Anne Marie Slaughter - International Law in A World of Liberal States

International Law in a World of Liberal States
Anne-Marie Slaughter *
International law and international politics cohabit the same conceptual space.
Together they comprise the rules and the reality of 'the international system', an
intellectual construct that lawyers, political scientists, and policymakers use to
describe the world they study and seek to manipulate. As a distinguished group of
international lawyers and a growing number of political scientists have recognized,
it makes little sense to study one without the other.
In keeping with this tradition, this article seeks to develop an integrated theory of
international law and international relations. Previous efforts in this vein fall into
several categories. Myres McDougal and Harold Lasswell, progenitors of the New
Haven school, used a theory of domestic politics and domestic law to rethink the
nature and definition of international law. The international legal process school,
pioneered by scholars such as Abram Chayes and Louis Henkin, sought to explore
and take account of the actual impact of international legal rules on international
political processes, from crises to routine decision-making. The task was to
determine to what extent law shapes 'how nations behave".1
Both of these approaches were developed in response to ongoing work in
political science. The young discipline of international relations surged to
respectability on the tide of Realism, proffering a hard-boiled code of conduct for
the Cold War and disdaining the dangerous moralism of international law.
International lawyers thus faced the 'Realist challenge*: the claim that law was
simply irrelevant to international politics. McDougal and his disciples offered a
theoretical response; international legal process scholars sought to establish more
*
Professor of Law, Harvard Law School. Formerly Anne-Marie Burley. I am grateful to Lea
Brilmayer, Walter Mattli, Andrew Moravcsik, Robert Keohane, Joseph Weiler, and David
Wippman for helpful comments. Sarah Fandell provided her customary excellent research
assistance. Finally, thanks are due to the Russell Baker Scholars' Fund and the Herbert and
Marjorie Fried Faculty Research Fund at the University of Chicago Law School. A more fully
documented version of this essay will appear in a forthcoming volume of recent contributions to
the European Journal of International Law.
For an overview of these earlier efforts to integrate international law and international relations,
see Slaughter Burley, 'International Law and International Relations Theory: A Dual Agenda', 87
AJIL (1993) 205.
6 EJIL (1995) 503-538
Anne-Marie Slaughter
empirical connections between international legal rules and foreign policy decisionmaking.
A third approach - the one pursued here - turns back to the discipline of
international relations itself for inspiration. It takes the 'law in context' injunction
seriously and acknowledges the capacity of many international lawyers for nuanced
political analysis. Nevertheless, instead of canvassing the political dimensions of
various international legal problems, it looks first to the discipline charged with
thinking and theorizing systematically about State behaviour in the international
system. Neither law nor politics may be a science, but international relations
theorists have a comparative advantage in formulating generalizable hypotheses
about State behaviour and in conceptualizing the basic architecture of the
international system.
This approach would look first to the congruence between the image or model of
the international system that implicitly or explicitly informs international law and
the models used by international relations theorists. As political scientists, these
scholars are concerned with the empirical validation of these models. Who are the
primary actors in the international system? What are the primary determinants of
their behaviour? To the extent that the resulting evidence disconfirms assumptions
embedded in the models used by international lawyers, international law and
international politics will become increasingly divorced. If, for instance, the primary
actors in the system are not States, but individuals and groups represented by State
governments, and international law regulates States without regard for such
individual and group activity, international legal rules will become increasingly
irrelevant to State behaviour.
The inquiry in this essay thus begins not with classical international law, but
with the dominant positive analytical framework shared by both international
lawyers and political scientists - Realism. Part I outlines the basic tenets of Realism
and introduces the principal alternative to Realism in international relations
scholarship Liberalism. Liberalism and Realism proceed from different
fundamental assumptions about the international system: assumptions about the
identity of the primary actors in that system, the relationship of those actors to State
institutions, and the primary determinants of State relations with one another.
International lawyers seeking to develop integrated theories of international law and
international relations must take the Liberal critique seriously, examining the ways
in which Liberal assumptions conflict with assumptions underlying traditional
international law.
The most distinctive aspect of Liberal international relations theory is that it
permits, indeed mandates, a distinction among different types of States based on
their domestic political structure and ideology. In particular, a growing body of
evidence highlights the distinctive quality of relations among liberal democracies,
evidence collected in an effort to explain the documented empirical phenomenon
that liberal democracies very rarely go to war with one another. The resulting
behavioural distinctions between liberal democracies and other kinds of States, or
504

more generally between liberal and non-liberal States, cannot be accommodated
within the framework of classical international law.
The project here, consistent with an overall commitment to a new generation of
interdisciplinary scholarship, is to reimagine international law based on an
acceptance of this distinction and an extrapolation of its potential implications. Part
II distills various factors that political scientists have correlated with the 'liberal
peace', factors that can be translated into assumptions about political and economic
relations among liberal States. Part III introduces the concept of a world of liberal
States, acknowledging the distance between such a world and the present
international system but arguing that the hypothesis may nevertheless describe an
important dimension of the current system.
Part IV constructs a model of international law based on a hypothetical world of
liberal States, integrating assumptions about relations among such States with the
broader assumptions of Liberal international relations theory. It focuses first on
relations among individuals and groups in transnational society, hypothesizing a set
of voluntary norms selected by these actors but facilitated by States. The second
level of law assumes the disaggregation of the State into its component political
institutions - courts, legislatures, executives and administrative agencies - and
examines the principles governing transnational interactions among these
institutions. The third level examines the origin, form, negotiations and enforcement
of inter-State agreements among liberal States. At each level the model seeks to
define the relevant body of rules and doctrines that would be included in a definition
of international law in a world of liberal States, and to introduce concepts and tools
of analysis specific to relations among liberal States.
The model developed is advanced as a hypothetical positive model. To the extent
it holds, however, it poses a set of normative challenges for international lawyers.
Part V concludes with a preliminary effort in this vein, re-examining the norm of
sovereignty in a world of liberal States. A central pillar of the positive model is the
conceptualization of the State as a disaggregated entity composed of its component
political institutions. Could the norm of sovereignty be similarly disaggregated? The
discussion in this part sketches the potential form and substantive bases for
complementary norms of judicial, legislative, and executive sovereignty in a world
of liberal States.
The project in this essay is a thought experiment - a largely deductive effort
supplemented with inductive illustrations - designed to generate a hypothetical
model of international law based on a set of assumptions about the composition and
behaviour of specific States. Its ultimate value must await empirical confirmation of
specific hypotheses distilled from this model. At the same time, however, the
explicit articulation of this model may cast a different light on current phenomena
identified as exceptions to the classical model. Christoph Schreuer, for instance, has
recently proposed a new paradigm for international law. He acknowledges,
however, that in articulating his proposed paradigm he draws on examples from
505
Western Europe, the archetypal community of liberal States.2 He may thus be
proposing a paradigm that assumes underlying conditions prevailing only in those
States.
The very idea of a division between liberal and non-liberal States may prove
distasteful to many. It is likely to recall 19th century distinctions between 'civilized'
and 'uncivilized' States, rewrapped in the rhetoric of Western political values and
institutions. Such distinctions summon images of an exclusive club created by the
powerful to justify their dominion over the weak. Whether a liberal/non-liberal
distinction is used or abused for similar purposes depends on the normative system
developed to govern a world of liberal and non-liberal States. Exclusionary norms
are unlikely to be effective in regulating that world.
More generally, however, these concerns raise serious questions, questions that
must and will be addressed if the insights generated by hypothesizing a world of
liberal States prove capable of capturing significant aspects of actual relations
among liberal States. For the moment, to the extent that a distinction is empirically
supported, rather than normatively proclaimed, international lawyers have an
obligation at least to assess its implications. A genuine commitment to
interdisciplinary scholarship - to improving the conceptual fit between international
law and politics - demands no less.
I. Liberalism and Realism in International Relations Theory

Scholars of international relations generate a wide range of theories to solve the
problems and puzzles of State behaviour. Each 'theory' offers a causal account of a
particular outcome or pattern of behaviour in inter-State relations in a form that
isolates independent and dependent variables sufficiently precisely to generate
testable hypotheses. At a higher level of generality, however, these theories can be
grouped into different families or 'approaches' based on a set of positive
assumptions about the international system as a whole. 3 From these very general
assumptions spring a host of more specific theories seeking to explain specific
international events or phenomena, from the causes of war to the dynamics of
international negotiation. Each of these theories can be disputed on its own terms,
on the basis of inaccurate empirical data or faulty logic. Alternatively, the entire
family of theories can be challenged on the ground that the underlying assumptions
about the international system are either wrong or unhelpful.
This part sets forth two of the most common approaches in international relations
theory: Realism and Liberalism. Political scientists would find the versions
2
3
506
Schreuer, 'The Waning of the Sovereign State: Towards a New Paradigm for International Law?',
4 //. (1993) 447,469.
The nature of these assumptions privileges the relative explanatory power of broad classes of
causal factors, such as the distribution of power in the international system, international
institutions, national ideology and domestic political structure.

