The Power of Opposition: How Legislative Organization Influences Democratic Consolidation 9780367437312, 9781032282459, 9781003005360

Proposing a novel way to look at the consolidation of democratic regimes, this book presents important theoretical and e

209 24 5MB

English Pages 201 [203] Year 2022

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Half Title
Series Page
Title Page
Copyright Page
Table of Contents
List of Figures
List of Tables
Acknowledgments
PART I: The Moderating Role of Parliamentary Procedures
1 Institutions, the Role of Opposition, and Democratic Consolidation
1.1 The Role of Citizens
1.2 The Role of Legislatures
1.3 Key Questions and Contribution
1.4 Outline of the Book
2 The Role of Parliamentary Procedures
2.1 The Importance of Legislatures – Moving beyond Governmental Organization
2.2 The Importance of Winners and Losers
2.3 The Influence of Parliamentary Procedures on Democratic Consolidation
2.4 Summary
3 Measuring Opposition Power
3.1 Policy-Making Power of Opposition Players
3.2 Method of Data Collection – Policy-Making Power of Opposition Players
3.3 Summary
PART II: Policy-Making Power of Opposition Players and Its Consequences for Democratic Consolidation
4 Methods and Data
4.1 Empirical Strategy
4.2 Independent Variable: Policy-Making Power of Opposition Players, Winners and Losers
4.3 Dependent Variable: Consolidation – Political Support
4.4 Control Variables
4.5 Sample
4.6 Summary
5 Policy-Making Power of Opposition Players around the World
5.1 How Much Policy-Making Power Do Opposition Players Have?
5.2 What Kind of Policy-Making Power Do Opposition Players Have?
5.3 Conclusion
6 Policy-Making Power of Opposition Players and Democratic Consolidation
6.1 Democracy as the Best Form of Government
6.2 Satisfaction with Democracy
6.3 Satisfaction with Government Performance
6.4 Conclusion
7 Why Should We Care About Opposition Power? Conclusion and Implications
7.1 Main Findings
7.2 The Study of Institutions, Oppositions, and Democratic Consolidation – Limitations and Future Research
7.3 Last Thoughts
Appendix
Index
Recommend Papers

The Power of Opposition: How Legislative Organization Influences Democratic Consolidation
 9780367437312, 9781032282459, 9781003005360

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

“This is an eye-opener on a topic of major political importance. The book demonstrates that successful functioning of democracy depends not only on the performance of the government but also on the political postures of the opposition, which are shaped in turn by the institutional prerogatives of the opposition in different political systems. The evidence is rich, the analysis is persuasive, and the conclusions are crucial for understanding the fate of democracies around the world. If you have time to read only one book, read this one.” Adam Przeworski, Carroll and Milton Professor Emeritus, New York University “Why do electoral losers accept defeats and continue to support democracy? In this important new study, Simone Wegmann emphasizes the role of legislative institutions, showing that voters in countries with strong opposition rights are more satisfied with democracy than elsewhere. This is an original, thought-provoking contribution both to comparative legislative studies and democratic consolidation research.” Radoslaw Zubek, Associate Professor of European Politics, University of Oxford “This is a path-breaking book about democratic consolidation. It shows that allegedly arcane details of legislative organization affect the opposition’s relative power in policy making, which in turn affects voters’ support for democracy. Wegmann connects two fields that have developed more or less independently: democratization and legislative studies. Her book is bound to open up important avenues for future research.” José Antonio Cheibub, Mary Thomas Marshall Professor, Texas A&M University

The Power of Opposition

Proposing a novel way to look at the consolidation of democratic regimes, this book presents important theoretical and empirical contributions to the study of democratic consolidation, legislative organization, and public opinion. Theoretically, Simone Wegmann brings legislatures into focus as the main body representing both winners and losers of democratic elections. Empirically, Wegmann shows that the degree of policy-making power of opposition players varies considerably between countries. Using survey data from the CSES, the ESS, and the LAPOP and systematically analyzing more than 50 legislatures across the world and the specific rights they grant to opposition players during the policy-making process, Wegmann demonstrates that neglecting the curial role of the legislature in a democratic setting can only lead to an incomplete assessment of the importance of institutions for democratic consolidation. The Power of Opposition will be of great interest to scholars of comparative politics, especially those working on questions related to legislative organization, democratic consolidation, and/or public opinion. Simone Wegmann is a post-doctoral researcher at the University of Potsdam, Germany. She currently researches the role of legislatures for democratic consolidation and the influence of legislative organization on public opinion and government formation. Her research has appeared or is forthcoming in a variety of journals, including the American Political Science Review, Democratization, the International Journal of Human Rights, the Journal of Legislative Studies, and West European Politics.

Routledge Research in Comparative Politics

Transparency and Secrecy in European Democracies Contested Trade-offs Edited by Dorota Mokrosinska Climate Governance across the Globe Pioneers, Leaders and Followers Edited by Rüdiger K.W. Wurzel, Mikael Skou Andersen and Paul Tobin The Future of Political Leadership in the Digital Age Neo-Leadership, Image and Influence Edited by Agnieszka Kasińska-Metryka and Tomasz Gajewski Electoral Participation in Newly Consolidated Democracies Turnout in Africa, Latin America, East Asia, and Post-Communist Europe Elvis Bisong Tambe Exploring Organized Interests in Post-Communist Policy-Making The “Missing Link” Edited by Michael Dobbins and Rafał Riedel Minimum Wage Regimes Statutory Regulation, Collective Bargaining and Adequate Levels Edited by Irene Dingeldey, Damian Grimshaw and Thorsten Schulten The Power of Opposition How Legislative Organization Influences Democratic Consolidation Simone Wegmann For more information about this series, please visit: https://www.routledge​.com/ Routledge-Research-in-Comparative-Politics/book-series/CP

The Power of Opposition

How Legislative Organization Influences Democratic Consolidation

Simone Wegmann

First published 2023 by Routledge 605 Third Avenue, New York, NY 10158 and by Routledge 4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2023 Taylor & Francis The right of Simone Wegmann to be identified as author of this work has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in- Publication Data A catalog record for this title has been requested ISBN: 978 - 0 -367- 43731-2 (hbk) ISBN: 978 -1- 032-28245-9 (pbk) ISBN: 978 -1- 003- 00536 - 0 (ebk) DOI: 10.4324/9781003005360 Typeset in Times New Roman by codeMantra

Contents

List of Figures ix List of Tables xi Acknowledgments xiii PART I

The Moderating Role of Parliamentary Procedures

1

1 Institutions, the Role of Opposition, and Democratic Consolidation 3 1.1  The Role of Citizens 5 1.2  The Role of Legislatures 8 1.3  Key Questions and Contribution 13 1.4  Outline of the Book 15 2 The Role of Parliamentary Procedures 28 2.1  The Importance of Legislatures – Moving beyond Governmental Organization 31 2.2  The Importance of Winners and Losers 32 2.3  The Influence of Parliamentary Procedures on Democratic Consolidation 33 2.4 Summary 40 3 Measuring Opposition Power 47 3.1  Policy-Making Power of Opposition Players 48 3.2  Method of Data Collection – Policy-Making Power of Opposition Players 58 3.3 Summary 59

viii Contents PART II

Policy-Making Power of Opposition Players and Its Consequences for Democratic Consolidation

65

4 Methods and Data 67 4.1  Empirical Strategy 67 4.2  I ndependent Variable: Policy-Making Power of Opposition Players, Winners and Losers 70 4.3  Dependent Variable: Consolidation – Political Support 71 4.4  Control Variables 75 4.5 Sample 84 4.6 Summary 88 5 Policy-Making Power of Opposition Players around the World 95 5.1  How Much Policy-Making Power Do Opposition Players Have? 96 5.2  W hat Kind of Policy-Making Power Do Opposition Players Have? 110 5.3 Conclusion 118 6 Policy-Making Power of Opposition Players and Democratic Consolidation 122 6.1  Democracy as the Best Form of Government 123 6.2  Satisfaction with Democracy 129 6.3  Satisfaction with Government Performance 142 6.4 Conclusion 150 7 Why Should We Care About Opposition Power? Conclusion and Implications 156 7.1  Main Findings 157 7.2  The Study of Institutions, Oppositions, and Democratic Consolidation – Limitations and Future Research 161 7.3  Last Thoughts 164 Appendix 167 Index 185

Figures

2.1 Democratic Consolidation: The Winner-Loser Gap 34 2.2 The Moderating Role of Parliamentary Procedures 35 3.1 Policy-Making Power of Opposition Players 49 5.1 Regional Distribution of Regime Types 97 5.2  Distribution of Policy-Making Power of Opposition Players 98 5.3  Policy-Making Power of Opposition Players by Regime Type 98 5.4  Policy-Making Power of Opposition Players by Democratic Experience 100 5.5 Bill Introduction by Regime Type 102 5.6 Agenda Setting by Regime Type 103 5.7 Amendments by Regime Type 105 5.8 Committee Structure by Regime Type 106 5.9 Committee Procedures by Regime Type 107 5.10 Executive Power by Regime Type 108 5.11 Referendum by Regime Type 109 Correlations of Initiation, Debate, and Veto Power of 5.12  Opposition Players 111 Distribution of Policy-Making Power of 5.13  Opposition Players 111 Distribution of Policy-Making Power of Opposition 5.14  Players – Slovenia 113 Distribution of Policy-Making Power of Opposition 5.15  Players – Chile 114 5.16  Distribution of Policy-Making Power of Opposition Players - Czech Republic 116 Predicted Probabilities: Democracy as Best Form of 6.1  Government 127 Winner-Loser Gap: Democracy as Best 6.2  Form of Government 129 Predicted Probabilities: Satisfaction with 6.3  Democracy. CSES 136

x Figures

6.4  Winner-Loser Gap: Satisfaction with Democracy. CSES 137 6.5  Predicted Probabilities: Satisfaction with Democracy. ESS 138 6.6  Winner-Loser Gap: Satisfaction with Democracy. ESS 139 Predicted Probabilities: Satisfaction with Democracy. 6.7  LAPOP 140 6.8 Satisfaction with Government Performance. CSES 146 Winner-Loser Gap: Satisfaction with Government 6.9  Performance. CSES 147 6.10 Satisfaction with Government Performance. ESS 148 Winner-Loser Gap: Satisfaction with Government 6.11  Performance. ESS 149 A.3.1 Policy-Making Power of Opposition Parties – 167 Aggregation Rules

Tables

4.1 Correlations – Variables of Political Support 75 4.2 Sample Selection: Chapter 5 85 4.3 Sample Selection: Chapter 6 87 6.1 Democracy as Best Form of Government 125 6.2 Satisfaction with Democracy 132 6.3 Satisfaction with Government 144 A.4.1 Overview Datasets – Country Variables 169 A.4.2 Overview Datasets – Individual Variables 172 A.5.1 Sample Information 177 A.5.2 Correlations of Variables 178 A.6.1 Descriptive Statistics: Democracy is Preferable 179 A.6.2 Descriptive Statistics: Satisfaction with Democracy 180 A.6.3 Descriptive Statistics: Satisfaction with Government Performance182

Acknowledgments

The desire to understand why some countries experience democratic stability and others do not and what can be done to encourage democratic consolidation motivated me to write this book. Nine years ago, I began to think about this question, and without the support of many people this book would not have been possible. First of all, I thank Simon Hug, my thesis supervisor, for giving me the opportunity to write a PhD thesis which was the starting point of this book. I am immensely grateful for all of his support and patience. Not only did he provide very constructive feedback but he always encouraged and supported me throughout my academic career. I also want to thank the other members of my thesis committee: Stefanie Bailer, Catherine de Vries, and Jonas Pontusson. I am honored to have received their thoughts and suggestions that improved my PhD thesis and thereby also this manuscript tremendously. Throughout the writing process, I have benefited greatly from conversations with and support from colleagues. I especially wish to thank Damien Bol, Cristiane Carneiro, José Antonio Cheibub, Holger Döring, Aurélien Evequoz, Anna Fruhstorfer, Steffen Ganghof, France Gaudreault, Nathalie Giger, William Heller, Johan Hellström, Carl Henrik Knutsen, Philip Manow, Shane Martin, Valentin Schröder, Kaare Strøm, Or Tuttnauer, Aiko Wagner, and Radoslaw Zubek. Thank you for your very helpful comments and encouraging words that gave me the energy to read and code more than 150 parliamentary standing orders and constitutions. I thank Christoph Koch who provided invaluable research assistance during the data collection process. I am indebted to Agustina Rayes and Fabian Schmid who assisted me in collecting data based on official documents in Spanish. Without their support, I would not have been able to include Latin American countries in my analyses. I would also like to thank Anda Filip from the Inter-Parliamentary Union for her help to get access to parliamentary standing orders. This manuscript has definitely benefited from comments by Marianna Hudson and Cade Roberts. Thank you for making this book more reader friendly.

xiv Acknowledgments

I  also thank the two anonymous reviewers for their helpful comments and suggestions. During this scientific journey, I have also received very generous financial support. I thank the Swiss National Science Foundation for providing generous financial support for my research stay at Columbia University in 2015 and the University of Bremen in 2018 and 2019. I also thank the Bureau of Equality of the University of Geneva for granting me a subside tremplin from January to June 2016. In this regard, a special thank you to Mireille Betrancourt who provided invaluable support and assistance. I am incredibly thankful to my family and friends. Without them, this book would not have been possible: Andreas, Aurélien, Danny, Fabien, France, Kathrin, Kä, Lawrence, Lisa, Lucile, Nadja, and Rebekka. Alexandra and Heidrun, thank you for coffee, chocolates, and incredible support during this journey. Romana, Migel, Gabriel, and Tiago, thank you for always helping me focus on what really counts. Maike and Ferdi, thank you for a home away from home. Claudia, Sven, Eliah, Marie, Robert, and Valentina, thank you for your perspective. Michael, Brisch, and Sophie, thank you for all you have done and still do. You are truly amazing! Grosi, you are not with us anymore, but your care packages always arrived when needed most. Beny and Rolf, thank you for haircuts and so much more. Pa, Ma, and Hanspeter, thank you for your unconditional support. Originally, I planned to submit the final version of this manuscript in May 2020 when the world was turned upside down. I want to emphasize that the reason why I finally managed to complete this book during a pandemic is not some sort of superpower but the help of many people: I am incredibly thankful to Natalja Mortensen and Charlie Baker who, beyond their amazing support, granted so many extensions for submitting this manuscript. I also thank Steffen Ganghof and all of my colleagues at the chair of comparative politics at the University of Potsdam: Sebastian Eppner, Anna Fruhstorfer, Alexander Pörschke, and Aiko Wagner. Without their kindness and flexibility, I would not have been able to finish this manuscript. Most importantly, I thank my daughter Lea who managed this difficult time so amazingly (like so many other children. You all are the true heroes of our time!). There are no words to express how much it meant to me every time she sat beside me at the desk, both of us working and she doing an incredible job at not talking. What an amazing four-year-old girl you are! I also thank my unborn son for doing his part by waiting to see the world until this manuscript was finished. I cannot wait to finally hold you in my arms! I thank my mother who moved in with us during the last weeks before submission while I tried to handle a pandemic, online teaching and grading, a pregnancy in its final weeks, moving to Berlin, and finalizing a book manuscript all at the same time. Without her, I would have failed miserably. Finally, I thank Jan who never stopped believing in me. This one is for you.

Part I

The Moderating Role of Parliamentary Procedures

Chapter 1

Institutions, the Role of Opposition, and Democratic Consolidation

Democracies are fragile. In recent times, the classical question of democratic stability and democratic support among citizens has again become widely discussed. Within these discussions, it has become clear that even established democracies are by no means safe from threats of democratic decline or breakdown. Some contributing works in these discussions point to increases in dissatisfaction with democracy and anti-democratic opinions among citizens in democratic regimes. While some authors point to a decreasing level of support for democratic regimes in well-established and stable democracies and a wave of autocratization and democracies in crisis around the globe (see, e.g. Lührmann et al. 2019; Przeworski 2019; Foa and Mounk 2016, 2017), others emphasize that no such decrease in the level of support for democracy exists (Voeten 2016). Undoubtedly, recent debates have increasingly been oriented toward the question of whether democracy is in crisis and whether the level of dissatisfaction with democracy among citizens has increased. But, for the survival and stability of democracies, it is essential to not only question if the level of dissatisfaction with democratic regimes has increased but also why such dissatisfaction occurs in the first place. Research questions related to the organization and performance of different political systems, including democracies, are embedded in one of the most traditional fields in political science. But despite this long research tradition, recent developments show that questions about the determinants of successful democratic consolidation and stability remain highly relevant. This is due to the fact that an initiated process of democratization is by no means a sufficient condition for democratic consolidation and stability. Also, as highlighted above, even established democracies may face increasing levels of dissatisfaction which may ultimately result in threats to their stability. Furthermore, the factors to explain democratic consolidation and stability in political science research are varied and, so far, have not led to clear results. In democracies, citizens are represented and, to varying degrees, included in the decision-making process. So why do citizens in some polities DOI: 10.4324/9781003005360-2

4  Moderating Role of Parliamentary Procedures

express dissatisfaction with the functioning of the democratic regime or even prefer non-democratic forms of government while others do not? This question lies at the heart of the present book. It thereby takes on a classical question in political science that has again become widely discussed in recent times. However, I propose a possible answer to this question that lies within the different institutional settings of democracies themselves but has been neglected by most research: the role of the power granted to opposition players to influence the policy-making process. Focusing on the relationship between institutions and democratic consolidation is of particular relevance. It relates to the extent to which institutions influence public opinion (i.e. dissatisfaction with democracy) which, in turn, might lead to protest and instability. Research has repeatedly pointed to the existence of a winner-loser gap when it comes to satisfaction with democracy (see, e.g. Bernauer and Vatter 2012; Anderson et al. 2005; Anderson and Guillory 1997). If we see losers as crucial actors when it comes to democratic consolidation, assessing how political institutions might affect citizen support for democracy is merely impossible without a detailed look at the roles of these losers, both within democratic institutions and among citizens. So far, however, the research on the role of institutions for democratic consolidation and stability neglects two important aspects: the role of individual citizens and the legislature. On the one hand, studies on the role of institutions for democratic consolidation often neglect the role of individual citizens. This is extremely surprising as the democratization literature repeatedly mentions this dimension of the individual in definitions of democratic consolidation and stability. For example, Linz and Stepan (1996: 15) define a consolidated democracy as “(…) a political regime in which democracy as a complex system of institutions, rules, and patterned incentives and disincentives has become, in a phrase, ‘the only game in town’.” The authors go on and describe what “the only game in town” means from a behavioral, attitudinal, and constitutional perspective. All three perspectives point toward the crucial role of political actors and citizens to adhere to democratic rules (Linz and Stepan 1996: 15–16). Similarly, Diamond (1994: 15) has emphasized that “(…) consolidation is the process by which democracy becomes so broadly and profoundly legitimate among its citizens that it is very unlikely to break down” (see also Diamond 1999; Pridham and Vanhanen 1994; Linz 1990b). According to these definitions, democratic consolidation is inherently about individual attitudes and opinions. On the other hand, although recognizing the important role of minorities, the research on institutional determinants of democratic consolidation has largely focused on different procedures of government formation (i.e. the difference between parliamentary and presidential regimes or majoritarian and consensus democracies) and neglected the

Institutions, Role of Opposition, and Consolidation  5

legislature as a policy-making institution and central body representing citizens. But, if consolidation is inherently about citizen attitudes and opinions – and democratic consent of losers is especially important for democratic consolidation – focusing on the role of the institution that actually represents citizens (i.e. the legislature) seems imperative. To further elaborate on these shortcomings, this introductory chapter proceeds as follows: In the first step, I outline a brief overview of the concept of democratization as used in the current literature and the role of citizens for democratic consolidation. In the second step, I outline a short overview of the existing work on institutional determinants of democratic stability as well as legislative studies, and outline some key shortcomings. In the third step, I present the key questions this book aims to answer. I finish the chapter with an overview of the outline of the book.

1.1  The Role of Citizens The term democratization has been used in democracy studies for two distinct developments. Beckman (2012: 162), for example, differentiates between democratization as the transition from dictatorship to democracy and democratization as a “gradual development towards the perhaps never fully achievable ideal of democracy” (see also O’Donnell 1992: 19). Hence, democratization does not necessarily only describe a regime change from a non-democratic to a democratic form of government but can also take place within an already democratic regime (see also Schedler 1998: 94). However, the research of the former conception of democratization (i.e. the transition from dictatorship to democracy) has also relied on the concept of democratization further differentiating between two related but somewhat different processes (Haerpfer et al. 2009; Pridham and Vanhanen 1994)1 On the one hand, the concept of democratization can be used to label the process of a democratic transition, meaning a regime change from dictatorship toward democracy (O’Donnell and Schmitter 1986: 6). Following Pridham and Vanhanen (1994: 17), the democratic transition “runs from the point at which the previous authoritarian system begins to be dismantled, through the constituent phase of the new democracy to its inauguration and early operation.” On the other hand, the process of democratization is also about what is referred to as consolidation, meaning the solidification of democratic rules as “the only game in town” (Linz and Stepan 1996; Linz 1990b: 5). Thus, whereas democratic transition involves changing political institutions, democratic consolidation involves “the full institutionalization of the new democracy, the internalization of its rules and procedures and the dissemination of democratic values” (Pridham and Vanhanen 1994: 2). Put differently, “consolidation is the process by which democracy becomes so broadly and profoundly

6  Moderating Role of Parliamentary Procedures

legitimate among its citizens that it is very unlikely to break down” (Diamond 1994: 15). In a subsequent work, Diamond (1999:65) presents a more detailed definition: (…) consolidation is most usefully constructed as the process of achieving broad and deep legitimation, such that all significant political actors, at both the elite and mass levels, believe that the democratic regime is the most right and appropriate for their society, better than any other realistic alternative they can imagine. (…) At the mass level, there must be a broad normative and behavioral consensus – one that cuts across class, ethnic, nationality, and other cleavages – on the legitimacy of the constitutional system, however poor or unsatisfying its performance may be at any point in time. A consolidated democracy will ultimately also be a stable democracy. Hence, a stable democracy is a consolidated democracy in which the probability of democratic backsliding is minimized, or put differently, the probability of preventing democratic breakdown is maximized. Following O’Donnell (1992: 19), such democratic breakdown not only comes as “rapid deaths” of democracy in the form of classical coups but also in the form of “slow deaths” in the form of “successive authoritarian advances.” Hence, democratic stability here refers to the “negative notion of consolidation” (Schedler 1998: 103) meaning the prevention of democratic breakdown and not the deepening of democracy, in general (positive consolidation). Now, the minor difference between democratic stability and consolidation becomes apparent. Consolidation refers to its “(…) original mission of rendering democracy ‘the only game in town’ (…)” (Schedler 1998: 91). Democratic stability, however, refers to the reduction of the “(…) probability of [democratic] breakdown to the point where they [proponents of democracy] can feel reasonably confident that democracy will persist in the near (and not-so-near) future” (Schedler 1998: 95). On the individual level, a consolidated democracy is a political regime in which individuals agree that democracy is the best form of government (“the only game in town”). Moreover, the satisfaction with the way the democratic regime is working in a given country also influences the willingness or unwillingness of its citizens to defend it. This willingness or unwillingness to defend democracy, in turn, influences the level of consolidation of democracy and, ultimately, its stability.2 It becomes evident that considering political attitudes is of utmost importance for the studies of democratic consolidation and stability. Unsurprisingly, research on the influence of mass beliefs on democratic stability (see, for example, Claassen 2019; Welzel and Inglehart 2009; Diamond 1999; Lipset 1959) identifies individual political support as a

Institutions, Role of Opposition, and Consolidation  7

crucial element for successful democracy. Various studies focus on this role of individual-level democratic legitimacy in the democratic consolidation and stability of countries. For example, Schmitter and Karl (1991) have emphasized that neither detailed definitions of democracy and democratization nor theories of democratization and democratic stability help to understand how democracy actually functions. According to the authors, “[t]he simplest answer [to this question] is ‘by the consent of the people’” (Schmitter and Karl 1991: 82). Similarly, Welzel and Inglehart (2009: 126) argue that “mass beliefs are of critical importance for a country’s chances to become and remain democratic” (see as well Welzel 2006; Inglehart and Welzel 2005; Inglehart 1997). “When democracy is valued as a good in itself” (Welzel and Inglehart 2009: 129), citizens are likely to be able to influence elites, both in the direction to democratic transition or to democratic stability (see also Welzel and Inglehart 2009; Welzel 2007; Bratton and Mattes 2001; Anderson and Guillory 1997; Inglehart 1997; Evans and Whitefield 1995). But, this argument about the importance of citizens’ consent to democracy is by no means a recent one. Lipset (1959) has already highlighted the importance of the legitimacy of the political system. The shared belief among different groups that the existing institutions are the most appropriate is a crucial aspect in defining the success of a democratic regime (Lipset 1959: 86). According to Eckstein (1966: 1), political regimes can only become stable when there is congruence between the authority pattern of a system and the authority beliefs of the people (see as well Eckstein 1998). This statement is in line with Diamond (1999: 168) highlighting that “it is by now a cardinal tenet of empirical democratic theory that stable democracy also requires a belief in the legitimacy of democracy.” Hence, understanding democratic consolidation or examining the factors contributing to it seems incomplete without considering the dimension of individual attitudes toward democracy itself. In this regard, previous research especially emphasizes the role of losers of democratic elections. Accordingly, the consolidation of democracy is not only dependent on winners of elections but the consent of losers is also crucial for the legitimacy and the survival of a democratic regime. Przeworski (1991: 15), for example, has emphasized that the central question in consolidation studies is why losers “(…) comply with the outcomes [of elections] and continue to participate rather than subvert democratic institutions” (see as well Anderson et al. 2005)? It is no surprise, then, that a long research tradition has focused on the influence of different institutional settings on public support for democracy. Anderson et al. (2005), for example, attribute several institutional features to more legitimacy expressed by electoral losers. However, this research has also largely focused on broad categorizations of countries,

8  Moderating Role of Parliamentary Procedures

such as considering different electoral systems, the parliamentary-­ presidential distinction (e.g. Anderson et al. 2005), or direct democracy (see, e.g. Bernauer and Vatter 2012). Yet, so far, the literature on individual political opinions has not considered the legislature among the institutions that influence democratic consolidation. This is surprising because, if we identify democratic consent of losers to be of crucial importance for democratic consolidation, we should also include the institution where these losers are represented in our analyses, which is the legislature. I turn to this point in the next section.

1.2  The Role of Legislatures Research on institutional determinants of democratic consolidation and stability has widely been organized around the question of government formation. In this connection, research has largely focused on the influence of presidential and parliamentary regimes on democratic stability and the question of which of these two types of government formation better accommodates cleavages and different preferences in society (and thereby influences democratic consolidation among citizens). The literature on the influence of these two types of executive-legislative relations, however, shows rather conflicting results on whether parliamentary or presidential systems favor more stability (see, e.g. Shugart and Carey 1992; Linz 1990a, Linz 1990b). Following Elgie (2005), the research on presidential regimes and the stability of democracy can be categorized into several waves. Linz’s contribution on the “perils of presidentialism” marked a first wave focusing on the question of why presidential regimes are less stable than parliamentary ones (Linz 1994, 1990a; Linz and Valenzuela 1994). Linz (1990a: 53) attributes the instability of presidential systems to the fact that, in a presidential system, the president and the majority in parliament can represent different parties which can both claim democratic legitimacy as both have been elected by citizens. This situation leads to conflict between the executive and the legislature. Furthermore, Linz (1990a: 53) criticizes the fixed term of presidents in presidential regimes “leaving no room for the continuous readjustments that events may demand.” However, Linz (1990a) has also highlighted that parliamentary regimes, too, can be unstable and that this is especially true under conditions of ethnic conflict. Still, overall, Linz concludes that parliamentary regimes are more conducive to democratic stability than presidential regimes, even in states with deep political cleavages and multiple political parties (Linz 1990a: 52). Similarly, the comparison of parliamentary and presidential systems by Stepan and Skach (1994; 1993) shows several features of parliamentary systems that make them more stable than presidential systems. These

Institutions, Role of Opposition, and Consolidation  9

features include, among others, a greater “propensity of governments to have majorities to implement their programs,” a “greater ability to rule in a multi-party setting,” and a “greater tendency to provide long party or government careers, which add loyalty and experience to political society” (Stepan and Skach 1993: 22). Similarly, Lijphart (2008; 1994) and Przeworski et al. (1996) have argued that parliamentary systems are more likely to lead to stable democracies than presidential systems. Or, put differently, that “presidentialism seems to involve greater risk for stable democratic politics than contemporary parliamentarism” (Linz 1994: 70). Later work, focusing on the presidentialism-parliamentarism distinction, has gone beyond the problem of the dual legitimacy inherent in presidential regimes as emphasized by Linz’s work. Scholars in this second wave do not focus on the sole characteristic of presidentialism and parliamentarism as an explanatory factor of stability. Mainwaring and Shugart (1997b: 449), for example, show that presidential systems are likely to function better “where presidencies have weak legislative powers, parties are at least moderately disciplined, and party systems are not highly fragmented” (see also Shugart and Carey 1992). However, Mainwaring and Shugart (1997c) and Power and Gasiorowski (1997) do not find any support for this thesis. Beliaev (2006: 394) shows that “[s]trong executive powers of presidents corresponded with lesser advances in terms of democratic consolidation” (see also Van Cranenburgh 2008). Hence, it seems to be that presidential systems, especially, in combination with some other features, such as multi-party systems (Mainwaring 1993: 210–222), are less likely to lead to stable democracy. Similarly, Cheibub (2007, 2002) has argued that it is not presidential systems per se that are less likely to be stable democracies, but rather the fact that presidential systems “have existed in countries where the environment is inhospitable for any kind of democratic regime” (Cheibub 2007: 136). Furthermore, critics of Linz’s conclusion that presidential systems are less stable than parliamentary ones have emphasized that coalition formation is not only possible in parliamentary regimes but equally so in presidential ones (see, e.g. Horowitz 1990: 75). Indeed, Cheibub (2002: 165) shows that the party of the president in presidential regimes and the prime minister in parliamentary ones show a similar distribution of legislative seats (for coalitions in parliamentary and presidential systems, see also Chaisty, Cheeseman and Power 2014; Raile, Pereira and Power 2011; Colomer and Negretto 2005; Cheibub, Przeworski and Saiegh 2004). Several scholars have also emphasized that the relationship between regime type and democratic stability is not due to the specific democratic institutions but to other factors. Shugart and Mainwaring (1997) mention that presidential systems predominantly exist in Latin America, whereas

10  Moderating Role of Parliamentary Procedures

parliamentary systems exist in Europe and Africa making it difficult to isolate the effect of these specific institutions. Przeworski et al. (2000) furthermore highlight that parliamentary systems are more frequent among the wealthier countries where, in turn, democratic stability is more likely. However, in the end the authors conclude that “[p]residential democracies are simply more brittle under all economic and political conditions” (Przeworski et al. 2000: 136). Elgie (2005) identifies a third wave including more recent contributions of “more general theories of political science” (Elgie 2005: 107), where the presidentialism-parliamentarism distinction is no longer the sole focus. Tsebelis’ work on veto players is one of the most well-known among these contributions. Through their influence on policy stability, veto players also influence government stability which can ultimately also affect democratic stability (see Tsebelis 2002: 209–221). Tsebelis’ approach goes beyond the usual presidentialism-parliamentarism dichotomy or the one between two-party and multi-party systems. What is more important is the number of veto players as well as their ideological distance from one another (Tsebelis 1999: 591). This is why Tsebelis (1995: 315) argues that similar features of policy stability are possible across different constitutional regimes. Furthermore, the research has increasingly pointed to the fact that the United States and the United Kingdom are among the few countries that come close to the ideal types of parliamentarism and presidentialism, respectively. Therefore, “collapsing the notion of executive-legislative relations into two categories (…) assumes away much of what is essential to understanding how the chain of democratic delegation and accountability is characterized by degrees of hierarchy and transaction” (Shugart 2006: 360).3 Cheibub, Elkins, and Ginsburg (2014) analyze national constitutions with reference to different characteristics of parliamentary, semi-presidential, and presidential regimes. These authors conclude that “[m]any countries, it seems, are veritable hybrids, showing absolutely no resemblance to the classic types across a long list of constitutional provisions concerning the power of executives and legislatures” (Cheibub et al. 2014: 539). Hence, the classification into either parliamentary or presidential systems, even when including the category of semi-presidential systems, does not account for the fact that the executive-legislative power relation within these systems can vary considerably and that a focus on government formation neglects the importance of a second crucial body in functioning democracies: the legislature.4 Although the importance of the legislature for democratic consolidation and stability has been highlighted repeatedly,5 surprisingly little research has connected the literature on democratic consolidation and legislative organization. Rather, research on democratization has primarily been concerned with

