Georgetown Law Journal 95: 565–590.
Smith, Adam. 1937. The Wealth of Nations (1776). New York: Modern Library.
Smythe, Elizabeth, and Peter J Smith. 2006. Legitimacy, Transparency, and Information
Technology: The World Trade Organization in an Era of Contentious Trade Politics. Global
Governance: A Review of Multilateralism and International Organizations 12 (1): 31–53.
Stasavage, David. 2004. Open-Door or Closed-Door? Transparency in Domestic and
International Bargaining. International Organization 58: 667–703.
Steinberg, Richard H. 2002. In the Shadow of Law or Power? Consensus-based Bargaining and
Outcomes in the GATT/WTO. International Organization 56: 339–374.
Venzke, I. 2012. How the Practice of Interpretation Makes International Law:  On Semantic
Change and Normative Twists. Oxford: Oxford University Press.
World Trade Organization. 2011. Decisions and Recommendations Adopted by the WTO
Committee on Technical Barriers to Trade Since 1 January 1995. Committee on Technical
Barriers to Trade, G/TBT/1/Rev.10, June 9.
 Pa rt  V I

I S SU E L I N KAG E S
 Chapter 22

Trade and  Wa r

Erik Gartzke a nd Jia k un Jac k  Z han g

Introduction

The relationship between trade and war has been both a cornerstone of statecraft and
a subject of debate for centuries. In recent decades the conventional wisdom that com-
merce brings peace has had a major impact on American foreign policy toward a rising
China and has served as an intellectual impetus behind the formation of the European
Economic Community. The belief that heightened economic interdependence inhibits
conflict rests on a widely cited empirical literature that documents the pacifying effect
of trade at the cross-national level. Yet the causal mechanisms that drive the empirical
relationship between economic interdependence and reduced military conflict have yet
to be fully established. Existing empirical studies suffer from important shortcomings
related to endogeneity and the measurement of key constructs. Much of this literature
has been motivated by, and framed in terms of, stylized debates between and among
traditional paradigms in international relations. Only in the last decade have research-
ers begun to establish a common set of microfoundations. Our analytical understanding
of trade and war has also progressed from the system level, to dyad, and more recently
to network-based theories. Considerable room remains for theoretical and empirical
development. Scholarship on interdependence can be improved through adoption of
common assumptions about the causes of war and through better understanding of the
domestic sources of trade and foreign policy.
This chapter begins by outlining the bargaining theory of war in order to offer a
set of working assumptions about the nature of conflict initiation that can guide
future research on war and trade. Because many theoretical priors used in the exist-
ing literature come from paradigms, the following section reviews and critiques the
ways in which liberals, realists, Marxists, and others have approached the relation-
ship between trade and war. The next section surveys the methodology and empiri-
cal findings of the existing literature. Moving beyond paradigmatic approaches, the
authors advocate that future work focus on theory development and empirical tests
420   Trade and War

of three broad mechanisms that could potentially link trade with war and/or peace:
constraints, information, and the transformation of actors’ preferences. The chapter
concludes with a discussion of how these three broad mechanisms are likely to benefit
future research.

A Theory of War

A discussion on the relationship between trade and war must begin with a theory of
conflict. Theories that are intended to explain how a given process is thought to mod-
ify conflict behavior must function by interacting with the underlying conflict process,
whatever it might be. Simply being explicit about why nations are believed to fight will
go a long way toward whittling down the number of possible ways that trade is likely to
have an impact on decisions about war and peace.
Advances in the study of interstate conflict have produced a theoretical perspective in
tension with the logic and validity of established theories of war. Once-dominant struc-
tural explanations asserting that balances of power promote peace and imbalances lead
to conflict have come under particular scrutiny (Niou et al. 2007). The problem with
using indicators of power as determinants of war and peace is that the conflicts they
might produce are resolvable through negotiation; theories of war fall short if they fail to
take into consideration the role and effect of endogenous bargaining as a more efficient
way of resolving who gets what (Powell 1999). Warfare is only one way in which states
can pursue their interests. Leaders who negotiate and obtain the settlements before
fighting begins are better off than those who must pay the high costs of war (Wagner
2007). It is thus not simply the distribution of power that determines war and peace,
but the degree to which power and benefits overlap or correspond in the global system.
Nations compete not because of power imbalances, but because of imbalances between
the distribution of power and the distribution of benefits.
Even then, it is not power relations per se that lead to conflict, but the incompatibility
of actors’ perceptions of them. Fearon (1995) places in a rationalist and internally consis-
tent framework Blainey’s (1988) basic insight that the causes of war reside not in dispari-
ties of power but in incompatible beliefs about power. Rationalist explanations assume
purposive action on the part of leaders when they are considering the use of force.
Opponents have incentives to feign strength and conceal weakness. Because informa-
tion about an opponent’s resolve or capabilities is generally incomplete, rational leaders
must make decisions in an environment of uncertainty. The centrality of uncertainty to
rationalist explanations for war means that the advent of war is itself stochastic (Gartzke
1999). Reasoning leaders generally are trying to avoid war because it is a costly outcome.
But they can still find themselves at war, for three reasons.
First, war can occur as competitors mistake relative resolve or capabilities, and
because competition generates incentives for actors to conceal true information about
these variables. In an uncertain world, egoistic leaders can benefit by bluffing. This
 Erik Gartzke and Jiakun Jack Zhang    421

“asymmetric information” argument encourages researchers to seek ways that states or

other actors may be able to communicate more or less credibly (Schultz 2001), though
this is difficult.
Second, actors can fight because they face a “commitment problem” created by time
inconsistency. If power relations are shifting over time, so that one actor is declining
in power relative to another, then agreements between them are unstable, as the ris-
ing power will often have an incentive to insist on better terms in the future, when it
becomes more capable. The declining power in turn has incentives to blunt the ascent
of the rising state. Since fighting a winning war today may be better than losing one
tomorrow, commitment problems can cause conflict. Powell (2006) shows that large,
rapid shifts in the distribution of power can lead to war in three classes of commitment
problems: preventive war, preemptive attacks due to offensive advantage, and conflicts
produced by bargaining over issues that affect future bargaining power.
Finally, actors can fight over issues because those at stake are not readily divisible.
This “indivisibility problem” appears to be an infrequent cause of war, as it is typi-
cally possible to fashion bargains through side payments that resolve this motive for
war. However, the fact that indivisibility problems can be resolved does not mean that
actors invariably have an interest in resolving them; competitors may choose to retain
or even develop indivisibility problems for strategic advantage. This in turn could
lead to fighting in a chicken-type scenario if actors are unable to unravel previous
indivisibilities.

The Theories of Trade

The literature on trade and war has remained paradigmatic and is not yet well integrated
with the evolution in thinking about the nature of conflict, such as the bargaining theory
of war. Though various unresolved debates in the paradigmatic literature might well
prove important, they are somewhat less likely to generate the most innovative future
research agendas. The next section discusses the literature on trade and war, using the
paradigmatic framework to present arguments and evidence. We then offer a nonpara-
digmatic approach in the subsequent section.
Liberal theories have dominated discourse on economic interdependence. But liber-
alism is not without its theoretical rivals. Realists and Marxists make starkly contrasting
predictions about the effects of trade on war and peace. Whereas liberals believe that
trade creates virtuous interdependencies that tend to dampen down conflict tendencies,
realists view trade more harshly, believing that it creates vulnerabilities and imbalances
of power, both of which make war more likely. Marxists contradict the realist perspec-
tive on the relationship between trade and war at the level of major powers, believing
instead that class and power collude to suppress and rule. At the same time, Marxists are
in agreement with realists on the more general point that trade tends to increase conflict
at the systemic level, mostly between the core and the periphery.
422   Trade and War

Liberalism
The view that trade has a dampening effect on war is almost a liberal article of faith.
As such, it is in some tension with the liberal perspective’s emphasis on rationality and
evidence-based inference. Advocacy of a commercial peace appears prominently in
the writings of Montesquieu, Rousseau, Kant, Adam Smith, Thomas Paine, and other
Enlightenment figures. Classical liberal political economy espoused policies that would
restrict the war-making power of the aristocratic elite and increase the autonomy of
the commercial classes. The key mechanisms driving liberal trade theories about peace
involve opportunity costs, domestic interest groups, and constitutional republicanism.
Commerce is seen as creating bonds of mutual benefit between countries that are costly
to sever. War threatens to disrupt these beneficial ties, and so the liberal logic predicts
that with increased trade, the incentives to fight will recede. In modern economic terms,
trade raises the opportunity cost of war. A notable articulation of this logic can be found
in Norman Angell’s The Great Illusion (1910), in which he criticized the jingoistic nation-
alism of turn-of-the-century Europe and argued that war, even when victorious, was
socially and economically futile, because wealth in the modern era is tied to credit and
commercial contracts, not to war.
The view that trade increases the opportunity costs of fighting has been revived in
recent years to account for the statistical linkage between trade interdependence and
conflict (Copeland 1996; Oneal et al. 1996; Oneal and Russet 1997, 1999; Oneal et al.
2003; Hegre et al. 2010). According to the classical liberal account, however, the accu-
mulation of commercial interest groups with increasing levels of economic interdepen-
dence provides a check on sovereigns who are otherwise inclined to go to war. Trading
states become populated with commercial interests, which seek to prevail on leaders
to refrain from resorting to force (Bentham 1781; Cobden 1867). Schumpeter (1962) in
particular emphasizes the role of democracy in facilitating the expression of commer-
cial interests in national politics, an approach also championed by Doyle (1983, 1986),
and for which Mousseau (2000, 2003, 2010) provides further theoretical and empirical
justification.1
The channel through which commercial interests express themselves could be greater
in democracies. As Immanuel Kant emphasized in Perpetual Peace, economic interde-
pendence may dampen the risk of war between states if the governments of those states
are responsive to, and representative of, wider rather than narrower social interests. For
Kant, “republican constitutions,” “a commercial spirit,” and a federation of interdepen-
dent republics would provide the basis for “perpetual peace.” Oneal and Russett have
provided much of the systematic empirical justification for the “Kantian peace.” Gelpi
and Grieco (2003, 2008) apply the logic of selectorate theory (De Mesquita and Smith
2005), appearing to show that trade and democracy are each contingent upon the other.
While it is plausible on its face, there are at least three reasons to believe that the
impact or applicability of the classical perspective may be limited. First, to deter conflict,
potential trade losses must be substantial relative to the (subjective) valuations nations
 Erik Gartzke and Jiakun Jack Zhang    423

place on issues or resources in dispute. War is already a highly costly enterprise, while
the portion of trade forfeit in a contest will often be modest in scale. Where commercial
losses can be expected to be small relative to the value of the issues at stake, there is no
reason to believe that interdependence will achieve what amounts to mutual deterrence.
Second, conflict is strategic. The most immediate effect of increasing the cost of warfare
for one actor is to make it more appealing for opponents to behave more aggressively
toward the state that is inhibited. States vary in their dependence on trade; if trade makes
war too costly for one actor, then political competitors—even states also dependent on
trade—have incentives to behave aggressively and to seek further concessions. Whether
raising the cost of fighting leads to cooperation or conflict depends on whether states
can capitalize on the leverage created by trade to achieve more satisfactory diplomatic
solutions, or whether the additional friction that results from increased leverage spills
over from economic to political and military conflict. Interdependence can make war
less, or more, likely depending on other factors (Morrow 1999). Some actors with lim-
ited aims may choose to enjoy increased leverage in the form of a more stable peace, but
others will not, given that most conflicts involve actors with considerable differences in
preferences or interests. Again, mutual linkages will best deter conflict where the incen-
tives to fight are already marginal from the outset.
Finally, conflict is highly competitive by its nature. The very point of competition is
to take advantage of the vulnerabilities of an opponent. The fact that vulnerabilities are
mutual does not in itself ensure peace, any more than common mortality guarantees
that humans never fight. As such, trade is not only compatible with, but sufficient for,
practicing conflict. Nations contemplating war are already anticipating considerable
harm and inconvenience. The additional burden of economic loss must be substantial
before this can do more than marginally affect the willingness to fight. The notion that
nations will avoid war because trade is valuable ignores the fact that war itself involves
imposing and suffering substantial loss. It is the search for value that ignites competition
and the destruction of value that makes war coercive. Indeed, war in modern times typi-
cally involves economic conflict, as well as the forfeiture of human life.

Realism
The realist perspective on trade is no less venerable. Thucydides attributed the origins of
the Peloponnesian War to the rise of Athens as a commercial empire and the alarm that
this inspired in Lacedaemon. One can point to many other wars that have been fought
over trade. Liberal theory makes a strong case that nations are better off trading than
fighting. Yet history is replete with examples in which nations appeared to prefer fight-
ing to trading. The Anglo-Dutch wars of the seventeenth and eighteenth centuries wit-
nessed two trading states with republican governments fighting for control of European
and world commerce. The Anglo-French wars of the eighteenth and nineteenth cen-
turies shared many of the same dynamics of commercial rivalry that did not appear to
be mitigated by the rise of republicanism in France. While these and other historical
424   Trade and War

examples are familiar to contemporary researchers, fighting over trade has hardly been
reconciled with liberal theory.
Realist theories of war rely on economic development as the primary internal mecha-
nism for increasing state power. The connection is indirect; economic growth allows
for more internal balancing and the accumulation of military power (Morgenthau
1948; Waltz 1959; Mearsheimer 2001). Trade is generally believed to increase economic
efficiency and therefore the size of the surplus. To the degree that commerce aug-
ments the resources available to a state, commercial powers should be more capable
and—depending on variables like system structure, the local or regional balance, or
other relationships—trading states may then be better equipped for foreign adventures.
The realist view of trade dates back to Jean-Baptiste Colbert, Alexander Hamilton, and
Frederich List, and has been closely identified with mercantilism, protectionism, and
statism.
Realists view the effects of interdependence as at odds with the competitive logic
of politics under anarchy (Carr 1964; Krasner 1976; Waltz 1979; Mastanduno 1998).
Kenneth Waltz maintains that “close interdependence means closeness of contact and
raises the prospect of at least occasional conflict” (1970, 250). Elsewhere, Waltz argues
that the rise of globalization has widened inequalities between rich and poor states, pro-
ducing dependencies rather than interdependencies. “A world in which a few states can
take care of themselves quite well and most states cannot hope to do so is scarcely an
interdependent one” (1979, 159). To realists like Waltz, states differ from each other pri-
marily in terms of power. Trade has the effect of exacerbating imbalances by changing
relative capabilities, usually in favor of those states that already wield disproportionate
influence in world affairs. Operating under anarchy, states worry that a partner today
can become a threat in the future. Building on this logic, Gowa (1989) shows that eco-
nomic ties between countries can have large security-related consequences because the
real income from trade is a source of security externalities.
More recent research challenges key assumptions and findings of realist scholarship.
As discussed earlier in the chapter, realist theories do not provide logically coherent
mechanisms for how differences in relative capabilities or shifts in the balance of power
lead to war. More generally, realists have had trouble explaining the broader empirical
correlation between trade and peace. In addition, the logic of relative gains (Grieco 1988,
1990; Mastanduno 1991) has been criticized by Snidal (1991) and Powell (1991), who show
that concerns about relative gains break down as the number of actors involved grows.
In a world with many states, preoccupation with relative gains within one pair of states (a
dyad) disproportionately advantages third parties. Finally, recent scholarship (Gartzke
and Lupu 2012) challenges the realist account of World War I, the most often cited case of
the failure of interdependence to maintain peace. The “Great War” began in the Balkans,
where trade was local and vestigial and where disputes regularly flared into open war-
fare. Fighting then spread to states where trade had played an important role in manag-
ing previous conflict, but where commitments to the warring factions took precedence.
 Erik Gartzke and Jiakun Jack Zhang    425

Marxism
Marxists and realists both view trade as exacerbating political competition. However,
Marxism (and radicalism in general) focuses on imbalances of power not among great
powers but between weak states and the powerful centers of trade (core and periphery).
Marxists see the modern industrial state as captured by expansionist capitalist interests.
At first blush, the history of European colonialism seems to validate the Marxist per-
spective. The nations of Europe spent a considerable portion of the sixteenth through
the eighteenth centuries fighting over the very lucrative East Asia trade or fighting with
indigenous groups to impose mercantilist commercial systems. Private commercial
firms created armies and navies and fought wars to consolidate markets, protect their
profits, and prevent foreign competition. These efforts were then augmented by hybrid
“crown companies” that drew sovereigns directly into private enterprise. Crown compa-
nies were later eclipsed by the power of sovereign governments that had grown rich on
global commerce.
Socialists (Hobson 1905) and Marxists (Lenin 1916) believed that the capitalist need
for markets propelled class warfare abroad. Hobson (1905) argued that imperialism
results from the structural problem of overproduction and underconsumption. As
capitalists continue to reinvest their wealth in production, they soon exhaust demand
for goods in their domestic markets and must look for foreign markets to absorb the
surplus goods and capital that they can’t use at home. Lenin (1916) built on Hobson’s
idea of excess production to argue that capitalism is the primary source of interna-
tional wars, as more powerful nations exploit weaker ones for economic gains. Marxists
point to World War I as an example of a capitalist war, in which business interests pros-
pered from the war while millions of ordinary people lost their lives in the trenches.
Paradoxically, imperialism was also used to explain delay in the advent of revolution
in Europe, since foreign trade and investment put off the crisis between domestic class
interests.
Reacting to modernization theory, dependency theorists (Santos 1970; Wallerstein
1974) revived Marxist arguments during the Cold War. Dependency theory has since
been heavily criticized on economic and historical grounds. Exploitation of periph-
eral countries by the core exists, but appears more uneven than dependency theory
would imply. More important, successful economic modernization in East Asia, and
the manifest failure of import-substituting industrialization (ISI) in Latin America
appeared to refute many of the policy recommendations of dependency theory. Gilpin
argues that developing countries are dependent because they are “weak in a world of
the strong,” suffering not from external dependence but from internal inefficiency
(Gilpin 2011). Marxism was also discredited by the collapse of the Soviet empire.
However, normative concerns about the capture of state policy by business interests as
well as the role of international capital in war suggest the need for, and value of, further
inquiry.
426   Trade and War

Empirical and Formal

Theoretical Findings

Debate over the effects of trade on conflict has generated a tremendous amount of schol-
arship. A large volume of empirical work in the 1990s and 2000s produced contradictory
findings; various scholars demonstrated positive, negative, and indeterminate relation-
ships between trade and war. The liberal view that trade pacifies conflict dominates con-
temporary thinking, even though it has proven difficult to find definitive evidence for
the mechanisms proposed by available theories.
Realists have used similar data in an attempt to demonstrate that trade actually
increases conflict (Barbieri 1996, 2002). It is also possible that trade has no net effect on
warfare (Gartzke et al. 2001; Gartzke and Li 2003; Gartzke 2007), or that the effect of
commerce is ambiguous, given contrasting incentives and the impact of rational expec-
tations in conditioning trade prior to actual fighting (Morrow 1999). Ambiguity about
trade and conflict stems from the very forces highlighted by available perspectives. States
in competition have motives to fight, while trade provides a rationale for restraint, or at
least for moderation in managing tensions. The net effect of these contrasting effects is
difficult to divine. States enmeshed in intense trading relationships may find that these
present a costly or informative barrier to conflict. Other states may fight, despite trading
ties, if the issues at stake are sufficiently important. Still other nations will find that trade,
though valuable, is seldom sufficiently important in any given relationship to interrupt
the path to war. Table 22.1 offers a summary of available findings in the current empirical
literature.
The first section in the table lists studies finding that increased trade decreases con-
flict. Central to this set of studies is the finding that increased levels of trade are asso-
ciated with a significant reduction in interstate disputes. Polachek’s (1980) pioneering
empirical work first documents a negative statistical correlation between dyadic trade
and peace. Oneal and colleagues (1996, 1997, 1999, 2003, 2010) approach interdepen-
dence from the perspective of the Kantian peace. Their analysis shows that higher levels
of trade and democracy tend to lead to peace. They argue that the mechanism behind
the pacifying effect of trade is the classical liberal expectation that state participation in
interstate conflict is motivated by expected utility; trade increases the benefits of refrain-
ing from war and decreases incentives to fight.
The second section reviews studies finding that trade increases conflict among states.
Realists contradict the Kantian peace findings with empirical results that seem to show
that heightened trade increases the likelihood of country dyads engaging in militarized
interstate disputes but has no influence on the occurrence of wars (Barbieri 1996, 2002;
Barbieri and Levy 1999). Barbieri argues that there is a curvilinear relationship between
interdependence and conflict; low to moderate degrees of interdependence reduce the
likelihood of dyadic disputes, and extensive economic linkages increase the probability
of disputes.
Table 22.1  Summary of Empirical Findings on the Trade-War Relationship
Main Findings Author(s) Methodology and Unit of Analysis

Trade Lupu and Traag (2012) Regression, network

Decreases Dorussen and Ward (2010) Regression, network
Warfare
Hegre, Oneal, and Russett (2010) Regression, dyads
Oneal and Russett (2003) Regression, PRD
Oneal and Russett (1999) Regression, PRD
Polachek, Robst, and Chang (1999) Regression, dyads
Oneal and Russett (1997) Regression, PRD
Oneal and Ray (1997) Formal model, regression, dyads
Polachek (1997) Regression, PRD
Oneal, Oneal, et al. (1996) Regression, PRD
Copeland (1996) Case study, system
Mansfield (1994) Regression, dyads
Gasiorowski and Polachek (1982) Granger causality, system
Polachek (1980) Regression, dyads
Trade Barbieri (2002) Regression, dyads
Increases Barbieri and Levy (1999) Regression, dyads
Warfare
Barbieri (1996) Regression, dyads
Domke (1988) Case study, system
Indeterminate Li and Reuveny (2011) Regression, dyads
or No Effect Ward and Hoff (2007) Prediction, network
Ward, Siverson, and Cao (2007) Bayesian model, dyads
Keshk, Pollins, and Reuveny (2004) Regression, dyads
Gartzke and Li (2003) Replication, dyads
Gartzke (2007) Regression, dyads
Gartzke, Li, and Boehmer (2001) Regression, dyads
Dorussen (1999) Formal model, >2 actors
Morrow (1999) Formal model, 2 actors
Dorussen (1999) Granger causality, dyads
Reuveny & Kang (1996) Formal model, 2 actors
Gowa (1994) Regression, dyads
Gasiorowski (1986)

PRD = Politically relevant dyads.
428   Trade and War

A final section of Table 22.1 deals with studies whose findings are inconclusive or in
which only the null hypothesis (that there is no relationship between trade and conflict) is
substantiated. We discuss these nonrelationships more extensively in the next subsection.

Artifacts of Variable Selection

Dyad level quantitative studies from the 1990s on the effects of trade on conflict appear
to yield discrepant findings. However, these differences may just be the result of fea-
tures inherent in the variable selection used by the competing approaches (Gartzke and
Li 2001). Oneal and Russet (1997) operationalize interdependence as whether a par-
ticular trade relationship matters relative to a state’s overall economic performance
(trade/GDP—trade salience), while Barbieri (1996) operationalizes interdependence
as a function of whether trade is valuable relative to other trade relationships (trade/
trade—trade share). Gartzke and Li show that the effect of trade share is inversely
related to the consensus measure of openness (monadic trade interdependence). In
effect, the Barbieri measure captures disconnectedness from the world economy or
dependency aspects of trading relations. Table 22.2 illustrates the steps for constructing
each dyadic measure.
Gartzke and Li show that trade share is inversely related to trade openness. Trade
share thus tends to reflect a country’s lack of integration into the world economy. Hence,
the positive effect of trade share and the negative effect of openness do not necessarily

Table 22.2  Comparisons of Two Measures of Trade Interdependence

Studies Measure

Barbieri (1995,
(1) tradesharei =
(
importsij + exportsij
=
)
tradeij
1996, 1998b)
 
( importsi + exportsi ) tradei

(2) trade salienceij = trade sharei ∗ trade share j

 

(3) tradesymmetry ij = 1 − trade sharei − trade share j

 
(4) trade interdependenceij = trade salienceij ∗ trade symmetryij
 
Oneal & Russett
(5) trade dependenceij =
(imports ij + exportsij ) = trade ij
(cf. 1997, 1999)
GDPi GDPi
 

(6) trade dependence ji =

(imports ji + exports ji
=
)
trade ji
GDPj GDPj
 
(7)  trade interdependenceij = lower of (dependenceij and dependenceji)
(8)  trade asymmetryij = higher of (dependenceij and dependenceji)
 Erik Gartzke and Jiakun Jack Zhang    429

indicate inconsistencies in theory, data, or other sources. Instead, it is possible to attri-

bute the discrepancy to variable construction alone. The inverse relationship between
trade share and openness helps to resolve the puzzle of why Barbieri and Oneal and
Russett and others achieve discrepant findings. There need be no theoretical contradic-
tion between the positive correlation of the trade share measure and the negative cor-
relation of the trade dependence measure in analyzing interstate disputes, at least not at
the dyadic level. More economic linkages correlate with fewer disputes.

Problems of Endogeneity
Much of the empirical literature relies on cross-national time series data and does not
effectively deal with endogeneities between trade and conflict. While a robust correlation
exists between dyadic trade and interstate disputes, it is difficult to determine causality. It
is plausible that trade reduces war as the liberals claim, but it is equally likely that peace
leads to an increase in trade. Formal theoretical work pays more attention to the issue of
endogeneity than existing empirical research. James Morrow (1999) points out that the
endoegeneity problem in interdependence has been overlooked by empiricists and links
interdependence to conflict bargaining. Outcomes of peace and conflict hinge on infor-
mation decision makers have about each other’s relative bargaining positions. Because
actual trade flows (and, by inference, other observable economic relations) only provide
a sliver of relevant information about players’ relative resolve, Morrow finds economic
links have an indeterminate effect on conflict. The failure to endogenize decision makers’
choice to trade is a critical shortcoming of empirical studies. If trading states are integrated
because they anticipate the likelihood of peace, then one would expect to find a correla-
tion between trade and peace even without a causal connection (Benson and Niou 2007).

Beyond Dyads
Maoz and Russett (1993) started a trend of dyad year designs using militarized inter-
state dispute (MID) data that has been widely adopted in subsequent literature (see
column 2 in Table 22.1). However, the empirical relationships established using this
mode of analysis have been challenged by studies that point out problematic assump-
tions about spatial and/or temporal independence (Beck et al. 1998; Ward et al. 2007).
Because these observations are temporally dependent—a peaceful dyad today is likely
to persist in being peaceful tomorrow—Beck and colleagues advocate using a grouped
duration model over logit or probit regressions. Ward and colleagues show that the
statistical association in the literature stems from three components: (1) geographical
proximity, (2) dependence among MIDs with the same initiator or target, and (3) higher
order dependences in dyadic data. They conclude that the Kantian peace tripod loses its
predictive power once these factors are controlled for statistically.
Other scholars have pushed the analysis of interdependence beyond dyads. Commercial
enterprises can generally continue through third parties, even when conflicts interfere
430   Trade and War

with direct bilateral ties. Recognition of this fact has led to attempts to estimate the demand
elasticity of substitution present in sourcing goods and services elsewhere (Polachek et al.
1999; Crescenzi 2003). Others have tried to model the effects of trade networks on whether
states fight (Ward and Hoff 2007; Dorussen and Ward 2008, 2010; Lupu and Traag 2012;
Ward et al. 2013), suggesting that it is the cumulative indirect effects of trade and conflict
that are most important. The network approach makes more realistic theoretical assump-
tions about trade and allows for a better test of interdependence. At the same time, how-
ever, it obscures particular tendencies, making causal inference even harder.

Retheorizing Interdependence

As Beth Simmons notes, “the empirical project has largely been stylized as a ‘realist’ ver-
sus ‘liberal’ horse race. When formulated in this way, scholars are missing rich oppor-
tunities to specify the nature of state-society relations that link private profiteering with
national foreign (even military) policy” (Simmons 2003, 32). Focusing on mechanisms
rather than paradigms allows for a more eclectic research agenda capable of resolving
inconsistencies found in existing empirical work. Borrowing from Scott Kastner (2009),
we outline three ways in which trade can interact with war. Interdependence can con-
strain, inform, or transform political relationships.

Trade Constrains
Much of the literature relies on constraint as the key mechanism behind commercial
peace. Scholars in this tradition view trade as generating efficiency gains. Constraint
mechanisms begin with the assumption that military conflict disrupts valuable com-
mercial ties between economic partners that happen to be sovereign, independent
states. Because the disruptions caused by military conflict are costly for domestic actors
(export-oriented firms and consumers), these groups appeal to their government to
refrain from escalating crises to open conflict or war.
Constraints, also referred to as “opportunity costs,” are the mechanism most fre-
quently used by liberal scholars to account for commercial peace (Levy 2003). The
theory is often tied to democracy, since popular rule is reasonably believed to be more
sensitive to the needs of domestic constituents. As with the Kantian conception of lib-
eral political restraint on warlike monarchs, leaders may still want to go to war, for ter-
ritorial or nationalistic reasons. However, the disruption of commerce associated with
war leads domestic constituents to oppose these other objectives (McDonald 2009).
As a result, the leader avoids or de-escalates fighting, despite his or her initial prefer-
ences. The argument suggests in particular that one should see evidence of consider-
able lobbying at the firm level against war and associated with crises or escalation.
While opportunity costs could inhibit conflict, they need to be large enough to alter
the calculus of war. Marginal increases in the overall cost of fighting can at most have a
 Erik Gartzke and Jiakun Jack Zhang    431

marginal effect on whether conflict occurs. At the same time, factors that increase war
costs create leverage that opponents, even trade partners, can use to extract additional
concessions or increase the odds that an adversary concedes rather than fighting. Having
more reasons not to fight makes it easier for other states, even other trade partners, to make
more extractive demands, since the nominal risk that the opponent will refuse is lower.
Finally, introducing domestic politics further complicates the constraint argument. Some
economic actors will favor peace, but others may prefer to fight (Brooks 2013). Indeed, the
practice and study of tariffs and other trade barriers suggests that import-competing fac-
tors or sectors might favor using military force under some circumstances. War can act as
a tariff, allowing import-competing industries and workers to receive higher returns on
their inputs to production. The net effect of these tendencies is again unclear.

Trade Informs
A second set of mechanisms focuses on the role of information and attempts to explain
the relationship between commerce and conflict within a bargaining framework
(Fearon 1995; Gartzke 1999; Morrow 1999; Powell 2002). As discussed previously in this
chapter, wars often arise due to bargaining failures driven by uncertainty about an oppo-
nent’s resolve or capabilities. In these models, military disputes result when leaders mis-
judge the relative commitment of their opponents. Both sides benefit from overstating
the level of their commitment, hoping that the opponent will back down. Wars result
when at least one side underestimates an adversary’s commitment, assuming that the
adversary is bluffing when in fact he is resolved. To overcome the problem of incomplete
information, leaders must demonstrate their commitment through costly acts such as
tying hands or sinking costs (Fearon 1997). Costly signals avoid the cheap talk problem,
backing up words with action.
Cutting off trade is one way that leaders can communicate resolve during crises.
Without economic interdependence, threats have relatively little cost until one side
escalates to military violence. Economic interdependence creates a middle step in the
escalation ladder between war and peace. Making threats or taking actions that harm
commerce is costly to both parties in an interdependent relationship. Therefore, as the
degree of economic interdependence increases, the costs involved in threatening war
rise as well (as merchants and investors abandon markets when and where war becomes
more likely), ensuring that leaders more credibly communicate resolve. Economies that
are well integrated into the global markets face the risk of capital flight when conflict is on
the horizon. Markets are thus a credible mechanism for revealing information, because
they offer leaders a way to signal resolve that is costly but also short of military violence.
A few scholars have studied the informational effects of trade on conflict (Gartzke
et al. 2001; Gartzke 2007). However, the foundations of this mechanism remain under-
theorized and in need of additional empirical support. This is thus a fertile area for
future work. At the same time, the impact of informational effects of interdependence
may (also) be marginal, as the reduction in uncertainty will at most be partial, and there
are many reasons that states do not fight.
432   Trade and War

Trade Transforms
A final set of mechanisms involves the potential for economic ties to transform the pref-
erences of political actors. Transformation is at the core of Richard Rosecrance’s (1986)
notion of the trading state, in which he argues that changes in the world economy have
led modern states to become less reliant on territory than on commerce. A related argu-
ment notes that trade is more efficient than military conquest as a way to acquire goods
and services, just as buying things at the store is often more efficient than stealing them,
even for thieves and bank robbers.
Whereas the first two mechanisms treat the interests of leaders as exogenous and
fixed, the transformation predicts that heightened economic exchange will change not
just the payoffs, but also the preferences of decision makers. In other words, economic
integration harmonizes the goals and interests of interdependent states. The most
salient example of this is the integration of Europe after World War II. In advancing
the Marshall Plan, American policy makers argued for rebuilding Germany’s economy
alongside the rest of Western Europe, in order to tie German interests to peace and pros-
perity in the West. Together, the Marshall Plan and the rise of European economic and
political institutions transformed European geopolitics. Economic integration ensured
development and growth, enabled economic cooperation, reduced strategic mistrust,
and created bonds of common interest between historical rivals. International com-
merce catalyzed fundamental changes to the culture, civil society, and political institu-
tions of these states. As the exchange of capital, goods, people, and ideas across borders
increases, the preferences of leaders are said to change, such that conquest is no lon-
ger considered a legitimate tool of foreign policy. Indeed, despite recent turmoil in the
Eurozone, European leaders and their populations would find another continental war
like those of the twentieth century inconceivable.
The transformation interpretation of trade and conflict scores points for being
optimistic and appearing to offer the potential for dramatic change. However, both
Rosecrance’s view and later arguments leave it unclear whether preferences or payoffs
are changing. In a sense the latter is a better option, since this does not require the chang-
ing of hearts and minds, but only the filling of pockets. It would, for example, be easier to
imagine that stability in Asia is possible because most states benefit intensely from trade,
than to argue that preferences must converge and that nations such as China, Japan, and
the Koreas are no longer willing even to contemplate the use of force.

Directions for Future

Research and Inquiry

The relationship between trade and war remains an important and appealing area for
future research. Despite considerable attention and a substantial amount of plausible
conjecture and speculation, there is much that remains in doubt. Growing consensus
 Erik Gartzke and Jiakun Jack Zhang    433

about the theory of war can play a critical role in building consensus about cause and
effect. Trade is most likely to impact peace where it most closely influences the causes of
war. Explanations of the effects of trade on conflict that do not address a coherent theory
of conflict must thus be held in some doubt. This creates an opportunity for refining
theory by evaluating existing claims in terms of causal logic.
The study of interdependence can also be improved by more rigorous assessment of
trade at different levels of analysis. Earlier studies measure the cross-national effects
of  trade on war, but existing causal mechanisms make unsubstantiated assumptions
about the preferences of domestic actors. Scholars can begin to better understand the
domestic microfoundations of interdependence by incorporating the latest research on
public opinion, firm level preferences toward trade, and distributional effects of trade
on local politics. Rather than rely on national level measures of trade share or trade
salience, scholars may choose to examine the behavior of key domestic political actors
for evidence of causal mechanisms. Evidence that firms are not significantly harmed by
conflict would cast doubt on the assumption that export-oriented firms constrain con-
flict by lobbying for peace. Eliminating mechanisms with no empirical grounding at the
subnational level would allow scholars to hone in on the drivers of commercial peace.
Existing theories must also be shaped through dialogue with other strands of research.
Current thinking on trade and peace pays very little attention to the best information
about trade economics, corporate strategy, or political economyresearch. Assumptions
in existing studies tend to be crude and the models of trade and politics tend to be overly
simplistic. For example, much of the literature uses aggregate bilateral trade to make
claims about the preferences of private actors rather than studying these actors directly.
Studies of interdependence must engage more with emerging scholarship in the inter-
national political economy literature using firm-level and product-level data.
As Kastner (2009) points out, an understanding of interdependence requires a thor-
ough understanding of the domestic determinants of trade policy. However, existing
studies of trade and conflict rarely focus on domestic actors and their preferences.
Michael Plouffe, in this volume, presents an analysis of the political economy of trade
emphasizing the political role of heterogeneity in firm preferences. Firms’ preferences
over trade policy are dictated by their ability to engage foreign markets. His data show
that in comparative advantage industries, low productivity firms pursue import protec-
tion, while in comparative disadvantage industries, high productivity firms seek trade
liberalization. Both of these firm level behaviors contradict classical models of trade
politics, which ignore firm level differentiation within industries. These firm-level dif-
ferences matter for interdependence because conflict would disrupt trade for some
firms more than others. Whether or not the constraint mechanism works should there-
fore depend on which set of firms can sway policy in their favor.
The literature on the domestic determinants of foreign policy is also informative.
To establish common microfoundations for trade and war, the preferences of relevant
domestic actors (policy makers, firms, consumers, etc) must be linked with foreign pol-
icy outcomes. This is easier to do for some types of outcomes, particularly in democra-
cies. Kleinberg and Fordham (2013) investigate whether domestic distributional effects
of trade affect support for hostile foreign policies toward China in the U.S. Congress.
434   Trade and War

They find that the export orientation and import dependency of congressional districts
influence many members’ decisions to take a tough stance against China in cosponsor-
ship and roll-call voting. More work is needed to examine the influence of economic
interests arising from trade on other aspects of the foreign-policy-making process. The
role that domestic politics play in the decisions made by chief executives, foreign min-
istries, and militaries is not systematically understood. But because trade clearly has
important domestic distributional consequences the interdependence literature should
incorporate domestic politics to help better understand how, and when, trade is able to
constrain, inform, or transform conflict.
In summary, students of interdependence will benefit by being more explicit about the
unit of analysis and the agency of relevant actors. Existing research take the nation-state
as the unit of analysis and does not often account for the agency of domesticactors, be
they trading firms or policymakers,. However, as other chapters in this volume indi-
cate, scholars of trade are turning increasingly to consumer, firm, and product level
data. The challenge ahead for the trade and war literature is to make sense of these new
findings about trade at the sub-national level and connect them to foreign policy at the
international level.

Conclusion

Existing paradigmatic approaches to economic interdependence remain underspecified

and the causal mechanisms untested. Too much of the literature consists of empirical
findings intended to resolve theoretical debates, but since the evidence is inconclusive,
and theories are imprecise, less has been learned than many would like. As Mansfield
and Pollins noted in their review of the literature over a decade ago, “existing research
has focused too much on addressing whether there is a relationship between interde-
pendence and conflict and too little on identifying the underlying micro-foundations
of any such relationship” (2003). This chapter has argued that future work can improve
upon the existing literature by adopting common assumptions about the causes of war
and building a better understanding of the domestic sources of trade and foreign policy.
A microfoundations approach will unify these two disparate literatures and offer new
insights about the nature of trade and war. Three basic mechanisms—constraints, infor-
mation, and transformation—appear to be likely candidates as drivers of the key find-
ings in the literature. The chapter also offered some ideas about how to operationalize
and test these mechanisms, and their strengths and weaknesses.
It should also be noted that the existing literature has focused largely only on the
impact of economic interdependence on conflict onset. Studying the effects of trade
within a rationalist framework can open up new areas for research. The bargaining per-
spective promises to link war onset, initiation, prosecution, termination, and conse-
quences into a single overarching theoretical framework. The role of trade could also be
examined in each of these stages (not just onset).
 Erik Gartzke and Jiakun Jack Zhang    435

Looking beyond trade, future research might also examine other dimensions of
commerce and their effects (positive and negative) on war and peace. Rosecrance and
Thompson (2003) argue that foreign direct investment (FDI) represents a link that is
more costly than trade to break; thus FDI links between countries are more likely to
reduce conflict than trading links. Other interdependencies such as capital markets
(equity, bond, forex, etc.) may play even greater roles in constraining state discretion
(Gartzke 2007). After all, investment is much more volatile than the exchange of goods
and services, and the volume of exchange is also much larger.
Finally, while this chapter has focused on the positive impact of trade on war and
peace, disruptions to the flow of commerce also merit study. The role of economic sanc-
tions, such as those deployed by the Western powers against Russia in 2014, and con-
sumer boycotts, such as those organized in China against Japanese goods in 2012, in
conflict are other areas ripe for further research. Examining the role of commerce in
war, not just the positive impact of trade, greatly broadens the potential for future study.

Note
1. For a discussion of the role of international economic institutions in this process, see
Gowa chapter in this volume.

References
Angell, S. N. 1910. The Great Illusion. New York: The Knickerbocker Press.
Barbieri, K. 1996. Economic Interdependence:  A  Path to Peace or a Source of Interstate
Conflict? Journal of Peace Research 33 (1): 29–49.
Barbieri, K. 2002. The Liberal Illusion: Does Trade Promote Peace? Ann Arbor: University of
Michigan Press.
Barbieri, K., and J. S. Levy. 1999. Sleeping with the Enemy: The Impact of War on Trade” Journal
of Peace Research 36 (4): 463–479.
Beck, N., J. N. Katz, et al. 1998. Taking Time Seriously: Time-Series-Cross-Section Analysis
with a Binary Dependent Variable. American Journal of Political Science 42 (4): 1260–1288.
Benson, B. V., and E. M. Niou. 2007. Economic Interdependence and Peace: A Game-Theoretic
Analysis. Journal of East Asian Studies 7 (1): 35–59.
Blainey, G (1988). Causes of War. New York: Simon and Schuster.
Brooks, S. G. 2013. Economic Actors’ Lobbying Influence on the Prospects for War and Peace.
International Organization 67 (04): 863–888.
Carr, E. H. 1964. Twenty Years’ Crisis, 1919–1939. New York: Harper Perennial.
Copeland, D. C. 1996. Economic Interdependence and War: A Theory of Trade Expectations.
International Security 20 (4): 5–41.
Crescenzi, M. J. 2003. Economic Exit, Interdependence, and Conflict. Journal of Politics 65
(3): 809–832.
De Mesquita, B. B., and A. Smith. 2005. The Logic of Political Survival. Cambridge, MA: MIT
Press.
Domke, W. K. 1988. War and the Changing Global System. New Haven, CT: Yale University Press.
436   Trade and War

Dorussen, H. 1999. Balance of Power Revisited: A Multi-country Model of Trade and Conflict.

