100% found this document useful (3 votes)

1K views266 pages

Public Policy A New Introduction

Uploaded by

Sameat Sovansak

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

100% found this document useful (3 votes)

1K views266 pages

Public Policy A New Introduction

Uploaded by

Sameat Sovansak

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 266

T E X T BO OK S I N

P OLI C Y S T U DIE S
S E R I E S E D I T O R : PAU L C A I R N E Y

A NEW INTRODUCTION
Copyright © 2020. Macmillan Education UK. All rights reserved.

‘The first edition of Public Policy rapidly became a modern classic in the field, a must-read
for all those working on public policy – whether students or established scholars. The
epithets elegant, concise, profound and clear easily attach themselves to this well-written
and useful book that stands head and shoulders above all its competitors. It is very wel-
come indeed to have the second edition that retains all the strengths of the first, but is
fully updated to reflect recent thinking and scholarship in public policy as well as taking
account of the profound changes in politics and policy that have occurred in recent years.
Christoph Knill and Jale Tosun have pulled off a tremendous feat.’
– Peter John, King’s College, London, UK

‘This book is an outstanding introduction to studies of public policy. It combines both

traditional approaches with innovative perspectives on the design and implementation of
public policies. It is the best textbook for policy studies in Europe.’
– Karin Ingold, University of Bern, Switzerland

‘There is no better policy textbook with a global perspective than Knill and Tosun’s. It com-
bines the leading paradigms with innovative insights in a well-written, comprehensive and
state-of-the-art account of the policy process. This volume belongs in the personal library
of any student, scholar and practitioner who wants to master the essentials of public
policy-making around the world.’
– Raymond Tatalovich, Loyola University Chicago, USA

‘The radically updated second edition of Public Policy engages with empirical developments
Copyright © 2020. Macmillan Education UK. All rights reserved.

– such as climate change, digitization and the questioning of the global international
order – that have a path-changing significance for public policy. The authors successfully
integrate theoretical, analytical and methodological advances in the book, which reflects
that Jale Tosun and Christopher Knill are leading scholars in the field. I strongly recommend
this book to scholars and educators of public policy and related fields.’
– Caroline de la Porte, Copenhagen Business School, Denmark

‘This is an excellent textbook that covers the core of policy research from a political science
and public administration perspective in a concise, clear and analytically structured way. It
is one of the few textbooks in the field that brings a distinctively European perspective to
the field. This new edition will strengthen the volume’s status as one the top textbooks in
the field – I will continue using it in my classes as core reading.’
– Kai Wegrich, Hertie School of Governance, Germany

‘Knill and Tosun’s Public Policy has proved to be an excellent text for introducing my stu-
dents to fundamental ideas in public policy and policy making. It structures theoretical
content in a clear and thoughtful way, and uses relevant real world examples to demon-
strate the complexities of policy making across national and international contexts.’
– Alex Mankoo, University of Sussex, UK

‘Knill and Tosun have written a highly accessible and comprehensive textbook that not
only discusses traditional policy theories but also highlights the importance of public-
private governance and policy-making beyond the state. This text forms an excellent
introduction for students and the new edition’s increased attention to climate change in
particular makes this book even more indispensable than before.’
– Stefan Renckens, University of Toronto, Canada

‘Public Policy is a highly accessible book introducing public policy through examples drawn
from all over the world. The authors survey and present a wide range of theoretical per-
spectives in a clear and concise fashion. It is a tremendous introduction to public policy
and points the reader where to go next for detailed work on each topic.’
– Martin Hansen, Brunel University London, UK

‘Public Policy continues to provide an excellent foundation in key theories and debates
regarding policy processes. This revised edition not only reflects development in the fields
of academic debate but also political changes to provide an excellent introduction for new
students. The revised conclusion provides useful practical guidance for students and prac-
Copyright © 2020. Macmillan Education UK. All rights reserved.

titioners alike.’
– Lee Gregory, University of Birmingham, UK

‘Christoph Knill and Jale Tosun offer students and scholars in their second edition a
sophisticated synthesis of public policy studies. They present theories with clarity and
integrate them with practical relevance through contemporary vignettes. This book is ideal
for students wanting a comprehensive and engaging introduction to public policy and why
it matters’.
– Christopher Weible, University of Colorado Denver, USA

2nd edition

Christoph Knill and Jale Tosun

Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:21:32.
© Christoph Knill and Jale Tosun, under exclusive licence to Macmillan Education Limited 2012, 2020
All rights reserved. No reproduction, copy or transmission of this publication
may be made without written permission.
No portion of this publication may be reproduced, copied or transmitted
save with written permission or in accordance with the provisions of the
Copyright, Designs and Patents Act 1988, or under the terms of any licence
permitting limited copying issued by the Copyright Licensing Agency,
Saffron House, 6–10 Kirby Street, London EC1N 8TS.
Any person who does any unauthorized act in relation to this publication
may be liable to criminal prosecution and civil claims for damages.
The authors have asserted their rights to be identified as the authors of this
work in accordance with the Copyright, Designs and Patents Act 1988.
This edition published 2020 by
RED GLOBE PRESS
Previous editions published under the imprint PALGRAVE
Red Globe Press in the UK is an imprint of Macmillan Education Limited,
registered in England, company number 01755588, of 4 Crinan Street,
London, N1 9XW.
Red Globe Press® is a registered trademark in the United States,
the United Kingdom, Europe and other countries.
ISBN 978-1-137-57330-8 hardback
ISBN 978-1-137-57329-2 paperback
This book is printed on paper suitable for recycling and made from fully
managed and sustained forest sources. Logging, pulping and manufacturing
processes are expected to conform to the environmental regulations of the
country of origin.
A catalogue record for this book is available from the British Library.
A catalog record for this book is available from the Library of Congress.
Copyright © 2020. Macmillan Education UK. All rights reserved.

List of Figures x
List of Tables xi
List of Boxes xii
Foreword to the Second Edition xiii

1 Introduction 1
What Is a Public Policy? 3
Polity, Politics and Policy 3
Elements of a Public Policy 4
Differences in Scope: Sectors, Targets and Instruments 5
Analytical Perspectives on the Policy-Making Process 6
Basic Theoretical Perspectives on Public Policies: Rational Process Design,
Muddling Through or Just Chance? 6
Stages of the Policy Process 7
Opportunities and Challenges for Policy-Making 9
Studying Public Policy: The Approach and Structure of This Book 10
Basic Analytical Tools, Concepts and Theories 10
Stages of the Policy Process 11
Cross-Cutting Themes 12

2 The Nature of Public Policies 13

Introduction 13
Typologies of Public Policies 14
Classification by Implications for Politics 14
Classification by Governance Principles and Instruments 19
Copyright © 2020. Macmillan Education UK. All rights reserved.

Alternative Schemes 23
Policy Dimensions 25
Policy Dimensions in the Sequence of the Policy Cycle 25
Policy Content Dimensions 26
Policy Styles 28
Conclusions 31
Further Reading 32

3 The Context for Policy-Making: Central Institutions and Actors 33

Introduction 33
National Institutions: Defining the Rules of the Political Game 34
Constitutions and Constitutional Courts 34
Division of Powers 36
Electoral Institutions and Party Systems 39
Intergovernmental Organizations: Global Public Policy 42
The United Nations System 42
The World Trade Organization 43

v
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:21:32.
vi Contents

Key Actors 45

Public Actors 46
Political Parties 49
Private Actors 51
Conclusions 54
Further Reading 55

4 Theoretical Approaches to Policy-Making 56

Introduction 56
Structure-Based Models 57
The Cleavage Approach 57
Varieties of Capitalism 60
Institution-Based Models 61
The Classic Approach to Political Institutions 62
The New Approaches to Political Institutions 63
Interest-Based Models 67
Concepts of Rationality 67
Strategic Choices of Actors 68
Actors and Institutions: Rational Choice Institutionalism 73
Overview of the Models 74
Conclusions 75
Further Reading 76

5 Problem Definition and Agenda-Setting 77

Introduction 77
Problem Definition: The Path to Agenda-Setting 78
The Process of Problem Definition 78
The Characteristics of Problems 81
Analytical Concepts of Agenda-Setting 84
Agenda Types 85
The Process Perspective 85
Copyright © 2020. Macmillan Education UK. All rights reserved.

The Power Distribution Perspective 87

The Institution-Based Perspective 88
The Contingency Perspective 89
Summary 90
Actors and Interests in the Agenda-Setting Process 91
Elected Public Officials 91
Judges 92
Bureaucrats 93
Interest Groups 94
Citizens 96
International Organizations 96
Media 97
Conclusions 98
Further Reading 99

6 Decision-Making 100
Introduction 100
Determinants of Policy Formulation 102
Expertise, Information and Ideas 102
International Organizations 103
Interest Groups 105
Political Preferences Stemming from Partisan Ideology 107
Pursuit of Private Interests 108
Institutional and Procedural Dimensions of Decision-Making 109
How Are Policy Preferences Formed? 110
How are Policy Preferences Transformed into Actual Policies? 111
Decision-Making in The United Kingdom, Australia, Canada and The United States 114
The United Kingdom 114
Australia 116
Canada 117
The United States 118
Conclusions 120
Further Reading 121

7 Implementation 122
Introduction 122
Who Is Involved in Policy Implementation? 123
Implementation Structure 123
Decision-Making within Agencies 125
Target Group Behaviour 126
Analytical Perspectives in Implementation Research 126
Top-Down Models of Policy Implementation 127
Bottom-Up Models of Policy Implementation 129
Hybrid Models of Policy Implementation 130
Summary 131
Implementation Success: Criteria and Determinants 132
Copyright © 2020. Macmillan Education UK. All rights reserved.

Criteria for Implementation Success 132

Determinants of Implementation Success 134
Conclusions 142
Further Reading 143

8 Evaluation 144
Introduction 144
Types and Methods of Evaluation 145
Types of Evaluation 145
The Main Elements of an Evaluation 151
Research Designs for Scientific Policy Evaluation 152
The Conventional Attribution Problem: Coping with Complex Context Conditions 153
The More Challenging Attribution Problem: Evaluating Complex Policy Mixes 154
Main Insights 156

The Political Dimension of Policy Evaluation 157

Policy Evaluation versus Political Evaluation 157
Evidence-Based Policy-Making 159
Conclusions 162
Further Reading 163

9 Governance: A Synoptic Perspective on Policy-Making 164

Introduction 164
Central Concepts and Modes of Governance 165
Governance Concepts 165
Modes of Governance 166
Changes in Governance Modes: From Hierarchies to Markets to Networks? 169
Four Ideal Types of Governance 172
Interventionist Governance 173
Regulated Self-Governance 173
Cooperative Governance 173
Private Self-Governance 174
Which Model for Which Season? The Relevance of Institutional
and Political Capacities 174
When Is Governance Good? 177
Decision-Making Capacity 177
Implementation Effectiveness 177
Democratic Legitimacy 178
Conclusions 180
Further Reading 181

10 Public Policies beyond the Nation State 183

Introduction 183
Public Policies beyond the Nation State: General Assessment 184
Public Policy beyond the Nation State: Underlying Rationales 186
Overview: Organizational Forms and Areas of International Public Policy 188
Copyright © 2020. Macmillan Education UK. All rights reserved.

Policy Formulation: Typical Interest Constellations and Interaction 191

The Problem Type 192
The Innovative Potential of Leader Countries 196
Deliberation and Diffusion 197
Log Jam or Progress? Conflict in International Policy-Making 199
Policy Implementation 199
Deficient Policy Design 200
Control Deficits 201
Preferences and Strategies of the Involved Actors 202
Conclusions 204
Further Reading 205

11 Policy Change and Policy Convergence 207

Introduction 207
Policy Change: Theories, Measurement and General Patterns 208
Theories of Policy Change 208
Policy Change as Change in Preferences and Beliefs 209
Linking Policy Change and Institutional Change 211
Policy Change by Chance 213
Measurement of Policy Change 213
Cross-National Policy Convergence: Concept, Measurement and Dimensions 220
Related, but Different: The Concepts of Policy Convergence, Policy Transfer,
Policy Diffusion and Institutional Isomorphism 220
Types and Measurement of Policy Convergence 222
Convergence Dimensions 224
Causes and Conditions of Cross-National Policy Convergence 225
Causal Mechanisms 225
Facilitating Factors 228
Conclusions 229
Further Reading 230

12 Conclusions: Theoretical Insights and Practical Advice 231

What Have We Learned? 231
Comparative Research on Public Policy 235
Developing Own Research Questions 237
Doing Policy Analysis 238

241
References
Index 282
Copyright © 2020. Macmillan Education UK. All rights reserved.

3.1 Stylized institutional setup of a polity 35

5.1 Relationship between media articles and press releases on nitrate
pollution in Germany 95
7.1 Institutional adaptation pressure and implementation effectiveness 139
11.1 Types of censoring 214
Copyright © 2020. Macmillan Education UK. All rights reserved.

x
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:21:32.
List of Tables

1.1 Basic research topics of public policy analysis 2

1.2 Policy stages and related research topics 8
2.1 Lowi’s policy typology 15
2.2 Wilson’s typology 19
2.3 Policy classification by governance principles 24
2.4 Potential determinants of policy styles 29
4.1 Basic theoretical approaches to the study of public policy 75
5.1 Four models for the study of agenda-setting 91
7.1 Overview of implementation models 132
7.2 Criteria for measuring implementation success 134
7.3 Instrument choices in the light of capacity and subsystem characteristics 135
9.1 Overview of governance concepts 166
9.2 Four ideal types of governance 172
11.1 Dimensions and indicators of policy change 217
11.2 Policy convergence and related concepts 222
11.3 Four forms of policy convergence 223
11.4 Dimensions of policy convergence 225
11.5 Causes of cross-national policy convergence 228
Copyright © 2020. Macmillan Education UK. All rights reserved.

xi
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:21:32.
List of Boxes

2.1 Politics of same-sex marriage 17

3.1 Composition of the legislature in South Africa 37
3.2 World Wildlife Fund: an international interest group 53
4.1 Structure of the prisoner’s dilemma 72
5.1 Perceptions of reasons for vaccine hesitancy 81
5.2 Sexual harassment as a policy problem 83
5.3 Issue attention cycle 86
6.1 Drafting the Kenyan biosafety law 104
6.2 (Super-)wicked problems: when incrementalism is intentional 113
7.1 Improving policy implementation in the United Kingdom 124
7.2 Implementation of Mexican environmental policy 140
8.1 Basic income experiment in Finland 148
9.1 Experimental governance in China and Europe 167
9.2 Governance in states with limited statehood 176
10.1 Polycentric climate governance 185
10.2 Administrative styles of international organizations 193
10.3 The Universal Postal Union 194
10.4 Indian ban on Chinese toys 195
10.5 Implementation of international agreements in federal systems 204
11.1 Changes in the Finnish policy on biofuels 210
11.2 Policy accumulation and the democratic responsiveness trap 219
11.3 Environmental policy convergence in Europe 227
Copyright © 2020. Macmillan Education UK. All rights reserved.

xii
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:21:24.
Foreword to the Second Edition

When we first considered publishing a second edition of this book, we estimated that we
would simply update the previous topics by adding new literature and providing additional
clarifications where necessary. However, once we got started, we realized that a much greater
effort would be needed, which also resulted in a significantly lengthier revision process than
we or the publisher had wished for. While on the one hand we wanted to get the revision
done as quickly as possible, on the other we realized that we were observers to some changes
in real-life politics and the study of public policy that warranted our attention.
Let us start with the positive developments in politics that we could witness and on
which we elaborate in this new edition. When we wrote the first edition of this book in 2010
and 2011, Europe and North America were still under the impression of the economic and
financial crisis that had unfolded in 2007 and 2008. A decade later, almost all countries hit
severely by the crisis have recovered. One of the countries that was affected particularly
severely by the crisis was Greece. After years of economic hardship and painful reforms, the
Greek economy is resurgent and policy-makers can shift their attention to other important
policy issues, of which there are many. The second positive development relates to the grow-
ing attention paid to fighting climate change. A decade ago, climate policy scholars felt as if
their findings would never feature prominently in real-life politics. Today, we can safely state
that climate change has reached the political agendas of countries in all world regions. Chile,
for instance, has placed climate action at the top of its political agenda and aspires to become
a climate leader in Latin America. A third development relates to growing demand from
policy-makers for policy analysis performed by public policy scholars. In European coun-
tries, for example, research funding has risen for projects that strive to inform policy-
making through providing a detailed analysis of policy problems and potential policy
solutions. In response to these developments, we condensed the discussion of the policy
implications of the economic and financial crisis, while conversely discussing climate policy
Copyright © 2020. Macmillan Education UK. All rights reserved.

and politics more extensively. In addition, we modified the conclusion to provide guidance
on how to do policy analysis.
However, the last decade was also marked by numerous less positive developments, to
which we allude in this book. From a European perspective, to many academic as well as
political observers, the outcome of the 2016 British referendum on whether to stay in or
leave the European Union (EU) came as a shock. Few expected that a majority of British vot-
ers would vote to leave the EU. Since the referendum, we have been able to observe how
difficult it has been for British policy-makers to deliver Brexit. We could also see how the
Brexit vote changed the political climate in both the United Kingdom and the EU. After
many years of considering the possibility of enlargement only, now for the first time a mem-
ber state might leave the EU, which will have far-reaching and yet not easy to anticipate
consequences for all parties concerned. A second development, which is also related to
Brexit, is the rise of populism, which can also be found in almost all world regions. Inherent
to populism is a fundamental mistrust in the capability of political elites to govern, which
calls into question the essence of representative democracy. Populism has changed how we
talk to each other and about what, and how public policies are being made. Clearly, without

xiii
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:21:17.
xiv Foreword to the Second Edition

digitization, populism would not have been as successful at it is today. From this, it follows
that policy studies need to pay more attention to modern forms of communication. Third,
and related in part to the rise of populism, is a tendency to question (economic) globaliza-
tion and multilateralism. We could witness a ‘my country comes first’ logic which was per-
haps absent or simply not visible when we wrote the first edition of this book. This change
in the perception of globalization has important implications for the functioning of interna-
tional organizations such as the World Trade Organization, which was not designed to gov-
ern ‘trade wars’. We have updated and expanded the discussion of the relevant parts of the
book in order to reflect these developments.
Turning to the changes in the study of public policy, we have observed the following three
advances. First, policy process theories have become increasingly popular and have experi-
enced important conceptual refinements. Further, the empirical approaches to the study of
public policy from the perspective of theories of the policy process have become very sophis-
ticated, leading to a rich body of research that includes many comparative studies. More
generally, we can see more and more comparative work on public policy being published.
Second, there have been attempts to integrate public policy with political economy, political
communication, political psychology and additional subdisciplines of political science.
Policy studies have become more diverse from both a theoretical and an empirical view-
point, but also more connected to other disciplines such as economics, legal studies, psy-
chology or sociology. Third, the methodologies used in studying public policy have advanced
and the state-of the-art methodological toolkit facilitates innovative research perspectives
such as the evaluation of policy measures by means of social media data. We have striven to
expand our analytical purview to capture these developments in the academic literature.
We received extremely helpful feedback from many colleagues, as well as the editor of the
Textbooks in Policy Studies series, Paul Cairney. We thank Lloyd Langman and Peter
Atkinson of Red Globe Press for their guidance and support. Dominic Afscharian, Lucas
Leopold, Laura Lindner, Julian Rossello, Simon Schaub, Paul Thalmann and Laura Zöckler
deserve credit for valuable research assistance.
If you have read the first edition, we thank you, and hope you will enjoy the second edi-
tion even more. If you are new to this book, we hope that you find the topics covered rele-
vant and stimulating enough to consider continuing to study public policy.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Christoph Knill and Jale Tosun

Munich and Heidelberg, November 2019

Public policies are omnipresent in our daily lives. Newspapers, the television, the internet
and social media provide a constant supply of information on public policies, from the regu-
lation of genetic engineering or blood alcohol limits for the operation of a vehicle to the
financing of infrastructure projects like new roads or the level of income tax. Many people
may not be aware that a great deal of information on politics is actually about ‘who gets
what, when, how’, as famously described by Harold D. Lasswell (1936), one of the founders
of policy research. The opening sentence of the book by Lasswell that carries this title posits
that politics is the study of ‘influence’ and ‘the influential’.
This book offers an introduction to public policy (‘who gets what, when, how’), defined in
Lasswell’s sense by concentrating on how politics (‘influence’ and ‘the influential’) deter-
mines public policy. In other words, we provide an exploration of how public policy comes
about and how it is changed by drawing from concepts and theories in comparative politics
(see Caramani 2020a).
For example, why do some governments provide specific welfare schemes (‘what’) to sin-
gle mothers (‘who’) whereas others abstain from doing so? Why is the retirement age for
females in Turkey 58 years but 64 in Switzerland (‘when’)? And why are welfare benefits
means-tested in the United Kingdom but universal in Sweden (‘how’)? These examples show
that public policy affects almost every aspect of our daily life. They also show that there exist
differences between countries that can only be explained by policy-making processes taking
place there. In other words, public policy cannot be understood without reference to policy-
making, which refers to a series of decisions made by public actors such as governments,
parliaments and public administrations, which are, however, affected by other public actors
as well as private actors such as interest groups.
Learning more about public policy and the making of it must be seen as an important
Copyright © 2020. Macmillan Education UK. All rights reserved.

complement to all other themes addressed by the discipline of political science. Concerning
much more than the analysis of political institutions, government behaviour, political par-
ties, interest groups or elections, the study of public policies covers the whole process of
public decision-making (John 2006). It also includes the feedback policy-makers receive on
their decisions (Mettler and SoRelle 2017), for example in the form of changes in welfare
politics and levels of trust (see, e.g., Kumlin et al. 2018).
The study of public policies seeks to understand their production and effects. Students of
public policy are interested in the processes and decisions that define the outputs of a politi-
cal system, such as higher education policies, social services, environmental regulations, the
administration of the health system or the organization of a police force. But policy research
also highlights the broader effects resulting from such policy decisions. To what extent do
environmental regulations result in environmental quality improvements? What are the
effects of budget cuts in the university system on educational performance? To what extent
do changes in the organization of the public health system affect health levels?

1
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:21:10.
2 Public Policy

Table 1.1 Basic research topics of public policy analysis.

Analytical
dimension Sectors Countries Time
Basic research Explanation of policy Explanation of policy Explanation of policy
question variation/similarity variation/similarity change/stability over
across sectors across countries time
Potential • Explanation of policy variation/similarity across sectors and countries;
combinations assessment of relative effect of sectoral versus national factors
• Explanation of change in cross-national and/or cross-sectoral policy variation
over time; assessment of policy convergence versus policy divergence or
persistence of existing differences

If policy outputs and policy effects are the core topics of public policy, their study generally
focuses on two fundamental issues: policy variation and policy change (see, e.g., John 2003;
Knill et al. 2012; Tosun 2013a; Weible and Sabatier 2017). Policy variation refers to the
explanation of differences between public policies across sectors and countries. To what
extent does policy-making differ in relation to, say, health policies, on the one hand, and
environmental or transport policies, on the other? And how can these differences be
explained? Which factors account for similarities and differences in public policies between
countries? As we will see in Chapter 2 when discussing the concept of policy styles (see
Richardson 1982, 2018a, 2018b), there is an ongoing debate in the literature as to whether
and to what extent the shape of public policies is affected by sector- or country-specific fac-
tors (see Table 1.1).
With regard to policy change, the central focus is on the explanation of stability and
change. Often public policies remain highly stable over time, although their functionality
and effectiveness have constantly been questioned, as has been the case, for instance, for
the agricultural policy of the European Union (EU). Despite some changes such as the ‘green-
ing’ of agricultural policy, the 2013 reform of the EU’s Common Agricultural Policy has re-
established income support for farmers as a core element (Daugbjerg and Swinbank 2016;
Copyright © 2020. Macmillan Education UK. All rights reserved.

Feindt 2017). But studying policy stability too can be intellectually stimulating, since
observing no changes to existing public policy does not mean that it was not affected by a
policy-making process (see Dodge and Lee 2017).
How can we explain the sometimes surprising stability of public policies? At the same
time, policies can be subject to swift and fundamental transformations. In the field of cli-
mate policy, for instance, the last few years have brought some significant changes in both
developed and developing countries, in the sense that the number of climate policies has
grown significantly and the policies adopted cover both the mitigation of climate change
and measures to adapt to it (Dubash et al. 2013; Lachapelle and Paterson 2013; Fankhauser
et al. 2015, 2016; Lesnikowski et al. 2016; Fleig et al. 2017). Why do stable periods of
decision-making sometimes give way to flux and unpredictability?
In responding to these questions, public policy analysis often adopts a comparative
research perspective and examines policy changes not only over time, but also across coun-
tries and different policy sectors such as environmental or social policy (see Adam et al.
2017; Tosun and Workman 2017). In this context, one particular area of interest is whether

cross-country variations in public policies remain stable or conversely become more or less
pronounced over the course of time. For example, the internationalization of certain issues
such as environmental protection has resulted in an increasing cross-national similarity of
policy arrangements (Holzinger et al. 2008a, 2008b).
It is the objective of this book to introduce the major analytical concepts, theories and
state-of-the-art tools of public policy analysis. In so doing, the book is designed to reach well
beyond a mere description of public policies and the political processes producing them. It
aims to invoke curiosity for analysing ‘empirical puzzles’ on the basis of well-informed theo-
retical considerations, sound research designs and informative data.
This chapter provides an entrée to the study of policy-making in five steps. First, we need
to get a basic understanding of the nature of the subject under study. What is a public policy,
and how can it be defined? Second, we outline the central analytical questions that have to
be addressed in order to learn more about the processes, structures and problems character-
izing the making of public policies. Third, we explain the challenges policy-makers face in
modern societies. Fourth, we outline the three central research questions guiding this text-
book. Finally, this chapter concludes with an overview of the book’s structure.

WHAT IS A PUBLIC POLICY?

In this section we elaborate on the characteristics of public policies to prepare the ground for
the concepts, theories, methods and data to be presented in this book.

Polity, Politics and Policy

In political science, we generally find that there are three major subject areas that cut across
the different subdisciplines: polity, politics and policy. While polity refers to the institu-
tional structures characterizing a political system, the study of politics concentrates on
political processes, such as party political cleavages and voting behaviour in legislative bod-
ies. The analysis of public policy, by contrast, puts the content of policies centre stage. Rather
Copyright © 2020. Macmillan Education UK. All rights reserved.

than focusing on institutions or processes, the research interest is on the analysis of the
outputs of a political system, i.e. the decisions, measures, programmes, strategies and
courses of action adopted by the government or the legislative body (e.g. parliament).
The focus on public policies does not imply that polity and politics are not taken into
account. On the contrary, these dimensions play an important role in explaining policy
change as well as policy differences across countries. So an important question of public
policy analysis concerns the policy implications of a country’s political institutions (its poli-
ties). Do polities have a crucial impact on policy-making? Which polities perform better
regarding the solving of a howsoever defined societal problem?
Likewise, the decision-making processes (politics) in a country are important for public
policy choices. In this regard, a prominent argument is that countries reveal distinctive pol-
icy styles, i.e. more or less stable patterns that characterize the policy process (politics),
which in turn affect the nature and design of their policies (Richardson 1982, 2018a, 2018b;
Knill et al. 2015; Howlett and Tosun 2018). By reversing this assumed causal relationship,
Theodore Lowi (1964) developed his classic distinction of different policy types, arguing
that different policy areas tend to be characterized by different politics which involve at

times more and at times less conflicting interactions between political actors. Legislative
studies have also shown that the characteristics of members of parliament (e.g. gender) can
matter for how they vote on policy proposals (see, e.g., Debus and Hansen 2014).
So, even if our analytical focus is on the study of public policy, the polity and politics
dimensions play an important role either as factors explaining public policies (see Richardson
1982, 2018a, 2018b; Knill et al. 2015; Howlett and Tosun 2018) or as phenomena that are
determined by policy types (see Lowi 1964). Policy studies must not exclude the dimensions
of polity and politics when describing and explaining policy decisions.

Elements of a Public Policy

Having introduced the linkages between polity, politics and policies, we still have to clarify
what the term ‘public policy’ actually means. In the scholarly literature, there is a general
consensus that a public policy can be defined as a course of action (or non-action; see
Bachrach and Baratz 1962) taken by a government or legislature with regard to a particular
issue. Although this definition is very broad, it emphasizes two constitutive elements. First,
public policies refer to actions of public actors (typically governments), although societal
actors might to some extent be involved or participate in public decision-making, which cor-
responds to the basic idea of governance (see, e.g., Kohler-Koch and Rittberger 2006).
Second, governmental actions focus on specific issues, implying that the scope of activities
is restricted to addressing a certain aspect or problem (such as air pollution control, animal
protection, internet content or the liberalization of the telecommunications sector).
Although this general definition seems straightforward, there is considerable variance in
its concrete specification. With regard to public actions, for example, some authors insist
that the presence of a policy requires the adoption of a larger number of related legislative
and administrative activities (see Knoepfel et al. 2007: 24). Other scholars adopt a narrower
definition and consider single governmental decisions or legal acts as public policies. This
perspective has been prominent in the study of changes to welfare state expenditure (for an
overview, see Green-Pedersen 2007). In this book, we will argue that the extent to which the
presence of a public policy requires only one type of action or a system of (interrelated)
actions – also known as policy bundles or policy mixes (see, e.g., Howlett and Mukherjee
Copyright © 2020. Macmillan Education UK. All rights reserved.

2018) – strongly depends on the issue at hand. Sometimes a single legal act might be very
encompassing and entail a broad range of different activities, while at other times it might
only constitute one out of several important elements of a public policy. It is hence not pos-
sible to specify generally a threshold number of courses of action as defining a ‘public
policy’.
In addition, we find different conceptions with regard to the nature of public policies as
issue related. On the one hand, such policies are seen as governmental activities made in
response to given societal or political problems. In other words, policy-making is conceived
as a problem-solving activity (Lasswell 1956; Birkland 2010: 7–11). On the other hand,
policy-making can be regarded as a means of exerting power by one social group over another
(Knoepfel et al. 2007: 21–2). According to this perspective, the existence and particular
design of policies are intended to protect the interests of certain groups, while disadvantag-
ing others. For example, studies of political clientelism deal with this aspect, i.e. the distri-
bution of selective benefits to individuals or clearly defined groups in exchange for political
support (see Kitschelt and Wilkinson 2007). It is certainly true that ‘all governments … give

greater weight to the preferences of those citizens with more political power than to the
preferences of those with less political power’ (Miller 2004: 20). However, this does not
mean that public policies represent ends in themselves. While the design of a given public
policy might reflect the differential power resources of social groups, this need not call into
question the requirement for a specific problem to exist or to be perceived in order for the
policy-making process to be initiated. As a consequence, the problem-solving and power
perspectives on public policy-making seem in practice to be compatible with each other.

Differences in Scope: Sectors, Targets and Instruments

So far, we have gained a basic understanding of the central elements that constitute a public
policy. However, this still leaves us with a broad range of activities that might constitute
public policies. This can be illustrated by the fact that the term ‘policy’ is used for activities
of very different scope.
First, it is often used to cover a whole range of different measures in a certain sector, such
as environmental policy, social policy, economic policy or fiscal policy. Used in this way, the
term grasps more than one legal act or political programme that belongs to the whole range
of legal and administrative activities that are related to a particular distinctive policy field.
Second, a similar approach is used to describe public activities in policy subfields. In most
instances, it is possible to classify the public activities in a field along certain subthemes that
cover functionally related measures. With regard to environmental policy, for example, sub-
sectors refer to water policy, clean air policy, climate change policy and waste policy. In social
policy, examples of subsectors include pension policy, unemployment policy and child ben-
efits (Knill et al. 2012).
Third, even within policy subfields, distinctive policy issues or targets can be identified.
Taking clean air policy as an example, such targets include industrial discharges of different
pollutants, urban air quality and car exhaust emissions. Hence a legal act – even if it is a mark-
edly specific one – can address more than one policy target (Steinebach and Knill 2018). For
example, in 2018 the European Commission adopted a proposal to update the drinking water
rules in a manner which aims to empower public authorities in managing water supply and
water quality, as well as consumers by giving them more oversight of the efficiency and effec-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tiveness of water suppliers. A third target group is business, since the new policy seeks to
boost a circular economy (aimed at continual use of resources) with regard to plastic bottles.
A fourth – and the least abstract – usage of the term refers to its connection with regulatory
instruments. While policy targets refer to what a legal act regulates, policy instruments define
how those targets are regulated. For example, one of the main tasks of immigration policy is
to regulate which individuals may legally enter a country. The most common instrument is the
establishment of a preference system based on quotas, targets or a ceiling, which allocates a
number of visas to certain groups of immigrants. Another – and clearly much stricter – instru-
ment would be an official ban on immigration (Adolino and Blake 2011: 103–4).
To sum up, public policy can be understood in any of the following four ways:

• Sector-specific measures (e.g. energy policy).

• Subfield-specific measures (e.g. renewable energy policy).
• Specific issues in the subfields (e.g. feed-in tariffs).
• Regulatory instruments connected to the issues (e.g. level of feed-in tariffs).

ANALYTICAL PERSPECTIVES ON THE POLICY-MAKING PROCESS

So far, we have approached public policies from a merely substantive perspective, focusing
on the central components that define a public policy. In this section, we adopt a procedural
perspective on public policies and take a closer look at the specific phenomena characteriz-
ing the policy-making process. First, we will briefly present the different theoretical perspec-
tives on the evolution of public policies. We will then distinguish the different stages of the
policy process, which entail distinct research questions and which hence might serve as a
basic analytical starting point for studying public policies.

asic Theoretical Perspectives on Public Policies: Rational Process Design,

B
Muddling Through or Just Chance?
In the literature, we find different attempts at analysing how public policies typically evolve
or should evolve. While the rationalist approach defines an ideal conception of how policies
should develop, the incrementalist perspective provides an explanation for the fact that in
reality this ideal is hardly ever reached. A more radical view is the one advanced by the ‘gar-
bage can’ model, which emphasizes that public policies often reveal the opposite pattern to
that envisaged by rationalist models (Cohen et al. 1972). Drawing on this model, Kingdon
(2003/2011) has developed the Multiple Streams Framework, which we will revisit and dis-
cuss at greater length in subsequent chapters.
The rationalist approach conceives of policy-making as a process of problem-solving.
Rather than seeking to explain the policy process, this approach prescribes an ideal concep-
tion of how policy-making should be organized and evolve in order to achieve optimal solu-
tions to the underlying policy problems. Hence it entails a normative (how policies should
evolve) rather than a positive (how policy can be explained) perspective on policy-making.
This question of how to optimally develop public policies was at the heart of Lasswell’s
(1956) thinking. He argued that ideally the policy process should be based on different steps
that follow a logical sequence: intelligence (collection and processing of all relevant knowl-
edge and information), promotion (identification and support of selected alternatives), pre-
Copyright © 2020. Macmillan Education UK. All rights reserved.

scription (imposition of a binding decision), invocation (policy enforcement), termination

(abrogation of policy) and appraisal (evaluation of policy effects against the backdrop of
initial objectives and intentions).
This prescription of how policies should be made was soon brought into question by its
overly ambitious view of rational decision-making. The theory of incrementalism explicitly
rejected the idea of public policy being made on the basis of a fully rational decision-making
process (Hayes 2006: 19). It originates in Lindblom’s (1959) path-breaking article ‘The
Science of Muddling Through’ and was further formalized by Hirschman and Lindblom
(1962), Braybrooke and Lindblom (1963) and Lindblom (1965). Public policy is regarded as
the political result of the interaction of various actors possessing different types of informa-
tion. These actors need to make concessions, and therefore policy-makers primarily concen-
trate on aspects that are less controversial and more technical. This process of ‘partisan
mutual adjustment’ can only lead to one outcome: incremental policy change. Empirically,
the basic aim of incrementalism – i.e. incremental policy change – has often been tested by
using budgetary data (see Wildavsky 1964; Berry 1990; Breunig et al. 2010).
Rather than an ideal, incrementalism purports to be a realistic description of how policy-
makers arrive at their decisions. This implies that policy-makers act within the context of

limited information, the cognitive restrictions of their minds and the finite amount of time
available for policy-making – all of which are largely congruent with the concept of ‘bounded
rationality’ (i.e. rationality is limited when individuals make decisions) put forward by
Simon (1955, 1957) and Cyert and March (1963). As a consequence, decision-makers apply
their rationality only after having greatly simplified the choices available, turning them into
‘satisficers’ who seek a satisfactory solution rather than the optimal one: ‘policy making is
serial and remedial in that it focuses heavily on remedial measures that happen to be at
hand rather than addressing itself to a more comprehensive set of goals and alternative poli-
cies’ (Rajagopalan and Rasheed 1995: 291). In this way, policy-makers are not confronted
with uncertain consequences, which may result from more radical reforms.
While incrementalism still presumes that policy-making is characterized by an – albeit
bounded – rational process in which solutions are developed in response to existing prob-
lems, the garbage can model (Cohen et al. 1972) questions even these less strict rationality
assumptions and disconnects problems, solutions and decision-makers from each other.
In contrast to the rationalist and incrementalist perspectives, decisions do not follow
an orderly process from problem to solution, but are the outcomes of several relatively
independent streams of events, namely problems, solutions, choice opportunities and
participants.
In contrast to the conventional view that problems trigger decision-making processes,
the garbage can model assumes that usually the involved actors within an organization go
through the ‘garbage’ first and look for a suitable fix, i.e. the ‘solution’. Hence, solutions exist
and develop independently of problems. The actors involved may have pre-existing ideas
about possible solutions – they may be attracted to specific ones and volunteer to be their
advocates. In many instances, solutions are prepared without knowledge of the problems
they might have to solve. Organizations thus tend to produce many solutions which are later
discarded due to a lack of appropriate problems. The independence of the problem and solu-
tion streams is further enhanced by the fact that participants may differ between the two
streams: the actors involved in the development of solutions might be different from those
who discuss the definition of the respective policy problems. Finally, the extent to which an
existing solution might actually be linked to a problem is affected by whether or not a choice
of opportunities exists, i.e. occasions when organizations are expected or perceived to be
Copyright © 2020. Macmillan Education UK. All rights reserved.

expected to produce a decision.

Stages of the Policy Process

So far, we have seen that the policy process can be approached from varying analytical angles
that entail highly different assumptions with regard to the sequencing of policy-making. In
the following it is not our objective to discuss the strengths and weaknesses of the different
approaches or to rank them in light of their analytical relevance. Rather, we ask if and to
what extent it is analytically useful to distinguish between different stages of the policy
process – notwithstanding the ambiguous picture that emerges from the previous theoreti-
cal discussion. In the literature, the most common approach is the distinction of different
policy stages that can be integrated into a process model: the famous ‘policy cycle’. This
approach models the policy process as a series of political activities, which basically consists
of the following phases: (1) problem definition and agenda-setting, (2) policy formulation
and adoption, (3) implementation and (4) evaluation (with the potential consequence of
policy termination or reformulation).

In many cases, the policy cycle model is interpreted as a sequential development, hence
following closely the idea underlying the rationalist approach. Accordingly, the policy pro-
cess starts with the identification of a societal problem and its placement on the govern-
ment’s agenda. Subsequently, various policy proposals are formulated, from which one will
be adopted by the decision-makers. In the next stage, the adopted policy is enacted, before
finally its impacts are evaluated. This last stage leads straight back to the first, indicating
that the policy cycle is continuous and unending.
In view of the previous theoretical discussion, it is not surprising that the idea of a cycli-
cal model based on a sequence of policy stages has been criticized. In particular, it has been
emphasized that the model hardly corresponds to empirical reality, as the different stages
might often overlap (see, e.g., Colebatch 2006; Jann and Wegrich 2007: 43–5). For instance,
a reformulation of policies might occur during the implementation stage, or some phases
might be skipped – policies are not always subject to a systematic evaluation. Notwithstanding
this criticism, the cycle model is employed as a standard approach for structuring the theo-
ries and concepts of public policy-making.
In this book, we will depart from the classic cycle approach by not assuming a sequential
model of the policy process, apart from the fact that current policy decisions are not inde-
pendent of decisions taken before and that policies under discussion today may have ‘knock-
on effects’ leading to further policies tomorrow (Newton and van Deth 2010: 266). In line
with many other authors (see, e.g., Fischer et al. 2007; Knoepfel et al. 2007; Hill 2009;
Howlett et al. 2009; Anderson 2010; Birkland 2010), we consider it more useful to conceive
of the different policy stages as potential analytical lenses on the policy-making process.
Depending on the specific lens, we focus on distinctive questions and apply distinctive con-
cepts and theories in order to explain observed patterns. If we focus on implementation, our
major goal is to analyse and explain the extent to which we can observe deviations from the
original policy goals. Looking at policy adoption, by contrast, the central objective is to
explain adoption or non-adoption; and in the case of policy formulation we want to account
for the specific design of a policy. In other words, we consider the distinction of different
stages as a heuristic tool that helps us to investigate the process of policy-making from dif-
ferent analytical angles (see Table 1.2).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Table 1.2 Policy stages and related research topics.

Problem definition and Policy formulation and

agenda-setting adoption Implementation Evaluation
Why do perceptions How do policy Why do certain policies How can policy
and definitions of decisions come fail? effects (outcomes
policy problems change about? and impacts) be
over time/vary across measured?
countries?
Why are certain problems How can policy Which factors account Which factors
ignored while others are outputs be for the variance in explain variation in
placed on the agenda? explained? policy implementation? policy effects?

OPPORTUNITIES AND CHALLENGES FOR POLICY-MAKING

In this book, we want to offer a realistic description of policy-making in modern societies.

Consequently, the subsequent chapters will elaborate on developments that can represent
both opportunities and challenges for policy-making. Amongst these, we concentrate on
three developments that have noticeably affected policy-making.
The first one is the budgetary constraints that governments face and which affect state
organization and policy-making. In North America and Europe, the financial crisis that
unfolded in 2007 and 2008 and the subsequent recession represented a significant challenge
(Schäfer and Streeck 2013). In this context, Armingeon and Baccaro (2012) argued that the
economic hardship experienced by European countries forced governments to implement
similar sets of austerity policies. The result of this was that to voters it seemed to make no dif-
ference who was in government, because all policy-makers proposed and adopted the same
measures. The economic and financial crisis has, however, also facilitated far-reaching reforms
and the adoption of new policies that would not have been in place, such as the European
Financial Stabilisation Mechanism, which allows countries to use bond issues on behalf of the
EU when experiencing severe financial difficulties (see, e.g., Tosun et al. 2014; Steinebach and
Knill 2018).
The second challenge is the ongoing process of economic globalization, which poses limits
on what governments can do in order to address policy problems. Climate change is a well-
known example where the emission of greenhouse gases (GHG) by some countries can
potentially affect all other countries and where economic interests shape countries’ policy
approaches to mitigate climate change. Given the negative externalities of GHG emissions
on the world’s climate, for a long time the dominant policy arena for addressing climate
change was the international one (see, e.g., Falkner et al. 2010). Due to the continued failure
to develop an effective international regime on climate governance, a new structure has
emerged that offers an arena for new actors such as cities and city networks (see, e.g.,
Bansard et al. 2017).
The third development that is crucial for policy-making is digitalization, i.e. the growing
importance of the internet and social media for supplying information on policy proposals
and their effects. The new information and communication media are a challenge for policy-
Copyright © 2020. Macmillan Education UK. All rights reserved.

making since they have changed how people inform themselves, which can lead to a gap
between what policy-makers think is important and what the public thinks policy-makers
should address. At the same time, social media can help policy-makers to assess how the
public feels about the measures proposed and what their acceptance would be (see, e.g.,
Ceron and Negri 2016).
Of course, there exist far more developments that affect policy-making, such as the rise
of populist parties like the People’s Party in Denmark, the Party for Freedom in the
Netherlands or in Italy the League and the Five Star Movement (e.g. Mudde and Kaltwasser
2012; Rooduijn 2018; Caiani and Graziano 2019), and new forms of public pressure as
exerted by the Fridays for Future Movement, for instance. However, the three we alluded to
are those that will frequently re-appear in the subsequent chapters. Moreover, these chal-
lenges (and potential opportunities) manifest themselves in different forms, which is con-
ducive to the purpose of this book and the topics it covers.

S TUDYING PUBLIC POLICY: THE APPROACH AND STRUCTURE

OF THIS BOOK
The study of public policy refers to the analysis of the contents, causes and conditions as well
as the outcomes and impacts of governmental activities. It constitutes a highly diverse and
dynamic research area. As already emphasized, it is nevertheless possible to identify domi-
nant themes that guide the study of public policy: the explanation of policy variation and
policy change across countries and policy sectors. These themes form the basic background
for this book. At the same time, however, it is obvious that policy variation and change can
be studied from very different perspectives, focusing, for instance, on the different elements
of public policies or on the causal effects of different factors (such as socio-economic condi-
tions, political and institutional factors or international influences). Hence, even if we iden-
tify overarching research themes, there is still a need for additional analytical considerations
that help to structure the analysis of public policy.
We formulate three guiding questions that we will repeatedly address throughout this
book:

• How do public policies come about?

• When does policy change happen?
• What are the effects of public policies?

These questions reflect the state of research in policy studies. The first research question
taps into the origins of public policy and allows a wide range of topics to be covered, includ-
ing questions related to path dependency and whether policy-making is under all circum-
stances initiated by the need to resolve a problem, howsoever defined. The second research
question aligns with the main research interest of an influential strand of policy studies
known as the theories of the policy process (see Weible and Sabatier 2017). While advancing
different arguments, this literature is coherent to the extent that all theories of the policy
process address causes and patterns of policy change. Lastly, public policies can only become
effective when they are implemented, and even then, they can have numerous intended and
unintended consequences. Some topics addressed here deal explicitly with that perspective
Copyright © 2020. Macmillan Education UK. All rights reserved.

and link the actual or perceived effects of public policies back to the previous two research
questions. In this book, we differentiate between three analytical steps.

Basic Analytical Tools, Concepts and Theories

In the first step, we familiarize the reader with the basic analytical tools, concepts and theo-
ries that form the background to the study of public policy. In Chapter 2, we focus on the
central actors and institutions that are involved in the development, formulation, execution
and evaluation of public policies. As already emphasized, public policies are affected by many
factors, including in particular polity (the institutional context in which policies are devel-
oped and implemented) and politics (the interests of the involved public and societal actors
and the cleavages and relationships between them). In order to shed more light on potential
interlinkages between polity, politics and policy, we discuss the institutions and actors that
are most central for understanding the making and shaping of public policies.

In Chapter 3, we address two key research questions relating to the nature of public pol-
icy. They include, first of all, the classic question of whether and to what extent it is possible
analytically to distinguish different policy types. Second, a long-standing debate centres on
the topic of policy styles. Is it possible to identify country- or sector-specific process pat-
terns that characterize the making of public policies? Which factors account for differences
in policy styles across countries or sectors?
Chapter 4 provides an overview of general theories that are applied in order to explain
policy variation and policy change. What are the central explanatory factors that are of rel-
evance in this regard? In the literature, varying approaches can be identified. They include
structural explanations (in which socio-economic structures and the pressure of problems
constitute the basic explanatory variables); interest group models (emphasizing the crucial
role of the power and resources of societal interests); political economy accounts involving
self-interested, vote-seeking politicians and budget-maximizing bureaucrats; models focus-
ing on the ideological and programmatic differences between political parties; and institu-
tional theories (in which major explanatory relevance is attached to the policy implications
of a country’s political institutions).

Stages of the Policy Process

Having gained an understanding of the basics of public policy analysis, we shift the focus to
the second analytical step: the specific analysis of the central stages of the policy process. So,
different actors might be involved to different degrees during the stages of problem defini-
tion, agenda-setting, policy formulation and adoption, implementation and evaluation –
implying that constellations of strategic interaction as well as institutional rules affecting
strategic opportunities and constraints for the involved actors will vary. If and to what
extent theoretical models and explanatory approaches need to take account of and be
adjusted in the light of these aspects form an important research topic. This is further
emphasized by the fact that the distinction of different stages typically goes hand in hand
with the different research questions that are addressed. For example, with regard to prob-
lem definition, the focus is on the factors that account for why a certain problem is perceived
in a certain way. For agenda-setting, by contrast, we are interested in the reasons that
Copyright © 2020. Macmillan Education UK. All rights reserved.

explain why a certain problem is placed on the political agenda when others are not. With
regard to policy adoption, the major focus is on the explanation of the specific policy design
that is selected; while for the implementation stage, scholarly attention shifts to the analy-
sis of the potential deviations from the original policy objectives and the underlying reasons
for these shifts. Policy evaluation is about the effects of public policies.
Chapters 5, 6, 7 and 8 take up the phases of the policy cycle. Chapter 5 addresses the
phase of problem definition and agenda-setting. The second phase of the policy cycle
(decision-making) is explored in Chapter 6. Decision-making consists of two related con-
cepts – policy formulation and policy adoption – which we will discuss at length. Of these
two processes, policy adoption is more concerned with institutional arrangements, whereas
policy formulation focuses more explicitly on the various forms of political power. Chapter
7 addresses the implementation of policies that have passed the adoption process. We out-
line why at the implementation stage bureaucrats might deviate from the goals originally
formulated by policy-makers, and we discuss how the likelihood of ‘bureaucratic drift’ can

be effectively reduced. Once we have clarified how public policies come about, we shift our
attention to policy evaluation in Chapter 8.

Cross-Cutting Themes
In the third analytical step, we shift from a stage-centred analysis of the policy process to a
study of the key topics of public policy that cut across different policy stages. In Chapter 9
we consider the topic of governance – a term that has been much developed since the early
1990s. Governance typically refers to the distinctive analytical perspective on policy-making
that emphasizes the sharing – sometimes even blurring – of competencies between public
and private actors. It cuts across the policy stages model by focusing on the exchange rela-
tionships between public and private actors when they are confronted with political prob-
lems. By explicitly emphasizing the potential problem-solving contributions of private
actors and the potential restrictions for public problem-solving capacities as a result of
growing international and domestic-level interdependence, the governance perspective
shifts the analytical focus from a purely state-centred perspective to one which deals with
the patterns of state–society relations.
In Chapter 10 we take account of the fact that the development and implementation of
national policies are increasingly overlapped by similar activities at the international level:
the public policies which exist beyond the nation state. There are many reasons that favour
the development of joint policies at the international level. At the same time, however, both
the adoption and implementation of international policies are characterized by specific chal-
lenges that are increasingly addressed by scholars of public policy.
In Chapter 11, the focus is on the analysis of policy change and policy convergence. The
study of policy change is a demanding one. Areas of policy change investigated by public
policy scholars have included welfare state retrenchment and race-to-the-bottom scenarios
in the fields of environmental protection and labour standards. Policy change is closely
related to the concept of cross-national policy convergence. Convergence research investi-
gates whether public policies in various countries become increasingly similar over time.
Finally, in Chapter 12 we summarize the main characteristics of policy-making on the
basis of the explanations we have given in the course of this book. The aim of this concluding
Copyright © 2020. Macmillan Education UK. All rights reserved.

chapter is to determine what we have learned throughout the book, which also includes
returning to the three basic questions formulated earlier. Then, this chapter offers an inte-
grated presentation of how public policies can be compared – comparison will be addressed
at multiple instances in the book, but in the concluding chapter we will draw together the
different aspects. The very final topic to be covered by this book is the relationship between
the theory of policy analysis and the practice of doing policy analysis. To this end, we will
formulate some guidelines and recommendations that the readers may find helpful for such
an endeavour.

Reader’s guide
The pervasive importance of public policies in ordinary life is manifest in such diverse
areas as defence, education, environmental protection, health care, unemployment
benefits, motorway construction, monetary issues and taxes. For analytical purposes,
however, it is useful to think of ways of categorizing public policies into specific
groups, as this allows us to make more general statements. The objective of this chap-
ter is to introduce and discuss criteria employed in the policy-analytical literature for
categorizing public policies. The most widely applied instrument for achieving this
refers to the construction of typologies, though the actual idea underlying these
typologies is that specific groups of public policies entail equally specific patterns of
policy-making. Thus, conceiving of these typologies as tools for descriptively catego-
rizing public policies would not do them justice. We will suggest alternative approaches
to categorizing public policies, which include considerations of how they bring about
changes in the behaviour of those at whom the policy is aimed and how they shed light
on different policy dimensions. While typologies represent simple analytical frame-
works, the term ‘dimensions’ concerns the different aspects of a public policy which
can refer to the stage of its development or its contents. We will also introduce the
concept of policy styles, which refer to ways in which governments make and imple-
ment public policies: a perspective which further underscores the fact that the nature
of public policies cannot be fully understood without paying attention to politics.
Copyright © 2020. Macmillan Education UK. All rights reserved.

More generally, the topics addressed in this chapter represent the key issues of ‘classi-
cal’ public policy analysis, and serve as the basis for subsequent explanations by pro-
viding a better understanding of what public policies are and how their particular
characteristics might influence the process by which they are made.

INTRODUCTION
As outlined in Chapter 1, political science is mainly concerned with three major analytical
dimensions: polities, politics and policies. Polities are the institutional arrangements char-
acterizing a political system. Politics refer to the patterns of the policy-making process,
involving various types of actors and their interactions. (Public) policies refer to the output
of a political system as it is realized in practice, including the laws, regulations, decisions,
plans, programmes and strategies that follow a particular purpose: they are designed to
achieve defined goals and present solutions to societal problems.

13
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:21:03.
14 Public Policy

These specifications of what constitutes a public policy still leave us with an immense
field of governmental actions. A public policy can as much relate to the minutiae of unem-
ployment benefits and the requirements of air pollution control as to the legalization of
drugs or the government’s decision to use daylight saving time. To grasp the nature of public
policies, identifying nominal policy types is insufficient, as there exist many different dimen-
sions that define the specific content of a policy. For example, a policy aiming to limit air
pollution might be composed of many varying measures, usually referred to as policy instru-
ments, such as maximum permissible limits for all possible pollutants emitted into the air
by mobile sources, e.g. passenger cars, stationary sources or production plants. Even if
described in an identical manner, such as ‘clean air policy’, public policies are likely to differ
with regard to their design from one jurisdiction to another. So how could we possibly make
generalizable statements about clean air policy? Or more generally, what are the typical
characteristics of different policies? Given the huge variety of different policies, many
attempts to classify policies are based on predefined analytical criteria. The latter include
the characteristics of the policy-making process, the effects of a policy with regard to the
allocation of costs and benefits concerning the relevant actors, or the way in which a policy
seeks to affect the behaviour of individuals in concert with certain policy objectives.

TYPOLOGIES OF PUBLIC POLICIES

The policy-analytical literature has strongly relied on typologies for classifying different
types of public policies. Most essentially, a typology allows for grouping entities by similar-
ity. Initial attempts at classification followed rather obvious lines of similarity in descriptive
terms, such as nominal typologies based on the subject matter, e.g. environmental policy,
social policy, economic policy or agricultural policy. Other nominal modes of classification
draw on periods (e.g. post-war policy), ideology (e.g. conservative policy) or institutions
(e.g. regional policy, local policy, European policy; McCool 1995: 175). With the maturation
of public policy analysis, typologies have increasingly emphasized analytical features. In this
section, we present the most influential analytical classifications of public policies. Each of
these classifications adopts a different analytical lens, stressing either the possibility that
Copyright © 2020. Macmillan Education UK. All rights reserved.

public policies can be grouped with regard to their implications for the politics dimension or
how they affect the behaviour of the target group.

Classification by Implications for Politics

The classical approaches to the study of public policy have often used typologies as ‘analyti-
cal shortcuts’ for the underlying decision-making processes. The two most influential typol-
ogies have been put forward by Theodore J. Lowi and James Q. Wilson.

Lowi’s Policy Typology

The classification proposed by Lowi (1964, 1972) corresponds to long-recognized differ-

ences amongst public policies and their making, leading to the famous causal statement that
‘policies determine politics’. The categorization consists of (1) measures which affect the
distribution of resources from the government to particular recipients (distributive policies),

(2) measures which transfer resources from one societal group to another (redistributive poli-
cies), (3) measures which specify conditions and constraints for individual or collective
behaviour (regulatory policies) and (4) measures which create or modify the state’s institu-
tions (constituent policies).
Of these four policy types, Lowi expects the politics of constituent policy to attract the
lowest degree of public attention, as only the top political stratum tends to be concerned.
Distributive policies use public funds to assist particular societal groups. Those groups
that benefit from distributive measures usually do not directly compete with one another.
The costs are assigned to the public (i.e. all taxpayers) rather than a specific group (Anderson
2010: 9–17). In this way, distributive policies appear to create only winners and no specific
group of losers. Consequently, the policy process reflects a rather consensual pattern, as the
potential for conflicts between winners and losers is very limited.
Regulatory policies define rules for human behaviour and only indirectly affect the distri-
bution of costs in society. As Levi-Faur (2011) explains, the costs associated with regulatory
policies are compliance costs, which are borne by the regulated parties and not by the gov-
ernment budget. Yet, depending on who is regulated and how, regulatory policies can still
entail political conflict. Therefore, it is difficult to predict the degree to which regulatory
politics is contested.
For redistributive policies, by contrast, such conflicts are always more likely, as they
involve the reallocation of costs and benefits between different societal groups. A typical
cleavage in this regard refers to resource reallocations between rich and poor people (e.g.
by introducing progressive taxation or income-based benefits). Another cleavage underly-
ing redistribution might be observed between younger and older generations (e.g. a pen-
sion scheme in which the contributions of younger people are used to pay the pensions of
retired people). Any change in the pension level or the contribution level might influence
the allocation of costs and benefits between these groups, hence often inducing a highly
conflictual political process characterized by the broad mobilization of societal actors (see
Table 2.1).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Table 2.1 Lowi’s policy typology.

Type of policy Definition Examples

Regulatory Policies specifying conditions and Environmental protection; migration
policies constraints for individual and policy; consumer protection
collective behaviour
Distributive Policies distributing new (state) Farm subsidies; local infrastructure
policies resources such as highways and schools
Redistributive Policies modifying the distribution of Welfare; land reform; progressive
policies existing resources taxation
Constituent Policies creating or modifying the Changes of procedural rules of
policies state’s institutions parliaments; creation of new
governmental agencies

Sources: Based on Lowi (1972: 300) and Knill and Tosun (2011).

While initially compelling, on closer inspection the clarity of the typology dissolves and
conceptual ambiguities amongst the policy types can be identified, which Lowi himself also
acknowledged (1964: 690). The typology’s main flaw is that it taps into different analytical
considerations. While the distinction between redistributive and distributive policies is
based on different policy effects – i.e. the (re)distribution of resources – Lowi emphasizes
the regulation of human behaviour as the central criterion of the regulatory policy type. This
creates the false impression that regulatory policies have no distributive or redistributive
effects, which the rich literature on regulation and regulatory governance has falsified (see
Levi-Faur 2011; Baldwin et al. 2012). Most importantly, the emergence and spread of regu-
latory impact assessments demonstrate that regulatory policies are subject to cost–benefit
considerations of governmental agencies (see Dunlop and Radaelli 2016a, 2016b). For con-
stituent policies, the analytical focus shifts again to the subject matter of a certain policy, in
this case the polity dimension.
There are also problems concerning the distinction between redistributive and distrib-
utive policies. First and most importantly, it seems hardly possible ex ante to define a
policy as distributive or redistributive. This classification depends on the individual per-
ceptions of the actors concerned: the allocation of transfers to a certain group might trig-
ger strong opposition, even though the money is taken from the general budget; or
taxpayers may mobilize against such transfers if they fear future tax increases. In other
words, in many instances one can classify a policy as distributive or redistributive only ex
post. Second, governments might try to influence the perception by societal actors of the
costs and benefits by strategically labelling a policy as distributive notwithstanding its
redistributive effects. Third, the perception of a policy as distributive or redistributive
might vary over time. For example, in Bismarckian welfare states (a social insurance sys-
tem financed by employers and employees through payroll deduction that goes back to
German Chancellor Otto von Bismarck; see Seeleib-Kaiser 2016), when there are high
austerity pressures and high unemployment, any increase in contribution levels is more
likely to be perceived as redistributive activity, as would be the case during an economic
boom. Finally, we should emphasize that the effects of a policy are not the only factor
influencing policy-making patterns. In addition to the policy type, many other variables –
such as institutional arrangements, the party system or general relationships between
Copyright © 2020. Macmillan Education UK. All rights reserved.

state and society – might have an impact on policy-making. In light of these reflections,
we must conclude that it is not just policies determining politics.
Another issue raised in this context is whether the typology is complete. A case in point
is the recent discussion on the extent to which ‘morality policy’ can be considered as an
additional policy type, hence supplementing Lowi’s distinctions. This particular policy,
amongst other things, comprises ‘abortion, capital punishment, legalized gambling,
homosexual rights, pornography [and] physician-assisted suicide’ (Mooney 1999: 675). It
is argued that the distinctive feature of morality policies is that they entail the regulation
of conflicts amongst social values (Baumgartner et al. 2008a; Engeli et al. 2012; Knill
2013; Knill et al. 2015). This feature makes this policy type distinct from policies of distri-
bution and redistribution or social and economic regulation, where conflicts over tangible
resources dominate political processes. And by no means can morality issues be assigned
to the category of constituent policies (Box 2.1).

Box 2.1 Politics of same-sex marriage

An example of morality policies is the legalization of same-sex marriage. In the last
few years, there have been highly controversial debates as to whether or not marriage
between two persons of the same biological sex should be recognized. These debates
have involved arguments based on civil rights, political, social, moral and religious
considerations. In fact, in many countries gay people are still confronted with dis-
crimination and cannot live in a way that corresponds with their sexual orientation.
Since the early 2000s, countries across the world have legalized same-sex marriages or
have established civil unions as an alternative for same-sex couples. According to the
2019 dataset on sexual orientation laws of the International Lesbian, Gay, Bisexual,
Trans and Intersex Association, South Africa is the only African country where same-
sex marriages are legal. In Latin America and the Caribbean, Argentina, Brazil, Chile,
Colombia, Ecuador, Mexico and Uruguay offer marriage or civil unions to same-sex
couples. In North America, both Canada and the United States have pertinent regula-
tions in place. Israel is the only Asian country where civil unions are possible, whereas
many European countries (too many to be listed individually here) have legalized
same-sex marriage or civil unions, as did Australia and New Zealand.
Following Lowi’s typology, the legalization of same-sex marriage would correspond
to regulatory policy as it specifies the conditions for individual behaviour. The policy-
making process in most countries, however, has been more characterized by high
degrees of conflict, the involvement of powerful interest groups such as churches, and
numerous instances of failed policy proposals due to unfeasible compromises. In
Canada, for example, which legalized same-sex marriage nationwide with the enact-
ment of the Civil Marriage Act in 2005, numerous bills tabled in parliament between
1997 and 2004 did not pass the first reading due to the elevated degree of conflict and
the impossibility of reaching a compromise. Similarly, in Argentina, which liberalized
same-sex marriage in 2010, there was an intense battle of words between the govern-
ment and the Roman Catholic Church, which organized large protests in the tens of
thousands. In Germany, the plenary debates and the subsequent roll-call voting on
Copyright © 2020. Macmillan Education UK. All rights reserved.

same-sex marriages in 2017 demonstrated that members of parliament are split on

this issue, eventually still leading to a majority of yes-votes in parliament.

Even though morality policies from Lowi’s perspective could still be viewed as regulatory
policy, it is argued in the literature that such policies should be conceived of as an additional
policy type, as one particularly prone to political salience, societal mobilization, controver-
sial political debate, the responsiveness of politicians to public opinion, political conflict
dominated by value-based rather than rational-egoistic orientations and taboos. All this
should lead to a very specific form of politics, i.e. morality politics, which can hardly be
expected to result in compromise, as the subject of political debate concerns conflicts
amongst fundamental values (see Meier 1994; Mooney 1999, 2001; Patton 2007; Hurka
et al. 2017).

Wilson’s Policy Typology

Who benefits from a policy? Who has to carry its costs? These questions underlie the
alternative approach to classification developed by Wilson (1989, 1995), who rejects
ambiguous policy types and distinguishes instead between policies on the basis of whether
the related costs and benefits are either widely distributed or narrowly concentrated. Each
of the four possible combinations yields different implications for policy-making. When
both costs and benefits of a certain policy are widely distributed, a government may
encounter no or only minor opposition, indicating majoritarian politics as the likely out-
come, with policy-makers following very closely the preferences of the electorate when
making their decisions. An example of majoritarian politics is universal health care, as it
spreads both benefits and costs across relatively large segments of the population (Oliver
2006: 211). This type of politics involves the basic ideological beliefs of the political actors,
and therefore legislative debates are very important for a visible expression of the respec-
tive view on the issue concerned. In a similar vein, interest groups are involved in the
policy-making process, and they use multiple avenues to express their preferences and
win support for them.
When, by contrast, both costs and benefits of a certain policy are concentrated, a govern-
ment may be confronted with opposition from rival interest groups, signalling interest group
politics. In this case, the expectation is that policy decisions will be strongly affected by the
positions and resources of relevant interest associations. Where there are clear winners and
clear losers, the level of conflict is high and the outcome of any single proposal is highly
unpredictable – which should give way to incrementalism: the making of minimal adjust-
ments to existing policy arrangements. For example, the traditional approach to agri-food
policy was characterized by the adoption of policies that concentrated exclusively or pre-
dominantly on the interests and needs of farmers (Skogstad 1998). However, agri-food poli-
tics has shifted towards new institutional arrangements, which Daugbjerg and Feindt (2017)
label ‘post-exceptionalism’, and which comprise incremental adjustments to the previous
agri-food policy regime.
On the other hand, if costs are concentrated and benefits diffuse, a government may
encounter opposition from dominant interest groups. In this case, entrepreneurial politics is
Copyright © 2020. Macmillan Education UK. All rights reserved.

the probable outcome: that policy change requires the presence of ‘political entrepreneurs’
who are willing to develop and put through political proposals despite strong societal resis-
tance. Opportunities for entrepreneurial politics usually come with special events, e.g. natu-
ral disasters. The organization of nationwide demonstrations against nuclear power and
environmental degradation by environmental groups in Taiwan is an example of entrepre-
neurial politics, as these protests aimed to introduce environmental policy that would pro-
mote the benefit of the public (Tang et al. 2011).
The fourth and final scenario consists of a situation in which costs are diffuse and bene-
fits concentrated. In such a case, governments are likely to be confronted with a relevant
interest group that is favourable to their reform endeavours, indicating that clientelistic poli-
tics is the likely outcome. Examples of clientelistic politics are health programmes that ben-
efit special groups (Oliver 2006). Another examples relates to agriculture politics, where
policymaking was constrained to actors based in this specific policy sector (foremost
amongst them agriculture ministries and farmers’ associations), resulting in corporatist
structures of interest intermediation (Skogstad 1998; Daugbjerg and Feindt 2017).

Table 2.2 Wilson’s typology.

Benefits
Costs Concentrated Diffuse
Concentrated Interest group politics (‘zero-sum game’) Entrepreneurial politics
Diffuse Clientelistic politics Majoritarian politics

Sources: Based on Wilson (1989) and Knill and Tosun (2011).

Clientelistic politics mostly takes place behind the scenes and is the result of a consensus
between political actors and interest groups, which allows both sides to pursue their respec-
tive interests (see Table 2.2).
Compared with Lowi’s typology, the approach by Wilson is analytically more compelling,
since it is more precise about the characteristics of policy-making and the actors involved in
it. Another advantage is that it is more dynamic, since formulating a particular politics may
shift from one type of politics to another. For example, in its very early stage environmental
politics may be regarded as entrepreneurial politics, since those who are regulated still rep-
resent a small group. With the proliferation of environmental policies, however, the corre-
sponding debate moves from entrepreneurial to majoritarian politics. Another example
shows that clientelistic politics can be replaced by majoritarian politics. With democratiza-
tion in Ghana, the government changed its approach to health policy and broadened the
group of recipients of health services (Carbone 2012).

Classification by Governance Principles and Instruments

A second approach to classifying public policies is based on underlying governance princi-
ples and instruments. As we will discuss in detail in Chapter 9, governance is about the
political steering of public and private actors (Héritier 2002a, 2002b). Accordingly, gover-
nance instruments specify the means or techniques by which governments seek to achieve
their policy goals (such as emission limits for environmental pollutants), whereas gover-
Copyright © 2020. Macmillan Education UK. All rights reserved.

nance principles refer to specific classes or groups of governance instruments. The analytical
and political relevance of governance principles and instruments constitutes the central
focus of various classification attempts in the literature (e.g. Hood 1986; Peters and van
Nispen 1998; Knill and Lenschow 2000; Salamon 2002; Bemelmans-Videc et al. 2003;
Eliadis et al. 2005; Holzinger et al. 2006; Hood and Margetts 2007; Goetz 2008).
While many typologies apply a highly differentiated scheme, ending up with long lists of
different instruments, the NATO scheme developed by Hood (1986; see also Hood and
Margetts 2007) represents a more parsimonious approach. It classifies governance princi-
ples based on four central resources of government, namely Nodality, Authority, Treasure
and Organization (NATO). These resources of government illustrate the different ways ‘to
get people to do things that they might not otherwise do; or it enables people to do things
that they might not have done otherwise’ (Schneider and Ingram 1990: 513).

Nodality

Nodality refers to the central role governments enjoy with regard to the use and distribu-
tion of information within political systems. Governments constitute large institutions
involved in a wide range of activities, and therefore they have more expertise and infor-
mation than most other societal actors (O’Toole and Meier 1999: 511; O’Toole 2007: 218–
21). This property of being at the centre gives rise to nodality, placing governments in a
strategic position from which to spread information to society and to detect information.
Typical examples of policy instruments based on this resource are the publication of data
and information, education, advice, recommendation and persuasion (Vedung 1998: 33).
These instruments are typically based upon an indirect governance logic that attempts to
change the beliefs and perceptions of public and private actors, hence changing individual
or collective behaviour in order to achieve political objectives. The most commonly
observed type of tool is the government information campaign. This includes such cam-
paigns as that to encourage citizens to receive vaccination against certain diseases (e.g.
seasonal flu). Another type of this tool relates to communication activities aimed at
changing producer behaviour by providing consumers with product and production pro-
cess information. A well-known example is ecolabelling for food and other consumer
products. Ecolabels are intended to make it easy for consumers to take environmental
concerns into account when shopping, as they certify compliance with a set of minimum
requirements.
The advantage of nodality tools lies in the relatively low cost of their application and
implementation; their weakness lies in their often limited and uncertain effectiveness.
Following Howlett (2009a: 31), many scholars argue that these tools are restricted to
policies that do not require complete compliance in order to be effective, and that only
apply when government and public interests coincide so that government appeals are
likely to be favourably received. Though they also appear to be appropriate in relatively
short-term crisis situations (e.g. flu epidemics), this is the case only when it is other-
wise difficult to impose sanctions and where the issue in question can be reduced to the
level of advertising slogans. Consequently, nodality instruments in reality are often
combined with other tools.
Copyright © 2020. Macmillan Education UK. All rights reserved.

A recent development in research on information-based policy instruments concerns the

use of behavioural insight for policy design (Shafir 2013), which is connected with an influ-
ential book by Thaler and Sunstein (2009) on structuring choices so that governments can
get people to make better choices in a manner without limiting their freedom. These changes
in the choice architecture are brought about by nudges, which must fully preserve freedom
of choice and operate through providing information. A 'nudge' could, for example, consist
in informing people about the calories they would burn by taking the stairs instead of the
lift. Nudges can be used to influence individuals’ behaviour with regard to road safety, health
care, environmental protection and tax policy. Nudges have been in use in the United States
for some years, where they have been used, for example, by the Department of Agriculture
to promote better nutritional choices. The British government even launched a Behavioural
Insights Team to inform policy-making by nudging (Sunstein and Reisch 2018). Together
with Australia, Canada, Denmark, France, Germany, the Netherlands, Peru and Singapore,
the United Kingdom and the United States are countries where behavioural insights are
systematically used for designing nudges (see World Bank 2019).

Authority

Authority is defined by the use of the law as the central resource for governmental interven-
tion. Authority implies the legitimacy of legal or official power and gives the government the
ability to force societal actors to follow legal rules. Typical instruments applied in this con-
text are command-and-control instruments, demanding, forbidding, guaranteeing and
judging whether the authoritative rule may or may not be followed by negative sanctions.
Authority can be equated to what Lowi (1964, 1972) called regulatory policies. The central
characteristic of these policies is to achieve behavioural changes directly by altering the legal
conditions in which public and private actors operate. Authority-based instruments change
the strategic opportunities and constraints for the involved actors (Knill and Lehmkuhl
2002a).
Authority-based policy can take many forms. The most common is certainly that of so-
called command-and-control regulation. Here the government hierarchically prescribes
requirements that must be fulfilled by the regulatees, with failures to comply usually involv-
ing penalties. In addition to this interventionist approach, the formulation and implemen-
tation of authority-based policies might occur in more cooperative forms, entailing a more
or less far-reaching participation by, and delegation of power to, private actors. These pat-
terns might be characterized by forms of regulated self-regulation, where governmental
activity is restricted to setting up a broad framework for private self-regulation (Knill and
Lehmkuhl 2002b). An example is the regulation of illegal and harmful content on the inter-
net, which in many countries is based on regulatory frameworks for the self-regulation of
internet providers (Akdeniz 2008).
A further important distinction is made between economic regulation and social regula-
tion. While for economic regulation the focus is on the conditions specifying access to and
operation within markets (such as cartel regulations, price regulations, taxes and licensing
requirements), social regulation is concerned with the reduction and avoidance of negative
effects that emerge from economic market activities. Classical fields of social regulation are
health and safety at work, environmental pollution or consumer protection (May 2002).
A central advantage of authority-based policies is the high predictability of policy effects,
at least if effective implementation of the legal rules is assumed. Regulation may be politi-
Copyright © 2020. Macmillan Education UK. All rights reserved.

cally appealing, especially when the public expects quick and committed action on the part
of the government (Howlett et al. 2009: 120). Furthermore, especially with regard to risky
or dangerous sectors (such as nuclear safety), regulation constitutes the most feasible gov-
ernance option and one that can hardly be substituted with other approaches that rely on
effecting voluntary compliance or behavioural change (Holzinger et al. 2006). A final advan-
tage of regulation from the perspective of governments is that they have only minor budget-
ary implications. Potential costs usually have to be carried by those affected by a policy, i.e.
the policy addressees.
These advantages, however, come with several problems of authority-based instruments.
First, regulation generally entails high costs with regard to controlling and monitoring its
proper enforcement. Second, it is argued that regulation implies there are no incentives for
policy addressees to go beyond legal requirements; so there is no stimulus for innovation
(Holzinger et al. 2006). Third, there is a danger of ‘regulatory capture’, which occurs when a
public authority or agency created to act in the public interest instead acts in favour of the
commercial or special interests that it has been charged with regulating. Regulatory capture
is particularly likely when the design of regulatory rules requires detailed scientific and

technical information. As the policy addressees (i.e. the regulated industries) often have a
deeper knowledge than the regulating authorities, there is a certain probability that the
regulators become dependent upon the regulated, which gives the latter an important lever-
age in influencing regulatory decisions (see Chapter 3).

Treasure

Treasure resources are based on money, or more precisely the various financial tools of gov-
ernments. These tools encompass anything that can be freely exchanged, which may mate-
rialize as rewards as well as fines. These kinds of tools are, however, different from
authoritative rules in the sense that they are voluntary; i.e. policy addressees are not legally
obliged to adopt the measures involved. Typical instruments based on treasure involve posi-
tive and negative financial incentives. A prominent form of treasure-based instrument is the
grant. Grants are offered to producers in order to stimulate them to produce more of a cer-
tain good or service than they would otherwise (such as research grants for universities or
grants for public transportation). Other instruments include tax subsidies (i.e. deductions,
special rates or exclusions) in response to a certain activity or behaviour (e.g. energy-saving
measures in private households), government loans at an interest rate below the market
rate and financial disincentives (taxes and user charges; Howlett et al. 2009: 125).
The advantage of treasure-based policies is that they are easier to implement than con-
trol- and monitoring-intensive regulatory policies. By allowing target groups to devise
appropriate responses, the provision of financial incentives is generally expected to stimu-
late innovative behaviour on the side of policy addressees. In addition, in many instances
financial tools are characterized by high levels of political acceptance, as benefits are rela-
tively concentrated on certain societal groups and the costs are dispersed rather widely
across all taxpayers (see Wilson’s typology above).
The disadvantages associated with these tools, by contrast, refer to the fact that they
strongly affect the public budget. This holds true as long as these policies are designed to be
distributive rather than redistributive. Second, it is often difficult to calculate the level of
incentives in such a way that they actually unfold their expected effects. Third, financial
incentives might be redundant, entailing a windfall on the side of the recipients: it might be
Copyright © 2020. Macmillan Education UK. All rights reserved.

the case that some of the policy addressees would have pursued these actions anyway (i.e.
also without governmental stimulation).

Organization

Organization refers to the reliance on formal organizational structures in order to achieve

policy objectives. Instead of relying on information, legal requirements or fiscal incentives
or disincentives to influence the behaviour of societal actors, in this case governments
directly provide public goods or services (Mayntz 1979). Classic examples include national
defence, diplomatic relations, education or road construction and maintenance. A very
prominent form of direct state provision is the establishment of public or state-owned
enterprises, including railway companies and telecommunications or electricity suppliers.
Although many countries from the 1980s onwards witnessed an overwhelming trend of
privatization, indicating a shift from the state as a provider to the state as a regulator of the
provision of goods and services (see Majone 1996), the public sector still plays a

considerable role as a governing body. This holds true in particular if we consider the huge
number of publicly owned companies that exist at the local level, such as local transporta-
tion or electricity and water supply companies. In this context, it should be noted that the
existence of publicly owned companies applies more to some countries (e.g. Germany) than
to others (e.g. England).
While nodality, authority and treasure are characterized by the governmental activities
of either directly or indirectly influencing target group behaviour, organization-based tools
lack this intermediary step. Rather, they rest on the direct provision of public goods by the
state itself. This approach bears several advantages. In particular, problems of indirect provi-
sion are avoided, including political conflicts, long processes of negotiation or ineffective
implementation. In many cases, public services or goods (such as railway networks) are pro-
vided, although private actors would not regard them as profitable to provide.
But there are also disadvantages associated with organization-based tools. First, public
enterprises might lead to inefficient operations because poor performance does not lead to
bankruptcy, but is compensated for by the taxpayer. Second, political conflicts can affect the
provision of public goods and services, implying that contemporary political needs (emerg-
ing from elections, for instance) become more important than serving the public as a whole.
There is evidence from Italian politics that more infrastructure expenditures were given to
those districts electing the politically more powerful deputies from the governing parties as
a reward for their core voters (Golden and Picci 2008). In the public choice literature, such
behaviour is discussed under the term ‘pork-barrel politics’, which essentially refers to
spending that benefits the constituents of a politician in return for their political support
(see Ferejohn 1974; Cox and McCubbins 1986). Third, principal–agent problems might
emerge in the sense that the management of public enterprises, i.e. the agent, can pursue
different objectives than the political principals, i.e. the government. In these cases princi-
pals are unable to ensure compliance with their goals because of an inability to monitor
agency activity and/or sanction non-compliance. Finally, many state-owned companies
enjoy a monopolistic position (e.g. in the area of railway transportation or electricity sup-
ply), enabling them to pass the costs of their inefficiency on to consumers. To that degree,
public monopolies display similar pathologies to private monopolies (Bovens et al. 2001a).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Alternative Schemes
While the NATO scheme represents the most widely used policy classification by gover-
nance principles (see Table 2.3), many similar approaches have emerged which focus in more
detail on certain characteristics. For example, the categorization of Schneider and Ingram
(1990: 514–22) sheds light more explicitly on the behavioural assumptions underlying a
particular policy type. Their typology is based on the following categories:

• Authority tools involving statements that permit, prohibit or require action under des-
ignated circumstances.
• Incentive tools inducing compliance via negative or positive tangible payoffs.
• Capacity tools involving information, training and resources to enable individuals,
groups or agencies to carry out expected activities.
• Symbolic and hortatory tools based on the assumption that people are motivated to take
policy-related actions on the basis of their beliefs and values.

Table 2.3 Policy classification by governance principles.

Governance
principle Nodality Authority Treasure Organization
Basic Information Law Money Structures and
resource capacity
Governance Indirect stimulation Direct Indirect stimulation Provision of public
logic of behavioural change prescription of of behavioural good or service by
through information behavioural change through the state or public
and persuasion rules financial incentives enterprise
Typical Information Prohibitions Taxes Public companies
instruments campaigns Bans User charges
Suasion Permits Grants
Research inquiries Standards Tax deductions

Source: Based on Hood and Margetts (2007).

• earning tools promoting the drawing of lessons from experience by the target popula-
L
tion and/or the policy-makers. This represents an open-ended process, which might be
appropriate when a social problem is recognized but not well understood.

A different way of thinking about classifying policies is put forward by Jordan et al. (2015b).
Recognizing existing classification attempts (see, e.g., Lascoumes and Le Galès 2007;
Howlett 2009a), the authors contend that there has been a bias in the debate towards policy
tools and instruments adopted for the implementation of policy goals that have been decided
upon, which correspond to the examples given above. Therefore, the authors invite policy
analysts to pay more attention to policy tools and instruments that seek to influence the
formulation of public policy. The toolbox of policy formulation includes tools for forecasting
(for example, by developing scenarios), recommending policy options (for example, by using
cost–benefit analysis) and for conceptualizing and framing policy problems (for example,
citizens’ forums or focus groups). In the same volume, Howlett et al. (2015b) give a useful
overview of policy formulation tools as used by the Canadian government, which shows that
Copyright © 2020. Macmillan Education UK. All rights reserved.

brainstorming was the most popular and the organization of focus groups the least popular
tool for conceptualizing policy problems. In sum, Jordan et al. (2015b) conclude that this
perspective can facilitate a better understanding of the policy stage of policy formulation
(see Chapter 6), by also offering insights into how scientists, policy consultants and think
tanks influence the process of policy tool selection.

Key points
•• Policy typologies vary with the respective analytical criteria used for their classification, such as the
effects of public policies on policy-making or the governance principles which motivate them.
•• Typologies of policy effects have in common that they base their expectations about whether
the policy-making process will be adversarial or consensual on the characteristics of the policy
measure in question.
•• Typologies of governance principles refer to specific groups of policy instruments and shed light
on how they might bring about changes in the behaviour of the target group.

POLICY DIMENSIONS

An alternative way of practically specifying public policies is provided by differentiating

between their dimensions. On the one hand, we can distinguish between policy dimensions
that vary with the stage of development of a policy. On the other hand, policy dimensions
refer to the different contents of a policy. These two concepts complement the previous
approaches to the extent that they do not aim to provide guidance for grouping public policies
that belong to different policy domains, but to highlight different levels of analysis that apply
equally to all policy types. This contributes an additional layer to the nature of public policies.

Policy Dimensions in the Sequence of the Policy Cycle

While it is acknowledged that the different stages of the policy cycle overlap and that the
scheme is of more heuristic than analytical value, it does serve as a good starting point for
identifying different policy dimensions in the sequential development of a policy. In this
regard, there are three dimensions that are usually employed when making reference to a
public policy and that involve different analytical concepts closely related to the individual
stages of the policy cycle:

• Policy outputs
• Policy outcomes
• Policy impacts
Policy outputs are the direct result of the decision-making process, which usually involves
the adoption of a certain programme, law or regulation. They are defined by the content of
a public policy, as it is fixed in legal or administrative documents, and can encompass both
substantive and procedural aspects (Howlett 2017). Substantive outputs, for instance, refer
to the specification of claims, prohibitions, bans or service levels. Examples are emission
limits for environmental pollutants, minimum requirements for obtaining transport
licences, authorization conditions for constructing industrial plants or registration require-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ments for the production and distribution of chemicals. In short, substantive policy outputs
directly allocate advantages or disadvantages (i.e. costs and benefits) to people. Procedural
aspects, in contrast, specify rules guiding the interaction between the implementing author-
ities and the target groups. They define, for instance, the rules governing private access to
public documents or the participation of private actors in the implementation process. They
may also define deadlines to be observed or possibilities for appeal proceedings. Hence, pro-
cedural policy outputs define how something is going to be done or who is going to take
action.
Policy outcomes are closely related to the stages of policy implementation and evaluation
(see Chapters 7 and 8). Here the focus is on the way policies induce behavioural change on
the side of the targeted actors. How do policy addressees respond to a given policy? Do they
alter their previous behaviour in line with the objectives of a public policy? For example, a
policy that increases taxes on cigarettes affects, and has as its two main target groups, smok-
ers and those who might consider taking up smoking. The outcome of this policy refers to
changes in smoking behaviour (i.e. lower consumption) because of the increased taxes on
cigarettes. Another example is the prescription of stricter emission control technologies for

industry. The outcome of this policy should be that the regulatees modernize their produc-
tion facilities in order to comply with the requirements of the policy.
Policy impacts focus on the extent to which a policy decision and its subsequent imple-
mentation have actually brought about the expected results, indicating that they are mainly
assessed at the evaluation stage (see Chapter 8). Did the introduction of emission limits for
environmental pollutants actually lead to environmental quality improvements? Did the
liberalization of the telecommunications sector actually increase competition? Did the
privatization of public enterprises actually increase their profitability? What matters most
in this context is that the focus lies not only on the changes in the regulatees’ behaviour, but
on the implications of these changes for achieving the predefined goals of a particular public
policy. For example, the political goal of a policy to increase taxes on cigarettes is to reduce
smoking-related diseases. In a similar vein, a policy imposing the need to modernize
pollution-control equipment effectively aims to improve air quality.
As the concepts of policy outputs, policy outcomes and policy impacts refer to different
stages of the policy cycle, they must also be analytically approached by adopting different
theoretical lenses and research methods.