Institutions, Role of Opposition, and Consolidation  11

questions of government formation and other institutional aspects and their influence on public support and regime stability.6 However, not only have democracy studies paid little attention to the role of the legislature for democratic consolidation, but legislative studies have also paid little attention to the influence of legislative characteristics on extra-legislative outcomes (see Fortunato 2021). Legislative scholars have largely focused on the US Congress and legislatures in established democracies7 and have been concerned with explaining the origins of different legislative rules as well as the working of specific parliamentary institutions and consequences for legislative behavior. This latter string of research, for example, has focused on the committee system (e.g. Fernandes, Goplerud and Won 2019; McGrath and Ryan 2019; Fernandes 2016; Khmelko, Wise and Brown 2010), voting procedures (e.g. Crisp and Driscoll 2012; Carey 2009, 2007), question time (e.g. Martin 2011b; Rozenberg and Martin 2011), speaker selection (e.g. Spicer and Nater 2013), or agenda setting, more generally (e.g. Bräuninger and Debus 2009). More important for the research question addressed in this book, however, is the question about the specific organization of legislative chambers. Without constraining rules in the legislative chamber, we would observe an impassable plenary bottleneck (Cox 2006). As plenary time is scarce, legislatures have resorted to such constraining rules to make themselves efficient (Cox 1987). Consequently, whereas voting power in legislatures is equal, proposal and veto power are not (Cox and McCubbins 2005: 9). But, legislative chambers around the world have adopted different rules to regulate their organization and work. As these legislative procedures are chosen and can be modified by the chambers themselves, we indeed see a great variety of such rules (see, e.g. Sieberer et al. 2016; Diermeier, Prato and Vlaicu 2015; Müller and Sieberer 2014; Martin 2011a; Sieberer, Müller and Heller 2011; Döring 1995). Different explanations have emerged in the US context to explain why MPs accept constraints on their rights in legislative chambers. Later work has also adopted these explanations to the European context (for an overview see, e.g. Martin 2014; Müller and Sieberer 2014). Early work on the US context, for example, has explained the existence of legislative committees with distributive goals of MPs. Accordingly, committees exist to allow MPs to distribute policies and benefits to their constituencies and thereby aid MPs to get re-elected (see, e.g. Shepsle and Weingast 1981,1987; Shepsle 1979). Challenging the distributional theory of legislative organization, the informational theory focuses on policy expertise of the legislative chamber. Gilligan and Krehbiel (1990), for example, suggest that legislative committees are organized to achieve informational needs and policy expertise of the legislative chamber (see also Krehbiel 1992). A third string of explanations has focused on partisan goals of the majority party. Accordingly, legislative chambers are organized in a

12  Moderating Role of Parliamentary Procedures

way as to benefit a party’s needs to adopt policies and minimize defection (Cox and McCubbins 2005, 1994, 1993). Finally, research has also focused on the effect of the ideological balance of power in legislative chambers and the role of second chambers (Schickler 2000; Diermeier and Myerson 1999). A large literature exists, building on these explanations, to analyze different rules of procedure in the European context. A vast part of this research remains based on case studies focusing on legislative organization and development in individual countries (e.g. Goet 2021; Goet, Fleming and Zubek 2020; Shomer 2015; Eggers and Spirling 2014; Capano and Flinders 2007; Binder 2006; Leston-Bandeira and Freire 2003; Giuliani 2003; Loewenberg 2003; Norton 2003). However, more recent contributions show a comparative character (e.g. Sieberer and Höhmann 2021; Wegmann 2022; Koß 2018, 2015; Garritzmann 2017; André, Depauw and Martin 2016; Hug, Wegmann and Wüest 2015; Sieberer and Müller 2015; Zubeck 2015; Carroll, Cox and Pachón 2006; Taylor 2006; Döring 1995). One recurring result from this research points to the influence of the electoral system and the size of the legislative chamber on legislative organization (see also Martin 2018; Carey 2009; Taylor 2006; Powell 2000). I discuss implications of these findings for the degree of opposition power in more detail in Chapter 5. However, a third research area within legislative studies which focuses on the extra-legislative effect of legislative rules has only recently emerged. As Fortunato (2021: 3–4) has highlighted, legislative studies have largely been concerned with the understanding of the legislative process and its consequences for legislative behavior and governance and less concerned with extra-legislative consequences. Recent research in legislative studies that actually considers such extra-legislative effects focuses on policy outcomes (Fortunato, Martin and Vanberg 2017; Diermeier, Prato and Vlaicu 2015), government formation (Cheibub, Martin and Rasch 2021; Rasch, Martin and Cheibub 2015; Martin and Vanberg 2011; Strøm 1990), or electoral consequences (Bouteca et al. 2019; Williams and Indridason 2018; Däubler, Bräuninger and Brunner 2016; Bowler 2010). Most importantly, this research shows that legislative organization actually does influence extra-legislative areas. It is thereby in line with the two other research areas – focusing on the origins of legislative rules and the understanding of the legislative process and its consequences for legislative behavior – pointing to the importance of legislative organization. But, as these examples nicely show, the implications of legislative organization for the larger institutional environment of democracies (such as their consolidation) have not gained much attention in the current literature. Hence, neither democracy nor legislative studies have focused on the potential consequences of legislative organization for democratic consolidation.

Institutions, Role of Opposition, and Consolidation  13

This is surprising as the legislature fulfils several important tasks without which no democracy is able to work properly. The literature has repeatedly mentioned the following three core functions, among others.8 First, the legislature is the institution in which the varied and conflicting interests in a society are represented (Barkan, 2008: 125; Carey, 2006: 431). As Holden (2006: 171) emphasizes “[t]he legislature is the vehicle by which, in theory, everyone has some representative, at least if the design is right.” In this connection, a second function of legislatures is the one of “checking majority and executive power” (Carey 2006: 431) or, put differently, exercising oversight of the government (Barkan 2008: 125). Third, “legislatures legislate” (Barkan 2008: 125) or, as Carey (2006: 431) has emphasized, legislatures constitute the “principal policy-making institutions” in a polity. Even though the majority in parliament plays a more prominent role in this process – as majorities are able to make decisions – policy-making involves all members of the legislature. Therefore, detailed knowledge about the organization of legislatures seems to be of crucial importance to the understanding of the role of different groups for democratic consolidation. As discussed above, the literature in the field of legislative studies has shown that the organization of legislative chambers varies significantly between countries and that different legislative rules affect legislative and extra-legislative behavior. Failing to consider the specific organization of legislatures therefore neglects an important differentiating feature of democratic regimes and excludes one of the most crucial groups for democratic consolidation: oppositions (i.e. losers of democratic elections).

1.3  Key Questions and Contribution This book addresses the gaps in the current research outlined in the previous sections. This book’s goal is, first, to combine the research on democratic consolidation and legislative organization. It aims to use the insights generated by the legislative studies literature to understand the importance of legislatures for democratic consolidation. It, thereby, contributes to research in legislative studies analyzing the effect of legislative procedures on extra-legislative outcomes. It further presents a detailed account of opposition rights and thus contributes to the comparative analysis of legislative organization. Second, the book aims to preserve the important distinction between winners and losers of democratic elections and to analyze the question of consolidation from the perspective of individuals. Thereby, it proposes to analyze democratic consolidation closely related to the definition of the concept itself (namely the individual opinions about democracy). The aim is to bridge the presented gaps and contribute to finding an answer to the general question about the determinants of successful democratic consolidation.

14  Moderating Role of Parliamentary Procedures

As the previous sections demonstrate, the legislative studies literature has shown that parliamentary institutions influence behavior (see, e.g. Fortunato 2021, Cox 2000). Furthermore, the literature on political opinion has shown that institutions (such as parliamentary and presidential regimes or electoral systems) shape individual opinions about democracy. In addition, parliamentary institutions vary considerably between countries (see, e.g. Wegmann 2022; Sieberer et al. 2016). Hence, these differences in legislative institutions should also be reflected in different levels of political support (i.e. democratic consolidation). It might be the specific institutional setting of legislatures itself that, to some extent, explains why citizens – especially losers – in some polities express dissatisfaction with the functioning of democracy while others do not. Therefore, in this book I, more specifically, investigate the question to what extent different characteristics of legislative organization affect the individual perception of democracy and consent to democratic rules. Connecting political institutions and individual attitudes about the political system is particularly relevant as it allows to address the potential for instability caused by dissatisfaction of the losers of the electoral game (Anderson and Guillory 1997: 67). As previously mentioned, the central question in consolidation studies is why losers accept election outcomes and continue to comply with democratic rules (Przeworski 1991: 15). Knowing whether a particular design of institutions can contribute to the moderation of potential protests and increase the likelihood of losers to accept democracy as “the only game in town” is highly relevant to the success of democratic regimes around the world. On the theoretical side, I bring legislatures into focus by moving beyond the traditional concern about government formation largely applied in studies on democratic consolidation and stability. Institutions organize the polity and “[t]hey affect how political actors are enabled or constrained and the governing capacities of a political system” (March and Olsen 2006: 8; see as well Weaver 1993; March and Olsen 1984). I build upon the vast literature in legislative studies and look at the legislature as a central institution in the policy-making process and its link to the success of democratic consolidation. Whether or not losers are willing to accept electoral outcomes might depend on their rights of opposition in the legislature. I argue that parliamentary procedures change the degree of political equality between opposition and government in legislatures thereby influencing the cost of electoral losses. I combine the literature on rational choice approaches to democratization, government formation, and losers’ consent to develop a rational accounting for the consolidation of democratic regimes. The focus on national legislatures as the central arena where opposition players can influence the policy-making process is crucial. As Carey (2006: 431) has emphasized, “legislatures are (…) the principal policy-making

Institutions, Role of Opposition, and Consolidation  15

institutions in modern democracies. The most fundamental policy decisions (…) all must be approved by legislatures.” So far, research on democratic consolidation and stability does not account for the different opposition rights in legislatures. Similarly, research on winners and losers, so far, does not take into consideration the fact that different polities might guarantee different levels of influence to opposition players (and thereby losers). The level of losers’ consent might be linked to the degree of rights losers (i.e. opposition actors) are granted in the legislature. Therefore, considering the different parliamentary procedures defining the rights of opposition players might generate new insights regarding the importance of institutions for democratic consolidation. On the empirical side, this book shows in the first part that the degree of power of opposition players varies considerably between countries. For example, parliamentary, semi-presidential, and presidential regimes do not show the degrees of power of opposition players clearly distinguishing these regime types. Rather, parliamentary regimes, especially, show a wide range of degrees of power of opposition players. Even though the government formation process varies across these regime types (i.e. parliamentary and presidential regimes), they might have much in common when looking at the rights of opposition players in the policy-making process. By presenting detailed data on legislative organization from the perspective of opposition players, this book contributes to the knowledge about differences in legislative rules among democratic regimes. The results highlight the great variance in legislative organization between countries and are in line with previous research in legislative studies pointing to such differences in legislative rules. This book shows in the second part that the different levels of political power of opposition players influence political support of citizens. Results indicate that the strong power of opposition players positively influences both losers and winners. Through its influence on political support of these groups, strong political power of opposition players contributes to the consolidation of democratic regimes. Results emphasize that neglecting the crucial role of the legislature in a democratic setting can only lead to an incomplete assessment of the importance of institutions for democratic consolidation. It thereby contributes to democracy studies by highlighting the important role of the legislature for democratic consolidation. Furthermore, the research presented in this book contributes to the emerging literature in legislative studies emphasizing the extra-­ legislative effect of legislative rules.

1.4  Outline of the Book This book is organized into two main parts. In Part I The Moderating Role of Parliamentary Procedures, Chapters 2 and 3, I include the

16  Moderating Role of Parliamentary Procedures

theoretical aspects of the book. In Part II about Policy-Making Power of Opposition Players and Its Consequences for Democratic Consolidation, Chapters 4–6, I focus on the results of the data collection effort to assess different levels of policy-making power of opposition players and its influence on democratic consolidation. In Chapter 2, I present the theoretical argument of the book. The chapter includes detailed discussions of the key concepts and findings in the extant literature upon which the theoretical argument is based. The theoretical argument presented in Chapter 2 combines the insights from legislative studies and those from the research on democratic stability and public support. The theoretical argument is based on the assumption that the existence of highly dissatisfied groups is dangerous to democratic regimes. Furthermore, the theoretical argument follows the logic of different costs of losing democratic elections for different groups in society (see, e.g. Przeworski 2019: 154). I argue that one element in determining the consequences of losing elections is the degree of political power granted to losers. Among others, it is parliamentary procedures that grant different degrees of such power to opposition players (i.e. losers). The degree of power opposition players can exert in the legislature can reduce the inequality between winners and losers, leading to more political support for the democratic regime. Hence, this specific allocation of power in the legislative chamber (determined by parliamentary procedures) to winners and losers influences the level of support these groups express and thereby the degree of democratic consolidation. Chapter 3 is dedicated to the question of how to measure opposition power. I propose an index to measure policy-making power of opposition players. The measure presented in Chapter 3 combines two core functions of legislatures: policy-making and checking majority and executive power. The operationalization is based on Strøm´s (1990) concept of a policy influence differential (Strøm 1990: 42) resulting from internal parliamentary structures and procedures. A higher policy influence differential indicates more power of the government as compared to the opposition. Hence, payoffs for policy-seeking oppositions are bigger in a polity with a lower policy influence differential. The structure of the concept of policy-making power of opposition players follows three stages of the policy-making process: initiation, debate, and veto. A large research tradition exists on the specific rules and procedures of the policy-­making process which serves as a guideline to develop the final index. Even though most of this research focuses on the role of the government, it allows for the identification of the policy influence differential and thereby the assessment of the power granted to opposition players. The final index combines seven variables: bill introduction, agenda setting, amendments, committee structure, committee procedures, executive veto, and referendums. Data collection was based on documents officially published by the state

Institutions, Role of Opposition, and Consolidation  17

institutions. Basic reference documents for each country are the official rules of procedure of the parliamentary chamber and the national constitution. These were complemented by other data sources where necessary.9 In Chapter 4, I present the empirical strategy and data used to analyze the influence of policy-making power of opposition players on democratic consolidation. I have chosen to devote a separate chapter to this topic in order to not interrupt the empirical chapters with discussions of methodological strategies and choices of operationalization of variables. The chapter includes five parts. First, I present the empirical strategy to test the connection between policy-making power of opposition players and democratic consolidation. Second, I present the independent variable (policy-making power of opposition players) and discuss the approach to identify winners and losers. Third, I present a detailed discussion of the operationalization of the dependent variable (democratic consolidation). The approach followed in this book is based on Easton’s (1975) differentiation between diffuse and specific support. The fourth part presents all control variables included in the empirical chapters. Control variables were included based on theoretical arguments and findings in the extant literature on public support and democratic consolidation. Finally, I discuss the criteria of sample selection. The aim of Chapter 4 is to provide all necessary information for the empirical Chapters 5 and 6. In a first descriptive analysis, in Chapter 5 I present results with reference to the data collection effort to measure policy-making power of opposition players. In the first part of the chapter, I ask how much policy-­ making power opposition players have around the world and present a detailed descriptive analysis of all seven variables included in the final index. Results show that the variance of policy-making power granted to opposition players is considerable. Hence, bringing the studies of legislative organization into research on democratic consolidation – and thereby going beyond the central question of government formation – seems to be a promising approach. If descriptive statistics showed little variance in the policy-making power of opposition players, then legislative organization of the policy-making process could not be a key factor in determining democratic consolidation. However, as these descriptive results show considerable variance, looking more closely at how legislatures are organized, and how they include oppositions in the legislative process might be a promising approach to explain different levels of democratic consolidation. In the second part of Chapter 5, I ask the question of what kind of policy-making power opposition players have. I present three countries in more detail: Chile, the Czech Republic, and Slovenia. I focus on these countries as they most closely represent the mean in the sample (Chile) as well as legislative chambers that show interesting combinations of weak and strong opposition power in different dimensions/ variables (the Czech Republic and Slovenia).

18  Moderating Role of Parliamentary Procedures

In Chapter 6, I present the results of the influence of policy-making power of opposition players on democratic consolidation. I include several multi-level regression models on the influence of strong policy-making power of opposition players on individual support for the political regime. The objective of this chapter is to analyze the influence of strong policy-making power of opposition players on winners and losers alike and to assess whether strong policy-making power of opposition players contributes to democratic consolidation. Based on Easton (1975), I present both results of the influence of policy-making power of opposition players on diffuse as well as specific support. More specifically, I include survey questions on democracy as a preferable form of government, satisfaction with democracy, and satisfaction with government performance. To provide as detailed a picture as possible, I include several waves of the Comparative Study of Electoral Systems (CSES 2019), the European Social Survey (ESS 2019), and the Latin American Public Opinion Project (Vanderbilt University 2019). Results show that strong policy-making power of opposition players is significantly correlated with democratic consolidation, especially in older democracies. This book concludes with Chapter 7, where I present the main findings as well as limitations and thoughts about future research.

Notes 1 Some authors even differentiate between more than two processes. Carothers (2002: 7) differentiates between three distinct phases of democratization: opening, breakthrough, and consolidation. Schneider and Schmitter (2004) identify three parts of democratization: liberalization of autocracy, the mode of transition and the consolidation of democracy. 2 This link between satisfaction and stability is largely based on the assumption of Rueschemeyer, Stephens and Stephens (1992: 57) that (…) those who have only to gain from democracy will be its most reliable promoters and defenders, those who have the most to lose will resist it and will be most tempted to roll it back when the occasion presents itself. 3 Similarly, Mainwaring and Shugart (1997a: 435) argue that presidential systems vary so greatly in the powers accorded to the president, the types of party and electoral systems with which they are associated, and the socio-­ economic and historical context in which they were created that these differences are likely to be as important as the oft-assumed dichotomy between presidential and parliamentary system. Haggard and McCubbins (2001: 4) emphasize that “[a]lthough the relative merits and consequences of presidential and parliamentary systems have received substantial attention, we argue that this distinction between macro institutions is inadequate; explaining political outcomes requires greater focus on the details of institutional structure.” 4 I develop this reasoning in more detail in Chapter 2.

Institutions, Role of Opposition, and Consolidation  19 5 For example: “(…) democratic consolidation – the highest stage of ­democratization – cannot be achieved without a developed and powerful legislature” (Barkan 2009a: 2). “Stronger legislatures, stronger democracies” (Fish 2006). 6 See e.g. Barkan (2008: 124) highlighting that “(…) apart from an essay by M. Steven Fish in the Journal of Democracy, neither this journal nor Democratization has published a single article on the relationship between legislative development and democratization.” See also Barkan (2009b: 5). 7 “But the scholarly world of legislative studies is, overwhelmingly, a world that studies the U.S. Congress” (Gamm and Huber 2002: 313). “Another distinguishing feature of the literature on legislative studies is that very little of it, probably no more than 5 percent, is genuinely comparative in scope and approach” (Barkan 2009b: 5). 8 Barkan (2008: 125–126) has mentioned a total of four functions: representative governance, legislation, oversight of the executive, and constituency service. Carey (2006: 431) has mentioned five functions: Representation of diversity, deliberation, cultivation of information and expertise, decisiveness, and checking majority and executive powers. 9 The Comparative Constitutions Project (Elkins, Ginsburgh and Melton 2014) is used for information about bill introduction of different actors. However, as this dataset considers national constitutions only, parliamentary standing orders were consulted to complement the information in case of the actors that did not have the right to introduce legislation according to the national constitution. The Parliamentary Power Index (Fish and Kroenig 2009) is used to determine whether the executive has decree power and the Institutions and Elections Project (Regan and Clark 2014) provides information about executive veto power. Finally, the Direct Democracy Database (IDEA 2014) provides detailed information about the possibilities of referendum and citizen initiatives.

Bibliography Anderson, Christopher J., Andre Blais, Shaun Bowler, Todd Donovan and Ola Listhaug. 2005. Losers’ Consent: Elections and Democratic Legitimacy. Oxford and New York: Oxford University Press. Anderson, Christopher J. and Christine A. Guillory. 1997. “Political Institutions and Satisfaction with Democracy: A Cross-national Analysis of Consensus and Majoritarian Systems.” American Political Science Review 91(1): 66–81. André, Audrey, Sam Depauw and Shane Martin. 2016. “‘Trust Is Good, Control Is Better’ Multiparty Government in Legislative Organization.” Political Research Quarterly 69(1): 108–120. Barkan, Joel D. 2008. “Legislatures on the Rise?” Journal of Democracy 19(2): 124–137. Barkan, Joel D. 2009a. African Legislatures and the “Third Wave” of Democratization. In Legislative Power in Emerging African Democracies, ed. Joel D. Barkan. Boulder, CO: Lynne Rienner Publishers: 1–32. Barkan, Joel D. 2009b. Legislative Power in Emerging African Democracies. Boulder, CO: Lynne Rienner Publishers. Beckman, Lugvig. 2012. Democratization and Inclusion. In Routledge Handbook of Democratization, ed. Jeffrey Haynes. London and New York: Routledge: 161–174.

20  Moderating Role of Parliamentary Procedures Beliaev, Mikhail V. 2006. “Presidential Powers and Consolidation of New Postcommunist Democracies.” Comparative Political Studies 39(3): 375–398. Bernauer, Julian and Adrian Vatter. 2012. “Can’t Get No Satisfaction with the Westminster Model? Winners, Losers and the Effects of Consensual and Direct Democratic Institutions on Satisfaction with Democracy.” European Journal of Political Research 51(4): 435–468. Binder, Sarah A. 2006. “Parties and Institutional Choice Revisited.” Legislative Studies Quarterly 31(4): 513–532. Bouteca, Nicolas, Jef Smulders, Bart Maddens, Carl Devos and Bram Wauters. 2019. “A Fair Day’s Wage for a Fair Day’s Work´? Exploring the Connection between the Parliamentary Work of MPs and their Electoral Support.” The Journal of Legislative Studies 25(1): 44–65. Bowler, Shaun. 2010. “Private Members’ Bills in the UK Parliament: Is There an ‘Electoral Conncetion’? The Journal of Legislative Studies 16(4): 476–494. Bratton, Michael and Rober Mattes. 2001. “Support for Democracy in Africa: Intrinsic or Instrumental?” British Journal of Political Science 31(3): 447–474. Bräuninger, Thomas and Marc Debus. 2009. “Legislative Agenda-Setting in Parliamentary Democracies.” European Journal of Political Research 48(6): 804–839. Capano, Giliberto and Marco Giuliani. 2003. “The Italian Parliament: In Search of a New Role?” Journal of Legislative Studies 9(2): 8–34. Carey, John M. 2006. Legislative Organization. In The Oxford Handbook of Political Institutions, eds. Sarah A. Binder, R.A.W. Rhodes and Bert A. Rockman. Oxford and New York: Oxford University Press: 431–454. Carey, John M. 2007. “Competing Principals, Political Institutions, and Party Unity in Legislative Voting.” American Journal of Political Science 51(1): 92–107. Carey, John M. 2009. Legislative Voting and Accountability. Cambridge: Cambridge University Press. Carothers, Thomas. 2002. “The End of the Transition Paradigm.” Journal of Democracy 13(1): 5–21. Carroll, Royce, Gary W. Cox and Mónica Pachón. 2006. “How Parties Create Electoral Democracy, Chapter 2.” Legislative Studies Quarterly 31(2): 153–174. Chaisty, Paul, Nic Cheeseman and Timothy Power. 2014. “Rethinking the ‘Presidentialism Debate’: Conceptualizing Coalitional Politics in Cross-regional Perspective.” Democratization 21: 72–94. Cheibub, José Antonio. 2002. “Minority Governments, Deadlock Situations, and the Survival of Presidential Democracies.” Comparative Political Studies 35(3): 284–312. Cheibub, José Antonio. 2007. Presidentialism, Parliamentarism, and Democracy. Cambridge: Cambridge University Press. Cheibub, José Antonio, Adam Przeworski and Sebastian Saiegh. 2004. “Government Coalitions and Legislative Success under Presidentialism and Parliamentarism.” British Journal of Political Science 34: 565–587. Cheibub, José Antonio, Shane Martin and Bjorn Erik Rasch. 2021. “Investiture Rules and Formation of Minority Governments in European Parliamentary Democracies.” Party Politics 27(2): 351–362.

Institutions, Role of Opposition, and Consolidation  21 Cheibub, José Antonio, Zachary Elkins and Tom Ginsburg. 2014. “Beyond Presidentialism and Parliamentarism.” British Journal of Political Science 44(3): 515–544. Claassen, Christopher. 2019. “Does Public Support Help Democracy Survive?” American Journal of Political Science 64(1): 118–134. Colomer, Josep M. and Gabriel L. Negretto. 2005. “Can Presidentialism Work Like Parliamentarism?” Government and Opposition 40: 60–89. Cox, Gary W. 1987. The Efficient Secret: The Cabinet and the Development of Political Parties in Victorian England. Cambridge: Cambridge University Press. Cox, Gary W. 2000. “On the Effects of Legislative Rules.” Legislative Studies Quarterly 25(2): 169–192. Cox, Gary W. 2006. The Organization of Democratic Legislatures. In The Oxford Handbook of Political Economy, eds. Barry R. Weingast and Donald A. Wittman. Oxford: Oxford University Press: 141–161. Cox, Gary W. and Matthew D. McCubbins. 1993. Legislative Leviathan. Party Government in the House. Berkeley: University of California Press. Cox, Gary W. and Matthew D. McCubbins. 1994. “Bonding, Structure, and the Stability of Political Parties: Party Government in the House.” Legislative Studies Quarterly 19(2): 215–231. Cox, Gary W. and Matthew D. McCubbins. 2005. Setting the Agenda. Responsible Government in the U.S. House of Representatives. New York: Cambridge University Press. Crisp, Brian F. and Amanda Driscoll. 2012. “The Strategic Use of Legislative Voting Procedures.” Legislative Studies Quarterly 37(1): 67–97. CSES. 2019. “The Comparative Study of Electoral Systems. CSES Modules 1 to 4 Full Release. Dataset. December 15, 2015 version. DOI Module 1: 10.7804/ cses.module1.2015-12-15; DOI Module 2: 10.7804/cses.module2.2015-12-15; DOI Module 3: 10.7804/cses.module3.2015-12-15; DOI Module 4: 0.7804/cses. module4.2015-03-20. CSES Module 5 First Advance Release. Dataset. May 21, 2019 version. doi:10.7804/cses.module5.2019-05-21.” Online: www.cses.org. Däubler, Thomas, Thomas Bräuninger and Martin Brunner. 2016. “Is Personal Vote-Seeking Behavior Effective?” Legislative Studies Quarterly 41(2): 419–444. Diamond, Larry J. 1994. “Toward Democratic Conslidation.” Journal of Democracy 5(3): 4–17. Diamond, Larry J. 1999. Developing Democracy: Toward Consolidation. Baltimore, MD: Johns Hopkins University Press. Diermeier, Daniel, Carlo Prato and Razvan Vlaicu. 2015. “Procedural Choice in Majoritarian Organizations.” American Journal of Political Science 59(4): 866–879. Diermeier, Daniel and Roger B. Myerson. 1999. “Bicameralism and Its Consequences for the Internal Organization of Legislatures.” American Economic Review 89(5): 1182–1196. Döring, Herbert. 1995. Parliaments and Majority Rule in Western Europe. Frankfurt: Campus Verlag. Easton, David. 1957. “An Approach to the Analysis of Political Systems.” World Politics 9(3): 383–400.

22  Moderating Role of Parliamentary Procedures Eckstein, Harry. 1966. A Theory of Stable Democracy. Princeton, NJ: Princeton University Press. Eckstein, Harry. 1998. Congruence Theory Explained. In Can Democracy Take Root in Post-Soviet Russia? Explorations in State-Society Relations, eds. Erik P. Hoffmann, Harry Eckstein, Frederic J. Fleron and William M. Reisinger. Oxford: Rowman and Littlefield: 3–33. Eggers, Andrew C. and Arthur Spirling. 2014. “Ministerial Responsiveness in Westminster Institutional Choices and House of Commons Debate, 1831– 1915.” American Journal of Political Science 58(4): 873–887. Elgie, Robert. 2005. “From Linz to Tsebelis: Three Waves of Presidential/Parliamentary Studies?” Democratization 12(1): 106–122. ESS. 2019. “European Social Survey. Multilevel Data.” Online: http://www.europeansocialsurvey.org/. Accessed: October 2019. Evans, Geoffrey and Stephen Whitefield. 1995. “The Politics and Economics of Democratic Commitment: Support for Democracy in Transition Societies.” British Journal of Political Science 25(4): 485–514. Fernandes, Jorge M. 2016. “Intra-party Delegation in the Portuguese Legislature: Assigning Committee Chairs and Party Coordination Positions.” The Journal of Legislative Studies 22(1): 108–128. Fernandes, Jorge M., Max Goplerud and Miguel Won. 2019. “Legislative Bellwethers: The Role of Committee Membership in Parliamentary Debate.” Legislative Studies Quarterly 44: 307–343. Fish, Steven M. 2006. “Stronger Legislatures, Stronger Democracies.” Journal of Democracy 1: 5–19. Fish, Steven M. and Matthew Kroenig. 2009. The Handbook of National Legislatures: A Global Survey. Cambridge: Cambridge University Press. Flinders, Matthew. 2007. “Analysing Reform. The House of Commons, 2001– 05.” Political Studies 55(1): 174–200. Foa, Roberto Stefan and Yascha Mounk. 2016. “The Democratic Disconnect.” Journal of Democracy 27(3): 5–17. Foa, Roberto Stefan and Yascha Mounk. 2017. “The Signs of Deconsolidation.” Journal of Democracy 28(1): 5–15. Fortunato, David. 2021. “Legislative Scholars Should Study Extralegislative Outcomes.” Legislative Studies Quarterly 46(1): 1–9. Fortunato, David, Lanny W. Martin and Georg Vanberg. 2019. “Committee Chairs and Legislative Review in Parliamentary Democracies.” British Journal of Political Science 49(2): 785–797. Gamm, Gerald and John D. Huber. 2002. Legislatures as Political Institutions: Beyond the Contemporary Congress. In Political Science: State of the Discipline, eds. Ira Katznelson and Helen V. Milner. New York: Norton: 313–341. Garritzmann, Julian. 2017. “How Much Power Do Opposition Have? Comparing the Opportunity Structures of Parliamentary Opposition in 21 Democracies.” The Journal of Legislative Studies 23(1): 1–30. Gilligan, Thomas W. and Keith Krehbiel. 1990. “Organization of Informative Committees by a Rational Legislature.” American Journal of Political Science 34(2): 531–564.