Journal of Peace Research 36 (4): 443–462.
Dorussen, H., and H. Ward. 2008. Intergovernmental Organizations and the Kantian Peace
A Network Perspective. Journal of Conflict Resolution 52 (2): 189–212.
Dorussen, H., and H. Ward. 2010. Trade Networks and the Kantian Peace. Journal of Peace
Research 47 (1): 29–42.
Doyle, M. W. 1983. Kant, Liberal Legacies, and Foreign Affairs. Philosophy & Public Affairs 12
(3): 205–235.
Doyle, M. W. 1986. Liberalism and World Politics. American Political Science Review 80(4):
1151–1169.
Fearon, J. D. 1995. Rationalist Explanations for War. International Organization 49: 379–379.
Fearon, J. D. (1997). Signaling Foreign Policy Interests Tying Hands versus Sinking Costs.
Journal of Conflict Resolution, 41(1): 68-90.
Gartzke, E. 2007. The Capitalist Peace. American Journal of Political Science 51 (1): 166–191.
Gartzke, E. 1999. War Is in the Error Term. International Organization 53 (3): 567–587.
Gartzke, E., and Q. Li. 2003. Measure for Measure: Concept Operationalization and the Trade
Interdependence-Conflict Debate. Journal of Peace Research 40 (5): 553–571.
Gartzke, E., Q. Li, et  al. 2001. Investing in the Peace:  Economic Interdependence and
International Conflict. International Organization 55 (02): 391–438.
Gartzke, E., and Y. Lupu. 2012. Trading on Preconceptions: Why World War I Was Not a Failure
of Economic Interdependence. International Security 36 (4): 115–150.
Gasiorowski, M., and S. W. Polachek. 1982. Conflict and Interdependence East-West Trade and
Linkages in the Era of Detente. Journal of Conflict Resolution 26 (4): 709–729.
Gasiorowski, M. J. 1986. Economic Interdependence and International Conflict:  Some
Cross-National Evidence. International Studies Quarterly 30 (1): 23–38.
Gelpi, C., and J. M. Grieco. 2003. Economic Interdependence, the Democratic State, and the
Liberal Peace. Economic Interdependence and International Conflict: New Perspectives on an
Enduring Debate: 44–59.
Gelpi, C. F., and J. M. Grieco. 2008. Democracy, Interdependence, and the Sources of the
Liberal Peace. Journal of Peace Research 45 (1): 17–36.
Gilpin, R. 2011. Global Political Economy:  Understanding the International Economic Order.
Princeton, NJ: Princeton University Press.
Gowa, J. 1989. Bipolarity, Multipolarity, and Free Trade. American Political Science Review 83
(4): 1245–1256.
Grieco, J. M. 1988. Realist theory and the problem of international cooperation: analysis with
an amended prisoner’s dilemma model. The Journal of Politics 50 (3): 600-624.
Grieco, J. M. 1990. Cooperation among nations:  Europe, America, and non-tariff barriers to
trade. New York: Cornell University Press.
Hegre, H., J. R. Oneal, et al. 2010. Trade Does Promote Peace: New Simultaneous Estimates of
the Reciprocal Effects of Trade and Conflict. Journal of Peace Research 47 (6): 763–774.
Hobson, J. A. 1905. Imperialism: A Study. Great Britain: Spokesman Books.
Kastner, S. L. 2009. Political Conflict and Economic Interdependence across the Taiwan Strait
and Beyond. Stanford, CA: Stanford University Press.
Keshk, O. M., B. M. Pollins, et al. 2004. Trade Still Follows the Flag: The Primacy of Politics
in a Simultaneous Model of Interdependence and Armed Conflict. Journal of Politics 66
(4): 1155–1179.
 Erik Gartzke and Jiakun Jack Zhang    437

Kleinberg, K. B., and B. O. Fordham. 2013. The Domestic Politics of Trade and Conflict.
International Studies Quarterly.
Krasner, S. D. 1976. State Power and the Structure of International Trade.” World Politics 28
(3): 317–347.
Lenin, V. I. 1916. Imperialism: The Highest Stage of Capitalism. Australia: Resistance Books.
Levy, J. S. 2003. Economic Interdependence, Opportunity Costs, and Peace. Economic
Interdependence and International Conflict: New Perspectives on an Enduring Debate: 127–147.
Li, Q., and R. Reuveny. 2011. Does Trade Prevent or Promote Interstate Conflict Initiation?
Journal of Peace Research 48 (4): 437–453.
Lupu, Y., and V. Traag. 2012. Trading Communities, the Networked Structure of
International Relations and the Kantian Peace. Journal of Conflict Resolution. doi:
10.1177/0022002712453708
Mansfield, E. D. (1994). Power, Trade, and War. Princeton, NJ.: Princeton University Press.
Mansfield, E. D., and B. M. Pollins. 2003. Economic Interdependence and International Conflict:
New Perspectives on an Enduring Debate. Ann Arbor: University of Michigan Press.
Maoz, Z., and B. Russett. 1993. Normative and Structural Causes of Democratic Peace,
1946–1986. American Political Science Review 87 (3): 624–638.
Mastanduno, M. 1991. Do relative gains matter? America’s response to Japanese industrial pol-
icy. International Security 16 (1): 73-113.
Mastanduno, M. 1998. Economics and Security in Statecraft and Scholarship. International
Organization 52 (4): 825–854.
McDonald, P. J. 2009. The Invisible Hand of Peace:  Capitalism, the War Machine, and
International Relations Theory. Cambridge, UK: Cambridge University Press.
Mearsheimer, J. J. 2001. The Tragedy of Great Power Politics. United States of America: W.W.
Norton.
Morgenthau, H. J. 1948. Politics Among Nations: The Struggle for Power and Peace. New York:
Alfred Knopf.
Morrow, J. D. 1999. How Could Trade Affect Conflict? Journal of Peace Research 36 (4):
481–489.
Mousseau, M. 2010. Coming to Terms with the Capitalist Peace. International Interactions 36
(2): 185-213.
Mousseau, M. 2000. Market Prosperity, Democratic Consolidation, and Democratic Peace.
Journal of Conflict Resolution 44 (4): 472–507.
Mousseau, M. 2003. The Nexus of Market Society, Liberal Preferences, and Democratic Peace:
Interdisciplinary Theory and Evidence. International Studies Quarterly 47 (4): 483–510.
Niou, E. M., P. C. Ordeshook, et  al. 2007. The Balance of Power:  Stability in International
Systems. Cambridge, UK: Cambridge University Press.
Oneal, J., and B. Russett. 1999. The Kantian Peace. World Politics 52 (1): 1–37.
Oneal, J. R., F. H. Oneal, et  al. 1996. The Liberal Peace:  Interdependence, Democracy, and
International Conflict, 1950–85. Journal of Peace Research 33 (1): 11–28.
Oneal, J. R. and J. L. Ray. 1997. New Tests of the Democratic Peace: Controlling for Economic
Interdependence, 1950–85. Political Research Quarterly 50 (4): 751–775.
Oneal, J. R., and B. Russett. 1999. Assessing the Liberal Peace with Alternative Specifications:
Trade Still Reduces Conflict. Journal of Peace Research 36 (4): 423–442.
Oneal, J. R., and B. M. Russet. 1997. The Classical Liberals Were Right:  Democracy,
Interdependence, and Conflict, 1950–1985. International Studies Quarterly 41 (2): 267–294.
438   Trade and War

Oneal, J. R., B. Russett, et al. 2003. Causes of Peace: Democracy, Interdependence, and Inter­
national Organizations, 1885–1992. International Studies Quarterly 47 (3): 371–393.
Polachek, S. W. 1980. Conflict and Trade. Journal of Conflict Resolution 24 (1): 55–78.
Polachek, S. W. (1997). Why Democracies Cooperate More and Fight Less: The Relationship
between International Trade and Cooperation. Review of International Economics 5 (3):
295–309.
Polachek, S. W., J. Robst, et  al. 1999. Liberalism and Interdependence:  Extending the
Trade-Conflict Model. Journal of Peace Research 36 (4): 405–422.
Powell, R. 1991. Absolute and Relative Gains in International Relations Theory. American
Political Science Review 85 (4): 1303–1320.
Powell, R. 2002. “Bargaining Theory and International Conflict. Annual Review of Political
Science 5 (1): 1–30.
Powell, R. 1999. In the Shadow of Power: States and Strategies in International Politics. Princeton,
NJ: Princeton University Press.
Powell, R. 2006. War as a Commitment Problem. International Organization 60 (01): 169–203.
Reuveny, R., and H. Kang. 1996. International Trade, Political Conflict/Cooperation, and
Granger Causality. American Journal of Political Science 40 (3): 943–970.
Rosecrance, R., and P. Thompson. 2003. Trade, Foreign Investment, and Security. Annual
Review of Political Science 6 (1): 377–398.
Rosecrance, R. 1986. The Rise of the Trading State:  Commerce and Conquest in the Modern
World. New York: Basic Books.
Santos, T. D. (1970). The structure of dependence. The American Economic Review. 60 (2):
231-236.
Schultz, K. A. (2001) Democracy and Coercive Diplomacy. Vol. 76. United Kingdom: Cambridge
University Press.
Schumpeter, J. A. 1962. Capitalism, Socialism and Democracy. New York: Harper & Row.
Simmons, B. (2003). ‘Pax Mercatoria and the Theory of the State’. Economic Interdependence
and International Conflict: New Perspectives on an Enduring Debate: 31–43.
Snidal, D. 1991. Relative Gains and the Pattern of International Cooperation. American Political
Science Review 85 (3): 701–726.
Wagner, R. H. 2007. War and the State. Ann Arbor: University of Michigan Press.
Wallerstein, I. (1974). The rise and future demise of the world capitalist system: concepts for
comparative analysis. Comparative studies in society and history 16(4): 387-415.
Waltz, K. 1970. The Myth of National Interdependence. In The International Corporation,
edited by C. P. Kindleberger, Cambridge, MA: MIT Press.
Waltz, K. N. 1959. Man, the State, and War:  A  Theoretical Analysis. New  York:  Columbia
University Press.
Waltz, K. N. 1979. Theory of International Politics. New York: McGraw-Hill.
Ward, M. D., J. S. Ahlquist, et al. 2013. Gravity’s Rainbow: A Dynamic Latent Space Model for
the World Trade Network. Network Science 1 (01): 95–118.
Ward, M. D., and P. D. Hoff. 2007. Persistent Patterns of International Commerce. Journal of
Peace Research 44 (2): 157–175.
Ward, M. D., R. M. Siverson, et  al. 2007. Disputes, Democracies, and Dependencies:
A Reexamination of the Kantian Peace. American Journal of Political Science 51 (3): 583–601.
 Chapter 23

Tr ade and Env i ronme nt

J. Samue l Bark in

This chapter examines two of the ways in which international trade is linked to the
natural environment in the literatures on both trade and environment. The first way
addresses the question of whether trade in general is good or bad for the environment.
At the general level, there is no clear answer to this question. The answer depends on the
specific version of the question asked and on the counterfactual posited. Nonetheless,
reviewing the manners in which the question can be asked and what the counterfactuals
might be is a useful exercise in thinking about the mechanisms through which interna-
tional trade affects the natural environment.
The second way looks more specifically at the relationship between international
trade institutions and environmental regulation. This is the terrain onto which much
of the politics of the relationship between trade and environment has been mapped.
Critics of globalization claim that these institutions constrain the ability of national gov-
ernments to make environmental policy. Furthermore, they fear that international trade
rules will undermine the use of trade sanctions to enforce multilateral environmental
agreements (MEAs). However, a careful look at the restrictions that international trade
institutions and agreements place on environmental regulation, whether national or
international, suggests a more nuanced reading. International rules do indeed constrain
national autonomy to use trade restrictions, but in ways that do not necessarily under-
mine effective environmental regulation. And while there is no proof that these rules
will not undermine MEAs, there is also no evidence that they are likely to.
The third section of this chapter looks at the politics of current negotiations over the
trade and environment relationship. There is little movement in these negotiations, as
has been the case since the beginning of this century. This is so both in trade negotia-
tions, such as the Doha Round negotiations, and (to a lesser extent) in the creation of
new environmental rules that have potential trade-restricting effects. This relative stasis
has happened in part because of the structure of the relevant international institutions,
which are biased toward the trade status quo. In large part, however, the stasis reflects
the current state of international politics. North-South tensions hinder change in the
trade-environment relationship, and there are too few states interested in prioritizing
440   Trade and Environment

international environmental regulation to overcome these tensions to create major new

regulatory mechanisms.

The Political Economy of Trade

and the Environment

The effects of international trade on the natural environment are complicated. At first
glance, the question of what these effects are might seem to be one purely of economics
and environmental science. But the question itself is a political one; what kind of trade
is considered and how, what environmental effects are focused on, and what counter-
factuals are chosen are all political statements. Proponents of international trade argue
that it makes economies more efficient and therefore results in more efficient uses of
the natural environment (Bhagwati 2000). Opponents argue that since trade increases
economic activity, and economic activity in turn degrades the environment, trade con-
tributes to environmental crisis (Princen, Maniates, and Conca 2002). Both positions
are simplistic (and most scholars of the relationship recognize that it is more nuanced).
But at the same time, both contain some truth.
One way to disaggregate the environmental effects of trade on the natural environ-
ment is to look at two distinctions. The first is between pollution and resource use. In
economic terms, resource use is a primary economic activity, and pollution is a nega-
tive externality. Resources, furthermore, can be both renewable and nonrenewable.
Somewhat counterintuitively, renewable resources are more prone to be overexploited.
As nonrenewable resources become scarcer, their price increases, leading to less quan-
tity demanded and greater investment in new sources and new technologies for extrac-
tion. The system, in other words, self-equilibrates. Renewable resources, however, can
be exploited too quickly, undermining their ability to reproduce themselves. Fish and
forests, in other words, are easier to use up than are iron or petroleum.
The second distinction is between local and global effects. Trade can relieve the pres-
sure on a resource base in a particular place by generating access to resources from afar,
for example by making food available in places that are suffering from drought. Or it
can add to pressure on a local resource base by allowing a global market access to those
resources. Allowing international trade in American shale gas, for instance, would
increase both extraction rates and prices in the United States. Similarly, trade can affect
global pollution levels by increasing levels of emissions of pollutants such as carbon
dioxide. But it can also have the effect of moving pollution from place to place, making
some places (usually richer ones) cleaner and others (usually poorer ones) dirtier.
The key argument that increasing international trade is benign or even actively ben-
eficial for the environment is based on the efficiency gains from trade. Trade allows pro-
duction to happen where it can be done most cost-efficiently. Some of this efficiency
comes from different comparative advantages, both natural and infrastructural, and
 J. Samuel Barkin    441

from efficiencies of scale. Therefore, for any given level of economic activity, production
is likely to happen more resource-efficiently the greater the level of international trade
(Bhagwati 2007). This argument clearly applies more directly to problems of resource
use than to pollution.
A related argument is that the economic growth that is generated by increasing lev-
els of international trade will lead countries to clean up their local environments. This
argument takes as its starting point what is called the environmental Kuznets curve,
which suggests that there is an inverted U-shaped relationship between national levels
of economic activity and pollution. As countries develop from lower- to middle-income
status, local pollution levels tend to get worse. But they level off in the middle-income
levels, and as countries develop from there they tend to focus more on cleaning up
their local environments. To the extent that international trade generates economic
growth, therefore, increases should be expected to make poor countries dirtier, but
middle-income countries cleaner. It should be noted, however, that this relationship
applies primarily to local pollution levels, such as soil pollution or smog. More global
forms of pollution, such as greenhouse gas emissions, do not seem to follow an environ-
mental Kuznets curve (Dinda 2004).
A final potential mechanism for trade to have positive environmental effects is
through the application of consumer power. On the one hand, international trade allows
consumers to affect the behavior of corporations or governments abroad. Examples
range from organized consumer boycotts of Chilean sea bass to opposition among
European consumers to genetically modified food crops. In the former case the boycott
helped force fishers in the southern oceans, generally not from the boycotting coun-
tries, into international regulatory mechanisms (Jacquet et al. 2010). In the latter case,
the aversion has kept some producers in commercial agriculture outside of Europe
from using such crops, when they otherwise would likely have adopted them (Paarlberg
2009). While consumer power can be effective in specific instances, however, it is not
viable as a systematic mechanism for generating global environmental governance.
On the other hand, increasing international trade might hurt the environment in a
number of ways. Broadest among them is the possibility of limits to growth. This argu-
ment begins with the key selling point of international trade, the equation between
increased trade and economic growth. Both the planet’s natural resource base and its
ability to absorb pollution are finite. To the extent that increased economic activity uses
more resources and generates more pollution, there must eventually be limits to eco-
nomic growth (Meadows et al. 1972). To the extent that increased trade, or freer trade,
generates increased economic activity, it brings us closer to those limits. This argument
is part of a broader critique of modernization in critical environmental thinking and is
distinct from the arguments (discussed below) that identify the regulation of economic
activity, rather than the activity itself, as the problem (Eckersley 1992).
A final argument that links increases in international trade with environmental deg-
radation focuses on the process of trade rather than production and final consump-
tion, and in particular on the transportation of goods between countries. Simply put,
international trade generally means shipping goods greater distances than would be the
442   Trade and Environment

case with domestic production (e.g., Gallagher 2005). This argument has made its way
into the popular discourse, most notably in the local food movement, which speaks of
“food miles,” the distance any particular bit of food has traveled between its original
source and the table, as necessarily bad, in part because transportation over greater dis-
tances leads to higher levels of carbon emissions in transportation (e.g., MacGregor and
Vorley 2007).
Which of these various arguments is right? There have been many empirical studies of
the effects of trade on the natural environment, but the conclusions of these studies tend
to be quite sensitive to the precise question being asked and specification of the model
being tested (Frankel and Rose 2005). In other words, the world is complicated enough
that one can generally find some evidence for most arguments linking trade and the
environment if one looks in the right places. More broadly, both the economic efficiency
and the limits to growth arguments are right, in different ways. International trade in
all likelihood does make the global use of resources more efficient, other things being
equal. But it also increases the aggregate level of economic activity. So resources can be
used more efficiently and more intensively at the same time.
The efficiency argument is focused on the resource-use part of the environmen-
tal effects of trade. It does not speak as clearly to the pollution effects. Furthermore, it
depends on what kinds of effects one is speaking of. Local economic activity can cause
critical levels of local pollution absent international trade, and in many instances there
is no clear connection between increased trade and pollution even when they covary
(Frankel and Rose 2005). However, it is also the case that local pollution from industrial
activities in various places can cause global effects. For example, the global climate in the
middle of the twentieth century was probably kept cooler than would otherwise have
been the case because of the accumulated effects of localized particulate pollution.
The evidence with respect to the transportation effects of trade, such as it is, is sim-
ilarly equivocal. The relationship between the distance goods travel from point of
production to point of consumption and the amount of pollution generated in trans-
portation is not linear. Different forms of transportation have very different levels of fuel
efficiency per amount transported—shipping goods by sea between ports half a world
apart can use less fuel than trucking those goods a few hundred miles within a country
(Barkin 2003). And the efficiency gains, energy and otherwise, generated by interna-
tional trade can more than offset the effects of additional transportation. For example,
growing tomatoes in North Africa in winter and shipping them to northern Europe can
be more energy efficient than growing those same tomatoes in a greenhouse next door to
the point of consumption (on food miles more broadly, see Weber and Matthews 2008).
Ultimately, economic modernization and industrialization are driving general global
levels of resource use and pollution. The amount of, and rules of, international trade
may affect these levels at the margins, but they are not the driving force behind the eco-
nomic expansion of the last two centuries and its attendant environmental effects. By
the same token, changing levels of and rules governing international trade can exac-
erbate or mitigate the environmental effects of global economic activity at the margins
but are unlikely to change basic patterns of pollution or resource exploitation. There are
 J. Samuel Barkin    443

exceptions to this generalization, as discussed below, but these tend to be for specific
pollutants or resources, not for more general patterns of economic activity.
One exception to the generalization is the possibility that increased trade inhibits
national environmental regulation. This can happen through both regulatory races to
the bottom and pollution havens. The regulatory race to the bottom argument is that
trade puts pressure on countries to attract investment in a globalized market. One way
that governments can attract investment is by lowering regulatory costs. This should
lead states to compete with each other to lower these costs, and the more open the inter-
national economy becomes, the greater this competition should be. Arguments about
regulatory races to the bottom can be found with respect to a variety of international
standards, including labor as well as environmental (Rodrik 1997).
Rather than a generalized pattern of laxer environmental regulation in response to
international trade, the pollution haven argument posits that increased economic glo-
balization generally, and international trade specifically, will generate dramatically
increased pollution (or, in some instances, unsustainable levels of resource extraction)
in specific sites (e.g., Clapp 2002). By this logic, international trade allows the negative
externalities of economic activity to be concentrated in those locations that are either
most desperate or have the least effective regulatory structures: the poorest and least
well governed of countries.
How big are these regulatory effects? The evidence here is, again, mixed. Evidence for
environmental regulatory races to the bottom, in which increased international trade
or globalization more generally causes a generalized reduction in environmental stan-
dards across countries, is scarce (Drezner 2001). Environmental compliance is gener-
ally a small enough proportion of companies’ total costs that regulatory competition
on this front is not an effective strategy for states to attract investment. There is even
some evidence of the opposite effect, in which globalization leads transnational corpo-
rations to push local regulatory standards up toward international norms, as the corpo-
rations seek regularization of regulatory levels across countries where they operate (e.g.,
Garcia-Johnson 2000).
Pollution havens, conversely, are a real problem, as poorer countries accept the rich
world’s industrial detritus, either for disposal or for recycling (Clapp 2001). Examples of
this effect can be found with used electronic equipment and parts sent to China, retired
ships sent to India, or toxic chemicals sent to West Africa. A similar effect can be found
with some forms of resource use, such as depleting forests for the international market,
as has happened, and continues to happen, in Southeast Asia. It should be noted, how-
ever, that while pollution havens are an effect of international trade, they are not neces-
sarily an effect of freer trade.
This distinction, between the effects of international trade in general and the effects
of freer trade in particular on the natural environment, is an important one. Many of
the most environmentally damaging aspects of international trade are those least cov-
ered by free trade agreements, particularly primary products such as agricultural goods,
minerals, and fuels. This is ironic inasmuch as much of the criticism of trade from an
environmental perspective happens within the context of trade institutions such as
444   Trade and Environment

the World Trade Organization (WTO) and North American Free Trade Agreement
(NAFTA). These agreements in turn have been much more effective at freeing trade in
manufactured than in primary products.
And yet much of the environmental harm related to international trade is gener-
ated by trade in primary products, such as natural resources and agricultural products.
International demand for petroleum or timber, for example, can lead to a far greater
level of exploitation of these resources in countries with poor environmental regula-
tion. The widespread deforestation in countries such as the Philippines would likely not
have happened absent international trade, and far less oil would have been spilled in the
Niger delta without an international market. In both cases, the culprit is incompetent
government, as is made clear by recent improvements in regulatory standards in these
two countries and a resultant decrease in environmental externalities (although stan-
dards in both countries could still do with considerably more improvement).
In both of these cases, the architecture of international trade governance is more
or less irrelevant. The goods in question are often sold on a global market and are dif-
ficult to trace, making it difficult for foreign governments to force up environmental
standards in producer countries, either unilaterally or through collective governance.
Effective reduction in the environmental externalities of trade in primary products
is more likely to happen through efforts to improve governance in source countries
than through changes in international trade rules. And as was the case in the broader
efficiency-versus-limits-to-growth argument, the real driving factor behind these sorts
of examples of environmental degradation is modernization and global economic
expansion generally, rather than the freeing of international trade more specifically.
A final category of international trade that has environmental effects is illegal trade.
This category encompasses illicit movements of goods that are not themselves contra-
band, as well as trade in contraband goods. Illicit movement of legal goods can result in
extraction processes that are less well regulated than those in the legal trade. Examples
include minerals such as coltan, luxury goods like diamonds, and food resources such
as fish (Nest 2011; Bieri 2010; DeSombre 2005). It can also include the export of waste. In
these cases, the illicit activity can result from efforts to avoid the costs of environmen-
tal regulation (as with fish caught outside of regulatory agreements) or the need to pay
royalties to governments (as with conflict diamonds). Trade in illicit goods is likely as
a general rule to generate greater negative environmental externalities than legal trade
because the goods are extracted or produced in a manner specifically designed to avoid
effective regulation. Even when it is not environmental regulation that is being inten-
tionally avoided, it seems unlikely that producers and traders of illicit goods will pay as
much attention to such regulation as producers and traders of licit goods.
From an environmental perspective, as well as from a trade governance perspective,
the most effective remedy to the negative externalities of illicit international trade is to
bring as much of it as possible into the formal international trade system. This can be
done by improving governance at the point of extraction or production and by using
markets to force as much of the trade as possible into the legal system. Improving gov-
ernance at the point of extraction generally means improving capacity in poor, weak,
 J. Samuel Barkin    445

or underperforming governments and improving cooperative international governance

for resources extracted in the international commons, such as is sometimes the case
with fisheries. Using markets to force trade within the legal system means having gov-
ernments of countries whose markets account for a significant proportion of the trade
of the good in question take measures to prevent the importation of illicit goods. In the
case of diamonds, many of the world’s countries have done this through the Kimberly
process (Wright 2004). In the case of fish, it is being done through various catch docu-
mentation schemes (DeSombre 2005). This use of market access to affect environmental
behavior abroad might seem to be an interference in international trade that would run
into difficulties with WTO rules. But as discussed in the next section, this is not the case.
Multilateral cooperation to enforce environmental standards is an accepted use of mar-
ket access policy.
Whereas illicit trade is illegal trade in goods that are themselves legal, illegal trade is
trade in goods that are themselves illegal (or illegal to trade internationally). Illegal trade
can have a negative impact on the natural environment through a number of routes. The
cultivation of illegal crops, such as cocaine in South America, is generally done outside
of regulatory control and in a manner that does not prioritize environmental sustain-
ability. Efforts by governments to control or eradicate those crops can also be environ-
mentally harmful, particularly if these efforts involve large-scale use of herbicides. The
exportation of certain toxic chemicals and waste products, trade in which is banned by
international law, leads directly to toxic pollution, generally in those countries least well
placed to clean it up. Other chemicals such as chlorofluorocarbons (CFCs), while not
themselves toxic, are banned for other environmental damage that they do. The illegal
trade in CFCs, for example, hurts efforts to restore the earth’s ozone layer (DeSombre
2000). And finally, the process of trading illegal goods can undermine governance and
therefore effective environmental governance, not only in source and destination coun-
tries, but also in the countries through which the goods transit. This problem is particu-
larly acute with the narcotics trade.

International Trade Institutions

and the Environment

Whereas many of the effects of trade on the environment occur, and can be remediated,
independently of international trade institutions, much of the politics of the relation-
ship between trade and the environment is focused specifically on those institutions.
On a broad historical scale it is patterns of economic development generally rather than
levels and forms of international trade specifically that drive most anthropogenic envi-
ronmental effects. But the specific rules of international trade can affect both the ability
of governments to make effective environmental policy and the forms that such pol-
icy takes. And opposition to the growth of international trade from an environmental
446   Trade and Environment

perspective often focuses on the institutions of, rather than general patterns of, interna-
tional trade (e.g., Eckersley 2004).
The WTO, as the only site for global authoritative rule-making on international
trade, is clearly a central institution for the discussion of trade and environment. Its
rules allow for exceptions to both general principles such as nondiscrimination and
specific national commitments in support of environmental policy. These exceptions,
which are discussed in more detail below, are limited and are adjudicated by the WTO’s
Dispute Settlement Mechanism (DSM), which has heard several environment-related
complaints. The WTO has a standing committee on trade and environment, and several
environment-related issues are on the Doha Round negotiating agenda, for example the
reduction of fisheries subsidies (Grynberg 2003).
Environmental concerns played a bigger role in the creation of NAFTA’s rules and
institutional structures than those of the WTO (Steinberg 1997). NAFTA itself incor-
porates the WTO’s environmental exceptions (NAFTA 1993, Article 2101), but from an
environmental perspective does not go much beyond them. However, the agreement
was signed in conjunction with the North American Agreement on Environmental
Cooperation, which has two key features in this context. The first is a legally binding
commitment by all three signatory governments to actually enforce their existing envi-
ronmental rules and to allow interested parties from within the NAFTA area to sue if
those rules are not being enforced (NAAEC 1993, Article 24). This commitment was
designed to discourage a regulatory race to the bottom in effective environmental
standards.
The second key feature is the creation of the Commission for Environmental Coop­
eration (CEC). The CEC’s remit is to encourage environmental cooperation among the
three NAFTA signatories by encouraging information-sharing and making recommen-
dations for environmental action and regulation both on transborder and continental
issues and on the enforcement of existing laws. The CEC is also charged with “consider-
ing on an ongoing basis the environmental effects of the NAFTA” (NAAEC 1993, Article
10[6]‌). In practice this has meant a set of reports over time compiling existing research
on the environmental effects of NAFTA, as well as commissioning original research on
the question.
This research has generally concluded that NAFTA has not had a systematically neg-
ative effect on the North American environment (e.g., CEC 2008). It has had locally
negative effects, however, particularly in places where the local enforcement of envi-
ronmental policy has not been able to keep up with NAFTA-driven economic change
(this is particularly true of the maquiladoras on the Mexican side of the border with
the United States). NAFTA may nevertheless have had a dampening effect on the cre-
ation of new environmental policy in some cases. This is particularly true of the effects
of its Chapter 11, which allows companies to sue governments for compensation for the
costs of unexpected policy changes. This particular rule has been the object of much
environmentalist ire (e.g., Mann and Von Moltke 1999). Although Chapter 11 rulings
cannot require changes to existing policy (as can WTO DSM rulings), the costs of settle-
ment can be high. Although few large settlements related to environmental policy have
 J. Samuel Barkin    447

in practice been ordered, the threat of Chapter 11 may nonetheless have an intimidating
effect on governments.
NAFTA is an important case not only in its own right, but also because it serves as
a template for other bilateral agreements that the United States negotiates. As a result,
these agreements often have stricter environmental controls than other trade agree-
ments and often have specific environmental requirements. For example, the bilateral
trade agreement between the United States and Peru has the strongest environmental
controls of any US trade agreement, which has in turn had a positive effect on the imple-
mentation of environmental regulation in Peru (e.g., Jinnah 2011).
The European Union (EU) is not usually thought of as primarily a trade institution,
but it does have a common market at its core. Unlike other free or managed trade areas,
however, the EU also has a common environmental policy. As such, the primary envi-
ronmental complaint concerning trade agreements, that they can lead to competition
among governments to promote trade by reducing environmental standards, does not
apply. Nor does the fear that international trade rules will interfere with the creation of
more stringent environmental rules at the national level, for the same reason. In a way, it
makes more sense to think of the EU as more akin to commerce across states within the
United States than to international trade (although the reality is of course more compli-
cated; see, e.g., Jordan and Lenschow 2000).
One can still ask what effects the increased trade that has resulted from the creation
of the common market and a common environmental policy within the EU has had
on environmental outcomes. The overall answer is that it has probably improved out-
comes, through two mechanisms. The first is efficiency gains from trade, which prob-
ably outweigh increased consumption in this case. The second is that the EU has had
a ratcheting-up effect on environmental policy in most of its member states. This is
certainly not true across the board—the Common Fisheries Policy, for example, has
from an environmental perspective been a complete disaster (Khalilian et  al 2010).
It is perhaps most true with respect to the Central and Eastern European countries
that have joined the EU since the beginning of the twenty-first century (Carmin and
VanDeveer 2005).
The relationship between international trade agreements generally and the environ-
ment has changed significantly over the years. While mechanisms for environmental
exceptions to trade rules were built into the original General Agreement on Tariffs and
Trade (GATT) in 1947 (as discussed below), the application of these exceptions did
not really become a political issue until the early 1990s. Mexico brought a complaint
against the United States to the GATT dispute resolution mechanism, arguing that US
rules designed to protect dolphins that prevented the importation of Mexican-caught
tuna broke GATT rules. The arbitration panel that heard the case ruled against the
United States, although Mexico decided not to pursue the ruling for fear of derailing
then-ongoing negotiations for NAFTA. The following year the European Economic
Community (EEC) launched a second case against the same US practice. The panel
ruled in the EEC’s favor, but the ruling was not adopted by the GATT. Both sets of rul-
ings were notable, among other things, for not recognizing protection of the global
448   Trade and Environment

environmental commons as a legitimate use of national trade policy (DeSombre and

Barkin 2002).
By the time the tuna-dolphin issue had played out, however, the institutions of inter-
national trade had begun to adopt a more formal environmental consciousness. The
agreement establishing the WTO has the goal of sustainable development written into
the first paragraph of its preamble and created a committee on trade and the environ-
ment. NAFTA created the CEC as part of its institutional structure. The tuna-dolphin
decisions represent a low-water mark for environmental consciousness in the interpre-
tation of international trade rules, and the situation has changed significantly in the past
two decades. Whether or not it has made any difference in practice, the language of sus-
tainable development has come to pervade the discourse on international trade. What
has made a difference in practice is a significant change in interpretation by interna-
tional trade tribunals, particularly the WTO’s DSM, of the legitimacy of environmental
protection as a role for trade policy.
The key basis for DSM decisions on environmental exceptions to the WTO’s trade
rules is Article XX of the original GATT, “General Exceptions.” The introduction to
this article (called the “chapeau” in international legal parlance) allows member states
to adopt and enforce several categories of measures, as long as “such measures are not
applied in a manner which would constitute a means of arbitrary or unjustifiable dis-
crimination between countries where the same conditions prevail, or a disguised restric-
tion on international trade.” The two relevant categories for environmental purposes are
in Article XX (b), measures “necessary to protect human, animal or plant life or health,”
and (g), measures “relating to the conservation of exhaustible natural resources if such
measures are made effective in conjunction with restrictions on domestic production or
consumption” (GATT 1947, Article XX).
Further guidance on environmental exceptions to trade rules can be found in the
Agreements on Sanitary and Phytosanitary Measures (SPS) and on Technical Barriers to
Trade (TBT), both of which are among the Uruguay Round agreements that are the legal
basis of the WTO. These agreements commit member states to ensuring that measures
related to human, animal, and plant health that interfere with trade be based on science,
and that where possible, environmental measures that interfere with trade be based on
international standards. Member states, in other words, can take measures to protect the
environment that interfere with trade and/or discriminate against other members of the
WTO, as long as they can make a plausible case that the environmental or health goal
is real, that the discrimination is necessary for that goal, and that it is compatible with
international standards, where these exist and are applicable.
Disagreements about how these exceptions should work in practice are adjudicated
by the DSM in much the same way as other disputes within the WTO, as discussed by
Busch and Pelc (this volume). Several cases related to the environment have been heard,
the best known of which was a complaint launched by four South and Southeast Asian
states concerning US rules requiring that shrimp imported into the United States be
caught only with nets that contained turtle-excluder devices (TEDs), which are basi-
cally holes in shrimp nets that allow marine turtles to escape. The original panel ruling,
 J. Samuel Barkin    449

in 1998, went against the United States and became a bit of a cause célèbre among envi-
ronmentalists. For example, several environmental protestors in the big protests against
the WTO in Seattle in 1999 wore turtle hats, expecting that television audiences would
understand the reference (DeSombre and Barkin 2002).
The case eventually went through five panel rulings, including two at the appellate
level, between 1998 and 2001. The first two rulings went against the United States on
the basis that its rules as applied to the complainants were arbitrary, that they did not
sufficiently take into account attempts by the complainants to protect turtles other than
through TEDs, and that the United States had made no good-faith effort to address
turtle protection multilaterally rather than unilaterally. A  second round of rulings,
responding to claims by the complainants that the United States had not adequately
addressed the first round, found that the United States had made some of the necessary
changes, but not enough of them. The third round, in 2001, found largely in favor of the
United States: that its revised rules interfered with the international trade in shrimp no
more than was necessary to protect turtles.
A clear set of precedents came out of this case, which has since been reinforced by
a variety of other rulings by the DSM on the trade-environment relationship. The
most notable of these is that the protection of the global environmental commons is
an accepted basis for national laws that interfere with international trade, even if the
environmental good in question is abroad. In other words, the DSM’s interpretation
of the WTO’s rules is that trade does not trump the environment. The environment
trumps trade, subject to a set of specific conditions, which were clearly articulated in the
shrimp-turtle findings and have been reinforced by other DSM panels since then. These
conditions are there to ensure that states are not using environmental protection as an
excuse to favor domestic industry over imports (DeSombre and Barkin 2002).
There are three such conditions. A Mexican complaint about US tuna import restric-
tions designed to protect dolphins has (perhaps ironically, given the history of the issue)
served to reinforce these conditions. Mexico launched the complaint in 2008, leading to
an appellate panel report in 2012 that referenced all three as accepted precedent (WTO
2012). The first condition is that the mechanisms of the policy be clearly related to, and
effective in achieving, the environmental goal. An innovation of the first appellate panel
in the shrimp-turtle case was that it heard testimony from turtle conservation experts
and other interested parties about whether the US rules were actually an effective means
of saving turtles in South and Southeast Asia (Appleton 1999).
The second condition is that the policy must be applied in a fair and nondiscrimina-
tory manner. There are a number of ways in which environmental rules can be unfair or
discriminatory. In the shrimp-turtle example, US rules required that the complainant
states be certified by the State Department as having TEDs installed throughout their
fleets, but the State Department did not have the personnel necessary to do the certify-
ing. The panel found this to be unfair, because it left the complainants unable to comply
with the rules through no fault of their own. In another case, brought by Brazil against
the United States and relating to pollution levels from refineries, the rules in question
allowed many US refineries to be dirtier than what was required for Brazilian refineries
450   Trade and Environment

(WTO 1996). There was no environmental logic for this distinction—it was simple pro-
tectionism. In both cases the DSM accepted the goal of the environmental regulations
but required that they be applied in a fairer way that would nonetheless leave their envi-
ronmental effectiveness intact.
The third condition is that states make a good-faith effort to address environmental
issues in the global commons multilaterally prior to, or at minimum concurrent with,
taking unilateral actions that affect trade. Thus, the United States was taken to task in
the first set of shrimp-turtle rulings for not attempting to negotiate turtle conserva-
tion efforts with the complainant states, even though US law mandated efforts to do so
(DeSombre and Barkin 2002). The corollary of this condition is that DSM precedent is
to accept multilateral environmental cooperation on an issue as prima facie evidence
that environmental action on that issue is legitimate. In the shrimp-turtle case, the fact
that some of the species of turtle that the United States was trying to protect were listed
as endangered by the Convention on International Trade in Endangered Species of
Wild Fauna and Flora (CITES) was taken by the panel to obviate further discussion of
the issue.
This discussion of multilateralism leads to one final concern of many environmental-
ists, about the relationship between the WTO and MEAs (e.g., Palmer and Tarasofsky
2007). Some MEAs explicitly use trade mechanisms to enforce environmental com-
mitments by member states or to coerce nonmembers either to join or to accept MEA
rules anyway. CITES, for example, explicitly bans international trade in those species it
lists as endangered and limits trade in those species listed as threatened. Other MEAs,
such as the Rotterdam and Basel Conventions, restrict international trade in hazard-
ous chemicals and waste. A complete ban on international trade in certain categories
of hazardous waste has been negotiated under the auspices of the Basel Convention,
although it has not yet come into force (on the hazardous chemicals regime complex,
see Selin 2010).
Another type of MEA uses restrictions on trade in particular goods to encour-
age nonmembers to participate in production limits on those goods. Examples range
from ozone-depleting substances to fish. The Montreal Protocol on Substances that
Deplete the Ozone Layer allows member states to prohibit the importation from non-
members of a wide range of goods that use CFCs and other ozone-depleting chemicals
(DeSombre 2000). A number of regional fisheries management organizations, such as
the Convention for the Conservation of Antarctic Marine Living Resources (CCAMLR)
and the International Convention for the Conservation of Atlantic Tunas (ICCAT),
use catch documentation schemes to identify fish caught legally and within quotas and
require members to prevent the importation of undocumented fish (DeSombre 2005).
Since these mechanisms in effect call for embargoes on imports of certain kinds of goods
from states that are not members of the relevant MEA, even if they are members of the
WTO, these rules are clearly discriminatory.
But at the same time, all of these MEA trade mechanisms fit well into the three condi-
tions for environmental restrictions to trade rules. They are clearly related to the envi-
ronmental issues at hand. They are not needlessly discriminatory—nonmember states
 J. Samuel Barkin    451

are welcome to join the relevant MEAs, and once they have joined, they can reasonably
be expected to follow the rules. And they are (by definition) multilateral. DSM panels
have tended, without recent exception, to defer to MEAs on environmental issues and
to accept MEAs as evidence in and of themselves that the environmental issues that they
deal with are appropriate objects of environmental policy, even if such policy interferes
with trade. There has been no indication in the history of DSM jurisprudence that the
dispute resolution body would rule against an MEA, and precedent suggests that it is
highly unlikely that it will do so in the future. This particular fear about the WTO and
the environment is misplaced.

The Politics of Trade and Environment

In fact, much of the criticism of the institutions of international trade from an environ-
mental perspective is misplaced. There are certainly some specifics of particular treaties
that may in particular cases have a dampening effect on environmental policy making,
such as NAFTA’s Chapter 11. But there is little evidence that this effect is widespread.
Meanwhile, changes in production patterns caused by trade agreements do affect pollu-
tion patterns. Where the changes move production from developed to developing coun-
tries, the net effect, at least in the short term, can be an increase in levels of pollution. But
trade agreements also often have the effect of ratcheting up environmental standards in
less well-regulated countries toward those in better-regulated countries. Furthermore,
the patterns of international trade that tend to cause the most environmental damage
are those that are least well-regulated by international trade institutions. This includes a
range of products from primary goods to illegal goods. Whether increased environmen-
tal harm results from inadequate governance at the national or the international level,
the answer to environmental harm from trade is likely to be better governance rather
than less cooperation.
Better governance can better manage the environmental effects of economic growth.
If one’s complaint about trade institutions is that they promote economic growth and
therefore necessarily promote increased pollution and resource use, then a focus on
trade specifically, and even more so on the institutions of international trade, misses
the core of the problem. International trade should generate more efficient use of
resources for any given level of economic activity and should not necessarily increase
pollution. The core problem in this view is with an economic system that is based on
an ethic of constant expansion rather than with particular patterns of trade within this
system, and a focus on trade will do little to change the system (Princen, Maniates, and
Conca 2002).
Improving environmental governance at the national level is beyond the scope of
this chapter, beyond noting that efforts to do so are included in some international
trade agreements, and that the use of trade mechanisms to force improvements in
environmental governance in other countries is allowed within the rules of the WTO,
452   Trade and Environment

particularly when done through MEAs. Improving environmental governance at the

international level can only be done multilaterally. Preventing changes in existing trade
institutions, or preventing the creation of new ones, leaves us with the existing status
quo. It does not allow governments to address weaknesses in the current rules and
therefore can hinder environmental cooperation rather than help it.
Preventing change in the existing institutional structure is, however, easily done.
There are two key stumbling blocks to improving international cooperation at the
trade-environment nexus:  political structure and a waning of interest by states in
international environmental cooperation. As a result of these two obstacles, there is
little movement on trade and environment issues at the multilateral level. The major
exception to this generalization is cases in which major importing nations can use their
market power to coerce participation in new environmental rules. Examples of such
coercion include bilateral trade deals between the United States and smaller countries
and the use of trade mechanisms in MEAs (which tend to be effective only when sup-
ported by the largest market economies, particularly the United States).
Beyond these examples of market power supporting the creation of new forms of
regulation at the trade-environment nexus, progress is often stifled by the structure of
international law in general and of existing institutions in particular. International law
in general binds only those countries that explicitly agree to new regulation, meaning
that it is easy for a minority of countries to undermine negotiations for new agreements.
This is a problem particularly with pollution issues, for which the explicit cooperation
of major polluters is needed for cooperation to be effective. Absent effective coercion
by importers, there is often little incentive for polluters to cooperate. This pattern of
behavior can be clearly seen, for example, in climate change negotiations. The United
States will not agree to anything that does not put it on an equal footing with China, and
China will not agree to anything that puts it on the same footing as the United States.
The result has been almost two decades of deadlock.
Existing institutions, particularly the WTO, often have rules requiring unanimous
agreement for effective change in rules. In the case of the WTO, this requirement has
left the Doha Round of negotiations stalled, with no successful end in sight. Since WTO
practice is for comprehensive rather than piecemeal changes in the rules, change in
environmental rules, beyond changes driven by the adjudication process of the DSM,
are held hostage to negotiations about issues that have nothing to do with the envi-
ronment. Other institutions deal with unanimity requirements by adopting volun-
tary guidelines rather than rules, or in other words soft law rather than hard law. For
example, negotiations toward reducing subsidies to fishers are being negotiated, or have
recently been negotiated, in a number of international forums. But the talks at the WTO
are held hostage to the broader failure of the Doha Round. An agreement at the UN’s
Food and Agriculture Organization is voluntary, and separate talks under the auspices
of the Organisation for Economic Co-operation and Development and the Convention
on Biological Diversity failed outright (Barkin and DeSombre 2013).
This failure connects with the second key stumbling block to multilateral coop-
eration at the trade-environment nexus, a waning of interest by states in prioritizing
 J. Samuel Barkin    453

multilateral environmental cooperation (Thomas 2004). In part, this change reflects

a decline of environmental issues in the international priorities of the global North.
While different parts of the North remain committed to international action on differ-
ent environmental issues, the range of such issues on which the developed world as a
whole agrees and is willing to pursue aggressively through trade measures has shrunk
since the high point of international environmental cooperation in the last quarter of
the twentieth century.
In part, a subset of a broader North-South tension in international politics (Neumayer
2004) has hampered progress at the WTO’s Committee on Trade and Environment
since its inception. The simple version of this tension is that the global North argues
that the environment is worth protecting for its own sake, and the global South worries
that environmental regulation, whatever the intensions of the North, will in practice
become a disguised form of protectionism. Most parties accept the principle of “com-
mon but differentiated responsibilities,” the idea that all countries much participate in
environmental governance, but the developed world should pay a greater share of the
costs of doing so (e.g., Stone 2004). But the gulf in interpretation can be unbridgeable,
with countries of the North arguing that effective environmental governance requires
active participation by the global South (or at least its larger and richer members), and
countries of the South arguing that development must be prioritized over managing the
global commons.
The good news is that existing international trade institutions are likely to continue
to defer to MEAs on environmental issues and to defer to national environmental pol-
icy making when rules are not unnecessarily discriminatory or unilateral. A potential
source of worry is new trade agreements. Bilateral agreements between the United
States and the EU on the one hand and smaller markets on the other will likely follow
existing models, but this is less clear with respect to regional multilateral negotiations.
More broadly, the state of the global environment will be affected much more by pat-
terns of multilateral environmental cooperation and of international economic devel-
opment than by details of trade agreements, whether existing or new.

Conclusion

The relationship between trade and environment is complex and multifaceted. Trade
can be beneficial for the natural environment to the extent that it contributes to eco-
nomic efficiency and generates preferences for stricter environmental policy. However,
increased global consumption and production may stress finite resources and the abil-
ity of natural systems to absorb pollutants. Meanwhile, the institutions that govern
international trade and those that address global environmental protection are not,
as many have assumed, inherently at odds. To the extent that they have seemed to be
so, it has often been because trade policies used environmental protection as a cover
for intentionally protectionist measures. International trade institutions allow a wide
454   Trade and Environment

scope for environmental policy, particularly if it is multilateral and not unnecessarily

discriminatory.