presented here overly simplified and distilled. Yet as presented, each approach gives
rise to a distinct 'mental map' of the international system, specifying the principal
actors within it, the preferences (or motives) driving those actors, and the constraints
imposed on those actors by the nature of the system itself. The following discussion
presents Realism and Liberalism in terms of their competing assumptions along
these three axes.
A. Realism
The dominant approach in international relations theory for virtually the past two
millennia, from Thucydides to Machiavelli to Morgenthau, has been Realism, also
known as Political Realism. Realists come in many stripes, but all typically share
the following assumptions. First, they believe that States are the primary actors in
the international system, rational unitary actors who are functionally identical.
Second, they assume that State preferences, ranging from survival to
aggrandizement, are exogenous and fixed. Third, they assume that the anarchic
structure of the international system creates such a degree of either actual conflict or
perceived uncertainty that States must constantly assume and prepare for the
possibility of war. In this context, outcomes of State interactions are typically zerosum and thus are determined by relative power. For Realists, power is the currency
of the international system. States interact with one another within that system like
billiard balls: hard, opaque, unitary actors colliding with one another.4
To grasp the defining characteristics and theoretical force of Realism, it is
necessary to understand not only what it includes within its analytical framework,
but also what it excludes: national ideologies, from nationalism to fascism to
communism; domestic regime type, from democracies to dictatorships; and
transnational actors, from multinational corporations to non-governmental
organizations (NGOs). International norms serve only an instrumental purpose, and
are likely to be enforced or enforceable only by a hegemon. The likelihood of
positive-sum games in which all States will benefit from cooperation is relatively
low.
B. Liberalism
A principal alternative to Realism among international relations theorists is
Liberalism.5 As in the domestic realm, Liberal international relations theories have
4
5
This is the classic Realist metaphor first used by Arnold Wolfers. A. Wolfers, Discord and
Collaboration: Essays on International Politics (1962) 19-24.
I use 'Liberalism' here and throughout this paper as a term of art to refer to Liberal international
relations theory. As Andrew Moravcsik has argued, the elements of this theory do indeed flow out
of the political theory and philosophy that we call 'liberalism'. However, the transposition of
liberal analytical assumptions from the domestic to the international realm is complicated. For
present purposes it makes more sense to try and understand international Liberal theory on its own
terms as a self-contained alternative to Realism. See A.M. Moravcsik, Liberalism and
507
been characterized repeatedly as normative rather than positive theories. The best
known Liberal theory in this category is Wilsonian 'liberal internationalism',
popularly understood as a program for world democracy. As used here, however,
Liberalism denotes a family of positive theories about how States do behave rather
than how they should behave. Efforts to reduce Liberalism to a set of core
assumptions that can be stated as succinctly as their Realist counterparts are ongoing
among a growing group of contemporary political scientists.6 I draw here primarily
on one particular version developed by Andrew Moravcsik.7
If Realists focus on States as monolithic entities in their interaction with other
States within an anarchic international system, Liberals focus primarily on Statesociety relations. The first Liberal assumption is that the primary actors in the
international system are individuals and groups acting in domestic and transnational
civil society. Thus where Realists look for concentrations of State power, Liberals
focus on the ways in which interdependence encourages and allows individuals and
groups to exert different pressures on national governments.8 Second, Liberals
assume that the 'State' interacts with these actors in a complex process of both
representation and regulation. Goverments are assumed to represent some subset of
individual and group actors. The fact and process of representation, however, entails
regulation of the activities of all social actors, both those represented and those that
are not represented. Thus where Realists assume 'autonomous' national decisionmakers, Liberals examine the 'nature of domestic representation ... [as] the decisive
link between societal demands and state policy'. 9 Third, Liberals assume that the
nature and intensity of State preferences, determined as the aggregation of the
preferences of individual and group actors represented in a particular State, will
determine the outcome of State interactions. Thus where Realists model patterns of
strategic interaction based on fixed State preferences. Liberals seek first to establish
the nature and strength of those preferences as a function of the interests and
purposes of domestic and transnational actors.
7
8
508
International Relations Theory (Center for International Affairs, Harvard University, Working
Paper No. 92-6, 1992).
See Moravcsik, supra note 5; See, e.g., Deudney, 'Binding Powers, Bound States: The Logica and
Geopolitics of Negarchy', Paper presented at the International Studies Association, Washington,
D.C., 28 March - 2 April 1994; Zacher, Matthew, 'Liberal International Theory: Common
Threads, Divergent Strands', Paper presented at American Political Science Association, Chicago,
IL, 3-6 September 1992; Keohane, 'International Liberalism Reconsidered', in J. Dunn (ed.). The
Economic Limits to Modern Politics (1990) 155; Nye, 'Neorealism and Neoliberalism', 40 World
Politics (1988) 235; Risse-Kappen, 'Ideas Do Not Float Freely: Transnational Coalitions,
Domestic Structures, and the End of the Cold War', 48 International Organization (1994) 185;
Powell, 'Anarchy in International Relations Theory: The Neorealist-Neoliberal Debate', 48
International Organization (1994) 313.
Moravcsik, supra note 5.
The phenomenon of 'interdependence', defined as a situation in which two or more nations each
depend on the other, whether symmetrically or not, by virtue of trade and investment patterns,
population flows, or even cultural and other social exchanges, can be analyzed from either a
Realist or a Liberal perspective. Realists focus only on the impact of interdependence on the power
differential between the nations concerned, whereas Liberals analyze it as an international social
phenomenon.
Moravcsik, supra note 5, at II.
Internationa] Law in a World of Liberal States
EL. Mapping the Attributes of the Liberal Peace

Liberal international relations theory applies to all States. Totalitarian governments,
authoritarian dictatorships, and theocracies can all be depicted as representatives of
some subset of actors in domestic and transnational society, even if it is a very small
or particularistic slice. The preferences of such States are likely to differ from the
preferences of States with more representative governments and more diverse and
complex societies, but not necessarily and not on all issues. Thus, like Realism,
Liberalism is a comprehensive theory of the international system.
Notwithstanding this universal applicability, however, Liberal theory explicitly
takes domestic regime-type into account in its analysis of State behaviour. If the
relevant universe for scholarly analysis is State-society relations, then the scope and
density of domestic and transnational society, as well as the structure of government
institutions and the mode and scope of popular representation, will be key variables.
These variables will be interrelated: a strong civil society, for instance, is better able
to support representative government institutions. Conversely, non-representative or
oppressive governmental institutions can stunt the growth of domestic and
transnational civil society. By taking account of these variables, in contrast to the
uniform assumptions of State identity made by Realists, Liberal theory permits more
general distinctions among different categories of States based on domestic regimetype.
The best-documented empirical distinction between different types of States by
scholars working within the Liberal paradigm concerns the frequency of war among
liberal States, as compared to war between liberal and non-liberal States or among
non-liberal States alone. Liberal States are States with some form of representative
democracy, a market economy based on private property rights, and constitutional
protections of civil and political rights.10 These States are far less likely to go to war
with one another than they are to go to war with non-liberal States, giving rise to
what some scholars have termed the 'liberal peace'. 11 The claim is not that liberal
States are more pacific by nature, only that a variety of factors converge to reduce
the likelihood of military conflict between them.' 2
To date, the phenomenon of the liberal peace is better documented than
explained. Scholars have put forth a wide range of explanations focusing on
variables related to the structure of liberal government and the nature and impact of
10
11
12
This is the definition used by Michael Doyle in his pioneering work on the phenomenon of peace
among liberal States. See Doyle, 'Kant, Liberal Legacies, and Foreign Affairs', 12 Philosphy and
Public Affairs (1983) 205, 207-208 (hereinafter Doyle, 'Liberal Legacies').
Scholars and policymakers use both the terms 'democratic peace* and 'liberal peace'. I use the
term 'liberal peace', meaning the peace among liberal States, as a more accurate description of the
empirical phenomenon.
The strongest form of the claim is that liberal States have never fought a war with one another.
Many scholars, however, defining the criteria for both liberal State and war more broadly, would
say 'hardly ever'. See B. Russett, Grasping the Democratic Peace: Principles for a Post-Cold War
World(\993) 11-23.
509
liberal norms. 1 3 Other researchers argue that these are not causally relevant
correlations, that the likelihood of peace and liberal government both correlate with
other factors such as a high level of economic development, a high level of
economic interdependence, a particular cultural tradition, or simply the presence of
a common threat. 14 These scholars no longer deny that the liberal peace exists as an
empirical phenomenon, but they dispute any causal analysis that posits a direct
relationship between liberal governments and societies and the likelihood of war.
Proponents of the liberal peace counter that factors such as economic
interdependence have an additional impact on peace and may reinforce the
beneficial influence of democracy.15
For present purposes, however, the precise mapping of cause and effect does not
matter. That is a job for political scientists, who must determine whether democracy
leads to peace, or economic prosperity leads to democracy and to peace, or peace
leads to economic development which in turn strengthens democracy. In the
meantime, international lawyers can canvass the available political science literature
and note the presence of a set of attributes that correlate with regard to relations
among a particular subset of States. These attributes provide a basis for a more
generalized distinction between liberal and non-liberal States, a distinction that is
positive rather than normative.
The following list of correlative attributes includes peace, liberal democratic
government, a dense network of transnational transactions by social and economic
actors; 'multiple channels' of communication and action that are both transnational
and transgovernmental rather than formally inter-State; and a blurring of the
distinction between domestic and foreign isssues. From both a Realist and a
classical international legal perspective, it is jarring to mix descriptions of the
domestic characteristics of the States that function within this system with
descriptions of the general conditions under which such States interact with one
another. But from the perspective of Liberal international relations theory, these
various characteristics are all important dimensions of the relations between States
and domestic and transnational society.
13
14
15
510
Explanations put forth include the following: 1) democratically elected leaders do not need to turn
to democratic ambition as a means of legitimating their rule; 2) those who bear the costs of war are
those who decide whether to go to war, 3) citizens in a democratic State will respect the political
structure of other democratic States and be hesitant to go to war with them; 4) electoral process
produces risk-averse elites and centrist policies - attributes militating against war, 5) States willing
to submit to the rule of law and civil society domestically are more likely to submit to their
analogues internationally; 6) democratic debate exposes policy to the marketplace of ideas, thereby
allowing unsound ideas to be critically evaluated and challenged. This summary is borrowed from
Kupchan, Kupchan, 'Concerts, Collective Security, and the Future of Europe', 16 International
Securityi 1991) 114.
See K.T. Gaubatz, Still Hazy After All These Years: Kant's Secret Plan of Nature and the
Expansion of Democratic States in the International System (1993) (unpublished manuscript, on
file with author); Gowa, Mansfield, 'Power Politics and International Trade', 87 American
Political Science Review (1993) 408.
J.R. Oneal, F.H. Oneal, Z. Maoz and B. Russett, The Liberal Peace: Interdependence, Democracy,
and International Conflict, 1950-1986 (1994) (unpublished manuscript, on file with author).
A. Peace
The assurance of peaceful relations is the assurance that conflict will not escalate
into military conflict. It is important to recognize that this assumption does not posit
automatic harmony of State interests. Far from it, conflicts of interest are an
inevitable and important part of the international landscape. The posited difference
affects only the means of resolving those conflicts. The assured choice of nonmilitary means in turn establishes a different psychological and political context in
which to interpret economic and political measures taken on either side of the
conflict.
B. Liberal Democracy
Liberal democracy can be defined in many ways. As used here, it denotes some
form of representative government secured by the separation of powers,
constitutional guarantees of civil and political rights, juridical equality, and a
functioning judicial system dedicated to the rule of law. These particular features of
domestic political structure are important determinants of the interaction between
the State and individual and group actors in domestic and transnational society. 16
Representative government assumes that the government must be responsive to a
wide range of social actors; the guarantee of civil and political rights assures
individuals and groups the opportunity to interact in 'civil society' free of undue
interference from State organs;17 and the existence of a judicial system independent
of political direction makes available a neutral arbiter for private disputes arising in
domestic and transnational society.18
16
17
18
A world of liberal democracies is particularly susceptible (and may indeed particularly require)
Liberal analysis because it is among such States that domestic and hence transnational civil society
should be most developed.
The causal links between liberal democracy and civil society are confused and contested. On the
one hand, the civil and political rights that are the hallmark of liberal States are both the rights that
protect the individual from the State sufficiently to demarcate the State from society, and the rights
that permit and encourage voluntary associations among individuals. See R. Beddard, Human
Rights and Europe (3d ed., 1993) 2-3. Liberal States are thus the States that allow maximum room
for the development and flourishing of civil society, within and across territorial lines. On the other
hand, both democracy and the prosperity needed to support it appear to require not only specific
institutional forms, but also a measure of civic engagement. Haas, 'Beware the Slippery Slope:
Notes toward the Definition of Justifiable Intervention', in L.W. Reed, C. Kaysen (eds), Emerging
Norms of Justified Intervention (1993) 63, 79; R.D. Putnam, Making Democracy Work: Civic
Tradition in Modern Italy (1993).
Judges are both agents and shapers of domestic and transnational civil society. They are the
curbers of the State, creating the breathing space for individuals and groups to flourish; but they
are also the agents of individuals, resolving disputes, stabilizing expectations. The definition of an
'independent judiciary' is a judiciary that is not the handmaiden of State power, that answers to
law rather than to the individuals who make it. Such a judiciary can set itself against the State, but
can also regulate a realm in which the State does not intrude. In civil law systems it operates selfconsciously in a 'private law' sphere, applying codes designed to protect and foster private
activity; in common-law systems it responds to and regulates such activity.
511
C. Market Economies
Market economies based on private property rights also assure at least the existence
of an economic sphere distinct from the State, even if supported by State-created
rights and subject to State regulation. The relatively unconstrained ability to pursue
economic interest is an engine of social interaction, which in turn produces a climate
of trust that facilitates economic expansion.19 The pursuit of economic interest in a
market economy is also a source of demand for legal rules and institutions, for the
enforcement of property rights and the provision of the certainty and predictability
necessary to minimize risk and permit calculation of future gain.
D. A Dense Network of Transnational Transactions