Institutions, Role of Opposition, and Consolidation  23 Goet, Niels D. 2021. “The Politics of Procedural Choice: Regulating Legislative Debate in the UK House of Commons, 1811–2015.” British Journal of Political Science 51: 788–806. Goet, Niels D., Thomas G. Fleming and Radoslaw Zubek. 2020. “Procedural Change in the UK House of Commons, 1811–2015.” Legislative Studies Quarterly 45(1): 35–67. Haerpfer, Christian W., Patrick Bernhagen, Ronald F. Inglehart and Christian Welzel. 2009. Democratization. Oxford and New York: Oxford University Press. Haggard, Stephan and Mathew D. McCubbins. 2001. Introduction. Political Institutions and the Determinants of Public Policy. In Presidents, Parliaments, and Policy, eds. Stephan Haggard and Mathew D. McCubbins. Cambridge: Cambridge University Press: 1–20. Holden, Matthew. 2006. Exclusion, Inclusion, and Political Institutions. In The Oxford Handbook of Political Institutions, eds. Sarah A. Binder, R.A.W. Rhodes and Bert A. Rockman. Oxford and New York: Oxford University Press: 163–190. Horowitz, Donald L. 1990. “Comparing Democratic Systems.” Journal of Democracy 1(4): 73–79. Hug, Simon, Simone Wegmann and Reto Wüest. 2015. “Parliamentary Voting Procedures in Comparison.” West European Politics 38(5): 940–968. Inglehart, Ronald. 1997. Modernization and Postmodernization. Cultural, Political and Economic Change in 43 Societies. Princeton, NJ: Princeton University Press. Inglehart, Ronald and Christian Welzel. 2005. Modernization, Cultural Change, and Democracy: The Human Development Sequence. Cambridge: Cambridge University Press. Khmelko, Irina S., Charles R. Wise and Trevor L. Brown. 2010. “Committees and Legislative Strengthening: The Growing Influence of Committees in Ukraine’s Legislative Process.” The Journal of Legislative Studies 16(1): 73–95. Koß, Michael. 2015. “The Origins of Parliamentary Agenda Control: A Comparative Process Tracing Analysis.” West European Politics 38(5): 1062–1085. Koß, Michael. 2018. Parliaments in Time: The Evolution of Legislative Democracy in Western Europe, 1866–2015. Oxford: Oxford University Press. Krehbiel, Keith. 1992. Information and Legislative Organization. Ann Arbor: University of Michigan Press. Leston-Bandeira, Cristina and André Freire. 2003. “Internalising the Lessons of Stable Democracy: The Portugese Parliament.” Journal of Legislative Studies 9(2): 56–84. Lijphart, Arend. 1994. Presidentialism and Majoritarian Democracy: Theoretical Observations. In The Failure of Presidential Democracy. Comparative Perspectives Volume 1, eds. Juan J. Linz and Arturo Valenzuela. Baltimore, MD: The Johns Hopkins University Press: 91–105. Lijphart, Arend. 2008. Thinking about Democracy: Power Sharing and Majority Rule in Theory and Practice. New York: Taylor & Francis. Linz, Juan J. 1990a. “The Perils of Presidentialism.” Journal of Democracy 1(1): 51–69.

24  Moderating Role of Parliamentary Procedures Linz, Juan J. 1990b. “The Virtues of Parliamentarism.” Journal of Democracy 1(4): 84–91. Linz, Juan J. 1994. Presidential or Parliamentary Democracy: Does It Make a Difference? In The Failure of Presidential Democracy. Comparative Perspectives Volume 1, eds. Juan J. Linz and Arturo Valenzuela. Baltimore, MD: The Johns Hopkins University Press: 3–87. Linz, Juan J. and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation. Baltimore, MD: Johns Hopkins University Press. Linz, Juan J. and Arturo Valenzuela. 1994. The Failure of Presidential Democracy: Comparative Perspectives. Baltimore, MD: Johns Hopkins University Press. Lipset, Seymour Martin. 1959. “Some Social Requisites of Democracy Economic Development and Legitimacy.” American Political Science Review 53: 69–105. Loewenberg, Gerhard. 2003. “Agenda-setting in the German Bundestag: Origins and Consequences of Party Dominance.” The Journal of Legislative Studies 9(3): 17–31. Lührmann, Anna, Sandra Grahn, Richard Morgan, Shreeya Pillai and Staffan I. Lindberg. 2019. “State of the World 2018: Democracy Facing Global Challenges.” Democratization 26(6): 895–915. Mainwaring, Scott. 1993. “Presidentialism, Multipartism, and Democracy: The Difficult Combination.” Comparative Political Studies 26(2): 198–228. Mainwaring, Scott and Matthew SøbergShugart. 1997a. Conclusion: Presidentialism and the Party System. In Presidentialism and Democracy in Latin America, eds. Scott Mainwaring and Matthew Søberg Shugart. Cambridge: Cambridge University Press: 394–439. Mainwaring, Scott and Matthew Søberg Shugart. 1997b. “Juan Linz, Presidentialism, and Democracy: A Critical Approach.” Comparative Politics 29(4): 449–471. Mainwaring, Scott and Matthew Søberg Shugart. 1997c. Presidentialism and Democracy in Latin America. Cambridge: Cambridge University Press. March, James G. and Johan P. Olsen. 1984. “The New Institutionalism: Organizational Factors in Political Life.” The American Political Science Review 78(3): 734–749. March, James G. and Johan P. Olsen. 2006. Elaborating the “New Institutionalism”. In The Oxford Handbook of Political Institutions, eds. Sarah A. Binder, R.A.W. Rhodes and Bert A. Rockman. Oxford and New York: Oxford University Press: 3–20. Martin Lanny W. and Georg Vanberg. 2011. Parliaments and Coalitions: The Role of Legislative Institutions in Multiparty Governance. Oxford: Oxford University Press. Martin, Shane. 2011a. “Electoral Institutions, the Personal Vote, and Legislative Organization.” Legislative Studies Quarterly 36(3): 339–361. Martin, Shane. 2011b. “Parliamentary Questions, the Behaviour of Legislators, and the Function of Legislatures: An Introduction.” The Journal of Legislative Studies 17(3): 259–270. Martin, Shane. 2014. Committees. In The Oxford Handbook of Legislative Studies, eds. Shane Martin, Thomas Saalfeld and Kaare W. Strøm. Oxford and New York: Oxford University Press: 352–368.

Institutions, Role of Opposition, and Consolidation  25 McGrath, Robert J., and Josh M. Ryan. 2019. “Party Effects in State Legislative Committees.” Legislative Studies Quarterly 44: 553–592. Müller, Wolfgang C. and Ulrich Sieberer. 2014. Procedures and Rules in Legislatures. In The Oxford Handbook of Legislative Studies, eds. Shane Martin, Thomas Saalfeld and Kaare W. Strøm. Oxford and New York: Oxford University Press: 312–331. Norton, Philip. 2003. “Playing by the Rules. The Constraining Hand of Parliamentary Procedure.” Journal of Legislative Studies 7(3): 13–33. O’Donnell, Guillermo. 1992. Transitions, Continuities, and Paradoxes. In Issues in Democratic Consolidation: The New South American Democracies in Comparative Perspective, eds. Scott Mainwaring, Guillermo O’Donnell and J. Samuel Valenzuela. Notre Dame, IN: University of Notre Dame Press: 17–56. O’Donnell, Guillermo and Philippe Schmitter. 1986. Transitions from Authoritarian Rule: Tentative Conclusions about Uncertain Democracies. Baltimore, MD and London: The Johns Hopkins University Press. Power, Timothy J. and Mark J. Gasiorowski. 1997. “Institutional Design and Democratic Consolidation in the Third World.” Comparative Political Studies 30(2): 123–155. Pridham, Geoffrey and Tatu Vanhanen. 1994. Democratization in Eastern Europe: Domestic and International Perspectives. London: Routledge. Przeworski, Adam. 1991. Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America. Cambridge: Cambridge University Press. Przeworski, Adam. 2019. Crises of Democracy. Cambridge: Cambridge University Press. Przeworski, Adam, Michael E. Alvarez, José Antonio Cheibub and Fernando Limongi. 2000. Democracy and Development. Political Institutions and Well-­ Being in the World, 1950–1990. Cambridge: Cambridge University Press. Przeworski, Adam, Michael M. Alvarez, Jose Antonio Cheibub and Ferdinand Limongi. 1996. “What Makes Democracies Endure?” Journal of Democracy 7(1): 39–55. Raile, Eric D., Carlos Pereira and Timothy J. Power. 2011. “The Executive Toolbox: Building Legislative Support in a Multiparty Presidential Regime.” Political Research Quarterly 64(2): 323–334. Rasch, Bjørn Erik, Shane Martin and José Antonio Cheibub. 2015. Parliaments and Government Formation: Unpacking Investiture Rules. Oxford: Oxford University Press. Regan, Patrick and Dave Clark. 2014. “The Institutions and Elections Project.” Online: https://havardhegre.net/iaep/. Accessed: January 2014. Rozenberg, Olivier and Shane Martin. 2011. “Questioning Parliamentary Questions.” The Journal of Legislative Studies 17(3): 394–404. Rueschemeyer, Dietrich, Evelyn Huber Stephens and John D. Stephens. 1992. Capitalist Development and Democracy. Chicago, IL: The University of Chicago Press. Schedler, Andreas. 1998. “What Is Democratic Consolidation?” Journal of Democracy 9(2): 91–107.

26  Moderating Role of Parliamentary Procedures Schickler, Eric. 2000. “Institutional Change in the House of Representatives, 1867–1998: A Test of Partisan and Ideological Power Balance Models.” The American Political Science Review 94(2): 269–288. Schmitter, Philippe C. and Terry Lynn Karl. 1991. “What Democracy Is…and Is Not.” Journal of Democracy 2(3): 75–88. Schneider, Carsten Q. and Philippe C. Schmitter. 2004. “Liberalization, Transition and Consolidation: Measuring the Components of Democratization.” Democratization 11(5): 59–90. Shepsle, Kenneth A. 1979. “Institutional Arrangements and Equilibrium in Multidimensional Voting Models.” American Journal of Political Science 23: 23–57. Shepsle, Kenneth A. and Barry R. Weingast. 1981. “Structure-induced Equilibrium and Legislative Choice.” Public Choice 37: 503–519. Shepsle, Kenneth A. and Barry R. Weingast. 1987. “The Institutional Foundation of Committee Power.” American Political Science Review 81: 85–104. Shomer, Yael. 2015”. Electoral Incentives and Individual Parliament Memebers Rights.” West Eurpean Politics 38(5): 1106–1127. Shugart, Matthew Søberg. 2006. Comparative Executive-Legislative Relations. In The Oxford Handbook of Political Institutions, eds. R.A.W. Rhodes, Sarah A. Binder and Bert A. Rockman. Oxford and New York: Oxford University Press: 344–365. Shugart, Matthew Søberg and John M. Carey. 1992. Presidents and Assemblies Constitutional Design and Electoral Dynamics. New York: Cambridge University Press. Shugart, Matthew Søberg and Scott Mainwaring. 1997. Presidentialism and Democracy in Latin America: Rethinking the Terms of the Debate. In Presidentialism and Democracy in Latin America, eds. Scott Mainwaring and Matthew Søberg Shugart. New York: Cambridge University Press: 12–54. Sieberer, Ulrich and Daniel Höhmann. 2021. “Do Party System Parameters Explain Difference in Legislative Organization? Fragmentation, Polarization, and the Density of Regulation in European Parliaments, 1045–2009.” Party Politics: 1–14. Sieberer, Ulrich, Peter Meißner, Julia F. Keh and Wolfgang C. Müller. 2016. “Mapping and Explaining Parliamentary Rule Changes in Europe: A Research Program.” Legislative Studies Quarterly 41(1): 61–88. Sieberer, Ulrich and Wolfgang C. Müller. 2015. “Explaining Reforms of Parliamentary Minority Rights: A Theoretical Framework with Case Study Application.” West European Politics 38(5): 997–1019. Sieberer, Ulrich, Wolfgang C. Müller and Maiko Isabelle Heller. 2011. “Reforming the Rules of the Parliamentary Game: Measuring and Explaining Changes in Parliamentary Rules in Austria, Germany, and Switzerland, 1945–2010.” West European Politics 34(5): 948–975. Spicer, Zachary and John L. Nater. 2013. “Legislative Dissent without Reprisal? An Alternative View of Speaker Selection.” The Journal of Legislative Studies 19(4): 505– 525. Stepan, Alfred and Cindy Skach. 1993. “Constitutional Frameworks and Democratic Consolidation. Parliamentarism versus Presidentialism.” World Politics 46(1): 1–22.

Institutions, Role of Opposition, and Consolidation  27 Stepan, Alfred and Cindy Skach. 1994. Presidentialism and Parliamentarism in Comparative Perspective. In The Failure of Presidential Democracy. Comparative Perspectives Volume 1, eds. Juan J. Linz and Arturo Valenzuela. Baltimore, MD: The Johns Hopkins University Press: 119–136. Strøm, Kaare. 1990. Minority Government and Majority Rule. New York: Cambridge University Press. Taylor, Andrew J. 2006. “Size, Power, and Electoral Systems: Exogenous Determinants of Legislative Procedural Choice.” Legislative Studies Quarterly 31(3): 323–345. Tsebelis, George. 1995. “Decision Making in Political Systems: Veto Players in Presidentialism, Parliamentarism, Multicameralism and Multipartyism.” British Journal of Political Science 25(3): 289–325. Tsebelis, George. 1999. “Veto Players and Law Production in Parliamentary Democracies: An Empirical Analysis.” American Political Science Review 93(3): 591–608. Tsebelis, George. 2002. Veto Players: How Political Institutions Work. Princeton, NJ: Princeton University Press. Van Cranenburgh, Oda. 2008. “’Big Men’ Rule: Presidential Power, Regime Type, and Democracy in 30 African Countries.” Democratization 15(5): 952–973. Vanderbilt University. 2019. “Latin American Public Opinion Project (LAPOP).” Online: http://www.vanderbilt.edu/lapop/. Accessed: October 2019. Voeten, Erik. 2016. “Voeten, Erik, Are People Really Turning Away from Democracy?” Online: https://ssrn.com/abstract=2882878. Accessed: March 2017. Weaver, R. Kent and Bert A. Rockman. 1993. Do Institutions Matter? Government Capabilities in the United States and Abroad. Washington, DC: Brookings. Wegmann, Simone. 2022. “Policy-Making Power of Opposition Players: A Comparative Institutional Perspective.” The Journal of Legislative Studies: 28(1): 1–25. Welzel, Christian. 2006. “Democratization as an Emancipative Process: The Neglected Role of Mass Motivations.” European Journal of Political Research 45(6): 871–896. Welzel, Christian. 2007. “Are Levels of Democracy Sffected by Mass Sttitudes? Testing Attainment and Sustainment Effects on Democracy.” International Political Science Review 28(4): 397–424. Welzel, Christian and Ronald F. Inglehart. 2009. Political Culture, Mass Beliefs, and Value Change. In Democratization, eds. Ronald F. Inglehart, Christian W. Haerpfer, Patrick Bernhagen and Christian Welzel. Oxford and New York: Oxford University Press: 126–144. Williams, Brian D. and Indridi H. Indridason. 2018. “Luck of the Draw? Private Members’ Bills and the Electoral Connection.” Political Science Research and Methods 6(2): 211–227. Zubek, Radoslaw. 2015. “Coalition Government and Committee Power.” West European Politics 38(5): 1020–1041.

Chapter 2

The Role of Parliamentary Procedures

In this chapter, I present a theoretical argument of the relationship between parliamentary procedures and democratic consolidation. Most contributions in the field of democratic consolidation have neglected the central role of the legislature. Even though a substantial research tradition exists on legislative organization, the literature on democratic consolidation has largely neglected the work in legislative studies that highlights the consequences of differences in legislative organizations. The argument developed in this chapter follows the logic of different costs of losing democratic elections for different groups in society (see, e.g. Przeworski 2019: 154). I argue that one element in determining the impact of elections on democratic consolidation is the degree of political power granted to losers. In particular, parliamentary procedures grant different degrees of power within the legislature to opposition players. Therefore, parliamentary procedures are among the crucial factors in determining how much is at stake in democratic elections. The degree of power opposition players can exert in the legislature can reduce the inequality between winners and losers, leading to more political support for the democratic regime. Research on both legislative studies and democratization presents promising approaches to answer the question of the extent to which parliamentary procedures might influence democratic consolidation. For example, rational choice arguments linked to the process of government formation have pointed to different incentives for political actors when deciding whether or not to participate in the executive (see, e.g. Strøm 1990a, 1990b). Moreover, economic approaches in the field of democratization have identified different incentives for and benefits from democratization for different groups, chiefly the poorer majority and a richer minority (see, for example, Acemoglu and Robinson, 2012, 2006 or Boix, 2006, 2003). I build on this research to illustrate my reasoning about the relationship between opposition power and political support of winners and losers. DOI: 10.4324/9781003005360-3

Role of Parliamentary Procedures  29

This chapter proceeds as follows. In the first section, I discuss why moving beyond the characteristic of government formation is important and how considering legislative organization in more detail can contribute to the understanding of the determinants of democratic consolidation. In the second section, I discuss why it is not only important to focus on public opinion but also to differentiate between winners and losers when explaining democratic consolidation. In the third section, I present the theoretical reasoning of why parliamentary procedures may influence democratic consolidation. I focus on the role of winners and losers as well as parliamentary procedures in determining the potential costs of losing democratic elections. To illustrate my argument, I link it to the literature on rational choice approaches to democratization and government formation. The roles of the majority and the minority in the process of democratization are used to understand preferences for democracy or autocracy, respectively. These roles, in the process of government formation, are also applied in order to understand the roles and the incentives of opposition players in democracies. In the fourth section, I present a summary of the theoretical argument. Before presenting the theoretical argument in more detail, the following paragraphs briefly recap definitions of the key concepts and how they are used in this book. Winners and Losers The argument presented in the following sections relies on the different behaviors of winners and losers of democratic elections. The definitions of the two groups conform to the literature on political support. Winners and losers exist on both the elite and citizen levels. Winners on the citizen level are individuals that voted for the party (or parties) that constitute the government after an election. This comprises the party of the president in presidential regimes and that of the prime minister in parliamentary regimes, respectively. Furthermore, parties of a government coalition are also considered to be part of the winners group. Winners on the elite level constitute all members of the government as well as MPs of government parties. Consequently, losers on the citizen level constitute the group of voters that voted for a party (or parties) not exercising executive power after an election. Individuals that voted for a party that is part of the opposition in the legislative chamber as well as individuals that voted for a party not represented in the legislature are both considered losers. On the elite level, MPs of opposition parties constitute the group of losers. Hence, the determining characteristic of winners and losers is not necessarily the feature of being in the minority or majority, according to legislative seats, but the exercise of executive office. Political Support The theoretical argument presented in the following sections focuses on the individual component of democratic

30  Moderating Role of Parliamentary Procedures

consolidation. A vast literature exists on different dimensions of individual political support. Among the most prominent distinctions is the one between specific and diffuse support (see Easton 1975, 1965a, 1965b, 1957). Whereas specific support refers to the satisfaction with a particular outcome of a regime, diffuse support relates to the agreement with the basic principles upon which such a regime is based. Public support therefore allows measuring democratic consolidation at the level of individual citizens. In the present case, diffuse support relates to the basic agreement with a democratic form of government whereas specific support relates to the satisfaction with how democracy works.1 Democracy and Democratic Consolidation The theoretical argument developed in this chapter refers to democracy in its minimal definition. The theoretical model of democracy is largely based on the holding of elections, as democratic elections constitute the starting point as well as the consequence of loser acceptance of previous electoral outcomes (see Figures 2.1 and 2.2). Therefore, it seems appropriate to define democracy in a minimalist sense. Accordingly, “the democratic method is that institutional arrangement for arriving at political decisions in which individuals acquire the power to decide by means of a competitive struggle for the people’s vote” (Schumpeter 1942: 169). Democratic consolidation is understood as the solidifying of democratic rule as “the only game in town” (Linz and Stepan 1996: 15–16). According to Diamond (1999: 65), consolidation at the mass level refers to “(…) broad normative and behavioral consensus (…) on the legitimacy of the constitutional system, however poor or unsatisfying its performance may be at any point in time.” This definition of consolidation refers to the diffuse system support specified earlier. However, specific system support can also influence willingness to defend democracy as the best form of government and, thereby, threaten democratic consolidation. Parliamentary Procedures Parliamentary procedures are defined as the rules established to overcome the “legislative state of nature” (Cox 2006: 143). What results from a “legislative state of nature” where every member has equal probability of recognition is an impassable plenary bottleneck. For a legislative chamber to be able to fulfil its function of policy-making, regulating plenary sessions is of crucial importance. Such regulations are defined by particular procedures. In turn, this results in “(…) the universal creation of offices endowed with special agenda-setting powers. Thus, while legislators are everywhere equal in voting power, they are everywhere unequal in agenda-setting power” (Cox 2006: 142). Therefore, in general terms, parliamentary procedures are the rules regulating the functioning of the legislative chamber(s). The theoretical reasoning presented in the following sections focuses particularly on parliamentary procedures regulating the legislative decision-making process.2

Role of Parliamentary Procedures  31

2.1  T he Importance of Legislatures – Moving beyond Governmental Organization Much of the research on institutional determinants of democratic consolidation has focused on broad categorizations such as the distinction between parliamentary and presidential systems or majoritarian and consensus democracies. Parliamentarism and presidentialism differ in their relationship of the executive and the legislative. Whereas a hierarchical relationship characterizes parliamentary systems, a more transactional relationship can be found in presidential systems (Shugart 2006: 344). However, a somewhat closer look at the relationships between executives and legislatures shows that most countries do not fully meet the criteria for the ideal type of either system (Shugart 2006). Similarly, majoritarian and consensus democracies differ in the question of who governs. In majoritarian democracies it is “the majority of the people,” whereas, in consensus democracies it is “as many people as possible” (Lijphart 1999: 2). Again, however, most countries cannot be categorized as clearly majoritarian or consensus oriented but show a combination of both types. Therefore, analyzing the influence of these constitutionally defined systems on democratic consolidation is unlikely to be able to determine which characteristics, in particular, are driving differences in public support for democracy. Hence, a different approach, moving beyond governmental organization, such as those proposed by Tsebelis (2002, 1999; 1995) or Fish and Kroenig (2009), might be more suitable to analyze the institutional determinants of democratic consolidation. Looking, for example, at Fish and Kroenig’s (2009) Parliamentary Power Index shows that the degree of parliamentary power is not necessarily higher in parliamentary systems than in presidential ones. Consequently, the distinction between parliamentarism and presidentialism does not automatically indicate different levels of parliamentary power. However, it is important to assess the power of particular institutions (Fish 2006: 6–7) and to not classify countries according to their constitutional systems in order to analyze the influence of different governmental arrangements. Even when considering the distinction between the constitutional systems of presidentialism and parliamentarism, each of these systems is characterized by a wide variation in the capacity of legislatures (Fish and Kroenig 2009: 2). Therefore, the distinction between parliamentary and presidential systems might not account for the most decisive mechanisms – which is “(…) where power really resides” (Fish 2006: 6) – when it comes to explaining democratic consolidation. This research suggests that there are at least two reasons not to classify political systems in terms of the manner of government formation. First, not all countries clearly match the ideal type of a parliamentary system

32  Moderating Role of Parliamentary Procedures

with a legislative agenda controlled by the government or a presidential system with a separation of power between the executive and the legislature, respectively. There are large differences, for example, in the degree of legislative powers granted to the president in presidential systems (Shugart and Carey 1992). Similarly, the difference in the frequency of coalition governments between parliamentary and presidential systems is not as large as one would expect if presidentialism did not provide any incentives for such coalition formation (Cheibub 2007). Second, such a classification does not consider the central role of electoral losers for the consolidation of democracies as they focus on the question of government formation. Institutions can favor democratic consolidation if designed in such a way as to encourage significant groups to express political support and discourage them from becoming disaffected (­Diamond 1999: 104). Categorizing countries based on government formation (i.e. a distinction between parliamentary and presidential systems or majoritarian and consensus democracies) does not necessarily account for such different institutional designs. Electoral losers exist in all of these systems as they exist in all democracies. Focusing on government formation, however, denies the crucial role of these losers for the consolidation of a democracy. In sum, rather than focusing on government formation, it seems crucial to focus on the different legislative rights and resources of legislative actors, bringing legislatures and especially electoral losers into focus. Neglecting the organization of the decision-making process is tantamount to neglecting one of the most important characteristics of what democracy is all about. Legislative procedures influence the structure of decision-making as well as the weight of individual legislators in policy decisions (Cheibub 2002: 168) and are therefore a vital factor in explaining political support (i.e. democratic consolidation). Hence, it is important to take a closer look at legislatures and at how the power inside legislative chambers is distributed among different actors, both winners and losers alike.

2.2  The Importance of Winners and Losers Even though the specific institutional design of democratic regimes varies considerably, they share one characteristic: all democracies create inequalities among voters through elections. One crucial characteristic of democracies is therefore the fact that although (ideally) all citizens have equal access to electoral participation, electoral outcomes produce inequalities among voters. Democracy is not only about winning elections, but equally about losing them (see, e.g. Przeworski 1991: 10). Although emphasized in earlier work on oppositions (e.g., Ionescu and de Madariaga 1968; Dahl 1966; Lipset 1963) and part of the definitions of

Role of Parliamentary Procedures  33

democracy (e.g. Przeworski 1991),3 this defining feature of democracy is often neglected in research on democratic consolidation. The research on support for democracy which distinguishes between winners and losers of democratic elections has advanced various factors explaining satisfaction with democracy, especially of losers. Accordingly, winners and losers gain different utilities from elections. Among the first authors are Gabriel (1989) and Nadeau and Blais (1993) who have emphasized that winners of elections are more likely to be satisfied with democracy than losers of elections are (see also, e.g. Singh, Karakoç and Blais 2012; Singh, Lago and Blais 2011; Henderson 2008; Bowler and Donovan 2007; Craig et al. 2006; Anderson et al. 2005). However, winners not only show more satisfaction with democracy but also more positive attitudes toward the government (Anderson and LoTempio 2002; Anderson and Tverdova 2001). Although most research has focused on individual support after elections, Blais, Morin-Chassé and Singh (2017) have confirmed different levels of system support by winners and losers testing for the relationship by looking at the level of satisfaction before and after elections. Results also show that the winner-loser gap is a “relatively stable phenomenon” (Dahlberg and Linde 2015: 19). Explanatory factors for democratic support have, again, largely focused on the role of the government and, consequently, on the winning parties (see, e.g. Anderson et al. 2005 or Bernauer and Vatter 2012). However, losers of elections hold an exceptionally important role within a democratic regime as without their consent to adhere to democratic rules stable democracy is impossible (Riker 1983; Lipset 1959). Accordingly, losers are at least as decisive for democratic consolidation and stability as winners are. Surprisingly, however, the vast majority of the literature analyzed political institutions from the winner’s point of view (Helms 2004: 23). At the heart of the reasoning presented in the next section lies the fact that the analysis of the institutional determinants of democratic consolidation is barely possible when the role of the losers is ignored or neglected (both on the elite and the citizen level).

2.3  T  he Influence of Parliamentary Procedures on Democratic Consolidation As research on political support for democracy has shown, a gap typically exists between the support expressed by winners and losers of democratic elections. Closing this gap or achieving a higher level of political support of losers – and thereby increasing the level of consolidation – should be of benefit to policy-makers and democratic governments alike. Figure 2.1 shows this basic setting, as the extant literature depicts it. This setting serves as a point of departure for the theoretical reasoning presented hereafter.

34  Moderating Role of Parliamentary Procedures

Figure 2.1  Democratic Consolidation: The Winner-Loser Gap

The starting point of the reasoning is the holding of democratic elections at time t1. What follows from democratic elections is two groups: winners and losers. Consequently, not all citizens are equally satisfied with the election outcome and, not all citizens express the same degree of political support, thus, leading to a winner-loser gap in political support where winners express more support than losers (see, e.g. Anderson et al. 2005). But, the consolidation of the democratic regime is not only dependent on the support of winners but is also highly dependent on the consent of losers to accept the outcome of the election at time t1, to wait until the next election at time t2 to try to access executive power, and to not overthrow the current democratically elected government in order to access power. I argue that inclusive parliamentary procedures that grant power to losers can positively influence the level of political support that they express after elections. Figure 2.2 shows a modification of Figure 2.1 and presents the central reasoning that lies at the heart of this book. Again, the starting point is an election at time t1 that creates two groups: winners and losers. As outlined before and as documented in the literature, in general, winners express more political support than losers, resulting in a winner-loser gap in support for democracy. So far, this reflects the basic setting generated by every democratic election (i.e. the situation as Figure 2.1 depicts it). The central question is: under which conditions do citizens express more political support, deepening democratic consolidation? I argue that parliamentary procedures determine the roles of different actors in the policy-making process, winners and losers alike. As the legislative studies literature shows, the organization of legislative chambers varies considerably. Different parliamentary procedures grant different degrees of power to the representatives of winners and losers, i.e. government and opposition players in the policy-making process. Hence, the main intervening variable in the reasoning presented in Figure 2.2 is parliamentary procedures that grant political power to opposition players and thereby influence the degree of political equality.4

Role of Parliamentary Procedures  35

Figure 2.2  The Moderating Role of Parliamentary Procedures

As presented in Chapter 1, MPs accept parliamentary procedures constraining their power in legislative chambers to overcome the ´legislative state of nature´ that would result in an impassable plenary bottleneck (Cox 2006). Hence, legislative rules guarantee that legislative chambers can work efficiently (Diermeier, Prato and Vlaicu 2015). However, these rules are also determined by the majority of MPs themselves. Consequently, legislative rules are likely to mirror the needs of these MPs and the larger institutional setting in which the legislature is embedded (McCubbins 2005). As I will discuss in more detail in Chapter 5, research has shown that strong opposition rights are more likely in more heterogeneous legislatures (Wegmann 2022). Hence, it seems that MPs indeed try to accommodate specific needs resulting from the institutional environment in which the legislature is embedded. This result fits into the vast research on the origin and determining characteristics of legislative rules (see Chapter 1). However, the central goal here is not to explain the differences in the degree of power granted to oppositions but to focus on the effect of such differences. As research has shown, legislative procedures undoubtedly and decisively influence the power distribution between winners and losers in legislatures. The central question is whether these differences have wider extra-legislative effects and also influence democratic consolidation. I discuss in the following paragraphs why this particular distribution of power in the legislature is likely to influence democratic consolidation. Theoretical arguments based on power distribution have been among the most well-known explanations of democratization. Accordingly, different groups have different political interests (e.g. different preferences for the level of redistribution), and they will only accept democracy if the system does not pose a threat to their interests (Acemoglu and Robinson

36  Moderating Role of Parliamentary Procedures

2006: 34–35). In these models, income inequality matters for both elites and non-elites alike as they perceive that democracy causes progressive redistribution (see, e.g. Acemoglu and Robinson 2006; Benhabib and Przeworski 2006; Boix 2003). Boix argues that as “each political regime has different redistributive consequences, every individual supports the political arrangement that maximizes his welfare” (Boix 2006: 10). Therefore, since the majority sets policies in a democratic setting, the poorer majority will vote for the redistribution of income. It follows that poor masses prefer democracy (where redistribution is expected) and rich elites prefer autocracy (where no such redistribution is expected). Consequently, inequality impacts the prospects for democratization by affecting the relative gains and costs of democracy for these two groups, and thus their willingness to fight for or against democracy. As these contributions show, democratization is inherently about the transformation of de jure political power (see, e.g. Acemoglu and Robinson 2006). I argue that within a democratic setting, the allocation of political power to different groups is what determines the degree of influence these groups have. Consequently, the allocation of political power also determines the benefit these groups can obtain from a particular democratic system, and especially from winning elections. In many models of democratization, political conflict is about redistribution; whether the solution of this conflict reflects a group’s preferences depends on their level of de jure political power. As Acemoglu and Robinson (2006: 118) emphasize for non-democratic countries, “(…) elites typically live in fear of being replaced by different social groups or by other individuals within the same group.” Therefore, the authors argue that for a non-democratic regime to survive, it must ensure that no group is dissatisfied to the extent that it might overthrow the regime. Although the authors limit their reasoning to an authoritarian setting, their argument can be adapted to the relationship between different groups in a democracy (i.e. winners and losers). The allocation of political power to different groups determines the degree of influence these groups have and the degree of benefit they can obtain from a particular political system. Thus, the existence of highly dissatisfied groups is dangerous to non-democratic and democratic regimes alike. Being dependent on the consent of losers, a democracy must assure that these losers are not dissatisfied to such an extent that they would prefer to overthrow the democratically elected government or support non-democratic alternatives. However, government formation processes are not the only element that defines the level of political power of different groups. The degree of political power granted to losers is likely to play a crucial role in determining the stakes of losing an election. However, as Acemoglu and Robinson (2006: 35) put it,

Role of Parliamentary Procedures  37

[i]f the elites have too much power in democracy, democracy will do little to improve the welfare of the majority. In this case, democracy is not a solution to social conflict, and the result will either be revolution or an elite that keeps itself in power through repression. Thus, winners and losers in a democracy (both on the elite and the citizen level) can be compared to the richer minority and the poorer majority in Acemoglu and Robinson’s (2006) model of democratization, respectively. In this model, the granting of more de jure political power to the poorer majority, in the democratic setting, helps to satisfy their preferences for a higher level of redistribution. Similarly, more de jure political power for opposition players (who represent losers)5 in a democracy allows them to influence the legislative decision-making notwithstanding their preferences lacking direct representation in government, which, on the citizen level, might lead losers to express more political support. In the opposite case, when opposition actors have only weak political powers, losing elections is costlier as they are denied access to the policy-making process. In such a situation, democratic consolidation seems extremely difficult as losers are more likely to express low levels of both diffuse and specific support. Hence, and most importantly for the explanation of democratic consolidation: successful democracy depends on a stable balance of power between different groups in society. It is parliamentary procedures that define which groups are included in the legislative policy-making process and to what degree. The inclusive nature of parliamentary procedures grants some parliamentary power not only to winners (i.e. the government) but also to losers (i.e. the opposition players).6 Although Acemoglu and Robinson (2006: 17) emphasize that “democracies generally approximate a situation of political equality relative to non-democracies (…),” the fact that different degrees of such political equality exist in democracies must not be neglected. The degree to which de jure political power approximates political equality is determined by opposition rights or, rather, by parliamentary procedures. The more political inequality that exists between different groups of society, generated by parliamentary procedures, the more that is at stake in elections (i.e. the prospect of not being part of the government). Stated differently, giving fewer rights to the opposition to influence the policy-­ making process in the legislature means that opposition groups have more to lose from elections. The probability of winning within the democratic institutions is what determines the degree of consent of different groups (Przeworski 1991: 30). This emphasizes the crucial aspect of political inequality, as the probability of winning within democratic institutions, or of influencing the decision-making process, determines what losing elections actually means. It defines whether being a loser means living in a relationship in the “(…) form of [a] dictatorship by group A over group

38  Moderating Role of Parliamentary Procedures

B” (Holden 2006: 172) or whether the costs of losing an election are bearable. What might follow is that opposition players who are given “more power than their numbers alone merit” may favor democratic consolidation (Acemoglu and Robinson 2006: 248). However, the link between specific parliamentary procedures and the organization of the decision-making process and how aware citizens are of these rules is not straightforward. How can citizens, both winners and losers, know exactly how much they and their representatives can influence the legislative arena? Some institutional settings speak to citizens in a very direct way, for example, allowing for the possibility of citizen initiatives in direct democratic institutions. Others might be less known among citizens but still grant opposition MPs access to the policy-making process. I do not argue that voters will observe parliamentary procedures and legislative behavior of MPs or political parties directly. The assumption here is that the legislative behavior of MPs and parties is mediated by mass media and that political actors themselves have an incentive to communicate their legislative activities to their potential voters. Research increasingly confirms this assumption. Based on survey data among legislatures in Latin America,7 Carey (2009) shows that legislators report paying more attention to voters in their district than to any other factor when making political decisions (Carey 2009: 38). Analyzing parliamentary questions, for example, Saalfeld (2011) also shows that MPs respond to electoral incentives. Moreover, several scholars have highlighted the crucial role of political parties as actors linking citizens and the government (see, e.g. Randall 2012; Sartori 2005[1976]; Schattschneider 1942). Furthermore, recent research shows that political elites actively communicate their legislative activities to the public (Tuttnauer and Wegmann 2022; Huber et al. 2022). Research also suggests that voters are more adept than previously thought at finding the political information most useful to them (Dalton 2021; Fortunato and Stevenson 2021). ­Several contributions reflect this connection by showing that parliamentary work has an impact on votes (Bouteca et al. 2019; Marcinkiewicz and Stegmaier 2019; Däubler, Bräuninger and Brunner 2016; Bowler 2010). Tuttnauer and Wegmann (2022), for example, show that voters rate opposition parties more favorably and as more distant from the government party if they show a higher conflict rate with the government in their legislative voting behavior. Similarly, McKay (2020) shows that voters perceive MPs as more constituency-focused and express higher levels of trust if MPs focus more on their constituency in their parliamentary activities. This research suggests that the activity of political actors does reach voters. Consequently, as parliamentary procedures determine the possibilities for action by representatives, they should also influence how individuals evaluate democracy.