References
Appleton, Arthur E. 1999. Shrimp/Turtle:  Untangling the Nets. Journal of International
Economic Law 2: 477–496.
Barkin, J. Samuel. 2003. The Counterintuitive Relationship between Globalization and Climate
Change. Global Environmental Politics 3: 8–13.
Barkin, J. Samuel, and Elizabeth R. DeSombre. 2013. Saving Global Fisheries: Reducing Fishing
Capacity to Promote Sustainability. Cambridge, MA: MIT Press.
Bhagwati, Jagdish. 2007. In Defense of Globalization. With a New Afterword. New York: Oxford
University Press.
Bhagwati, Jagdish. 2000. On Thinking Clearly about the Linkage between Trade and the
Environment. Environment and Development Economics 5: 483–529.
Bieri, Franziska. 2010. From Blood Diamonds to the Kimberley Process: How NGOs Cleaned up
the Global Diamond Industry. London: Ashgate Publishing.
Carmin, JoAnn, and Stacy D. VanDeveer. 2005. EU Enlargement and the Environment:
Institutional Change and Environmental Policy in Central and Eastern Europe. New York:
Routledge.
Clapp, Jennifer. 2001. Toxic Exports:  The Ttransfer of Hazardous Wastes from Rich to Poor
Countries. Ithaca, NY: Cornell University Press.
Clapp, Jennifer. 2002. What the Pollution Havens Debate Overlooks. Global Environmental
Politics 2: 11–19.
Commission for Environmental Cooperation (CEC). 2008. Environmental Assessment of
NAFTA: Lessons Learned from CEC’s Trade and Environment Symposia. Montreal: Secretariat
of the Commission for Environmental Cooperation.
DeSombre, Elizabeth R. 2000. The Experience of the Montreal Protocol:  Particularly
Remarkable, and Remarkably Particular. UCLA Journal of Environmental Law & Policy
19: 49–81.
DeSombre, Elizabeth R. 2005. Fishing under Flags of Convenience: Using Market Power
to Increase Participation in International Regulation. Global Environmental Politics
5: 73–94.
DeSombre, Elizabeth R., and J. Samuel Barkin. 2002. Turtles and Trade: The WTO’s Acceptance
of Environmental Trade Restrictions. Global Environmental Politics 2: 12–18.
Dinda, Soumyananda. 2004. Environmental Kuznets Curve Hypothesis: A Survey. Ecological
Economics 49: 431–455.
Drezner, Daniel W. 2001. Globalization and Policy Convergence. International Studies Review
3: 53–78.
Eckersley, Robyn. 2004. The Big Chill: The WTO and Multilateral Environmental Agreements.
Global Environmental Politics 4: 24–50.
Eckersley, Robyn. 1992. Environmentalism and Political Theory: Toward an Ecocentric Approach.
London: UCL Press.
Frankel, Jeffrey A., and Andrew K. Rose. 2005. Is Trade Good or Bad for the Environment?
Sorting Out the Causality. Review of Economics and Statistics 87: 85–91.
 J. Samuel Barkin    455

Gallagher, Kevin P. 2005. International Trade and Air Pollution:  Estimating the Economic
Costs of Air Emissions from Waterborne Commerce Vessels in the United States. Journal of
Environmental Management 77: 99–103.
Garcia-Johnson, Ronnie. 2000. Exporting Environmentalism:  US Multinational Chemical
Corporations in Mexico and Brazil. Cambridge, MA: MIT Press.
General Agreement on Tariffs and Trade (GATT). 1947.
Grynberg, Roman. 2003. WTO Fisheries Subsidies Negotiations: Implications for Fisheries
Access Arrangements and Sustainable Management. Marine Policy 27: 499–511.
Jacquet, Jennifer, John Hocevar, Sherman Lai, Patricia Majluf, Nathan Pelletier, Tony Pitcher,
Enric Sala, Rashid Sumaila, and Daniel Pauly. 2010. Conserving Wild Fish in a Sea of
Market-based Efforts. Oryx 44: 45–56.
Jinnah, Sikina. 2011. Strategic Linkages:  The Evolving Role of Trade Agreements in Global
Environmental Governance. Journal of Environment & Development 20: 191–215.
Jordan, Andrew, and Andrea Lenschow. 2000. “Greening”the European Union:  What Can
Be Learned from the “Leaders” of EU Environmental Policy? European Environment
10: 109–120.
Khalilian, Setareh, Rainer Froese, Alexander Proelss, and Till Requate. 2010. Designed for
Failure: A Critique of the Common Fisheries Policy of the European Union. Marine Policy
34: 1178–1182.
MacGregor, James, and Bill Vorley. 2007. “Fair Miles”: The Concept of “Food Miles” Through
a Sustainable Development Lens. London:  International Institute for Environment and
Development.
Mann, Howard, and Konrad Von Moltke. 1999. NAFTA’s Chapter  11 and the Environment:
Addressing the Impacts of the Investor-State Process on the Environment. Winnipeg, MB:
International Institute for Sustainable Development.
Meadows, Donella H., Donella H. Meadows, Jorgen Randers, and William W. Behrens. 1972.
The Limit s to Growth: A Report to The Club of Rome. New York: New American Library.
North American Agreement on Environmental Cooperation (NAAEC). 1993.
North American Free Trade Agreement (NAFTA). 1993.
Nest, Michael. 2011. Coltan. Cambridge, UK: Polity.
Neumayer, Eric. 2004. The WTO and the Environment: Its Past Record Is Better Than Critics
Believe, but the Future Outlook Is bleak. Global Environmental Politics 4: 1–8.
Paarlberg, Robert. 2009. Starved for Science: How Biotechnology Is Being Kept Out of Africa.
Cambridge, MA: Harvard University Press.
Palmer, Alice, and Richard Tarasofsky. 2007. The Doha Round and Beyond: Towards a Lasting
Relationship between the WTO and the International Environmental Regime. London:
Chatham House.
Princen, Thomas, Michael Maniates, and Ken Conca. 2002. Confronting Consumption.
Cambridge, MA: MIT Press.
Rodrik, Dani. 1997. Has Globalization Gone Too Far? Washington, DC: Peterson Institute.
Selin, Henrik. 2010. Global Governance of Hazardous Chemicals. Cambridge, MA: MIT Press.
Steinberg, Richard H. 1997. Trade-Environment Negotiations in the EU, NAFTA, and
WTO: Regional Trajectories of Rule Development. American Journal of International Law
91: 231–267.
Stone, Christopher D. 2004. Common But Differentiated Responsibilities in International Law.
American Journal of International Law 98: 276–301.
456   Trade and Environment

Thomas, Urs P. 2004. Trade and the Environment: Stuck in a Political Impasse at the WTO after
the Doha and Cancún Ministerial Conferences. Global Environmental Politics 4: 9–21.
Weber, Christopher L., and H. Scott Matthews. 2008. Food-miles and the Relative Climate
Impacts of Food Choices in the United States. Environmental Science & Technology
42: 3508–3513.
World Trade Organization (WTO). 2012. United States—Measures Concerning the Importation,
Marketing and Sale of Tuna and Tuna Products. Report of the Appelate Body, WT/DS381/
AB/R. Geneva: World Trade Organization.
World Trade Organization (WTO). 1996. United States—Standards for Reformulated and
Conventional Gasoline. Report of the Appellate Body, WT/DS2/AB/R. Geneva: World Trade
Organization.
 Chapter 24

B rid ging th e  Si l o s

Trade and exchange rates
in international political economy

Ma rk S . C ope l ov itc h

a nd Jon C. W. Pev e h ouse

Introduction

The inextricable link between trade and exchange rates is a central feature of the inter-
national economy. Any traveler who crosses a border considers the effect of exchange
rate levels on a decision to purchase goods. As the cost of traversing international bor-
ders has decreased, as production of goods has become globalized, and as exchanging
currencies has become easier, the effect of exchange rates on the price of goods and ser-
vices has become more visible to the average citizen. Trade and financial liberalization
are both key components of globalization, and they are intimately linked through the
conduit of exchange rates.
The vital connection between trade and exchange rate policies is also illustrated by a
variety of well-known historical and contemporary examples. The question of whether
or not to adhere to the gold standard mobilized tradable goods producers and domi-
nated political debates about economic policy in the United States and elsewhere dur-
ing both the pre-World War I and interwar periods (Eichengreen 1992; Frieden 1993;
Simmons 1994). Similarly, the large current US account deficits in the late 1960s and
early 1970s, coupled with concerns of American exporters about the loss of competitive-
ness vis-à-vis Europe and Japan, heavily influenced the Nixon administration’s decision
to close the “gold window” and end the Bretton Woods era (Odell 1982; Gowa 1983). In
the mid-1980s the trade-related implications of the dollar’s 50 percent appreciation rela-
tive to the West German deutsche mark and Japanese yen were a major factor leading
to the Plaza and Louvre Accords, in which G-7 central banks engaged in coordinated
foreign exchange intervention to stabilize their exchange rates (Destler and Henning
458   Bridging the Silos

1989; Frankel 1994). Over the last decade the extensive academic and policy debate
about Chinese currency undervaluation has further illustrated the relationship among
exchange rate regimes, levels, and trade flows (Bergsten 2006, 2010).
Despite this widespread acknowledgment of the links between trade and exchange
rates, international political economy (IPE) scholars have largely treated the two policy
areas as analytically distinct realms of inquiry. A large literature discusses the political
determinants of states’ trade policy and trade flows, focusing on a host of international
and domestic political factors.1 Likewise, a robust literature examines the determinants
and effects of the choice of exchange rate policies.2 Yet in the vast majority of the writ-
ings in both areas, there is little discussion of exchange rates as determinants of trade
policy and of trade policy as a predictor of exchange rate and monetary policy choices.
We defer our discussion of why this “silo-ing” of IPE has developed to our final sec-
tion. For now, we simply note that while the political economy literature on the nexus
between trade and exchange rates is certainly not an empty set, it is substantially smaller
than one might expect given the widespread economic understanding of the intimate
linkages between trade and exchange rates.
This chapter highlights some of the existing work that has begun to explore these link-
ages between trade and exchange rate policies, takes stock of the state of our knowl-
edge about the overlap between these policies, and suggests fruitful avenues for future
research. We begin our review of existing work on the relationship between trade and
exchange rates by examining how trade policy and trade flows may influence exchange
rates and monetary policy. We then discuss how monetary and exchange rate policy
choices may influence trade policy, briefly reviewing an extensive number of stud-
ies that have examined this question. Finally, we consider how research might better
address these policies together and bridge the “silos” of trade, money, and finance to
better understand the complex and interactive relationship between these policies in the
contemporary global economy.

Trade as a Determinant
of Exchange Rates

Economists have long been aware of the trade implications of exchange rate regimes,
exchange rate levels, and monetary policy choices. Indeed, while more recent work
has focused on the adoption of fixed exchange rates as a nominal anchor for monetary
policy (Svensson 1999; Calvo and Vegh 1994), the canonical literature in economics on
exchange rates emphasizes the reduction of exchange rate risk—and the corresponding
increase in cross-border trade and financial flows—as one of the key reasons that coun-
tries choose fixed exchange rates over more flexible regimes (Mundell 1961; McKinnon
1962; Kenen 1969). In developing the theory of optimum currency areas (OCA),
Mundell emphasized trade ties as the first of four key criteria determining whether it is
 Mark S. Copelovitch and Jon C. W. Pevehouse    459

optimal for countries to share a common currency. McKinnon encapsulates this logic
nicely: “[T]‌here is less need for independent monetary policies within the bloc and a
strong case to be made for imposing the uniform discipline that a common currency
system would provide” (1973, 86). Independent national policies are neither necessary
nor desirable if exchange rate changes can upset carefully negotiated tariff, tax, and pric-
ing policies.
Building on this body of work in economics, the IPE literature on interest groups and
exchange rates also emphasizes the importance of trade as a reason that some actors
favor fixed exchange rates over floating regimes. Pegging the exchange rate is generally
favored by tradables producers and international investors, since it reduces or elimi-
nates exchange rate risk and facilitates cross-border trade and exchange (Frieden 1991).
In contrast, currency volatility creates uncertainty about cross-border transactions,
adding a risk premium to the price of traded goods and international assets (Frieden
2008). At the same time, the level of the exchange rate also has important implications
for trade, as it affects the relative price of traded goods in both domestic and foreign
markets. Fluctuations in exchange rates can have substantial effects on domestic pro-
ducers’ competitiveness in world markets: “In the case of a real appreciation, domestic
goods become more expensive relative to foreign goods; exports fall and imports rise
as a result of the change in competitiveness. Real depreciation has the opposite effects,
improving competitiveness” (Frieden and Broz 2001, 331). Consequently, exchange
rate movements have significant domestic distributional consequences. All else being
equal, exporters and import-competing industries lose from currency appreciation,
while the nontradables sector and domestic consumers gain (Frieden 1991). Currency
depreciations have the opposite effect, helping exporters and import-competing firms
at the expense of consumers and the nontradables sector (Frieden and Broz 2001). The
degree to which firms and interest groups are sensitive to exchange rate movements
and levels, however, depends heavily on the degree of exchange rate “pass-through”: the
degree to which prices of goods respond rapidly to exchange rate fluctuations (Campa
and Goldberg 2005; Goldberg and Knetter 1997; Devereaux and Engel 2002). In gen-
eral, standardized products and commodities exhibit more pass-through than highly
differentiated, more specialized products (Frieden and Broz 2006). At the same time,
exporters’ preferences regarding the level of the exchange rate also depend on the degree
to which they import intermediate inputs of production; when firms are more heavily
dependent on such inputs, the benefits to competitiveness of a more depreciated cur-
rency may be more than offset by the higher cost of factors of production purchased
from overseas providers (Campa and Goldberg 1999). Although firm-level tests of the
political economy hypotheses are limited, the evidence to date suggests that produc-
ers’ preferences about currency stability and depreciation fall broadly in line with these
expectations (Broz, Frieden, and Weymouth 2008).
More recent work (Walter 2008, 2013) shows that firms’ and households’ balance
sheet exposure to foreign currencies also creates domestic constituencies for or against
devaluation or depreciation. Thus, the trade side of the equation, though important,
is only “one half of the story” (Walter 2013). When firms and consumers have assets
460   Bridging the Silos

and liabilities denominated in foreign currency—such as interbank loans or home

mortgages—these financial linkages also influence preferences regarding exchange rate
regime choice and macroeconomic policies. These societal preferences, Walter argues,
strongly influence policy makers’ decisions to delay macroeconomic adjustments and to
defend (or abandon) currency pegs. Drawing on both qualitative and quantitative anal-
yses of the Asian financial crisis and the global economic crisis, Walter finds substantial
evidence that these firm- and consumer-level balance sheet considerations also influ-
ence governments’ exchange rate policy choices. Further research into how domestic
actors and governments weigh the relative importance of the trade and financial aspects
of their interests is a promising route forward for scholars.

Beyond Flows: Trade Policy, Trade

Institutions, and Exchange Rates
Thus, the relationship between trade flows and exchange rate regimes and levels is widely
understood. A less well-understood and more interesting political question, however,
involves the effect of trade policy and international trade institutions on monetary and
exchange rate policy choices. Here the literature is far less developed. A good starting
point is the literature on economic regionalism. Within this body of work, studies focus-
ing on questions involving sequencing in regional integration directly address the issue
of linkages between trade and exchange rate policy. The foundational work of Balassa
(1961) identified five stages of regional integration: 1) a free trade area (among members
only); 2) a customs union (establishing a common external tariff); 3) a common market
(including free movement of labor and capital); 4) an economic union (which includes
harmonization of national economic policies); and finally 5) a political union, including
the unification of social, fiscal, and monetary policies. Balassa argued that significant
supranational political institutions would be needed to facilitate these final stages of
integration. To coordinate the adoption and enforcement of trade, social, and monetary
integration, states would have to agree to an extensive set of common rules and regula-
tions. These in turn would require the establishment of legal and political institutions at
the supranational level. In Balassa’s early formulation, the form, timing, and impetus for
cooperation were largely assumed or left implicit. However, work by neofunctionalist
scholars of European integration identified several causal mechanisms driving the move
from one stage of regional integration to another (Haas 1958; Schmitter 1969): spillover
(frustration in one issue area leading to efforts in cognate areas); externalization (the
need to deal with third parties and the externalities arising from cooperation); and
politicization (the mobilization of domestic actors in response to more controversial
integration efforts). In the neofunctionalist view, any or all of these factors could lead
to an expansion (or contraction) of cooperative efforts from one area (trade) to another
(monetary policy).
In contrast to the OCA literature in economics, the logic underlying this approach is
not simply economic efficiency. Rather, the political and economic interests of private
 Mark S. Copelovitch and Jon C. W. Pevehouse    461

economic agents and supranational bureaucrats play a critical role in determining when
and why trade cooperation and regional trade agreements evolve into broader mone-
tary cooperation (Haas 1958; Sandholtz and Stone Sweet 1998). Clearly the European
Union (EU) is the most illustrative case of these stages of regional integration and the
linkage between trade institutions and monetary integration. Indeed, of the eight cases
of regional trade integration leading to extensive monetary cooperation identified by
Cooper (2007, 631), three are related to the European Monetary Union (EMU) and its
two predecessors: the Snake agreement under the EEC and the EMS agreement under
the EC. Given Europe’s unique success (at least until 2008) in the realm of monetary
integration, this is not surprising. Nonetheless, there are and have been other cases
of regional monetary integration (e.g., the CFA Franc zone in West Africa, the East
Caribbean dollar zone, the German Zollverein, the Latin and Scandinavian monetary
unions).
Moreover, given the vast growth in bilateral and regional agreements in the last two
decades (Mansfield and Milner 1999), the broader question of the relationship between
trade policies/institutions and exchange rates applies to many more countries than
those within the EU. Indeed, Mansfield and Pevehouse (2013) show that every region
has experienced an expansion in regional trade agreements of various types since World
War II, with nearly every country now involved in at least one such agreement. Moving
from trade cooperation to institutionalized monetary cooperation, however, has been
relatively rare. Cooper (2007) hypothesizes several reasons for the difficulty in moving
from trade to money; most involve the higher costs to governments of losing monetary
policy autonomy as well as the comparative difficulty in reversing monetary versus trade
cooperation. Moreover, as McNamara (1999) has argued, the roots of monetary policy
lie not only in interest-based factors, but also in ideas. Without significant ideational
convergence among policy makers, states are unlikely to give up the ability to print their
own money—an important sign of national sovereignty.
While the literatures on regionalism and European integration offer important
insights into the relationship between trade and exchange rate policies, another impor-
tant starting point is the literature on “exchange rate protection” (Corden 1982), the
manipulation of the exchange rate as a substitute for trade policies that are prohibited
under a country’s international trade commitments in the General Agreement on Tariffs
and Trade (GATT)/World Trade Organization (WTO) or preferential trade agreements
(PTAs). The logic here starts from the observation that a real currency depreciation
is tantamount to both a tax on imports and an export subsidy for domestic producers
(McKinnon and Fung 1993). In other words, a currency depreciation alters the terms of
trade in precisely the same way as protectionist trade policies. Coupled with the large
body of research that has demonstrated the correlation between flexible exchange rates
and both nominal and real depreciation (International Monetary Fund 1997; Frieden
2002; Blomberg, Frieden, and Stein 2005), this logic suggests that countries that have
tied their hands more on trade policy through international trade agreements will
be more likely to engage in competitive devaluations or depreciations to enhance the
competitiveness of domestic exporters. Recent empirical work strongly supports this
462   Bridging the Silos

conjecture. Copelovitch and Pevehouse (2013a) argue that PTAs influence both a coun-
try’s choice of exchange rate regime and its propensity to allow its currency to depre-
ciate. They find that countries that have signed a PTA with their “base” country—the
country most likely to be their anchor currency based on current and past experience
with fixed exchange rates, regional proximity, and trade ties (Klein and Shambaugh
2006)—are less likely to adopt a fixed exchange rate and tend to have more depreciated/
devalued currencies (Copelovitch and Pevehouse 2013a).
These results suggest that the linkages between exchange rate and trade policies pres-
ent policy makers with a trade-off. On the one hand, fixed exchange rates and trade lib-
eralization are complementary, as articulated in the logic of regionalism and Balassa’s
stages of integration. On the other, flexible exchange rates and competitive devaluation
may allow a government to circumvent its commitments to liberalize trade by pursu-
ing monetary policies that are close substitutes and cater to the interests of domes-
tic actors (e.g., import-competing sectors) hurt by the policies imposed under PTAs.
These “conflicting interests”—to stabilize exchange rates to stimulate further trade or
to retain monetary and exchange rate policy autonomy to influence the competitive-
ness of domestic export interests—are ever more salient, given the proliferation of bilat-
eral, regional, and multilateral trade institutions in the global economy (Pomfret and
Pontines 2013). This could be yet another explanation for the finding that trade agree-
ments rarely move into monetary cooperation: if trade and monetary policy are sub-
stitutes, policy makers will want to keep some policy autonomy to respond to domestic
pressures for adjustment during difficult economic times.
And while IPE scholars have only begun to examine the relationship between trade
and monetary institutions, further work will no doubt assess how specific elements of
these trade institutions affect countries’ exchange rate and monetary policy choices. For
example, how might the presence or absence of escape clauses (Rosendorf and Milner
2001) influence a government’s desire to use monetary policy to respond to pressures,
given the ability to temporarily get relief within the trade realm? Will the presence of
dispute settlement mechanisms (Busch 2007) in trade institutions deter leaders from
undertaking exchange rate protectionism, given an institution to adjudicate accusations
of such behavior from other members? (See also Pelc and Busch, this volume.)

Exchange Rates and Trade Policy

Thus far we have focused on the influence of trade and trade policy on exchange rates.
However, much of the existing empirical literature reverses the causal arrow and
explores the effects of exchange rates and monetary policy choices on trade. Most of this
work has been done in economics, where scholars have extensively studied the effects
of exchange rate regime choice on the level and volatility of trade flows (IMF 1984,
2004; De Grauwe 1988; Baccheta and van Wincoop 2000; Rose 2000; Lopez-Cordova
and Meissner 2003; Klein 2005; Bahmani-Oskoosee and Hegerty 2007; Auboin and
 Mark S. Copelovitch and Jon C. W. Pevehouse    463

Ruta 2011; Klein and Shambaugh 2006, 2010; Huchet-Bourdon and Korinek 2011). The
theoretical predictions and empirical results of this literature are largely indeterminate
and highly sensitive to assumptions about price elasticities and model specification, as
well as to the specific time period or country sample selected (Auboin and Ruta 2011;
Pomfret and Pontines 2013).
A subset of this literature specifically explores the effects of currency unions on trade.
Frankel and Rose (2002) find that countries that share a currency trade three times
more than similarly related countries that do not; Engel and Rogers (1996) found that
US states (specifically North and South Dakota) trade twenty-two times more with each
other than with neighboring Canadian provinces. These results, and the logic under-
lying them—permanently eliminating currency risk through the adoption of a single
currency—provided a clear economic rationale and aspiration for European monetary
union and are the driving force behind other regional experiments with monetary
integration.
In spite of this extensive literature in economics on the effects of exchange rates on
trade flows, work on the specific link between exchange rate and trade policies is far
less extensive. One productive line of research in this vein focuses on the link between
exchange rate regime choice and trade policy. In an influential article and book,
Eichengreen and Irwin (2010) and Irwin (2012) show that the degree to which coun-
tries engaged in protectionist trade policies during the Great Depression was driven
primarily by their exchange rate policies. Those countries that remained on the gold
standard—thereby foregoing the opportunity to employ monetary and fiscal policy to
stimulate domestic employment and growth—were more likely to impose protectionist
trade policies and capital controls as “a second-best macroeconomic tool” (Eichengreen
and Irwin 2010, 872). Thus, with their hands tied on monetary policy by the commit-
ment to a fixed exchange rate regime, Depression-era governments resorted to trade
protection as an imperfect substitute for macroeconomic adjustment policies. This his-
torical finding may also have contemporary relevance. In preliminary work, Copelovitch
and Pevehouse (2013b) find a similar phenomenon within the WTO. Drawing on
Eichengreen and Irwin’s logic and linking it to the Mundell-Fleming “trilemma,” we find
that governments are more likely to impose legalized protectionist measures (e.g., safe-
guards, antidumping duties) within the WTO when they have adopted fixed exchange
rates and liberalized their capital accounts. As a consequence, such countries are also
more likely to find themselves named by other member states as defendants in WTO
disputes.
A second line of research focuses on the impact of exchange rate levels and fluctua-
tions on trade policy choices. This work indicates that protectionism has been great-
est at the regional level during periods of sharp exchange rate fluctuations, such as the
1992–1993 European Monetary System (EMS) crisis and the 1999 Brazilian real devalu-
ation within Mercosur (Fernandez-Arias et. al. 2004; Eichengreen 1993; Pearce and
Sutton 1985).3 Similarly, Frieden (1997) shows that protectionist demands in the United
States during the nineteenth century correlated closely with the strength of the dollar.
In addition, a number of studies suggest that real exchange rate appreciations have led
464   Bridging the Silos

to increases in antidumping filings in the United States and other industrialized coun-
tries (Oatley 2010; Niels and Francois 2006; Irwin 2005; Knetter and Prusa 2003; Grilli
1988; Bergsten and Williamson 1983). Broz and Werfel (2013) extend this work to the
firm level, arguing that several industry-specific characteristics determine the protec-
tionist response to changes in the exchange rate, including the degree of exchange-rate
pass-through, the level of import penetration, and the share of imported intermediate
inputs in total industry inputs. Jensen, Quinn, and Weymouth (2013) go further, show-
ing that currency undervaluation promotes demands for protection in some cases but
not others. They find clear differences in the international investment positions and the
composition of trade flows among US firms that file antidumping petitions and those
that do not: firms with vertical foreign direct investment (FDI) and high related-party
trade with host countries are less likely to initiate antidumping disputes against those
countries when their currencies are undervalued (Jensen, Quinn, and Weymouth
2013). In short, undervaluation’s effects are mediated through a firm’s investments in the
undervalued currency or its imports from the country with the undervalued exchange
rate. Thus, at both the country and firm levels, there appears to be a strong link between
exchange rate levels and trade protection.

Trade and Exchange Rate

Policy: Moving Beyond
the “Silos” of IPE

As this brief review of the existing literature illustrates, we now know quite a lot about
the political economy of trade and exchange rates, albeit less about the precise rela-
tionship between policy choices in each issue area. And yet in many ways we have only
scratched the surface of the possibilities, particularly within IPE. Vast swathes of the
trade literature within the subfield, for example, ignore exchange rates and monetary
policy entirely. The burgeoning literature on mass public opinion and trade policy pref-
erences (e.g., Scheve and Slaughter 2001; Hainmueller and Hiscox 2006; Mansfield and
Mutz 2009; Kuo and Naoi, this volume) has not, to our knowledge, addressed the issue
of the trade-exchange rate nexus. Likewise, the sizeable literature on the determinants
of WTO disputes completely ignores exchange rate levels and regime choices as possible
determinants of conflict between member states (e.g., Bown 2005; Busch and Reinhardt
2000).4 Finally, the rich literature on the design and effects of bilateral and regional
trade agreements (e.g., Haftel 2013; Kucik 2012) also does not incorporate exchange rate
regime choices or levels into the analysis. We single out these specific bodies of research
not because they are exceptional, but because they are representative. Despite wide-
spread general recognition of the inextricable relationship between these two policy
areas, the vast majority of IPE scholars continue to act as if trade policy and exchange
rate/monetary policy can be analyzed separately in “silos.”
 Mark S. Copelovitch and Jon C. W. Pevehouse    465

To some extent, this strategy is not particularly surprising. The literature on the
political economy of trade policy has developed largely from cases and data within
the United States, where exchange rates are substantially less salient as a political
issue than in smaller, more open economies and in those countries whose money
is not the global reserve currency.5 On the other hand, the literature on the political
economy of exchange rates over the last decade has focused far more extensively on
the adoption of fixed exchange rates as a solution to the time-inconsistency prob-
lem confronting monetary policy makers (e.g., Bernhard, Broz, and Clark 2002;
Hallerberg 2002; Keefer and Stasavage 2003; Guisinger and Singer 2010), rather than
on the trade implications of regime choice and exchange rate levels. The recent shift
back toward a focus on the trade policy implications, as discussed earlier, is a wel-
come and overdue adjustment in the literature. Thus, the first step to moving beyond
the current silos of IPE is to more closely incorporate the “other” policy area into
existing lines of research in the field. Trade scholars need to take more seriously the
exchange rate preferences of domestic actors, as well as the domestic and interna-
tional monetary commitments of governments. At the same time, monetary scholars
need to pay closer attention to trade institutions—particularly PTAs and the WTO—
and how they might affect governments’ choices of monetary policy and exchange
rate regimes.
Bridging the trade/monetary silos, however, raises a second issue in need of further
research: the direction of causality. As discussed above, there is substantial evidence
from existing work that the causal arrow points in both directions (trade policy deter-
mining exchange rate policy; exchange rate regime choice and monetary policy influ-
encing trade policy). Clarifying the conditions under which each relationship holds
presents both theoretical and empirical challenges, yet this must be a critical area of
focus for future work. One potential way to do this is to work toward developing a typol-
ogy of policy “baskets” (combinations of trade, monetary, and exchange rate policies),
rather than adopting the usual strategy of holding one policy or institution constant
while seeking to explain variation in the other.
Another way forward is to more closely look at lobbying on monetary policy and
exchange rates to see whether the same actors are involved in pressuring govern-
ments for both types of policies and whether these actors view trade and exchange
rate policies as complements or substitutes. Clearly there is an extensive empirical
literature on lobbying and trade policy (e.g., Gawande and Hoekman 2006; Baldwin
and Magee 2000; Busch and Reinhardt 2000), yet even much of that literature
makes strong assumptions about lobbying as necessarily correlated with the loca-
tion of industry or congressional votes as evidence of lobbying. And although there
are good reasons to believe that domestic actors will lobby less extensively on mon-
etary issues (Gowa 1988), such domestic pressure is the bread-and-butter of many
political economy models of exchange rate policy choice (e.g., Leblang 2003), and
more work is clearly warranted in this arena. Indeed, early researchers of exchange
rate policies were quite puzzled by the lack of open lobbying (Odell 1982), especially
given that extant models predicted the preferences of firms for various exchange rate
466   Bridging the Silos

policies under various political and economic circumstances. More recently, such
lobbying behavior has become more visible. For example, the US Congress spent
two years debating the Currency Reform for Fair Trade Act, meant to cajole China
into easing its peg on the Renminbi.6 This bill had more than two hundred cospon-
sors in the House and was the subject of intense lobbying, meant to link US trade
policy to currency issues. These types of instances are among those that could allow
IPE scholars to closely assess the presence and effectiveness of lobbying on exchange
rate policy.
A third way forward for the literature is to focus more clearly on the bureaucratic
agencies and actors involved in trade and monetary policy making. The trend toward
micro-level experimental approaches and the focus on legislative votes in the trade
literature masks the fact that—as has long been the case with monetary and exchange
rate policy, where central bankers play a major role in policy making—trade policy
is increasingly made by officials in executive agencies (e.g., the US Trade Repre­
sentative’s office, the European Commission). Yet we know almost nothing about
the political economy of these actors’ behavior. Instead, the vast majority of our
domestic politics models of trade and exchange rates continue to adopt a simple
“Grossman-Helpman” style policy-making model, in which office-seeking elected
officials trade off the interests of the median voter against those of organized sectors
or lobbies (“votes for campaign contributions”) (Grossman and Helpman 1994; for a
clear summary see Goodhart, this volume). A better understanding of the relation-
ships between trade and monetary policy-making bureaucrats—as well as the poten-
tial conflicts these actors have within the executive branch—would substantially
improve our understanding of the ways in which these two sets of policies are com-
plementary or in conflict with each other in particular countries. A focus on agencies
and bureaucrats could also answer the “puzzle” of why so little lobbying has appeared
visible to scholars to date (see Frieden and Broz 2001, 333): if lobbying occurs between
private agents and bureaucrats, there would be much less chance of discovering this
behavior.
Finally, the longer-term goal in terms of bridging the silos of IPE needs to be not just
linking the study of trade policy with that of exchange rates, but rather to move toward
simultaneously modeling trade, monetary policies, and financial policies. In recent
years IPE scholars have reached a broad consensus on the use of the Mundell-Fleming
“trilemma” or “impossible trinity” (Mundell 1960; Fleming 1962) as the workhorse
framework for analyzing the political economy of monetary policy and exchange rate
regime choice in an open economy. According to this framework, countries can achieve
only two of three policy goals simultaneously: a fixed exchange rate, full capital mobility,
and domestic monetary policy autonomy—the ability to adjust interest rates in reac-
tion to exogenous shocks or domestic economic downturns. The logic is straightfor-
ward: in a world of mobile capital, a discrepancy between domestic and world interest
rates causes capital to flow toward higher returns. Under floating exchange rates, capital
inflows will lead to an exchange rate appreciation, while outflows will result in deprecia-
tion. Under fixed exchange rates, however, interest rate differentials will be arbitraged
 Mark S. Copelovitch and Jon C. W. Pevehouse    467

away by these capital flows. Consequently, the combination of capital mobility and fixed
exchange rates makes it impossible for a government to adopt an independent monetary
policy.
The trilemma has proven enormously useful in identifying the trade-offs facing mac-
roeconomic policy makers in a world of global finance. And yet IPE scholars have not
really fully explored the implications of the trilemma. Rather, the common assump-
tion is to take full capital mobility as a given and analyze the political economy of the
remaining dilemma: the choice between monetary policy autonomy and fixed exchange
rates. Since the onset of the global financial crisis in 2007–2008, however, a wide range
of countries have imposed capital controls in an attempt to stem currency appreciation,
capital outflows, or capital inflows (e.g., Brazil, Iceland), and the International Monetary
Fund (IMF) has even revised its long-standing blanket opposition to the use of controls.
These developments make clear that capital account openness is also a policy choice,
and that policy makers around the world evaluate the pros and cons of financial liberal-
ization differently.
Of course there are nontrivial challenges in combining the study of trade and
exchange rates. First is the question of timing and time horizons. Exchange rates fluctu-
ate daily, and while large market-driven devaluations may take time, these changes still
happen more quickly than most legislative or bureaucratic processes. Thus, to discuss
how shifts in exchange rates influence government behavior, it is helpful to think about
timing. Changes in trade policy require legislation or administrative hearings. World
Trade Organization disputes involve complex legal procedures. Will the proposed pol-
icy actually help provide relief fast enough for those lobbying for it? How quickly does
the proposed policy take to enact?
The answers to such questions clearly depend on the causal mechanisms and the
model of politics employed. For example, if the trade policy remedy arising from
exchange rate fluctuations is enacted to provide actual relief to aggrieved parties, one
should probably focus on remedies arising out of the executive branch or emergency
relief policies. If the remedy is enacted to provide political cover—that is, so it appears
the government is “doing something” in response to lobbying—timing may be less
important. These two causal processes, however, are very different: in the former politi-
cal actions are the means, whereas in the latter political actions are the ends. Both of
these mechanisms also ignore the other states: Are tariffs raised as a bargaining chip in
international negotiations? The inevitable two-level game involving competing govern-
ments and their domestic constituencies becomes all the more complex once one theo-
rizes about multiple policy levers and domestic constituencies.
This leads to a second challenge: what we label the “blunt instrument” problem. Trade
policy can be narrowly tailored. In theory, tariffs can be placed on a very small or very
large number of goods—as specific as the six-digit product code in the Harmonized
System of Codes. Thus, political actors could respond to lobbying by aggrieved par-
ties with very focused trade policy measures. Adopting a new exchange rate policy or
even attempting to intervene in a managed float exchange rate system influences terms
of trade for all goods. Would a government alter exchange rate policy or intervene in
468   Bridging the Silos

exchange markets due to strong complaints from important exporters even if those poli-
cies would impose costs on all domestic actors through currency devaluation? Is the
cure worse than the disease?
Third is a question of political agency. Governments set tariff and exchange rate poli-
cies. Firms trade and lobby for trade policy. State actors and private actors (firms and
individuals) engage in currency exchange. Markets determine exchange rate levels
in floating exchange rate regimes; government agents do so in pegged and (to a lesser
extent) in managed float regimes. In the trade realm, it is largely assumed that state
policies will change the behavior of economic actors. But in the monetary realm, states
must compete with actors who, in the aggregate, possess resources that are stronger than
those of all but the largest states. Indeed, the size of transaction flows in each area is
significantly different: $18 trillion of goods are exported in a year; over $5 trillion per
day changes hands in FOREX markets. Can governments undertake policies, short of
major exchange rate regime changes, to influence monetary policy? While this is not a
new question in IPE, its implications for questions of trade and exchange rate policy are
significant: one reason governments may want to pull trade policy levers in response to
noncompetitive currency positions is their inability to influence their currency without
very costly interventions or policy changes. Still, any theory of linkages between trade
and exchange rate policies must confront the varying ability of states to influence trade
policy versus monetary policy change.
Fourth, if the causal story linking trade and monetary policy involves lobbying, find-
ing evidence of such lobbying can be difficult (Bearce 2003). As previously discussed,
while there are a myriad of papers on lobbying in the trade policy realm, few tackle the
difficult question of whether it is the lobbying that influences votes or simply economic
fundamentals. Put differently, with lobbying in any issue area, it is difficult to show the
counterfactual:  that without lobbying a government or legislator would behave dif-
ferently. Demonstrating that a legislator behaved consistently with her or his district’s
economic interests in a vote on tariffs is not evidence of lobbying, but simply the recog-
nition of the interests within that legislator’s district.
We have a final point on the opportunities and challenges of integrating the literatures
on trade policy and exchange rates. Note that empirically, the most voluminous work
exists on questions involving trade and monetary flows. This is particularly true in the
economics literature, where politics is often considered a nuisance or exogenous. And
this state of affairs is hardly surprising: armed with a theory of firm behavior, holding
constant policy and changes in policy, one can find how firms that create flows respond
to the incentives created by policy. Moving forward, empirically assessing the relation-
ship between trade and monetary policies requires a careful theory of political actors,
their preferences, and an aggregation mechanism to move from firm or constituent
demands to policy. In other words, economics has provided us with a common theory of
the firm from which a set of expectations about behavior flows. We have no equivalent
theory of politics, nor is such a theory desirable: it is exactly the variations in politics
that make policy outcomes interesting to study and that is the comparative advantage of
political science.
 Mark S. Copelovitch and Jon C. W. Pevehouse    469

Conclusion

For consumers and firms, the relationship between the trade in goods and exchange
rates has long been apparent. Unfortunately, most scholars of international politi-
cal economy have developed theories and models of each largely in isolation from the
another. This chapter reviewed the many ways in which trade and exchange rate policies
as well as trade and monetary flows influence one another.
We first examined the implications of trade policy for exchange rates and exchange rate
regime choice. While the least amount of work has occurred in this area, it is certainly not
an empty set. In particular, exploration of the evolution of trade integration agreements
to encompass monetary policy stands out as an important strand of literature examining
this question. Yet given that so few trade agreements evolve into monetary agreements,
there is unfortunately a paucity of systematic empirical work on this question.
More common are studies reversing the causal arrow: the influence of exchange rate
regimes on trade. Here, economists have dominated the study of the linkages between
finance and trade, yet little consensus has been reached on when exchange rate regimes
influence trade flows. A new set of research has recently emerged examining fluctua-
tions in exchange rates as determinants of trade policy choices such as targeted protec-
tion. This newer literature is a promising avenue to link finance to trade policy.
Finally, we outlined some of the problematic aspects of integrating theories of trade
policy and exchange rate choices. While these hurdles are surmountable, they are not
trivial, involving problems of blunt policy instruments, preference aggregation, and
political agency. Efforts to integrate the worlds of trade and finance should be under-
taken with care, paying equal attention to the many levers that can be pulled by policy
makers in each area.
As globalization continues to bring together actors in high-frequency interactions
across the trade and finance world, it is incumbent on IPE scholars to explore the link-
ages between actions taken by political actors to regulate and govern those interactions.
Here we have outlined several avenues that we think can be fruitful for exploring the
relationship between trade and finance.

Notes
1. See Milner (1999); Frieden and Martin (2002); and Lake (2009) for reviews of this vast
literature.
2. See Frieden and Broz (2001, 2006) and Frieden (2008) for reviews.
3. Such episodes of currency instability and protectionism were precisely why Europeans
pushed for completion of the single market in the 1992 Maastricht Treaty and moved
toward economic and monetary union in 1999. Within the Southern Cone, tensions aris-
ing from the Brazilian devaluation, the subsequent Argentine protectionist response,
and the ensuing financial crises in Argentina and Uruguay nearly led to the collapse of
Mercosur in the early 2000s.
470   Bridging the Silos

4. See Copelovitch and Pevehouse (2013b) for a preliminary attempt to address this gap in
the literature.
5. Indeed, it is not an accident that initial theories of trade and exchange rates were pioneered
by Robert Mundell, a Canadian economist!
6. Evans-Pritchard (2010).

References
Auboin, M., and M. Ruta. 2011. The Relationship between Exchange Rates and International
Trade: A Review of Economic Literature. WTO Staff Working Paper ERSD-2011-17.
Baccheta, P., and E. van Wincoop. 2000. Does Exchange Rate Stability Increase Trade and
Welfare? American Economic Review 93: 42–55.
Bahmani-Oskoosee, M., and S. Hegerty. 2007. Exchange Rate Volatility and Trade Flows:
A Review Article. Journal of Economic Studies 34 (3): 211–55.
Baldwin, Robert E., and Christopher S. Magee. 2000. Is Trade Policy for Sale? Congressional
Voting on Recent Trade Bills. Public Choice 105 (1–2): 79–101.
Bearce, David H. 2003. Societal Preferences, Partisan Agents, and Monetary Policy Outcomes.
International Organization 57 (2): 373–410.
Belassa, Bela. 1961. Towards a Theory of Economic Integration. Kyklos 14 (1): 1–17.
Bergsten, C. Fred. 2010. Correcting the Chinese Exchange Rate: An Action Plan. http://www.
voxeu.org/index.php?q=node/4876.
Bergsten, C. Fred. 2006. The US Trade Deficit and China. Testimony at the Hearing on
US-China Economic Relations Revisited. Committee on Finance, United States Senate,
March 29. http://www.iie.com/publications/papers/paper.cfm?ResearchID=611.
Bergsten, C. Fred, and John Williamson. 1983. Exchange Rates and Trade Policy. In Trade Policy
in the 1980s, edited by William R. Cline, 99–120. Washington, DC: Institute for International
Economics.
Bernhard, William, J. Lawrence Broz, and William Roberts Clark. 2002. The Political Economy
of Monetary Institutions. International Organization 56 (4): 693–723.
Blomberg, S. Brock, Jeffry Frieden, and Ernesto Stein. 2005. Sustaining Fixed Rates:  The
Political Economy of Currency Pegs in Latin America. Journal of Applied Economics 8
(2): 203–225.
Bown, Chad P. 2005. Participation in WTO Dispute Settlement:  Complainants, Interested
Parties, and Free Riders. World Bank Economic Review 19 (2): 287–310.
Broz, Lawrence, Jeffry Frieden, and Stephen Weymouth. 2008. Exchange Rate Policy
Attitudes: Direct Evidence from Survey Data. IMF Staff Papers 55 (3): 417–444.
Broz, Lawrence, and Seth Werfel. 2013. Exchange Rates and Industry Demands for Trade
Protection. International Organization 68 (2): 393–416.
Busch, Marc. 2007. Overlapping Institutions, Forum Shopping, and Dispute Settlement in
International Trade. International Organization 61 (4): 735–761.
Busch, Marc, and Eric Reinhardt. 2000. Bargaining in the Shadow of the Law: Early Settlement
and GATT/WTO Disputes. Fordham International Law Journal 24: 158–165.
Calvo, Guillermo, and Carlos Vegh. 1994. Inflation Stabilization and Nominal Anchors.
Contemporary Economic Policy 12 (2): 35–45.
Campa, Jose Manuel, and Linda Goldberg. 2005. Exchange Rate Pass-Through into Import
Prices. Review of Economics and Statistics 87 (4): 679–690.
 Mark S. Copelovitch and Jon C. W. Pevehouse    471

Campa, Jose Manuel, and Linda Goldberg. 1999. Investment, Pass-through, and Exchange
Rates: A Cross-Country Comparison. International Economic Review 40 (2): 287–314.
Cooper, Scott. 2007. Why Doesn’t Regional Monetary Cooperation Follow Trade Cooperation?
Review of International Political Economy 14 (4): 626–652.
Copelovitch, Mark S., and Jon C.  W. Pevehouse. 2013a. Ties That Bind? Preferential Trade
Agreements and Exchange Rate Policy Choices. International Studies Quarterly 57
(2): 385–399.
Copelovitch, Mark S., and Jon C.  W. Pevehouse. 2013b. The Trilemma and Trade Policy:
Exchange Rates, Financial Openness, and WTO Disputes. Unpublished manuscript,
University of Wisconsin.
Corden, W. Max. 1982. Exchange Rate Protection. In The International Monetary System
Under Flexible Exchange Rates—Global, Regional, and National: Essays in Honor of Robert
Triffin, edited by Richard Cooper, Peter Kenen, Jorge Braga de Macedo, and Jacques Van
Ypersele, 17–34. Cambridge, MA: Ballinger.
De Grauwe, P. 1988. Exchange Rate Variability and Slowdown in International Trade. IMF Staff
Papers 35: 63–84.
Destler, I. M., and C. Randall Henning. 1989. Dollar Politics: Exchange Rate Policymaking in the
United States. Washington, DC: Institute for International Economics.
Devereaux, M. B. and Charles Engel. 2002. Exchange Rate Pass-Through, Exchange Rate
Volatility and Exchange Rate Disconnect. Journal of Monetary Economics 49 (5): 913–940.
Eichengreen, B. 1992. Golden Fetters: The Gold Standard and the Great Depression, 1919–1939.
Oxford, UK: Oxford University Press.
Eichengreen, B. 1993. The Crisis in the EMS and the Transition to EMU:  An Interim
Assessment. CIDER Working Paper C93-022. University of California–Berkeley.
Eichengreen, B., and D. Irwin. 2010. The Slide to Protectionism in the Great Depression: Who
Succumbed and Why? Journal of Economic History 70 (4): 871–97.
Engel, Charles, and John Rogers. 1996. How Wide Is the Border? American Economic Review
86: 1112–1125.
Evans-Pritchard, Ambrose. 2010. Currency Wars Are Necessary If All Else Fails. The Telegraph,
October 10. http://www.telegraph.co.uk/finance/comment/ambroseevans_pritchard/ 8054
066/Currency-wars-are-necessary-if-all-else-fails.html.
Fernández-Arias, Eduardo, Ugo Panizza, and Ernesto Stein. 2004. Trade Agreements,
Exchange Rate Disagreements. In Monetary Unions and Hard Pegs:  Effects on Trade,
Financial Development, and Stability, edited by Volbert Alexander, George M. von
Furstenberg, and Jacques Melitz, 135–150,. New York: Oxford University Press.
Fleming, Marcus. 1962. Domestic Financial Policies under Fixed and under Floating Exchange
Rates. IMF Staff Papers 9: 369–380.
Frankel, Jeffrey. 1994. The Making of Exchange Rate Policy in the 1980s. In American Economic
Policy in the 1980s, edited by Martin Feldstein, 293–341. Chicago: University of Chicago Press.
Frankel, Jeffrey, and Andrew Rose. 2002. An Estimate of the Effect of Common Currencies on
Trade and Income. Quarterly Journal of Economics 117 (2): 437–466.
Frieden, Jeffry. 1991. Debt, Development and Democracy: Modern Political Economy and Latin
America 1965–1985. Princeton, NJ: Princeton University Press.
Frieden, Jeffry. 1993. The Dynamics of International Monetary Systems:  International and
Domestic Factors in the Rise, Reign, and Demise of the Classical Gold Standard. In Coping
with Complexity in the International System, edited by Jack Snyder and Robert Jervis,
137–162. Boulder, CO: Westview Press.
472   Bridging the Silos

Frieden, Jeffry. 2008. Globalization and Exchange Rate Policy. In The Future of Globalization,
edited by Ernesto Zedillo, 344–357. New York: Routledge.
Frieden, Jeffry. 1997. Monetary Populism in Nineteenth-Century America: An Open-Economy
Interpretation. Journal of Economic History 57 (2): 367–395.
Frieden, Jeffry. 2002. Real Sources of European Currency Policy:  Sectoral Interests and
European Monetary Integration. International Organization 56: 831–860.
Frieden, Jeffry A., and J. Lawrence Broz. 2006. The Political Economy of Exchange Rates. In
Oxford Handbook of Political Economy, edited by Barry Weingast and Donald Wittman,
587–600. New York: Oxford University Press.
Frieden, Jeffry A., and J. Lawrence Broz. 2001. The Political Economy of International
Monetary Relations. Annual Review of Political Science 4: 317–343.
Frieden, Jeffry A., and Lisa L. Martin. 2002. International Political Economy:  Global and
Domestic Interactions. In Political Science:  The State of the Discipline, edited by Ira
Katznelson and Helen Milner, 118–146, New York: W.W. Norton.
Gawande, Kishore, and Bernard Hoekman. 2006. Lobbying and Agricultural Trade Policy in
the United States. International Organization 60 (3): 527–561.
Goldberg, Pinelopi, and Michael Knetter. 1997. Goods Prices and Exchange Rates: What Have
We Learned? Journal of Economic Literature 35: 1243–1292.
Gowa, Joanne. 1983. Closing the Gold Window: Domestic Politics and the End of Bretton Woods.
Cornell Studies in Political Economy. Ithaca, NY: Cornell University Press.
Gowa, Joanne. 1988. Public Goods and Political Institutions: Trade and Monetary Processes in
the United States. International Organization 42 (1): 15–32.
Grilli, Enzo. 1988. Macroeconomic Determinants of Trade Protection. World Economy 11
(3): 313–326.
Grossman, Gene M., and Elhanan Helpman. 1994. Protection for Sale. American Economic
Review 84 (4): 833–850.
Guisinger, Alexandra, and David Andrew Singer. 2010. Exchange Rate Proclamations and
Inflation-Fighting Credibility. International Organization 64 (1): 313–337.
Haas, Ernst B. 1958. The Uniting of Europe. Notre Dame, IN: Notre Dame Press.
Haftel, Yoram. 2013. Commerce and Institutions: Trade, Scope, and the Design of Regional
Economic Organizations. Review of International Organizations 8: 389–414.
Hainmueller, Jens, and Michael Hiscox. 2006. Learning to Love Globalization:  Education
and Individual Attitudes Toward International Trade. International Organization 60
(2): 469–498.
Hallerberg, Mark. 2002. Veto Players and Monetary Commitment Technologies. International
Organization 56 (4): 775–802.
Huchet-Bourdon, M., and J. Korinek. 2011. To What Extent Do Exchange Rates and Their
Volatility Affect Trade? OECD Trade Policy Papers No. 119. Paris:  OECD. http://dx.doi.
org/10.1787/5kg3slm7b8hg-en.
International Monetary Fund. 2004. Exchange Rate Volatility and Trade Flows—Some New
Evidence. IMF Occasional Paper 235.
International Monetary Fund. 1984. Exchange Rate Volatility and World Trade. IMF Occa­
sional Paper 30.
International Monetary Fund (IMF). 1997. World Economic Outlook. Washington, DC:
International Monetary Fund.
Irwin, Douglas A. 2005. The Rise of U.S. Anti-Dumping Activity in Historical Perspective.
World Economy 28 (5): 651–668.
 Mark S. Copelovitch and Jon C. W. Pevehouse    473