Another factor that has been correlated with a reluctance to use force among liberal
democracies is a high level of transnational social and economic relations among
individuals and groups. Karl Deutsch argued in the 1950s that the density and level
of transnational contacts, to be measured through communication flows, was a
reliable indicator of a 'pluralistic security community', a group of States whose
members share a sense of common identity and who rule out the use of force in
resolving disputes. 2 ^ A second approach focuses more on economic contacts,
particularly on levels of economic interdependence. These scholars seek empirical
confirmation of claims dating back to Kant and the Manchester School about the
pacific consequences of trade. Yet they must contend with those who note that
interdependence can breed friction and suspicion as easily as harmony and trust. 21
The most promising recent scholarship in this debate suggests that economic
interdependence does have a positive impact on the likelihood of military conflict
among States, an impact independent from and in addition to the impact of
republican constraints on executive decision-making and democratic norms of
peaceful dispute resolution. 22
Yet a third approach suggests that democratic norms and structures and
economic interdependence reinforce one another in contributing to peaceful conflict
resolution. In a book that has become a classic in the American international
relations curriculum, Robert Keohane and Joseph Nye developed a model of
19
20
21
22
512
Again, the causal arrows here are unclear. Putnam argues that civic engagement permits economic
expansion; the converse observation that economic interest generates social interaction is easily
apparent from the expense budgets and golfing proclivities of business executives the world over.
Compare Susan Strange's discussion of the 'international business civilization' in Strange, 'The
Name of the Game', in N. Rizopoulos (ed.), Sea-Changes (1990) 238,260-265.
K. Deutsch et. al., Political Community and the North Atlantic Area (1957).
For a revival of the original Kantian argument, see Doyle, supra note 10, at 230-32; see also R.
Cobden, Political Writings (1867). Many scholars have countered with the example, inter alia, of
the high level of interdependence between Great Britain and Germany just prior to World War I.
See, e.g., S. Hoffmann, The State o/War(1965). Seeking to move past this debate, recent work has
emphasized different types of interdependence. Art, 'A Defensible Defense: America's Grand
Strategy after the Cold War', 15 International Security (Spring 1991) 5.
See Oneal et. al, supra note 15.

'complex interdependence' as an ideal-type of international relations.23 This model
correlated a reluctance to resort to the use of force among a group of States with
'multiple channels of contact connecting] societies'. 24 More recent work supports
the proposition that
[p]olitical and economic freedoms allow individuals to form transnational associations
and to influence policy in light of the resulting interests, inhibiting their governments
from acting violently toward one another.2^
The point here is not the mutual vulnerability that economic interdependence
implies, but rather the corollary network of transnational transactions and
communications among individual and group actors in liberal States. This is the
common denominator of all three schools.
E. Transgovernmental Communication
A subset of Keohane and Nye's model of complex interdependence is the
phenomenon of 'transgovernmental communication', the existence of 'informal ties
between governmental elites' and direct meetings and communications between
bureaucrats from different countries.26 These contacts coexist with 'formal foreign
office arrangements'. 27 As they recognized, this dimension of inter-State relations
cannot be accommodated within the traditional Realist conception of States as
unitary actors. 28 It suggests instead an image of what I will call 'disaggregated
sovereignty', the recognition of multiple actors exercising different types and modes
of governmental authority. And when combined with the more general phenomenon
of transnational communications among individuals and groups, it limits 'the ability
of foreign offices tightly to control governments' foreign relations'.29
23
24
25
26
27
28
29
R.O. Keohane, J.S. Nye, Power and Interdependence (1977) 25.

Keohane, Nye, Jr., 'Power and Interdependence Revisited', 41 International Organization (1987)
725, 731. Russett characterizes an extension of this approach as 'transnationalism', the claim that
'individual autonomy and pluralism within democratic States foster the emergence of transnational
linkages and institutions - among individuals, private groups, and governmental agencies'. He
argues, however, that this factor cannot be isolated as a variable independent of democracy itself in
explaining the liberal peace. Russett, Grasping the Democratic Peace, supra, note 12, at 26.
Keohane and Nye themselves originally wrote before the phenomenon of the liberal peace had
been identified, thus they made no effort to connect their model of complex interdependence to
relations among liberal States per se. They did note, however, that the reluctance to use military
force, a critical component of their model, was particularly observable 'among industrialized,
pluralist countries'. Power and Interdependence, at 27. I have thus included aspects of their model
as factors generally correlated with the liberal peace.
Oneal et. at, supra note 15, at 4.
Keohane and Nye, Power and Interdependence, supra note 23, at 25-26.
Ibid., at 26.
Keohane and Nye, Power and Interdependence Revisited, supra note 24, at 740.
Ibid., at 738.
513
F. Collapse of the Foreign/Domestic Distinction

The final attribute of inter-State relations that political scientists have correlated
with peaceful relations among a particular group of States is the absence of a
hierarchy of foreign affairs issues, a hierarchy that traditionally ranked the 'high
politics' of security above the 'low politics' of economics.30 The relative parity of
security with economic and environmental issues in turn breaks down the distinction
between foreign and domestic politics, as many issues reach the foreign policy
agenda that have a direct impact on domestic actors. 31 The ability of those actors to
influence policy decisions further decreases the traditional insulation of foreign
policy, a trend that is strengthened by the assurance of peace itself.
The standard equation of foreign policy with security policy is reflected in a host
of legal and constitutional arguments for the autonomy of executive control of
foreign relations, free of the normal political and legal constraints on policymaking.
Arguments justifying an autonomous executive sphere are typically based on the
need for secrecy, speed and flexibility in foreign policy, reflecting an underlying
assumption that national security is at stake. Conversely, the factors underlying the
disintegration of the foreign/domestic distinction in relations among liberal States
suggest that these arguments are of decreasing relevance. Foreign policy among
such States should be subject to the same constraints as domestic policy.
III. Hypothesizing a World of Liberal States

Taken together, the above six attributes describe a hypothetical world of liberal
States, a world of peace, democracy, and human rights. By combining these
attributes with the levels of analysis generated by Liberal international relations
theory, we can generate a model of international law in such a world. This part will
address the reasons for undertaking such an exercise and sketch the methodology for
conducting a self-professed 'thought experiment'.
Why bother even hypothesizing a world of liberal States? It is manifestly not the
world of traditional international law, a world that accepts many Realist assumptions
about States as functionally identical unitary actors seeking primarily to preserve
their own sovereignty. It is not the world of contemporary international politics, of
Bosnia, Haiti, Rwanda or China. Nor is it a world likely soon to emerge. Do we then
hypothesize such a world solely as a Utopia? Is the point simply to perform an
abstract interdisciplinary exercise?
The answers are several. First, part of the world, a growing part, is composed of
liberal States. The attributes listed above may be a more accurate description of
relations among countries in this part of the world than the traditional model of the
international system on which classical international law is based. If so, a model of
30
31
514
Keohane and Nye, Power and Interdependence, supra note 23, at 25.
Ibid., at 26-27.
international law derived from these attributes will capture more of the legal and
political reality of relations among these countries. Legal relations among States
such as the United States, Canada, the Member States of the European Union, Japan,
Australia, and New Zealand are most likely to fit this model, but it could also at
least provide a point of departure for conceptualizing the legal relations among
Argentina, Chile, Brazil, Ecuador, and Mexico; Poland, Hungary, and the Czech
Republic; Taiwan, South Korea, and the Philippines; India, Israel, and, with luck,
South Africa.
Second, to the extent that this model shapes our expectations of how law
operates among liberal States, it will also generate a corollary set of expectations
concerning legal relations between liberal and non-liberal States. These twin sets of
expectations will in turn provide the conceptual tools to grasp the differential
significance of apparently universal phenomena. As a formal matter, for instance, all
supranational tribunals are expected to wield the same authority. The judgments of
the International Court of Justice and the European Court of Justice should thus
have an equal impact on State behaviour. In practice, of course, our expectations of
the effectiveness of these two tribunals vary considerably, but without justification
or explanation in the language and conceptual framework of classical international
law. A model of law among liberal States that analyzes the enforcement of
international agreement in light of the underlying political configuration and the
nature of individual and group interests in the States involved can instead generate a
set of conditions in which we might expect supranational tribunals to be relatively
more or less successful.
In both of the above categories, the model will provide a basis for empirical
testing of specific hypotheses concerning legal relations among liberal States and
between liberal and non-liberal States. If these hypotheses hold, the next step will be
to develop a corresponding set of norms within each category. The positive model
cannot itself give rise to normative propositions. Yet to the extent that the positive
model gives rise to a different conceptualization of the principal actors engaged in
legal relations and the nature of the relations between them, it will define the
subjects of new norms and the type of activity such norms are designed to regulate.
For instance, I will argue that legal relations among liberal States are characterized
by 'disaggregated sovereignty', in which the 'State' is disaggregated into its
component political institutions, each of which is bound to one another in a
constitutionally determined set of relationships. This conceptualization would then
require us to generate norms defining the content and limits of the sovereignty of
each of these component institutions.
The third potential advantage of developing a hypothetical model of legal
relations among liberal States flows from this normative project. We may find that
in some instances it will be more attractive to use the model to generate a universal
set of concepts and norms, applicable to liberal and non-liberal States alike. In many
cases we are likely to find that relations between liberal and non-liberal States
display some of the features of the model but not others. This congruity is to be
515
expected to the extent that non-liberal States display some of the political,
economic, and social attributes described above but not others. In such cases, it may
be preferable to accept the necessary fiction inherent in applying a positive model
and its corollary norms to States that only partially fit the model than to sacrifice the
principle of universality.
Assume, then, a world of liberal States. Assume further that the vision of the
international system developed by Liberal international relations theory is correct:
that the primary actors in this system are individuals and groups operating in
domestic and transnational society; that States seek to regulate these individuals and
groups but simultaneously represent some aggregation of individual and group
preferences; that the strength of State preferences determines the outcome of interState interactions. What would international law in such a world look like? What
rules would we expect to govern transnational and inter-State transactions?
'International law', as defined here, comprises all the law that regulates activity
across and between territorial boundaries. It can include the law of peoples and the
law of nations, the jus gentium and the jus inter gentes. Existing categories and
distinctions such as public and private, domestic, transnational and international are
immaterial. The identifying element that qualifies a rule or set of rules for inclusion
in this category is the potential for contribution to international order, whether by
constraining domestic forces that might otherwise escalate international disputes
into military or severe economic conflict, by strengthening or regulating transactions
in transnational society, or by directly regulating inter-State relations. The resulting
body of 'law' is defined not according to subject or source, but rather in terms of
purpose and effect, in conformity with a particular body of international relations
theory.
IV. International Law in a World of Liberal States