Policy Content Dimensions

Even if we restrict our focus to the analysis of policy outputs or the content of a policy, we
are confronted with broad lists of elements that can be distinguished. Dolowitz and Marsh
(1996: 349–50), for instance, differentiate between ‘policy goals, structure and content,
policy instruments or administrative techniques, institutions, ideology, ideas, attitudes and
concepts’. A widely accepted typology has been suggested by Hall (1993), who distinguishes
between three components of policy outputs: (1) policy paradigms (i.e. the overarching goal
that guides a policy in a particular field), (2) policy instruments (i.e. the means used to
achieve these goals) and (3) the precise setting or calibration of these instruments.
Hall developed this typology in his analysis of economic policy change in Britain. In this
context, so-called first-order changes refer to constellations in which only the calibration of
policy instruments is adjusted (instrument settings). Examples of such patterns are the
tightening or weakening of environmental emission standards or the increase or decrease of
Copyright © 2020. Macmillan Education UK. All rights reserved.

unemployment benefits. Changes in instrument settings typically imply only minor adjust-
ments to existing policies, as they do not entail changes either in policy goals or in policy
instruments that are already in place. If a second-order change occurs the type of instru-
ment is affected, implying that new instruments are introduced or existing instruments are
abolished or replaced by others. An example is the shift from command-and-control regula-
tion (emission standards, limit values) towards market-oriented instruments (emission
trading, eco taxes) that can partially be observed in many European countries (Holzinger
et al. 2006; Knill and Liefferink 2007).
The most fundamental category of change – i.e. a third-order change – takes place with
regard to policy paradigms. These refer to the goals guiding a policy in a particular field,
including how the problem at hand should be understood, i.e. in terms of its basic interpre-
tations and perceptions (Hall 1993: 279; Campbell 1998). British environmental policy, for
instance, has traditionally been characterized by a paradigm according to which the environ-
ment is considered to be well capable of absorbing a certain amount of dangerous sub-
stances. The perception was that there were no harmful pollutants, but only harmful
concentrations of pollutants. Accordingly, policy solutions were selected that entailed a

highly flexible approach allowing the regulatory bodies to define regulatory requirements in
the light of given pollutant concentrations, scientific evidence about their detrimental
impact and the state of environmental degradation at the local level. This paradigm, how-
ever, was fundamentally challenged during the 1990s by European Union (EU) policies that
followed the German paradigm of the precautionary principle. According to this approach,
pollutant emissions have to be reduced as far as technologically possible, even if there is no
scientific proof of their detrimental impact (Knill and Liefferink 2007: 78–9; Tosun 2013b).
While this distinction was originally intended to provide an explanation for different forms
of policy change, Hall’s typology has also become an important analytical tool for assessing the
question of whether it is possible to identify typical policy patterns across countries or policy
sectors. In this regard, the scholarly debate has predominantly focused on the dimension of
policy instruments. To what extent can we observe typical instrument choices (such as com-
mand-and control approaches or market-based instruments) in individual countries? To what
extent are such choices, by contrast, affected by the specifically perceived nature of the policy
problems in a given policy sector? These are just some of the questions exemplifying the rele-
vance attached to the choice of policy instruments in the literature.
The discussion of the role of policy instruments dates back to the work of Dahl and
Lindblom (1953: 6), which recognized that the capacity of modern societies to solve prob-
lems crucially depends on the policy instruments chosen. Schneider and Ingram (1990:
522–5) elaborated a theoretical approach to choosing an instrument. They argued that the
characteristics of the policy process (e.g. partisanship) and the extent to which a political
system is dominated by elites may have an impact on the policy instruments chosen.
Furthermore, they show that different historical periods often show a bias towards particu-
lar instruments because of the underlying rationale in government action.
Similarly, Lascoumes and Le Galès (2007: 10–11) argue that policy instruments are not
purely technical, but that they tend to produce original and sometimes unexpected effects.
In this regard, they stress three main effects of instrument choice. First, the instrument
chosen creates inertial effects, resulting in a resistance to outside pressures such as con-
flicts of interest between the actors involved in the policy-making process. Second, the
instrument chosen produces a specific representation of the issue it is handling. Third, the
instrument leads to a particular problematization of the issue. In short, all these
Copyright © 2020. Macmillan Education UK. All rights reserved.

approaches emphasize that an insightful study of policy instrumentation should entail

considerations related to social relations and the balance of power between actors (see
Kassim and Le Galès 2010: 5).

Key points
•• Policy outputs, outcomes and impacts are three categories used in public policy analysis.
•• Policy outputs refer to the content of a public policy and relate directly to the policy-making
process. Policy outputs can be analysed by examining the underlying paradigms, the specific
policy instruments and instrument settings.
•• Policy outcomes are the effects of a public policy in terms of changes in the behaviour of the
target groups. Studies of policy implementation and evaluation are often interested in policy
outcomes.
•• Policy impacts are the effects of a public policy in terms of problem resolution. Policy impacts
represent the central research interest of evaluation studies.

POLICY STYLES
A third prominent approach in the policy-analytical literature to understanding the nature of
public policies refers to policy styles. The concept of policy styles refers to the ‘standard operat-
ing procedures’ of governments in the making and implementing of public policies (Richardson
et al. 1982: 2; Richardson 2018a). While the term ‘policy style’ has been used as the most
general and widespread to identify such process patterns, some authors (in particular Vogel
1986; Vogel and Kagan 2004) refer to similar phenomena as styles of regulation. While styles
of regulation hence concentrate on a specific policy type, the concept of administrative styles
(Knill 1998, 2001; Knill et al. 2019) explicitly focuses on traditional behavioural patterns of a
specific player in the political administrative system, namely the public administration.
The concept of national policy styles was first introduced by Gustafsson and Richardson
(1979), Richardson et al. (1982) and Richardson and Jordan (1983). According to Richardson
et al. (1982), there are two dimensions determining national policy styles. The first dimen-
sion relates to a government’s approach to problem-solving, ranging from anticipatory/
active to reactive. The second dimension is about a government’s relationship to other actors
in the policy-making and implementing process, characterized by their inclination either to
reach consensus with organized groups or to impose decisions on them. Based on these two
axes, four ideal typical policy styles for (Western) European countries are identified: (1) the
German rationalist consensus style, (2) the British negotiation style, (3) the French concert-
ing style and (4) the Dutch negotiation and conflict style.
The empirical testing of the concept of national policy styles, however, has not provided
much support. Most importantly, the comparative volume edited by Richardson (1982)
revealed that there is much more similarity in policy-making styles than anticipated. This
induced Richardson to conclude that there is indeed a common (Western) European policy
style (see also Cram and Richardson 2004). More precisely, the case studies showed that there
was a common demise of anticipatory problem-solving and a trend towards more consensus-
oriented policy-making. The rise of a less hierarchical and consensual style of policy-making
was also confirmed by the comparative volume edited by Bovens et al. (2001a). Furthermore,
this compilation underscores variations in policy styles within countries, which also concurs
with more recent empirical studies. Cairney’s (2009: 671) analysis of the policy style regarding
Copyright © 2020. Macmillan Education UK. All rights reserved.

mental health policy in the United Kingdom suggests that there is ‘more than one picture of
British styles’. Jordan and Cairney (2013) argue that the British policy style, on average, cor-
responds to a consensual governance model. Recent empirical work on the British policy style
by Richardson (2018a, 2018b) challenges this finding and highlights that the policy style has
shifted back towards the traditional Westminster model of top-down policy-making as a con-
sequence of the public vote to leave the EU (Brexit) in June 2016.
In an attempt to transport the concept to countries beyond Western Europe, the volume
edited by Howlett and Tosun (2018) argues that a policy style is best thought of as a set of
political and administrative routines and behaviours that are in turn heavily influenced by
the rules and structures of the civil service and political system in which it is situated.
Defined in this manner, the concept becomes applicable to representative democracies (for
example, Korea), participatory democracies (for example, Brazil), closed-centralist systems
(for example, Vietnam) and competitive autocracies (for example, Togo). Similar to the orig-
inal volume by Richardson (1982), this collection shows that in many countries policy styles
vary across sectors. However, some countries stand out in terms of how they make policies.
For example, as Qian (2018) shows, policy-making in China is characterized by an

‘experimentalist’ style (see Box 9.1). Since the 1980s, some local governments have consid-
erable discretion to implement pilot reforms for local economic and social development.
After sufficient experiences from local experiments are accumulated, the central govern-
ment decides on whether to adopt the policy in question at the central level. This institu-
tionalized use of policy experimentation is a unique feature of policy-making in China.
A focus on sectoral rather than national factors was also advocated by scholars such as
Freeman (1985) and Rüdig (1987). Some scholars have argued that alternative typologies are
needed. The one put forward by van Waarden (1992), for example, concentrates on the degree
to which public–private interactions are formalized and whether societal interest groups par-
ticipate in the formulation and implementation of public policies. The combination of high
formalization (i.e. strong state) and low participation (i.e. weak strength of societal interests)
yields an étatist policy style (i.e. a state-centred model with top-down policy-making and imple-
mentation), as could be observed in France. Low formalization and low participation opportu-
nities, by contrast, result in a pluralist policy style, such as associated with the United States.
High formalization and high participation options – as provided in Austria, the Netherlands
and Sweden – favour social corporatist or meso-corporatist policy patterns (with the latter involv-
ing a more long-term and very strong involvement of state actors as compared with corporat-
ism), while low formalization and high participation opportunities correspond to clientelism as
well as liberal corporatism (see, for example, Switzerland).
To advance the debate on the concept of policy styles as put forward by Richardson et al.
(1982), we suggest that a promising approach includes both national characteristics and
characteristics of policy sectors. Depending on the specific practical constellation of these
factors, policy styles might reflect more either national or sectoral peculiarities. In this way,
it is conceivable that empirical findings will provide evidence for both country- and sector-
specific influences, which is empirically supported by the country case studies presented by
Howlett and Tosun (2018). These factors can also be divided into those that are relatively
volatile and those that remain rather stable over time. This approach enables us to account
for the variation and change of policy styles across countries, sectors and time (see Table 2.4).
With regard to national factors, the socio-economic development of a country might influ-
ence patterns of policy-making. We can expect more conflictive and adversarial patterns the
less developed a country is. This mainly stems from the restrictions in the resources that can
Copyright © 2020. Macmillan Education UK. All rights reserved.

be (re)distributed by means of public policy. The extent to which this structural factor affects

Table 2.4 Potential determinants of policy styles.

High stability Low stability

Country • Socio-economic development status • Economic situation
specific • Cultural orientations • Public opinion
• Institutional arrangements • Government coalition/government
• General relationship between state change
and society • Current interactions between public
and private actors
Sector • Nature of problem • Pressure of problem
specific • Policy paradigms • Current experience
• Policy legacies and path • Current policy developments
dependencies • Current conflicts and bargaining
• Typical cleavages and conflict processes

patterns of policy-making might also be influenced by the current economic situation. In addi-
tion, dominant cultural orientations (e.g. with regard to accepted patterns of governmental
intervention and relationships between state and society) might exert an influence on policy
styles. These orientations stem from the state and legal tradition of a country (Dyson 1980;
Knill 2001), but can also be affected by current developments in public attitudes and opinions.
Patterns of national policy-making are moreover strongly affected by institutional arrange-
ments, which define the strategic opportunities and constraints that public and private actors
face during the formulation and implementation of public policies.
These structures, for instance, strongly affect the extent to which policy styles reflect
more consensual or more adversarial patterns (Lijphart 1994, 1999/2012). Notwithstanding
the stability of these structural aspects, short-term developments, in particular changes in
government, might bring about changes in the strategic opportunities of the involved
actors. Finally, institutionally entrenched patterns of state–society relationships, such as
more corporatist or pluralist patterns, leave their mark on national policy styles (van
Waarden 1992, 1995). Again, the effects of these structures might vary because of context-
specific interactions between public and private actors.
Turning to sector-specific factors, the nature of the policy problem, as well as short-
term changes in the pressure of the problem, can have an impact on policy styles. For
instance, it makes a difference whether the problem refers to the regulation of risks (such
as nuclear plants or genetically modified organisms), of markets, or of social and environ-
mental issues. Depending on the problem type, different actors and conflicts of interest
might be involved, implying different policy styles. At the same time, dominant policy
paradigms – i.e. dominant perceptions of a problem and ideas of how the problem can be
solved – influence which actors might have more or less influence in the policy process.
Previous policy developments (policy legacies) can create path dependencies (the continued
use of a policy based on historical use; see Fioretos et al. 2016), as changes in established policy
patterns can induce high costs (both economically and in terms of institutional adjustment
needs). For example, Myles and Pierson (2001) demonstrate how much old-age pension systems
are affected by path-dependent processes. Pensions in most countries correspond to a pay-as-
you-go system: current workers pay contributions that finance the previous generation’s retire-
ment. Pay-as-you-go systems are highly resistant to radical reform since they generate unfunded
Copyright © 2020. Macmillan Education UK. All rights reserved.

pension commitments for future retirees. Shifting to private pension arrangements would place
an untenable burden on current workers, as they would have to finance the previous generation’s
retirement while simultaneously saving for their own. Thus, the best predictor of privatization
outcomes of pension systems is the size of the unfunded pension commitments already in place.
Finally, of course, policy-type cleavages will have an important impact on observable policy
styles. This is the argument developed by Lowi (1964, 1972), who stated that distinctive policy
types are characterized by distinctive process patterns.

Key points
•• The concept of policy styles seeks to identify stable country- or sector-specific patterns of
policy-making and implementation.
•• Empirical studies of policy styles reach contradictory conclusions about the nature and stability
of policy styles.
•• A promising avenue for explaining policy-making and implementation concentrates on national
and sectoral characteristics.

CONCLUSIONS
In this chapter, we have sought to specify the nature of public policies in the real world by
presenting the main policy-analytical approaches to classifying them. The most popular
way of structuring the many different types of public policies is to use typologies. We have
distinguished between typologies that concentrate on policy effects and those that rely on
governance principles. A second approach to explaining public policies focuses on differ-
ent analytical dimensions, which bring the differences between policy outputs, policy out-
comes and policy impacts to the fore. This approach also corresponds to Hall’s (1993)
classification of policy paradigms, policy instruments and instrument settings. The con-
cept of policy dimensions emphasizes the importance of the individual stages of the pol-
icy cycle that analyses public policies, including whether or not they should be studied
from a more abstract or more concrete perspective. This perspective does not attempt to
generate groups of public policies according to whether they are able to make predictions
as to how they affect policy-making or the behaviour of the addressees. Instead, it seeks
to make intelligible the notion that, in order to explore effectively the nature of public
policies, one must take into account the different aspects constituting the individual pol-
icy measures. This also provides a promising way of being able to make generalizable
descriptive or causal statements. Thus, we can now state that focusing on specific policy
dimensions can facilitate comparative analysis.
In the third and final step, we illustrated the research on policy styles, which again
combines considerations about public policies with stable processes of policy-making and
implementation. This perspective further corroborated the fact that it is very difficult to
assess the nature of public policies without the politics dimension. As a result, we have
drawn the overall conclusion that the classic policy-analytical literature has perhaps ironi-
cally more to say about the nature of politics than public policies per se. At the same time,
however, this indicates that the study of public policy cannot be successful without includ-
ing the politics and the polity dimensions. In what follows we will attempt to refine these
approaches.
Copyright © 2020. Macmillan Education UK. All rights reserved.

WEB LINKS

www.apsc.gov.au. The Australian government explains various policy instruments

and discusses their respective strengths and weaknesses.
www.bi.team. This is the website of the Behavioural Insights Team of the British
government.
blogs.lse.ac.uk/usappblog. This blog seeks to increase the public understanding of
social science in the context of American politics and policy-making.
www.policypointers.org. At this website various think tanks provide access to
research and conclusions from their work.
www.who.int. This is the most comprehensive information resource regarding public
health; it gives a good overview of the different governance principles and instru-
ments used in this policy field.

FURTHER READING

Adolino, J. and Blake, C. (2011). Comparing Public Policies: Issues and Choices in Six
Industrialized Countries. Washington, DC: CQ Press. This book provides a concise
overview of decision-making in various national contexts in the fields of immi-
gration, fiscal, health care, social, education and environmental policy.
Balla, S. J., Lodge, M. and Page, E. C. (eds.) (2015). The Oxford Handbook of Classics in
Public Policy and Administration. Oxford: Oxford University Press. This collection
offers a broad overview of the central theories in public policy and is helpful if
readers are looking for a theoretical framework for their own research.
Dunlop, C. A., Radaelli, C. M. and Trein, P. (eds.) (2018). Learning in Public Policy:
Analysis, Modes and Outcomes. London: Palgrave. Drawing from a wealth of case
studies, this collection offers an insightful treatise on how policy actors learn in
different settings.
Howlett, M. and Tosun, J. (eds.) (2018). Policy Styles and Policy-Making: Exploring
the Linkages. London: Routledge. This volume offers a generalized conceptualiza-
tion of policy styles and applies it to democracies and transformation countries
around the world.
Richardson, J. (2018). British Policy-Making and the Need for a Post-Brexit Policy Style.
London: Palgrave. A very timely update on the concept of policy styles during a
phase when the country was preparing to leave the EU.
Thaler, R. H. and Sunstein, C. R. (2009). Nudge: Improving Decisions about Health,
Wealth, and Happiness. London: Penguin. This is the high-impact book that
started new thinking on the behavioural foundations of policy instruments.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Reader’s guide
Policy choices are affected by both the polity (i.e. the institutional arrangements char-
acterizing a political system) and the politics (i.e. the policy-making process). This
chapter provides an overview of the most central institutions and actors participating
in the policy-making process and paves the way for the analytical approaches pre-
sented later. Our understanding of institutions corresponds to rules that shape the
choices individuals make in different institutional settings. We define actors as
(groups of) individuals who participate in policy processes and whose preferences will
ultimately determine the policy choice. The most central institution of any political
system is represented by the constitution and – where this exists – the constitutional
court protecting the fundamental principles of government it defines. This is followed
by the horizontal division of power between the executive, the legislature and the
judiciary, as well as the vertical division of power determining whether a state is uni-
tary or federal. This group of fundamental institutions is complemented by the elec-
toral and party system. In addition to national institutions, supranational and
intergovernmental institutions are increasingly influential in domestic policy-mak-
ing. Taking this into account, we outline the main characteristics of two such intergov-
ernmental organizations: the United Nations (UN) and the World Trade Organization
(WTO). We then shift the focus from institutions to the key policy-making actors,
including the executive, the legislature, the judiciary, bureaucracies, political parties,
Copyright © 2020. Macmillan Education UK. All rights reserved.

interest groups and experts.

INTRODUCTION
Public policy is determined by many factors. Amongst the most crucial are a country’s polity
and its politics. But before we shed light in subsequent chapters on how polity and politics
affect policy, we first provide a characterization of the main policy-relevant institutions and
actors. Of the numerous existing definitions of institutions, we will draw on the influential
work of Douglass C. North (1990, 1999), who posits three elements of institutions: formal
rules, informal constraints (norms of behaviour and conventions) and the enforcement char-
acteristics of formal rules and informal constraints. Concerning the third element, enforce-
ment is about establishing identity (to the greatest possible degree) between the objectives
of the institutional constraints and the choices individuals make in that institutional set-
ting. While we acknowledge the importance of informal rules, in this chapter we concentrate
on institutions defined as formal rules, which also corresponds to organizations.

33
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:20:28.
34 Public Policy

We illustrate this general, organization-based context of policy-making before we present

(in Chapter 4) theoretical arguments that also build on wider definitions of institutions. As
we will show in the subsequent chapter, one type of new institutionalism – i.e. sociological
institutionalism – conceives of institutions in terms of both formal and informal rules (see,
e.g., Hall and Taylor 1996; Peters 2019).
We conceive of actors as individuals, corporations or other collective entities (organiza-
tions) who possess policy preferences and the desire to realize them through their participa-
tion in the policy-making process (see Scharpf 1997a: 43). In our understanding, institutions
and actors are two different things and fulfil different functions in political systems.
Institutions determine the extent to which actors can transform their preferences into pub-
lic policy (Mayntz and Scharpf 1995; Ingram and Clary 2000; Tsebelis 2002; König et al.
2010). For example, if a policy-maker for whatever reason wishes to promulgate an employ-
ment policy that gives younger workers preference over older workers, most constitutions
would effectively prevent such a proposal from becoming law, as it would mean discriminat-
ing against some groups for the sake of others.
However, distinguishing between actors and institutions is not always straightforward.
Indeed, the actors we present here are predominantly collective actors (e.g. the executive),
which are at the same time formal organizations, i.e. they are constituted through institu-
tional rules (Jackson 2010). For example, the government as institution is defined by its formal
powers, how it is composed, which procedures exist for its internal decision-making, etc. The
government as collective actor is composed of persons who have preferences regarding the policy
area they are responsible for and which they express in front of the other actors (mostly the
legislature) in order to turn their preferences into public policy. Hence, when referring to a
government as an actor, we are stressing that its role for policy-making stems from its prefer-
ences and not its internal organization. At the same time, it is the government’s internal orga-
nization – the government as institution – that will determine to what extent the members of
the government will be able to realize their policy preferences.
What is the role of institutions and what is the role of actors in making public policy?
These questions guide this chapter.
Copyright © 2020. Macmillan Education UK. All rights reserved.

ATIONAL INSTITUTIONS: DEFINING THE RULES

N
OF THE POLITICAL GAME
This section introduces the most essential institutions of polities, which have important
repercussions for the process and output of policy-making. Over the last four decades, insti-
tutions at the international level have become more influential. We will discuss these and
outline how they may affect the public policies in their respective member states.

Constitutions and Constitutional Courts

The most essential institution in a political system is its constitution, which is a set of fun-
damental principles and formal rules according to which a state is governed. In some polities
(e.g. New Zealand and the United Kingdom), the constitutions are flexible, meaning that
they can be changed by regular majorities in the legislative body. In other polities (e.g.
Australia, Canada and the United States), constitutions are rigid and any changes must be

approved by super-majorities (i.e. two-thirds or more; Lijphart 1999/2012). The reason for
having rigid constitutions is that they cannot easily be modified to deal with short-term
political conflicts. High thresholds for change – quorums – ensure that politicians are bound
by their own decisions, which should impede governmental encroachment on the rights of
individuals and ensure that the provisions of the constitution do not become modified in
order to serve the interests of particular governments (for instance by changing the rules for
re-election).
Put simply, constitutions define the most basic rules of the game in any political system by
structuring and restricting the exercise of government power (see Figure 3.1). In this regard,
the constitution determines the centralization of power, that is whether a state is organized in
accordance with unitarism (i.e. sovereignty is concentrated at the level of a single central gov-
ernment) or federalism (i.e. sovereignty is shared across several levels of government). For
example, the Australian Constitution establishes a federal system of government, in which
powers are distributed between a national government (i.e. the Commonwealth) and the six
states (i.e. New South Wales, Queensland, South Australia, Tasmania, Victoria and Western
Australia). The Australian Constitution also defines that three territories (i.e. the Australian
Capital Territory, the Northern Territory and Norfolk Island) have self- government
arrangements.
Likewise, constitutions establish the three branches of government and assign them for-
mal rights to make, implement and interpret laws. And yet, constitutions do not only out-
line how a state is organized. They are also crucial for determining procedural rules that are
indispensable for making a political system work. For example, constitutions ‘lay down leg-
islative procedures; and they tell us how legislative authority is constituted (through elec-
tions, for example), and what the legislature can do (through enumerating powers)’ (Stone
Sweet 2011: 164). Beyond this, constitutions generally limit the exercise of government
power in the name of individual rights, e.g. by prohibiting some acts of discrimination. In
addition, constitutions may explicitly define public rights (Shane 2008: 193–4). Article 20 of
the German Basic Law (the equivalent of a constitution) defines that Germany is a ‘social
federal state’. This ‘social-state principle’ compels policy-makers at each level of government
to protect the social welfare of the German people.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Constitution

Vertical
Horizontal
power separation
power separation
National
Executive level
Legislative
Subnational
Judiciary
level

Electoral and party system

Figure 3.1 Stylized institutional setup of a polity

In most polities, the constitution is protected by a supreme or constitutional court. In

some countries, however, constitutional courts are missing and ordinary courts are in charge
of fulfilling this task, such as in Denmark. Detecting constitutional violations occurs through
judicial review, which mainly occurs in three forms (Stone Sweet 2011: 167–8): abstract
reviews, concrete reviews and constitutional complaints. Abstract reviews are made before
legislation enters into force (or in some systems after promulgation but before application).
The purpose of an abstract review is to check that a legislative bill is compatible with the con-
stitution. In most cases they are initiated by other government institutions, such as a parlia-
mentary minority. Concrete reviews, by contrast, are carried out in accordance with
procedures for ordinary litigation and are hence initiated by ordinary judges, who activate the
process by sending a constitutional question to the constitutional court. The presiding judge
will initiate a concrete review if the constitutional question is decisive for determining who
wins or loses the case and if there are reasonable doubts about the constitutionality of the
legal act in question. The third form, the constitutional complaint, can be directly activated by
individuals who can prove that their constitutionally guaranteed rights have been violated by
a public authority, usually after judicial remedies have been exhausted or are not available.

Division of Powers
The functioning of modern states is determined by an old principle that was laid down by
the French political philosopher Charles de Montesquieu in his famous book On the Spirit of
Laws (Cohler et al. 1989). In this book, Montesquieu argues that the best government would
be one in which power was balanced amongst three groups of officials. By taking eighteenth-
century England as the benchmark, the three branches of government were identified as the
executive, the legislature and the judiciary. Often, these three together are referred to as the
horizontal division of powers of the state. Of these branches, the executive and the legisla-
ture can be conceived of as the principal actors in the process of policy-making, although as
we have seen the judiciary can in some countries veto public policy in cases where it violates
constitutional provisions.
The executive branch is, in principle, in charge of implementing public policy and has the
authority to administer the bureaucracy. It corresponds to what people tend to call ‘the
Copyright © 2020. Macmillan Education UK. All rights reserved.

government’. In real-life politics, however, governments are not only in charge of imple-
menting public policies, but are also actively involved in their formulation (see Chapter 6).
There are three forms of democratic government: parliamentary, presidential and semi-
presidential. In parliamentary systems, the head of government – who can have many dif-
ferent official titles, such as prime minister or chancellor – and his or her cabinet (i.e. a
collective executive) are derived from and are responsible to the legislature. In a presiden-
tial system, the president as the head of government (i.e. a non-collective executive) is
directly or indirectly elected by the citizenry and cannot, or can only on very demanding
conditions, be removed from office by the legislature. In a semi-presidential system, the
government must respond both to the legislative assembly and to an elected president.
Typically, these systems are characterized by a president who is elected for a fixed term with
some executive powers and a government that serves at the discretion of the legislature.
While we have presented the established types of political systems, Ganghof et al. (2018)
identify Australia as a specific type described as a semi-parliamentary system, which repre-
sents a hybrid form of a parliamentary and a presidential system. Semi-parliamentary sys-
tems are characterized by voters selecting two fully equal legislative agents, and therefore

these systems have, thus the authors, a specific potential to balance competing modes of
representation.
The legislature (often referred to as ‘parliament’) is the branch of government endowed
with the competency to make legislation. Legislative assemblies have many different names
(e.g. House of Commons, Diet, Bundestag, Senate, Congress). In most systems (one excep-
tion would be Estonia) they consist of two houses: a lower and an upper. Generally, the lower
house is elected directly and is more influential in terms of policy-making than the upper
house. There is, however, considerable variation across countries in the degree of power
exercised by the different houses (see Lijphart 1999/2012; Kreppel 2017; Box 3.1).
The judiciary is composed of the various levels of courts (generally with a supreme or
constitutional court as the court of final appeal) that interpret and apply the law and
resolve disputes emerging amongst private and public actors. The judiciary is indepen-
dent of the other branches of government. In a similar vein, judges should only be

Box 3.1 Composition of the legislature in South Africa

The parliament of South Africa is composed of two houses: the national assembly and
the national council of provinces. The national assembly is the lower house and con-
sists of 400 members. The members of the national assembly are directly elected,
which is in line with the procedure of most lower houses in parliamentary systems.
The number of seats that a political party has in the national assembly is in proportion
to the number of votes it received in an election. The national council of provinces is
the upper house and consists of 90 provincial delegates, that is, 10 delegates for each
of the nine provinces of the republic to ensure that they are all equally represented.
The purpose of the national council of provinces is to represent the governments of
provinces, rather than directly representing the people – this is a feature that can be
observed in many other (federal) states as well. Therefore, election to the national
council of provinces is indirect: citizens vote for provincial legislatures, and each leg-
islature then nominates a delegation of ten members.
Copyright © 2020. Macmillan Education UK. All rights reserved.

The Constitution of the Republic of South Africa sets out the following functions
for the national assembly:
• Selection and election of the president.
• Provision of a nationwide forum for the public consideration of issues.
• Passage of legislation.
• Oversight and control of executive action.
• Ensuring the accountability of the members of cabinet to parliament for the exer-
cise of their powers and the performance of their functions.
The national council of provinces is involved in legislation by means of proposing,
considering, amending, passing and rejecting legislative bills. However, it may not
initiate or prepare money bills.
In addition, the South African parliament is in charge of promoting the values of
human dignity, equality, non-racialism, non-sexism, the supremacy of the constitu-
tion, universal adult suffrage and a multiparty system of democratic government.

influenced by the law and the facts of a case. As the judiciary can affect policy-making
through its sentences, in many countries the selection of judges for supreme or consti-
tutional courts is a political process. In the United States, for instance, judges are nomi-
nated by the president, who then places his or her ideological stamp on the courts.
Whether the candidate is eventually appointed depends on the US senate (O’Connor
and Sabato 2009: 350).
In some political systems, the horizontal division of powers is complemented by a verti-
cal division of power between the state as a whole and its geographically defined constituent
units. Such political systems are known as federal polities. The core idea of federalism is to
help prevent the abuse of power, a considerable share of which is exercised at regional levels
(e.g. in states, provinces or regions). In such systems, certain powers are exercised by the
federal or general government (shared rule) and other powers by the regional governments
of the constituent states (self-rule; Hooghe et al. 2008; Marks et al. 2008). In marked con-
trast, in unitary states, such as Estonia (see Tatar 2011), there is only one level of
government.
Generally, the United States is regarded as the prototype of a federal state, but Australia,
Belgium, Canada, Germany and Switzerland – to name but a few – are also federal systems.
To be precise, the United States represents one specific model of federalism, i.e. coordi-
nated federalism, where the federal government and regional governments are financially
and politically independent from one another within their own spheres of responsibility.
The complementary model is cooperative federalism, in which the different levels of gov-
ernment interact cooperatively, as is the case in Germany (see Keating 2012; Benz and
Broschek 2013).
It should be noted that unitary states can also experience decentralization, i.e. a transfer
of decision-making powers from the federal government to subnational units, which subse-
quently display a number of ‘federal’ characteristics (see Hooghe et al. 2008). In 1998, the
United Kingdom, for instance, launched devolution reforms to transfer powers in areas like
education and health to Scotland, Wales and Northern Ireland (see, e.g., Keating et al. 2009;
Cairney 2012). From this, it follows that decentralization does not necessarily correspond
with federalism. Likewise, a unitary state is not synonymous with a centralized state, i.e. a
concentration of decision-making powers in the hands of government institutions. Rather,
Copyright © 2020. Macmillan Education UK. All rights reserved.

both federal and unitary states can be more or less centralized or decentralized (Lijphart
1999/2012). In order to able to judge whether a state is federal or unitary, the division of
labour between the national and subnational levels of government – as defined in the con-
stitution – is the main criterion.
Federalism certainly represents an effective means of mutual control between differ-
ent elements of government and may produce desirable public policies, as decision-mak-
ers at the subnational level should have a better understanding of regional problems
than in unitary states. Moreover, the variety of different political arenas in federal sys-
tems might favour policy innovation and learning (see, e.g., Benz 2012; Keating and
Cairney 2012). Policies that turned out to be successful in one subnational unit are
often emulated by other units – a phenomenon that has been extensively studied in the
US context (see Gray 1973; Volden 2006) as well as in Australia (see Painter 1991;
Chappell 2001; Hollander and Patapan 2007) and Canada (see Poel 1976; Nykiforuk
et al. 2008; for comparative assessments, see Hueglin and Fenna 2006; Gamkhar and

Vickers 2010). By the same token, however, the two levels need to seek compromises
which can prevent or at least delay (necessary) policy reforms. Problems emerge in par-
ticular in constellations in which decisions at the federal level require the consent of the
subnational units. This situation often entails the persistence of ineffective policies as
long as only one state objects to policy change. Finally, there is an economic aspect to
federalism and decentralization: maintaining individual parliaments, governments and
administrations at various levels of government is expensive.
More generally, there are hints in the literature that unitary states tend to change policies
to a greater extent and more rapidly. Treisman (2000), for instance, shows that unitary
states as compared with federal states often change monetary policies more swiftly and radi-
cally, moving quickly from high to low inflation and interest rates and vice versa. This find-
ing is not very surprising considering that in unitary states the number of actors whose
consent is needed to enable a policy to pass (also known as veto points or veto players) is
lower than in federal states, which is likely to increase the speed of policy-making as well as
the ‘radicalism’ of the measures proposed, which can be good or bad depending on the issue
concerned and one’s point of view.

Electoral Institutions and Party Systems

Elections are one of the most central features of democracies, as they legitimize policy-
makers and determine who is going to have the greatest political influence for a certain
period of time. Electoral institutions vary from one country to another. They include provi-
sions about who can be elected (e.g. age, residency requirements), specifications of election
types (e.g. general elections, regional elections, presidential elections), the time that may
elapse between elections, and issues of suffrage and compulsory voting. Another impor-
tant aspect is how votes are transformed into parliamentary seats, which is understood as
the electoral system. There are numerous ways in which electoral systems can be catego-
rized. Here we follow Gallagher and Mitchell (2008) and Gallagher (2017), who categorize
electoral systems in accordance with the magnitude of the constituencies (i.e. electoral dis-
tricts) in which seats are allocated to candidates and, amongst other things, their propor-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tionality. The two most basic types are majoritarian systems and systems of proportional
representation.
Majoritarian systems are usually based on single-member constituencies, in which the
strongest party in each constituency wins the seat. This is a very straightforward way of vot-
ing: only one candidate can be chosen, and the one who receives the highest number of votes
is awarded the seat. In some polities, e.g. Australia, this system is in use in a slightly modi-
fied version to fill a single seat as voters can rank-order the different candidates, with their
ordering being taken into account when there is no clear overall winner. Another modifica-
tion (used in French elections, for example) is the two-round system: if no candidate wins a
majority of the votes cast, the two top-ranked candidates participate in a second election
round; whoever wins this is declared the winner.
Proportional representation systems are characterized by multimember constituencies, in
which the seats are shared amongst parties in proportion to the share of votes they have

received. The basic idea is to give each party the same share of seats as it won votes, which can
be achieved in various ways. In some countries (e.g. the Netherlands) the entire country is
regarded as one constituency. Despite the high level of proportionality guaranteed by this
system, there is no possibility of choosing local members of parliament – this does, however,
favour extremist parties. Therefore, in many countries (e.g. Brazil) there are multiple con-
stituencies within which seats are awarded proportionally. Another refinement is the use of
so-called tiers: seats are first assigned in a lower tier (i.e. individual constituencies) and then
in a higher tier (i.e. the whole country) to avoid disproportional representation (e.g. Denmark,
Germany, New Zealand, Sweden and the Scottish parliament; see Gallagher 2017).
Electoral institutions are considered to be an important explanatory variable for a
wide range of political phenomena. Typically, they have been used to explain party
systems. In this regard, Duverger’s (1954) very influential book argues that majoritar-
ian electoral systems favour two-party systems, whereas proportional representation
leads to multiparty systems (a rare counter-example is provided by India; see Diwakar
2007).
Conversely, the party system has important repercussions on policy-making. Party sys-
tems are a concept developed by scholars of comparative political science for categorizing
enduring institutional characteristics of the party landscape (see Blondel 1990). More pre-
cisely, they describe the system of interaction between multiple parties that are engaged in
competition (Sartori 2005: 39). This research perspective has primarily focused on charac-
terizing party systems and determining changes to them over time. Following Caramani
(2020b), we distinguish four types of party system:

• Dominant-party system
• Two-party system
• Multiparty system
• Bipolar system

The first type, dominant-party systems, characterizes a situation in which one very large
party (with an absolute majority of well above 50% of parliamentary seats) dominates all
others over long periods. This is not to say that there is no party competition in these sys-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tems, just that no other party has received enough votes to come close to 50%. As a conse-
quence, the dominant party holds a hegemonic position and does not need to enter into a
coalition with other parties to form a government. A well-known case of a dominant-party
system was Mexico, where the Institutional Revolutionary Party was the hegemonic party
from the mid-1930s to the late 1990s (see Lehoucq et al. 2008).
Two-party systems are characterized by two equally strong parties (with vote shares
of 35–45% each) which dominate the party system and alternate in holding power. In
these systems even a small number of votes changing from one party to the other can
lead to a change of majority, which makes alternations in power fairly frequent. As a
result, such systems tend to be very competitive. Since both parties have high vote
shares, the winning one is likely to hold the absolute majority of seats and therefore be
able to form a single-party government. Yet, there are also often other, smaller parties
(i.e. with a very low share of votes) in these systems. According to Caramani (2020b),
only the United States provides a perfect example of this party system, which is essen-
tially composed of the Democrats and the Republicans (for an overview, see O’Connor

and Sabato 2009). Another example would be the United Kingdom, with the Conservative
Party and the Labour Party as the dominant political forces.
Multiparty systems are the most frequent type of party system, although there is
remarkable variation in the number and size of the parties. Relatively small multiparty
systems (up to five parties) can be found in Canada, Ireland, Japan (especially in recent
times) and Norway. Party systems with more than five parties in parliament exist in
Belgium, the Netherlands and Switzerland. Regardless of the actual number of parties,
what multiparty systems have in common is that none of the parties is majoritarian (i.e.
holds 50% of the votes or seats), leading to the necessity of forming coalitions. Usually,
all of these parties run individually in elections; governmental coalitions are negotiated
after the elections. Due to the particular election dynamics and the need to form coali-
tions, government change rarely takes place through electoral change. Instead, govern-
ment change is likely to occur by swaps of coalition partners.
The final type is the bipolar party system, which combines elements of multi- and two-
party systems. As in multiparty systems there are many parties, of which none is majoritar-
ian, making coalition governments the rule. The major difference, however, is that parties
form relatively stable electoral alliances. In most systems, there are two large electoral coali-
tions running in elections and alternating in power, making electoral competition look like
a two-party system. An appropriate example of a bipolar system is Germany, where in usual
circumstances two electoral alliances oppose each other. On the one side, there are the Social
Democrats and the Green Party; on the other, there is the coalition formed by the Christian
Democratic Union and its Bavarian sister party the Christian Social Union as well as the
Liberal Party. ‘Grand coalitions’ consisting of the Christian Democratic Union, the Christian
Social Union and the Social Democrats are only formed if there is no other option for form-
ing a government at the national level (see, e.g., Bolgherini and Grotz 2010).
Policy-making can be expected to vary across different party systems. Dominant-party
systems should be able to respond immediately to emerging policy problems, whereas the
policy-making process can be expected to take longer in multiparty systems due to the
need to receive the consent of the other parties for a policy proposal. As in two-party and
bipolar systems, policy-making might occur relatively swiftly, but a change in the majority
situation could entail drastic changes in existing policy arrangements (for an overview,
Copyright © 2020. Macmillan Education UK. All rights reserved.

see, e.g., Martin and Vanberg 2005, 2011).

Key points
•• Constitutions define the most fundamental principles of government in a political system. They are
often protected by special courts through means of abstract judicial review, concrete judicial review
and constitutional complaints.
•• Horizontal division of power refers to the three ‘classic’ branches of government: the executive, the
legislature and the judiciary. In some countries, various levels of government are responsible for
decision-making. In such cases, the horizontal division of power is complemented by a vertical one.
•• Electoral systems determine how votes are transformed into seats in legislative assemblies.
These can generally be assigned into two categories: majoritarian systems and proportional
representation systems.
•• Electoral systems have repercussions on party systems, which can be categorized accordingly:
dominant-party systems, two-party systems, multiparty systems and bipolar systems. These, in
turn, are expected to have an impact on policy-making.

INTERGOVERNMENTAL ORGANIZATIONS: GLOBAL PUBLIC POLICY

Organizations operating in an international context are classified as intergovernmental

organizations. A special form of intergovernmental organization is a supranational organi-
zation, which has powers that its member states do not have because they have delegated a
limited amount of their sovereignty to it. In this way, the supranational organization may
enact legislation that pre-empts the laws and regulations of its member states (see Mitchell
and Sheargold 2009: 5). In other words, supranational organizations can adopt public poli-
cies that are legally binding for their member states. To date, there is only one intergovern-
mental organization that corresponds mostly – yet not fully – to these criteria, namely the
European Union (EU; for an overview, see Hix and Høyland 2011).
An intergovernmental organization also consists of member states, but they remain
independent and engage in voluntary cooperation and coordination. The crucial point is
that in this type of organization the member states do not surrender any sovereignty to it.
There are many examples of intergovernmental organizations, such as the North Atlantic
Treaty Organization (NATO), the International Union for Conservation of Nature, the
International Maritime Organization, the International Atomic Energy Agency, the
Organisation for Economic Co-operation and Development (OECD) and the Organization of
Petroleum Exporting Countries, to name just a few (see Chapter 10).
Here we will limit our discussion to two intergovernmental organizations: the United
Nations (UN) and the World Trade Organization (WTO). The UN is a multifunctional orga-
nization with an impressive breadth of tasks and scope of membership. The WTO – together
with the World Bank and the International Monetary Fund (IMF), also known as the Bretton
Woods institutions – plays an essential role in promoting free trade. At the same time, both
of these organizations possess formal legal powers. The UN can take collective security
action, whereas the WTO possesses an instrument of formal legal suasion over its member
states by means of its dispute settlement mechanism (Higgott 2008: 616). However, in con-
trast to a supranational organization, the WTO does not have the right to adopt rules that
are superior to national law. These rules have no direct effect on domestic actors, unless they
are formally transposed into national law.
Copyright © 2020. Macmillan Education UK. All rights reserved.

The United Nations System

The UN was founded in 1945 to maintain international peace and security, develop friendly
relations amongst nations and promote social progress and human rights. Since it com-
prises a large number of organizations and carries out many different functions, one usually
speaks of the UN system. The organization works on many different issues, including disas-
ter relief, refugee protection, sustainable development, the promotion of democracy, gender
equality and the advancement of women, economic development, international health,
clearing landmines and expanding food production.
At the centre of this system is the General Assembly, which is the main deliberative
organ of the UN and is composed of representatives of all its 193 member states. The
Secretariat supports the activities of the General Assembly and the other UN organiza-
tions by carrying out various tasks. The Economic and Social Council is the principal organ
that coordinates the economic, social and related work of the UN and its numerous spe-
cialized agencies and institutions, such as the Food and Agriculture Organization (FAO),

the World Health Organization (WHO), the World Bank Group and the IMF. The
International Court of Justice is the chief judicial organ. It settles legal disputes between
states and gives advisory opinions to the UN and its specialized agencies. Another impor-
tant institution is the Security Council: this has primary responsibility for the mainte-
nance of international peace and security (for an overview, see Hanhimaki 2008; Weiss
and Daws 2008).
The UN is usually associated with conflict prevention and peace missions as well as
humanitarian assistance (see Thakur 2006). There are many other ways in which the UN
may affect public policy. This mainly happens through the UN’s specialized agencies or funds
and programmes such as the United Nations Environment Programme (UNEP) or the
United Nations Children’s Fund (UNICEF). These organizations can affect public policy in
the UN member states through data gathering and the dissemination of information, the
definition of benchmarks and good practice, and financial or other forms of aid. With regard
to peace-keeping and state-building, the UN, of course, has a much more far-reaching
impact, as it can alter the polity of countries and thus profoundly modify ways of policy-
making. More generally, however, the UN’s impact on national public policies varies from
one UN organization to another as well as across issue areas. Therefore, it is not easy to
make general statements about its various possible impacts.
Under usual conditions, i.e. if there is no armed conflict or a humanitarian or natural
disaster occurring in a country, the numerous UN conventions play a decisive role in shaping
public policy in its member states. A particularly well-known instance of this is the
Convention on Climate Change, which came into force in 1994 (see, e.g., Keohane and Victor
2011). This sets an overall framework for intergovernmental efforts to tackle climate change.
Under the Convention, participating governments are expected to:

• Gather and share information on greenhouse gas emissions, relevant national policies
and best practice for reducing emissions
• Develop strategies for lowering greenhouse gas emissions and adapting to expected
impacts
• Cooperate in preparing for adaptation to the various impacts of climate change
Copyright © 2020. Macmillan Education UK. All rights reserved.

There are many more UN conventions which target, for example, corruption, children’s
rights, the law of the sea and the rights of persons with disabilities. Signing up to these
instruments of international law clearly impacts on national policies that are affected by the
issues targeted by the respective conventions and can entail significant constraints to
policy-making. For example, membership of the UN Convention against Torture and Other
Cruel, Inhuman or Degrading Treatment or Punishment requires states to take effective
measures to prevent torture within their borders. Additionally, it forbids states from return-
ing foreign residents or asylum seekers to their home country if there is reason to believe
that they will be tortured there.

The World Trade Organization

One of the most important international institutions is the Geneva-based WTO, which is the
successor to the General Agreement on Tariffs and Trade (GATT; Martin 2008). The WTO cur-
rently has 164 member states and in this way directly affects well above 90% of total global

trade. Its main objective is to reduce or completely eliminate trade barriers in order to facilitate
free trade. To this end, it negotiates agreements with its member states, and these define
legally binding rules for international trade. The organization is governed by a ministerial con-
ference, which meets every two years, and a general council that is in charge of implementing
the ministerial conference’s policy decisions and is responsible for day-to-day administration.
Additionally, the WTO has a director general, who is appointed by the ministerial conference.
The monitoring of the national trade policies of its member states is the WTO’s most
fundamental activity. Central to this task is the trade policy review mechanism. All
WTO members are subject to review, with the frequency depending on the country’s
size. Furthermore, the WTO asks its member states to notify it of new or modified trade
measures. For example, details of any new anti-dumping legislation, technical standards
affecting trade and changes to regulations which impact on commercial services all have
to be communicated to the appropriate body of the WTO. As a consequence of the con-
tinuous reviewing process and the obligation to communicate relevant modifications of
trade policy, the WTO does indeed have the potential to affect domestic policy
arrangements.
Even more important is the impact of the WTO’s dispute settlement mechanisms.
These help to resolve trade disputes between member states, which usually involve
questions about the legality of a particular trade policy or practice (see Grando 2010;
Josling 2010). Disputes start with a request for consultations in which the member
government bringing the case to the WTO (i.e. the complainant) explains why it objects
to the trade policy of the other member state (i.e. the defendant).
The rules determine that the complainant and defendant must engage themselves in
consultation activities for 60 days to try to find a solution that satisfies them both and
ends the dispute. If these efforts fail, the complainant can request a panel proceeding, i.e.
enter formal litigation. The panels vary in their composition and typically comprise two
rounds of testimony. At this point, the complainant and defendant can still negotiate a
settlement. If this does not happen, the panel issues a final report.
This leads to three possible scenarios. The first is that the final report will be adopted by
the WTO. The second is that both the complainant and defendant agree not to adopt the
report for whatever reason. The third and most likely scenario is, however, that the com-
Copyright © 2020. Macmillan Education UK. All rights reserved.

plainant or the defendant or both appeal against the panel’s report. These appeals are con-
sidered by the appellate body, i.e. a standing body of jurists, which then arrives at a final
decision. The body scrutinizes the case again by hearing testimony and checking ways in
which the panel might have arrived at the wrong conclusion. It can either confirm the
entirety or parts of the panel’s report or overturn it. If the body decides that the measure in
question is inconsistent with the terms of the relevant WTO agreement, the member state
concerned will be asked to bring the measure into conformity with that agreement. Mitchell
and Sheargold (2009: 18) illustrate this well by making reference to the Australia – Subsidies
Provided to Producers and Exporters of Automotive Leather dispute. The WTO panel ruled that
a subsidy, which had been provided to a private company, had to be repaid in full to the
Australian government. Remarkably, this decision had to be implemented even though no
domestic legal basis existed to force the company to repay the subsidy, since under domestic
law it had already been legally granted.

Key points
•• Supranational organizations may enact rules that pre-empt the laws and regulations of their
member states. To date, there is only one international organization that corresponds to these
criteria, namely the EU.
•• Intergovernmental organizations do not require their member states to surrender any sover-
eignty; there exist many different types of this kind of organization.
•• The UN system is the most encompassing intergovernmental organization in terms of its
organizational structure as well as its tasks.
•• The WTO can equally affect the trade policies of its member states. An effective way of doing
this is by reviewing the member states’ trade policies and the WTO’s dispute settlement
mechanism.
•• Both the UN and the WTO impose limitations on domestic actors’ room for manoeuvre, thus
affecting national policy choices.

KEY ACTORS
Actors are individuals, collectives or corporations involved in the policy-making process,
who seek to turn their preferences into public policy. While there can also be influential
individual actors, policy-making is usually characterized by collective and corporate actors.
Collective actors such as social movements are composed of individuals coordinating their
actions; they are dependent on and guided by the preferences of their participants. Corporate
actors are also composed of participants, but they ‘are typically “top-down” organizations
under the control of an “owner” or of a hierarchical leadership representing the owners or
beneficiaries’ (Scharpf 1997a: 56; see also Engel 2010). In the literature, firms are often
conceived as corporate actors.
For understanding how actors might determine policy outputs, there are three character-
istics to which we must pay attention: their capabilities, perceptions and preferences
(Scharpf 1997a: 43). First, actors are characterized by their capabilities, i.e. the action
Copyright © 2020. Macmillan Education UK. All rights reserved.

resources at their disposition to influence the policy-making process in a way to make one
specific outcome more likely than another. For example, firms and companies are widely
regarded to have high capabilities with regard to influencing policy decisions. As Walker and
Rea (2014) show, in the United States corporate actors use different avenues to affect public
policy. They provide financial support to electoral campaigns, they engage in business cam-
paigns in civil society, they carry out collective action with the support of associations and
they lobby policy-makers directly.
Second, actors can be characterized by the way they perceive a particular social problem.
For example, citizens of different ideological orientations (as measured in terms of a left–
right scale or party support) diverge in their perceptions of reality. Research has shown that
supporters of Republicans and Democrats cannot agree on their perception of macro-
economic indicators, and that political parties are trying to manipulate the very perception
of policy issues (see Bisgaard and Slothuus 2018). Depending on their respective, ideology-
filtered perceptions, citizens will review potential solutions to the present problem and
decide which ones they find most desirable.
Third, actors are characterized by their specific preferences, i.e. the choice they make
between alternatives. To fight climate change, one actor prefers a carbon tax over an

emissions trading scheme if she would rather choose the carbon tax than the emissions trad-
ing scheme. Preferences depend on the feasible set of alternatives and whether they are
either stable over time or changeable.
In addition, actors can be distinguished according to whether they are public (acting on
behalf of the state) or private (acting on behalf of their own preferences).

Public Actors
In representative democracies, government is based on elected individuals representing citi-
zens. These elected officials, in turn, appoint additional actors who equally represent the
state and are involved in various aspects of policy-making. This section gives an overview of
these public actors.

The Executive
When we speak of the executive as an actor, we are referring to the people who compose it.
In parliamentary and semi-presidential systems, the executive consists of the head of gov-
ernment (prime minister, chancellor) and the ministers who form the cabinet. The number
and denomination of ministers vary from country to country, as does the extent to which
the head of government dominates the other cabinet members. The relationship between
the individual cabinet members and the cabinet as a whole can take two distinct forms. One
is the ‘principle of collective responsibility’ that characterizes Westminster systems. This
means that all cabinet members must publicly support decisions made in cabinet, regardless
of whether they privately agree with them or not. The other form is represented by the ‘prin-
ciple of ministerial autonomy’, which is, for instance, applied in Germany. It gives each cabi-
net member the freedom and responsibility to supervise departmental operations and
prepare legislative proposals without cabinet interference.
As a rule, cabinets must be formed in such a way that they can be supported or at least
tolerated by a parliamentary majority – the latter can be observed quite frequently in
Scandinavian countries and is known as a ‘minority government’ (see Hansen 2017). Against
this backdrop, there are two ways in which the cabinet can be composed: a one-party govern-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ment or a coalition cabinet. In the first case, one party alone holds the absolute majority in
parliament and it is also this party to which the head of government and the cabinet mem-
bers belong. In the second case, parties must form a coalition in order to enjoy the support
of the parliamentary majority. Such coalition cabinets can take three different forms. First,
they can form a minimal winning cabinet, implying that the cabinet will include just as
many parties as are necessary for gaining control of a majority of parliamentary seats.
Second, a cabinet can be ‘oversized’, insofar as it includes more parties than is necessary to
provide the government with majority support in the legislature. Third, a cabinet can include
fewer parties than is necessary for majority support in the legislature. Such minority cabi-
nets need to be tolerated by the parliament (Lijphart 1999/2012). It is clear that a coalition
cabinet’s leeway in proposing legislation increases with the lowering of the number of par-
ties participating in it. At the same time, however, forming an oversized cabinet can be use-
ful to ensure that controversial policies will be passed in the legislature.
In presidential systems, by contrast, the executive is not drawn from the legislative
majority and only needs its support for the president’s legislative proposals. Within the

executive domain the president is sovereign and decides autonomously the extent to which
he or she will rely on the cabinet. Some US presidents, for instance, used their cabinet mem-
bers either as advisers or for carrying out their orders, but they were never regarded as equal
participants (Müller 2011: 145). Therefore, in presidential systems the whole policy-making
process is to a much stronger degree characterized by the president’s preferences.

The Legislature
Legislatures are important actors in policy-making and fulfil three principal tasks (Kreppel
2017). First, legislatures provide legitimacy for the political system. In this context, it is
important that the members of the legislature facilitate communication between citizens
and government. Strongly related to the communication function is representation: citizens
expect that their preferences are expressed by ‘their’ members of the legislature. The repre-
sentation of these preferences primarily occurs through debating, since legislatures are
forums for deliberation (Carey 2008: 431).
Second, legislatures have control and oversight functions. Their control rights mainly
refer to control over the executive. These tend to be relatively limited in presidential sys-
tems, as the executive is neither drawn from nor responsible to the legislature. Also in these
systems, the policy agenda of the executive is not subject to legislative control. In parlia-
mentary and semi-presidential systems, by contrast, executives are responsible to the legis-
lature for their policy agenda. There is, however, no difference between the regime types
regarding the right of legislatures to oversee policy: they possess a variety of tools for moni-
toring the executive. Committee hearings, hearings in plenary sittings, investigative com-
mittees, question times, special inquiries and hearings, interpellations, ombudsmen and
the preparation of reports on specific topics are amongst the most common tools. They can
be used during policy formulation or after the government has enacted a policy and the
legislature wants to check whether that policy has been implemented correctly (see Pelizzo
and Stapenhurst 2008: 9–13). Another option is the indirect overseeing of the executive by
controlling the budgetary process.
The third group of activities refers to what legislatures should, in theory at least, prin-
cipally do, i.e. legislate. In most political systems, however, it is the executive which tends
Copyright © 2020. Macmillan Education UK. All rights reserved.

to dominate the policy-making process (see Chapter 6). Even so, there are still various
ways in which legislatures can participate in policy-making. The weakest way to influence
public policy is through consultation; i.e. presenting an opinion about a legislative pro-
posal introduced by the executive branch. Another option is to delay the passage of a
policy proposal, at times by using the power of a formal veto. In most cases, this will lead
to bargaining between the executive and the legislature. While delaying and blocking pol-
icy-making represent a ‘negative’ power, legislatures also possess the ‘positive’ power of
initiating and amending policy proposals. The extent to which this positive power is pos-
sible very much depends on the political system that is being studied (Kreppel 2017).
Generally, however, legislatures tend to be more actively involved in policy-making in
presidential systems than in the other two regime types, namely semi-presidential and
parliamentary systems.
When conceiving of the legislature as an actor, it is important to note that these functions
are not equally carried out by the entirety of the legislature. In parliamentary and semi-
presidential systems, the majority of the legislature and the executive form an entity, which

drastically reduces the chances of policy-related conflict between them. In marked contrast,
in presidential systems there is a clear-cut separation between the president and the con-
gressional majority. As a result, the likelihood of policy-related conflict tends to be higher.

The Judiciary
The judiciary is the third branch of government and mainly refers to constitutional courts.
For a long time the judiciary was not considered to play an important role in policy-making.
This seems to have changed, as other political actors have been affected by the policy-
modifying power of the judiciary (Adolino and Blake 2011: 71). In this sense, abstract judi-
cial review, for instance, has increasingly been used by parliamentary minorities for strategic
reasons (Vanberg 1998). In such cases, the parliamentary minority, also known as the ‘oppo-
sition’, can ask the constitutional court to annul a constitutionally dubious bill introduced
or passed in the legislature by the governing party. A resultant negative decision could force
the government and/or governing majority in the legislature to modify its policy.
However, abstract review is not the only way in which constitutional courts can affect
public policy: they can become involved in policy-making by declaring a certain piece of
legislation to be unconstitutional. Yet, as Stone Sweet (2007: 87–9) explains, there are two
ways in which constitutional judges exercise indirect authority over the legislature. The
first way is ‘self-limitation’ and refers to the exercise of self-restraint on the part of the
government and its parliamentary majority when it is seeking an annulment by the consti-
tutional court. This is observed when the government and the parliamentary majority take
decisions that sacrifice previously held policy objectives in order to lower the probability
that a bill will be either referred to the court or found unconstitutional. The second way,
‘corrective revision’, takes place after a bill has been annulled. It refers to the reworking
of a censured text in conformity with the court’s decision, so as to secure promulgation.
Constitutional court decisions therefore can have a profound impact on the content and
design of public policies. The most important policy-making power of the courts stems from
their authority to interpret and apply the constitution and other laws. In this way, they can
help to resolve disagreements about abstract goals enshrined in a country’s constitution,
such as the main characteristics of its economic regime (Dahl 1957: 280). Policy decisions can
Copyright © 2020. Macmillan Education UK. All rights reserved.

be subject to judicial review, and when they are found to violate constitutional rules decision-
makers need to start a new policy process, one involving the elaboration of another policy
design and the need to receive broad support for it. For example, the US Supreme Court in
2010 ended a policy banning self-declared homosexuals from serving in the military (known
as the ‘don’t ask, don’t tell’ policy) by declaring it unconstitutional. The previous policy had
prohibited any homosexual or bisexual person from disclosing his or her sexual orientation.
The judiciary can also affect public policies by acting as an agenda-setter (see Chapter 5).
If the key policy-makers, i.e. the legislature and the executive, neglect certain social prob-
lems, filing lawsuits can bring this problem onto the policy agenda and force policy-makers
to address it appropriately. Indeed, litigation has been successfully employed to reduce indus-
try practices that may be harmful to public health, such as smoking (Daynard et al. 2002).
There are also many indirect ways in which the judiciary can influence policy-making.
One way is by excluding certain actors from the policy-making process. Constitutional
courts can, for example, ban political parties. By excluding particular actors from the policy
process, their preferred policy proposals cannot enter the political arena, and so are unlikely
ever to materialize as public policies.

In this context, it should be noted also that the courts can become quite political. Hönnige
(2009) shows that judges in France and Germany have political preferences of their own and
that the ideological position of the court is determined by the selection of the pivotal judge.
So if a court is mainly staffed by supporters of the government, it is likely to favour the ideo-
logical preferences and policy proposals of the government rather than those of the parlia-
mentary minority that is assumed to have opposing preferences.

The Bureaucracy
For implementing public policy, the executive relies on the bureaucracy, i.e. the layers of
hierarchically appointed officials (Kettl 2008: 371). However, bureaucrats are not only
involved in policy implementation but also policy formulation due to their procedural and
specialist knowledge. Therefore, it is reasonable to distinguish between the implementing
bureaucracy and the ministerial bureaucracy, as it is the latter type that might play a role in
policy formulation. The important role of the bureaucracy has been particularly emphasized
by Max Weber (1947), who attributed to it three main characteristics that he deemed were
needed to make it an effective instrument of government:

• Personnel stability: attracting and maintaining qualified bureaucrats working in a govern-

mental organization, enabled by the provision of suitable career paths.
• Organization: organizing bureaucracies by their functional speciality, and assigning their
members to units based on the type of work they do. Bureaucracies are marked by hier-
archical authority, with higher-level officials supervising lower-level workers.
• Procedure: having a high degree of standardization of decision-making and the applica-
tion of rules, which are also achieved by management keeping written documents.

There are two ideal-typical systems of how bureaucrats are appointed. With the ‘spoils’ system,
individuals are appointed on the basis of their political loyalty, which they are expected to
demonstrate in all aspects of their work. This kind of clientelistic bureaucracy displays a low
level of autonomy from elected politicians and results in staff with reduced technical capaci-
ties. With the meritocratic system, bureaucrats are appointed because of their expertise and
Copyright © 2020. Macmillan Education UK. All rights reserved.

experience – and this form is now adopted in most advanced democracies. Meritocratic
bureaucracies are characterized by high degrees of autonomy and technical capacities. In addi-
tion to these two extreme forms, there are ‘administrative bureaucracies’ that have autonomy
but low capacity and ‘parallel bureaucracies’ that have high capacity but low autonomy
(Scartascini 2008: 64–5). Weber (1947) acknowledged the fact that bureaucracies – in light of
their characteristics – did not just constitute an instrument of the government, but could also
exert an autonomous influence on policy formulation and implementation. If and to what
extent this is still the case will be discussed in Chapters 6 and 7. In general, we find a broad
variance across countries regarding the political autonomy of the bureaucracy (Peters 2010).

Political Parties
So far, we have shed light on the elected and appointed officials involved in the process of
policy-making. However, the members of legislatures, executives and in some countries the
bureaucracy must be recruited from somewhere. This is one task of political parties, which, in

seeking to influence government policy, usually nominate their own candidates and try to seat
them in political office. According to Katz (2017), political parties perform four core tasks:

• Coordinating
• Conducting electoral campaigns
• Structuring competition
• Representing

Coordinating involves many different dimensions. Political parties are in charge of coordinat-
ing functions in party caucuses (e.g. maintaining discipline; coordination of action of the par-
liamentary caucus in support of or opposition to the cabinet), organizing the political activity
of citizens and linking them with elected officials. Another important field of action is elec-
toral competition. Political parties provide candidates, formulate policy programmes and
organize electoral campaigns. This aspect is strongly linked with the recruitment of person-
nel, which parties do by selecting candidates for elections and/or appointed offices. Finally,
parties represent their members within government institutions as well as in society.
More directly with respect to policy-making, the role of political parties is to influence
the ideas and beliefs of citizens about public policy and so affect their electoral decisions,
which themselves then define the strength of a given party in the legislature and/or the
executive. However, since elections in democracies take place on a regular basis, the policies
adopted by the political parties will shape the electoral decisions of citizens (Aldrich 2008:
571; see also Gilardi 2010). The policy options that parties offer citizens vary as they are
strongly affected by their fundamental beliefs. These differing ideological views can be used
to categorize parties into party families (von Beyme 1985). This notion has since been fur-
ther developed. For example, Mair and Mudde (1998) identify the following party families,
which have certain very general policy orientations (see Bale 2011):

• Communist parties have their ideological roots in Marxism–Leninism and seek to secure
a more equitable distribution of income and opportunity in society (see, e.g., Bull and
Heywood 2016).
• Social democratic parties or labour parties seek to align capitalism with social justice (see,
Copyright © 2020. Macmillan Education UK. All rights reserved.

e.g., Schulman 2015).

• Christian democratic parties apply Christian principles to public policy. They are often
considered conservative on social and moral issues, and moderately market-liberal on
fiscal and economic issues (see, e.g., Mainwaring 2003).
• Conservative parties are market-liberal without having a religious background (see, e.g.,
Bale 2011).
• Agrarian parties typically combine a commitment to decentralized policy-making and
environmental protection, as well as maintaining the interests of small businesses (see,
e.g., Tosun 2017).
• Liberal parties are a difficult category, as they can be centre left, centrist or centre right,
depending on the national party system, though they all strongly support civil rights
(see, e.g., Clarkson 2014).
• Left-libertarian parties are essentially liberal parties supporting some form of income
redistribution (see, e.g., Otjes and Krouwel 2015).
• Green parties foster the principles of environmental protection, social justice, reliance on
grassroots democracy and opposition to war (see, e.g., Spoon et al. 2014).

• Nationalist parties put the well-being of their ‘own’ people to the fore. Often their policy
agenda is defined by nationalist separation from a state in which they are presently sub-
sumed. Examples of these can be found in Northern Ireland, Scotland and Wales (see
Moran 2005: 336); the Basque region; and to an extent in Belgium.
• Right-wing populist parties are associated with a political ideology that rejects the existing polit-
ical consensus and combines economic liberalism and anti-elitism (Mudde and Kaltwasser
2012; Akkerman et al. 2016). Examples of this party group include the United Kingdom
Independence Party (which could also be assigned to the previous group of nationalist par-
ties), Geert Wilders' Party for Freedom in the Netherlands and the French National Rally.
• Left-wing populist parties, similar to right-wing populist parties, emphasize socio-
economic issues in maintaining that the elites do not pay enough attention to the needs
of common working people (March and Mudde 2005; Otjes and Krouwel 2015). Examples
include the Socialist Party in the Netherlands and the Syriza Party in Greece.
• Racist parties or extreme right-wing parties seek to attract votes by racist and anti-system
statements (see Ennser 2010).

Private Actors
Despite not being elected or appointed by elected officials, private actors are central to pol-
icy-making, on the grounds that they bring valuable information to policy-makers that
might eventually guide the way to solving social problems.

Interest Groups
Interest groups are organizations that make policy suggestions to governments in order to
bring public policies more in line with the interests of their members. In the literature they go
by numerous names, such as lobby groups, interest associations, pressure groups and non-
governmental organizations (NGOs). The study of interest groups concerns two major topics,
namely how interest groups organize and the ways in which they affect public policy. Regarding
organization, Olson’s (1965) famous study emphasized the problems typical of collective
Copyright © 2020. Macmillan Education UK. All rights reserved.

action. Based on the assumption that individuals are rational and possess perfect information,
he contended that joining an interest group is irrational, since individuals can enjoy the ben-
efits secured by participants in other interest groups, i.e. by ‘free riding’. As a consequence,
smaller groups face fewer problems of collective action, since they are susceptible to a higher
degree of social pressure that reduces the risk of free riding. Concerning the formation of large
interest groups that are particularly subject to free riding, Olson argues that only the provision
of ‘selective’ incentives – i.e. a benefit reserved strictly for group members – can stimulate
individuals in a latent group to act in a group-oriented way (see also Moe 1981).
Wilson (1974) identifies three reasons for individuals’ becoming members of interest
groups. The first is that they want to enjoy the benefits that are exclusively accessible to
members (selective incentives). For example, when joining an automobile club, members
benefit from first-hand legal and technical information and support. Wilson’s second reason
is that individuals join an interest group to pursue their specific goals. For example, firms
and companies join associations to be able to push collectively for their preferred policy
arrangements, such as lower corporate tax rates or longer transition periods for old tech-
nologies that need to be replaced by new ones. A third motivation is that membership can
be seen as a (political) statement, e.g. being against the use of nuclear energy.

We can distinguish between private and public interest groups, (though both are technically
private actors). Private interest groups are organizations that seek to affect policy-making in a
way that pursues the special – often economic – preferences of their members. Due to their
relatively small size, the homogeneity of their members and the concentration of potential
benefits from their activities, private interest groups are on average less likely to face free-
riding problems. These groups often come in various forms, representing, for instance, enter-
prises or professions (e.g. doctors, farmers or teachers). In most cases, however, they are
equated with business groups and trade unions (see Dye and Zeigler 2006: 235–9).
Public interest groups, by contrast, may be defined as organizations that seek to pursue
goals that are not exclusively limited to their members. They are usually non-profit organi-
zations and lobby for greater government regulation, for example to protect the environ-
ment, such as Greenpeace or Friends of the Earth. Like private interest groups, public ones
come in many forms. The areas in which they are particularly active include environmental
issues, peace, human rights, animal rights and gender equality. Generally, they are con-
fronted with larger organizational problems than private interest groups. The large size and
the prospect of diffuse benefits make these groups relatively vulnerable to free riding.
Scholars in EU policy-making have pointed out that the EU’s multilevel system offers
multiple points of access or opportunity structures to both private and public interest
groups (Princen and Keremas 2008). Some interest groups will have the capacity (i.e. fund-
ing and personnel; Eising 2007) to be active at the various levels of the EU political system
(Eising 2008), whereas those with limited capacity must choose at which institutional level
they want to participate (Varone et al. 2018a).
It is important to delineate public interest groups from social movements. Kriesi (2017)
defines social movements as (1) a group of individuals with a conflictual orientation towards
an opposing group, (2) with a collective identity and sharing common beliefs and objectives
and (3) with a repertoire of collective actions. What predominantly distinguishes a social
movement from an interest group is that the first is composed of a network of individuals
and organized groups of individuals who engage in non-institutionalized coordinated
actions to achieve common goals while maintaining their autonomy. Interest groups, by
contrast, are more formally constituted and mostly engage in institutionalized actions, as
they have regular access to different decision-making arenas.
Copyright © 2020. Macmillan Education UK. All rights reserved.

The scope of action of a social movement is also much greater. Social movements target, for
example, women’s rights, gay rights, peace, anti-nuclear and environmental policies. One of the
most famous social movements of the twentieth century is the American Civil Rights Movement,
which sought to end racial discrimination against African Americans. Interest groups and politi-
cal parties may be part of a social movement, though the latter cannot be reduced to interest
groups or political parties. A case in point is the (global) environmental movement, which
encompasses Green Parties and green interest groups such as Greenpeace (see Leonard 2007 on
the green movement in Ireland). In light of these characteristics, Kriesi (2017) argues that social
movements cannot be regarded as simply another type of interest group (Box 3.2).
How can interest groups affect policy-making? Following Erne (2017), there are four ways
in which interest groups and policy-makers interact. The first is known as inside lobbying,
which is based on personal contacts between interest groups and politicians, and involves
various strategies such as meetings and participation in legislative committees. As a rule,
inside lobbying tends to be more effective when interest groups possess generous resource
endowments. This strategy can be complemented by outside lobbying, namely activities of
interest groups taking place outside the actual policy-making arenas. Outside lobbying
includes media activities and campaigning in favour of or against a particular policy. Some

Box 3.2 World Wildlife Fund: an international interest group

The World Wildlife Fund (WWF) is a public interest group working towards attaining the
conservation of nature, founded in 1961 at the headquarters of the International Union
for the Conservation of Nature and Natural Resources (IUCN) in Switzerland. Today, the
organization is active in more than 80 countries. The WWF has adopted a strategy for
influencing conservation policy in the countries in which it is active that differentiates it
from other environmental groups such as Greenpeace. The WWF forms partnerships with
governments and business actors and seeks to influence public and company policies
through communication and dialogue. Supermarket chains, for instance, collaborate with
the WWF to demonstrate that they are committed to corporate social responsibility. The
WWF is willing to enter such arrangements since it can influence decision-making and
attempt to raise the environmental awareness of consumers (see, e.g., Kopnina 2016).

interest groups (predominantly labour and business associations) may try to establish a spe-
cific structure for political exchange with policy-makers. The idea of this perspective – which
we discuss in detail when introducing the concept of (neo-)corporatism in Chapter 4 – is that
interest groups must be given access to the policy-making process by politicians, who are will-
ing to ‘open doors’ on the condition that they receive something in exchange. Consequently,
not all kinds of interest groups have the same chances of affecting public policy; their success
depends on the resources they can exchange and itself is shaped by the institutional context.
These resources include, amongst other things, valuable policy-relevant information, i.e.
expert knowledge and a guarantee of achieving social consent over public policies, which is
expected to turn policy implementation into an uncontroversial issue (see Bouwen 2004).
The flipside to this relationship is that ‘regulatory capture’ might emerge, i.e. a situation in
which interest groups abuse their access to policy-making in order to shape policies in such a
way that they serve their private benefits (Stigler 1971). Recent research in comparative poli-
tics concentrates on exactly this possibility and compares the policy demands made by inter-
est groups with the policy supply eventually offered by policy-makers (see, e.g., Klüver 2013).
Copyright © 2020. Macmillan Education UK. All rights reserved.

The final form in which interest groups can become involved in policy-making has been
described by Streeck and Schmitter (1985) as ‘private interest government’. This describes a
situation in which private interest groups, primarily business representatives, make legally
binding decisions, thereby taking over a central function of the state (see Knill and Lehmkuhl
2002b: 50–1). According to Veldman and Willmott (2016), a case in point is the UK Code of
Corporate Governance, which defines standards of good practice in relation to board leader-
ship and effectiveness, remuneration, accountability and shareholder relations.

Experts
Not all individuals who can provide knowledge and expertise to policy-makers are members of
interest groups. We use the term expert for individuals or groups of individuals that can have
an impact on policy-making on the grounds of the information they supply to policy-makers.
In contrast to interest groups, experts do not necessarily pursue any specific policy goal, though
they are expected to deliver unbiased knowledge that is required for effective policy-making.
In most cases, experts can be understood as part of ‘epistemic communities’, i.e. net-
works of recognized specialists with policy-relevant knowledge in a particular issue area

(Haas 1992: 5). The International Panel of Climate Change can be regarded as such an epis-
temic community, for example.
However, following this classic definition, we can speak of an epistemic community only
where there is a structured network of experts, who share a common interest and task and
possess diverse knowledge – which explains why it is useful for them to collaborate. Some
examples of these communities include the World Futures Studies Federation or the Global
Facilitation Network for Security Sector Reform. One problem with the notion of epistemic
communities, however, is that it concentrates on knowledge elites who possess scientific
expertise. Yet there are many other kinds of knowledge-based actors, such as individual
scientists, consultants and practitioners, who can equally help to change policy-makers’
ideas about a certain issue (see Head 2008; Howlett 2009b). Ideas, in turn, can be under-
stood as causal beliefs produced by cognitive processes. They posit causal connections, thus
providing guides for action (Béland and Cox 2011: 3–4).
In this context, the role of political advisers is particularly interesting, since they have
become relevant in the formulation of public policy in the many Westminster systems. In
New Zealand, for example, a minister’s office will typically ‘include officials seconded from
departments or agencies for which he or she is responsible … a Senior Private Secretary
responsible for overall office administration, other administrative staff, at least one desig-
nated Press Secretary, and one or more “political” ministerial advisors’ (Eichbaum and Shaw
2007: 610–11). Usually, political advisers are partisan rather than neutral, which distin-
guishes them from experts.
The increasing importance of political advisers is sometimes viewed with suspicion (see
Maley 2000; Tiernan 2007). According to Eichbaum and Shaw (2008: 338), however, there
is too little empirical data available to make reliable statements about how influential politi-
cal advisers actually are. This does not rule out that there is, in principle, a political dimen-
sion to the provision of policy-related information and advice.

Key points
•• Numerous actors participate in policy-making; they can be categorized into public actors and
private actors.
Copyright © 2020. Macmillan Education UK. All rights reserved.

•• The most central public actors are the executive, the legislature, the judiciary and the bureau-
cracy. With the general exception of the judiciary and parts of the bureaucracy, these are mostly
recruited from political parties with different ideologies. In accordance with this, political par-
ties offer specific policy suggestions, which are subject to evaluation by the electorate.
•• Interest groups and experts are particularly influential types of private actors.

CONCLUSIONS

In this chapter, our main objective has been to familiarize the reader with the main institu-
tions and actors affecting policy-making. We defined institutions in a minimalist way as
rules determining the extent to which the policy preferences of actors can be transposed
into public policies. In shedding light on institutions, we did not limit ourselves to domestic
institutions only, but also outlined the functioning of the UN and the WTO as two influen-
tial intergovernmental organizations.
We defined actors as individuals or groups of individuals forming a collective entity that
participate in policy processes and whose preferences can ultimately determine policy

choice. We have shown that there are numerous actors participating in policy-making. Yet
we have restricted our presentation to the most essential ones who are involved at each of
the various phases of policy-making. Indeed, there are additional actors to which we will
turn in Chapter 4 when discussing the various theoretical concepts of policy-making. Most
importantly, however, we have not introduced citizens as a specific category of actor. This
stems from the fact that in representative democracies the direct involvement of citizens is
restricted to elections and in a few cases to referendums. As a result, citizens’ preferences
are mediated by the actors presented here.

WEB LINKS

www.comparativeconstitutionsproject.org. Comprehensive data about the world’s

constitutions and the constitution-making process.
ec.europa.eu/transparencyregister/public/homePage.do?locale=en. The transparency
register from the European Parliament and the European Commission shows
what interests are being pursued, by whom and with what budgets in the EU.
www.electionguide.org. This website provides information about upcoming and past
elections.
www.ipu.org. The Interparliamentary Union supplies information on the organiza-
tion of parliaments around the world.
www.ituc-csi.org. The website of the International Trade Union Confederation gives
a useful overview of the policy themes on which trade union position themselves.
www.lobbyfacts.eu. This website answers questions related to the characteristics of
lobbying in the EU.

FURTHER READING

Biela, J., Kaiser, A. and Hennl, A. (2013). Policy Making in Multilevel Systems:
Federalism, Decentralisation, and Performance in the OECD Countries. Colchester:
Copyright © 2020. Macmillan Education UK. All rights reserved.

ECPR Press. One of the few books that systematically connects federalism and
decentralization with policy-making.
Choudhary, S. K. (2019). The Changing Face of Parties and Party Systems: A Study of
Israel and India. London: Palgrave. A stimulating book that traces back the devel-
opment of parties and party systems as well as elaborates on coalition politics.
Hague, R., Harrop, M. and McCormick, J. (2019). Comparative Government and
Politics. London: Red Globe Press. A very established source that gives an instruc-
tive overview of the core concepts of comparative politics.
Lijphart, A. (2012). Patterns of Democracy: Government Forms and Performance in
Thirty-Six Countries. New Haven, CT: Yale University Press. The second edition of
one of the most influential books on comparing political institutions.
Peters, B. G. (2019). Institutional Theory in Political Science: The New Institutionalism.
Cheltenham: Edward Elgar. An instructive read that gives an excellent overview of
the varieties of institutionalism.
Petracca, M. P. (ed.) (2018). The Politics of Interests. Interest Groups Transformed.
New York: Routledge. A comprehensive volume with a strong section on the rela-
tionship between interest groups and political institutions.

Reader’s guide
The objective of this chapter is to provide a basic idea of the processes underlying
policy-making and to present different theoretical perspectives to explain why gov-
ernments institute, modify and terminate public policy. We differentiate between
structure-based models, institution-based models and interest-based models.
Structure-based models emphasize the possible policy consequences of enduring
conflicts and constellations in societies. Institution-based models outline how for-
mal and informal institutional arrangements may influence public policy. Thus, the
theoretical perspectives to be presented complement the description of formal insti-
tutions that we gave in Chapter 3. Interest-based models, by contrast, emphasize the
role of actors and their respective policy preferences. These three theoretical per-
spectives are complementary as they address different levels of analysis. Cleavage-
based models offer an explanation at the macrolevel, whereas interest-based models
are at the microlevel. Institution-based models can relate to both levels, although in
most cases they produce mesolevel explanations; i.e. they illuminate how macrolevel
pressures shape microlevel activity.

INTRODUCTION
Copyright © 2020. Macmillan Education UK. All rights reserved.

Policy-making is undoubtedly a complex process – often one factor that stimulates the
promulgation of a public policy in one political system does not necessarily lead to the same
outcome in another. For example, many countries are confronted with demographic changes.
In particular, the proportion of senior citizens is growing, creating increased demands for
certain public services, such as social care. Despite having common demographic changes,
countries have adopted different policies for addressing this problem. Germany and Korea,
for instance, have established long-term care insurance to cover the costs resulting from
elderly persons’ losing their ability to look after themselves (see, e.g., Kim et al. 2013).
Another approach, adopted in Scotland, is to introduce personal and nursing care that is free
at the point of use for older people in care homes and in their own homes. But there are also
many countries that have not yet adjusted their policy arrangements to deal with the ageing
population, and so non-governmental organizations (NGOs) and other private actors have
to step in. In Tanzania, for example, the Swiss NGO Kwa Wazee grants pensions to grand-
mothers that are funded through donations. And in other countries (e.g. Turkey; see Yilmaz

56
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:20:21.
Theoretical Approaches to Policy-Making 57

2015), public policy tackles this issue but in order to strengthen the role of families in giving
care to elderly people.
Why are some countries more aware of the need to reform their social care policies in
order to meet changes in demography than others? Why do the adopted policies diverge
across countries? To answer these and similar research questions theoretical considerations
are needed – some of which we have already introduced in the previous chapters, but to
which we will add in this one.
Key to explanatory success is a simplification of the various processes accompanying the
making of one particular policy. This requires leaving out any distracting details and focus-
ing on the essential features of an empirical phenomenon – and this is best attained by rely-
ing on theoretical approaches. Essentially, a theory is a set of statements comprising various
hypotheses, i.e. the expected causal relationship between a dependent variable (what we
want to explain) and an independent variable (the factor that explains something) that may
or may not be true. An example of a hypothesis is: if the Green Party is elected to govern (the
independent variable), then environmental policies (the dependent variable) will become
more stringent. In the field of public policy analysis there are several theoretical approaches
that can explain why some governments are more likely to introduce or modify a public
policy than others. It would become very difficult, if not to say impossible, to provide
answers to important research questions without the use of theories (Peters 2011: 40). As
we will show in what follows, there are a number of basic theoretical perspectives that help
to give us a simplified idea of what shapes policy-making.

STRUCTURE-BASED MODELS

Structure-based models draw attention to the most basic socio-economic problems present
in societies, which provide decision-makers with the incentive to create or modify public
policy. In this context, it is important to note that these models concern the structure of the
society or capitalism rather than the structure of problems. Here we introduce the social
cleavage approach put forward by Lipset and Rokkan (1967) and the Varieties of Capitalism
approach according to Hall and Soskice (2001). While the latter directly connects the role of
Copyright © 2020. Macmillan Education UK. All rights reserved.

the economy to policy-making, the cleavage approach, by contrast, outlines a broader range
of social problems that are related to policy-making by means of political parties that define
their policy positions in accordance with them. Essentially, it represents the foundation of
why different political parties exist, namely because they address a lasting division between
societal groups on policy problems. The cleavage approach hence represents a complement
to the party families presented in Chapter 3.

The Cleavage Approach

The cleavage approach draws on the idea that certain enduring socio-economic problems
exist in societies and affect policy choices by means of creating lasting divisions between
social groups (i.e. social cleavages) which possess different perceptions about these prob-
lems and the ways of solving them. These social cleavages are important for policy choices in
a country, as they have been taken up by political parties that offer policy proposals in accor-
dance with those problems that they deem most important.

There are four major social cleavages identified by Lipset and Rokkan (1967) that emerged
in the context of national revolutions in the nineteenth century and of industrialization in
the nineteenth and twentieth centuries, which created profound and lasting socio-economic
and cultural divisions amongst social groups. This already suggests that the cleavage
approach is a concept that emphasizes historical developments. The traditional approach
recognizes the following cleavages:

• Centre–periphery
• State–church
• Rural–urban
• Workers–employers

Caramani (2020b) identifies two ‘modern’ cleavages, which refer to the following
dimensions:

• Materialists–post-materialists
• Open–closed societies

Concerning the traditional cleavage approach, national revolutions created the division
between the centre and the periphery: a division that consists on the one hand of social
groups supporting the view that political power and administrative structures in a nation
state should become centralized and, on the other hand, of social groups which assert
their traditional autonomy against such centralizing pressures. Individuals attaching
importance to this particular cleavage will either support or oppose policies strengthen-
ing the centralization of the state, for instance by means of a centralized use of tax reve-
nues. This cleavage is mainly associated with separatist nationalist parties, which are
sometimes also described as ‘regionalist’ parties (see Massetti and Schakel 2013;
Mazzoleni and Mueller 2016).
The state–church cleavage is about social groups seeking to separate secular from reli-
gious authority and social groups in favour of preserving the connection between them.
Depending on whether they are religious or not, individuals will differ in their views with
Copyright © 2020. Macmillan Education UK. All rights reserved.

respect to various policy issues. The state–church cleavage has been found to be of particular
relevance for so-called morality policies that touch upon fundamental values and hence dis-
pose high affinity to religious doctrines. Intuitively, this state–church cleavage is of impor-
tance to Christian democratic parties and to a certain extent to conservative and social
democratic parties. For instance, the voting patterns on the regulation of pre-implementation
diagnostic in the German parliament show that the religiosity of members of parliament
matters (see Baumann et al. 2015; Engler and Dümig 2016; Euchner and Preidel 2018). In a
similar vein, Fink (2008) in his study on embryo research laws in Organisation for Economic
Co-operation and Development (OECD) countries finds that regulatory strictness increases
with the power of the Catholic Church in a country. Yet more systematic assessments of the
impact of the state–church cleavage in different areas of morality policies show that reli-
gious influence reflects a brake on rather than a barrier to the adoption of more liberal
arrangements. In other words, religion causes a delay effect in reform developments com-
pared with countries in which religious positions are less pronounced and powerful (Budde
et al. 2017a, 2017b).

The Industrial Revolution created two further structural divisions. First, there is the
cleavage between rural and urban societies, i.e. between social groups whose survival
depends on traditional activities and social groups that endeavour to remove traditional
constraints in order to foster new economic activities. The rural–urban cleavage yields far
more policy implications than one might suspect at first glance. With respect to trade poli-
cies, for example, the rural population usually favours trade barriers for protecting national
agricultural products, whereas the urban population tends to prefer trade liberalization to
enhance economic well-being. Usually, agrarian parties, such as the Latvian Farmers Union,
the Polish People’s Party or the Swedish Centre Party, are associated with this cleavage (see
Batory and Sitter 2004; Varley 2010; Arter 2017; Tosun 2017).
The second cleavage to come out of the Industrial Revolution concerns the opposition
between ‘capital’ and ‘labour’. On one side there are the employers who favour a low level of
state intervention in the economy, and on the other side there are the workers who demand
job security and social protection. This cleavage takes into account social changes as a con-
sequence of industrialization. In fact, this division between workers and employers is still
present in every modern society and has important implications for policy choices, such as
the privatization of infrastructure, including railways, telecommunications or water ser-
vices (see Fink 2011; Obinger et al. 2014; Schiffler 2015; Schmitt and Zohlnhöfer 2017;
Zohlnhöfer et al. 2018), or welfare policies (see Häusermann et al. 2013; Manow et al. 2018).
It is the classic domain of communist parties (e.g. the Portuguese Communist Party), social
democratic parties (e.g. the Chilean Party for Democracy) and labour parties (e.g. the Irish
Labour Party).
The materialist–post-materialist cleavage describes what priority social groups give to
new values. This cleavage is strongly interlinked with Inglehart’s (1977) value concepts of
materialism and post-materialism. These terms explain the way in which political values rise
out of individual needs during the process of socialization, i.e. learning how to behave
according to cultural norms. Materialist values are the concern of those who have experi-
enced political and economic insecurity, leading them to give priority to political stability
and economic strength. Those who have post-materialist values have been exposed to
greater security and are oriented towards satisfying other needs. As a result, post-materialists
place a higher value on ideas such as equality of opportunity, citizen involvement in decision-
Copyright © 2020. Macmillan Education UK. All rights reserved.

making at government and community levels (also known as ‘direct democracy’) and envi-
ronmental protection (Inglehart 1977, 1997; Welzel 2013). In modern societies, Green
Parties are predominantly associated with post-materialism (see Bomberg 2005; Carter
2006, 2013; Spoon 2009; Tosun 2017). The cleavage between materialism and post-
materialism also matters for countries’ stances on morality policies. The permissiveness in
the regulation of moral issues is not only affected by the state–church cleavage, but also by
the extent to which political parties representing post-materialist positions are represented
in government and the legislature (Budde et al. 2017a).
With national boundaries becoming more and more permeable, an open–closed society
cleavage has emerged, in which social groups are either in favour of or against the opening
of markets. Those who feel disadvantaged by economic globalization favour trade barriers to
protect local manufacturing and ‘locals first’ policies in the labour market. The economically
defensive attitude of these groups is reinforced by anti-immigration stances, which stress
religious and national values against a multi-ethnic society (see Kriesi 2010; Kriesi et al.
2012; Burgoon 2013; Zürn and de Wilde 2016; Walter 2017). Often nationalist parties,

right-wing populist parties, racist parties and extreme right parties address this social cleav-
age (see Bornschier 2018).
The cleavage approach is important for understanding how party families have formed.
The various political parties, in turn, address existing social problems by proposing spe-
cific solutions to them, which may eventually become public policy. In this context, it
should be noted that the cleavage approach as outlined here mostly corresponds to the
European context. Nevertheless, many of these social cleavages can be found elsewhere.
Australian rural policy, for instance, is marked by the rural–urban cleavage (see Botterill
2009). Croissant and Völkel (2012) and Croissant and Lorenz (2018) show that party sys-
tems in Indonesia, Taiwan and South Korea are structured along dominant cleavage lines.
The dominant cleavage in South Korea and Indonesia is the one on religion and seculariza-
tion, whereas in Taiwan political parties are structured along conflict over national iden-
tity forms.
Outside Europe there are other cleavage structures. In Canada, for example, there is a
linguistic cleavage between English-speaking and French-speaking Canadians that is likely
to influence public policy (see Béland and Lecours 2007; Lambert and Curtis 2008; Bickerton
and Gagnon 2014). To give another example, in East Timor the occupation by Indonesia
(1975–1999) resulted in the emergence of a specific cleavage concentrating on people who
left the country during occupation and those who stayed behind to fight (Croissant and
Völkel 2012).
Irrespective of the exact cleavage structures, what matters is that there are deep and lasting
divisions between groups in societies that generate certain political parties that are then
involved in the policy-making process and aim to realize their preferences when proposing
public policies.