Irwin, Douglas A. 2012. Trade Policy Disaster:  Lessons from the 1930s. Cambridge, MA:
MIT Press.
Jensen, J. Bradford, Dennis Quinn, and Stephen Weymouth. 2013. Global Supply Chains,
Currency Undervaluation, and Firm Protectionist Demands. NBER Working Paper No. 19239.
Keefer, Philip, and David Stasavage. 2003. Checks and Balances, Private Information, and the
Credibility of Monetary Commitments. International Organization 56 (4): 751–774.
Kenen, Peter. 1969. The Theory of Optimum Currency Areas: An Eclectic View. In Monetary
Problems in the International Economy, edited by Robert Mundell and Alexander Swoboda,
41–60. Chicago: University of Chicago Press.
Klein, Michael W. 2005. Dollarization and Trade. Journal of International Money and Finance.
24 (6): 935–43.
Klein, Michael W., and Jay Shambaugh. 2010. Exchange Rates in the Modern Era. Cambridge,
MA: MIT Press.
Klein, Michael W., and Jay Shambaugh. 2006. Fixed Exchange Rates and Trade. Journal of
International Economics 70 (2): 359–383.
Knetter, Michael M., and Thomas J. Prusa. 2003. Macroeconomic Factors and Antidumping
Filings: Evidence from Four Countries. Journal of International Economics 61: 1–17.
Kucik, Jeffrey. 2012. The Domestic Politics of Institutional Design: Producer Preferences over
Trade Agreement Rules. Economics and Politics 24 (2): 95–118.
Lake, David A. 2009. Open Economy Politics:  A  Critical Review. Review of International
Organizations 4: 219–244.
Leblang, David. 2003. To Devalue or to Defend? The Political Economy of Exchange Rate
Policy. International Studies Quarterly 47 (4): 533–560.
López-Córdova, Jose Ernesto, and Christopher M. Meissner 2003. Exchange-Rate Regimes and
International Trade: Evidence from the Classical Gold Standard Era. American Economic
Review 93(1): 344–353.
Mansfield, Edward D., and Helen Milner. 1999. The New Wave of Regionalism. International
Organization 53 (3): 589–627.
Mansfield, Edward D., and Diana Mutz. 2009. Support for Free Trade: Self-Interest, Sociotropic
Politics, and Out-Group Desire. International Organization 63 (3): 425–457.
Mansfield, Edward D., and Jon C. W. Pevehouse 2013. The Expansion of Preferential Trade
Agreements. International Studies Quarterly 57 (3): 592–604.
McKinnon, Ronald I. 1973. The Dual Currency System Revisited. In The Economics of Common
Currencies, edited by H. G. Johnson and A. K. Swoboda, 85–92,. Cambridge, MA: Harvard
University Press.
McKinnon, Ronald I. 1962. Optimum Currency Areas. American Economic Review 53: 717–725.
McKinnon, Ronald I., and K. C. Fung. 1993. Floating Exchange Rates and the New Interbloc
Protectionism. In Protectionism and World Welfare, edited by Dominick Salvatore, 221–244.
Cambridge, UK: Cambridge University Press.
McNamara, Kathleen. 1999. The Currency of Ideas: Monetary Politics in the European Union.
Princeton, NJ: Princeton University Press.
Milner, Helen. 1999. The Political Economy of International Trade. Annual Review of Political
Science 2: 91–114.
Milner, Helen V., and B. Peter Rosendorff. 2001. The Optimal Design of International Insti­
tutions: Uncertainty and Escape. International Organization 55 (4): 829–857.
Mundell, Robert. 1960. The Monetary Dynamics of International Adjustment under Fixed and
Flexible Exchange Rates. Quarterly Journal of Economics 74: 227–50.
474   Bridging the Silos

Mundell, Robert. 1961. A Theory of Optimum Currency Areas. American Economic Review
51: 657–664.
Niels, Gunnar, and Joseph Francois. 2006. Business Cycles, the Exchange Rate, and Demand
for Antidumping Protection in Mexico. Review of Development Economics 10 (3): 388–399.
Oatley, Thomas. 2010. Real Exchange Rates and Trade Protectionism. Business and Politics 12
(2): 1–17.
Odell, John. 1982. US International Monetary Power: Markets, Power, and Ideas as Sources of
Change. Princeton, NJ: Princeton University Press.
Pearce, Joan, and John Sutton. 1985. Protection and Industrial Policy in Europe. London:
Routledge & Kegan Paul.
Pomfret, M., and V. Pontines. 2013. Exchange Rate Policy and Regional Trade Agreements:
A Case of Conflicted Interests? ADBI Working Paper 436. Tokyo: Asian Development Bank
Institute. http://www.adbi.org/working-paper/2013/10/08/5909.exchange.rate.policy.trade.
agreements/.
Rose, Andrew. 2000. One Money, One Market: The Effect of Common Currencies on Trade.
Economic Policy: A European Forum 30: 7–33.
Sandholz, Wayne, and Alec Stone Sweet, eds. 1998. European Integration and Supranational
Governance. Oxford: Oxford University Press.
Scheve, Kenneth, and Matthew Slaughter. 2001. What Determines Individual Trade Policy
Preferences? Journal of International Economics 54 (2): 267–292.
Schmitter, Phillipe C. 1969. Three Neo-functional Hypotheses About International Integration.
International Organization 23 (1): 161–166.
Simmons, Beth. 1994. Who Adjusts? Domestic Sources of Foreign Economic Policy During the
Interwar Years. Princeton, NJ: Princeton University Press.
Svensson, Lars E. O. 1999. Inflation Targeting as a Monetary Policy Rule. Journal of Monetary
Economics 43 (3): 607–654.
Walter, Stefanie. 2008. A New Approach for Determining Exchange-Rate Level Preferences.
International Organization 62 (3): 405–438.
Walter, Stefanie. 2013. Financial Crises and the Politics of Macroeconomic Adjustment.
Cambridge, UK: Cambridge University Press.
 Chapter 25

Tr ade and Dev e l op me nt

Ma rk S . Ma nge r an d
Ke nneth C. Shadle n

Introduction

Since the latter years of the twentieth century developing countries have become increas-
ingly integrated in international trade. There is near consensus that engagement in inter-
national trade is a necessary condition for economic development, reduction of poverty,
and improved living standards. But what steps should countries undertake to increase
their participation in international trade? Under what terms should countries engage?
These questions remain debated. In this chapter we attempt to shed light on the political
underpinnings of the trade policy choices that countries face regarding these issues.
The chapter is organized around two themes. First is the role of trade in development
policy, and in particular the relationships between trade, industrialization strategy, and
economic development. We maintain that trade needs to be understood in the context
of broader constellations of economic policies, and that scholars need to keep in mind
the overarching commonalities with regard to the role of the state in late development.
Importantly, we suggest that not enough work has gone into analyzing the political
economy conditions that underpin variants of state intervention and “developmental”
trade policies.
The second theme is the challenges developing countries face when they adopt more
explicitly export-oriented strategies to secure greater integration into the global trade
regime. Here we emphasize that following a period of rapid, unilateral liberalization in
the aftermath of major economic crises, many developing countries have chosen a par-
ticular route of trade opening: preferential trade agreements between North and South.
While accession to the World Trade Organization (WTO) is often a laborious process
(see Davis and Wilf, this volume), nearly all countries have decided to go this route. By
contrast, with whom to negotiate a preferential trade agreement (PTA) and what liberal-
ization to agree to in doing so is a policy choice that varies across countries.
476   Trade and Development

Before we address these two themes, we highlight three overarching issues that give
the politics of trade a distinct flavor in developing countries: the relative importance of
trade to national economies, the importance of trade taxes as a source of government
revenue, and a set of structural characteristics that complicate expected patterns of
adjustment to changing trade flows. These conditions arguably make trade more promi-
nent in policy debates in developing than developed countries, and often more salient in
the minds of citizens.
First and foremost, trade typically makes up a large share of gross domestic product
(GDP) for smaller, developing economies. In 2011 the total of exports and imports was
equal to more than 30 percent of GDP in every developing country (World Bank 2013).
For nearly a third of countries, combined exports and imports exceeded GDP. Even dur-
ing the period of most restrictive import policies in the 1960s, larger developing econo-
mies like Argentina and Mexico had trade shares in GDP around 15 percent, a figure the
United States did not attain until the mid-1970s. The archetype of a small trading state,
the Netherlands, only reached a trade-to-GDP share greater than 100 percent in 1980.
Individual developing countries may contribute little to global trade, but they are none-
theless buffeted by the winds of trade. At the same time, this means that export growth
can potentially contribute more to national income than in larger, less open, and more
mature economies.
Likewise, taxes on external trade make up a much larger share of government rev-
enue in developing countries, with an average of 10 percent in 2011. Least-developed
countries relied on tariffs on imports and exports for up to 30 percent, and this figure
is probably understated given that many of the poorest countries do not report any
data. By comparison, the average for Organisation for Economic Co-operation and
Development (OECD) countries was a mere 0.85 percent. Trade liberalization thus has
manifestly different effects on the revenue stream of developing country governments,
with important implications for trade policy choices (see Kono, this volume). Economic
growth, gains from trade, and the generation of new sources of tax revenue could of
course more than compensate for lost tariff revenues, but this presupposes the capacity
of states to tax the economy by means other than tariffs. This has historically been—and
remains—a challenge for most developing countries.
Trade liberalization, in addition to compelling states to find new sources of revenue,
also requires societal actors (firms, workers) to adjust. Yet in settings marked by lower
levels of education and human capital development, poorer infrastructure, less mature
financial and credit systems, and, in general more recurrent market failures—all dis-
tinguishing characteristics of developing countries—adjustment becomes complex and
difficult. Rarely is adjusting to international trade a simple and smooth process, but the
abilities of actors to redeploy their resources in other sectors where they are more pro-
ductive should be regarded as at least in part a function of these structural and institu-
tional characteristics that may systematically differ in developing countries. This is not
to deny the welfare gains that can be generated by adjustment and resource redeploy-
ment; the point is simply that trade liberalization raises different and potentially more
complex political challenges in developing countries. A prominent argument in political
 Mark S. MANGER AND Kenneth C. SHADLEN    477

economy is that the welfare state has expanded in industrialized countries in response
to trade opening—but in developing countries, welfare states, where they exist, have
tended to be primarily accessible for urban citizens employed in the formal sector, plac-
ing greater burdens of adjustment on those outside. Moreover, the presence of a large
pool of reserve labor can keep wages at minimal levels (the classic argument by Lewis
1954) and lead to an erosion of existing welfare states as a consequence of trade opening
(Rudra 2002). In short, the social and economic costs of trade liberalization in devel-
oping countries are often enormous, with little state capacity to soften the blow—and
there remains insufficient research on the relationship between trade policy and social
protection aside from the important contributions by Rudra (2004) and Fernández de
Córdoba and Laird (2006).

Trade, Industrialization,
and Development

Throughout much of the post–World War II era the contribution of trade policy to eco-
nomic development has been seen through the prism of competing approaches to the
challenges of industrialization. The vast majority of developing countries started out
as commodities exporters and, as postulated by development economists such as Raúl
Prebisch (1949) and Hans Singer (1950), the terms of trade for such countries were per-
ceived to be continuously worsening and subject to endemic volatility, necessitating
sustained efforts at industrialization. This premise, the empirical accuracy and substan-
tive effects of which were subject to considerable debate (Spraos 1980), has continued to
frame policy making and inspire efforts to diversify exports in many developing coun-
tries. The dominant question, then, was not so much the relationship between trade and
development directly, but rather indirectly, via trade’s effects on the industrialization
process.
To build their own industries, most developing countries aimed to replace imports
with domestically produced goods and to structure their economic relationships with
the developed world to facilitate such industrialization strategy efforts; otherwise, it
was feared, developing countries would continue to specialize in primary commodity
exports and remain vulnerable to declining and volatile terms of trade. Thus, virtually all
developing countries’ industrialization efforts began with efforts to substitute imports, a
characteristic that makes patterns of “late-late industrialization” (Hirschman 1968) dis-
tinct from the experience of “late industrialization” described by Gerschenkron (1962).
Key elements of “import-substitution industrialization” (ISI) strategies included
tariffs and nontariff measures (NTMs) to shield local producers of consumer goods
from international competition, along with subsidies and tax incentives to encourage
investment in industry. In its initial phases, ISI tended to focus on producing basic con-
sumer goods, for which demand existed and which did not require significant capital
478   Trade and Development

expenditure or technological prowess and thus were within reach of local investors.1
Capital goods that could not (yet) be produced in developing countries would have to
be imported, however, and with only the earnings from commodity exports available
to pay for said imports, balance-of-payments difficulties were sure to arise (Chenery
1958; Diaz-Alejandro 1965). In response, many countries made efforts to “deepen” their
industrial structures, sometimes referred to as moving to “secondary” stages of ISI, and
produce more complex producer goods, capital goods, and capital-intensive consumer
durables locally. To do so, the array of trade policies (e.g., tariffs and NTMs, tax incen-
tives, subsidies) introduced in the early phases of ISI were extended to new sectors.
Importantly, policies to regulate imports were also supplemented by a broad complex
of “industrial policy” instruments. By imposing local content requirements on foreign
investors, for example, many developing countries sought to compel transnational firms
to source more of their inputs locally and thus increase demand for domestically manu-
factured goods (and also increase foreign firms’ incentives to work with and raise the
quality of local suppliers). Countries often conditioned foreign investors’ participation
in the local market on forming joint ventures with local partners, sharing technology,
reinvesting revenues in the domestic economy, and employing nationals in key manage-
ment positions.2 Beyond regulating the activities of foreign investors, “strategic” sec-
tors were often reserved outright for national firms. Many countries also had significant
levels of state ownership in key sectors (e.g., energy, electricity, finance) to provide local
firms with critical inputs. Preferential treatment of national firms in government pro-
curement was often used as a tool to promote local industry. States actively regulated
technology markets, often requiring licensing and technology transfer agreements to be
registered with and approved by trade and industry officials. Many countries introduced
lax intellectual property (IP) regimes that made patents difficult to obtain (and, where
obtainable, difficult to assert), to facilitate local actors’ efforts to access and reverse-engi-
neer foreign technologies. Obviously developing countries displayed important differ-
ences in the relative priorities and emphases given to different instruments. The point
of this brief—and by no means exhaustive—review of industrial policy instruments is
to drive home a frequently overlooked point that development policies in this period
included both “trade” and a multitude of “trade-related” industrial policies. Industrial
policy was not only about reserving markets for local firms, “infant industry” protec-
tion, but about creating incentives for investment and capabilities in new productive
sectors.
Notwithstanding broad commonalities in trade and industrial policy instruments,
they were often deployed toward different ends. While the prevailing tendency in Latin
America (particularly in the larger countries in the region) was to utilize these instru-
ments toward deepening of ISI, the tendency in East Asia was to complement ISI with
efforts to promote manufactured exports (Gereffi and Wyman 1990; Haggard 1990).
East Asian countries (most notably South Korea and Taiwan) also sought to deepen
their industrial structures through the promotion of local manufacturing capabilities,
but industries that benefited from government promotion were encouraged to achieve
export proficiency. They were often required to do so:  according to Amsden’s (1989)
 Mark S. MANGER AND Kenneth C. SHADLEN    479

prominent account of Korea, meeting export targets was the quid pro quo for receiving
state support. Indeed, the enforcement of “reciprocal control mechanisms” (Amsden
2001) was the hallmark of successful “developmental states” (Woo-Cumings 1999; Wong
2004). In these countries ISI and export-led growth became not opposites, as typically
portrayed, but flip sides of a coin.
As the discussion above reveals, countries across the developing world tended to
actively attempt to direct resources into particular sectors, utilizing not just trade poli-
cies per se but also investment, technology transfer, IP, and other “trade-related” instru-
ments to create new capabilities. What distinguished countries from each other was not
so much whether they were “liberal” or “statist,” but the extent to which state interven-
tions included measures to create and sustain export-competitive firms and sectors.
Though the extent of the differences must not be exaggerated, as even the archetypal ISI
poster children, large Latin American countries such as Argentina, Brazil, and Mexico,
were putting significantly greater efforts into manufactured exports in the 1970s, all
in all it is clear that the array of policy tools to support the local manufacturing base
and generate industrial exports was deployed in systematically different ways across
regions (Gereffi and Wyman 1990). The result of these two different trajectories is that
industrialization in Latin America was more prone to balance-of-payments deficits and
thereby became more dependent on external financing (which ultimately contributed
to debt crises in the 1980s), whereas industrialization in East Asia was more balanced as
imported inputs needed for subsequent industrialization could more easily be financed
from manufactured export revenues.3
Understanding the political conditions underlying these different trajectories is
essential. Why were some countries, particularly in East Asia, so much more able
to combine policies designed to promote rapid industrial development with greater
engagement in global trade? Attributing East Asian dynamics to the existence of “strong
states” that allegedly were insulated from societal pressures and thus able to manage
trade and other economic policies effectively is unsatisfactory without an explanation of
the social and political underpinnings of state strength (Robinson 2011).
Scholars of the political economy of trade have proffered many explanations for the
differences. Some focus on distinct external conditions, such as different forms of for-
eign aid and financing available (Stallings 1990) and the imperatives that conditions of
“systemic vulnerability” provided East Asian governments to construct more efficacious
state institutions to promote industrial upgrading and export-competitive firms (Doner,
Ritchie, and Slater 2005). Scholars have also suggested that Latin American countries’
earlier period of economic development and wealth had the perverse effect of making
shifts of the sort witnessed in Asia more challenging, either because of the greater attrac-
tiveness of local markets for transnational firms seeking to deepen ISI (Maxfield and
Nolt 1990) or because greater wealth and union density would have made the costs of
shifting economic strategies politically prohibitive (Mahon 1992).
Another explanation has to do with the effects of large-scale land reforms (Kay 2002;
Schrank 2007). Late-late industrialization requires significant resource transfers from
the rural sector (agriculture, natural resources) to the urban sector. In countries that
480   Trade and Development

had land reforms that weakened agricultural elites, the transfer of resources could be
made directly via taxation. In contrast, in countries without significant land reforms,
where entrenched landlords or agricultural elites could resist direct taxation of this
sort, intersectoral transfers tended to be achieved indirectly, via overvalued exchange
rates (Hirschman 1968). The pernicious effects on industrial development of persis-
tently overvalued exchange rates are well known and need not be elaborated on here. We
simply add that, in as well as providing industrialists with strong disincentives against
trying to engage in exports, the environment created by overvalued real exchange
rates (RERs). Overvalued RERs made it more difficult for governments to establish the
sorts of institutional arrangements and enforce the array of performance requirements
and reciprocal control mechanisms that were the cornerstone of developmental states
(Amsden 2001). Quite simply, intervention to promote local industry in the context of
overvalued exchanges must depend more on tariffs, but because tariffs also serve a pur-
pose of providing governments with sources of revenue, threats to withdraw the sup-
port (i.e., to enforce the performance requirement) are less credible. Subsidies and tax
incentives, in contrast, impose costs on the public purse, and thus the threat of removal
(“meet this export target within five years or the subsidy will no longer be available”) is
more credible.
The debt crises in the 1980s set off a sea change in trade policy. With lack of access
to external finance making acquisitions of key imports virtually impossible, many
countries’ trade and industrialization strategies became unsustainable. Unprecedented
debt-servicing obligations for many countries made generating foreign exchange rev-
enues through increased exports an imperative. At the same time, technological change
that facilitated the fragmentation of production created new opportunities for coun-
tries to participate in international trade networks. Thus, the 1980s and 1990s witnessed
shifts, throughout the developing world, toward new strategies of economic develop-
ment based on more liberalized trade and export orientation, with a principal objec-
tive of trade and industrialization strategies to build capabilities to integrate into global
value chains (GVCs).
New and shifting economic environments—internal in terms of the sustainability
of prevailing models of development and external in terms of new opportunities and
challenges created by technological change and the emergence of GVCs—were accom-
panied by changes in the nature of international economic regimes—that is, the for-
mal and informal rules and regulations that structure the global trading system. These
changes include, of course, a new approach toward trade, industrialization, and devel-
opment promoted by the World Bank, other international financial institutions, and
foreign donors (Woods 2007)—what has been called the “Washington Consensus,”
with an emphasis on liberalization and reliance on international price signals to allo-
cate resources within economies. So, for example, countries that turned to interna-
tional financial institutions such as the International Monetary Fund and World Bank
for support during episodes of debt restructuring were typically subject to “structural
adjustment” conditionalities that included, among other things, trade liberalization and
deregulation of foreign investment.
 Mark S. MANGER AND Kenneth C. SHADLEN    481

This period also witnessed the conclusion of the Uruguay Round and the creation of
the WTO to supplant the General Agreement on Tariffs and Trade (GATT). The WTO
introduced a significant expansion in the scope of international trade governance. In
contrast to the GATT, which had a narrow remit focused largely on tariffs of manu-
factured goods and left most policy areas unaffected and not subject to international
constraints,4 the WTO includes a multitude of agreements affecting a nearly all aspects
of international trade. Agreements on tariff levels for manufactured goods are supple-
mented by rules on, for example, subsidies, health and safety measures, product stan-
dards, agricultural trade, and customs valuation. Moreover, the WTO now includes
rules that affect national policies in a range of other areas, such as services (shipping,
telecommunications, finance, movement of people), regulation of foreign investment,
and IP. In the simplest terms, this is a shift from an international trade regime designed
to facilitate “shallow integration” to one that attempts to dismantle barriers to “deep
integration” (Lawrence 1996).
Changes in the nature of production and trade and in the international rules govern-
ing trade go hand in hand. After all, the construction of GVCs based on fragmented
and modularized production is only feasible to the extent that barriers—both “trade”
and “trade-related”—to markets and restrictions on economic activity are eliminated or
reduced. Technological change creates opportunities for the creation of integrated pro-
duction networks and GVCs, but to go further, to make production networks truly inte-
grated (to put the global in GVCs), national rules and regulations regarding the entry of
firms to markets and the activities that firms undertake in markets must be addressed.
Firms creating GVCs may seek to take advantage of cost and locational advantages
and thus source different stages of production in different settings (directly or via sub-
contracting); to conduct different elements of R&D or design in different countries; to
locate billing, finance, and customer service operations in yet other locations; and so
forth. But the ability to take these steps depends on national policies. Thus, firms cre-
ating (or seeking to create) GVCs become concerned not just with developing coun-
tries’ tariffs and NTMs that directly affect imports, but with regulations and policies that
affect investment, IP, services, and other areas that affect interfirm relations. In short,
broadening of the scope of the international trade regime—from shallow integration to
deep integration—creates the political underpinnings to facilitate the globalization of
production and the expansion of GVCs.
At the same time that the GATT/WTO’s rules became more expansive, they also
became more popular. Prior to the 1980s membership in the GATT was far from univer-
sal, but the new emphasis on exports and, relatedly, attracting efficiency-seeking foreign
investors to help integrate national economies in GVCs, made countries much more
concerned with the international trade regime. After all, only as members of the GATT/
WTO can countries be assured that their exports will benefit from most-favored nation
(MFN) treatment in developed countries’ markets. Thus, with more countries now
members of an international organization with such a wide remit as the WTO, virtu-
ally the entire array of trade and “trade-related” policies that had constituted countries’
economic development strategies became subject to new sets of international rules. The
482   Trade and Development

establishment of a new complex of rules on both “trade” and “trade-related” issues was
controversial, and their existence presents developing countries with a new interna-
tional environment. The following section examines the political economy of trade and
development policies in the contemporary international context.

Development and Trade

Policy Choices

Until the late 1990s, developing countries had very little influence on the institutions of
the global trade regime, and they had to accept its rules upon joining. While many for-
mer colonies availed themselves of GATT provisions in Article XXVI:5(c) that allowed
automatic accession with very little liberalization, others delayed accession until after
the creation of the WTO and thus faced arduous negotiations (Copelovitch and Ohls
2012). Late entrants had to liberalize their trade policies and dismantle other regula-
tory policies much more deeply and rapidly than founding members. And yet while
the WTO certainly places constraints and limitations on developing countries’ policy
choices, eliminating some important tools from the arsenal of available measures, most
analysts tend to agree that strategies to promote key sectors, build new capabilities, and
upgrade industries remain feasible for WTO members (Amsden and Hikino 2000;
Shadlen 2005).5
Much more constraining than the WTO, however, are the disciplines in North-South
PTAs, which have multiplied at a rapid pace. Indeed, following the unilateral changes
in the trade policy of developing countries in the late 1980s and early 1990s, subsequent
policy changes have almost exclusively been the cause and consequence of PTA nego-
tiations. While “deep integration” (see Kim, this volume) is becoming common among
PTAs in general, at least for agreements among developed countries this appears as a
response to slow progress at the WTO. Although, for example, the Australia-US PTA
covers much more than tariffs and trade in goods, the required policy changes were
limited. Since the turn of the millennium, South-South PTAs have also become more
common, but these sorts of agreements typically do not entail much liberalization: The
“Enabling Clause” agreed upon in the 1979 Tokyo Round of the GATT allows develop-
ing countries to postpone liberalization or exclude whole sectors. Some South-South
agreements may be meaningful (e.g., most PTAs involving the ASEAN countries), but
others (e.g., Morocco-Pakistan) neither cover much trade nor prescribe any tariff reduc-
tions. By contrast, North-North and North-South PTAs fall under GATT Article  24
and thus participating states have to eliminate nearly all barriers on trade within ten to
fifteen years.
Given this context, North-South PTAs often require the creation of regulation from
scratch and transform long-standing policies in developing countries.6 In a sense the
exchange implicit in many North-South PTAs might be regarded as asymmetrically
 Mark S. MANGER AND Kenneth C. SHADLEN    483

enhanced versions of the WTO. Developing countries receive somewhat improved

market access to developed countries’ markets, in that they receive better than MFN
treatment on more stable terms than would be available otherwise (e.g., through the
Generalized System of Preference, GSP), though many developing countries’ exports
remain encumbered under the WTO (e.g., agriculture) and are not covered in PTAs. At
the same time, developing countries commit to undertaking reforms, in both trade and
trade-related realms, beyond what is required by the WTO.
Notwithstanding these asymmetries, North-South PTAs have proliferated in recent
decades. Developing countries have sought PTAs with developed countries for a num-
ber of (not mutually exclusive) reasons. And yet, especially the literature in economics
has long remained attached to a focus on trade in goods as the driver of trade policy
even in developing countries. This contrasts with accounts that, in particular in Latin
America, the initial motivation was to attract foreign direct investment (Cameron
and Tomlin 2000, 59–63). PTAs support this goal in two ways: first, by committing the
developing country to a specific set of trade policies that cannot be overturned without
abrogating the agreement, and second, by facilitating a coalition that can implement
further trade policy reforms. PTAs with big, developed economies are obviously attrac-
tive because these countries remain the most important sources of foreign direct invest-
ment (FDI) and offer large export markets, but North-South PTAs also have political
advantages: The bigger and more powerful the partner, the more capable it is of moni-
toring and policing compliance with the rules of a PTA. This applies in particular to the
European Union (EU) and the United States. Baccini and Urpelainen (2014) provide
rich evidence that developing countries have implemented much “deeper” PTAs with
these two actors, often changing trade policy prior to, during, or shortly after the con-
clusion of negotiations.
Two key issues need to be understood in discussing the proliferation of North-South
PTAs. First, we need to understand the reasons countries want to sign such agreements.
Second, we need to understand the distinct nature of the politics of trade (and in par-
ticular the politics of PTAs) within developing countries.
One reason developing countries seek PTAs is to secure stable and preferential mar-
ket access. One of the cornerstones of the global trade regime regulating North-South
trade has been the GSP, whereby on a unilateral, nonreciprocal basis, developed
countries can offer better market access to developing countries than is granted
under MFN status. The GSP originated in United Nations Conference on Trade and
Development (UNCTAD) demands for such nonreciprocal concessions in the 1960s.
By the mid-1970s most developed economies offered such programs. Since then, several
countries have implemented related programs that make more goods eligible for pref-
erential treatment and higher tariff preference margins. The United States supplements
GSP with the Caribbean Basin Initiative (CBI), the Caribbean Basin Trade Partnership
Act (CBPTA), the Andean Trade Promotion and Drug Eradication Act (ATPDEA),7
and the African Growth and Opportunity Act (AGOA). The EU’s preferential programs
include a GSP+ scheme and the “Everything But Arms” initiative for least developed
countries.
484   Trade and Development

While the GSP itself excludes textiles and apparel, related schemes permit their
export, so that they make up a large share of trade under the GSP+ programs.
Production of light manufactured goods is generally the first step on the ladder toward
industrialization, so that prima facie the GSP and related schemes should be ideal com-
plements to a development-friendly trade policy. Less appreciated is that reliance on
preferential schemes can induce considerable dependence on access to individual mar-
kets, which in turn leads to a phenomenon we have elsewhere referred to as “political
trade dependence” (Manger and Shadlen 2014).8 Although the GSP and related schemes
are preferential and do not require reciprocal trade concessions, they are by no means
nonreciprocal or unconditional. Developed countries retain nearly complete discretion
with regard to the design and implementation of these schemes. Virtually all aspects
of the schemes’ design and implementation—which products are covered, how large
quotas are, what ceilings (if any) are imposed, what procedures are used to waive ceil-
ings, and so on—are determined by the importing country. Eligibility is often explicitly
linked to a range of political criteria. The United States, for example, makes GSP prefer-
ences conditional on the enforcement of labor and environmental standards, improve-
ments in the protection of IP rights, participation in the “War on Drugs” in the form
of drug eradication, and other, often less explicit forms of foreign policy cooperation
(Mason 2004, 524–525). The EU requires the ratification of international human rights
conventions and compliance with a variety of other governance standards as a precondi-
tion for GSP+ and other programs. Both make decisions on compliance unilaterally and
internally, with authority respectively lying with the US Trade Representative and the
European Commission.
Most problematic in terms of economic development is that the establishment of a
competitive, successful export industry is often rewarded with the withdrawal of trade
preferences. In the case of the United States, “competitive need limitations” automati-
cally limit preferences at the request of import-competing industries. In the EU, mem-
ber states can at any time request the reintroduction of MFN tariffs. Because the GSP
and enhanced preferential schemes are unilateral concessions, developing countries
have no recourse if they lose market access in this way. In particular, they cannot access
the WTO dispute settlement procedures, unlike in the case of trade remedies that tem-
porarily suspend MFN.
To avail themselves of the tariff benefits of these schemes in the first place, exporters
often have to comply with onerous documentation requirements. In the case of the EU,
some studies estimate that half of the tariff preferences go unused because of complex
rules of origin (Brenton 2003). Exporters therefore have to make investments that are
specific to the relationship; that is, to the market access to a particular country. Yet they
do so in industries such as apparel where capital is mobile, barriers to entry are low, and
production is quickly redeployed across the globe. Developing countries that use pref-
erential schemes can therefore become particularly dependent on their continued exis-
tence, because it is difficult to sustain any other advantage over competing locations. In
our own research (Manger and Shadlen 2014), we have found evidence that in response,
they often seek fully reciprocal PTAs with the countries whose markets they also have
 Mark S. MANGER AND Kenneth C. SHADLEN    485

access to through the GSP and other programs. Although PTA negotiations are unlikely
to yield further tariff reductions, whatever is implemented in a PTA is at least locked in
and no longer completely at the discretion of the importing country.
The Central American countries provide examples of this dependence. Prior to
the conclusion of the Dominican Republic-Central American Free Trade Agreement
(DR-CAFTA) with the United States, a majority of exports to the United States were
textiles and apparel under preferential schemes. While some import-competing indus-
tries opposed the agreement in the Dominican Republic, well-organized exporters in
the apparel sector provided much of the political support for the country’s accession
(Schrank 2008). As we discuss below, export interests in developing countries are
often  acutely aware of trade policy developments, politically mobilized, and able to
attain their political goals.
Political trade dependence contributes to the larger issue of asymmetrical interde-
pendence in a trade relationship. Simply put, only the largest developing countries have
much leverage in bilateral negotiations to extract tariff concessions from developed
countries, insofar as trade negotiations aim at opening export markets. In this case,
asymmetry is simply a function of market size, or more specifically the size of the econ-
omy, income per capita, and expectations of future growth. While for the United States
not much is lost from walking away from the table when negotiating with a small coun-
try, this is not true for the other side. Consequently, in particular in North-South PTAs,
developing countries have therefore accepted many of the “deep integration” clauses in
return for limited market access gains.
Discussion of how trade structure affects countries’ desire for PTAs needs to be com-
plemented by analysis of domestic politics. Analysis of trade policy since the 1980s is
by and large a story of the dismantling of many of the policies and institutions intro-
duced during the postwar period of state-led industrialization. As indicated, under
the rubric of “reform,” most developing countries followed the prescriptions of the
“Washington Consensus” to tear down barriers. A plethora of research has studied this
process, with particularly insightful contributions by Rodrik (1994), who argues that
trade liberalization is inherently difficult because it results in enormous redistribution
for very limited overall efficiency gains. However, the situation can be different in situ-
ations of profound macroeconomic crisis, such as high levels of inflation, where liber-
alization becomes more broadly perceived as part of the solution to right the economy
(e.g., Pastor and Wise 1994). Trade reform is thus most likely to occur when it takes
place concurrently with economic stabilization programs. This helps explain why such
extensive trade liberalization in developing countries took place in the aftermath of
debt crises.
Moving from the politics of trade per se to the politics of PTAs requires yet another
shift. Most of the literature on trade builds on a standard model in which the likely
beneficiaries of liberalization are expected to be unaware of their potential gains, geo-
graphically diffuse, and unorganized, while those who stand to be adversely affected are
expected to be keenly aware of their likely fate, concentrated, and organized (Destler
1995; Haggard 1998). The result, then, is that trade liberalization according to this model
486   Trade and Development

occurs more or less in spite of politics—it requires mobilization of “silent majorities”

who may be in a position to secure privileges ex post but are unlikely to mobilize to push
for policy change in the first place, or of politicians who hope for the overall economic
growth that trade liberalization promises to generate (Grossman and Helpman 1994).
These are the political underpinnings of a “status quo” bias (Fernandez and Rodrik
1990), and only a few authors have underscored how trade policy is influenced by export
interests (Milner 1988; Dür 2010; Gilligan 1997).
Research on the politics of trade in developing countries reveals different patterns of
political organization and mobilization. Potential beneficiaries are often exceptionally
well-organized, actively mobilized, and politically influential (Solís 2013; Shadlen 2008;
Thacker 2006). For example, exporters organized in associations in export-processing
zones (EPZs) are likely to be keenly and acutely aware of the benefits to be gained from
securing a PTA. At the same time, because trade policy reform (i.e., liberalization, PTA
formation) often occurs in the wake of economic crisis, where import-competing pro-
ducers are already suffering on account of diminished purchasing power in the domestic
market, the potential losers tend to be already weakened and politically disarticulated.
In sum, we witness a distinct model of trade politics in developing countries, where the
actors who stand to benefit directly from liberalization are better positioned politically
than those who stand to lose.
The structure of trade and investment also affect trade politics in profound ways. As
developing countries engage in trade negotiations with the aim to attract FDI, the inter-
ests and lobbying of investing multinational firms change the relationship, and as stages
of manufacturing are located in developing countries, new coalitions emerge that span
the North-South divide. Using Dunning’s (2002) well-established typology, foreign
investors may seek markets, assets, or efficiency. If a developing country is primarily a
market, barriers to FDI matter as much as tariffs. If, on the other hand, FDI aims at effi-
ciency gains—mostly in manufacturing industries—then the endowments of a develop-
ing country, its human capital, and its geographic location may give it leverage in trade
negotiations.
Consider the case of Mexico. Geographic proximity to the United States, long seen
as more of a burden than a boon, made the country attractive for foreign investment
in manufacturing with the goal to serve an integrated North American market. In par-
ticular after the entry into force of NAFTA, Mexico therefore gained leverage in trade
negotiations with the EU and Japan—both under pressure from multinational firms
that sought to establish parity with US firms in the conditions for foreign investment in
Mexico and in the export of intermediate and capital goods to facilitate this (see respec-
tively Dür 2007; Manger 2005)—and obtained market access concessions that went con-
siderably beyond existing GSP and GSP+ preferences.
Furthermore, Mexico is an example of the growing integration of middle-income
developing countries into global manufacturing networks. While the notion of GVCs
is often more metaphor than substantiated by data, the “processing trade” with the
United States in EPZs, in the Mexican case called “maquiladoras,” created substantial
support for trade talks with Mexico in the United States (Cameron 1997; Chase 2003).
 Mark S. MANGER AND Kenneth C. SHADLEN    487

FDI then subsequently led to the production of goods for sale in developed mar-
kets, a process that is called vertical specialization in intra-industry trade (Krugman
1981). Less sophisticated goods are produced in developing countries, while highly
capital-intensive, high-value-added goods are produced in developed economies. In
both countries, however, such manufacturing efforts are characterized by economies of
scale, so that a developed and a developing country can negotiate mutual tariff reduc-
tions by “exchanging” market access concessions for these specialized goods (see Milner
1997, for the original statement; also Manger 2012 for PTAs and vertical specialization).
As the same firms undertake manufacturing FDI and trade, we no longer observe the
traditional split along industry lines predicted by specific-factor models (Alt et  al.
1996). General Motors, its labor force, and its suppliers are as much actors in Mexico’s
political economy as in the United States, much like Panasonic in Japan and Thailand.
A renewed interest in the role of multinational firms in issues of trade and development
could yield plentiful research.
Finally, it is worth noting how the second issue we have discussed here—the domestic
politics of PTAs—reinforces the first—the proliferation of such agreements. While the
first North-South PTA in general requires profound adjustment in the developing coun-
try, the coalition of PTA supporters only has to overcome opposition in initial negotia-
tions. Not only does the existence of trade officials whose mission is to negotiate such
agreements and already-mobilized business organizations create inertia for additional
trade agreements, but subsequent PTAs become easy to negotiate: each additional PTA
may add to market access but not require much more domestic adjustment.

The New International Political

Economy of Trade and Development

While ISI was a more or less universal strategy pursued by developing countries, the
various exit paths of developing countries represent different challenges and attempts
to harness trade for developmental purposes. We close this chapter with a typology of
developing countries that we hope will generate future research that fills the gap between
small-N, within-region comparisons and the broad sweep of large-N studies.
First, small countries with limited market potential never stood much chance of devel-
oping “national” industries primarily serving a domestic market. While the dynam-
ics of change vary across time and space, in general the move toward more open and
export-oriented economic strategies is hardly surprising. Yet smaller countries’ integra-
tion into the global trade regime creates particular development challenges: there sim-
ply are many—perhaps too many—countries whose comparative advantage is low labor
cost. Individually, each country has little influence on the global trade regime and is
generally a rule-taker rather than rule-maker. Asymmetrical interdependence therefore
often leads to trade policy choices that closely approximate the Washington Consensus.
488   Trade and Development

These countries move from first-stage ISI to an export-oriented trade policy based on
foreign direct investment. Yet there are precious few examples of countries (among
them Mauritius) that have managed to “upgrade” technologically, moved up the value
chain, or diversified their exports away from textiles and apparel (Heron 2012). Whether
the causes for this stagnation lie in the lack of domestic capabilities, as in the domain of
traditional development economics, or in international or domestic political economy
remains an unanswered question that we hope future research will address.
Second, the “East Asian” trade policy rested on decisive steps to complement ISI with
the deliberate promotion of exports by key manufacturing industries. This policy was
the more successful, the fewer alternatives there were; Taiwan did better than Malaysia,
perhaps because Malaysia had agricultural products and natural resources to export.
While these states relied to some extent on foreign aid or foreign capital, they gener-
ally did not focus on attracting FDI. And yet the demonstrated role of the state and the
remarkable success in combining rapid economic growth with relatively low income
inequality points to the continuing importance of the East Asian model. Indeed, for
many developing countries, efforts to reinvent industrial policy remain inspired by the
Asian experience.
Third, some countries moved through a second stage of ISI, but with domestic mar-
kets that are big enough to attract market-seeking FDI. Archetypal examples are Brazil,
India, and Turkey. Compared to our first type, they often retain much higher trade bar-
riers (with the exception of Turkey, due to its 1996 customs union with the EU), but have
undertaken considerable efforts at liberalizing their domestic markets. Aside from being
in a less asymmetrical position in trade negotiations, we do not have many accounts for
why these countries are more reluctant to liberalize than smaller economies, especially
if the benefits of increased trade might be considerable.
Finally, we submit that no discussion of trade and development is complete without
special reference to China. Several of our categories map onto Chinese provinces and
Chinese policies in different times. One the one hand, the coastal region successfully
industrialized based on FDI for export, turning whole provinces into the equivalent of
export-processing zones. Simultaneously, the growing sizable Chinese market provides
the Chinese government with enough leverage to require local content quotas, joint ven-
tures, and technology transfer to an extent not seen since the heyday of ISI. The Chinese
state is authoritarian and has been called strong, yet corruption is rampant. None of this
is usefully explained with existing categories of the developmental state literature. We
do not believe that China is a case sui generis—but comparisons need to dig deeper than
is possible in large-N studies.
Neither is it the case that research in general has ignored these questions. In particu-
lar Rodrik’s work (as a sample, consider Rodrik 1994, 2006; Hausmann, Hwang, and
Rodrik 2007) offers rich inspiration but has not been taken up sufficiently by political
economy scholars. Political economy has an inherent tendency to endogenize all forces,
with the result that accounts appear to describe the inevitable—and yet choices clearly
matter for development outcomes.
 Mark S. MANGER AND Kenneth C. SHADLEN    489

Notes
1. It is worth noting that in many countries ISI began not as a policy choice but as a result of
supply shortages during the 1940s, as industrial plants in developed countries were ori-
ented toward war efforts.
2. These requirements were often imposed informally, in terms of inspectors or tax authori-
ties treating foreign investors differently depending on their relationships with local
suppliers.
3. Note that industrialization in South Korea, the archetypal East Asian “export-oriented”
state, also relied heavily on financing from commercial lenders in the 1970s (Frieden 1981),
but because of a greater ability to generate revenues through manufactured exports, Korea
could borrow on fixed rates. The result is that the substantial increase in international
interest rates of the early 1980s did not translate into higher debt-servicing obligations.
4. The GATT did include additional agreements that expanded beyond tariffs, particularly
following the Tokyo Round of the 1970s, but most of these were in the form of optional
“codes.”
5. For a contrarian view, which depicts the WTO as placing severe limits on policy choices
and thus essentially freezing the international division of labor, see Wade (2003).
6. Though not the focal point of this chapter, it is worth underscoring here that the promi-
nence of deep integration in PTAs means that conventional metrics used to assess the rela-
tive costs and benefits of such agreements may be misleading. That is, most analyses aim to
estimate the relative differences between trade creation and diversion, yet the flow of goods
and services may be just a part (and a small part) of what matters in these agreements.
7. The ATPDEA replaced the Andean Trade Preference Act (ATPA) in 2002.
8. The broader point that the manipulation of asymmetrical trading relations can function as
an instrument of coercion was anticipated in Hirschman (1945).