To develop a model of international law in a world of liberal States, I combine the
levels of analysis generated by Liberal international relations theory with the
political, economic, and social attributes of a world of liberal States. As discussed
above, Liberal international relations theory offers a set of fundamental assumptions
about the international system: about its primary and secondary actors, their
preferences, and the constraints imposed by the system itself. In accordance with
these assumptions, an effort to think about legal relations within the international
system would focus first on relations among individuals and groups in transnational
society; second on State institutions in relation to these social actors; and third on
inter-State interactions where State preferences are a changing function of individual
and group interests as those interests are themselves defined in domestic and
transnational society. Within each of these categories, the social actors and State
institutions involved are assumed to interact with one another as dictated by the
ideological and structural principles of a liberal State.
516

The first level comprises a network of transnational relationships among
individuals and groups governed largely by rules of their own choosing. These rules
may not be 'law' per se, but rather voluntary norms adopted by a wide range of
professional associations. Alternatively, individuals and groups may choose preexisting bodies of national law to govern transnational contracts. They may also
choose where they would like their disputes resolved, and by which tribunals, either
arbitral or judicial. Individuals and groups are the primary actors within this
network. State institutions are relevant primarily to the extent that they promote or
hinder these voluntary choices, through doctrines governing personal jurisdiction,
forum selection, parallel litigation, evidence-gathering, and enforcement of arbitral
agreements and awards. States may also decide to codify voluntary norms. The
policies underlying these doctrines and decisions bear directly on the depth and
breadth of transnational society, with resulting impact on the pressures that social
groups can exert on representative institutions, the transmission of knowledge about
foreign legal systems, and the socialization of individuals and groups subject to
multiple bodies of national regulation.
The second level is the level of domestic governmental institutions, primarily
courts and legislatures, interacting directly with one another in the process of
making, selecting and enforcing the law governing transnational transactions. The
'State' at this level is disaggregated into its component parts: legislative, executive,
administrative and judicial. The assumption of liberal States permits further
assumptions about the representativeness and relative independence of these
institutions within their respective spheres, resting on a common foundation of core
protection of individual rights. The legal relations among these institutions
primarily comprise questions of overlapping national jurisdiction, embodied in
doctrines governing choice of law and extraterritorial application of law. By
departing from the traditional conception of States as unitary actors on the
international stage, it is possible to understand these relations in the context of a
transjudicial and trans-legislative dialogue, in which courts and legislatures
acknowledge and evaluate foreign law. It also facilitates the conceptualization of a
transnational legal process encompassing disaggregated judicial, legislative and
executive interaction.
The third level is the level of inter-State interaction, in which States must present
themselves as unitary actors at least to the extent of having a sole representative in
international negotiations. The difference between the analysis presented here and
traditional public international law is that the States involved are presumed to be
embedded in transnational society. The unitary facade is thus in fact very porous,
open to penetration by individuals and groups. Further, the 'aggregation' of the
political institutions themselves for purposes of international negotiation will be
quite loose. The executive, as negotiating representative, will be accountable to the
legislature; acts of the legislature approving and implementing the results of
executive negotiation will be enforceable by the courts. These conditions affect the
origins, form, negotiation and enforcement of international agreements.
517
Before proceeding to a more detailed analysis within each of these categories, a
final methodological note is in order. The methods used to imagine or conceptualize
legal relations in a world of liberal States are both deductive and inductive. They are
deductive to the extent that they follow logically from the various assumptions and
attributes already discussed. They are inductive to the extent that they draw on
anecdotal examples of relations among liberal States, such as the States of the
European Union or the OECD. Whether these examples will in fact prove
representative of the larger phenomena postulated here is a subject for empirical
testing; I offer them here to add concrete form to abstract propositions and to
stimulate further expansion of existing categories of thought.
A. The Voluntary Law of Individuals and Groups in Transnational Society

Much of the law governing relations among economic and social actors operating
across borders is a decentralized network of self-selected and customized rules.
These actors choose both the substantive law governing their association and the
mode and forum for dispute resolution. Territorial boundaries impose no
restrictions, but only delimit the menu of choices. Participants are free to choose the
product of one legislature and the procedures and competence of another tribunal. A
contract might thus specify that disputes will be tried in London under New York
law. The resulting freedom and flexibility is particularly important for the efficient
functioning of market economies.
1. Formal and Informal Voluntary Norms
The voluntary law governing transnational society divides into several categories.
First is a category of proto-law generated by a wide range of business and
professional organizations. In the domestic context, Robert Cooter has described
these rules as the 'new law merchant', voluntary norms adopted by corporate
networks, self-regulating professions, and business associations.32 Many of these
networks and associations extend transnationally as well, generating an
accompanying network of transnational voluntary norms. These norms may not
seem like law at all. Yet scholars and practitioners seeking to predict actual
behaviour must take them into account as empirical facts that guide action. Further,
as will be discussed below, these bodies of rules may be templates for future law.
A second category of rules governing transnational commerce is the law selected
by individual actors to govern the interpretation and application of bilateral
commercial agreements and the mode of resolving disputes arising out of those
agreements. The actors enter into contractual relations. Their interaction is to be
32
518
Examples include rules promulgated by the VISA network, the American Bar Association ethical
guidelines, and standards promulgated by the Association of Home Appliance Manufacturers.
Cooter, 'Structural Adjudication and the New Law Merchant: A Model of Decentralized Law',
International Review ofLaw and Economics (1994) 143, 144.

governed by the contract. But they also determine what law shall govern the contract
and where and how disputes arising out of or related to the contract will be resolved.
They can choose either a judicial or an arbitral forum. If they choose arbitration,
they can also write their own rules of procedure governing the dispute resolution
process.
Within international commercial arbitration, individuals and groups may find
themselves, either by specification or not, governed by an independent body of law
developed by international commercial arbitrators on the basis of customary
transnational business practice. These rules have also been referred to as the 'new
law merchant', or lex mercatoria. The arbitrators derive their authority from the
arbitral agreement. They themselves form an informal network,33 such that they can
effectively draw on their collective experience in distilling and applying these rules.
Formally, this body of law has no more status than voluntary norms developed and
adopted by professional associations. It similarly evolves from social and economic
practice in transnational society.
2. State Facilitation of Individual Choice
Thus far, the law of transnational society has been presented as essentially stateless.
This is an accurate depiction in that individuals and groups are the primary actors;
State functions are ancillary to individual choices. But the role of the State is
nevertheless critical to the functioning of the system. It appears in several guises.
First, of course, the State provides the bodies of rules available for selection by
individual and group actors. State legislatures thus provide a public good. Second,
States may ultimately adopt and codify the norms that individuals and groups have
already developed through voluntary associations. Merchant custom, for instance,
formed the basis for much of the common law. Today scholars from law and
economics often urge the adoption of these voluntary codes by national legislatures.
In this regard, developments in transnational society lay a foundation for subsequent
State action.
Third, States determine to what extent bilateral commercial agreements will be
enforced, thereby facilitating or blocking individual choice. This point has been
most frequently recognized in the arbitral context; the availability of judicial
enforcement - the coercive apparatus of the State - undergirds the entire system of
international commercial arbitration. Thus States wishing to promote and facilitate
commercial arbitration authorize their domestic courts both to compel arbitration at
the request of one of the parties to the arbitral agreement and to enforce the ensuing
award.34 In the multilateral context, Michael Reisman observes that the New York
33
34
See Dezalay, Garth, 'Grand Old Men vs. Multinationals: The Routinization of Charismatic
Arbitration into Off-Shore Litigation', Sees. IV, VI (American Bar Foundation Working Paper,
No. 9317, 1994).
In the absence of State action, private trade associations can also provide enforcement mechanisms
by enhancing reputation costs. This was the purpose of the associations arising out of the medieval
'champagne fairs'. Similar mechanisms, albeit less effective, are provided by Better Business
519
Convention 35 is effectively an agreeement whereby each State acknowledges the
legitimacy of each other State's control of the arbitral process and thus agrees to
enforce the results. 36
Equally true, although less often observed, is the role of the State in facilitating
individual choice in transnational litigation. The parallel is easiest to see in doctrines
governing forum selection and choice of law where a particular forum and
governing law is specified by agreement. Similarly, doctrines governing parallel
litigation, such as the issuance of antisuit injunctions, often function to prevent one
party in transnational litigation from circumventing the forum and law specified in a
contract. Yet doctrines governing service of process, personal jurisdiction, forum
non conveniens, transnational discovery of evidence, and enforcement of foreign
judgments can also be analyzed from the perspective of facilitating or hindering
dispute resolution in transnational society. The interpretation and application of
these doctrines typically arise where one party to an international transaction
chooses a forum to which another party objects. Party A seeks to invoke the power
of the forum State to conduct the litigation on its terms; Party B may in turn assert
the power of a foreign State as a counterweight. But the forum itself, in hearing and
pronouncing on these arguments, is less an agent of one particular sovereign State
than one of a number of multiple centres of authoritative dispute resolution available
to actors in transnational society.
3. Theoretical Implications
From a traditional Realist 'billiard ball' perspective, all of the above doctrines
ultimately pose questions of State power. The issue is whether a particular
transaction or the parties to it fall within one or another sovereign sphere. In case of
disagreement on this score, the only solution is to mediate conflicting claims of
power by multiple competing sovereigns. Assuming a world of liberal States,
however, a number of very different features emerge. First, transnational economic
and social interactions are an important and enduring dimension of relations among
liberal States. Second, all potential dispute resolution fora within the realm of liberal
States, whether judicial or arbitral subject to judicial supervision, are assumed to be
neutral and independent of direct political influences. Third, the core rights of
individual litigants are assumed to be constitutionally protected in every potential
forum. One such right is the right of a litigant not to be forced to defend in a forum
with which she has no connection. Given these preconditions, participants in
transnational transactions have multiple potential fora from which to choose. To the
extent that the provision of dispute resolution mechanisms is regarded as a core
function of a unitary State, the proliferation of such choices contributes to the
35
36
520
Bureaus. See generally Milgrom et al, "The Role of Institutions in the Revival of Trade: The Law
Merchant. Private Judges, and the Champagne Fairs', 2 Economics and Politics(1990) 1.
Convention on the Recognition and Enforcement of Foreign Arbitral Awards, opened for
signature, lOJune 1958,21 UST2517, TIAS No. 6997,330UNTS 38.
M. Reisman, Systems of Control in International Adjudication and Arbitration (1992) 139.