Varieties of Capitalism
The Varieties of Capitalism (VoC) approach is closely connected with a subdiscipline of
political science known as political economy, which can very broadly be defined as a per-
spective that examines how economic processes influence policy-making. John (2018)
shows that the perspectives of political economy and public policy analysis lend them-
Copyright © 2020. Macmillan Education UK. All rights reserved.

selves to integration because they share an analytical interest in the same concepts, such
as blame avoidance (see Weaver 1986; Hood 2007; Bonoli 2012; Mortensen 2012;
Wenzelburger 2014; Hinterleitner and Sager 2015; Wenzelburger and Hörisch 2016;
Howlett and Kemmerling 2017).
In the comparative political economy literature, scholars have distinguished between dif-
ferent forms of capitalistic economies, which vary according to the coordination of public
and private actors. This literature has referred notably to coordinated (Germany, Sweden)
and liberal (United Kingdom, United States) market economies (CMEs and LMEs; Hall and
Soskice 2001; Hancké et al. 2007). As their respective names suggest, in LMEs strong mar-
ket forces dominate, while in CMEs coordination mechanisms play a major role between the
various market participants. In LMEs, companies coordinate their activities in all subsys-
tems of the economy primarily by orientation on price signals, free competition on the mar-
ket and marginal cost. In CMEs, non-market-based relations play a crucial role in the
development of corporate core competencies. Therefore, the LME-typical short-term profit-
ability indicators have a limited influence on decisions to cooperate or not in CMEs.

Furthermore, CMEs are characterized by the existence of coordination mechanisms that

include economic groups such as employers’ associations or unions and the state.
Hall and Soskice (2001) find that CMEs can produce goods that stem from incrementally
innovative production, which requires especially industry- and firm-specific human capital,
more efficiently than liberal market economies. One of the reasons for this is the nature of
intercompany relations giving rise to joint research, product differentiation and niche produc-
tion, rather than direct product competition. Another reason is based on the human capital
argument. In CMEs, employees are provided with sophisticated training schemes and appren-
ticeships (see Busemeyer and Trampusch 2012), so that they rely on sufficiently specialized
knowledge to propose and implement incremental changes in the production line. This
requires a ‘mix of company-specific and more general technical skills’ in the workforce (Hall
and Soskice 2001: 36). These differences between CMEs and LMEs shape policy-making in
areas such as (vocational) educational training and child-care policies as well as policy responses
to economic crises and austerity (see Bonoli and Reber 2010; Jensen 2011; Thelen 2012;
Breunig and Busemeyer 2012; Dobbins and Busemeyer 2014; Hörisch and Weber 2014).
More recent developments in the literature point to the existence of additional types of
market economies (see Hancké et al. 2007). One of the newer types refers to mixed-market
economies (France, Italy; Molina and Rhodes 2007) as well as Central European (King 2007)
or dependent market economies (Nölke and Vliegenthart 2009).

Key points
•• The cleavage approach focuses on basic lines of societal conflicts as the main determinants of policy-
making. Historical cleavages include conflicts between centre and periphery, church and state, rural
and urban areas and employers and workers.
•• The two newer types of cleavages address the division between materialists and post-
materialists and supporters and opponents of an open society.
•• VoC offers a systemic view that emphasizes linkages across all of the major institutions that
define capitalist political economies and the policy decisions they make.
Copyright © 2020. Macmillan Education UK. All rights reserved.

INSTITUTION-BASED MODELS

In Chapter 3 we introduced the most important formal institutions to provide an overview

of the context in which policy-making takes place. In this section, we move beyond the mere
description of the characteristics of institutions and introduce theoretical approaches that
explain how they matter for policy outputs. The different theoretical perspectives all con-
ceive of institutions as the central independent variable for explaining policy-making, which
itself represents the dependent variable.
The major difference between institutional approaches is their respective definition of
institutions. The ‘classic’ approaches to institutional analysis rely on a conception of ‘insti-
tutions’ as formal-legal arrangements, similar to the understanding presented in Chapter 3.
‘New institutionalism’, by contrast, offers more complex or broader notions of institutions
and the ways in which they matter for policy-making. Despite being termed institutionalist,
it should be noted that both approaches explicitly acknowledge the importance of actors.
Indeed, the most general way of explaining how institutions affect policy outputs is by

stating how it is that they empower and constrain actors (see March and Olsen 2008: 3). In
the following section we will illustrate this causal relation.

The Classic Approach to Political Institutions

For a long time, classic institutionalist analyses of public policy have drawn on the formal-
legal approach, which emphasizes the role that governmental organizations play. This classic
form of institutionalism can be characterized in two ways. First, the basic independent vari-
able is given by the legal rules and procedures in a political system. The different functions
of the state, including policy-making, and how they are performed represent the dependent
variable. Second, legal rules are regarded as behavioural prescriptions. For example, policy-
makers in some nations would not think about formulating legislation with a discrimina-
tory character because they know that the constitution prohibits this.
Against this backdrop, classical institutionalism contends that the functioning of the state
not only depends on economic and social conditions, but also on the design and effectiveness
of political institutions (March and Olsen 1984: 734). A similar perspective is adopted by
Arend Lijphart (1999/2012) in his influential book Patterns of Democracy. The author scruti-
nizes the relationship between government forms and policy performance, i.e. how effectively
a government can solve a social problem. In so doing, he argues that political systems of
advanced democracies in all their complexity can be basically assigned to two competing cate-
gories: ‘majoritarian’ and ‘consensus’. Majoritarian democracies typically have majority or plu-
rality electoral systems, only two major political parties, single-party cabinets, effective
unicameralism and a unitary and centralized government. Consensus democracies, by con-
trast, are characterized by proportional electoral systems, having more than two major par-
ties, coalition cabinets, bicameralism and decentralized or federal political systems.
Accordingly, the United Kingdom would represent a majoritarian democracy and
Switzerland a consensus democracy. Applying the logic of classical institutionalism, British
public policy would have to be significantly different from Swiss public policy because of the
different institutional arrangements in the two countries (for discussions, see Armingeon
2002; Schmidt 2002; Roller 2005; Vatter and Stadelmann-Steffen 2013; Vatter et al. 2014;
Bernauer et al. 2016). To be sure, Lijphart’s approach somehow parallels the idea about
Copyright © 2020. Macmillan Education UK. All rights reserved.

national policy styles which we introduced in Chapter 2. Lijphart’s theoretical argument

suggests that policy styles flow from electoral systems and the distribution of power. So
policy styles in countries corresponding to a majoritarian democracy are expected to be dif-
ferent from those in consensus democracies, as the former produces a concentration of
power at the centre and encourages majoritarian, top-down government, whereas the latter
diffuses power and encourages the formation of coalitions and the pursuit of consensus.
Another important classic institutionalist analysis is represented by the work of Francis Castles
(1998), which analyses variations in public policy by combining political-institutional variables and
socio-economic indicators. The study reveals the existence of four ‘families of nations’, which differ
in respect of policy-making, particularly in the areas of social and economic policy:

• An English-speaking family of nations, including Australia, Canada, Ireland, New

Zealand, the United Kingdom and the United States.
• A Continental family of nations, consisting of Austria, Belgium, France, Germany, Italy
and the Netherlands.
• A Scandinavian family of nations, consisting of Denmark, Finland, Norway and Sweden.
• A Southern family of nations, comprising Greece, Portugal and Spain.

Again, this analytical approach corresponds with the notion of national policy styles, espe-
cially to the extent that political-institutional variables – including a categorization of
countries in accordance with Lijphart’s concept of majoritarian and consensus democra-
cies – represent an essential component of the theoretical framework. Moreover, it is the
assumption that public policy is the result of stable and family-specific policy-making pro-
cesses which parallel the concept of national policy styles.
While the core argument put forward by classic institutionalism (that formal-legal insti-
tutions matter) seems intuitive, two major points of criticism have been raised. First, it was
argued that classical institutionalism would not open the ‘black box’ between formal institu-
tions and policy choices. As Schmidt (2002: 160) puts it: ‘types of democracy, like other
constitutional structures, are both constraints on policy choice and at the same time condi-
tions enabling choice between policy alternatives. However, types of democracy do not
determine these choices: nor do they determine their outcomes.’ Second, political institu-
tions in a country tend to be stable even though public policy is constantly being modified.
These and other points of criticism of classic institutionalist approaches have given way to
‘new institutionalism’.

The New Approaches to Political Institutions

Strictly speaking, new institutionalism does not constitute a single and coherent body of
theory, but comprises many different streams of argument. Although sharing a basic com-
mon assumption – namely, that institutions do matter – there are a variety of conceptions
of how, why and to what extent institutions make a difference (Knill 2001). Essentially, new
institutionalism contends that institutions matter because they exert an influence on public
policy (Thelen and Steinmo 1992: 7). In this regard, they can be independent variables or
‘intervening’ variables, i.e. have some impact on the relationship between the main inde-
pendent variables and the dependent variable.
Following Hall and Taylor (1996), three different varieties of new institutionalism can be
distinguished: sociological institutionalism, historical institutionalism and rational choice
institutionalism. Here we will focus on sociological and historical institutionalism, since
they represent so-called institution-based approaches; i.e. they attach a determining explan-
Copyright © 2020. Macmillan Education UK. All rights reserved.

atory role to institutional factors. With rational choice institutions, the actors come to the
fore, so we shall discuss this form in the next section on interest-based models.

Sociological Institutionalism
At the heart of sociological institutionalism is a very broad understanding of institutions,
incorporating symbol systems, cognitive scripts and moral templates that provide meaning
to action. Hall and Taylor (1996: 947) argue that this definition abrogates the divide between
‘institutions’ and ‘culture’. Sociological institutionalism has a distinctive understanding of
the relationship between institutions and individual action. It stresses the way in which
institutions influence behaviour by providing the cognitive concepts and models that are
indispensable for action. In this way, the relationship is between institutions, and individual
action is thought to be interactive and mutually constitutive.
Central to this perspective is the notion that action is tightly bound up with interpreta-
tion, which corresponds to the idea of social constructivism. This means that sociological
institutionalists perceive actors as purposive, though purposive action is itself socially con-
stituted in accordance with the ‘logic of appropriateness’ (March and Olsen 1996: 252). The

logic of appropriateness states that individuals make their choices according to what they
view as socially valuable. Rules, for example, are followed because they are regarded as the
‘right thing’ (see Börzel and Risse 2010: 121). Put differently, individuals’ actions are deter-
mined by their sense of obligation as structured by the ‘appropriate’ institutional rules and
routines rather than by self-interest. This adds another understanding to what institutions
are, as they are not only regarded as enabling and constraining actors’ preferences, but also
as influencing the way actors conceive of their preferences in the first place. It is hence by
influencing actors’ preferences and perceptions that institutions shape a polity and thereby
the context in which policy-making occurs.
Furthermore, sociological institutionalism argues that many of the institutional forms
and procedures used by modern organizations are culturally specific practices rather than a
means to enhance efficiency and effectiveness. On this basis, the research question that soci-
ological institutionalists typically pose is: why do organizations take on specific sets of insti-
tutional forms, procedures or symbols? They also stress that organizational practices are
prone to diffuse through organizational fields or across nations. In this context, sociological
institutionalists are interested, for instance, in explaining the striking similarities between
organizational forms and practices. With regard to this latter point, the concept of legitimacy
is crucial, for it implies that the main logic of action guiding organizational behaviour is to
increase legitimacy in the organizational environment rather than functional efficiency. A
consequence of seeking legitimacy is that, when creating new institutions, these will most
likely be patterned after existing institutional templates. This is expected to lead to institu-
tional isomorphism, i.e. the emergence of similar organizational structures. According to
DiMaggio and Powell (1991), an organization will engage in institutional isomorphism in
response to three types of pressures: coercive, mimetic and normative.
According to coercive isomorphism, organizations adjust their structures and procedures
to be in line with those organizations on which they are dependent. However, organiza-
tional adjustment to the environment is not only expected to take place as a result of coer-
cive pressures, but may also occur in constellations of high uncertainty, e.g. in the presence
of ambiguous goals or confrontations with new problems. In such constellations organiza-
tions are likely to practise mimetic isomorphism by imitating the structures of other orga-
nizations which they perceive as particularly successful so as to ensure their legitimacy by
Copyright © 2020. Macmillan Education UK. All rights reserved.

emulation. Normative isomorphism is based on similar dominant normative orientations of

staff members. In this regard, similar professional backgrounds and the role of professional
organizations in spreading mutual understanding of policy problems and solutions are
important (for an overview, see Knill and Balint 2008).
Regardless of the type of isomorphism concerned, the central argument is that legitimacy
rather than functional efficiency is the main driving force of organizational change. For
example, institutional isomorphism can be seen as the main mechanism underlying changes
in higher education policies. In the last decade, many countries started to change their
higher education policies and the organizational structures of their universities and colleges
in order to pattern them after the US or the British model, as these are broadly perceived as
legitimate (see Dobbins and Knill 2009; Dobbins 2011).

Historical Institutionalism
At the core of historical institutionalism is the argument that ‘history matters’. More spe-
cifically, historical institutionalism emphasizes the fact that former choices about policies or
institutions affect the range of subsequent policy options, for instance by generating certain

ideas about public policy (see Cox 2004). This structuring effect of existing arrangements is
conceived as so far-reaching that existing institutions or policies are considered as indepen-
dent rather than as merely intervening variables in explaining policy choices, which distin-
guishes historical institutionalism from the other two new institutionalisms. A central
emphasis is hence placed on institutional path dependence and policy legacies.
Historical institutionalists point out that for explaining policy choices it is more impor-
tant to recognize that ‘history matters’ than to figure out whether path dependence is the
result of lock-in effects, the distribution of bargaining power or the institutional reproduc-
tion of standard operating procedures (Hall 1986: 19; Thelen and Steinmo 1992: 9).
Institutions can be seen as the points of ‘critical juncture’ (i.e. the adoption of a particular
institutional arrangement) in a historical path analysis, since political battles are fought
inside institutions and over the design of their futures (Steinmo et al. 1992). Thus, critical
junctures represent rare events in the development of an institution. Instead, the normal
state of an institution is either one of stability or one of limited adaptive change (Mahoney
2000; Capoccia and Kelemen 2007: 367).
Historical institutionalists stress the structuring impact of institutional ‘lock-ins’, where
deviations from the initial path become increasingly costly or difficult as a result of the insti-
tutionally structured distribution of power between different actors (North 1990; Pierson
2000, 2004). However, this does not mean that public policy must remain stable. Streeck
and Thelen (2005) argue that policy and institutional change takes place, but usually in a
gradual manner. This suggests that institutional change is not necessarily the result of a
critical juncture; it can also be the result of an incremental process. Therefore, historical-
institutionalist analyses have to make sure that they conceive of policy and institutional
change as a long-term process (see Pierson 2004: 82).
The puzzles that have typically been addressed by historical institutionalists are mostly
empirical, corresponding to questions such as: why was one policy chosen rather than
another (Immergut and Anderson 2008: 351)? The concepts of critical junctures and path
dependence as the central components of historical institutionalism have been applied to
many policy areas, including social welfare policies, constitutional law, budgetary law, regu-
lation of competition in product markets and banking, and the development of regimes or
entire regions (Capoccia and Kelemen 2007: 345).
Copyright © 2020. Macmillan Education UK. All rights reserved.

A well-known example of a historical-institutionalist approach is given by Esping-

Andersen’s research (1990). Pursuing the analytical objective of explaining welfare state
development, Esping-Andersen argues that the following three factors are relevant: the
mobilization of the working class, class-political action structures and the historical legacy
of regime institutionalization. These considerations induce Esping-Andersen to categorize
welfare state regimes into three groups: the Anglo-Saxon liberal model, the Scandinavian
social democratic model and the Continental conservative model. More recent research
points to the existence of a South European model (see Ferrera 2010; Manow 2015), a post-
communist regime and former-USSR welfare regimes (see Campos-Matos and Kawachi
2015; for a discussion, see Jensen and van Kersbergen 2016).
Following the logic of welfare regimes, modifications to welfare policy may occur within
these types, but they do not cut across them. In other words, it is highly unlikely that the
Scandinavian countries will give up their historically developed generous social democratic
model and replace it with the less generous Anglo-Saxon liberal one. Rather, they are
expected to preserve the essential characteristics of their welfare policy and make only mar-
ginal modifications to it (see Taylor-Gooby et al. 2017).

Discursive Institutionalism
The newest form of institutionalism was put forward by Schmidt (2008a, 2010) and is called
discursive institutionalism. From this perspective, institutions are ‘simultaneously structures
and constructs internal to agents whose “background ideational abilities” within a given “mean-
ing context” explain how institutions are created and exist and whose “foreground discursive
abilities”, following a logic of communication, explain how institutions change or persist’
(Schmidt 2008a: 303). Thus, compared with the previously discussed types of new institutional-
ism, this definition of institutions in discursive institutionalism has the widest scope of all.
The main goal of this type of institutionalism is to explain institutional change by stress-
ing the importance of ideas and discourse in policy-making. Ideas represent the substantive
content of discourse; they can be both cognitive and normative and exist at the levels of
policies, programmes and philosophies. Discourse is the process by which individuals inter-
act in order to exchange ideas. Schmidt differentiates between a coordinative discourse
amongst policy actors (e.g. members of a coalition government) and communicative dis-
course between policy-makers and the public (e.g. policy programmes communicated by
means of election manifestos).
Schmidt (2010: 5) explains that the main analytical advantage of this new perspective is that
it conceives of institutional change as an endogenous process (i.e. one that takes place within
institutions) through ‘background ideational and foreground discursive abilities’. She contends
that the other types of new institutionalism are more apt to explain institutional stability than
institutional change. The concept has also been developed further various times, for example to
theorize ideational power (see Béland 2010; Béland and Cox 2011). In this context, Carstensen
and Schmidt (2016) proposed a threefold typology of power and ideas: first, ‘power through
ideas’ (the capacity of actors to persuade other actors to accept their views); second, ‘power over
ideas’ (the imposition of ideas on actors); and finally, ‘power in ideas’ (the establishment of
institutions imposing constraints on what ideas can be considered).
In contrast to the new institutionalisms that emerged in the 1990s, discursive institu-
tionalism has been challenged by policy scholars. Bell (2011), for instance, argues that insti-
tutional change can also be explained by historical institutionalism if it conceives of
institutions as flexible and not static, which makes the need for this form of institutional-
ism obsolete.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Regardless of the criticism, discursive institutionalism has been embraced by policy

scholars and is applied to explain a wide range of themes in public policy, such as changes in
school curricula (see Wahlström and Sundberg 2018) and the EU’s Emissions Trading
System (see Fitch-Roy et al. 2019).

Key points
•• Classic institutionalist approaches conceive of institutions in strictly formal terms, i.e. the
‘rules’ of the political ‘game’. This perspective regards policy outputs as the result of exist-
ing formal political institutions.
•• Sociological institutionalism defines institutions very widely and argues that they shape actors’
perceptions, which, in turn, affects their behaviour, so that it becomes based on the logic of
appropriateness and/or on legitimacy.
•• Historical institutionalism emphasizes the structuring impact of institutional lock-ins and
path dependence.
•• The newest form of new institutionalism is discursive institutionalism, which emphasizes the
importance of ideas and discourse.
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:20:21.
Theoretical Approaches to Policy-Making 67

INTEREST-BASED MODELS

This section introduces the most influential interest-based models. First, we outline the con-
cepts of full and bounded rationality, as these represent the behavioural assumptions under-
lying many interest-based models. Next, we introduce models focusing on the strategic
actions of public and private actors. This is followed by a concise presentation of the main
characteristics of game theory, which represents a useful device for analysing the interac-
tion of actors. Finally, we will shed light on how institutions can structure the interaction of
actors. To this end, rational choice institutionalism will be introduced.

Concepts of Rationality
When focusing on actors, it is necessary to decide whether to model them as ‘fully’ rational
or ‘boundedly’ rational. A rational actor is broadly perceived to make choices about public
policy that tend to maximize his or her expected utility. For a choice to be rational, an actor
must be faced with alternatives that can be rank-ordered according to some degree of utility
or desirability. A rational policy choice requires preferences to display at least two character-
istics. First, they should be complete; i.e. all alternatives must be ranked in order. Second,
they should be transitive, implying that, if alternative A is preferred to B and alternative B
is preferred to C, then A must also be preferred to C.
Let us assume that a policy-maker can choose between the following three policy instru-
ments in order to lower the emission of greenhouse gases: (1) expansion of nuclear power
generation, (2) imposition of a tax on energy-intensive industries and (3) offering insula-
tion subsidies for home owners. In real life there exist many more possible policy instru-
ments, but to illustrate the concepts of completeness and transitivity we will constrain
ourselves to these options. A preference order is complete if the policy-maker can place each
of these three policy instruments in a rank order. Such a preference order could look as fol-
lows. The most preferred policy instrument is the insulation subsidies, followed by the tax
on energy-intensive industries, followed by an expansion of nuclear power generation. The
preferences are closely related to the respective utility of each instrument, which may as
Copyright © 2020. Macmillan Education UK. All rights reserved.

much stem from the likelihood of being re-elected as from believing that one policy instru-
ment outperforms the others on account of its superior effectiveness. Now transitivity
comes into play, since if the subsidies are preferred over the tax, then the subsidies must
also be preferred over an expansion of nuclear power generation. Only if this condition
holds can a policy-maker’s preference order be regarded as rational.
The concept of a fully rational actor is demanding, as it rests on the strong assumptions
that he or she possesses the best information available and can use it in a way to choose
from amongst all other possible policy options that alternative which is most likely to maxi-
mize his or her utility (for a discussion, see Shafir 2013). When looking at the immense
number of tasks that policy-makers are confronted with every day, doubts arise as to
whether these assumptions are realistic. In response to the dissatisfaction with the assump-
tions underlying fully rational behaviour, Herbert Simon (1957) developed the concept of
bounded rationality (for an overview, see Jones 2001).
Similar to the concept of full rationality, bounded rationality assumes that actors are goal
oriented, though it equally takes into account their cognitive limitations. Most importantly, the
amount of information actors can gather and process in a meaningful way is thought to be

limited. While the concept of full rationality pays attention to the notion of optimizing, i.e. the
selection of the best possible action, bounded rationality mostly relies on the concept of ‘satisfic-
ing’, which describes the situation where people settle on a solution that is satisfactory.
Recalling our previous example about the three possible policy instruments for reducing
greenhouse gas emissions, a decision-maker practising satisficing might choose the insula-
tion subsidies simply because this is an instrument that has already been applied in the
country’s past and/or in other countries and produced satisfactory results. Hence, the policy
instruments are not rank-ordered to meet the criteria of completeness and transitivity –
rather, the instrument chosen is the one for which information is available concerning the
most likely satisfactory outcome.

Strategic Choices of Actors

As we demonstrated in Chapter 3, there are numerous actors participating in the policy-
making process. So it is impossible to introduce at once all the theoretical approaches that
name the relevant policy actors and the ways in which they affect public policy. Rather, here
we limit ourselves to outlining briefly some of the most important theoretical approaches
for explaining the behaviour of elected officials and bureaucrats on the one hand, and of
interest groups as key private actors on the other hand. In so doing, we present theories that
are based on the assumption that the actors are fully rational – ‘rational choice approaches’.
In subsequent chapters, we will also introduce some theories that rest on the assumption of
bounded rationality.

Explaining the Behaviour of Public Actors

Rational choice theorists have formulated theories to explain the behaviour of elected offi-
cials. Often individual behaviour is explained on the basis of competitive party behaviour.
Strøm (1990) explains that political parties and therefore elected officials can be seen as (1)
vote-seeking, (2) office-seeking or (3) policy-seeking. The first model is based on Downs’
(1957) work on electoral competition, in which parties maximize their electoral support for
Copyright © 2020. Macmillan Education UK. All rights reserved.

the purpose of controlling government. Office-seeking parties want to maximize their con-
trol over political office, for instance by acquiring the most influential or prestigious minis-
tries. This theoretical perspective on the intrinsic reward of office is based on the work of
Riker (1962). Finally, parties can be policy-seeking; i.e. they participate in politics in order
to further particular policy objectives, often reflecting their underlying ideological positions
(Budge and Laver 1986: 485).
When exploring the impact of political parties on public policy, the ‘party difference
hypothesis’ put forward by Hibbs (1977) is crucial. In its original form, the model explains
political parties’ different views on macro-economic policies, notably the alleged trade-off
between inflation and unemployment. The basic argument is that left-wing parties would
solve this trade-off by preferring lower unemployment over inflation, whereas right-wing
parties would tend to keep inflation at the lowest possible level. Rational voters should
hence support left-wing parties when high unemployment is expected and right-wing par-
ties when high inflation is expected. According to partisan theory, ‘parties do matter’ for all
areas of policy-making (see Schmidt 1996). This means that parties compete for votes by
offering diverse policy options.

The role of bureaucracy in the policy-making process has most prominently been
addressed by Niskanen’s (1971) model of a budget-maximizing bureaucracy. It contends
that bureaucrats will try to maximize their department’s budget in order to increase their
salary and prestige. The basic idea is that the government defines a department’s budget on
the grounds of the quantity of service it supplies. The more services the department sup-
plies, the higher will be its budget. Therefore, the bureaucrat’s objective will be to maximize
the quantity of services supplied. This can only work due to information asymmetries
between the bureaucracy and the government. The latter cannot directly observe the level of
service output, which gives the bureaucracy a preferential position in this relationship.
Further, the dead-weight loss of the excessive production of services must never be higher
than the government’s consumer surplus, otherwise the government would notice irregu-
larities in the bureaucrats’ activities. Despite its far-reaching analytical impact, Niskanen’s
model has also provoked criticism, which has led to the development of alternative models
(see also Chapter 6).
One of these alternative approaches is the ‘bureau-shaping model’ as put forward by
Dunleavy (1991). In a nutshell, it argues that senior bureaucrats are most interested in max-
imizing the status and quality of their work, which is best pursued through bureau-shaping
rather than budget maximization. To pursue bureau-shaping, bureaucrats can employ up to
five strategies:

1. They can refer to internal reorganizations in which the policy-making role of senior
bureaucrats is strengthened.
2. Internal work practices can become transformed so as to increase the status of monitor-
ing activities.
3. Bureaucrats can redefine relationships with external partners in order to maximize their
control over policy and rid themselves of routine issues.
4. They may compete with other bureaux over the defence of policy-oriented

responsibilities.
5. The bureau-maximizing actor seeks to transfer to other bodies (such as executive agen-
cies) all functions and tasks which do not fit with their preferred bureau shape (see Marsh
et al. 2000).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Recall that in Chapter 3 we hinted that bureaucrats might be more than just an instrument
of government. Above, we have reinforced this argument by underscoring the fact that
bureaucrats also have their own individualistic preferences that they seek to realize in the
policy-making process. In other words, despite Weber’s (1947) prognosis of a concern for
the collective future of government, the preferences of bureaucrats can deviate from those
of their political masters and have implications for public policy. We will return to this in
subsequent chapters.

Explaining the Role of Private Actors

While elected officials are central for taking policy decisions and bureaucracies are essential
for the drafting of public policies and policy implementation, interest groups can also exert
considerable influence on policy-making through their specialist knowledge, financial con-
tributions and political campaigns (Howlett et al. 2009: 96). The relative influence of an
interest group principally depends on three factors. First, the internal organization matters,

which includes factors such as income and organizational characteristics (e.g. sanctioning
mechanisms). In this context, the size of membership and the internal cohesion as well as
the recruitment of new members are equally relevant factors. Second, the very topic on
which the interest group works might make a difference, as with complex and highly techni-
cal issues interest groups possess more leverage over policy-making since the necessary
information may not otherwise be provided. Third, the relative influence of an interest
group depends on the degree of competition with other interest groups (see Coen and
Richardson 2009; Maloney and van Deth 2010; Newton and van Deth 2010: 170). The inten-
sity of competition in turn is determined by the access of various interest groups to decision-
makers, which leads us to the concepts of pluralism, corporatism and neo-corporatism.
The theory of pluralism suggests that in democratic polities power is widely distributed
amongst different groups, creating a ‘marketplace’ with more or less perfect competition for
influence over public policy (Dahl 1958). Pluralism is based on four premises. It assumes
equal access to the policy-making arena between different groups, a fragmentation of the
‘marketplace’ of interest groups, a competitive process for determining policies and the neu-
trality of government. From this it follows that the relative influence of a group is a function
of its resources and ‘decibel rating’ (Hill 2009: 30). In pluralist systems – according to the
criteria defined above – such as the United States or the United Kingdom, interest groups
are viewed as making a vital contribution to policy-making ‘because (1) they provide a more
effective voice for citizens who are competing for resources, (2) they reduce the anxiety pro-
duced by feelings of powerlessness, and (3) they provide an element of stability’ (Dye and
Zeigler 2006: 234).
In corporatist polities, the central actors are still organized interests, but the policy-
making process is cooperative rather than competitive, and is closed to all but certain privi-
leged interest groups, mainly those representing business and labour (Schmidt 2008a,
2008b: 305; see also Lehmbruch and Schmitter 1982). Following Schmitter’s (1977: 9) defi-
nition, corporatism is a system of interest intermediation in which the interest groups are
organized into single, compulsory, non-competitive and hierarchically ordered categories
that have been given a representational monopoly by the state in return for their exercising
control over their supporters. According to Siaroff (1999), an ideal, typical, corporatist sys-
tem would exhibit:
Copyright © 2020. Macmillan Education UK. All rights reserved.

• A high degree of unionization

• A concentration of labour unions
• A business community dominated by large export-oriented firms
• Centralized bargaining of wages
• The existence of work councils in industrial firms and codetermination in important
industry sectors, and a centralized state that is involved in the economy

This means that fewer interest groups participate in policy-making, but the ties between
them and the decision-makers are institutionalized. Often, these interest groups are incor-
porated officially into policy-making as members of a committee.
Neo-corporatist polities differ from corporatist ones with respect to the greater constitu-
tional autonomy of the groups involved and the voluntary nature of involvement in policy-
making. Neo-corporatism also refers to social arrangements dominated by tripartite
bargaining between unions, the private sector (capital) and government (Lehmbruch 1982;
Streeck 1982). This model of interest intermediation principally evolved in the post-war

period in connection with the increased state intervention in the economy and the growing
involvement of the central trade union and employers’ associations in economic planning
and incomes policy.
Common to both corporatist and neo-corporatist models of interest intermediation is a
stable pattern of consultation and cooperation between one or a few dominant interest
groups and policy-makers, indicating that public policy is a ‘negotiated’ outcome between
both actor groups (Adolino and Blake 2011: 51). As a result, we do not further distinguish
between corporatism and neo-corporatism, but employ the term ‘(neo-)corporatism’
instead. Another reason is that ‘corporatism’ can have a negative connotation, since it has to
an extent been associated with semi- or non-democratic regimes, especially in Latin America.
Therefore, we have a preference for the term ‘(neo-)corporatism’ to underline that we exclu-
sively focus on democratic forms of interest intermediation.

Understanding the Interaction of Actors: Game Theory

So far, we have mainly discussed theoretical approaches that focus on individual actor
groups. An exception was the presentation of the concepts of pluralism and (neo-)corporat-
ism, which address the relationship between interest groups and the government. However,
to understand policy-making, one cannot only concentrate on the actors in isolation, but
must pay equal attention to the ways in which they interact with one another. This is neces-
sary as, akin to institutions, the presence of other actor groups can either help or impede
policy-makers’ interests from becoming directly transposed into public policy. The most
general way of theoretically addressing interactions is provided by game theory.
Game theory is a device to analyse decision-making in situations in which two or more
rational players interact, and where the outcome depends on the choices made by each.
Actors involved in games are referred to as players. Each player in a game faces a choice
amongst two or more possible strategies. A strategy tells the player which actions to take in
response to every possible strategy other players might use. A crucial aspect of the specifica-
tion of a game involves the information that players have when they choose strategies. The
simplest games are those in which the players have perfect information, meaning that at
every point where each player’s strategy tells him or her to take an action, he or she knows
Copyright © 2020. Macmillan Education UK. All rights reserved.

everything that has happened in the game up to that point. More realistically, however,
players generally have imperfect information about each other’s strategy (for an overview,
see McCarthy and Meirowitz 2007).
The most famous game of incomplete information is the prisoner’s dilemma (see Box
4.1). This game is representative of an important class of structural situations involving
conflict and cooperation amongst actors. Despite the availability of better outcomes, the
suboptimal one is chosen due to a lack of trust in the other player. It is a dilemma since it
illustrates how, even when rational actors appreciate the importance of cooperation, they
still find themselves trapped in situations where they cannot get to their desired goal
through cooperation.
The theory of regulatory competition in environmental policies represents an illustrative
example of a prisoner’s dilemma. In brief, this theory argues that states have incentives to
lower their environmental protection standards in order to attract mobile capital. The idea
underlying this reasoning is that environmental protection standards (e.g. the need to meet
criteria for industrial emissions into air and water) cause costs to industries that they seek

Box 4.1 Structure of the prisoner’s dilemma

Suppose that the police have arrested two men, whom they know have committed
an armed robbery together, but lack enough admissible evidence to get a jury to con-
vict them. They do, however, have enough evidence to send each prisoner away for
two years for the theft of the getaway car. The inspector now makes an offer to each
prisoner that if he confesses to the robbery – and his partner in crime does not – he
will go free, whereas the partner will be sent to prison for ten years. However, if both
confess, each will get five years in prison. If neither of the suspects confesses, then
each will have to go to prison for two years for the car theft.
We can now rank these options with regard to their utility for each of the two sus-
pects. The following numbers in parentheses are used to express each player’s payoffs
in the various possible outcomes. Payoffs are the values assigned to each individual
outcome with regard to their desirability: the higher the numerical value of a payoff,
the more a certain outcome is desired. The best option is to go free (3), followed by a
two-year stay in prison (2), then a five-year stay (1) and then a ten-year stay (0). The
matrix below represents this situation of strategic interaction in the so-called normal
form. We refer to the two suspects as Player 1 and Player 2:

Player 2
Confess Keep silent
Player 1 Confess 1, 1 3, 0
Keep silent 0, 3 2, 2

Each cell of the matrix indicates the payoffs to both players for each combination of
actions. Player 1’s payoff appears as the first number, and Player 2’s as the second. If
both players confess, each gets a payoff of 1. If both keep silent, each gets a payoff of
2. If one player confesses but not his partner, that player gets a payoff of 3, whereas
the partner gets a payoff of 0. Taking into consideration that one player might always
Copyright © 2020. Macmillan Education UK. All rights reserved.

have an incentive to confess, this action represents the best strategy regardless of
what the partner does. In more technical terms, this means that confessing beats
keeping silent for both players. Both players know this about each other, thus entirely
eliminating any temptation to depart from the dominant strategy. This solution rep-
resents the so-called Nash equilibrium of the game, which is named after the Nobel
Laureate John Nash who proposed it.

to evade by shifting to another jurisdiction. Assuming that governments are rational actors
interested in optimizing their benefits, the theory of regulatory competition predicts that
there will be a continued lowering of environmental protection standards across countries
to the level of the least stringent country. This will ultimately result in suboptimal environ-
mental protection standards, simply because countries do not trust one another, which pre-
vents them from collectively attaining higher standards (see Holzinger 2008).

In political science, the provision of common goods (i.e. goods that one cannot be effec-
tively excluded from using) is frequently modelled as a prisoner’s dilemma. In this regard,
the best-known pathology is the ‘tragedy of the commons’ (Hardin 1968; Ostrom 1990),
meaning that if there is no regulation of a certain public resource, e.g. a clean environment
or public roads, individuals are likely to consume more than their fair share of it. The long-
term effect of this free-riding is that at some point in time the amount of public resource will
decline, entailing negative consequences for all. In such a case, we are confronted with a
so-called negative externality, since the cost to society as a whole is greater than the cost to
the individual who is paying for it. Again, this point can be illustrated by drawing on an
example from environmental policy. When pollution crosses borders countries are often
reluctant to exercise full control. This happens mostly in the case of air pollution, which
never stops at national borders. In the past, countries tried to avoid stricter regulation and
the ensuing higher costs to industry by building high stacks or locating plants close to bor-
ders. In other words, they externalized their pollution to their neighbouring countries and
benefited from free-riding.
More recent work, however, shows how the occurrence of the prisoner’s dilemma can be
circumvented. In this context, the most influential work has been done by Axelrod and
Hamilton (1981) and Axelrod (1984). Based on computer simulations, they show that in
repeated games the best strategy is ‘tit for tat’, implying that an actor plays whatever her
opponent played in the previous round. Tit-for-tat is a common practice in trade policies
and is even backed by the WTO’s dispute settlement process within multilateral trade inte-
gration (Dluhosh 2016). For example, in summer 2019, the United States imposed tariffs on
exports from India under its Generalized System of Preferences programme. India responded
to the end of this programme by imposing higher retaliatory tariffs on exports from the
United States. Ending duty-free access to the US market aligns with a general change in US
trade policy since 2017, which has been marked by ‘trade wars’ with several countries such
as China and Mexico (see Bouët and Laborde 2018).

Actors and Institutions: Rational Choice Institutionalism

In the previous section on institution-based models, we presented two of the three
Copyright © 2020. Macmillan Education UK. All rights reserved.

approaches to new institutionalism, namely sociological and historical institutionalism.

Now we add the third, which is based on the notion of actors’ making rational decisions.
Rational choice institutionalism is based on the ‘logic of consequentialism’ (March and
Olsen 1989, 2008): that rational actors engage in strategic interactions using their resources
to maximize their utility on the basis of given, fixed and ordered interests (also known as
‘preferences’). In marked contrast to sociological institutionalism, rational choice institu-
tionalism conceives of actors’ preferences as exogenously given: the preferences are not
affected by institutional factors and not subject to social construction.
According to Hall and Taylor (1996: 945), rational choice institutionalism regards politics
as individuals acting to maximize their utility in ways that are likely to produce an outcome
that is collectively suboptimal. Typically, what hinders actors from taking a collectively
superior course of action is the absence of institutional arrangements that would guarantee
complementary behaviour by all concerned. Classic examples of such suboptimal collective
action include the prisoner’s dilemma and the tragedy of the commons.
Accordingly, rational choice institutionalism is not only about actors and their interests,
but also about institutions, as they ultimately structure the opportunities for strategic

choice and interaction. Existing institutional structures restrict or provide opportunities for
policy-making either by affecting the cost–benefit calculations of the actors involved or by
defining a certain distribution of powers and resources between them. In this context, the
concepts of ‘veto points’ and ‘veto players’ are important. A veto point ‘is defined as a politi-
cal arena with the jurisdictional power to veto a government legislative proposal, in which
the probability of veto is high’ (Immergut 2008: 567). The veto player approach is a further
development of the concept of veto points. This approach principally concentrates on the
role of veto players for policy-making. Veto players are most essentially defined as any insti-
tutional or partisan actor whose agreement is required to adopt a policy proposal (Tsebelis
2002).
Most importantly for rational choice institutionalists, institutions affect the strategies
of actors in order to get what they want, though they have no impact on the formation of
their interests, which are conceived as exogenous to institutional analysis. So rational
choice institutionalists focus on interests as the primary, independent explanatory factor:
institutions are considered to be a secondary, intervening factor for explaining political
outcomes (Shepsle and Weingast 1987; Weingast 1996; Shepsle 2008).
Rational choice institutionalists have also developed a distinctive approach to the prob-
lem of explaining how institutions originate (Hall and Taylor 1996: 945). The rules of formal
institutions are explained by reference to the way in which they minimize certain costs, such
as decision, transaction and information costs. Decision costs are incurred when actors try
to come to an agreement, including costs related to bargaining. Transaction costs are
incurred after an agreement is reached, involving costs that ensure compliance with con-
tractual agreements. There are also costs incurred by searching for relevant information for
decision-making (see Jones and Baumgartner 2005: 151–2). These and other kinds of costs
can be reduced by establishing appropriate institutional structures.
As a closing note, institutions have regained currency in policy studies, since they allow
for a comparative assessment of policy formation and policy change. Tosun and Workman
(2017), for instance, demonstrate that this perspective allows for introducing a comparative
dimension to the theories of the policy process (see Weible and Sabatier 2017).

Key points
Copyright © 2020. Macmillan Education UK. All rights reserved.

•• Fully rational behaviour aims to optimize actions, whereas bounded rational behaviour is about
finding a satisfactory solution.
•• Game-theoretic approaches are useful for modelling strategic interactions.
•• The most prominent game-theoretical model is the prisoner’s dilemma.
•• Rational choice institutionalism is based on the ‘logic of consequentialism’, meaning that ratio-
nal actors engage in strategic interactions that are structured by institutional arrangements.

OVERVIEW OF THE MODELS

Table 4.1 summarizes the three perspectives discussed in this chapter, which refer to struc-
tures, institutions and interests. The table outlines the main analytical interests of the mod-
els that fall into these three perspectives.

Table 4.1 Basic theoretical approaches to the study of public policy.

Structure-based models Institution-based models Interest-based models

Cleavage approach: Classic institutionalism: Focuses on the Game theory: Explains
Explains the design and functioning of institutions and how rational actors
formation of political how they shape public policy interact with each other
parties and their in a decision-making
preferences situation
Varieties of New institutionalism: Explains the behaviour Rational choice
Capitalism: Explains of actors within political institutions institutionalism: Focuses
the relationship following a ‘logic of appropriateness’ on rational actors who
between the state and (sociological institutionalism), the pursue their preferences
market actors, development of political institutions that are following a ‘logic of
including the role of subject to path dependence (historical calculation’
business interests institutionalism), and stresses the importance
of ideas and discourse (discursive
institutionalism)

CONCLUSIONS

This chapter offered an overview of the most important theoretical perspectives on policy-
making, concentrating on structure, institutions and actors. The theories introduced help to
explain agenda-setting, policy formation and policy change. What is important to note is
that the theoretical perspectives introduced in this chapter are complementary rather than
competing. They allow for the examination of a certain empirical phenomenon from many
different angles. This also means there are no objective criteria for judging which perspec-
tive produces the ‘most correct’ explanation. However, while this complementarity is desir-
able for fully understanding policy-making, it also implies that looking at one and the same
subject of interest might lead to different conclusions, depending on the analytical perspec-
tive that is taken. Therefore, for the purposes of scientific discussion and the avoidance of
unjustified criticism, the theoretical approach chosen for analysis should always be explicitly
Copyright © 2020. Macmillan Education UK. All rights reserved.

spelled out and motivated by referring to the pertinent literature.

WEB LINKS

www.acm.org/webinar-cybersecurity-and-public-policy. The Association of Computing

Machinery produced a webinar that explains the issue of cybersecurity and identi-
fies needs for policy-making.
www.iterated-prisoners-dilemma.net. This web page offers an interactive online
game of the prisoner’s dilemma, consisting of several rounds. The game is useful
for demonstrating the effects of repeated interactions and the tit-for-tat strategy
for the evolution of cooperative behaviour.
www.kwawazee.ch. This is an NGO implementing pilot programmes in Tanzania to
experiment with different types of welfare policies.

www.newinstitutionalism.org. An international platform for scholars and students

interested in new institutionalism.
www.oecd.org/education. The OECD provides excellent data on education systems,
which according to the Varieties of Capitalism approach are affected by the type
of capitalism.

FURTHER READING

Cairney, P. (2020). Understanding Public Policy: Theories and Issues. Basingstoke:

Palgrave Macmillan. The book provides an overview of the most important theo-
ries in public policy.
Ho, S. (2019). Thirsty Cities – Social Contracts and Public Goods Provision in China and
India. Cambridge: Cambridge University Press. This book sheds light on infor-
mal institutions and how these affect formal institutional design, which in turn
explains the provision of public goods.
Howlett, M. and Mukherjee, I. (eds.) (2018). Routledge Handbook of Policy Design.
London: Routledge. This edition offers a complementary perspective on policy
theories by concentrating on questions of policy design.
Lowndes, V. and Roberts, M. (2013). Why Institutions Matter: The New
Institutionalism in Political Science. Basingstoke: Macmillan International Higher
Education. Provides a valuable complement to the material presented in this
chapter, since the book elaborates on a wide range of themes pertinent to the for-
mation, functioning and change of institutions.
Weible, C. and Sabatier, P. (eds.) (2017). Theories of the Policy Process. New York:
Routledge. This is the fourth edition of the authoritative volume.
Zohlnhöfer, R. and Rüb, F. (eds.) (2016). Decision-Making under Ambiguity and Time
Constraints: Assessing the Multiple-Streams Framework. Colchester: ECPR Press.
A stimulating collection of conceptual and empirical discussions of the Multiple
Streams Framework.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Reader’s guide
This chapter addresses the initial moments of the policy-making process, namely
problem definition and agenda-setting. Most citizens and organizations possess some
concerns – real or perceived – that they believe merit government attention. Our first
goal in this chapter is to show that problem definition is subject to different interpre-
tations, which also entails that there is competition over how a problem is defined.
Although there are many problems, in the end only a small number of them will be
given attention by legislators and executives. Problems chosen by decision-makers for
further consideration constitute the policy agenda. Our second goal is to outline the
dynamics by which ideas, policy proposals and new understandings of problems are
formulated. We will learn that the possibility of influencing the policy agenda is an
important source of power, since decision-making grants an advantage to those who
address a problem and propose a solution to it first. Several groups of actors compete
with one another in order to set the agenda in accordance with their respective policy
preferences.

INTRODUCTION
How are issues selected for government attention? Why are some issues taken up while oth-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ers are left unconsidered? Which actors are involved in defining policy problems and bring-
ing them onto the agenda? These are the questions guiding the discussion in this chapter on
the first stage of the policy-making process, namely problem definition and agenda-setting.
Fundamental to these two processes is a focus on the dynamics by which new ideas and
understandings of social problems emerge. Although problem definition and agenda-setting
are in reality closely interlinked, for analytical purposes we will discuss these issues
separately.
We will begin with the stage of problem definition. Here we not only show that the way a
problem is defined has important repercussions for the subsequent decision process, but
that current and differing factors influence the definition process. While this process
appears quite intuitive, in many cases conflicts emerge as to how a given issue should be
defined and whether it actually needs to be addressed by policy-makers.
We then turn to the stage of agenda-setting. Put simply, this stage is about getting an
issue onto the ‘agenda’. Here our analysis starts to become problematic already, as there
exist many different agendas. However, we can make a basic differentiation between

77
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:19:51.
78 Public Policy

systemic and institutional agendas (Cobb and Elder 1983). The former consists of all issues
that are commonly perceived as meriting public attention, whereas the latter is composed of
those issues that are clearly of concern to decision-makers. Having defined in more detail
the various forms of agendas, we then introduce the theoretical approaches to agenda-
setting that can be found in the relevant literature, and finally shift our attention to the
actors who are involved in the agenda-setting process.

PROBLEM DEFINITION: THE PATH TO AGENDA-SETTING

Problem definition can be understood as a causal story which (1) identifies harm, (2)
describes what causes it, (3) assigns blame to those causing it and (4) claims that the govern-
ment is responsible for stopping it (Stone 1989: 282). Problem definition is consequential
for policy-making since the way a problem is defined to a certain extent determines the
feasible set of policy responses through which the problem can be resolved (Hogwood and
Peters 1983; Dery 1984). For instance, it makes a big difference whether unemployment is
conceived of as a problem in terms of a lack of education, an economic recession or a lack of
willingness to look for a job. Depending on the definition used, solutions might vary consid-
erably, from improving the public funding of the education system, to intervening at a
macro-economic scale to stimulate economic growth, to cutting unemployment benefits in
order to motivate individuals to look for a job.
This example illustrates the fact that in most cases there is no ‘obvious’ conception of a
problem. Different actors might have different perceptions of the same empirical phenom-
enon. In this regard, it generally holds that the actor proposing the first definition of the
problem has a first-mover advantage, since subsequent normalizing of a ‘deviant’ definition
will require considerable effort by competing actors. This effort stems from the fact that not
only must another and equally persuasive definition of a problem be presented, but also the
first definition needs to be discredited, which its proponents will oppose. Depending on the
outcome of this process, we may find highly different policy approaches across countries,
even though the objective characteristics of the problem are very similar (for the opposite
scenario, policy convergence, see Chapter 11). However, the definition of a problem is not
Copyright © 2020. Macmillan Education UK. All rights reserved.

only affected by the interests and resources of the involved actors, but also by the distinctive
characteristics of the problem that render its definition in a certain way more or less likely.
In the following we will first take a closer look at the aspects of the process that character-
izes problem definition, before we discuss the relevance of the different characteristics that
problems can possess.

The Process of Problem Definition

Problem definition is essentially characterized by two features: objective data (e.g. poverty
levels) and the question of whether and to what extent these objective criteria are actually
acknowledged. To be sure, problem definition is not about a neutral realization of the exis-
tence of problems, but rather about their ‘social construction’ (Lindblom and Cohen 1979;
Wildavsky 1979; Dery 1984, 2000; Fischer 1993; Bacchi 2009). In fact, a ‘problem’ might
only exist because people think that it exists. Studies of risk regulation, for instance, have
revealed that there tends to be a significant difference in how experts assess the existence

and severity of a risk (e.g. health and environmental risks related to genetically modified
plants) and the public perception of it (see Kahneman et al. 1982; Sunstein 2005). Many
studies have observed that ‘lay people’ tend to overestimate risk in comparison with experts.
More precisely, the public tends to understand risk as danger and to relate risk to risk fac-
tors such as toxins or an unhealthy diet (Wöhlke et al. 2019).
This split in risk perception between experts and lay people has important implications
for policy-making, since public demand might lead to the initiation of a policy-making pro-
cess despite the absence of a ‘genuine’ social problem that needs to be solved. Maor (2012:
231) refers to such a situation as a policy over-reaction, which is ‘“too much” and/or “too
soon” patterns of policy’ (see also Gillard and Lock 2017; Howlett and Kemmerling 2017;
Maor et al. 2017; Peters et al. 2017).
In situations in which it is uncertain whether or not a social problem really exists, policy-
makers may still prefer to address it. There are two main reasons for this. First, even if a
certain social problem may seem ‘irrational’, it can still result in a loss in social welfare. For
example, if people are afraid of eating beef treated with growth hormones, this fear should
be taken seriously by policy-makers as it could lead to a lowering of beef consumption,
which, in turn, would reduce the income of farmers (see Chang 2003: 7–8) and require spe-
cific policies to support farmers by subsidies – regardless of whether or not growth hor-
mones are likely to impact on the population’s health. Secondly, policy-makers who do not
address an issue that is broadly perceived as a social problem run the risk of not being re-
elected or even being forced to resign. For example, in 2013, Bulgaria’s government had to
resign after mass protests against high electricity prices – an issue the government had not
attended to.
That said, governments can equally be responsible for creating such ‘problems’. According
to Bacchi (2009), unemployment through a lack of education and training, for instance, can
be conceived of as a government-induced problem, with the solution ready to hand in the
form of life-long learning to continue the education and training of adults. This logic of rea-
soning ties in with the Multiple Streams Framework, which acknowledges that ‘solutions’
may be looking for ‘problems’ (Kingdon 2003/2011; see also, e.g., Cairney and Zahariadis
2016; Herweg et al. 2017).
While the previous explanations give the impression that the problems addressed by
Copyright © 2020. Macmillan Education UK. All rights reserved.

policy-makers might be inflated, there is also the opposite dimension to problem definition,
i.e. that policy-makers refuse to see an issue as a problem that warrants a policy response
(Crenson 1971; Lukes 1974). While the United Kingdom, similar to many other states, has
experienced poverty and inequality, these have only sporadically been defined as policy
problems (Dorey 2005: 11). One reason for this is that fighting poverty is a challenging
endeavour as well as one that attracts considerable attention when the endeavour fails. If a
problem is (for strategic reasons) not recognized as one, it will also be denied access to the
policy agenda (for an overview, see Rochefort 2016).
How can a social problem – real or perceived – turn into a political problem? First, this
can only happen if the social problem can potentially be addressed by public policies
(Wildavsky 1979: 42). Earthquakes, for example, do not constitute a political problem, as
governments cannot affect the likelihood of their occurrence. The picture is different, by
contrast, if we focus on disaster management, which of course does fall within the realm of
public policy (on this, see Boin et al. 2016). This argument, however, does not eliminate the
possibility that there are also political problems that can only be solved temporarily and in

an imperfect manner. Rittel and Webber (1973) refer to these as ‘wicked’ problems; they
entail conflicts that are value based and difficult to define and that resist resolution (see Box
6.2).
Second, turning a social problem into a political problem is characterized by conflict.
According to Schattschneider (1960), politics is a means to resolve conflicts about how to
govern a community. Thus, the organization of conflict is important and is achieved through
the scope (who is involved) and the bias (how the audience is involved) of the conflict. The
interplay of these two dimensions determines what happens. Schattschneider deems the
scope of conflict to be more decisive for the outcome than the bias, and thus a rational
policy-maker should address the scope dimension. This can occur through ‘privatization’, i.e.
restricting the scope of the conflict to its current boundaries, or through ‘socialization’, i.e.
appealing to a previously uninvolved group of bystanders and inviting new actors. In this
context, interest groups and – and even more so – political parties play a key role in privatiz-
ing and socializing conflict.
In this regard, Cobb and Coughlin (1998: 417) distinguish between two types of actors.
The first is the so-called expander, who seeks to publicize the issue to different societal
groups in order to get them involved in the policy debate. In this regard, the expander can
be viewed as an innovator, since he or she brings up an issue that is not currently under
consideration by decision-makers. However, the content of this issue will be evaluated by
other individuals – ‘containers’ – who might be adversely affected if the issue enters the
agenda stage and eventually becomes a law. These containers try to prevent the expanders
from reaching the agenda stage. If they are successful, then no action will be taken and the
status quo will prevail. Accordingly, we can characterize the problem-definition process as a
battle between expanders and containers, in which the expanders must redefine the issue so
that the public, who previously did not care about it, now become concerned. Each of the
two groups tries to frame the issue for the public and politicians in a way which makes action
or inaction likely, according to their interests – which indicates the importance of public
rhetoric (Riker 1986).
However, it is not only persuasion from expanders and containers that matters, but also
the interpretation of these activities against the background of dominant beliefs. As a result,
the status quo of a given social perception can be a powerful force making it difficult to move
Copyright © 2020. Macmillan Education UK. All rights reserved.

people away from the prevailing perception of a condition as not being problematic.
Nevertheless, occasionally forces may emerge that produce movement towards a new social
perception.
Against this background, Wood and Vedlitz (2007) illustrate how problem definition is
affected by both stable and dynamic factors. First, the manner in which an individual defines
an issue depends on a set of relatively fixed factors such as values. Second, the more uncer-
tain is the information environment surrounding an issue, the more likely it is that an indi-
vidual will rely on these established values than on new information that becomes available
for defining the issue. Conversely, the more trust individuals possess in the existing infor-
mation surrounding an issue, the more likely they are to define the issue in a manner con-
sistent with that information rather than redefining it in the light of new evidence. In this
context, trust in information increases if, for example, it is supplied by epistemic communi-
ties (see Chapter 10). Third, an information signal of sufficient strength can produce shifts
in definitions of specific issues, regardless of the individual’s predispositions and knowl-
edge. Information signals producing such redefinitions can relate directly to the issue itself,

Box 5.1 Perceptions of reasons for vaccine hesitancy

The World Health Organization regards refusal to vaccinate despite the availability of
vaccines as a threat to the progress made in reducing vaccine-preventable diseases. A
prominently discussed example in this context is measles, which is a very serious and
highly contagious disease caused by a virus and accompanied by a widespread skin
rash. While many countries were close to eliminating measles, it has seen a resurgence
and an increase of infections by 30% at the global level. While the vaccine for measles
is very safe, there has been a misperception in online communication that there is a
link between childhood immunization and autism (Edy and Risley-Baird 2016). This
raises the question of what measures should be taken in order to overcome vaccine
hesitancy, which, however, depends on how this specific problem is perceived. If vac-
cine hesitancy is interpreted as a problem of knowledge deficit, an apt measure to
increase vaccine rates could be to launch information campaigns. But if vaccine hesi-
tancy is considered a problem of public mistrust of scientific experts and institutions,
the measures to be adopted for increasing vaccine rates will have to be different. The
focus then has to be on increasing the credibility of scientific and governmental bod-
ies and investment in outreach to vaccine opponents (Goldenberg 2016).

or they can somehow alter an individual’s interpretation of the social forces surrounding the
issue. In this context, the mass media plays a key role, since it represents the main means of
communication and persuasion (Jerit 2006; Box 5.1).
For example, the influence of the mass media became particularly apparent with the
nuclear emergency that Japan faced in the spring of 2011. Reacting to the comprehensive
media coverage of this incident, a global debate about the perils of nuclear power was
started. In Germany, for instance, this particular event increased the general level of atten-
tion paid to nuclear power policy, strengthened the position of advocates and further dimin-
ished support for nuclear power. In Japan where the accident had happened, public attention
Copyright © 2020. Macmillan Education UK. All rights reserved.

initially centred on disaster management doubts regarding the capacity of governmental

authorities to provide disaster relief. Only later was nuclear power called into question and,
in contrast to Germany, the accident did not result in a phasing out of nuclear power
(Rinscheid 2015).

The Characteristics of Problems

The definition of a problem is not only affected by the interests and beliefs of different
actor coalitions, but also by the specific characteristics of a problem. Again, these char-
acteristics can be socially constructed. Some attributes of a problem may be highlighted
as they lend themselves to a particular solution, while other attributes are downplayed.
Such a specific way of presenting problems is commonly referred to as ‘framing’. It
‘involves the selective use of knowledge and information about a problem and the causal
relationships surrounding it, to give it meaning and render it manageable’ (Ward et al.
2004: 92). It is about emphasizing certain aspects of a problem and de-emphasizing

others. Frames are strategically employed by actors (e.g. interest groups) as a key way to
influence the policy-making process (Riker 1986; Stone 1988; Rochefort and Cobb
1995; Baumgartner et al. 2008, 2009a; Baumgartner and Jones 2009; Eising et al. 2015;
Klüver and Mahoney 2015).
Rochefort and Cobb (1995: 15–23) have advanced a series of categories which influence
how a problem can be framed. The seven categories refer to causality, severity, proximity,
‘crisis character’, incidence, novelty and problem populations. The category of causality
deals with the question of which factors or actors have caused a problem in the first place. In
many instances, issues of culpability and allocation of blame play an important role in this
context. For many problems, however, the underlying causes are not easy to diagnose. The
political debate about problem causes can therefore constitute a highly contested area of
problem definition. An illustrative example is given by the determinants of state debts and
decreasing tax revenues. Usually, excessively low or high tax levels lead to low revenues.
Depending on which of these two views is supported, the policy proposals will differ
substantially.
Severity concerns how serious an issue and its consequences are perceived to be. The
more a problem can be framed as severe and serious, the more it will capture the attention
of decision-makers and the mass media. Often, different actors have different views regard-
ing the severity of a problem, depending on their interest positions. The severity of nuclear
accidents, for instance, will be judged very differently by the affected population on the one
hand and the power company on the other.
In order to increase the severity perception of an issue, expanders might try to attach the
‘crisis’ label to it (see Boin et al. 2016). In other words, strategic actors can combine two
framing strategies. Labelling a situation as a crisis signals the public that a problem requires
urgent action. With some situations (e.g. epidemics or pandemics) this label is straightfor-
ward, whereas with others the ‘crisis image’ must be actively ‘constructed’ (see, e.g., Van
Wijk and Fischhendler 2017).
Proximity implies that the issue is directly or indirectly affecting the personal interests of
a broad public. If a problem can be defined in such a way that a broad range of people are
potentially affected, the chances of political mobilization are higher. For example, to increase
problem proximity, it is more reasonable to link a cut in the salaries of teachers to a reduc-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tion in the quality of education than to highlight the disapproval of this measure by the
teachers’ union.
Incidence refers to the actual prevalence of the problem. How many actors are potentially
affected? In this regard, the expanders must show that the problem is of huge scope and
causing misery for many people, and argue that the problem is getting worse. Although many
societal problems are persistent, their perception as prevalent is likely to change. It is natural
to think in terms of thresholds here – up to a given critical value the public’s attention to a
certain issue is negligible, but once this value has been reached public pressure on policy-
makers becomes too strong and they must act.
Novelty as the sixth category is also an important device for winning attention. Labelling
an issue as something new and unprecedented can result in wide media coverage. This was
the case with the Coronavirus outbreak in China in late 2019 and in other world regions in
early 2020 and the drastic policy responses to it, which were everywhere in the news.
Another example is provided by the promotion of nanotechnology in the United States.
Lindquist et al. (2010) argue that the strategy of the supporters of nanotechnology, by

emphasizing its novelty, was an important factor in attracting media attention. However,
when using this strategy one has to be careful: issues that have not yet been experienced are
difficult to tackle since they lack familiar solutions and therefore policy-makers might refuse
to pick them up. Thus, novelty has an ambivalent impact on problem definition.
The seventh and final category concerns the characteristics of a problem popula-
tion. It deals with how the image of target groups is manipulated by contending par-
ties. The expanders seek to place the target group in a positive light. This can, for
instance, be done by portraying the problem population as helpless. To gain public
support for the controversial Economic Opportunity Act of 1964, which included vari-
ous social programmes to promote the health, education and general welfare of the
poor, the US government constructed the target population as ‘hopeless victims’.
With this understanding, the act made sense as one which would provide tools for
ending poverty. Thus, a very particular image was created to attain Congressional
consent for the act. Consistent with the views of who the poor are, the goal of the act
was framed as giving new hope to the hopeless (see Guetzkow 2010: 175–83). A prom-
ising strategy for the containers would have been to emphasize that the poor should
not be regarded as ‘victims’ but as people who took advantage of the possibilities
already given to them. These considerations suggest that a helpful way of thinking
about political problems is to illuminate how they are linguistically constituted (see
Fischer 2003: 46; Box 5.2).

Box 5.2 Sexual harassment as a policy problem

One topic that has certainly become more salient over the last two decades is sexual
harassment. In their study, Wood and Doan (2003) examine the declining tolerance
of sexual harassment in the United States. Their empirical illustration refers to the
case of Judge Clarence Thomas of the US Supreme Court. In 1991, President George
Bush nominated Thomas to the Supreme Court. On this occasion, the Senate
Judiciary Committee examined charges by a law professor at the University of
Oklahoma, Anita Hill, who accused Thomas of sexual harassment. Testimony and
Copyright © 2020. Macmillan Education UK. All rights reserved.

debate on the charges were broadcast nationwide on television. Prior to these hear-
ings, widespread social tolerance prevented most individuals from active and open
opposition to sexual harassment. With the hearings emphasizing the costs of sexual
harassment and the negative tone of the media coverage following the hearing, indi-
vidual perceptions of the phenomenon started to change, leading to a new social
interpretation of the issue. While the accusation did not prevent Clarence Thomas
from receiving a seat on the US Supreme Court, US policy-makers felt that they had
to react to the increasing public disapproval of sexual harassment and immediately
adopted legislation that resulted in an increased number of charges. The topic of
sexual harassment entered the discussion agenda when in mid-2017 the #MeToo
campaign started on the social media platform Twitter. Supported by US celebrities,
both male and female, #MeToo has become an international movement against sex-
ual harassment and assault. It has drawn enormous attention to this issue and
immediately resulted in changes in company policies. Whether public policies will
follow and what exactly the policy consequences will be remain to be seen.

Overall, the analysis of the specific characteristics of policy problems represents a prom-
ising research strategy to better understand the process of problem definition. While the
categories outlined above already provide a framework for insightful research, the analytical
approach of Bacchi (2009) further helps to explain problem definition. This approach stimu-
lates critical thinking about this step of the policy-making process, and is achieved by fram-
ing the problems as questions:

• What is the problem to be represented by a specific policy?

• What presuppositions or assumptions underlie this representation of the problem?
• How has this representation of the problem come about?
• What is left unproblematic in this problem representation? Where are the silences? Can
the problem be thought about differently?
• What effects are produced by this representation of the problem?
• How has this representation of the problem been produced, disseminated and defended?
How could it be questioned, disrupted and replaced?

Key points
•• The way in which a policy problem is defined is consequential for all subsequent policy stages.
•• Problem definition is not predominantly about pointing to the existence of an objective
issue, but is an endogenous construct and thereby the subject of social conflict between
supporters and opponents.
•• Public attention can be increased or decreased by framing, which can refer to the causality,
severity, proximity, ‘crisis character’, incidence, novelty and problem populations for any
particular policy problem.

ANALYTICAL CONCEPTS OF AGENDA-SETTING

The fact that a certain societal problem is perceived as a problem that needs to be addressed
by the government is only a necessary not a sufficient condition for political action. Not
every political problem ends up on the agenda. Policy-making generally occurs in the pres-
ence of multiple constraints, e.g. shortages of time and resources (see, e.g., Bevan 2015). As
a consequence, there are always more public problems warranting action than there is space
on the policy agenda. The central research question with regard to the agenda-setting stage
hence concerns the conditions that affect whether a public problem actually reaches the
policy agenda or not.
To answer this question, varying models have been developed that analyse agenda-setting
using different analytical lenses. Basically, the diverse literatures can be grouped into four
approaches that focus on different aspects of agenda-setting phenomena and that empha-
size process patterns, power distribution, institutional factors and contingency. However,
before we enter the theoretical discussion, we will first have a closer look at different agenda
types, as agenda-setting can refer to quite different phenomena which entail different theo-
retical explanations and perspectives.

Agenda Types
Agenda types are usually defined in accordance with Cobb and Elder’s (1983) distinction
between the ‘systemic agenda’ and the ‘institutional agenda’. The systemic agenda refers to
all societal problems that demand public attention and form the discussion agenda. Most
items on it are general and still lack a precise definition. Systemic agendas exist for all levels
of the political system: national, state and local. Some items may appear on multiple sys-
temic agendas, while others will appear on only one.
The institutional agenda contains a set of problems that are up for serious consideration
by decision-makers. The institutional agenda is the action agenda, which is more specific
and concrete than the systemic agenda. Remarkably, most issues on the institutional agenda
are hardly recognized by the general public, since they predominantly require detailed, tech-
nical knowledge about the already existing public policy. To be sure, some issues will attract
much attention from the general public, but usually the number of highly salient issues is
rather limited. Similar to the systemic agenda, there is a multitude of institutional agendas
referring to the different levels of government – i.e. the national, state and local institu-
tional agendas – and within these levels there is a further distinction amongst branches –
i.e. the legislative, executive, administrative and judicial institutional agendas (Anderson
2010: 90–3). Only if a public problem reaches one or more institutional agenda(s) will it be
considered as needing a solution.
Taking into consideration Kingdon’s (2003/2011) Multiple Streams Framework, we fur-
ther suggest differentiating between the ‘drafting agenda’ and the ‘decision agenda’. The
drafting agenda – which Kingdon terms the ‘governmental agenda’ – is a list of subjects that
are getting attention within government. As will be shown in Chapter 6, the adoption of a
policy in a first step requires the development of policy proposals. The government needs to
take a basic decision that a certain policy problem has to be addressed and start drafting
potential measures that relate to the problem. Those issues for which such a decision has
been taken constitute the drafting agenda. The decision agenda, by contrast, is based on
those issues for which the government has agreed on a draft proposal and hence has decided
to put the issue on the agenda of the responsible decision-making body (usually the legisla-
ture), which has to take the final decision of whether or not to adopt the policy.
Copyright © 2020. Macmillan Education UK. All rights reserved.

In the literature there is a major debate about the extent to which the setting and shaping
of the decision agenda might affect the chances of policy adoption. It is generally acknowl-
edged that those actors who have the power to influence the decision agenda enjoy a power-
ful position in the decision-making process (see Anderson 2010: 90–3).

The Process Perspective

Process-based assessments all focus on general patterns characterizing agenda-setting
dynamics. Agenda-setting is hence described and explained from a more abstract perspec-
tive by focusing on systemic factors that shape political processes. The most influential
descriptive account of agenda-setting processes has been developed by Cobb et al. (1976),
who differentiate between three process models: (1) the outside-initiative model, (2) the
mobilization model and (3) the inside-access model.

The outside-initiative model describes a situation in which an individual actor or a group of

actors outside the governmental structures initiates an issue for reform which they then
seek to expand into the public domain. This can take place through various means: media
coverage, lobbying of politicians or other forms of public action which pressure decision-
makers to place the issue on the institutional agenda. In this context, the dynamics of the
public attention paid to issues are of key relevance.
The mobilization model describes a process of agenda-setting where policy-makers move
issues from the systemic to the institutional agenda in order to muster the support needed to
attain their objectives. In contrast to the outside-initiative model, the mobilization model
takes into account the possibility of top-down agenda-setting, as issues are placed on the insti-
tutional agenda directly by decision-makers. While there may have been extensive debates
about the issue in questions amongst decision-makers prior to the announcement, the public
may have little knowledge of the issue. This process is often triggered by a systems failure or
exogenous shock – also known as ‘focusing events’ (see Birkland 2006) – followed by the com-
missioning of a panel of experts. The decision-makers may then formally adopt the experts’
recommendations in the decision process. Hence, this model takes into account the fact that
expert opinions can make an important contribution to constructing agendas. After the initia-
tion of the issue, decision-makers then elucidate it and explain what they seek from the public.
Mobilization of the issue within the public domain is necessary in circumstances where the
implementation of a policy requires widespread voluntary acceptance. Decision-makers seek
to expand an issue from the institutional to the systemic agenda in the same manner as pro-
ponents seek to attain a formal agenda under the outside-initiative model.
In contrast to the two previous models, the inside-access model describes a pattern of
agenda-setting that seeks to exclude public participation. Under this model, the initiators
seek to place an issue on the institutional agenda. If the issue is expanded into the public
domain, it is usually limited to a few selectively influential groups who are considered impor-
tant to the passage of the policy. So we are mainly confronted with a situation in which the
initiators seek a ‘private’ decision within government, who are ‘insiders’ such as a govern-
ment department or a regulatory agency (Box 5.3).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Box 5.3 Issue attention cycle

What are the characteristics of the public perception of ‘crises’ or ‘problems’? This
question underlies Downs’ (1972) seminal concept of the dynamics of the ‘issue atten-
tion cycle’. The main argument is that public perception does not reflect real condi-
tions as much as it reflects the operation of a cycle of increasing interest and then of
rising boredom with issues. This cycle has five stages, which may vary in duration
depending upon the particular issue concerned.

1. The pre-problem stage: some undesirable social conditions exist, but at this stage
they only capture the attention of experts or interest groups.
2. Alarmed discovery and euphoric enthusiasm: for certain reasons (e.g. accidents, riots,
epidemics) the public becomes alarmed about a particular problem. This discovery
of a problem is accompanied by a euphoric enthusiasm about society’s ability to
solve it within a relatively short time.

3. Realization of the costs of significant progress: the public gradually realizes the high
costs of solving the problem. These costs are not only monetary, but also refer to
sacrifices some groups in the population would need to make.
4. Gradual decline of intense public interest: as more and more people realize the costs and
other difficulties of solving the problem, they get discouraged. Some just suppress
thinking about it; others become bored by the issue. Additionally, by this time
another issue has usually entered the stage of alarmed discovery and euphoric
enthusiasm.
5. Post-problem stage: an issue moves to receiving less attention, but because it has
gone through the cycle it almost always receives a higher level of attention than in
the pre-problem stage. This results from the new institutions and policies that may
have been created to help solve it. These entities often persist and can have a long-
lasting impact on public attention. Another reason for the higher average level of
attention might stem from the success of the institutions and policies in address-
ing some aspects of the problem.

The Power Distribution Perspective

While the process model described above provides an important starting point for assessing
how processes of agenda-setting unfold, it cannot explain why certain issues are moved
onto the agenda while others are not. To address this, existing theories emphasize in par-
ticular the distribution of power between the involved actors.
The origins of research into agenda-setting relate to the pluralism–elitism debate of the
1950s and 1960s. The key aspect of this debate concerned the notion of power, especially its
use and distribution in society (see Truman 1951; Schattschneider 1960; Dahl 1961; Polsby
1963). Regarding agenda-setting, the debate asserted that different groups of actors seek to
increase the probability that an issue will receive collective attention by attempting to raise
an issue’s salience and/or its support. In a similar vein, but from an inverse perspective,
Copyright © 2020. Macmillan Education UK. All rights reserved.

Bachrach and Baratz (1962) argued that the political power of decision-makers could be
exercised through banning from the agenda certain issues, i.e. as ‘non-decisions’. In other
words, they could use tactics to decrease the probability of agenda inclusion for an issue,
thus helping to preserve the status quo (Gerston 2004: 67).
Schattschneider (1960) emphasized that, rather than the actors actively participat-
ing in agenda-setting, it is the wider public domain which is key to understanding poli-
tics. In this regard, Schattschneider argued that strategically it is in the best interest of
the contestant who is losing a battle over an issue to bring more and more fence-sitters
into the conflict, thus socializing them, until the balance of forces changes. Conversely,
it is advantageous for the one who is winning to contain the scope of the conflict so as
not to upset this favourable balance of power. Hence, issue battles are frequently won or
lost over the combatants’ success at either getting the public or other bystanders
involved or excluding them. Baumgartner and Jones (1993/2009) refer to the scope of
conflict as policy images that are a mixture of empirical information and emotive
appeals (Jones and Mortensen 2017).

The importance of redefining or reframing issues is illustrated by Sheingate’s (2000)

analysis of agricultural retrenchment in the United States and the European Union (EU).
The author argues that advocates of retrenchment must redefine the issue of agricultural
subsidies in a manner that highlights the negative externalities associated with farm policy.
Second, they must also exploit opportunities for strategic changes of venue, so that policy
decisions in agriculture do not rest solely with those who benefit from the status quo. In the
terminology of Baumgartner and Jones (1993/2009), this corresponds to ‘venue shopping’.
A similar argument is put forward by Daugbjerg and Studsgaard (2005) in their analysis of
radical agricultural policy reforms in Sweden and New Zealand in the 1980s and 1990s. The
authors show that the redefinition of agricultural policy – from constituting a matter of
finding the balance between budgetary costs and farmers’ income to considering agricul-
tural policy as a part of macro-economic policies – was an important factor in venue change
and eventually successful reform. In a similar vein, Pralle (2006) illustrates how Canadian
anti-pesticide activists mobilized otherwise apathetic members of the public by changing
the image of garden pesticides from useful and seemingly harmless to superfluous and
dangerous.
An additional insight concerning issue definition is presented by Timmermans and
Scholten (2006), who deal with the different roles that scientific knowledge can play in
shaping policy images. Focusing on immigrant policy and assisted reproductive technol-
ogy policy in the Netherlands, they show how science and the structural arrangements
through which it is produced and disseminated can aid depoliticization – or, on the con-
trary, fuel emerging policy disputes (see also Fischer 2006). Drawing from the analysis of
the media discourse on forest fires in Indonesia, Ekayani et al. (2016) argue that theimpact
of scientists depends on their active involvement with media, which complements the
argument about the role of scientists in agenda-setting put forward by Timmermans and
Scholten (2006).

The Institution-Based Perspective

In addition to the ways in which an issue can be mobilized, early theories of agenda-setting
Copyright © 2020. Macmillan Education UK. All rights reserved.

also emphasized the role of institutional structures in defining the lines of conflict and the
balance of power: ‘all forms of political organizations have a bias in favour of the exploita-
tion of some kinds of conflicts and the suppression of others because organization is the
mobilization of bias. Some issues are organized into politics while others are organized out’
(Schattschneider 1960: 719).
One of the best-known theoretical approaches that takes into account institutional
arrangements is the Punctuated Equilibrium Theory, developed by Baumgartner and Jones
(1993/2009) and True et al. (2007). We have already referred to some components of the
model, but when looking at the model as a whole, we believe that its strong institutional
dimension makes it an institution-based approach. In fact, this institutionalist logic is the
reason why the Punctuated Equilibrium Theory has been embraced by comparative politics
and why it was successfully developed into a body of comparative research (Tosun and
Workman 2017).
Essentially, this framework seeks to explain why political processes are generally char-
acterized by stability and incrementalism – i.e. by minimal adjustments to existing policy

arrangements (see Lindblom 1959, 1979) – but occasionally produce fundamental shifts
(labelled ‘policy punctuations’). According to this approach, both policy stability and pol-
icy change are mostly determined by the presence of policy venues and policy images.
Policy venues are institutional arenas in which decision-making takes place. Policy images
are the shared views of policy communities concerning the characteristics of a given pub-
lic problem and the ways of solving it. It is through agenda-setting that policy images
become changed. As a rule, this happens when a new policy image is received well by the
actors of a new policy venue. What bears the highest potential to affect both policy images
and venue promotion is an external shock, such as the Japanese nuclear accident in the
spring of 2011 (see, e.g., Rinscheid 2015), the global credit crunch in 2008–2009 (see,
e.g., Aliber and Zoega 2019) or the 2015 asylum and migration crisis in Europe (see, e.g.,
Scipioni 2018).
The influence of institutional rules is of particular relevance when it comes to the
specification of the decision agenda. The question of who sets this is determined by the
institutional rules that characterize the legislative process. Actors who have the power
to shape the decision agenda are more likely to have their proposal adopted than those
whose influence is basically confined to reacting to policy. This can be traced to the fact
that agenda-setters, in anticipation of the preferences of the other actors, can frame a
proposal in such a way that it is more likely to be adopted. This aspect was first discov-
ered by Shepsle and Weingast (1987), who showed that institutionally determined dif-
ferences in agenda-setting power implied that certain committees of the US Congress
were more influential than others when it came to getting their proposals accepted.
This institutionalist perspective on agenda-setting is taken by agenda-setting studies
rooted in comparative politics, which concentrate on the role of institutional veto point
or veto players (see, e.g., Rasch and Tsebelis 2013). The institutionalist perspective
also facilitates a more differentiated analysis of the role of political regimes, which was
already one of the goals of Cobb et al. (1976) when they put forward their three process
models.

The Contingency Perspective

In contrast to models that emphasize the central explanatory role of the distribution of
power between the involved actors and institutional arrangements, contingency models
conceive of agenda-setting as a process that is affected by chance rather than rational calcu-
lation and the characteristics of politics. The central argument is that policy-making is basi-
cally about the coupling of policy problems and policy solutions, which are considered as
evolving independently of each other. This view also underscores the arguments presented
in the section on problem definition about the social construction of problems by the public
as well as policy-makers. On the one hand, within political systems there is a continuous
discussion over policy problems and how to define them; on the other hand, at the same
time, there is an ongoing debate and development of solutions which take place rather
autonomously from the definition of the problem. It can even be the case that solutions are
developed for problems that do not yet exist or are perceived not to exist. This is the basic
argument of the garbage can model (Cohen et al. 1972). These solutions are then in need of
problems to which they can be applied.

According to this approach, it is only when pre-existing solutions can be successfully

coupled with perceived problems that an issue is put on the agenda. As argued by Kingdon
(2003/2011), the inventor of the Multiple Streams Framework, the extent to which cou-
pling is possible is affected not only by independent streams of problems and proposals,
but also by the politics stream. The latter refers to events such as elections, changes in
government or the emergence of new social movements. The three independent streams
of problems, proposals and politics need to come together at a critical time in order to
open up the famous ‘policy window’, resulting in an issue being put on the agenda. The
process of coupling is not considered as something that can be influenced by strategic
behaviour, but as a highly contingent phenomenon. Yet it should be noted that newer
conceptual developments of the Multiple Streams Framework – aside from offering more
differentiated models, such as Howlett et al. (2015) who emphasize interactions between
the streams and the corresponding outcomes – elaborate on the possibility of strengthen-
ing the role of institutions in explaining agenda-setting. In this context, Zohlnhöfer et al.
(2016) acknowledge that the original conceptualizations of the framework by Kingdon
(2003/2011) and Zahariadis (2007) attribute institutions’ relevance to the extent that
they allow for a more efficient processing of information and defining of the position and
influence of policy entrepreneurs who can couple the streams and open the opportunity
window. Furthermore, the authors contend that the politics stream is essentially about
institutions. However, to make the institutional dimension of the Multiple Streams
Framework more prominent, Zohlnhöfer et al. (2016) propose to differentiate between
coupling for agenda-setting and coupling for decision-making, which stands in contrast to
the dominant approach of collapsing both coupling processes (see Zahariadis 2003). By
adopting this differentiated perspective, the role of institutions becomes evident, espe-
cially but not exclusively for decision-making. As shown above, Kingdon (2003/2011)
himself distinguishes between the agenda-drafting and the decision agenda, which makes
this proposal intuitive.

Summary
The agenda-setting models discussed share a number of similarities, while at the same time
Copyright © 2020. Macmillan Education UK. All rights reserved.

they focus on different aspects. Table 5.1 summarizes the main features of the four models
and discusses their respective strengths and weaknesses for analysing agenda-setting.

Key points
•• We can analytically distinguish between different agenda types, including the systemic (public)
and the institutional (formal) agenda (Cobb et al. 1976), and the drafting and the decision agenda
(Kingdon 2003/2011).
•• The systemic (public) agenda refers to all societal problems that demand public attention and
form the discussion agenda. The institutional (formal) agenda, by contrast, contains a set of
problems that are up for serious consideration by decision-makers.
•• The drafting agenda is a list of subjects currently receiving attention within government. The deci-
sion agenda is based on those issues for which the government has agreed on a draft proposal and
hence has decided to put the issue on the agenda of the responsible decision-making body.
•• Agenda-setting can be analysed from different theoretical perspectives, including models emphasiz-
ing process patterns, the distribution of power, the role of institutions and contingency. These
perspectives are complementary rather than competing.
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:19:51.
Problem Definition and Agenda-Setting 91

Table 5.1 Four models for the study of agenda-setting.

Contingency
Process patterns Power distributions Institutional factors models
Focus Analysis of generic Analysis of who Analysis of how Coupling of
forms of has the power to institutions (veto streams and the
agenda-setting frame and reframe points, veto player, opening of an
issues political regimes opportunity
more generally) window
affect
agenda-setting
Analytical Abstract Actor-centred Integrated analysis Realistic
strengths conceptualization approach that of actors and modelling of
that facilitates facilitates the institutions agenda-setting
comparative analysis of
research strategic
approaches
Analytical Static and Overemphasis on Overemphasis on The individual
weaknesses reductionist concept actors’ strategies actors’ strategies analytical
and little and little attention components (the
attention to to non-strategic streams) cannot
non-strategic behaviour be separated
behaviour from each other

ACTORS AND INTERESTS IN THE AGENDA-SETTING PROCESS

The theories presented in the previous section map out the theoretical state of the art in the
agenda-setting literature and contribute to our understanding of the agenda-setting pro-
cess. Nevertheless, they are still rather abstract and lack precise statements regarding the
role that the different groups of actors play in the process. Therefore, we will now reduce the
level of abstraction and concentrate in more detail on the following types of actors that can
Copyright © 2020. Macmillan Education UK. All rights reserved.

play a central role in agenda-setting processes, namely:

• Elected public officials

• Judges
• Bureaucrats
• Interest groups
• Citizens
• International organizations
• Media

Elected Public Officials

Elected public officials (e.g. the president, prime minister or chancellor; ministers; mem-
bers of parliament) are the most obvious agenda builders, since their position enables
them not only to make decisions about policies, but also to place certain issues on the
agenda. The battle for control over legislative agenda-setting is predominantly waged

between the executive and legislative branches, but in coalition government it also applies
to the political parties that are members of that government (Green-Pedersen et al. 2018).
However, actual agenda-setting is related to the larger political game in terms of the
power and intensity of ideological conflict both within and between government and par-
liament. In this regard, Green-Pedersen’s (2006) study of the politicization of the issue of
euthanasia in Denmark, the Netherlands and Belgium shows how the possibilities of link-
ing issues to pre-existing conflicts in party systems can determine whether such issues are
placed on the agenda.
Notwithstanding a high degree of variation in the rules and practices of agenda-build-
ing in Western European parliaments (Döring 1995), setting the agenda for parliament is
the most significant institutional weapon for governments to shape policy results. As
argued by Rasch and Tsebelis (2013), governments with significant agenda-setting pow-
ers (e.g. France or the United Kingdom) are able to produce the outcomes they prefer,
while governments that lack these powers (e.g. Italy or the Netherlands) see their projects
being significantly altered by their parliaments (for a detailed discussion, see Rasch and
Tsebelis 2013). More precisely, this power stems from the fact that in some political sys-
tems it is easier for policy-makers in the legislature to accept the government’s proposals
and demand minor amendments to them than to come up with their own separate pro-
posal, due to the difficulties of attaining a necessary majority for it. As a rule, the execu-
tive’s agenda-setting powers are more pronounced in political systems in which there is a
partisan congruence between the government and the parliamentary majority. However,
even in polities characterized by a strong executive, the legislature might to some extent
affect the agenda-setting process via parliamentary bills, as shown by Baumgartner et al.
(2009a) and Bräuninger and Debus (2009).
For elected officials, agenda-setting power might emerge from various tools, such as con-
trol over the legislative calendar, the amendment process and the overall timing of the leg-
islative process – and whichever actors control them will largely control the entire
policy-making process within a given polity (Kreppel 2009: 185–8; see also Jennings et al.
2011; Bevan and Jennings 2014; Bevan 2015). In most cases no single actor controls all
aspects of agenda-setting, although there is usually a tendency towards a dominant role for
the executive or the legislature, depending on the general character of the political system.
Copyright © 2020. Macmillan Education UK. All rights reserved.

So the greater the degree of centralization of agenda-setting processes, the higher the level
of agenda-setting power in the hands of either the legislative majority or the executive (see
Tsebelis and Aléman 2005). In coalition governments, the party of the prime minister has
the agenda power, which is, however, limited by the issues that the coalition partners ‘own’.
The constraints on the prime minister’s party are lower when the prime minister can dis-
solve parliament, but higher when the coalition government comprises many coalition part-
ners (Green-Pedersen et al. 2018).