Role of Parliamentary Procedures  39

In this relation, research in legislative studies has repeatedly shown that parliamentary procedures influence the behavior of political actors. Among the most prominent contributions is Strøm’s reasoning about minority government formation. Accordingly, minority government formation is a result of a party leader’s rational behavior coming to the conclusion that participating in government does not imply more benefits than joining the opposition (Strøm 1990b: 23). If the principal objective of political parties is not necessarily office-seeking but policy-­ making, the exercise of executive power is not necessary to achieve their goals. Typically, political parties participating in the government can exert more influence than opposition parties. However, the degree of policy-making influence opposition parties can exert depends on the political system and varies considerably among different polities (Strøm 1990b: 38; 1984: 211). According to Strøm (1984, 1990b), payoffs from government participation are lower when the parliamentary opposition has more opportunity to influence policy-making.8 Under these circumstances, losing elections is less costly for both citizens and elites as they still have access to the policy-making process. Therefore, again, different levels of political power for the opposition define the degree to which a political system is characterized as “(…) a form of dictatorship by group A [the winners, i.e. the government] over group B [the losers, i.e. the opposition]” (Holden 2006: 172). If the difference in payoffs for government and opposition parties is relatively small due to the influence opportunities granted to opposition parties by the political system, losers on the citizen level9 should express more political support for the overall democratic regime and the outcome produced by democracy (i.e. diffuse and specific support). Consequently, losers’ level of political support is again dependent on the degree of political power they can exert compared to the power of the winners.10 Therefore, I argue that the specific allocation of power to winners and losers in the legislative chamber (determined by parliamentary procedures) influences the level of support these groups express. However, I also argue that the political power of opposition players influences the groups of winners and losers differently. If opposition players have more political power, winners and losers (on both the citizen and the elite level) are politically more equal. Under these circumstances, losing an election does not necessarily imply high costs for losers on the elite level (i.e. MPs of opposition parties), as they may still be able to influence the policy-­ making process (see Strøm’s, 1990b, reasoning about incentives to not participate in the executive). The same can be true for losers on the citizen level. Even though not represented by the executive, granting them and their representatives more power in the legislative arena11 might alleviate feelings of grievance or frustration about both the outcomes of the

40  Moderating Role of Parliamentary Procedures

democratic regime (i.e. specific support) as well as the principles upon which it is based (i.e. diffuse support). It follows that more political equality – generated by inclusive parliamentary procedures – for losers should lead such groups (on a citizen level) to express a higher level of political support (both diffuse and specific). From a procedure-oriented perspective as well as an outcome-­ oriented perspective, losers at the citizen-level benefit when opposition players are granted more power. At the same time, more political equality due to opposition players having more power would not necessarily lead to more political support expressed by winners on the citizen level. Outcome-oriented winners will be likely to express less specific support if opposition players have more power to influence the decision-­making process. However, granting more power to opposition players does increase representation in a democratic system. Thus, granting more equality in the procedures and access to policy-making can be seen as moving closer to an ideal type of democracy in which every citizen is “entitled to participate” (Dahl 2006: 9). This effect may lead procedure-­ oriented winners to still express more diffuse support (i.e. support for the democratic principles).

2.4 Summary In this chapter, I have elaborated on the importance of moving beyond a focus on the mechanism of government formation, and toward differentiating between winners and losers to fully explain democratic consolidation. Much of the extant literature on institutional determinants of democratic consolidation categorizes countries based on their government formation processes. I argue that such broad categorizations with a focus on government formation may lead to an incomplete picture of the determinants of democratic consolidation for two reasons. First, only a few countries clearly match the ideal types of parliamentary and presidential or majoritarian and consensus democracies. Second, such classifications do not consider the crucial role of losers for democratic consolidation as they focus on the side of the winners. Democracies share the fact that through elections, they create inequalities among voters. Explaining democratic support by differentiating between these broad categories of government formation neglects the important role of losers. I therefore argue that when explaining democratic consolidation, it seems crucial to assess the different legislative rights and resources of legislative actors, bringing legislatures and especially losers into focus. The central question arising from these observations is the extent to which different parliamentary procedures influence democratic consolidation. The proposed answer focuses on the political support of winners and losers which is determined by the different levels of political power

Role of Parliamentary Procedures  41

parliamentary procedures grant to opposition players. The starting point for this argument is the holding of democratic elections, and the creation of winners and losers. The existing research has repeatedly shown that winners are generally more satisfied with democracy than losers are. However, democratic consolidation is highly dependent on the consent of losers. I have argued that it is inclusive parliamentary procedures, granting power to losers and thereby influencing the degree of political equality between winners and losers, that positively influence the level of support expressed by losers. Economic approaches to democratization and rational choice approaches to minority government formation have identified political power as a crucial element in determining political support of individuals and political actors. Hence, even though the focus on political power is not new, the crucial intervening element presented here is specific parliamentary procedures that create more political equality between winners and losers. In both Acemoglu and Robinson’s (2006) economic approach and Strøm’s (1990b) reasoning about minority governments, parliamentary procedures (respectively, de jure political power) determine the role of opposition players in a polity. Consequently, parliamentary procedures determine the degree of benefit these groups may obtain from a particular democratic system (i.e. the costs of losing elections). The more political inequality that exists between winners and losers, the more is at stake when losing elections. Inclusive parliamentary procedures (granting power to losers) can therefore reduce the costs of losing elections. In addition to the reasoning presented in the extant literature, the theoretical reasoning presented here differentiates between the effect of political power of opposition players on diffuse and specific support, respectively. From a procedure-oriented perspective as well as an outcome-­ oriented perspective, losers benefit when more power is granted to opposition players. Therefore, I expect that they will express higher levels of both diffuse and specific support. From a procedure-oriented perspective, I expect that winners will express more diffuse support, as more political equality may be associated with more democratic representation. On the contrary, from an outcome-oriented perspective, winners should express less specific support if opposition players have more power to influence the decision-making process.

Notes 1 See Chapter 4 for a detailed discussion of political support. 2 See Chapter 3 for the detailed operationalization of the concept. 3 See for example Przeworski’s (1991: 10) definition of democracy: “Democracy is a system in which parties lose elections”. 4 These parliamentary procedures are defined in standing orders of legislative chambers and national constitutions.

42  Moderating Role of Parliamentary Procedures 5 The definition of opposition players follows the one of losers presented before. Hence, opposition players are all actors that are not part of the executive or represented by it. 6 Inclusive parliamentary procedures can be seen as what Strøm calls “the potential influence of the parliamentary opposition” (Strøm 1984: 213) and what Acemoglu and Robinson call “de jure political power” of opposition parties (Acemoglu and Robinson 2006: 21). 7 Carey (2009: 37) considers the following countries from the Proyecto de Elites Latinoamericanas project: Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Peru, Venezuela. For further information about the project see PELA (2020). 8 What Strøm (1984: 213) defines as “the potential influence of the parliamentary opposition” can be linked to what Acemoglu and Robinson (2006: 21) define as “de jure political power” of the majority and the minority in a polity. 9 Opposition parties in Strøm’s terms. 10 Government parties in Strøm’s terms. 11 Citizens themselves can have access to the policy-making process through, for example, direct democratic institutions. See Chapter 4 for a detailed discussion of the operationalization of the concept of political power granted to opposition players.

Bibliography Acemoglu, Daron and James Robinson. 2006. Economic Origins of Dictatorship and Democracy. Cambridge: Cambridge University Press. Acemoglu, Daron and James Robinson. 2012. Why Nations Fail. The Origins of Power, Prosperity and Poverty. London: Profile Books. Anderson, Christopher J., Andre Blais, Shaun Bowler, Todd Donovan and Ola Listhaug. 2005. Losers’ Consent: Elections and Democratic Legitimacy. Oxford and New York: Oxford University Press. Anderson, Christopher J. and Andrew J. LoTempio. 2002. “Winning, Losing and Political Trust in America.” British Journal of Political Science 32(2): 335–351. Anderson, Christopher J. and Yuliya V. Tverdova. 2001. “Winners, Losers, and Attitudes about Government in Contemporary Democracies.” International Political Science Review 22(4): 321–338. Benhabib, Jess and Adam Przeworski. 2006. “The Political Economy of Redistribution under Democracy.” Economic Theory 29: 271–290. Bernauer, Julian and Adrian Vatter. 2012. “Can’t Get No Satisfaction with the Westminster Model? Winners, Losers and the Effects of Consensual and Direct Democratic Institutions on Satisfaction with Democracy.” European Journal of Political Research 51(4): 435–468. Blais, André, Alexandre Morin-Chassé and Shane P. Singh. 2017. “Election Outcomes, Legislative Representation, and Satisfaction with Democracy.” Party Politics 23(2): 85–95. Boix, Carles. 2003. Democracy and Redistribution. Cambridge: Cambridge University Press. Boix, Carles. 2006. The Roots of Democracy. Policy Review 135: 3–21.

Role of Parliamentary Procedures  43 Bouteca, Nicolas, Jef Smulders, Bart Maddens, Carl Devos and Bram Wauters. 2019. “‘A Fair Day’s Wage for a Fair Day’s Work’? Exploring the Connection between the Parliamentary Work of MPs and their Electoral Support.” The Journal of Legislative Studies 25(1): 44–65. Bowler, Shaun. 2010. “Private Members’ Bills in the UK Parliament: Is There an ‘Electoral Connection’?” The Journal of Legislative Studies 16(4): 476–494. Bowler, Shaun and Todd Donovan. 2007. “Reasoning about Institutional Change: Winners, Losers and Support for Electoral Reforms.” British Journal of Political Science 37(3): 455–476. Carey, John M. 2009. Legislative Voting and Accountability. Cambridge: Cambridge University Press. Cheibub, José Antonio. 2002. “Minority Governments, Deadlock Situations, and the Survival of Presidential Democracies.” Comparative Political Studies 35(3): 284–312. Cheibub, José Antonio. 2007. Presidentialism, Parliamentarism, and Democracy. Cambridge: Cambridge University Press. Cox, Gary W. 2000. “On the Effects of Legislative Rules.” Legislative Studies Quarterly 25(2): 169–192. Cox, Gary W. 2006. The Organization of Democratic Legislatures. In The Oxford Handbook of Political Economy, eds. Barry B. Weingast and Donald A. Wittman. Oxford and New York: Oxford University Press: 141–161. Craig, Stephen C., Michael D. Martinez, Jason Gainous and James G. Kane. 2006. “Winners, Losers, and Election Context: Voter Responses to the 2000 Presidential Election.” Political Research Quarterly 59(4): 579–592. Dahl, Robert A. ed. 1966. Political Oppositions in Western Democracies. New Haven, CT and London: Yale University Press. Dahl, Robert A. 2006. On Political Equality. New Haven, CT and London: Yale University Press. Dahlberg, Stefan and Jonas Linde. 2015. “The Winner-Loser Gap in Satisfaction with Democracy Over Time. Evidence from a Swedish Citizen Panel.” QoG Working Paper Series 2015:1. University of Gothenburg: The Quality of Government Institute. Dalton, Russell. 2021. “The Representation Gap and Political Sophistication: A Contrarian Perspective.” Comparative Political Studies 54(5): 889–917. Däubler, Thomas, Thomas Bräuninger and Martin Brunner. 2016. “Is Personal Vote-Seeking Behavior Effective?” Legislative Studies Quarterly 41(2): 419–444. Diamond, Larry J. 1999. Developing Democracy: Toward Consolidation. Baltimore, MD: Johns Hopkins University Press. Diermeier, Daniel, Carlo Prato and Razvan Vlaicu. 2015. Procedural Choice in Majoritarian Organizations. American Journal of Political Science 59(4): 866–879. Easton, David. 1957. “An Approach to the Analysis of Political Systems.” World Politics 9(3): 383–400. Easton, David. 1965a. A Framework for Political Analysis. Englewood Cliffs, NJ: Prentice Hall. Easton, David. 1965b. A Systems Analysis of Political Life. New York: Wiley.

44  Moderating Role of Parliamentary Procedures Easton, David. 1975. “A Re-Assessment of the Concept of Political Support.” British Journal of Political Science 5(4): 435–457. Fish, Steven M. 2006. “Stronger Legislatures, Stronger Democracies.” Journal of Democracy 1: 5–19. Fish, Steven M. and Matthew Kroenig. 2009. The Handbook of National Legislatures: A Global Survey. Cambridge: Cambridge University Press. Fortunato, David and Randolph T. Stevenson. 2021. Party Government and Political Information. Legislative Studies Quarterly 46(2): 1–45. Gabriel, Oscar W. 1989. “Regierungswechsel und politische Unterstützung: Implikationen des Parteienwettbewerbs für die Struktur politischer Unterstützung in der Demokratie.” Politische Vierteljahresschrift 30(March): 75–93. Helms, Ludger. 2004. “Five Ways of Institutionalizing Political Opposition: Lessons from the Advanced Democracies.” Government and Opposition 39(1): 22–54. Henderson, Ailsa. 2008. “Satisfaction with Democracy: The Impact of Winning and Losing in Westminster Systems.” Journal of Elections, Public Opinion and Parties 18(1): 3–26. Holden, Matthew. 2006. Exclusion, Inclusion, and Political Institutions. In The Oxford Handbook of Political Institutions, eds. Sarah A. Binder, R.A.W. Rhodes and Bert A. Rockman. Oxford and New York: Oxford University Press: 163–190. Huber, Lena Maria, Anita Bodlos, Elisabeth Graf and Thomas M. Meyer. 2022. “Disseminating Legislative Debates: How Legislators Communicate the Parliamentary Agenda.” Party Politics 28(2): 365–376. Ionescu, Ghita and Isabel de Madariaga. 1968. Opposition: Past and Present of a Political Institution. London: C.A. Watts. Lijphart, Arend. 1999. Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. New Haven, CT and London: Yale University Press. Linz, Juan J. 1990. “The Virtues of Parliamentarism.” Journal of Democracy 1(4): 84–91. Linz, Juan J. and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation. Baltimore, MD: Johns Hopkins University Press. Lipset, Seymour Martin. 1959. “Some Social Requisites of Democracy Economic Development and Legitimacy.” American Political Science Review 53: 69–105. Lipset, Seymour Martin. 1963. “The Value Patterns of Democracy: A Case Study in Comparative Analysis.” American Sociological Review 28(4): 515–531. Marcinkiewicz, Kamil and Mary Stegmaier. 2019. “Speaking Up to Stay in Parliament: The Electoral Importance of Speeches and Other Parliamentary Activities.” The Journal of Legislative Studies 25(4): 576–596. McCubbins, Matthew D. 2005. Legislative Process and the Mirroring Principle. In Handbook of New Institutional Economics, eds. Claude Ménard and Mary M. Shirley. Dordrecht: Springer 123–147. McKay, Lawrence. 2020. “Does Constituency Focus Improve Attitudes to MPs? A Test for the UK.” The Journal of Legislative Studies 26(1): 1–26. Nadeau, Richard and Andre Blais. 1993. “Accepting the Election Outcome: The Effect of Participation on Losers’ Consent.” British Journal of Political Science 23(4): 553–563.

Role of Parliamentary Procedures  45 PELA. 2020. “Latin American Elites project of the University of Salamanca (PELA-­USAL)”. Online: https://oir.org.es/pela/en/. Accessed: December 2020. Przeworski, Adam. 1991. Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America. Cambridge: Cambridge University Press. Przeworski, Adam. 2019. Crises of Democracy. Cambridge: Cambridge University Press. Randall, Vicky. 2012. Political Parties. In Routledge Handbook of Democratization, ed. Jeffrey Haynes. London and New York: Routledge: 222–236. Riker, W.H. 1983. Political Theory and the Art of Heresthetics. In Political Science: The State of the Discipline, ed. Ada W. Finifter. Washington, DC: American Political Science Association Washington: 47–67. Saalfeld, Thomas. 2011. “Parliamentary Questions as Instruments of Substantive Representation: Visible Minorities in the UK House of Commons, 2005– 10.” The Journal of Legislative Studies 17(3): 271–289. Sartori, Giovanni. 2005[1976]. Parties and Party Systems: A Framework for Analysis. Colchester: ECPR Press. Schattschneider, Elmer Eric. 1942. Party Government. New York: Rinehart. Schumpeter, Joseph A. 1942. Socialism, Capitalism and Democracy. New York: Harper and Bros. Shugart, Matthew Søberg. 2006. Comparative Executive-Legislative Relations. In The Oxford Handbook of Political Institutions, eds. R.A.W. Rhodes, Sarah A. Binder and Bert A. Rockman. Oxford and New York: Oxford University Press: 344–365. Shugart, Matthew Søberg and John M. Carey. 1992. Presidents and Assemblies Constitutional Design and Electoral Dynamics. New York: Cambridge University Press. Singh, Shane, Ignacio Lago and Andre Blais. 2011. “Winning and Competitiveness as Determinants of Political Support.” Social Science Quarterly 92(3): 695–709. Singh, Shane P., Ekrem Karakoç and André Blais. 2012. “Differentiating Winners: How Elections Affect Satisfaction with Democracy.” Electoral Studies 31(1): 201–211. Strøm, Kaare. 1984. “Minority Governments in Parliamentary Democracies. The Rationality of Nonwinning Cabinet Solutions.” Comparative Political Studies 17: 199–227. Strøm, Kaare. 1990a. “A Behavioral Theory of Competitive Political Parties.” American Journal of Political Science 34: 565–598. Strøm, Kaare. 1990b. Minority Government and Majority Rule. New York: Cambridge University Press. Tsebelis, George. 1995. “Decision making in political systems: Veto players in presidentialism, parliamentarism, multicameralism and multipartyism.” British Journal of Political Science 25(03): 289–325. Tsebelis, George and Jeannette Money. 1997. Bicameralism. Cambridge: Cambridge University Press. Tsebelis, George. 2002. Veto players: How political institutions work. Princeton, NJ: Princeton University Press.

46  Moderating Role of Parliamentary Procedures Tuttnauer, Or and Simone Wegmann. 2022. “Voting for Votes: Opposition Parties’ Legislative Activity and Electoral Outcomes.” American Political Science Review: 1–18. Wegmann, Simone. 2022. “Policy-Making Power of Opposition Players: A Comparative Institutional Perspective.” The Journal of Legislative Studies 28(1): 1–25.

Chapter 3

Measuring Opposition Power1

The purpose of this chapter is to present the concept of the policy-­making power of opposition players and its operationalization. As Figure 2.2 in Chapter 2 shows, the theoretical framework presents two groups that are represented in the legislature: winners and losers of democratic elections (represented by MPs of the government party/parties and MPs of opposition parties). Furthermore, the theoretical argument suggests that inclusive parliamentary procedures define the degree of power allocated to these different actors in the legislative arena. Therefore, to determine the degree of political power granted to opposition players, parliamentary procedures must be assessed and categorized with a special focus on the role and power granted to actors from the opposition. However, there has not been a great deal of guidance for operationalizing the concept as relatively few contributions in the field of legislative studies have focused on the role of the opposition in the legislature.2 Among the earliest contributions in the literature on legislative studies is the edited volume on political oppositions in Western democracies by Dahl (1966b). In his concluding remarks, Dahl (1966a) proposes six differentiating features of oppositions: organizational cohesion, competitiveness, identifiability, goals, strategies, and the site of encounter between the opposition and the government. Blondel (1997: 470–471) proposed merging the features of oppositions proposed by Dahl (1966a) in order to create a two-dimensional categorization of oppositions: the distance between the goals of the agents of opposition from those of the government and the relative strength of the bodies of the opposition. However, these earlier contributions did not focus on specific rights of the opposition to influence the policy-making process. Beyond the research on different types of oppositions and the features used to distinguish between them, most of the existing studies on parliamentary oppositions focus on relatively few cases (see, e.g. Church and Vatter 2009; Andeweg, Winter and Müller 2008; Christiansen and Damgaard 2008; Inoguchi 2008; Kaiser 2008; Kopecky and Spirova 2008; Schrire 2008; Mujica and Sanchez-Cuenca 2006; Gel’man 2005; DOI: 10.4324/9781003005360-4

48  Moderating Role of Parliamentary Procedures

Helms 2004) or do not look at political regimes beyond one specific type of democracy. Several studies either focus on parliamentary (see, e.g. Garritzmann 2017; Sieberer 2011; Schnapp and Harfst 2005) or on presidential democracies (see, e.g. Morgenstern, Negri and Pérez-Liñán 2008). Furthermore, research often depicts parliamentary opposition as inherent to parliamentary regimes and unimportant to presidential ones. However, considering that policy-making is one of the most important tasks of legislatures in democracies, opposition to the government party or parties can play an important role in both parliamentary and presidential regimes. Moreover, most of the contributions to the research on oppositions focus on the characteristics of oppositions, how they organize, or how they behave. However, the literature rarely addresses the specific rights that the rules of procedure of legislative chambers grant to the opposition.3 The conceptualization presented in the following sections combines the two core functions of legislatures discussed earlier4: policy-making and checking majority and executive power.5 The first core function of parliaments – policy-making – links extensively to political support. Citizens are likely to evaluate the democratic system based on the outcomes it generates. Different policies influence citizens in different ways as some benefit, while others do not, thereby affecting their view of both democracy and the government. Policy-making is therefore likely to be one of the central concerns of citizens.6 The second core function of parliaments – checks – refers to the “oversight and limitations on the ability of policy-­ makers to take action” (Carey 2006: 447). These checks can manifest in two ways: more opposition rights or fewer government rights. Therefore, such checks define the inclusiveness of parliamentary procedures. These procedures determine how much action the government and opposition parties can take. Specifically: few opportunities for the opposition to take action and/or be included in the decision-making process indicates a context with exclusionary parliamentary procedures (from the point of view of the opposition) whereas many such opportunities for the opposition indicate inclusionary parliamentary procedures and point to more policy-making power for opposition players. This chapter proceeds as follows. In the first section, I present the operationalization of the policy-making power of opposition players. This includes a detailed presentation of seven variables that are combined into a final index of policy-making power of opposition players. In the second section, I present information on data collection to measure opposition power, and in the third section, I present a summary of the operationalization and data-collection effort.

3.1  Policy-Making Power of Opposition Players Figure 3.1 presents an overview of the operationalization of policy-­ making power of opposition players. Coding decisions are deduced from

Measuring Opposition Power  49

Figure 3.1  Policy-Making Power of Opposition Players Source: Adopted from Wegmann (2022: 4)

the existing research on legislative organization. As highlighted by Strøm (1990: 42) on the power of oppositions, a policy influence differential results from internal parliamentary structures and procedures. A higher policy influence differential indicates more power of the government party as compared to the opposition. Hence, payoffs for policy-seeking oppositions are bigger in a polity with a lower policy influence differential. I apply this logic about the policy influence differential to be used to think about the role and specific power of the opposition to influence the policy-­making process. The policy influence differential can manifest itself in two aspects: more opposition rights or less government rights. Hence, such rights define the policy influence differential in the sense that they determine how much action the government and opposition players can take. An extensive literature exists on the specific rules and procedures of the policy-making process. Even though most of this literature focuses on the

50  Moderating Role of Parliamentary Procedures

role of the government, it allows for the identification of the policy influence differential and thereby the assessment of the power granted to opposition players. Thus, the existing literature on parliamentary rules and procedures of the policy-making process serves as a guideline for which variables should be included in an index measuring opposition power. The structure of the concept follows three stages of the policy-making process: initiation, debate, and veto. At the stage of initiating policies, I consider bill introduction and agenda setting. Amendments, the committee structure, and the committee procedures represent the stage of debate power. Finally, executive power and referendums measure the veto power of opposition players. All the variables are formulated such that higher scores indicate a reduction in the policy influence differential and thus more power for opposition players as compared to the government. Thus, the policy-making power of opposition players refers to the “capacity for checking majority action within legislatures” (Carey 2006: 433) during the policy-making process. In this sense, the power of opposition players is the de jure political power meaning the power allocated by political institutions. Therefore, institutions that grant opposition players a greater ability to influence the legislative process are important. Following Carey (2006), the index presented here considers two types of “checks” or political power. On the one hand, internal checks refer to de jure political power granted to different actors within the legislature, including opposition MPs and parties. On the other hand, external checks refer to de jure political power granted to different actors outside the legislature, including citizens or executive power (Carey 2006: 448). Internal checks are most visible in the initiation and debate stage of the policy-making process. The more actors that are entitled to introduce legislative bills, propose amendments to bills, and the fewer restrictions to do so, the more power opposition players have in comparison with the government. Similarly, parliamentary committees structured in a decentralized way that allow participation of opposition players give the latter more power compared to committee systems dominated by the government parties (see also Strøm 1990). The influence the government has as external actor figures into all three stages of the policy-making process. This is due to the underlying logic of the policy influence differential as it captures the power granted to opposition players in relation to the dominance of government powers.7 The government is one of the potential actors that can introduce bills and dominate the agenda in the initiation stage of the policy-making process. Furthermore, the government is also a potential actor able to introduce amendments and to dominate the committee system. Finally, executive power can considerably influence the final stage of the policy-making process. Other external actors such as citizens and subnational entities can influence the policy-making process in the initiation and the veto stage. Both are among the possible

Measuring Opposition Power  51

actors who have the right to introduce bills, while citizens have the potential to veto bills through referendums. In the following sections, I discuss each of the variables in more detail (see Figure 3.1). The first section discusses the initiation power considering the power to propose legislation and agenda setting. The second section presents the stage of debate power looking at the rights of opposition actors during the deliberation process. The third section deals with the veto power of different actors and their relation to the power of opposition players. Finally, the fourth section presents the final index of the policy-making power of opposition players. 3.1.1  Initiation Power “[W]ithout parliamentary rights (…) a legislator cannot make noteworthy contributions to the legislative product” (Krehbiel 1991: 2). The first step to enable such contributions is the right to propose legislation. Hence, as presented in Figure 3.1, the first variables measuring the policy-making power of opposition players relate to initiation power and capture the ability of opposition players to initiate the policy-making process. However, actors that exert power on the parliamentary agenda can limit the power of individual actors to propose legislation. Initiation power is therefore determined by bill introduction and agenda setting power. 3.1.1.1  Bill Introduction As shown by Mattson (1995: 455) and Taylor (2006: 326), representative democracies allow individual members of parliament to influence the policy-making process through the right to introduce legislative bills. However, such rights are not only granted to MPs. The executive often dominates the right to initiate legislation (Mattson 1995: 455), giving it considerably more power than the opposition. Even so, actors other than the executive or individual MPs may also have the right to introduce legislation. Parliamentary committees or parliamentary parties that have the right to introduce legislation, for example, grant additional power to the opposition. Furthermore, citizens may propose legislation through popular initiatives (Mattson 1995: 455–456). As the executive and individual MPs should have the right to introduce bills in all representative democracies (see, e.g. Taylor 2006; Mattson 1995), the initiation power of opposition players largely depends on the number of other actors (parliamentary parties, committees, subnational entities, and citizens) that also have the right to introduce bills. Importantly, procedural restrictions and potential veto players determine the right to introduce bills. The literature has identified two different kinds of restrictions that limit the power of opposition players. First,

52  Moderating Role of Parliamentary Procedures

to prevent the introduction of bills that lack substantial support some legislative chambers apply numerical restrictions. For example, they require a specific number or percentage of MPs to support a bill in order for it to be introduced (Mattson 1995: 457). This considerably reduces the power of individual MPs and may create an obstacle for opposition players.8 Second, some chambers require other actors to accept the introduction of a bill9 before it can be introduced thereby also reducing the potential power of opposition players. Consequently, the variable 1a. bill introduction (see Figure 3.1) considers both the number of actors that are allowed to introduce bills as well as restrictions and veto players limiting the power to do so. The opposition is weak in chambers where only the executive and MPs are able to introduce bills and MPs face restrictions and potential veto players. The opposition has more (though still quite limited) power in chambers where at least one other actor (i.e. committees, parties, subnational entities or citizens) is allowed to introduce bills but are also faced with restrictions and/or veto players.10 Opposition players are granted strong power if MPs face neither restrictions nor veto players. The strongest opposition power is granted in chambers in which at least one other actor than the executive and MPs can introduce bills without restrictions or veto. An additional obstacle to overcome, however, is the actors that decide on the plenary agenda of the legislative chamber which might be dominated by the government. 3.1.1.2  Agenda Setting Power Control of the plenary timetable is linked to control of what will be debated and decided. Döring points to the fact that “[o]pposition bills may be stopped in some instances by simply keeping them off the agenda fixed by the government” (Döring 1995: 224; see as well Cox and McCubbins 2011; Tsebelis and Rasch 2011). Therefore, being able to define this timetable and setting the order of the day implies considerable power (Laver and Shepsle 1994: 295). If this power is in the hands of the government, it reduces the initiation power of opposition players. Two elements are considered here. First, the government may control policy-making if it can determine the plenary timetable (Mattson 1995: 459–461). The degree of influence the government has on the plenary agenda depends on whether or not the government alone can set the agenda (Döring 2001, 1995). Full government control of the plenary timetable obviously reduces the chance of the successful introduction of a bill by opposition players. However, the initiation power of opposition players may be stronger when collective actors including opposition MPs such as a steering committee or the legislative chamber itself, set the plenary agenda.