dispersal of the core functions of the State across a particular transnational
community.
Even doctrines that appear to enforce one party's choice of forum over another
party's objection look quite different in the context of transnational transactions
among liberal States. Given the assumed procedural and substantive safeguards, the
primary import of these doctrines is the translation of economic interdependence
into a set of diagonal relationships between individuals and groups and States other
than their home State for the limited sovereign function of dispute resolution.37 The
contacts with a foreign legal system that either give rise to an economic or social
transaction in the first instance, or that flow from such a transaction undertaken for
other reasons, become the basis for a limited relationship with a foreign sovereign, a
sovereign that can be guaranteed to provide the same minimum safeguards in the
performance of this sovereign function that the individual's home sovereign would
provide.38 The equation is not a straightforward transformation of economic power
into legal power, but rather a decision in favour of one individual's choice of forum
over another's based on an assessment of a relationship between the individuals
involved and each of their connections with different States.
The density and velocity of transnational transactions among liberal States is
related to the phenomenon of the 'liberal peace', although the precise causal
relationships have not been mapped out. We may speculate that cross-fertilization
among legal systems may play some role; knowledge of multiple legal means to
similar ends in systems with similar safeguards might dampen perceptions of
difference and 'otherness' so likely to escalate conflict; opportunities for
transnational formulation of norms through business and professional associations
may facilitate legal convergence. It may also be that subjection to multiple bodies of
law has an important socializing dimension, transferring individual expectations to
more than one source of sovereign power. Finally, the picture of transnational
society presented here has an important self-governing dimension. Whether through
professional associations or bilateral contracts, individuals and groups in
transnational society are often able to make and enforce their own rules and to
customize their own modes of dispute resolution, choosing among a number of
disaggregated State 'products'. Power and responsibility are equally dispersed,
laying the foundation for a hypothetical transnational polity with multiple centres of
authority.
37
38
See L. Brilmayer, Justifying International Acts (1989) 55-57, 84-87 for a discussion of 'diagonal
relationships' crossing the 'horizontal' relationships among States and the 'vertical' relationships
between citizens and their home State.
In a similar vein, Lea Brilmayer uses jurisdictional analysis to determine when a sovereign owes
the same duties to a foreign citizen as it does to its own. Brilmayer, supra note 37, Chapter I.
521
B. The Law of Transnational Governmental Institutions

If the first level of law among liberal States focuses on the rules encouraging and
strengthening the formation and development of transnational society, the second
level is the law governing relationships between governments and individual and
group actors in transnational society. This law subdivides into two general
categories: 1) the substantive law directly regulating these relationships, typically
the domestic law of one or more States, and 2) the legal or quasi-legal principles
regulating the interaction of governments themselves concerning the appropriate
division of regulatory jurisdiction. At present, the relevant doctrines that would fit
within these categories are an amalgam of private choice of law rules, public
international law rules governing jurisdiction to prescribe and enforce, diverse
domestic principles governing statutory interpretation, and various procedural rules
governing the treatment of foreign governments in domestic court.
The traditional public international law doctrines governing State jurisdiction to
prescribe and enforce conduct focus on the demarcation of horizontal relationships
among competing sovereigns, relying on principles of territoriality and nationality
as the legal corollaries of the physical limits of political power. Private international
law doctrines, on the other hand, encompass a wide range of potential approaches,
from territoriality to State interest-balancing to efforts to quantify and qualify the
relationship between an individual or a dispute and a particular regulatory authority.
Given its focus on State representation and regulation of individual interests, Liberal
international relations theory would fit most comfortably with the amalgamated
approach originally developed by Andreas Lowenfeld and adopted by the Third
Restatement of American Foreign Relations Law, an approach combining links
between the individual and her conduct and the States seeking to regulate that
conduct, State interests, and individual expectations.39
The assumption of a world of liberal States would refine this analysis further by
disaggregating the State itself. In such a world, each State is composed of roughly
symmetrical and homogeneous domestic political institutions: representative
legislatures, separate representative executives and administrative agencies, 40 and
independent judiciaries. Each of these institutions, in turn, has a roughly similar
relationship with individuals and groups in domestic society, performing the same
functions within a broadly similar set of constraints designed to safeguard individual
rights. It thus becomes possible to conceive of the State not as a unitary actor, but
rather as a constellation of governmental institutions performing the three basic
functions of law-making, execution, and enforcement, on the assurance that each
State in our hypothetical world will replicate this structure.
39
40
522
See Lowenfeld, 'Public Law in the International Arena: Conflict of Laws, International Law, and
Some Suggestions for their Interaction', 163 Hague Academy of International Law (1979-11) 311;
Restatement of Foreign Relations Law of the United States (Third), paras. 402-403.
Even in parliamentary systems, the 'government' is distinct from the legislature; cabinet ministers
in charge of different government departments often act on their own initiative.

Within each liberal State, these institutions form multiple centres of power, each
reinforcing and checking the other. In international matters, however, the perceived
threat to the State as a whole has often led legislatures and courts to line up behind
the executive, the branch of government best equipped to engage in high-stakes
international diplomacy. The assurance of peace in a world of liberal States, coupled
with a presumption of stable and continuous relations with other liberal States,
would be likely to remove these traditional impediments to independent judicial and
legislative action. Executives and courts, legislatures and courts, legislatures and
administrators could thus all interact across borders in multiple ways. Each would
represent a facet of a particular State, but national allegiance would be counterbalanced by links to counterpart institutions performing the same functions in
transnational society. The result would be a web of relations among multiple centres
of State authority within and across borders.
Interactions among these institutions will typically be triggered by the interaction
of individuals and groups in transnational society. Thus, for instance, a dispute
between two citizens of different States regarding a transaction that took place in
both States is likely to bring the courts of both States into contact with one another,
directly or indirectly, to the extent that the two individuals involved each prefers to
litigate the dispute in her home State. The same dispute may also bring the
legislatures of the two States into contact with one another, directly or indirectly, to
the extent that the two courts involved disagree over which legislature's
prescriptions should apply to the conduct at issue. Alternatively, if only one court is
involved, its disregard of a foreign legislature's prescriptions may invite a response
from the foreign legislature itself or the foreign executive charged with
implementing those prescriptions.41
The substantive legal rules applicable to a particular class of individuals or
groups or of conduct in transnational society will thus be determined in the context
of an interaction between the individuals and groups involved and two or more
governmental institutions: courts, legislatures, executives and administrative
agencies. The question then arises as to what principles should govern the
interaction not of 'States', but of these specific institutions in interactions with one
another in the larger context of regulating transnational society. To answer this
question, it is helpful to begin first by analyzing actual and potential relations among
like institutions across borders - court to court, legislature to legislature - in terms
both of the principles that might govern an ongoing 'dialogue' between these
institutions and the functions that such a dialogue might perform as part of a process
of transnational governance. The final section examines interactions among all three
institutions.
41
See Lowenfeld, supra note 39, at 330-331 (describing clashes of regulatory jurisdiction as
involving two courts, one executive and one legislature).
523
1. A Transjudicial Dialogue
Courts in a world of liberal States would recognize each other as like units,
dedicated to and legitimated by the same core principles of the rule of law:
assumptions of impartiality in adjudication, of the separation of the political and
judicial branches, and of the equality of all citizens before the law. They are likely
to interact with and take account of one another in a form of transjudicial dialogue.
This dialogue can be initiated by individuals, in the sense that the multiple
transnational contacts of parties to a particular case raise choice of law questions.
Alternatively, courts themselves can initiate the dialogue by seeking information as
to how their counterparts in foreign countries resolve similar questions. This
dialogue can take several forms, for several different purposes.42
First, the application of traditional choice of law rules can be conceptualized as a
reciprocal dialogue in which courts of different States are engaged in a common
endeavour to make transnational relations among individual and groups more certain
and predictable while taking account of multiple State interests. Some conceptions
of this dialogue would go further and envision a common effort to select and
harmonize substantive rules. The significance of postulating a dialogue is that a
court of one State may assume that its foreign counterpart will respond to signals of
accommodation or confrontation in an ongoing conversation. Some conflicts-of-law
scholarship has gone so far as to model this dialogue in game theoretic terms; other
scholars argue that traditional notions of comity and reciprocity have long captured
this phenomenon. 43
The assumption of a world of liberal States means that this dialogue takes place
in a setting in which national courts face fewer constraints on their application of
foreign law. All courts can assume that a genuine 'choice of law' exists, in the sense
of a choice between two rules arrived at by a process that both polities and societies
would recognize as legitimate. The paradigm is set forth by a US court in the case Bi
v. Union Carbide Chemicals & Plastics Co.,44 in which the court recognized and
gave effect to an Indian law on the ground that 'India is a democracy'.45 The court
looked only to affirm that the law in question had been 'passed by [India's]
democratic parliament' subject to review by an independent judiciary.46
42
43
44
45
46
524
For a fuller exposition of different forms of transjudicial communication, see Slaughter, 'A
Typology of Transjudicial Communication', University of Richmond Law Review (forthcoming
1995).
See, e.g., L. Brilmayer, Conflict of Laws: Foundations and Future Directions (1991) 155-167
(discussing game theoretical approaches in which actors have both common interests that they
attain through cooperation and conflicting interests they realize at each other's expense); Larry
Kramer, 'Rethinking Choice of Law', 90 Colum. L Rev. (1990) 277,339-344 (discussing choice of
law in terms of prisoner's dilemma hypothetical); see also Weinberg, 'Against Comity', 80
Georgia Law Journal (1991) 53 (asserting that game theory as applied to conflicts of law
reincarnates traditional notions of comity and reciprocity).
984 F.2d 582 (2d Cir. 1993).
Ibid., at 586.
Ibid.