Judges
Agenda-setting might result from judicial review of existing legislation, thus awarding judi-
cial institutions the status of agenda-setters. The agenda-setting power of courts comes as
no surprise, as they are often asked to decide on controversial political issues. The main
focus of research on judicial institutions has been the study of the determinants of

decision-making. Of course, judges may privately have policy preferences, but they cannot
act as independent agents (Gibson 2008: 521). Instead, the expression of their preferences
is strictly limited to cases in which they have been authorized to act in accordance with the
constitution. Yet within this scope judges can certainly instigate a political debate about
issues with which policy-makers have previously dealt.
The 2010 decision of the German Constitutional Court on the constitutionality of welfare
and unemployment benefit levels is a case in point. In the context of the so-called Hartz
reforms – named after the head of the responsible commission, Peter Hartz – these benefits
had been merged into one scheme and at the same time considerably reduced. Despite ongo-
ing criticism, the issue of the appropriateness of benefit levels had disappeared from the
policy agenda. The court, however, ruled that the benefit levels were not in line with consti-
tutional requirements and obliged the legislature to undertake reforms to them. This judg-
ment brought the controversial issue of benefit levels back onto the German policy agenda.
Empirical evidence from the United States illustrates that judges of the Supreme Court
are more willing to review public policy and so possibly set the agenda for a new policy-
making process when they believe that the new policy output will be closer to their prefer-
ences than the status quo. However, these judges tend to deny review when they are in
favour of the status quo policy (see Perry 1991; Hammond et al. 2005; Black and Owens
2009). From this, it follows that policy preferences are a predictor of US Supreme Court
agenda-setting.

Bureaucrats
Bureaucracies have an impact on policy-making, not only at the formulation and implemen-
tation stage, but also in terms of placing issues on the agenda. This results from the fact that
bureaucracy makes up the largest part of most governments (Bevan 2015), which explains
why bureaucracy can easily influence the government agenda. In this context, Niskanen
(1971) posits that the bureaucrats impose upon a passive legislature their most preferred
alternative. His argument rests on two assumptions. First, the bureaucracy has a virtual
monopoly on true supply-cost information. Second, the bureaucracy knows the demand of
the legislature for the services it delivers. Based on these assumptions, it is expected that
Copyright © 2020. Macmillan Education UK. All rights reserved.

bureaucratic agencies are able to make all-or-nothing offers to the legislature (concerning
both the budget and the policy output) and in this way shape the government agenda.
Legislatures, however, might not always be as uninformed as is assumed in Niskanen’s
model; rather, information asymmetries might vary from issue to issue. Accordingly, Eavey
and Miller (1984) argue that the imposition of bureaucratic agendas should be replaced by
theories of bureaucratic bargaining, which emphasize the key structures of the relationship
between the bureaucracy and the elected officials. In this context, Schnapp (2000) speaks of
bureaucracies as ‘stand-in’ agenda-setters in situations where politicians do not have the
time and/or technical knowledge to set the agenda for all the problems that may arise in a
polity (see also Suleiman 1974; Aberbach et al. 1981). Here the bureaucracy can stand in,
take up problems and advance proposals for their solution, thereby becoming the agenda-
setter for a certain policy. In this vein, Kato (1994) shows that bureaucratic agenda-setting
was crucial for accomplishing a major tax reform in Japan which entailed the introduction
of the value added tax in 1989. Likewise, Page (2003) illustrates for the United Kingdom

how important civil servants are for placing issues on the decision agenda and initiating
policies. Aunphattanasilp’s (2018) analysis of energy agenda-setting in Thailand addition-
ally highlights that it is not the bureaucracy per se that sets the agenda, but the networks
formed by bureaucratic actors. For example, to explain why the Thai government first liber-
alized energy policy in 1987, only to centralize (2006–2008) and then re-nationalize it, it is
useful to direct attention to the networks formed by the Prime Minister’s Office and the
Ministry of Energy.
Bevan (2015) makes an interesting point by drawing attention to the fact that bureau-
crats are subject to the same resource constraints as politicians, which forces them to define
priorities (see also Workman et al. 2009; Workman 2015). The definition of priorities
depends on how they relate to the bureaucracy’s main goal of survival. In that sense, Bevan
argues that in the United Kingdom the bureaucratic agenda is formed by using information
about the known or anticipated priorities of the elected government in order to avoid scru-
tiny and sanction that could endanger the bureaucracy’s survival (see also John et al. 2013).
Therefore, it can be concluded that in the United Kingdom bureaucracy responds to elected
government.
In sum, we can conclude that there exists a close relationship between the government
agenda and the bureaucratic agenda, which is worth studying systematically by taking into
account the autonomy of the bureaucracy in individual countries.

Interest Groups
Similar to the mass media, the logic underlying the impact of interest groups relates to gen-
erating awareness of an issue and striving for public support (Wasieleski 2001). At the same
time, interest groups are involved in the process by framing issues and providing elected
officials with information (Cobb and Elder 1983; Haider-Markel 1999).
Baumgartner and Jones (1993/2009) present a comprehensive theoretical model relating
interest groups to agenda-setting. The authors contend that interest groups help to destroy
policy monopolies by means of ‘conflict expansion’ (Schattschneider 1960). Put differently,
a given societal problem must be perceived as such by those who are responsible for policy-
making. The politicization of an issue facilitates its inclusion on the policy agenda. To
Copyright © 2020. Macmillan Education UK. All rights reserved.

achieve this, the opponents of the prevalent image of the issue must mobilize and increase
the scope of the conflict in order to include actors outside of the policy monopoly. If the
mobilization efforts are successful, they will alter the dominant image of the policy through
agenda-setting.
In this regard, Baumgartner et al. (2009b) distinguish between three forms of lobbying.
The first is given by inside advocacy, in which information is supplied to those involved in
the policy-making process. This can, for instance, happen though (in)formal meetings with
civil servants or ministers, contacts with local government authorities, participation in
commissions and government advisory committees, contacts with members of parliament
or parliamentary committees, and contacts with officials of political parties. The second
strategy is outside advocacy, where information is communicated to actors outside the
policy-making process, such as through press conferences. The third form refers to grass-
roots advocacy, entailing the mobilization of the masses.

The extent to which interest groups are able to influence the agenda-setting process
depends on various factors. First is their resources, in terms of budget, staff and expertise.
These affect whether and to what extent interest groups might be able to mobilize public
support for an issue. Second, interest group influence is not only a matter of resources, but
also a matter of access, i.e. the extent to which such groups are actually consulted by politi-
cians and the bureaucracy. The more that an interest group enjoys an ‘insider status’, imply-
ing close and institutionalized relationships with political and administrative actors, the
less it needs to rely on mobilization strategies when trying to influence political agenda-
setting processes. Access to the media is also important, albeit depending on the strategy of
the interest group. Binderkrantz et al. (2015) explain that interest groups interested in
agenda-setting will direct their attention to the media, which stands in contrast to interest
groups seeking influence on the formulation or implementation of policies, which are more
likely to seek access to bureaucrats. From this, it follows that for agenda-setting by interest
groups to be a success, support from the media is an important factor. Therefore, interest
groups will produce press releases and other pertinent information that the media can pick
up and communicate to the wider public.
Figure 5.1 shows the number of press releases (in black) produced by environmental
groups in Germany on the contamination of water with nitrate and the number of articles
published in the newspaper Süddeutsche Zeitung (in grey) published between January 2014
and December 2018. Both media reporting and the number of press releases by environ-
mental groups on this issue was rather low until October 2016. The situation changed when
the EU Commission filed a lawsuit against Germany at the beginning of November 2016.
After that, the number of newspaper articles and press releases increased considerably. A
second peak can be observed in June 2018, when the European Court of Justice finally ruled

12
Absolute number of publications

8
Copyright © 2020. Macmillan Education UK. All rights reserved.

0
14 014 014 014 015 015 015 015 016 016 016 016 017 017 017 017 018 018 018 018
20 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q

Newspaper articles Press releases

Figure 5.1 Relationship between media articles and press releases on nitrate pollution in Germany
Source: Own data.

that Germany had breached the Nitrates Directive. While we see an increase, in many
instances there is a relationship between the number of press releases published and the
number of newspaper articles on that topic, and there are instances where the newspaper
did not cover it despite the press releases.
An interesting argument with regard to globalization and interest group strategies is put
forward by Weil (2017), who alludes to the possibility of ‘exporting’ lobbying structures from
one country to another. The author shows how Western interest groups – as a consequence of
growing foreign investment in the country – use China’s corporatist system to shape public
policy. Since Western interest groups predominantly target Western policy-makers to influ-
ence public policy in China, the mechanisms in place for monitoring the activities of Chinese
interest groups do not affect the lobbying activities of Western interest groups.

Citizens
The Fridays for Future movement and the reactions of the public and policy-makers to it
exemplify that citizens can influence the policy agenda. What is remarkable about this
particular movement is that it consists of school children who do not go to school on
Fridays to demand immediate and effective policy responses to climate change. In addi-
tion to demonstrations, protests and strikes, citizens can have an impact on the policy
agenda by organizing citizens’ initiatives. While the procedures and consequences of citi-
zens’ initiatives vary across countries (see Setälä and Schiller 2012), this format does cre-
ate media attention for the respective causes and can have a tangible impact on the policy
agenda. For example, the European Citizens’ Initiative ‘Water and sanitation are a human
right!’ (Right2Water) succeeded in placing the issue of ensuring access to drinking water
on the policy agenda of the EU Commission (Schiffler 2015). The possibility of influencing
agenda-setting also helps to promote inclusiveness. An analysis of a citizens’ initiative in
Finland shows that this instrument includes young people in political processes
(Christensen et al. 2017).

International Organizations
Copyright © 2020. Macmillan Education UK. All rights reserved.

A notable share of the work of international organizations such as the Organisation for
Economic Co-operation and Development (OECD) and the numerous organizations belong-
ing to the United Nations (UN) system explicitly aims to influence agenda-setting in the
member states (Majone 2008: 247). The World Health Organization (WHO), for example,
has been influential in bringing vaccination issues onto the political agenda worldwide (see
Fedson 2005). Likewise, the Programme for International Student Assessment of the OECD,
an internationally standardized assessment of 15-year-olds in schools, brought education
issues onto political agendas – especially in those countries that scored unexpectedly low
(see Ertl 2006). Furthermore, the annual Conference of the Parties to the United Nations
Framework Convention on Climate Change – an international climate summit – has been
influential in increasing the amount of climate change-related coverage in the media (Schäfer
et al. 2014). There are many more examples that could be given here to underline the agenda-
setting power of international organizations. Altogether, agenda-setting through interna-
tional organizations is a particularly interesting area of study, since the information supplied
is the same, but the reactions of national policy-makers in the individual countries may be
quite diverse.

Media
Research has started in the field of communication research regarding the role of the mass
media in agenda-setting (see Cohen 1963; Lippmann 1964), with emphasis being placed on
the relevance of the mass media in forming the image of issues (also discussed previously as
framing). Here the focus has been on the assessment of links between the salience of issues
on the media’s agenda and those on the systemic agenda. McCombs and Shaw’s study (1972)
is widely regarded as the first empirical analysis of public agenda-setting. The authors dem-
onstrate a relationship between what survey respondents feel are the most important issues
and the coverage these issues are given in primary news sources. Numerous studies have
examined the influence of news media coverage on the public’s issue priorities. The vast
majority of these have indeed found strong correlations between the media coverage of
issues and the public’s level of concern with them (see, e.g., Boomgaarden and de Vreese
2007), which is, for example, measurable via the Most Important Problem dataset (see
Heffington et al. 2019).
Subsequent research has investigated the relationship between the mass media and insti-
tutional agendas. For example, Walgrave and van Aelst (2006: 99–101; see also Walgrave
et al. 2008) argue that media coverage is associated with public opinion. Whether the media
really affects public opinion or not is less relevant than the fact that political actors believe
that it does. Political actors do not primarily react to the media coverage itself, but to the
presumed public opinion about issues. This implies that they anticipate the expected media
impact on the public and build their political strategy on that premise to maximize their
electoral benefits in accordance with rational choice logic (see Chapter 4). Green-Pedersen
and Stubager (2010) further explain that media attention to an issue generates attention
from political parties when the issue is one that parties have an interest in politicizing in the
first place.
Another strand of research systematically relates the agenda-setting power of the mass
media to the type of issue concerned. For example, Soroka’s (2002) study of agenda-setting in
Canada differentiates between prominent, sensational and governmental issues. Prominent
issues such as inflation have concrete effects on citizens; they can be experienced without the
media’s providing information about them. Sensational issues are less connected with daily
Copyright © 2020. Macmillan Education UK. All rights reserved.

experience, but still have tangible effects for most of the population. Often they are marked by
dramatic events which draw massive media attention. Examples of sensational issues are the
spread of an epidemic and oil spills from grounded tankers, but also the assassination of a poli-
tician (see Boomgaarden and de Vreese 2007). Governmental issues are unobtrusive and often
without specific effects on the population (e.g. budgetary policies). Of these three issue types,
the media are expected to be most influential in the case of sensational issues. For prominent
issues with strong and specific effects, the political effects of the media are expected to be
much more modest, as people and politicians can rely on their own direct observation. Finally,
the impact of the media should be lowest in the case of governmental issues.
In summary, agenda-setting is essentially about attention dynamics and, as a conse-
quence of it, the coverage of issues by the media is intimately connected to policy-making
(Wolfe et al. 2013: 179). From that perspective, it is beneficial for the study of agenda-
setting to improve our understanding of media dynamics and to better understand why
certain issues are covered by the media and others are not. Having said that, we must also
recognize that media dynamics today comprise both traditional and social media. The co-
existence of different types of traditional and social media means that there exist rival

frames of issues (see Guggenheim et al. 2015). Another important aspect is that with the
pluralization of media in the digital era comes selective exposure to news, which is also
likely to affect agenda-setting (see Muddiman et al. 2014). However, the co-existence of
different media can also have reinforcing effects. As Feezell (2018) shows, mainstream
media can also have more leverage in influencing issue salience and therefore the discus-
sion agenda when channelled through social media.
In conclusion, it should be noted that agenda-setting via media can be motivated both
by the media itself and by the actors discussed in this chapter that make use of the media
to expand the scope of conflict and seek to mobilize supporters for their cause. While
agenda-setting also comprises the containment of the scope of conflict and the de-empha-
sizing of issues, we must acknowledge that, in agenda-setting via increasing public atten-
tion, the media play a significant role. Consequently, compared with the other actors
involved in the agenda-setting process, the media are of enhanced importance, and under-
standing their role is key to understanding agenda-setting. And in this context, it should
be noted that access to the media varies across the different actor groups, but also within
them. As Chaqués Bonafont (2016: 209) and Binderkrantz et al. (2017) contend, media
attention to interest groups, for example, is concentrated on a small group characterized
by high status, abundant resources and advocacy for legitimate causes.

Key points
•• Agenda-setting includes activities by elected public officials, judges, bureaucrats, interest
groups, citizens, international organizations and the media.
•• The media stand out in this group of actors since they can engage in agenda-setting because of
their own motivation or because the other actors seek access to the media.
•• Agenda-setting research is increasingly recognizing the existence of both traditional and social media.
•• The outcome of agenda-setting efforts is often a combination of what different actors do and
how they interact.
Copyright © 2020. Macmillan Education UK. All rights reserved.

CONCLUSIONS

In this chapter we have shed light on the stage of policy initiation and examined problem
definition and agenda-setting. We have seen that problem definition is about making issues
public and turning them into problems that are perceived as important. In general, a social
problem may turn into a political problem only if it meets the following three criteria: (1)
the issue must receive attention from a large number of people; (2) a sizeable proportion of
the population must demand action; and (3) the government must be able to tackle it. Yet
we have also highlighted that problem identification is rarely a mere matter of objective
data. Rather, the perception of problems is constructed socially, and so there will be a strug-
gle over the definition of problems at various stages.
The perception of the population is, however, not a sufficient condition for policy-makers
to include an issue on their institutional agenda. If and to what extent this is the case is
addressed by various analytical approaches to agenda-setting. Models emphasizing process
patterns, the distribution of power between the involved actors, institutional factors and con-
tingency have been discussed. Regarding the role of different actors in the agenda-setting
process, particular emphasis was placed on elected public officials, judges, bureaucrats,

interest groups, citizens, international organizations and the media. All in all, the success of
agenda-setting depends on the levels of attention, the image of the issue to be brought onto
the agenda, and the extent to which conflict can be expanded or contained. If too many frames
are invoked and too many participants are brought into the agenda-setting process, subse-
quent policy-making can become quite adversarial (Rochefort 2016: 49). This may result in
the adoption of policies that leave many details unspecified, which eventually have to be
defined at the implementation stage.

WEB LINKS

www.agendasetting.com. This site offers some intriguing case studies and

announces international conferences on agenda-setting.
www.comparativeagendas.org. This is the website of a transnational research project
on agenda-setting. It provides links to country-specific research groups as well as
publications and data.
ec.europa.eu/citizens-initiative/public/welcome?lg=en This website covers all citi-
zens’ initiatives launched using the format of the European Citizens’ Initiative.
www.newspaperarchive.com. This is a very comprehensive online newspaper archive
featuring articles from newspapers around the United States and some other
countries.
www.parl.ca/legisinfo/Agenda.aspx. This website presents the legislative agenda of
the Canadian House of Commons and the Senate.
www.theconversation.com. The Conversation is an independent source of news and
views from the research community delivered to the public.

FURTHER READING

Baumgartner, F. R. and Jones, B. D. (2009). Agendas and Instability in American
Politics. Chicago, IL: University of Chicago Press. Recommended reading for every-
Copyright © 2020. Macmillan Education UK. All rights reserved.

one working on agenda-setting and policy dynamics.

Engeli, I., Green-Pedersen, C. and Larsen, L. T. (2012). Morality Politics in Western
Europe: Parties, Agendas and Policy Choices. Basingstoke: Palgrave Macmillan. A con-
cise collection of studies of how morality issues are placed on the political agenda.
Green-Pedersen, C. and Walgrave, S. (eds.) (2014). Agenda Setting, Policies, and Political
Systems: A Comparative Approach. Chicago, IL: University of Chicago Press. This
edited volume summarizes the main findings of the Comparative Agendas Project.
John, P., Bertelli, A., Jennings, W. and Bevan, S. (2013). Policy Agendas in British
Politics. New York: Palgrave Macmillan. This book offers an instructive overview
of different types of agenda-setting in the United Kingdom.
Kingdon, J. W. (2011). Agendas, Alternatives, and Public Policies. New York:
HarperCollins. Informative reading for those who wish to gain a more complete
picture of policy-making.
Zahariadis, N. (ed.). (2016). Handbook of Public Policy Agenda Setting. Cheltenham:
Edward Elgar. A comprehensive volume that examines how individuals and insti-
tutions are involved in agenda-setting processes.

Reader’s guide
When a specific policy problem has become a part of the institutional agenda, the
relevant actors will address it in the course of the decision-making process. This
stage consists of two theoretically separable – albeit empirically closely related –
actions. The first action concerns the drafting of a legislative proposal, which often
involves a debate about the more specific nature of the social problem to be resolved
by it. Based on this definition, official and unofficial policy-makers discuss policy
design, which also includes decisions about the instruments to be employed and
their specific settings. Central to policy formulation are the executive and the min-
isterial bureaucracy; the latter in particular often has a dominant position. The way
in which the ministerial bureaucracy develops a policy proposal might be affected by
external expertise, policy recommendations of international organizations, inter-
est group preferences, partisan ideology and bureaucratic self-interests. The second
action related to decision-making refers to the actual adoption of the policy proposal
in order to turn it into binding law. At this stage, the number of actors involved
diminishes and executive–legislative relations are brought to the fore. Depending
on the institutional and procedural characteristics of the country concerned, in the
formal adoption process the executive and the legislature can either be on an equal
footing or have a relationship in which one dominates the other, bringing us back to
the characteristics of polities discussed in Chapter 3.
Copyright © 2020. Macmillan Education UK. All rights reserved.

INTRODUCTION

Who decides on the design and adoption of public policy? How is the decision-making pro-
cess structured? When speaking about decision-making, most of us would think of mem-
bers of parliament debating draft legislation introduced by the government. These
parliamentary debates that criticize or support the policy proposal and the discussion of
alternative suggestions are often instigated by the opposition, i.e. parliamentary groups of
parties holding a minority of seats. Though debates are often long and controversial, alter-
native proposals are rarely picked up by the majority parties and the government. Rather,
these debates serve the objective of sending signals to the electorate that the current minor-
ity parties also have policy solutions to offer. Of course, the idea underlying these signals is
that at the next election the electorate will remember parliamentary debates and help the
now minority parties to gain the majority of parliamentary seats.
As we have already seen in Chapter 2, there is a plethora of policy instruments; their
employment varies significantly across individual policy fields. For example, if decision-makers

100
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:19:41.
Decision-Making 101

aim to reduce traffic jams, they can choose between several options. First, they could invest in
new road construction or the maintenance of existing roads. Second, they could decide to
develop a parking guidance system. Third, they could change road users’ attitudes about how
to use their cars. It might be useful to promote the idea of car-sharing or to provide incentives
for using public transport more frequently. Fourth, they could charge road fees. Which of
these instruments might they propose? Which of the different existing proposals have a real-
istic chance of being turned into binding law?
These are exactly the questions around which decision-making revolves. Decision-making
involves considerations about the technical and/or financial feasibility and effectiveness of
each of these options, as well as considerations about how the electorate would respond to
them. Some actors benefit from a certain policy design, while the same design disadvantages
others. Policy-makers – who generally seek to become re-elected – have to keep all these
points in mind when they develop a policy proposal and seek to take it to adoption.
Which factors determine the design of a certain policy proposal? Why are the chances of
a specific policy proposal being adopted higher than for another one? These are the two main
research questions lying at the heart of this chapter on decision-making. Fitzgerald et al.
(2019) contend that policy-making is determined by the interaction between interests, ide-
ologies and institutions, which affect the incentives for policy-makers and the information
they receive. The levels of uncertainty and complexity also matter for how policy decisions
are taken. As we will show in this chapter, these concepts do indeed matter and are worth
discussing in greater detail.
Technically speaking, decision-making consists of two processes, which are empirically
interlinked but which for analytical reasons should be differentiated. The first process is
about the drafting of a piece of legislation, whereas the second is about its formal adoption.
Both of these phases are characterized by procedural and substantial restrictions. Procedural
restrictions emerge from the country’s respective polities and politics. Substantial restric-
tions refer to the policy problems that need to be resolved. While policy-making in some
cases can be considered routine work, many policy problems refer to much more complex or
‘wicked’ constellations; they present themselves in so many diverse facets or are simply so
extraordinary that the elaboration of a sound policy is hampered (Peters 2017; Head 2019).
Examples of such ‘wicked’ policy problems are the prevention of terrorist attacks and gover-
Copyright © 2020. Macmillan Education UK. All rights reserved.

nance in natural disasters (see later Box 6.2).

Along with the government and parliament, the ministerial bureaucracy is a key player at
the drafting stage. Experts, interest groups and ideas promoted by international organiza-
tions assist the ministerial bureaucracy in developing policy proposals. In some polities or
specific legislative situations, the public is explicitly encouraged to participate in the formu-
lation of draft legislation within consultation processes.
Once we move to the stage of policy adoption, the number of actors declines, as now only
public actors can decide whether to accept a bill or not. Nevertheless, this stage entails con-
troversial parliamentary debates and a formal voting process in at least one chamber. In
many political systems, however, a second chamber is a co-legislator and needs to approve
of legislative bills. Depending on the specific institutional characteristics of the country con-
cerned, the need to get a majority in both chambers might lead to delays and additional
modifications of the policy proposal. In some countries, referendums have an impact on
whether the legislation is adopted or not. Hence, drafting a bill that effectively solves the
policy problem in question and is likely to be adopted by the relevant official actors – and
accepted by the public – is a challenging task.

DETERMINANTS OF POLICY FORMULATION

In parliamentary systems, the formulation of public policy is mostly a task of the executive.
In presidential systems, by contrast, there is no executive monopoly with respect to policy
formulation. Instead, legislative committees play a vital role in proposing legislation – how-
ever, they also tend to cooperate closely with the ministerial bureaucracy (O’Connor and
Sabato 2009: 322), again giving the executive notable discretion over policy proposals
(Kreppel 2009: 183). There are, however, also presidential systems, such as Chile’s, in which
the executive is directly responsible for policy formulation (for details and a discussion, see
Alemán and Tsebelis 2005, 2011).
These considerations suggest that the executive is crucial for the stage of policy formu-
lation, though it needs – albeit to various degrees – support from the ministerial bureau-
cracy (see Aberbach et al. 1990). Recognizing the importance of ministerial bureaucracy
for drafting public policy, in this section we will discuss the ways in which bureaucrats
form their preferences. All of the factors to be presented can, in principle, also affect leg-
islative actors preparing policy proposals; and in some countries (e.g. Italy) this would
represent a more accurate perspective. However, we will frame the explanations in a way
that accounts for the central role of bureaucrats in preparing legislation in many countries
(Peters 2010).

Expertise, Information and Ideas

The first way of thinking about how the ministerial bureaucracy formulates a policy proposal
emphasizes the importance of expertise. Bureaucrats typically possess special insights into
the relation between programmes and their effects, due to their professional training and
the fact that in most political systems ministerial bureaucrats do not regularly change their
positions and, hence, are able to accumulate expert knowledge over time (Bendor et al.
1985: 1041; see also Wildavsky 1979; Hammond 2003; Peters 2010; see also Balla et al.
2015 for an overview). In this regard, bureaucrats also play an essential role in managing
unusual situations and crises, such as the 2014–2016 Ebola outbreak in West Africa (see
Copyright © 2020. Macmillan Education UK. All rights reserved.

Boin and Lodge 2016). Another important function bureaucrats fulfil is to act as so-called
policy brokers or mediators in situations of political conflict (Varone et al. 2018b). Several
empirical studies demonstrate that bureaucrats do indeed opt for policy designs that they
believe will produce desirable results (see Eisner et al. 2006: 21). While it is true that proce-
dural knowledge – i.e. knowledge about how to carry an action out – stems from inside the
bureaucracy, it is likely that the ministerial bureaucracy’s own expertise needs to be enriched
by additional information sources for specific policy contents.
With policy problems becoming increasingly complex (Aberbach et al. 1981: 2), it is likely
that bureaucrats cannot provide themselves with all the relevant information about a cer-
tain social problem. For example, what are the long-term effects of genetically manipulated
crops on public health and the environment? This question is difficult to answer, as the issue
of genetically manipulated crops is a relatively recent one. Even more recent is the possibil-
ity of genome editing, which makes the long-term effects of this technology significantly
more difficult to assess. In such cases, bureaucrats can – and depending on the legal rules,
must – use information provided by experts such as scientists, think tanks, consultancy
firms and (political) foundations (see Stone 2004).

Haas (1992) outlines exactly how experts can influence policy choices. First, they can give
advice about the likely results of different courses of action following an event such as a
shock or a crisis. Second, they might help decision-makers to grasp complex interlinkages
between issues. Third, they can help to develop fundamental policy principles such as ‘work-
fare’, i.e. an alternative model to conventional social welfare systems that became popular in
many countries from the 1970s onwards, where recipients must demonstrate certain par-
ticipation efforts to continue to receive their benefits (see Dingeldey 2007). Neo-liberal
think tanks were indeed highly influential in establishing the idea of ‘welfare-for-work’ (also
known as ‘workfare’) in Australia, Canada, the United Kingdom and the United States in the
late 1970s (Stone 2001: 342), which is about requiring welfare recipients to work for welfare
payments in jobs assigned to them by government (see Lodemel and Moreira 2014). Finally,
experts might provide support for choosing amongst policy alternatives through framing
them in accordance with certain norms. In view of these opportunities to affect public pol-
icy, (internationally operating) experts have been informing national policy-makers about
successful policy ideas adopted elsewhere, thus stimulating processes of ‘lesson-drawing’
(Rose 1991) and ‘policy learning’ (see, e.g., Braun and Gilardi 2006; Holzinger et al. 2008a,
2008b; Gilardi et al. 2009; Meseguer Yebra 2009; Gilardi 2010; Dunlop and Radaelli 2016a).
While these arguments clearly underline the many different ways in which experts can
affect policy design, it must be noted that the institutionalized participation of experts var-
ies strongly across countries. In some countries like the United Kingdom, experts tend to
play an important role in policy drafting. In Israel, for example, a central administrative
position in various ministries is the ‘chief scientist’, which is in charge of organizing and
reviewing expertise with a view to supporting evidence-based policy-making (Schmidt et al.
2018). In low-income countries, development agencies often act as experts and advise min-
isterial bureaucracy (see, e.g., Rahman et al. 2016).
In many Continental European countries, the role of experts is limited. In Germany,
institutionalized circles of ministerial bureaucrats and scientific advisers formally linked to
the government are decisive for policy formulation; this makes it difficult for independent
experts to gain access to ministerial bureaucrats (see Knill 1999; Strassheim and Kettunen
2014). Similar findings of a limited role for independent experts are reported for Austria
and Denmark (see Lindvall 2009). However, the economic and financial crisis that unfolded
Copyright © 2020. Macmillan Education UK. All rights reserved.

in Europe from 2008 onwards entailed institutional reforms in which experts featured
prominently (Tosun et al. 2014). The most marked case is Greece, where the crisis led to far-
reaching changes in public administration and introduced experts to policy-making, espe-
cially in the realm of economic and fiscal policies (see Ladi 2014; Featherstone 2015).

International Organizations
The proposals of ministerial bureaucrats can also be affected by ideas promoted by interna-
tional organizations. There are, however, many ways in which such organizations can be
influential (see Chapter 10; Knill and Bauer 2016; Knill et al. 2018). First, ministerial
bureaucrats can voluntarily employ information provided by international organizations.
Virtually all these organizations dedicate considerable resources to gathering data and pre-
paring reports on specific policy situations or the presentation of policy recommendations.
The air quality guidelines published by the World Health Organization (WHO) since 1987
represent such a compilation of information. These recommendations are employed by

countries all over the world when defining their air quality standards (see Biesenbender and
Tosun 2014). Many developing countries are in particular grateful to have this source since
it spares them from carrying out costly scientific studies. This turns the formulation of clean
air policy into a more efficient process. However, the degree to which ministerial bureau-
crats in individual countries finally choose to follow these recommendations varies.
The involvement of international organizations in policy drafting can also be of a more
coercive character. This is particularly the case when governments turn to organizations
such as the International Monetary Fund (IMF) or the World Bank for financial help. When
these organizations get involved they often make very specific policy recommendations,
which the governments have to take into consideration (Dolowitz and Marsh 2000: 11; see
also Holzinger and Knill 2005). Generally, the more drastic the financial and/or economic
situation of a country, the stronger will be the IMF or World Bank’s influence on the eco-
nomic and/or financial policy proposals to be implemented by national ministerial bureau-
crats. The World Bank has, for example, had a great impact on economic and financial policy
in Argentina after the country became dependent on loans as a consequence of the currency
crisis of 2002 (see Teichman 2004). More generally, during the 1990s these same organiza-
tions affected policy formulation by vigorously promoting the idea of allowing private-sector
participation in water and sanitation in Africa, Asia and Latin America (Schiffler 2015).
Policy recommendations in these regions in the area of water provision and sanitation cor-
respond with the broader ‘neo-liberal’ reform policies that the IMF and the World Bank
encourage (Budds and McGranahan 2003). In the case of Greece, the European Commission,
the European Central Bank and the IMF – the so-called Troika – made the provision of finan-
cial assistance conditional on far-reaching institutional and policy reforms (Ladi 2014;
Featherstone 2015). Therefore, one of the most recent examples where international orga-
nizations applied conditionality to a particular country is Greece. The last official request for
a government loan was in July 2015.
What is also worth noting is that in some areas international organizations even compete
for influence over policy-making. De Francesco (2016) explains that this is the case for the
‘good governance’ indicators that countries adopt by following the guidelines by the OECD
or the World Bank (Box 6.1).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Box 6.1 Drafting the Kenyan biosafety law

Biosafety issues focusing on ecology and human health have been receiving burgeon-
ing public and political attention. Most prominently, biosafety legislation defines
the rules for the cultivation, manufacture, use, import, export, storage and research
of genetically modified organisms such as biotech cotton, maize or soya beans. The
way in which the Kenyan biosafety law was drafted is particularly interesting, as a
large number of national and international stakeholders participated in the process.
According to Karembu et al. (2010), at the initial stage of policy drafting in 2001,
workshops were held which were attended by members of the Kenyan executive and
legislature, universities, non-governmental organizations (NGOs), religious groups,
media, different branches of industry, farmers’ associations, foreign development
organizations and United Nations (UN) agencies. Some of the topics discussed were
about the merits and demerits of modern biotechnology, the question of how to

integrate the law into the existing body of public policy and the government depart-
ment to be made responsible for regulating biosafety.
In 2002, the first draft was finished and circulated amongst the most relevant
stakeholders for comments and input. A more formal discussion of the bill took place
in 2003. Although the bill was then, in principle, ready to be presented to the legisla-
ture, some fresh problems arose concerning the country’s preparedness, which signifi-
cantly delayed the drafting process. It took until 2005 to finalize the drafting of the
bill. At this last stage, the bill was rewritten in order to incorporate all unsettled issues
and the experiences that other countries (e.g. Australia, Canada, India, South Africa
and the Philippines) had had while regulating for biosafety. Technical and financial
support was provided by some industrialized countries to improve further the regula-
tory and scientific quality of the bill.
It then took the Kenyan policy-makers until 2009 finally to enact the biosafety law
due to unforeseen complications and some particularities of Kenyan politics. While
the involvement of experts and in particular of the scientific community had a great
impact on the design of the policy (Kingiri 2010), the reform may not have material-
ized without the intervention of a government agency that acted as a policy broker
and facilitated policy-making (Kingiri 2014).

Interest Groups
In parliamentary systems the executive branch of government – i.e. cabinet members plus
the ministerial bureaucracy – represents the primary focus of interest groups (Werner and
Wilson 2008). This also holds true for presidential systems such as the United States,
although in these the relevant congressional committees are also addressed by interest
groups. In fact, for a long time US policy-making has been associated with ‘iron triangles’,
consisting of closely knit relationships between the responsible ministerial bureaucracy,
congressional committees and interest groups (see Jordan 1981: 96).
Copyright © 2020. Macmillan Education UK. All rights reserved.

There are good reasons why ministerial bureaucrats are willing to cooperate with interest
groups when drafting legislation. The first is that interest groups are able to supply valuable
information concerning the effects of a policy to be proposed, as well as how it might be
received by their members. Accordingly, in addition to the issue-related information pro-
vided by experts, interest groups are able to provide information about the ‘political’ dimen-
sion of a law proposal. This turns them into a particularly valuable resource for those drafting
legislation. Second, interest groups engage in a two-way information-mediation process,
which means that they also supply information to their members. In so doing, they can
frame policy proposals so as to make their acceptance more or less likely. Consequently,
interest groups can, in principle, act as advocates of public policy and thereby help to increase
the chances of their implementation and – more generally – the popularity of a government.
Grossman and Helpman (2001) further argue that interest groups can attempt to educate
the public. The typical voter lacks the expertise to evaluate alternative policy proposals. For
their part, interest groups are willing to serve as educators, because by doing so they can
shape public opinion in a way that will be beneficial to themselves.

On the other hand, interest groups aim to pursue the objectives of their members and so
are likely to present information in a way that is most instrumental to their members. As a
result, administrative actors might not receive information about the entirety of feasible
policy alternatives, but only about those that correspond to the organizational goals of
interest groups. In extreme scenarios involving areas of regulatory policy characterized by
high technical and scientific complexity, information asymmetries between interest groups
and the bureaucracy can lead to problems of ‘regulatory capture’ (Stigler 1971; Peltzman
1976). This implies that administrative actors no longer serve the public interest, but
advance the private interests of those actors they are charged with regulating. There is, how-
ever, evidence in the literature that an interest group’s effectiveness at influencing public
policy is determined by its ability to establish a reputation as a provider of reliable and com-
plete information (see Grossman and Helpman 2001; Coen 2007). It should also be noted
that much lobbying activity is concentrated on only a few issues, while the majority of policy
proposals are only addressed by one or a few interest groups (Baumgartner and Leech 2001;
Baumgartner et al. 2009b).
Depending on the general patterns of interest intermediation, the exchange relation-
ships between the bureaucracy and interest groups might vary considerably. In pluralist sys-
tems, the bureaucracy might be confronted with a broad range of different interest groups.
This means that the administration can select from very different sources of information.
The problem is, however, that in view of rather diverse interest groups, it might be difficult
for bureaucrats to discern the accuracy of information. In (neo-)corporatist systems, by con-
trast, the state usually grants representational monopolies to a limited number of associa-
tions, hence reducing the number of interest groups that are formally consulted (see, e.g.,
Weil 2017). This more structured approach might bear the disadvantage that the provision
of policy-relevant information becomes further reduced if only a limited number of interest
groups gain access to the policy formulation process.
There is another dimension along which the interest groups involved in policy-making in
(neo-)corporatist systems can be distinguished, namely the type of exchange relationships
they maintain with political actors. More precisely, we can identify ‘multipartite’ relation-
ships, ‘clientela’ relationships and ‘parantela’ relationships. Multipartite structures mani-
fest themselves in various ways. The ‘classic’ (neo-)corporatist mode of political exchange is
Copyright © 2020. Macmillan Education UK. All rights reserved.

tripartite, including unions, employers (i.e. the social partners) and the government. Many
studies, however, argue that even in (neo-)corporatist systems there is the tendency that
political exchange involves more extensive ‘issue networks’ (Heclo 1978).
A ‘clientela’ relationship describes the mode of political exchange within a (neo-)corpo-
ratist setting where one single-interest group manages to become the key cooperation part-
ner of a ministry (La Palombara 1964; Peters 2010: 181–5). For example, Enjolras and
Waldal (2007) demonstrate that the making of Norwegian sport’s policy is in some instances
characterized by a clientela relationship between a monopolistic interest group (i.e. the
Norwegian Olympic Committee and Confederation of Sports) and the ministerial
bureaucracy.
‘Parantela’ relationships refer to a situation of close ties between an interest group and
the government or dominant political party (La Palombara 1964). Interest groups obtain
access to administrative decision-making through the willingness of a hegemonic party to
intercede on their behalf with the bureaucracy and therefore, in essence, control bureau-
cratic policy-making. This type of relationship today cannot be found in Europe or North
America, but it still exists in some African and Latin American states (Peters 2010: 185–8).

Regardless of the specific form of relationship between government and interest groups,
the latter must be regarded as essential for policy-making. For example, when Helen Clark
became New Zealand’s prime minister following the 1999 election, she adopted a pro-
interest-group rhetoric and signalled that she would accept the legitimacy of interest groups
and appreciate their support in policy-making. It should also be noted that Clark’s approach
was intended to ‘correct’ the previous government’s attempts at eliminating interest groups
from policy-making. In fact, this decision had created some practical problems for the New
Zealand government, especially with regard to the role of interest groups to increase the
acceptance of government policy amongst those they represent (Tenbensel 2003: 353). It is
particularly this potential to mobilize support for political decisions that turns interest
groups into key actors in the policy-making process.

Political Preferences Stemming from Partisan Ideology

A way of limiting the range of alternatives when developing policy proposals is to impose an
ideological filter corresponding to the party membership of ministerial bureaucrats or that
of their political masters. For example, education policy in Germany is a highly politicized
issue. With regard to the secondary school system, there are two models: one is to distin-
guish between three levels of secondary schools, and the other is to integrate them into a
comprehensive school. Generally, members of the Social Democratic Party tend to prefer
comprehensive schools, whereas members of the Christian Democratic Union and the
Christian Social Union are in favour of the three-tiered structure. In consequence, policy
proposals developed by the ministerial bureaucracy under a Social Democratic minister
would most certainly entail modifications to the existing comprehensive schools rather than
their replacement through the three-tiered model. In other words, ideological consider-
ations may impose restrictions on the policy alternatives that are taken into account by
bureaucrats.
However, being in the same party does not mean that all party members have the same
policy preferences as, for example, Hansen and Rasmussen (2013) show for Danish parties.
Further, the shaping effects of partisan ideology strongly depend upon the extent to which
political actors control bureaucrats; that is, how strongly they are politicized. The reason for
Copyright © 2020. Macmillan Education UK. All rights reserved.

politicization is that political actors want to ensure that ministerial bureaucrats possessing
an informational advantage do not deviate from political guidelines when drafting legisla-
tion. Therefore, in the last three decades there has been a tendency towards politicizing
bureaucratic actors. Yet not all political systems are willing to take this particular step.
The British political system, for instance, is characterized by the separation of politics and
administration (Campbell and Wilson 1995: 14; see also Sausman and Locke 2004). This
separation is not at all obvious, as the role of the British ministerial bureaucracy is essen-
tially a political one. Notwithstanding the emphasis placed on a non-partisan, permanent
civil service, the Whitehall model demands not that ministerial bureaucrats be non-political,
but that they be politically promiscuous (Rose 1987). In fact, many of the classic activities of
the ministerial bureaucracy are essentially political, as it has always worked in a situation
where the need for high technical ability has been low and the demand for politically attuned
advice high and frequent (Campbell and Wilson 1995: 29). From this perspective, British
ministerial bureaucracy potentially constitutes an independent source of political influence
(for a discussion, see Page and Jenkins 2005).

The fact that a politically powerful ministerial bureaucracy has always been an important
characteristic of the British political–administrative system does not automatically imply
that the degree of this influence is absolute. Thus, the Conservative governments under
Margaret Thatcher and John Major demonstrated that political influence is nothing to be
taken for granted, but is contingent upon the preferences of political leaders (Knill 1999:
132). Thisis also underlined by the creation of ‘special advisers’, who are members of the
governing political party and on the civil service payroll.
In contrast to the United Kingdom, the German ministerial bureaucracy is strongly polit-
icized (Schröter 2004: 75). The politicization basically occurs in two ways. First, administra-
tive positions at the very top of the hierarchy, i.e. state secretaries and division heads in the
ministries, are filled with ‘political civil servants’ who have no lifetime tenure, but can be
temporarily retired (Derlien 1995). A second trend is the increasing party politicization of
the ministerial bureaucracy. Mayntz and Derlien (1989: 397) have observed a correlation
between membership of the governing party and employment in the federal bureaucracy,
which significantly reduces the advancement opportunities for non-party members and
those with the wrong party sympathies (for an overview, see Knill 2001: 92–5).
In the United States, bureaucrats are also appointed for political reasons. Yet Peters
(2004: 127) argues that the politicization of the ministerial bureaucracy in that country is
different from the situation in other countries, as both Congress and President control the
administrative actors. Congressional control in particular ensures that patronage appoint-
ments by the executive are low in number and that a functioning merit system is in place.

Pursuit of Private Interests

It should have become clear by now that bureaucrats possess an information advantage over
elected officials, which might allow them to exercise discretion over policy decisions. There
are various theoretical approaches that make use of public choice theory (i.e. the economic
theory of political choices) to explain the problem of the control of administrative organiza-
tions by their political masters, and that consider that bureaucratic pathologies have much
to do with problems related to information flow (for the most essential approaches, see
Parkinson 1958; Tullock 1965; Downs 1967; Niskanen 1971). Theoretical considerations
Copyright © 2020. Macmillan Education UK. All rights reserved.

building on sociological institutionalism are unlikely to predict that bureaucrats would nec-
essarily deviate from organizational priorities, but instead they would assert that institu-
tions socialize their members in such ways that they define an institution’s goals and values
as their own. Thus, it should be noted that the idea of bureaucrats indulging in ‘bureau-
shaping’ or expansionist tendencies is constrained to rational choice theory.
What kinds of pathologies exist? The most important one is that bureaucrats – akin to all
other individuals – might aim to increase their income and prestige by climbing up the
career ladder and by seeking to realize their private interests when drafting legislation
(Müller 2011). The influential work of Dunleavy (1991) contends that senior bureaucrats are
more interested in utility maximization through bureau-shaping than in budget maximiza-
tion (Niskanen 1971). Dunleavy contends that senior bureaucrats seek to enhance their
functions as policy advisers. In the event of budgetary cuts or other requirements for reor-
ganization, senior bureaucrats reshape their bureaux into small staff agencies. This strategy
ensures organizational survival and allows them to concentrate on their role as policy advis-
ers, which grants them high status and quality of work.

Against this backdrop, Downs (1967) puts forward an argument that is different from the
public choice perspective by stating that the extent to which self-interested bureaucratic behav-
iour dominates depends on the personality types present in administrative agencies. The author
differentiates between five types of bureaucrats who are more or less likely to follow their private
interests in preparing legislation. In this sense, ‘climbers’ and ‘conservers’ are interested in maxi-
mizing or conserving personal power and prestige, which entails that they are relatively likely to
pursue their self-interests. ‘Zealots’, by contrast, are unlikely to be driven by private interests, as
they are loyal to narrow aspects of the organization’s goals, e.g. the specific programme for which
they work. In a complementary vein, ‘advocates’ are devoted to the organization’s broader policy
goals and are equally unlikely to pursue their own interests. Finally, there are ‘statesmen’ who
seek to make sure that the organization to which they belong follows the public interest, which
also reduces the chances of self-interested behaviour.
Although theoretically compelling, one must wonder whether policy decisions are likely to
be affected by such reasoning. This question is addressed by Egeberg (1995), who suggests that
we should differentiate between ‘substantive’ policy-making on the one hand and ‘administra-
tive’ policy-making on the other. The first type is concerned with formulating proposals in the
various policy fields, e.g. transport policy. While bureaucrats might have their private prefer-
ences regarding the policy proposals to be set out, Egeberg argues that the coupling between a
certain policy proposal and its utility for different bureaucrats is rather low. Things are, how-
ever, expected to look different with the second type of policy-making, which deals with aspects
of the administrative system itself, e.g. the definition of levels of reward for top civil servants.
In this case, the coupling between bureaucrats’ self-interests and policy choices should be much
stronger and thus lead to a situation in which private interests move to the fore. On balance,
however, bureaucrats are predominantly engaged in substantive policy-making, thus indicating
that there is little risk of self-interested behaviour in regular political business.
The pursuit of self-interests more generally represents a situation in which an agent (here,
the bureaucracy) deviates from the guidelines defined by the principal (here, the politicians).
Such situations are known as agency problems. The politicization of bureaucracy presented in
the previous section aims to minimize this risk of shirking, i.e. taking actions that maximize
the benefits of the agent while imposing costs on the principal (Huber and Shipan 2002: 84).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Key points
•• Ministerial bureaucracies are heavily involved in drafting policy.
•• Ministerial bureaucrats can rely on their own expertise or on ideas put forward by experts. They
can also reduce the feasible set of policy options in accordance with certain criteria. One way of
doing this is to rely on only that kind of information provided by interest groups. Another way is
by using an ideological filter corresponding to that of their political masters.
•• Bureaucrats can also be perceived to act in an egoistic way when they propose legislation that
primarily promotes their own well-being.

I NSTITUTIONAL AND PROCEDURAL DIMENSIONS

OF DECISION-MAKING
We now turn to the theoretical approaches to formal policy adoption that, first, clearly dis-
tinguish between policy formulation and adoption and, second, underscore the institutional
and procedural dimension of policy adoption.

How Are Policy Preferences Formed?

Legislation proposed by the executive is subject to discussion on the parliamentary floor(s)
and scrutiny in the relevant legislative committees. Depending on the political system, leg-
islative committees can possess high degrees of specialization and expertise, and so disagree
with a legislative bill introduced by the executive if they deem another policy to be more
effective (see Strøm 1995: 65). As a result, legislative actors might ask for (far-reaching)
amendments before accepting a bill. Likewise, disagreement with the policy proposal can be
based on information supplied by experts.
Arguably more relevant, however, is the collaboration of decision-makers with interest
groups (see Austen-Smith and Wright 1992). As already outlined, the involvement of inter-
est groups in the policy process not only means that they can provide relevant information,
but also that they are able to ensure acceptance of a policy amongst their members. In this
regard, Fischer et al. (2019) show that the networks parliamentarians form with interest
groups are important for their legislative behaviour. Members of parliament who have simi-
lar networks of interest groups are also more likely to support similar policy proposals.
In fact, the degree to which an actor is lobbied is a good indicator of his or her policy-
shaping power. From this perspective, the increasing attention the European parliament
receives from interest groups clearly signals its growing political influence (see Lehmann
2009). For a long time, the European parliament was not regarded as a central player in mak-
ing European policy. Due to the strengthening of its formal decision-making powers, this
situation has changed. For example, in 2010 it rejected the introduction of ‘traffic light’ food
labelling. Many observers claimed that this was a direct consequence of the successful lob-
bying activities of the food industry (see The Guardian, 16 June 2010; The Daily Telegraph, 16
June 2010).
Hall and Deardorff (2006: 70) argue that the actual impact of interest groups very much
depends on the actors’ policy predispositions. This means that only if the policy suggestions
made by interest groups correspond to the views of politicians are they likely to affect policy
decisions. Consequently, the influence of knowledge-based actors or interest groups is likely
to strengthen the views that elected officials have already developed. How can one learn
about these views? The best indicator is ideology, which in most cases correlates strongly
Copyright © 2020. Macmillan Education UK. All rights reserved.

with a politician’s party affiliation.

A political actor’s mindset as to what the political world ought to be – also known as ideol-
ogy – is very likely to guide policy choices. For example, a person endorsing a social demo-
cratic ideology is likely to support the use of government to (re)distribute private resources
in order to equalize economic and political opportunities. Members of parliament with simi-
lar world views are very likely to support proposals introduced by the parliamentary party
group of the Social Democrats or a government that is composed of them. Further, mem-
bers of parliament will evaluate any policy proposal brought in through this particular ideo-
logical lens, and may or may not arrive at the conclusion that it corresponds to their
preferences.
In most political systems, members of a legislative group have almost perfect discipline,
so that, for instance, the members of parliament belonging to the majority parties normally
vote for any policy proposal made on the floor by the government (see Sieberer 2006;
Depauw and Martin 2009; Stecker 2015a). For example, if the Social Democratic party is a
member of government, the corresponding legislative group will probably accept proposals
advanced by the executive or legislative committees. By the same token, in most systems

opposition parties automatically vote against any proposal brought in by the government or
the majority parties (see Martin and Vanberg 2005; Bräuninger and Debus 2009). They do
this to signal that they can also offer policy alternatives, which is also one of the rationales
for delivering speeches in parliament (e.g. Proksch and Slapin 2014; Baumann et al. 2015;
Bäck and Debus 2016). As a result, ‘agreement may be thwarted by pressures to compete’
(Scharpf 1997a: 192). This notion of party competition is central to the stage of policy
adoption.

How are Policy Preferences Transformed into Actual Policies?

The second set of factors determining policy adoption is the institutional and procedural
arrangements in a given political system. In this context, the concept of veto points has
received some scholarly attention (Immergut 1992). Veto points refer to the fact that
policy decisions need the agreement of several, constitutionally generated, institutional
points (or links) in a chain of decisions. The adoption of legislative proposals depends
upon the number and location of opportunities for veto along this decision-making chain.
For example, a parliament represents a veto point if it can block the decision of the execu-
tive. The logic here is straightforward: the more veto points in a given political system, the
more difficult it is to gain approval for a policy proposal. While the veto point perspective
already provides an improved understanding of policy adoption, some scholars (e.g.
Weaver and Rockman 1993) argue that a more complete explanation would require tack-
ling the question of why veto points are willing to use their formal power in some circum-
stances and not in others.
This question guides a refinement of the veto point perspective that is almost exclusively
associated with the work of George Tsebelis (1995, 2000, 2002). His veto player theory
holds that policy adoption can be explained by the institutions governing the decision-
making process and the preferences of the actors involved in it, thus following the logic of
rational choice institutionalism (see Chapter 4). More precisely, however, the general expec-
tation of this model is policy stability, i.e. that it is very difficult to change existing policies.
Veto players are defined as ‘individual or collective actors whose agreement is necessary
for a change of the status quo’ (Tsebelis 2002: 19). There are two types of veto players: insti-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tutional and partisan. Institutional players are those established by a country’s constitu-
tion. So the executive and the legislature form the key veto players in any political system,
though there are additional institutions that can impede the adoption of a policy, for exam-
ple constitutional courts or referendums. Partisan veto players correspond to a more
dynamic concept, as they are formed in the course of decision-making. Tsebelis concentrates
on partisan veto players whenever possible. For the strictly illustrative purpose of this chap-
ter, however, we will only discuss institutional veto players.
For analysing veto players and their impact on policy adoption, three pieces of informa-
tion are necessary. First, one has to determine the number of players. It generally holds that
higher numbers of veto players reduce the odds of policy adoption. This reasoning becomes
quite straightforward if we compare a situation in which a government proposes a legisla-
tive bill with either a unicameral or bicameral parliament. Excluding any considerations
about the partisan composition of the parliament, policy adoption should be more likely in
the first case, as the government would simply need to ensure the approval of one potential
veto player instead of two.

Second, one has to assess the ideological distance between the actors possessing formal
veto power. This dimension generally refers to the veto players’ partisan composition, and is
calculated as the distance on a policy continuum between the most extreme player on the
left and the most extreme one on the right. For example, when conceiving of a parliament
as a veto player cooperating with a coalition government in a parliamentary system, the
ideological distance across the range of policy positions of coalition parties is decisive. The
likelihood that the parliamentary majority – which is formed by the coalition parties –
agrees with a legislative proposal should increase with lower ideological distance between
the individual factions.
Finally, as veto players are often collective actors (e.g. parliamentary party groups), it
is important to assess how cohesive they are concerning their policy interests (e.g. Stecker
2015a). High cohesion implies that, for instance, all members of a parliamentary group
have the same policy preferences and hence form a single collective actor with a marked
veto potential. In cases of low cohesion in which individual veto players have differing
preferences, their veto power can be considered to be lower, as they are likely to fail to
speak with ‘one voice’. Numerous empirical analyses have demonstrated that these three
main expectations of veto player theory essentially hold true (for an overview, see
Hallerberg 2010). Moreover, the design of the existing policy and the allocation of
agenda-setting power may be of importance for the outcome of decision-making (see
Ganghof 2017).
Which kind of veto players exist at the stage of decision-making? There is no uniform
answer to this question. In cases of minority cabinets, for instance, the cabinet does not
have sufficient seats in parliament to change the status quo. Thus, as the cabinet depends on
parliamentary support, it may or may not be confronted with the parliamentary majority
acting as a veto player, depending on the particular piece of legislation (Ganghof and
Bräuninger 2006). In this regard, Green-Pedersen (2001) explains that policy adoption has
become a smoother process in Denmark as minority governments have become more open
about bargaining and entering agreements with changing legislative groups. In contrast, in
semi-presidential and presidential systems, the president is usually a relevant veto player
because his or her assent is needed for a bill to become law. Further, constitutional courts
can act as powerful veto players (see Hallerberg 2010).
Copyright © 2020. Macmillan Education UK. All rights reserved.

There are cases that are more difficult to predict. For instance, in multiparty govern-
ments, each coalition partner can, in principle, be a veto player. However, coalition gov-
ernments usually emerge as a collective actor and do not veto policy proposals brought in
by their coalition partners (Birchfeld and Crepaz 1998), in particular if the coalition is
built on a detailed policy agreement (Müller and Strom 2008; Bowler et al. 2016). Such
behaviour is also achieved by ‘logrolling’, i.e. a situation in which parties give successive
decisions different priorities (Crepaz 2002: 174). Logrolling occurs in many legislative
assemblies in which two (or more) legislators agree to trade their own vote on one bill to
which they attach a lower importance, in exchange for the other’s vote on a bill that is
much more important to them. This practice is particularly common when the legislators
are only sporadically controlled by their national party leaders and are seeking to secure
votes for bills that concentrate benefits on their home districts, while spreading the costs
out over taxpayers in the rest of the country. This corresponds with the logic of pork-
barrel politics outlined in Chapter 2 in the context of the NATO scheme (see also Ferejohn
1974; Cox and McCubbins 1986).

Similarly, a second legislative chamber is not necessarily a veto player, but can become
one if its party composition varies from that of the first chamber, which happens quite
often in federal states (see Vatter 2005; Stecker 2016; Gagnon and Keil 2016; Däubler
et al. 2018). Another type of veto player is represented by citizens who give or refuse their
consent to a policy proposal in referendums (Hug and Tsebelis 2002). In 2009, the ‘anti-
smacking law’, which would have imposed sanctions on parents who smacked their chil-
dren, was stopped by a New Zealand referendum. In 2011, the residents of the Canadian
province of British Columbia used a referendum to stop a new government tax policy, i.e.
a new sales tax. In the same year, Italians rejected the government’s plans for nuclear
power and water privatization through referendums. In 2016, the British public voted to
leave the European Union (EU), which will have a very drastic impact on future public
policies in the United Kingdom (see Richardson 2018a, 2018b). These random examples
illustrate that in some jurisdictions referendums can be an effective means for stopping
policy proposals that are not supported by the public.
Whether the public acts as a veto player or not depends on the public’s opinion on the
policy proposal on the table. In this c ontext, deliberation scholars have emphasized the
power of deliberative instruments such as citizens’ assemblies for changing opinions (e.g.
Fishkin 1995, 2009). Analysing policy decisions in Ireland, O’Malley et al. (2019) show that
public opinion can indeed change because of participation in deliberation processes, which
comprises both an information and a deliberation effect (Box 6.2).

Box 6.2 (Super-)wicked problems: when incrementalism is intentional

Policy reform is widely conceived as desirable in policy studies and public discourse.
However, when encountered with ‘wicked’ or ‘super-wicked’ policy problems, rapidly
changing policies may not be a desirable outcome. Wicked policy problems are char-
acterized by ambiguous and contested definitions and require collaboration by mul-
tiple organizations (Termeer et al. 2015; Candel et al. 2016), as they often demand
the integration of different types of sectoral policies (Varone et al. 2013). Super-
Copyright © 2020. Macmillan Education UK. All rights reserved.

wicked policy problems are characterized by urgency, the identity of those who cause
and seek to solve a problem, weak or absent central authority, and policy responses
that discount the future irrationally (Levin et al. 2012).
The most prominent example of a (super-)wicked policy problem is the governance
of climate change. Yet it should be noted that wickedness can affect all policies, since
in essence it is about competition to resolve problems that involve a degree of ambigu-
ity. With policy problems that possess such features, it appears reasonable to trigger
‘sticky’ interventions that can endure changes in political preferences. Jordan and
Matt (2014) identify the voluntary agreement on carbon dioxide emissions from new
cars adopted by the EU in 1998 as a policy instrument designed to provide stickiness,
which allowed the European Commission to implement gradually its preferred policy
agenda on promoting a new generation of cleaner cars. Not only the content of a pol-
icy requires attention when (super-)wicked policy problems are concerned, but also
the broader policy design such as built-in revision mechanisms.

To set the implications of the veto player theory into perspective, we point to the theo-
retical concept of incrementalism which we introduced in Chapter 1 (see Lindblom 1959,
1979). While the veto player theory is based on the concept of rational choice and incremen-
talism, in contrast, relies on bounded rationality, the predictions of both theories concern-
ing decision-making are remarkably similar. An important means by which incremental
policy-making occurs is through governments practising legislative self-restraint.
Governments usually have an interest in avoiding complete legislative failure, because,
apart from anything else, failure ‘looks bad’ to the electorate, giving the impression that the
government is unable to get its preferred policies through. If a government has to give up
too much of its preferred policy position in order to avoid legislative failure, then it might
prefer to delay introducing the bill in the first place (Manow and Burkhart 2007). This
behaviour is more likely in multilevel systems with complex patterns of policy competition
(e.g. Stecker 2015b; Däubler et al. 2018). Overall, the result of legislative self-restraint is
fewer policy-making activities and relative overall policy stability.

Key points
•• During the policy adoption stage, interest groups seek to influence the relevant formal actors
to modify a policy proposal so that it fits their preferences better.
•• The actual impact of interest groups depends on the actors’ policy predispositions; only if the
policy suggestions made by interest groups somehow correspond to the views of politicians are
they likely to affect policy decisions.
•• Veto players are defined as actors whose agreement is necessary for a change in the status quo.
•• The relationship between the executive and the legislature is central to understanding decision-
making. However, this relationship changes substantially from one country to another.

ECISION-MAKING IN THE UNITED KINGDOM, AUSTRALIA, CANADA

D
AND THE UNITED STATES
In this section we illustrate decision-making in the four major English-speaking countries:
Copyright © 2020. Macmillan Education UK. All rights reserved.

the United Kingdom, Australia, Canada and the United States. We will see that there is vari-
ation with regard to the powers of the executive and the legislature, which has repercussions
for how public policy is made.

The United Kingdom

The United Kingdom is a parliamentary democracy based on a unitary structure and plural
electoral system that often produces a two-party system. The head of the executive is the
prime minister and his or her cabinet. The prime minister takes on a wider range of respon-
sibilities, which cannot be listed in detail here. Most importantly with regard to decision-
making, the prime minister ‘leads’ the government and is also the leader of the majority
party, which involves close cooperation with the parliamentary party (Budge et al. 2001:
203). This also indicates that the prime minister as the party leader has a specific standing
and is likely to influence the party’s preferences regarding public policy in a way that corre-
sponds to his or her own preferences. The prime minister and the cabinet members are
responsible for government policies and actions. They are supported by the Cabinet Office,

which is in charge of structuring government departments, allocating ministerial responsi-

bilities, developing policies and defining the broader legislative programme in parliament.
The members of the ministerial bureaucracy working inside the British executive are not
politically appointed and usually do not change when a new government comes in. Due to
their long-standing experience, ministerial bureaucrats are very important to ministers for
guiding policy decisions. As already mentioned, there are also temporary civil servants –
special advisers – who are political appointees and whose loyalties are claimed by the gov-
erning party and often particular ministers with whom they have a close relationship.
The British parliament consists of two houses: the directly elected lower House of
Commons and the upper House of Lords. Of these two, it is mainly the House of Commons
that fulfils the ‘classic’ parliamentary tasks: members can propose new laws and can scruti-
nize government policies by asking ministers questions about current issues. Party unity is
high in the House of Commons. Even when members of parliament disagree with their
party leadership, they normally vote with their party for the sake of unity.
The British government faces few legal and constitutional restraints. Despite the name
‘parliamentary democracy’, the parliament plays only a limited role in decision-making in
the British Westminster model. Policies are largely determined by the executive and,
together with the high degree of party cohesion in the House of Commons and the fact that
the majority party and the executive are strongly interlinked, draft legislation introduced
by the government is likely to pass with few changes. The opposition parties, however, are
ideologically different from the majority party and thus affect decision-making through
criticizing and opposing government proposals. The opposition parties also put forward
their own policy proposals to improve their chances of winning the next election. However,
the likelihood that these alternative proposals will be preferred over the government’s are
low. In parliamentary committees the majority party also holds the committee majority
and often controls all committee chairmanships (Strøm 1995: 65). As a result, the British
parliament can be seen as a ‘policy-modifying body, rather than a policy-making one’
(Dorey 2005: 162). In sum, the British political system is based on the notions of strong
government and leadership and a top-down policy style (Marsh and Hall 2007; Richardson
2018a; Hall et al. 2018).
The devolution process that began in the late 1990s is an important characteristic of the
Copyright © 2020. Macmillan Education UK. All rights reserved.

British polity. Under the Scotland Act 1998, the Scottish parliament can pass acts and the
executive can make secondary legislation in areas other than those which are reserved to the
Westminster government in London. Health policy, for example, is an important area in
which the Scottish government can adopt specific legislation (see Keating 2005; Cairney
2009, 2011). Likewise, Scottish social policy displays some specific traits. It is interesting to
note that the prospect of formulating more generous social policy played a key role in
nationalist campaigns for devolution and the creation of the Scottish parliament (Béland
and Lecours 2008: 139), although in the end it was created with cross-party support and not
by the nationalists.
However, policy-making by the devolved institutions is still de facto constrained by the
common security area, common market and common welfare state arrangements through
the whole of the United Kingdom (see Keating 2005: 33–43). Principally, the Westminster
parliament can legislate in devolved policy areas, but under the Sewel Convention it
explicitly needs to be asked by the Scottish parliament to do so (Leeke et al. 2003: 3).
While the policy-making powers of the Scottish legislature and executive must be
regarded as far-reaching, devolution is less advanced in Wales and Northern Ireland. Under
the Government of Wales Act 1998, powers in devolved areas have been passed to the

National Assembly for Wales, which can make delegated or secondary legislation in these
areas, though primary legislation for Wales is still made by the Westminster parliament.
However, in some areas there is wide scope for the details of the policy to be made under
secondary legislation, allowing the National Assembly for Wales some autonomy. The
Northern Ireland Assembly can make primary and delegated legislation in those policy areas
which are transferred, such as environment and health policies (Leeke et al. 2003: 29).
The devolution process is important to the extent that in devolved policy areas it is now
possible to have diverging legislation across the United Kingdom. While devolution gener-
ally does not change how the Westminster government drafts and adopts public policies, it
offers an opportunity structure for a different way of policy-making in specified areas (see
Cairney and Rummery 2018). The devolution of decision-making powers suggests that in
some areas different policies might exist in the United Kingdom due to the different prefer-
ences of key policy-makers and/or specific politics (see Keating et al. 2009).

Australia
The Australian polity resembles the British Westminster system in many ways. In Australia,
as in the United Kingdom, there are two dominant parties: the centre-left Australian Labor
Party and the centre-right Liberal Party of Australia. The Australian government – consist-
ing of a prime minister and his or her cabinet – is drawn from the legislature, and due to this
the executive tends to dominate the working of parliament. However, recent research in
comparative politics argues that Australian bicameralism (a legislature with two separate
assemblies) mixes competing models of democracy and corresponds to a ‘semi-parliamentary
government’, where ‘[t]he assembly is divided into two directly elected parts, and the cabi-
net’s survival in office depends on the confidence of only one of them’ (Ganghof et al. 2018:
214; for a discussion, see Elgie 2018; Weale 2018).
The Australian federal parliament is composed of the Senate (upper house) and the House
of Representatives (lower house). The House of Representatives performs a function corre-
sponding to the British House of Commons. However, as in the Commons, the opportuni-
ties for the House of Representatives to control the government are reduced, because the
party or coalition of parties forming the government holds a majority of seats.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Unlike the British House of Lords, which is mainly appointed, the Senate is composed of
elected representatives of each State, with representation being proportional to population.
The Senate has virtually the same legislative powers as the House of Representatives. In the
literature it is argued that the Australian Senate has become the ‘House of Review’ (Mulgan
1996), which scrutinizes the government. This results from the system of proportional rep-
resentation, which often leads to a balance of power amongst the political parties in the
Senate. There are some examples of cases in which the government has been involved in
long negotiations with a few senators over policy proposals that they did not want to pass
through parliament (Maddison and Denniss 2009: 27). Consequently, government tends to
adjust policy proposals so that they will receive the Senate’s consent. In so doing, the gov-
ernment often seeks the support of a ‘neutral’ party, e.g. the Australian Democrats, in the
Senate, that is, minor parties which are not in the House of Representatives and are there-
fore not opposition parties in the narrow sense of the term (Ganghof and Bräuniger 2006).
For example, the Senate played an important role in the policy process related to the inten-
tion of Tony Abbott’s coalition government, consisting of the Liberal Party and National
Party, to repeal carbon pricing, which was introduced by the preceding Labor government in

2012. Initially the Senate blocked the corresponding bills, but when it was newly consti-
tuted in 2014, at the second attempt, it adopted them with amendments (Crowley 2017).
The Senate is one of the key features of the federal system in Australia. The second key
feature is a division of powers and responsibilities between the federal government and the
State governments. While there are specific powers granted to the federal government,
most are concurrent and can be exercised by both the Commonwealth and the States. The
States are the entities that mostly provide public services and the regulations experienced
by citizens. Since both levels of government possess policy-making powers, the Australian
High Court fulfils an important task in settling disagreements about which government has
power over particular issues.
Judicial review represents the third key feature of Australian federalism (Parkin and
Summers 2006: 52–3). Summers (2006: 138) contends that, on balance, the effect of judicial
review has been to increase the power of the Commonwealth at the expense of the States,
especially with regard to financial issues. Notwithstanding these shifts in power, the States
retain their power to make public policy on a wide range of issues such as education and
health services.
Thus, for understanding decision-making in Australia, it is important to investigate pro-
cesses at both levels of government. On the one hand, this entails that the political constel-
lations in each single State are of interest for explaining policy outputs; on the other hand,
the interplay of the two levels is worth studying in detail, especially when some or all State
governments have policy preferences that diverge from those of the federal government.
The literature argues that the federal system in Australia provides an opportunity structure
for policy experiments, which can trigger a more effective policy response (Chappell and
Curtin 2013). It should be kept in mind that the Senate as a veto player directly results from
the country’s federal polity (see Bach 2003), which represents a marked contrast to the
decision-making situation in the United Kingdom.

Canada
A Westminster-style polity co-existing with federalism is also the major characteristic of the
Canadian political system. When appointing ministers to a cabinet, i.e. the supreme policy-
Copyright © 2020. Macmillan Education UK. All rights reserved.

making body, the Canadian prime minister usually selects at least one per province. The
ministers forming the cabinet make policy decisions together and are collectively responsi-
ble for them. As in the United Kingdom, the lower chamber is the House of Commons,
where seats are distributed roughly in proportion to the population of each province and
territory. The Senate consists of 105 members appointed by the governor general, i.e. the
representative of the Canadian monarch, on the advice of the prime minister. Seats are
assigned on a regional basis. On paper, the Canadian Senate has a veto power. In reality,
however, most analysts argue that it is reluctant to use this power due to its lack of demo-
cratic legitimacy (see Russell 2001); hence, it rarely rejects bills passed by the directly elected
House of Commons.
Compared with the United Kingdom and Australia, there is a higher number of political
parties. In the current legislature, five are represented in the House of Commons: the
Conservative Party (right wing), the New Democratic Party (left wing), the Liberal Party
(centre left), the Bloc Québécois (nationalist) and the Green Party. Especially in recent years,
party discipline in the House of Commons has been very strictly enforced (see Penner et al.
2006: 1008), which obviously affects policy-making because majority parties have an

advantage over those holding only a minority of seats. Further, there have been several
minority governments, which have had to rely on the support of other parties to stay in
power. Concerning policy-making, minority governments are likely to engage in incremen-
tal policy change as they need to gain the consent of the other parties. This situation repre-
sents another major difference vis-à-vis the United Kingdom.
In Canada, federalism is constitutionally entrenched against unilateral modification by
either level of government. In addition, limits on the authority of both the federal and the
provincial governments are judicially enforceable. Howlett and Lindquist (2004: 234) main-
tain: ‘It is difficult to overstate the complexity of Canadian federalism and its supporting
policy institutions in such a huge, regionally and linguistically diverse country, with prov-
inces and territories of starkly different fiscal, population and land bases.’ Executive federal-
ism has long been considered the defining characteristic of Canadian intergovernmental
relations, i.e. relations dominated by the executives of the different governments within the
federal system (Watts 1989: 3), and in recent years it has been increasingly informed by a set
of practices known as ‘collaborative federalism’, which is characterized by a codetermination
of broad national policies. While codetermination in the Canadian context generally involves
the two orders of government working together as equals, it can also entail provincial and
territorial governments taking their own initiative – acting collectively in the absence of the
federal government – to formulate national policy (Cameron and Simeon 2002). In this con-
text, a survey administered by Howlett and Wellstead (2012) revealed that the individual
provinces offer different opportunity structures for professional policy work.
By and large, policy-making in Canada is characterized by the absence of a veto power being
exerted by the provinces, as the Senate usually does not reject policy proposals that have been
passed by the House of Commons. Nevertheless, policy adoption can become complicated if
the government is a minority one, dependent on the support of the other parties, as this may
require giving up preferred policy positions (see Dobuzinskis and Howlett 2018).

The United States

Legislative decision-making in the United States is highly structured by the strong degree of
Copyright © 2020. Macmillan Education UK. All rights reserved.

federalism on the one hand, and by the dualism between the president and Congress on the
other. Despite the fact that the office of the president – who serves as the head of state and
the executive, is the commander-in-chief of the military and nominates judges to the
Supreme Court – is seen as the most important in the world, many of the president’s deci-
sions need the approval of both houses of Congress, i.e. the House of Representatives (the
lower house) and the Senate (the upper house). More generally, US politics is characterized
by a very well-developed system of checks and balances, which gives each of the branches of
government – the legislative, the executive and the judiciary – some degree of oversight and
control over the actions of the others.
At the national level, the power of the president is limited with regard to the process of
legislative decision-making. First, he or she cannot introduce a law proposal in both houses of
Congress directly: he or she needs a member of the House of Representative or the Senate to
insert a bill. As a result of the institutional relevance of the legislature in the US system, legis-
lative committees are strongly specialized and possess high expertise in a wide range of areas
(Strøm 1995: 65). Congress rather than the president and his or her ministers is the target of
interest group activities (Werner and Wilson 2008). However, each law adopted by Congress
has to be signed by the president. If he or she refuses to sign a bill approved by the majority in

the House of Representatives and the Senate, the proposal goes back to both houses of
Congress. Only if the president’s veto is overridden by a two-thirds majority, inside both the
Senate and the House of Representatives, can a bill rejected by the president become law
(O’Connor and Sabato 2009: 258–62).
The at times conflictive relationship between the president and the Congress is exemplified
by the Affordable Care Act launched in 2014 by the Obama administration. Following the pas-
sage of this law, the House of Representatives advanced over 50 bills to repeal it in whole or in
part. When President Obama was in office he vetoed the bills introduced to challenge the law.
With the entering into office of President Trump and the unified Republican government that
held power during the first two years of his administration, the legislative challenges to the
Affordable Care Act became serious (McIntyre and Song 2019).
Another interesting observation in this regard is that the US cabinet, which consists of so-
called secretaries (i.e. heads of executive departments) who assist the president in executing
laws and making decisions, is only an informal institution, which again distinguishes the US
political system from those presented previously.
This clarifies that in terms of legislative decision-making not only do the policy prefer-
ences and thus the partisan affiliation of the president matter, but also the partisan compo-
sition of both houses of Congress and the degree of majority possessed by the two main
parties – i.e. the Republicans and Democrats. Although there also exist some other political
parties, the electoral system in which the political party that receives most votes wins the
election – known as the ‘winner-takes-all system’ – favours their concentration. Moreover,
it should be kept in mind that US political parties are more loosely organized and there is
more intra-party variation concerning policy positions compared the ideologically cohesive
political parties in other countries. However, party politics matters in the United States, and
this becomes most apparent in decision-making in Congress, which has far-reaching compe-
tences regarding budgetary issues.
In contrast to the four-year term of the president, the complete House of Representatives
and one-third of the Senate face re-election every second year. During ‘mid-term’ elections,
voters tend to vote for the candidates of the party that does not control the presidency at
that time. As was the case after the 2010 elections for Congress, what is called ‘mid-term
loss’ (Gaines and Crombez 2004) often results in majorities for the opposition parties in at
Copyright © 2020. Macmillan Education UK. All rights reserved.

least one chamber of Congress. These patterns of divided government can result in legisla-
tive gridlock between the executive and the legislature, which often results in policy com-
promises between both partisan camps and, therefore, in moderate policies being adopted.

Key points
•• The British executive is particularly powerful due to its predominantly two-party system and
the high degree of party coherence.
•• In Australia, the Senate, which is composed of the representatives of each state, can force the govern-
ment to modify its policy proposal before agreeing to give its consent to it.
•• In Canada, the Senate is a weak player, with federalism unlikely to complicate policy-making at the
national levels. Policy adoption can, however, require the government to give up policy positions if it is
a minority government relying on the support of other parties.
•• In the United States, the number and power of veto players are much higher. On the whole, this makes
policy adoption more complicated, often leading to substantial modification between the initial policy
proposal and the form of law that is adopted in the end.

CONCLUSIONS

Which factors determine the design of a policy proposal? When does a policy proposal more
easily gain adoption? These have been the research questions guiding this chapter. In
response to the first question, we note that in parliamentary systems the (initial) design of
a policy proposal is mostly determined by the executive, comprising both the cabinet and
the ministerial bureaucracy. While cabinet members are important for placing a certain
issue on the policy agenda, it is the ministerial bureaucracy that develops the actual pro-
posal. In presidential systems, however, legislation is introduced by the legislature, entailing
that the ministerial bureaucracy and the responsible legislative committees have to cooper-
ate to a greater degree than is the case in parliamentary systems.
The actual design of a policy proposal depends on the preferences of competent actors.
This can result from their own expertise, information supplied by experts, international
organizations and/or interest groups. Further, with regard to the dominance of ministerial
bureaucrats in drafting legislation in parliamentary systems, we have stressed that they
might, in principle, deviate from political guidelines and propose legislation that corre-
sponds to their own interests. This does not occur with elected officials, as they are demo-
cratically accountable for their actions. Additionally, party discipline can prevent individual
members of the legislature from coming up with policy proposals that do not correspond to
the interests of the legislative groups. One way to limit the risk of shirking by bureaucratic
actors is through politicization.
The legislative proposal prepared by the executive or legislature is then subject to parlia-
mentary approval. In this process, draft legislation might still undergo more or less funda-
mental changes. The extent of these changes strongly depends on whose consent is required
for passing a law – in other words, the existence of veto points or veto players. Usually, hav-
ing a smaller number of actors possessing formal veto powers should speed up the decision-
making process. Likewise, the ideological distance between veto players and – in the case of
collective veto players – their internal cohesion are important for being able to predict the
outcome of the policy adoption process.
Copyright © 2020. Macmillan Education UK. All rights reserved.

WEB LINKS

www.internationalbudget.org. This website provides all kinds of information regard-

ing budgets and the making of budgets.
www.dpmc.govt.nz/our-programmes/policy-project. The Policy Project of the govern-
ment of New Zealand aims to build a high performing policy system.
www.parliament.uk. Here information is supplied about aspects of decision-making
in the British parliament.
www.pippanorris.com. Professor Norris maintains a very instructive website on
which she provides empirical data on the institutional characteristics of political
systems.

www.polidoc.net. On this website election manifestos and other policy-relevant

documents of political parties are presented and can be downloaded.
www.mgmt.wharton.upenn.edu/faculty/heniszpolcon/polcondataset/. Professor
Henisz has developed a system for measuring political constraints. The resulting
Political Constraint Index dataset can be downloaded from this website.

FURTHER READING

Aberbach, J. D., Putnam, R. A. and Rockman, B. A. (1981). Bureaucrats and Politicians
in Western Democracies. Cambridge, MA: Harvard University Press. A classic read
about the relationship between elected and appointed policy-making actors.
Bickerton, J. and Gagnon, A. G. (eds.) (2014). Canadian Politics. Toronto: University
of Toronto Press. This authoritative collection gives a useful entrée into the polit-
ical system of Canada and how policy-making happens there.
Croissant, A. and Lorenz, P. (2018). Comparative Politics of Southeast Asia: An intro-
duction to governments and political regimes. Cham: Springer. This is a recom-
mended complement to the discussion of the political systems presented in this
chapter.
Dobuzinskis, L. and Howlett, M. (eds.) (2018). Policy Analysis in Canada. Bristol:
Policy Press. This edited volume is an ideal complement to the politics perspective
on the political system of Canada provided by Bickerton and Gagnon.
Gagnon, A. G. and Keil, S. (2016). Understanding Federalism and Federation.
New York: Routledge. This volume provides a comprehensive overview of differ-
ent types of federalism and how they affect decision-making.
Jordan, A. J. and Turnpenny, J. R. (eds.) (2015). The Tools of Policy Formulation:
Actors, Capacities, Venues and Effects. Cheltenham: Edward Elgar. This volume
examines the analytical value of policy formulation tools such as indicators and
cost–benefit analysis for public policy.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Reader’s guide
In Chapter 2 we argued that a public policy is intended to solve a certain social prob-
lem that has reached the institutional agenda. As a rule, the problem that initiated the
policy-making process can only be solved effectively if the adopted policy is properly
put into practice. If a given policy is introduced but insufficiently implemented, it is
possible that the ultimate result will be even less desirable than the previous state.
Implementation research has demonstrated that it is anything but a straightforward
task to put public policies into practice. We will approach the implementation of pub-
lic policy from different analytical angles, including a clarification of which actors
implement public policy and a presentation of the major theoretical perspectives on
implementation activities. We then propose categories for assessing the degree to
which a policy can be implemented effectively and identify factors that might hamper
or facilitate implementation.

INTRODUCTION
Policy implementation is the stage in the policy-making process where a policy is put into
effect by the responsible actors and agencies. In more technical terms, implementation
Copyright © 2020. Macmillan Education UK. All rights reserved.

involves the transformation of a policy output into a policy outcome. The attainment of the
intended policy outcome is a necessary condition for bringing about the desired policy
impact. However, it should be kept in mind that only the policy outcome can be directly
affected by the implementers, not the policy impact, since the latter might be affected by
additional factors. Therefore, all arguments made in this chapter solely concern the process
of how policy outputs are turned into policy outcomes, leaving to one side considerations
about policy impacts.
Analysing policy implementation always involves a judgement about the intended policy
outcomes on the one hand, and about those actually achieved on the other. Indeed, Pressman
and Wildavsky (1973) regarded policy implementation as the ‘missing link’ between policy-
making and evaluation. In their classic study, they expected there to be widespread ‘imple-
mentation problems’, i.e. when a policy cannot be put in place as intended by policy-makers
due to the implementers’ failure to overcome obstacles.

122
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:19:08.
Implementation 123

What happens to a policy after it is officially passed from the government to the legisla-
ture? How do the formal transposition and the practical application of legal acts take shape?
Which problems and deviations from the initial objectives can be observed? These questions
lie at the heart of this chapter.

WHO IS INVOLVED IN POLICY IMPLEMENTATION?

The policy implementation process involves a multitude of actors and organizations.

Vancoppenolle et al. (2015) consider the existing implementation structure, decision-
making within implementation agencies and target group behaviour as the three consti-
tutive elements of the implementation process. This distinction is inspired by the
integrated implementation model set forth by Winter (2012), which regards the imple-
mentation process as influenced by the previous policy process and the selection of a
policy design.

Implementation Structure
Implementation structure refers to the formal organizational arrangements that have been
set up for implementing a policy. In most polities, policy implementation is carried out by
different levels of state bureaucracy. At the central level, there are the various national min-
istries (education, defence, trade etc.), which form the core of the executive branch. Also at
the central level, there are (quasi-)autonomous agencies located outside of the ministries
that are charged with implementing public policy. These are set up with specialized expertise
for dealing with complex or new policy areas. For example, in 1970 the US government
founded the Environmental Protection Agency (EPA), which is responsible for developing
and enforcing environmental regulations. In the United Kingdom since 1995, the
Environment Agency has performed similar tasks with regard to the protection of the envi-
ronment. Likewise, the Copenhagen-based European Environment Agency, which started to
Copyright © 2020. Macmillan Education UK. All rights reserved.

operate in 1994, provides information on the environment and guides the environmental
policy of the European Union (EU).
This international trend of delegating implementation – and often also policy formula-
tion – competencies to (quasi-)autonomous agencies is based on the assumption that they
will improve overall implementation performance and efficiency (Bouckaert and Peters
2004). This is also related to the idea of separating politics from public administration and
insulating certain decisions from political considerations (Jordan 1997), which is expected
to minimize deviations from the original intentions of policy-makers and prevent delays
(see Pollitt and Talbot 2004; Pollitt et al. 2004). Regulatory agencies are a subgroup of such
(quasi-)autonomous agencies, and one of their main tasks is to implement public policy in
market-related areas, such as electricity, telecommunications or water supply. These bodies
carry out regulation using their own delegated power, resources and responsibilities (see
Gilardi 2008, Maggetti 2012; Box 7.1).

Box 7.1 Improving policy implementation in the United Kingdom

In the United Kingdom, policy-makers have paid particular attention to how pol-
icy implementation – also known as policy delivery – can be improved, culminat-
ing in two prominent government programmes. The starting point of this
reflection exercise was the fragmentation or ‘pillarization’ of public administra-
tion, making the delivery of policies and services ineffective and inefficient.
Citizens needed to file requests with multiple agencies in order to benefit from
services. As a result, the Blair government (1997–2007) advanced the concept of
holistic government in the mid-1990s, predominantly in the areas of housing policy
and disaster management, in order to increase cooperation amongst the respon-
sible agencies.
The second government concept that emerged in the United Kingdom in the
early 2000s with the publication of the report Wiring it up: Whitehall’s manage-
ment of cross-cutting policies and services, replacing holistic government, was
joined-up government. This concept introduced new forms of administrative coop-
eration, enabling public authorities to work together, and new types of account-
ability and incentives in order to effectively align different public authorities. In
2001, Tony Blair also created the Prime Minister’s Delivery Unit, which was in
charge of tracking and improving performance in priority areas such as health
and education.
These governance principles have been put into question by subsequent gov-
ernments calling for stronger decentralization and autonomy of local service
delivery, in particular the coalition government of Conservatives and Liberal
Democrats (2010–2015). In 2010, the new government axed the Prime Minister’s
Delivery Unit as it was perceived as a mechanism for driving top-down targets,
which collided with Prime Minister David Cameron’s decentralization agenda,
only to replace it with the Implementation Unit in 2012, which resembles the old
unit in many aspects (Matthews 2015; Gold 2017). While challenged in the United
Kingdom, the Prime Minister’s Delivery Unit was copied by countries around the
Copyright © 2020. Macmillan Education UK. All rights reserved.

world as diverse as Albania, Canada, Jordan, Pakistan, Paraguay and Sierra Leone,
just to name a few (Gold 2017).

While centrally located ministries and autonomous agencies play an important role in
policy implementation, a large part of national policies is also implemented by public
entities at the local level. Employment and welfare policy is usually carried out by local
employment agencies (see Vancoppenolle et al. 2015). In Germany, for example, employ-
ment agencies do not only support job seekers in finding employment; they are also
responsible for processing applications for many kinds of welfare benefits, including
child benefits, business start-up and part-time working allowances, unemployment ben-
efits and insolvency payments, for providing vocational and further training, and for the
integration of older and disabled people into the labour market (Shore and Tosun 2019).

In a similar vein, the 20 District Health Boards in New Zealand, created by the Public
Health and Disability Act 2000, are responsible for implementing national health policy
(Ashton et al. 2008). The objectives of the District Health Boards include the integration
and delivery of health services, the delivery of personal health services or disability sup-
port, the inclusion and participation in society of people with disabilities, and the reduc-
tion of health outcome disparities. The District Health Boards fund and own the public
hospitals.
There are also public policies that are implemented by multiple organizations, which
might even be located at different levels of government. Recent empirical studies have
pointed out that a considerable share of legislation requires multi-organizational struc-
tures (see Tosun and Lang 2017; Trein et al. 2019). In this context, the role of interor-
ganizational networks has been emphasized, as they can provide an effective means of
coordination (Peters 2015). Such networks are relatively stable structures through
which individuals and organizations act in situations in which they are dependent on
the contributions of others (Hanf and O’Toole 1992). Drug policy is an area that typi-
cally requires the collaborative efforts of different ministries and agencies to implement
it properly due to its intersectoral character, which can involve the ministry of health
and local health offices as well as other ministries and/or agencies working in the fields
of education, finance, economy, trade, foreign affairs and criminal justice (see, e.g.,
Percival 2009).
Another scenario is policy implementation by private actors, which can include arrange-
ments in which policies are delivered by private actors only or by ‘hybrid governance’ in
which private and public actors collaborate (Knill and Lehmkuhl 2002a, 2002b; Thomann
et al. 2018). For example, in some countries such as England and Wales, drinking water
and sanitation are provided exclusively by for-profit organizations (Schiffler 2015).
Another example refers to the supply of car fuels that consist of a mix between conven-
tional fuel and bioethanol, which can be purchased, for example, in Finland, France,
Germany and Sweden (Tosun 2018). In this case, the filling stations have to implement
government policy on the promotion of such fuel blends by selling the products. And in
poor and tribally populated areas of India private welfare organizations such as Save India
provide health and education services (Mukherji 2017). A hybrid governance arrange-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ment can be found in Switzerland with the legal requirements for veterinary medicinal
products. Both public inspectors and private veterinarians are responsible for the imple-
mentation of the policy (Sager et al. 2014).