References
Alt, James E., Jeffry A. Frieden, Michael Gilligan, Dani Rodrik, and Ronald Rogowski. 1996.
The Political Economy of International Trade: Enduring Puzzles and an Agenda for Inquiry.
Comparative Political Studies 29 (6): 689–717.
Amsden, Alice H. 1989. Asia’s Next Giant: South Korea and Late Industrialization. New York:
Oxford University Press.
Amsden, Alice H. 2001. The Rise of “the Rest”: Challenges to the West from Late-Industrializing
Economies. New York: Oxford University Press.
Amsden, Alice H., and Takashi Hikino. 2000. The Bark Is Worse Than the Bite: New WTO Law
and Late Industrialization. Annals of the American Academy of Political and Social Science
570 (4): 104–114. doi: 10.2307/1049243
Baccini, Leonardo, and Johannes Urpelainen. 2014. International Institutions and Domestic
Politics: Can Preferential Trading Agreements Help Leaders Promote Economic Reform?
Journal of Politics 76 (1): 195–214. doi: 10.1017/S0022381613001278
Brenton, Paul. 2003. Integrating the Least Developed Countries into the World Trading
System: The Current Impact of European Union Preferences under “Everything but Arms”.
Journal of World Trade 37 (3): 623–646.
490   Trade and Development

Cameron, Maxwell A. 1997. North American Trade Negotiations:  Liberalization Games

Between Asymmetric Players. European Journal of International Relations 3 (1): 105–139.
Cameron, Maxwell A., and Brian W. Tomlin. 2000. The Making of NAFTA: How the Deal Was
Done. Ithaca, NY: Cornell University Press.
Chase, Kerry A. 2003. Economic Interests and Regional Trading Arrangements: The Case of
NAFTA. International Organization 57 (1): 137–174.
Chenery, Hollis B. 1958. The Role of Industrialization in Development Programmes. In The
Economics of Underdevelopment, edited by A. N. Agarwala and S. P. Singh, 450–471.
Oxford: Oxford University Press.
Copelovitch, Mark S., and David Ohls. 2012. Trade, Institutions, and the Timing of GATT/
WTO Accession in Post-Colonial States Review of International Organizations 7 (1): 81–107.
doi: 10.1007/s11558-011-9129-2
Destler, I. M. 1995. American Trade Politics. 3rd ed. Washington, DC: Peterson Institute.
Diaz-Alejandro, Carlos F. 1965. On the Import Intensity of Import Substitution. Kyklos 18
(3): 495–511. doi: 10.1111/j.1467-6435.1965.tb00986.x
Doner, Richard F., Bryan K. Ritchie, and Dan Slater. 2005. Systemic Vulnerability and the
Origins of Developmental States: Northeast and Southeast Asia in Comparative Perspective.
International Organization 59 (02): 327–361.
Dunning, John H. 2002. Theories and Paradigms of International Business Activity.
Edited by John H. Dunning. The Selected Essays of John H. Dunning vol. 1. Cheltenham,
UK: Edward Elgar.
Dür, Andreas. 2007. EU Trade Policy as Protection for Exporters: The Agreements with Mexico
and Chile. Journal of Common Market Studies 45 (4): 833–855.
Dür, Andreas. 2010. Protection for Exporters: Power and Discrimination in Transatlantic Trade
Relations, 1930–2010. Ithaca, NY: Cornell University Press.
Fernandez, Raquel, and Dani Rodrik. 1990. Why Is Trade Reform so Unpopular? On Status
Quo Bias in Policy Reforms. NBER Working Paper No. 3269. Cambridge, MA: National
Bureau for Economic Research. http://www.nber.org/papers/w3269.
Fernández de Córdoba, Santiago, and Sam Laird. 2006. Coping with Trade Reforms:
A Developing Country Perspective on the WTO Industrial Tariff Negotiations. Houndmills,
Basingstoke, UK: Palgrave Macmillan.
Frieden, Jeffry A. 1981. Third World Indebted Industrialization:  International Finance and
State Capitalism in Mexico, Brazil, Algeria, and South Korea. International Organization 35
(3): 407–431. doi: 10.2307/2706430
Gereffi, Gary, and Donald L. Wyman. 1990. Manufacturing Miracles: Paths of Industrialization
in Latin America and East Asia. Princeton, NJ: Princeton University Press.
Gerschenkron, Alexander. 1962. Economic Backwardness in Historical Perspective. Cambridge,
MA: Belknap Press.
Gilligan, Michael J. 1997. Empowering Exporters: Reciprocity, Delegation, and Collective Action
in American Trade Policy. Ann Arbor: University of Michigan Press.
Grossman, Gene M., and Elhanan Helpman. 1994. Protection for Sale. American Economic
Review 84 (4): 833–850.
Haggard, Stephan. 1990. Pathways from the Periphery:  The Politics of Growth in the Newly
Industrializing Countries. Ithaca, NY: Cornell University Press.
Haggard, Stephan. 1998. The Political Economy of Regionalism. In The Post-NAFTA Political
Economy: Mexico and the Western Hemisphere, edited by Carol Wise, 302–338. University
Park: University of Pennsylvania Press.
 Mark S. MANGER AND Kenneth C. SHADLEN    491

Hausmann, Ricardo, Jason Hwang, and Dani Rodrik. 2007. What You Export Matters. Journal
of Economic Growth 12 (1): 1–25.
Heron, Tony. 2012. The Global Political Economy of Trade Protectionism and Liberalization: Trade
Reform and Economic Adjustment in Textiles and Clothing. Milton Park, Abingdon, Oxford
(UK): Routledge.
Hirschman, Albert O. 1945. National Power and the Structure of Foreign Trade. Expanded
ed. The Politics of the International Economy. Berkeley and Los Angeles:  University of
California Press.
Hirschman, Albert O. 1968. The Political Economy of Import-Substituting Industrialization in
Latin America. Quarterly Journal of Economics 82 (1): 1–32. doi: 10.2307/1882243
Kay, Cristóbal. 2002. Why East Asia Overtook Latin America:  Agrarian Reform,
Industrialisation and Development. Third World Quarterly 23 (6): 1073–1102. doi: 10.1080/0
143659022000036649
Krugman, Paul R. 1981. Intraindustry Specialization and the Gains from Trade. Journal of
Political Economy 89 (5): 959–973.
Lawrence, Robert Z. 1996. Regionalism, Multilateralism, and Deeper Integration. Integrating
National Economies. Washington, DC: Brookings Institution.
Lewis, W. Arthur. 1954. Economic Development with Unlimited Supplies of Labour.
Manchester School 22 (2): 139–191. doi: 10.1111/j.1467-9957.1954.tb00021.x
Mahon, James E., Jr. 1992. Was Latin America Too Rich to Prosper? Structural and Political
Obstacles to Export-led Industrial Growth. Journal of Development Studies 28 (2): 241–263.
doi: 10.1080/00220389208422231
Manger, Mark S. 2005. Competition and Bilateralism in Trade Policy:  The Case of Japan’s
Free Trade Agreements. Review of International Political Economy 12 (5):  804–828.
doi: 10.1080/09692290500339800
Manger, Mark S. 2012. Vertical Trade Specialization and the Formation of North-South PTAs.
World Politics 64 (4): 622–658.
Manger, Mark S., and Kenneth C. Shadlen. 2014. Political Trade Dependence and North-South
Trade Agreements. International Studies Quarterly 58 (1): 79-91. doi: 10.1111/isqu.12048
Mason, Amy M. 2004. The Degeneralization of the Generalized System of Preferences
(GSP): Questioning the Legitimacy of the US GSP. Duke Law Journal 54 (2): 513–547.
Maxfield, Sylvia, and James H. Nolt. 1990. Protectionism and the Internationalization of
Capital: U.S. Sponsorship of Import Substitution Industrialization in the Philippines, Turkey
and Argentina. International Studies Quarterly 34 (1): 49–81. doi: 10.2307/2600405
Milner, Helen V. 1997. Industries, Governments, and the Creation of Regional Trade Blocs. In
The Political Economy of Regionalism, edited by Edward D. Mansfield and Helen V. Milner,
77–106. New York: Columbia University Press.
Milner, Helen V. 1988. Resisting Protectionism: Global Industries and the Politics of International
Trade. Princeton, NJ: Princeton University Press.
Pastor, Manuel, and Carol Wise. 1994. “The Origins and Sustainability of Mexico Free-Trade
Policy.” International Organization 48 (3): 459–489.
Prebisch, Raúl. 1949. O Desenvolvimento Econômico Da América Latina E Seus Principais
Problemas. Revista Brasileira de Economia 3 (3): 47–1111.
Robinson, James A. 2011. Industrial Policy and Development: A Political Economy Perspective.
In Annual World Bank Conference on Development Economics Global 2010: Lessons from East
Asia and the Global Financial Crisis, edited by Justin Yifu Lin and Boris Pleskovic, 61–79.
Washington, DC: The World Bank.
492   Trade and Development

Rodrik, Dani. 1994. The Rush to Free Trade in the Developing World:  Why So Late? Why
Now? Will It Last? In Voting for Reform: Democracy, Political Liberalization, and Economic
Adjustment, edited by Steven B. Webb and Stephan Haggard, 61–88. New  York:  Oxford
University Press.
Rodrik, Dani. 2006. What’s So Special about China’s Exports? China & World Economy 14
(5): 1–19.
Rudra, Nita. 2002. Globalization and the Decline of the Welfare State in Less-Developed
Countries. International Organization 56 (2): 411–445.
Rudra, Nita. 2004. Openness, Welfare Spending, and Inequality in the Developing World.
International Studies Quarterly 48 (3): 683–709. doi: 10.1111/j.0020-8833.2004.00320.x
Schrank, Andrew. 2007. Asian Industrialization in Latin American Perspective:  The Limits
to Institutional Analysis. Latin American Politics & Society 49 (2):  183–200. doi:  10.1353/
lap.2007.0025
Schrank, Andrew. 2008. Export Processing Zones in the Dominican Republic:  Schools or
Stopgaps? World Development 36 (8): 1381–1397.
Shadlen, Kenneth C. 2005. Exchanging Development for Market Access? Deep Integration and
Industrial Policy under Multilateral and Regional-Bilateral Trade Agreements. Review of
International Political Economy 12 (5): 750–775.
Shadlen, Kenneth C. 2008. Globalisation, Power and Integration: The Political Economy of
Regional and Bilateral Trade Agreements in the Americas. Journal of Development Studies
44 (1): 1–20.
Singer, Hans W. 1950. The Distribution of Gains between Investing and Borrowing Countries.
American Economic Review 40 (2): 473–485. doi: 10.2307/1818065
Solís, Mireya. 2013. Business Advocacy in Asian PTAs:  A  Model of Selective Corporate
Lobbying with Evidence from Japan. Business and Politics 15 (1):  87–116. doi:  10.1515/
bap-2012-0045
Spraos, John. 1980. The Statistical Debate on the Net Barter Terms of Trade Between Primary
Commodities and Manufactures. Economic Journal 90 (357): 107–128. doi: 10.2307/2231659
Stallings, Barbara. 1990. The Role of Foreign Capital in Economic Development. In
Manufacturing Miracles: Paths of Industrialization in Latin America and East Asia, edited by
Gary Gereffi and Donald L. Wyman, 55–89. Princeton, NJ: Princeton University Press.
Thacker, Strom C. 2006. Big Business, the State, and Free Trade:  Constructing Coalitions in
Mexico. Cambridge, UK: Cambridge University Press.
Wade, Robert H. 2003. What Strategies Are Viable for Developing Countries Today? The World
Trade Organization and the Shrinking of “Development Space”. Review of International
Political Economy 10 (4): 621–644.
Wong, Joseph. 2004. The Adaptive Developmental State in East Asia. Journal of East Asian
Studies 4 (3): 345–362.
Woo-Cumings, Meredith. 1999. The Developmental State. Ithaca, NY: Cornell University Press.
Woods, Ngaire. 2007. The Globalizers: The IMF, the World Bank, and Their Borrowers. Ithaca,
NY: Cornell University Press.
World Bank. 2013. World Development Indicators. Washington, DC: The World Bank.
 Chapter 26

A Match M ade i n H e av e n?
T he Wedding of T ra de
a nd Hum an Ri g h ts

Susan Arie l Aaron son

As long and men and women have traded, they have wrestled with questions of human
rights. Not surprisingly, throughout history policy makers have tried to marry trade and
human rights policies throughout history. For example, archaeological evidence shows that
ancient civilizations traded at great risk to their freedoms. According to economist Peter
Temin (2003), the ancients shipped a wide range of goods, from wheat to wine. But these
traders often lived in fear; when they engaged in trade they risked being captured, sold as
slaves, or enslaved by pirates. The sea may bring contact with strangers who could enhance
national prosperity, but these same strangers might threaten the security of the nation and
its people.
Many years later, during the Age of Exploration, theologians, scholars, and royal
advisors debated whether they had the right to exploit the land and wealth of indig-
enous populations. The economic historian Douglas Irwin notes that Vitoria, one of the
“founders of international law,” contended that the right to trade is “derived from the law
of nations… . Foreigners may carry on trade, provided they do no hurt to citizens.”1
The truth is that although scholars, policy makers, and activists have long debated the
relationship between trade and human rights, scholars today still know very little about that
relationship or about causality. Do trade and human rights make an effective and enduring
match? Figure 26.1 shows the big questions that scholars should be asking about the relationship
between trade and human rights. This chapter discusses whether scholars have asked and
answered these questions and whether they have examined why countries are demanding and
accepting such human rights provisions. It also considers whether scholars have examined the
“so what” question: Have these provisions led to changes in human rights policies, conditions,
and the ability of citizens to demand their rights?
The chapter is organized in three sections. The first defines human rights and dis-
cusses human rights language within existing trade agreements. In the second, I dis-
cuss the state of scholarship on the relationship between trade and human rights. Many
494    A Match Made in Heaven?

3. Does the protection of

1. Does expanded trade 2. Does expanded trade
specific human rights make
stimulate or undermine encourage or discourage
trade liberalization or specific
government responsibility for public demand for specific
trade agreements possible (e.g.,
respecting human rights? human rights?
property rights protection)?

5. Do governments use human

4. Do governments use trade
rights conditions to justify
agreements to promote a
trade agreements or
particular human right,
relationships
and is that strategy
(or nonrelationships)?
effective (conditionality)?
Should they be linked?

Fig. 26.1  The Big Questions in the Complex Relationship between Trade and Human Rights.

scholars have used empirical tools to examine the trade and human rights relationship,
but have paid little attention to what trade agreements actually say about human rights.
Moreover, as I will show, when scholars use empirical tools to examine the effects of
trade on human rights, they tend to focus on one grouping of human rights: personal
integrity. Personal integrity rights include freedom from physical abuse by authorities,
such as torture, arbitrary imprisonment, or rape. I argue that although this scholarship
has reached a critical mass in scale and scope, academics may not be asking the right
questions.2 Scholars have asked whether trade leads to human rights improvements,
whether governments use trade to promote specific human rights, and whether policy
makers use human rights conditions to justify trade agreements. But they are not ask-
ing if expanding trade empowers people to demand their rights. Nor are they asking
whether government respect for specific human rights (an aspect of good governance)
might be a precondition for trade liberalization; for example, establishing the protec-
tion of property rights. From my vantage point, while past studies have provided new
insights, they often seem disconnected from the real world where policy makers are
increasingly linking trade and specific human rights (although these officials are not
linking trade and personal integrity rights).
The third section offers some conclusions as to why governments are increasingly
linking trade and human rights, whether such links are effective, and whether these pol-
icy unions will thrive. I then make suggestions for further research.

The Rights and Responsibilities

of States and What They Mean
for Trade Policy

Under international law, states are supposed to do everything in their power to respect,
promote, and fulfill human rights. But advancing human rights is not easy. As noted pre-
viously, many states are unable or unwilling to meet all of their “international” human
 Susan Ariel Aaronson    495

rights responsibilities. Moreover, few officials win or maintain office on the basis of their
efforts to promote the human rights of non-citizens.
Not surprisingly, most policy makers do not make human rights a top priority for
trade policy making. In most countries, they develop trade policies as if these poli-
cies were strictly commercial policies. These officials weigh the interests of their pro-
ducers and consumers, and although they may consider national security or political
concerns, they rarely introduce the interests of the global community into such delib-
erations. Although policy makers are well aware of the human rights consequences
of some of their trade decisions, they have few incentives to ensure that trade policies
advance the human rights outlined under the Universal Declaration of Human Rights.
Moreover, trade policy makers are generally not charged with ensuring that trade pol-
icies or flows do not undermine specific human rights at home or abroad. In trade
negotiations, governments are charged with pursuing national commercial interests,
not global interests (Commission on Human Rights 2004; 3D 2005; Aaronson and
Zimmerman 2007).
However, many people are not comfortable knowing that other human beings lack
basic rights in their own countries or live in countries where government officials
undermine human rights. These individuals may demand that policy makers take
action to protect human rights in other countries. Trade policy is not the best means of
extending such protection, but policy makers have few options short of military force
for changing the behavior of leaders in other countries. Hence, when a country such as
the United States wants to affect human rights in another country such as Syria (not a
member of the World Trade Organization [WTO]), it can use access to its markets as
leverage.

A Brief Overview of the

Rules Governing Trade
and Its Intersection

Policy makers first began to regulate state behavior regarding trade and human rights
in the nineteenth century. In some instances they used the incentive of trade expan-
sion; at other times they used trade sanctions—the disincentive of lost trade—to
punish officials from other countries that undermined human rights. For example,
after the United Kingdom and the United States outlawed the slave trade in 1807,
the United Kingdom signed treaties with Denmark, Portugal, and Sweden to rein-
force its own ban. When the United States banned goods manufactured by convict
labor, Australia, Canada, and the United Kingdom followed with similar measures.
These efforts stimulated international cooperation, and in 1919 the signatories of the
Treaty of Versailles formed the International Labour Organization (ILO) to establish
fair and humane rules regarding the treatment of labor producing goods for trade
(Aaronson 2001).
496    A Match Made in Heaven?

Although policy makers had long used trade policy to affect human rights conditions
abroad, it was not until the twentieth century that trade diplomats first negotiated spe-
cific human rights language within trade agreements. In the 1980s and 1990s, US and
European Union (EU) officials began to include human rights conditionality clauses in
their preference programs, which were justified as a waiver to WTO rules (Charnovitz
2005, 29, n. 103–105). The 1993 North American Free Trade Agreement (NAFTA),
signed by Canada, Mexico, and the United States, included language on labor rights pro-
visions in a side agreement. More than 131 countries (some 70 percent of all countries)
participated in a trade agreement containing human rights language or requirements
(Aaronson 2011, 443). Moreover, as Table 26.1 illustrates, trade agreements today cover a
diverse range of human rights.
While some of these human rights provisions are aspirational and simply expressed in
hortatory language in a preamble, other human rights are delineated in actionable pro-
visions in the body of a trade agreement. As Table 26.2 show, some governments even
make their human rights commitments disputable—in other words, they may challenge
a country that fails to live up to the obligations in a trade dispute. Finally, middle-income
and developing-country governments such as Chile, Brazil, and Uruguay have included
human rights provisions in their free trade agreements (FTAs).
Some analysts see these provisions as “legal inflation” and assert that governments are
using trade agreements to globalize their social policies or regulatory approaches. Many
argue that trade agreements are not the right place to address human rights issues, and
they point out that trade agreements do not need to have explicit human rights language
to have positive human rights spillovers.
But others argue that trade policy makers pay insufficient attention to human rights,
which are necessary to achieve sustainable development (Sen 1999). Sen asserts that
governments must actively provide citizens with public goods such as education, prop-
erty rights, credit, and justice; these public goods allow citizens to meet their potential
and in turn to act in ways that grow the economy (UNDP 2000).
Trade and human rights is not only a controversial development issue; even Western
policy makers disagree about how to protect human rights and what human rights
should be discussed in trade agreements. In July 2013 the twenty-eight nations of the
EU and the United States began negotiating a free trade agreement. These twenty-nine
nations include the world’s most technologically advanced and richest democracies,
which have made steady progress on human rights. Yet stakeholders on both sides of
the Atlantic have expressed concern that the potential trade agreement could actually
undermine human rights and are worried that the agreement has no chapter on human
rights.3 Moreover, the negotiations were colored by revelations that the United States
had not adequately protected personal online data that it had gathered. Many EU politi-
cians complained that EU privacy rights were not respected by the United States; some
officials even threatened to cancel the trade talks to illustrate how important privacy
rights are to the EU populace.4
Table 26.1  Human Rights Embedded in Free Trade Agreements and the Countries Embedding Them
Access to Political
Democratic Affordable Right to Cultural Freedom of Indigenous Participation and
Labor Rights Rights Medicines Participation Movement Minority Rights Due Process Privacy

Canada The Members of Costa Rica Canada The Members of Canada Canada Canada
Mercosur EFTA/EEA
Chile United States New Zealand The Members of New Zealand EU
CARICOM

Note: CARICOM = Caribbean Community; EEA = European Economic Area; EFTA = European Free Trade Agreement; EU = European Union;
Mercosur = Southern Cone Common Market (Mercado Común del Sur).
498    A Match Made in Heaven?

Table 26.2  Aspirational vs. Binding Language in Free Trade

Agreements: Comparing the EFTA, the EU, the United States, and Canada
EFTA EU United States Canada

Strategy Universal Universal human Specific human Specific human

human rights rights and specific rights rights
rights
Which rights? Labor rights, Transparency, due Transparency, due
transparency, due process, political process, political
process, political participation, participation, labor
participation, and access to affordable rights, privacy
privacy rights medicines, and labor rights, cultural and
rights indigenous rights
How enforced? No Human rights Since 2007 (e.g., in Only labor rights
enforcement violations lead Colombia), labor rights (monetary
to dialogue and can be disputed under penalties). Use
possible suspension dispute settlement dialogue first.
depending on body affiliated with
nature of violation. the agreement.
Process begins with
bilateral dialogue to
resolve issues.
Any challenge? First
challenge: Guatemala
Review of Jordan

Source:  Aaronson (2011).

Note: EEA = European Economic Area; EFTA = European Free Trade Agreement; EU = European Union.

GATT and WTO Provisions

on Human Rights

From 1948 to 1995 the General Agreement on Tariffs and Trade (GATT) governed mul-
tilateral trade relations. GATT grew from 23 nations in 1948 to 128 in 1995. In 1995 the
members of the GATT agreed to join a new international organization, the WTO, which
contains the GATT agreement and has a stronger system of dispute settlement. Today,
the 160 GATT/WTO members must adhere to two key principles: most favored nation
(MFN) principle, and national treatment principle. The MFN principle requires that the
best trading conditions extended to one member by a nation must be extended auto-
matically to every other nation. The national treatment principle provides that once a
product is imported, the importing state may not subject that product to regulations
less favorable than those that apply to like products produced domestically. The GATT/
WTO does not explicitly prohibit countries from protecting human rights at home
 Susan Ariel Aaronson    499

or abroad, but its rules do constrain the behavior of governments. The WTO (and the
GATT) require that when member states seek to promote human rights at home or
abroad, they must not unnecessarily or unduly distort trade by discriminating among
producers and products as noted above. Hence, member states are not supposed to use
trade to distinguish among those nations that may undermine the human rights of their
citizens and those that strive to advance these rights.
The GATT/WTO do not directly address how governments relate to their own citi-
zens, and they say very little about human rights. But human rights are seeping into the
workings of the WTO (Aaronson 2007). Some WTO members have used the GATT/
WTO exceptions to advance human rights abroad or to protect human rights at home.
Under Article XX, nations can restrict trade when necessary to “protect human, animal,
or plant life or health” or to conserve exhaustible natural resources. This article also states
that governments may restrict imports relating to the products of prison labor. Although
it does not refer explicitly to human rights, the public morals clause of Article XX is
widely seen as allowing GATT/WTO members to put in place trade bans in the interest
of promoting human rights at home or abroad ( Howse 2002; Charnovitz 2005).
The national security exception, Article XXI, states that WTO rules should not pre-
vent nations from protecting their own security. Members are not permitted to take
trade action to protect another member’s security or to protect the citizens of another
member. If, however, the United Nations Security Council authorizes trade sanctions
because of human rights violations, WTO rules allow countries to comply with such
authorizations, such as when sanctions were instituted against South Africa’s apartheid
regime in the 1980s (Levy 1999; Aaronson and Zimmerman 2007, 19).
Members of the GATT/WTO can use other avenues to protect human rights at home
and within other member states (see Table 26.3). In recent years member states have
used temporary waivers of GATT rules to promote human rights. For example, after
the United Nations (UN) called for a ban on trade in conflict diamonds, WTO member
states agreed to temporarily waive WTO rules to allow trade in only those diamonds
certified by the Kimberly process to be free of conflict (Aaronson and Zimmerman
2007, 43). In addition, some members bring up human rights during accessions, when
new members are asked to make their trade and other public policies transparent,
accountable, and responsive. They have also discussed human rights issues at trade
policy reviews, when member states review the trade and governance performance of
other member states. Finally, WTO members have discussed some human rights issues
during recent trade negotiations; examples include food security, intellectual property
rights (IPRs), and public health (Aaronson 2007; Aaronson and Abouharb 2011).
Although the GATT/WTO contains no explicit human rights provisions, it does refer
to human rights implicitly. Some of these provisions relate to economic rights, such as
the right to property, while others refer to democratic and political rights. For exam-
ple, under GATT/WTO rules, member states give economic actors “an entitlement to
substantive rights in domestic law including the right to seek relief; the right to submit
comments to a national agency, or the right to appeal adjudicatory rulings” (Charnovitz
2001). Member states must also ensure that “[m]‌embers and other persons affected, or
500    A Match Made in Heaven?

Table 26.3  Examples of Avenues and Actions at the WTO Related to Human

Rights, 2003–2013
Avenue Human Right Affected

Accessions Labor rights, access to information, due process (Vietnam, Saudi Arabia,
Cambodia) 2003–present
Trade policy reviews Members discussed labor rights, women’s rights, access to medicines
2003–present
Disputes Right to health (Brazil tires) 2005–2009
Public morals (Antigua gambling) 2003–2013
Freedom of speech, access to information (China-Audiovisual Services case)
WTO, 2007–2010
Negotiations Access to safe, affordable food 2003–present.
Access to medicines, 2003–present

Source:  Aaronson (2007), updated.

likely to be affected, by governmental measures imposing restraints, requirements and

other burdens, should have a reasonable opportunity to acquire authentic information
about such measures and accordingly to protect and adjust their activities or alterna-
tively to seek modification of such measures.”5 These rules can be rephrased as promot-
ing due process and access to information, or more generally as “political participation
rights.”
Although these provisions in the GATT/WTO were not directly intended to
advance human rights per se, these provisions could have important human rights
spillovers (indirect effects). Under WTO rules, member states must apply the same
rules and privileges to domestic and foreign actors. These provisions may prod pol-
icy makers to provide access to information and enforce rights to public comment in
countries where governance is not transparent and participatory. In repressive states,
WTO rules may empower domestic market actors (consumers and taxpayers, as well
as producers) who may not have been able to use existing domestic remedies to obtain
information, influence policies, or challenge their leaders (Aaronson and Abouharb
2011). In WTO countries without a strong democratic tradition, member states may
make these changes because they want to signal investors that they can be trusted to
enforce property rights; uphold the rule of law; and act in an evenhanded, impartial
manner (Dobbin, Simmons, and Garrett 2007; Büthe and Milner 2008; Mansfield and
Pevehouse 2008, 273).
In sum, the WTO regime is not hostile to human rights, and human rights discus-
sions are clearly part of the trade agenda. Trade policy makers have become more will-
ing to accommodate human rights as the public became more aware of their human
rights and more demanding of policies that advance and protect those rights.
 Susan Ariel Aaronson    501

An Overview of the Literature

on Trade and Human Rights

At first glance, this might be the golden age of scholarship on the marriage of trade and
human rights. Economists, political scientists, philosophers, and law professors have
joined the discussion and at times have transcended traditional academic boundar-
ies. Moreover, they have examined the trade and human rights relationship from many
angles with a variety of tools, from empirical analysis to case studies.
Some academics have focused on the long run, arguing that trade is a means to the
end of improving human welfare. They note that when trade contributes to economic
activity, more people can improve their quality of life (Dollar 1992; Sachs and Warner
1995). As trade expands, individuals exchange ideas, technologies, processes, and cul-
tural norms and goods. With more trade, people in countries with fewer rights and
freedoms become aware of conditions elsewhere, and with such knowledge they may
demand greater rights (van Hees 2004). Still others argue that trade stimulates an
export-oriented middle class, which will use its increasing economic clout to demand
political freedoms and press for openness and good governance (Rodrik: 2000). Blanton
and Blanton conclude that there is a “mutually reinforcing relationship between trade
openness and human rights. Trade openness contributes to societal changes conduc-
tive to respect for human rights … conversely, respect for human rights enables states to
better compete in the global trading order” (2008, 93). Some scholars directly assert that
trade may help to encourage guarantees of civil and political liberties. In this regard, law
professor Joel Paul cites the example of Mexico, where greater trade with Canada and
the United States helped the country mature into a multiparty democracy (Paul: 2003).
Thus, many of these scholars conclude that policy makers need not include human
rights provisions in trade agreements (Bhagwati 1996, 1; Sykes 2003, 2–4). However,
other analysts disagree; they believe that human rights are trade issues and cite history
and the increasing number of human rights provisions in preferential trade agreements
(PTAs) as evidence for their perspective.
However, others have a less benign view of the marriage of trade and human rights.
As an example, some gender rights scholars argue that because the WTO restricts the
ability of member states to adopt protectionist policies, these states are less able to pro-
tect their citizens, especially women, from harm. They also believe that because many
women are small farmers who can’t compete effectively with farmers from middle
income and industrialized nations, they are effectively penalized by the WTO (Seguino
and Grown 2006). Moreover, women often find it harder to compete with men for
skilled jobs created by trade (Korinek 2005, 11).
But trade and trade agreements can also increase access to resources and opportuni-
ties for women. In both developing and industrialized states, trade-related globalization
may provide more and better jobs for some citizens, but job loss for others—both women
502    A Match Made in Heaven?

and men. Scholars don’t yet know if job loss can be attributed solely to globalization or if
unemployment and underemployment are caused by other factors such as technological
progress.6 Whatever the reason, given demographic changes, many countries will need
to find strategies to encourage lasting female participation in the workforce.7 Hence,
countries with rapidly aging populations and relatively low female workplace participa-
tion may benefit by gaining a broader understanding of how to make their labor, gender,
and trade policies more coherent (Jansen and Lee 2007, 90).
Although growing numbers of scholars are examining gender and trade and other
topics, the bulk of scholarship has focused on one subset of human rights: physical integ-
rity rights such as the right to life and the inviolability of the human person. President
Franklin Roosevelt first enunciated this concept in his “Four Freedoms” speech (1941)
when he mentioned freedom from fear. Although Roosevelt was referring directly to
freedom from the fear caused by war and violence, personal integrity rights are a sub-
set of human rights that include freedom from rape, torture, and arbitrary imprison-
ment. Apodaca (2001, 597–598) and Harrelson-Stephens and Callaway (2003, 151–152)
argue that increased trade enhances government respect for physical integrity rights,
even after substituting country exports for trade openness. Harrelson-Stephens and
Callaway conclude that “increases in both imports and exports have a substantial impact
on human rights performances of a state” (2003, 152). Hafner-Burton and Tsutsui (2007)
and Hafner-Burton (2005) find that trade and/or foreign direct investment is positively
related to stronger respect for physical integrity rights. Hafner-Burton (2009, 160–164)
finds that about 82 percent of the countries that have a PTA with the EU improve their
human rights protection, as opposed to 75 percent of countries without a trade agree-
ment. Wesley Milner also finds that trade openness enhances physical integrity rights
(2002, 90–91). In an intriguing study, Brian Greenhill finds that membership in an
international organization such as a trade agreement may “socialize” policy makers and
encourage them to respect human rights (Greenhill 2010).
Political scientists Cao, Greenhill, and Prakash (2013) sought to test whether trade
can improve human rights standards. They hypothesized that if a country trades with
another country that has strong human rights practices, it may be encouraged to
adopt stronger standards itself. They also posited that there may be a tipping point at
which outside pressure (from business, governments, nongovernmental organizations
[NGOs], trade unions, and consumer groups) can prod a government with a relatively
poor human rights record to improve its performance. Although many of their qualita-
tive examples relate to labor rights (2013, 135–141), Cao, Greenhill, and Prakash tested
their hypothesis using physical integrity rights. They conclude that “trading relation-
ships can, under certain conditions, serve as ‘transmission belts’ for the diffusion of
human rights standards,” but only at a certain tipping point. They note that trade not
only connects firms, but also “connects the practices of importing jurisdictions to those
of the governments of exporting states.” Therefore, they conclude that scholars should
pay greater attention to such norm diffusion, and activists should see trade as a tool
to help developing countries improve governance (Cao, Greenhill, and Prakash 2013,
154–155). This phenomenon was evident in Bangladesh in 2013. After a series of fires
 Susan Ariel Aaronson    503

and a building collapse killed over a thousand garment workers who toiled in unsafe
conditions, activists pressed the US government to suspend trade preferences and sev-
eral companies organized to do a better job monitoring work conditions among their
suppliers.8
Although these studies have helped us better understand the marriage of trade and
human rights, many of these analyses have two important flaws. First, they conflate
physical integrity rights with human rights in general. As a result, they assert broader
human rights effects than they are actually finding. Hafner-Burton and Ron (2009,
365–366) essentially argue that scholars have little choice; using personal integrity
rights makes sense because they allow scholars to study “the state’s actual treatment
of its population … Personal integrity rights are vital … [because] states are obliged
under international treaties to immediately cease and desist from their abuse.” Second,
scholars who rely on personal integrity data don’t explain how trade or trade agreements
affect physical integrity rights directly or indirectly. We can find no trade agreements
that mention physical integrity rights, so scholars need to map out how, when, and why
trade agreements and policies affect personal integrity rights. To put it differently, how
does enhanced trade or a specific trade agreement affect whether or when states torture,
rape, or use arbitrary imprisonment? Until scholars can show the mechanism of trans-
mission, it seems strange that they are examining personal integrity rights to illuminate
how trade affects human rights, rather than other human rights directly affected by or
mentioned in trade agreements, such as labor, freedom of speech, access to information,
or due process.
Scholars can increasingly choose from other data sets.9 CIRI, as an example, has data
on women’s rights, political and civil liberties, and labor rights for the bulk of the world’s
countries.10 The Social and Economic Empowerment Rights Initiative (SERF) has two
years of useful data examining country performance on the fulfillment of economic
and social rights obligations.11 Global Integrity, an NGO that works on anticorruption
counterweights, prepares an annual survey that looks at various democratic rights in
principle and in practice, including access to information and freedom of speech and
assembly. The information is comparable and has been available since 2004, but not
for every country in every year.12 Finally, the World Bank has good data on property
rights,13 while UNCTAD has the Human Development Index and the Multidimensional
Poverty Index, which examine the many dimensions of how people experience develop-
ment and poverty, respectively.14
Some scholars are trying to examine human rights that are directly affected by trade
and/or mentioned in trade agreements. In contrast with those working on physical
integrity rights, scholars working on the trade-labor rights relationship have an easier
time showing why they chose to examine labor rights. In a survey, Elliott and Freeman
(2003) demonstrate that labor rights can improve under globalization. Mosley and Uno
(2007) hypothesize that the impact of globalization on labor rights depends not only
on the overall level of economic openness, but also on how a country participates in
global production networks. They find that although foreign investment inflows are
positively and significantly related to the rights of workers, trade competition generates
504    A Match Made in Heaven?

downward “race to the bottom” pressures on collective labor rights. The authors also
find that domestic institutions and labor rights in neighboring countries are important
correlates of workers’ rights. Mosley (2008) examines the specific case of Costa Rica,
finding that local conditions influence its ability to improve labor standards. Greenhill,
Mosley, and Prakash (2009) examine whether high trade standards in the importing
country prod the developing country to improve labor conditions. They conclude that
labor standards in developing countries are influenced by the labor standards of their
exporting destinations, not by their overall levels of trade openness. Instead of exporters
pushing down labor standards of the importing countries, as the race to the bottom lit-
erature suggests, importers can influence—positively or negatively—the collective labor
laws and practices of trade partners. They also find it is not trade openness per se that
matters for labor standards in the developing world, but rather the status of activism and
labor standards in the importing country.
Other scholars have tried to understand the trade-human rights relationship by
examining the activists who have pushed for greater balance between trade and human
rights objectives. Cambridge law professor Andrew Lang believes activists have taught
policy makers that they must pay greater attention to human rights. Hebrew University
professor Tomer Broude (2013), in contrast, believes activists have finally realized that
the WTO is not directly causing human rights problems. He stresses that during “the
Battle in Seattle” in 1999, activists blamed the WTO for human rights abuses, but dur-
ing “Occupy Wall Street” some ten years later, the WTO was rarely, if ever, mentioned.
The Occupy protestors blamed the banks and complicit governments. He concludes
that activists now recognize that member states, which make WTO decisions, are the
problem.15 Shareen Hertel (2005) has a very different take. She believes that protes-
tors made human rights arguments against WTO rules, but did not have a clear argu-
ment or understanding of human rights. For example, she stresses that labor rights
and human rights were discussed in separate spheres, as if labor rights are not human
rights.16
In lieu of reviewing activism, other scholars have decided to follow the money—to
examine the role of business as a human rights actor conducting trade (Spar 1998;
Ottoway 2001; Elliott and Freeman 2003; Aaronson and Higham 2013). University of
Memphis professors Robert Blanton and Shannon Blanton have done several studies
that examined how human rights conditions might affect business and trade relation-
ships (Blanton and Blanton 2001). They argue that repressive states are tricky trade part-
ners. Because corporations are conscious of their image, executives may seek locales
where there is good governance and respect for human rights. Clearly a growing num-
ber of firms have been under pressure to monitor how their practices affect human
rights, especially in nations with inadequate governance (Cassel 2001; Chandler 2003).
Blanton and Blanton argue that “respect for human rights may be the foundation for a
‘virtuous cycle’ in which citizens have both the skills and motivation to succeed in the
global economy” (2008, 103). However, many citizens around the world may feel that
globalization is undermining the value of their skills and expertise, and that the cycle is
less virtuous than vicious.
 Susan Ariel Aaronson    505

The Effects of the Marriage of PTAs

and Human Rights: Are Human Rights
Conditions Improving?

Although the wedding of trade and human rights in specific trade agreements is a new
policy phenomenon, it seems the marriage has been consummated. Policy makers in
a growing number of countries are changing how they think about human rights. As
a result, more policy makers are respecting human rights norms, as Hafner-Burton
(2009) and Bartels (2005a, 2008) have noted.

Mauritania
Mauritania provides an example of how the EU used an FTA to ensure that a government
met its commitments to civil and political rights. In 2005, after the military seized power
in a coup, the European Commission warned the new government that its behavior had
breached that country’s commitments under the Cotonou Agreement, the most com-
prehensive partnership agreement among seventy-nine developing countries in Africa,
the Caribbean, and the Pacific. Soon thereafter the new Mauritanian regime pledged to
hold free and fair elections and initiated a process to establish an independent National
Commission for Human Rights. In 2007 the government held free and fair elections.
Political scientist Hafner-Burton concludes that these provisions gave the EU leverage
over the government’s progress toward reforms (Hafner-Burton 2009, 151–160).

Thailand
Thailand provides a good example of how citizens are monitoring policy makers’ adher-
ence to international human rights norms when they make trade policy. In 2003 Thailand
began negotiating an FTA with the United States. In 2005 the Thailand National Human
Rights Commission drafted a report on the human rights implications of the proposed
agreement. Thailand was the first country to do a thorough human rights review of an
FTA, expressing concern about traditional knowledge and intellectual property rights.
Traditional knowledge can be defined as a knowledge system embedded in the cultural
traditions of regional or indigenous communities. It can include knowledge about hunt-
ing or agriculture, midwifery, ethnobotany, or ecological knowledge. In recent years
indigenous peoples have sought to prevent the patenting of traditional knowledge
and resources where they have not given explicit consent. The commission concluded
that if the FTA was approved, Thailand might undermine its citizens’ human rights.17
As the report was publicized, growing numbers of Thai citizens began to oppose the
agreement. The Thai prime minister pledged to ensure greater public involvement in
506    A Match Made in Heaven?

the negotiating process to shape the agreement.18 However, the negotiations ended after
a military coup on September 19, 2006. In recent years, as Thailand’s neighbors have
become active participants and demandeurs of FTAs, Thailand has become more will-
ing to negotiate these agreements with the EU, EFTA, and United States. Various NGOs,
including human rights NGOs, are closely monitoring these efforts and plan to make
sure policy makers pay attention to the human rights implications for Thailand, should
FTA negotiations recommence.19

Mexico
Mexico provides a real-world example of how trade agreements may encourage policy
makers to accept human rights norms. Since joining NAFTA, Mexican trade policy
has become more responsive to public concerns about labor rights. For example, the
Mexican government, which was long chided for its unwillingness to respect labor
rights, began to work internationally to protect its citizens’ labor rights. In September
2009 Mexican consulates attempted to educate Mexican guest workers in the United
States regarding their labor rights. Moreover, NAFTA leaders meet regularly, and human
rights and the rule of law have become important parts of their discussions. Thus, at
their meeting in August 2009 the leaders of the three NAFTA countries discussed public
health, border controls, and public safety—all issues relating to human rights.20

Next Steps: What Should

Scholars Examine?

1.  While policy makers have examined how some human rights are affected by trade
and how some trade agreements affect some human rights, they do not yet know
how protecting human rights may facilitate trade or what human rights should be
in place to encourage trade. Policy makers should ask whether respect for human
rights, such as property rights, needs to be in place for trade to be conducted.
2.  While scholars have extensively studied trade and personal integrity rights, work-
ers’ rights, and gender, they have paid little attention to other important rights
such as privacy or access to information, which are directly affected by trade
agreements. In fact, the United States has struggled to balance Internet freedom,
the need to surveil Internet users to protect national security and the competitive
advantage of its Internet firms to find an effective balance between expanding trade
and respecting human rights.

Edward Snowden’s revelations about US (and other governments’) surveillance of

cross-border communications has undermined international efforts to balance privacy,
 Susan Ariel Aaronson    507

public security, and information flows. Given the importance of the Internet as a plat-
form for trade as well as a medium to enhance human rights, we find it surprising that so
few scholars have examined this area where trade and human rights intersect.
When individuals exchange information across borders, they are trading. As citizens,
consumers, and producers of information, individuals have a right to privacy when their
personal information is exchanged between third parties. Citizens also have the right to
have their data be protected from theft or misuse. The EU, Canada, Australia, and New
Zealand, among other countries, have strong privacy laws. The EU and Canada include
privacy regulations in their trade agreements, although that language is aspirational and
not binding.
However, under authority delineated in the USA Patriot Act of 2001, the US National
Security Agency routinely examines communications between foreign citizens and US
citizens when officials suspect possible links to terrorist activity. In so doing, the United
States has violated the privacy of these foreign citizens without their knowledge or per-
mission. American officials claim they are trying to protect human security by moni-
toring the Internet for terrorists. Although other democracies have acted in a similar
manner (monitoring their citizens’ online communications),21 many American allies
have not responded positively to Snowden’s revelations. For example, some Canadian
agencies have refused to send information to the United States through e-mail or data
flows; they are concerned that such outsourcing could undermine Canada’s security.22
Several Canadian provinces including British Colombia (BC) and Nova Scotia have
enacted legislation to restrict the disclosure of personal information outside Canada
and expanded the scope of personal liability and sanctions for contraventions this
legislation.23
Meanwhile, the United States has become a strong advocate for including language
encouraging the free flow of information in its new trade agreements, the Trans-Pacific
Partnership (TPP) and the Trans-Atlantic Trade and Investment Partnership (TTIP)
(Aaronson and Maxim, forthcoming). The United States wants its trade partners to
eliminate barriers to the free flow of information across borders such as restrictions on
server location, egregious privacy standards, or censorship and filtering. The United
States is the largest source of Internet services and seeks to maintain its comparative
advantage. Should it succeed in including these policies in these new trade agree-
ments, the country could simultaneously enhance access to information, facilitate free-
dom of expression, and maintain the market power of the American Internet industry.
However, international support for these provisions seems shaky in light of Snowden’s
revelations. Other nations are determined to protect their citizens from US surveillance
but they also may see an opportunity to gain market share by developing regulations
that require information be housed in local servers or maintained by domestic firms.
While other governments have become more determined to protect their citizens’
online privacy, US policy makers (at the behest of major firms) have become increas-
ingly concerned that privacy protections can undermine both the free flow of informa-
tion and Internet industry competitiveness. For example, in its 2013 report on foreign
trade barriers, the US Trade Representative (USTR) argued that British Columbia’s and
508    A Match Made in Heaven?

Nova Scotia’s privacy laws discriminate against US suppliers because they require that
personal information be stored and accessed only in Canada.24 The USTR claims these
laws prevent public bodies from using US services when personal information could be
accessed from or stored in the United States.25 In its 2012 report the United States also
cited Australia’s approach to privacy, noting Australia’s unwillingness to use US com-
panies for hosting due to concerns about privacy violations.26 In 2013 the USTR noted
that negative messaging about US privacy is on the decline, but has not disappeared.27
The US government also complained about Japan’s uneven approach to privacy and
Vietnam’s unclear approach.28 Ironically, the United States also argues that China’s fail-
ure to enforce its privacy laws stifles e-commerce.29
The US government is also concerned that some governments have restricted infor-
mation flows to the United States because of the Patriot Act.30 However, because of the
National Security Agency privacy leaks, the United States may struggle to convince
other governments and its own citizens that it fully respects foreigners’ privacy rights.
Finally, the US government is closely monitoring how some countries regulate the
Internet and the effects of such domestic policies on Internet freedom. For example,
in December 2012 the United States extended normal trade relations to Russia and
Moldova. The law contains a provision added by the House of Representatives that
would expand the scope of the Special 301 report issued by the Office of the US Trade
Representative each year. This provision mandates that the report include a description
of laws, policies, or practices by the Russian Federation that deny “fair and equitable
treatment” to US digital trade.31
To some observers, the US approach to digital trade and human rights appears con-
tradictory and ineffective. On the one hand, the United States is pushing for language
in trade agreements to facilitate the free flow of information, which could enable more
people access to more information. American officials have been forceful and vocal
advocates of Internet freedom. Yet at the same time, the United States tries to name
and shame governments that are developing data localization policies or local server
requirements in the interest of protecting their citizens’ privacy. In so doing, the US may
encourage other governments to enact policies that favor local firms which respect local
privacy standards. Finally, Snowden’s revelations have led to distrust among netizens
and doubt about U.S. efforts to promote digital rights.

3.  Surprisingly, although the WTO’s GATT is the main international trade agree-
ment, very few scholars have examined how WTO rules may have direct or indi-
rect human rights spillovers. Moreover, there are few studies that examine how
WTO membership affects the behavior of governments regarding human rights.

Mary C. Cooper tried to test the relationship between WTO membership and democ-
ratization (as opposed to protecting a particular human right) for the period 1947–1999.
She could not determine whether democratic states were more likely to join the WTO
or WTO membership makes countries more likely to become or remain democratic
(Cooper 2003). Aaronson and Abouharb (2011) tested whether WTO norms, which
 Susan Ariel Aaronson    509

replicate civil and political rights, due process, access to information (what the WTO
calls transparency), and nondiscrimination (evenhandedness) spill over into the polity
of member states. Although today policy makers seem more focused on their bilateral
and regional FTAs, they are still working with the WTO. We need more information on
its influence on human rights.

4.  We need a greater understanding of how trade disputes may facilitate greater gov-
ernmental respect for human rights (or empower individuals to demand their
rights).

In 2010 European Commission vice president Neelie Kroes said that China’s Internet
censorship qualified as a trade barrier that could be challenged in a trade dispute.
China’s Web filtering and firewall act as trade barriers, and they aren’t applied in a uni-
form and impartial way. In late 2011 the US government sent a letter to the Chinese gov-
ernment asking it to explain inconsistencies in its Internet policies. Under paragraph 4
of Article II of the General Agreement on Trade in Services (GATS), the United States
asked China to explain why some foreign sites were inaccessible in China, who decides
if and when a foreign website should be blocked, and if China has an appeal proce-
dure for such blockage. Although China is required to respond under the GATS, the
US government supposedly did not receive a formal reply. The US Trade Representative
has also researched whether it could challenge Chinese Internet restrictions as a viola-
tion of WTO rules.32 However, the United States is unlikely to take this route, as policy
makers would not want to create precedents that could limit its or its allies’ ability to
restrict access to the Internet for national security reasons.33 Tomer Broude and Holger
Hestermeyer (forthcoming) conclude that a trade dispute would not be a good idea, as
it is unclear what specific WTO rules have been breached by such censorship. Scholars
might provide new insights into whether trade disputes should be utilized in this man-
ner, and if so, how.

5.  Is the strategy of linking trade and human rights supportive of internationally
accepted human rights? Here again the United States provides a contradictory
case study.