Measuring Opposition Power  53

Second, the existence of specific rules which give priority to government business diminish the opposition players’ power to introduce legislation. As emphasized by Cox’s (2006) bottleneck and the necessity of regulating plenary time, free access to the plenary agenda is highly unlikely (cf. Cox and McCubbins, 2011: 458). However, the degree to which the government is given priority access varies considerably. Prioritizing government business during plenary time gives more power to the government and less power to the opposition. Special times for opposition or individual member’s business grant stronger initiation power to opposition players. In the situation with the highest level of initiation power for opposition players, no priority is given to the government. The variable 1b agenda setting (see Figure 3.1) combines both, the power to organize the plenary timetable and the prioritizing of government business. Opposition players are granted strong power if either the legislative chamber itself or a collective body, such as a steering committee, sets the plenary timetable11 and there is no prioritizing of government business. 3.1.2  Debate Power The successful introduction of a legislative bill is followed by the debate phase in the legislative chamber. Again, without specific rights to do so, MPs cannot meaningfully contribute to this phase of the policy-making process (see Krehbiel 1991). The literature has repeatedly pointed to the crucial role of legislative committees at this stage of policy-making as “much of the real deliberation takes place away from the plenary arena in much smaller groups of legislators such as legislative committees” (Mattson and Strøm 1995: 249). Committees support “(…) the development and review of policy proposals in their domains, and [draw] on the expertise of their members and staff to make recommendations to the full assembly” (Carey 2006: 441). Contributing to the work of committees, therefore, is an important way of influencing policy-making. However, the influence of opposition players in such committees depends on the specific structure and the procedures of the committee system. The measure of debate power of opposition players includes these three variables: the strength of the committee structure, the strength of the committee procedures, and the power to introduce amendments.12 3.1.2.1 Amendments Similar to bill introduction, the right to introduce amendments is subject to restrictions and limited to certain actors (Mattson 1995: 473). The variable follows the operationalization of bill introduction presented before. The more actors, including opposition players, that are entitled to

54  Moderating Role of Parliamentary Procedures

propose amendments to bills the more power the opposition has, compared to the power of the government. As previously stated, restrictions affecting the power of opposition players can be numerical, meaning that legislative chambers require the support of a specific number or percentage of MPs for an amendment to be proposed. Furthermore, other actors can be required to accept the proposal of an amendment before it is introduced.13 Consequently, the variable 2a amendments considers who, besides the government, can introduce amendments and the restrictions and veto players limiting the power to do so. If only one actor (e.g. MPs) may introduce amendments but they face restrictions and veto players, the power of opposition players will be weak. The opposition has most power when at least two actors can introduce amendments in the chambers without restriction or veto (see Figure 3.1). 3.1.2.2  Committee Structure The specific structure of legislative committees determines whether opposition MPs have access to such committees or whether they are dominated by government MPs. The variable of committee structure aims to assess how much access the opposition has to parliamentary committees. Opposition parties can exert more influence in a “strong and decentralized committee structure” (Strøm 1990: 70). I follow Strøm’s “five-point index of the potential for oppositional influence” by considering the following (Strøm 1990: 71): (i) the possibility of the opposition to influence the deliberation requires a minimum number of standing committees. According to Strøm “more than 10 standing committees [are] necessary for high oppositional influence” (Strøm 1990: 71). (ii) Similarly, fixed areas of specialization that (iii) correspond to ministerial departments guarantee more deliberative legislatures and more effective oversight which leads to higher levels of influence by opposition parties. (iv) Furthermore, a numerical restriction of committee assignments per MP leads to more specialized legislators and fewer opportunities for MPs of government parties to manipulate membership allocations to its advantage. Finally, (v) a proportional distribution of committee chairs, according to parliamentary parties, guarantees that government parties do not control all committee chairs and guarantees more political power to opposition parties (Mattson and Strøm 1995: 277; Strøm 1990: 71). As Figure 3.1 shows, the variable 2b committee structure is the sum of all positive answers regarding the presented characteristics and ranges from 0 (weak opposition power) to 5 (strong opposition power). 3.1.2.3  Committee Procedures The possibility for opposition MPs to become members of parliamentary committees is important in order to influence the policy-making process

Measuring Opposition Power  55

but specific procedures of these committees are also important as they can considerably shape the policy influence differential. Committee procedures grant power to MPs of government and opposition parties based on the degree of influence opposition MPs have within these committees (Mattson and Strøm 1995: 274). The literature has identified several characteristics that determine the power of committees. The following three indicators especially relate to the power granted to opposition MPs and therefore are considered here: (i) the right to submit minority reports along with the committee report to the chamber (Mattson and Strøm 1995: 283–284), and (ii) the degree to which the government can influence the agenda of the committees. Both of these indicators vary considerably across polities (Döring 1995: 237–238). Hence, the strength of the committee procedures is determined by whether the committees themselves can set their own agenda or whether the government can exert influence over the agenda. In countries where the government cannot influence the agenda of committees, the policy influence differential is reduced and opposition MPs may have more political power compared to the government. (iii) A third characteristic considers the right of committees to rewrite government bills. Where committees are free to rewrite bills “…, government prerogatives in the procedure for passing legislation would be considerably curbed” (Döring 1995: 235). Therefore, as Figure 3.1 shows, the variable 2c committee procedures is the sum of all positive answers regarding the three characteristics and ranges from 0 (weak opposition power) to 3 (strong opposition power). 3.1.3  Veto Power What follows the successful initiation and debate of a legislative bill is its adoption or rejection. In case of an adoption, many polities have specific rights for certain actors to veto the adoption or to issue laws without the necessary approval of the legislature. Hence, to assess the specific powers of opposition players in the policy-making process, we need to consider that “legislatures everywhere are embedded in broader institutional environments in which policy-making decisions depend on multiple actors” (Carey 2006: 433). Actors that have veto power are part of such an environment.14 Therefore, veto power is a crucial aspect that, depending on which actors hold veto power, may grant or restrict the power of opposition players to influence the decision-making process. Certainly, the executive is one actor that may influence the policy-­ making process to a high degree if provided with certain rights. The level of influence the executive may exert in this final stage of the policy-­ making process is an important element in determining the policy influence differential. Furthermore, in some countries the executive as well as other actors have the rights to challenge the adoption of particular

56  Moderating Role of Parliamentary Procedures

bills by asking for a referendum. Therefore, the measure of veto power of opposition players includes the two variables of executive power and referendums. 3.1.3.1  Executive Power Tsebelis (1995: 305) states that in some presidential systems the president is not given the power of veto and therefore not all presidents can be considered veto players. Similarly, Mainwaring and Shugart have emphasized that the “dynamics of presidential systems vary according to president’s formal powers” (Mainwaring and Shugart 1997: 464). Such powers define the degree to which executive domination of the political process is possible. Veto power which allows an actor to “defend the status quo against attempts to change it” demonstrates reactive powers, whereas, decree power which allows the executive “to attempt to establish a new status quo” demonstrates proactive powers (Mainwaring and Shugart, 1997: 464; see as well Carey and Shugart, 1998: 8). Such executive decree and/or veto power influences the policy influence differential in the sense that the government is granted considerably more power than opposition players. Thus, as Figure 3.1 shows, the variable 3a executive power measures whether the executive has both veto and decree power, (weak opposition power), one of the two powers, or neither power (strong opposition power). 3.1.3.2 Referendums However, the executive is not the only potential actor that can veto the adoption of a bill. The possibility of referendums grants this power to citizens or parliamentary oppositions. To be able to assess the degree of power opposition players can exert using referendums, a simple look at legal provisions whether mandatory and/or optional referendums are possible is not enough. Even though citizens may act as veto players when granted power through a required referendum or a referendum requested by another actor, not all polities that know referendums also grant their initiation to opposition players. However, through optional referendums, opposition players have the potential to influence the policy-making process to an important extent if they have corresponding rights. Taking a closer look at referendums, Hug and Tsebelis differentiate between the two questions of “(1) who asks the question and (2) who triggers a referendum” (Hug and Tsebelis 2002: 466). Following these authors, the veto power of opposition players increases if they can actually trigger a referendum. Therefore, the variable 3b referendum differentiates between four degrees of opposition power (see Figure 3.1): (i) neither mandatory nor optional referendum possibilities (weak opposition power),

Measuring Opposition Power  57

(ii) optional referendum possibilities that may not be triggered by opposition actors, (iii) mandatory referendums, and (iv) optional referendums that may be triggered by opposition actors (strong opposition power). 3.1.4  M easuring Policy- Making Power of Opposition Players Although recent contributions to legislative studies have started to look at the role of the opposition, overall, oppositions have not figured prominently in the legislative studies literature. The question of the power of oppositions in different systems, especially, has not taken up much attention. One reason for the lack of comparative research on the topic might be the lack of data necessary for such an endeavor. The proposed measure of policy-making power of opposition players is one such attempt to measure the power granted to opposition players in a comparative perspective and might thereby contribute to our understanding of legislative organization and behavior, more generally. To measure the power of opposition players during the policy-making process, the different variables described above are combined to form an index. Factor analysis supports an additive index as results show all seven variables to load on one ­factor15 – however, executive power shows factor loadings below 0.3. These results also indicate that the three stages of policy-making (i.e. initiation, debate, and veto) are not independent dimensions of the concept of the policy-making power of opposition players but different stages of one process. This result leads to the decision not to apply weights to these different stages when composing the overall index, but to base the aggregation rule on the seven variables. The final index treats these seven variables as different aspects of policy-making power of opposition players that can substitute each other. None of the variables is a necessary condition for strong opposition power. As Figure 3.1 shows, for each variable, higher scores mean more policy-­making power for opposition players. However, due to the different ranges of the variables, a simple additive index would give more weight to variables with larger ranges (such as committee structure or bill introduction) compared to variables with smaller ranges (such as agenda setting or executive power). Therefore, the seven variables are standardized before aggregation. These standardized variables are then combined in an additive index, which is again standardized. Thus, the resulting index of the policy-making power of opposition players is the average of the seven variables and takes values between 0 (weak opposition power) and 1 (strong opposition power). Figure A.3.1 of the Appendix shows correlations between the final index (as described above) and indices based on alternative aggregation rules. The alternatives depicted in the figure are a simple additive index without standardized variables,

58  Moderating Role of Parliamentary Procedures

an index based on both standardized variables and standardized stages of the policy-making process, and an index based on factor analysis. As Figure A.3.1 shows, the correlations among the different indices are very high (between 0.96 and 0.99). Therefore, in the following chapters the additive index with standardized variables as discussed here will serve as a measure of the policy-making power of opposition players.

3.2  M  ethod of Data Collection – Policy-Making Power of Opposition Players The inclusiveness of parliamentary procedures (which define the policy-­ making power of opposition players) is determined by the official rights during the policy-making process attributed to government and opposition players. Therefore, data collection was based on documents officially published by the state institutions. The essential reference documents for each country are the official rules of procedure of the legislative chamber and the national constitution. However, these two documents do not always specify rules and procedures for all the variables discussed in the previous section. Most notably, rules on the organization of committees are often not specified in the constitution or the parliamentary standing orders. Therefore, to code this information, when it is not available in constitutions or standing orders, the official parliamentary and/or government websites of a given country are used as a source of information. The criteria to decide whether a document is used for coding information is its publication by an official state institution (i.e. the legislative chamber or the government). Besides official documents published by state authorities, several existing data sources were used to complement some of the information. The Comparative Constitutions Project (Elkins, Ginsburgh and Melton 2014) is used for information about bill introduction by different actors. However, as this dataset only considers national constitutions, parliamentary standing orders were consulted to complement the information in cases where the actors with the right to introduce legislation were not fully listed in the national constitution. The Parliamentary Power Index (Fish and Kroenig 2009) is used to determine whether the executive has decree power and the Institutions and Elections Project (Regan and Clark 2014) provides information about executive veto power. Finally, the Direct Democracy Database (IDEA 2014) provides detailed information about the possibilities of referendum and citizen initiatives. Coding decisions are based on the most recent versions of constitutions and parliamentary standing orders available in 2015 for lower chambers in 54 democracies around the world.16 All countries with at least one year of democratic experience after 1990 as measured by Marshall, Jaggers and Gurr (2014) were considered.17

Measuring Opposition Power  59

3.3 Summary In this chapter, I have presented the operationalization of the concept of the policy-making power of opposition players. In the past, relatively few contributions in the field of legislative studies have focused on the role of the opposition. Hence, not much guidance exists for a task such as measuring opposition power. However, a vast literature exists on the specific rules and procedures of the policy-making process from the perspective of the government. I apply the logic of the policy influence differential (Strøm 1990) to measure the specific powers of the opposition players to influence the policy-making process based on this literature. As Strøm (1990) highlights, internal parliamentary structures and procedures result in a policy influence differential which can manifest itself in two closely related ways: more opposition rights or less government rights. Hence, the policy influence differential indicates how much action the government can take compared to the opposition. A high-policy influence differential indicates more power for the government as compared to the opposition whereas a low policy influence differential indicates more payoffs for policy-making opposition players. Hence, a low-policy influence differential points toward more inclusive parliamentary procedures from the perspective of opposition players. The conceptualization of the policy-making power of opposition players presented in this chapter refers to two core functions of legislatures: policy-making and checking majority and executive power. I have argued that opposition players can influence the policy-making process in three different stages, namely: the initiation, debate, and veto stages. At the stage of initiating policies, I have considered the rights to introduce bills and agenda setting powers. At the debate stage, I have included the power to introduce amendments, the committee structure, and committee procedures. Finally, at the veto stage, I have considered executive power and the rights to ask for referendums. To measure the policy-making power of opposition players, these seven variables are combined into an overall index. As higher scores indicate more policy-making power of opposition players (see Figure 3.1), the final index is the average of the seven variables and takes values between 0 (weak opposition power) and 1 (strong opposition power).

Notes 1 This chapter is largely based on Wegmann (2022). 2 See e.g. also Helms’ (2008) introduction to the special issue of The Journal of Legislative Studies on Parliamentary Opposition in Old and New Democracies. The author mentions that “[r]egrettably, there is no equivalent research regarding the sub-field of parliamentary opposition. (…) Indeed, a large proportion of contributions to the field could be described as ‘occasional papers’

60  Moderating Role of Parliamentary Procedures by authors specialising on other areas and aspects of legislative research” (Helms 2008: 7). 3 Garritzmann (2017), Kaiser (2008), and Schnapp and Harfst (2005) are notable exceptions. 4 See Chapter 1. 5 The conceptualization does not directly address the third core function (representation) since the theoretical framework excludes the selection process of MPs but instead takes the situation with two groups (winners and losers) after a democratic election as the point of departure. 6 However, I do not claim that looking at policy-making can assess the whole range of potential influence of opposition players (see also Sieberer 2011). 7 Therefore, the rights considered to measure opposition power are not necessarily confined to opposition players. Government actors might enjoy the same rights which constitutes a higher policy-influence differential (Strøm 1990). 8 For example, in Germany only parliamentary parties or 5 percent of all MPs are entitled to propose a legislative bill (Bräuninger and Debus 2009: 809). 9 In Belgium, for example, “the Speaker must agree with the proposal before it can be translated, printed and distributed amongst the members of the House…” (Mattson 1995: 462). 10 As it is natural to require a certain number of signatures for citizen initiatives, only cases in which the number of signatures is particularly high (equals 10 percent or more of voters) are considered to constitute restrictions. 11 As Figure 3.1 shows, I labelled the agenda-setting actor collective actor if either the legislative chamber itself or a steering committee set the plenary timetable. Situations in which the government controls the plenary timetable are labelled single actor, even though the government is also formally a ’collective actor’. 12 Mattson and Strøm highlight “committee powers” including the power to initiate bills and amendments as well as the power to set their own agenda as important characteristics of committee systems (Mattson and Strøm 1995: 285). However, as parliamentary committees were considered as potential actors to have the right to introduce bills the element of bill introduction is not considered in this section. Similarly, committees might not be the only actors to have the right to submit amendments during the debate phase of the policy-making process. For example, the influence MPs can have on proposed bills also considerably depends on their right to propose amendments to these bills. Therefore, amendments are considered separately. 13 The French government, for example, is permitted “to group articles and amendments selectively, excluding the amendments to which it is opposed” (Huber 1992: 676). 14 A veto player is “(…) an individual or collective actor whose agreement (by majority rule for collective actors) is required for a change in policy” (Tsebelis 1995: 301). 15 But, as the sample size is small, results of factor analysis should be interpreted with caution. 16 See Table 4.2 of Chapter 4. 17 However, six countries were not classified as democracies at the time the standing orders came into force. These countries are: Bangladesh, Benin, Guatemala, Malawi, Pakistan, and Sierra Leone. For more information on the sample, see Chapter 4.

Measuring Opposition Power  61

Bibliography Andeweg, Rudy B., Lieven De Winter and Wolfgang C. Müller. 2008. “Parliamentary Opposition in Post-Consociational Democracies: Austria, Belgium and the Netherlands.” The Journal of Legislative Studies 14(1/2): 77–12. Blondel, Jean. 1997. “Political Opposition in the Contemporary World.” Government and Opposition 32(4): 462–486. Bräuninger, Thomas and Marc Debus. 2009. “Legislative Agenda-Setting in Parliamentary Democracies.” European Journal of Political Research 48(6): 804–839. Carey, John M. 2006. Legislative Organization. In The Oxford Handbook of Political Institutions, eds. Sarah A. Binder, R.A.W. Rhodes and Bert A. Rockman. Oxford and New York: Oxford University Press: 431–454. Carey, John M. and Matthew Soberg Shugart. 1998. Calling Out the Tanks or Filling Out the Forms? In Executive Decree Authority, eds. John M. Carey and Matthew Soberg Shugart. New York: Cambridge University Press: 1–32. Christiansen, Flemming Juul and Erik Damgaard. 2008. “Parliamentary Opposition under Minority Parliamentarism: Scandinavia.” The Journal of Legislative Studies 14(1/2): 46–76. Church, Clive H. and Adrian Vatter. 2009. “Opposition and Consensual Switzerland: A Short but Significant Experiment.” Government and Opposition 44(4): 412–437. Cox, Gary W. 2006. The Organization of Democratic Legislatures. In The Oxford Handbook of Political Economy, ed. Barry B. Weingast and Donald A. Wittman. Oxford, New York: Oxford University Press: 141–161. Cox, Gary W. and Mathew D. McCubbins. 2011. Managing Plenary Time: The U.S. Congress in Comparative Context. In The Oxford Handbook of the American Congress, eds. Eric Schickler and Frances E. Lee. Oxford and New York: Oxford University Press: 451–472. Dahl, Robert A. 1966a. Patterns of Opposition. In Political Oppositions in Western Democracies, ed. Robert Dahl. New Haven, CT and London: Yale University Press: 332–347. Dahl, Robert A. ed. 1966b. Political Oppositions in Western Democracies. New Haven, CT and London: Yale University Press. Döring, Herbert. 1995. Time as a Scarce Resource: Government Control of the Agenda. In Parliaments and Majority Rule in Western Europe, ed. Herbert Döring. Frankfurt: Campus Verlag: 223–246. Döring, Herbert. 2001. “Parliamentary Agenda Control and Legislative Outcomes in Western Europe.” Legislative Studies Quarterly 26(1): 145–165. Elkins, Zachary, Tom Ginsburgh and James Melton. 2014. “Characteristics of National Constitutions, Version 2.0.” Online: http://www.comparativeconstitutionsproject.org. Accessed: June 2013. Fish, Steven M. and Matthew Kroenig. 2009. The Handbook of National Legislatures: A Global Survey. Cambridge: Cambridge University Press. Garritzmann, Julian L. 2017. “How Much Power Do Opposition Have? Comparing the Opportunity Structures of Parliamentary Opposition in 21 Democracies.” The Journal of Legislative Studies 23(1): 1–30.

62  Moderating Role of Parliamentary Procedures Gel’man, Vladimir. 2005. “Political Opposition in Russia: A Dying Species?” Post-Soviet Affairs 21(3): 226–246. Helms, Ludger. 2004. “Five Ways of Institutionalizing Political Opposition: Lessons from the Advanced Democracies.” Government and Opposition 39(1): 22–54. Helms, Ludger. 2008. “Studying Parliamentary Opposition in Old and New Democracies: Issues and Perspectives.” The Journal of Legislative Studies 14(1/2): 6–19. Huber, John D. 1992. “Restrictive Legislative Procedures in France and the United States.” The American Political Science Review 86(3): 675–687. Hug, Simon and George Tsebelis. 2002. “Veto Players and Referendums Around the World.” Journal of Theoretical Politics 14(4): 465–515. IDEA. 2014. “Institute for Democracy and Electoral Assistance. Direct Democracy Database.” Online: https://www.idea.int/data-tools/data/direct-­ democracy. Accessed: January 2014. Inoguchi, Takashi. 2008. “Parliamentary Opposition under (Post-)One-Party Rule: Japan.” The Journal of Legislative Studies 14(1/2): 113–132. Kaiser, André. 2008. “Parliamentary Opposition in Westminster Democracies: Britain, Canada, Australia and New Zealand.” The Journal of Legislative Studies 14(1/2): 20–45. Kopecky, Petr and Maria Spirova. 2008. “Parliamentary Opposition in Post-Communist Democracies: Power of the Powerless.” The Journal of Legislative Studies 14(1/2): 133–159. Krehbiel, Keith. 1991. Information and Legislative Organization. Ann Arbor: The University of Michigan Press. Laver, Michael and Kenneth A. Shepsle. 1994. Cabinet Government in Theoretical Perspective. In Cabinet Ministers and Parliamentary Government, eds. Michael Laver and Kenneth A. Shepsle. Cambridge: Cambridge University Press: 285–310. Mainwaring, Scott and Matthew S. Shugart. 1997. “Juan Linz, Presidentialism, and Democracy: A Critical Approach.” Comparative Politics 29(4): 449–471. Marshall, Monty G., Keith Jaggers and Ted Robert Gurr. 2014. “Polity IV Project. Political Regime Characteristics and Transitions, 1800–2013.” Online: http://www.systemicpeace.org/polity/polity4.htm. Accessed: October 2015. Mattson, Ingvar. 1995. Private Members Initiatives and Amendments. In Parliaments and Majority Rule in Western Europe, ed. Herbert Döring. New York: St. Martins: 448–487. Mattson, Ingvar and Kaare Strøm. 1995. Parliamentary Committees. In Parliaments and Majority Rule in Western Europe, ed. Herbert Döring. New York: St. Martins: 249–306. Morgenstern, Scott, Juan Javier Negri and Anibal Pérez-Liñán. 2008. “Parliamentary Opposition in Non-Parliamentary Regimes: Latin America.” The Journal of Legislative Studies 14(1/2): 160–189. Mujica, Alejandro and Ignacio Sanchez-Cuenca. 2006. “Consensus and Parliamentary Opposition: The Case of Spain.” Government and Opposition 41(1): 86–108. Regan, Patrick and Dave Clark. 2014. “The Institutions and Elections Project.” Online: https://havardhegre.net/iaep/ Accessed: January 2014.

Measuring Opposition Power  63 Schnapp, Kai-Uwe and Philipp Harfst. 2005. “Parlamentarische Informationsund Kontrollressourcen in 22 westlichen Demokratien.” Zeitschrift für Parlamentsfragen 36(2): 348–370. Schrire, Robert A. 2008. “Parliamentary Opposition after Apartheid: South Africa.” The Journal of Legislative Studies 14(1–2): 190–211. Sieberer, Ulrich. 2011. “The Institutional Power of Western European Parliaments: A Multidimensional Analysis.” West European Politics 34(4): 731–754. Strøm, Kaare. 1990. Minority Government and Majority Rule. New York: Cambridge University Press. Taylor, Andrew J. 2006. “Size, Power, and Electoral Systems: Exogenous Determinants of Legislative Procedural Choice.” Legislative Studies Quarterly 31(3): 323–345. Tsebelis, George. 1995. “Decision Making in Political Systems: Veto Players in Presidentialism, Parliamentarism, Multicameralism and Multipartyism.” British Journal of Political Science 25(3): 289–325. Tsebelis, George and Bjørn Erik Rasch. 2011. Introduction. In The Role of Governments in Legislative Agenda Setting, eds. George Tsebelis and Bjørn Erik Rasch. New York: Routledge: 1–20. Wegmann, Simone. 2022. “Policy-Making Power of Opposition Players: A Comparative Institutional Perspective.” The Journal of Legislative Studies 28(1): 1–25.

Part II

Policy-Making Power of Opposition Players and Its Consequences for Democratic Consolidation

Chapter 4

Methods and Data

In this chapter, I present methods and data to test the theoretical argument presented in Chapter 2. The analyses presented in this second part of the book aim to shed light on democratic consolidation by bringing the crucial role of legislatures into focus. More specifically, the results answer the question of the extent to which opposition power in policy-­ making influences democratic consolidation. This chapter details the empirical strategy and the data used to answer this question. The goal of the chapter is to provide all necessary information for the empirical analysis that is presented in Chapters 5 and 6. The chapter proceeds as follows. In the first section, I present the overall empirical strategy. The second section briefly discusses the operationalization of the independent variable of policy-making power of opposition players1 and presents the data used to categorize winners and losers of elections. The third section presents the operationalization of the dependent variable of democratic consolidation as measured by political support. The fourth section presents all the control variables considered in the analyses. In the fifth section, I discuss the criteria for sample selection and present the final sample of countries considered in the analyses. Section six presents a summary of the empirical strategy and data included in the analyses. Detailed overviews of all the datasets and the respective variables considered in the analyses are presented in Tables A.4.1 and A.4.2 in the Appendix.

4.1  Empirical Strategy The research strategy presented in this chapter aims to answer the question of whether or not inclusive parliamentary procedures contribute to democratic consolidation. Chapters 1 and 2 have shown the importance of differentiating between winners and losers of democratic elections and the influence different levels of policy-making power of opposition players might have on the political support expressed by these two groups. Hence, the analytical chapters of Part II of this book aim to test whether DOI: 10.4324/9781003005360-6

68  Power of Opposition Players and Consequences

or not specific parliamentary procedures (i.e. strong policy-making power of opposition players) are linked to more political support expressed by winners and/or losers – thereby contributing to democratic consolidation. I organize the chapters in Part II (Chapters 5 and 6) accordingly, including one descriptive chapter and one analytical chapter. However, analyzing the specific effect of institutions on behavior is difficult due to their endogenous character. Institutions are “(…) neither neutral reflections of exogenous environmental forces nor neutral arenas for the performance of individuals driven by exogenous preferences and expectations” (March and Olsen 1984: 742). Instead, the challenge is “(…) to distinguish effects of institutions from those of the conditions that give rise to them” (Przeworski 2004: 527). Therefore, in Chapter 5 I present an overview of the power granted to opposition players around the world and also address the question of the possible determinants of policy-making power of opposition players in more detail. More specifically, in Chapter 5 I present the results of the data collection effort to measure the policy-making power of opposition players according to the operationalization presented in Chapter 3. To analyze the effect of inclusive parliamentary procedures (i.e. strong policy-­making power of opposition players), a comprehensive dataset of legislative organization from the perspective of opposition players is necessary. Chapter 5 makes two important contributions to existing research. First, it shows that the policy-making power of opposition players varies considerably between countries. This finding contributes to the rich literature in legislative studies pointing to the different forms of parliamentary rules. Second, the results of this data collection effort show that bringing studies of legislative organization into research on democratic­ ­consolidation – and thereby looking beyond the central question of government ­formation – is a promising approach. If the descriptive statistics showed little variance in the policy-making power of opposition players, then the legislative organization of the policy-making process could not be a key factor in determining democratic consolidation. However, as these descriptive statistics show considerable variance, looking more closely at how legislatures are organized and how they include oppositions in the policy-making process may be a promising approach to explain the different levels of democratic consolidation. In Chapter 5, I include both general descriptive results on the degree of policy-making power of opposition players around the globe, and several more in-depth analyses of average and exceptional cases. Furthermore, I address the question of endogeneity and its consequences for conclusions and implications from the results. In Chapter 6, I present a number of regression models that test the theoretical expectations described in Chapter 2. I utilize multi-level

Methods and Data  69

regression models, accounting for the particular structure of the data of individuals nested in different countries. Here, the combined measure of the policy-making power of opposition players presented in Chapter 3 serves as an explanatory variable.2 As the presented expectations refer to the influence of particular national institutions on individual behavior, the unit of analysis is the individual citizen of a country. I opt for models which only include the main effect of losers and interaction effects of losers and policy-making power of opposition players. Hence, interaction effects show for both, losers and winners, whether policy-making power of opposition players has an effect on political support. However, to make interpretation of the results more accessible, all results will be discussed based on predicted probabilities. To make it possible to draw conclusions from the results of a possible effect of policy-making power of opposition players on individual political support, some questions arise regarding which countries and survey waves to include in the sample. I include a sample as large as possible to present a broad comparative analysis. However, for identifying a possible causal link, the question of sequence is important. Therefore, the countries under investigation are all included in the sample3 that fulfil the following three conditions. First, measurement of the policy-making power of opposition players was possible for all seven variables of the concept (i.e. complete information on the level of policy-making power of opposition players is available). Second, survey data on individual political support is available for each relevant country at a point in time after their standing orders and constitutions, considered for coding policy-­making power of opposition players, came into force. Third, if more than one survey was available between two elections, only the one that immediately followed an election was considered. Even though I rely on survey data to measure political support and therefore cannot observe a causal effect directly, this strategy ensures that the rules determining the level of policy-making power of opposition players came into force before the surveys that measured individual political support were conducted. Different surveys provide the necessary individual-level data: the analyses will primarily focus on the Comparative Study of Electoral Systems CSES (CSES 2019) and the European Social Survey (ESS 2019). The Latin American Public Opinion Project LAPOP (Vanderbilt University 2019) also provides questions to measure democratic consolidation, but only provides a different question to identify winners and losers of elections. As the following sections on the operationalization of all variables show, the surveys provide slightly different questions making it impossible to merge all of them. Hence, in Chapter 6, I present separate analyses for the different surveys. At the same time, this strategy of considering different surveys may serve as a further robustness check.