Decision-Making within Agencies

Decision-making within agencies is about the formal authority to make decisions in order to
put a policy into practice (Vancoppenolle et al. 2015). Depending on the specific policy to be
implemented, agencies can have differing degrees of decisional discretion. For example, in the
case of unemployment benefits, specific conditions are defined for eligibility, the levels of ben-
efits and the duration of their payment. Therefore, most bureaucratic decisions on the imple-
mentation of unemployment benefits involve application of the conditions specified in the
legislation to the facts of the case at hand.

Bureaucracies might, however, have greater leeway in the interpretation of policy

outputs if these are characterized by vague policy contents or goals. An example of such
a policy goal that is difficult to define is the achievement of sustainable development,
which is a very broad concept that involves an ecological, a social and an economic
dimension. At the United Nations (UN) Sustainable Development Summit in 2015,
more than 150 countries adopted the 2030 agenda for sustainable development, includ-
ing specific Sustainable Development Goals (SDGs), with the aim of making the dimen-
sions of sustainable development more concrete and facilitating their implementation.
Yet even the scope of the concept as reflected by the SDGs still entails challenges for
how to put them into practice (Nilsson and Persson 2017; Tosun and Leininger 2017).

Target Group Behaviour

The third group that participates in the implementation process is the target group.
Only if the target group changes its behaviour in the intended manner can a policy be
said to be implemented effectively (Vancoppenolle et al. 2015). The target groups of a
policy vary strongly, ranging from individuals who file a request to receive welfare ben-
efits to companies and their respective production processes. Public entities themselves
can also be the target groups of a policy. For example, policies aiming to promote New
Public Management (Hood and Dixon 2013; Christensen and Lægreid 2016) target pri-
marily public entities and seek to bring about behavioural changes amongst civil ser-
vants and other members of these organizations.
All in all, policy implementation involves multiple actors and the focus of implementa-
tion research can be on the implementation structure, decision-making within agencies
and/or target group behaviour.

Key points
•• Policy implementation involves multiple actors.
•• Policy implementation can be understood as a process during which policy outputs are transformed
into policy outcomes, including in particular implementation structures, decision-making within
Copyright © 2020. Macmillan Education UK. All rights reserved.

agencies and target group behaviour.

•• The achievement of intended policy outcomes does not necessarily mean that a given policy also
achieves its intended policy impacts, such as the reduction of poverty levels.
•• Most analytical perspectives regard public entities as implementing agencies, but they can also
be a target group of policies and therefore need to change their behaviour in order to comply with
the content of the policy concerned.

ANALYTICAL PERSPECTIVES IN IMPLEMENTATION RESEARCH

At the most basic level, implementation is about putting public policy into practice. While at
first glance this definition is straightforward, the empirical analysis is complex as it encom-
passes – as seen in the previous section – various actions by public and private actors that
are directed at the achievement of certain goals specified in terms of their policy outputs
(Van Meter and Van Horn 1975: 447).
To illustrate this point, let us assume that a government has adopted a policy that requires
food products to be labelled in a transparent and easy-to-understand manner in order to

increase consumer trust (i.e. the policy objective). The food industry (i.e. the target group)
may not comply with the requirements unless its products are to be scrutinized by the
responsible administrative agency (e.g. the food standards agency). Thus, to achieve the
intended policy outcome, compliance with the relevant policy has to be monitored and if
necessary also enforced, for example by issuing fines. However, an alternative policy could,
in principle, demand that a specific website is established on which consumers can complain
about food products that are insufficiently labelled. Again, the target group of the policy
would be the food industry and the policy outcome a modification of product labelling. This
time, however, the activity of the implementers would involve the establishment and main-
tenance of the requested website. Many more examples could be given here, but what mat-
ters is that there is not only one way of implementing policy outputs and often implementation
requires multiple actions, which further complicates analysis.
The diversity of activities is reflected in the theoretical perspectives adopted to describe
and explain policy implementation. The traditional approach is characterized by a top-down
perspective, which concentrates on policy outputs and investigates the extent to which the
intended objectives have been achieved over time and why. In the late 1970s and early
1980s, the bottom-up perspective emerged. This perspective analyses the multitude of
actors who interact at the operational level of a particular policy issue. Bottom-up models
usually stress the strong interlinkages between the stages of policy formulation, implemen-
tation and reformulation (Sabatier 1986: 22). More recently, hybrid models of implementa-
tion have been advanced, which seek to overcome the divide between the other two
approaches by incorporating elements of top-down, bottom-up and other theoretical mod-
els (Pülzl and Treib 2007: 90). In this section, we present the main characteristics of these
three approaches.

Top-Down Models of Policy Implementation

In their path-breaking study on policy implementation, Pressman and Wildavsky (1973)
demonstrated that great deviations and shifts in policy goals can occur during the imple-
mentation phase. The subtitle of their book concisely summarizes the central finding of the
analysis: How great expectations in Washington are dashed in Oakland; or, Why it’s amazing that
Copyright © 2020. Macmillan Education UK. All rights reserved.

federal programs work at all. According to the authors, deviations in policy goals are likely if
action depends upon a number of actors who are required to cooperate. Hence, the longer
the implementation chain and the greater the number of actors involved in the process, the
more difficult implementation becomes.
The approach of Pressman and Wildavsky corresponds to rational choice theories, as they
conceive of implementation as purposive action by different groups of actors with different
preferences. Since the success of implementation depends on the cooperation of actors,
there is a relatively high chance that problems will occur, which implies a mismatch between
intended policy objectives and actual outcomes. Put this way, implementation can be anal-
ysed by means of game theory, as outlined in Chapter 4. So implementation can be modelled
as a ‘one-shot’ or a repeated game (see also Bardach 1977). The reasoning about the impor-
tance of the number of actors involved in the implementation process (i.e. the length of the
implementation chain) resembles the argument underlying the theories of veto points and
veto players presented in Chapter 6. In both cases, during decision-making and implementa-
tion, a greater number of actors can lead to conflict over policy goals, as well as delays and
deviations from what policy-makers originally intended.

The central finding of Pressman and Wildavsky that shifts in policy objectives can be fre-
quently observed during the implementation stage is closely associated with the top-down
approach (see Smith 1973; Van Meter and Van Horn 1975; Hood 1976; Sabatier and
Mazmanian 1980; Mazmanian and Sabatier 1983; Hogwood and Gunn 1984; Sabatier 1986).
Most importantly, this perspective makes a clear distinction between the stages of policy for-
mulation and implementation (Hill 2009: 196). It is only on this basis that an actual compari-
son between policy requirements and their degree of actual implementation is possible. The
degree of goal attainment serves as an indicator of implementation success; and effective
implementation corresponds to a match between policy objectives and outcomes. If the objec-
tive of national legislation is, for instance, to set a certain standard for industrial emissions
into the air, effective implementation is achieved as soon as the prescribed emission levels are
met by industry. Likewise, an effective implementation of bans on smoking in public buildings
requires that appropriate steps be taken by the authorities in these buildings to make sure that
nobody smokes inside. This can be achieved by no-smoking signs, the removal of ashtrays,
monitoring and the announcement of sanctions.
In this context, Van Meter and Van Horn (1975) put forward a model in which they com-
bine the characteristics of the policy to be implemented, institutional characteristics of the
implementation agencies and how these are interrelated with other relevant organizations,
contextual factors (including the economic, social and political environment) and the
response of the implementers to all these factors. Based on this model, the authors hypoth-
esize that implementation will be most successful when the policy output only requires mar-
ginal changes compared with the status quo and when goal consensus amongst the public
and private actors involved is high.
The logic of the top-down perspective is also well illustrated by the four-step model
suggested by Sabatier and Mazmanian (1980). The model first addresses the extent to
which the actions of implementing officials and target groups were consistent with the
objectives and procedures outlined in a public policy. Next, they focus on the extent to
which the objectives were attained over time. Third, they evaluate the principal factors
affecting policy outcomes. Finally, they suggest analysing whether and how the policy was
reformulated on the basis of experience. Even though the main characteristics of policy
evaluation will only be introduced in Chapter 8, it is apparent that the analytical perspec-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tive advanced by Sabatier and Mazmanian is ‘evaluative’ (see Laws and Hajer 2008: 411),
especially when looking at the fourth step, which points out a feedback process (Hill and
Hupe 2009: 49). This observation holds true more generally for the various top-down
models, as they all compare actual with expected policy outcomes in order to assess the
degree to which they are congruent.
While the top-down perspective has produced many interesting insights into implemen-
tation processes, it has also been met with three sets of criticisms (Matland 1995: 147–8).
First, this perspective takes policy outputs as the starting point of analysis and disregards
actions taken earlier in the process, especially during policy drafting.
Second, top-down models tend to see implementation as a purely administrative pro-
cess that ignores political aspects. However, as already discussed in Chapter 4, bureau-
cratic actors may not be ‘Weberian’ in nature and always make independent decisions
based on technical criteria, but might also pursue their own interests in accordance with
the argument put forward by Niskanen (1971) and other theorists of public choice (for an
overview, see Mueller 2003). The criticism concerning the lacking ‘political’ perspective on
policy implementation is addressed by newer research on policy implementation that

examines blame management strategies of politicians who fail to deliver a certain policy.
Large-scale infrastructure projects are particularly susceptible to implementation failure,
and politicians have adopted a specific strategy in communication of these: they rely on
the stepwise announcement of delays and cost overruns (Hinterleitner 2019).
Third, top-down models have been criticized for not taking into account local actors and
the particular conditions for policy implementation at the ‘street level’. These points of criti-
cism paved the way for bottom-up models of policy implementation.

Bottom-Up Models of Policy Implementation

Bottom-up models regard effective implementation in a process-oriented way that aban-
dons the divide between policy formulation and implementation (see Berman 1978;
Lipsky 1980; Hjern and Porter 1981; Hjern 1982; Hjern and Hull 1982). Policy objec-
tives and instruments are no longer defined as benchmarks to be reached; instead, it is
expected that they may undergo modifications during the process of implementation.
Implementers have flexibility and autonomy to adjust policy in the light of particular local
requirements and changes in the perception or constellation of policy problems, as well
as new scientific evidence on the causal relationships between means and ends. Hence,
effective implementation is not measured by the attainment of a certain centrally defined
objective, but judged by the extent to which the perceived outcomes correspond with the
preferences of the actors involved. The crucial question for evaluating implementation
success is the extent to which a certain policy allows for processes of learning, capac-
ity-building and support-building in order to address problems associated with it in a
decentralized way, consistent with the interests of the actors involved (see Wilson 1989;
Schneider and Ingram 1997).
It is important to note that policy implementation occurs at two levels. First, there is
the macrolevel comprising central actors that devise a policy output. Second, local actors
at the microlevel react to macrolevel policies, develop their own programmes and imple-
ment them. According to Berman (1978), implementation problems stem from the fact
that macrolevel implementation cannot influence microlevel implementers, leading to a
variation in how the same national policy is implemented at the local level (Matland
Copyright © 2020. Macmillan Education UK. All rights reserved.

1995: 148). This corresponds to the logic of the principal–agent problem, which we
outlined in Chapters 2 and 6: the agent (here, the local implementers) can be inclined to
deviate from the principal (here, the centrally located actors). We will return to this
point later in this chapter, since agency problems represent an important source of
imperfect implementation.
In this regard, Lipsky (1980) argues that the likelihood of local implementers or
street-level bureaucrats’ deviating from centrally defined policy objectives stems from
pressures imposed on them and how they cope with them. They develop methods of
providing a service in a relatively routine way. These local implementers are oppressed
by the bureaucracy within which they work, and yet they possess discretionary freedom
and autonomy. Against this background, street-level bureaucrats make choices about
the use of scarce resources under pressure. So increasing the monitoring of local imple-
menters would not reduce the odds of imperfect implementation, but increase the ten-
dency to provide routine services and variations at the local level, as routines might
vary from one local unit to another (see Hill and Hupe 2009: 52–3). Therefore, Hjern

and Porter (1981), Hjern (1982) and Hjern and Hull (1982) suggest that bottom-up
studies of implementation will be particularly insightful if they focus on microlevel
implementers and their goals and preferences as well as the constraints they face (see
also Grohs et al. 2016; Shore and Tosun 2019).
This conception challenges the simplifying assumptions of the top-down perspective
and tries to take into account the complexity of implementation processes. Thus, it is
emphasized that the formulation of clear-cut objectives often contrasts with the interests
of politicians, who have a preference for vague and ambiguous objectives in order to facili-
tate a positive evaluation later and to make detection of potential failures more difficult.
In addition, the bottom-up perspective accounts for the fact that implementation pro-
cesses are rarely characterized by a clear delineation of competencies between the political
and administrative actors involved at different institutional levels. Hence, implementa-
tion is based less on hierarchically defined and controlled requirements, and instead can
be understood as bargaining between a great number of public and private actors as well
as administrative agencies participating in the implementation process.
The precise mapping of the complexities of the implementation process inherent in the
bottom-up approach also introduces problems when it comes to the measurement of suc-
cess. Since effective implementation is not measured on the basis of a comparison between
initial objectives and actual achievements, but on the extent to which goals have been
reached by taking into account the specific conditions ‘on the ground’, general and com-
parative assessments of effectiveness are difficult (Knill and Lenschow 2000). In addition,
there are two further frequently expressed criticisms of bottom-up models (Matland
1995: 149–50). First, there is the normative criticism that in democratic polities local
implementers should be subject to central control. Second, bottom-up models tend to
overemphasize the level of local autonomy, as often it is the policy itself that defines how
it should be implemented.

Hybrid Models of Policy Implementation

In view of the specific problems of both top-down and bottom-up models, there have been
Copyright © 2020. Macmillan Education UK. All rights reserved.

increasing efforts to combine the two perspectives and enrich them with additional theo-
retical approaches (see Mayntz 1979; Windhoff-Héritier 1980; Elmore 1985; Sabatier 1986;
Matland 1995; Winter 2003; O’Toole 2003). As a rule, hybrid models seek to integrate the
‘macro world’ of the policy-makers with the ‘micro world’ of the implementers (McLaughlin
1987: 177). One possibility is to integrate both approaches into one model (see Goggin et al.
1990; Winter 2003). Other authors (see Sabatier 1986; Matland 1995) identify additional
conditions of scope that in a given constellation render a distinctive approach more or less
analytically promising or suitable. These conditions include the ambiguity of political objec-
tives, the level of political conflict surrounding a policy decision, the complexity of policy
networks and government capacity (see Linder and Peters 1989; Knill and Lenschow 2000;
Pülzl and Treib 2007).
Particularly insightful is the hybrid model of policy implementation advanced by
Matland (1995), which analyses the ambiguity and conflict levels of policies to determine
whether a top-down or a bottom-up approach is more appropriate for explaining a par-
ticular implementation process. Policy ambiguity refers to a lack of clarity of goals and/or
means of achieving them. Policy conflict is the difference between the most preferred

outcome of an implementation agency and the output that the agency has to implement.
The conception of these two dimensions as being high or low gives way to four ideal-typi-
cal implementation processes: administrative, political, experimental and symbolic.
Administrative implementation involves low policy ambiguity and low policy conflict.
Policy outcomes are determined by resources, and the process is compared to a machine,
which is the central policy-making authority. All in all, administrative implementation fits
with the top-down model.
Political implementation involves low policy ambiguity and high policy conflict. In such
constellations, actors have clearly defined goals, though there is dissent when those goals
are incompatible or a conflict occurs over the means of achieving them. Therefore, policy
outcomes are determined by power or bargaining, which clearly indicates the top-down logic
underlying this implementation process.
The remaining two implementation processes build on the bottom-up approach.
Experimental implementation refers to a situation of high policy ambiguity and low policy
conflict. Following this model, policy outcomes depend on the resources and actors present
in microlevel implementation, which are likely to vary from context to context. As this
implementation process emphasizes the relevance of contextual conditions and the role of
chance, it parallels the garbage can model (Cohen et al. 1972) and the multiple streams
approach (Kingdon 2003/2011). From this it follows that policy outcomes are hard to
predict.
Symbolic implementation involves a situation in which there is high policy ambiguity and
high policy conflict. The central principle is that coalitions of actors at the local level exist who
control the available resources. However, the power of the various actor coalitions is again
determined by contextual conditions. The preferences of actors’ groups are likely to be based
on their professional training. Amongst groups trained in different ways, there will be dis-
agreement over proposals for policy implementation, leading to long battles and significant
delays in attaining outcomes.

Summary
Table 7.1 summarizes the three types of implementation models just introduced. It alludes
Copyright © 2020. Macmillan Education UK. All rights reserved.

to their main criteria for assessing the outcomes of the implementation process, as well as
pointing out advantages and disadvantages when using these models for examining policy
implementation.

Key points
•• There are three types of implementation models: top-down, bottom-up and hybrid.
•• Top-down models primarily emphasize the ability of policy-makers to produce specific policy
outcomes; most are characterized by an ‘evaluative’ strategy for analysis.
•• Bottom-up models stress the characteristics of the actors implementing policy outputs; (local)
implementers make ongoing choices themselves about the appropriate courses of action in spe-
cific contexts.
•• Hybrid models integrate elements of both top-down and bottom-up models and other theoreti-
cal models.

Table 7.1 Overview of implementation models.

Implementation type Assessment criterion Advantages Disadvantages

Top-down Degree of goal Allows for formulation of Disregards actions
attainment to assess clear-cut expectations taken during policy
implementation about the outcome of the drafting; ignores
success implementation process; strategic bureaucratic
can be expanded to action; excludes local
include additional aspects implementing actors
such as policy feedback
Bottom-up Degree to which Dual focus on formulation Complicates the analysis
policy outcomes are and implementation; the of implementation due
influenced by knowledge and to flexibility in the policy
learning, capacity- preferences of the actors outcomes desired;
building and who actually deliver public normative argument
support-building policy are modelled; more about autonomy of
realistic view on local-level implementers;
implementation overestimation of the
autonomy of local-level
implementers
Hybrid Ambiguity of policy Allows for a more realistic The different models do
objectives is and yet analytically not treat the
recognized parsimonious modelling of perspectives equally –
implementation; puts sometimes the
forward analytical top-down elements
categories to explain dominate and
differing policy outcomes sometimes the
bottom-up elements do

Source: Own elaboration.

IMPLEMENTATION SUCCESS: CRITERIA AND DETERMINANTS

In this section we first outline how implementation effectiveness can be empirically assessed.
The criteria we suggest are intended to provide a better understanding of the different activ-
ities related to policy implementation. We then provide some explanations for possible
variations in the effectiveness of implementation.

Criteria for Implementation Success

The assessment of implementation is not only affected by the analytical perspective adopted,
but also by the criteria that are applied. Based on the logic of top-down models, we suggest
distinguishing between formal transposition and practical application (see Knill and
Lenschow 1998; Weale et al. 2000; Winter 2003; Knill 2006; Hartlapp and Falkner 2009;
Robichau and Lynn 2009). While there are some valid points of criticism concerning the
top-down logic that we have already outlined, it is clearly more suitable for providing a mea-
surement of implementation than the bottom-up models.
Formal transposition focuses on the entirety of the specific provisions of a given public
policy and their incorporation into the existing legal and administrative system. In so doing,

bureaucrats in ministries or (quasi-)autonomous agencies have to ensure that this is done in a

complete manner and within the time frame specified by the legal act by which a policy is
adopted. Put differently, this stage is about the steps necessary to make a policy ‘imple-
mentable’ so that it can actually be put into practice. In some polities policy-makers adopt
detailed and specific laws, whereas in others they adopt general and vague laws that leave
bureaucrats with considerable discretion to fill in the policy details (Huber and Shipan 2002).
Although formal transposition is mostly unproblematic, it can sometimes still be a
demanding task. This holds particularly true for federal states when the implementa-
tion of a (new) federal policy requires that the constituent units make adjustments to
their respective policy arrangements (see Haider-Markel 1998; Huque and Watton
2010). Formal transposition can also be difficult when it is related to international
agreements, such as the UN Convention on Climate Change, or supranational law as it
is produced by the EU. Though we will discuss the EU’s role in policy-making beyond
national borders in detail in Chapter 10, here we want to point out that there is a size-
able literature acknowledging that the transposition of EU policies by member states is
marked by persistent deficits (see Knill and Lenschow 1998; Knill 2001, 2006; Falkner
et al. 2005; Steunenberg and Toshkov 2009; Thomson 2009).
Practical application corresponds to the actual putting into practice of a policy. Depending on
the content of a policy, practical application comprises different activities. It can refer to the
provision of services defined by legislation, such as welfare benefits which are given to those
entitled to them. Likewise, the actual provision of education, health care, infrastructure or equal
opportunities for job applicants is the result of the practical application of corresponding public
policies. However, this dimension is not the only aspect of practical application.
In most – if not all – cases, the implementation of a policy entails that the policy addressees
or target groups have to modify their behaviour in ways that are fully in line with the obliga-
tions stemming from the public policy. An ideal outcome of practical application would be one
in which organizations and individuals that are targeted by legislation voluntarily demonstrate
the behaviour that conforms to the legislation in question; i.e. they comply with it. However, as
non-compliance might occur, actions are required to induce compliance with a policy.
There exists a wide range of non-coercive measures to achieve compliance. Governments
can, for instance, make appeals to target groups to comply by underlining the respective
Copyright © 2020. Macmillan Education UK. All rights reserved.

policy’s benefits. Likewise, implementers can provide information about a policy. More gen-
erally, it is the perception of the legitimacy of the policy by those to whom it is addressed
that is crucial for compliance. As a rule, compliance is likely to be higher when non-
compliance is widely perceived as socially unacceptable.
In most cases, however, and especially when command-and-control policies are employed,
practical application involves monitoring and enforcement activities. Monitoring is about
surveillance, increased transparency and gathering information on how well the target
group complies with the requirements of a given public policy. This can be achieved in many
different ways, for example by requesting compliance reports or carrying out announced or
unannounced on-site inspections. Widespread monitoring activities include, for instance,
speed controls on motorways or controls on drivers’ blood alcohol concentration. If such
activities reveal non-compliance, the next step is enforcement to ensure that the non-
compliance stops and remedial action is taken. Enforcement powers available to implement-
ers include prohibition notices, suspension of operational licences, injunctions and the
carrying out of remedial works. For example, a driver exceeding road speed limits or being
found to be driving while drunk may be punished by means of a fine or the temporary or
even permanent loss of his or her driver’s licence (see Table 7.2).

Table 7.2 Criteria for measuring implementation success.

Focus Criteria
Formal Legal and administrative provisions for the • Time frame
transposition transposition of requirements into the existing • Completeness
legal and administrative system • Correct integration
into the regulatory
context
Practical Organizational and administrative structures and • Provision of policy-
application procedures related services
• Provision of non-
coercive incentives for
compliance
• Monitoring and
enforcement

For effectively assessing implementation success, either of these two dimensions can be
employed. Concerning the analytical insights, focusing on formal transposition is particu-
larly promising for policies that are intersectoral and require collaborative efforts by many
different organizations (see, e.g., Tosun and Leininger 2017; Trein and Tosun 2019), and
might yield interesting insights in federal states. With regard to practical application, the
study of this dimension certainly offers a better understanding of the more substantive
aspects of public policy that go beyond what is written in the law, though it is more complex
to assess due to the various types of activities related to it.

Determinants of Implementation Success

We now turn to the most important factors that explain variation in implementation effective-
ness, beginning with those factors that refer to the characteristics of policy outputs and then
moving onto institutional factors and the strategies for increasing their social acceptance.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Choice of Policy Instruments

An important theoretical debate centres on the question of whether the choice of policy
instruments makes a difference regarding effective implementation (Mayntz 1979; Bressers
and Klok 1988; Linder and Peters 1989; Howlett 1991; Knill and Lenschow 2000; May
2003). To illustrate this argument, we present the model of instrument preferences advanced
by Howlett et al. (2009: 173–5). The assumption underlying the model is that certain policy
instruments are better equipped than others to bring about the intended policy outcomes,
because they are easier to implement. According to Howlett et al., the appropriateness of the
choice of policy instruments depends on two dimensions: the complexity of the policy envi-
ronment, also known as the ‘policy subsystem’; and the capacity of the state to effect changes
in the light of institutional constraints. Governments with a high capacity for facing com-
plex policy environments are able to use directive instruments, including measures such as
a government reorganization in order to create or modify policy subsystems. The next con-
stellation refers to high-capacity governments faced with simple policy environments; these
can achieve effective implementation by using authoritative instruments, involving the

Table 7.3 Instrument choices in the light of capacity and subsystem characteristics.

Complexity of policy environment

High Low
Capacity of the state High Directive instruments Authoritative instruments
Low Subsidy instruments Informative instruments

Source: Adapted from Howlett et al. (2009: 175).

creation of specialized independent regulatory commissions or advisory committees, which

help to cultivate ideas that favour compliance with public policy. In contrast, governments
possessing a low capacity to cope with complex policy environments can rely on subsidy
instruments such as grants or specific funding schemes to induce policy addressees to com-
ply with the provisions of a policy. However, when low-capacity governments are confronted
with simple policy environments, they can use information instruments, such as campaigns
or information disclosure, to induce compliance (see Table 7.3).
The logical consequence of this model is that implementation problems stem from the
‘wrong’ policy instruments being chosen by policy-makers who are restricted by the ability
of governments and the characteristics of the policy environment. From this it follows that
a government that fails to evaluate correctly its capacity or the characteristics of the policy
subsystem is likely to experience problems in achieving compliance.
In this regard, a special case is the selection of policy instruments under uncertainty or
ambiguity. There exist policy problems that are not sufficiently well understood to immedi-
ately allow for an optimal design. A case in point is the micro-pollution of surface waters,
which is a highly complex phenomenon that requires a better understanding of the various
cause–effect chains. In such cases, guiding principles such as the precautionary principle pro-
vide guidance for policy instrument selection (Tosun 2012, 2013b; Metz and Ingold 2017).

Precision and Clarity of Policy Design

The second policy-related explanation for implementation problems concerns deficient policy
Copyright © 2020. Macmillan Education UK. All rights reserved.

design – a literature that has recently begun to experience a revival (see, e.g., Howlett et al.
2015; Colebatch 2018; Howlett and Mukherjee 2018; Tosun and Treib 2018). Such problems
of policy design can first of all be the result of vague and ambiguous policy objectives and
requirements. For example, the US State of Montana enacted a law to legalize medical mari-
juana in 2005, but it is defined so vaguely that it could not be implemented (see Pacula et al.
2015). In 2012, the law was changed to fix the flaws so that it can now be implemented, albeit
making legal access to medical marijuana to relieve pain more difficult (Pacula et al. 2014).
Often, it is only on the basis of such imprecise formulations that the adoption of a policy
is politically feasible at all. High degrees of distributional conflict and politicization favour a
negotiation context that is dominated by bargaining rather than problem-solving (see, e.g.,
Scharpf 1997b); i.e. the involved actors are primarily more concerned with potential losses
and gains than with analysing more thoroughly the extent to which there actually exists a
sound causal relationship between policy objectives and the suggested policy instruments.
As a result, distributional conflicts might lead to the formulation of ill-designed policies that
are characterized by inaccurate assumptions about the causal relationship between policy
problems and politically adopted remedies. This relationship between politics and the

implementation process has already been acknowledged by Bardach (1977), who contended
that conflicts that are not sufficiently resolved during the formulation stage bear a high risk
of popping up again during the implementation process, where they possess far-reaching
deficits (see also Jordan 1997; Winter 2003; Knill 2006).
The likelihood of deficient policy designs is not only affected by the degree of distribu-
tional conflicts characterizing a certain policy area. It also varies with the number of actors
that are involved in the decision-making process. As was illustrated in Chapter 6, the greater
the number of actors with veto power and the more complex the overall decision-making
structure, the higher the probability of compromises based on open formulations and incon-
sistencies in policy design.
A further source of design problems emerges from the fact that policy spaces are rarely
empty. This means that newly designed policies typically are added to an already existing
stock of multiple policy targets and instruments in a given policy area. This pattern of policy
accumulation (Adam et al. 2019) entails that the design implications of new policies can
hardly be fully understood without considering their potential interactions with the effects
of existing policy elements. With the increasing size of the policy stock, the design of new
policies hence turns into an increasingly complex activity.

Control Structures

Principal–agent theories constitute an important starting point for the explanation of

implementation deficits. In these theories, it is assumed that implementation problems
result from the differences between policy objectives and their actual implementation
through the responsible administrative agencies. This difference is seen as an unavoidable
consequence emerging from the configuration of modern political-administrative systems
that are characterized by the delegation of competencies to subordinate administrative
authorities. This delegation is of particular relevance with regard to the distinction between
the tasks of policy formulation (usually taking place within central ministries) and the
implementation of these policies (often delegated to agencies at the subnational or local
level; Howlett et al. 2009: 160–3).
Delegation entails the problem of bureaucratic drift, one which is inherent to the configu-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ration of political-administrative systems and which is further aggravated by two factors.

On the one hand, high organizational complexity might increase the number of agents and
government levels that are involved in the implementation process and hence increase the
potential for bureaucratic drift. Schnapp (2000) shows that bureaucratic drift is likely in
countries with coalition governments that have a high number of coalition parties, namely
Finland, Switzerland, Belgium and Japan. By contrast, it seems less likely in Spain, Canada,
New Zealand, Greece and the United Kingdom, since their governments are normally com-
posed of one party only. The occurrence of bureaucratic drifts during implementation in
governments consisting of several parties is hypothesized to be more likely because one of
the coalition partners that could not realize its preferences during policy formulation might
be willing to accept policy change if it knows it can attempt to modify policy outcomes dur-
ing the implementation phase (Hammond and Knott 1996). On the other hand, a high
degree of scientific or technological complexity of the underlying policy problem will
increase the chances of different interpretations of policy objectives by principals and
agents. The more specialized the knowledge needed to implement a public policy, the more
likely the implementer is to possess an information advantage vis-à-vis policy-makers,
which facilitates deviations from the original policy guidelines.

There are two ways in which politicians can control the bureaucracy and the way it imple-
ments public policy. The first form is formal oversight, in which the parliament directly
monitors agency behaviour to gain the information it needs to correct undesirable behav-
iour (see Aberbach 1990). This principally takes place by means of committee hearings and
investigations (see Meier 2000). In the United States and United Kingdom, for instance,
members of Congress or parliament are frequently involved in ‘casework’, which concerns
the handling of problems that occurred to citizens during the stage of policy implementa-
tion. Such cases typically involve delays in the provision of different kinds of social benefits
or an unsatisfactory provision of certain services and/or infrastructure. Members of
Congress and parliament devote their resources to this casework since it is in this way that
they can demonstrate their responsiveness to the needs and problems of their electorate,
which is expected to increase their chances of re-election (Anderson 2010: 219).
The second form is statutory control, in which the executive or the legislature designs the
agency’s structure and processes to favour some policies over others. While oversight occurs
after the bureaucratic actors have implemented a policy, statutory controls are established
before they act (Bawn 1997: 102). There are two forms of statutory controls: those designed
around ‘fire alarms’ and those centring on ‘stacked decks’. ‘Fire alarms’ are defined as a system
in which the parliament establishes rules and informal practices that enable individual citizens
and interest groups to examine administrative decisions and to ‘raise the alarm’ should they
disagree strongly with specific decisions (McCubbins and Schwartz 1984: 427). Alternatively,
legislation that delegates policy decisions to an agency may specify in great detail how the
agency decisions are to be made. McCubbins et al. (1987) argued that these issues of agency
structure and process can be designed strategically by legislators to ‘stack the deck’ in favour of
groups that the legislators want to help. Additionally, courts can play an important role in
ensuring that administrative agencies do not exploit their implementation powers.
So far, we have only concentrated on ways of controlling administrative agencies.
However, Newton and van Deth (2010: 124–5) point out additional possibilities for reduc-
ing the risk of bureaucratic drift. First, politicians can appoint bureaucrats on the grounds
of political considerations such as their ideological proximity to the party or parties to which
the government and/or the parliamentary majority belong. Second, the potential of bureau-
cratic drift could be reduced by training bureaucrats in a manner that develops a profes-
Copyright © 2020. Macmillan Education UK. All rights reserved.

sional ethos of public service. Third, financial controls could be used as one form of oversight.
Fourth, an increasing share of open government might help to reduce delegation problems.
The authors suggest the employment of ‘sunshine laws’; i.e. a type of law that requires
administrative agencies to do their work in public through open meetings. Finally, malad-
ministration might be reduced through the presence of ombudsmen.
It should be noted that most of the points mentioned above include assumptions about the
behaviour of bureaucrats that correspond more to Niskanen than to Weber. Following Weber,
bureaucrats should be guided by their professional ethos and thus behave in ways that serve
their political masters. In contrast, the view promoted by Niskanen stresses the self-interest
of bureaucrats and their objective of maximizing it. However, the second suggestion by
Newton and van Deth essentially approximates to the Weberian ideal of a bureaucrat.

Institutional Design

Except for very rare cases where policies to some extent are ‘self-implementing’ – i.e. the
declaration of policy requires no further actions to bring about the desired policy outcomes –
the implementation of policies generally requires institutional structures and arrangements

(O’Toole 2003: 234). In other words, policies generally have institutional implications, i.e.
requirements for the establishment of appropriate structures and procedures for their
proper implementation. A distinction can be made between policies that can be imple-
mented by single organizations or authorities (Torenvlied 1996) and measures whose proper
implementation entails horizontal and vertical coordination across several administrative
units and levels (Hjern and Porter 1981). It is obvious that in the latter case much greater
challenges to effective implementation exist than in the case of an integrated implementa-
tion structure: ‘between or among organizations, the differing routines and specialized lan-
guages, not to mention distinct ways of seeing the world, mean that interorganizational
implementation poses particularly daunting challenges’ (O’Toole 2003: 235).
These challenges are based on the fact that the implementation of a policy in this way requires
major changes in existing institutional structures. This aspect is of empirical importance with
regard to implementation in federal polities. For example, Canada possesses one of the most
decentralized frameworks for environmental policy implementation in the world. The prov-
inces have supremacy over most environmental matters and are relatively free to set their own
standards and carry out implementation activities (Huque and Watton 2010: 77–8). As a con-
sequence of the well-developed competencies of the provinces and the difficulties in achieving
institutional cooperation across the different levels of government, the federal government has
faced problems in the nationwide implementation of environmental policy. Therefore, policy
implementation is mainly carried out by the provinces, with federal policy only defining ‘soft’
environmental measures that can be implemented without coordination efforts.
The central argument here is that it is less the choice of the instrument per se that affects
the implementation success of policies, but rather the extent of necessary institutional
modifications that arise from public policy (see Knill and Lenschow 1998; Knill 2001).
Although policies are generally directed at the specification of policy contents and
instruments rather than institutional arrangements, there is often a tight linkage between
policy content and the corresponding requirements of institutional implementation.
Therefore, decisions on policy design to a certain extent always entail decisions on the
corresponding institutional arrangements for their proper application. Consequently,
implementation problems can be conceived of as problems of institutional change (Knill
and Lenschow 1998; see also Tosun and Treib 2018).
Copyright © 2020. Macmillan Education UK. All rights reserved.

The role of institutions has long been acknowledged in the literature on implementation.
However, institutions were initially analysed mainly from the perspective of adequate insti-
tutional design. Analysts coming from the top-down perspective developed optimal struc-
tural and organizational arrangements that would permit effective implementation of a
certain policy (see Pressmann and Wildavsky 1973). This reasoning relies on the implicit
assumption that existing institutions would easily adapt to the suggested ‘model’ structure.
Problems of institutional change were ignored. The bottom-up perspective assumes a simi-
lar malleability of existing institutional factors. Here, analysts are interested in the impact
of varying institutional designs on the skills, resources and capacities of relevant actors.
They are interested in the perfect design that serves to equip the implementing authorities
with sufficient financial, legal and personal resources.
Without denying the importance of adequate institutional design, such a perspective
remains incomplete as long as it ignores the problems associated with the process of adjust-
ing the existing institutional arrangements to the defined ‘ideal’ arrangements. It is one of
the few generally accepted findings in the literature on institutions that change rarely takes
place in a smooth and unproblematic way. Existing institutions ‘matter’, and they do so
mainly by constraining the options for future change and adaptations.

An emphasis on institutional stability and continuity is, however, not synonymous with
an entirely static understanding of institutional development. Rather, institutions often
find themselves in a permanent process of adaptation to their environment. However, the
scope of these adaptations is restricted by the structuring effects of existing institutional
arrangements. Institutional change is hence often limited to aspects that do not question
the ‘identity’ of an institution (March and Olsen 1989; Thelen and Steinmo 1992). This
abstract argument is of limited explanatory value so long as we do not have any criteria to
judge which particular institutional requirements stemming from public policy are likely to
exceed the adaptation capacity of existing institutions and which are not.
Knill and Lenschow (1998) suggest a distinction between three levels of adaptation pres-
sure, each of them linked to different expectations with regard to implementation effective-
ness. The first scenario refers to constellations where there is low pressure for institutional
adaptation. In this case, the institutional implications of new policies are completely in line
with existing arrangements; i.e. no or only marginal changes are demanded. Implementation is
therefore expected to be relatively unproblematic, as adjustment requirements are very limited
or completely absent. In the second scenario of high adaptation pressure, new requirements
exceed the adjustment capacities of existing institutions. Ineffective implementation is the
probable consequence. Such constellations can be expected, for instance, when new require-
ments contradict the strongly entrenched institutional elements of policy arrangements.
In this regard, Krasner (1988) differentiates between two dimensions of institutional
anchorage: ‘institutional depth’, which refers to the extent to which institutional arrange-
ments are embedded in normative orientations and dominant belief systems; and ‘institu-
tional breadth’, which refers to the extent to which institutional arrangements are
interlinked with their environment. The third scenario of moderate adaptation pressure
refers to constellations in which new policies require substantive adjustments of existing
institutions, though without challenging well-entrenched core patterns within the political,
legal and administrative system. While in such cases there is a higher probability of an effec-
tive implementation of policies, it cannot be taken for granted. In contrast to the two other
scenarios, an institutional perspective is insufficient to develop hypotheses on expected
implementation performance (in terms of institutional adjustments to new requirements).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Institutional adaptation Implementation

pressure

Low EFFECTIVE

‘Favourable’ context EFFECTIVE

Moderate

‘Unfavourable’ context INEFFECTIVE

High INEFFECTIVE

Figure 7.1 Institutional adaptation pressure and implementation effectiveness

Source: Knill and Lehmkuhl (2002a).

Box 7.2 Implementation of Mexican environmental policy

Mexico has emulated various of the United States’ environmental protection stan-
dards. This transfer of policies was primarily motivated by the country’s wish to
become a member of the North American Free Trade Agreement (NAFTA). To avoid
criticism from the US public concerning environmental issues, the Mexican govern-
ment displayed notable efforts to bring its legislation in line with that of its northern
neighbour. Despite legal reforms, environmental degradation is still occurring in
Mexico due to insufficient policy implementation. Of course, there are many reasons
why the authorities face such implementation problems. One of the main reasons is
that the transfer of US environmental protection standards turned out to be the
wrong choice, as they were technically too demanding for the organizational arrange-
ments in place in the country.
This point is best illustrated by Mexican wastewater standards, which were origi-
nally sector specific, as is the case in the United States. However, since it became clear
that the Mexican authorities lack the necessary institutional infrastructure for moni-
toring such a sophisticated regulatory approach, wastewater standards have been
modified in a manner to make them ‘implementable’ (Knill et al. 2008a). One of the
reasons for the failure to implement sophisticated regulatory standards can be seen in
the weak financing of water and wastewater infrastructure, which made it difficult to
maintain drinking water networks and wastewater treatment plants (Pacheco-Vega
2018). This entailed the abolition of the ambitious initial standards and the establish-
ment of a new regulatory system that does not differentiate between industry sectors.
The modification of the regulatory approach reduced the extent of implementation
failure, but the overall state of compliance is still far from perfect (Pacheco-Vega 2015).

We have to complement our analysis with a second explanatory step which considers the
particular interest constellation and institutional opportunity structures. To what extent is
there sufficient public support for adjusting to new requirements? To what extent have
actors who support regulatory change sufficient powers and resources to realize their inter-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ests? As Figure 7.1 shows, institutional adaptation and hence effective implementation can
only be expected if they are facilitated by favourable contexts (Knill and Lenschow 1998;
Knill and Lehmkuhl 2002a; Box 7.2).

Administrative Capacities

While the factors discussed so far are focused on the willingness of actors to achieve effective
implementation, the focus on administrative capacities entails a different perspective that is
concerned with the ability rather than willingness to comply with given policy requirements
(see, e.g., Mukherji 2017). In other words: varying implementation effectiveness can also be
explained by different capacities, which affect the opportunities available for effective formal
and practical implementation. As already mentioned, implementation is generally carried out
by a designated government agency that has responsibility for the new policy measure.
Theoretically, the responsible agency should be equipped with the necessary resources to ensure
that the policy is carried out as intended, though in reality this does not always occur. For suc-
cessful implementation, the entity concerned must possess sufficient resources to be able to
translate the policy objectives into an operational framework. Particular emphasis is placed on

human capacity (administrative and technical expertise) as well as financial, technical and orga-
nizational resources (see Gerston 2004: 103). The less developed these capacities, the more
important the allocation of existing resources in the light of political priorities becomes.
As shown by Adam et al. (2019), the presence of sufficient administrative capacities for
proper implementation can hardly be taken for granted in modern democracies since political
legitimacy often rests more on the delivery of new policy outputs than on their implementa-
tion. A continuously growing stock of policies directly translates into the accumulation of
administrative burdens when it comes to the practical application of public policies. Thus, to
cope with increasing burden loads, we should have observed considerable expansions in
administrative capacities, especially in terms of personnel and budgetary resources. Yet gov-
ernments face fundamental fiscal and ideological constraints for public-sector expansion in
times of globalized financial markets, austerity and ideas of New Public Management.
Statistical data on public-sector employment does indeed reveal a trend of stagnation and
partial decline for members of the Organisation for Economic Co-operation and Development
(OECD), despite considerable variation across countries (ILO 2017). Even if we assume that
the efficiency of administrative processes has somewhat increased over time, there is the risk
of an increasing gap between accumulating policies and stagnating or even declining imple-
mentation capacities. Once individual laws and regulations are adopted, they move off the
desks of policy-makers and onto the desks of lower-level frontline bureaucrats, where imple-
mentation burdens accumulate, very often without adequate financial and staff resources to
handle the additional workload and complexity. As implementation burdens continue to accu-
mulate, the prevalence of administrative backlog and selective implementation increases
(Adam et al. 2017; Adam and Knill 2018; Knill et al. 2018).

Social Acceptance

As already outlined, the main purpose of policy implementation is to modify the behaviour of
policy addressees. While there are many policy-specific and institutional factors that might
hamper this goal, it is ultimately the social acceptance of public policy that matters for success-
ful implementation. There are certainly some policy measures that are welcomed by policy
addressees, such as increases in social benefit levels. By the same token, there is legislation that
Copyright © 2020. Macmillan Education UK. All rights reserved.

is less well received, either because it imposes costs on the target group or because its implica-
tions are not well understood (see, e.g., Tosun 2018). In such cases, implementers might
approach interest groups to seek their support actively for the policy measure in question
(Anderson 2010: 227). Interest groups can communicate the exact content of the law to their
members and convince them of the advantages it entails. From this perspective, interest groups
can faciliate the implementation of public policies by increasing their social acceptance.

Key points
•• When measuring implementation effectiveness, it is useful to distinguish between formal
transposition and practical application.
•• Formal transposition is about taking the necessary steps with regard to the accommodation of
a policy into the legal and administrative system to make it ‘implementable’.
•• Practical application refers to what is usually conceived of as policy implementation. It involves the
provision of services on the one hand and monitoring and enforcement activities on the other.
•• There are six principal factors affecting implementation effectiveness: choice of policy instruments,
policy design, control structure, institutional design, administrative capacity and social acceptance.

CONCLUSIONS

Once the government has taken a decision on a public policy, the stipulations of that policy
must be put into action to bring about the behavioural changes intended by the policy-
makers. Therefore, unless the stipulations of a given policy are actually implemented, the
problem originally initiating the policy process will persist. At first glance, implementation
appears as an automatic continuation of the policy-making process. Yet there often exists a
substantial gap between the passage of new legislation and its application, which is addressed
by implementation research. The bureaucracy is delegated a significant degree of power dur-
ing the implementation stage of the policy process because of its discretion in interpreting
the actual intent, method and scope of a policy decision. During implementation, adminis-
trative agencies by no means always follow political guidelines unrestrictedly, and even if
they do so, in some cases results deviate remarkably from political expectations. The role of
bureaucracy during implementation reveals a contradictory picture of great interest. On the
one hand, bureaucracies are essential for making policies work; but, on the other hand,
senior bureaucrats are often more experienced and better trained than their political mas-
ters, which paves the way for bureaucratic drift.
Previous considerations have made clear that both the formal transposition and practical
application of a policy in general are based on complex processes. These can be analysed
from different perspectives (top down versus bottom up) and by shedding light on various
factors that are expected to affect implementation effectiveness. The implementation stage
of the policy process is by definition an operational phase, where policy is actually translated
into action with the desire to solve some public problem. We have identified six key chal-
lenges which we believe can routinely impede the effective implementation of public policy.
In this context, the question of whether and under what conditions the implementation of
policies can be characterized as effective or ineffective cannot be answered in terms of a
single cause, but needs to take into account different aspects.

WEB LINKS
Copyright © 2020. Macmillan Education UK. All rights reserved.

www.apsc.gov.au/publications09/devolvedgovernment.htm. This is a link to an elec-

tronic book that gives an overview of challenges regarding policy implementation
through the federal government in Australia.
www.epa.gov/air-sensor-toolbox. The US EPA encourages citizens to participate in
the monitoring of air quality. This website explains how it works and how citizens
can support the EPA’s work with their measuring activities.
www.healthpolicyinitiative.com/policyimplementation. This is a useful tool for train-
ing people in how to interview policy-makers about influences on health policy
implementation.
www.inece.org/. This is the website of the International Network for Environmental
Compliance and Enforcement. It is a valuable source of information about the
implementation of environmental policy.

www.instituteforgovernment.org.uk. The Institute for Government regularly pub-

lishes global trends in policy delivery units and is a good source of data on all
institutional aspects related to policy implementation.
www.nibusinessinfo.co.uk/content/what-expect-food-safety-inspection. This is a prac-
tical guide to how a food safety inspection works in the United Kingdom.

FURTHER READING

Bardach, E. (1977). The Implementation Game: What Happens after a Bill Becomes a
Law. Boston, MA: MIT Press. A classic book about the complex process of policy
implementation.
Hill, M. and Hupe, P. (2009). Implementing Public Policy: An Introduction to the Study
of Operational Governance. Thousand Oaks, CA: Sage. This book is a very compre-
hensive treatise on all aspects related to policy implementation.
Howlett, M. and Mukherjee, I. (eds.) (2018). Routledge Handbook of Policy Design.
New York: Routledge. This is an up-to-date and comprehensive overview of the
study of policy design.
Lipsky, M. (1980). Street-Level Bureaucracy. New York: Russell Sage Foundation. This
is another classic book that we recommend for further reading.
Pressman, J. and Wildavsky, A. (1973). Implementation. Berkeley, CA: University of
California Press. An influential book with respect to policy implementation that
is helpful in understanding how implementation actually works.
Scala, F. (2019). Delivering Policy - The Contested Politics of Assisted Reproductive
Technologies in Canada. Vancouver: UBC Press. An instructive book that engages
with the politics perspective on policy implementation.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Reader’s guide
Policy-making does not end with the passage and implementation of legislation.
Several questions emerge afterwards. Has the policy attained its objectives? What
are its unintended effects? Is a failure to meet the policy goals related to the design
of the public policy or its implementation? Policy evaluation tackles these and
related questions about expected and unexpected policy outcomes and impacts. By
definition, evaluation studies make judgements about the quality of public policies,
which implies that negative findings can, in principle, re-initiate the policy-making
process with the objective of improving existing policy arrangements. While this
definition might give the impression that evaluation studies are only carried out by
experts who possess the required knowledge and techniques for making such judge-
ments, many actors are in fact involved in the process. The large number of potential
stakeholders at this policy stage results from the fact that there is a ‘political’ compo-
nent to policy evaluation, meaning that statements about the success and failure of
a given public policy are likely to be used for generating positive or negative images
of those in power. Therefore, to understand policy-making fully, the evaluation stage
cannot be left out. To illustrate the central topics, we will first give an overview of
the different types and methods of policy evaluation, before moving on to provid-
ing ideas about research design for evaluating policies. This is followed by a discus-
sion of the political characteristics of policy evaluation and the role of ‘evidence’ in
policy-making.
Copyright © 2020. Macmillan Education UK. All rights reserved.

INTRODUCTION

When is a policy measure successful? Policy evaluation provides answers to this and addi-
tional questions. It is essentially about generating information in order to compare the
intended and actual effects of public policies and can refer to insights regarding policy out-
puts, outcomes and/or impacts. The information generated during the evaluation process
can then be used for many different purposes, such as improving public policy, supporting
the views of proponents or critics, or responding to political pressure. While evaluation is
mostly about analysing policy decisions, including the consideration of (scientific) evidence
(see Cairney 2016), it is also an extension of the political debate to determine policy success
and policy failure (Bovens et al. 2001a, 2001b; McConnell 2011, 2015; Bovens and ‘t Hart
2016; Hinterleitner 2018; FitzGerald et al. 2019). It is possible or even desirable that a new
policy cycle will begin if the overall verdict is that the evaluated public policy does not meet
its objectives (see policy change as discussed in Chapter 11). While evaluation is a useful and

144
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:19:00.
Evaluation 145

even necessary device of policy-making in modern states, it is seldom a straightforward

task. Some of the factors that can complicate evaluation activities include:

• The identification of policy goals

• The appropriate definition and measurement of performance indicators
• The isolation of a policy’s effects from other factors
• The political context

The first three challenges can be resolved by a carefully developed research design, which
concerns making decisions regarding the structure and strategy of investigation, and which
we will consider later in this chapter. In contrast, the fourth issue, about the political con-
text in which evaluation takes place, cannot be fully controlled. Although evaluation research
attempts to assess a policy in the most objective manner, these activities occur in a political
environment, meaning that policy-makers might be interested in emphasizing those find-
ings that help to portray them in a positive light.
Besides the potential threat of instrumentalization, there are also practical con-
straints stemming from the political context. As such, policy-makers often want imme-
diate information on policy effects, though many public policies have long-term effects
that will not be known in the short term, forcing researchers to project effects rather
than actually measuring them. Accordingly, ‘all public policy evaluations … are projects
in both political science and political science’ (King et al. 2007: 480). In this chapter, we
provide an introduction that highlights the scientific and political characteristics of
policy evaluation.

TYPES AND METHODS OF EVALUATION

As already hinted, policy evaluation can be conducted in many ways and for different pur-
poses. We will now outline the main types of evaluation and provide some background
knowledge regarding the methodologies that can be used to assess the effects of public
policies.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Types of Evaluation
Actors involved in evaluation are diverse, ranging from researchers, consulting firms,
think tanks and non-governmental organizations (NGOs) to courts, political bodies
(such as parliamentary commissions) and the executive. The diversity of actors that
might participate in or conduct policy evaluation reflects the different types of evalua-
tion, which comprise:

• Administrative evaluation
• Judicial evaluation
• Political evaluation
• Scientific evaluation
• Economic evaluation
• Performance evaluation

A dministrative Evaluation

Administrative evaluation is usually carried out within government bodies and examines the
delivery of public policy and government services (Hessing et al. 2005: 249). It is performed by
specialist agencies; financial, legal and political overseers; or private consultants. The main objec-
tive of administrative evaluation is to ensure that public policies attain their goals at the least
possible cost and least burden on the target groups (Howlett et al. 2009: 185). There is a direct
connection between administrative evaluation and the introduction of New Public Management
(see Dahler-Larsen 2005, 2012). In essence, New Public Management is about opening up the
public sector to private-sector management principles (see Christensen and Lægreid 2016),
which has exerted considerable influence on the practice of public administration and its effec-
tiveness and efficiency in providing public services (Thomas 2017: 8; see also Box 7.1).
There exist many different techniques for administrative evaluation, including process, effort,
efficiency and effectiveness evaluation (Howlett et al. 2009: 186). Process evaluation is about
exploring possibilities for making operating procedures more efficient. Effort evaluation assesses
the amount of effort governments put into attaining their policy objectives (e.g. in terms of
budgets or personnel resources). Efficiency evaluation is about a public policy’s costs and the
ways of accomplishing the same goals at lower costs. Effectiveness evaluation compares the
intended goals of a policy with the ones actually achieved. These four forms have recently been
complemented by performance indicators and benchmarks that are designed to allow public-
sector efforts to be compared. In a comparative study, Pollitt (2006) showed that British admin-
istrations could push such performance indicators faster and further than in other European
countries, due to the characteristics of their political system and their more individualistic and
risk-accepting administrative culture.
While administrative evaluation is predominantly concerned with the ex post (after the
event) assessment of the delivery of government services, there also exist ex ante (before the
event) techniques, known as policy appraisal, that seek to inform decision-makers by predict-
ing the effects of policy proposals (see Turnpenny et al. 2008, 2009; Jordan et al. 2015b). The
most widely used form of policy appraisal is regulatory impact analysis, which includes various
methods to assess ex ante the impact of proposed regulatory policies on target populations.
The aim of this administrative procedure is to increase the empirical basis of political deci-
Copyright © 2020. Macmillan Education UK. All rights reserved.

sions, to make the regulatory process more transparent and to increase accountability (Radaelli
2004: 723). Thus, regulatory impact analysis is directly related to the stage of problem defini-
tion, with the objective of enhancing the quality of regulation, i.e. of achieving ‘better regula-
tion’. The Organisation for Economic Co-operation and Development (OECD) and its 1997
guidelines on regulatory impact assessment played a central role in the international diffusion
of this evaluation technique (Radaelli 2009: 31–2), and today it is in place in most of its mem-
ber countries (Sager and Rissi 2011: 151) by means of policy diffusion (De Francesco 2012; see
Chapters 10 and 11).

Judicial Evaluation

Judicial evaluation is clearly different from administrative evaluation, as it is principally con-

cerned with legal issues relating to the way in which policies are developed and implemented.
Depending on the political system, judicial evaluation is either initiated by the judiciary or when
requested by individuals or organizations presenting a legal complaint against any given policy
measure. In the most extreme case, a judicial evaluation can declare a policy (or rather the legal

act on which it is based) unconstitutional and ask the legislature to develop modifications to it
(see Chapter 3). Thus, judicial evaluation represents one of the few means by which private actors
can challenge the activities of public actors. Moreover, public actors can utilize judicial evaluation
as a means for achieving compliance with public policy (Hessing et al. 2005: 197).
In the United States especially, the courts also evaluate whether a policy adopted by the govern-
ment is in compliance with the intent of the law in question. Thus, in some polities ‘judicial evalua-
tions may move from the determination of the scope of government authority to the determination
of whether such authority was used appropriately in a given situation’ (Adolino and Blake 2011:
26). In this context, Howlett et al. (2009: 189) point out that, in parliamentary systems, judicial
evaluation focuses on whether or not courts, tribunals or government agencies have acted within
their powers, indicating that the evaluation is mostly confined to procedural issues. In political
systems with a constitutionally entrenched division of powers, such as the United States, judicial
evaluation concentrates more directly on legislative and executive actions as such. In other words,
judicial evaluation is about substance, which provides an opportunity structure for pursuing policy
change. For example, in Peru, a semi-presidential system, instead of creating a political party, indig-
enous groups rely on litigation and judicial review to strengthen indigenous rights (Merino 2019).

Political Evaluation

Howlett et al. (2009: 189–91) identify political evaluation as a third category, in addition to admin-
istrative and judicial evaluation. This evaluation form is based on a rather unsystematic and techni-
cally not very sophisticated way of gathering and interpreting information about public policies.
Thus, strictly speaking, political evaluation is not an evaluation activity in the classic sense, but
rather a tool for depicting (framing) public policy in a positive or negative way (see Fischer 1995,
2003). The purpose of this kind of evaluation is strongly related to aspects of party competition.
Political parties which had supported the adoption of a certain public policy have a strong interest
in declaring it a success, while opposing parties will strive to draw a negative picture of the policy in
question. However, political evaluation is not confined to political parties. Many other actors par-
ticipating in policy-making, such as think tanks or interest groups, can use it for their own pur-
poses. Thus, there is a certain risk that evaluations are used as political instruments rather than for
measuring the effects of policy decisions. In this way, evaluations can form the basis for politicizing
Copyright © 2020. Macmillan Education UK. All rights reserved.

policies (see Wood 2016), which can result in either policy change or stability, depending on the
respective findings of the evaluation exercise and how they are presented.
Varone et al. (2018b) illustrate the empirical relevance of political evaluation. Concentrating
on the relationship between members of parliament and interest groups, the authors show that
affiliations of the former with the latter do indeed influence requests for policy evaluation. What
is particularly interesting about the empirical findings is that party membership does not matter
for this relationship. Regardless of party membership, members of parliament who are affiliated
with an interest group are more likely to request policy evaluations than parliamentarians with-
out such connections. This shows that elite networks do not only matter for policy formulation
and policy adoption (see, e.g., Fischer et al. 2019), but also for how the policies adopted are per-
ceived and whether the policy process starts anew in light of the evaluation findings obtained.

Scientific Evaluation

Policy evaluation can also constitute a social scientific activity which aims to provide neutral
answers to the question of whether a given public policy is effective or not. In this context,

we can differentiate between formative and summative evaluation (Scriven 1967). Formative
evaluation has the purpose of improving a certain policy measure by providing advice to
implementing actors and other stakeholders. This evaluation type is useful for fostering
internal development and improvement. Summative evaluation takes place at the end of
policy implementation and assesses whether the policy has attained the intended objec-
tives. It represents an assessment intended for policy-makers who are involved in oversee-
ing the policy and who may utilize the findings to make decisions about its continuation.
Scientific evaluation can only be carried out by experts who are trained in a specific way and
are familiar with the required research techniques.
A systematic scientific evaluation requires resources that are often not available to policy-
makers. Most importantly, it requires time and, especially in situations where the findings
are to be used for the formulation of public policy, is difficult to provide. Time can also be an
issue for summative evaluation, since the time frame of policy-makers is limited by election
periods and the fact that they want their policy successes to be backed by evidence before
elections. However, in order to make valid statements about policy outcomes and/or
impacts, an observation period of a certain length is needed. Unless this is granted, scien-
tific evaluation cannot produce an accurate assessment of policy effects.
More generally, scientific evaluation is part of evaluation research, which represents a
distinct discipline in social science. This body of research aims to discuss methodological
issues related to the assessment of policy success, the establishment of causality and tech-
niques of analysis. There has been a lively debate in evaluation research concerning how to
integrate theoretical considerations and improve theory construction. Although such con-
siderations about the practice of policy evaluation also form part of the scientific debate, it
is clear that evaluation research is only loosely related to the political process that produces
public policies, which distinguishes it from the previous three evaluation types. However,
we want to stress that there have been instances in which scientific evaluation research has
informed subsequent policy-making. The negative income tax experiments that were con-
ducted in the 1960s and 1970s in the United States, for instance, informed later welfare
policies by projecting how elastic labour supply would be for varying amounts of welfare
guarantees. Box 8.1 presents a very recent and ongoing scientific exercise.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Box 8.1 Basic income experiment in Finland

The conservative government of Prime Minister Juha Sipilä implemented a basic
income experiment in Finland in 2017–2018, in which 2,000 randomly selected
unemployed persons (aged 25–58) received a basic income of €560 per month
regardless of their other income and their effort to find a job. They served as the
treatment group, whereas the rest of the unemployed in Finland served as the con-
trol group, thus mimicking a laboratory experiment. The preliminary evaluation
results as reported by the project leader, Hiilamo (2019), show that the labour mar-
ket effects of basic income are very small. Basic income recipients did not work more
days or have higher incomes than the persons in the control group. Therefore, the
initial findings question whether basic income gives people a stronger incentive to
work, which makes it unlikely that the Finnish social security system will be modi-
fied to offer basic income.

Likewise, scientific evaluation is part of the literature on evidence-based policy-making (for

overviews, see Oliver et al. 2014; Wellstead et al. 2018), which we discuss in detail later in
this chapter. This literature concentrates on how policy-makers make use of evidence when
addressing policy problems. Cairney (2016) argues that policy-makers do not consider all
evidence supplied on policy problems, but rather use the information provided to define
their goals and priorities (the ‘rational’ shortcut), or they make decisions quickly without
paying attention to the evidence by following their beliefs and emotions (the ‘irrational’
shortcut). The main implication from this perspective, which brings the political character-
istic of scientific evaluation to the fore, is that scientists have to become policy entrepre-
neurs to increase the chances that the scientific evidence will be taken up by policy-makers
(see Cairney 2018). This conclusion is underlined by an empirical study by Newman et al.
(2017), which shows that policy-makers have difficulty in accessing scientific evidence.
Often policy-makers rely on ‘best practices’ rather than insights provided by scientific evalu-
ation, which is problematic since they feature place-bound characteristics and do not lend
themselves to simple policy emulation (Nagorny-Koring 2019). Therefore, if research evi-
dence is to guide policy-making, it will need to be presented in a concise, straightforward
and non-technical fashion.

Economic Evaluation

Economic evaluations involve the identification, measurement, evaluation and then com-
parison of the costs (inputs) and benefits (outcomes/impacts) of two or more alternative
policies. The costs and consequences of alternative interventions or scenarios are com-
pared to find the best use of scarce resources, thus focusing on the efficiency of a policy.
This approach is predominantly used for administrative evaluations and differs according
to scope and intent. It can have a very narrow focus, in which evaluators are only con-
cerned about the resource consequences for the implementing agency (for instance, in the
case of child benefits, for the ministry for family affairs or its equivalent). In these evalu-
ations, a new intervention which shifts costs to another agency may be preferred.
Alternatively, economic evaluations can examine wider social costs. In these evaluations,
a new intervention that shifts relative costs but does not reduce total costs may not be
Copyright © 2020. Macmillan Education UK. All rights reserved.

desirable.
A country that has adopted economic evaluation as a formal tool for guiding the design of
health policies is South Korea. In 2001, it used economic evaluations for decisions on pricing
and the extent of health insurance coverage of new medical technologies, encompassing
pharmaceuticals, equipment and diagnostic technology, that were necessary within the con-
text of amending the act on national health insurance (Yang et al. 2008: 183). The number
of Asian countries preparing for the adoption of economic evaluation for making health
policy is rising: China, Taiwan and Thailand have already elaborated their own guidelines in
this policy area (Yothasamut et al. 2009).
The three main types of economic evaluation are cost–benefit analysis, cost-effective-
ness analysis and cost–utility analysis. They all address efficiency issues with regard to
social interventions. Cost–benefit analysis involves weighing the total expected costs
against the total expected benefits of one or more actions in order to choose the best
option. To this end, benefits and costs are often expressed in monetary terms and are
adjusted for time value, so that all flows of benefits and costs over time are expressed on

a common basis in terms of their present value (see Shapiro 2011). Cost-effectiveness
analysis compares the relative costs and outcomes of two or more policies. It is distinct
from cost–benefit analysis, which assigns a monetary value to the measure of effect. Cost-
effectiveness analysis is often used in the field of health services, where it may be inap-
propriate to monetize health effects. Typically it is expressed in terms of a ratio, where the
denominator is a gain in health from a measure (e.g. years of life) and the numerator is the
cost associated with the health gain. Similarly, cost–utility analysis involves looking at
whether an action should be undertaken. Due to the strong overlap between the concepts
of cost-effectiveness and cost–utility analysis, the latter is usually referred to as a specific
case relating to health care and life expectancy.

Performance Evaluation

In contrast with the previous evaluation types, performance measurement can take into
account either effectiveness or efficiency considerations. Performance measurement dif-
fers from the previous forms of evaluation with respect to the use of the information
gathered, since implementing actors and programme managers are expected to be account-
able for achieving the intended goals. Moreover, the analytical focus is mostly at the man-
agerial level. Performance measurement is about the use of organizational resources
relative to a predefined goal, characterized by ongoing monitoring and reporting of policy
accomplishments (see Pollitt 2006). This definition clearly indicates that reporting is a
central activity to performance measurement. It is typically conducted by programme or
agency management. In this sense, it and programme evaluation are complementary in
gathering information to reduce the uncertainty around a policy measure. The fundamen-
tal purpose behind the measurement is – akin to policy evaluation – to improve interven-
tions: ‘performance is not only a concept, but also an agenda’ (van Dooren et al. 2010: 4).
Originally, performance measurement emerged in the United States at the level of local
government, where it was possible to assess inputs and outputs of local government ser-
vices. In the United Kingdom, it was an essential component of the general reform of the
civil service in the 1980s under the Thatcher government. Government publications of that
time made explicit the hypothesized links between increased managerial efficiency, a fore-
Copyright © 2020. Macmillan Education UK. All rights reserved.

cast decline in public spending and economic growth (Pollitt 1986: 159).
However, this idea of improving the effectiveness of public services through the use of pri-
vate-sector principles is not confined to conservative or liberal governments. During the New
Labour period in the United Kingdom (1997–2010), there was a clear drive towards increasing
transparency in the performance of public services through the introduction of targets in all
areas of the public sector (Radnor and McGuire 2004: 245). One of many examples of this was
the Modernising Social Services reform, which advanced a particular form of performance
measurement in the area of social care. It should, however, be noted that the Conservative–
Liberal Democrat coalition, in power in the UK between 2010 and 2015, removed a number of
these targets after assuming office. Yet these changes only entailed incremental rather than
more fundamental departures from the status quo (Matthews 2015).
More generally, in the last three decades performance measurement has emerged as a
major trend amongst many OECD countries (Kouzmin et al. 1999). The growing popularity
of this type of measurement can be substantively attributed to the reform process known as
New Public Management (McDavid and Hawthorn 2006: 300). Again, this indicates that
performance measurement has become a mainstream tool of administrative evaluation.

The Main Elements of an Evaluation

There is no fundamental incompatibility between the evaluation types outlined above; each
of them addresses a valuable dimension and thus complements our understanding of a
given policy’s effects. In recent years attention has increasingly turned to how one might
integrate results from evaluations that use different strategies, carried out from different
perspectives and using different methods. Regardless of which strategy a researcher adopts,
there are some elements that are inherent to each evaluation. Loosely following Rossi et al.
(2004) and McDavid and Hawthorn (2006), eight steps can be regarded as essential for car-
rying out an evaluation.
The first step requires identification of the clients of the evaluation (e.g. policy-makers,
scientific community, beneficiaries of the policy measure). Second, clear-cut evaluation
questions or criteria have to be formulated. The third step is about assessment of the
resources available to carry out the evaluation project (time, personnel, organizational
infrastructure etc.), which is followed by the need for a detailed analysis of the intended
effects of the policy measure as the fourth step. The fifth step is selection of the most appro-
priate evaluation strategy (e.g. a [quasi-]experimental versus a qualitative design). The next
two steps are about the development of measures, the gathering of data and carrying out
the data analysis.
As concerns data analysis, other methods than those discussed here can be employed,
including techniques for describing the distributional characteristics of the data (see
Burnham et al. 2008). Such descriptive techniques include bar charts, line graphs and
tables reporting the frequency of observations for the individual categories, thus allowing
for a comparative assessment of the data. For example, one could prepare a table that
shows the consumption of alcoholic beverages before and after the adoption of an
increased alcohol tax, broken down by age groups and/or gender. In this way, the expected
effect (i.e. lower consumption) can be investigated as to whether or not it is observable in
the entire target population or only parts of it. The eighth and final step is about publica-
tion of the results and presentation of recommendations for improving the existing policy
measure.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Key points
•• We can distinguish between administrative, judicial, political, scientific and economic evalua-
tion as well as performance evaluation; of these evaluation forms, political evaluation is the least
systematic.
•• In essence, evaluations – regardless of whether they are carried out before a policy is put in
place (ex ante) or afterwards (ex post) – are about producing information; the consequences of
the information for policy-making are determined by the political process.
•• There exists a strong link between the concept of (scientific) evaluation and evidence-based
policy-making; scientists who want to influence policy decisions according to the research evi-
dence they have found must take a proactive role.
•• All evaluation activities must meet certain criteria. Thus, despite the different evaluation
methods, eight general steps can be identified for conducting policy evaluation.

RESEARCH DESIGNS FOR SCIENTIFIC POLICY EVALUATION

Having provided an overview of the different forms of evaluation discussed in the pertinent
literatures, we now turn to a more detailed discussion of scientific evaluations. The latter
can be considered an important subfield of public policy and policy analysis. Contrary to
other forms of evaluation, scientific evaluations are based on the adoption of social science
research methodologies in order to provide an accurate assessment of the extent to which a
policy achieved its intended impacts and the causes of potential deviations from initial pol-
icy goals.
Scientific policy evaluations generally pursue two basic objectives. First, the goal is to
find out to what extent a certain policy actually achieved its intended effects. Second,
evaluators are interested in uncovering the factors that caused potential deviations from
the initial goals. In this regard, a major source for such deviations is deficient policy design
(e.g. Howlett et al. 2015; Tosun and Treib 2018). Policies might be based on incorrect
assumptions about cause–effect relationships. Moreover, policy-makers might have not
sufficiently taken into account potential interaction effects of new policy instruments
with existing ones, implying that the different instruments might operate in a contradic-
tory manner and weaken each other (e.g. Howlett and Del Rio 2015). A second source of
goal deviations refers to deficient implementation arrangements, such as lacking person-
nel and financial resources for the bodies in charge of implementation, as discussed in
Chapter 7.
Although scientific evaluations are rarely conducted on a regular basis, they provide
information that is of high relevance for policy-makers. As emphasized by Capano and Lippi
(2017), policy-makers tend to be driven by two basic motives that guide their policy choices,
namely the search for effectiveness (instrumentality) and the construction of a shared sense
and common acceptance of a policy (legitimacy). Although instrumentality and legitimacy
reflect distinctive rationales guiding policy choices, they both essentially rely upon evidence.
On the one hand, evidence about cause–effect relationships is crucial for governments seek-
ing to enhance the effectiveness of their policies. On the other hand, providing evidence for
the effectiveness of policy choices is one of the cornerstones of legitimate policy-making
(Adam et al. 2018).
Copyright © 2020. Macmillan Education UK. All rights reserved.

What sounds straightforward in theory can turn out to be a demanding exercise in prac-
tice, however. Even if we ignore potential political obstacles to conducting proper evalua-
tions, as discussed in the previous section, important methodological challenges prevail.
These challenges stem from varying sources. Apart from the fact that evaluations often
require costly and time-consuming endeavours of data collection, the major problem refers
to the development of appropriate research designs that tell us how likely it is that changes
in policy outcome are in fact attributable to distinctive choices of policy outputs. Even if we
have valid measures and data, we are still confronted with the problem of attributing
changes in effects or outcomes to causes. Are the observed outcome changes actually caused
by public policy? Which elements of the policy portfolio contributed by how much to the
observed outcome changes?
Addressing these questions can be challenging in practice. Two problems stand out in this
regard. First, policies might unfold their effects only under certain context conditions.
Although sometimes certain policy instruments might be discussed as panaceas (such as the
increase of tobacco taxes in order to prevent people from smoking), their effects strongly

depend on the context in which they are employed (Howlett and Lejano 2012; Head 2015).
Second, policy spaces are rarely empty, but consist of policy mixes (Howlett and Rayner
2013; Maor et al. 2017) that have been accumulating over time (Adam et al. 2017, 2019).
Evaluators hence no longer only face the challenge of handling many non-policy parameters
as control variables; the many parameters within policy mixes have to be handled adequately
as well. Thus, the challenge is not only to integrate a relatively high number of control vari-
ables, but also to cope with the increasingly complex nature of the independent variable –
the increasingly complex nature of accumulating policy portfolios. In sum, evaluators face
problems of attributing certain effects to specific causes; i.e. policies or policy instruments.
These attribution problems have different sources and vary in the extent to which they can
be effectively addressed.

e Conventional Attribution Problem: Coping with Complex Context

Th
Conditions
Policy problems are usually highly complex, with efforts to solve one aspect of a problem
potentially creating other problems due to interdependencies (Rittel and Webber 1973;
Alford and Head 2017). For instance, local governments facing high levels of debt might
close down public services, such as public swimming pools or child-care facilities, in order
to cut back spending levels. These measures, however, might reduce the overall life quality
in the given area, hence reducing the capacity of the region to attract new investments
and industrial settlements. As a result, local tax incomes may shrink, implying that the
government’s saving activities may even lead to an increase rather than a decrease in bud-
getary deficits. This example illustrates that we have to control for a broad range of factors
if we want to evaluate the isolated effect of a certain saving policy on public debt levels in
a given setting.
It is generally considered that the source of this attribution problem is that we typically
operate outside of experimental settings with controlled environments and treatments that
are only provided to one group (the treatment group) and not to the other (the control
group). Instead, we tend to rely on observational research designs and naturally occurring
Copyright © 2020. Macmillan Education UK. All rights reserved.

policy experiments. This implies that we need to control for all influences on the policy out-
come other than just the public policy in question.
Yet these difficulties characterizing policy evaluation have long been recognized and have
been accompanied by a continuous refinement of methodological tools. The conceptual and
methodological toolkits for evaluating public policy all reflect this attempt to isolate policy
effects by controlling for as many other influences as possible. Thereby, they help to alleviate
the ‘attribution problem’, i.e. the question of which factors are responsible for an observed
effect.
Conceptual tools for policy design (see, e.g., Howlett and Mukherjee 2018), such as strat-
egy maps (Kaplan and Norton 2004), reflect hypotheses about cause-and-effect relation-
ships that ideally try to include all factors affecting the relevant outcome. This approach
explicitly tries to identify interdependencies between different problems, i.e. the goal of
attracting investment and the goal of cutting public spending levels in our previous exam-
ple. The goal of strategy maps is to obtain an encompassing picture of different problems,
the potential solutions to them and their interactions. Yet such approaches can primarily be

seen as a broad list of hypotheses on different cause-and-effect relationships that might

potentially impact upon policy design. Strategy maps provide no systematic test of these
expected relationships.
Apart from such conceptual approaches, the methodological toolkit for policy evalu-
ation – including comparative case study designs and quantitative methods – attempts
to isolate the actual impact of public policy by controlling for potentially confounding
influences. In comparative case study designs, a very common approach is to compare
the most similar systems in order to hold context conditions constant as far as possible
(see,e.g., Toshkov 2016). In our example, we would compare the impact on public ser-
vice cuts on budgetary deficits for local governments that are similar on all other aspects
that might affect this relationship, such as legal arrangements defining the competen-
cies and discretion of local governments, the degree of industrialization, the structure
of the population, the level of urbanization and so on. Quantitative methods, by con-
trast, can correspond to ‘quasi-experimental’ designs that elicit the causal effects of
interventions. By comparing observations lying closely on either side of the threshold,
it is possible to estimate the average treatment effect in environments in which ran-
domization is unfeasible. Comparing the outcome of the treatment group (for which the
intervention applies) to the outcome of the non-recipient (control) group will hence
deliver the local treatment effect.
All of these approaches have certain limitations, strengths and weaknesses. Nevertheless,
if carefully applied, they provide an appropriate toolkit that might allow for overcoming the
problem of controlling for the high contextual diversity in which public policies are applied.
This way, we should be capable of isolating the causal effects of a given policy on a certain
outcome, at least if we have a sufficiently broad, balanced and valid database for applying the
different methodological tools. In short, complexity in context conditions constitutes a
well-known and conventional research design challenge for evaluation studies that can – in
principle – be handled.

e More Challenging Attribution Problem: Evaluating Complex

Th
Policy Mixes
Copyright © 2020. Macmillan Education UK. All rights reserved.

While the evaluation literature has been primarily concerned with problems of controlling
for variation in context conditions, more recent contributions highlight research chal-
lenges that emerge from a different source; namely, the fact that policies are not only
applied in different contexts, but are also added to different instrument mixes or policy
portfolios (Adam et al. 2018). Policy spaces are rarely empty, but typically represent con-
figurations of complex policy mixes (Howlett and del Rio 2015) that are continuously
accumulating over time (Knill et al. 2012; Adam et al. 2019). While this insight in some
ways reflects a truism in the public policy literature, this development of policy accumula-
tion bears important challenges for evaluation studies, highlighting a different dimension
of the abovementioned attribution problem. Evaluators are not only required to handle
many non-policy parameters as control variables; the many parameters within policy
mixes have to be handled adequately as well. Policy elements do not exert their influence
independently of each other, but their effects are often conditional on the configuration

of other policy elements. Thus, the challenge is not only to integrate a relatively high num-
ber of control variables, but also to cope with the increasingly complex nature of the inde-
pendent variable – the increasingly complex nature of public policy. Evaluation studies
have to answer the question of how much the individual elements within policy mixes
account – independently or in combination with other elements – for the observed level
of outcome achievement.
If we accept the argument that complex (and potentially growing) policy mixes pose chal-
lenges for policy evaluation, the immediate question that arises is how such challenges
might be addressed. We can distinguish three approaches that are applied in response to
these problems: (1) theoretical ex ante assessments of interactions, (2) the modelling of
interactions and (3) attempts to keep interaction factors stable. Yet each of these approaches
can be considered merely as a partial remedy to address the challenges emerging from com-
plex policy mixes (Adam et al. 2018).
First, while the relevance of interactions between different elements within the policy
mix is well known to the literature on policy design and policy mixes, this literature assumes
that the effects of policy mixes can be assessed ex ante on theoretical or conceptual grounds
(Linder and Peters 1984; McConnell 2010a, 2010b). For example, where policy mixes were
found to be coherent, consistent or congruent, they were assumed to have the intended
impacts (Howlett and Rayner 2013). The key assumption underlying this approach is that
with ‘careful examination’ we would be able to ‘improve policy designs and outcomes’
(Howlett and Lejano 2012: 362).
However, such ex ante theoretical assessments are insufficient for constellations in which
policy mixes continue to grow and accumulate, as governments continuously respond to
societal demands by adding new policies to the existing stock of policy targets and instru-
ments (Adam et al. 2019). This development renders any assessment of instrument choice
and related outcomes more challenging. We need more time, resources and knowledge to
identify all measures related to the outcome of interest.
Second, instead of ex ante theoretical considerations of interaction effects, we could try to
explicitly model interactions within complex policy mixes. According to Majone, such inter-
action effects are almost inevitable whenever we are confronted with a ‘policy space’ which
is of limited size and a ‘population of policies that grows relative to the size of the space’
Copyright © 2020. Macmillan Education UK. All rights reserved.

(Majone 1989: 169). From this perspective, policy interaction is a result of ‘congestion’
(Majone 1989: 159) due to accumulating policy mixes (Knill et al. 2012; Adam et al. 2019).
While some interactions will lead to an unintended weakening of individual policy compo-
nents’ effects, other components will reinforce each other’s effectiveness. If we want to
identify the accurate effect size of individual components within the policy mix, we need to
account for both types of such interaction effects accurately. By treating policy mixes as
variables whose interactions are explicitly considered within statistical models of outcome
changes, the independent variable problem of evaluating policy mixes might be reduced.
While this might very well be true for policy mixes of low and moderate size, this strategy
soon becomes unviable in the light of more complex policy mixes.
Third, and in response to these problems, sophisticated approaches have been devel-
oped to avoid this bias, including experimental and quasi-experimental designs. Quasi-
experimental approaches to policy evaluation in particular perform rather well in
ensuring the internal validity of findings. These approaches treat policy mixes as

background factors that are kept constant across research subjects. This works espe-
cially well in studies that use panel data on individual behaviour before and after spe-
cific policy reforms of interest. With such approaches, the behaviour of individuals
affected by the policy can be compared with a quasi-control group of unaffected indi-
viduals. Furthermore, the complexity of the policy mix within which the policy reform
took place is identical for both groups. This ensures the internal validity of claims about
the effect of being exposed to a certain policy. Even as the continuous process of policy
accumulation makes policy mixes increasingly complex, this research design should pro-
tect the internal validity of findings. In a similar vein, experimental rollouts of public
policies are increasingly applied. This strategy explicitly builds on controlling who is
affected by the reform and who is not.
Still, the strategy of keeping background factors stable comes along with several limita-
tions. In particular, this strategy only helps to address problems of internal validity of
outcome-based approaches, but fails to address problems of external validity. External valid-
ity refers to the generalizability of claims and is thus a crucial prerequisite for policy transfer
between countries and jurisdictions. Simply put, showing that a specific policy was effective
within the context of one particular policy mix hardly allows for predictions about how this
policy will perform within a different policy mix. Accordingly, while the identification of
policy effects is possible in such research designs, we should be careful in extrapolating
these findings to other contexts. External validity problems of policy-based evaluations are
thus closely linked to problems of policy transfer (Dolowitz and Marsh 2000).
For instance, let us consider evaluations that find that, for the United States, a 10%
price increase in the cost of cigarettes results in as much as a 13% decrease in cigarette
consumption amongst youth and young adults (Ross and Chaloupka 2003). Because of the
research design choices made, the internal validity of this finding is arguably rather high.
The studies show that, given the existing policy portfolio, an increase in tobacco taxes
comes along with a clearly quantifiable decrease in smoking prevalence. Yet the problem
of policy-based evaluations of this kind is their limited external validity. Whether a
tobacco tax increase will be similarly effective in another country depends on the extent
to which that country’s policy portfolio is similar to the policy portfolio of the United
States at the time that the evaluation studies were conducted. For instance, whether or
Copyright © 2020. Macmillan Education UK. All rights reserved.

not tax increases are complemented by smoking or advertising might crucially affect their
effectiveness (Adam et al. 2019). In a similar vein, a comparable reduction in unemploy-
ment replacement rates in Sweden and the United Kingdom will interact with highly dif-
ferent social policy portfolios in both countries, given their different welfare state
traditions (Esping-Andersen 1990). As a consequence, knowledge about effects of such a
policy change on unemployment rates in Sweden can hardly be transferred to the British
context, even when effectively controlling for confounding factors, such as socio-economic
or demographic conditions.

Main Insights
We have seen that the development of appropriate research designs for policy evaluations is
a challenging task. These challenges emerge from two sources that are related to the so-
called attribution problem. This problem consists of assessing the causal link between policy
intervention and outcome effect. While conventional discussions of the attribution problem

see the source of this problem in the multitude of non-policy factors that influence out-
comes, the phenomenon of accumulating and increasingly complex policy mixes represents
an additional source of the attribution problem.
The need to estimate conditional effects within increasingly complex policy mixes creates
theoretical and methodological challenges. Bringing all elements within the policy mix to
the forefront by explicitly modelling their conditional impacts implies the handling of a high
number of parameters, even for relatively simple policy mixes. Depending on the research
design and data structure, this can quickly lead to a situation where too many variables are
tested on too few cases. The resulting challenges to the internal validity of results could be
mitigated with the help of (quasi-)experimental evaluation approaches, which single out
individual elements from the policy mix while treating the rest of this mix as constant back-
ground factors. Yet since the evaluative strategy consists of focusing on policy effectiveness
within one specific policy mix, advocates of policy transfer will nevertheless have to deal
with the conditionality of the identified effect in the foreign policy mix in order to anticipate
how the policy will perform within their domestic mix (Bovaird 2014).
Overall, we seem much better equipped to evaluate whether policy mixes as such are
associated with improvement of deteriorating outcomes than to assess which elements
within these mixes work effectively and why. This makes it very difficult to reform these
mixes in a meaningful way on the basis of evidence, and it leads to a situation that is
skewed towards amending existing mixes and against simplifying existing mixes by get-
ting rid of ineffective or inefficient policy elements (Adam et al. 2018).

Key points
•• Policy evaluation requires the use of appropriate research methods, which can be both experi-
mental and non-experimental; the corresponding literature tends to stress methodological issues
more than other literatures in public policy.
•• Policy evaluation must also react to the adoption of policy mixes.
•• Recent empirical findings report that policies tend to accumulate over time; the accumulation
of policies complicates the evaluation process due to the attribution problem.
Copyright © 2020. Macmillan Education UK. All rights reserved.

•• The scientific evaluation of policies requires the overcoming of the attribution problem.

THE POLITICAL DIMENSION OF POLICY EVALUATION

The results of policy evaluation can be utilized to achieve certain political goals. Moreover,
the decision to start a new policy process with the objective of replacing an existing public
policy with a new one is based on political considerations. In this section, we address these
two political dimensions of policy evaluation.