On the one hand, it prioritizes certain human rights and makes adherence binding
in the trade agreement. On the other hand, it undermines a central ethos of the human
rights community: that human rights are universal and indivisible. The EU approach is
more supportive of human rights as universal and indivisible, but it appears less effective
at achieving human rights improvements. Scholars should examine the pros and cons of
each approach.
Moreover, by focusing on specific human rights in its FTAs, the United States ignores
the internationally accepted notion that human rights are universal and indivisible, yet
it works hard to promote specific human rights and now makes some of these rights
binding. In this sense, the United States does more than any other government to link
510    A Match Made in Heaven?

FTAs and human rights. However, it is essentially saying to its partner nations: make
the rights we value top priorities. In this way, the US strategy for linking human rights
is insensitive toward other cultures that may have different human rights priorities. This
strategy may inspire US trade partners to do more to advance some human rights, but
it is unlikely to inspire these governments to devote more resources to human rights in
general.
In contrast, the EU includes language supportive of universal human rights as
delineated in the Universal Declaration. It requires its developing-country trade
partners to accept a clause stipulating that human rights be protected to receive
trade benefits. If its partners don’t adhere, the EU can suspend trade concessions
and reduce foreign aid. However, the EU seems reluctant to use the power inher-
ent in these agreements. It rarely suspends trade concessions, because policy mak-
ers believe that the process of dialogue can change hearts and minds (Aaronson and
Zimmerman 2007).

6.  Countries have a wide range of views about whether they should accept human
rights provisions in trade agreements. Scholars should examine what is behind
these differing attitudes.

Some nations, such as Australia, have actually refused to negotiate trade agreements
with human rights provisions. Yet China has accepted human rights provisions in its
PTAs with Chile and New Zealand. How do we explain these differences? Some coun-
tries are comfortable using their economic power to promote political change, some are
neutral and noninterventionist, and others may be using their acceptance of these pro-
visions to signal investors and funders. Perhaps China sees adopting these provisions as
a way to signal foreign investors that it is evenhanded and is attempting to promote the
rule of law. Scholars should endeavor to explain why countries respond so differently to
the marriage of trade and human rights.

7.  Scholars don’t know if trade-related incentives work better than disincentives to

encourage policy makers to respect and advance human rights. What do incen-
tives or disincentives do for the ability of citizens to demand their rights?

Some countries have decided to use disincentives such as trade sanctions or fines
as a means of advancing the human rights embedded in particular trade agree-
ments. However, sanctions or fines can do little to build demand for human rights
or to train government officials or factory managers in how to respect such rights.
Isolating a government or punishing it will do little to increase the targeted coun-
try’s commitment to human rights over time. Other countries rely on dialogue
to prod changes, but dialogue may do little to encourage a country to change its
behavior. Still others rely on incentives. Scholars can do more to help policy makers
understand whether these incentives are really effective, and if so, when such incen-
tives should be offered.
 Susan Ariel Aaronson    511

8.  Human rights provisions in trade agreements have costs as well as benefits.

The world and its people benefit when more governments protect, respect, and
realize human rights. Yet some human rights provisions are expensive for developing
countries to implement. These governments often have few resources, yet under many
recent trade agreements they are tasked with protecting intellectual property, providing
access to affordable medicines, and investing in education, among other goals. Often
developing-country governments cannot achieve all of these goals and instead must pri-
oritize them. Trade agreements may prod policy makers to turn the human rights prior-
ities of their trade partners into their own human rights priorities. We don’t know if this
strategy ultimately increases the supply of, and demand for, human rights. To gain better
understanding of the costs and benefits of the association of trade and human rights,
scholars, policy makers, and activists could use qualitative studies, empirical studies, or
human rights impact assessments. Human rights impact assessments are relatively new;
they are designed to measure the potential impact of a trade agreement on internation-
ally accepted human rights standards. Trade and human rights policy makers should
collaborate with scholars, NGOs, and others to develop a clear and consistent method-
ology for evaluating such impacts (3D 2009; Walker 2009).

9.  We have little insight into whether governments use human rights conditions to
justify a trade agreement or relationship.

Some scholars tried to make this case for China, when that country sought to become
a member of the WTO. Did human rights improve in China as a result of expanded
trade? Which human rights? How did membership in the WTO directly or indirectly
affect the ability of Chinese citizens to demand their rights, or the government’s willing-
ness to respect particular rights (such as due process or access to information)? Or has
expanded trade perpetuated a repressive regime?

Conclusion

Scholars have made great progress in trying to understand the relationship of trade and
human rights, but we don’t know if it is an enduring match made in heaven, or a match
in name only. Scholars have asked and attempted to answer only some of the big ques-
tions. They have not researched whether expanded trade encourages (or discourages)
public demand for specific human rights. They have not done much work on whether
the protection of specific human rights facilitates trade and trade agreements (such as
property rights). Moreover, we have little insight into whether governments use human
rights conditions to justify a trade agreement or relationship. Empirical scholars need to
move beyond labor rights, women’s rights, and physical integrity rights to examine other
human rights such as privacy, property rights, and access to information. Moreover,
512    A Match Made in Heaven?

because a growing number of countries are linking human rights and trade objectives,
scholars need to pay greater attention to what governments are actually doing and which
strategies work.
In the world beyond the ivy tower, more countries are joining trade and human
rights. If this marriage is to be sustainable, we need greater understanding of whether
this union is effective and can or should endure.

Acknowledgment
I am grateful to Rob Maxim and K. Daniel Wang for their editing and research assis-
tance and to Lisa Martin for her advice on this chapter.

Notes
1. Irwin (1996, 14–16, 21).
2. I asked my research assistant, K. Daniel Wang, to do a general count of articles and books
on the trade and human rights relationship since 2001. He found approximately one hun-
dred books and articles.
3. European Commission (2013); Remarks of Industrial Trade Union, p.  5 (http://trade.
ec.europa.eu/doclib/docs/2013/july/tradoc_151656.pdf).
4. Reaching for the Clouds, The Economist, July 20, 2013, http://www.economist.com/
news/europe/21582015-europe-wants-tougher-data-privacy-rules-deter-american-
snooping-reaching-clouds.
5. Paragraph 2(a) General 319, GATT Analytical Index, http://www.wto.org/english/res_e/
booksp_e/analytic_index_e/gatt1994_04_e.htm#articleXA.
6. International Monetary Fund (2007, 169–172); Jansen and Lee (2007).
7. OECD Secretariat (2005, 3, 9); OECD (2002, 2003, 2004, 2005).
8. On Bangladesh, see Bose (2013) and Greenhouse (2013).
9. Some of these options are listed at http://www.law.harvard.edu/library/research/data-
bases/international-relations.html and http://classguides.lib.uconn.edu/content.
php?pid=30942&sid=330266.
10. http://www.humanrightsdata.org/faq.asp#2.
11. http://www.serfindex.org/data/ and http://www.serfindex.org/research/.
12. http://www.globalintegrity.org/report.
13. http://www.doingbusiness.org/about-us.
14. http://hdr.undp.org/en/statistics/hdi/.
15. The “Battle in Seattle” refers to the protests that took place during the WTO’s Third
Ministerial Conference, held in Seattle, Washington, in November–December 1999.
16. Hertel (2005); Compa and Diamond (1996).
17. http://www.measwatch.org/autopage/show_page.php?t=5&s_id=3&d_id=7.
18. PM Calls for More Public Participation in FTA Deals, MCOT News, http://www.bilaterals.org/
article.php3?id_article=953&var_recherche=public+information (accessed August 3, 2006).
19. http://www.bilaterals.org/spip.php?rubrique117; http://www.ftawatch.org/; http://www.
ustr.gov/sites/default/files/US--Thailand%20TIFA.pdf; http://www.ustr.gov/about-us/
press-office/press-releases/2013/january/US-Thailand-engagement, all last searched 8/18/2013.
 Susan Ariel Aaronson    513

20. US Trade Representative (2009).

21. See for example, Britain, at http://www.theguardian.com/technology/2013/jun/07/
uk-gathering-secret-intelligence-nsa-prism?guni=Article:in%20body%20link, or
France, at http://www.nytimes.com/2013/07/05/world/europe/france-too-is-collecting-d
ata-newspaper-reveals.html.
22. Information Technology Association of Canada, Shared Services Canada Takes National
Security Exception, http://itac.ca/news/shared_services_canada_takes_national_security_
exception.
23. Jason Young, BC Attempts to Regulate Outsourcing of Personal Information,” November
4, 2004, http://www.dww.com/?page_id=1052; Cate (2008, 1–2); and Fred H.  Cate,
“Provincial Canadian Geographical Restrictions on Personal Data in the Public Sector,
Center for Information Leadership, Hunton and Williams, http://www.hunton.com/files/
Publication/2a6f5831-07b6-4300-af8d-ae30386993c1/Presentation/PublicationAttachme
nt/0480e5b9-9309-4049-9f25-4742cc9f6dce/cate_patriotact_white_paper.pdf. And Cate,
“Private Data in Public Hands,” and http://www.hunton.com/files/Publication/
caa02b36-10f1-41eb-8237-90699b4d0777/Presentation/PublicationAttachment/
d63a384c-c51f-47ee-a660-bb9dd470f232/Private_Data_in_Public_Hands.pdf.
24. US Trade Representative (2013, 60–61).
25. USTR Flags Procurement, Data Flow Issues as New Barriers in Canada, Inside US Trade,
April 27, 2012, http://insidetrade.com/Inside-US-Trade/Inside-US-Trade-04/27/2012/
ustr-flags-procurement-data-flow-issues-as-new-barriers-in-canada/menu-id-710.html.
26. US Trade Representative (2012).
27. US Trade Representative (2013, 31).
28. US Trade Representative (2012, 216).
29. US Trade Representative (2012, 96).
30. US Trade Representative (2012, 166).
31. Sec. 203—Reports on Laws, Policies, and Practices of the Russian Federation That
Discriminate Against United States Digital Trade,” Russia and Moldova Jackson-Vanik
Repeal and Sergei Magnitsky Rule of Law Accountability Act of 2012, Public Law 112-208,
http://www.gpo.gov/fdsys/pkg/PLAW-112publ208/html/PLAW-112publ208.htm.
32. Obama Acts on FAC Petition against China’s “Great Firewall,” FAC, October 19, 2011,
http://www.firstamendmentcoalition.org/2011/10/obama-acts-on-fac-petition-against-
chinas-Internet-censors/.
33. Brendan Greeley and Mark Drajem, China’s Face- book Copycats Focus US on Trade as
Well as Rights, Bloomberg/BusinessWeek, March 10, 2011; letter from Ambassador Michael
Puncke, US Ambassador to the WTO, to Ambassador Yi Xiaozhun, Chinese Ambassador
to the WTO, and attachment, October 17, 2011, http://insidetrade.com/iwpfile.html?file=o
ct2011%2Fwto2011_2996a.pdf.

References and Recommended Reading

Aaronson, Susan A. 2011. Human Rights. In Preferential Trade Agreement Policies for
Development: A Handbook, edited by Jean-Pierre Chauffour and Jean-Christophe Maur,.
Washington, DC: World Bank Publications, pp. 443–466.
Aaronson, Susan Ariel. 2007. Seeping in Slowly: How Human Rights Concerns Are Penetrating
the WTO. World Trade Review 6 (3): 413–449.
514    A Match Made in Heaven?

Aaronson, Susan Ariel. 2002. Taking Trade to the Streets: The Lost History of Public Efforts to
Shape Globalization. Ann Arbor. University of Michigan Press.
Aaronson, Susan Ariel, and M. Rodwan Abouharb. 2011. Unexpected Bedfellows: The GATT,
the WTO and Some Democratic Rights. International Studies Quarterly 55 (2): 379–408.
Aaronson, Susan Ariel, and Ian Higham. 2013. “Re-righting Business”: John Ruggie and the
Struggle to Develop International Human Rights Standards for Transnational Firms.
Human Rights Quarterly 35 (2): 333–364.
Aaronson, Susan Ariel, and Robert Maxim. Forthcoming. “Trade and the Internet.” In
International Political Economy of Trade, edited by David Deese, Edward Elgar.
Aaronson, Susan Ariel, and Jamie M. Zimmerman. 2007. Trade Imbalance:  The Struggle
to Weigh Human Rights Concerns in Trade Policymaking. Cambridge, U.K. Cambridge
University Press.
Apodaca, Clair. 2001. Global Economic Patterns and Personal Integrity Rights after the Cold
War. International Studies Quarterly 45 (4): 587–602.
Bagwell, Kyle, and Robert W. Staiger. 1998. Will Preferential Agreements Undermine the
Multilateral Trading System? Economic Journal 108: 1162–1182.
Bartels, Lorand. 2008. The Application of Human Rights Conditionality in the EU’s Bilateral Trade
Agreements and Other Trade Arrangements with Third Countries. Brusel: Evropský parlament.
Bartels, Lorand. 2005a. Human Rights Conditionality in the EU’s International Agreements.
Oxford: Oxford University Press.
Bartels, Lorand. 2005b. The Legality of the EC Mutual Recognition Clause under WTO Law.”
Journal of International Economic Law 8 (3): 691–720.
Bhagwati, Jagdish. 1996. The Demands to Reduce Domestic Diversity among Trading Nations.”
Fair Trade and Harmonization: Prerequisites for Free Trade 1: 9–40.
Blanton, Robert G., and Shannon L. Blanton. 2008. Virtuous or Vicious Cycle? Human Rights,
Trade, and Development. In North and South in the World Political Economy, edited by Rafael
Reuveny and William R. Thompson, 91–103. Malden, MA; Oxford: Blackwell Publishing.
Blanton, Robert. G., and Shannon Lindsey Blanton. 2001. Democracy, Human Rights, and
US-Africa Trade. International Interactions 27 (3): 275–295.
Bose, Nandita. 2013. Stronger Labor Law in Bangladesh after Garment Factory Collapse.
Reuters, July 13. http://www.reuters.com/article/2013/07/15/us-bangladesh-labour-idUSBRE
96E05R20130715.
Broude, Tomer. 2013. From Seattle to “Occupy”: Trade, Human Rights and the Shifting Focus
of Social Protest. in Daniel Drache and Les Jacobs, eds., The Changing Global Landscape
for International Trade and Human Rights Linkages (app.  16 pages) (forthcoming)
(Oxford: Routledge, 2013).
Broude, Tomer, and Holger P. Hestermeyer. Forthcoming. The First Condition of Progress?
Freedom of Speech and the Limits of International Trade Law. Virginia Journal of
International Law. (Abstract at http://ssrn.com/abstract=2260969.)
Büthe, Tim, and Helen V. Milner. 2008. The Politics of Foreign Direct Investment into
Developing Countries: Increasing FDI Through International Trade Agreements? American
Journal of Political Science 52 (4): 741–762.
Cao, Xun, Brian Greenhill, and Aseem Prakash. 2013. Where Is the Tipping Point? Bilateral
Trade and the Diffusion of Human Rights. British Journal of Political Science 43 (1): 133–156.
Cassel, Douglass. 2001. Human Rights and Business Responsibilities in the Global Marketplace.
Business Ethics Quarterly: 261–274.
 Susan Ariel Aaronson    515

Cate, Fred H. 2008. Provincial Canadian Geographic Restrictions on Personal Data in the
Public Sector. Centre for Information Policy Leadership, Submitted to the Trilateral
Committee on Transborder Data Flows.
Cate, Fred H. 2010, “Private Data in Public Hands,” Speech to 11th Annual Privacy and Security
Conference,” Victoria, British Columbia.
Chandler, Geoffrey. 2003. The Evolution of the Business and Human Rights Debate.” In
Business and Human Rights:  Dilemmas and Solutions, edited by Rory Sullivan, 22–32.
Greenleaf Publishing.
Charnovitz, Steve. 2005. The Labor Dimension of the Emerging Free Trade Area of the
Americas. In Labor Rights as Human Rights, edited by Philip Alston, Oxford:  Oxford
University Presspp. 143–176.
Charnovitz, Steve. 2001. The WTO and the Rights of the Individual. Intereconomics 36
(2): 98–108.
Compa, Lance, and Stephen Diamond, eds. 1996. Human Rights, Labor Rights and International
Trade. Philadelphia: University of Pennsylvania Press.
Cooper, Mary Comerford. 2003. International Organizations and Democratization: Testing
the Effect of GATT/WTO Membership. Asia Pacific Research Center, April. http://www.
kenanflagler.unc.edu/KI/kiWashington/humanRightsAndTrade/hrtLibrary.
Dobbin, Frank, Beth Simmons, and Geoffrey Garrett. 2007. The Global Diffusion of Public
Policies:  Social Construction, Coercion, Competition, or Learning? Annual Review of
Sociology 33: 449.
Dollar, David, 1992. “Outward-Oriented Developing Economies Really Do Grow More
Rapidly: Evidence from 95 LDCs, 1976-1985,” Economic Development and Cultural Change,
University of Chicago Press, 40(3), pp. 523–44.
Elliott, Kimberly Ann, and Richard B. Freeman. 2003. Can Labor Standards Improve under
Globalization? Washington: D.C. Peterson Institute Press/All Books.
European Commission. 2013. Update on the Transatlantic Trade and Investment Partnership
(TTIP)—First Negotiation Round. Civil Society Dialogue, July 16.
Greenhill, Brian. 2010. The Company You Keep: International Socialization and the Diffusion
of Human Rights Norms. International Studies Quarterly 54 (1): 127–145.
Greenhill, Brian, Layna Mosley and Aseem Prakash 2009 “Trade-Based Diffusion of labor
rights: A Panel Study, 1986-2002. American Political Science Review 103 (4): 669–691.
Greenhouse, Stephen. 2013. Bangladesh Adopts New Labor Law. New  York Times, July 16.
http://www.nytimes.com/2013/07/17/world/asia/under-pressure-bangladesh-adopts-new-l
abor-law.html?_r=0.
Hafner-Burton, Emilie. 2009. Forced to Be Good: Why Trade Agreements Boost Human Rights.
Ithaca, NY: Cornell University Press.
Hafner-Burton, Emilie M. 2005. Right or Robust? The Sensitive Nature of Repression to
Globalization. Journal of Peace Research 42 (6): 679–698.
Hafner-Burton, Emilie M., and James Ron. 2009. Seeing Double:  Human Rights Impact
Through Qualitative and Quantitative Eyes. World Politics 61 (2): 360–401.
Hafner-Burton, Emilie M., and Kiyoteru Tsutsui. 2005. Human Rights in a Globalizing
World: The Paradox of Empty Promises. American Journal of Sociology 110 (5): 1373–1411.
Harrelson-Stephens, Julie, and Rhonda Callaway. 2003. Does Trade Openness Promote
Security Rights in Developing Countries? Examining the Liberal Perspective. International
Interactions 29 (2): 143–158.
516    A Match Made in Heaven?

Hertel, Shareen. 2005. What Was All the Shouting About: Strategic Bargaining and Protest at
the WTO Third Ministerial Meeting. Human Rights Review 6 (3): 102–118.
Howse, Robert. 2002. Human Rights in the WTO: Whose Rights, What Humanity? Comment
on Petersmann. European Journal of International Law 13 (3): 651–659.
International Monetary Fund. 2007. Chapter 5: The Globalization of Labor. World Economic
Outlook (October). http://www.imf.org/external/pubs/ft/weo/2007/01/pdf/c5.pdf.
Irwin, Douglas A. 1996. Against the Tide:  An Intellectual History of Free Trade. Princeton,
NJ: Princeton University Press.
Jansen, Marion, and Eddy Lee. 2007. Trade and Employment: Challenges for Policy Research;
A  Joint Study of the International Labour Office and the Secretariat of the World Trade
Organization. International Labour Organization.
Korinek, Jane. 2005. Trade and Gender: Issues and Interactions. OECD Trade Policy Working
Paper No. 24. Paris, France: OECD Publishing.
Lang, Andrew T. F. 2007. The Role of the Human Rights Movement in Trade Policy-Making:
Human Rights as a Trigger for Policy Learning. NZJPIL 5: 77–102.
Levy, Philip I. 1999. Sanctions on South Africa: What Did They Do? Yale Economic Growth
Center, Discussion Paper No. 796.
Mansfield, Edward D., and Jon C. Pevehouse. 2008. Democratization and the Varieties of
International Organizations.” Journal of Conflict Resolution 52 (2): 269–294.
Milner, Wesley T. 2002. Economic Globalization and Rights. In Globalization and Human
Rights, edited by Alison Brysk,. Oakland: CA. University of California Press.
Mosley, Layna. 2008. Workers’ Rights in Open Economies Global Production and Domestic
Institutions in the Developing World. Comparative Political Studies41 (4–5): 674–714.
Mosley, Layna, and Saika Uno. 2007. Racing to the Bottom or Climbing to the Top? Economic
Globalization and Collective Labor Rights. Comparative Political Studies 40 (8): 923–948.
OECD. 2002. Babies and Bosses: Reconciling Work and Family Life. Volume 1, Reconciling Work
and Family Life in Australia, Denmark and the Netherlands. Paris: OECD. (Volume 1, 2002;
Volume 2, 2003; Volume 3, 2004; Volume 4, 2005).
OECD Secretariat. 2005. Aging Populations:  High Time for Action. Background Paper
Prepared for the Meeting of G-8 Employment and Labour Ministers, London, 10–11 March.
http://www.oecd.org/dataoecd/61/50/34600619.pdf.
Ottaway, Marina. 2001, “Reluctant Missionaries. http://www.foreignpolicy.com/articles/2001/
07/01/reluctant_missionaries.
Paul, Joel R. 2003. Do International Trade Institutions Contribute to Economic Growth and
Development. Virginia Journal of International Law 44: 285.
Rodrik, Dani. 2000. Trade Policy Reform as Institutional Reform. Cambridge, MA: Harvard
University Press.
Roosevelt, Franklin D. 1941. Four Freedoms Speech: Annual Message to Congress on the State
of the Union, January 1. http://www.fdrlibrary.marist.edu/pdfs/fftext.pdf.
Sachs, Jeffrey D., and Andrew M. Warner. 1995. Economic Convergence and Economic Policies.
No. 5039. Cambridge: MA., National Bureau of Economic Research.
Seguino, Stephanie, and Caren Grown. 2006. Gender Equity and Globalization:
Macroeconomic Policy for Developing Countries. Journal of International Development 18
(8): 1081–1104.
Sen, Amartya. 1999. Development as Freedom. Oxford University Press.
Spar, Debora L. 1998. The Spotlight and the Bottom Line: How Multinationals Export Human
Rights. Foreign Affairs 77, 2: 7–12.
 Susan Ariel Aaronson    517

Sykes, Alan O. 2003. “International Trade and Human Rights: An Economic Perspective.” Law
and Economics Working Paper No 188, University of Chicago Law School.
Temin, Peter. 2003 Mediterranean Trade in Biblical Times. In Eli Heckscher, International
Trade, and Economic History, edited by Ronald Findlay, 141–156.
3D. 2009. Insights on Human Rights Impact Assessments of Trade Policies and Agreements.
Geneva. http://www.humanrights.ch/upload/pdf/100607_09_HRIA_3D_berblick.pdf.
US Trade Representative. 2009. Joint Statement of the 2009 NAFTA Commission Meeting.
http://www.ustr.gov/about-us/press-office/press-releases/2009/october/joint-statement-2
009-nafta-commission-meeting.
US Trade Representative. 2012. 2012 National Trade Estimate Report of Foreign Trade Barriers.
http://www.ustr.gov/sites/default/files/NTE Final Printed_0.pdf.
US Trade Representative. 2013. 2013 National Trade Estimate Report on Foreign Trade Barriers.
March. http://www.ustr.gov/sites/default/files/2013%20NTE.pdf.
van Hees, Floris. 2004. Protection V. Protectionism: The Use of Human Rights Arguments in the
Debate for and Against the Liberalisation of Trade. Åbo akademi Turku, Finland.
Walker, Simon. M. 2009. The Future of Human Rights Impact Assessments of Trade
Agreements. School of Human Rights Research Series 35: (2009).
World Trade Organization. 2013. General Agreement on Tariffs and Trade 1994. Geneva: WTO
WTO Analytical Index.
 Chapter 27

Trade and M i g rat i on

Ma rgaret E . Pete r s

As citizens, policy makers, and scholars, we do not often think about trade and migra-
tion as linked: trade in goods seems to be an activity of profit-motivated firms, whereas
migration is an act of people for a myriad of economic and personal reasons. Yet from
some of our earliest models of trade, trade and migration have been linked through
their effects on prices for goods and the returns to the factors of production (people and
capital). In the Stolper-Samuelson (1948) model of trade and factor mobility, trade leads
to the equalization of goods prices and from there to the equalization of factor prices.
Likewise, factor mobility leads to the equalization of factor prices and from there to the
equalization of goods prices. Mundell (1957) argues that under this model, free trade
should decrease the demand for migration—if wages are the same in two countries
there would be little reason to migrate—and migration should dampen the demand
for trade—as migration would lead to more equal wages and more equal goods prices.
Further trade barriers should stimulate factor mobility, assuming the factors can move,
and barriers to the movement of capital or people should lead to increased trade, assum-
ing there are few barriers to trade. Thus, our classic economic models of trade suggest
that it is deeply tied to issues of both capital mobility and migration.
These classic models, of course, rely on many assumptions about the nature of trade
between two countries, some of the most important assumptions being that the coun-
tries have identical technologies; consumers in both countries have identical prefer-
ences driven by cost rather than love of variety; production is characterized by perfect
competition; and there are no domestic economic distortions (e.g., taxes, subsidies)
in either country (Markusen 1983, 342). Markusen argues that when we relax these
assumptions, we find that trade and factor movements can be complements instead
of substitutes. Similarly, theories of intra-industry trade (e.g., Melitz 2003) argue
that consumers gain from trade due to an increase in variety instead of simply from
decreased prices of goods. Other scholars (e.g., Gould 1994) argue that a desire for
variety can also make migration and trade complementary. I discuss below when it is
likely that greater trade leads to more migration, when it leads to less migration, and
vice versa.
 Margaret E. Peters    519

While scholars have devoted considerable attention to the issues of trade and capital
mobility (see Copelovitch and Pevehouse 2014 in this volume Should there be a link to
the Copelovitch and Pevehouse piece and should it be listed at 2015?), they have devoted
less attention to connections between trade and migration. This lacuna is due in part to
the lack of attention paid to migration in the international political economy (IPE) lit-
erature in general. For example, Keohane and Milner note that “since labor moves much
less readily across national borders than goods or capital, we have not considered migra-
tion as part of internationalization… . [I]‌n future work, serious attention should be given
to including migration in the analysis of internationalization” (1996, 258). Lake’s (2009)
review of open economy politics (OEP) mentions “trade” seventy-eight times, “capital”
twelve times, and “immigration” three times. Oatley’s (2011) critique of OEP mentions
tariffs, monetary and exchange rate policies, and investment flows, but not migration.
Migration has been relatively understudied because it was not seen as an important
policy issue when the field of IPE began in the 1960s and 1970s. At the time migra-
tion was at historically low levels for several reasons: emigration was restricted from
the Communist bloc countries; most of Western Europe was relatively wealthy, giving
Europeans few reasons to migrate; and most of the rest of the world was still poor enough,
given transportation costs, to make migration almost impossible. Today, of course,
migration has become a much larger issue as the flows of migrants have increased.
In response to the increased attention paid to migration by policy makers and the
public, scholars have also increased their study of migration. There are three main areas
of study that have linked trade and migration. The first seeks to understand public opin-
ion on trade and immigration. Given that trade and immigration should have similar
effects on the wages people earn and the prices they pay for goods, many have argued
that public opinion on these two flows should be similar. The literature has consistently
found that this is not the case and has explored two main explanations: the fiscal effects
of immigrants (Hanson, Scheve, and Slaughter 2007) and the cultural effects of immi-
grants (Hainmueller and Hiscox 2007, 2010).
Second, scholars have sought to understand how trade and immigration policy affect
each other. They have found that trade policy and immigration policy are substitutes.
Hatton and Williamson (2005, 2007) focus on how individual level preferences, driven
by the fiscal costs of immigrants and/or immigrants’ effects on inequality, lead to policy
change. Peters (2014, 2015) examines how trade affects business preferences for immi-
gration. Closed trade leads to an increase in businesses that use much immigrant labor,
and open trade leads to the closure of these businesses. Trade policy thus affects the
average business demand for immigration openness. Similarly, immigration policy may
affect trade policy: businesses that are uncompetitive internationally can be made more
competitive using immigrant labor (Peters 2014, 2015). This makes these firms more
supportive of trade openness than they would have been in the absence of immigration
openness.
Finally, scholars have examined how immigration and trade flows affect each other. It
has long been hoped by policy makers that open trade will lead to decreased migration.
Here, it is thought that open trade will lead to increased trade flows, which in turn will
520   Trade and Migration

lead to increased development and a lower demand for migration. This logic was used
to help sell the North American Free Trade Agreement (NAFTA) as a way to decrease
Mexican migration (Uchitelle 2007). Similarly, European policy makers have often
opened trade with new members of the European Union (EU) prior to opening their
labor markets to migration in hopes that open trade would increase development in new
member states and decrease migration (Geddes and Money 2011). While this may be
true in the long run, scholars have found that in the short run trade can be disruptive to
local economies, paradoxically resulting in more migration (Bacon 2012; Sassen 1988).
Scholars have also studied how increased migration can increase trade flows (e.g.,
Gould 1994). In this line of research, it is argued that migrants can increase trade
because they want goods from home that are not available in their new country, which
increases imports, and are parts of networks that can facilitate overseas trade, increasing
both imports and exports.
Trade and migration thus are linked in important ways: at times they act as a sub-
stitute for each other, and at times they are complements. These different effects can
influence how the public perceives these flows; how interest groups, including firms,
approach these flows; and how policy makers respond to these flows. The first section
of this chapter discusses the classic approaches to trade and migration from economics.
I then review the public opinion literature on trade and migration. In the third section
I examine how trade and migration policies can interact. The fourth section discusses
how trade and migration flows affect each other. Finally, I conclude with some thoughts
on new areas for research.

Standard Approaches
to Trade and Migration

Trade and migration have been linked by economists through the Stolper-Samuelson
model of trade (e.g., Samuelson 1948). In this model, there are two countries (A
and B) and two factors of production, capital and labor (or land and labor). One coun-
try (Country A) is relatively abundant in capital but scarce in labor, and the other
(Country B) is relatively abundant in labor but scarce in capital. The important thing
to note is that the countries are relatively abundant or scarce in the factors; one country
may have an absolute advantage in both labor and capital. Because Country A is rel-
atively abundant in capital, capital is relatively cheap and labor is relatively expensive
(wages are high); similarly, because Country B is relatively abundant in labor, labor is
relatively cheap (wages are low) and capital is relatively expensive. Goods made with
capital (labor) are relatively cheap (expensive) in Country A, and goods made with labor
(capital) are relatively cheap (expensive) in Country B.
If the two countries were to open migration, people would migrate from Country
B to Country A, because the wages in Country A are higher than they are in Country
 Margaret E. Peters    521

B. Wages would decrease in Country A as the supply of labor increased; wages would
increase in Country B as the supply of labor decreased. People would continue to
move until wages equalized. At this point, both countries would have the same ratio
of capital to labor, the same wages, and the same returns to capital. Goods would
be the same price, as the inputs—capital and labor—would cost the same in the two
countries.
If the two countries were to open trade instead, goods made mostly with labor
(labor-intensive goods) would be exported from Country B to Country A, since they
are cheaper in Country B, and goods made mostly with capital (capital-intensive goods)
would be exported from Country A to Country B, since they are relatively cheaper in
Country A.  Because labor-intensive goods imported from Country B are cheaper
than those made in Country A, companies in Country A  that make labor-intensive
goods would go out of business (or reinvest their capital in the production of the
capital-intensive good) and lay off their workers. This would increase the supply of
workers in Country A relative to the number of jobs, leading to a decrease in wages.
In Country B, the reverse would happen. Because companies that make labor-intensive
goods are increasing production to export goods to Country A, their demand for labor
would go up. This would decrease the supply of workers in Country B relative to the
number of jobs, leading to an increase in wages. Trade would therefore eventually lead
to the equalization of wages.
Under the Stolper-Samuelson model, trade and migration have the same effect
on wages and prices. Opening trade or migration decrease wages in Country A and
increases wages in Country B, leading to equalization of the prices of goods. Closing
trade and migration has the reverse effect: wages increase in Country A and decrease in
Country B, and prices would diverge.
Given these similar effects, Mundell (1957) argued that closing trade could spur factor
movements, assuming that factors are allowed to move. Although he examined the effect
of trade and capital movement, his argument could easily apply to migration. If trade is
closed, wages in Country A will be higher than in Country B. Assuming that there is no
impediment to moving, people should move from Country A to Country B until wages
equalize. Thus, if a country closes trade, it should expect more migration—emigration
if it is relatively labor abundant and immigration if it is relatively labor scarce. Over
time, wages will tend to converge. At this point, trade would decrease regardless of the
country’s openness to it. In this model, to produce one capital-intensive widget takes
the same amount of capital and labor in Country A as in Country B; similarly, it takes the
same amount of labor and capital to produce one labor-intensive widget in both coun-
tries. Because the costs of labor and capital are the same in both countries, the prices of
the capital- and labor-intensive goods are equal in both countries. Note, however, that
these results rely on the assumptions that goods are made with the same technology in
both countries, so that the equal price of inputs leads to equal costs; that production has
constant returns to scale, so that having one country specialize in the production of one
good does not lead to cost savings; and that there is perfect competition in each country,
so that marginal price equals the marginal cost. Under these strong assumptions, there
522   Trade and Migration

is no reason to trade. In practice, wages will tend to converge over the long run, but it is
unlikely that there would be no trade, because these assumptions do not hold fully.
Similarly, if a state closes migration, trade should increase, assuming that there are
no impediments to trade. Labor-intensive goods would be exported from Country B to
Country A, and capital-intensive goods would be exported from Country A to Country B.
Wages would eventually equalize; at this point the two states could remove any barriers
to migration without their removal leading to much migration, as wages would be the
same. Thus, in the Mundell model trade and migration are substitutes.
While the Stolper-Samuelson model and the Mundell model are helpful for under-
standing much of the relationship between trade and migration, the results of these
models rely on several assumptions that may or may not hold. Markusen (1983) pro-
vides several reasons why trade and migration might be complements instead of sub-
stitutes; among the most important are different production technologies, production
taxes, monopolies, economies of scale, and distortions in the factor (labor or capital)
markets. In each case, one industry has a special advantage when it comes to trade that
it would not otherwise have. For example, if labor-intensive goods have a special advan-
tage in Country A, say due to technology that leads to increased productivity, it will
export those goods instead of capital-intensive goods. Factor mobility will lead to more
inputs going into the sector with the special advantage instead of the disadvantaged sec-
tor. In the example above, migration will lead to increased labor in the labor-intensive
sector in Country A, decreasing the price of labor-intensive goods further and leading
to more trade.
Similarly, as long as consumers like variety, states with similar factor endowments
trade goods in the same industry (intra-industry trade). For example, the United States
and Germany, both relatively capital abundant states, trade Fords for Volkswagens.
Migration could be a complement to trade if, as we will see below, migration leads to a
desire for a greater variety of products. Thus, there are numerous reasons why trade and
migration would be complements rather than substitutes.
The importance of this discussion is that whether trade and migration are substitutes
or complements is likely to affect the politics of these two flows differently, as we will
see below.

Opinions on Trade and Migration

One area of research that examines the interplay of trade and migration is public opin-
ion. Most scholars start from the Stolper-Samuelson model, which suggests that respon-
dents should view trade and migration similarly. In many of these studies, the model
relies on a different set of factors of production than the canonical model; namely,
high-skill and low-skill labor, instead of capital and labor, as the main factors of produc-
tion. Nonetheless, the results are the same: in states that are high-skill labor abundant
(developed states), opening trade or migration should lead to an increase in wages for
 Margaret E. Peters    523

high-skill labor and a decrease in wages for low-skill labor. In states that are low-skill
labor abundant (developing states), opening trade or migration should lead to an
increase in wages for low-skill labor and a decrease in wages for high-skill labor.
Survey respondents should favor trade and/or immigration based on these effects,
and their responses on these two issue areas should be similar. In a country like the
United States, high-skill survey respondents should favor open trade and open immi-
gration for low-skill workers and immigration restrictions for high-skill workers.
Low-skill survey respondents should favor restrictions on trade and low-skill immigra-
tion and favor openness for high-skill immigration. However, the survey data do not
show this (Goldstein and Peters 2014; Hainmueller and Hiscox 2007, 2010; Hanson,
Scheve, and Slaughter 2007; Malhotra, Margalit, and Mo 2013; Mayda 2008). Instead,
high-skill respondents tend to like trade and immigration of both high- and low-skill
immigrants more than low-skill respondents. Further, over the entire population, trade,
even with developing countries, is usually greatly preferred to immigration, especially
low-skill immigration.
Scholars have posited several different reasons for this difference on opinions between
trade and immigration. Hanson, Scheve, and Slaughter (2007) argue that the difference
is caused by the fiscal effects of immigrants. Unlike goods, immigrants use the social
welfare system. Thus, high-skill natives may oppose low-skill immigration to a greater
degree than trade, because low-skill immigrants are more likely to use the social wel-
fare system, leading to an increase in taxes. Low-skill natives may also dislike low-skill
immigrants for their use of the social welfare system, as it may lead to crowding out.
Nonetheless, trade may also have these effects: increased trade may lead to fewer job
opportunities and increasing natives’ use of social welfare programs, leading to crowd-
ing out or increased taxation.
Mayda (2008) argues that the difference in views is driven by the difference in expo-
sure to trade and immigration in the labor market. She finds that respondents who work
in nontradable sectors are more pro-free trade than people who work in tradable sec-
tors. Because these respondents do not face job competition from trade, they view trade
more positively, as consumers who benefit from it. In contrast, immigrants can pene-
trate all sectors of the economy, which leads to the greater fear of labor market competi-
tion from immigrants and greater opposition to immigration. Yet immigration also does
not have an impact on most respondents. While in the United States the overall share
of the foreign born population is 12 percent, immigrants are disproportionately repre-
sented in a few industries, including agriculture, construction, textiles, and low-skill
services. Similar patterns of employment are found in European countries. Moreover,
many studies have found that immigration leads natives to up-skill, giving them more
economic opportunity than they had before (e.g., Strauss 2012). Many natives thus face
little labor market competition from immigration or trade, making their divergence on
these issues puzzling.
Hainmueller and Hiscox (2007, 2010) argue that it is probably differences in the cul-
tural costs of trade and immigration that lead to the differences in responses. They argue
that education, usually used to measure skill level, also measures tolerance. As part of
524   Trade and Migration

education, we are also taught tolerance; thus, those respondents who are more educated
will express more tolerance, or at least be unwilling to express intolerance. Thus it is the
cultural animus of the less educated that is driving their different preferences for immi-
gration and trade.
Recently, two studies have taken a more nuanced view. Malhotra, Margalit, and Mo
(2013) show that cultural animus and economic threat are both a factor. They find that
information technology (IT) workers are more likely to oppose immigration of other IT
workers but are no more opposed to immigration of other immigrants than other simi-
larly skilled natives. Goldstein and Peters (2014) argue that cultural animus can affect
the baseline preferences people have about immigration, but that shocks to economic
circumstances can lead to changes in opinion. They examine a six-year panel survey
conducted during the Great Recession to see how opinions change with large shocks to
the economic system. As expected, those who have suffered during the Great Recession
are more anti-immigrant than those who did not. Thus, it is likely that both cultural ani-
mus (or tolerance) and economic threat affect opinions on immigration and, to a lesser
extent, on trade.
In all, the puzzle of why survey respondents are more anti-immigrant than anti-trade
has not yet been resolved. Yet perhaps it does not matter what opinions respondents
hold on these issues if their opinions do not affect policy outcomes. While there is some
research on whether/when public opinion affects policy (Peters and Tahk 2011), there is
much room for more research on this topic as well.

Ties Between Trade and

Migration Policy

Researchers have also focused on the link between trade and migration policies in devel-
oped countries. Immigration and trade policy in developed countries have tended to
move in opposite directions: trade closure was paired with immigration openness prior
to World War II, and trade openness has been paired with immigration restrictions
since World War II. These policy combinations are somewhat surprising from a politi-
cal standpoint. The Stolper-Samuelson model shows that openness (closure) to trade
or immigration will lead to lower (higher) wages for low-skill labor and higher (lower)
wages for high-skill labor and higher (lower) returns to capital. Thus, if a policy maker
wants to protect native low-skill labor, she or he should restrict both trade and immigra-
tion. On the other hand, if the policy maker wants to help native high-skill labor and
capital, she or he could open either or both flows. By this logic, states’ choices of policy
should be idiosyncratic, or all should open in tandem, since opening any one flow would
lead to the same distributional consequences as opening the other. Yet empirically, trade
and immigration policy are rarely opened together, nor do states seem to choose these
policies idiosyncratically. To explain the patterns of policy, scholars have focused on
 Margaret E. Peters    525

three explanations: the rise of the welfare state, the extension of the franchise, and the
effects of trade on firms.
Hatton and Williamson (2005, 2007) focus on two explanations for the pattern
of trade and immigration openness: the rise of the welfare state and the expansion of
the franchise. The rise of the welfare state argument is similar to that of the fiscal costs
argument in the public opinion literature. Once states had welfare systems, immi-
grants became costly to society, as immigrants are (or are thought to be) more likely
than natives to use the welfare system. Natives opposed immigration more because they
feared having to pay higher taxes or face crowding out. There are two problems with this
argument. First, as noted above, trade has also caused much, if not more, job dislocation,
leading to increased use of the welfare system by natives. Thus, if natives are concerned
about the costs of the welfare system or crowding out, they should also oppose trade
openness. Second, there were concerns about immigrants’ use of the social welfare sys-
tem as early as the colonial period in the United States. Neuman (1993) argues that many
US states passed laws against immigration in part due to immigrants’ use of poor relief.
Thus, immigrants’ use of the welfare system was nothing new in the mid-twentieth cen-
tury, when many of the new restrictions on immigration were put into place and trade
was opened.
Hatton and Williamson’s (2005, 2007) second argument focuses on how immigration
increases inequality and how that, combined with the increase in the franchise, leads to
immigration restrictions. As the franchise expanded, those who competed directly with
immigrants were more likely to vote for immigration restrictions, and/or middle-class
voters did not like the increasing inequality from immigration and voted for immigra-
tion restrictions as well. Again, the problem with this argument is that trade also causes
inequality in developed countries and may have increased inequality to a greater degree
than immigration. Thus, if voters are worried about inequality, they should also have
been against free trade.
Hatton and Williamson’s (2005, 2007) research, like the public opinion literature,
is based on how individuals perceive trade and immigration; however, they apply a
median voter model to these opinions to say something about policy. Instead of focus-
ing on how trade and migration affect individuals, Peters (2014, 2015) examines how
trade and migration affect firms and in turn, how this affects firm lobbying on these two
issues.
Peters (2014, 2015) also uses Stolper-Samuelson as a starting point. In labor scarce
states, opening trade would lead to the export of capital-intensive goods and the import
of labor-intensive goods. Labor-intensive industries would then shrink in size as they
could not compete with the imported goods. As the Stolper-Samuelson model predicts,
these industries would lay off their workers, which would not be absorbed entirely by
the export industries. These layoffs would lead to a decrease in wages across society.
Overall, as the demand for labor in open labor-scarce states decreases, so too does the
incentive for firms to lobby for immigration. Some labor-intensive firms that had lob-
bied for open immigration no longer exist, having been forced to close due to trade pres-
sure. These firms lay off their workers, which decreases the wage that all other firms have
526   Trade and Migration

to pay. Existing firms may in general want increased immigration, but they have limited
resources to spend on lobbying; money (effort) spent on lobbying could always be spent
in other aspects of the business or spent lobbying on other, now more pressing issues.
As their labor costs decrease due to the layoffs caused by increased trade, firms decrease
their lobbying on immigration. Given the existence of groups that oppose immigration,
the decrease in lobbying allows policy makers to restrict immigration.
Similarly, closure to trade would lead to an increase in support for immigration and
immigration openness. Closure to trade increases the size of the import-competing sec-
tor, as consumers can no longer buy imported goods and instead turn to domestically
produced goods. The import-competing sector is labor intensive and will thus increase
the overall economy’s demand for labor, pushing up wages throughout the economy.
As demand for labor increases, so too should lobbying for immigration by firms, which
leads policy makers to open immigration (Peters 2014, 2015).
Immigration may also affect support for trade, at least in the short run. Immigration
makes import-competing firms in labor-scarce states more competitive, by driving
down labor costs. This may make import-competing firms more supportive of free
trade, as it is less likely that they will be forced out of business with trade openness
(Peters 2015).
While an open immigration policy may support free trade in the short run, free trade
will likely undermine open immigration in the long run. As trade opens further—either
through decreasing tariffs and nontariff barriers more or decreasing them on more
goods—immigration would have to open further to keep import-competing firms com-
petitive. In order to keep firms competitive as trade opens, policy makers would have
to open immigration enough so that, eventually, wages would equalize. As discussed
above, immigration is not politically popular for several reasons, which makes it hard
for policy makers to open immigration enough to keep firms competitive. At this point,
some firms will close, reducing the economy-wide need for labor and allowing the pol-
icy maker to restrict immigration (Peters 2015).
This experience of immigration policy supporting trade policy may help explain why
Western European countries were able to open their economies to trade after World
War II. Many of these countries, notably West Germany, France, and the Netherlands,
had guest worker programs that allowed their firms to hire thousands of foreign work-
ers. Many of these migrants worked in sectors that were relatively uncompetitive. Even
with the immigrant labor, many of these industries closed, as they could not compete as
trade opened further. At this point, support for immigration declined, and labor migra-
tion was largely curtailed throughout Europe (Peters 2015).
Thus the empirical regularity we see of open immigration, closed trade and open
trade, closed immigration is driven by firms. Firms are affected by trade, which in turn
affects their preferences for immigration. The level of trade openness affects the average
firm-level demand for immigration and the amount of lobbying by firms for immigra-
tion. Immigration policy may also affect trade policy by making relatively uncompeti-
tive firms more competitive, but it is unlikely that immigration will be able to support all
uncompetitive firms, especially as trade opens further.
 Margaret E. Peters    527

Trade and Migration Flows

Finally, we can consider how trade and migration flows affect one another. The
Stolper-Samuelson model predicts that increased trade flows should lead to less migra-
tion and vice versa. Increased trade flows should lead to equalization of wages across
borders, reducing much of the demand to migrate, and increased migration flows
should lead to the equalization of prices, reducing the demand to trade most goods. Yet
the data are inconsistent with the Stolper-Samuelson model to some extent. In the short
run, trade flows seem to lead to more migration, but in the long run trade flows do seem
to lead to equalization of wages and less migration. In contrast, so far the data suggest
that migration leads to increased trade in the short and long runs.
In the short run, trade flows affect the desire and ability of potential migrants to
emigrate by disrupting the traditional economy and removing the cost constraint on
migrating. Trade flows often disrupt the traditional economy. One recent example of
this was Mexico’s decision, due to its commitments under NAFTA, to reduce its tariffs
on US corn. This decision severely disrupted small-scale agriculture in Mexico, result-
ing in a loss of jobs for agricultural laborers (Bacon 2012). These laborers, then, were
forced to find employment elsewhere, including migrating to the United States.
This phenomenon has not been limited to Mexico. World systems scholars, most
notably Sassen (1988), have argued that increased globalization has led to increased
immigration. Trade, as in the Mexican case, leads to the displacement of workers who
were employed in the import-competing sector in developing countries. This increases
the benefits of migration, as there are fewer economic opportunities at home. Further,
increased trade also often comes with increased foreign direct investment in many
developing countries. These investments, often in factories, lead to “stepping-stone
migration.” Poor migrants from the countryside first move to cities within their country
(or a neighboring developing country) to work in the factories; once they have enough
money saved for the journey to a developed state, they migrate internationally. For
example, rural Mexicans have long migrated first to the factories on the border with the
United States and then to the United States, once they have saved enough money for the
journey (Fernandez-Kelly 1983).
In addition to disrupting the traditional economy, trade can also increase the
resources that potential migrants have for migrating. Lopez and Schiff (1998) focus on
how trade can release the cost constraint on migration. Migration is relatively costly for
most potential migrants. They have to have the funds to pay the cost of transportation to
the new location, the cost of a visa if they migrate legally or the cost of smuggling if they
migrate illegally, and the cost of supporting themselves both during the journey and
in their new location until they get a new job. Further, potential migrants often live in
countries that are credit constrained; as such, many migrants have to save for their jour-
ney before they can migrate. Trade helps to release the cost constraint by providing an
increase in income for low-wage workers in developing nations, which allows them to
528   Trade and Migration

save more and have a greater ability to migrate (Lopez and Schiff 1998). Thus, trade may
increase the desire to migrate by reducing job opportunities in the traditional economy
in developing countries as well as increasing the ability to migrate by releasing the cost
constraint.
In the long run, however, there is evidence that trade helps to decrease the incentives
to migrate, as predicted by the Stolper-Samuelson Model. Hatton and Williamson (1998,
2005) and O’Rourke and Williamson (1999) provide evidence that European wages
increased throughout the late nineteenth and early twentieth centuries, converging to
American and other New World country wages. Part of the convergence was driven by
migration, which drastically lowered unemployment and underemployment, and part
of it was driven by development of European economies, which in turn was driven in
part by trade. Thus, trade and migration helped lead to the convergence of wages in the
Atlantic economy.
Policy makers, using the Stolper-Samuelson model and the evidence of convergence
in wages in the Atlantic economy, have often sold free trade as a way to decrease immi-
gration. For example, it was argued by US policy makers that NAFTA would help curb
Mexican migration to the United States (Uchitelle 2007). While Mexican migration
increased directly after NAFTA, recently it has been falling and now is at net-zero lev-
els. Part of the reason for the decline in Mexican migration has been the poor state of
the US economy in recent years, but the other part has been the growth of the Mexican
economy, in part driven by increased trade. Similarly, the EU has always kept restric-
tions on freedom of movement from new, poorer member states longer than restrictions
on free trade. It was argued in the EU case that free trade (and free capital movement)
would help these new EU members develop, leading to less migration once the borders
were opened (Geddes and Money 2011). In the case of the EU, migration too has often
increased in the short term but has decreased over time as new members have become
more developed, thanks in part to trade.
Migration flows have also been found to increase trade flows (e.g., Dunlevy and
Hutchinson 1999; Dunlevy 2006; Felbermayr and Jung 2009; Felbermayr and Toubal
2012; Gould 1994; Hatzigeorgiou 2010; Head and Ries 1998; Herander and Saavedra
2005; Rauch and Trindade 2002). There are two main reasons why migration increases
trade:  preferences for home goods and increased information. Migrants bring with
them preferences for goods that are only produced at home. Once they migrate, they
increase demand for these goods, which increases imports in the country that they
migrated to and exports from their home country. This effect decreases as the migrant
network ages—the longer migrants have been in the immigrant-receiving country, the
less they import goods from home (Dunlevy and Hutchinson 1999). Instead they may
find other sources of them in the immigrant receiving country, create sources of these
products in the receiving country, or simply lose their taste for them.
The second, and likely the more important reason, is the informational mechanism.
Trade requires information about arbitrage opportunities, language, preferences in both
their home and host countries, business contacts, business norms and practices, and
the legal framework, as well as trust. Migrants thus can increase intra-industry trade
 Margaret E. Peters    529

by identifying varieties of goods from the home (host) country that would appeal to
consumers in the host (home country). Migrants can also increase trade more gener-
ally by making trade contracts more secure. Many trade contracts rely on a difference in
time (and space) between delivery of goods and payment. For trade to occur, merchants
in one country must find trustworthy partners in other countries. Migrant networks
have been able to overcome these informational and trust problems for a long time. For
example, Greif (2006) shows how, in the Middle Ages, the Maghrabi traders, who were
migrants, used their networks to facilitate trade and punish those who did not abide
by their contracts. Similarly, Rauch and Trindade (2002) argue that ethnic Chinese
migrants play a similar role today. Further, several empirical studies show a robust link
between migration flows and trade flow creation (Felbermayr and Toubal 2012; Gould
1994; Head and Ries 1998; Rauch and Trindade 2002).
Migrant networks are especially important in facilitating trade in situations where
enforcement of contracts is likely to be weak or where more knowledge about goods
is needed. For example, Dunlevy (2006) finds that the effect of migration is greater for
those countries with relatively high levels of corruption. Migrant networks can help cir-
cumvent the problems associated with corrupt governments by providing information
about the business environment and providing trustworthy trading partners. Rauch
and Trindade (2002) find that the effect of migration networks on trade is stronger for
differentiated goods. It is relatively easy to gain knowledge about prices and demand
for homogeneous goods, such as raw materials. It is more difficult to gain informa-
tion about prices and demand for differentiated goods, as they typically have a smaller
market. Thus migrant networks help fill in these informational gaps and increase trust
between trading partners, leading to increased trade.