Further, the process of choosing and applying a foreign law takes place within
what I have called a 'zone of legitimate difference'.47 Courts will presume the
legitimacy of a wide range of different means to achieve similar ends.48 Where they
do refuse to apply an otherwise applicable foreign law, it is likely to be based on the
perception that although the societies in question adhere to the same fundamental
values, the potentially applicable laws reflect different choices as to which of those
values should trump. Such a refusal is to be contrasted to a more general refusal
even to assess the validity of the foreign law in cases in which the societies in
question diverge not on specific choices between competing values, but on the
identity and validity of the underlying values themselves.
Finally, in all of these instances the institutions involved can be expected to
recognize, either implicitly or explicitly, a common core of political and economic
values that will preserve roughly the same boundary between 'political' and 'legal'
questions as would exist in purely domestic cases. In a world in which courts and
lawmakers in one country can find no recognizable counterpart in another, or in
which foreign policy concerns connected to a particular dispute might raise even the
distant spectre of military conflict, courts may understandably cede the field to the
political branches. But where such counterparts are readily ascertainable and peace
is assured, disputes raising questions of international politics should be as amenable
to judicial resolution as disputes raising questions of domestic politics. The degree
of that amenability will vary by country, depending on the degree to which the
judicial system in question recognizes some domestic variant of the American
political question doctrine.
A second version of transjudicial dialogue may occur where the courts of liberal
States seek to improve the quality of their decision-making on important public
issues. Mary Ann Glendon has documented the increasing willingness of European
constitutional courts to take account of the solutions crafted by their foreign
counterparts on issues ranging from free speech to abortion. 49 The German
constitutional court, for instance, maintains a library of constitutional decisions
from courts around the world. ^ The presupposition here is of multiple societies,
similarly structured and dedicated to the same basic values, grappling with the
resolution of conflicts of those values. The institutions charged with the resolution
of those conflicts communicate the fruits of their common experience to one
another.
Third, courts may communicate with one another to protect or persuade their
home governments. Joseph Weiler has documented this phenomenon among
47
48
49
50
Burley, 'Law among Liberal States: Liberal Internationalism and the Act of State Doctrine', 92
Colum. L Rev. (1992) 1907, 1919, 1946.
The American judge (later Justice) Benjamin Cardozo put it best: 'We are not so provincial as to
say that every solution of a problem is wrong because we deal with it otherwise at home...', Loucks
v. Standard Oil Co., 120 N.E. 198, 201 (N.Y. 1918).
M.A. Glendon, Rights Talk (1991) 158-159.
D. Kommers, The Federal Constitutional Court (American Institute for Contemporary German
Studies, Key Institutions of German Democracy No. 2, 1994) 17.
525
national courts in the European Community, observing that in the context of
common supranational obligations courts seek to ensure that acceptance of these
obligations as a matter of national law will not disadvantage their governments
relative to other governments. They thus canvass comparable decisions by their
foreign counterparts. Alternatively, where a national court wants to urge compliance
with a supranational obligation, it may argue that it is just keeping pace with the
decisions of its foreign counterparts. 51 The result is a transjudicial dialogue that
shapes domestic judicial conversations with national executives and legislatures.
2. A Transnational Legislative Dialogue
The assumption of a world of liberal States allows us to assume competing spheres
of public regulation of represented individuals. Moreover, we assume a dense
transnational society, such that laws passed by one legislature will inevitably have
an impact on individuals and groups outside the national territory and citizenry. And
we assume that those affected individuals and groups will have access to their own
legislatures.
The upshot of these assumptions is that, like courts, national legislatures should
also be presumed to be engaged in a reciprocal dialogue with their foreign
counterparts. Legislative measures that adversely affect individuals and groups
outside the national polity will invite retaliation where those individuals and groups
have sufficient power in their home countries to secure a foreign legislative
response.
These assumptions should inform judicial analysis of extraterritorial application
of national legislation. This analysis includes an initial determination of legislative
power to legislate extraterritorially and a secondary determination of whether in fact
the legislature intended to exercise that power with regard to the statute before the
court. The result may well be a reinforced presumption of 'legislative comity', the
deference that one legislature gives to another.52 At the very least, courts might be
expected to adopt principles of statutory interpretation designed to ensure that a
legislature has taken account of retaliatory possibilities.
In the normative realm, the conception of legislatures operating in dialogue with
one another in a world of liberal States could be used to generate a set of principles
that could function as an informal transnational constitution. Common principles of
limited government, individual liberty, and the certainty and predictability inherent
in the rule of law could be generalized across borders to mandate presumptions in
51
52
526
Weiler, 'A Quiet Revolution: The European Court of Justice and Its Interlocutors', 26
Comparative Political Studies (1994) 510, 521-522.
Compare Justice Scalia's distinction between legislative and judicial comity in Hartford Fire
Insurance Co. v. California, 113 S.Ct. 2891, 2920 (1993) (Scalia, J., dissenting). Scalia
distinguishes 'the comity of courts, whereby judges decline to exercise jurisdiction over matters
more appropriately adjudged elsewhere' and '"rescriptive comity": the respect sovereign nations
afford each other by limiting the reach of their law'. Id. Prescriptive comity is 'exercised by
legislatures when they enact laws, and courts assume it has been exercised when they come to
interpreting the scope of laws their legislatures have enacted'. Id.

favour of extraterritorial application of any statutes designed to limit the exercise of
governmental authority and against the extraterritorial application of statutes
restricting individual liberty where the individuals in question are demonstrably
subject to competing regulation by their home legislatures. Further, common
principles of representative government could lead to the adoption of principles
designed to protect distortions of national legislative processes by rent-seekers
trying to exploit transnational regulatory gaps. 53
3. Disaggregated Transnational Judicial, Legislative, and Executive Interaction
A final aspect of this second level of law among liberal States is the interaction of
the three domestic branches of government in each State transnational^ with one
another in much the same way that they would interact with their co-branches of
government at home. Thus, for instance, executive branches have to take account
not only of their own courts, but of the decisions of foreign courts as having an
impact on the concerns and constraints of diplomacy. 54 They might choose to
respond to a foreign judicial decision by pressuring a fellow executive to urge a
reversal or reinterpretation. Alternatively, however, they can be expected to
intervene directly in foreign litigation by presenting their views to the foreign court
itself. Finally, a struggle between two national courts may require executive
intervention on both sides.
Conversely, courts may choose to treat foreign executive branches in the same
way that they relate to their own national executive. The United States Court of
Appeals for the Seventh Circuit, for instance, chose to defer to a ruling by the
French executive branch under the same US doctrine that requires deference to US
administrative agencies.55 In the process, it acknowledged the common function of
the two entities rather than differentiating between them by drawing a line between
foreign and domestic affairs and waiting for a communication from the branch of its
own government charged with the conduct of foreign affairs.
With regard to executives and legislatures, executive branches can lobby foreign
legislatures directly through national law firms. Legislatures can also conduct their
own diplomacy through delegations of national legislators visiting foreign
executives and consultations with foreign parliamentarians. Alternatively, a
legislature might respond to the actions of a foreign court, such as the British and
53
54
55
See Turley, 'Dualistic Values in the Age of International Legisprudence', 44 Hastings Law
Journal (1993) 185, 243-245 (arguing for the elimination of the presumption against
extraterritoriality to prevent exploitation of legislative processes by multinational rent-seekers).
To take one recent example, the executive and legislative branches of the European Community
found themselves waiting in suspense to see how the US Supreme Court would rule on the validity
of a unitary corporate tax imposed by the State of California. US Supreme Court to Hear Unitary
Tax Case, 5 Eurecom 1 (No. 10, November 1993) (reporting the decision of the US Supreme Court
to grant certiorari in Barclays Bank Pic v. Franchise Tax Board of California in accordance with
'EC and UK hopes, and disagreeing with the Clinton Administration').
In the Matter of: Oil Spill by the Amoco Cadiz Off the Coast of France on March 16, 1978, 954
F.2d 1279, 1312-1313 (7th Cir. 1990) (percuriam).
527
Australian legislatures did in enacting 'clawback' statutes to blunt US courts'
extraterritorial application of US antitrust laws. 56 Finally, legislatures can pass
resolutions requiring their executives to take a particular position in negotiations
with foreign executives.
All of these interactions are sharply at variance with the classical 'billiard ball'
image of inter-State relations. Each of the governmental institutions involved acts
with a degree of autonomy determined as much by its specific function with regard
to individuals and groups in domestic and transnational society as by its State
allegiance. Moreover, each of these institutions acknowledges the autonomous
functioning of counterpart and complementary functions. As discussed in Part V, the
result challenges our conception of State sovereignty itself.
C. The Law of Inter-State Relations

The third level of law in a world of liberal States is the law governing inter-State
relations, the traditional province of public international law. This body of law
includes inter-govemmental agreements, customary international law, and the law of
international institutions. Liberal international relations theory combined with an
assumption of liberal States generates insights and predictions in all three areas. The
discussion here, however, will focus only on inter-governmental agreements.
Assuming a world of liberal States leads to predictions about the origins, nature,
incidence, negotiation and enforcement of such agreements. The baseline
assumption is that governments will interact with one another within a web of
individual and group contacts in transnational civil society. This larger context will
affect inter-governmental agreements in a variety of ways. For instance, the plethora
of individual and group transactions will obviate the need for some kinds of
agreements altogether, such as treaties providing for cultural and educational
exchanges and certain kinds of economic assistance. On the other hand, a need for
national regulation is likely quickly to become a need for transnational regulation. In
addition, the negotiators to all inter-governmental agreements in this world will be
responsible to national legislatures, which themselves must be responsive to
domestic and transnational group pressure. Finally, all States parties to all intergovemmental agreements will have well-functioning domestic legal systems, with
national courts accustomed to laying down the law.
1. Origins of Inter-Governmental Agreements among Liberal States
Regarding the origins of inter-govemmental agreements, imagine an idealized
spectrum running from a world composed only of non-liberal States on one end and
only of liberal States on the other. In a world only of non-liberal States, the initiative
56
528
See G. Bom, D. Westin, International Civil Litigation in US Courts (2d. ed., 1992) 600-603 for a
discussion of various foreign responses to extraterritorial application of US antitrust laws.

for inter-State transactions should come from State actors, on the simplified
assumption that they control all important aspects of the polity, economy and
society.57 In the middle of the spectrum are relations between liberal States and nonliberal States, in which individual and group actors within the liberal State might
push for inter-governmental agreements to foster their interests with respect to the
non-liberal State. Such interests might include family reunification and visitation,
communications of various sorts, humanitarian aid, and targeted economic
assistance. Individuals and groups in the non-liberal State might also encourage
economic agreements with the non-liberal State designed to encourage that State to
liberalize its economy so as to create opportunities for foreign investors and traders.
Finally, at the far end of the spectrum, we would expect a dense network of
individual and group activity. These actors do not require State intervention to spur
or substitute for transnational activity, but only to regularize and channel it. Most
agreements would be designed to address specific issues or problems arising out of
ongoing activity.
Another important category of inter-governmental agreements among two or
more liberal States should flow from transnational regulatory activity. Domestic
regulators must follow the actors they seek to regulate. The assumption of a dense
transnational society assumes that regulated activities will spill over borders,
requiring regulators to interact with their foreign counterparts. Both conflict and
informal cooperation are likely to result. In the case of regulatory conflict, an intergovernmental agreement may be necessary to coordinate or harmonize regulatory
action. Where a range of informal cooperative practices develop, an agreement may
be useful to institutionalize them as the basis for further cooperation.
In all cases involving liberal States, however, a fairly lengthy period of relatively
decentralized transgovernmental interaction is likely to precede and shape the
agreement. The parties involved can count on long-term, relatively stable
relationships, and at least a presumption of trust and good faith. Counterpart
agencies may proceed by trial and error as they grope for common formulae. Less
cooperatively, multiple branches of government may become involved across
boundaries, as in a confrontation between a court in one nation and the executive
branch of another, or a stand-off between two legislatures. Even as the rhetoric heats
up, however, all parties are likely to assume that the pulling and hauling will
ultimately result in a stable informal equilibrium or a negotiated solution. Where an
inter-governmental agreement results, it will be not the result of a deliberate
'foreign' policy decision, but rather of the evolution, coordination or harmonization
of pre-existing domestic policy.
57
'Non-liberal' State here refers to a State that has neither a representative government nor a market
economy.
529
2. Form of Agreements
Agreements among liberal States could be expected to come in three broad generic
types. First are specific problem-solving agreements, in which a large number of
private individual and group actors already have an interest. Such agreements are
likely to be detailed and direct, intended either for direct judicial enforcement or
immediate incorporation into domestic implementing legislation. Their specific
provisions will probably be shaped in the halls of private firms as much as at the
inter-State negotiating table, as lobbying groups pressure their respective
governments to support particular interests.
Second are relatively vague and general agreements designed primarily to create
a framework for further transgovemmental cooperation. The generality of such
agreements is, paradoxically, an indicator of trust and cooperation. Increased
interaction breeds mutual confidence, allowing further interaction to take place on
the basis of imprecise and open-ended agreements, to be filled in in good faith. In
some cases a formal and relatively specific agreement will help spark the process of
regulatory cooperation, which will then develop in ways that will outstrip the
agreement and make the specified procedures seem slow and cumbersome.
Extradition treaties, for instance, typically concluded primarily between States
confident of the quality of each other's judicial systems, were originally intended to
create the possibility of rendition of fugitives where no such possibility previously
existed. Over time, however, they have come to co-exist with a plethora of informal
procedures. The initial agreement thus provides an umbrella for multiple modes of
informal cooperation in the shadow of the treaty. At the same time, each
renegotiation of the treaty becomes progressively more general in form, moving
from a specification of crimes for which fugitives could be extradited to a general
'criminality' clause, assuming that any crime in one State would be recognized in
the other. 58 Underlying such an evolution is an assumption of steadily increasing
domestic convergence, and growing trust.
Another example, although with a different genesis, is the US-EC Memorandum
of Understanding on Antitrust Enforcement. Here the intergovernmental agreement
was the culmination of a long-simmering inter-governmental conflict, spurred by
private actors. It is nevertheless very general in form, leaving it to the government
agencies involved to fashion their own modes of working together. A similar
example in the security field is the North Atlantic Treaty, the founding document of
NATO. 5 9 The relatively few and sparse phrases of its founding treaty provided the
framework for decades of effective cooperation.
A third category of agreements among liberal States are not actually agreements
at all, but rather model codes designed to spur harmonization through convergence
58
59
530
See generally, M.C. Bassiouni, International Extradition: United States Law and Practice (1987)
329-333 (noting that the 'contemporary trend in extradition treaties is to designate extraditable
offenses' purely on the basis that the penalty applied in both States meets 'an agreed degree of
severity').
North Atlantic Treaty, 4 April 1949, 63 Stat. 2241, TIAS No. 1964,401 UNTS 75.