70  Power of Opposition Players and Consequences

4.2  I ndependent Variable: Policy-Making Power of Opposition Players, Winners and Losers To test whether or not more inclusive parliamentary procedures relate to individual political support, the measure of the policy-making power of opposition players is considered as independent variable. As discussed in Chapter 3, the overall index is composed of seven variables4: bill introduction, agenda setting, amendments, committee structure, committee procedures, executive power, and referendums. Attributing equal weight to each variable in the overall index leads to a measure of the policy-­ making power of opposition players ranging from 0 (weak policy-making power of opposition players) to 1 (strong policy-making power of opposition players). To test the effects of the different levels of policy-making power of opposition players among winners and losers of democratic elections, these two groups need to be identified. As discussed in Chapter 3, the winners of democratic elections are defined as the group of individuals that voted for a party that constitutes the government after an election. This may be a party, or multiple parties in the case of a coalition government. Hence, to determine whether an individual is part of the group of winners or losers, two pieces of information are necessary. First, information about which political party/parties constitute the government and second, which party an individual voted for in the election. Combining this information allows for the identification of individuals within the electorate as either winners or losers of democratic elections. The Comparative Study of Electoral Systems (CSES 2019) provides all of the necessary information to identify winners and losers, respectively. Information is available for the respondent’s vote choice for presidential candidates as well as party lists and district candidates. Furthermore, information is available for the number of government portfolios after the election for each political party competing in the election. This variable allows for the identification of government and opposition parties. Thus, respondents who voted for a party with at least one government portfolio after the election are coded as winners and individuals that voted for a party with no government portfolio after the election are coded as losers. Furthermore, the ESS (2019) also provides a question about the party a respondent voted for in the last national election. This information is combined with information about cabinet parties in EU and OECD countries provided by Döring and Manow (2018). Respondents that voted for a party that is a member of cabinet after the election are coded as winners and respondents that voted for a party without cabinet membership are coded as losers. Finally, the question in the LAPOP survey (Vanderbilt University 2019) is slightly different. The survey does not include a question about the vote

Methods and Data  71

in the last election. Instead, the survey includes a question about the vote intention in the next election. Respondents are asked to report whether they would vote for the incumbent candidate/party or for a candidate/ party different from the current administration. Compared to the questions of the ESS (2019) and the CSES (2019), the question of the LAPOP (Vanderbilt University 2019) does not directly allow for the identification of winners and losers. Nevertheless, individuals are asked to report whether they support the current government or would prefer a different one which translates into identifying as winner or loser at the moment the survey was conducted. For the analyses presented in this book, individuals who would not vote for the presidential party/government are coded as losers, and those who would vote for the current presidential party/government are coded as winners. However, the difference in the framing of the question must be considered when interpreting the results in Chapter 6.

4.3  D ependent Variable: Consolidation – Political Support As outlined in Chapters 1 and 2, I look at democratic consolidation from an individual-based perspective. This operationalization reflects definitions of democratic consolidation that have been presented in the literature on democratic consolidation (see, e.g. Diamond 1999). Accordingly, democratic consolidation refers to the solidification of democratic rules as “the only game in town” (Linz and Stepan 1996; Linz 1990: 5). Therefore, “consolidation is the process by which democracy becomes so broadly and profoundly legitimate among its citizens that it is very unlikely to break down” (Diamond 1994: 15). Consequently, democratic consolidation is inherently about political support and measured accordingly in Chapter 6. Among the early and most prominent contributions to the literature on how to properly define and measure political system support is Easton’s (1975; 1965a, 1965b; 1957) distinction between diffuse support and specific support of state performance. It is also among the most influential typologies of political support in the current literature. The difference between specific and diffuse support lies in the difference between satisfaction with the specific outputs and performance of a regime and with the basic principles upon which the regime is based (Easton 1975: 436–437; but see also Easton 1965a, 1965b, 1957). Specific support is the result of an individual’s evaluation according to the extent to which its demands are perceived to have been met. In the case of a democracy, it refers to the satisfaction with the outcome that a democratic system generates. On the contrary, diffuse support expresses itself through the conviction that a political object conforms “(…) to his own moral principles, his own sense

72  Power of Opposition Players and Consequences

of what is right and proper in the political sphere” (Easton 1965b: 278). In the case of a democracy, it relates to the support of the democratic institutions as a good/best form of government independent of the level of satisfaction with its outcome. Therefore, the difference between diffuse and specific support for a democracy lies in the difference between an individual’s satisfaction with the performance of a regime and satisfaction with the basic principles upon which a political regime is based (Easton 1975: 436–437). Linde and Ekman (2003: 395) have emphasized that a consolidated democracy requires citizens who support their democratic system. Ideally, this involves support for the regime principles (i.e., support for democracy, as a principle or an ideal, as the most appropriate form of government), as well as support for the performance of the regime (i.e., support for how it functions in practice). Accordingly, the variables considered in Chapter 6 to measure democratic consolidation include both diffuse and specific support. However, the literature has repeatedly emphasized the potential problems which occur when measuring political support using survey questions, and especially when relying on a single question to measure the concept. Norris (2011, see also Norris 1999), for example, has introduced a more fine-grained conceptualization of political support based on Easton’s distinction between the nation, the regime, and the authorities as three components of the political system (Easton 1965a). Norris (2011: 24) proposes five levels of support on a continuum from most diffuse to most specific: national identities (e.g. feelings of patriotism), approval of core regime principles (e.g. support for democratic ideals), evaluation of regime performance (e.g. satisfaction with democratic performance), confidence in regime institutions, and approval of incumbent office holders. On the one hand, scholars have especially criticized the frequently used question examining the level of “satisfaction with democracy” as a measure of diffuse support (see, e.g. Norris 2011; Linde and Ekman 2003; Canache, Mondak and Seligson 2001). Norris’ (2011: 24) conceptualization of political support illustrates this problem, placing the evaluation of regime performance (i.e. “satisfaction with democracy”) on the intermediate level between diffuse and specific support. Nonetheless, “satisfaction with democracy” is frequently used as an indicator of diffuse support. However, as several scholars have emphasized repeatedly, “the problem with the ‘satisfaction with democracy’ item is that it is not particularly useful if we are interested in monitoring the levels of democratic legitimacy (support for democracy as a principle)” (Linde and Ekman 2003: 396). The authors further emphasize that the performance of a democratic regime most likely also influences the legitimacy of the

Methods and Data  73

regime (Linde and Ekman 2003: 397). Therefore, results based on the question of “satisfaction with democracy” should actually lead to very similar results to a question investigating the performance of, for example, the government (Linde and Ekman 2003: 397). This view categorizes the question of “satisfaction with democracy” to measure specific support. Clarke, Dutt and Kronberg (1993: 1003) show evidence that satisfaction with democracy is an “(…) overall summary measure of satisfaction with existing democratic political systems,” again pointing to the question lying at the intersection of diffuse and specific support. On the other hand, however, scholars have argued that “[t]he indicator ‘satisfaction with democracy’ measures a felt discrepancy between democratic norms and the actual democratic process” (Thomassen 1995: 383). Similarly, Canache, Mondak and Seligson (2001: 511) mention that survey respondents are required to “contrast the actual nature of democracy with some standard of performance” when asked about their satisfaction with the way democracy works. The problem, however, is that the question does not state precisely any basis of comparison. Hence, some respondents might interpret the question to be about specific support (e.g. for the current government) whereas others might interpret the question to be about diffuse support (e.g. about the possible ideal of a democracy under different institutional settings).5 But, this view also relates to Norris’ (2011: 24) conceptualization of political support placing “satisfaction with democracy” on the intersection between diffuse and specific support. Furthermore, several scholars have highlighted the important difference between political support in older and more newly established democracies (or even regimes in transition). Mishler and Rose (2001: 303), for example, mention that “an idealist approach [which is measuring adherence to democratic principles] cannot capture the realities facing citizens in new democracies (…).” This is the case as citizens in new democracies have little knowledge of democratic principles, whereas citizens in older democracies are often more critical toward the government or the quality of democracy more generally (Mishler and Rose 2001: 304). Therefore, the authors suggest measuring political support by asking citizens to assess the current regime against the performance of the past one (see also Mishler and Rose 1996). Unfortunately, this strategy is not possible based on the available survey questions.6 However, I will control for the number of years of democratic experience in the analyses (see Section 4.4). Hence, even though frequently used in the literature, considering support for the performance of the regime (i.e. satisfaction with democracy) alone does not capture the whole concept of political support. I share the idea of a multidimensional approach to measure political support proposed by several authors (e.g. Norris 2011; Linde and Ekman 2003).

74  Power of Opposition Players and Consequences

Therefore, the analyses in Chapter 6 present three different dependent variables covering three levels on the continuum between diffuse and specific support according to Norris (2011).7 The question of whether democracy is preferable relates to the approval of regime principles and is linked to diffuse support. Satisfaction with government performance relates to the approval of incumbent officeholders and lies at the other end of the continuum of political support (i.e. specific support). The question of satisfaction with democracy relates to the evaluation of regime performance lying in between specific and diffuse support. The LAPOP (Vanderbilt University 2019) is the only one among the three surveys asking a question about democracy as best form of government. Respondents are asked to choose from the statements ‘We need a strong leader who does not have to be elected,’ or ‘electoral democracy is the best.’ All individual who responded ‘electoral democracy is the best’ are given a score of 1 indicating diffuse political support (i.e. support of the principle of democracy). All three surveys include a question about the level of satisfaction with democracy. However, the framing and especially the possible answer categories for this question differ across the surveys. The CSES (CSES 2019) and the LAPOP (Vanderbilt University 2019) provide a four-point scale ranging from ‘very satisfied’ to ‘very dissatisfied’/‘not at all satisfied.’ The variable considered in the analyses scores 1 for all individuals responding ‘very satisfied’ or ‘fairly satisfied’ and 0 otherwise. The ESS (ESS 2019) provides an eleven-point scale ranging from 0 (‘extremely dissatisfied’) to 10 (‘extremely satisfied’). For comparability, this variable is recoded scoring 1 for all individuals with a score of 7 (i.e. higher than the median in the sample) or more on the original scale and 0 otherwise. The original variable is used for further robustness checks. The CSES (2019) and the ESS (2019) both include a question about the satisfaction with the government performance. Again, the CSES (2019) provides a four-point scale ranging from a ‘very good job’ to a ‘very bad job.’ The variable considered in the analyses scores 1 for all individuals giving a positive answer (‘very good’ or ‘good job’), and 0 otherwise. The ESS (2019) again provides an eleven-point scale ranging from 0 (‘extremely dissatisfied’) to 10 (‘extremely satisfied’). This variable is recoded scoring 1 for all individuals with a score of 6 (i.e. higher than the median in the sample) or more on the original scale and 0 otherwise. The original variable is used for further robustness checks. Table 4.1 shows correlations between the different measures of political support available in the three surveys. Unfortunately, none of the surveys provide all three measures of political support (i.e. democracy is preferable, evaluation of government performance, and satisfaction with democracy). Nevertheless, the correlations in the following table still shed some light on whether the question of “satisfaction with democracy”

Methods and Data  75 Table 4.1  C  orrelations – Variables of Political Support

Democracy is preferable Government performance

Survey

Satisfaction with democracy

LAPOP CSES ESS

0.07 0.27 0.65

Sources: CSES: Comparative Study of Electoral Systems (2019); ESS: European Social Survey (2019); L APOP: Latin American Public Opinion Project (2019).

relates more to diffuse or specific support, respectively. Table 4.1 shows the strongest correlation (0.65) between the item for ‘satisfaction with democracy’ and that for ‘satisfaction with government performance’ (specific support) among respondents of the ESS (2019). In the CSES (2019), ‘satisfaction with democracy’ and ‘satisfaction with government performance’ (specific support) show a weaker correlation (0.27). However, in the LAPOP (Vanderbilt University 2019) there is a very weak correlation (0.07) between the ‘satisfaction with democracy’ question and diffuse support (i.e. democracy is preferable). Based on these results, ‘satisfaction with democracy’ does not seem to capture the concept of diffuse support very well. Nevertheless, this is only a vague indication that ‘satisfaction with democracy’ is more closely related to specific support than to diffuse support in the surveys considered in Chapter 6. The large difference in the correlations between ‘satisfaction with democracy’ and ‘satisfaction with government performance’ in the CSES (2019) and the ESS (2019) also points to differences between the surveys. Taking into account the average age of the democracies in these surveys, however shows that the average individual in the CSES (2019) lives in a democracy with 39.9 years of democratic experience, whereas the average respondent in the ESS (2019) lives in a democracy with 52.7 years of democratic experience. Therefore, it may be the case that individuals in more established democracies relate the question about ‘satisfaction with democracy’ more closely to specific support.8

4.4  Control Variables The theoretical argument presented in Chapter 2 states that there is a relationship between policy-making power of opposition players and political support. The direction of this influence is contingent upon individuals being winners or losers of democratic elections. However, characteristics besides being a winner or loser and living in a polity which grants weak or strong policy-making power to opposition players may influence the level of political support. Therefore, the analyses presented in Chapter 6

76  Power of Opposition Players and Consequences

include a number of control variables both on the individual level as well as the country level. On the country level, the analyses control for decentralization of a polity, the regime type, the electoral system, GDP/capita, the level of corruption, and democratic experience. On the individual level, control variables include an individuals’ background (education, economic situation) as well as their opinions (interest in politics, identification with a political party, satisfaction with life), which I discuss in the following paragraphs. Tables A.4.1 and A.4.2 of the Appendix show all variables including control variables, corresponding data sources, and recoding for the analyses in more detail. 4.4.1 Decentralization Research on losers’ consent has confirmed the finding that losers show less satisfaction with democracy but has also started to emphasize that different institutional settings affect these political opinions. For example, not only winning national elections but even winning regional elections contributes to a higher level of satisfaction with democracy beyond the regional level (Singh, Karakoç and Blais 2012; Blais and Gélineau 2007). Similarly, a bicameral organization of the legislature may contribute to more satisfaction with democracy (Henderson 2008: 13).9 Past research has shown both positive and negative effects of federalism and bicameralism on political opinions about democracy (see, e.g. Bernauer and Vatter 2012; Anderson et al. 2005). Yet, such decentralization of a polity may also lead to more policy-making power for opposition players since national governments may be constrained in their actions. Among the most prominent institutions that may constrain government actions are bicameralism and federalism. The measure of the policy-making power of opposition players included in Chapter 6 considers first chambers only. But, not all countries in the sample are unicameral. Therefore, controlling for decentralization, including bicameralism and federalism, is of crucial importance in being able to capture the influence of the policy-making power of opposition players. As this variable may not only influence individual political support but also add to the power of opposition players to a considerable extent, I briefly discuss possible types of bicameralism and federalism as well as implications for opposition players in the following two paragraphs. 4.4.1.1 Bicameralism Bicameralism is essentially about power-sharing relationships between the two legislative chambers (Uhr 2006: 474–475). Since bicameralism

Methods and Data  77

allows giving special representation to minorities and smaller sub-­ national entities in federal states, a strong bicameral system with two equally powerful chambers is part of Lijphart’s (1999) concept of consensus democracy (Lijphart 1999: 39). However, to allow for such different representation, the two chambers in a bicameral system10 need to be elected differently and the upper house must have real power (see, for example, Anderson et al. 2005: 127; Lijphart 1999: 39). Similarly, Tsebelis and Money (1997: 44) emphasize two central characteristics that distinguish the different types of bicameral systems. First, the two chambers in a bicameral system may represent different categories of citizens and, second, the power granted to the two chambers varies across different bicameral systems. As Lijphart (1999: 206) has emphasized, when members are not selected based on direct elections, second chambers only enjoy reduced legitimacy. Moreover, different electoral rules for selecting members of the lower and upper chambers allow for the over-representation of certain groups in the upper chamber (Lijphart 1999: 207). In federal systems, for example, subnational entities may be represented in the upper house. Such subnational entities are granted equal representation as some smaller entities are weighted more heavily than larger entities. Furthermore, some countries grant special representation to ethnic or linguistic minorities (Tsebelis and Money 1997: 53). Such a bicameral system applying different electoral rules to over-represent certain groups in the upper house is defined as incongruent (Lijphart 1999: 207). Second, direct elections of second chambers and over-representation of certain groups leads to potential influence of opposition parties only when second chambers have at least some power in the political system. Consequently, Lijphart (1999: 205) and Tsebelis and Money (1997: 54) highlight the crucial role that the respective power of second chambers plays as compared to first chambers. A subordinated second chamber cannot exert as much influence in the legislative process as a second chamber with equal power. Bicameral systems are symmetrical where both chambers have equal or only slightly unequal power and are popularly elected (Lijphart 1999: 206). Accordingly, strong bicameralism is a symmetrical and incongruent system, medium-strength bicameralism is either symmetrical and congruent or asymmetrical and incongruent, and weak bicameralism is both asymmetric and congruent (Lijphart 1999: 211). 4.4.1.2 Federalism A federal system divides power between levels of government. Accordingly, a federal system is one in which both regional governments and a central government make final decisions on some activities and,

78  Power of Opposition Players and Consequences

therefore, divide government activities between these two levels (Riker 1975: 101). Hence, federalism “can be considered the most typical and drastic method of dividing power (…)” (Lijphart 1999: 185). Similar to bicameralism, federalism is a characteristic that Lijphart (1999) attributes to the concept of consensus democracy. According to these definitions, federalism has two central features. First, the existence of both central and regional governments and second, the division of power between these two units. As Beramendi (2007: 753) has emphasized, “[t]he existence of several levels of government is a necessary, yet insufficient, condition for federalism to exist, (…).” What characterizes federal systems is the fact that “(…) both constituent units and the central government have constitutionally recognized autonomous powers to interact directly with citizens” (Beramendi 2007: 754). Hence, the federal structure of national and subnational levels of government must be anchored in the constitution (see also Galligan 2006: 268). Controlling for federalist structures is crucial as losers on the national level may be winners on a subnational level, and vice versa, which may influence the effect of policy-making power of opposition players on individual opinions. Furthermore, several authors mention bicameralism as a central feature of federalism (see, for example, Beramendi 2007 or Watts 1996). Bicameralism is important because second chambers allow for the representation of subnational units at the national level. This is why the analyses in Chapter 6 consider a control variable combining a measure of bicameralism and federalism. 4.4.1.3  The Measure of Decentralization Data to measure decentralization come from Gerring et al. (2005).11 The variable “Average of Nonfederalism and Nonbicameralism” allows for a fine-grained classification as the variable categorizes countries on a three-point federal and a three-point bicameral scale, respectively. The variable “Average of Nonfederalism and Nonbicameralism” classifies countries on a scale ranging from 0 (federalist countries with strong bicameralism) to 2 (unitary countries). Unfortunately, the authors do not provide the separate variables of nonfederalism and nonbicameralism but only the combined average variable.12 Nevertheless, it seems important to not only control for the existence of federalism and/or bicameralism but to also account for the different degrees as discussed in the two previous sections. Thus, decentralization is the combined measure of bicameralism and federalism. For the analyses, this variable is recoded into a binary variable with a score of 1 for all countries scoring 0 or 0.5 (indicating decentralization) on the original unitarism-variable and a score of 0 otherwise.13

Methods and Data  79

4.4.2  Regime Type Several scholars have analyzed the influence of particular democratic institutions on individual political support. Previous research has looked into the effect of parliamentary versus presidential regimes, consensual versus majoritarian systems, or the influence of direct democratic institutions (see, e.g., Bernauer and Vatter 2012; Wells and Krieckhaus 2006; Anderson et al. 2005; Rose, Mishler and Haerpfer 1998; Anderson and Guillory 1997). The main independent variable in this study (i.e. the policy-­making power of opposition players) looks at the possible influence of oppositions in the policy-making process adding a new aspect of democratic institutions to be considered when analyzing individual political support. To control for the fact that a more general characterization of regime types – and not necessarily the level of power granted to opposition players – may cause differences in political support, the variable regime type is added as country-level control variable. Regime type classifies countries as either parliamentary, semi-presidential, or presidential regimes.

4.4.3  Electoral System Another institutional aspect shaping the organization of elections and thereby influencing winners and losers is the electoral system itself. Anderson et al. (2005), for example, have shown that less proportional electoral systems reduce satisfaction with democracy. Furthermore, research has highlighted that individuals react to an imbalance between the seat share and the vote share. Accordingly, under-representation due to translation from vote into seat share leads to less satisfaction with democracy (Blais, Morin-Chassé and Singh 2017: 4; Berggren et al. 2004). Independent of the electoral formula, elections leading to single-party governments cause winners to be more satisfied if they won by a close margin, whereas losers are not affected by the margin of the election result (Howell and Justwan 2013: 340). Using experiments, Bowler and Donovan (2007) have shown that winners and losers react differently to reform proposals of electoral systems. For example, losers were largely supportive of proportional electoral formulas, whereas the majority of winners were opposed (Bowler and Donovan 2007: 464). As Singh (2014) was able to show, strategic voting of winners also influences their level of satisfaction. A possible link between the electoral formula and the occurrence of not voting for one’s first choice may then help explain the link between electoral systems and satisfaction with democracy. Results show that having voted for one’s first choice has a less pronounced effect on satisfaction in non-­majoritarian electoral systems than in majoritarian ones (Singh 2014: 311, 319). Similarly, in a multi-ethnic setting, proportional electoral systems also

80  Power of Opposition Players and Consequences

contribute to more satisfaction with democracy for ethnic minorities whose political party is small or is not represented at all. On the contrary, in the case of large ethnic parties, more majoritarian electoral systems increase satisfaction with democracy (Ruiz-Rufino 2013: 114). Since the extant literature has repeatedly shown that the type of electoral system influences the degree of political support, the analyses include the variable electoral system as an additional country-level control variable. Data on electoral systems come from Bormann and Golder (2013).14 Based on the basic type of electoral system used in elections, the original variable in the dataset classifies countries into majoritarian, proportional, or mixed. The variable is recoded into a binary variable scoring 1 for proportional electoral systems and 0 otherwise. 4.4.4  Political System In some of the surveys used for the analyses in Chapter 6, policy-­making power of opposition players, regime type, electoral system, and/or decentralization are strongly correlated. This is the case for the sample used to analyze satisfaction with democracy in the LAPOP survey (Vanderbilt University 2019). Here the regime type and policy-making power of opposition parties are correlated at 0.61. A closer look at the regime type shows that most countries are presidential regimes, whereas only two out of the ten countries in the sample are parliamentary and none is semi-presidential. Furthermore, only the two parliamentary regimes have majoritarian electoral formulas whereas all remaining countries use proportional or mixed systems. Due to this distribution of the two variables and the correlation of regime type and policy-making power of opposition players, regime type and electoral system are combined into an index of political system. This index takes a value of 1 for all parliamentary systems or for countries with a non-PR electoral formula, and 0 otherwise. This coding allows for the identification of countries deviating from the predominant institutions of presidentialism and PR electoral formula in the region. A model including only decentralization and PR electoral formula as control variables is estimated as a further robustness check. Furthermore, countries included in the CSES survey to analyze satisfaction with the government show a high correlation between decentralization and regime type (0.78). Among the countries in the sample only one is categorized as presidential and decentralized and only two countries use non-PR electoral formulas. Due to this distribution of the variables, the analyses presented in Chapter 6 include a combined index of political system. The index combines PR electoral formula, parliamentary or semi-presidential regime, and decentralization (strong federalism and bicameralism). Further robustness checks look at analyses including only

Methods and Data  81

PR electoral formula and regime type as control variables and at models excluding Switzerland (as only presidential and strongly decentralized regime and also oldest democracy in the sample). 4.4.5 GDP/Capita Several scholars have also emphasized the connection between individual behavior and the overall economic situation in a given country. Individuals in a prospering economy are more likely to be satisfied with how democracy works as well as with the government performance (see, e.g. Bernauer and Vatter 2012; Henderson 2008; Rose and Mishler 2002; Anderson and Guillory 1997). Hence, GDP/capita in the year of the election is considered in the analyses to account for this probable influence. 4.4.6 Corruption An individual’s evaluation of the political system may not only be influenced by the overall economic situation, but also by the level of corruption. However, previous research has not shown clear results for whether or not corruption significantly influences the level of political support (see, e.g. Singh 2014; Wells and Krieckhaus 2006; Rose and Mishler 2002). Nevertheless, corruption is added as a country-level control variable. Data on corruption comes from Coppedge et al. (2017).15 The variable is a combined index of corruption in four different government spheres, giving equal weight to all of them: (i) public sector corruption; (ii) executive corruption; (iii) legislative corruption; and (iv) judicial corruption (Dahlberg et al. 2019: 603). 4.4.7  Democratic Experience Research on individual opinions and political support has emphasized the importance of differentiating between new and established democracies (Conroy-Krutz and Kerr 2015; Curini, Lou and Memoli 2012; Anderson et al. 2005; Mishler and Rose 2001; Rose, Mishler and Haerpfer 1998). Accordingly, unlike citizens in established democracies, citizens in young democracies lack the experience to differentiate between the government and the regime when expressing their support of democracy (or lack thereof) (Mishler and Rose 2001: 557). Also, citizens in newer democracies express lower levels of satisfaction with democracy than citizens in established democracies (Aarts and Thomassen 2008: 15). Svolik (2013) has further emphasized that demands and expectations in new democracies are higher than in established democracies which also affects the level of satisfaction with democracy of citizens in these former regimes (see also Aarts and Thomassen 2008). In addition, Rose and

82  Power of Opposition Players and Consequences

Mishler (2002) have also pointed to the differences in regime support of citizens between democracies and non-democracies. Socialization in a non-democratic regime also influences the support for democracy once the regime has transformed into a democratic regime (Voicu and Peral 2013). In this relation, several authors have analyzed political values and support in former communist countries (see, e.g. Knutsen and Wegmann 2016; Pop-Eleches and Tucker 2014, 2012, 2011; Mishler and Rose 1996; Tóka 1995). Furthermore, not only might institutions affect the consolidation of democracies but also vice versa. Strong policy-making power of opposition players could only develop over time and therefore might be a mere effect of stable democracies. If this is the case, one should observe weak policy-making power of opposition players among all young democracies and stronger power among all older ones. However, as Wegmann (2022) has shown, democratic experience is not among the determinants of strong policy-making power of opposition players. Results do not show evidence that strong policy-making power of opposition players is a consequence of an institutional development occurring only over time in long-living democracies. Although rather weak policy-making power of opposition players exists among very young democracies, likewise does it in a large number of consolidated ones. Nevertheless, based on the extant literature, it is likely that the policy-­ making power of opposition players affects public opinion differently in older and newer democracies, respectively. To control for these possible effects of regimes with different levels of experience with democracy, a variable measuring democratic experience in years is added as interaction with policy-making power of opposition players in the analyses of Chapter 6. 4.4.8  Individual Background The level of political support expressed by citizens may be influenced not only by country-level features but also by several individual-level characteristics. Hence, the analyses presented in Chapter 6 include several individual-level control variables. Unfortunately, not all surveys provide the same variables and, therefore, not all variables can be included in all of the analyses. Table A.4.2 of the Appendix details all variables, data availability, and recoding decisions. The first set of variables looks at an individual’s background. The analyses therefore consider variables on the respondent’s education and the economic situation in their household as control variables. Previous research has repeatedly found a positive effect of the level of individual education on the level of political support (see, e.g. Bernauer and Vatter 2012; Hobolt 2012; Anderson et al. 2005; Anderson and

Methods and Data  83

Guillory 1997). Individuals with more education tend to be more satisfied with democracy. Two of the three surveys (ESS 2019 and CSES 2019) include variables on the level of the education of respondents. Low education is a re-coded binary variable attributing the score of 1 to all individuals with lower than secondary education. Similar to the influence of the overall economic situation on the assessment of the political situation, the personal economic situation may also influence an individual’s political support. Individuals living in less welloff households may express less political support (see, e.g. Bernauer and Vatter 2012; Henderson 2008; Anderson and Guillory 1997). Again, two of the three surveys provide variables about the respondent’s household income, however, the available categories differ slightly. Whereas the ESS (2019) assesses the deciles of the household’s total income, the CSES (2019) provides quintiles of a respondent’s household income. Therefore, for these two surveys household income is a recoded three-point scale measuring the respondent’s household income as either low, intermediate, or high, respectively. The LAPOP (Vanderbilt University 2019) does not provide a question about the household income in all countries, however, it does ask about the economic situation in the respondent’s household. The provided variable has the following four scores: “good enough and can save”, “good enough, with no major problems”, “not enough, and are stretched”, “not enough, and having a hard time”. The variable household income for this survey is a binary variable scoring 1 for all individuals reporting their household income to be good enough. 4.4.9  Individual Opinions and Interests A second set of variables focuses on an individual’s opinions and ­interests. Politically interested individuals on both sides – winners and losers – may be more affected by the influence of the policy-making power of opposition players than less interested respondents (see, e.g. Anderson and Guillory 1997). Therefore, the analyses add several variables accounting for different levels of political interest. Unfortunately, not all three surveys provide the same variables. The analyses based on the different surveys will therefore comprise a slightly different set of these control variables. Table A.4.2 of the Appendix details all variables, data availability, and recoding decisions. In two of the three surveys direct questions about the interest in politics are available. The ESS (2019) and the LAPOP (Vanderbilt University 2019) provide four-point scale variables based on the questions “how interested would you say you are in politics” (ESS 2019) and “how much interest do you have in politics?” (Vanderbilt University 2019). Both variables are re-coded into binary measures. The variable interest in politics therefore gives scores of 1 for more interested respondents (“very

84  Power of Opposition Players and Consequences

interested” and “quite interested” as well as “a lot” and “some”, respectively) and 0 otherwise. Other variables could be a less direct assessment of an individual’s interest in politics. The variable close to party asks whether an individual usually thinks of being close to any particular political party. All three surveys provide this information as binary variable (CSES 2019; ESS 2019; Vanderbilt University 2019). However, the questions are framed in slightly different ways. The ESS (2019) asks whether there is “(…) a particular political party you feel closer to than all the other parties?” The LAPOP (Vanderbilt University 2019) asks: “Do you currently identify with a political party?” Finally, the CSES (2019) asks whether “(…) you usually think of yourself as close to any particular party?” Both of these variables (interest in politics and close to party) may express, to some extent, the individual knowledge about the specific organization of the policy-making process in a polity. Individuals that are more interested in politics may also be more likely to follow the political debates and to have some knowledge about the political process. The same might be true for individuals that report to be close to a political party. These citizens may follow party activities more closely and therefore be more aware of parliamentary activities of party members. Finally, the ESS (ESS 2019) and the LAPOP (Vanderbilt University 2019) provide questions about satisfaction with life. Although not a measure of direct political interest, a higher level of overall satisfaction with one’s life might easily translate into higher satisfaction with the way democracy works or higher political support, in general (see, e.g. Anderson et al. 2005). Therefore, the variable life satisfaction controls for this possible effect. The questions in the two surveys ask respondents to indicate their overall life satisfaction on different scales. Whereas the LAPOP (Vanderbilt University 2019) asks to respond on a four-point scale, the ESS (ESS 2019) asks respondents to answer on a scale from 0 (extremely dissatisfied) to 10 (extremely satisfied). The variable considered for the analyses is re-coded attributing a score of 1 to all individuals with scores of 1 or 2 (i.e. very satisfied and somewhat satisfied) on the first scale and with a score of 6 or more on the latter.

4.5 Sample As I outlined in the theoretical argument presented in Chapter 2, the goal of this book is to analyze the influence of inclusive parliamentary procedures (i.e. policy-making power of opposition players) on democratic consolidation. The point of departure for the presented argument is the holding of democratic elections that generates winners and losers. Thus, the theoretical argument is set in a democratic environment and interested in the question of how democratic consolidation can be achieved

Methods and Data  85

once the transition of institutions from authoritarian to democratic has already taken place. Therefore, the population of cases consists of only those countries with at least some experience of democratic rule. In a first step of the analyses, Chapter 5 presents an overview of the level of policy-making power of opposition players around the world. When deciding whether or not to collect information on the policy-making power of opposition players, all countries with at least one year of democratic experience after 1990 as measured by Marshall, Jaggers and Gurr (2018) were considered. In line with the suggested classification of countries, I consider countries reaching a Polity score of 6 or more to fulfil the criterion of democratic experience (Marshall, Jaggers and Gurr 2018). Table 4.2 shows all countries of the sample based on this selection criterion. As Table 4.2 shows, a total of 110 countries meet the selection criteria of reaching a democracy score of 6 or more on the Polity IV scale in at least one year since 1990 (Marshall, Jaggers and Gurr 2018). These countries constitute the sample of cases included in the descriptive analyses presented in Table 4.2  S ample Selection: Chapter 5 Albania Argentina Armenia Australia Austria Bangladesh Belarus Belgium Benin Bolivia Botswana Brazil Bulgaria Burundi Canada Cape Verde Chile Colombia Comoros Costa Rica Croatia Cyprus Czech Republic Denmark Dominican Rep. East Timor Ecuador El Salvador

Estonia Fiji Finland France Gambia Georgia Germany Ghana Greece Guatemala Guinea-Bissau Guyana Haiti Honduras Hungary India Indonesia Ireland Israel Italy Jamaica Japan Kenya Kosovo Kyrgyzstan Lativa Lebanon

Lesotho Liberia Lithuania Luxembourg Macedonia Madagascar Malawi Malaysia Mali Mauritius Mexico Moldova Mongolia Montenegro Mozambique Namibia Nepal Netherlands New Zealand Nicaragua Niger Norway Pakistan Panama Paraguay Peru Philippines

Countries in italics: No of f icial documents are available. Source: Polity data (Marshall, Jaggers and Gurr 2018b).