Policy Evaluation versus Political Evaluation

There are two fundamentally different points of view on evaluation activities. On the one
hand, scientific realists argue for the possibility of an objective description. They advocate a
separation of facts and values with the aim of producing politically neutral knowledge. This
perspective is based on the assumption that there exists a set of clear-cut policy goals and that

there is agreement on the indicators on the basis of which the evaluation is to be carried out
(see Pawson and Tilley 1997). On the other hand, social constructivists – also sometimes
labelled post-positivists – claim that evaluation research is entirely subjective. They stress that
evaluation does not study public policy as such, but its interpretation, which entails that the
evaluation process itself is based on values and therefore is socially constructed (see Fischer
1995). However, both sides seem to agree that evaluation cannot be practised without being
somehow influenced by political judgement (Taylor and Balloch 2005: 1–2).
A helpful perspective for understanding the important difference between policy evaluation
and political evaluation is provided by McConnell (2010b). Even though in his book McConnell
concentrates on the concept of ‘policy success’, the reasoning can still be applied to what is of
interest here, namely the extent to which evaluation can be political. This distinguishes
McConnell’s work from other conceptualizations of policy success and policy failure that con-
sider these as ‘good’ or ‘bad’ policy decisions (see, e.g., FitzGerald et al. 2019). According to
McConnell (2010b), policy success can be assessed on the basis of four dimensions.
First, a policy is successful when the implementing actors achieve the objectives laid
down in a legal act. This dimension overlaps with implementation studies (see Chapter 7). A
more demanding approach is represented by the second dimension, which concentrates on
the question of whether a policy output produces the intended outcomes and/or impacts. A
third dimension is the benefit it brings to the target group. The fourth dimension is about
meeting the criteria of the respective policy domain. Different policy areas are characterized
by different values being widely held by their relevant policy community, such as precaution
in environmental policy. Thus, a policy may be deemed successful when it concurs with these
specific standards. From this discussion of policy success we can derive what policy evalua-
tion should ideally be about, i.e. an assessment of these four dimensions.
Political success is clearly different from policy success (see McConnell 2010b, 2011,
2015). The first dimension of political success refers to increasing electoral prospects. As
explained in Chapter 4, the central objective of elected officials is to stay in power. To attain
this, the governing party will try to employ evaluation to demonstrate that the public poli-
cies it made while in office were successful. The second dimension refers to the process of
policy-making. A policy measure can be successful in this regard if it involves a narrow defini-
tion of the policy problem that triggered the policy process, gives the appearance of dealing
Copyright © 2020. Macmillan Education UK. All rights reserved.

with the problem and helps to counter critics or gain support from the most relevant actors.
A policy corresponding to these criteria can help to keep more difficult-to-solve policy prob-
lems off the decision agenda and ease the business of governing. The third dimension is
about the values and direction of government. A policy measure might be politically success-
ful in helping to forge new approaches to policy-making. Political evaluation would provide
the information necessary to achieve success with regard to these three dimensions.
The understanding of the political success of a public policy makes political evaluation
different from objectively assessing expected and unexpected policy effects. In short, it is
about the purposeful use of information to stay in power. Bovens et al. (2008: 320) argue
that even the decision to make an evaluation could be politically motivated, e.g. by ‘the
replacement of key officials, elections that produce government turn-overs, incidents or fig-
ures that receive publicity and trigger political calls for an investigation, and so on’. This
concurs with the argument advanced by Rossi et al. (2004: 37) that actors initiating evalua-
tion might have hidden agendas: ‘occasionally, an evaluation is commissioned to provide a
rationale for a decision that has already been made behind the scenes to terminate a pro-
gram, fire an administrator, and the like’. Alternatively, evaluation can become political to

the extent that policy-makers ignore recommendations to discontinue programmes on the

basis of a negative evaluation (see Quinn and Magill 1994).
The politics of policy evaluation has also been discussed in terms of the diffusion of regu-
latory impact assessment. According to Turnpenny et al. (2009: 645), the US-based litera-
ture posits that elected politicians want regulatory impact assessment performed so as to
receive information on whether ministerial departments and agencies are disadvantaging
important political constituencies. Thus, the adoption of policy appraisal is perceived as a
means of politically controlling the bureaucracy. More generally, the institutional context of
polities has been found to be decisive for the actual use of this tool (Radaelli 2004). Sager
and Rissi (2011), for instance, demonstrate that the specific institutional characteristics of
the Swiss polity have prevented the success of regulatory impact assessment. At the Swiss
federal level, various direct democratic instruments (e.g. referendums) are used. When
drafting legislation it is more important to include the opinions of most actors than the
results of policy appraisal. Thus, the full potential of ex ante evaluations cannot be exploited.
Another way in which policy evaluation can become political is by telling evaluators what
they should find. This aspect specifically relates to the increasing dominance of evaluation
research as ‘contractual research’ (Wollmann 2007: 399). This form of evaluation entails an
agreement between the evaluator and the client. The client agrees to fund the evaluation,
defines the subject matter and the leading research questions, and approves the evaluation
method. The evaluator, in turn, agrees to complete the evaluation within a specified time
period and budget. By receiving funding from an external source, i.e. the client, the evalua-
tor loses independence regarding the subject matter, the research questions and methods.
While this is already undesirable in terms of academic research, contractual evaluation
might also bear the risk that the evaluator is asked to deliver a type of ex ante defined results.
It is important that policy evaluators are aware of their social responsibility.

Evidence-Based Policy-Making
In the traditional policy cycle introduced in Chapter 1, evaluation follows policy implemen-
tation. Based on the assessment of what works out and what does not, policy-makers might
improve public policy by initiating a new policy process. However, the decision to utilize the
Copyright © 2020. Macmillan Education UK. All rights reserved.

findings of evaluation is a political one. Classic policy analysis as it emerged in the 1960s
strongly built on the evaluative rationality applied to public problems, i.e. to improve policy
outcomes and impacts by applying analytical methodologies to policy assessment. Evidence-
based policy-making represents a recent effort to reform policy processes by prioritizing
decision-making criteria that are evidence based and to stimulate processes of policy-
oriented learning (Howlett 2009b: 154). As already briefly discussed in this chapter, the idea
underlying evidence-based policy-making is that governments can learn from experience
and avoid repeating the errors of the past. In other words, it is an approach that is designed
to help politicians make ‘well-informed’ decisions about policies by putting the best avail-
able evidence at the heart of policy formulation and implementation.
To understand better what evidence-based policy-making is about, we have to return to
Chapter 6, in which we discussed how policy proposals are drafted. In an ideal world, these
proposals would represent the most effective or, indeed, most efficient way of resolving a
social problem, howsoever defined. In reality, however, many factors influence the design of
public policy. These include experts, ideas, international organizations and interest groups.
In a complementary vein, the partisan ideology of elected and appointed politicians and

their private interests can be decisive for the design of a policy. This suggests that a policy
proposal is not necessarily drafted in a way to ensure that the solution to the social problem
is attained. It is this possible flaw in designing public policy that the concept of evidence-
based policy-making seeks to remedy.
In the United Kingdom, for instance, under the New Labour government (1997–2010)
the profile of evidence-based policy was raised (Wells 2007; Boaz et al. 2008). Numerous
publications acknowledge this drive. In 2000, for instance, a report by the Performance and
Innovation Unit called for policy analysis to be placed at the heart of policy-making (Wilson
2008: 159). Another publication showing the growing importance of evaluation is the 2003
Magenta book: Guidance notes for policy evaluation and analysis.
At the international level, the use of evaluation – as one principal component of evidence-
based policy – varies strongly across countries. Canada, Germany, Sweden and the United
States are broadly considered ‘pioneers’ in the use of evaluation. A bit later, in the 1970s and
1980s, countries such as Denmark, France, the Netherlands, Norway, Switzerland and the
United Kingdom also institutionalized evaluation. Yet a sizeable group of countries only
adopted such an evaluation culture during the 1990s, including Finland, Ireland, Italy and
Spain (Viñas 2009: 459). Other European countries such as Belgium still have a notably less-
developed evaluation culture (Varone et al. 2005).
Exactly what constitutes the ‘evidence’ in evidence-based policy-making remains conten-
tious in the literature. It is, however, clear that evaluation is only one instrument for inform-
ing policy decisions. More generally, Head (2008) suggests that three types of evidence are
especially relevant in modern states: scientific knowledge, policy management knowledge
and political knowledge. Scientific knowledge – to which we can also attribute evaluation if
based on a scientific methodology – is the product of the systematic analysis of current and
past conditions and trends, and the analysis of the causal interrelationships that explain
them. Policy management knowledge is the ‘practical wisdom’ of professionals and the orga-
nizational knowledge associated with managing programme implementation. The political
form of knowledge resides primarily in politicians, political parties, organized groups and
the media. Political knowledge is related to a wide range of activities, including considering
and adjusting strategies or tactics, undertaking agenda-setting, determining priorities and
building coalitions of support. Head’s central argument is that evidence-based policy cannot
Copyright © 2020. Macmillan Education UK. All rights reserved.

exclusively rely on scientific knowledge, but must be complemented by the other two forms
of evidence in order to be feasible and effective in policy formulation.
While evidence-based policy-making appears to be a sensible approach in theory, there
are several practical limitations (Oliver et al. 2014; Wellstead et al. 2018). Governments are
under constant pressure to solve problems and characteristically do not have the time to
study things as much as they want. Consequently, they are often required to make a policy
decision based on an ‘informed guess’, instead of waiting for a more complete answer. The
second limitation is that many policy initiatives rest on assumptions about the future that
are not testable. Policy-makers have to make predictions about future impacts against the
credibility of the sources of the information and their own common sense.
The capacity of governments actually to practise evidence-based policy-making is also cru-
cial (Howlett 2009b). Cohen and Levinthal (1990: 133) introduced the term ‘absorptive capac-
ity’ for the ability of a firm to innovate. As described by the authors, absorptive capacity has
two constituent components: the capacity to adopt ideas from the outside world (i.e. adoption
capacity) and the capacity to create new inventions (i.e. invention capacity).
Adolino and Blake (2011: 27) adapted the logic of this model to governmental learning on
the basis of evaluation. According to these authors, governments are most likely to learn from

the past when they have expertise and good information acquired through communication
with so-called policy networks. If the exchange between government and policy networks is
intense, learning will take place in government and society. By the same token, in the case of a
high-capacity government with minimal links to policy networks, learning based on evalua-
tion findings will be restricted to the government itself. In constellations where governments
have little expertise or insufficient information, learning will take place at much lower levels,
even if the government in question interacts intensely with policy networks.
Evidence-based policy-making is a rationalist concept that stresses the possibility of
objective information. An important criticism of it has been the constructivist or interpre-
tative position, which argues that even scientific knowledge is socially constructed. From
this perspective, scientific evidence cannot inform the design of public policy without being
affected by beliefs and principles as well as particular theoretical frames and interpretations
of the world (see Sanderson 2002: 6). Another line of criticism is advanced by Marston and
Watts (2003: 158), who argue that there is a risk that evidence-based policy can be misused
by political elites to increase their strategic control over the definition of the nature of policy
problems, which could turn into a threat to open and democratic policy-making.
Finally, it is important to note that there are different ‘levels’ of evidence that might
inform policy-making (Camasso 2003). In health policy, for instance, Brownson et al. (1999:
90) state that there are at least two levels of evidence: one focusing on the importance of a
particular health condition and its link with some preventable risk factor, and another that
concentrates on the relative effectiveness of specific interventions to address a particular
health condition. Depending on which level of evidence is taken into consideration, differ-
ent or even contradictory policy recommendations can be derived. This further underlines
the fact that basing policy decisions on evidence does not eliminate the need to choose from
a variety of policy options.
In reality, evidence-based policy-making is only practised in a few policy areas. One area
in which evidence is systematically taken into consideration when making policy is the
health sector, but even there Shimkhada et al. (2008) found that this does not happen very
extensively. One of the factors impeding the more systematic use of evidence in policy-
making refers to the structural barriers to communication between researchers and decision-
makers. Even in the United Kingdom, where the political will to establish evidence-based
Copyright © 2020. Macmillan Education UK. All rights reserved.

policy-making was considerable, the policy areas in which this approach was advanced were
limited (Boaz et al. 2008: 242).
Moreover, evidence-based policy-making is often an impossible quest. This can be traced
to the fact that policy spaces are rarely empty, implying that new policies need to be assessed
against the backdrop of a complex bundle of already existing policy targets and policy instru-
ments. In view of these constellations, the sound collection of evidence is restricted to the
assessment of the policy effects of whole policy bundles rather than singular policies. We are
only able to assess the marginal effects of a new policy added to this bundle, rather than
precisely mapping potential interaction effects between different policy elements.

Key points
•• It is important to distinguish between policy evaluation and political evaluation. The first is
about comparing the expected and observed effects of a public policy; the second is about utiliz-
ing formally or informally derived information about public policy for political purposes.

•• There are many different ways in which an evaluation can be political. What they all have in
common is the objective of helping elected officials to stay in power or to induce a change in
the partisan composition of government.
•• Evidence-based policy-making is about using experience or new information to avoid a repetition
of the errors of the past or to find better resolutions to policy problems by means of policy-oriented
learning. It is a political decision as to whether or not base policy formulation on evidence.
•• The ‘evidence’ to be used when formulating public policy does not only refer to findings of evalu-
ation studies, but also includes other types of information. However, there are many different
definitions in the literature.

CONCLUSIONS

Evaluation is about assessing the expected and unexpected outcomes and impacts of pub-
lic policy. There are multiple strategies for designing an evaluation study. Most impor-
tantly, the evaluator should be clear about whether the principal interest is in assessing
the effectiveness or efficiency of a policy measure. Once this decision has been taken, he
or she can choose from various evaluation methods. Another important question relates
to the purpose of the evaluation. The findings will depend on the way the evaluation is
realized, since an exclusively objective assessment of policy effects is hardly possible.
From this it follows that the findings of evaluation studies – regardless of how well they
are designed – should always be seen in the social context in which they were
undertaken.
We have further learned in this chapter that evaluation is an integrative part of evidence-
based policy-making and is thus important for providing feedback to policy-makers. The
recent popularity of this approach can be seen to reflect the growing uncertainties sur-
rounding policy-making and the increasing complexity of policy decisions; scientific evi-
dence appears to offer a way of controlling this uncertainty and reducing complexity. At
least, this is the perspective that the great majority of studies on evidence-based policy-
making take.
Depending on the results of the evaluation process, a new policy cycle can be initiated
Copyright © 2020. Macmillan Education UK. All rights reserved.

with the aim of terminating or modifying the newly established policy. Thus, evaluation
presents an important tool for policy-making, because it might bring an issue back to the
political agenda and motivate policy actors to improve the policy design. However, whether
lessons for future policy decisions are actually drawn from evaluation findings depends on
numerous factors, of which some lie outside the control of policy-makers. For example, the
occurrence of a natural disaster can result in more prominence given to (scientific) evalua-
tion results. Conversely, the entering in office of a government with an ideology that is at
odds with the evaluation findings could lead to evaluation findings’ being ignored when
proposing and adopting public policy.

WEB LINKS

www.aes.asn.au. This is the website of the Australasian Evaluation Society.

www.eval.org. This is the website of the American Evaluation Society.
www.evaluation.org.uk/. This is the website of the UK Evaluation Society.
www.evaluationcanada.ca. This is the website of the Canadian Evaluation Society.
www.oecd.org/gov/policy-monitoring-evaluation.htm. This website details the
OECD’s approach to policy monitoring and evaluation.
www.socialresearchmethods.net. On this website there are many resources and links
to other websites that deal with applied social research methods.

FURTHER READING

Bardach, E. (2008). A Practical Guide for Policy Analysis: The Eightfold Path to More
Effective Problem Solving. Washington, DC: CQ Press. This book provides a helpful
complement to this chapter, as it systematically links the first stage of the policy
cycle with the final one.
Cairney, P. (2016). The Politics of Evidence-Based Policy Making. London: Palgrave. A
stimulating read that shows the contested nature of evidence and how it affects
policy-making.
Dahler-Larsen, P. (2012). The Evaluation Society. Palo Alto, CA: Stanford University
Press. A book that elaborates on the effects of policy evaluation on society and
presents some surprising findings.
McConnell, A. (2010). Understanding Policy Success: Rethinking Public Policy.
Basingstoke: Palgrave Macmillan. This book provides a holistic understanding of
policy evaluation by framing it in the context of policy success.
Rossi, P. H., Lipsey, M. W. and Freeman, H. E. (2004) Evaluation: A Systematic
Approach. Thousand Oaks, CA: Sage. A useful resource that provides insights into
a wide range of evaluation-related topics.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Toshkov, D. (2016). Research Design in Political Science. London: Palgrave Macmillan.

A very accessible and useful book for those interested in obtaining an encompass-
ing introduction to research designs that can also be applied to evaluation studies.

Reader’s guide
The governance concept focuses on patterns of political steering, i.e. the institution-
alized relationship between public and private actors, in order to resolve social prob-
lems. It concerns both hierarchical and non-hierarchical modes of political steering,
and it is particularly the recognition of the non-hierarchical forms that distinguishes
governance from government, i.e. the traditional mode of hierarchical intervention,
which we have implicitly employed as the point of reference in previous chapters.
The second innovative feature of the governance concept relates to the fact that it
allows for a more holistic view of policy-making that cuts across policy stages. This
is attained by placing the forms of cooperation between public and private actors at
centre stage. In this way, the governance perspective can be employed to scrutinize
processes of problem definition and agenda-setting as well as decision-making,
implementation and, to a certain extent, even evaluation. The chapter begins with
an introduction of the central concepts and modes of governance. We also address
the question of whether there has been a shift from hierarchical to non-hierarchical
forms of political steering over time. We then present a typology of governance
types, to provide a better understanding of how public and private actors might
cooperate in the policy-making process and to what extent their relationship is char-
acterized by a dominant position of the state. Finally, we discuss and apply the dif-
ferent criteria for evaluating the extent to which governance might be characterized
Copyright © 2020. Macmillan Education UK. All rights reserved.

as ‘good’ or ‘bad’.

INTRODUCTION

The term ‘governance’ has had an impressive career from the early 1990s onwards. It has
developed into the catchword for an ever-growing number of studies in the social sciences.
At the same time, it is used as a ‘magic formula’ – often in terms of ‘good governance’ – in
political speeches and documents, at both the domestic and international levels. However,
as is often the case with new concepts, there is no consensus regarding its meaning and
specific applicability (Kooiman et al. 2008: 2). The definition of the term varies considerably
across different subfields and research strands of the social sciences (see Pierre and Peters
2000: 7; Kohler-Koch and Rittberger 2006; Levi-Faur 2012). This lack of a common meaning
can be traced to the fact that the concept is used not only in an analytical way, but also in a
normative sense (Doornbos 2001).

164
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:18:24.
Governance: A Synoptic Perspective on Policy-Making 165

This basic assessment, however, does not mean that governance and related research are
not of relevance for the study of public policy. On the contrary, it implies a new perspective
on public policy that cuts across the distinction of policy stages made in the previous chap-
ters. In placing patterns and forms of coordination between public and private actors at
centre stage, a governance focus implies that processes of problem definition, agenda-
setting, decision-making, implementation and, to a certain extent, evaluation are inter-
preted through a different analytical lens. The governance perspective sheds light on specific
policy regimes understood as institutional forms and instruments shaping processes of col-
lective action. The focus is hence on general patterns of policy-making in a given policy sec-
tor and country, and potential changes in these patterns over time. What types of governance
exist, and when can we consider governance to be ‘good’? We now address these questions
in detail.

CENTRAL CONCEPTS AND MODES OF GOVERNANCE

When taking a closer look at the rich literature on governance, we are confronted with dif-
ferent, partially overlapping definitions and concepts. To get a clearer understanding of the
relevance of the governance concept, we need to differentiate and classify these approaches.
On this basis, we can turn to a more detailed discussion of different modes of governance
from a policy-analytical perspective.

Governance Concepts
On the basis of various classification attempts (Mayntz 2010: 37–8; see also Rhodes 1996,
1997: 46–60; Benz and Dose 2010), two general conceptions of governance can be
distinguished.
The first – rather broad – definition conceives of governance as political steering and
hence purposeful attempts at coordinating individual action in order to achieve certain pol-
icy goals. Thus, governance refers to the collective settlement of social affairs in a polity,
Copyright © 2020. Macmillan Education UK. All rights reserved.

including a broad range of different modes, such as hierarchical intervention and non-
hierarchical steering, based on cooperation between public and private actors or patterns of
private self-governance.
In contrast to this conception of governance as a classification of different modes of polit-
ical steering, the term is also widely used to describe a distinctive mode of steering, namely
non-hierarchical approaches (Peters and Pierre 1998). Rhodes (1996: 660), for instance,
refers to governance as self-organizing, interorganizational networks. These networks
encompass not only public but also private actors, and cooperation between these different
types of actors is based on the need to exchange resources and negotiation over shared pol-
icy objectives and solutions.
This narrower definition of governance is emphasized in particular in the literature
on global governance (Czempiel and Rosenau 1992; Biermann et al. 2010; Biermann and
Pattberg 2012) or transnational governance (Hale and Held 2011). In this debate, gov-
ernance is restricted to non-hierarchical modes of political steering; i.e. it is explicitly
delineated from ‘government’ as the traditional mode of hierarchical intervention. This dis-
tinction is based on the understanding that beyond the nation state attempts at collective

Table 9.1 Overview of governance concepts.

Definition of governance Description

Governance as classification of Deliberate collective coordination of individual action,
modes of political steering including both hierarchical and non-hierarchical modes
Governance as a distinctive mode Non-hierarchical modes of political steering
of political steering

coordination necessarily have to be based on non-hierarchical modes, as the option of hier-

archical intervention is restricted to the nation state (Zürn 2004).
It is in the context of this latter distinction between governance and government that an
emphasis on the governance concept, in the sense of ‘good governance’, has emerged. This
development has been particularly driven by international organizations like the World
Bank, who have stressed the positive effects of non-hierarchical patterns of political steer-
ing that emerge from broad societal participation in policy-making, especially in the field of
development policy (Knill 2004; see Table 9.1).

Modes of Governance
Basically, we find a distinction between three governance modes in the literature, namely
hierarchy, markets and networks (Héritier 2002b; Knill and Lenschow 2003, 2005; Lütz
2010). Before outlining these different modes, it is important to stress that they are all
mutually present in modern states. The different modes can vary in the same country regard-
ing their dominance in the individual policy areas. Moreover, the dominance of a mode can
change, as we will show in the next section.
Governance by hierarchy stresses the role of formal rules and procedures that are binding
for both public and private actors. Most importantly, however, the state has a monopoly on
the use of force to bring private actors into compliance with public policy. From this it fol-
lows that the state has a sanctioning power that exceeds that of private actors to a great
extent, indicating that the relationship between public and private actors is asymmetrical.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Compared with other types of coordination such as markets or networks, the state plays a
key role in policy-making within this mode (see Box 9.1). It hierarchically intervenes to
produce and supply common goods (e.g. infrastructure, education or clean air) and defines
the legal framework without which no economic activities could be realized.
By supplying common goods, the state generates fair conditions for all market partici-
pants and a reliable framework for economic activities, which represents the main advan-
tage of hierarchical governance. In contrast, the disadvantage of this mode of governance is
a decoupling of the state from the policy needs of private actors and an insufficient knowl-
edge of the ‘actual’ problems they are confronted with. This has important implications for
our previous discussion of problem definition and agenda-setting in Chapter 5. In constella-
tions that are characterized by hierarchical governance, it is possible that the problems that
receive attention from policy-makers might deviate from what citizens regard as problems,
leading to public policy that may not be accepted or is difficult to implement.
Governance by markets represents the opposing model to hierarchical governance, as it is
based on the idea that goods and services are allocated efficiently without intervention by the
state. Markets provide individual and corporate actors with an ideal setting for exchanging

Box 9.1 Experimental governance in China and Europe

China has embraced a specific governance tool that consists of designating regional
or local pilot zones where specific policies or instruments for implementing policies
are put in place (Qian 2018). The lessons from policy outcomes and impacts from
the experimental zones are intended to inform future policy decisions. McDowall
et al. (2017) contend that this form of experimental governance has been used to
improve the implementation of environmental policy, which is hampered by a tax
system that encourages local government to focus on economic growth and pays
insufficient attention to environmental degradation. To reduce the implementation
gap, the central government has provided the provincial government with direct
incentives to meet environmental goals. The authors also point out similarities
between the Chinese approach to experimental governance and the European one,
as they both rely on drawing lessons from pilots or demonstrations at the level of
one member state (in the case of the European Union [EU]) or regions (in the case
of China and the EU). In this context, the European semester (a process by which
experiences with policies are assessed across Europe) and the cohesion policy (which
facilitates policy diffusion across regions through funding) are identified as impor-
tant tools for experimental governance in the EU.

resources based on price. This argument is founded on the assumption that the prices of goods
and services contain all the information relevant for their exchange and, therefore, that
exchange relationships can emerge between all kinds of market participants. Based on the
price, market participants can judge whether it is advantageous to agree to a transaction or not.
Another central assumption of market governance is that actors are rational, i.e. that
they seek to maximize their own well-being, so that they are likely to produce negative
externalities that cause harm to third persons and hence lower society’s overall welfare level.
Usually, the most effective way of internalizing negative externalities is provided by state
intervention, which is also needed to define the basic rules required to make sure that mar-
Copyright © 2020. Macmillan Education UK. All rights reserved.

kets function at all. Thus, the logical consequence of trying to remedy the disadvantages of
market governance would be to strengthen governance further by increasing the established
hierarchy. At the same time, the hierarchical definition and enforcement of the rules of the
game (including such elementary things as property rights) constitute basic requirements
for the emergence and functioning of markets.
The third mode of coordinating the relationship between public and private actors is net-
work governance. The literature on this challenges the conventional wisdom that the market
is the only efficient system for non-hierarchical coordination of societal actors. An impor-
tant discussion unfolded as to whether networks are only a combination of elements of
market and hierarchy, or whether they would be better understood as a specific form of
governance in their own right (Provan and Kenis 2008: 232; Raab and Kenis 2009: 205).
Policy networks are defined as stable sets of interdependent public and private actors
who interact informally to achieve distinctive but interdependent goals (Laumann and
Knoke 1987; Schneider 1992, 2006). The members of a policy network are linked to one
another through the exchange of policy-relevant information, expertise and additional
resources (Rhodes and Marsh 1992: 10). The basic idea is that policy networks provide the

governing resources to intervene in policy areas, which are not available within the state
apparatus (Kenis and Schneider 1991: 41). Policy networks can participate in the prepara-
tion of decisions taken by the executive or the legislature, i.e. policy drafting, and help to
implement them. Governments will, in principle, work most closely with those groups offer-
ing the most valuable resources (see Daugbjerg 1998).
Patterns of cooperation in policy networks may take many different forms, including
not only forms of horizontal coordination between public and private actors, but also pat-
terns of societal self-governance, for instance so-called private interest governments
(Streeck and Schmitter 1985). With such private interest governments, opposing interests
are not represented by independent organizations, but are internalized within regulatory
regimes that subject the activities of private agents to a self-imposed discipline. Minimum
quality standards imposed collectively on an industry by producers, a situation common
to agricultural industries, are often achieved by self-regulation. Another area in which
self-regulation is applied refers to situations in which the industry limits or controls its
activities. For example, companies producing breakfast cereals, snacks or beverages
increasingly take measures to address obesity, such as issuing mission and/or educational
commitment statements, practising nutrition labelling, adopting a marketing code of con-
duct, and offering education initiatives aimed at professionals (Jensen and Ronit 2015).
All these different patterns of non-hierarchical governance in policy networks are based
on negotiations between the involved actors. Decisions are not imposed in a top-down man-
ner, but are negotiated. It is this importance of negotiations for the development and imple-
mentation of public policy that characterizes this mode of governance and hence constitutes
its distinctive feature relative to ‘classic’ approaches to policy-making (Kooiman et al. 2008:
2). Against this background we can understand why many scholars restrict the term ‘gover-
nance’ in order to bring out this distinctive feature – implying that we are confronted with
two different meanings of the term, as the discussion has already highlighted.
Network governance emphasizes mutual trust and complementarity of resources that are
expected to result in reciprocal actions. The fact that networks are both relatively loosely
coordinated and stable brings about the exchange of resources that cannot be reflected in
prices, i.e. in intangible goods. These resources can be tacit knowledge or competencies that
only emerge when the resources of different actors are shared. If conflicts between network
Copyright © 2020. Macmillan Education UK. All rights reserved.

participants arise, these are resolved by discussion and negotiations rather than forcing the
affected actors to leave the network.
In this context, the perspective of polycentric governance that is closely associated with the
Institutional Analysis and Development Framework (Ostrom 2011) is instructive. According
to polycentric governance, groups acting at relatively small scales such as the local level can
successfully organize collective action and therefore pool resources in a sustainable manner
(Ostrom 2010a; McGinnis and Ostrom 2012; see also Abbott 2012; Jordan et al. 2015a;
Jordan et al. 2018). The reason for this success is that, at this level, the participants interact
directly and repeatedly with each other and build trust through reputational rewards and
sanctions as well as reciprocity. Furthermore, local communities develop a specific kind of
knowledge and are prone to policy learning (Ostrom 2010b).
The disadvantage of network governance is the emergence of a distinction between
‘insiders’ (those actors actually incorporated into and consulted in decision-making and
implementation) and ‘outsiders’ (those interests to which access is denied). Those who are
members of the network benefit from having insider status, though this comes at the price
of systematically excluding other actors (Lütz 2010: 139). This has important implications
for fundamental principles of democratic politics that stress the importance of inclusion

for policy-making. In extreme cases, network governance might become ‘undemocratic’ and
produce policy outputs that, for society as a whole, are less desirable than those attained by
hierarchical or market governance. Bäckstrand (2008), for example, discusses in detail the
accountability of transnational climate governance by networks.

Changes in Governance Modes: From Hierarchies to Markets to Networks?

An important topic within the governance debate centres on the question of whether there
has been a shift in governance modes from hierarchy towards non-hierarchical approaches
to political steering. So far, however, there is a lack of empirical studies that would allow a
systematic answer to this question. Against this background, it is hard to judge whether the
political and, to some extent, scientific emphasis on new modes of governance is ‘rhetoric or
reality’ (Holzinger et al. 2006). It should also be emphasized that changes in modes or the
scientific discovery of new modes of governance might to a lesser extent be driven by actual
empirical changes, and more by analytical innovations, meaning that the same empirical
phenomenon is being seen through different lenses. So it is very difficult to disentangle
empirical and analytical developments in the governance debate.
Notwithstanding these limitations, some general developments can be identified. The
first pattern refers to attempts at increasing and improving hierarchical governance capaci-
ties during the 1960s and 1970s. For a long time, governance corresponded to government,
i.e. the hierarchical steering of society by state interventions. With the emergence of unex-
pected societal conflicts and economic problems in the 1960s and 1970s, the interventionist
state came under attack. The main argument advanced by political scientists and politicians
was that the interventionist state based on hierarchical governance is ill equipped to resolve
the newly emerging social and economic problems of modern states. These developments
strongly challenged the trust placed in the ability of governments to steer and plan society –
an aspect that became increasingly apparent with the detection of far-reaching deficits in
implementation. Consequently, top-down policy-making based on command-and-control
instruments and coercive techniques of policy implementation was perceived as old-
fashioned and ineffective for increasing public welfare.
As a first consequence, the wish to reduce the hierarchical components in policy led to a
Copyright © 2020. Macmillan Education UK. All rights reserved.

new relationship between political and administrative actors. While hierarchical governance
assigns to the state bureaucracy the sole function of implementing public policy, criticism of
the interventionist state favoured the evolution of a new approach that regarded political
and administrative actors as interrelated in one system, i.e. the political administrative sys-
tem (Jann and Wegrich 2010: 178). The reforms were summarized under the heading of the
‘active state’, since reformist efforts departed from the old bureaucratic styles of merely
administering the public sector and found a more proactive style of encompassing policy
planning and engineering social reforms (Reichard 2003: 347; see also Seibel 2001).
This pattern of extended hierarchical governance was increasingly challenged by implemen-
tation and evaluation research that had revealed far-reaching deficits with regard to the effec-
tiveness of this technocratic approach (Taylor and Balloch 2005: 4). At the same time, the
takeover by neo-liberal parties in many Western democracies – in particular the United
Kingdom and the United States – implied that the encompassing involvement of government
was increasingly being questioned. This coincided with activities of market liberalization and
privatization of state monopolies at the international and EU level. These developments
favoured extensive reforms, determining that the state should delegate tasks that it could not
fulfil in a satisfactory manner to private actors, including industry and societal organizations.

Thus, since the 1980s, we observe a second pattern of changes in governance modes that
is characterized by political attempts to reduce hierarchical governance and instead to rely
on market coordination. This implies that governments no longer provided public goods and
services themselves, but reduced their role to establishing regulatory frameworks for the
provision of these goods and services by private actors. The ‘interventionist state’ has been
increasingly replaced by the ‘regulatory state’ (Majone 1994).
The regulatory state thesis postulates that the provision of common goods has basically
been left to the market, with the role of the state being confined to defining the rules and
incentive structures of private actors in such a way that socially desirable outcomes are
achieved. This is well illustrated by the global wave of privatization in areas such as the
postal service, telecommunications, railways and many other common goods previously
provided and managed by the state. Often, privatization went hand in hand with the cre-
ation of regulatory agencies. This ‘rolling back’ of the state, as a result of deregulation, priva-
tization and administrative reforms, has partly changed the functional role of interventionist
policies: it no longer provides common goods, but rather enables them to be provided.
At the administrative level, the growing influence of market mechanisms led to the rise
of New Public Management, which governments since the 1980s have used to modernize
the public sector. In a nutshell, New Public Management assigns to elected officials only the
role of defining long-term policy goals; apart from that they are expected to delegate compe-
tencies to operative agencies and institutions. Further, New Public Management regards the
public–private dichotomy as obsolete. Consequently, the idea of competition can also be
applied to the delivery of public services, which is theorized as making the entire process
more efficient. Introducing competition also provides benchmarks and other bases for com-
parison, which is at odds with the classic understanding of public administration. This is also
underlined by a preference for output control (e.g. definition of policy or programme tar-
gets) over input control (e.g. the selection of personnel; Peters and Pierre 1998: 227–31; see
also Hood 1991; Pollitt and Bouckaert 2000; Norman 2003), which amongst other things
resulted in a fragmentation of policy delivery (see Box 7.1).
It is questionable though if and to what extent these developments entail fundamental or
gradual shifts in governance modes. On the one hand, it is well acknowledged that privatiza-
tion, deregulation and the shift to New Public Management were not similarly pronounced
Copyright © 2020. Macmillan Education UK. All rights reserved.

across countries. For example, Wilkins et al. (2019) show that New Public Management reforms
have influenced school governance differently in Australia, England, Spain and Switzerland. In
the latter two countries, the reforms resulted in limited decentralisation of governance struc-
tures as compared to the first two due to path-dependent processes. On the other hand, it
should not be overlooked that – regardless of the country in question – the public provision of
common goods and the dominance of interventionist governance havehardly been completely
reversed. The emphasis on markets basically entailed a shift rather than a reduction in hierar-
chical intervention. Hierarchical provision has been replaced by hierarchical market regulation.
A third pattern in the discussion about changes in modes of governance refers to the
emphasis on policy networks that can be observed from, especially, the 1990s onwards. This
discussion developed independently of the emphasis on market mechanisms, though it was
primarily driven by the acknowledgement of increasing needs for cooperation between pub-
lic and private actors. The emergence and growth of policy networks as a new form of gover-
nance are interpreted as the result of two developments. On the one hand, policy networks
emerge from a loss of internal sovereignty of the state, arising from growing organizational
complexity and societal differentiation and fragmentation. Consequently, governments can
no longer effectively intervene in societal processes, but need to rely on cooperation with

societal actors. On the other hand, the growing integration of nation states into interna-
tional institutions and globally integrated markets leads to a loss of external sovereignty.
National governments need to coordinate their activities with a broad range of public and
private actors operating at the domestic and international levels.
As a consequence, it is argued that over time a shift away from hierarchical intervention
towards more cooperative forms of interaction between public and private actors has
occurred (Héritier 2002a: 3). The growth and emergence of these patterns are generally
referred to as ‘modern governance’ (Kooiman 1993) or ‘governance in the modern state’
(Mayntz 2010). An even newer concept is that of ‘public value governance’, in which the
government is given the role of the guarantor of democratic values (Stoker 2006; Bryson
et al. 2014). Regardless of the term used, recent empirical analyses stress that the ‘public’
has returned to governance arrangements, especially in global governance, and affects how
issues as different as climate change and security provision are dealt with (see Best and
Gheciu 2014).
It should, however, be noted that modes of governance co-exist and that different policy chal-
lenges might require the combination of different modes of governance. This point is well made by
the encompassing volume on governance success edited by Bovens et al. (2001a). In the introduc-
tion, the editors underscore the point that a government coping with industrial decline must apply
a different governance mode than a government that, for example, wants to achieve success in the
reform of health care, the liberalization of financial markets or the containment of the AIDS
epidemic.
More generally, notwithstanding the fact that public and private contributions are equally
relevant in more cooperative forms of governance, it is important to emphasize that the over-
all responsibility for providing common pool goods still lies with the state (Börzel and Risse
2010). The state plays a central and active role and disposes of powers and resources which are
not available to societal actors. In particular, governments may provide important incentives
(the state may offer financial support, or delegate power, or it may refrain from direct and
potentially less effective state intervention) in order to stimulate and increase the integration
and organization of societal interests (Knill and Lehmkuhl 1998). Ultimately, it is also the
state and its public entities that monitor the delivery of policies and services.
Finally, it is questionable if and to what extent policy networks are really a new develop-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ment, or only indicate that political steering is analysed through a distinctive analytical per-
spective that explicitly takes account of the fact that public and private actors cooperate in the
formulation and implementation of public policies. The emphasis of governance without gov-
ernment and the role of policy networks might stem from analytical as well as real-world
changes. In other words, policy networks might be a new scientific paradigm as much as a new
empirical development.

Key points
•• Governance has different meanings in the literature; here, we conceive of it as the purposeful
collective coordination of individual action to achieve policy goals.
•• The literature has identified three main governance modes, namely hierarchy, markets and net-
works. These modes of governance are present in each polity, but their dominance might vary
across policy areas and change over time.
•• The extent to which hierarchy has been replaced by markets and policy networks as dominant
governance modes is difficult to assess empirically. This can be traced to the fact that market
governance need not mean a reduction in hierarchical governance.

FOUR IDEAL TYPES OF GOVERNANCE

We have seen that the distinction between different modes of governance might to some
extent be problematic. This holds true in particular for the distinction between markets
and hierarchies. The reliance on market governance need not mean a reduction of hierar-
chical governance, but may even depend to some extent on hierarchical intervention. The
reliance on market governance might lead to a shift in patterns of hierarchical gover-
nance; i.e. from the state as the provider to the state as the regulator of public goods. At
the same time, policy networks are often used as a catch-all category emphasizing that
public and private actors cooperate in the policy-making process. This, however, would be
to overlook the fact that this cooperation might take rather different forms, with public
and private actors playing rather different roles. So far we have only limited insights into
the causes of changes in the reliance on different governance modes across countries and
policy sectors.
In an attempt to address these problems, Knill and Lenschow (2003) identify four ideal
types of governance (see also Knill and Lehmkuhl 2002a, 2002b; Knill 2004, 2008). This
typology is based on two analytical distinctions. On the one hand, the focus is on the degree
of cooperation between public and private actors in the policy-making process; on the other
hand, a basic distinction is made between hierarchical and non-hierarchical modes. This dis-
tinction is assessed through the degree of legal obligation that characterizes collective policy
solutions.
Depending on the specific configuration of these two dimensions, four basic governance
types can be distinguished:

• Interventionist governance
• Regulated self-governance
• Cooperative governance
• Private self-governance

The first pattern of interventionist governance corresponds to hierarchy or government and

mechanisms through the hierarchical definition of regulatory frameworks governing mar-

ket interactions. The three remaining ideal types cover different forms of governance in
policy networks. While in reality one might often observe a more variegated picture than is
offered by ideal types, they still have the virtue of providing a standard against which real-
world systems can be compared and potential differences explained (see Table 9.2).

Table 9.2 Four ideal types of governance.

Cooperation of public and private actors

High Low
Degree of legal obligation High Regulated Interventionist governance
self-governance (government)
Low Cooperative governance Private self-governance

Source: Knill and Lehmkuhl (2002b: 49).

Interventionist Governance
Interventionist governance refers to constellations which reflect the classical scenario of
policy-making, namely the limited governance capacity of private actors in view of the
underlying incentive structure, which can only be compensated for by external power; i.e. it
requires the hierarchical intervention of the state. Although this scenario does not exclude
the involvement of private actors, the overall responsibility for the provision of public goods
lies with the state. In general, this pattern is characterized by a hierarchical relationship
between public and private actors, with the state intervening ‘from above’ in society through
highly detailed and legally binding requirements, i.e. on the basis of clearly defined rules and
regulations which have to be complied with by the public and private actors concerned
(‘command and control’).

Regulated Self-Governance
Regulated self-governance refers to constellations in which hierarchical intervention
through legally binding rules is accompanied by more cooperative relationships between
public and private actors during the formulation and implementation of public policies.
Regulated self-governance implies that the participation of society takes place on the basis
of clearly formalized and institutionalized procedures, although the state still plays a domi-
nant role in the final decision on policy contents and regulatory arrangements. In this con-
text, the relationship between public and private actors might be arranged in various ways:
private actors might participate in policy-making and implementation; competencies might
be delegated to private organizations; or regulatory frameworks for private self-governance
might be cooperatively developed. Often, regulated self-governance can only be achieved
under the ‘shadow of hierarchy’ (Mayntz and Scharpf 1995; Börzel and Risse 2010), the
state being capable of relying on traditional forms of intervention should there be gover-
nance failures (Peters 1998).

The remaining governance patterns – cooperative governance and private self-governance –

crucially differ from the previous patterns with respect to two factors: the voluntary charac-
ter of policy-making, and the fact that private actors rather than the state play a dominant
role in policy formulation and implementation. In the case of cooperative governance, the
definition and application of instruments do not occur on the basis of legally binding
requirements, but through negotiations and voluntary agreements between public and pri-
vate actors. Policies are the result of bargaining processes, in which both public and private
actors participate on an equal footing. While regulated self-governance still means that gov-
ernments hierarchically define the rules of the game for private governance, cooperative
governance means that public and private actors cooperate in the development of such
rules.
Instead of hierarchical intervention through legally binding instruments, the focus is on
cooperation between state and society. On the one hand, it is the objective of these arrange-
ments to allow for the negotiation of cooperative arrangements by including a broad range
of public and private actors. On the other hand, a second feature of this pattern of ‘joint

policy-making’ is the objective of replacing hierarchical intervention by voluntary agree-

ments between public and private actors, such as industry associations.

Private Self-Governance
While in cooperative governance, the definition and implementation of public policies are
based on close cooperation between state and society, these tasks are completely in the
hands of private actors in private self-governance. Similar to cooperative governance, gover-
nance here is based on voluntary rather than legally binding instruments. An example is the
declaration of the car industry in many countries to reduce car exhaust emissions to a cer-
tain level within a given period of time.
In this scenario, the provision of public goods basically depends on the governance
capacity of private actors. Nevertheless, in such constellations states might still play a
role in providing complementary governance contributions, hence ‘refining’ and guid-
ing societal self-governance (Börzel and Risse 2010). For instance, public actors can
increase the legitimacy of private governance by officially acknowledging the outcomes
of private governance (Ronit and Schneider 1999; Lehmkuhl 2000) or by mediating and
moderating between conflicting interests, stimulating communication and coordina-
tion between different actors (Willke 1995). Finally, the activities of private actors
might restrict the role of public actors (Knill and Lehmkuhl 1998; Knill 2001) and vice
versa (Mills 2016).
In conclusion, these considerations indicate that governance patterns might vary strongly
across countries and policy sectors, depending on the level of legal obligation present in
political steering activities, as well as the degree of cooperation between public and private
actors in policy-making. Governance by public or private actors should not be seen as exclu-
sive alternatives, but as mutually reinforcing.

hich Model for Which Season? The Relevance of Institutional and Political

It would be wrong to assume that there is a free choice between the different governance
models identified above, regardless of the particular context in which a certain political
problem has to be addressed. Governance models cannot be understood in the light of
mere ideological orientations or political preferences of the governing party, but against
the backdrop of specific institutional and political structures, which might vary across
countries and policy sectors and so determine the governance capacity of public and pri-
vate actors, i.e. their formal and factual ability to shape the formulation and implementa-
tion of public policies. By affecting the cost–benefit calculations of the actors involved and
by defining a certain distribution of powers and resources between them, the existing
institutional structures have an important impact on the capacity for governance by pub-
lic or private actors.
The basic factor affecting the governance capacity of national governments is the struc-
tural potential for policy adjustments that aim at coping with new problems. For instance,
economic and technological challenges may imply that common goods can no longer be pro-
vided if the existing policy arrangements are relied upon. Rather, fundamental policy adjust-
ments at the national level might be necessary. Hence, the governance capacity of national

governments can be expected to increase with the structural potential for such adjustments.
The reform capacity also may vary from country to country and from policy to policy.
The potential for policy-making depends on the particular institutional arrange-
ments characterizing a country’s legal, administrative and political system (Knill 1999).
It decreases with the number of formal and factual institutional veto points that affect
the opportunities for national governments to initiate and push through institutional
reforms against political and societal resistance. Although the level of reform capacity
does not make it possible to predict the timing or the concrete content and direction of
policy reforms, it does indicate the structural potential of national governments to
maintain their governance capacity by adjusting policy arrangements in the light of the
challenges emerging from economic internationalization. These considerations suggest
that, the more governments need to rely on broad political support when developing
their policies, the more governance models which incorporate private actors in the for-
mulation and implementation of public policies (regulated self-governance, cooperative
governance or even private self-governance) constitute appropriate alternatives.
However, as for public actors, the governance capacity of private actors cannot simply
be taken for granted. Rather, the participation of private actors in governance itself is
crucially dependent on certain institutional preconditions. The most important factor
refers to their organizational structures; namely, that the level of private governance
capacity will increase with both the strength and the degree of organization of private
actors (see, e.g., Tosun et al. 2016).
Organizational strength defines the extent to which organizations are able to influ-
ence, monitor and sanction the behaviour of their members, i.e. the extent to which
organizations have sufficient autonomy to make decisions on behalf of their members
and are capable of ensuring their compliance with these decisions (see Box 9.2). The
level of organizational strength is generally expected to increase with certain organiza-
tional properties such as centralization and the degree of organization within a specific
domain (Streeck and Schmitter 1981).
The degree of organization refers to the extent to which private actors are organized or
willing to contribute to the provision of common goods by private organizations. As shown
in Chapter 3, the size of the group and the extent to which organizations might offer ‘selec-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tive incentives’ for cooperation might play an important role in this context. The degree of
organization may have important repercussions on the resources of the actors involved,
including financial, personnel and technological capacities as well as scientific expertise.
Examples of effective private governance reveal that – particularly with respect to com-
plex technological problems – private actors must, surely, have more appropriate resources
for developing corresponding solutions than do bureaucracies (Knill and Lehmkuhl 1998;
Cutler 1999; Donahue and Zeckhauser 2011; Mills 2016; Tosun et al. 2016). The successful
incorporation of private actors in policy-making therefore cannot be assumed to constitute
a feasible governance mode without prior consideration of the organizational structures in
which these actors operate. Private self-governance or cooperative governance will not work
if the involved associations have no representative monopoly or are not able to ensure the
rule-following behaviour of their members when it comes to the implementation of policy
decisions. In other words, private governance is of little help if societal structures are weakly
developed.
In summary, determining the extent to which public and private actors will contribute to
the solution of policy problems requires a detailed analysis of the particular context that

Box 9.2 Governance in states with limited statehood

The governance concept lends itself particularly well to connecting policy analysis with
comparative politics and conflict research. More precisely, the modes of governance
and their performance in so-called fragile states have received increased attention, for
example in the context of the Sustainable Development Goals adopted in 2015 (see,
e.g., Persson et al. 2016; Tosun and Leininger 2017). According to the Fund for Peace,
the fragility of states can be assessed on the basis of cohesion, economic, political,
social and cross-cutting indicators (Messner 2018). A country that scores highly on
the fragility score is Afghanistan, with particularly high institutional fragility scores
for the functioning of the security apparatus and public services as well as state legiti-
macy, which is aggravated by the influence of external actors in the functioning of the
state, demographic pressure and the displacement of large communities. In such coun-
tries with dysfunctional state institutions, governance is still possible, mostly at the
local level where groups have developed trust and reciprocity (Börzel and Risse 2016).
In the case of Afghanistan, both national and international non-governmental organi-
zations (NGOs) are actively involved in governance, mostly in the field of education
(see Mitchell 2017). The NGOs form networks with foreign donors (e.g. the United
States Agency for International Development) and public entities at the local level in
order to provide schooling to children.

characterizes the strategic constellation underlying the provision of a certain common good.
As Box 9.2 shows, an analysis of governance modes and their expected and/or actual out-
comes benefits from examining the respective institutional setup and how well the existing
institutions function.

•• The combination of the degree of hierarchy and the extent of cooperation between public and
private actors produces four ideal types of governance: interventionist governance, regulated self-
governance, cooperative governance and private self-governance.
•• Interventionist governance refers to the classic understanding of policy-making through gov-
ernment, which grants the state a dominant role vis-à-vis private actors.
•• The state is still dominant in situations of regulated self-governance, but private actors partici-
pate in policy-making in a formalized and institutionalized manner.
•• Cooperative governance involves voluntary agreements between public and private actors.
•• With private self-governance, the provision of public goods completely depends on the govern-
ance capacity of private actors; the state only provides guidance in this process.
• • The appropriateness of these governance types for policy-making depends on the insti-
tutional and political context. In the case of the opposite scenario (weak government,
strong society), private self-governance or cooperative governance is the more viable
alternative.

WHEN IS GOVERNANCE GOOD?

While the institutional and political context affects the appropriateness of governance pat-
terns in the light of different problem constellations, they are not the only aspects that inform
the choice of governance arrangements. In addition, the decision as to whether a certain mode
of regulation reflects ‘good governance’ or not requires clarity about the normative bench-
marks of such an evaluation (Baldwin et al. 2012: 76). When considering the current political
and scientific discussion on ‘good governance’, it becomes obvious that these criteria are
debatable. The main conflict deals with the question of whether the input or the output legiti-
macy of public policy should serve as the primary evaluation criterion (Scharpf 1999).
With respect to output factors, particular attention is paid to two aspects: (1) the extent
to which a political system has the capacity of taking political decisions in a certain area; and
(2) the extent to which these decisions are actually implemented and complied with. With
respect to input legitimacy, the focus is primarily on the democratic quality of the gover-
nance process (for a detailed discussion, see Knill 2004; Knill and Lenschow 2005).

Decision-Making Capacity
A necessary, albeit not sufficient, condition for effective governance is the capability of gov-
ernments to take a policy decision. As already mentioned in the previous section, this capac-
ity cannot be taken for granted, but varies across different political systems, depending on
the specific institutional rules characterizing the decision-making process.
However, the capacity of governments to take political decisions is not only affected by
the institutional configuration in which they operate, but also by the underlying mode of
governance. In other words, there exists a close linkage between this capacity and the respec-
tive levels of discretion and obligation implied by different modes of governance. As a gen-
eral rule, one can expect the decision-making capacity of political systems to increase the
more that powers and discretion are delegated to lower institutional levels (decentraliza-
tion) and to private actors. The more the precise specification of policies is left to other
actors or institutional levels, the less difficult it is to adopt these policies at the central level.
Copyright © 2020. Macmillan Education UK. All rights reserved.

The more policy discretion is delegated to potential veto players or policy opponents, the
less they will mobilize against the policy in question.
Consequently, the ability to reach decisions can be expected to be lowest when gover-
nance follows the pattern of hierarchical intervention, with prescribed requirements that
are detailed and legally binding and affect the interests of many public and private actors
who might resist the adoption of corresponding decisions. By contrast, a consensus between
conflicting interests and the adoption of a certain policy will be less cumbersome and time-
consuming when the degree of detailed governmental intervention into society ‘from above’
is lower. For instance, if the actors addressed by a certain decision have no obligation to
comply with it, there is a relatively low probability of political resistance.

Implementation Effectiveness
Good governance not only depends on legislative decisions, but also on the extent to which
these decisions are actually implemented and complied with. Generally, implementing agen-
cies or private actors have to take the necessary steps to fulfil the objectives spelled out in

legislation or underlying agreements in both formal and practical terms. To what extent do
governance modes differ with respect to these aspects?
A specific advantage of hierarchical intervention lies in the fact that precise and obliga-
tory rules have a higher potential for effective implementation, as the force of law can be
used to impose fixed standards or objectives (Baldwin et al. 2012: 35). In the absence of
legally binding requirements, by contrast, this ‘push factor’ is lacking and compliance rests
solely on the ‘goodwill’ of the implementers. Hence, from this perspective, hierarchical
intervention and regulated self-governance achieve a higher ranking than private self-
governance and cooperative governance.
However, hierarchical ‘push’ is not the only factor affecting the implementation of policies.
Of similar importance are ‘pull factors’, i.e. aspects which influence the willingness of imple-
menting bodies and policy addressees to comply with regulatory rules. It has been argued that
governance patterns that are responsive to the motivations and interests of implementers and
the target actors contribute to implementing the regulation in question effectively. Analysing
the different modes of governance from this perspective, we find that such patterns are par-
ticularly relevant with respect to regulated self-governance and co-governance. At first sight,
private self-governance seems to rely positively on pull factors, as the incentive to escape top-
down policies causes private regulators to formulate and comply with their own rules. This
incentive, though, depends on the presence of a coercive threat. If it is perceived as weak,
industry may respond to the opposite incentive to cheat. Private actors might implement rules
in a rather light-handed way, as the threat of enforcement or later top-down intervention, in a
case of self-governance failure, is low (Baldwin et al. 2012: 58).
These considerations suggest that implementation effectiveness will be highest in cases
of regulated self-governance. Here, effective compliance is not only driven by hierarchical
‘push’, but also by societal ‘pull’. For all other governance modes, at least one factor is miss-
ing. Most problematic in this respect is the pattern of private self-governance, at least in
those constellations in which governmental threats of hierarchical intervention in the case
of weak compliance are not feasible or credible.

With regard to democratic legitimacy, the focus is generally on questions of due process and
the accountability of governance patterns. ‘Due process’ relates to the decision-making and
implementation phase and claims public support on the basis of equal and wide participa-
tory rights granted to those affected by policy decisions.
The issue of participatory rights refers to the scope of participation in the formulation of
public policy by those affected or involved in the implementation. At first glance, it seems to
be quite obvious that private self-governance, co-governance as well as regulated self-
governance constitute better alternatives than hierarchical intervention with regard to this
criterion. These approaches all allow for the incorporation and participation of private
actors in the formulation and implementation of public policies.
A closer look, however, reveals that this initial evaluation overlooks the problem of
unequal access opportunities. On the one hand, and as already emphasized by Olson (1965),
not every societal interest group is equally powerful. Hence, the opportunities to exert polit-
ical influence might vary, with economic interests being typically represented much more
strongly than ‘public’ interests, such as environmental or consumer protection. Participation
therefore need not automatically coincide with equality of representation.

In addition, participatory patterns of policy-making usually imply a distinction

between ‘insiders’ and ‘outsiders’. This problem is particularly severe, for instance, in
arrangements of regulated self-governance. While certain societal actors are granted a
representational monopoly, other interests are excluded. This problem is even more
pronounced in private self-governance. In this case, there is a high possibility that regu-
lation will only take into account the interests of the actors directly affected (e.g. the
industry in a certain sector), while the input of the general public (consumers, residents
etc.) remains very limited. The question of bringing outsiders into policy-making pro-
cesses has also gained prominence in the literature on democracy promotion in aid-
dependent states. As Leininger (2010), for example, shows for the case of Haiti and
Mali, the networks that foreign donors and local domestic actors form actually under-
mine the goal of democracy promotion, since they develop specific institutions that
compete with the existing state institutions.
In terms of equal participation and access, neither regulated self-governance nor co-
governance nor private self-governance necessarily constitutes a better alternative to hier-
archical intervention. If and to what extent this is the case is crucially dependent on the
specific rules regulating the access of interested actors to policy-making.
With respect to the second dimension of due process – substantive equality – decen-
tralization and legal discretion (as inherent in private self-governance and co-governance)
may turn out to be a problem. Discretion opens up room for the unequal treatment of the
target groups, inconsistencies within or between policies and distortions of the market
due to different local policy patterns. The insistence on substantive equality is not uncon-
troversial, however, as it may go hand in hand with a great insensitivity to the circum-
stances experienced by those affected by policy decisions. Uniform rules do not take
account of different administrative structures, established technologies or the salience of
problems at subnational levels. Hence, we need to distinguish between policy content that
requires uniform application due to the presence of certain problem types and policy con-
tent that can be achieved flexibly.
In other words, while decentralization is likely to impact positively on the level of access
and participation, discretion does not contribute to substantive equality. The relative impor-
tance of this latter criterion varies with the policy intent. Second, the choice of steering
Copyright © 2020. Macmillan Education UK. All rights reserved.

mechanisms also affects due process. The hierarchical model contributes to substantive
equality, while treating the openness of procedures and involvement of stakeholders as sec-
ondary; the more cooperative models emphasize the openness of the process more than
equal outcomes.
Policy-makers may claim democratic legitimacy, even in cases where the public has not
been involved in the rule-making process, if they have, nevertheless, the opportunity of
exercising public control over the policy authority (accountability). Both parliamentary and
direct public control are enhanced by decentralizing policy tasks to regional or local public
authorities or private actors through the use of discretionary instruments. The fact that in
practice this may result in some confusion as to the placing of responsibility does not negate
their potential to bring about a higher level of control, especially in a multilevel system. On
the other hand, self-governance is clearly most problematic in terms of public control.
Besides the remote sanctioning powers of public authorities, self-governance systems risk
being captured by groups who are not representative of the general public – or even those
affected by the policy – and becoming isolated from public oversight.

Key points
•• Whether governance is ‘good’ or not can be evaluated on the basis of three dimensions: a politi-
cal system’s decision-making capacity, the implementation of its decisions and its democratic
legitimacy.
•• Decision-making capacity might be expected to be low in the case of hierarchical intervention,
but to increase with more cooperative forms of governance.
•• Effective policy implementation is a function of hierarchical ‘push’ and societal ‘pull’.
Based on this reasoning, effective implementation is most likely in the case of regulated
self-governance.
•• Democratic legitimacy centres on the participation in and accountability of policy-making.
With regard to the former, all governance types are problematic to some extent. The
same holds true for accountability, although self-governance systems appear particularly
problematic.

CONCLUSIONS

The governance concept provides a new perspective on policy-making, as it moves

beyond the policy cycle and patterns of how public and private interact. Governance is
at the same time a complex concept that is applied differently in the literature. It is not
only used to classify different modes of political steering (including hierarchy, markets
and networks), but also as a notion representing a particular mode of steering – namely
non-hierarchical patterns – which yields a distinction between government (hierarchy)
and governance.
Based on the degree of cooperation between public and private actors and the degree of
legal obligation characterizing policy outputs, we have further developed the discussion of
different governance modes and identified different ideal types of governance. On this basis,
we further discussed potential factors that account for varying governance patterns across
countries and sectors. We have also shown that the classification of governance as ‘good’ or
Copyright © 2020. Macmillan Education UK. All rights reserved.

‘bad’ strongly depends on the normative benchmark that is applied. In this regard, we dif-
ferentiated between decision-making capacity, implementation effectiveness and demo-
cratic legitimacy.
The governance perspective is suitable for investigating the initial policy stages that con-
sist of problem definition and agenda-setting, as it might indicate whether the perspectives
of all participants in the policy process have the same chances of being acknowledged during
policy-making. Additionally, utilizing the governance concept also provides a clearer under-
standing of policy formulation and adoption. As we have illustrated, the general scope of
participation and the extent to which policy-making is characterized by the creation of
‘insiders’ and ‘outsiders’ have far-reaching consequences for policy outputs. In this context,
a central finding we made in Chapter 7 was that a poorly designed policy that lacks broad
public acceptance is likely to fall short of its goals. It is also likely that the findings of evalu-
ation studies are one source for changing forms of governance.
The evaluation of different governance modes shows that what is seen as good gover-
nance strongly varies with the respective evaluation criterion that is applied. Rather,

effective and legitimate governance is crucially dependent upon the involvement of the
state. It is this latter perspective that makes the governance concept attractive for combin-
ing it with theories in comparative politics and conflict research. We have seen that the
question of governance in fragile states is a prolific area of scholarship that bridges the
divide between policy analysis and comparative politics.

WEB LINKS

www.asrcreviews.org. This is the website of the US Advertising Self-Regulatory

Council, which presents the industry’s target group-specific self-regulatory
systems.
www.fundforpeace.org/fsi/. This website hosts the Fragile States Index, which is
updated annually.
www.gppi.net The Global Public Policy Institute is an independent non-profit think
tank that seeks to improve global governance through research, policy advice and
debate.
www.sgi-network.org. The Sustainable Governance Index by the Bertelsmann
Foundation is based on a cross-national comparative survey designed to identify
successes in effective policy-making. The Index comprises the following three
dimensions: policy performance, democracy and governance.
www.wcfia.harvard.edu/research_activities/harvard_mit_private_governance/
overview. The Harvard–MIT Private Governance Working Group is a forum for
scholars to discuss research on private governance. It hosts some interesting
publications and might be consulted for learning about relevant conferences and
other events.
info.worldbank.org/governance/wgi/. This website hosts aggregate and individual
governance indicators for over 200 countries and territories over the period
1996–2018.
Copyright © 2020. Macmillan Education UK. All rights reserved.

FURTHER READING

Börzel, T. A. and van Hüllen, V. (eds.) (2015). Governance Transfer by Regional
Organizations. London: Palgrave Macmillan. A stimulating collection on the
role of regional organizations for changes in their member states’ governance
arrangements.
Eliadis, P., Hill, M. and Howlett, M. (eds.) (2005). Designing Government: From
Instruments to Governance. Montreal: McGill-Queens University Press. The
strength of this book is that it systematically illuminates the relationship
between policy instruments and governance, constituting an innovative approach
to the study of governance.

Esguerra, A., Helmerich, N. and Risse, T. (eds.) (2016). Sustainability Politics and
Limited Statehood: Contesting the New Modes of Governance. Cham: Springer. A
timely perspective on how to attain sustainable development when the state
capacity is weak.
Levi-Faur, D. (ed.) (2012). The Oxford Handbook of Governance. Oxford: Oxford
University Press. This volume covers the main research topics in the study of
governance.
Seabrooke, L. and Henriksen, L. F. (eds.) (2017). Professional Networks in
Transnational Governance. Cambridge: Cambridge University Press. This volume
offers an actor-level analysis of transnational networks and connects transna-
tional governance with the agenda-setting perspective.
Sørensen, E. and Torfing, J. (eds.) (2016). Theories of Democratic Network
Governance. London: Palgrave. An insightful volume that discusses the
dynamics, conditions and functions of network governance and its implica-
tions for democracy.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Reader’s guide
The development of public policies is not restricted to the nation state. Rather,
governments have always cooperated at the international level in order to address com-
mon problems that cannot be effectively addressed by individual countries. In fact, it
was the existence of transboundary effects and externalities that stimulated the cre-
ation of international regimes such as the Basel Convention governing the international
movement of hazardous waste, or the United States–Canada Great Lakes water quality
regime. As a result, there is an increasing number of public policies that reach beyond
the nation state. While interdependencies are certainly one of the most important driv-
ers of this process, social, political, economic and technological changes equally contrib-
ute to it. In light of these developments, it is our central objective in this chapter to
identify analytical factors that influence the formulation and implementation of public
policies beyond the nation state, i.e. international public policies. We will first offer a
general assessment of the rationales for international cooperation and provide an over-
view of the actors and institutions involved in policy-making beyond the nation state.
This step involves combining public policy analysis with the disciplines of international
relations and international political economy. We will then address interest constella-
tions and mechanisms affecting the dynamics of policy formulation at the international
level. Finally, we turn to the peculiarities and challenges that characterize the imple-
mentation of these policies. In this way, we take up once again two important stages of
Copyright © 2020. Macmillan Education UK. All rights reserved.

the policy-making process, namely decision-making and implementation.

INTRODUCTION

Are internationalization and globalization reflected in policy-making? How do policy stud-

ies incorporate factors external to a country’s political system? Since the 1990s there has
been a growing shift in scholarly attention from the national to the international level of
policy-making. Students of public policy, international relations and international political
economy have become more and more interested in the factors affecting the formulation
and implementation of public policies beyond the nation state, as well as potential interac-
tions between policy-making at the international and the domestic levels. In response to the
real-world developments of increasing international cooperation and growing economic and
technological interdependencies between states, as well as the considerable progress in the
process of European integration, scholars of public policy and international relations have

183
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:18:15.
184 Public Policy

begun to focus on theories and patterns of global governance (Brown 2001; Czempiel and
Rosenau 1992; Young 1994; Biermann et al. 2009; Stone 2008; Biermann and Pattberg
2012; Hale and Held 2012; see also Chapter 9).
This shift in scholarly attention has not only been characterized by a partial convergence in
terms of the analytical concepts and subjects of analysis of the different fields of political sci-
ence – i.e. public policy, international relations and international political economy – it has also
been characterized by the rise of the catchword of ‘globalization’, which has been seen as the
crucial driving force behind the development of public policies beyond the state. The problem
with this term, however, is that it still lacks a comprehensive and generally accepted definition.
On the one hand, globalization can be seen as an intensified version of interdependence
between states driven by economic liberalization, which has rendered nation states increas-
ingly sensitive and vulnerable to each other. This vulnerability, in turn, constitutes an impor-
tant driving force for international cooperation, i.e. the establishment of international
organizations and regimes to manage and regulate these interdependencies (Cooper 1968). On
the other hand, globalization can be characterized as a phenomenon that is qualitatively differ-
ent from mere interdependence, and even as being caused by the latter. According to this
understanding, instead of being linked to nation states, globalization refers to increasing cross-
border transactions by private actors. As this book is concerned with public policies, our focus is
on nation states rather than private actors. Hence, we will start from the problem of interde-
pendence between states and consider globalization in terms of the phenomenon of increasing
interdependence (see Braun and Gilardi 2006; Teney et al. 2014). More precisely, we follow the
conceptualization of globalization established by Dreher (2006) that globalization consists of
economic, social and political dimensions (see also Potrafke 2015; Gygli et al. 2018).
The starting point of this chapter therefore concerns the conditions and patterns of
international cooperation. Public policies beyond the nation state refer to the formulation
and implementation of policies via international organizations and regimes. This does not
lead to a neglect of the role of private actors in this process: they are integrated into the
analysis insofar as they participate in the making of public policies through transnational
governance (see, e.g., Hale and Held 2011; Abbott 2012; Kotzian and Steffek 2013; Stone
2013; Eberlein et al. 2014; Auld et al. 2015). Purely private policies, such as codes of conduct
or private arbitration, are not taken into account as they are beyond the scope of this book
Copyright © 2020. Macmillan Education UK. All rights reserved.

(but see, e.g., Berliner and Prakash 2015).

This analytical choice is further justified by the fact that, after the ‘big wave’ of globaliza-
tion research and debates on ‘governance without government’, it is generally acknowledged
that the nation state still matters a lot in the development of policies beyond it, more than
was assumed in the early stages of the global governance debate. It seems fair to say that the
‘state is back in’ in the globalization debate (Héritier 2002a; Holzinger et al. 2008a, 2008b;
Tömmel and Verdun 2009; Risse 2013; Börzel and Risse 2016). In light of these develop-
ments, we will identify analytical factors that are central to influencing the formulation and
implementation of international public policies.

PUBLIC POLICIES BEYOND THE NATION STATE: GENERAL ASSESSMENT

There is a consensus in the literature that globalization has significantly affected the condi-
tions for national policy-making. On the one hand, globalization created a range of prob-
lems that exceed the scope of national sovereignty and can therefore no longer be sufficiently
resolved by the unilateral action of national governments. Examples include the regulation

of electronic commerce and the protection of intellectual property rights over digital infor-
mation. On the other hand, the emergence of globally integrated markets puts pressure on
national governments to redesign national regulations to avoid excessive regulatory bur-
dens being imposed on domestic industries (Knill and Lehmkuhl 2002a, 2002b; Prakash
and Potoski 2006; Holzinger 2008; Rudra 2008; Tosun 2013a; Mukherji 2017). In view of
this constellation, national governments consider cooperation in order to establish interna-
tional regimes and international organizations that would allow them to maintain their
capacity to address problems that extend beyond the parameters of national sovereignty.
International regimes are generally defined as ‘implicit or explicit principles, norms, rules
and decision-making procedures around which actors’ expectations converge in a given area
of international relations’ (Krasner 1983: 2; see also, e.g., Young and Zürn 2006; Hafner-
Burton 2012; Nielson and Simmons 2015; Breitmeier 2016). International regimes can be
subject to differentiation and fragmentation, which can eventually result in a loosely cou-
pled set of specific regimes, which Keohane and Victor (2011) refer to as regime complexes
(see also Alter and Meunier 2009) and Ostrom (2010a, 2010b) as polycentric governance
(see Box 10.1). International organizations, by contrast, refer to stable forms of cooperation
founded on the grounds of an international agreement (Simmons and Martin 2002). Indeed,
the number, relevance and regulative activities of international regimes, regime complexes
and organizations have grown steadily over the past few decades (see Breitmeier et al. 2006;
Reinalda 2009). Issues such as climate change or terrorism can be read as ciphers for the
challenges posed by the globalization of our economies and societies. Collective political

Box 10.1 Polycentric climate governance

Numerous observers have noted that the climate regime has become increasingly
diverse – involving state and non-state actors as well as public–private partnerships –
and fragmented (see, e.g., Bäckstrand 2008; Keohane and Victor 2011; Abbott 2012;
Bulkeley et al. 2012, 2014; Jordan et al. 2015a). This development can be regarded as
problematic, since fragmentation in international regimes can increase transaction
costs and reduce their effectiveness. An alternative perspective is that this develop-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ment has produced a polycentric climate governance regime, which, following the
argument put forward by Ostrom (2010a, 2010b), is likely to produce positive out-
comes. Polycentric governance involves multiple, formally independent decision-
making units that operate at multiple scales. Since climate change is the result of
behaviour at multiple scales, governance must also take into consideration these
scales when designing tools for climate governance. With individuals and organiza-
tions at each scale taking climate-relevant actions, the chances of responding climate
change will increase. It follows from this particular theoretical perspective that uni-
fied international treaties governing climate change, such as the Kyoto Protocol or the
Paris Agreement, need to be flanked by actions at smaller scales. Examples of actions
at smaller scales are transnational city networks such as C40 or Energy Cities. These
city networks develop local-level policy tools to mitigate as well as to adapt to climate
change, and exchange their experience with the respective tools with a view to stimu-
lating policy learning and horizontal diffusion. Further, solutions elaborated at the
local level can diffuse vertically to upper levels (‘upscaling’) and inform policy decision
that national governments make.

responses to such problems require ever higher levels of coordination and preparation on
the part of international organizations (Knill and Bauer 2016). As a consequence, national
policy-makers delegate tasks and competencies to organizations that are in a better position
than national governments to develop policy responses to problems exceeding the scope of
national boundaries. For example, the Organisation for Co-operation and Development
(OECD) and the World Bank have been active in designing and promoting transnational
benchmarks for evaluating the quality of governance (De Francesco 2016).
As a result, the development of international public policies is characterized by the
involvement of a huge and still growing range of different governmental and non-
governmental organizations, whose relevance might vary from sector to sector and issue to
issue. To get a basic understanding of these developments, we first have to learn more about
the reasons driving general trends towards policy-making beyond the nation state. We will
then provide an overview of the various organizations and areas of international public
policy.

Public Policy Beyond the Nation State: Underlying Rationales

Why do national governments delegate powers to international regimes and organiza-
tions? Which factors account for the rise in international organizations and regimes and
their growing differentiation and institutionalization? To explain this development, two
aspects must be taken into account, namely constellations in which the scope of the
underlying problem exceeds the scope of territorially bounded regulatory structures, and
restrictions on national policy options due to economic interdependencies between states.

Incongruence between Transnational Problem Structures and Regulatory Structures

In many instances, the development of public policies beyond the nation state is driven by
the existence of transboundary problems that cannot be effectively addressed by the uni-
lateral action of individual countries. Problems such as climate change have implications
at a global scale and require far-reaching international cooperation (including more or less
all states) for the development of solutions that still have not yet been adopted. In addi-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tion, many transboundary problems exist that can be addressed by the cooperation of
smaller groups of countries, such as water pollution of transboundary rivers, where down-
stream countries might suffer from polluting activities in upstream countries (see Mitchell
2006). An example is the Great Lakes Water Quality Agreement of 1978, concluded by the
governments of Canada and the United States. Problems of transboundary scope, how-
ever, are not restricted to the transboundary effects of environmental pollution, but can
also be observed in many other areas, such as trade (e.g. Hoekman and Kostecki 2010),
global health (e.g. Gostin and Sridhar 2014; Shiffman et al. 2016; Smith and Shiffman
2016; Van de Pas et al. 2017; Sethia and Kumar 2018), finance (e.g. Walter 2013; Frieden
2016), migration (e.g. Sachs 2016), energy (e.g. Goldthau and Witte 2010; Chaban and
Knodt 2015; Knodt et al. 2016; Lesage et al. 2016; Goldthau et al. 2018) or cybersecurity
(e.g. Heinl 2014).
Since transboundary problems might vary in scope, the range of countries whose
cooperation is needed in order to develop policy responses might also differ. While
many problems (like climate change, financial market regulation or telecommunications
standardization) have a worldwide range, other problems can be addressed by setting up

regional regimes and organizations like the European Union (EU) or multilateral coop-
eration between a few countries (e.g. the Chiang Mai Initiative, which is a currency swap
arrangement amongst the Association of Southeast Asian Nations, the People’s Republic
of China, Japan and South Korea).
It is important to emphasize that not every problem triggered by globalization is
necessarily of global scope, and hence not every problem created by globalization
exceeds the regulatory scope of national governments. It might well be that a problem
created by economic and technological interdependence can still be sufficiently resolved
within the territorial boundaries of one nation state, while such solutions are no longer
feasible for other problems. Under certain conditions, even problems of global scope
might be effectively resolved within national boundaries. Such constellations are pos-
sible when the extent to which a good is provided is determined by the largest individual
contribution – that is, by the ‘best shot’ (Hirshleifer 1983; Holzinger 2002). Despite the
fact that the scope of the problem exceeds a single jurisdiction and thus impedes author-
itative rule-making, other actors might accept the contribution of a single actor. An
example of this scenario is the provision of a global system for the administration of
internet addresses and domain names. One nation state (the United States) has resolved
the problem for all other countries by developing an appropriate system (Knill and
Lehmkuhl 2002b).
However, this is only one of the basic options concerning the ways in which individual con-
tributions and the provision of a public good can be linked. In many cases, the level of provi-
sion is based on the sum of individual contributions, e.g. the activities of individual states
addressing the problem of global warming. It is conceivable that the provision of the good is
determined by the smallest individual contributions, i.e. the ‘weakest link’ (Holzinger 2002).
For instance, the control of illegal and harmful content on the internet is de facto deter-
mined by the country with the lowest regulatory standards, given that providers of such
material can move their services across national borders. It is becoming apparent that, in
both of these cases, activities of individual states are no longer sufficient for coping with
problems of global scope. Hence, there is a need for transnational solutions.

The need to develop public policies beyond the nation state emerges not only from the exis-
tence of problems that exceed the scope of territorially bounded regulatory structures, but
also from restrictions on governmental options to address domestic problems. These restric-
tions result from global market integration intensifying economic interdependencies
between states.
With the increasing abolition of national trade barriers, there is the potential that the
international mobility of goods, workers and capital puts pressure on nation states to rede-
sign domestic market regulations to avoid excessive regulatory burdens (Goodman and
Pauly 1993; Keohane and Nye 2000). Consequently, national governments compete over
the optimal design of domestic regulations in order to attract foreign capital and to improve
the competitive position of their economy. On the one hand, the presence of mobile capital
can induce governments to attract capital from elsewhere by lowering regulatory require-
ments (e.g. social or environmental standards); and, on the other, the threat to remove
domestic capital can exert pressure on governments to lower the level of environmental and
labour regulation (Drezner 2001: 57–9).

While much research effort has been dedicated to the almost classic question of the
extent to which regulatory competition induces races to the regulatory top or bottom
between countries (Vogel 1995; Scharpf 1997b; Holzinger 2002; Holzinger and Knill 2008;
Rudra 2008; Tosun 2013a), less emphasis has been placed on the fact that countries might
anticipate and reduce potential competition effects and their induced restriction on the
national set of feasible strategies by international cooperation. In other words, countries
that are strongly integrated economically may seek to avoid problems of collective action by
harmonizing regulatory standards. Hence, they may pre-empt potential effects of regula-
tory competition in order to maintain their ‘steering potential’ (see Knill and Lehmkuhl
2002a, 2002b; Knill et al. 2008b).
A case where these effects are certainly most pronounced is the EU. Although economic
integration has always been a core objective of European integration, the establishment of
the Single Market has been accompanied by enduring efforts to harmonize social, environ-
mental and consumer protection regulation. European integration in these areas at least
partially emerged because of intersectoral spillovers that demanded responses in policy
areas that did not originally belong to the Community’s field of activity. To put it more gen-
erally, developments in international cooperation do not always occur independently from
one another, but rather push and pull each other (Lindberg and Scheingold 1970; Burley and
Mattli 1993; Bang et al. 2007), which applies to cooperation between states, but also to
cooperation between state and non-state actors (see, e.g., Steffek 2013; Box 10.1).

Overview: Organizational Forms and Areas of International Public Policy

The making of public policies beyond the nation state is characterized by the involvement of
a huge number of different organizational forms and touches on a broad range of different
policy sectors. At a general level, these organizational forms can be distinguished along two
criteria: (1) whether they are governmental or non-governmental, and (2) whether they are
of international or national scope.
The role and influence of national governments are crucial for understanding the emer-
gence of policies beyond the nation state. Cooperation between national governments con-
stitutes the precondition of any international public policy. Often, but not always, this
Copyright © 2020. Macmillan Education UK. All rights reserved.

cooperation entails the establishment of international organizations. Similar to national

politics, the making of public policies beyond the nation state, in many instances, is charac-
terized by the involvement and participation of various non-state actors. With regard to
these, a general distinction is made between non-profit organizations representing public
interests, such as environmental protection, development aid or human rights; and organi-
zations representing private-sector interests, such as business associations or individual
companies. Again, these organizations can be both national and international in scope,
including international non-governmental organizations (NGOs) such as Amnesty
International, the International Red Cross, Médecins Sans Frontières or the International
Union for the Conservation of Nature, along with international interest groups, multina-
tional corporations and national organizations that seek to influence the policy position of
their governments in international negotiations. Moreover – and again comparable to
national politics – the making of international public policies can entail the widespread par-
ticipation of scientific experts and advisory boards operating at national and international

levels (see, e.g., Stone 2008; De Francesco 2016). Against this background, Haas (1992) pub-
lished his influential article on epistemic communities, i.e. networks of professionals with
recognized expertise and authority in a particular issue area.
As it is impossible to offer an exhaustive list of organizations and sectors in international
public policy, in this section we take a different point of departure by suggesting some basic
analytical criteria that help to structure this highly complex field. In so doing, our focus is
on three dimensions, namely the form, the geographical scope and the sector of interna-
tional cooperation.

Forms of International Cooperation

As already emphasized, any public policy beyond the nation state requires some form of
international cooperation. Three basic patterns can be distinguished: international organi-
zations, international regimes and coalitions or groupings. The most prominent form of
international cooperation is the establishment of international organizations. These are
established by a treaty that has to be ratified by the member states’ governments to provide
the organization with its own legal personality. The establishment of an international orga-
nization entails the delegation of specific competencies by member states and the establish-
ment of administrative structures and resources to fulfil these tasks (Snidal and Abbott
2000; Simmons and Martin 2002). An international organization that consists of states
only is known as an intergovernmental organization. Sometimes members of intergovern-
mental organizations are themselves an intergovernmental organization in their own right.
For example, the EU – as an intergovernmental organization – as well as the individual EU
member states are members of the World Trade Organization (WTO), which has implica-
tions for policy processes in the EU (see Oberthür and Gehring 2006; Daugbjerg and
Swinbank 2015).
Apart from these general characteristics, however, as we have already pointed out in
Chapter 3, international organizations strongly differ in their functions, competencies and
organizational features (Pevehouse and Warnke 2005; Bauer and Knill 2007; Volgy et al.
2009). While some primarily serve as a neutral forum for debate, others are characterized
by the far-reaching delegation of national policy competencies, with the EU being the most
Copyright © 2020. Macmillan Education UK. All rights reserved.

developed example. While some organizations, like the EU, the OECD or the United
Nations (UN), cover a broad range of different sectors, others are restricted to single issues.
These differences are reflected in varying administrative structures and resources (see
Bauer 2009). The Yearbook of International Organizations counts over 37,500 active and
approximately 38,000 dormant international organizations, including intergovernmental
and international NGOs (Union of International Associations 2018).
International cooperation, however, need not be exclusively based on the establishment
of an international organization, but might be restricted to setting up treaty-based regimes.
Many international treaties have not been set up as international organizations with their
own legal personalities and administrative structures, and instead rely on national adminis-
trations or ad hoc commissions. Some forms of international cooperation such as the Group
of Seven (G7) or Group of 20 (G20) lack even a treaty basis, implying that – in a legal sense –
they exist only as simple groupings or coalitions of states (Pevehouse and Warnke 2005; for
details, see Knaack and Katada 2013).

Geographical Scope

In addition to differences in the precise form and scope of responsibilities (single issue ver-
sus multi-issue) as well as the establishment of an administrative body, international coop-
eration might vary in its geographical scope. Four basic patterns can be empirically observed.
First, international cooperation can be of global scope, implying that membership is generally
open to all nations. Examples include the UN and its specialized agencies, the Universal Postal
Union, Interpol, the WTO and the International Monetary Fund (IMF). Second, cooperation can
be restricted to countries belonging to certain regions of the world. This category includes the
EU, the Council of Europe, the North Atlantic Treaty Organization (NATO), the African Union,
the Organization of American States, the Association of Southeast Asian Nations (ASEAN), the
Arab League and the Organization for Security and Co-operation in Europe (OSCE). Third, inter-
national cooperation might be based on cultural criteria, including member countries that are
characterized by similar linguistic, ethnic, religious or historical traditions. An illustrative exam-
ple is the Nordic Council, which was formed in 1952 and has 87 elected members from Denmark,
Finland, Iceland, Norway and Sweden, as well as from the three autonomous territories of the
Faroe Islands, Greenland and Åland. Fourth, cooperation can be based on functional consider-
ations, such as the Convention for the Protection of the Marine Environment of the North-East
Atlantic of 1972, of which the following countries are members: Belgium, Denmark, Finland,
France, Germany, Iceland, Ireland, Luxembourg, the Netherlands, Norway, Portugal, Spain,
Sweden, Switzerland and the United Kingdom.

Policy Sectors

International cooperation can be found in a broad range of different policy sectors. Some
international organizations like the UN cover a policy spectrum that is more or less compara-
ble to the scope of national governments. In addition to these encompassing organizations, we
find single-issue organizations in many different policy sectors. Basically, international coop-
eration is particularly pronounced in the following areas: security, human rights, social affairs
(humanitarian aid, health care, education), environmental protection, economic regulation
and technical standardization. These areas are characterized by the need to pool resources for
effectively solving existing problems or achieving a reduction in economic costs by means of
Copyright © 2020. Macmillan Education UK. All rights reserved.

harmonization. The second point predominantly holds true for international cooperation on
economic regulation and technical standardization.
So far, we have concentrated on general patterns regarding the rationales as well as the forms,
geographical scope and sectors of international cooperation. While these merely descriptive
accounts offer an initial overview of policy developments beyond the nation state, they still leave
open the important questions of why countries with heterogeneous interests actually agree on
common solutions and if and to what extent international agreements are actually complied
with. These issues will be investigated in the following sections.

Key points
•• The development of public policies beyond the nation state is driven by problems of transna-
tional scope and growing economic interdependence between states.
•• The EU – which has developed into a supranational organization – also started as a forum for
international cooperation (on energy and economic issues).
•• Types of international cooperation vary in their organizational forms, their geographical scope
and the policy sectors they cover.

OLICY FORMULATION: TYPICAL INTEREST CONSTELLATIONS

P
AND INTERACTION
The huge variety in forms and patterns of international cooperation implies that processes
of policy formulation differ widely. This holds true in particular with regard to the relevant
rules used for decision-making and the extent to which member states have delegated
autonomous powers to international organizations. In the case of the EU, the European
Commission constitutes a very powerful player at the stage of policy formulation, given the
Commission’s formal and exclusive right to initiate policy proposals, as well as its attempts
to build informal and close working relationships with national civil servants, policy experts
and interest associations (Egeberg 2006; Trondal 2010). As a result, processes of policy for-
mulation in the EU are not easily comparable to those of other international organizations,
where such autonomous powers are less developed.
Yet we should not conclude from this consideration of the formal autonomy of interna-
tional public administrations that these bodies are not influential in international policy-
making. Rather, the administrative bodies of international organizations have developed
informal routines in order to strengthen their role in the policy process. Such routines may
include, for instance, proactive search strategies in order to identify and frame new policy
problems and to push these policies onto the political agenda. Moreover, some international
bureaucracies have developed strategies of carefully mapping the political space for action
by anticipating member state preferences during the process of policy drafting. We can also
observe administrative routines directed at the mobilization of internal and external sup-
port for certain policies. A mere focus of formal powers of international administrations
thus hardly captures the full potential of bureaucratic influence in international politics.
Still, such informal patterns are not equally pronounced in every international organization
(Knill et al. 2018).
In light of this constellation, the focus of this section is on the more general discussion of
factors that render international cooperation more or less likely. When can we expect inter-
national cooperation to be successful? When are the involved countries able to achieve com-
mon policy solutions? When, by contrast, does the heterogeneity of member state interests
yield a less favourable scenario for international cooperation?
Copyright © 2020. Macmillan Education UK. All rights reserved.

To be sure, one could certainly argue that the joint development of policies at the inter-
national level is generally a highly difficult exercise. This can be related to two aspects that
characterize policy-making beyond the nation state: the potentially high heterogeneity of
national interest positions, and the very demanding quorums (the smallest number of
members needed before official decisions can be made) for adopting joint solutions.
In certain cases, these problems of joint decision-making can to some extent be reduced
by the creation of so-called package deals. The latter require various decisions to be linked
with one another during negotiations; concessions by individual countries in one area are
thus compensated for by the concessions of other states in other areas. A case in point is
Russia’s ratification of the Kyoto Protocol, which was requested by the EU in return for sup-
porting the country’s accession to the WTO (e.g. Henry and Sundstrom 2007). Another pos-
sibility which can sway countries to give up their resistance is (monetary) compensation for
the costs which arise for them during the implementation of planned measures (see Scharpf
1997b). The means of doing so are limited, though, as most international organizations do
not have any specific financial resources from which such compensatory payments can be
made. An example of this strategy is the so-called EU–Turkey Refugee Deal, which is about

returning any irregular migrant who is found to have entered the EU through Turkey with-
out having already undergone a formal asylum application process. In exchange, the EU
agreed to resettle Syrian migrants living in Turkey who had qualified for asylum and reset-
tlement within the EU, as well as to provide financial aid to Turkey (e.g. Baban et al. 2017).
Despite the theoretical claim that international cooperation is difficult to attain, empiri-
cally we observe that it has progressed significantly. Obviously, there are ways to success-
fully overcome the abovementioned difficulties, and three aspects are of particular
significance. First, national interest constellations may be defined to a great extent by the
underlying problem type. This implies that the chances for successful international coopera-
tion may vary from case to case. Cooperation was, for example, easily attained in the case of
the Montreal Protocol on Substances that Deplete the Ozone Layer (e.g. Epstein et al. 2014),
but it is more challenging when international migration is concerned (e.g. Geiger and Pécoud
2010). Second, there is evidence that international cooperation may be facilitated by the
active efforts of individual states at the international level to influence the content and form
of cooperation according to their interests. For example, the EU has replaced the United
States as the promoter of international environmental law (e.g. Kelemen and Vogel 2010).
These interests are particularly associated with minimizing the potential costs of adminis-
trative and institutional adaptation, and at the same time securing the competitiveness of
national industries. Third, it must not be underestimated that the interest constellations
during international negotiations can change as a result of learning processes and the diffu-
sion of ways of perceiving and solving problems between the involved states (e.g. Dunlop
et al. 2018). These processes, which are promoted by institutionalized forms of interaction
between the involved actors, have significant ramifications for the specific form and content
of international policies (Box 10.2).