toward a common focal point. Assuming States organized along similar economic
and political lines, we could expect a variety of regulatory means adopted to achieve
broadly similar aims. The resulting conflicts and contradictions would make all
States worse off, notwithstanding their common goals. Such difficulties should once
again be perceived by national regulators working informally with one another.
Rather than a formal inter-govemmental agreement, however, many such problems
can be solved by decentralized harmonization, a process by which each State
chooses independently to conform its domestic law to a model promulgated by a
neutral third party, such as the OECD. This model is much closer to the 'model
codes' promulgated for adoption by the States of the United States than to formal
international agreements.
3. Negotiating Agreements
Much of the above analysis dissolves the traditional model of States negotiating
agreements as unitary actors seeking to secure exogenous preferences through armslength bargaining, substituting instead the image of communication among a
network of national regulatory officials. Yet even where the more traditional model
holds, negotiations among liberal States should be more likely to conform to Robert
Putnam's 'two-level game' model than negotiations between liberal and non-liberal
States. 60 Because each liberal State is relatively transparent and porous concerning
its domestic political processes, legislative involvement will become increasingly
important at the international level. Each treaty party should thus become
increasingly familiar with its partners' domestic constraints; many should seek
directly to influence each other's domestic legislative process. Members of the
United States Congress have often been included in treaty delegations; in the
Uruguay Round of the GATT negotiators for a number of States were in direct
contact with US senators back in Washington.
Conversely, the host of political constraints on national executives in liberal
States is likely to cast a favourable light on international agreements as a way of
accomplishing goals through an international mechanism that, domestically, would
be politically impossible. National executives, for instance, can use the arena of an
international regime to achieve objectives that they can then sell to their domestic
constituencies as internationally necessary, or as a disagreeable but indispensable
part of an otherwise desirable international agreement. Challenges can always be
deflected in part by the cloak of special executive expertise in foreign affairs and
appeals to the need for a unified national voice in dealings with other nations. Over
time, the force of these appeals should diminish in a world of entirely liberal States,
in part because the assurance of peace blunts the spectre of war lurking in the
background of such traditional executive arguments. Arguments for a unified
60
Putnam, 'Diplomacy and Domestic Polities', 42 International Organization (1988) 427; P.B.
Evans et al. (eds), Double-Edged Diplomacy: International Bargaining and Domestic Politics
(1993).
531
national position will remain, but will have less resonance in the face of divergent
domestic interests and transnational coalitions of domestic interest groups.
4. Enforcing Agreements
Agreements concluded among liberal States are more likely to be concluded in an
atmosphere of mutual trust, a precondition that will facilitate any kind of
enforcement. In particular, however, the assumptions that these are agreements
reached with the participation of a network of individuals and groups in the
participating States, and that these States are committed to the rule of law enforced
by national judiciaries should lead to more 'vertical' enforcement through domestic
courts. This mode of enforcement contrasts with the traditional 'horizontal' mode
involving State responsibility, reciprocity, and countermeasures.
Enforcement through the mechanism of a neutral tribunal backed by coercive
force is an optimal way to ensure compliance with any agreement, whether between
individuals or States. It is the mode of dispute resolution established where possible
by societies the world over. 61 A system of horizontal counter-measures, by contrast,
inevitably entails each party being judge in its own case, with the attendant
illegitimacy of the resulting determination. International lawyers have long
recognized the inferiority of this system, and thus have sought to rely on a general
norm of enforcement of international obligations through domestic courts. 62 Yet
horizontal enforcement remains prevalent. Why? One reason is that absent a
minimum homogeneity of States, States cannot be certain that a treaty partner has
the kind of domestic judicial system capable of or willing to enforce an international
agreement against another branch of government.63 Indeed, the risk averse position
for all States in such a system is to assume that its treaty partners are subject to no
domestic constraints.
The assumption of liberal States ameliorates these concerns in a number of
ways. First, liberal States are States with governments of limited powers, powers
limited by law enforced by courts. Such governments are thus accustomed to the
application of a legal instrument to curtail asserted political power.64 Second, the
governments of liberal States are governments structured around a separation of
powers. The courts of liberal States are thus governed by a constitution that
specifies that they sit as a branch independent of the Executive. Third, liberal States
61
62
63
64
532
Martin Shapiro elegantly lays out the logic of such a system in his multi-country analysis of courts.
M. Shapiro, Courts: A Comparative and Political Analysis (1981).
See Wyatt, 'New Legal Order, or Old?', 7 EL Rev. (1982) 147.
Based on this model, we should expect to find that tax treaties, extradition treaties and other types
of treaties specifically negotiated on a bilateral basis are much more likely to be judicially
enforceable than multilateral treaties. Each Member State could size up the domestic judicial
system of its treaty partner, and design a mechanism of treaty enforcement accordingly.
Commenting on the power of the European Court of Justice's successful attempt to enlist national
courts in the enforcement of the Treaty of Rome, Joseph Weiler writes: 'A state, in our Western
democracies, cannot disobey its own courts'. Weiler, "The Transformation of Europe', 100 Yale
Law Journal (1991) 2403,2421.

guarantee a host of individual rights against the government, to be enforced through
legal action. Thus it is possible to imagine individuals as monitors of government
compliance with agreed rules, whether arrived at through a domestic or an
international legislative process. Fourth is a commitment to transparency as a key
cog in the mechanism of liberal government. Treaty partners can thus be reassured
that it will be relatively easy to monitor each other's political decisions. Fifth, liberal
States are more likely to be monist than dualist, as evidenced by international
'override' provisions in domestic constitutions that mandate the supremacy of
international over domestic law. These provisions are much more common in the
constitutions of liberal States.65
In short, the domestic constraints on liberal governments are more likely to
create the conditions in which States entering into an international agreement have
reason to believe that their co-parties are equally constrained by domestic courts,
such that domestic judicial enforcement would not handicap one party significantly
more than another. If all States are similarly situated regarding this mode of
enforcement, they will be prepared to acknowledge its relative efficiency in holding
all parties to their initial bargain. Further, to the extent that the agreements
concluded contain provisions of direct benefit to individuals and groups in
circumstances in which the contents of such agreements are relatively wellpublicized, private parties will have an incentive to use domestic courts to enforce
these agreements. Indeed, in such circumstances, given representative legislatures,
liberal governments may well need to hold out the prospect of domestic judicial
enforcement to secure domestic ratification of the agreement.
Domestic judicial enforcement of international agreements can come in several
modes. First is the requirement of domestic implementing legislation for an
international agreement; legislation that will then be enforced by domestic courts as
domestic law. The difficulty is monitoring comformity of the implementing
legislation with the international agreement. Second is the self-executing treaty, in
which the international agreement operates directly as law for domestic courts. Here
the problem is likely to be variations in interpretation and application of the treaty
provisions. In both these cases the possibility of having States be judge in their own
cases remains, on the assumption that even an independent court is likely to be
swayed more by the position of its executive versus that of a foreign executive. Thus
a third mode of enforcement is to add an international tribunal to ensure impartiality
and uniformity by deciding all questions of treaty law arising in cases brought to
national courts that involve issues both of national law and treaty law. This type of
system will give rise to a 'vertical dialogue' between national courts and the
supranational court. Opportunities for this dialogue will be facilitated where
members of the supranational tribunal are drawn from the ranks of well-respected
domestic lawyers, such that the legitimacy and respect accorded members of the
65
See Stein, 'International Law in Internal Law: Toward Internationalization of Central-Eastern

European Constitutions?', 88 AJIL (1994) 427, 428; Cassese, 'Modern Constitutions and
International Law', 192 RdC (1985-D) 331,352.
533
domestic bar in each country is transferred to the international tribunal. Finally,
such a dialogue can be facilitated on a non-binding basis, in which domestic courts
are instructed to 'take account' of the judgments of a supranational tribunal.
Vertical enforcement through domestic courts is not the only effective means of
ensuring compliance with international agreements. It is also possible to bypass
domestic courts altogether and rely on legislatures to take account of the judgments
of a supranational tribunal. Yet this mechanism works best when embedded in
domestic and transnational society, by empowering individuals and groups to bring
cases in the first instance so that they can monitor and publicize the results as a
means of pressuring domestic legislatures. Alternatively, a group of States can adopt
procedural rules making the application of horizontal counter-measures dependent
on the decision of a neutral third-party tribunal. Finally, a wide range of compliance
measures and mechanisms exist that rely primarily on diplomatic suasion and
informal dispute resolution. Nevertheless, the proposition here is that where
available, vertical enforcement is the most secure means of assuring compliance
with international agreements. And that means is most likely to be available in a
community of liberal States.
V. The Disaggregation of Sovereignty