Poland Portugal Romania Russia Senegal Serbia Sierra Leone Slovak Republic Slovenia Solomon Islands South Africa South Korea Spain Sri Lanka Sweden Switzerland Taiwan Thailand Trinidad and T. Tunisia Turkey Ukraine United Kingdom United States Uruguay Venezuela Zambia

86  Power of Opposition Players and Consequences

Chapter 5. However, for 32 of these countries no official rules of procedure could be retrieved. Therefore, no information on the policy-making power of opposition players could be collected for these countries. A vast majority of the selected countries are classified as democratic for all of the years considered. However, a closer look at developments over time shows some democratic breakdowns as well as some democratic transitions during the time period considered here. Based on Marshall, Jaggers and Gurr (2018) 20.7 percent of country-years (548 out of a total of 2,649 country-years) do not satisfy the threshold of 6 or more on the polity scale. A total of five countries experienced democratic breakdown between 1990 and 2014. Democratic breakdown here means a change from a score of 6 or more on the Polity IV democracy scale to a score of less than 6 on this scale. Based on Marshall, Jaggers, and Gurr (2018), Armenia, Belarus, and Gambia experienced a breakdown in the early 1990s, whereas Ecuador and Venezuela did so in the late 2000s. Additionally, the data include a number of countries that experienced a democratic transition (i.e. a development from a democracy score below 6 to one of 6 or more on the Polity IV democracy scale) between 1990 and 2014.16 A democratic transition from a non-democratic to a democratic regime occurred in a total of 26 countries. Such transitions happened in all regions of the world, most of them during the 1990s or early 2000s. Mozambique is the only country to have experienced a recent transition with a development from below 6 to 6 on the Polity IV democracy scale in 2014. However, countries do not only experience either a democratic transition or a breakdown: sometimes they experience both. In a total of 21 countries across the world both developments occurred during the period under study.17 But, the data show a slightly higher tendency for this pattern in Asian and African countries. Furthermore, a majority of the countries that were classified as non-democratic in 1990 experienced a democratic transition and a following democratic breakdown. In only six countries (the Dominican Republic, Haiti, Pakistan, Peru, the Solomon Islands, and Ukraine) a reversed development took place (i.e. democracy in 1990 followed by a breakdown and a new transition). A closer look at the data shows that even among the countries that were classified as a democracy (i.e. a score of 6 or more on the Polity IV democracy scale) during the entire time period considered, some developments toward being more or less democratic exist. A total of 17 countries meet the threshold of 6 on the Polity IV democracy scale to be classified as democratic, but do not receive the same score during the entire period between 1990 and 2014. Again, countries from all regions of the world fall into this pattern.18 This time, however, slightly more American and European countries seem to show some variation of their level of democracy. Among the 17 countries, a majority of 12 countries show a development toward higher levels of democracy. Three countries (Belgium, Bolivia, and Jamaica) show a decline in the level of democracy, whereas in the remaining two countries (Colombia and Turkey), results

Methods and Data  87

experienced changes in both directions. Colombia experienced a development toward higher levels of democracy in the early 1990s, followed by a decline in 1995. Turkey, however, shows a steady decline in the level of democracy until the late 1990s, followed by a stable period until 2010 and a new development toward higher levels of democracy in 2011. Thus, although a large number of countries in the sample are stable democracies during the entire time period, a closer look at the developments over time shows a considerable number of countries with democratic transitions and/or breakdowns as well as fluctuations in their levels of democracy. Table 4.3  S ample Selection: Chapter 6 Surveys Country

CSES

Argentina Australia Belgium

2013

Bolivia Brazil

ESS

2016 2008, 2010, 2012, 2014, 2016

2010, 2014

Chile Colombia

2017

Czech Republic Estonia Finland Germany Greece Guatemala Hungary Ireland Israel Jamaica Latvia Lithuania Mexico Panama Peru Poland Portugal Romania Slovakia Slovenia Spain Sweden Switzerland Trinidad and Tobago Turkey United Kingdom United States

2010, 2013 2011 2011, 2015 2013, 2017 2012, 2015 2016 2013 2014 2016 2012 2016 2011 2015 2012 2010 2011

L APOP

2010, 2014 2012 2008, 2012, 2016 2010, 2014 2010 2010, 2014 2012, 2016

2010, 2014, 2016 2008, 2010, 2012, 2014, 2016 2008, 2012, 2014 2008, 2012, 2014, 2016, 2018

2008, 2012

2012 2012, 2016

2008, 2012, 2016 2012

2014 2011

2010, 2012 2008, 2012, 2014 2008, 2012 2012 2008, 2012, 2016

2011, 2015 2015

2008 2012

Source: CSES (2019); ESS (2019); Vanderbilt University (2019).

2012, 2014 2014, 2016 2012, 2017

2010, 2012, 2014 2012

88  Power of Opposition Players and Consequences

In a second step of the analyses, Chapter 6 investigates the relationship between the policy-making power of opposition players and democratic consolidation. To this end, Table 4.3 shows an overview of all countries fulfilling the three necessary conditions for inclusion in the analyses. First, complete information on the level of the policy-making power of opposition players is available, and second, survey data on individual political support is available for each relevant country at a point in time after the standing orders and constitutions considered for coding came into force. Third, if more than one survey was available between elections, only those years that immediately followed an election are considered. Due to challenges with data availability, this second sample includes only American and European countries. Even though a considerable share of them are categorized as stable democracies, the sample also includes some of the less stable countries. For example, Argentina, Belgium, Bolivia, Chile, Guatemala, Israel, Jamaica, Mexico, Panama, Peru, Poland, Romania, Trinidad and Tobago, and Turkey show changes in the level of democracy after 1990.

4.6 Summary The goal of this chapter was to present all necessary information on case selection and operationalization of variables for the empirical Chapters 5 and 6. In Chapter 5, I present results of a comprehensive data collection effort to measure the policy-making power of opposition players. As results of this effort show a considerable variance in the level of the policy-­making power of opposition players, bringing the studies of legislative organization into research on democratic consolidation may be a promising approach. In Chapter 6, I present various multi-level regression models to analyze to what degree strong policy-making power of opposition players relates to the political support of winners and losers. The empirical chapters are based on slightly different samples. The effort to measure the policy-making power of opposition players (Chapter 5) considers all countries with at least one year of democratic experience since 1990 as measured by Polity (Marshall et al. 2018). The sample in Chapter 6 is narrowed by several additional conditions: (1) the availability of complete information on the policy-making power of opposition players; (2) the availability of survey data on individual political support for the relevant country at a point in time after the coming into force of the standing orders and constitution considered for coding policy-­ making power of opposition players; and (3) the consideration of only the survey closest to the time of an election if more than one survey was available between two elections. Even though I cannot observe a causal effect directly, this strategy ensures that the rules determining the level

Methods and Data  89

of policy-making power of opposition players came into force before the surveys that measured individual political support were conducted. The analyses in Chapter 6 include three dependent variables to measure political support: democracy as the best form of government, satisfaction with democracy, and satisfaction with government performance. I follow Norris (2011) and apply a multidimensional approach to measure political support. Therefore, different dependent variables are considered when analyzing both the link between the policy-making power of opposition players, and diffuse as well as specific support. Furthermore, the analyses in Chapter 6 include several country-level and individual-level control variables: decentralization, regime type, electoral system, GDP/ capita, corruption, democratic experience, education, an individual´s economic situation, interest in politics, feeling close to a political party, and life satisfaction. To present a broad comparative analysis, results are presented for three different surveys measuring the dependent variables (CSES, ESS, and LAPOP).

Notes 1 See Chapter 3 for a detailed discussion. 2 For a detailed presentation of the operationalization of this variable see Chapter 3. For a detailed discussion of the descriptive results of policy-­ making power of opposition players see Chapter 5. 3 See Table 4.3 in this chapter. 4 See also Wegmann (2022) and Chapter 3 for a more detailed discussion of the operationalization of policy-making power of opposition players. 5 The author’s suggestion is then also to refrain from using the “satisfaction with democracy” question. 6 To conduct the analyses, I am only able to consider surveys that allow for the identification of winners and losers of elections. 7 Ideally, a measure of political support considers all five levels of political support proposed by Norris (2011). The inclusion of only three instead of all five levels is due to data availability and not based on any theoretical reasoning. 8 A look at correlations for a split sample of older and younger democracies (20 years or less of democratic experience) within the Comparative Study of Electoral Systems (CSES 2019) and the European Social Survey (2019) gives some evidence to this notion. CSES: Correlations between ‘satisfaction with democracy’ and ‘government performance’: Young democracy: 0.21, Old democracy: 0.26. ESS: Correlations between ‘satisfaction with democracy’ and ‘government performance’: Young democracy: 0.54, Old democracy: 0.67. 9 Henderson (2008), however, finds a negative influence of federalism on the level of satisfaction with democracy. Accordingly, winning at the national level significantly contributed to more satisfaction with democracy whereas winning at the local level does not (Henderson 2008: 13, 18). 10 I follow Lijphart’s (1999) terms of bicameral parliaments by using the term first chamber or lower house (for the more important chamber or the one that

90  Power of Opposition Players and Consequences

11 12

13 14 15 16

17

18

is elected on the basis of population) and second chamber or upper house (Lijphart 1999: 201). This data is provided by Dahlberg et al. (2019). See Table A.4.1 of the Appendix for more detail. The authors argue that, empirically, “constitutional federalism is a necessary condition for strong bicameralism” and, conceptually, “the purpose of a strong second chamber is usually to protect the powers and prerogatives of subnational units” (Gerring, Thacker and Moreno 2005: 571, footnote 2). See Appendix Table A.4.1 for details. Dataset provided by Dahlberg et al. (2019). Dataset provided by Dahlberg et al. (2019). These countries are: Albania, Benin, Burundi, Cape Verde, Comoros, Croatia, Georgia, Ghana, Guatemala, Guyana, Indonesia, Kenya, Kyrgyzstan, Liberia, Mexico, Moldova, Mongolia, Mozambique, Paraguay, Poland, Romania, Senegal, Serbia and Montenegro, Sierra Leone, South Africa, and Taiwan. These countries are: Bangladesh, Dominican Republic, Fiji, Guinea-Bissau, Haiti, Lesotho, Madagascar, Malawi, Malaysia, Mali, Nepal, Niger, Pakistan, Peru, Russia, Solomon Islands, Sri Lanka, Thailand, Tunisia, Ukraine, and Zambia. These countries are: Argentina, Belgium, Bolivia, Botswana, Bulgaria, Chile, Colombia, El Salvador, Honduras, India, Israel, Jamaica, Nicaragua, Panama, South Korea, Trinidad and Tobago, and Turkey.

Bibliography Aarts, Kees and Jacques Thomassen. 2008. “Satisfaction with Democracy: Do Institutions Matter?” Electoral Studies 27(1): 5–18. Anderson, Christopher J., Andre Blais, Shaun Bowler, Todd Donovan and Ola Listhaug. 2005. Losers’ Consent: Elections and Democratic Legitimacy. Oxford and New York: Oxford University Press. Anderson, Christopher J. and Christine A. Guillory. 1997. “Political Institutions and Satisfaction with Democracy: A Cross-National Analysis of Consensus and Majoritarian Systems.” American Political Science Review 91(1): 66–81. Beramendi, Pablo. 2007. Federalism. In The Oxford Handbook of Comparative Politics, eds. Carles Boix and Susan C. Stokes. Oxford: Oxford University Press: 752–781. Berggren, Heidi M., Gregory A. Fugate, Robert R. Preuhs and Dennis R. Still. 2004. “Satisfied? Institutional Determinants of Citizen Evaluations of Democracy.” Politics & Policy 32(1): 72–96. Bernauer, Julian and Adrian Vatter. 2012. “Can’t Get No Satisfaction with the Westminster Model? Winners, Losers and the Effects of Consensual and Direct Democratic Institutions on Satisfaction with Democracy.” European Journal of Political Research 51(4): 435–468. Blais, André, Alexandre Morin-Chassé and Shane P. Singh. 2017 “Election Outcomes, Legislative Representation, and Satisfaction with Democracy.” Party Politics 23(2): 85–95. Blais, André and François Gélineau. 2007. “Winning, Losing and Satisfaction with Democracy.” Political Studies 55(2): 425–441.

Methods and Data  91 Bormann, Nils-Christian and Matt Golder. 2013. “Democratic Electoral Systems around the World, 1946–2011.” Electoral Studies 32(2): 360–369. Bowler, Shaun and Todd Donovan. 2007. “Reasoning about Institutional Change: Winners, Losers and Support for Electoral Reforms.” British Journal of Political Science 37(3): 455–476. Canache, Damary, Jeffery J. Mondak and Mitchell A. Seligson. 2001. “Meaning and Measurement in Cross-National Research on Satisfaction with Democracy.” Public Opinion Quarterly 65: 506–528. Cheibub, José Antonio, Jennifer Gandhi and James Vreeland. 2010. “Democracy and Dictatorship Revisited.” Public Choice 143: 67–101. Clarke, Harold D., Nitish Dutt and Allan Kronberg. 1993. “The Political Economy of Attitudes toward Polity and Society in Western Europe.” The Journal of Politics 55(4): 998–1021. Conroy-Krutz, Jeffrey and Nicholas Kerr. 2015. “Dynamics of Democratic Satisfaction in Transitional Settings: Evidence from a Panel Study in Uganda.” Political Research Quarterly 68(3): 593–606. Coppedge, Michael, John Gerring, Staffan I. Lindberg, Jan Teorell Svend-Erik Skaaning, with David Altman, Michael Bernhard, M. Steven Fish, Adam Glynn, Allen Hicken, Carl Henrik Knutsen, Pamela Paxton Kelly McMann, Daniel Pemstein, Jeffrey Staton, Frida Andersson Brigitte Zimmerman, Valeriya Mechkova and Farhad Miri. 2017. “V-Dem Codebook v5. Varieties of Democracy (V-Dem) Project.”. Online: http://www.v-dem.net/dsarchive.html. Accessed: January 2016. CSES. 2019. “The Comparative Study of Electoral Systems. CSES Modules 1 to 4 Full Release. Dataset. December 15, 2015 version. DOI Module 1: 10.7804/ cses.module1.201512-15; DOI Module 2: 10.7804/cses.module2.2015-12-15; DOI Module 3:10.7804/cses.module3.2015-12-15; DOI Module 4: 0.7804/cses. module4.2015–0320. CSES Module 5 First Advance Release. Dataset. May 21, 2019 version. doi:10.7804/cses.module5.2019-05-21.” Online: www.cses.org. Curini, Luigi, Willy Lou and Vincenzo Memoli. 2012. “Satisfaction with Democracy and the Winner/Loser Debate: The Role of Policy Preferences and Past Experience.” British Journal of Political Science 42(2): 241–261. Dahlberg, Stefan, Sören Holmberg, Bo Rothstein, Natalia Alvarado Pachon and Richard Svensson. 2019. “The Quality of Government Basic Dataset, version Jan19. University of Gothenburg: The Quality of Government Institute.” Online: http://www.qog.pol. gu.sedoi:10.18157/qogbasjan19. Accessed: October 2019. Diamond, Larry J. 1994. “Toward Democratic Conslidation.” Journal of Democracy 5(3): 4–17. Diamond, Larry J. 1999. Developing Democracy: Toward Consolidation. Baltimore, MD: Johns Hopkins University Press. Döring, Holger and Philipp Manow. 2018. “Parliaments and Government Database (ParlGov): Information on Parties, Elections and Cabinets in Modern Democracies.”. Online: https://parlgov.org/data-info/. Accessed: March 2018. Easton, David. 1957. “An Approach to the Analysis of Political Systems.” World Politics 9(3): 383–400. Easton, David. 1965a. A Framework for Political Analysis. Englewood Cliffs, NJ: Prentice Hall. Easton, David. 1965b. A Systems Analysis of Political Life. New York: Wiley.

92  Power of Opposition Players and Consequences Easton, David. 1975. “A Re-Assessment of the Concept of Political Support.” British Journal of Political Science 5(4): 435–457. ESS. 2019. “European Social Survey. Multilevel Data.” Online: http://www.europeansocialsurvey.org/. Accessed: October 2019. Galligan, Brian. 2006. Comparative Federalism. In The Oxford Handbook of Political Institutions, eds. Sarah A. Binder, R.A.W. Rhodes and Bert A. Rockman. Oxford and New York: Oxford University Press: 261–280. Gerring, John, Philip Bond, William Brandt and Carola Moreno. 2005. “Democracy and Growth: A Historical Perspective.” World Politics 57(3): 323–364. Gerring, John, Strom C. Thacker and Carola Moreno. 2005. “Centripetal Democratic Governance: A Theory and Global Inquiry.” The American Political Science Review 99(4): 567–581. Henderson, Ailsa. 2008. “Satisfaction with Democracy: The Impact of Winning and Losing in Westminster Systems.” Journal of Elections, Public Opinion and Parties 18(1): 3–26. Hobolt, Sara B. 2012. “Citizen Satisfaction with Democracy in the European Union.” Journal of Common Market Studies 50(1): 88–105. Howell, Patrick and Florian Justwan. 2013. “Nail-Biters and No-Contests: The Effect of Electoral Margins on Satisfaction with Democracy in Winners and Losers.” Electoral Studies 32(2): 334–343. IDEA. 2014. “Institute for Democracy and Electoral Assistance. Direct Democracy Database.” Online: http://www.idea.int/elections/dd/search.cfm. Accessed: January 2014. Knutsen, Carl Henrik and Simone Wegmann. 2016. “Is Democracy about Redistribution?” Democratization 23(1): 164–192. Lijphart, Arend. 1999. Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. New Haven, CT and London: Yale University Press. Linde, Jonas and Joakim Ekman. 2003. “Satisfaction with Democracy: A Note on a Frequently Used Indicator in Comparative Politics.” European Journal of Political Research 42(3): 391–408. Linz, Juan J. 1990. “The Virtues of Parliamentarism.” Journal of Democracy 1(4): 84–91. Linz, Juan J. and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation. Baltimore, MD: Johns Hopkins University Press. March, James G. and Johan P. Olsen. 1984. “The New Institutionalism: Organizational Factors in Political Life.” The American Political Science Review 78(3): 734–749. Marshall, Monty G., Keith Jaggers and Ted Robert Gurr. 2018. “Polity V Project. Political Regime Characteristics and Transitions, 1800–2018.” Online: http://www.systemicpeace.org/inscrdata.html. Accessed: October 2019. Mishler, William and Richard Rose. 1996. “Trajectories of Fear and Hope: Support for Democracy in Post-Communist Europe.” Comparative Political Studies 28(4): 553–581. Mishler, William and Richard Rose. 2001. “Political Support for Incomplete Democracies: Realist vs. Idealist Theories and Measures.” International Political Science Review 22(4): 303–320.

Methods and Data  93 Norris, Pippa. 1999. Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press. Norris, Pippa. 2011. Democratic Deficit. Critical Citizens Revisited. New York: Cambridge University Press. Pop-Eleches, Grigore and Joshua A. Tucker. 2011. “Communism’s Shadow: Postcommunist Legacies, Values, and Behavior.” Comparative Politics 43(4): 379–408. Pop-Eleches, Grigore and Joshua A. Tucker. 2012. “Post-Communist Legacies and Political Behavior and Attitudes.” Demokratizatsiya: The Journal of Post-Soviet Democratization 20(2): 157–166. Pop-Eleches, Grigore and Joshua A. Tucker. 2014. “Communist Socialization and Postcommunist Economic and Political Attitudes.” Electoral Studies 33: 77–89. Przeworski, Adam. 2004. “Institutions Matter?” Government and Opposition 39(4): 527–540. Riker, William H. 1975. Federalism. In Handbook of Political Science, 5: Governmental Institutions and Processes, eds. Fred I. Greenstein and Nelson W. Polsby. Reading, MA: Addison-Wesley: 93–172. Rose, Richard and William Mishler. 2002. “Comparing Regime Support in Nondemocratic and Democratic Countries.” Democratization 9(2): 1–20. Rose, Richard, William Mishler and Christian Haerpfer. 1998. Democracy and Its Alternatives. Baltimore, MD: John Hopkins University Press. Ruiz-Rufino, Ruben. 2013. “Satisfaction with Democracy in Multi-ethnic Countries: The Effect of Representative Political Institutions on Ethnic Minorities.” Political Studies 61(1): 101–118. Singh, Shane P. 2014. “Not All Election Winners Are Equal: Satisfaction with Democracy and the Nature of the Vote.” European Journal of Political Research 53(2): 308–327. Singh, Shane P., Ekrem Karakoç and André Blais. 2012. “Differentiating Winners: How Elections Affect Satisfaction with Democracy.” Electoral Studies 31(1): 201–211. Svolik, Milan W. 2013. “Learning to Love Democracy: Electoral Accountability and the Success of Democracy.” American Journal of Political Science 57(3): 685–702. Thomassen, Jacques. 1995. Support for Democratic Values. In Citizens and the State, eds. Hans-Dieter Klingemann and Dieter Fuchs. Oxford and New York: Oxford University Press: 383–416. Tóka, Gábor. 1995. Political Support in East-Central Europe. In Citizens and the State, eds. Hans-Dieter Klingemann and Dieter Fuchs. Oxford and New York: Oxford University Press: 354–381. Tsebelis, George and Jeannette Money. 1997. Bicameralism. Cambridge: Cambridge University Press. Uhr, John. 2006. Bicameralism. In The Oxford Handbook of Political Institutions, eds. Sarah A. Binder, R.A.W. Rhodes and Bert A. Rockman. Oxford and New York: Oxford University Press: 474–494. Vanderbilt University. 2019. “Latin American Public Opinion Project (LAPOP).” Online: http://www.vanderbilt.edu/lapop/. Accessed: October 2019.

94  Power of Opposition Players and Consequences Voicu, Malina and Edurne Bartolome Peral. 2013. “Support for Democracy and Early Socialization in a Non-Democratic Country: Does the Regime Matter?” Democratization 21(3): 554–573. Watts, Ronald L. 1996. Comparing Federal Systems in the 1990s. Kingston, Ontario: Institute of Intergovernmental Relations. Queen’s University. Wegmann, Simone. 2022. “Policy-Making Power of Opposition Players: A Comparative Institutional Perspective.” The Journal of Legislative Studies 28(1): 1–25. Wells, Jason M. and Jonathan Krieckhaus. 2006. “Does National Context Influence Democratic Satisfaction? A Multi-Level Analysis.” Political Research Quarterly 59(4): 569–578. World Bank. 2016. “World Development Indicators. The World Bank Washington DC.”. Online: http://data.worldbank.org/data-catalog/world-development-­ indicators. Accessed: October 2018.

Chapter 5

Policy-Making Power of Opposition Players around the World

In the literature on legislative studies relatively few contributions deal with the specific rights of the opposition.1 Therefore, we have only a very limited understanding of how inclusive parliamentary procedures actually are. In contrast, a vast literature exists on legislative organization more generally, indicating important differences between countries (see, e.g. Sieberer 2011; Döring 2001). Most importantly, as previous research has shown, legislative organization is more varied than traditional categorizations such as parliamentary and presidential regimes would suggest (see, e.g. Döring 2001). The descriptive results presented in the following sections provide first evidence that focusing on legislative organization related to opposition rights may be a promising new approach to understanding the role of legislatures and the dynamics of democratic consolidation. This chapter proceeds as follows. In the first part, I deal with the question of the degree of policy-making power accorded to opposition players. I present the results of the overall index of policy-making power of opposition players around the world including a discussion of the endogenous character of parliamentary procedures, followed by a more detailed look into the seven variables that constitute the overall index. In the second part of the chapter, I deal more specifically with the question of the kind of policy-making power opposition players have. To this end, I present the results of the policy-making power of opposition players in the lower chambers of Chile, the Czech Republic, and Slovenia. Whereas Chile is among the countries that best represents an average case, Slovenia shows the strongest policy-making power of opposition players in the sample. The lower house of the Czech Republic grants intermediate power to opposition players, however, a closer look at the individual variables nicely shows that this intermediate score is the result of a combination of very strong and very weak opposition power.

DOI: 10.4324/9781003005360-7

96  Power of Opposition Players and Consequences

5.1  H ow Much Policy-Making Power Do Opposition Players Have? As the previous chapters have shown, moving beyond the characteristics of government formation – focusing on winners of democratic elections – and considering the role of losers instead may be a promising approach to our understanding of the processes leading to more democratic consolidation. It is inclusive parliamentary procedures that grant more power to opposition players and may thereby increase political equality between winners and losers. More political equality may then contribute to more political support – especially of losers – and increase democratic consolidation.2 As presented in Chapter 3, I propose to measure such inclusive parliamentary procedures by looking more closely at the power opposition players have during the policy-making process. In a first step, it is important to know how much policy-making power opposition players actually have and whether this power differs between legislatures. The results in the following sections cover lower houses of all 77 countries for which data is available.3 Among the 77 countries, 29 are parliamentary systems, whereas 30 are presidential and 18 are semi-­presidential. The regional distribution of the sample shows a majority of European and American countries (30/39 percent and 22/29 percent, respectively), though, the sample includes 11 African countries (14 percent) and 14 Asian countries (18 percent).4 However, for theoretical reasons, the main focus of the chapter lies on the distribution of the policy-­making power of opposition players within and across different regime types rather than across different world regions. The main goal is to see whether different regime types result in different levels of policy-making power of opposition players. This is important as it will be evidence of whether or not moving beyond a focus on government formation – by looking more closely at the degree of power granted to opposition players – actually captures a new element that may help explain democratic consolidation. However, as Figure 5.1 shows, the distribution of parliamentary, semi-presidential, and presidential regimes also shows a regional pattern. Each regime type is predominantly present in one of four world regions (see also Cheibub, Elkins and Ginsburg 2014). A slight majority of parliamentary systems are in Europe, whereas about one-third of them are in Asia. Only a small minority of parliamentary regimes are in Africa (Mauritius) and the Americas (Canada, Jamaica, and Trinidad and Tobago). The regional pattern is more pronounced for semi-presidential and presidential regimes. Almost three-quarters of semi-presidential regimes are in Europe (13 out of 18) and none of them are in the Americas. On the contrary, almost four-fifths of the presidential regimes are American or African (25 out of 30). Hence, we see a clear distribution of regime types across different world regions.

Power of Opposition Players around the World  97

Figure 5.1  Regional Distribution of Regime Types.

In the remainder of this chapter, I follow the structure of the operationalization of policy-making power of opposition players presented in Chapter 3. In the first section, I show a detailed overview of the combined measure of policy-making power of opposition players before I present the results for the individual variables. 5.1.1  Policy- Making Power of Opposition Players Figure 5.2 presents the distribution of the policy-making power of opposition players across 54 legislative chambers around the world. With a mean score of 0.59 the overall picture of the policy-making power of opposition players points to intermediate power. None of the 54 countries, for which information on all indicators is available, show very weak opposition power (overall score of 0). Similarly, no country reaches the maximum score of 1. Slovenia is the country with the strongest policy-­ making power of opposition players (with a score of 0.92), whereas Bangladesh (with a score of 0.18) has the weakest power of opposition players. The highest number of countries (45 out of the 54 countries) is clustered around a level of policy-making power of opposition players between 0.4 and 0.8 (see also Wegmann 2022). However, the most interesting empirical question for the argument proposed in this book is to know whether strong policy-making power

98  Power of Opposition Players and Consequences

Figure 5.2  Distribution of Policy-Making Power of Opposition Players.

Figure 5.3  Policy-Making Power of Opposition Players by Regime Type.

of opposition players is linked to a specific regime type. Figure 5.3 shows the distribution of the policy-making power of opposition players by regime types. The mean of the policy-making power of opposition players is highest for semi-presidential regimes (0.69, median 0.72) compared to presidential regimes (0.59, median 0.60) and parliamentary regimes (0.55, median 0.51). But, as Figure 5.3 also illustrates, the variance in the policy-making power of opposition players for parliamentary regimes is significantly larger than for semi-presidential and presidential regimes. Contrary to parliamentary regimes, none of the semi-presidential or presidential regimes have a score of policy-making power of opposition players below 0.46. However, none of the presidential regimes show an overall score of more than 0.8 whereas one semi-presidential and three parliamentary regimes score above this threshold. Hence, parliamentary regimes show both, the weakest and the strongest policy-making power of opposition players. Even though we see differences in the policy-making power of opposition players between regime types, results show that by looking only at how governments are formed it is very unlikely that we will be able to grasp the differences in the power granted to opposition players. This result is in line with results of a more detailed analysis (Wegmann 2022) showing that regime type is not among the institutional determinants of strong policy-making power of opposition players. Parliamentary regimes, especially, show a wide range of power granted to opposition players. Characterizing parliamentary regimes, in particular, as granting

Power of Opposition Players around the World  99

either weak or strong policy-making power of opposition players seems impossible. Hence, the variance in policy-making power of opposition players reflects in previous research showing that, as parliamentary rules are adopted and modified by parliaments themselves, we see a wide range of different rules in legislative chambers (see, e.g. Zubek 2015; Müller and Sieberer 2014; Martin 2011; Sieberer, Müller and Heller 2011; Taylor 2006; Döring 1995). But, this also begs the question of the endogenous character of these rules and its consequences for research focusing on the effect of legislative institutions. Even if it is not strongly linked to different regime types, the level of policy-making power of opposition players may be connected to other institutional characteristics. The endogenous character of legislative rules is especially important, as some of the determining characteristics of such rules might also be relevant for the political support of individuals. According to the logic of the mirroring principle, parliamentary rules mirror the external environment of the “…separations of purpose and power within a polity” (McCubbins 2005: 123). Research has repeatedly pointed to two characteristics that are particularly mirrored in the rules of procedure of legislative chambers: the electoral system and the size of the chamber. A large literature exists on the connection between electoral rules and legislative power. Particularly, the strength of the committee system seems to be linked to the electoral system as a strong committee system mirrors the electoral need of MPs to cultivate a personal vote (see, e.g. Martin 2018; Carey 2009). Powell (2000: 39), for example, highlights that legislative chambers elected by PR electoral formula also show proportional decision rules (such as opposition influence in committees). Furthermore, research has shown that legislative procedures in larger chambers likely reflect their size in more restrictive rules (see, e.g. Taylor 2006; Wawro and Schickler 2006). Testing the mirroring principle for the policy-making power of opposition players partially confirms this existing research. Strong policy-­ making power of opposition players actually seems to mirror mixed and proportional electoral systems but not the size of the legislative chamber (Wegmann 2022). Furthermore, policy-making power of opposition players is weaker in Asian countries compared to European countries and is stronger in countries with high religious fractionalization. However, this also highlights that the considerable variance in policy-making power of opposition players around the world is not necessarily captured by institutional characteristics such as the size of the legislative chamber, democratic experience, or the regime type of the country (Wegmann 2022: 19). Nevertheless, the possible connection to democratic experience merits to be highlighted. This is particularly important regarding the literature speaking to an effect of democratic learning on the level of individual

100  Power of Opposition Players and Consequences

opinions (see, e.g. Svolik 2013; Anderson and Mendes 2005). As strong political support on the individual level might only develop over time and be particularly influenced by one’s past, the institutional characteristics of democracies might also develop over time. Accordingly, policy-­ making power of opposition players might develop only as democracies consolidate and therefore be a mere effect of stable democracies. If this were the case, however, one should observe weak policy-making power of opposition players among all young democracies and stronger power among older ones. Figure 5.4 shows the level of policy-making power of opposition players based on the years of democratic experience at the time that the standing orders, used for coding, came into force. Results in Figure 5.4 do not show evidence of such a democratic learning pattern. Therefore, strong policy-making power of opposition players does not seem to be a consequence of an institutional development occurring only over time in long-living democracies. Although very young democracies have rather weak policy-making power of opposition players, a number of older ones do as well. A very weak correlation (0.05) between policy-­making power of opposition players and democratic experience at the time the standing orders came into force reflects this result (see also Wegmann 2022). In sum, this evidence does not point toward a possible causal direction of democratic consolidation leading to the development of strong

Figure 5.4  P olicy-Making Power of Opposition Players by Democratic Experience.