Conclusion

Trade and migration have long been linked. Since the Middle Ages, migrants have often
been traders because they were the ones who had information about market opportuni-
ties and had the trustworthy contacts to make trade over long distances possible. Within
the study of international trade, trade and migration have also long been linked through
one of the canonical trade models: the Stolper-Samuelson model. While there are many
reasons the Stolper-Samuelson model may not hold, it has provided scholars with a use-
ful framework from which to examine the ties between trade and migration.
In the last two decades scholars have begun to examine the relationships between
trade and migration empirically and with new formal models. They have focused on
three main areas of study: pubic opinion, policy, and flows. The literature has started to
make headway, at least in establishing some empirical regularities: trade is preferred by
most survey respondents to immigration; trade policy and immigration policy appear
to be substitutes, with only one open at any given time in labor-scarce countries; trade
flows at first increase but later decrease migration; and migration flows increase trade
530   Trade and Migration

flows. As discussed above, scholars have given many different reasons for these empiri-
cal regularities but have also converged on a few causal mechanisms.
The results on public opinion have engendered some of the most scholarly debate.
Scholars have posited that it is the fiscal effects of migrants, their effects on the labor
market, and their cultural effects that drive the relationship between opinions on trade
and those on migration. Recent studies focus on teasing out these effects. Malhotra,
Margalit, and Mo (2013) use a carefully designed sample with oversampling of technol-
ogy workers to show that cultural animus against migration often serves as a baseline for
opinions on immigration, but that direct labor market competition can lead to intense
anti-immigrant sentiment. Similarly, Goldstein and Peters (2014), using a panel survey
during the Great Recession and the recovery, show that respondents also have a baseline
of cultural animus toward immigrants but that their support for immigration is affected
by their economic situation. Finally, Helbling and Kriesi (2014) use carefully designed
survey experiments to tease out how exactly the fiscal effect and cultural animus may
play out in different institutional environments.
More research on the differences in public opinion on trade and on migration could
be done. For example, instead of examining population averages, scholars could exam-
ine more closely which respondents favor (oppose) trade and migration and which have
differing opinions. What differences are there between these two groups? Further, like
Helbling and Kriesi (2014), they could undertake more careful analysis of the different
hypotheses in various institutional environments.
In addition to these empirical puzzles, scholars should examine why the dislike of
foreigners affects immigration more than it affects trade, why some groups are more dis-
liked than others, and why the dislike of foreigners is more salient at some times than at
others. For example, we know that American survey respondents dislike both Chinese
trade and Mexican immigration (Goldstein and Peters 2014), yet the effect of culture
seems to influence overall levels of support for immigration more than trade. Why
does nativism have a greater effect on immigration than nationalism does on trade?
Similarly, we know that respondents in the United States currently dislike Indian immi-
gration more than they dislike German immigration, and Chinese trade is more dis-
liked than German trade (Goldstein and Peters 2014). Yet while there are more Indian
immigrants, both Indian and German immigrants pose a threat to high-skill natives,
and both Germany and China have been running large trade surpluses at the expense
of the United States. So why then do Americans dislike Indian immigrants or Chinese
goods more than German immigrants or goods? Is it just racism, or is there an eco-
nomic logic to the opposition to these flows? Finally, how do we explain the salience
of cultural threat? For example, in the early 1900s an increase in nativism helped lead
to passage of the Literacy Act and the 1921 and 1924 quota acts in the United States and
anti-Asian immigrant acts in Australia and Canada, and increased nationalism led to
beggar-thy-neighbor trade and exchange rate policies during the Great Depression.
Increased nativism and nationalism during the Great Recession has not had the same
effect on either migration or trade. Why was nativism/ nationalism more effective in the
past than today? Is it the case that the public is less nativist/ nationalist, or that norms
 Margaret E. Peters    531

against nativism/ nationalism prevent policy makers from using it as a way to frame the
debate (see Freeman 1995), or are their other factors that have made nativism/ national-
ism less salient?
Moreover, there is a missing link between opinion and policy that public opinion
scholars should examine. Peters and Tahk (2011) offer one of the few studies to examine
whether public opinion translates to policy, finding little link between opinion on immi-
gration and policy. However, they only examine the United States and Canada. It could
be that these two countries are relatively special institutional environments and that
opinion links to policy in other countries. However, we do not know. Scholars should
examine whether opinion affects policy, as well as why or why not.
While scholars have found that trade and migration policy are substitutes in
labor-scarce countries, little research has been done on whether trade and emigration
are substitutes or complements in labor-abundant countries. Many developing coun-
tries, for example the former Communist countries but also Japan before 1868, kept
both trade and emigration relatively restricted. Some countries, like India and Mexico
until recently, tacitly or explicitly encouraged emigration while maintaining trade bar-
riers. Others, like China, have opened trade but discouraged emigration, especially
high-skill emigration. Finally, some countries, like South Korea in the 1960s and 1970s,
encouraged both trade and emigration. What explains these divergent patterns? If a
labor-abundant country keeps trade closed, it may restrict emigration to keep a cheap
supply of labor for capitalists or for security reasons. On the other hand, it may encour-
age emigration to decrease unemployment or serve as a safety valve against revolution.
Similarly if a labor-abundant country opens trade, it may restrict emigration as a way to
keep its low labor costs or may encourage emigration as another way to help the country
develop or as a way to decrease unrest during times of trade-related economic disloca-
tion. Thus, it is an open puzzle how trade and emigration policy might affect each other
in developing countries.
Finally, the scholarship on migration and trade flows has tended to focus on indi-
vidual market actors and not on how political leaders might respond to those market
actors. Yet if trade and migrant flows affect each other, we should expect that there
would be some political ramifications from these effects. For example, it appears that
Mexico used the issue of Mexican migration to the United States to help gain support for
NAFTA; do other emigrant-sending states use the issue of migration to lobby for pref-
erential trade access in immigrant- receiving states? Are they successful? What about
lobbying by migrants themselves? Bermeo and Leblang (2012) find that migrants are
relatively successful in lobbying for foreign aid. Are they also successful in lobbying for
trade openness with their home country? Or would this openness cut at their advantage
over other groups, making them less likely to lobby? In general, how do migrant net-
works protect their advantages in the world economy, if they do at all?
Stolper and Samuelson long ago argued that trade and migration are linked, yet this
scholarship was largely ignored in favor of studying trade and migration separately. In
large part this was because when the field of IPE began in the 1960s and 1970s, trade
was relatively open—at least in comparison to the interwar years—whereas migration
532   Trade and Migration

was relatively restricted, resulting in low levels of migration. Globalization, then, was
defined as the movement of goods and capital, but not people. In contrast, the nine-
teenth century was a time of high levels of migration as well as trade and capital move-
ments; globalization, as defined in this period by Keynes (1919), was the movement of
goods, capital, and people. Today, however, migration is becoming a more important
topic in the policy world and in scholarship as well. Scholars have begun studying the
ties between trade and migration, returning to the Stolper-Samuelson model and exam-
ining when and where it holds. This line of inquiry is still in its infancy but promises to
be a fruitful one for scholars.

References
Bacon, David. 2012. How US Policies Fueled Mexico’s Great Migration. The Nation, January 23.
Bermeo, Sarah Blodgett, and David A. Leblang. Forthcoming. Migration and Foreign Aid.
International Organization.
Dunlevy, James A. 2006. The Influence of Corruption and Language on the Pro-trade Effect
of Immigrants: Evidence from the American States. Review of Economics and Statistics 88
(1): 182–186.
Dunlevy, James A., and William K. Hutchinson. 1999. The Impact of Immigration on American
Import Trade in the Late Nineteenth and Early Twentieth Centuries. Journal of Economic
History 59 (4): 1043–1062.
Felbermayr, Gabriel J., and Benjamin Jung. 2009. The Pro-trade Effect of the Brain
Drain: Sorting Out Confounding Factors. Economic Letters 104: 72–75.
Felbermayr, Gabriel J., and Farid Toubal. 2012. Revisiting the Trade-Migration Nexus: Evidence
from New OECD Data. World Development 40 (5): 928–937.
Fernandez-Kelly, Maria Patricia. 1983. Mexican Border Industrialization, Female Labor Force
Participation and Migration. In Women, Men, and the International Division of Labor, edited
by June C. Nash and Fernandez-Kelly, Maria Patricia, 205–223. Albany: State University of
New York Press.
Freeman, Gary P. 1995. Modes of Immigration Politics in Liberal Democratic States.
International Migration Review 29 (4): 881–902.
Geddes, Andrew, and Jeannette Money. 2011. Mobility Within the European Union. In
Migration, Nation States, and International Cooperation, edited by Randall Hansen and Jobst
Koehler, 31–43. New York: Routledge.
Goldstein, Judith L., and Margaret E. Peters. 2014. Nativism or Economic Threat: Attitudes
Toward Immigrants During the Great Recession. International Interactions. 40 (3): 1–26.
Gould, David M. 1994. Immigrant Links to the Home Country: Empirical Implications for U.S.
Bilateral Trade Flows. Review of Economics and Statistics 76 (2): 302–316.
Greif, Avner. 2006. Institutions and the Path to the Modern Economy: Lessons from Medieval
Trade. Cambridge, UK: Cambridge University Press.
Hainmueller, Jens, and Michael J. Hiscox. 2010. Attitudes Toward Highly Skilled and
Low-skilled Immigration: Evidence from a Survey Experiment. American Political Science
Review: 1–24.
Hainmueller, Jens, and Michael J. Hiscox. 2007. Educated Preferences: Explaining Attitudes
Toward Immigration in Europe. International Organization 61 (2): 399–442.
 Margaret E. Peters    533

Hanson, Gordon H., Kenneth Scheve, and Matthew J. Slaughter. 2007. Public Finance and
Individual Preferences Over Globalization Strategies. Economics & Politics 19 (1): 1–33.
Hatton, Timothy J., and Jeffrey G. Williamson. 1998. The Age of Mass Migration: Causes and
Economic Impact. New York: Oxford University Press.
Hatton, Timothy J., and Jeffrey G. Williamson. 2007. “A Dual Policy Paradox:  Why Have
Trade and Immigration Policies Always Differed in Labor-Scarce Economies?” In The New
Comparative Economic History, edited by Hatton, Timothy J., Kevin H. O’Rourke, and Alan
M. Taylor, Chapter 9, 217–240. Cambridge, MA: MIT Press.
Hatton, Timothy J., and Jeffrey G. Williamson. 2005. Global Migration and the World Economy.
Cambridge, MA: MIT Press.
Hatzigeorgiou, Andreas. 2010. Migration as Trade Facilitation: Assessing the Links Between
International Trade and Migration. B.E. Journal of Economic Analysis & Policy 10
(1): Article 24.
Head, Keith, and John Ries. 1998. Immigration and Trade Creation: Econometric Evidence from
Canada. The Canadian Journal of Economics/Revue Canadienne d’Economique 31 (1): 47–62.
Helbling, Marc, and Hanspeter Kriesi. 2014. Why citizens prefer high-over low-skilled immi-
grants. Labor market competition, welfare state and deservingness”, European Sociological
Review. (forthcoming).
Herander, Mark G., and Luz A. Saavedra. 2005. Exports and the Structure of Immigrant-Based
Networks:  The Role of Geographic Proximity. Review of Economics and Statistics 87
(2): 323–335.
Keohane, Robert O., and Helen V. Milner. 1996. Internationalization and Domestic Politics.
Cambridge, UK: Cambridge University Press.
Keynes, John Maynard. 1919. The Economic Consequences of the Peace. New York: Harcourt,
Brace and Howe.
Lake, David A. 2009. Open Economy Politics:  A  Critical Review. Review of International
Organizations 4: 219–244.
Lopez, Ramon, and Maurice Schiff. 1998. Migration and the Skill Composition of the Labour
Force: The Impact of Trade Liberalization in LDCs. Canadian Journal of Economics/Revue
Canadienne d’Economique 31 (2): 318–336.
Malhotra, Neil, Yotam Margalit, and Cecilia Mo. 2013. Economic Explanations for Opposition
to Immigration:  Distinguishing Between Prevalence and Conditional Impact. American
Journal of Political Science 57 (2): 391–410.
Markusen, James R. 1983. Factor Movements and Commodity Trade as Complements. Jounral
of International Economics 14 (3–4): 341–356.
Mayda, Anna Maria. 2008. Why Are People More Pro-trade Than Pro-migration? Economic
Letters 101: 160–163.
Melitz, Marc J. 2003. The Impact of Trade on Intra-Industry Reallocations and Aggregate
Industry Productivity. Econometrica 71 (6): 1695–1725.
Mundell, Robert A. 1957. International Trade and Factor Mobility. American Economic Review
47 (3): 321–335.
Neuman, Gerald L. 1993. The Lost Century of American Immigration Law (1776–1875).
Columbia Law Review 93 (8): 1833–1901.
Oatley, Thomas. 2011. The Reductionist Gamble:  Open Economy Politics in the Global
Economy. International Organization 65 (2): 311–341.
O’Rourke, Kevin H., and Jeffrey G. Williamson. 1999. Globalization and History: The Evolution
of a Nineteenth-Century Atlantic Economy. Cambridge, MA: MIT Press.
534   Trade and Migration

Peters, Margaret E. 2015. Open Trade, Closed Borders: Immigration in the Era of Globalization.
World Politics. 67 (1). 114-154.
Peters, Margaret E. 2014. “Trade, Foreign Direct Investment and Immigration Policy Making
in the US.” International Organization 68 (4): 811–844.
Peters, Margaret E., and Alexander M. Tahk. 2011. Are Policy Makers Out of Touch with
Their Constituencies When It Comes to Immigration? Working Paper. University of
Wisconsin—Madison and Yale University.
Rauch, James E., and Vitor Trindade. 2002. Ethnic Chinese Networks in International Trade.
Review of Economics and Statistics 84 (1): 116–130.
Samuelson, Paul A. 1948. International Trade and the Equalisation of Factor Prices. The
Economic Journal 58 (230): 163–184.
Sassen, Saskia. 1988. The Mobility of Labor and Capital: A Study in International Investment and
Labor Flow. Cambridge, UK: Cambridge University Press.
Strauss, Jack. 2012. Does Immigration, Particularly Increases in Latinos, Affect African
American Wages, Unemployment and Incarceration Rates? Working Paper. University of
Denver—Reiman School of Finance.
Uchitelle, Louis. 2007. NAFTA Should Have Stopped Illegal Immigration, Right? New York
Times, February 18, The Nation section. http://www.nytimes.com/2007/02/18/weekinreview/
18uchitelle.html.
 Index

Aaronson, S. A.,╇ 14–15, 508–9 Anderson, J.,╇ 81, 93n8

Abouharb, M. R.,╇ 508–9 Anderson, K.,╇ 162, 172n9
Acemoglu, D.,╇ 272, 273, 301 Angell, N.,╇ 422
Acer,╇239 Anglo-French commercial treaty of╇ 1786,
Act for the Prevention and Suppression of 42–43, 47
Combinations Formed in Restraint of Antidumping laws,╇ 60, 61t, 67–69, 72n2, 73n5,
Trade of╇ 1889, 214 161–62, 392
Afghanistan,╇392 Antitrust/trade policies interactions
Africa adoption/content/stringency,╇220–2221,
antidumping laws, 61t 220f, 221t, 228n9
antitrust enforcement,╇ 222 aggregate national welfare variant,╇ 217–19
apartheid regime sanctions,╇ 499 anticompetitive behavior patterns,╇ 222–23
Cotonou Agreement,╇ 505 competition patterns,╇ 222–23
democracy formation by country,╇ 266t complementarity model,╇ 223–25, 225f,
democratization/trade reforms,╇ 264 228nn13–15
pollution effects,╇ 443 domestic political economy variant,╇ 218–19
regional monetary integration,╇ 461 empirical studies,╇ 219–23, 220f, 221t,
RTAs,╇ 371, 373–74 228nn9–11
transportation/efficiency relationships,╇ 442 enforcement,╇ 218, 222–25, 225f, 228nn10–11
WTO dispute participation,╇ 411 free trade as substitute,╇ 215–17, 227nn5–6
African Growth and Opportunity Act,╇ 62, 483 history,╇ 213–15, 227nn2–4
Agreements on Sanitary and Phytosanitary optimal tariff theory,╇ 218
Measures (SPS),╇ 401, 406, 448 ordo-liberalism,╇ 216, 226, 227n6
Agricultural protection,╇ 44, 111, 162–63, overview,╇ 13, 213, 226–27
172nn8–9 protectionism╇ vs., 217–19, 228nn7–8
Ahlquist, J. S.,╇ 105 Antràs, P.,╇ 70
Aitken, B.,╇ 244 Apodaca, C.,╇ 502
Albania,╇266t Apple,╇235
Allee, T.,╇ 385, 391 Aquinas, Thomas,╇ 38–39
Alliance Treaties and Obligations (ATOP),╇ 29 Aquino, A.,╇ 192n6
Alt, J. E.,╇ 160, 171 Ardanaz, M.,╇ 110, 321
Alvarez, M.,╇ 299 Ardelean, A.,╇ 282
Ambroziak, L.,╇ 247 Argentina
American National Electoral Study antidumping laws,╇ 60, 61t
(ANES),╇ 100, 103t currency manipulation,╇ 469n3
American Recovery and Reinvestment Act democracy formation,╇ 266t
of╇ 2009, 339 democratization/trade reforms,╇ 259–60, 262
Amsden, A. H.,╇ 478–79 financial crises,╇ 469n3
Andean Community,╇ 338 framing experiments,╇ 110, 114n3
536   Index

Argentina (Cont.) Baccini, L.,  139, 366, 483

free trade areas (FTAs),  62 Bagwell, P.,  64, 66, 67, 71
global production networks,  250 Baier, S.,  32n13, 80, 84, 85, 90, 94n14
government export promotion,  478–79 Bair, J.,  13
policy preference formation,  321 Baker, A.,  101, 120
trade shares in GDP,  476 Balassa, B.,  460
Asia-Europe Survey (ASES),  106 Baldwin, R. E.,  57, 72n4, 82, 162, 168, 170, 187,
Association of South-East Asian Nations 239, 241, 367
(ASEAN),  340, 482 Bandyopadhyay, U.,  67, 167
Athukorala, P.,  376n3 Bangladesh,  260, 266t, 502–3
Atlantic Charter,  23 Bannerman, G.,  12
Australia Barbieri, K.,  426, 428, 429
antidumping laws,  125 Barigozzi, M.,  386
human rights in trade agreements,  495, 510 Barkin, J. S.,  14
MFN tariff studies,  66 Barsoom, P. N.,  372, 389
privacy laws,  507, 508 Bartels, L.,  505
PTAs,  352, 352f Battle in Seattle,  504, 512n15
WTO dispute resolution,  401, 406, 407 Bauer, R. A.,  107, 197, 319
Authoritarian regimes Beaulieu, E.,  106, 165, 172n13
access point theory,  306 Bechtel, M.,  148
coalition size,  302–6 Beck, N.,  429
concepts, definitions,  299 Benefits targeting,  170–71, 173nn17–18,
coups vulnerability,  309–10 284–85, 294n3, 480
dispute settlement procedure (DSP),  145 Benin, 266t
economic structure,  299–302, 312nn6–7 Bergstrand, J. H.,  32n13, 80, 84, 85,
foreign direct investment,  132, 307, 308 90, 94n14
income inequality,  301 Bermauer, T.,  152n8
leader change effects,  145–46 Bermeo, S. B.,  152n7
leader mode of entry,  309–10 Bernard, A. B.,  198, 201
leader survival effects,  149–50 Bhagwati, J.,  148
lobbying, 122 Bhala, R.,  400
median voter model,  300–302 Binatli, A. O.,  247
military,  304, 306, 308 Blackhurst, R.,  216
monarchies,  304, 306, 308 Blainey, G.,  420
personalist,  304, 305, 308 Blair, Hugh,  42
PFS model,  300 Blanton, R.,  501, 504
power fragmentation,  305–6 Blanton, S.,  501, 504
property rights,  307 Blinder, A.,  130
protectionism, industry-level,  301–2 Blonigen, B. A.,  69
protectionism  vs. free trade, 305–6 Bobick, T.,  145
redistribution, 301 Boix, C.,  301
single-party, 304–7 Bombardini, M.,  169
subtypes, 304–5 Borga, M.,  233
tariffs,  298, 305, 306, 310 Borrus, M.,  238
time horizons,  307–9 Bown, C. P.,  12, 67, 69, 148
trade policy generally,  13–14, 263, 298–99, 311 Bradford, A.,  220, 223–25, 228n11
trade reforms,  262–63 Bradford, S.,  123, 124
 Index   537

Brazil protectionism, industry-level,  172n13

antidumping laws,  60, 61t WTO dispute resolution,  401, 406
currency manipulation,  463, 467, 469n3 WTO membership benefits,  386
democratization/trade reforms,  259–60, Cao, X.,  502
262, 264 Cape Verde,  266t
electoral system,  304 Cardenas, Lazaro,  242
foreign direct investment,  242, 245, 488 Carey, J. M.,  303
free trade areas (FTAs),  62 Caribbean Basin Trade Partnership Act
globalization/inequality relationships,  132 of  2000, 243
global production networks,  250 CARIFORUM, 352
government export promotion,  478–79 Carnegie, A.,  386
human rights in trade agreements, Cassing, J.,  121
495, 500t Ceccoli, S. J.,  104
import-substituting industrialization Central African Republic,  266t
(ISI), 488 Central America Free Trade Agreement
MFN tariffs,  60 (CAFTA),  102, 108, 113, 244, 250, 321
PTAs, 339 Central and Eastern Europe (CEE).  see
RTAs,  11, 62, 371 Europe, European Union
specific trade concerns (STCs),  406–7 Chaffour, J. P.,  373
tariffs,  60, 259–60, 271, 276n10 Chang, P. -L.,  386
technical barriers to trade (TBTs),  407 Chase, K. A.,  14, 121, 129, 322, 368
WTO dispute resolution,  407, 411, 449–50 Chaudoin, S.,  148, 388
Bretton Woods Agreements,  392 Chauffour, J. P.,  365
Broda, C.,  65–66 Chile, 266t, 294n8, 352, 441, 496, 497t, 510
Broude, T.,  504 China
Broz, L.,  464 antidumping laws,  60, 61t
Budetta, M.,  364, 372 economic structure,  300
Bueno de Mesquita, B.,  302, 303, 311 export-processing zones,  488
Buhr, R.,  142 free trade  vs. protectionism, 103–4
Bulgaria, 266t globalization/inequality
Burgoon, B.,  105 relationships, 132
Busch, M. L.,  14, 121–23, 143, 146, 152n10, 263, global production networks,  239–41
285, 318, 325, 326, 331n9, 387, 407 hostile foreign policies,  433–34
Bush, George W.,  126, 321 human rights in trade agreements,  510
Büthe, T.,  13, 220, 223–25, 228n11 immigration policy,  530
international cooperation in trade
CAFTA,  102, 108, 113, 244, 250, 321 agreements, 452
Callaway, R.,  502 Internet censorship,  509
Canada priming experiments,  111–12
antidumping laws,  61t, 125 RTAs, 371
antitrust law,  214, 222, 227n2 trade agreements generally,  109, 488
electoral systems,  294n10 trade models,  6
GATT (see GATT) trade policy historically,  4–5, 13–14
human rights in trade agreements,  495–96, Chow, W. M.,  309, 310
497–98t Chrysler, 240
MFN tariff studies,  66 CIRI, 503
privacy laws,  507–8 Clague, C.,  307
538   Index

Clayton, A. B.,  105 income inequality,  272–74, 277n18

Cobden, R.,  39, 41, 44, 46, 47, 50 individual attitudes,  113, 261, 276nn2–3
Collective action.  see lobbying institutional effects,  123–24, 261–63
Commission for Environmental Cooperation intra-industry trade,  188–90
(CEC),  446, 448 labor mobility,  269–71, 271f, 277n16
Common external tariff (CET),  361 leader change effects,  145–46
Common Fisheries Policy,  447 leader mode of entry,  309–10
Competition policy.  see antitrust/trade leader survival effects,  149–50
policies interactions lobbying,  122, 133, 261
Convention for the Conservation of media ownership, marketing effects,  108
Antarctic Marine Living Resources multinational corporations,  235–36
(CCAMLR), 450 nontariff barriers,  259, 263, 275
Convention on International Trade in political affiliation,  263
Endangered Species of Wild Fauna and post-transition governments,  260, 263–64,
Flora (CITES),  450 269–72, 271f, 276n5
Cooper, M. C.,  508 protectionism, industry-level,  162, 172n4,
Cooper, S.,  461 172n14, 259–60, 275, 301–2
Copeland, B. R.,  65 PTAs,  77, 108–9, 140–50
Copelovitch, M. S.,  14, 391–92, 462, 463 reg3hdfe assessment,  85, 86t, 93n9
Corn Laws,  3, 39, 43, 44, 46, 320, 324 skilled  vs. low skilled tariffs, 272,
Cotonou Agreement,  505 276nn9–10
Cox, G. W.,  290 supply-side studies,  261–62
Crowley, M. A.,  67 tariffs,  259–60, 262–72, 265f, 266–67t, 271f,
Currency Reform for Fair Trade Act,  466 275, 276nn6–14, 277n15
Cusumano, M. A.,  235 time horizons,  308
Czechoslovakia, 266t, 392 trade policy generally,  9, 13, 259–60,
274–75, 276n1
Davis, C. L.,  14, 144m 153n13, 145, 152n7, 387, trade reforms,  262–63
388, 389, 392 Deng Xiaoping,  237
De Benedictis, L.,  386 Design of Trade Agreements (DESTA)
de la Court, Pieter,  41 project, 365
De Melo, J.,  365 de Sola Poole, I.,  107, 197, 319
De Mesquita, B. B.,  284 Developing countries
Democratic regimes democratic regimes,  263–64, 271, 273–75,
coalition size,  302–3 276n5, 276n10, 303
collectivism,  49–50, 52–53, 262 development-trade relationships,  475–88
demand-side studies,  261, 269 dispute settlement procedure,  387–88,
developed  vs. developing countries, 400–401, 407–11, 414nn5–6
263–64, 271, 273–75, 276n5, 276n10, 303 environment-trade relationships,  441, 451
dispute settlement procedure (DSP),  142– GATT effects,  26–27
50, 188 globalization in,  132
electoral systems,  113, 261, 272, 276n1 human rights-trade relationships,  511
foreign direct investment,  132, 235 immigration effects,  527–28, 531
free trade,  49–50, 52–53, 259, 262–63, protectionism, industry-level,  161–62,
272–74, 277n18 172n12
global trends,  262, 276n4 WTO membership,  387–88
human rights,  500 Development-trade relationships
 Index   539

balance-of-payments deficits,  478, 479 design variation,  142–43, 152n6

benefits targeting,  170–71, 173nn17–18, DESTA data set,  139, 140f
284–85, 294n3, 480 developing countries,  387–88, 400–401,
dependency, 484–85 407–11, 414nn5–6
developing countries,  475–88 early settlement,  146
economic policy,  475, 477–82, 489nn1–4 effectiveness, 388
exchange rates,  480 enforcement,  148, 386–89, 402–4, 408–10
export-processing zones,  486–87 environmental policy,  446, 448
foreign direct investment,  478, 479, 483, feedback effect,  147
486–88, 489n2 fire alarm effect,  147
free trade,  476–77, 485–86 flexibility provisions,  141–42, 152nn3–5
global value chains,  235, 238, 480, 481, functioning, outcomes,  143–47
486–87 information, 149–50
government export promotion,  478–79, institutional bias,  408
488 intra-industry trade,  188
GSP, 483–85 judicial economy,  402
import regulation,  478 leader survival,  149–50
import-substituting industrialization leader turnover,  145–46
(ISI),  4, 242, 425, 477–79 , 487–88, 489n1 legal capacity,  400, 411
industrialization strategy,  475, 477–82, literature review,  402–5
489nn1–4 litigation,  403, 407
intellectual property,  478 overview,  138–39, 150–51, 400, 413
interdependence, asymmetrical,  485, 487 participation,  401, 407–11, 414nn5–6
market access,  483, 487–88 political uncertainty,  140–42, 152nn2–5
most-favored nation (MFN),  481–84 post-adjudicative bargaining,  146–48
overview,  14, 475–77 retaliation, 404–5
political organization, mobilization,  486 rulings,  401–2, 411–13
PTAs,  475, 482–87, 489nn5–8 sanitary, phytosanitary (SPS)
resource transfer,  479–80 measures,  401, 406, 448
state interventions,  478–79 selection bias,  144–45, 152nn8–11, 401, 405,
tariffs,  476, 480, 481, 484 408
Washington Consensus,  480, 487 settlement,  146–47, 400, 403–4, 409–11
welfare state,  476–77 specific trade concerns (STCs),  152n10, 401,
Dexter, L. A.,  107, 197, 319 405–8
Dispute settlement procedure (DSP).  see also technical barriers to trade (TBTs),  401,
preferential trade agreements (PTAs) 405–8
adjudication,  144–47, 152n7, 354, 452 third parties,  403, 409–11
appeals, 404 trade policy effects,  148–49
Appellate Body,  404 transparency,  401, 409
compliance,  148, 152n11, 353–54 turtle-excluder devices (TEDs),  448–50
concessions, 403–4 unfair eviction,  147
condemnation,  404, 410 voter actions,  149–50
consultations, 402–3 words,  401–2, 411–13
decentralization, 402 Dispute Settlement Understanding
democracy, development,  143–44 (DSU),  139, 150–51, 152n1, 374, 387, 400,
democratic regimes,  142–50, 188 408, 412
design,  139, 140f, 152n1, 342, 344 Dixit, A.,  126, 198, 216
540   Index

Domestic geography Drope, J.,  125

background,  317–19 330nn1–3, 331n4 DSP.  see dispute settlement procedure (DSP)
checkerboard problem,  318, 325 Dür, A.,  139, 366
close-group hypothesis,  325, 327 Dutt, P.,  122, 172n12, 301, 312n7, 385
clustering,  319, 322
coalition-building,  329, 331n8 Eaton, J.,  198
collective action,  323–26, 329, 331nn6–7 Egger, H.,  128
dartboard approach,  318 Ehrlich, S.,  106, 107, 108, 126, 290, 295n14,
dispersed-group hypothesis,  325, 327 306, 311
electoral systems,  325–26, 328 Eichengreen, B.,  23, 463
empirical analysis,  318, 331n5 Eicher, T. S.,  386
employment concentration,  328 Electoral systems
first nature,  317 alternative measures,  292
geographic central concentration,  318–19, benefits targeting,  284–85, 294n3
331n4 causal mechanisms,  287–93, 294nn8–11,
lobbying,  322, 324, 329, 331n5 295n14
location Gini,  318 concepts, definitions,  281
low  vs. high skilled workers, 322 constituency size,  303
modifiable areal unit problem,  331n4 consumer prices,  284
nontariff barriers,  325, 326 democratization,  113, 261, 272, 276n1
offshoring, 322 district size,  287–89, 294n8
partisan effects,  329, 331n8 domestic geography,  325–26, 328
policy pressures,  316, 319 electoral competition,  289–90
political dispersion,  328 import licenses,  285
political representation,  327–30, 331nn8–10 individual attitudes,  107–8
political system responsiveness,  328 institutional bundles,  293
preference formation,  319–23, 331n5 institutions, interests,  291–93, 294n13
redistribution, 329–30 majoritarian,  107–8, 280–86, 291–92,
research applications,  316–17, 330, 331n11 294n13
second nature,  317–18, 330nn1–2 mixed,  286–87, 294nn5–6
spatial component studies,  323, 326 nontariff barriers,  282, 284
tariffs,  320, 323, 324, 329, 331n10 party strength,  290–91, 294nn9–11
Domestic political institutions personal votes,  289–90
electoral systems,  291–93, 294n13 policy instrument selection,  285
exchange rates,  459–61, 465, 468 proportional,  280, 283–86, 291–92, 294n2,
human rights-trade relationships,  494–95 294n5
most-favored nation (MFN) status,  10, 21 protectionist bias,  281–83, 294n1
PTAs (see preferential trade agreements seat-vote elasticity,  283
(PTAs)) single-member district plurality system
reciprocity, 10 (SMDP), 281
RTAs,  366–68, 375, 489n3 subsidies,  71, 282–83, 285, 480
trade policy generally,  9–10 tariffs, 282–84
Dominican Republic-Central American Free time horizons,  308
Trade Agreement (DR-CAFTA),  485 trade policy generally,  280, 293–94
Downs, G. W.,  372, 389 winning coalitions,  284
Doyle, M. W,  422 WTO disputes,  283, 288, 294n3
Dreher, A.,  386 Elliott, K. A.,  503
 Index   541

Ellison, G.,  318 Germany (see Germany)

Elsig, M.,  139, 366, 411 global production networks,  245–50
Embedded liberalism,  107, 126 GSP+ programs,  483–84
Engel, C.,  463 human rights in trade agreements,  495–96,
Environment-trade relationships 497–98t, 510
consumer power,  441 immigration policy,  526, 528, 530
developing countries,  441, 451 income inequality/unemployment,  128
economic growth limits,  441, 444 MFN import tariffs,  66, 67, 484
efficiency gains,  440–42, 444 multilateral tariff reductions,  68, 73nn12–13
environmental exceptions,  447–49 multinational corporations,  239, 240
food miles,  441–42 nontariff barriers,  207, 325
free trade,  443–44 party organization,  304
global environmental commons policy preference formation,  319–21
protection, 449–53 political system responsiveness,  328
governance issues,  443–45, 451–53 privacy laws,  507
illegal trade,  444–45, 450 PTAs,  352–53, 352f, 483
international cooperation,  452–53 regional integration,  461
international trade institutions,  445–51 as RTA,  11
local  vs. global effects,  440, 442 RTAs,  360, 366, 367, 370–71, 461
market access,  445 tariffs, 321
multilateral environmental agreements trade agreement institutionalization,
(MEAs),  439, 450–52 341–42, 447
multilateralism, 448–51 trade policy historically,  3–4, 319–21
NAFTA, 443–48 WTO dispute resolution,  412–13
overview,  14, 439–40, 453–54 WTO membership benefits,  386
ozone-depleting chemicals,  450 European Free Trade Agreement
pollution,  440–44, 450, 451 (EFTA), 497–98t, 506
regulatory effects,  443–45, 452–53 Evans, C. L.,  282, 284
resource use,  440, 442–43 Exchange rates
transportation, 441–42 blunt instrument problem,  467–68
turtle-excluder devices (TEDs),  448–50 bureaucratic agencies, actors,  466, 468
Ernst, D.,  238 causality,  465, 467
Estevadeordal, A.,  68, 365 competition effects,  459
Europe, European Union currency manipulation,  459–62, 467
Agenda  2000, 246 currency unions,  463
Anglo-French commercial treaty of  1786, development-trade relationships,  480
42–43, 47 firm behavior,  468
antidumping laws,  61t, 125 fixed  vs. floating,  459, 463, 466–67
collective action,  324 foreign direct investment,  464
competition laws historically,  228n9 gold standard,  23, 457, 463
consumer power effects,  441 individual actors,  459–61, 465, 468
electoral systems,  294n10 institutions,  459–61, 465, 468
environmental policy in trade lobbying, 465–68
agreements, 447 Mundell-Fleming trilemma,  463, 466–67
foreign direct investment,  246–48 optimum currency areas (OCA)
France (see France) theory, 458–59
GATT (see GATT) overview,  14, 457–58, 469
542   Index

Exchange rates (Cont.) free trade  vs. protectionism,  42–43, 45,

pass-through,  459, 464 47, 49, 51
as policy instrument,  70–71, 458–62 GATT effects generally,  26, 29
political agency,  468 global production networks,  237
protectionism/labor interests,  121, 123, immigration policy,  526
463–64, 469n3 political systems,  47, 423
regional integration,  460–61 PTAs, 24
research strategies,  464–65 WTO membership,  386
tariffs, 467–68 Francois, J.,  162
timing, time horizons,  467 Frankel, J.,  463
trade policy,  459–61 Freeman, R. B.,  503
trade policy choices,  462–64 Free trade
alternatives to,  48–50
Fabbri, F.,  131 Anglo-French commercial treaty of  1786,
Factor-based trade theorem,  100–101 42–43, 47
Factor endowments (Heckscher-Ohlin) antitrust/trade policies interactions,  215–17,
models,  5–6, 120, 164–65, 172n13, 178 227nn5–6
Fagiolo, G.,  386 commercial policy,  42–48
Fariss, C. J.,  152n2 comparative advantage theory,  43–44
FDI.  see foreign direct investment (FDI) competition policy  (see antitrust/trade
Fearon, J. D.,  420 policies interactions)
Feenstra, R.,  129 concentrated interests,  46
Felbermayr, G.,  28 concepts, definitions,  38–40
Findlay, R.,  166, 173n18 decline, reformation of,  50–52
Finite mixture modeling,  101–2 democratic regimes,  49–50, 52–53, 259,
Fontagné, L.,  183 262–63, 272–74, 277n18
Fordham, B. O.,  106, 119, 320, 433 development-trade relationships,  476–77,
Foreign direct investment (FDI) 485–86
authoritarian regimes,  132, 307, 308 diminishing returns theory,  44
development-trade relationships,  478, 479, domestic altruism,  42
483, 486–88, 489n2 economic development,  44–48
exchange rates,  464 economic efficiency,  42, 63, 70, 73n12
Latin America,  242–44, 483 economic nationalism,  48–49
Mexico,  242–44, 486–87 environment-trade relationships,  443–44
migration-trade relationships,  527 GATT,  21, 32nn2–4, 51
multinational corporations,  234–36, government policy,  41, 46
242–43, 246–48 heterogeneous firms,  196–97, 202–3,
new new trade theory,  127 206–8
PTAs,  483, 489n2 import restrictions,  38
RTAs, 368 individual attitudes,  101–12, 114nn1–3, 126
war-trade relationships,  435 infant industry argument,  48
Framing experiments,  109–10, 114n3 international economy,  41–42, 52–53
France Iron Law of Wages,  44
alliance formation,  3 Keynesian demand stimulus,  51
Anglo-French commercial treaty of  1786, market access,  38
42–43, 47 mercantilism, 39–41
economic geography,  322 migration-trade relationships,  526, 528
 Index   543

moral economy,  49–50 human rights provisions,  498–500, 500t

most-favored nation (MFN) status,  47 Kennedy Round,  26, 362, 374
overview, 37–38 most-favored nation (MFN),  21, 23, 59–60,
political economy (classical),  40–42 62, 498
price-specie flow mechanism,  40 national security exception,  499
prosperity  vs. power (hegemony),   41, 42, national treatment principle,  498
46–48 political/military alliances effects,  28–31,
protectionism,  39, 41, 43–46, 49–52 30t, 33nn19–21
reciprocity treaties (trade agreements),  45, preferential trade agreements (PTAs)  (see
47, 49 preferential trade agreements (PTAs))
support for,  44, 46–49 principal-supplier rule,  25–26
tariffs,  38, 43, 45–48, 50–51, 63, 73nn7–8 protection instruments,  59–60, 72nn1–3,
Universal Economy doctrine,  39 489n4
war strategy,  43 provisional members,  27
WTO membership effects,  380–81, 381f, PTAs, 353
384–86, 391 reciprocity principle,  64
Freudenberg, M.,  183 RTAs (see regional trade agreements
Freund, C.,  68 (RTAs))
Frye, T.,  305–6 tariffs,  20, 21, 23, 26–27
temporary trade barrier policies,  60, 61t
Garrett, G.,  263 Tokyo Round,  361, 383, 482, 489n4
Gartzke, E.,  14, 428 trade institutions value, impacts,  20–22,
GATT.  see also World Trade 32nn1–7
Organization (WTO) Trade Policy Review Mechanism
accession processes,  390–91, 395n5 (TPRM), 24
compliance monitoring,  24 tuna-dolphin decisions,  407, 412, 447–49
creation,  22–25, 32n9 unforeseen developments
developing  vs. industrial countries,   26–27 safeguard, 412–13
dispute-resolution mechanism Uruguay Round,  24, 62, 66, 68, 139, 151,
(DRM), 24–25 376n4, 387, 448, 481
dispute settlement procedure  (see dispute voluntary export restraints (VERs),  61–62,
settlement procedure (DSP)) 72n4, 73n5
Doha Round,  11, 360, 362, 373–75, 392, 395, Gawande, K.,  67, 123, 160, 164, 166, 167, 169
439, 446, 452 Gawer, A.,  235
effects,  25–28, 31–32, 32n13, 33n15, Geddes, B.,  263, 298, 304, 305, 312n10
33n17, 392 Gelleny, R. D.,  104
Enabling Clause,  482 Gelpi, C.,  422
enforcement,  64, 67 General Agreement on Tariffs and Trade.
environmental policy,  447–48 see GATT
equilibrium,  21–22, 32n6, 64 General Agreement on Trade in Services
European Economic Community (GATS), 162
(EEC), 27 Generalized system of preferences (GSP)
exchange rate manipulation,  459–62 analysis methods,  87–90, 88t, 92
free trade,  21, 32nn2–4, 51 development-trade relationships,  483–85
gravity model  (see gravity model) overview,  29–31, 30t
history,  4, 12, 15n3, 19–20, 59–60, 72nn1–3, Geography.  see domestic geography
215, 380 Gereffi, G.,  235
544   Index