If the model of law among liberal States developed here holds as a positive model, it
will present international lawyers with a new set of normative challenges. We will
have to redefine existing principles of international law and develop new principles
to govern actors and processes that remain obscured when viewed through the
lenses developed by Realist international relations theory and traditional
international law to view the international system. The Grundnorm of sovereignty is
a natural point of departure for this normative project. I offer here some preliminary
thoughts on how sovereignty might be redefined as a foundational norm in a world
of liberal States.
According to the analysis above, a world of liberal States would be a world of
disaggregated States. It would be a world with the following attributes:
- 'the State' is composed of multiple centres of political authority - legislative,
administrative, executive, and judicial;
- each of these institutions operates in a dual regulatory and representative
capacity with respect to individuals and groups in domestic society. Each is
defined in terms of a specific set of functions it performs for the members of
domestic society, a set of functions that structures its interaction with its
coordinate branches as co-representatives of 'the people'.^ At the same time,
66
534
Although it may seem odd to speak of courts 'representing' the individuals who appear before
them in private disputes, a judge's conception of her duty to resolve such disputes as they are
presented before her also reflects a wider conception of the need to represent the interests of an
entire class of potential litigants

each of these institutions represents a facet of the exercise of State power making, implementing, and enforcing regulations against individuals on behalf
of the whole;
- the proliferation of transnational economic and social transactions creates links
between each of these institutions and individuals and groups in transnational
society. The development of links between individuals and groups in
transnational society with the political institutions of multiple States in turn
generates contacts among these institutions, either directly or indirectly;
- interactions among counterpart or coordinate institutions from different States court to court, court to legislature, legislature to legislature, executive to court are shaped by both an awareness of a common or complementary function
transcending a particular national identity, and a simultaneous recognition of an
obligation to defend and promote the interests of a particular subset of
individuals and groups in transnational society.
The State is disaggregated, but remains the State: a constellation of political
institutions bound together by territory, text, history and culture. What then of
sovereignty? If the State is disaggregated as a positive matter, can sovereignty
continue to attach to a unitary State as a normative principle designed to constitute
that State as a unitary entity?
A world of liberal States could be conceptualized as a transnational polity. The
organizing principle of this polity would mirror the organizing principle of liberal
States: the limitation of State power by establishing multiple institutions designed
both to overlap and complement one another. The resulting system of 'checks and
balances' - competition and coordination, division and duplication - creates
sufficient friction to curb the abuse of power. The result, to borrow a term coined by
political theorist Daniel Deudney, is a 'negarchy', a liberal political order between
anarchy and hierarchy in which power is checked horizontally rather than
vertically.67 These divisions and deliberately created frictions are further designed
to create space for individuals and groups to interact with and influence State
institutions, rather than being passive subjects of their rule.
Transposed to a transnational plane, the principle of negarchy would require the
vigorous interaction of the governmental institutions of participating States with one
another in as many combinations as possible. Assuming that 'sovereignty' is the
principle that both constitutes States and defines their rights and duties in the
international system consistent with an overarching principle of international order,
the task is then to redefine sovereignty to conform to the ordering principle of
negarchy. In practice, the norm of sovereignty would have to be constructed so as to
constitute and protect the political institutions of liberal States in carrying out their
individual functions and in checking and balancing one another.
The first element of such a redefined norm of sovereignty might thus be one of
non-interference with basic legislative, judicial, and executive functions on the part
67
Deudney, "The Philadelphian System: Sovereignty, Anns Control, and the Balance of Power in the
American States-Union, circa 1781-1861', 49 International Organization (forthcoming 1995).
535
of the component institutions of each State in the system. Each institution could
assert sovereignty both as a constitutive norm and as a shield to protect its ability to
represent and regulate a subset of individuals and groups in transnational society the individuals and groups principally associated with its State. If the flip side of the
constitutive norm of sovereign Statehood is currently non-intervention in matters
within a State's domestic jurisdiction, parallel norms of legislative, judicial, and
executive sovereignty in a world of liberal States would dictate non-interference
with the basic provision of legislative, judicial, and executive functions for the
nationals or territorial residents of a particular State. Viewed from a transnational
perspective, this norm of non-intervention as attached to these component
institutions would protect the integrity of each institution sufficient to allow it to
check and balance its counterpart and complementary institutions.
The second element of a redefined norm of sovereignty would concern the
interaction of these disaggregated State institutions with one another as centres of
authority in a transnational polity. Such interaction should be structured to achieve
two functions: the representation and regulation of particular subsets of individuals
and groups in transnational society. Each institution would thus encapsulate the dual
dynamic of State-society relations in liberal States, simultaneously standing for the
autonomous power of the State and the delegated authority of the people. As such,
each institution must have the capacity not only to engage with other institutions,
but also to receive communication from and to respond to the concerns of individual
and group actors. Each must have the capacity to represent as well as to regulate.
A more concrete expression of this second element could be a right of equal
participation in transnational legislative, judicial, executive and administrative
processes. Such a right would ground claims on the part of component State
institutions to be recognized and taken account of by fellow institutions involved in
the negotiation of an international agreement, the regulation of transnational
conduct, or the resolution of a transnational dispute. These claims would in turn rest
not on abstract conceptions of national power, but on the capacity to represent and
regulate a particular subset of individuals and groups in transnational society.
Competing claims would be resolved by assessing the relative capacity of these
institutions to perform the asserted function and the legitimacy of the claimed
representation.68
Abram and Antonia Chayes have recently written of 'the new sovereignty',
arguing that full membership in the international community no longer means the
prerogative of splendid isolation, but rather requires engagement in a network of
multilateral commitments. 69 Much current writing on the 'transformation of
sovereignty' adopts a passive voice, emphasizing the extent to which State borders
68
69
536
This new conception of sovereignty would exist side by side with more traditional conceptions,
which are still accurate and important in relations between liberal and non-liberal States. Where a
government controls its people absolutely, its 'sovereignty' is still largely unitary and subject to
violation primarily by coercive intervention. Sovereignty here is constrained less by individuals
and groups in transnational society than by other States or by international institutions.
Abram and Antonia Chayes, The New Sovereignty (forthcoming 1995).

have become permeable to forces beyond a particular government's control. The
Chayes and Chayes conception, by contrast, redefines the active component of
sovereignty in relation to increased mechanisms and systems of global governance.
Sovereignty becomes the capacity to participate in an international regulatory
process. The redefinition of sovereignty in a world of liberal States pushes this
redefinition one step further, devolving it onto the component institutions of
individual States and giving it substantive content with regard to the relationship
between these institutions and individuals and groups in transnational society.
VI. Conclusion
The purpose of this essay has been to reimagine international law from the
perspective of Liberal international relations theory in a hypothetical world of
liberal States. It has identified the relevant bodies of law governing different actors
in this system - rules governing the voluntary interactions of individuals and groups
in transnational society, rules selected and applied by national courts and
legislatures, and inter-State agreements. It has mapped the interactions of the
relevant actors and sketched new ways to interpret the results of those interactions.
The point of departure for this exercise was the proposition that international
lawyers must be more explicit about their underlying political science. The 17th and
18th century fathers of classical international law internalized deep assumptions
about the incidence of war and peace and the nature of States. These scholars lived
in a world in which war was endemic and domestic governance structures diverse; a
world in which furthering the domestic consolidation of power under an allpowerful sovereign and simultaneously delimiting that power in the international
sphere offered the most promising hope of reducing violent conflict in both
spheres. 70 The founding principle of the Westphalian system - cujus regio, ejus
religio (look only to the prince and no farther) was a formula for peace. A
prohibition on taking account of domestic differences among States thus converged
with an argument about the foundations of international security.
This convergence forged an analytical synthesis between classical international
law and Realist international relations theory. Yet within political science, Liberal
theorists consistently and successfully have challenged Realist assumptions about
the nature of the international system. Their alternative framework assumes that how
States behave depends on how they are internally constituted. Empirical research
within this framework further suggests a fundamental difference in the nature of
relations among liberal States as compared to relations between liberal and nonliberal States. My aim has been to consider this research and its underlying
assumptions seriously and to explore its implications for international law.
70
See E.F. Hinsley, Sovereignty (1986) Chap. 5.
537
Imagining a world of liberal States is not a purely hypothetical exercise. The
principles and postulates of classical international law have long been subject to
numerous exceptions and modifications that reflect departures from the underlying
positive assumptions of unitary and functionally identical States. Contemporary
human rights law, for instance, was founded on the recognition that domestic
political conditions have consequences for international security. To the extent that
the existing catalogue of fundamental human rights expands to include a right of
'democratic governance', a right Thomas Franck proposes based in part on
empirical evidence of peace among liberal States, international law will take the
first step toward an explicit distinction among States based on domestic regimetype. 71
Further, the record of bloodshed in the 20th century challenges the 18th century
paradigm of the sources of international and domestic conflict. In many cases, strife
appears to result more from the abuse than the absence of sovereign power.
Representative political institutions, the protection of minority rights, and the
furtherance of group autonomy short of Statehood appear more likely to further
long-term domestic and international peace than the raising of new Leviathans. At
the same time, the realm of peace and relative prosperity is no longer a
condominium of all-powerful princes but rather a domain of representative
governments embedded in a dense network of transnational economic and social
transactions. The perception of such seismic shifts, to the extent they hold, could
lead to the adoption of a new model of the international system, normatively
applicable to all States even if positively descriptive of only some. Alternatively, the
values of universalism could be sacrificed to the realism of recognizing that States
in the international system inhabit very different worlds.
The world of liberal States, as hypothesized here, is a world of individual selfregulation facilitated by States; of transnational regulation enacted and implemented
by disaggregated political institutions courts, legislatures, executives and
administrative agencies enmeshed in transnational society and interacting in
multiple configurations across borders; of double-edged diplomacy and intergovernmental agreements vertically enforced through domestic courts. Such a world
is neither a Utopia nor a panacea. It poses new problems and normative challenges,
requiring the extension of domestic principles and safeguards to transnational and
international governance processes. It requires a rethinking of the relationship
between public and private international law, a reconceptualization of the rules that
can be said to serve international order.
71
538
Franck, "The Emerging Right to Democratic Governance', 86 AJIL (1992) 46.

Anne Marie Slaughter - International Law in A World of Liberal States

Uploaded by

Copyright:

Available Formats

Anne Marie Slaughter - International Law in A World of Liberal States

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Anne Marie Slaughter - International Law in A World of Liberal States

Uploaded by

Copyright:

Available Formats

International Law in a World of Liberal States

6 EJIL (1995) 503-538

International Law in a World of Liberal States

I. Liberalism and Realism in International Relations Theory

International Law in a World of Liberal States

Internationa] Law in a World of Liberal States

EL. Mapping the Attributes of the Liberal Peace

International Law in a World of Liberal States

D. A Dense Network of Transnational Transactions

International Law in a World of Liberal States

R.O. Keohane, J.S. Nye, Power and Interdependence (1977) 25.

F. Collapse of the Foreign/Domestic Distinction

III. Hypothesizing a World of Liberal States

International Law in a World of Liberal States

IV. International Law in a World of Liberal States

International Law in a World of Liberal States

A. The Voluntary Law of Individuals and Groups in Transnational Society

International Law in a World of Liberal States

International Law in a World of Liberal States

B. The Law of Transnational Governmental Institutions

International Law in a World of Liberal States

International Law in a World of Liberal States

International Law in a World of Liberal States

C. The Law of Inter-State Relations

International Law in a World of Liberal States

International Law in a World of Liberal States

International Law in a World of Liberal States

See Stein, 'International Law in Internal Law: Toward Internationalization of Central-Eastern

V. The Disaggregation of Sovereignty

International Law in a World of Liberal States

International Law in a World of Liberal States

See E.F. Hinsley, Sovereignty (1986) Chap. 5.

Franck, "The Emerging Right to Democratic Governance', 86 AJIL (1992) 46.

You might also like