The Problem Type

In the literature, it is generally argued that the peculiarities of an underlying policy problem
have a significant impact on the politics involved in providing it. Basically, three distinct
constellations are identified: coordination, agreement and defection. For its part, each of
these is characterized by a specific problem in resolving conflicts of interest. Depending on
Copyright © 2020. Macmillan Education UK. All rights reserved.

underlying interest constellations, international cooperation between the involved coun-

tries might be more or less difficult to achieve (Knill and Lehmkuhl 2002a, 2002b).
Coordination problems refer to constellations in which the involved countries have a
strong common interest in the development of joint solutions and in which there is agree-
ment between the countries upon the specific policy solution (Box 10.3). Hence, interna-
tional cooperation presumes communication between the involved countries. In the
absence of strong interest conflicts between the involved negotiation partners, agreements
may be achieved rather swiftly. For each individual country, international cooperation con-
stitutes rational behaviour.
As soon as international cooperation aims at redistribution, however, it becomes more
difficult to achieve an international agreement between states or collective action between
private actors. Generally, such agreement problems are characterized by a common interest
in the development of a common solution and by a pronounced disagreement about the
kind of solution to be selected. In many instances, such constellations can be observed when
it comes to the international harmonization of standards, for example with regard to digital

Box 10.2 Administrative styles of international organizations

In an attempt to conceptualize informal patterns of administrative influence, Knill
et al. (2018) developed a typology of administrative styles. Depending on dominant
strategic orientations shaping administrative routines, they identified four ideal
types: a servant style, an advocacy style, a consolidator style and an entrepreneurial
style.
A servant administration presumes a rather reactive and instrumental role. The
administration refrains from any attempt to intervene with politics beyond its formal
duties and does not try to influence the policy process. In this case, the bureaucracy
follows a routine pattern of operating ‘by the books’, strictly adhering to the formal
procedural and legal arrangements that define its tasks and functions.
Advocates focus their activities on influencing those aspects that are directly related
to the quality, internal consistency and effectiveness of their policies. Such bureaucra-
cies will place less emphasis on behavioural patterns directed at safeguarding or
advancing their institutional position, but advocate for their approach towards the
achievement of substantive policy goals.
Consolidators concentrate primarily on further strengthening their political auton-
omy, status, size and competencies. Their main interest lies with the increase of com-
petencies as such; i.e. the growth of the policy portfolio is given priority over policy
consistency. Consolidators are hence primarily driven by positional rather than policy
interests.
An entrepreneurial style is the combination of the two previous types and thus pre-
supposes administrative routines that entail intensive bureaucratic advocacy in pol-
icy-making, as well as a strong orientation towards institutional consolidation in
order to strengthen the administration’s position.
The bureaucracy of the IMF, for instance, exhibits a style that comes close to the
entrepreneurial ideal type. Whereas the secretariat of the Bank of International
Settlements largely corresponds to the advocatory ideal, its counterpart at the Food
and Agriculture Organization comes close to that of a consolidator. The International
Copyright © 2020. Macmillan Education UK. All rights reserved.

Labour Organization, finally, has a bureaucracy that predominantly pursues a servant

style.

broadcasting. To ensure the compatibility and interconnectivity of their products, producers

are generally interested in common standards. For reasons of economic competitiveness,
however, they might prefer different options; e.g. they may try to promote their own prod-
uct as the ‘solution’ to which other companies would have to adjust.
A typical example that illustrates international agreement problems refers to the harmo-
nization of environmental product standards in integrated markets. Product standards
specify the quality and technical characteristics of a certain product, such as emission stan-
dards for passenger cars or the lead content of fuels. They have to be distinguished from
process standards that typically define restrictions on the use of specific inputs or specify
requirements, technologies or processes for industries. An example is the specification of
emission standards for combustion plants (see Knill et al. 2008a).

Box 10.3 The Universal Postal Union

The Universal Postal Union (UPU) offers an instructive example of the international
coordination of national postal policies. International postal communications were
originally governed by bilateral agreements which answered the particular needs of
each country. This system, involving as it did a great variety of rates calculated in
different currencies and according to different units of weight and different scales,
made it complicated to operate the service and hampered its development. The
invention of steam navigation and the railway brought about a change in the postal
system. Countries began to realize that, if international communications were to
keep pace with the means of transport, formalities would have to be standardized
and reduced. To simplify the complexity of this system, several countries (in par-
ticular the United States and Germany) called for an international solution, which
led to the establishment of the UPU in 1874. In 1964, it was reorganized into the
‘modern’ UPU (see Lyall 2016).
The UPU is one of the oldest international organizations within the UN system and
currently has 192 member states. Its general purpose lies in the coordination of the
activities of national postal authorities. Amongst the principles governing its opera-
tion, as set forth in the Universal Postal Convention and the General Regulations, two
of the most important were the formation of a single territory by all signatory nations
for the purposes of postal communication; and uniformity of postal rates and units of
weight. An important achievement of the treaty was that it ceased to be necessary, as
it often had been previously, to affix the stamps of any country through which one’s
letter or package would pass in transit; the stamps of member nations would be
accepted for the entire international route.

In the case of product standards, we can expect all states to have a shared interest in
common standards (Box 10.4). Different national product standards and authorization
Copyright © 2020. Macmillan Education UK. All rights reserved.

procedures would pose restraints on industries in all countries and be at odds with the
purpose of integrated markets. We are thus faced with a situation in which all states have
an interest in harmonization and thus a common interest in international regulation
(Holzinger 2002: 69).
The conflict of interest mainly relates to the level of regulation: poorer countries tend to
prefer lower standards than rich countries. In developed countries, citizens may attach a
high priority to issues like environmental quality. Accordingly, these countries may be will-
ing to bear the costs of a more ambitious environmental policy. In developing member
states, these issues are assumed to be of only secondary priority, since the population is less
willing to bear the costs brought about by strict regulations (Scharpf 1997b; Holzinger
2002; Knill et al. 2008a).
In the literature, it is argued that those states that are interested in higher standards
enjoy a more favourable negotiating position. This can be attributed to the fact that, under
certain conditions, high-regulating countries have the opportunity to introduce trade
restrictions for products that do not adhere to their national levels. This is the case if the
member states are entitled to enforce market segmentation, for health or environmental
reasons for example, under the rules of the EU and WTO (Holzinger and Knill 2005, 2008).

Box 10.4 Indian ban on Chinese toys

China is a major manufacturer and exporter of toys. On many occasions, countries or
regions with high consumer protection standards, such as Australia, Canada, the
United States and the EU, have temporarily banned Chinese toys from their markets,
as some of them were found to pose health and safety risks. In 2009, India also
imposed such a ban. This was lifted six months later, but the Indian trade ministry
explained that Chinese toys, in order to be imported, would need to obtain certificates
showing that they conformed to standards prescribed by safety bodies such as the
International Organization for Standardization or the American Society for Testing
and Materials. In 2017, the Indian government adopted and enforced new testing
requirements for toys that focus on their chemical content and flammability. While
the new rules apply to domestic and foreign manufacturers, the market dominance of
China in the toy sector means they will mostly affect foreign trade with China and are
likely to reduce Chinese imports to India.

The constellation of national interests facilitates the international harmonization of

product standards at a high level of regulation. As a consequence, the establishment of inte-
grated markets such as the European Single Market rarely result in the often-feared races to
the bottom when it comes to the regulation of product standards. Rather, the abovemen-
tioned constellation might bring about races to the top, in which the dynamics of market
integration cause the member states collectively to elevate their national standards (Vogel
1995; Scharpf 1997b; Holzinger and Knill 2005, 2008).
While, in principle, bargaining between actors can still resolve agreement problems,
the prospects for international cooperation are gloomier for defection problems. The basic
difference between problems of coordination or agreement and problems of defection is
that, notwithstanding their common interest in the provision of a certain public good
(e.g. a reduction of transboundary air pollution) and corresponding cooperation agree-
ments, when there are defection problems the involved countries prefer to free-ride, tak-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ing advantage of the contributions of the others. Amongst states, the risk of defection
might either hamper the emergence of an international agreement or cause serious com-
pliance problems.
This scenario is typical for the international harmonization of regulations that define
common standards for industrial processes. In contrast to product standards, countries
have no common interest in international harmonization. For poor countries, in light of the
lower level of economic development, harmonization at a high level would severely threaten
existing industrial sectors. Harmonization at a low level would not be an attractive option
either, because national industries would thus be exposed to increased competition from
highly productive companies from the rich countries, who could offer their products at
much cheaper prices. Rich countries, by contrast, would prefer international harmonization
at their high regulatory level. If they cannot achieve this objective, harmonization at a lower
level would be the second-best solution from an economic standpoint, in order to improve
their competitive position vis-à-vis their competitors from the poorer member states.
However, in light of the high priority attached to social and environmental protection in
many rich countries, politically asserting the second solution is not viable. Thus, national

governments prefer the continuation of different national standards over international

regulations at a low level (Holzinger and Knill 2005, 2008).
Compared with product standards, we should expect a less dynamic development with
regard to the international harmonization of process regulations. First, for process stan-
dards there is no common harmonization interest across countries. Second, potential
dynamics emerging from market segmentation are absent. No country can restrict the
import of products which were produced under conditions that do not correspond to their
own regulations on air quality control or water protection. Hence, if the regulation of pro-
duction processes implies an increase in the costs of production, potentially hampering the
international competitiveness of an industry, regulatory competition will exert downward
pressures on economic regulations (Holzinger 2002). It is expected that governments will
lower taxes, social or environmental standards in the face of lobbying or threats to exit made
by the respective industry.
However, this sceptical view is frequently challenged by empirical findings. For example,
the EU has passed very strict process standards in important areas such as air and water
pollution control, chemical control and waste policy, which were not always at the level of
the most ambitious member state, though in fact they did go far beyond the standards of
the least regulated states (Sbragia 2000; Knill and Liefferink 2007; Holzinger et al. 2008a,
2008b). It appears that member states often accept measures whose implementation is
associated with significant economic and institutional costs (Héritier et al. 1996; Jordan
1999; Knill et al. 2008a). In other words, there is a considerable gap between theoretically
forecasted cases of failure to achieve harmonization and empirical observations of success-
ful international cooperation in many instances.
Consequently, the distinction of different problem types and related interest constella-
tions cannot provide a complete account of international cooperation. Nevertheless, this
theoretical model contains important hypotheses which can serve as a point of departure
for analysis and can be modified and further developed according to empirical findings.
There are two patterns that characterize the formation of public policies beyond the nation
state: the role of leader countries when it comes to transferring their regulatory patterns to
the international level, and institutionalized forms of learning and policy diffusion (see
Braun and Gilardi 2006; Dunlop and Radaelli 2016b; Dunlop et al. 2018). We will examine
Copyright © 2020. Macmillan Education UK. All rights reserved.

these factors more closely in the following sections.

The Innovative Potential of Leader Countries

Regardless of the specific problem type and constellation of national interests, the dynamics
of international cooperation may be affected by the pioneering activities of leader countries.
Such countries are typically characterized by very advanced patterns of domestic regulation
in a certain policy field, such as environmental or social policy. Compared with the develop-
ments in other countries and at the international level, they are first movers. They adopt
innovative regulations, even though these policies might entail competitive disadvantages
for their industries and they cannot presume that others might follow them. In light of this
constellation, they seek to incorporate their national regulations into the international leg-
islative process to the greatest possible extent. The activities of leader countries may thus
constitute an important driving force for international cooperation, even in the case of
interest constellations that at first glance hardly seem to provide a sufficient basis for the
adoption of joint solutions.

Such patterns have been observed, in particular, for harmonization activities at the level
of the EU. The fact that we observe European harmonization towards the most rather than
the least stringent of existing member state regulations has been explained by the dynamics
resulting from the activities of pioneer countries in influencing EU policies (Héritier et al.
1996; Jänicke and Weidner 1997; Liefferink and Andersen 1998; Jänicke and Jacob 2004;
Knill and Liefferink 2007).
These dynamics emerge from the interest of national governments in minimizing the
institutional costs of adjusting domestic regulatory arrangements to EU policy require-
ments. In particular, highly regulating countries with a comprehensive and consistently
developed regulatory framework might face considerable problems of adjustment if
European policies reflect regulatory approaches and instruments that depart from domestic
arrangements.
As a result, these countries have a strong incentive to promote their own approaches at
the European level. The most promising way to do this is to rely on the strategy of the ‘first
mover’, i.e. to try to shape European policy developments at the stages of problem defini-
tion and agenda-setting rather than later in the process. This requires that member states
have to win the support of the EU Commission, which has the formal monopoly over initiat-
ing policies at the EU level. The Commission, in turn, is generally interested in strengthen-
ing and extending supranational policy competencies. As a consequence, only those domestic
initiatives that fit with these objectives have a chance of succeeding. This specific interaction
of national and supranational interests favours the development of innovative and ambi-
tious policies at the EU level, hence driving EU harmonization towards stricter regulation.
What impact do these dynamics have on policy-making and the level of regulation at the
international level? First, we should bear in mind that this mechanism to a considerable
extent accounts for the broad spectrum of instruments and patterns of regulation, which
reflect different regulatory traditions at the national level, as it is not always the same coun-
tries that assert themselves in situations marked by regulatory competition. Secondly, the
influence of leader countries can help to prevent the often feared race to the bottom in social
or environmental policy.
Against this background, there is considerable potential for an ongoing expansion of
more stringent regulations at the international, and particularly the EU, level. Of course,
Copyright © 2020. Macmillan Education UK. All rights reserved.

this general trend does not rule out the possibility of negotiations breaking down in indi-
vidual cases or international regulatory approaches partially lagging behind those of indi-
vidual member states. Nevertheless, the role of leader countries might enhance the scope
and level of international regulatory cooperation in the long term.

Deliberation and Diffusion
Up to now we have presumed that the interests of nation states are consistent and defined
clearly during international negotiation processes. An agreement can only be expected when
the preferences of national governments are compatible or distributional conflicts can be
avoided by concessions, package deals or compensation payments. In other words, a rela-
tively static constellation of national interests is assumed, which defines the options for the
design of policy measures at the international level.
However, Joerges and Neyer (1997) have demonstrated that this form of intergovern-
mental bargaining is by no means the only mode of interaction characterizing international
negotiations. Other instances can be observed in which the patterns of interaction are

influenced to a lesser extent by actors defending and asserting national policy positions
than by a collective problem-solving orientation. In such cases, national representatives
develop a common understanding of problems and solutions, such that national ideas and
interests are not regarded as static.
Such processes are facilitated by a specific form of interaction, which is described by
Joerges and Neyer as deliberation. With deliberation, the main focus is placed on discussion
and reasoning on the basis of scientific and technical insights, rather than on strategic bar-
gaining to assert national interests. This ‘deliberative problem-solving’ facilitates learning
processes between negotiating partners (see Stone 2008; Howlett and Joshi-Koop 2011).
Transnational networks of experts or epistemic communities emerge (Haas 1992) in which
converging ideas, assumptions and convictions develop by means of the collective profes-
sional orientation and socialization of the participating actors. This in turn provides the
basis for a convergence of positions of national interest (see Chapter 11). However, the
development of this kind of problem-solving orientation cannot be taken for granted for all
negotiations at the international level.
We can identify three factors which can facilitate deliberative problem-solving. First, the
chances of deliberative problem-solving increase with the uncertainty surrounding the pol-
icy’s possible distributive effects. In such constellations, national interests and problem
definitions are less structured in advance and can be modified more easily. Second, institu-
tionalized interactions between national representatives over a longer period enhance the
diffusion of scientific and technical expertise between the member states of an interna-
tional organization. Third, and related to the latter aspect, the stimulation of such processes
of policy learning presupposes the existence of institutional arrangements which allow for
the regular and continuous exchange of arguments and ideas between national experts and
representatives. The development of these structural arrangements varies across interna-
tional organizations and is most developed in the EU where, amongst others, the Open
Method of Coordination stimulates policy learning (see De la Porte and Pochet 2012).
Deliberative problem-solving processes not only lead to the convergence of national pol-
icy positions and so to agreement at the international level, but also to a facilitating of the
international diffusion of innovative policy concepts, independently of the adoption of
international measures. For example, analysts frequently observe the emulation and trans-
Copyright © 2020. Macmillan Education UK. All rights reserved.

fer of policies and regulatory instruments that have proven to be particularly successful in
one country (see Simmons and Elkins 2004; Volden 2006; Gilardi et al. 2009). The trend
towards internationally converging structures and regulatory patterns that are frequently
obtained (Bennett 1991; Holzinger and Knill 2005; Knill 2005) can in turn contribute to a
decrease in national conflicts of interests over decisions on international measures.
Kern et al. (2001), for instance, show that international organizations play an impor-
tant role in accelerating and facilitating cross-national policy learning (see also, e.g.,
De Francesco 2016). These organizations constitute important channels for multilateral
communication and policy diffusion. Kern et al. demonstrate that – compared with policy
exchange resting on bilateral and horizontal communication between countries – policy
models spread much more widely and quickly if these countries are members of the same
international organization. These results are confirmed by a macro-quantitative analysis
on environmental policy convergence by Holzinger et al. (2008a, 2008b). Analysing policy
developments in 24 Western countries between 1970 and 2000 for a broad range of differ-
ent policy measures and standards, they find not only an impressive degree of cross-
national convergence, but also that this development has to a considerable extent been
driven by processes of communication in international organizations.

Log Jam or Progress? Conflict in International Policy-Making

The analysis up to now has shown that international policy-making is marked by different
constellations of national interests and patterns of consensus-building. Depending on the
analytical perspective, we arrive at different assessments of the possibilities for successful
cooperation. Nevertheless, we can conclude with a number of general observations.
As a rule, there are more favourable conditions for extensive international harmoniza-
tion in product regulation than in process regulation. This can be traced back to two factors:
on the one hand, the harmonization of product standards facilitates market integration and
thus is associated with advantages for all involved countries; on the other hand, those states
that advocate a high level of regulation are in a better negotiating position, because they
have the legal means to enforce high standards single-handedly, if need be. Both of these
preconditions exist in the area of process regulation.
However, we should not infer from this that all attempts to harmonize process regulation
will fail or result in suboptimally low regulatory standards. Depending on how countries
anticipate they will be affected in economic and ecological terms by regulatory proposals,
the constellation of national interests can have a different impact on the capacity for action
at the international level. An important factor that tends to help overcome political impasses
results from the first-mover activities of pioneer countries.
Finally, the interest constellation between countries cannot always be viewed as a static
factor. Policy may be altered because of the deliberation and diffusion that may take place,
under certain conditions, during institutionalized cooperation at the international level.
Deliberation and diffusion in turn facilitate the convergence of national understandings of
problems and solutions. By these means, failures in negotiation can be avoided or overcome.

Key points
•• The chances of successful international cooperation are strongly affected by the underlying
problem type and the related constellation of national interests. A basic distinction can be
made between problems of coordination, agreement and defection.
•• The pioneering activities of leader countries that adopt innovative domestic policies can stimu-
Copyright © 2020. Macmillan Education UK. All rights reserved.

late international cooperation.

•• Processes of transnational policy learning driven by deliberation and diffusion can improve the
basis for international cooperation.

POLICY IMPLEMENTATION

The fact that countries have adopted common policies at the international level by no means
guarantees that the very same countries are also willing and able to implement them. In
Chapter 7, we saw that implementation problems – even in merely domestic settings – are
fairly widespread. This is even more the case when it comes to the formal and practical trans-
position of international or supranational law. Compared with national policies, the imple-
mentation of international policies is characterized by features that further reduce the
likelihood of effective compliance. These features particularly aggravate implementation
problems with regard to those factors we mentioned in Chapter 7, namely deficient policy
design, control deficits and the preferences and strategies of the involved actors. We apply a

perspective where we concentrate on national capacity to implement global policy. It is

important to note that there also exists a complementary perspective that stresses transna-
tional administrations (e.g. multistakeholder initiatives) and how they deliver global policy
(see Stone and Ladi 2015).

Deficient Policy Design

International policies are generally more vulnerable to design deficits than national poli-
cies. This can be traced to the more cumbersome and complex decision-making process
involved, entailing the participation of many actors with often highly heterogeneous
interests. In many negotiation constellations at the international level, options for resolv-
ing distributional conflicts between the involved countries by compensation payments or
packages are rather limited. As a consequence, agreement on common policies can often
only be achieved by compromise formulae that grant national governments sufficient lee-
way to adjust international policy requirements in the light of domestic conditions and
interest constellations.
The first pattern that can be observed comprises policy designs that allow member states
high discretion with regard to the interpretation of policy requirements. This is with good
reason, as it permits member states to accommodate their specific interests. At the same
time, however, this freedom leads to cross-country differences in the application of interna-
tional rules which were originally intended to be uniform. A case in point is the implementa-
tion of the Sustainable Development Goals adopted in 2015, which ask for the attainment
of policy coherence, i.e. an approach for integrating the economic, social and environmental
dimensions of sustainable development at all stages of policy-making (Tosun and Lang
2017). However, it is up to the individual states to define what exactly policy coherence
means to them and how exactly they intend to attain it (e.g. Persson et al. 2016; Tosun and
Leininger 2017).
A second way of balancing conflicting national interests is by the use of vague and open
legal terms that leave a lot of room for interpretation during national implementation. In
international environmental regulation, for instance, we find two recurring and vague legal
terms, where approval requirements are defined based on the ‘technological state of the art’
Copyright © 2020. Macmillan Education UK. All rights reserved.

and in line with ‘economically reasonable’ technical obligations. Very different judgements
can be made when deciding whether a measure is ‘economically reasonable’ or if the require-
ments linked to the current ‘state of technology’ are fulfilled.
Third, member states opposing a certain policy are often appeased by granting them spe-
cific exemptions. These can refer to both formal and substantive regulatory requirements. A
common example of the former category is the possibility of exceeding compliance dead-
lines. International policies normally define a point in time when certain provisions (e.g.
maximum permissible values) come into effect, or when certain standards must be achieved.
In some cases, however, member states are given the opportunity to deviate from this dead-
line if certain conditions are fulfilled. Exemptions referring to the latter category of sub-
stantive aspects typically relate to the scope and level of regulatory requirements and might
vary from case to case, dependent on the specific regulatory issue. These specific exemptions
sometimes grant individual countries extensive means to deviate from international
requirements. The EU directive on large combustion plants, for example, contains various
special agreements accommodating the interests of individual member states. Thus, new

plants with a capacity of more than 400 megawatts are allowed to exceed twice over the
stipulated emission limit for sulphur dioxide, as long as they are in operation for less than
2,200 hours per year. This provision was included following pressure from France, where
so-called peak-load plants are used to cover peaks in demand which cannot be accommo-
dated at short notice by nuclear power (see Knill and Liefferink 2007).
These various exceptions, extensions and vague legal terms entail that individual coun-
tries often have considerable room for manoeuvre in implementing international policies,
despite often seemingly ‘uniform’ policy design. This, in turn, results in a differentiation of
international policy effects at the national level. In most cases, such differentiation leads to
a dilution of original policy objectives. Hence, regulations allowing for deviations from ini-
tial policies can be characterized as design deficits. They are a considerable part of the story
behind implementation deficits in international policy.

Control Deficits
In Chapter 7 we saw that many implementation problems are linked to control deficits, i.e.
constellations in which political principals are not able to control the agents that implement
policy. For several reasons, implementation of international policy is particularly vulnerable
to agency drift, which typically entails more or less far-reaching deviations from the formal
and practical requirements of any policy.
First, the implementation of international policies usually involves more institutional
levels and actors than is typically the case for national political systems. As each of these
actors (international bureaucrats, national politicians, administrative agencies, street-level
bureaucrats, transnational and national interest groups) might pursue different objectives,
there is a high likelihood of deviation from initial policy goals (Pressman and Wildavsky
1973; Winter 2003; Howse and Teitel 2010).
Second, in monitoring the implementation of international law, international bureau-
cracies possess very limited formal resources for enforcing the cooperation of the public
and private actors participating in the implementation process. On the one hand, respon-
sibility for the execution of international policies generally lies with the member states.
This means that international bodies have no competence to intervene directly in the
Copyright © 2020. Macmillan Education UK. All rights reserved.

implementation process at the domestic level. The defining characteristic of international

law is the lack of a centralized enforcement mechanism. When international law is
enforced, this is done by states themselves (Brewster 2009). On the other hand, a major
obstacle to effectively controlling the implementation of international law is linked to
problems with data and monitoring. International organizations are strongly dependent
on information provided by the member states. Implementation deficits can hence only
be discovered if they are reported as such by member states or if failures are reported to
the international body concerned by national interest groups or citizens. Though in this
way potential deficits in formal transposition can still be detected rather easily, the detec-
tion of implementation deficits is much more difficult when it comes to the practical appli-
cation of international law.
In view of their limited competencies and resources, international bodies are unable to
monitor and control the implementation of international policies as they are practised in
member states. Instead, they have to rely on information provided by national authorities
or complainants. This entails problems for comparing implementation performance across

countries. The more developed the administrative monitoring capacities and structures of
interest organizations in a country, the more likely that potential implementation problems
will be detected and reported. This leads to a paradoxical constellation where countries with
higher implementation capacities face a higher risk of being accused of failing to implement
than countries in which such capacities are less developed.
In summary, the implementation of international policy can generally be expected to suf-
fer from far-reaching control deficits. The latter can be traced not only to the huge number
of institutional levels and actors that are involved in the implementation process, but also
to the lack of comprehensive data on implementation effectiveness and the underdeveloped
ability of international bodies to sanction both formal and practical non-compliance.

Preferences and Strategies of the Involved Actors

The high dependence on the willingness and capacity of national governments to implement
international law effectively underlines the crucial relevance of the interests of national actors
in accounting for varying degrees of implementation. The central role of national actors and,
in particular, national governments in ensuring compliance is underscored by the fact that
international law has no direct effect at the member state level; i.e. in order to become effec-
tive, international law has to be transposed into national law by national legislation. In a simi-
lar vein, the fact that international law is not superior to national law entails that national
transposition legislation can be annulled by new national law at any time. This leads us to the
general question of when we might expect national governments to transpose international
policies and apply them effectively. Several factors should favour effective implementation.
First, it may be argued that international law and organizations are endogenous to the collec-
tive preferences of states, as are their membership criteria and decision-making rules. Hence,
high rates of compliance can be due to the fact that international agreements may not depart
much from what states would have done in their absence (Bueno de Mesquita 2009). Second,
there is evidence that states are generally concerned about their international reputation. A
state that violates international law develops a bad reputation, which leads other states to
exclude the violator from future cooperative opportunities. Anticipating a potential loss of
future gains, states have an incentive to implement international policies effectively, even if
Copyright © 2020. Macmillan Education UK. All rights reserved.

the latter are not in their immediate interests (Brewster 2009). Compliance with international
law, according to this view, can be instrumental in building leaders’ reputations and increasing
their chances of political survival (Bueno de Mesquita 2009: 353).
However, these arguments can be challenged on various grounds. First, as emphasized by
Martin and Simmons (1998: 743), once established, institutional arrangements will con-
strain and shape domestic policy choices, even as they are constantly challenged and
reformed by member states. International institutions are therefore not only subject to the
decisions of states, but at the same time impact on subsequent governmental activities.
Even if we assume that, at the time of adoption, a certain international policy was fully in
line with the interests of national governments, subsequent developments at the domestic
level might put this initial congruence into question.
Such developments can be linked to changing policy salience at the national level or to
changes in government. It may not be taken for granted that the government in office when
a certain international policy was negotiated and the government that is in office when it
comes to the implementation of these policies have similar interests with regard to the regu-
lation of the issue at stake. In addition, swift changes in international rules are often

prevented by the fact that international decision-making processes usually require high
quorums (Pierson 1996). In these situations of so-called joint decision traps, a few countries
or even only one state that is in favour of the status quo can block reform attempts by other
countries (Scharpf 1988).
Second, the reputation argument raised above needs modification. The extent to which
reputation concerns will lead to compliance will depend on how national governments weigh
the potential costs of being excluded from future cooperative agreements and the degree to
which they are tempted to renege on international agreements to play to the domestic audi-
ence, and therefore to increase their chances of re-election.
Yet studies on the role of international public administrations – i.e. the bureaucracies of
international organizations – might rely on a range of informal tools of governance that can
affect the cost–benefit calculations of member states in order to stimulate effective compli-
ance with international rules (Knill and Bauer 2016; Bauer et al. 2016). Although
international administrations in general have very limited enforcement powers, some of them
have developed elaborate routines of strengthening the implementation of their policies.
Some international administrations hence engage not only in monitoring and enforcement,
but also in enhancing the impact and effectiveness of their programmes through capacity-
building and policy promotion. Moreover, they try to strengthen the implementation capaci-
ties of the relevant actors, engaging in proactive communication with member states that goes
beyond their formal duties. Moreover, research has shown that many international bodies
adopt a highly strategic approach with regard to implementation. This implies that enforce-
ment and monitoring routines are guided by the strategic considerations of the administra-
tion, either attempting to go beyond formal duties (in order to foster effective implementation)
or deliberately shying away from using their formal powers (in order to avoid member states
reducing their support for the development of new policies at the level of the international
organization in question; Knill et al. 2018; Box 10.5).
In conclusion, there is a considerable potential that – even if they initially agree with
international regulations and laws – countries deviate from international policy objec-
tives during the implementation stage. The major reasons for such deficits are linked to
inadequate policy designs and control deficits, as well as the preferences and strategies of
the domestic actors involved in the implementation process, in particular national gov-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ernments. Although these problems are to a certain extent also present when it comes to
the implementation of merely national policies, they can be considered rather fundamen-
tal with regard to compliance with international ones. At the same time, however, many
bureaucracies of international organizations have developed a range of informal routines
over time in order to strengthen the implementation of international rules and to over-
come their formally rather limited enforcement and monitoring powers.

Key points
•• The implementation of international policy often suffers from deficient policy design (such as
vague legal terms, extensions and exceptions).
•• Control deficits constitute a crucial problem for the implementation of international policy,
given that international organizations generally lack enforcement powers.
•• The specific multilevel constellation characterizing the implementation of international law
facilitates domestic deviations from original policy objectives in the light of changing interests
and the strategic behaviour of national governments.

Box 10.5 Implementation of international agreements in federal

systems
Numerous empirical analyses have demonstrated that implementation of the
provisions of international agreements is likely to depend on domestic actor and
institution constellations. Against this backdrop, Paquin (2010) asks whether fed-
eral political systems affect implementation by examining Belgium and Canada.
The Belgian system is characterized by substate actors who have a co-decision role,
in which pronounced intergovernmental mechanisms are in place. In Canada, by
contrast, the decision-making process is more centralized and intergovernmental
mechanisms are poorly institutionalized. The analysis reveals that the Belgian
system displays higher implementation effectiveness, largely because its substate
actors have an important role at every step of the conclusion of an international
treaty. The power of the Belgian regions became visible in 2016, when Wallonia
threatened to block the Comprehensive Economic and Trade Agreement between
Canada and the EU. Belgium only agreed to provisional ratification of the agree-
ment when it was granted its wish that the dispositions on investment would be
excluded and examined by the European Court of Justice. The agreement includ-
ing the investment dispositions will only enter into force when all EU member
states have ratified it.

CONCLUSIONS

We have seen that the development of public policies beyond the nation state can be
driven by varying rationales, including economic interdependencies between countries
as well as the emergence of problems of transnational scope that cannot be effectively
addressed by unilateral responses. As a result of these factors, states cooperate interna-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tionally in many different organizational forms and policy sectors, although the devel-
opment of international policy competencies varies from sector to sector and issue to
issue.
Given the fact that international policy formulation typically requires the accommo-
dation of many, often highly heterogeneous, interests, joint agreement amongst the
involved countries cannot be taken for granted. The fact that we can nevertheless some-
times observe far-reaching international cooperation can be explained by three fac-
tors – the underlying problem type, the facilitating role of pioneer countries, and
cross-national policy learning and diffusion. Finally, we have seen that – compared with
national policies – the implementation of international laws is much more vulnerable to
non-compliance and deviation from initial policy objectives. This can be traced, in

particular, to deficiencies in policy design and control deficits, as well as the interests of
domestic actors in charge of implementation.

WEBLINKS

www.c40.org. This website presents the C40 transnational city network, which con-
sists of major cities.
www.energy-cities.eu. This website introduces another transnational city network
called Energy Cities.
www.gppi.net. The Global Public Policy Institute provides policy analysis and policy
advice to improve global governance.
www.inogov.eu/outputs/mooc. This website hosts a massive open online course on
the governance of climate change.
www.kof.ethz.ch/en/forecasts-and-indicators/indicators/kof-globalisation-index.
html. This website hosts the KOF Globalization Index, which can be downloaded
for many different years.
www.uia.org/yearbook. On this website all international organizations are
listed.

FURTHER READING

Börzel, T. A. and Risse, T. (eds.) (2016). The Oxford Handbook of Comparative
Regionalism. Oxford: Oxford University Press. This collection provides a compre-
hensive overview of different forms of regional integration, covering a wide range
of geographical areas and themes.
Hoekman, B. M. and Kostecki, M. M. (2010). The Political Economy of the World Trading
System. Oxford: Oxford University Press. An informative introduction to interna-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tional trade and the nature and functioning of the WTO.

Jordan, A., Huitema, D., van Asselt, H. and Forster, J. (eds.) (2018). Governing Climate
Change: Polycentricity in Action?. Cambridge: Cambridge University Press. This book
is the basis for the MOOC on climate governance presented above in the section on
relevant weblinks.
Knodt, M., Piefer, N. and Müller, F. (eds.) (2016). Challenges of European External
Energy Governance with Emerging Powers. New York: Routledge. This is a stimulat-
ing collection of case studies on how the EU interacts with emerging powers such
as China in the domain of foreign energy relations.
Sethia, B. and Kumar, P. (eds.) (2018). Essentials of Global Health. Philadelphia, PA:
Elsevier Health Sciences. An authoritative and accessible introduction to the foun-
dation of global health.

Walter, S. (2013). Financial Crises and the Politics of Macroeconomic Adjustments.

Cambridge: Cambridge University Press. This book makes a strong case for the
interconnectedness between the national and the international levels, which
became particularly apparent during the economic and financial crisis unfolding
from 2008 onwards.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Reader’s guide
We now turn to one of the most central topics in the study of public policy, namely the
patterns and sources of policy change. Policy change is again an issue that cuts across
the different stages of the policy cycle and therefore allows us to combine different
analytical concepts and theoretical perspectives. The causes of policy change can be
found in the outcomes of the processes taking place at individual stages such as
agenda-setting, implementation problems or (negative) evaluation outcomes. In addi-
tion, the theories discussed in the context of decision-making may be equally helpful
in scrutinizing the events of policy change. The institutional characteristics of polities
again prove to be important for facilitating or impeding policy change. However, there
also exists a specific set of policy process theories that specifically aim at explaining the
occurrence of policy change. In the second step, we will turn to cross-national policy
convergence, where we address the question of the extent to which national policies
become more similar over time. We discuss basic types and dimensions of the conver-
gence concept and differentiate it from related concepts, including policy transfer,
policy diffusion and isomorphism. Based on this discussion, the third step focuses on
the causes and conditions that trigger cross-national policy convergence.
Copyright © 2020. Macmillan Education UK. All rights reserved.

INTRODUCTION

Policy change is a central topic in policy studies. Under which conditions can we expect
adjustments or transformations of existing policies? Why do some policies remain
untouched, while others are changed? Do policies follow a sequential path of development?
Why do policies generally change incrementally, but sometimes radically? These are only
some of the questions that have guided research on policy change over recent decades.
Notwithstanding the considerable progress made in sustained research activities
related to these issues, we are still left with open questions and challenges. On the one
hand, we face a broad variety of theoretical frameworks that assess policy change from
different analytical perspectives and that pursue different research interests. On the other
hand, the measurement of policy change is not always comparable across different stud-
ies, which results from the many elements and dimensions of policies that may be affected
by change (see, e.g., Hall 1993).
In recent years, the study of policy change has gained additional impetus from research
activities that concentrate on the cross-national aggregate effects of change. The central

207
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:17:28.
208 Public Policy

question here is the causes and conditions that lead to different forms of cross-national
policy convergence. To what extent do policy changes at the domestic level result in increas-
ing or decreasing policy homogeneity across countries? Answers to these questions not only
depend on the way convergence is conceptualized and measured, but also on the specific
factors that affect national policy developments in individual cases.
In this chapter, we offer an integrated discussion of policy change and policy conver-
gence. We begin with policy change, since it is one of the key topics in public policy, and then
shift our attention to policy convergence as a newer item on the research agenda.

OLICY CHANGE: THEORIES, MEASUREMENT AND GENERAL

P
PATTERNS
The study of policy change may be considered a core area of public policy, with the major inter-
est being in the description and explanation of changes in dominant policy patterns in differ-
ent fields. In the literature, we find many ways by which policy change is measured. So far, no
dominant and generally accepted approach has emerged. As a consequence, the theoretical
and empirical findings of varying studies are not commensurable (i.e. they cannot be discussed
by using a shared nomenclature), as the dependent variable (i.e. the phenomenon we aim to
explain) is conceptualized and measured very differently. This conceptual diversity is reflected
in a broad variety of basic theoretical approaches to the study of policy change.
In this section we will provide an overview of the key approaches to the study of policy
change. We will proceed in three steps. After a discussion of the central theoretical frame-
works that account for policy change and policy stability, we turn to key issues associated
with the measurement and operationalization of policy change, before presenting expecta-
tions about the overall direction of policy change, i.e. the relationship between policy expan-
sion and policy dismantling.

Theories of Policy Change

In Chapter 4, we discussed a broad range of basic theoretical approaches that serve as a starting
point for analysing public policy, broadly distinguishing between structure-based, institution-
based and interest-based frameworks. Frameworks used to explain policy change and policy sta-
bility can also be classified along these lines and hence be linked to the broader concepts
mentioned in Chapter 4. However, theories of policy change differ not only in the extent to
which they emphasize the role of structure, institutions and interests as explanatory factors, but
also in the extent to which they link these factors into more complex theoretical frameworks.
More specifically, theories of policy change can be classified according to their underlying
causal logic. On the one hand, we find approaches that are based on a ‘linear-additive’ view
of causality, in which a clear separation between dependent and independent variables is
assumed (Capano 2009: 16). Such approaches typically address the question of why policy
change occurred or not. Independent variables generally refer to macro factors, such as
socio-economic conditions, the strength of societal interest groups, changes in government,
party political positions and international influences (see Schmidt 2015).
On the other hand, there are approaches characterized by the ‘logic of combinative causal-
ity’, in which possible combinations of causal conditions capable of generating a specific

result are sought (Ragin 2006). The theories in this tradition are primarily concerned with
the question of how different variables can be associated with each other according to
Boolean logic rather than in a simple positive or negative fashion (Capano 2009: 16). Studies
of policy change that embrace the logic of combinative causality typically put forward theo-
retical models that combine necessary and sufficient conditions. Concerning the empirical
testing of these models, the method of choice is qualitative comparative analysis (for an over-
view, see Schneider and Wagemann 2012). An apt illustration of the theoretical and method-
ological approach that is based on combinative causality is provided by Fischer (2014), who
shows that in Switzerland major policy change is brought about in two constellations that are
characterized by a combination of different conditions: (1) actor coalition structures with low
conflict and strong collaboration amongst coalitions, as well as (2) structures with dominant
coalitions and weak collaboration, which both facilitate major policy change.
In the following, we will neither adopt a linear-additive view of causality nor embrace the
logic of combinative causality, but will present theoretical approaches and individual ele-
ments of these that place emphasis on the policy process underlying or leading to policy
change (for an overview, see Weible and Sabatier 2017).

Policy Change as Change in Preferences and Beliefs

The Advocacy Coalition Framework (ACF) developed by Sabatier (1988) and Sabatier and
Jenkins-Smith (1993, 1999), and modified by Sabatier and Weible (2007), is one of the most
comprehensive approaches to policy change (see Jenkins-Smith et al. 2017). It employs the
structure of beliefs in the governing coalitions of so-called policy subsystems – semi-
autonomous networks of policy participants that focus on a particular policy issue, usually
within a geographical boundary – to predict changes in shared beliefs that lead to changes in
policy. The framework posits a three-tiered hierarchical belief system, including deep-core
beliefs, which are ontological and normative; basic political values, which constitute the
‘glue’ that ties coalitions together; and policy- or instrumental-level positions. In this sense,
an advocacy coalition consists of actors from a variety of market, state and civil society insti-
tutions at all levels, which share the abovementioned basic beliefs (i.e. on policy goals and
other perceptions) and coordinate their activities over time.
Copyright © 2020. Macmillan Education UK. All rights reserved.

As Markard et al. (2016) conveniently summarize, according to the ACF, there are different
paths to policy change that also come along with more or less far-reaching departures from the
status quo. The first is policy-oriented learning, which corresponds to a process in which new
information on the policy issue concerned becomes available and induces actors who are part
of the dominant advocacy coalition to change their beliefs according to the new information.
Typically, policy-oriented learning is routine based and produces only minor policy change.
Coalitions adapt the secondary aspects of their beliefs in light of new information.
The second path refers to rare constellations in which external events cause subsystem
instability and increase the potential for rapid and more far-reaching policy change. External
events may include, for instance, the election of a new government with beliefs that favour
one coalition over the other, or focusing events like crises or shocks (economic recessions,
environmental disasters, rampage shootings etc.). Such external events may lead to both
external and internal shocks. An internal shock refers to event-triggered changes within a
coalition’s belief system, prompting a coalition to revisit its policy core beliefs, perhaps fol-
lowing a realization by many of its actors that existing policies have failed. An external shock,
by contrast, has the added element of competition – another coalition uses the experience

of a major event to reinforce its position within the subsystem, largely by demonstrating that
its belief system is best equipped to interpret new information and solve the policy problem.
In other words, the external event is not enough to cause an external shock; it also has to be
exploited successfully by a competing coalition (Weible and Ingold 2018) (Box 11.1).
The importance of learning as a trigger of policy change is not restricted to the ACF. Indeed,
numerous explanations of policy change are based on notions of learning, including the
concepts of ‘political learning’ (Heclo 1974), ‘government learning’ (Etheredge 1981), ‘social
learning’ (Hall 1993) and ‘lesson-drawing’ (Rose 1991; for an overview, see Bennett and
Howlett 1992: 275). Advancement on the lesson-drawing approach includes the more
recently developed concepts of ‘rational learning’ and ‘Bayesian learning’ (see Meseguer
Yebra 2009). Both conceive of governments as rational actors looking for appropriate policy
solutions that work well abroad (see Gilardi et al. 2009; Gilardi 2010). In the case of Bayesian
learning, governments additionally update their beliefs about the consequences of policies
with all the available information about policy designs and choose the one that is expected
to yield the best results.
In the literature, particular attention has been paid to Hall’s (1993) theoretical argument
about social learning, which conceptualizes policy-making as a process that is about making
decisions along three dimensions: the first is the overarching goal that guides a policy in a
particular field; the second is the policy instrument used to attain those goals; and the third
is the precise setting of this instrument. These three elements jointly form a ‘policy para-
digm’. According to Hall’s reasoning, policy change is most likely for the policy setting (i.e.
‘first-order change’), but becomes more difficult when it comes to policy instruments (i.e.
‘second-order change’) and even more so with goals (i.e. ‘third-order change’). Changes in

Box 11.1 Changes in the Finnish policy on biofuels

The European Union (EU) has been attempting since 2003 to promote the produc-
tion and use of biofuels in its member states in order to reduce greenhouse gas
(GHG) emissions and to mitigate climate change. One of the countries that was
initially against promoting biofuels is Finland. The government of Finland even
Copyright © 2020. Macmillan Education UK. All rights reserved.

openly criticized the EU for this policy and declared that it would not follow it. In
2011, however, the Finnish government changed its approach to biofuels and intro-
duced fuels blended with biofuels. According to Lovio and Kivimaa (2012), the rea-
son for this policy change can be seen in the country’s large forestry industry and
the development of technologies that allowed the production of biodiesel from
wood-based feedstock. In this particular case, policy change was stimulated by
policy-oriented learning, since the availability of new technologies changed the cal-
culus of the government. With the technological possibilities, the production of bio-
diesel emerged as an efficient instrument to generate additional revenue from the
forestry industry, as well as to take a measure against climate change. In the mean-
time, the promotion of biofuels as a means to reduce GHG emissions has become
institutionalized. In 2016, the Finnish government announced that it would
increase the share of biofuels in the fuel pool to 30% by 2030. In addition, the gov-
ernment embraced a policy agenda that fosters the implementation of the bioecon-
omy in the country, which aims to convert biological resources and waste into
value-added products (see McCormick and Kautto 2013).

instruments can be attributed to ‘incrementalism, satisficing, and routinized decision mak-

ing’ (Hall 1993, 280). The adoption of a new policy instrument is likely to be the outcome of
strategic action. Changes in policy paradigms are associated with experimentation with new
policies and lesson-drawing from policy failures.
Learning also represents an important factor in the policy subsystem adjustment model
put forward by Howlett and Ramesh (2002). The authors’ model builds on the critique that
the Advocacy Coalition Framework treats endogenous factors – i.e. the advocacy coalitions
themselves and exogenous factors such as macro-economic shocks – individually. As a result
of this analytical distinction, there is little to prevent the conflating of exogenous and
endogenous processes for explaining the nature of policy change. The authors argue that
exogenous factors can induce subsystem adjustments by producing new policy regimes, i.e.
by affecting endogenous factors. The four resulting types of adjustments include ‘policy
learning’, ‘venue change’, ‘systemic perturbations’ and ‘subsystem spillovers’.
With the first two types, actors located within the existing subsystem use exogenous fac-
tors, such as new information, to pursue desired policy goals. In these cases, exogenous
factors initiate policy change, but subsequently the goals and strategies of actors internal to
the subsystem become more influential. Systemic perturbations can generally be seen as
exogenous events that draw new attention to a policy sector, such as the occurrence of a
financial crisis. Crises bring new actors or new ideas and policy goals to the forefront of
policy-making and can produce dynamic policy change. Much like a systemic perturbation,
a subsystem spillover is generated by exogenous factors, which promote the merging or
integration of previously separate subsystems into a single policy domain.

Linking Policy Change and Institutional Change

The Punctuated Equilibrium Theory (PET) advanced by Baumgartner and Jones (1993/2009)
seeks to explain why political processes are usually characterized by stability and incremen-
talism, but occasionally produce fundamental shifts from the past. As we have already seen
in Chapter 5, where we discussed the model with a specific perspective on agenda-setting,
this framework conceives of the given institutional setting as crucial in influencing policy
dynamics and the degree of change.
Copyright © 2020. Macmillan Education UK. All rights reserved.

PET addresses the challenge of explaining policy change and persistence within one theo-
retical framework. It starts from the premise that policy-makers suffer from bounded ratio-
nality and cannot pay attention to all demands for political adjustment at the same time.
Typically, the existence of a policy subsystem allows for the parallel processing of several
demands and compensates for this cognitive limitation. At the same time, however, the
processing of new information by subsystems typically reinforces the policy status quo
rather than promoting policy change. This is because subsystems tend to provide a fixed
venue in which stable actor constellations interact and, over time, create a dominant policy
image or paradigm. This policy image will influence the processing of new incoming infor-
mation and thereby attenuate the perceived demands for policy adjustment. Only when
information is able to attract attention beyond the relevant subsystem can this negative
feedback mechanism attenuating the perceived demand for change turn into a positive feed-
back mechanism amplifying the perceived demand for adjustment. This happens when the
dominant policy image is effectively challenged, the number of venues dealing with the issue
expands and new actors become part of the policy debate. When policy monopolies in sub-
systems break down this way and issues move out of the subsystem to attract attention at

the macro-political level, the opportunity for major policy change arises. These mechanisms
of negative and positive feedback mean that political systems tend to process new informa-
tion disproportionately. They ‘shift from underreacting to overreacting to information’
(Jones and Baumgartner 2012: 7).
In turn, one central claim of PET is that policy output reflects a pattern of punctuated
equilibrium where policy stability is disrupted every now and then by substantial punctua-
tions. In other words, the steady flow of political demands is typically translated into patterns
of enduring policy stability interrupted by rare instances of major change. Where the design
of political institutions introduces more or less friction into this process of translating policy
demands into policy adjustments – by defining procedural rules and hurdles for change – pat-
terns of punctuated equilibrium will be more or less pronounced (Hurka et al. 2017).
Closely linked to historical institutional accounts of policy-making (Thelen and Steinmo
1992) are models that interpret policy change as path-dependent development. Such models
conceive of policy change as sequential development, with previous decisions having a deci-
sive impact on subsequent ones. The sequential development can be driven by two logics. On
the one hand, self-reinforcing sequences can account for the long-term reproduction of exist-
ing policy patterns as a result of increasing returns, which once a policy is adopted deliver
increasing benefits to the actors concerned. Hence, over time it becomes increasingly difficult
to transform existing arrangements or select previously available options, even though the
existing approach has been widely recognized as suboptimal (Pierson 2000; Hacker 2004;
Howlett and Rayner 2006; Peters et al. 2017). On the other hand, reactive sequences conceive
of policy development as involving temporally ordered and causally connected events. Each
event within the sequence is in part a reaction to previous events. With reactive sequences,
the final event is typically the dependent variable under investigation, and the overall chain
of events can be seen as a path leading up to this outcome (Mahoney 2000: 509).
Depending on the underlying sequential logic, we might arrive at different expectations
about policy change and policy stability. According to the self-reinforcing sequence, the
expected pattern is policy stability, with change being restricted to incremental adjustments
of existing arrangements. Path-breaking changes can only be expected in constellations of
institutional destabilization through some kind of exogenous shock (Howlett and Cashore
2009: 35). If policy development is analysed as a reactive sequence, by contrast, it is likely
Copyright © 2020. Macmillan Education UK. All rights reserved.

that path dependency will coincide with major changes.

Path dependency is interlinked with the concept of policy feedback (e.g. Mettler and
SoRelle 2017). This scholarship mostly connects public policy to future policy change or
individual behaviour through the mechanism of feedback. Positive feedback is associated
with stability and negative feedback with policy change (see also Jordan and Matt 2014). In
this context, Weaver (2010) observes a growing interest in negative feedback in the litera-
ture, which he regards as a correction to the emphasis that the historical institutionalism
and path dependency literature placed on positive feedback. Therefore, the burgeoning
attention paid to negative feedback concurs with the general increase in interest in the
explanation of policy change.
The importance of institutional factors is also highlighted by theoretical work on veto
points (Immergut 1990; Immergut and Anderson 2007) and veto players (Tsebelis 2002;
König et al. 2010), which have developed into central analytical concepts for explaining pol-
icy stability. Both perspectives emphasize ways in which political institutions can obstruct
policy change. In doing so, the veto point theory highlights the relevance of the legislative
process, which is conceived as a chain of political decisions taken in a series of policy arenas.
However, veto player theory goes a step further by taking the policy preferences of the

players into account. In this context, the volume edited by König et al. (2010) shows that
policy change – there conceived as policy reforms – is still attainable despite the existence of
veto players and their support for or opposition to the policy projects concerned.

Policy Change by Chance

The Multiple Streams Framework (MSF) put forward by Kingdon (2003/2011) underscores
the possibility of policy change through agenda-setting. The MSF’s starting point is the
notion of three independent streams, known as ‘problems’, ‘policy’ and ‘political’ streams.
For an issue to be placed on the political agenda, these independent streams need to come
together at some point. The opportunity to bring these streams together arises if a ‘policy
window’ or ‘window of opportunity’ opens. For this to happen, a policy entrepreneur has to
bring the streams together, which is known as the process of ‘coupling’.
Herweg et al. (2017) offer a detailed discussion of the concept of the policy window and
explain that some policy windows (e.g. elections) are predictable, whereas many other windows
(e.g. natural disasters) are less predictable. It is analytically difficult to decide ex ante whether or
not these events constitute a policy window for a given policy at all (see also Béland 2016;
Cairney and Zahariadis 2016; Zohlnhöfer and Rüb 2016). What is perhaps more important, as
Herweg et al. (2016) argue, is that policy entrepreneurs utilize such events as policy windows.
A similar argument is advanced by Carter and Jacobs (2014), who explicitly contend that
policy windows can remain open longer than generally expected and that, for example, party
competition can help to extend the opportunity for policy change in parliamentary systems.
What is particularly interesting about this approach is that it is based on the coupling of the
problem, politics and policy streams by policy entrepreneurs according to the MSF, but it
also adds to that the importance of changes in policy image and institutional venues, as sug-
gested by the PET. In other words, this approach combines elements of the MSF and the PET
(for a similar endeavour to integrate different theoretical perspectives, see, e.g., Breunig
et al. 2016). More broadly, it shows that the concepts summarized above lend themselves to
be utilized individually or in combination with the others.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Measurement of Policy Change

A thorough explanation of policy change presupposes – as a necessary condition – the devel-
opment of a sound concept of measurement for the phenomenon under study. However,
while students of public policies are increasingly concerned with the analysis of policy
change, conceptual choices and problems related to the assessment of change are rarely
made explicit. As Capano (2009: 14) correctly observes, ‘the definition of what is change
represents a strategic issue for the researcher. It really makes a substantial difference if pol-
icy change is defined in terms of the transformation of the definition of the issues in ques-
tion, or as the structure and content of the policy agenda, or in terms of the content of the
policy programme, or as the outcome of implementation of policy.’
Nevertheless, specifying the object of change is only one, albeit important, challenge to be
overcome. There are several additional problems deserving analytical attention (see, e.g., Tosun
and Schnepf 2020). First, policy change is a dynamic concept and thus attention must be paid to
the time frame within which processes of change are observed. Second, the evaluation of the
degree of policy change (e.g. radical versus incremental) is affected by different levels of abstrac-
tion. Third, policy change is a multifaceted concept with respect to the direction as well as the

dimensions of change. Finally, the likelihood of policy expansion relative to policy dismantling
must be specified. From this, it follows that the researcher must outline how he or she conceives
of the occurrence, direction and intensity of policy change. In this section, we take up these chal-
lenges and provide some guidance for more accurately analysing policy change.

Temporal Issues Regarding Policy Change

When addressing the phenomenon of policy change, we need to observe at least two points in
time in order to assess that a policy has changed its characteristics. Therefore, studies of policy
change must be based on data which – depending on the specific research focus – covers peri-
ods of years or even decades (Howlett and Cashore 2009: 35). To receive the most convincing
empirical evidence of policy change, the number of observations should be as large as possible.
Ideally, this would imply observations for each year over a sufficiently long period of time. As
such, the selection of the time axis represents one of the most challenging tasks in the study
of policy change, since the conclusions to be drawn from the analysis are highly contingent on
the observation period. The main problem arises from ‘censoring’, i.e. the incomplete record-
ing of information at the beginning or end of the event that is of interest.
Figure 11.1 illustrates four frequent forms of censoring. We randomly defined an obser-
vation period starting in 1990 and ending in 2010. The two crosses (X1 and X2) symbolize an
episode of policy change, taking place during an infinite period of time. Yet the only avail-
able empirical information refers to the observation window, which is marked by two verti-
cal lines. Observations A to D are used to differentiate between episodes of change and the
accuracy of their empirical assessment.
Observation A is completely censored on the left, implying that both moments of policy
change occurred before the beginning of the observation. Hence, we would merely observe
the prevalence of the status quo and therefore underestimate the likelihood of any policy
change. Observation B is partially censored on the left so that we merely observe one
moment of change instead of two. Accordingly, we might overestimate the scope of the
change, since we are missing information about the previous level of change. Observation C
is completely censored on the right, implying that the start and end times of the episode of
change are located after the end of the observation. Again, we would wrongly conclude that
Copyright © 2020. Macmillan Education UK. All rights reserved.

no change has occurred. What we in fact want is a time frame which enables us to observe

A ´1 ´2
B ´1 ´2
´1 ´2 C

´1 ´2 D

1990 2010 Time

Figure 11.1 Types of censoring

Source: Own elaboration based on Blossfeld et al. (2007: 40).

the entire episode of change in order to arrive at correct conclusions with regard to the
occurrence of change, as shown by observation D.
To reduce the risk of censoring, it is essential to select the observation period in accor-
dance with the empirical characteristics of the object of analysis or – even better – in line
with theoretical considerations. For example, if we are interested in changes to public policy
arrangements in Canada due to the entering into office of the government headed by Justin
Trudeau (Liberal Party), we would need to start the observation with the 2015 elections that
brought Trudeau into power. If we started to measure policy change any later than 2015, it
would be possible to underestimate the extent to which public policies have changed due to
the new partisan composition of the government. Accordingly, the identification of the rel-
evant time axis is a crucial step for examining policy change and should receive enhanced
attention within the development of the research design.

Policy Change at Different Levels of Abstraction

When is change fundamental and when is it incremental? Sabatier and colleagues, for
instance, approach the issue by focusing on changes in the beliefs of the coalitions of actors
concerned. On this basis, they distinguish between two degrees of change, namely changes
in core beliefs and changes in secondary aspects, with the former being much more difficult
to achieve than the latter (Sabatier and Jenkins-Smith 1993, 1999; Sabatier and Weible
2007). Another widely used distinction, which we have already mentioned at various times,
is made by Hall (1993), who differentiates between three levels of policy change: (1) policy
paradigms (the overarching goal that guides policy in a particular field), (2) policy instru-
ments (the means used to achieve these goals) and (3) the precise setting or calibration of
those instruments. Baumgartner and Jones (1993/2009) also distinguish between different
magnitudes of policy change, namely exceptional large-scale changes (punctuations) and –
as they assume is the general rule – incremental or minor adjustments (for examples, see,
e.g., Breunig et al. 2010; Tosun and Schnepf 2020).
However, the assessment of whether changes affect the policy core or only secondary
aspects strongly depends on the evaluation benchmark and the detailed analysis and inter-
pretation of the underlying case by the individual researcher. Respective judgements can
Copyright © 2020. Macmillan Education UK. All rights reserved.

only be made by taking account of the specific policy issue. The same problem applies to the
identification of policy paradigms and the distinction between incremental and fundamen-
tal change that Baumgartner and Jones make. In other words, the evaluation of the degree
of policy change crucially depends on the measuring rod against which change is assessed. It
makes a big difference in terms of theorizing as to whether we explain the change of a cer-
tain policy (e.g. with regard to the atmospheric emissions of large combustion plants) ‘on
the spot’ (i.e. by focusing on the reasons that caused the change at a certain point of time in
a given country), or whether we adopt a more abstract perspective on one and the same
change, interpreting it in the light of long-term changes or changes taking place in many
other countries (Knill and Lenschow 2001; Capano 2009; Tosun and Schnepf 2020).

Policy Density and Policy Intensity: Directions and Dimensions of Policy Change

Change is defined as any departure from the status quo. What is often overlooked is that
policy change involves two directions, namely expansion and dismantling (Knill et al. 2009,
2010, 2012; Tosun and Schnepf 2020). Policy expansion implies that a new policy is

introduced or an existing one is intensified. By the same token, policy dismantling means
that a policy or parts of it have been abolished without substitution. This still leaves us with
limited guidance for actually assessing the characteristics of policy change. To remedy this
situation, we suggest a differential measurement of policy change based on two basic dimen-
sions: policy density and policy intensity. By doing this, we distinguish different types of
change that refer to varying policy dimensions. So we pay attention to the presence of a
policy, the instruments a policy utilizes and the setting of those policy instruments. Our
framework represents a refinement and extension of these categories.
This way, we deliberately restrict our focus to changes in policy outputs; hence we exclude
other potential objects of policy change (such as problem definition or agenda-setting). This
is not to say that the latter are not worth studying. Rather, we use this specific focus in order
to highlight how problems of measuring the direction and intensity of policy change can be
addressed.
The dimension of policy density describes the extent to which a certain policy area is cov-
ered by governmental activities. It tells us something about the legislative penetration and
internal differentiation of a policy field or subfield. Hence, it measures the extensiveness or
breadth of governmental intervention. Any increase in it indicates policy expansion; any
decrease can be interpreted as policy dismantling.
Changes in policy density can be assessed by two indicators: the number of policy targets
and the number of policy instruments that are applied in a given policy field or subfield. The
number of policy targets offers a more general measurement of the breadth of policy activ-
ity; i.e., the higher the number of policy targets, the broader the policy involvement of a
country in a given sector.
This point can be illustrated by the example of clean air policy – a subfield of environmental
policy. To reduce air pollution, governments might rely on a variety of different measures. On
the one hand, they can regulate the amount of pollutant emissions into the air: policies would
refer to the different substances that are targeted; i.e., there might be a policy on the reduction
of emissions of carbon dioxide. On the other hand, governments can adopt policies that define
minimum requirements for air quality. Again, we can conceive of different measures that pre-
scribe such standards for different pollutants. These types of legislative activities are only
examples that indicate the universe of potential policy options within clean air policy. Any
Copyright © 2020. Macmillan Education UK. All rights reserved.

increase in the number of policy items in a given policy field or subfield indicates an increasing
breadth of policy involvement and hence a higher density of regulation – i.e. policy expan-
sion – regardless of the strictness of the policy requirements. Any decrease, by contrast, means
a lower level of state involvement and hence a dismantling of policies.
The second indicator for assessing policy density is defined by the number of policy instru-
ments that are applied. This measures policy density on a less general level than the number of
policies. This can be traced to the fact that a change in the number of instruments need not
coincide with a change in the number of policies, as a policy can be characterized by different
instruments. Knowing that a certain policy has been adopted tells us nothing about the spe-
cific means by which the intended effects of the policy could be achieved. To reduce the amount
of carbon dioxide emissions into the air, for instance, governments can rely on a broad array
and combination of instruments, including command-and-control approaches (i.e. the defini-
tion of legally binding emission standards), economic incentives (such as environmental taxes
or emission trading systems) or industrial self-regulation and voluntary agreements. Even if
the number of policies in a given field remains constant over time, the number of policy instru-
ments can increase or decrease, implying respective changes in policy density.

Given this approach, it is important to emphasize that not every change in the composi-
tion of policies or instruments has to coincide with dismantling or expansion. Any substitu-
tion of policies or instruments that will not affect the number of the respective measures
cannot be interpreted as a policy change in these terms. For instance, if a command-and-
control instrument is replaced by a market-based approach, we would not interpret this as a
policy change, but rather as a substitution and preservation of the status quo in terms of
policy density.
Policy intensity, by contrast, refers to the level of policy intervention. On the one hand, it
is defined by the settings of the applied policy instruments, i.e. the levels of policy require-
ments. The characterization of changes in instrument settings depends on the nature of the
item in question. For tax rates, for instance, a lowering of the setting implies policy disman-
tling; whereas for environmental pollution standards, the lowering of the maximum permis-
sible limits would be interpreted as policy expansion, since it increases the overall level of
environmental protection.
On the other hand, policy intensity varies with the scope of policy intervention. The scope
of the latter increases with the number of cases, constellations or targets that are covered by a
certain policy. If, for instance, a government altered the conditions for welfare benefits in such
a way that not only national citizens but also foreigners living in the country might apply, this
would mean an increase in policy scope and policy expansion. In a similar vein, the scope of
regulation of carbon dioxide emissions from power plants would increase if the benchmark
(e.g. the generated power per year) for plants falling under the regulation was lowered, imply-
ing that more companies are covered by the respective legislation.
Based on these considerations, we can construct an aggregate measure of changes in pol-
icy density by calculating the cumulative difference between policies introduced and abol-
ished, and adding this information to the cumulative difference between instruments
introduced and abolished. Likewise, a combined assessment of the cumulative difference
between increases and decreases in policy levels and scope allows us to come up with an
aggregate measure of changes in policy intensity. Note, however, that a summative combi-
nation of these two dimensions cannot be achieved, since they address different aspects of
policies. Table 11.1 summarizes our conceptual approach by outlining the dimensions, sub-
dimensions and indicators of policy change.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Table 11.1 Dimensions and indicators of policy change.

Dimension Subdimensions Indicators

Policy Policy target (Cumulative) difference in the number of policy targets which
density density were introduced or abolished
Instrument (Cumulative) difference in the number of instruments which were
density introduced or abolished
Policy Intensity level Development of policy strictness/policy generosity over time
intensity (difference in number and/or degree of measures with increasing
and decreasing effects)
Intensity scope Development of scope of a policy over time (difference in number
and/or degree of measures with increasing and decreasing effects)

Source: Based on Knill et al. (2010: 417).

General Patterns of Policy Change: Policy Expansion versus Policy Dismantling

Various theoretical considerations suggest that policy dismantling and policy expansion
are not equally likely. Rather, we should expect that – independent of the country, pol-
icy area, change indicator and time period in question – instances of expansion should
occur more frequently than those of dismantling (see, e.g., Bauer et al. 2012). The cen-
tral argument behind this is that – for various reasons – the adoption of new as well as
stricter or more generous policies generally meets less political resistance than termi-
nating or cutting back regulations or public services. As stated already by Bardach
(1976: 129), ‘the American political system, like most others, rewards novelty and
innovation’.
The first factor that might reduce the likelihood of dismantling refers to sunk costs:
efforts spent on setting up legal and organizational structures for implementing a cer-
tain programme. Increasing returns on path dependency might lead to constellations
where suboptimal policies are continued rather than replaced (North 1990). Second,
anti-termination or anti-dismantling coalitions consisting of actors who benefit from
the status quo might mobilize strong political resistance (Bardach 1976; deLeon 1997).
The formation of such coalitions is favoured by so-called ratchet effects that emerge
from individuals adjusting, out of habit, more easily to increasing than sinking benefits
(Duesenberry 1949).
A basic expectation, then, is that we should observe a continuous growth of policy den-
sity and policy intensity over time, regardless of the policy field and country under investi-
gation. If the number of instances of expansion is regularly higher than the number of
instances of dismantling, there should be a clear tendency for continuous policy expansion
over time.
In this context, Jensen et al. (2014) show that public spending and policy production
do not always go hand in hand. Protection against social risks is popular amongst vot-
ers, and governments are aware of that. At the same time, public spending on social
policies entails substantive costs, which sometimes are too high to allow for the con-
tinuation of programmes. Therefore, retrenchment occurs rather frequently in advanced
welfare states without this systematically leading to electoral punishment, as govern-
Copyright © 2020. Macmillan Education UK. All rights reserved.

ments practise different types of blame-avoidance strategies (see, e.g., Wenzelburger

and Hörisch 2016). Following Jensen et al. (2014), one blame-avoidance strategy con-
sists of ‘expansionary dismantling’, where new policies and instruments are used to
compensate reform losers and to obfuscate cutbacks (see also Pal and Clark 2015).
Steinebach and Knill (2018) show that the choice of governmental strategies of policy
dismantling and policy expansion during economic crises is crucially affected by the
party in government and the type of government. Their analysis reveals not only that
dismantling becomes a more likely scenario in crisis times, but also that dismantling is
particularly likely for right-wing governments in consensual and left-wing governments
in majoritarian democracies. These findings indicate that it either needs consensus and
broad coalition governments, or support from and promotion by left-wing parties, for
dismantling to become a valid policy option under the impression of economic hardship
(Box 11.2).

Box 11.2 Policy accumulation and the democratic responsiveness trap

Knill et al. (2018) argue that modern democracies are potentially caught in a respon-
siveness trap. On the one hand, it is a major asset of democratic governments that
they are responsive to societal demands. Citizens clamour for cleaner environments,
better social protection or more and better education. Governments typically
respond to these demands by adopting new policy outputs. As existing policies are
dismantled or terminated only very rarely, over time policy outputs continuously
pile up in modern democracies. In many ways, this accumulation of public policies is
a direct reflection of societal modernization and progress. Most citizens are happy
that they do not live in a country that still trusts in the social policy portfolio of the
1870s, or the environmental policy portfolio of the 1950s. On the other hand, merely
adopting new policies reflects nothing but symbolic politics as long as the respective
policy outputs do not also reduce the problems that they are supposed to solve. This
presupposes – as a minimum requirement – that governments have sufficient admin-
istrative capacities for implementing an ever-growing stock of policies. Policy accu-
mulation hence directly translates into the accumulation of administrative
implementation burdens. Yet governments face fundamental fiscal and ideological
constraints on public-sector expansion in times of globalized financial markets and
austerity. Consequently, there is the risk of an increasing gap between accumulating
policies and stagnating or even declining implementation capacities. This scenario
indicates the potential responsiveness trap of modern democracies. The more gov-
ernments respond to societal demands via policy accumulation, the more they face
the risk of ‘empty policy growth’, or symbolic politics that result from an increasing
mismatch between implementation burden and implementation capacities, and
hence ineffective policy responses that undermine the legitimacy of democratic sys-
tems (Adam et al. 2017, 2018, 2019).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Key points
•• Theories of policy change can be divided into three basic approaches: (1) models that ana-
lyse policy change in terms of the changing preferences and beliefs of the actors involved,
(2) models analytically linking institutional arrangements and policy change and (3) models
emphasizing the role of chance.
•• The measurement of policy change is a complex task and is confronted with several chal-
lenges, including (1) selection problems regarding the time period under investigation, (2)
the benchmarks for assessing the degree of change and (3) the concepts and indicators for
measuring the direction of change. These challenges also offer opportunities for innovative
research.
•• We suggest policy density and policy intensity as concepts for assessing policy change.
Policy density describes the extent to which a certain policy area is covered by govern-
mental activities. Policy intensity refers to the level of policy intervention.
•• The general expectation on the aggregate direction of change is that policy expansion should be
more likely than policy dismantling.

ROSS-NATIONAL POLICY CONVERGENCE: CONCEPT, MEASUREMENT

C
AND DIMENSIONS
So far, we have analysed key issues concerning policy change from a purely country-specific
perspective. We have presented different concepts and theoretical approaches that seek to
describe and explain the occurrence, degree and direction of policy change in a given coun-
try at a certain point in time. The focus on individual countries, however, does not allow for
systematically comparing the dynamics of policy change across countries. What are the con-
sequences of national change in comparison with developments in other countries? Do
national policy changes entail growing policy similarity across countries and hence policy
convergence? Or can we observe an opposite pattern of policy divergence, with national
policies moving further apart over time?
With regard to these questions, there is a long-standing scholarly debate as to
whether national public policies become more similar over time. On the one hand, sev-
eral studies emphasize a clear trend towards policy convergence, implying the develop-
ment of similar or even identical policies across countries (for an overview, see Heichel
et al. 2005). For example, there is empirical evidence that national environmental policy
arrangements have become increasingly similar over time (see, e.g., Dobbins and Knill
2014; Holzinger et al. 2008a, 2008b; Jörgens et al. 2014; Jahn 2016). On the other
hand, there are many studies – typically in the tradition of the new institutionalism –
that modify or even challenge the general expectation of cross-national policy conver-
gence. Emphasizing important differences in national institutions and opportunity
structures for domestic actors, these studies expect diverging or at least parallel, rather
than converging, policy developments across countries. For instance, cross-national
policy convergence is of a notably limited magnitude in the case of welfare state arrange-
ments (Esping-Andersen 1990; Starke et al. 2008), but it is still observable (Schmitt and
Starke 2011; Obinger et al. 2013).
In the discussion on policy convergence, reference is often made to closely related con-
cepts like policy diffusion, policy transfer and institutional isomorphism. So we first have to
differentiate convergence from these concepts, before focusing on different types of policy
convergence that are distinguished in the literature. In a third step, we need to show that, in
Copyright © 2020. Macmillan Education UK. All rights reserved.

addition to these different types, different dimensions of convergence can be identified,

which refer to the degree, direction and scope of convergence (see Knill 2005; Holzinger and
Knill 2005; Holzinger et al. 2007, 2008a, 2008b).

elated, but Different: The Concepts of Policy Convergence, Policy Transfer,

R
Policy Diffusion and Institutional Isomorphism
Policy convergence is generally defined as increasing policy similarity over time (Kerr 1983:
3). Notwithstanding this broadly accepted understanding, policy convergence is often
equated with partially overlapping concepts, such as isomorphism, policy transfer or policy
diffusion. While policy convergence is related to results, policy transfer is concerned with
processes and describes a development that might, but need not, lead to cross-national pol-
icy convergence. Policy transfer is not restricted to merely imitating the policies of other
countries, but can include profound changes in the content of the exchanged policies (for an

overview, see Marsh and Evans 2012; Stone 2012). Accordingly, Dolowitz and Marsh (2000:
5) define policy transfer as ‘processes by which knowledge about policies, administrative
arrangements, institutions and ideas in one political system (past or present) is used in the
development of policies, administrative arrangements, institutions and ideas in another
political system’.
Similar to transfer, policy diffusion refers to processes (rather than effects) that might
result in increasing policy similarities across countries, hence leading to policy conver-
gence (Elkins and Simmons 2005: 36). Diffusion is generally defined as the socially medi-
ated spread of policies across and within political systems, including communication and
influencing processes which operate both on and within the populations of those who
adopt the policies (Rogers 2003: ch. 2). Diffusion studies typically start out from the
description of adoption patterns for certain policy innovations over time (Marsh and
Sharman 2009; Berry and Berry 2017). In a subsequent step, they analyse the factors that
account for the empirically observed spreading process (for an overview, see Maggetti and
Gilardi 2016).
Policy diffusion and policy transfer share a number of similarities, such as their focus on
process, which can make it difficult analytically to distinguish them from one another (see
Karakhanyan et al. 2011). Yet they are conceptually distinct inasmuch as they have differing
empirical foci and dependent variables. Diffusion studies typically start out from a rather
general perspective and concentrate on patterns. They often reveal a rather robust adoption
pattern, with the cumulative adoption of a policy innovation over time following an S-shaped
curve (Gray 1973). Relatively few countries adopt an innovation during the early stages.
Over time, the rate of adoption increases, until the process gets closer to saturation and
then slows down again. Analyses of policy transfer investigate the underlying causes and
contents of singular processes of bilateral or multilateral policy exchange, often by using
process tracing (Marsh and Sharman 2009).
From these considerations, it also follows that policy transfer and policy diffusion dif-
fer from policy convergence in important ways. First, differences exist with respect to
their underlying analytical focus. While diffusion and transfer are concerned with process
patterns, convergence studies place a particular emphasis on effects. Transfer and diffu-
sion thus reflect processes that under certain circumstances might result in policy conver-
Copyright © 2020. Macmillan Education UK. All rights reserved.

gence. This does not imply, however, that the empirical observation of converging policies
must necessarily be the result of transfer or diffusion (Drezner 2001). It is conceivable
that policy convergence is the result of similar but relatively isolated domestic events.
Second, the concepts differ in their dependent variable. Convergence studies typically
seek to explain changes in policy similarity over time. By contrast, transfer studies inves-
tigate the content and process of policy transfer as the dependent variable, while the focus
of diffusion research is on the explanation of adoption patterns over time (Elkins and
Simmons 2005; Jordana and Levi-Faur 2005; Gilardi 2008; Marsh and Sharman 2009;
Maggetti and Gilardi 2016).
The particular focus underlying the analysis of policy convergence places it in close prox-
imity to the concept of isomorphism, which has been developed in organizational sociology
(see Chapter 4). The central question underlying studies on isomorphism refers to the rea-
sons which lead organizations to become more similar over time. For example, Bellé (2010)
demonstrates that the adoption of performance-related pay in Italian municipalities can be
explained by isomorphism. Another area where isomorphism has been discussed

Table 11.2 Policy convergence and related concepts.

Policy
Dimensions convergence Isomorphism Policy transfer Policy diffusion
Analytical Effects Effects Process Process
focus
Empirical Policy Organizational Policy characteristics Policy
focus characteristics structures characteristics
Dependent Similarity Similarity change Transfer content; Adoption
variable change transfer process pattern

Source: Knill (2005: 768).

extensively is the spread of quality assurance regimes in higher education. In this context,
Jarvis (2014) claims that the emergence of a regulatory regime for quality assurance in the
Hong Kong higher education sector is an outcome of coercive isomorphism.
There is thus a broad overlap between studies on policy convergence and those on
isomorphism, with the major difference between the two concepts constituting their
empirical focus. The literature on isomorphism concentrates on the increasing similar-
ity of organizational and institutional structures and cultures. Studies on policy conver-
gence, transfer or diffusion, by contrast, focus on changes in national policy characteristics
(see Table 11.2).
Following these considerations, policy convergence can be defined ‘as any increase in the
similarity between one or more characteristics of a certain policy (e.g. policy objectives,
policy instruments, policy settings) across a given set of political jurisdictions (suprana-
tional institutions, states, regions, local authorities) over a given period of time. Policy con-
vergence thus describes the end result of a process of policy change over time towards some
common point, regardless of the causal processes’ (Knill 2005: 768).
Copyright © 2020. Macmillan Education UK. All rights reserved.

Types and Measurement of Policy Convergence

In the literature, four basic convergence types are distinguished which are associated with
different concepts of measurement, namely sigma (σ), beta (β), gamma (γ) and delta (δ)
convergence (see Sala-ì-Martin 1996; Heichel et al. 2005; Holzinger et al. 2008a; Jörgens
et al. 2014). The most common type is σ-convergence. Following this approach, convergence
occurs if there is a decrease in variation of policies amongst the countries under consider-
ation over time. Measures of σ-convergence are usually based on the total variation, stan-
dard deviation, measures of association and algebraic distances. These measures are applied
in both quantitative and qualitative designs. With regard to the latter, Pierson and Humpage
(2016), for instance, show that Australia, New Zealand and the United Kingdom have con-
verged in their welfare policies, with the most noticeable degree of convergence in ‘welfare-
to-work’ policies.
The study of β-convergence focuses on mobility. This form occurs when laggard coun-
tries catch up with leader countries, implying, for instance, that the former strengthen

their regulatory standards more quickly and fundamentally than the latter (Sala-ì-
Martin 1996). In the literature, this mode of policy convergence is also discussed under
the heading of ‘catching up’. The volume edited by Jörgens et al. (2014) gives a number
of examples of where β-convergence occurred and led to the adoption of more ambitious
environmental protection standards by countries that originally had low regulatory
standards.
β-convergence only allows for the identification of processes of catching up, but gives
no information on the extent to which catching up is accompanied by processes of over-
taking between countries. In response to this problem, the concept of γ-convergence is
applied (Boyle and McCarthy 1999). For the analysis of γ-convergence, country rankings
for different points in time are compared to assess the pattern, and not just the speed, of
mobility of countries. If countries in the first ranks fall behind or countries in the last
ranks catch up over time, convergence occurs. Policy change is analysed by simple mea-
sures of association. A low degree of similarity between rankings indicates high mobility
of countries over time. The concept of γ-convergence adds an additional perspective to the
study of policy convergence, as it may occur where other approaches do not detect changes.
Country rankings may change without a significant decrease in cross-country variation
(Holzinger et al. 2011). Again, the analysis of various environmental policies in the vol-
ume edited by Jörgens et al. (2014) offers an overview of cases where γ-convergence could
be observed.
In contrast to the measurement of homogeneity and mobility changes, the final conver-
gence type concentrates on changes in the distance between policy outputs and an exem-
plary policy model, e.g. a model promoted by an international organization or a
front-runner country. Based on the algebraic notation of ‘distance’ (δ), this concept is
referred to as δ-convergence (Heichel et al. 2005). With its reference to an exemplary
policy model or benchmark, δ-convergence adds aspects of the direction of convergence to
the analysis (Heichel et al. 2008: 83; see Table 11.3). Noy and Sprague-Jones (2016), for
instance, investigate trends and convergence in public health spending in Latin America
and the Caribbean compared with Organisation for Economic Co-operation and
Development (OECD) countries. The authors find a period of δ-convergence followed by
divergence in Latin America.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Table 11.3 Four forms of policy convergence.

Convergence type Concept Measurement

σ-convergence Decrease in the variation of national Changes in total variation or
policies standard deviation
β-convergence Mobility of policies Group-wise comparison of changes
in total variation or standard
deviation
γ-convergence Changes in country performance Association measures
rankings
δ-convergence Changes in the distance between policy Algebraic distances
outputs and an exemplary policy model

Convergence Dimensions
Even if we restrict our focus to the conception of convergence as the growing similarity of
policies over time, i.e. σ-convergence, this still leaves a broad range of options for how to
assess empirically and evaluate increasing or decreasing policy similarity (Heichel et al.
2005; Holzinger and Knill 2005: 778–9). Three dimensions can be distinguished: the degree,
direction and scope of convergence. With regard to degree, the focus is on the extent to
which we can actually observe an increase in policy similarity across countries over time. In
other words, if the degree of convergence is higher, the more the variation across countries
has decreased over time.
The direction of convergence, by contrast, indicates the extent to which convergence
coincides with an upward or downward shift of the mean policy position from time t1 to
t2. Convergence at the top or bottom presupposes therefore both a decrease in variation
and a shift of the mean (Botcheva and Martin 2001: 4). The direction of convergence is
usually related to the extent of state intervention or to the strictness of regulation. Lax
standards or laissez-faire policies are identified with the ‘bottom’, strict standards or
interventionist policies with the ‘top’ (Drezner 2001: 59–64). However, the direction of
convergence can only be measured when the policies under consideration can be viewed
as ‘scalar’, or can be measured, which can be associated with a normative judgement on
the extent of an intervention. Typical examples are the levels of environmental and
consumer protection or the level of welfare benefits. Nevertheless, it is not always easy
to identify what the top and the bottom are in a policy, because this may vary in the
light of different value judgements. For example, in media regulation there are two com-
peting goals: restricting harmful content on the one hand and promoting freedom of
information on the other hand (Holzinger and Knill 2005: 779).
The scope of convergence describes the range of countries that are potentially affected by
a certain convergence factor. For example, if we talk about compliance with international
law, we would not expect any convergence effects on countries which are not members of the
international institutions in which harmonization efforts take place. The scope of conver-
gence increases with the number of countries and policies that are actually affected by a
certain factor, with the reference point being the total number of countries and policies
Copyright © 2020. Macmillan Education UK. All rights reserved.

under investigation (see Table 11.4).

Key points
•• Policy convergence is defined as an increase in the similarity between one or more character-
istics of a certain policy across a given set of political jurisdictions over a given period of time.
•• There are three concepts that are closely related to policy convergence: policy transfer, policy
diffusion and isomorphism.
•• The literature distinguishes between four basic convergence types which are associated with
different concepts of measurement: sigma (σ), beta (β), gamma (γ) and delta (δ). These con-
vergence types entail distinctive analytical perspectives on the same empirical phenomenon,
and hence might also yield different evaluations of convergence or divergence.
•• In addition to different types, convergence can also be assessed along different analytical
dimensions, namely degree, direction and scope.

Table 11.4 Dimensions of policy convergence.

Convergence dimension Research question Measurement

Convergence degree How much convergence can Changes in total variation or standard
be observed? deviation (for σ-convergence)
Convergence direction Where does convergence δ-convergence
occur (top or bottom)?
Convergence scope Which countries and policies Relationship of affected countries/policies
are affected? in comparison with the whole sample

AUSES AND CONDITIONS OF CROSS-NATIONAL POLICY

C
CONVERGENCE
The literature on convergence and its related concepts offers a broad range of causal factors
in order to explain changes in the similarity of policies across countries. At a very general
level, these factors can be grouped into two categories: (1) causal mechanisms triggering
convergent policy changes across countries, and (2) facilitating factors that influence the
effectiveness of these mechanisms (Knill 2005).

Causal Mechanisms
With respect to causal mechanisms, five central factors can be found in the highly diverse
literature (see Bennett 1991; DiMaggio and Powell 1991; Dolowitz and Marsh 2000; Drezner
2001; Hoberg 2001; Knill 2005; Holzinger and Knill 2005, 2008; Braun and Gilardi 2006;
Holzinger et al. 2011; Schmitt and Starke 2011; Obinger et al. 2013). First, cross-national
policy convergence might simply be the result of similar but independent problem-solving of
different countries when responding to parallel pressure from problems (e.g. ageing of soci-
eties, environmental pollution or economic decline); i.e. policy convergence is caused by
countries reacting to similar policy problems. Just as individuals open their umbrellas
Copyright © 2020. Macmillan Education UK. All rights reserved.

simultaneously during a rainstorm, governments may decide to change their policies in the
presence of tax evasion, environmental pressures such as air pollution or an ageing popula-
tion. As Bennett notes, the analyst of policy convergence ‘must avoid the pitfall of inferring
from transnational similarity of public policy that a transnational explanation must be at
work’ (1991: 231).
Second, emphasis is placed on convergence effects emerging from compliance with interna-
tional law. As shown in Chapter 10, there are several reasons that favour the setting up of
international organizations and regimes, including economic interdependencies and prob-
lems of transnational scope that cannot be effectively addressed by unilateral action.
International cooperation typically involves the adoption of common rules by signatory
countries; i.e. the member states are legally obliged to comply with international law. While
the specific design of international rules is of course determined in negotiations between
member countries (rather than hierarchically defined by international organizations), these
rules, once adopted, constitute an important source of domestic policy convergence.