Germany bonus vetus approach,  84, 85, 88t, 90,

alliance formation,  3 94n13
electoral system,  286 border effects,  81
foreign direct investment,  237, 246 collinearity, 92
free trade,  47–49, 522 control variables,  83
GATT effects generally,  26, 29, 31, 83–84 country-year effects,  82–84, 93n9
global production networks,  246–47 cross-sectional form,  78–79
HOSS model,  7 distance,  78, 93n3
immigration policy,  526, 530 dummy variables,  80, 83–85, 92
Marshall Plan,  432 dyadic factors,  79, 80, 82, 92, 93n5, 93n9
protectionism,  49–51, 319–20 equations, 78–79
WTO membership,  386 GATT effects,  26–27, 32n13, 77–78, 86–91,
Gerschenkron, A.,  319 88t, 94n20
Ghana, 260 history,  78–81, 93nn3–7
Gilligan, M. J.,  179 institutional effects,  80–83, 92, 93nn4–6
Gilpin, R.,  235, 425 interwar trade study,  83–85, 86t
Gladstone, W. E.,  45 logged trade values,  79
Glaeser, E. L.,  318 multilateral trade resistance,  81–85, 88t,
Global Integrity,  503 90–92, 93nn8–9, 94nn10–14
Globalization.  see also multinational ordinary least squares,  78, 85
corporations (MNCs) panel data,  79
in developing countries,  132 PTAs,  77–78, 86–91, 88t, 94n20
environment-trade relationships,  442–44 reg3hdfe, 85, 86t, 94nn15–18
exchange rates  (see exchange rates) tetrad approach,  84, 85, 94nn11–12
global production networks,  233–36 trade agreements,  80, 82
global value chains,  235, 238, 480, 481 trade costs,  78, 94n10
history,  3–5, 15n3, 19–20, 39–40 trade-diverting effects,  80–81
immigration effects,  527–28 trade flows,  79
income inequality,  132 WTO membership analysis,  385
labor political influence,  131 year fixed effects,  81
labor rights effects,  503–4 Great Britain.  see Europe, European Union
redistribution, 127 The Great Illusion (Angell), 422
union bargaining,  131 Greece, 246
war-trade relationships,  424 Greenhill, B.,  502, 504
Global production networks (GPNs), Greif, A.,  529
233–36.  see also multinational Grieco, J. M.,  422
corporations (MNCs) Grix, J.,  246
Global value chains (GVCs),  235, 238, 480, Grossman, G.,  8, 63, 65, 122, 166, 263, 276n1,
481, 486–87 282, 300, 311
Goldberg, P. K.,  67, 132, 167 Grossman-Helpman model.  see PFS model
Goldstein, J.,  27, 29, 33n19, 80, 87, 105, 123, 385, (protection-for-sale)
388, 395n5, 524, 530 Grubel, H. G.,  184, 192n6
Goodhart, L. M.,  13, 170 GSP.  see generalized system of preferences
Gowa, J.,  12, 25, 26, 29, 72n1, 83, 85, 180, 189, (GSP)
190, 386, 424 GSP+ programs,  483–84
Gravity model Guisinger, A.,  112
application recommendations,  92–93 Gulotty, R.,  207
 Index   545

Güngör, H.,  247 offshoring, 208

Gutmann, J.,  220 PFS model (protection-for-sale),  197
Guzman, A.,  152n8 policy implications,  206–7
policy preferences,  202–5
Hadenius, A.,  304 producer associations,  208–9
Hafner-Burton, E. M.,  152n2, 502, 503, 505 producers in trade politics,  197, 208
Haftel, Y. Z.,  139, 142 profitability,  199, 202
Hainmueller, J.,  106, 523 protectionism/labor interests,  127–29
Hankla, C. R.,  287, 288, 303, 305–8, 311, 328 tariffs,  203–5, 207
Hansen, W.,  125 total factor productivity (TFP),  198–99,
Hanson, G.,  129, 523 201–2
Harrelson-Stephens, J.,  502 trade liberalization,  196–97, 202–3, 206–8
Harrison, A.,  244 trading behaviors,  200, 208
Hart, J. A.,  235 Hicks, R.,  83, 85, 108, 113, 338, 366
Haskel, J.,  131 Hillman, A.,  166, 173nn17–18
Hatch, W.,  13, 238, 241 Hiscox, M. J.,  101, 105, 106, 110, 121, 123, 320,
Hathaway, O.,  123 523
Hatton, T. J,  519, 525, 528 Hobson, J. A.,  425
Hauk, W. R.,  288 Hoda, A.,  26
Hawkins, H. C.,  25–26 Hoekman, B.,  31, 162, 222, 370, 372, 376n4
Hayakawa, K.,  240 Hofmann, T.,  139, 142, 143, 148, 388
Hays, J. C.,  107, 126 Hollyer, J. R.,  150
Head, K.,  84, 90, 94n12, 94n18 Horn, H.,  338, 365, 366, 406, 407
Hearn, E.,  107 Hudec, R.,  387, 402
Heckscher, E.,  5 Hull, Cordell,  22, 380
Heckscher-Ohlin (factor endowments) Human Development Index,  503
models,  5–6, 120, 164–65, 172n13, 178 Human rights-trade relationships
Heiduk, G.,  13 activism, 504
Helbling, M.,  530 causality, 503
Helfer, L. R.,  152n2 data sets,  503
Helpman, E.,  8, 63, 65, 122, 128, 166, 201, 263, developing countries,  511
276n1, 282, 300, 311, 385 digital trade,  507–8
Henderson, J.,  236 GATT/WTO provisions,  498–500, 500t
Henisz, W. J.,  123–24, 263 globalization effects,  503–4
Henn, C.,  386 information access,  506–7
Hertel, S.,  504 labor rights,  503–4
Heterogeneous firms literature review,  501–4
comparative advantage theory,  205–6, market access,  495, 499
210n37 multinational corporations,  504
export behaviors,  200, 202–3 NAFTA,  496, 506
foreign direct investment,  200–201 national treatment principle,  498
import behaviors,  200 overview,  14–15, 493–94, 494f
intra-industry product differentiation,  206, personal integrity,  494, 502–3
210n37 privacy, 506–8
lobbying,  203, 204, 207 PTAs,  502, 505–6
new new trade theory,  127, 171n3, 198–200 quality of life,  501
nontariff barriers,  203–4, 207 state rights, responsibilities,  494–95
546   Index

Human rights-trade relationships (Cont.) collective actions,  100, 105–7

trade agreements covering,  495–96, consumer goods consumption,  101
497–98t cosmopolitanism, 104
trade encouragement,  506 cultural threat perception,  111
trade sanctions, fines,  510 democratic regimes,  113, 261, 276nn2–3
women's rights,  501–2 economic self-interest,  100–106, 103t,
WTO membership,  508–9 114n1
Hume, David,  40, 41 education, 106
Hungary,  246, 247, 266t, 392 electoral systems,  107–8
Hutcheson, Frances,  41 embedded liberalism,  107, 126
Hwang, W.,  126 experimental approaches,  109–12, 114n3
Hymer, S. H.,  234–35 factor-based trade theorem,  100–101
finite mixture modeling,  101–2
Iacobucci, E. M.,  227n4 framing experiments,  109–10, 114n3
Imai, K.,  101–2, 385 free trade  vs. protectionism, 101–12,
Import-substituting industrialization (ISI), 114nn1–3, 126
4, 242, 425, 477–79, 487–88, 489n1 gender, 105–6
Income inequality homeownership, 101
authoritarian regimes,  301 ideational factors,  104–5
aversions, individual attitudes,  111–12 identification strategies,  106–7
democratic regimes,  272–74, 277n18 income inequality aversions,  111–12
Europe, European Union,  128 international politics,  108–9
globalization, 132 issue framing,  108–10
inter-industry trade,  128 legislators ideology,  104, 107–8
intra-industry trade,  128 marriage, 105
migration-trade relationships,  525 media market structure,  108
new new trade theory,  128–29, 132 mobility, 101
protectionism/labor interests,  128–29, 132 occupational scores,  114n3
United States,  128 outer-group anxiety,  104–5
war-trade relationships,  424 political origins,  107–9
India political persuasion,  108
antidumping laws,  60, 61t priming experiments,  111–12
foreign direct investment,  488 protectionism, industry-level,  159–60,
globalization/income inequality 163–65, 172nn10–15
relationships, 132 religious beliefs,  105
global production networks,  239 research limitations,  102–4, 103t, 114n2
immigration policy,  530 research on generally,  99–100
import-substituting industrialization risk orientation,  106–7
(ISI), 488 service trade liberalization,  102–3
MFN import tariffs,  67, 276n10 skill specificity,  101
PTAs, 353 socialization through group
WTO dispute resolution,  353 membership, 105–7
Individual attitudes socio-tropic formation,  105–7
agricultural protection,  111 Stolper-Samuelson theorem,  100–101
attitude/behavior relationships,  112–13 survey instruments,  102, 103t, 112, 114n2,
attitude/policy outcome 114nn4–5
relationships, 113–14 unionization, 105
 Index   547

Indonesia, 61t, 237, 266t, 271 Jackson, Henry,  327

Industry interests.  see protectionism, Jackson, J. H.,  384
industry-level James, S.,  320
Inter-industry trade. Japan
see also intra-industry trade antidumping laws,  61t
alliances, 190 EPAs, 360
comparative advantage theory,  121, 180 female protectionism,  106
income inequality,  128 free trade  vs. protectionism,
intra-industry  vs., 190 103–4, 204
patterns,  177–80, 183 global production networks,  236–41
International Convention for the Kodak-Fuji dispute,  376n6
Conservation of Atlantic Tunas MFN tariff studies,  66
(ICCAT), 450 priming experiments,  111
International Social Survey Program privacy laws,  508
(ISSP),  100–101, 103t, 106–8, 164 proportional representation,  327
Intra-industry trade. RTAs, 376n6
see also inter-industry trade survey instruments,  103t
alliances, 189–91 voluntary export restraints (VERs),
benefits, 180 61–62, 72n4, 73n5
comparative advantage theory, 178–79, 192n1 WTO dispute settlement,  144
conflict, 189 Jennings, K.,  124
democratic regimes,  188–90 Jensen, J. B.,  198, 322
dispute settlement procedure (DSP),  188 Ji, Z.,  240
heterogeneous firms product Johns, L.,  14, 142, 146, 152n6
differentiation,  206, 210n37 Jones, K.,  391
horizontal  vs. vertical, 183–84, 192n5 Judkins, B.,  122
income inequality,  128
inter-industry trade  vs., 190 Kaltenthaler, K.,  104
international politics models,  184–85, Kant, I.,  422, 430
192nn6–7 Karacaovali, B.,  68
lobbying, 179–80 Karol, D.,  288
migration-trade relationships,  518, 528–29 Kastner, S. L.,  433
new trade theory,  121, 123, 160–61, 171n3 Katzenstein, P. J.,  236, 281
overview,  13, 177–78, 191–92 Kayser, M. A.,  107, 283
product-level aggregation,  181–83, 182t, Kee, H. L.,  222
192nn2–4 Keohane, R. O.,  519
protectionism,  121, 123, 160–61, 171n3, 179–80 Keynes, J. M.,  51
PTAs,  186–88, 192n8 Kim, S.- Y.,  14, 25, 26, 29, 139, 142, 143, 148, 203,
rivalry, 189–91 204, 206, 235, 338, 366, 367, 383, 385, 386,
state  vs. dyad level, 184 388, 391
Stolper-Samuelson theorem,  179 Kimura, F.,  241
third parties,  187–88 Kindleberger, C. P.,  319, 383
trade partner specialization,  187 Kleinberg, K. B.,  119, 433
Iraq, 392 Kletzer, L. G.,  322
Irwin, D. A.,  23, 27, 59, 69, 73n8, 112, 216, 321, Klienberg, K.,  106
463, 493 Knowles, V.,  246
Itskhoki, O.,  128 Kodak-Fuji dispute,  376n6
548   Index

Kohler, W.,  28 Lawrence, R. Z.,  361, 375

Kokkinou, A.,  246 Lee, H.,  126
Kono, D. Y.,  13–14, 179, 263, 308, 309, 310 Lee, J.-W.,  161
Kortum, S. S.,  198 Lee, M. -J.,  386
Kostecki, M. M.,  31, 370 Lenin, V. I.,  425
Kotschwar, B.,  364, 376n5 Lesotho, 266t
Krasner, S.,  22 Levi, M.,  105
Kreickemeier, U.,  128 Li, Q.,  123, 235, 273, 428
Kriesi, H.,  530 Life and Writings of Adam Smith
Krishna, P.,  160, 164, 166, 169 (Stewart), 43
Kronthaler, F.,  220 Lim, H.,  372
Krugman, P.,  7, 15n8, 201, 317, 318, 330nn1–2 Limão, N.,  65–66, 67, 68
Kubota, K.,  9, 262, 303, 305, 306 Lindert, P. H.,  162
Kucik, J.,  142, 148, 338, 388, 392 Lipset, S. M.,  272
Kume, I.,  103, 106, 107, 111 Lipsey, R.,  244
Kuno, A.,  108 List, Friedrich,  48
Kuo, J.,  12 Lithuania, 266t
Kuthy, D.,  305–8, 311 Liu, X.,  386
Lloyd, D. J.,  184, 192n6
Labor interests.  see protectionism/labor Lobbying
interests authoritarian regimes,  122
Lake, D.,  320, 383 democratic regimes,  122, 133, 261
Lang, A.,  504 domestic geography,  322, 324, 329, 331n5
Latin America exchange rates,  465–68
Brazil (see Brazil) heterogeneous firms,  203, 204, 207
Chile, 266t, 294n8, 352, 441, 496, 497t, 510 intra-industry trade,  179–80
democracy formation by country,  266t migration-trade relationships,  525–26
electoral systems,  294n8 protectionism, industry-level,  160, 163, 165,
foreign direct investment,  242–44, 483 166, 168–69, 172n11, 172n15, 173n19
free trade areas (FTAs),  11, 62 protectionism/labor interests,  124, 125
globalization/income inequality United States,  204
relationships, 132 United States trade unions,  124
global production networks,  242–45, 250 WTO membership,  384
import-substituting industrialization Lobbying Disclosure Act of  1995, 204
(ISI),  4, 242, 425, 477–79, 487–88, 489n1 Lohmann, S.,  9, 303, 306
Mexico (see Mexico) Londregan, J.,  126
multilateral tariff reductions, Long, A. G.,  190
68, 73nn12–13 Lopez, R.,  527
Peru, 266t, 387 Lopez-Cordova, E.,  273
policy preference formation,  321 Lu, X.,  103–4, 111, 172n4
PTAs,  352, 352f, 483 Ludema, R.,  66
RTAs,  11, 62, 368, 371, 373–74
technical barriers to trade (TBTs),  407 Madagascar, 266t
trade policy historically,  4, 262 Madeira, M. A.,  179
Uruguay,  62, 266t, 469n3, 496 Maestas, C.,  106
WTO dispute resolution,  407 Magee, C.,  123, 169
Latvia, 266t Maggi, G.,  65, 67, 73n10, 167
 Index   549

Malawi, 266t government export promotion,  478–79

Malaysia,  237–38, 240, 241, 488 human rights in trade agreements,  495–96,
Malhotra, N.,  524, 530 497–98t, 506
Mali, 266t party organization,  304
Manger, M. S.,  14 RTAs, 368
Mangioni, G.,  386 technical barriers to trade (TBTs),  407
Mankiw, N. G.,  130 trade shares in GDP,  476
Mansfield, E. D.,  104, 105, 121, 123–24, 149, 180, tuna-dolphin decisions,  407, 412, 447–49
189, 190, 263, 284, 285, 305–6, 338, 374, WTO membership,  393–94
386, 434, 461 MFN.  see most-favored nation (MFN)
Maoz, Z.,  429 Migration-trade relationships
Marchetti, J.,  372 capital-intensive goods,  521, 522, 525
Margalit, Y.,  104, 105, 111, 112, 126, 130, 173n23, complementarity, 522
322, 524, 530 developing countries,  527–28, 531
Markusen, J. R.,  518, 522 economic threat,  524
Marshall, A.,  318 education/tolerance, 523–24
Marshall Plan,  432 factor mobility model,  518, 521–22
Martin, L.,  21 foreign direct investment,  527
Marx, Karl,  44 free trade,  526, 528
Mateev, M.,  247 high  vs. low-skilled labor, 523, 524
Matschke, X.,  124 home goods preferences,  528
Matsushita, 235 immigration policy,  519, 524–26
Maur, J. C.,  365, 373 import-competing sector,  526, 527
Mauritania, 505 income inequality,  525
Mavroidis, P. C.,  23, 27, 59, 338, 365, 366, increased information effects,  528–29
406, 407 intra-industry trade,  518, 528–29
Mayda, A. M.,  66, 100, 106, 164, 170, 523 labor-intensive goods,  521, 522, 525
Mayer, T.,  84, 90, 94n12, 94n18, 300–302, 311 lobbying, 525–26
Mayer, W.,  164 migrant networks,  529
McCallum, J.,  81 NAFTA,  520, 528
McGillivray, F.,  276n1, 287, 288, 294n11, nontradable  vs. tradeable sectors,  523
294n13, 328 open economy politics,  519
McKeown, T.,  121 overview,  15, 518–20, 529–32
McKinley tariff,  214 public opinion,  519, 522–24, 530–31
McKinnon, R. I.,  459 Stolper-Samuelson model,  518, 520–21, 524,
McNamara, K.,  461 527, 528, 531–32
Meissner, C.,  273 trade contracts,  529
Melitz, M. J.,  102, 127, 171n3, 198–201, 385 trade flows,  519–20, 527–32
Mercosur,  11, 62, 351, 469n3, 497t wages, prices effects,  521–23, 526, 528
Mexico welfare state argument,  525
antidumping laws,  60, 61t Mihov, I.,  385
democracy formation,  266t, 501 Mill, John Stuart,  44, 48
development-trade relationships,  486–87 Milner, H. V.,  9, 73n9, 108, 113, 114n1, 122, 123,
export-processing zones,  486–87 149, 162, 172n4, 172n14, 196, 197, 262, 303,
foreign direct investment,  242–44, 486–87 305, 306, 338, 519
globalization/income inequality Milner, W.,  502
relationships, 132 Mitra, D.,  122, 128, 172n12, 300–302, 301, 312n7
550   Index

Mitsubishi, 240 Murillo, M. V.,  110, 321

MNC.  see multinational corporations Mutz, D.,  104, 105
(MNCs)
Mo, C.,  524, 530 NAFTA
Mongolia, 266t environment, 443–48
Montinola, G. R.,  308 global production networks,  243, 244, 250
Montreal Protocol on Substances that Deplete human rights,  496, 506
the Ozone Layer,  450 migration,  520, 528
Moore, M. O,  331n8 as policy instrument,  62
Moran, T. H.,  235 protectionism/labor interests,  124
Morrow, J.,  429 as PTA,  337–38, 340
Mosley, L.,  235, 503, 504 as RTA,  11, 360, 368
Most-favored nation (MFN) Naoi, M.,  12, 103, 106, 107, 108, 111, 125
benefits, WTO membership,  385 Napier, M.,  106
development-trade relationships,  481–84 National Development Plan of  1982, 394
GATT,  21, 23, 59–60, 62 National Security Agency,  507
human rights provisions,  498 Nepal, 266t
PTAs, 353 Neuman, G. L.,  525
WTO membership, import tariffs,  66, 385 Neumayer, E.,  391
Mousseau, M.,  422 New new trade theory
Mozambique, 266t development-trade relationships,  121
Mukherjee, B.,  13, 123, 162, 172n4, 172n14 foreign direct investment,  127
Multidimensional Poverty Index,  503 heterogeneous firms,  127, 171n3, 198–200
Multi-Fiber Arrangement (MFA),  62 income inequality,  128–29, 132
Multilateral environmental agreements intra-industry trade,  121, 123, 160–61, 171n3
(MEAs),  439, 450–52 as policy instrument,  69–70
Multinational corporations (MNCs) principles,  7, 15n8, 121
automobile production,  240, 244 protectionism/labor interests,  123
democratic institutions,  235–36 Nielson, D.,  263, 303
East Asia,  236–41, 250 Nike, 235
electronics industry,  239, 244 Nontariff barriers (NTBs).  see also tariffs
Europe,  245–49, 250 antidumping laws,  60, 61t, 67–69, 72n2,
foreign direct investment,  234–36, 242–43, 73n5
246–48 benefits targeting,  170–71, 173nn17–18,
global commodity chains,  235, 238 284–85, 294n3, 480
global production networks,  233–36 countervailing duties,  341
global value chains,  235, 238, 480, 481 democratic regimes,  259, 263, 275
host country policies,  235–36, 240, 241 domestic geography,  325, 326
Latin America,  242–45, 250 electoral systems,  282, 284
nontariff barriers,  207 Europe, European Union,  207, 325
overview,  13, 15n7, 249–50 heterogeneous firms,  203–4, 207
producers in trade politics,  197 import licenses,  285
production networks generally,  233–36 multinational corporations,  207
RTAs, 368 as policy instrument,  58, 66–67
tariffs,  240, 241 protectionism, industry-level,  160–61, 164, 167
wintelism, 235 protectionism/labor interests,  123, 124
Mundell, R.,  234, 458–59, 518, 521 PTAs, 339–40
 Index   551

specific trade concerns (STCs),  152n10, Peritz, L.,  14, 152n6

401, 405–8 Peru, 266t, 387
subsidies,  71, 282–83, 285, 480 Peters, M. E.,  15, 519, 524, 525, 530, 531
technical barriers to trade (TBTs),  364, 366, Petersen, N.,  220, 222
367, 375, 376n2, 401, 405–8, 448 Peterson, T. M.,  13, 181, 184, 188, 192n6
United States,  325 Pevehouse, J. C. W.,  14, 142, 461, 462, 463
voluntary export restraints (VERs),  61–62, PFS model (protection-for-sale)
72n4, 73n5 heterogeneous firms,  197
North American Agreement on MFN import tariffs,  67
Environmental Cooperation,  446 principles,  8, 13, 63, 122–23
North American Free Trade Agreement. protectionism, industry-level,  165–71,
see NAFTA 173nn16–22
NTBs.  see nontariff barriers (NTBs) protectionism/labor interests,  122–23
Nunn, N.,  267, 276n10 trade policy variation,  300
validation, 123
Oatley, T.,  388, 519 Philippines,  259–60, 264, 266t, 444
Obashi, A.,  240, 241 Piermartini, R.,  364
Occupy Wall Street,  504 Pincus, J. J.,  324, 331n7
Ochs, J.,  121 Pinto, P. M.,  110, 321
Of the Balance of Trade (Hume),   40 Plouffe, M.,  13, 201, 203, 204, 433
Of the Jealousy of Trade (Hume),   40 Polachek, S. W.,  426
O’Halloran, S.,  9, 303, 306, 320 Poland, 266t, 392
Ohlin, Bertil,  5 Policy instruments.  see also specific
Ohls, D.,  391–92 instruments
Olarreaga, M.,  169 antidumping laws,  60, 61t, 67–69,
Olson, M.,  168, 323, 325 72n2, 73n5
Oneal, J.,  422, 426, 428, 429 electoral systems selection,  285
Optimal tariff theory,  218 enforcement,  64, 67
Optimum currency areas (OCA) exchange rates as,  70–71, 458–62
theory, 458–59 GATT/WTO system,  59–60, 72nn1–3
Orefice, G.,  366, 369 (see also GATT)
Ornelas, E.,  68 hold-up problem,  70
O’Rourke, K. H.,  100, 106, 172n13, 528 Multi-Fiber Arrangement (MFA),  62
Osgood, I.,  206, 207 NAFTA, 62
Ossa, R.,  69 nontariff barriers,  58, 66–67
Owen, E.,  12, 130 PTAs, 62
quantitative restriction (quotas),  63, 72n4
Pakistan, 266t, 482 retaliation capacity,  69
Panagariya, A.,  365 subsidies, 71
Papaioannou, E.,  262, 276n4 tariffs,  57–58, 63–68, 73nn7–13
Paraguay,  62, 266t temporary trade barrier policies,  60, 61t,
Paul, J.,  501 72nn2–3
Pavcnik, N.,  132 trade agreements,  58, 63–67, 73nn9–10
Peinhardt, C.,  107, 126 voluntary export restraints (VERs),  61–62,
Pelc, K. J.,  14, 148, 150, 152n3, 152n10, 355n8, 72n4, 73n5
388, 389, 391 WTO dispute jurisprudence,  69, 73nn14–15
Péridy, N.,  183 Polity V database,  276n7
552   Index

Pollins, B. M.,  434 North-South,  475, 482–87, 489nn5–8

Porche, S.,  162 overview,  14, 62, 138
Portillo, Jose Lopez,  393–94 political-economy model, small
Powell, R.,  421, 424 countries, 65
Powers, W.,  240 proliferation, expansion,  349–51, 350–51f
Prakash, A.,  502, 504 reciprocity, 338
Prebisch, R.,  477 rigidity,  141–42, 152nn3–5, 341, 348–49,
Preferential trade agreements (PTAs).  see also 348f
specific agreements safeguards, 341
antidumping mechanisms,  142 scope, 339–40
antitrust provisions,  224–25 state behavior impacts,  342–49, 345–48f,
countervailing duties,  341 355nn7–8
deep integration provisions,  360–65, tariffs,  344–49, 345–48f, 355nn7–8
374–76, 376nn1–2 treaty commitments,  340–41
democratic regimes,  77, 108–9, 140–50 WTO membership,  87, 152n1, 338, 353
depth,  339, 346–47, 347f, 485, 489n6 Program on International Policy Attitudes
design elements,  339–42, 344 (PIPA),  102, 103t
design generally,  337–38, 355, 485 Protectionism
development-trade relationships,  475, antitrust  vs., 217–19, 228nn7–8
482–87, 489nn5–8 free trade  vs., 42–43, 45, 47, 49, 51, 101–12,
dispute settlement procedure  (see dispute 114nn1–3, 126, 305–6
settlement procedure (DSP)) individual attitudes,  101–12,
domestic political pressure,  343–46, 345f 114nn1–3, 126
domino effect,  350 intra-industry trade,  121, 123, 160–61,
equilibrium behavior,  344–46, 345f, 171n3, 179–80
355nn7–8 leader turnover,  145–46
Europe, European Union,  352–53, 352f, 483 Protectionism, industry-level
exchange rate manipulation,  459–62 agriculture,  162–63, 172nn8–9
fair trade agreements,  341 antidumping laws,  161–62
flexibility provisions,  141–42, 152nn3–5, 341, authoritarian regimes,  301–2
348–49, 348f benefits targeting,  170–71, 173nn17–18
foreign direct investment,  483, 489n2 countervailing duties,  161–62
gravity model,  77–78, 86–91, 88t, 94n20 democratic regimes,  162, 172n4, 172n14,
history,  4, 11, 23–24 259–60, 275, 301–2
human rights effects,  502, 505–6 developed  vs. developing countries,  161–
institutionalization, 341–42 62, 172n12
intellectual property provisions,  340 empirical analysis,  160–63, 171n3,
international politics,  108–9 172nn4–9
intra-industry trade,  186–88, 192n8 factor endowments model,  164
membership, 340–41 import penetration,  161
most favored nation (MFN),  353 individual attitudes,  159–60, 163–65,
multilateralism, 354 172nn10–15
multilateral tariff reductions,  68, 73nn12–13 likelihood of receiving,  160–61, 172n12
NAFTA,  337–38, 340 lobbying,  160, 163, 165, 166, 168–69, 172n11,
network complexity,  351–54, 352f, 356n12, 172n15, 173n19
356n17 low-wage/low-skilled labor,
nontariff barriers,  339–40 161, 163, 172n4
 Index   553

mobility, 164 Przeworski, A.,  274

nontariff barriers,  160–61, 164, 167 Psycharis, I.,  246
overview,  159, 171n1 PTAs.  see preferential trade agreements
PFS model (protection-for-sale),  165–71, (PTAs)
173nn16–22
political support function,  166 Quinn, D.,  126
risk/loss aversion,  169–70 Qvortrup, M. S.,  286
specific factors model,  164–65,
172n14 Ranjan, P.,  128
status quo conditions policy,  161, 168–69, Rauch, J. E.,  529
172nn5–6 Rausser, G.,  172n9
Stolper-Samuelson theorem,  164 Real exchange rates.  see exchange rates
tariffs,  160, 164, 166–67 Reciprocal Trade Agreements Act (RTAA)
voting patterns,  164, 165, 172n12, of  1934, 23
172nn14–15 Redding, S. J.,  201
Protectionism/labor interests Regional trade agreements (RTAs).  see also
antidumping laws,  125 specific agreements
collective action modeling,  122–23 Africa,  371, 373–74
compensation, 126–27 baseline studies,  372–73, 376nn4–5
countervailing duties,  125 Brazil,  11, 62, 371
elasticity of demand,  130–31 China, 371
exchange rates,  121, 123, 463–64, 469n3 common external tariff (CET),  361
factor mobility,  120–21, 123 deep integration measurement,  365–66
factors of production,  121 deep integration provisions,  360–65,
firm-specific wages,  128 374–76, 376nn1–2
free trade agreements,  125, 130 effects, 369–71
heterogeneous firms,  127–29 Europe, European Union,  360, 366, 367,
income inequality,  128–29, 132 370–71, 461
industry cleavages,  120–23 foreign direct investment,  368
international  vs. domestic firms, 123 governance, regulatory disarray in,  373–74,
lobbying,  124, 125 376n6
macroeconomic trends,  123–24 governance hubs,  370–71, 375
NAFTA, 124 harmonization,  363–65, 367
nontariff barriers,  123, 124 history, 4
offshoring, 128–30 investment provisions,  364
partisanship, 122 Japan, 376n6
politician's incentives,  122, 124 Latin America,  11, 62, 368, 371, 373–74
redistribution, 124–27 liberalization, 363–65
research on generally,  119–20 Mexico, 368
services offshoring,  129–30 multilateral environmental agreements
Stolper-Samuelson theorem,  120, 128 (MEAs), 374
tariffs,  123, 125 multilateralism,  369–70, 373–74
trade policy effects,  121–24 multinational corporations,  368
trade theory models,  120–21 NAFTA,  11, 360, 368
unemployment,  121, 123–24, 127–28 production networks,  367–68, 376n3
unionization,  124, 131 protection, 363–65
Prusa, T. J.,  372 roles, structures,  11, 14, 362, 363f
554   Index

Regional trade agreements (Cont.) Ruggie, J. G.,  107

second unbundling,  367 Russett, B.,  422, 426, 428, 429
state motivations,  366–68, 375, 489n3 Russia (Soviet Union),  266t, 392, 395, 425, 508
technical barriers to trade (TBTs),  364, 366,
367, 375, 376n2 Sachs, J. D.,  310
United States,  366, 367, 370–71, 376n6 Salamon, L. M.,  327
WTO classification,  365 Salinas de Gortari, Carlos,  243
Regional Trade Agreements Database,  338 Samsung, 239
Rehm, P.,  126 Samuelson, Paul,  6
Reinhardt, E.,  122, 142, 143, 318, 325, 326, 331n9, Sapier, A.,  338, 365, 366
374, 386, 387, 392, 407 Sassen, S.,  527
Remmer, K.,  263 Sattler, T.,  148, 152n8
Resmini, L.,  247 Scalera, J. F.,  385, 391
Resnick, A.,  235 Schattschneider, E. E.,  324
Reuveny, R.,  273 Schelling, T. C.,  404
Ricardo, D.,  41, 43–44, 178, 214, 226 Scheve, K.,  100, 101, 103–4, 109, 111, 127, 131,
Ricardo-Viner model (specific factors 164, 165, 172n4, 523
approach),  7, 112, 120, 163–65, 172n13 Schiff, M.,  365, 527
Rickard, S. J.,  13, 152n11, 263, 284, 285, 291, Schiller, W. J.,  329
294n1 Schonhardt-Bailey, C.,  321
Ries, J.,  84, 94n12 Schott, P. K.,  201
Rieselbach, L. N.,  320 Schwab, Susan,  392
Rivers, D.,  27, 29, 33n19, 80, 87, 105, 385, 388, 395n5 Scott, A. J.,  322
Robert-Nicoud, F.,  168, 170 Sen, A.,  496
Robertson, G.,  132 Service trade liberalization,  102–3
Robinson, J. A.,  272, 273, 301 Shadlen, K. C.,  14
Rocha, N.,  366, 369 Sherlund, S.,  124
Rocke, D. M.,  372, 389 Sherman, J.,  213–14
Rodríguez-Clare, A.,  65, 73n10 Shirato, Y.,  145
Rodrik, D.,  100, 114n1, 160, 164, 166, 170, 171, Shrimp-turtle case,  448–50
171n1, 485, 488 Shugart, M. S.,  303
Rogers, J.,  463 Shughart, W. F.,  222
Rogowski, R.,  9, 107, 120, 121, 130, 131, 190, 263, Siegfried, J. J.,  327
276n1, 283, 287, 288, 303, 306, 327, 328 Siemens, 235
Romania,  247, 392 Silverman, J. D.,  222
Ron, J.,  503 Simmons, B.,  21, 152n8
Roosevelt, Franklin,  502 Simmons, E.,  430
Rose, A.,  25, 27, 32n13, 79, 86–87, 384, 385, Singer, H.,  477
395n2, 463 Sinnott, R.,  100, 106
Rosecrance, R.,  435 Siourounis, G.,  262, 276n4
Rosendorff, B. P.,  12, 73n9, 144, 145, 149, 150, Slaughter, M.,  100, 101, 103–4, 109, 111, 127, 131,
343, 392 164, 165, 172n4, 523
Roy, M.,  372 Slovak Republic,  266t
RTAs.  see regional trade agreements (RTAs) Slovenia,  211, 221, 266t
Rubinstein, Y.,  385 Smith, A.,  144, 145, 284
Rudra, N.,  236, 273 Smith, Adam,  5, 38, 40–43, 48, 214, 226, 404
Rueda, D.,  122 Smith, D.,  123
 Index   555

Smith, J.,  142 Taagepera, R.,  286

Smoot-Hawley tariffs,  59, 69, 73n8, 204–5, 320 Tabellini, G.,  73n10
Smuckler, R. H.,  320 Taglioni, D.,  82
Snidal, D.,  424 Tahk, A. M.,  531
Snowden, Edward,  506–7 Taiwan,  238–39, 488
Social and Economic Empowerment Rights Tajoli, L.,  386
Initiative (SERF),  503 Takacs, W.,  123
Sony, 238 Tariffs.  see also nontariff barriers (NTBs)
South Korea ad valorem,  73n8
antidumping laws,  61t authoritarian regimes,  298, 305, 306, 310
democracy formation,  266t Brazil,  60, 259–60, 271, 276n10
global production networks,  238–39 common external tariff (CET),  361
government export promotion,  478–79 defection,  344–46, 345f
industrialization, 489n3 democratic regimes,  259–60, 262–72, 265f,
MFN tariff studies,  66 266–67t, 271f, 275, 276nn6–14, 277n15
WTO dispute resolution,  401, 406 development-trade relationships,  476, 480,
Soviet Union (Russia),  266t, 392, 395, 425, 508 481, 484
Specific factors approach (Ricardo-Viner domestic geography,  320, 323, 324, 329,
model),  7, 112, 120, 163–65, 172n13 331n10
Specific trade concerns (STCs),  152n10, 401, electoral systems,  282–84
405–8 Europe, European Union,  321
Staiger, R. W.,  64, 66, 67, 70, 71, 73n10 exchange rates,  467–68
Steagall, J.,  124 GATT,  20, 21, 23, 26–27
Stewart, D.,  43 heterogeneous firms,  203–5, 207
Stiglitz, J.,  198 McKinley, 214
Stokes, S.,  263 MFN import tariffs, EU,  66, 67, 484
Stolper, W.,  6 multilateral reductions,  68, 73nn12–13
Stolper-Samuelson theorem multinational corporations,  240, 241
democratic regimes,  262, 273 optimal tariff theory,  218
individual attitudes,  100–101 as policy instrument,  57–58, 63–68,
intra-industry trade,  179 73nn7–13
migration-trade relationships,  518, 520–21, protectionism, industry-level,  160, 164,
524, 527, 528, 531–32 166–67
principles, 6 protectionism/labor interests,  123, 125
protectionism, industry-level,  164 PTAs,  344–49, 345–48f, 355nn7–8
protectionism/labor interests,  120, 128 retaliatory, 405
Story, D.,  393 settlement,  345–46, 345f
Stucki, P.,  411 skilled  vs. low skilled in democratic
Sturgeon, T.,  235 regimes, 272, 276nn9–10
Subramanian, A.,  26, 385 Smoot-Hawley,  59, 69, 73n8, 204–5, 320
Subsidies,  71, 282–83, 285, 480 treaty depth impacts,  346–47, 347f
Suominen, K.,  365 United States,  59, 69, 73n8, 204–5, 320, 324,
Svolik, M. W.,  309 329, 384
Swagel, P.,  130, 161 WTO membership,  384, 385
Swinnen, J.,  172n9 Technical barriers to trade (TBTs),  364, 366,
Switzerland,  29, 107 367, 375, 376n2, 401, 405–8, 448
Sykes, A. O.,  23, 27, 59, 69, 152n4 Teh, R.,  364, 365, 372
556   Index

Teitelbaum, E.,  132 World War II,  4

Temin, P.,  493 Trans-Pacific Partnership (TPP),  360, 362,
Teorell, J.,  304 363f, 507
Thailand Trebilcock, M. J.,  216
democracy formation,  266t Trefler, D.,  73n11, 161, 267, 276n10
global production networks,  237–40, 250 Trindade, V.,  529
human rights,  505–6 Tsekov, I.,  247
multinational firms role,  487 Tsutsui, K.,  502
Thies, C. G.,  13, 162, 181, 184, 188, 192n6 Tucker, D.,  39, 41, 48
Thomakos, D. D.,  300–302 Turkey
Thompson, P.,  435 antidumping laws,  60, 61t
Tinbergen, J.,  362 democracy formation,  266t, 300–302
Tingley, D.,  101–2, 108, 113 EC customs union,  353
Tollison, R. D.,  222 foreign direct investment,  488
Tomz, M.,  27, 29, 33n19, 80, 87, 385, 388, 395n5 GATT/WTO trade effects,  29
Tovar, P.,  67, 169 import-substituting industrialization
Toyota, 236–38 (ISI), 488
Trade Adjustment Assistance (TAA),  107, 126, MFN import tariffs,  67
322 PTAs, 353
Trade barriers.  see nontariff barriers (NTBs); WTO dispute resolution,  353
tariffs Turtle-excluder devices (TEDs),  448–50
Trade policy generally
comparative advantage theory,  5 Ulubasoglu, M. A.,  300–302
democratic regimes,  9, 13, 259–60, 274–75, UNCTAD,  27, 71, 483, 503
276n1 United States
factor-based trade theorem,  100–101 antidumping laws,  61t, 125, 464
finite mixture modeling,  101–2 antitrust law,  213–14, 222, 227n2, 228n11
framing experiments,  109–10, 114n3 bilateral investment treaties (BITs),  370
Heckscher-Ohlin (factor endowments) citizen trade preferences,  100
models,  5–6, 120, 164–65, 172n13, 178 coalition-building,  329, 331n8
history,  3–5, 15n3 competitive need limitations,  484
HOSS model,  6–7, 15n7 countervailing duties,  125, 394
industrialization, 3 currency unions,  463
international negotiations, elasticity of demand,  131
agreements, 10–11 electoral district size,  288, 302
mercantilism, 3 environmental policy in trade
optimal tariff theory,  218 agreements, 447
optimum currency areas (OCA) female protectionism,  105–6
theory, 458–59 framing experiments,  109–10
protectionism,  3, 8–10 GATT (see GATT)
Ricardo-Viner model (specific factors global production networks,  238–40, 242,
approach),  7, 112, 120, 163–65, 172n13 250
social institutions,  9 GSP preferences,  483, 484
Stolper-Samuelson theorem  (see Stolper- hostile foreign policies,  433–34
Samuelson theorem) human rights in trade agreements,  495–96,
trade barriers,  10–11 497–98t, 509–10
World War I,  3 immigration policy,  523, 525, 528, 530
 Index   557

income inequality/unemployment,  128 Urpelainen, J.,  483

international cooperation in trade Uruguay,  62, 266t, 469n3, 496
agreements, 452 USA Patriot Act of  2001, 507, 508
Kodak-Fuji dispute,  376n6 US v. Tuna I/II, 407, 412, 447–49
labor unions,  9
lobbying, 331n5 Van Wincoop, E.,  81, 93n8
lobbying legislation,  204 Van-Zandt, T.,  385
McKinley tariff,  214 Verdier, D.,  322
Mexico WTO membership case Vietnam,  387–88, 508
study, 393–94 Voigt, S.,  220
MFN tariff studies,  66 Voluntary export restraints (VERs),  61–62,
multilateral tariff reductions,  68, 72n4, 73n5
73nn12–13, 93n6
nontariff barriers,  325 Wage differences.  see income inequality
online communications monitoring,  507 Waked, D. I.,  220, 222
partisan effects,  329, 331n8 Waldkirch, A.,  244
policy preference formation,  319–22 Wallace, N. C.,  214
political persuasion,  108 Wallerstein, I.,  47, 49
political system responsiveness,  328 Walmart, 235
postwar trading regime formation,  383 Walter, S.,  107, 126, 459–60
priming experiments,  111–12 Waltz, K.,  424
privacy laws,  507–8 Wang, Z.,  240
proportional representation,  327 Ward, M. D.,  429
protectionism, demands for,  123, 164 Warner, A.,  310
protectionism  vs. free trade,   103–4, War-trade relationships
109–10, 126 analysis methods,  433
PTAs,  352, 352f, 483 anarchy, 424
RTAs,  366, 367, 370–71, 376n6 asymmetric information,  420–21
sectionalism, 320 bargaining theory,  420–21
Smoot-Hawley tariffs,  59, 69, 73n8, 204–5, capitalism, 425
320 causal mechanisms analysis,  433
specific trade concerns (STCs),  405–8 commitment problem,  421, 431
tariffs,  59, 69, 73n8, 204–5, 320, 324, 329, 384 competition,  421–23, 425, 426
trade agreements generally,  109, 110, 447 dependency theory,  424, 425
trade models,  6–7 economic development,  421, 423–24
trade policy historically,  3–4, 9–10, 49–51 empirical findings,  426–28, 427t
trade shares in GDP,  476 endogeneity, 429
trade union lobbying,  124 firm level differentiation,  433
tuna-dolphin decisions,  407, 412, 447–49 foreign direct investment,  435
voluntary export restraints (VERs),  61–62, foreign policy determinants,  433–34
72n4, 73n5 gains distribution,  32n6
WTO dispute resolution,  388, 401, 406, globalization, 424
411–13, 447–50 income inequality,  424
WTO membership benefits,  386 indivisibility problem,  421
Uno, S.,  503 information mechanisms,  431
Urata, S.,  108 interdependence analysis,  433
Urbatsch, R.,  113, 321 liberalism,  421–23, 426
558   Index

War-trade relationships (Cont.) 367, 375, 376n2, 401, 405–8, 448

Marxism (socialism),  421, 425 trade agreement flexibility provisions,
militarized interstate disputes (MIDs),  426, 141–42, 152nn3–5
429 World Value Survey (WVS),  100–101, 103t, 164
networks, 430 Wright, J.,  307, 308
opportunity costs,  430–31 WTO Analytical Index,  401
overview,  14, 419–20, 434–35 WTO membership
power relations,  420–21, 424, 425 accession processes,  390–91, 395n5
realism,  421, 423–24, 426 alliance ties,  29, 392
third parties,  429–30 analysis methods,  29, 385, 395n2
trade constraints,  430–31 antidumping laws,  392
trade share  vs. openness, 428–29, 428t benefits generally,  382
transformation mechanisms,  432 developing countries,  387–88
variable selection artifacts,  428–29, 428t dispute resolution  (see dispute settlement
Wei, S. -J.,  26, 240, 385 procedure (DSP))
Weinstein, D. E.,  65–66 economic, political determinants,  391–93
Wellisz, S.,  166, 173n18 effectiveness, 388–90
Werfel, S.,  464 effects generally,  25, 29, 66, 87, 388–89,
Weyland, K.,  262 394–95, 395n2
Wijkstrom, E. N.,  406, 407 enforcement,  383, 384, 386–89
Wilf, M.,  14, 392 expansion,  381, 382f
Williamson, J. G.,  519, 525, 528 free trade effects,  380–81, 381f, 384–86, 391
Wilson, B.,  387 gravity model analysis,  385
Winters, L. A.,  365, 372 human rights-trade relationships,  508–9
World Bank,  503 legalization, legitimacy,  384
World Trade Organization (WTO).  see lobbying, 384
also GATT market access,  391–92
Agreement on Antidumping,  73n5 Mexico, 393–94
Agreement on Safeguards,  73n5 MFN benefits,  385
antidumping laws,  60, 61t, 67–69, 72n2, MFN import tariffs,  66, 385
73n5, 227n4 motivations to form,  383–84
compliance monitoring,  24 motivations to join,  390, 392
dispute jurisprudence,  69, 73nn14–15 negotiation outcomes,  66
dispute settlement procedure  (see dispute network complexity,  353–54
settlement procedure (DSP)) ordinary least squares analysis,  25
Doha Round negotiations,  11 preference changes,  389–90
effects,  25–28, 32n13, 33n15, 33n17 PTAs,  87, 152n1, 338, 353
exchange rate manipulation,  459–62 selection bias,  389
history,  4–5, 19–20, 215, 380, 481 spillover effects,  389
human rights provisions,  498–500, 500t tariffs,  384, 385
membership (see WTO membership) trade openness effects,  380–81, 381f,
national security exception,  499 384–86, 391
roles,  10–12, 14, 124–25, 162, 380, 443–44, Wu, W. -C.,  301, 306
446, 448, 481
RTA classification,  365 Yamamura, K.,  238
subsidies, 71
sustainable development,  448 Zeile, W. J.,  233
technical barriers to trade (TBTs),  364, 366, Zhang, J. J.,  14