The degree to which compliance with international law triggers cross-national convergence
varies with the legal specification of international law. Specificity is particularly high if inter-
national law requires the harmonization of national standards. Compliance with European
law is such an example and lends itself to studying policy convergence (see, e.g., Kemmerling
2010; Nohrstedt and Hansén 2010). Convergence effects are less pronounced, by contrast, if
legal rules are defined in a less rigid way, leaving member states broad leeway for selecting
appropriate instruments to comply with international policy objectives.
Third, several studies emphasize convergence effects stemming from the imposition of
policies. This refers to constellations where countries or international organizations force
other countries to adopt certain policies by exploiting asymmetries in political or economic
power. While legal requirements are the crucial mechanism linking international policy-
making and national policy change, the domestic effects of imposition are based on asym-
metries of political power rather than legal obligation.
A typical scenario of imposition is the conditionality enforced by international organiza-
tions. First, conditionality can be based on the exchange of policies for loans. For instance,
it is argued that the spread of neo-liberal monetary and trade policies to developing coun-
tries was driven by such forms of conditionality. Governments pressed by international
financial institutions switched to liberal trade regimes (see Dolowitz and Marsh 1996: 347).
In a similar vein, the International Monetary Fund (IMF) required fundamental policy
reforms in Greece as a precondition for granting monetary support in the aftermath of the
global financial crisis of 2008 (see, e.g., Armingeon and Baccaro 2012). A second form of
conditionality is the exchange of policy adjustments for membership in international insti-
tutions. The EU accession of Central and Eastern European countries has been governed by
this principle (Schimmelfennig and Sedelmeier 2004; Knill and Tosun 2009; Tosun 2013a).
Those Central and Eastern European countries that wished to join the EU had to comply
with the acquis communautaire, i.e. the entirety of European law, and with the so-called
Copenhagen criteria, which require that a state has the institutions to preserve democratic
governance and human rights and has a functioning market economy in place. As a result,
in the pre-accession period agenda-setting and policy-drafting were directly affected by the
EU in these countries (Tosun 2013a).
Sometimes, it might be difficult to draw a clear analytical borderline between interna-
Copyright © 2020. Macmillan Education UK. All rights reserved.

tional negotiations characterized by symmetrical relations of power and those characterized

by asymmetrical ones. In negotiations at the international or European level, it will almost
always be the case that some states are more influential than others. It is hence not always
easy to decide whether the power constellations imply a voluntary or an imposed agreement
from the perspective of individual countries (Holzinger and Knill 2008: 41).
Fourth, regulatory competition emerging from the increasing economic integration of
European and global markets has been identified as an important factor that drives the
mutual adjustment of policies across countries. Besides legal obligations and political power,
internationalization can trigger national policy change via the route of market forces. In
general, theories of regulatory competition predict that countries adjust policy instruments
and regulatory standards in order to cope with competitive pressures emerging from inter-
national economic integration. The more exposed a country is to competitive pressures fol-
lowing high economic integration (emerging from its dependence on the trading of goods,
capital and services with other countries), the more likely it is that its policies will converge
with those of other states with international exposure. In other words, the degree of conver-
gence depends on the level of competitive pressures to which countries are exposed. Drawing

Box 11.3 Environmental policy convergence in Europe

The question of whether and to what extent different international factors trigger
convergence in environmental policy has been systematically addressed by the
ENVIPOLCON project (see Knill 2005; Holzinger et al. 2008a, 2008b, 2011; Holzinger
and Knill 2008; Jörgens et al. 2014). The empirical findings of this project strongly
suggest that environmental policy converged in Europe in the period 1970–2000. On
the one hand, environmental policies have on the whole grown more alike
(σ-convergence), though at the same time they have moved in an ‘upward’ direction,
thus becoming stricter. Hence, a ‘race to the bottom’ due to regulatory competition –
i.e. a lowering of environmental standards by countries as a consequence of engaging
in competitive markets, as often referred to in the literature – does not appear to have
taken place.
This astonishing rate of cross-national policy convergence can be explained as the
result of the effects of two international influences. First, the research findings pro-
vide evidence that compliance with international law constituted an important driv-
ing force for policy convergence. Second, communication within transnational
networks also turns out to have been a major driving force. In contrast, competition
effects play a negligible explanatory role. Regulatory competition has not led to envi-
ronmental races to the bottom.
The project findings also show that environmental leader countries are able to pull
along the laggards. This applies, first, to environmental standard-setting through
international law. The establishment of legally binding agreements at the interna-
tional level typically implies that low-regulating countries adjust their standards to
the level of the environmental leader countries. In other words, the leaders are gener-
ally able to set the pace in international environmental harmonization. In addition,
this effect is also relevant in the absence of legally binding agreements. Communication
and information exchange alone can induce laggard countries to raise their standards
as they seek to avoid being blamed as ‘pollution havens’.
Copyright © 2020. Macmillan Education UK. All rights reserved.

on this logic, Jensen (2011), for instance, shows that social spending patterns have con-
verged due to economic globalization and competitive pressure (Box 11.3).
Finally, cross-national policy convergence can simply be caused by transnational communi-
cation. Under this heading, there are several mechanisms which all rest on communication
and information exchange amongst countries (see Holzinger and Knill 2005, 2008). They
include lesson-drawing (where countries deliberately seek to learn from successful problem-
solving activities in other countries), joint problem-solving activities within transnational
elite networks or epistemic communities, the promotion of policy models by international
organizations with the objective of accelerating and facilitating cross-national policy trans-
fer and the emulation of policy models.
The basic assumption here is that information exchange and common problem-solving
activities at the international level favour cross-national policy learning and so alter the
beliefs and expectations of domestic actors (Holzinger and Knill 2005: 782). One could
certainly argue that communication is also of relevance with regard to the other mecha-
nisms of imposition, international harmonization or regulatory competition. In these cases,

Table 11.5 Causes of cross-national policy convergence.

Factor Mechanism Effect

Independent Parallel problem Independent similar response
problem-solving pressure
Compliance with Legal obligation Direct ‘prescription’ of policy change
international law
Policy imposition Political power Direct ‘prescription’ of policy change
Regulatory competition Economic pressure Indirect change via redistribution of power
between actors
Transnational Learning Indirect change via alteration of actor beliefs
communication

Source: Holzinger and Knill (2008a: 42).

however, communication and information exchange are basically a background condition

for the operation of the mechanisms rather than the central factor that actually triggers
convergence.
The extent to which communication in transnational networks has repercussions for
national policies is affected by various factors. First, the frequency of interaction at the
international level is of importance. The more regularly national actors meet, the higher the
chances for policy learning. International organizations and regimes display considerable
variance in this regard. The same holds true for the second factor: organizational differentia-
tion and the institutionalization of information exchange. The higher the number of forums
and bodies in which national policy-makers and experts can meet, the higher the learning
potential (Heichel et al. 2008: 91; see also Table 11.5).

convergence? The first group of factors refers to characteristics or, more precisely, the
similarity of the countries under investigation. It is argued that converging policy devel-
opments are more likely for countries that are characterized by high institutional similar-
ity. Policies are transferred and properly implemented only insofar as they fit with existing
institutional arrangements (see Knill and Lenschow 1998; Knill 2001). In other words,
the adoption of similar policies across countries varies with the compatibility between
transnational concepts and domestic policy legacies. The degree of expected convergence
will decrease with the cost of adopting the policy concept in question (Knill 2001). The
same scenario applies to constellations in which adoption of the new policy entails high
economic costs or is likely to face strong political opposition (e.g. in the case of strong
redistributive effects amongst coalitions of national actors). Cultural similarity also plays
an important role in facilitating cross-national policy transfer. In their search for relevant
policy models, decision-makers are expected to look to the experiences of those countries
with which they share an especially close set of cultural ties (Strang and Meyer 1993).
Finally, similarity in socio-economic structures and development has been identified as a fac-
tor that facilitates the transfer of policies across countries (for instance on environmental
policy, see Jänicke 1988).

The second group of facilitating factors that can be analytically distinguished is com-
posed of the characteristics of the underlying policies. In this context, the type of policy
has been identified as a factor that influences the likelihood of convergence. The expecta-
tion is that policies involving high distributional conflicts between coalitions of domestic
actors will diffuse and so converge to a lesser extent than policies with comparatively
small redistributional consequences. A second argument about the impact of policy char-
acteristics on convergence concentrates on different policy dimensions. As mentioned
above, Hall (1993), for instance, distinguishes between policy paradigms, policy instru-
ments and settings, arguing that change (and consequently convergence) is most difficult
in relation to goals, given their deep embeddedness in the dominant beliefs of domestic
actors. Instruments and even more settings, by contrast, can be adjusted without neces-
sarily demanding ideational change; hence, convergence on the latter dimensions is more
likely than on paradigms. This view, however, is not uncontested in the convergence litera-
ture (see Lenschow et al. 2005; Radaelli 2005).

Key points
•• There are five major causes of national policy change: parallel problem pressure, compliance with
international law, policy imposition, regulatory competition and transnational communication.
•• The effects of these factors on cross-national policy convergence are conditioned by several factors,
including the institutional, cultural and socio-economic similarity between countries, as well as the
type and dimension of the underlying policy.

CONCLUSIONS

In this chapter we have focused on two central and related topics within the study of public pol-
icy, namely policy change and policy convergence. We have addressed the phenomenon of policy
change and have seen that there exists a broad range of different theoretical frameworks that are
used to account for policy change and policy stability. These approaches analyse and interpret
change from different analytical perspectives and hence should be viewed as complementary
Copyright © 2020. Macmillan Education UK. All rights reserved.

rather than contradictory. With regard to the assessment of policy change, we have made clear
that the researcher must pay attention to several issues in order to attain a correct assessment of
the characteristics of change. Most importantly, the findings are contingent on the observation
period. In addition, the evaluation of the degree of change is affected by the analytical level of
abstraction. We also emphasized the need to disaggregate the concept of policy change into pol-
icy density and stringency in order to provide an accurate empirical account. Finally, we have
explained why there is a differing expectation with regard to the likelihood of the occurrence of
policy expansion and dismantling. For a number of reasons, policy dismantling – and in the most
extreme case policy termination (see, e.g., Geva-May 2004) – is less likely than policy expansion.
From this it follows that the general expectation regarding policy change is one of policy expan-
sion. Regardless of the type of policy change, political reform agendas are more likely if policy-
makers, inter alia, compensate losers (see, e.g., Pal and Clark 2015).
In addressing the question of the potential aggregate effects of national policy develop-
ments, we focused on the extent to which national policy change can lead to cross-national
policy convergence. To answer this, we first discussed the concept of policy convergence and
how it relates to similar, partially overlapping concepts like diffusion, transfer and isomor-
phism. Second, we introduced the four major types of policy convergence, namely sigma (σ),

beta (β), gamma (γ) and delta (δ) convergence. Third, we provided insights into different
convergence dimensions (degree, direction and scope). We argued that these three dimen-
sions are helpful for further characterizing processes of policy convergence.
Finally, we turned to factors affecting the degree of cross-national policy convergence,
distinguishing between causal mechanisms (parallel problem pressure, international har-
monization, policy imposition, regulatory competition and transnational communication)
and facilitating factors (institutional, cultural and socio-economic similarity between coun-
tries, as well as policy type and policy dimension).

WEB LINKS

airlex.web.ua.pt. A worldwide database of air quality policies and legislation.

www.commonwealthfund.org. The Commonwealth Fund provides data on health care
issues and publishes the World Care Report.
www.devpolicy.org. The Devpolicy Blog provides information aid and development
policy with a focus on Australia, the Pacific and Papua New Guinea.
www.ourworldindata.org. Our World in Data makes worldwide challenges visible by
offering country-comparative data on them.
www.rug.nl/ggdc/productivity/pwt. The Penn World Table provides national accounts
data as provided by the University of California, Davis and the University of Groningen.
www.social-inclusion-monitor.eu/about-sim-europe/reform-barometer. The Reform
Barometer monitors policy reforms that aim to strengthen social cohesion in Europe.

FURTHER READING

Bianculli, A. C. and Hoffmann, A. R. (eds.) (2016). Regional Organizations and Social
Policy in Europe and Latin America: A Space for Social Citizenship? London: Palgrave.
This book adopts a stimulating approach to the study of policy diffusion and policy
convergence by systematically comparing two regions.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Elhiraika, A., Mukungu, A. and Nyoike, W. (eds.) (2016). Regional Integration and Policy
Challenges in Africa. London: Palgrave Macmillan. Provides a detailed and informative
analysis of how regional integration has affected the fiscal policies of African countries.
Holzinger, K., Knill, C. and Arts, B. (eds.) (2008). Environmental Policy Convergence in
Europe? The Impact of International Institutions and Trade. Cambridge: Cambridge
University Press. The studies brought together in this edited volume all represent
important contributions to the study of policy convergence.
Jörgens, H., Lenschow, A. and Liefferink, D. (eds.) (2014). Understanding Environmental
Policy Convergence: The Power of Words, Rules and Money. Cambridge: Cambridge
University Press. This is the case study extension of the ENVIPOLCON project.
Peters, B.G. and Fontaine, G. (eds.) (2020). Handbook of Research Methods and Applications
in Comparative Policy Analysis. Cheltenham: Edward Elgar. A very useful collection of
methods that can be used to study policy change and policy convergence.
Weil, S. (2016). Lobbying and Foreign Interests in Chinese Politics. New York: Palgrave
Macmillan. An elucidating book about the strategies of European and US business
groups to bring about policy change in China.

Reader’s guide
In this chapter we pursue four goals. The first is to summarize the main insights
gained from the analysis of the various themes in policy analysis and to bring together
the individual threads covered by this book. The second goal is to highlight the merit
of comparative research on public policy. The next two sections adopt a different per-
spective and seek to offer practical advice. In this context, the third goal we strive to
attain in this chapter is to offer the readers guidance on how they can find research
questions for their own analyses. The fourth goal is to make a bridge into the practical
application of the analytical skills conveyed in this book and to offer some guidance
on doing policy analysis.

WHAT HAVE WE LEARNED?

Policy-making is a complex process, which can best be defined by giving an answer to ‘who
gets what, when and how’ in social life, as Lasswell (1936) famously put it in his path-break-
ing book. For an improved understanding of the causal mechanisms that underpin policy-
making and how these work in different (country or policy sector) contexts, policy studies
usually focus on single stages of the policy-making process: problem definition and agenda-
Copyright © 2020. Macmillan Education UK. All rights reserved.

setting, policy formulation and adoption, policy implementation, and policy evaluation.
We have learned that the possibility of influencing the agenda – along with defining the
nature of a social problem by means of framing – is an important source of power, as
decision-making grants an advantage to those who address a problem and propose a solu-
tion to it first: those known as first movers. Due to this characteristic, several groups of
actors compete with one another in order to be able to set the agenda in accordance with
their preferences: politicians, bureaucrats, citizens and civil society groups and experts. In
addition, international organizations affect agenda-setting. Access to the media is one of
the factors that explains which groups dominate the agenda-setting stage. There are occa-
sions when bureaucrats and politicians are most influential in agenda-setting, and then
there are occasions when civil society or interest groups are successful in placing an issue on
the agenda. Furthermore, there exist situations when a problem is so visible or pressing that
the general public puts it on the agenda – this scenario is most likely in the presence of a
focusing event such as a disaster or political scandal (see Birkland 2006, 2010). The
Comparative Agendas Project brings together expertise on many countries and a wide range

231
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:17:21.
232 Public Policy

of policy areas, with a view to uncovering the sources, patterns and outcomes of agenda-
setting efforts (Baumgartner et al. 2008b; Green-Pedersen and Walgrave 2014).
With respect to policy formulation, in most political systems the executive and the min-
isterial bureaucracy feature as important actors, whereas with policy adoption executive–
legislative relations come to the fore. The formulation of a policy involves legal, technical
and procedural knowledge as it is accumulated in ministries and ministerial departments.
When examining the adoption of a policy proposal – in its original or an amended form –
public policy can benefit from concepts, theories, methods and data offered by the discipline
of comparative politics (see Caramani 2020a) or more specifically by legislative studies (see,
e.g., Martin and Vanberg 2011). In a nutshell, policy adoption depends on the willingness of
the actors involved to negotiate and to search for compromise, which gets more complicated
the more the political power is fragmented. An indicator of the fragmentation of political
power is the number of institutional veto points (Immergut 1990, 2008) and veto players
(Tsebelis 2002; König et al. 2010) in a political system. Conversely, policy adoption is less
challenging in political systems in which power is concentrated in the hands of a few actors.
Nevertheless, empirically we can observe that democracies – even those with complex insti-
tutional setups and fragmented power – manage to produce effective solutions to social
problems. A case in point is the improvement of environmental quality (in terms of reduced
environmental pollution) ever since this policy field became institutionalized in the 1960s
and 1970s.
Once the government has taken a decision on a public policy, the post-decision phase
(Jordan 1997) begins, which is about putting the policy concerned into action.
Implementation research is the strand of policy studies that examines this phase and
whether there are gaps between the policy as it is written in the book and how it is applied
in reality. In the broadest sense, with regulatory policies such as environmental policy, pol-
icy implementation involves compliance with the policy’s stipulations by the target group
and monitoring and enforcement by public entities (Tosun 2013a). With (re)distributive
policies such as social policy, policy implementation includes service delivery by the state,
such as labour market counselling or the provision of welfare benefits (Dingeldey 2007). The
experience people have when receiving a social service has important implications for trust
or confidence in government (Christensen and Laegreid 2005; Tosun and Shore 2019),
Copyright © 2020. Macmillan Education UK. All rights reserved.

which illustrates nicely the link between public policy, public administration and political
sociology. The choice of policy instruments, the policy design, the control structure, the
institutional design, the administrative capacity and social acceptance are important factors
for explaining how public policies are put into practice and to what extent – if any – the poli-
cies implemented deviate from the original policy with its intended outcomes and impacts.
Considerations regarding policy implementation are also an important aspect of the rich
body of literature on governance. Governance recognizes the participation of public and
private actors in policy-making as well as non-hierarchical forms of policy-making, which
include network or market governance (Knill and Lehmkuhl 2002a, 2002b; Kohler-Koch
and Rittberger 2006). Market governance is characterized by formally equal actors that
‘compete’ against each other concerning the best performance in policy-making, i.e. to be
regarded as having established the best practice. Network governance is about communica-
tion in negotiating systems between public and private actors. The symmetrical relations of
networks are constituted by mutual resource dependencies and/or informal norms of equal-
ity. The literature has shown that non-hierarchical forms of governance often exist in the
‘shadow of hierarchy’. For example, Börzel (2010) makes a plea for recognizing that any

interaction between public and private actors is affected by rules of the game defined, moni-
tored and, if necessary, enforced by public actors.
In the evaluation stage, the floor is opened to different types of ‘experts’ and their
appraisal of whether a policy performs well or poorly. Essentially, policy evaluation is a
systematic process for assessing the content, implementation and impact of public poli-
cies. In other words, policy evaluation is an enterprise that aims to offer a systematic
assessment of the effectiveness or efficiency of policy interventions, and to determine
their merit for improving the economic or social conditions of the policy target groups or
different stakeholders. To this end, evaluation studies use a wide range of social science
research methods, including various qualitative and quantitative techniques. Policy evalu-
ation begins with a description of the policy concerned, and then moves on to developing
an evaluation design, the collection and analysis of data, the interpretation of the find-
ings, and formulation of the conclusions to be drawn from the evaluation exercise.
Reflecting on the outcome of a policy evaluation, policy-makers can decide whether the
measure in question is effective or not. In the latter case, one of the consequences for
policy-makers can be to propose policy change.
We have seen that policy change – i.e. the adoption of a new policy or the amendment
or repeal of an existing one – has and continues to receive significant attention amongst
policy scholars. There exists a broad range of different theoretical frameworks to
account for policy change and policy stability. Most theories of the policy process
(Weible and Sabatier 2017), for example, strive to explain policy change (Knill et al.
2015; Tosun and Workman 2017). The interest in policy change is particularly visible
with the Advocacy Coalition Framework (Sabatier and Jenkins-Smith 1993), whereas
the Punctuated Equilibrium Theory (Baumgartner and Jones 1993/2009) predomi-
nantly concentrates on patterns of stability interrupted by rare events of change (for an
overview, see Weible and Sabatier 2017). A policy process theory that has an inherently
evaluative character is that elaborating on the role of policy feedback for policy change
(Cartwright and Hardie 2012: 107–9; see also Soss et al. 2007; Kumlin and Stadelmann-
Steffen 2014; Moynihan and Soss 2014).
In our view, there are three topics deserving particular attention when studying policy
change. First, policy change is a dynamic concept, and thus enhanced attention must be paid
Copyright © 2020. Macmillan Education UK. All rights reserved.

to the period within which the processes of change are observed. Second, assessments of the
degree of change might be affected by the level of analytical abstraction from which change
is evaluated. Third, policy change is a multifaceted concept, which entails that policy change
can have both a ‘positive’ and a ‘negative’ direction (see Knill et al. 2012) and concerns dif-
ferent elements of a policy, as stated by Hall (1993).
Policy change can be brought about by a whole range of factors and processes at the level
of the nation state. For instance, policy-makers may decide to reform a policy because it
is unpopular with voters and maintaining the policy in question may yield electoral costs.
A case in point is moral regulation. As a result of societal value change and processes of
secularization, many governments in Western countries face a growing demand for more
permissive regulations of moral issues like abortion, euthanasia or same-sex marriage
(Knill et al. 2015). To give another example, the Multiple Streams Framework (Kingdon
2003/2011) would argue that a policy entrepreneur was successful in coupling the policy
streams and managed to open up a policy window, which facilitated policy change taking
place (Zohlnhöfer et al. 2016; Herweg et al. 2017). Likewise, a whole range of interna-
tional factors can and do shape public policies adopted by national governments. There

is a consensus in the literature that political, social, economic and technological changes
generally discussed under the heading of globalization have significantly affected the
conditions for policy-making in nation states. In particular, the increasing integration of
markets and the emergence of transnational information and communication networks
challenge the autonomy and effectiveness of national governments in policy-making
(Holzinger et al. 2008a, 2008b).
On the one hand, economic and technological interdependencies have created a range of
problems that exceed the scope of national sovereignty, and so they can no longer be resolved
by the unilateral action of national governments. Perhaps the most widely discussed exam-
ple of such an issue is climate change, which is a problem at a global scale and which requires
coordinated – or as another strand of research claims, polycentric (Ostrom 2010a, 2010b,
2011; McGinnis and Ostrom 2012) – policy action. On the other hand, the emergence of
globally integrated markets poses new challenges for the regulation of domestic problems.
In view of this, national governments cooperate to establish international organizations
and international regimes in order to maintain their capacity to address social and political
problems that extend beyond the parameters of national sovereignty. As a result, the devel-
opment of global public policies is characterized by the involvement of a huge and still grow-
ing range of different international – both governmental and non-governmental – institutions
and the emergence of transnational administrations (Stone 2001, 2004, 2008; Stone and
Ladi 2015). Against this backdrop, we posed the question of whether there is a tendency for
public policies to become more similar across countries, i.e. for convergence. There is indeed
empirical evidence supporting the occurrence of convergence in various policy areas such as
environmental protection.
A prominent example is offered by the Sustainable Development Goals adopted in 2015,
which represent an attempt to foster a coordinated policy approach by the individual coun-
tries around the world towards sustainability. The Sustainable Development Goals comprise
17 goals and associated targets and indicators that are meant to be implemented during the
15 years starting in January 2016. As the implementation of the Sustainable Development
Goals goes on, the chances are that the national policy response to the global challenges
addressed will become more similar over time (see Tosun and Leininger 2017).
Lastly, we argued that public policies are primarily determined by the institutional configu-
Copyright © 2020. Macmillan Education UK. All rights reserved.

rations and the preferences of actors participating in the policy-making process. The institu-
tions establish formal rules and determine which actors participate in policy-making and what
role they play. Policy-analytical theories can be roughly divided into structure-based models,
institution-based models and interest-based models. All of these theories advance our under-
standing of how public policies come about by stressing different aspects. They represent an
important further development of the initial attempts to explain public policies, which were
characterized by the use of typologies and taxonomies. Altogether, we can state that public
policy analysis has experienced a steady expansion of sophisticated theories and empirical
research, which offer both country- and sector-specific insights, but also comparative ones.
In the introductory chapter, we formulated three general research questions in order to
tie together the various themes discussed in this book. We have already addressed these
questions, but now we turn to them and discuss them directly:

• How do public policies come about?

• When does policy change happen?
• What are the effects of public policies?

We learned that public policies may come about in response to several factors, including
the anticipation or manifestation of a (novel) problem; the emergence of new ideas, actors
or institutions; or sometimes even by chance, so to speak, as argued by the Multiple
Streams Framework. Public policies may change due to the same factors identified as
important to the beginning of the policy-making process. This finding is plausible and
aligns with the basic logic of the policy cycle, which contends that policy-making is an
ongoing process. Finally, the effects of public policy can be diverse, ranging from having
no substantial impact, to solving the social problem concerned and increasing public trust
in democratic government. With regard to the latter, in a similar vein, maintaining poli-
cies that are contested by the public is likely to lower trust in government. From this, it
follows that policy decisions and their effects at one or different points in time should be
regarded as a holistic unit.

COMPARATIVE RESEARCH ON PUBLIC POLICY

This book offered a comprehensive overview of concepts, theories, methods and data in
public policy. To this end, it approached public policy from different analytical angles and by
using different theoretical lenses. Despite the fact that policy studies are rich in concepts,
theories, methods and data, the body of research is remarkably coherent, since the great
majority of research concentrates on the comparative assessment of public policies. Some
prominent outlets of policy research such as the Journal of Comparative Policy Analysis under-
line their commitment to comparative research; other journals such as Policy Sciences or the
Policy Studies Journal do not stress the comparative dimension in their titles, but nonethe-
less most of the studies published in them are of a comparative nature. Having said that,
comparative public policy can be comparative on different dimensions (Dodds 2017), and in
this section we seek to give an overview of the different ways in which policy analysis can be
comparative. This overview is informed by the various empirical examples given throughout
the book in order to illustrate specific theories or arguments. We strive to bundle the points
raised explicitly or implicitly when discussing the individual topics to offer additional
insights regarding policy-making.
Copyright © 2020. Macmillan Education UK. All rights reserved.

Comparative policy research can include the following dimensions, which has implica-
tions for the choices to be made regarding research questions, the theoretical lens, the
empirical data and methodology:

• Policy outputs.
• Policy effects (including both policy outcomes and impacts).
• Comparison across geographical units.
• Comparison across policy sectors.
• Comparison across time.

Starting with policy outputs, the direct results of policy-making such as laws, ordinances and
other types of legally binding decisions, the literature has drawn heavily on the influential
conceptual framework developed by Hall (1993), which differentiates between policy instru-
ments, the setting of policy instruments and the policy paradigms on which these instru-
ments rest. These three categories have been applied jointly or separately for comparing
policies across countries (e.g. Holzinger et al. 2008a, 2008b; Knill et al. 2015; Tosun 2013a)

or across subnational units within countries (Gsir et al. 2016). Moreover, these categories
have been expanded or refined. Knill et al. (2012), for example, added a fourth category to
Hall’s framework, which denotes the scope of a policy instrument, i.e. how it governs its
target groups. In their edited volume, Howlett and Mukherjee (2018) offer a detailed discus-
sion of policy instruments and prepare the ground for a more refined comparison of coun-
tries with regard to the policy instruments they use. Another popular approach to comparing
policies across countries or subnational units is to analyse their spending patterns (e.g.
Dodds 2017). This approach has been particularly prominent in the comparative analysis of
welfare states, which makes sense, since higher spending levels can be intuitively associated
with more generous welfare states and lower spending levels with less generous welfare
states (see, e.g., Wolf et al. 2014).
Policy effects comprise policy outcomes and impacts and can be measured in many differ-
ent ways. For example, unemployment or environmental pollution levels are indicators that
can be used for comparative policy research (see, e.g., Knill et al. 2012; Jensen et al. 2014).
Depending on the nature of the policy outputs to be connected to policy effects, one can also
use other indicators such as public opinion on an issue (e.g. nuclear power phase-out) or
public trust in the government.
As already indicated in the previous discussion, the most prominent dimension of com-
parative research in public policy refers to geographical units such as nation states or entities
at the subnational level (e.g. regions, provinces or municipalities). Usually, when no further
clarifications are given on the way in which a study is comparative, it is reasonable to expect
a country comparison. At a higher level of aggregation, there also exist policy studies that
compare country groups with one another. For example, there is a body of research that
compares welfare state groups with each other rather than individual countries.
Another dimension on which public policies can be compared is the sector, which can, for
example, result in a systematic comparison of policy-making in the fields of environmental
and social policy. One of the comparative approaches that Dodds (2017) pursues in her book
is to determine how interests, ideas, institutions and globalization affect policy decisions
taken in the fields of economic, welfare, health, education and environmental policy. In
other words, the author adopts a sector-comparative approach to study policy-making,
which is particularly promising for learning about the circumstances under which or when
Copyright © 2020. Macmillan Education UK. All rights reserved.

interests, ideas, institutions and globalization matter for policy decisions. Sector-
comparative approaches can be applied for analysing one or several geographical units at
one or different points in time. This perspective can be used to gain a detailed understanding
of the role that specific actors such as political parties play. For example, one can analyse
whether the Green Party – being a member of the government or of the parliament – has an
impact on environmental policies only, or whether its involvement also affects other policy
areas, such as economic and commercial policies. Likewise, this design allows for learning
about the effect of interest groups, which are stronger in some policy sectors (e.g. health)
and weaker in others (e.g. internal affairs).
The temporal perspective is also addressed by comparative public policy. In terms of policy-
analytical concepts, comparison across time is closely associated with the concept of policy
change, and again Hall’s (1993) work on the three orders of policy change has been the founda-
tion for the corresponding literature. Similar to policy change, policy diffusion is a compara-
tive concept that explicitly includes a temporal perspective, as diffusion processes are defined
by determining a period of time during which the adoption of policy innovations takes place.
A third concept that is comparative on the temporal dimension is policy convergence.

As Tosun and Workman (2017) explain, there is one fundamental difference between
studies that are interested in policy change and those that explore policy diffusion and con-
vergence with regard to the comparative perspective they adopt. Studies of policy change
compare the same policy or set of policies over time in the same context, whereas policy
diffusion and convergence are comparative with regard to the same policy over time and in
different contexts. The feature they share is that all three concepts concentrate on the same
policy or set of policies, but research on diffusion and convergence is more comparative than
studies examining policy change, as it involves a greater number of observations. Studies of
policy diffusion and convergence can additionally exploit the comparative potential stem-
ming from different contexts. Regardless of which of the three analytical approaches one
chooses, it is necessary to select the observation period in accordance with theoretical
considerations.
To sum up, this book has pursued a comparative approach, even if it may not have been
explicit at all times. It presented theories that facilitate comparative research, it discussed
examples to offer comparative insights into the themes covered, and it pointed to method-
ological issues that primarily arise when carrying out comparative research.

DEVELOPING OWN RESEARCH QUESTIONS

The foremost goal of this textbook is to supply the readers with analytical tools that they can
use to evaluate their own research questions. In the previous two sections of this chapter we
attempted to give a concise summary of these analytical tools. Now we shift our focus and
develop advice for practising policy analysis. The starting point of policy analysis is the selec-
tion of a research question, which often presents the first challenge to students.
Zigerell (2011) brought together a number of strategies for developing research ques-
tions that are suitable for being addressed by students in the context of their own research
project. Zigerell’s suggestions take existing scholarship as the starting point and involve
reflecting on what has already been done there, and how a new project can develop further
the state of research. One possibility is to add one or a few new variables to the analysis. For
example, if a study analyses the role of Green Parties for the ambition of climate policy, an
Copyright © 2020. Macmillan Education UK. All rights reserved.

extension of this study could be to include other party families as well. The second idea is to
maintain the focus on Green Parties, but to argue that Green Parties only matter if they
form the government. Put differently, the variable of the Green Party would interact with a
variable gauging government participation. Continuing with the same example, a third
strategy would be to test the argument for different groups of countries. Let us assume that
the original study tests the effect of Green Parties on climate policy-making in Organisation
for Economic Co-operation and Development (OECD) countries. A student research project
could repeat the analysis, but differentiate between OECD countries that are European
Union (EU) member states and those that are not. A fourth possibility could be to concen-
trate on one federal polity such as Canada and to analyse whether the existence of a Green
Party at the provincial level affects climate policy strategies. To be sure, this strategy does
not only involve analysing the subnational level; such a disaggregation strategy can be
implemented in many different ways.
The next two recommendations include the improvement of how concepts are measured
and analysed. A promising student research project could propose and demonstrate the
added value of a different measurement approach for policies. The great majority of

comparative studies of social policy tend to use spending data for measuring policy change
(e.g. Dodds 2017), which is one but not the only way of measuring social policy. Therefore,
students interested in analysing social policy could use an alternative measurement such as
the number of specific social policy types adopted (see, e.g., Jensen et al. 2014; Steinebach
and Knill 2018). Likewise, a promising student research project could draw on an existing
study and modify the techniques for analysing the data. For example, instead of using the
classic unit of country-years for pooled data, a study could rely on cabinet-years (see, e.g.,
Schmitt 2016; Budde et al. 2017a).
The last four recommendations are based on the invitation to reflect on whether concepts
travel; if the response is affirmative, then this offers a good strategy for choosing a research
question. One possibility is to apply a concept discussed for one specific group of countries
(e.g. developed countries) to another group of countries (e.g. developing countries). Another
idea is to consider changing the time period and running the analysis again, or changing the
institutions that are at the centre of attention. The final idea is to look into other subdisci-
plines of political or fringe disciplines and to use their theories for policy analysis. In the last
few years, theories from (social) psychology have increasingly been adopted by policy stud-
ies, which has resulted in the emergence of the field of behavioural public policy (see, e.g.,
Shafir 2013; Sunstein and Reisch 2018; Strassheim and Beck 2019). A particularly interest-
ing perspective is put forward by the psychological theory of social identities, which Hornung
et al. (2019) propose to include in the study of policy processes. This approach emphasizes
the importance of individuals’ membership in social groups for their policy beliefs. Therefore,
a particularly promising and timely strategy for choosing a research question for a policy
study is represented by looking into psychology and exploring to what extent theories origi-
nating from that discipline can be used for analysing public policies.
Likewise, it appears promising to expand theorizing in policy studies to pay more atten-
tion to the role and characteristics of individuals, as for example legislative studies do. It
could be possible for politicians to support or oppose a certain policy because of personal
characteristics and experience. Pre-implementation diagnostics is an area where it is likely
that members of parliament do not only base their decisions on considerations regarding
vote maximization or responding to the policy demands of their constituencies, but per-
sonal characteristics and experience such as whether they have children of their own (see
Copyright © 2020. Macmillan Education UK. All rights reserved.

Baumann et al. 2015). The Multiple Streams Framework offers an analytical hook for accom-
modating such features of individuals by means of the concept of the policy entrepreneur,
but research has paid relatively little attention to the characteristics of this player in
policy-making.

DOING POLICY ANALYSIS

In this closing section, we want to bring together some practical guidelines for producing a
policy-analytic report that is written for practitioners rather than scholars. Students may
want to write such a report as part of the requirements for receiving credits in a course on
public policy, or because they want to reach out to a non-academic audience and offer advice
on what works and why. In the last few years, funding institutions have paid much attention
to the applicability of scientific findings to real life. In the context of the United Kingdom’s
Research Excellence Framework, an impact evaluation assessing the research of British

higher education institutions, for example, the impact of research outside of academia is
systematically assessed on the basis of criteria for rank-ordering research institutions with
regard to their effect on culture, economy, society and public policy or services. Similarly,
research projects funded by the EU in the context of Horizon 2020 ask for research findings
to be communicated and disseminated to non-academic audiences, as well as for possibili-
ties for the exploitation of findings by other users to be explored. In light of these develop-
ments, it appears necessary not only to train students in doing public policy research, but
also to give them guidance on how they can produce policy-analytical reports for a non-
academic audience.
For these audiences, policy analysis is typically motivated by some kind of problem that
needs to be addressed by policy measures. The document to be produced – for example, a
policy brief – presents some policy measures that might be taken and discusses each of these
measures with regard to their expected costs and benefits or other criteria of interest. The
document may contain only a discussion of the different policy measures, or also include a
recommendation as to which policy alternative should be chosen. While this may sound
straightforward initially, producing a good policy-analytical report is a demanding task and
requires some effort and patience, as several iterations may be necessary before the product
can be considered finished.
When producing such a policy-analytical report, we recommend proceeding in different
analytical steps (Cartwright and Hardie 2012; Bardach and Patashnik 2015), which rather
closely follow different stages of the policy cycle we have discussed in this book. More spe-
cifically, the following aspects should be addressed:

• Identify and demonstrate the underlying policy problem.

• Present potential solutions and criteria for their evaluation.
• Analyse effects of and trade-offs between alternatives on outcomes and impacts.
• Policy recommendation.

First of all, students need to define and specify the underlying policy problem, which is con-
sequential for the approach and ambition of the policy analysis exercise. In Chapter 5 of this
book, we discussed a range of analytical dimensions that are relevant with regard to the defi-
Copyright © 2020. Macmillan Education UK. All rights reserved.

nition and framing of policy problems. This also entails assembling evidence. Students
should collect data that can be transformed into information and subsequently into evi-
dence that has some bearing on the problem concerned. Second, students need to connect
the problem with potential policy solutions. At this stage, the policy analyst conceptualizes
alternative policy options that may, in principle, be appropriate for solving or mitigating the
problem concerned (see Chapter 6). The literature on policy design (see, e.g., Howlett and
Mukherjee 2018) can be a useful resource for this particular activity when doing policy anal-
ysis. Since policy analysis aims to offer guidance for policy-makers or other stakeholders,
this activity is inherently evaluative and needs to be based on certain criteria. On the basis
of this discussion, students hence also need to select and justify criteria for the assessment
of the expected policy effects, which then offer guidance on the selection of policy alterna-
tives. Commonly used criteria include effectiveness, efficiency, equity or fairness. Since
policy-analytical reports are written for practitioners, Bardach and Patashnik (2015) add
that often the evaluation criteria that really matter are legality, political feasibility, and
robustness and improvability.

The third step is then about projecting the effects of the policy alternatives in terms of
outcomes and impacts. The projection of the policy effects (including the side effects) can
refer to the policy problem concerned, specific target groups or both. At this stage of the
policy-analytical process, the policy analysts need to position themselves on how much
uncertainty they accept and how they generally handle uncertainty and ambiguity. This also
entails identifying and clarifying trade-offs, i.e. a situation where one of the policy alterna-
tives is expected to produce a better outcome than any of the other alternatives with regard
to every single evaluative criterion.
The fourth and final step refers to the recommendation of a decision based on the analy-
sis. If the policy analyst finds it difficult to choose amongst policy alternatives, this indicates
that the analysis has not been executed rigorously enough and that improvement is needed.
At this stage, there is a need to provide a story that brings together the problem definition,
the conceptualization of the policy alternatives, the evaluation criteria, the projected policy
effects and the discussion of trade-offs. There exist many different ways in which the story
can be told; most importantly, it should correspond to the characteristics of the target audi-
ence and it should be simple to understand and follow.
From these considerations it follows that the transfer of insights from policy analysis
from an academic to a non-academic sphere still requires a very analytical approach. In fact,
perhaps it requires even more advanced analytical skills, since policy-analytical reports,
especially when they are commissioned by clients, demand answers to questions and there-
fore ask the authors of such reports to adopt a comprehensive and yet rigorous approach.
Preparing a policy-analytical report can be seen as the litmus test for public policy scholars
as to whether they really understood the concepts, theories and methods of policy analysis.
Or in other words, only a profound analytical understanding of public policy can serve as a
basis for policy recommendations.
In this book we have sought to lay this foundation and to provide an up-to-date and
hands-on introduction to the study of public policy. We end the book by inviting students to
reflect critically on the topics covered and to think creatively of ways of advancing the state
of research in the future. However, this book also provides the tools necessary to do policy
analysis and to be in a position to liaise with policy-makers and other actors participating in
the policy-making process. It has provided insights into how problems are being defined and
Copyright © 2020. Macmillan Education UK. All rights reserved.

what features of problems make them more likely to be pursued by policy-makers. We stress
that there exist, in principle, multiple solutions to a problem and that the solutions depend
on how the problem is framed. Depending on the issue framing, the subsequent policy-
making process can become either more or less adversarial, resulting in policy outputs that
are either well-defined or vague. Once policy-makers have decided on the policy decision,
the criteria for evaluating the policy in question can be defined. We have presented a com-
prehensive discussion of the different ways in which policies can be evaluated, which can
refer to both their outcomes and their impacts. The scientific knowledge on these issues
conveyed by this book should enable students of public policy actually to make policy recom-
mendations if they wish to do so.

A B
Abbott, T., 116 Baccaro, L., 9
Actors and institutions, 73–4 Bacchi, C.L., 79, 84
Adam, C., 141 Bachrach, P., 87
Administrative bureaucracies, 49 Baratz, M.S., 87
Administrative capacities, 140 Bardach, E., 136, 239
Administrative evaluation, 146 Basic income experiment in Finland, 148
Administrative implementation, 131 Baumgartner, F.R., 87, 88, 92, 94
Administrative policy-making, 109 Behaviour of public actors, 68–9
Administrative styles, of international Belgian system, 204
organizations, 193 Bellé, N., 221
Adolino, J., 160 Bell, S., 66
Adopted policy, 8 Berman, P., 129
Advocacy Coalition Framework, 233 β−convergence, 223
Advocates, 193 Bevan, S., 93–4
Affordable Care Act, 119 ‘Big wave’ of globalization research, 184
Agency problems, 109 Binderkrantz, A.S., 95, 98
Agenda-setting, 77 Biosafety, 104
actors and interests in, 91–8 Bipolar party system, 41
analytical concepts of, 84–90 Blair, T., 124
models for, 90 Blake, C., 160
Agenda types, 85 Bonafont, C., 98
Agrarian parties, 50, 59 Börzel, T.A., 232
Agricultural policy of EU, 2 Bottom-up models of policy implementation,
Agri-food policy, 18 127, 129–30
American Civil Rights Movement, 52 Bottom-up perspective, 138
Anglo-Saxon liberal model, 65 Bounded rationality, 7
Anti-smacking law, 113 Bovens, M., 28, 158, 171
Copyright © 2020. Macmillan Education UK. All rights reserved.

Appraisal, policy process, 6 Bräuninger, T., 92

Armingeon, K., 9 Braybrooke, D., 6
Attribution problem, 154–6 Bretton Woods institutions, 42
Aunphattanasilp, C., 94 British environmental policy, 26
Australia, decision-making in, 116–17 British executive, 115, 119
Australian bicameralism, 116 British government, 115
Australian Constitution, 35 British House of Lords, 116
Australian federalism, 117 British parliament, 115
Australian federal parliament, 116 British political–administrative system, 108
Australian polity, 116 British political system, 107
Authority, 21–3 British public policy, 62
Authority-based policy, 21 British Westminster model, 115
Axelrod, R., 73 Brownson, R.C., 161

282
Knill, Christoph, and Jale Tosun. Public Policy : A New Introduction, Macmillan Education UK, 2020. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/ubnru-ebooks/detail.action?docID=6274048.
Created from ubnru-ebooks on 2021-04-03 16:16:55.
Index 283

Bureaucracies, 49, 126 Comprehensive Economic and Trade

Bureaucratic drift, 11–12 Agreement, 204
Bureaucrats, 69, 93–4, 102, 109 Concepts of rationality, 67–8
Bureau-maximizing actor, 69 Conceptual tools for policy design, 153
Bureau-shaping model, 69 Conditionality, 226
Conflict expansion, 94
C Congressional control, 108
Cabinet Office, 114 Consensus democracies, 62
Cairney, P., 28, 149 Conservative–Liberal Democrat coalition, 150
Cameron, D., 124 Conservative parties, 50
Canada, decision-making in, 117–18 Conservatives and Liberal Democrats, 124
Canadian context, 118 Consolidators, 193
Canadian federalism, 118 Constellations organizations, 64
Canadian government, 24 Constituent Policies, 15
Canadian monarch, 117 Constitutional court decisions, 48
Capacity tools, 23 Constitutions and constitutional courts, 34–6
Capano, G., 152 Context for policy-making, 34
Caramani, D., 40, 58 intergovernmental organizations, 42–5
Carstensen, M.B., 66 key actors, 45–54
Causal mechanisms, 225–8 national institutions, 34–41
Central advantage of authority-based Continental conservative model, 65
policies, 21 Continental European countries, 103
China, 195 Continental family of nations, 62
experimental governance in, 167 Contingency perspective, 89–90
Choice of policy instruments, 134–5 Contractual research, 159
Christian democratic parties, 50, 58 Control deficits, 201–2
Christian Democratic Union, 41, 107 Control rights, 47
Christian Social Union, 41, 107 Control structures, 136–7
Citizens, 96 Conventional attribution problem, 153–4
Civil Marriage Act in 2005, 17 Convention on Climate Change, 43
Clark, H., 107 Convergence dimensions, 224
Classic cycle approach, 8 Convergence research, 12
Classic institutionalist approaches, 66 Cooperative governance, 173–4, 176
Cleavage approach, 57–60 Copenhagen-based European Environment
Copyright © 2020. Macmillan Education UK. All rights reserved.

‘Clientela’ relationship, 106 Agency, 123

Clientelistic politics, 18–19 Copenhagen criteria, 226
Climate change, 9 Corporatism, 71
Coalition cabinets, 46 Corporatist policy pattern, 29
Cobb, R.W., 80, 82, 85 Corporatist polities, 70
Coercive isomorphism, 64 Corrective revision, 48
Cohen, M., 160 Cost–benefit analysis, 149
Collaborative federalism, 118 Cost-effectiveness analysis, 149–50
Command-and-control regulation, 21 Cost–utility analysis, 150
Communist parties, 50 Coughlin, J.F., 80
Comparative Agendas Project, 231 Critical juncture, 65
Comparative research, on public policy, 235–7 Croissant, A., 60
Complementary model, 38 Cross-cutting themes, 12
Complex constellations, 101 Cross-national policy convergence, 220–5
Complex context conditions, coping causes and conditions of, 225–9
with, 153–4 Cultural similarity, 228
Compliance with international law, 225 Cyert, R.M., 7

D Effectiveness evaluation, 146

Dahl, R., 27 Efficiency evaluation, 146
Daugbjerg, C., 18, 88 Egeberg, M., 109
Deardorff, A.V., 110 Eichbaum, C., 54
Debus, M., 92 Ekayani, M., 88
Decision costs, 74 Elder, C.D., 85
Decision-making, 3, 100–1 Elected public officials, 91–2
within agencies, 125–6 Elections, 39
in Australia, 116–17 Electoral institutions and party systems, 39–41
in Canada, 117–18 Elements, of public policy, 4–5
capacity, 177 Empirical evidence, 93
institutional and procedural dimensions of, English-speaking family of nations, 62
109–14 Enjolras, B., 106
policy formulation, determinants of, 102–9 Entrepreneurial politics, 18
in United Kingdom, 114–16 Entrepreneurial style, 193
in United States, 118–19 Environmental policy, 5
Deficient policy design, 200–1 convergence in Europe, 227
De Francesco, F., 104 Environmental Protection Agency (EPA), 123
Deliberative problem-solving processes, 198 Epistemic communities, 53
Democratic legitimacy, 178–9 Erne, R., 52
de Montesquieu, C., 36 Esping-Andersen, G., 65
Derlien, H.U., 108 EU policy-making, 52
‘Deviant’ definition, 78 Europe
Devolution process, 115, 116 environmental policy convergence in, 227
Diffusion, 197–8 experimental governance in, 167
Digitalization, 9 European Commission, 5, 113, 191
DiMaggio, P., 64 European Court of Justice, 95, 204
Direct democracy, 59 European Financial Stabilisation Mechanism, 9
Discursive institutionalism, 66 European integration, 183, 188
Distributive policies, 14, 15 European parliament, 110
District Health Boards in New Zealand, 125 European Single Market, 195
Division of powers, 36–9 European Union (EU), 113
Doan, A., 83 agricultural policy of, 2
Dodds, A., 236 Common Agricultural Policy, 2
Copyright © 2020. Macmillan Education UK. All rights reserved.

Dolowitz, D.P., 221 Emissions Trading System, 66

Dominant-party systems, 40, 41 policies, 27
Downs, A., 86, 109 EU–Turkey Refugee Deal, 191
Drug policy,125 Evaluators, 154
Due process, 178 Evidence-based policy-making, 159–61
Dunleavy, P., 69, 108 Ex ante techniques, 146
Duverger, M., 40 Executive, 46–7
Executive federalism, 118
E Experimental governance in China and Europe, 167
Eavey, C.L., 93 Experimental implementation, 131
Ecolabels, 20 Experts, private actors, 51–4
Economic and financial crisis, 9 Ex post assessment, 146
Economic and Social Council, 42 External validity problems, 156
Economic evaluation, 149–50 Extreme right-wing parties, 51
Economic globalization, ongoing process of, 9
Economic integration, 188 F
Economic interdependencies, 187–8 Facilitating factors, 228–9
Economic Opportunity Act of 1964, 82 ‘Families of nations’, 62

Federalism, 35, 38 Governance principles, 124

Federal systems, 38 policy classification by, 24
international agreements in, 204 Governance principles and instruments, 19
Feezell, J.T., 98 authority, 21–2
Feindt, P.H., 18 nodality, 20
Fink, S., 58 organization, 22–3
Finland, basic income experiment in, 148 treasure, 22
Fire alarms, 137 Government as collective actor, 34
First-order changes, 26 Government as institution, 34
Fischer, F., 110 Government of Wales Act 1998, 115
Fitzgerald, C., 101 Governments, 114
Focusing events, 86 Great Lakes Water Quality Agreement, 186
Food and Agriculture Organization, 193 Greenhouse gases (GHG) emission, 9
Formal transposition, 132 Green Parties, 41, 50, 59, 237
Formative evaluation, 148 Green-Pedersen, C., 92, 97, 112
Formulation of public policy, 24 Grossman, G., 105
Frames, 81 Group of Seven (G7), 189
Freeman, G.P., 29 Group of 20 (G20), 189
Fridays for Future Movement, 9 Gustafsson, G., 28
Fund for Peace, 176
H
G Haas, P., 103
Gallagher, M., 39 Hall, P.A., 26, 31, 57, 60, 61, 63, 73, 235, 236
γ–convergence, 222 Hall, R.L., 110
Game theory, 71–3 Hall’s typology, 27
Ganghof, S., 36 Hamilton, W.D., 73
General Agreement on Tariffs and Trade, 43 Hansen, M.E., 107
General Assembly, 42 Harmonization, 196
Generalized System of Preferences programme, 73 Hartz, P., 93
General Regulations, 194 Hawthorn, L.R.L., 150
Geographical scope, 190 Head, B.W., 160
Geographical units, 236 Helpman, E., 105
German Basic Law, 35 Hibbs, D., 68
German Constitutional Court, 93 Hierarchical governance, 169
Copyright © 2020. Macmillan Education UK. All rights reserved.

German parliament show, 58 Hierarchical intervention, 177, 178

Global Facilitation Network for Security Sector Hierarchical market regulation, 170
Reform, 54 Hiilamo, H., 148
Global governance, 165 Hill, A., 83
Globalization, 183–5 Hirschman, A.O., 6
Global public policy, 42 Historical institutionalism, 64–5
United Nations system, 42–3 Hjern, B., 129
World Trade Organization, 43–4 Holistic government, 124
Good governance, 164, 166, 177 Holzinger, K., 198
Governance, 12, 164–5 Homogeneity, 223
changes in, 169–71 Hönnige, C., 49
concepts of, 165–6 Hornung, J., 238
good governance, 177–9 House of Commons, 115, 117–18
ideal types of, 172–6 House of Representatives, 116, 119
modes of, 166–9 House of Review, 116
perspective, 180 Howlett, M., 20, 24, 28, 29, 90, 118, 134, 236
Governance by hierarchy, 166 Hull, C., 129
Governance by markets, 166 Humpage, L., 222

Hybrid governance, 125 Internal reorganizations, 69

Hybrid models of policy implementation, Internal work practices, 69
130–1 International agreements, in federal systems, 204
Hypothesis, 57 International bodies, 202
International cooperation, forms of, 189
I International Court of Justice, 43
Ideal types of governance, 172–6 International interest group, 53
Ideal-typical systems, 49 Internationalization, 183
Ideology, 110 International Labour Organization, 193
Implementation effectiveness, 177–8 International Monetary Fund (IMF), 42, 104, 226
Implementation models, 131 International organizations, 96, 103–4,
Implementation structure, 123–5 185, 189
Implementation success administrative styles of, 193
criteria for, 132–4 International Panel of Climate Change, 54
determinants of, 134–41 International policies, 200, 201
Imposition of policies, 226 International policy formulation, 204
Incentive tools, 23 International policy-making, 199
Incidence, 82 International postal communications, 194
Incrementalism, 7, 114 International public policy, areas of, 188–90
Incremental policy-making, 114 International regimes, 185
Independent problem-solving, 225 International solution, 194
Industrial Revolution, 59 International Union for the Conservation of
Information signals, 80 Nature and Natural Resources (IUCN), 53
Inglehart, R., 59 Interventionist governance, 173
Ingram, H., 23, 27 Interventionist state, 169, 170
Innovative potential of leader countries, Involved actors, preferences and strategies of,
196–7 202–3
Inside-access model, 86 Isomorphism, 221
Institutional agenda, 85 Issue attention cycle, 86–7
Institutional Analysis and Development
Framework, 168 J
Institutional breadth, 139 Jarvis, D.S., 222
Institutional depth, 139 Jensen, C., 227
Institutional design, 137–40 Joerges, C., 197, 198
Copyright © 2020. Macmillan Education UK. All rights reserved.

Institutional dimension, of decision-making, John, P., 60

109–14 Joined-up government, 124
Institutional Revolutionary Party, 40 Jones, B.D., 87, 88, 94
Institutional similarity, 228 Jordan, A., 113
Institution-based approaches, 63 Jordan, A.G., 28
Institution-based models, 61–6 Jordan, A.J., 24
Institution-based perspective, 88–9 Jordan, G., 28
Intelligence, policy process, 6 Jörgens, H., 223
Interaction of actors, 71–3 Journal of Comparative Policy Analysis, 235
Interest-based models, 67 Judges, 92–3
actors and institutions, 73–4 Judicial evaluation, 146–7
concepts of rationality, 67–8 Judicial review, 117
strategic choices of actors, 68–73 Judiciary, 37, 48–9
Interest groups, 94–6, 105–7
private actors, 51–3 K
Intergovernmental organizations, 41–4 Karembu, M., 104
United Nations system, 42–3 Kato, J., 93
World Trade Organization, 43–4 Katz, R.S., 50

Kenyan biosafety law, 104–5 M

Kenyan policy-makers, 105 Magic formula, 164
Keohane, R.O., 185 Mair, P., 50
Kern, K., 198 Majone, G., 155
Key actors, 45–6 Majoritarian democracies, 62
political parties, 49–51 Majoritarian politics, 18
private actors, 51–4 Majoritarian systems, 39
public actors, 45–9 Major, J., 108
Kingdon, J.W., 6, 85, 90 Maor, M., 79
Knill, C., 139, 172, 193, 236 March, J.G., 7
Knock-on effects, 8 Market governance, 232
Krasner, S.D., 139 Marsh, D., 221
Kriesi, H., 52 Marston, G., 161
Kyoto Protocol, 185, 191 Martin, L.L., 202
Marxism–Leninism, 50
L Materialist–post-materialist cleavage, 59
Labelling, 82 Materialist values, 59
Labour parties, 50 Matland, R., 130
Laissez-faire policies, 224 Matt, E., 113
Lascoumes, P., 27 Mayntz, R., 108
Lasswell, H.D., 1, 6, 231 Mazmanian, D.A., 128
Lax standards, 224 McCombs, M.E., 97
Learning tools, 24 McConnell, A., 158
Left-libertarian parties, 50 McCubbins, M., 137
Left-wing populist parties, 51 McDavid, J.C., 151
Le Galès, P., 27 McDowall, W., 167
Legislation, 110 Media, 97–8
Legislative assemblies, 37 Meritocratic bureaucracies, 49
Legislative committees, 102 Meritocratic system, 49
Legislative decision-making, 118 Meso-corporatist policy pattern, 29
Legislature, 37, 47–8, 93 Mexican environmental policy, implementation
in South Africa, 37 of, 140
Leininger, J., 179 Miller, G.J., 93
Lenschow, A., 139, 172 Ministerial bureaucracies, 102, 109
Copyright © 2020. Macmillan Education UK. All rights reserved.

Levi-Faur, D., 15 Ministerial bureaucrats, 103

Levinthal, D.A., 160 Mitchell, P., 39
Liberal parties, 50 Mobilization model, 85, 86
Lijphart, A., 62 Models, overview of, 74–5
Lindblom, C., 6 ‘Modern’ cleavages, 58
Lindblom, C.E., 6, 27 Modern governance, 171
Lindquist, E., 82, 118 Modernising Social Services reform, 150
Lippi, A., 152 ‘Modern’ UPU, 194
Lipset, S.M., 57, 58 Modes of governance, 166–9
Lipsky, M., 129 Montreal Protocol on Substances, 192
LME-typical short-term profitability Morality policies, 16, 17, 58
indicators, 60 Mudde, C., 50
‘Logic of appropriateness’, 63–4 Mukherjee, I., 236
‘Logic of consequentialism’, 73 Multipartite structures, 106
Logrolling, 112 Multiparty systems, 40, 41
Lorenz, P., 60 Multiple Streams Framework, 6, 79, 85, 90, 233,
Lowi’s policy typology, 14–17, 19 235, 238
Lowi, T.J., 3, 14, 21, 30 Myles, J., 30

N Organizational forms, 188–90

National Assembly for Wales, 116 Organizational strength, 175
National institutions, rules of political game Organizations, 7, 22–3
constitutions and constitutional courts, 34–6 Ostrom, E., 185
division of powers, 36–9 Outcome-based approaches, 156
electoral institutions and party systems, Outside-initiative model, 85
39–41
Nationalist parties, 51 P
National policy styles, 28 Package deals, 191
NATO scheme, 19, 23 Page, E.C., 93
Nature of public policies, 13–14 Paquin, S., 204
policy dimensions, 25–7 ‘Parantela’ relationships, 106
policy styles, 28–31 Paris Agreement, 185
typologies of, 14–24 Parliamentary systems, 102, 105
Neo-corporatism, 71 Participatory rights, 178
Neo-corporatist polities, 70 Partisan ideology, political preferences
(Neo-)corporatist systems, 106 stemming from, 107–8
‘Neo-liberal’ reform policies, 104 ‘Partisan mutual adjustment’ process, 6
Neo-liberal think, 103 Partisan theory, 68
Network governance, 167, 168, 232 Party difference hypothesis, 68
New institutionalism, 61, 63 Party for Freedom in the Netherlands, 9
New Labour government, 160 Patashnik, E.M., 239
Newman, J., 149 Patterns of Democracy (Lijphart), 62
New Public Management, 126, 141, 146, 150, 170 People’s Party in Denmark, 9
Newton, K., 137 Performance and Innovation Unit, 160
Neyer, J., 198 Performance evaluation, 150–1
Niskanen, W.A., 69, 93, 128 Performance measurement, 150
Nitrates Directive, 95 Peters, B.G., 108
Nodality, 20 Pierson, C., 222
Nodality, Authority, Treasure and Organization Pierson, P., 30
(NATO), 19 Pluralist policy style, 29
Non-governmental organizations (NGOs), 56, 188 Pluralist systems, 70, 106
Non-hierarchical governance, 168 Policies determine politics, 14
Normative isomorphism, 64 Policy, 5
Copyright © 2020. Macmillan Education UK. All rights reserved.

North American Free Trade Agreement Policy adoption, 232

(NAFTA), 140 Policy ambiguity, 130
North, D.C., 33 Policy-analytical literature, 14
Northern Ireland Assembly, 116 Policy-analytical theories, 234
Novelty, 82 Policy-analytic report, 238–9
Nudges, 20 Policy appraisal, 146
Policy brokers, 102
O Policy change, 12, 207, 223
Office-seeking parties, 68 by chance, 213
Olson, M., 51, 178 and institutional change, 211–13
O’Malley, E., 113 measurement of, 213–19
On the Spirit of Laws (de Montesquieu), 36 in preferences and beliefs, 209–11
Open–closed society cleavage, 59 theories of, 208–9
Open Method of Coordination, 198 Policy characteristics, 222
‘Opposition,’ 48 Policy classification by governance principles, 24
Organisation for Economic Co-operation and Policy conflict, 130
Development (OECD) countries, 58, 96, Policy content dimensions, 26–7
141, 146, 186, 223, 237 Policy convergence, 220–2

dimensions of, 224 Policy-making, theoretical approaches to, 56–7

forms of, 223 institution-based models, 61–6
types and measurement of, 222–4 interest-based models, 67–74
‘Policy cycle’ model, 7–8 models, overview of, 74–5
sequence of, 25–6 structure-based models, 57–61
Policy decisions, 48 Policy management knowledge, 160
Policy design, conceptual tools for, 153 Policy mediators, 102
Policy diffusion, 221 Policy networks, 161, 167–8, 171
Policy dimensions, 25–7, 229 Policy outcomes, 25, 27
policy content dimensions, 26–7 Policy outputs, 25, 27, 235
sequence of the policy cycle, 25–6 Policy problems, 153
Policy effects, 236 Policy process, 8
Policy evaluation, 144–5, 233 stages of, 7–8, 11–12
administrative evaluation, 146 Policy reform, 113
economic evaluation, 149–50 Policy Sciences, 235
elements of, 151 Policy sectors, 190
judicial evaluation, 146–7 Policy spaces, 154
performance evaluation, 150 Policy stages, 8
political dimension of, 157–62 Policy Studies Journal, 235
political evaluation, 147 Policy styles, 28–31
scientific evaluation, 147–9 Policy subsystem, 134
Policy formulation, 191–9 Policy success, 158
determinants of, 102–9 Policy transfer, 221
Policy impacts, 26, 27 Policy typologies, 24
Policy implementation, 122–3, 125–6, 199–200 Policy variation, 2
bottom-up models of, 129–30 Political civil servants, 108
control deficits, 201–2 Political conflicts, 23
decision-making within agencies, 125–6 Political dimension, of policy evaluation,
deficient policy design, 200–1 157–62
hybrid models of, 130–1 Political economy, 60
preferences and strategies of involved actors, Political entrepreneurs, 18
202–4 Political evaluation, 147
structure, 123–5 policy evaluation vs., 157–9
success, criteria and determinants, 132–41 Political implementation, 131
Copyright © 2020. Macmillan Education UK. All rights reserved.

target group behaviour, 126 Political institutions, 63

top-down models of, 127–9 classic approach to, 61–3
in United Kingdom, 124 discursive institutionalism, 66
Policy-makers, 7, 79, 101, 145, 152, 160, 179, historical institutionalism, 64–5
240 sociological institutionalism, 63–4
Policy-making, 7, 9, 12, 41, 84, 115, 231 Political parties, 49–51, 147
Policy-making, context for, 33–4 Political success, 158
intergovernmental organizations, 42–5 Politics, 3, 13
key actors, 45–54 of same-sex marriage, 17
national institutions, 34–41 Polities, 3, 13
Policy-making process, 69 Pollitt, C., 146
analytical perspectives on, 6–8 Polycentric climate governance, 185
Policy-making, synoptic perspective on, 12, 164–5 Polycentric governance, 168
changes in, 169–71 Porter, D.O., 129
concepts of, 165–6 Potential determinants of policy styles, 29
good governance, 177–80 Powell, W.W., 64
ideal types of, 172–6 Power distribution perspective, 87–8
modes of, 166–9 Practical application, 133

Pralle, S., 88 Radical view, 6

Prescription, policy process, 6 Rasch, B.E., 92
Presidential systems, 46–7, 102 Rasmussen, N.E.K., 107
Pressman, J.L., 122, 127 Rational choice theorists, 68
Prime Minister’s Delivery Unit, 124 Rational decision-making process, 6
Principal–agent problems, 23 Rationalist approach, 6, 8
Principal–agent theories, 136 Rational policy choice, 67
‘Principle of collective responsibility,’ 46 Rational voters, 68
Private actors Redistributive policies, 15
experts, 53–4 Reform capacity, 175
interest groups, 51–3 ‘Regionalist’ parties, 58
role of, 69–71 Regulated self-governance, 173
Private interest governments, 168 Regulation, 21
Private interests, pursuit of, 108–9 Regulatory agencies, 123
Private self-governance, 174, 176 Regulatory capture, 21–2
Privatization, 80 Regulatory competition, 226
Problem definition, 78 Regulatory policies, 15, 21
characteristics of, 78–81 Regulatory state, 170
process of, 78–81 Research designs for scientific policy evaluation,
Procedural dimension, of decision-making, 152–7
109–14 Richardson, J., 28
Procedural policy outputs, 25 Richardson, J.J., 28, 29
Procedural restrictions, 101 Right-wing populist parties, 51
Process-based assessments, 85 Riker, W., 68
Process evaluation, 146 Rissi, C., 159
Process perspective, 85–7 Rittel, H.W.J., 80
Promotion, policy process, 6 Rochefort, D.A., 82
Proportional representation systems, 39 Rokkan, S., 57, 58
Proximity, 82 Rossi, P.H., 151, 158
Public actors, 46–9 Rüdig, W., 29
bureaucracy, 49 Rural–urban cleavage, 60
executive, 46–7
judiciary, 48–9 S
legislature, 47–8 Sabatier, P.A., 128
Copyright © 2020. Macmillan Education UK. All rights reserved.

Public enterprises, 23 Sager, F., 159

Public Health and Disability Act 2000, 125 Same-sex marriage, politics of, 17
Public interest groups, 52 Scandinavian family of nations, 62
Public policies, 1, 184 Scandinavian social democratic model, 65
beyond the nation state, 184–90 Schattschneider, E.E., 80, 87
comparative research on, 235–7 Schmidt, M.G., 63
Public policy analysis, research of, 2 Schmidt, V.A., 66
Public policy-making, 8 Schmitter, P.C., 53, 70
Punctuated Equilibrium Theory, 88, 233 Schnapp, K.-U., 93, 136
Pursuit of private interests, 108–9 Schneider, A., 23, 27
Scholten, P., 88
Q Scientific evaluation, 147–9
Quantitative methods, 154 Scientific knowledge, 160, 240
Quasi-experimental approaches, 155 Scientific policy evaluation, research designs for,
152–7
R Scope of convergence, 224
Racist parties, 51 Scotland Act 1998, 115
Radicalism, 39 Scottish parliament, 115

Scottish social policy, 115 Stubager, R., 97

Secretariat, 42 Students, of public policy, 1
Secretaries, 119 Studsgaard, J., 88
‘Self-limitation,’ 48 Study of public policy, 10–12, 14, 75
Semi-parliamentary systems, 36 Substantial restrictions, 101
Semi-presidential system, 36 ‘Substantive’ policy-making, 109
Servant administration, 193 Substantive policy outputs, 25
Severity, 82 Summative evaluation, 148
Sexual harassment, as policy problem, 83 Sunstein, C.R., 20
Shaw, D.L., 97 Super-wicked problems, 113
Shaw, R., 54 Supranational organizations, 45
Sheingate, A.D., 88 Sustainable Development Goals (SDGs), 126,
Shimkhada, R., 161 176, 200, 234
Siaroff, A., 70 Swiss polity, 159
σ-convergence, 222–223 Symbolic and hortatory tools, 23
Signalling interest group politics, 18 Symbolic implementation, 131
Similarity, in socio-economic structures and Synoptic perspective on policy-making, 12,
development, 228 164–5
Simmons, B.A., 202 changes in, 169–71
Simon, H., 67 concepts of, 165–6
Simon, H.A., 7 good governance, 177–80
Single Market, 188 ideal types of, 172–6
Sipilä, J., 148 modes of, 166–9
Social acceptance, 141 Systematic scientific evaluation, 148
Social construction, 78 Systemic agendas, 85
Social Democratic Party, 41, 50, 107, 110
Socialization, 80 T
Social media, 9 Taiwan political parties, 60
Social policy, 5 Target group behaviour, 126
Social-state principle, 35 Taylor, R.C.R., 63, 73
Sociological institutionalism, 63–64 Temporal perspective, 236
Soroka, S., 97 Termination, policy process, 6
Soskice, D., 57, 61 Thaler, R.H., 20
South Africa, legislature in, 37 Thatcher, M., 108, 150
Copyright © 2020. Macmillan Education UK. All rights reserved.

Southern family of nations, 62 Thelen, K., 65

Sprague-Jones, J., 223 Theory of incrementalism, 6
Stacked decks, 137 Theory of pluralism, 70
Stages of policy process, 7–8, 11–12 Thomas, C., 83
State–church cleavage, 58 Tiers, 40
State-owned companies, 23 Timmermans, A., 88
Stone Sweet, A., 48 ‘Tit for tat,’ 73
Strategic choices of actors, 68 Top-down models of policy implementation,
behaviour of public actors, 68–69 127–9
interaction of actors, 71–73 Top-down policy-making, 169
role of private actors, 69–71 Tosun, J., 28, 29, 74, 237
Strategy maps, 153 Traditional approach, 58
Streeck, W., 53, 65 Traditional cleavage approach, 58
Strøm, K., 68 Transaction costs, 74
Structure-based models, 57 Transboundary problems, 186
cleavage approach, 57–60 Transitivity, 67
Varieties of Capitalism (VoC) approach, 60–1 Transnational governance, 165
Structure of prisoner’s dilemma, 72 Transnational networks, 198

Transnational problem structures and regulatory Van Horn, C.E., 128

structures, incongruence between, 186–7 Van Meter, D.S., 128
Treasure, 22 Varieties of Capitalism (VoC) approach, 57, 60–1
Treisman, D., 39 Varone, F., 147
Trigger decision-making processes, 7 Vedlitz, A., 80
Troika, 104 Veldman, J., 53
True, J.L., 88 Veto players, 74, 111–12, 114
Tsebelis, G., 92, 111 Victor, D.G., 185
Turnpenny, J., 159 Völkel, P., 60
Two-party systems, 40 von Bismarck, O., 16
Type of policy, 229 Vulnerability, 184
Typical interest constellations and interaction,
191–9 W
Typologies of public policies, 14 Waarden, F. van, 29
alternative schemes, 23–4 Waldal, R.H., 106
governance principles and instruments, Watts, R., 161
19–23 Weber, M., 49, 69, 80
Lowi’s policy typology, 14–17 Weil, S., 95
Wilson’s policy typology, 18–19 Western interest groups, 95
Westminster parliament, 115, 116
U Westminster-style polity, 117
UN Convention on Climate Change, 133 Westminster systems, 54
UN conventions, 43 ‘Wicked’ constellations, 101
‘Uniform’ policy design, 201 Wildavsky, A., 122, 127
Unitarism, 35 Wilkins, A., 170
United Kingdom Willmott, H., 53
decision-making in, 114–16 Wilson, J.Q., 14, 51
policy implementation in, 124 Wilson’s policy typology, 18–19
Research Excellence Framework, 238 Winner-takes-all system, 119
United Nations (UN), 42–3, 96 Winter, S., 123
United Nations Children’s Fund (UNICEF), 43 Wood, B.D., 80, 83
United Nations Environment Programme World Bank, 42, 104, 166, 186
(UNEP), 43 World Futures Studies Federation, 54
United Nations (UN) Sustainable Development World Health Organization (WHO), 81, 103
Copyright © 2020. Macmillan Education UK. All rights reserved.

Summit, 126 World Trade Organization (WTO), 42–5, 189

United States, decision-making in, 118–19 World Wildlife Fund (WWF), 53
Universal Postal Convention, 194
Universal Postal Union (UPU), 194 Y
US-based literature, 159 Yearbook of International Organizations, 189
US Supreme Court, 48
US trade policy, 73 Z
Zahariadis, N., 90
V Zigerell, L.J., 237
Vancoppenolle, D., 123 Zohlnhöfer, R., 90
van Deth, J.W., 137

Theories of The Policy Process - Christopher M - Weible - 5, 2023 - Taylor & Francis Group - 9781003308201
No ratings yet
Theories of The Policy Process - Christopher M - Weible - 5, 2023 - Taylor & Francis Group - 9781003308201
411 pages
Kashish Research Paper
No ratings yet
Kashish Research Paper
101 pages
Spicker, Paul - Social Policy - Theory and Practice-Policy Press (2014)
50% (2)
Spicker, Paul - Social Policy - Theory and Practice-Policy Press (2014)
514 pages
Untitled
100% (2)
Untitled
649 pages
Policy Analysis Concepts and Practice (David L. Weimer, Aidan R. Vining)
100% (3)
Policy Analysis Concepts and Practice (David L. Weimer, Aidan R. Vining)
494 pages
The Routledge Handbook of Policy Tool...
No ratings yet
The Routledge Handbook of Policy Tool...
620 pages
The Public Policy Theory Primer
50% (2)
The Public Policy Theory Primer
277 pages
GD121 Spare Parts Old
No ratings yet
GD121 Spare Parts Old
647 pages
Public: Administration
100% (3)
Public: Administration
520 pages
Research Methods in Public Administration and Public Management An Introduction by Sandra Van Thiel
100% (6)
Research Methods in Public Administration and Public Management An Introduction by Sandra Van Thiel
205 pages
Handbook - Research - Methods - Applications - Comparative - Policy - Analysis - Guy Peters - Fontaine PDF
100% (2)
Handbook - Research - Methods - Applications - Comparative - Policy - Analysis - Guy Peters - Fontaine PDF
425 pages
LIVRO HOOD The Tools of Government 1983 PDF
No ratings yet
LIVRO HOOD The Tools of Government 1983 PDF
187 pages
Implementing Public Policy
100% (2)
Implementing Public Policy
200 pages
Principles of Administrative Theory
No ratings yet
Principles of Administrative Theory
261 pages
The Public Administration Theory Primer PDF
100% (2)
The Public Administration Theory Primer PDF
274 pages
(Textbooks in Policy Studies) Christoph Knill - Public Policy - A New Introduction-Palgrave MacMillan (2012)
No ratings yet
(Textbooks in Policy Studies) Christoph Knill - Public Policy - A New Introduction-Palgrave MacMillan (2012)
376 pages
HOWLETT. Designing Public Policies
100% (3)
HOWLETT. Designing Public Policies
170 pages
Understanding Public Management
No ratings yet
Understanding Public Management
209 pages
Fred Halliday (Auth.) - Rethinking International Relations-Macmillan Education UK (1994)
80% (5)
Fred Halliday (Auth.) - Rethinking International Relations-Macmillan Education UK (1994)
301 pages
Philosophy and Public-Administration (An Introduction) by Edoardo Ongaro (2017)
100% (2)
Philosophy and Public-Administration (An Introduction) by Edoardo Ongaro (2017)
280 pages
En10272 PDF
100% (1)
En10272 PDF
42 pages
KNOEPFEL - Public Policy Analysis
80% (10)
KNOEPFEL - Public Policy Analysis
582 pages
Handbook of Theories of Public Administration and Management (Thomas A. Bryer)
No ratings yet
Handbook of Theories of Public Administration and Management (Thomas A. Bryer)
353 pages
2 - Knoepfel Et Al Public Policy Analysis
No ratings yet
2 - Knoepfel Et Al Public Policy Analysis
335 pages
Public Policy Analysis Models
73% (11)
Public Policy Analysis Models
56 pages
John McCormick - Carrots, Sticks and Sermons - Policy Instruments and Their Evaluation-Routledge (2017)
No ratings yet
John McCormick - Carrots, Sticks and Sermons - Policy Instruments and Their Evaluation-Routledge (2017)
313 pages
Peters 2e2021 - Advanced Introduction To PP
No ratings yet
Peters 2e2021 - Advanced Introduction To PP
232 pages
Good Practices and Innovations in Public Governance 2003-2011
100% (3)
Good Practices and Innovations in Public Governance 2003-2011
464 pages
(Political Analysis) Vivien Lowndes, Mark Roberts - Why Institutions Matter - The New Institutionalism in Political Science (2013, Palgrave)
100% (6)
(Political Analysis) Vivien Lowndes, Mark Roberts - Why Institutions Matter - The New Institutionalism in Political Science (2013, Palgrave)
247 pages
Making and Managing Public Policy
No ratings yet
Making and Managing Public Policy
302 pages
FULL Version Testbank Coordinate Geometry For JEE Advanced 3rd Edition G Tewani Multiple Formats
No ratings yet
FULL Version Testbank Coordinate Geometry For JEE Advanced 3rd Edition G Tewani Multiple Formats
409 pages
Public Administration
No ratings yet
Public Administration
178 pages
TOS TLE 8 Agricrop For Sharing
No ratings yet
TOS TLE 8 Agricrop For Sharing
2 pages
Isbn 978-1-29203-920-6
No ratings yet
Isbn 978-1-29203-920-6
240 pages
Public Policy and Public Admin
100% (1)
Public Policy and Public Admin
15 pages
Understanding Public Policy Theories and Issues 2nbsped 1137545194 9781137545190 Compress
100% (5)
Understanding Public Policy Theories and Issues 2nbsped 1137545194 9781137545190 Compress
503 pages
Peter Knoepfel Public Policy Resources Policy Press - 2018
100% (1)
Peter Knoepfel Public Policy Resources Policy Press - 2018
326 pages
Jurg Martin Gabriel-Worldviews and Theories of International Relations-Palgrave Macmillan (1994)
100% (1)
Jurg Martin Gabriel-Worldviews and Theories of International Relations-Palgrave Macmillan (1994)
203 pages
Toward A Comparative Institutional Analysis
100% (1)
Toward A Comparative Institutional Analysis
482 pages
An Institutional Framework For Policy Analysis and Design - Ostrom
No ratings yet
An Institutional Framework For Policy Analysis and Design - Ostrom
49 pages
Public Policy Dye Study Material
100% (6)
Public Policy Dye Study Material
40 pages
Bharat Petroleum Corporation Limited: Sv/Tv/Ta-Form
100% (2)
Bharat Petroleum Corporation Limited: Sv/Tv/Ta-Form
1 page
Philippine Education: Where We Are, Basic Characteristics, Issues and Concerns
No ratings yet
Philippine Education: Where We Are, Basic Characteristics, Issues and Concerns
56 pages
1845424360
100% (1)
1845424360
268 pages
Paul Cat
100% (1)
Paul Cat
70 pages
The Public Policy Theory Primer
100% (3)
The Public Policy Theory Primer
277 pages
Public Policy
100% (16)
Public Policy
462 pages
Models of Public Policy Formulation
67% (6)
Models of Public Policy Formulation
17 pages
Cairney 2020 - Understanding Public Policy - 2 Edition - Front Matter
100% (1)
Cairney 2020 - Understanding Public Policy - 2 Edition - Front Matter
20 pages
BYK E-Prospectus of PDF
No ratings yet
BYK E-Prospectus of PDF
9 pages
Theory in The Study of Public Policy
No ratings yet
Theory in The Study of Public Policy
7 pages
01 - Assignment TX Line Solutions
100% (2)
01 - Assignment TX Line Solutions
4 pages
Public Policy Making Process
83% (6)
Public Policy Making Process
24 pages
Policy Analysis
100% (13)
Policy Analysis
494 pages
Techno-Commercial Proposal (Without Price) (08!04!2025)
No ratings yet
Techno-Commercial Proposal (Without Price) (08!04!2025)
6 pages
Policy Making in The Real World Final
100% (1)
Policy Making in The Real World Final
105 pages
Political Science CC 3.3
No ratings yet
Political Science CC 3.3
85 pages
Fukuyama - What Is Governance
No ratings yet
Fukuyama - What Is Governance
22 pages
(Ebook) The Public Policy Theory Primer by Kevin B. Smith, Christopher W. Larimer ISBN 9780813347493, 9780813347509, 0813347491, 0813347505
No ratings yet
(Ebook) The Public Policy Theory Primer by Kevin B. Smith, Christopher W. Larimer ISBN 9780813347493, 9780813347509, 0813347491, 0813347505
81 pages
Comparative Public Policy
No ratings yet
Comparative Public Policy
1 page
Theories of Public Policy Making Process
100% (2)
Theories of Public Policy Making Process
7 pages
Merged PDF 20250712 12.14.37
No ratings yet
Merged PDF 20250712 12.14.37
22 pages
Spectroscopic Techniques
No ratings yet
Spectroscopic Techniques
38 pages
Chapter 1 To Students Public Policy Year 2 Pos 2023
No ratings yet
Chapter 1 To Students Public Policy Year 2 Pos 2023
7 pages
A Theory of World Politics
100% (6)
A Theory of World Politics
278 pages
What Is Public Policy
100% (1)
What Is Public Policy
57 pages
01 - Understanding Public Policy Basics and Importance
No ratings yet
01 - Understanding Public Policy Basics and Importance
8 pages
Final Paper
No ratings yet
Final Paper
6 pages
Werner 2018 Geographies of Production I Global Production and Uneven Development
No ratings yet
Werner 2018 Geographies of Production I Global Production and Uneven Development
11 pages
Banana - Mail Arte - Flue - v4 - n3-4 - 1984
No ratings yet
Banana - Mail Arte - Flue - v4 - n3-4 - 1984
3 pages
CT TIF Presentation For Kickoff-Final
No ratings yet
CT TIF Presentation For Kickoff-Final
13 pages
Result
No ratings yet
Result
1 page
Sample
No ratings yet
Sample
45 pages
What Is Weather in Canada
No ratings yet
What Is Weather in Canada
5 pages
Pre-Call Report 4 Template - MKT3403-S2020
No ratings yet
Pre-Call Report 4 Template - MKT3403-S2020
5 pages
Formato Aplicacion para Lanchas DE
No ratings yet
Formato Aplicacion para Lanchas DE
2 pages
Force of Friction
No ratings yet
Force of Friction
30 pages
Grammar Worksheets
No ratings yet
Grammar Worksheets
30 pages
Week 5 MODULE PURPOSIVE COMMUNICATION
No ratings yet
Week 5 MODULE PURPOSIVE COMMUNICATION
13 pages
Idsl 865 Leveraging Human Resources
No ratings yet
Idsl 865 Leveraging Human Resources
7 pages
Brosur Master Steel
No ratings yet
Brosur Master Steel
4 pages
Introduction To Public Policy: Course Description
No ratings yet
Introduction To Public Policy: Course Description
15 pages
Residential Plots For Sale in Wadakpally - Bheeramguda
No ratings yet
Residential Plots For Sale in Wadakpally - Bheeramguda
2 pages
Specifications-700-HC Relays: Relay and Timer Specifications
No ratings yet
Specifications-700-HC Relays: Relay and Timer Specifications
1 page
A Guidelines For Interviewing For The High School Newspaper
No ratings yet
A Guidelines For Interviewing For The High School Newspaper
4 pages
Title of Paper - Bending-Axis Effects On Load-Moment (P-M) Interaction Diagrams For Circular Concrete Columns Using A Limited Number of Longitudinal Reinforcing Bars
No ratings yet
Title of Paper - Bending-Axis Effects On Load-Moment (P-M) Interaction Diagrams For Circular Concrete Columns Using A Limited Number of Longitudinal Reinforcing Bars
8 pages
KYC Template Individual AnnexB1
No ratings yet
KYC Template Individual AnnexB1
1 page
Projectile Motion Quiz Review: Define
No ratings yet
Projectile Motion Quiz Review: Define
5 pages
Writing Effective Public Policy Papers Young Quinn
83% (6)
Writing Effective Public Policy Papers Young Quinn
121 pages
Public Policy Models
94% (54)
Public Policy Models
22 pages

Public Policy A New Introduction

Uploaded by

Public Policy A New Introduction

Uploaded by

T E X T BO OK S I N

‘This book is an outstanding introduction to studies of public policy. It combines both

Christoph Knill and Jale Tosun

2 The Nature of Public Policies 13

3 The Context for Policy-­Making: Central Institutions and Actors 33

Key Actors 45

4 Theoretical Approaches to Policy-Making 56

5 Problem Definition and Agenda-Setting 77

The Power Distribution Perspective 87

Criteria for Implementation Success 132

The Political Dimension of Policy Evaluation 157

9 Governance: A Synoptic Perspective on Policy-Making 164

10 Public Policies beyond the Nation State 183

Policy Formulation: Typical Interest Constellations and Interaction 191

11 Policy Change and Policy Convergence 207

12 Conclusions: Theoretical Insights and Practical Advice 231

3.1 Stylized institutional setup of a polity 35

1.1 Basic research topics of public policy analysis 2

2.1 Politics of same-sex marriage 17

Christoph Knill and Jale Tosun

Table 1.1 Basic research topics of public policy analysis.

WHAT IS A PUBLIC POLICY?

Polity, Politics and Policy

Elements of a Public Policy

Differences in Scope: Sectors, Targets and Instruments

• Sector-specific measures (e.g. energy policy).

ANALYTICAL PERSPECTIVES ON THE POLICY-MAKING PROCESS

 asic Theoretical Perspectives on Public Policies: Rational Process Design,

scription (imposition of a binding decision), invocation (policy enforcement), termination

expected to produce a decision.

Stages of the Policy Process

Table 1.2 Policy stages and related research topics.

Problem definition and Policy formulation and

OPPORTUNITIES AND CHALLENGES FOR POLICY-MAKING

In this book, we want to offer a realistic description of policy-making in modern societies.

S TUDYING PUBLIC POLICY: THE APPROACH AND STRUCTURE

• How do public policies come about?

Basic Analytical Tools, Concepts and Theories

Stages of the Policy Process

TYPOLOGIES OF PUBLIC POLICIES

Classification by Implications for Politics

Lowi’s Policy Typology

The classification proposed by Lowi (1964, 1972) corresponds to long-recognized differ-

Table 2.1 Lowi’s policy typology.

Type of policy Definition Examples

Box 2.1 Politics of same-sex marriage

same-sex marriages in 2017 demonstrated that members of parliament are split on

Wilson’s Policy Typology

Table 2.2 Wilson’s typology.

Sources: Based on Wilson (1989) and Knill and Tosun (2011).

Classification by Governance Principles and Instruments

A recent development in research on information-based policy instruments concerns the

Organization refers to the reliance on formal organizational structures in order to achieve

Table 2.3 Policy classification by governance principles.

Source: Based on Hood and Margetts (2007).

An alternative way of practically specifying public policies is provided by differentiating

Policy Dimensions in the Sequence of the Policy Cycle

Policy Content Dimensions

approaches emphasize that an insightful study of policy instrumentation should entail

Table 2.4 Potential determinants of policy styles.

High stability Low stability

www.apsc.gov.au. The Australian government explains various policy instruments

interest groups and experts.

We illustrate this general, organization-based context of policy-making before we present

 ATIONAL INSTITUTIONS: DEFINING THE RULES

Constitutions and Constitutional Courts

Electoral and party system

Figure 3.1 Stylized institutional setup of a polity

In most polities, the constitution is protected by a supreme or constitutional court. In

Box 3.1 Composition of the legislature in South Africa

Electoral Institutions and Party Systems

see, e.g., Martin and Vanberg 2005, 2011).

INTERGOVERNMENTAL ORGANIZATIONS: GLOBAL PUBLIC POLICY

Organizations operating in an international context are classified as intergovernmental

The United Nations System

The World Trade Organization

3 The Context for Policy-Making: Central Institutions and Actors 33

3.1 Stylized institutional setup of a polity 35

1.1 Basic research topics of public policy analysis 2

2.1 Politics of same-sex marriage 17

WHAT IS A PUBLIC POLICY?

Polity, Politics and Policy

Elements of a Public Policy

Differences in Scope: Sectors, Targets and Instruments

ANALYTICAL PERSPECTIVES ON THE POLICY-MAKING PROCESS

asic Theoretical Perspectives on Public Policies: Rational Process Design,

Stages of the Policy Process

OPPORTUNITIES AND CHALLENGES FOR POLICY-MAKING

S TUDYING PUBLIC POLICY: THE APPROACH AND STRUCTURE

Basic Analytical Tools, Concepts and Theories

Stages of the Policy Process

TYPOLOGIES OF PUBLIC POLICIES

Classification by Implications for Politics

Lowi’s Policy Typology

Wilson’s Policy Typology

Classification by Governance Principles and Instruments

Policy Dimensions in the Sequence of the Policy Cycle

Policy Content Dimensions

ATIONAL INSTITUTIONS: DEFINING THE RULES

Constitutions and Constitutional Courts

Electoral Institutions and Party Systems

INTERGOVERNMENTAL ORGANIZATIONS: GLOBAL PUBLIC POLICY

The United Nations System

The World Trade Organization

The Cleavage Approach

The Classic Approach to Political Institutions

The New Approaches to Political Institutions

A well-known example of a historical-institutionalist approach is given by Esping-

Strategic Choices of Actors

Explaining the Behaviour of Public Actors

Explaining the Role of Private Actors

Understanding the Interaction of Actors: Game Theory

Actors and Institutions: Rational Choice Institutionalism

PROBLEM DEFINITION: THE PATH TO AGENDA-SETTING

The Process of Problem Definition

The Characteristics of Problems

ANALYTICAL CONCEPTS OF AGENDA-SETTING

The Process Perspective

The Power Distribution Perspective

The Institution-Based Perspective

The Contingency Perspective

ACTORS AND INTERESTS IN THE AGENDA-SETTING PROCESS

Elected Public Officials

DETERMINANTS OF POLICY FORMULATION

Expertise, Information and Ideas

Political Preferences Stemming from Partisan Ideology

Pursuit of Private Interests

I NSTITUTIONAL AND PROCEDURAL DIMENSIONS

How Are Policy Preferences Formed?

ECISION-MAKING IN THE UNITED KINGDOM, AUSTRALIA, CANADA

The United Kingdom

The United States

WHO IS INVOLVED IN POLICY IMPLEMENTATION?