Power of Opposition Players around the World  101

policy-making power of opposition players over time. Nevertheless, results do highlight the endogenous character of legislative rules. Strong policy-making power of opposition players seems to be particularly linked to proportional and mixed electoral systems. The next sections show more detailed and descriptive analyses of the seven variables included in the overall index. As highlighted above, for theoretical reasons, the main focus lies on the distribution of policy-making power of opposition players within and across different regime types. 5.1.2  Bill Introduction Three different aspects determine how much power different actors have to actually introduce legislative bills. The first refers to the explicitly mentioned right to propose a bill. The second depends on numerical restrictions that have to be fulfilled in order to introduce bills, and the third refers to the right of other actors to veto the introduction of a bill. As I discussed in Chapter 3, the highest level of bill introduction power of opposition players exists in a legislative chamber in which actors other than the executive and MPs are able to introduce bills combined with no numerical restrictions or veto players. Figure 5.5 illustrates, separately, the results of the power to introduce bills for all legislative chambers in the sample and for each regime type. As expected, in all of the legislative chambers in the sample, the executive as well as individual MPs have the right to introduce bills. Therefore, whether any other actors, such as committees, parliamentary parties, or citizens also enjoy this right considerably influences the power granted to the opposition players. Results show that in a vast majority of the legislative chambers considered here, the power to introduce bills is strong or very strong. Hence, numerical restrictions or veto possibilities to bill introduction exist in only a small share of the legislative chambers (21.6 percent). Results, however, show an interesting pattern with reference to the distribution among regime types (lower part of Figure 5.5). The right of bill introduction indicates a similar pattern among presidential and semi-presidential regimes. In these two regime types, a majority of legislative chambers grants very strong power to opposition players to introduce bills (75 percent and 65.5 percent, respectively). On the contrary, this is the case in only 31 percent of parliamentary regimes. This result is in line with the legislative studies literature indicating that the government enjoys more agenda control in parliamentary systems. However, the variance within parliamentary regimes is considerable and indicates that simply considering regime types does not capture the degree of power granted to opposition players. Even among parliamentary regimes, a majority of legislative chambers (62 percent) grants strong or very strong bill introduction power to opposition players (i.e. no restrictions or veto to

102  Power of Opposition Players and Consequences

Figure 5.5  Bill Introduction by Regime Type.

bill introduction). Thus, the overall picture shows that only a minority of legislative chambers limits the right to introduce bills with numerical restrictions or veto power. 5.1.3  Agenda Setting However, the picture is not as pronounced for agenda setting powers. Two aspects define the power to set the plenary agenda from the perspective of opposition players. The first is whether or not a collective actor, including opposition players, determines the agenda. The second depends on the priority of government business. Collective actors, that include opposition players, setting the plenary agenda without priority of government business indicates the highest level of agenda setting power of opposition players. Figure 5.6 shows the distribution of agenda setting power of opposition players both overall and according to regime types. Results show that in nearly half of the legislative chambers (49.3 percent) a combination of either collective actors setting the agenda or no priority of government business exists.5 The number of countries that grant strong agenda setting power (i.e. a collective actor including oppositions setting the agenda and no priority of government business) and those that grant weak agenda setting power (i.e. a single actor setting the agenda and priority of government business) are nearly identical (26.0 percent and 24.7 percent, respectively).

Power of Opposition Players around the World  103

Figure 5.6  Agenda Setting by Regime Type.

Furthermore, results reveal that semi-presidential and presidential regimes, again, show almost identical patterns. In only a small minority of semi-presidential and presidential regimes does a single actor set the plenary agenda, and government business has priority (11.8 percent and 13.8 percent, respectively) whereas in parliamentary regimes 44.4 percent of legislative chambers show this pattern. The majority of the semi-presidential and presidential regimes combine either a single actor setting the agenda without priority of government business or a collective actor, including opposition players, setting the agenda with priority of government business. Across all three regime types, only a minority of countries is categorized as providing strong agenda setting power to opposition players (i.e. collective actors including oppositions that set the agenda without government business having priority). But here, we see no big differences between regime types. Parliamentary regimes even grant strong agenda setting power to opposition players in slightly more legislative chambers than presidential regimes do (25.9 percent compared to 24.1 percent, respectively). The literature on the traditional differentiation between parliamentary and presidential regimes would suggest different patterns between these regimes. For example, Mattson (1995: 449) has highlighted that the logic of bill introduction and agenda control differs significantly between presidential and parliamentary regimes, as the government has much more control in the latter than in the former. Results in this section show

104  Power of Opposition Players and Consequences

that the level of initiation power (including bill introduction and agenda setting) varies considerably across both regime types. Although results show a tendency of weaker initiation power of opposition players among parliamentary regimes, the level of initiation power does not clearly match regime types. Considerable variance exists especially within each regime type, and not only between them. Whereas most legislative chambers with weak initiation power are parliamentary, this regime type also shows a comparable share of countries with very strong initiation power. At the same time, not all presidential regimes grant strong initiation power to opposition players. Hence, the variance within regime types is considerably larger than the one between them. 5.1.4 Amendments Similar to the right to introduce bills, three different aspects determine the level of power different actors have to propose amendments. The first refers to the right to introduce amendments. The second aspect addresses the potential numerical restrictions to the proposition of an amendment, and the third refers to potential veto players to the introduction of an amendment. Figure 5.7 illustrates the distribution of the amendment power of opposition players. Among the 70 countries with detailed information about amendments, the majority (72.8 percent) grant the right to introduce amendments to at least two different actors.6 Similarly, in a vast majority of legislative chambers (77.1 percent), actors are able to introduce amendments without numerical restrictions or veto. Therefore, only a small number of legislative chambers limit opposition power by restricting access to the introduction of amendments. Even though this overall pattern resembles that of the power to introduce bills (see Figure 5.4), differences among regime types are less accentuated here. Opposition players are granted strong amendment power in 57.7 percent of the presidential regimes, (i.e. at least two actors can introduce amendments, no restrictions or veto), 50.0 percent of the semi-presidential regimes, and 46.4 percent of the parliamentary regimes, respectively. However, very weak or weak amendment power is granted to opposition players in only 15.4 percent of the presidential regimes, 25.0 percent of the semi-presidential regimes, and 28.6 percent of the parliamentary ones, respectively. Thus, we see slightly weaker power attributed to opposition players in parliamentary regimes. However, first, this difference is far less pronounced than in the case of the power to introduce bills. Second, results show again that categorizing countries according to their regime type is unlikely to grasp the power granted to opposition players.

Power of Opposition Players around the World  105

Figure 5.7  Amendments by Regime Type.

5.1.5  Committee Structure As presented in Chapter 3, the measure of the power granted to opposition players in committees is based upon Strøm (1990). Accordingly, the number of standing committees, a fixed area of specialization corresponding to ministerial departments, the restriction on the number of possible assignments per MP, and proportional distribution of committee chairs define the level of power the committee structure grants to opposition players. Figure 5.8 presents the results of this variable for all legislative chambers and for all three regime types, respectively. The results show a median of weak/intermediate opposition power of the overall measure of committee structure (category 2). In a majority of legislative chambers (52.1 percent), opposition players only benefit from very weak to weak/intermediate power (categories 0–2). Still, even though no legislative chamber reaches the maximum score of 5, 47.9 percent of legislative chambers grant strong/intermediate or strong committee power to opposition players (categories 3 or 4). Looking more closely at the distribution among regime types reveals that presidential regimes most closely match this overall pattern. But, legislative chambers in all three regime types cluster around intermediate categories of opposition power (categories 2 and 3). Nevertheless, parliamentary regimes show slightly fewer legislative chambers with strong opposition power (7.1 percent compared to 20 percent and 14.3 percent, respectively). Again,

106  Power of Opposition Players and Consequences

Figure 5.8  Committee Structure by Regime Type.

parliamentary regimes show slightly weaker opposition power in committee structure than presidential and semi-presidential regimes. However, the variation within all regime types is considerable and there is no indication that one could draw any conclusion about the degree of opposition power by simply considering the regime type of a given country. 5.1.6  Committee Procedures As presented in Chapter 3, the degree of power granted to opposition players is defined not only by the committee structure but also by the committee procedures. The ability to submit minority reports, to set its own agenda, and to rewrite bills determines the degree of power opposition players have in reference to the committee procedures. Figure 5.9 illustrates the overall results as well as the distribution of the power of committee procedures across regime types. The distribution of the power of committee procedures shows a slightly different pattern to that of the committee structure. The median tends toward strong committee procedures (category 2). Overall, half of the legislative chambers (50 percent) grant strong power (category 2) and another 12.9 percent of legislative chambers grant very strong power (category 3) of committee procedures to opposition players. However, in comparison with the distribution of the committee structure, 12.9 percent of the legislative chambers grant the weakest possible power with reference to committee procedures. An examination of the distribution across regime types indicates that the share of weak power of committee

Power of Opposition Players around the World  107

Figure 5.9  Committee Procedures by Regime Type

procedures (categories 0 and 1) is higher among presidential regimes than among parliamentary or semi-presidential regimes. In the presidential regimes, 52.4 percent of legislative chambers grant very weak or weak opposition power (categories 0 and 1) whereas this is only the case for 35.7 percent of parliamentary and 15.4 percent of semi-presidential regimes, respectively. Moreover, semi-presidential regimes show an interesting pattern as none of them grant very weak committee procedures (category 0). In addition, parliamentary regimes show the highest share of legislative chambers granting very strong committee procedures (17.9 percent). 5.1.7  Executive Power The two different aspects of decree power and veto power determine the level of executive power. Opposition players experience the most veto power if the executive can neither issue decrees nor veto the adoption of a bill. Opposition players experience the least power in a setting in which the executive has both decree and veto power. A combination of either strong executive veto power and weak decree power or weak executive veto power and strong decree power indicates an intermediate category. Figure 5.10 shows the descriptive results of executive power. Importantly, as presented in Chapter 3, the coding is conducted from the perspective of opposition players: strong power indicates more power granted to opposition players (as compared to the government). Thus, strong executive power indicates more power from the perspective of opposition players.

108  Power of Opposition Players and Consequences

Figure 5.10  Executive Power by Regime Type.

Among the 74 countries for which information about executive power is available, only a minority of legislative chambers (27 percent) grant weak opposition power (i.e. both executive decree and executive veto power). The majority of legislative chambers (59.5 percent) have either decree or veto power of the government, but not both. The remaining legislative chambers (13.5 percent) grant strong power to opposition players. Taking a closer look at the distribution between regime types shows similar patterns. Even though parliamentary regimes tend to grant more power to opposition players, results show strong opposition power for 16.7 percent of the semi-presidential and 3.6 percent of the presidential regimes. Among all three regime types, the majority of legislative chambers grants intermediate power to opposition players. Therefore, even if we see a pattern of more veto and/or decree power of the executive in presidential and semi-presidential regimes, looking at the regime type alone does not allow us to assess the degree of power granted to opposition players. 5.1.8 Referendums Finally, the possibility of referendums must be considered when measuring the policy-making power of opposition players. This variable is classified into four categories, higher scores indicating more power of opposition players: The first is countries in which neither mandatory nor optional referendums exist. The second category is countries in which

Power of Opposition Players around the World  109

legal provisions for optional referendums exist but opposition players7 have no right to ask for such a referendum. The third category is that of countries in which legal provisions for mandatory referendums exist. The final category is that of countries with optional referendums that allow at least one of the opposition players to ask for a referendum. As Figure 5.11 shows, in a large majority of countries provisions for either mandatory or optional referendums exist. Only 16 percent of the countries have no provision for either optional or mandatory referendums. In 14.7 percent of the countries there are mandatory referendums whereas there are provisions for optional referendums in 68.7 percent of the countries. However, only 38.7 percent of the legislative chambers grant not only referendum power to majority parties but also give opposition players the right to ask for a referendum. In sum, slightly more than half of the legislative chambers (53.4 percent) have strong or very strong opposition power. The distribution of referendums across regime types shows a different pattern than the distribution of executive power. Presidential and semi-presidential regimes tend to grant the right to ask for referendums to opposition players more often than parliamentary regimes. In 39.3 percent of the presidential regimes and 55.6 percent of the semi-­ presidential ones at least one opposition player (parliamentary minority, electors, or subnational entities) has the right to ask for an optional referendum. However, this is the case in only 27.6 percent of the parliamentary regimes. In addition, the number of countries which have no legal

Figure 5.11  Referendum by Regime Type.

110  Power of Opposition Players and Consequences

provisions for referendums is considerably higher in parliamentary regimes (27.6 percent compared to 5.6 percent of semi-presidential and 10.7 percent of presidential regimes, respectively). This indicates more veto power for opposition players in presidential, and especially in semi-presidential, regimes. However, as results for the other variables have already shown, considering regime types does not fully explain the degree of referendum power granted to opposition players.

5.2  W hat Kind of Policy-Making Power Do Opposition Players Have? As the descriptive results in the previous sections have shown, broad regime classifications such as parliamentary, semi-presidential, and presidential regimes do not allow us to understand the degree of power these regimes attribute to opposition players. This finding is confirmed by Wegmann (2022) showing that the regime type is not among the institutional characteristics explaining different degrees of the policy-making power of opposition players. Patterns across different regime types reveal that parliamentary regimes, especially, may not be classified as either granting weak or strong policy-making power to opposition players. Whereas semi-presidential and presidential regimes tend to guarantee at least intermediate levels of power, the distribution among parliamentary regimes ranges from very weak to very strong policy-making power of opposition players. These results may suggest that legislative chambers score either high or low on all seven variables included in the final measure of the policy-­making power of opposition players.8 Figure 5.12 shows correlations among each of the dimensions of the policy-making power of opposition players (initiation, debate, and veto power). For comparability, this figure is based on the same cases used in Figure 5.1, including all 55 countries for which complete information on all variables is available. Correlations between the three dimensions do not confirm such overall patterns. None of the three correlations between the separate dimensions is particularly strong. Correlations are highest between initiation and veto power (0.41) as reflected in the second panel in Figure 5.12. Legislative chambers that grant rather strong initiation power to opposition players also tend to guarantee strong veto power and vice versa. The picture in the first and the third panel of Figure 5.12 are not as clear. This is also reflected in the weakest correlation between debate and veto power of 0.26 (third panel) and an intermediate correlation of 0.34 between initiation and debate power (first panel).9 As these results show, a considerable number of legislative chambers combine strong opposition power on some of the variables with weak power granted to opposition players on others. Consequently, the question that follows from these observations is what kind of policy-making

Power of Opposition Players around the World  111

Figure 5.12  C orrelations of Initiation, Debate, and Veto Power of Opposition Players

Figure 5.13  Distribution of Policy-Making Power of Opposition Players

power do opposition players actually have? To get at least a partial answer to this question, I present three countries in more detail in the following sections. One way to select countries would be to look at the most prominent ones, i.e. the United States and the United Kingdom as they are also the countries representing presidential and parliamentary regimes most clearly. However, the literature has increasingly highlighted that these two countries are among the few that come close to the ideal types of parliamentarism and presidentialism, respectively (see, e.g. Cheibub, Elkins and Ginsburg 2014; Shugart 2006). Looking more closely at the United States and the United Kingdom would therefore present a biased picture of the policy-making power of opposition players in different regimes. Furthermore, the previous sections and Wegmann (2022) have shown that the way governments are formed does not translate into either weak or strong policy-making power of opposition players. Therefore, in the next section I focus on countries that most closely represent the mean in the sample as well as legislative chambers that show interesting combinations of weak and strong opposition power in different dimensions/variables. Figure 5.13, again, shows the overall distribution of the policy-making power of opposition players (as shown in Figure 5.2) but highlights the selected cases: Chile, the Czech Republic, and Slovenia.

112  Power of Opposition Players and Consequences

Chile is the country closest to the mean of the policy-making power of opposition players. Chile not only shows a score of policy-making power of opposition players closest to the mean, but also shows scores close to the mean of the different dimensions of the concept (initiation, debate, and veto). Therefore, Chile is among the countries that best represents an average case in the sample. Slovenia, on the contrary, is what could be labelled an extreme case. With an overall score of 0.92, Slovenia shows the strongest policy-making power of opposition players in the sample. This is also reflected in high scores on each of the three dimensions. Finally, the Czech Republic is one of the countries showing an interesting combination of the different variables. Even though the Czech Republic shows an overall score (0.65) relatively close to the mean (0.59), it represents the group of countries that combine weak and strong opposition power on the different variables. It is also among those showing the largest variance on the different dimensions (initiation, debate, and veto power). In the following sections, I present these three cases in more detail. 5.2.1  Slovenia – Lower House Državni zbor Figure 5.14 shows an overview of the policy-making power of opposition players in the lower house of Slovenia (Državni zbor). Among the countries in the sample, policy-making power of opposition players is strongest in Slovenia (with an overall score of 0.92). Opposition players are granted the strongest possible power during the initiation stage of the policy-making process (bill introduction and agenda setting) as well as the veto stage (executive power and referendums) of the policy-making process. Only during the debate stage do opposition players not benefit from the strongest power regarding the committee structure and committee procedures. However, the legislative chamber still scores above the overall mean on these two variables. According to Article 114 of the Rules of Procedure of the National Assembly, bills may be introduced “… by the Government, a deputy, the National Council, or at least 5,000 voters” (Državni zbor 2007: Article 114).10 Hence, the lower house of Slovenia grants the right to introduce bills to a broad range of actors and without restrictions leading to very strong opposition power regarding bill introduction. Similarly, the policy-­making power of opposition players is very strong regarding agenda setting power. A collective body (the Council and the National Assembly itself) determines the agenda which does not grant particular priority to government business (Državni zbor 2007: Articles 23 and 64). A closer look at the second stage of the policy-making process (i.e. debate power, panels 3–5 in Figure 5.14) shows some variation in the degree of power granted to opposition players. Similar to the power to introduce bills, the lower house of Slovenia grants the power to introduce

Power of Opposition Players around the World  113

Figure 5.14  D istribution of Policy-Making Power of Opposition Players – Slovenia

amendments to a broad group of actors. According to Article 129 of the Rules of Procedure of the National Assembly, amendments may be introduced by “deputies, a deputy group, the working body concerned, and the working body referred to in Article 124 of these Rules of Procedure, as well as the Government where it is not itself the proposer of the law” (Državni zbor 2007). However, the power granted to opposition players in committees is considerably weaker. Even though there are more than ten standing committees with fixed areas of specialization, they do not all clearly correspond to a ministerial department. Moreover, the Rules of Procedure of the National Assembly do not mention any restrictions on the number of possible committee memberships per MP. According to Articles 33(2) and 33(4) of the rules of procedure, committee seats are proportionally distributed among parties and committee chairs and deputy chairs are split between government and opposition parties. However, it is not explicitly mentioned that committee chairs are to be proportionally distributed according to party size (Državni zbor 2007).11 Finally, weaker power is granted to opposition players based on committee procedures. The Rules of Procedure of the National Assembly specify the possibility of minority reports (Državni zbor 2007: Article 42(1)) and the fact that committees themselves are able to decide on the agenda (Državni zbor 2007: Article 50). However, the rules of procedure do not mention that committees themselves have unconditional authority to redraft bills. As Figure 5.14 shows, opposition players again have the strongest possible power during the veto stage of the policy-making process. According to the Constitution, the president can only issue decrees if the National Assembly is unable to meet due to war or a state of emergency (Državni zbor 1991: Article 108). Before promulgation of a law by the President of the Republic, the National Council may require the National Assembly

114  Power of Opposition Players and Consequences

to vote a second time, however, the executive does not have veto power (Državni zbor 1991: Article 91). Finally, opposition players also have the possibility to ask for an optional referendum (i.e. the parliamentary minority and citizens) (IDEA 2014) leading to strong referendum power. In sum, the lower chamber in Slovenia grants very strong policy-­ making power to opposition players. This is reflected in five of the seven composing variables of the final index achieving the highest score possible. As Figure 5.14 shows, policy-making power of opposition players is only somewhat weaker regarding committee structure and committee procedures. Contrary to Slovenia, which represents an extreme case in the sample, the lower house in Chile represents the average case in the sample.  hile – Lower House Cámara de diputadas 5.2.2  C y diputados Figure 5.15 shows the results of the individual variables and the overall score of the policy-making power of opposition players in the lower house of Chile (Cámara de diputadas y diputados). As the right panel (PPOP) in Figure 5.14 shows, the Cámara de diputadas y diputados in Chile shows an overall score of policy-making power of opposition players of 0.59, the same score as the mean in the sample. Taking a closer look at the seven variables composing the overall index, Figure 5.15 reveals an interesting combination of very strong amendment power, intermediate bill introduction and committee procedures power and somewhat weaker agenda setting, committee structure, executive, and referendum power. The Constitution of the Republic of Chile specifies, in Article 65, that bills may be introduced either by the president or by any member of the legislative chamber (Cámara de diputadas y diputados 2010). Even though no other actors are entitled to introduce bills, the constitution and standing orders of the chamber do not specify any numerical restrictions or actors

Figure 5.15  D istribution of Policy-Making Power of Opposition Players – Chile

Power of Opposition Players around the World  115

entitled to veto a bill’s introduction. Therefore, even though the number of actors that are entitled to introduce bills is limited in the Chilean lower house, they may do so without facing restrictions, leading to strong opposition power (see first panel in Figure 5.15). According to the standing orders of the Chilean lower house (Cámara de diputadas y diputados 2014: Article 108), the agenda is set by the president of the chamber, the vice president of the chamber and the committees. However, the standing orders also specify the possibility of prioritizing government business leading to intermediate opposition power (second panel in Figure 5.15). Therefore, whereas opposition players benefit from rather strong power to introduce bills, this power is somewhat reduced by weaker agenda setting power. Overall, however, the legislative chamber grants intermediate power during this first-stage of the policy-making process. Looking more closely at the second stage of the policy-making process (i.e. debate power, panels 3–5 in Figure 5.15) also reveals intermediate power granted to opposition players. This intermediate power is the result of very strong amendment power, relatively weak power regarding the committee structure and intermediate power in reference to the committee procedures. First, the standing orders of the Chilean lower chamber (Cámara de diputadas y diputados 2014: Articles 274–275) indicate that individual MPs, committees, as well as the president are entitled to propose amendments to bills. Similar to the power to introduce bills, the rules of procedure do not indicate any numerical restrictions or actors entitled to veto the introduction of an amendment. Contrary to amendment power, the legislative chamber grants rather weak power to opposition players based on the committee structure. Even though the rules of procedure (Cámara de diputadas y diputados 2014: Article 216) mention 25 standing committees, they do not all clearly correspond to a ministerial department. Furthermore, the rules of procedure do not indicate any restrictions on the number of committee memberships per MP. Article 216 of the rules of procedure further specifies a proportional distribution of committee seats among members, but does not mention a corresponding distribution among committee chairs. Finally, more power is granted to opposition players based on committee procedures. The rules of procedure specify the possibility of minority reports (Cámara de diputadas y diputados 2014: Article 356) and the fact that committees themselves may set the agenda (Cámara de diputadas y diputados 2014: Article 264). However, the rules of procedure do not mention that committees themselves have unconditional authority to redraft bills. As Figure 5.15 shows, opposition players only have limited power during the final stage of the policy-making process (i.e. veto power). Even though the president lacks decree power (Article 64 of the Chilean constitution), he does benefit from veto power according to the constitution (Cámara de diputadas y diputados 2010: Article 73) as well as the rules of procedure

116  Power of Opposition Players and Consequences

(Cámara de diputadas y diputados 2014: Article 168–173). Finally, opposition players are only granted weak referendum power as there are optional referendums, but opposition players may not trigger them (IDEA 2014). In sum, the lower house in Chile grants intermediate policy-making power to opposition players and represents the average case in the sample (with a score of 0.59). Considering the three stages of the policy-making process (initiation, debate, and veto power), the lower house also grants intermediate power in all three stages. However, as Figure 5.15 shows, the policy-making power of opposition players is also characterized by some variations between the composing variables: whereas amendment power is very strong, referendum power is rather weak. The Czech Republic represents a case with an even more pronounced variation between the variables. 5.2.3  Czech Republic – Lower House Poslanecká sneˇmovna The right panel in Figure 5.16 shows a score of policy-making power of opposition players (PPOP) of 0.65, classifying the lower house of the Czech Republic as a case with intermediate opposition power and relatively close to the mean in the sample (0.59). However, a closer look at the variables shows particularly strong bill introduction and agenda setting power (first stage of the policy-making process). Opposition players benefit from intermediate power regarding the debate stage of the policy-making process (i.e. amendments, committee structure and committee procedures) and rather weak power during the veto stage of the policy-making process (executive veto and referendums). Hence, the Czech Republic perfectly shows that countries do not necessarily score high or low on all component variables of the final index. An intermediate score of the policy-making power of opposition players may be the result of a combination of very weak and very strong power during different stages of the policy-making process.

Figure 5.16  D istribution of Policy-Making Power of Opposition Players Czech Republic

Power of Opposition Players around the World  117

Similar to the lower house in Slovenia, the lower house in the Czech Republic grants strong power to opposition players during the first stage of the policy-making process. Opposition players are granted very strong power regarding both bill introduction and agenda setting. According to the Rules of Procedure of the Chamber of Deputies in the Czech Republic, “[a]ny Deputy, a group of Deputies, the Senate, the government or the councils of higher self-governing units (hereinafter only the ‘sponsor’) may present a draft bill” (Poslanecká sněmovna 2006: Article 86(1)). Furthermore, there is no mention in the rules of procedure of numerical restrictions or veto power for other actors in regards to bill introduction, and the agenda of the chamber of deputies is set by the steering committee with no priority given to government business (Poslanecká sněmovna 2006: Article 46). During the second stage of the policy-making process (i.e. debate power), opposition players benefit from intermediate power in the lower house of the Czech Republic. Contrary to the power to introduce bills, the range of actors that are allowed to introduce amendments is limited. Only individual MPs have the right to introduce amendments, however, they may do so without facing numerical restrictions or possible veto players (Poslanecká sněmovna 2006: Article 63(1)). In contrast to the lower houses in Slovenia and Chile, the rules of procedure of the lower house of the Czech Republic do specify a maximum number of committee memberships per MP (Poslanecká sněmovna 2006: Article 33(1)). Furthermore, the rules of procedure specify proportional representation of committee memberships, however it is not specific regarding committee chairs (Poslanecká sněmovna 2006: Article 115(2)). Similarly, even though there are more than ten standing committees, they do not clearly match ministerial departments. As the corresponding panel in Figure 5.16 shows, this leads to intermediate opposition power regarding the committee structure. Finally, opposition players are also granted intermediate power regarding committee procedures. There is a possibility to add minority reports and committees may set their own agenda (Poslanecká sněmovna 2006: Articles 36 and 40), however, they do not have unconditional authority to redraft bills (Poslanecká sněmovna 2006: Article 91). Finally, as Figure 5.16 shows, opposition players are only granted weak power during the last stage of the policy-making process (i.e. veto power). Whereas the executive lacks decree power, they do have veto power (Regan and Clark 2014) leading to intermediate opposition power with regards to executive power. Furthermore, there is no provision for mandatory or optional referendums that grant policy-making power to opposition players (IDEA 2014). Therefore, opposition players have no referendum power. In sum, the lower house of the Czech Republic nicely illustrates that a combination of very strong and very weak opposition power may lead

118  Power of Opposition Players and Consequences

to an overall intermediate score on the final index of the policy-making power of opposition players. In the case of the Czech Republic, the lower house combines very strong initiation power (i.e. bill introduction and agenda setting) with intermediate debate power (i.e. amendments, committee structure and committee procedures), and very weak veto power (i.e. executive power and referendums).

5.3 Conclusion In this chapter, I have presented the descriptive results of the newly constructed measure of the policy-making power of opposition players. The policy-­making power of opposition players varies considerably, not only across countries but also across regime types with regards to not only the separate dimensions (i.e. initiation power, debate power, and veto power) but also the overall index. Furthermore, a closer look at three different cases (Chile, the Czech Republic, and Slovenia) has shown more detail regarding possible combinations of the different variables composing the final index. The results presented in this chapter confirm the findings in the legislative studies literature that the organization of the legislative decision-­ making process varies considerably between countries. For example, legislative chambers grant the right to introduce bills to various actors, ranging from only a few to a very broad group of political players. In all legislative chambers, individual MPs and the government have the right to propose a bill. However, some chambers grant this right to a much broader group of actors, including, for example, political parties or citizens. Other countries grant the right to introduce bills to an even broader group of players. One example is Guatemala where, according to the constitution of the country, “[d]eputies to the Congress, the Executive Organism, the Supreme Court of Justice, the University of San Carlos de Guatemala, and the Supreme Electoral Tribunal have the initiative of law” (Article 174 of the Constitution of the Republic of Guatemala). Similarly, legislative chambers also vary considerably in their number of legislative committees. Data show that, among the countries considered here, the number of standing committees ranges from as few as four, such as in Mauritius, to as many as 58, as in the Philippines. The distribution across different regime types shows a slight tendency of semi-presidential regimes to grant strong policy-making power to opposition players more often than presidential and parliamentary regimes. Whereas presidential regimes show a cluster of policy-making power of opposition players ranging from 0.5 to 0.8 parliamentary regimes show a wide range of power granted to opposition players (see Figure 5.3). The large variance in policy-making power of opposition players, especially, within different regime types, clearly shows that the potential role and influence of opposition players may not be fully understood when only

Power of Opposition Players around the World  119

considering the process of government formation. These results strongly suggest that research that is interested in the role of institutional characteristics and its link to individual political support and democratic consolidation should also consider the actual rights granted to the representatives of not only winners (i.e. the government) but also losers (i.e. the opposition). This new measure of the policy-making power of opposition players marks a first step to moving beyond classical definitions of regime types and to more closely looking at the specific rights of opposition players and their influence on public support. In the next chapter, I analyze whether this variation in the level of policy-making power of opposition players actually relates to democratic consolidation from an individual perspective.

Notes 1 Among the few contributions on opposition power is Garritzmann (2017), showing that the power granted to oppositions varies considerably among parliamentary democracies. Results presented in this Chapter give further evidence to this pattern. 2 See Chapter 2 for a detailed presentation of the theoretical argument. 3 See Table A.5.1 in the Appendix for an overview of countries. 4 Due to the low number of countries, the category of Asia is broadly defined, including also Eurasian countries and Oceania. 5 Government priority is not only about absolute priority of all government business but also about rules indicating general priority of government business with exceptions and special times for opposition or private members’ business. 6 As for example the executive, individual MPs, parliamentary parties, or parliamentary committees. 7 Parliamentary minorities, citizens, or subnational entities, see Chapter 3. 8 For example, parliamentary regimes tend to grant weaker initiation and debate power and also show among the weakest overall results. 9 Table A.5.2 of the Appendix shows correlations between the separate variables of the final index. Results confirm the pattern that legislative chambers do not necessarily score high or low on all variables. 10 “The Council consists of the President and Vice-Presidents of the National Assembly, the leaders of the deputy groups, and the deputies of the national communities” (Državni zbor 2007: Article 21(2)). 11 “Article 33(2) Working bodies are composed of a chairman, as a general rule one deputy chairman, and the members. Their total number must be odd. The offices of chairman and deputy chairman of a given working body are, as a general rule, held by deputies of different deputy groups. If the chairman of a working body belongs to one of the deputy groups of the governing coalition, the office of deputy chairman will as a general rule be held by a member of an opposition deputy group, and vice versa” (Državni zbor 2007).

Bibliography Anderson, Christopher J. and Silvia M. Mendes. 2005. “Learsning to Lose: Election Outcomes, Democratic Experience and Political Protest Potential.” British Journal of Political Science 36: 91–11.

120  Power of Opposition Players and Consequences Cámara de diputadas y diputados. 2010. “Constitucion Politica de la Republica de Chile.” Online: https://www.camara.cl/camara/leyes_normas.aspx. Accessed: June 2014. Cámara de diputadas y diputados. 2014. “Reglamento de la Camara de Diputados de Chile.” Online: https://www.camara.cl/camara/leyes_normas.aspx. Accessed: June 2014. Carey, John M. 2009. Legislative Voting and Accountability. Cambridge: Cambridge University Press. Cheibub, José Antonio, Zachary Elkins and Tom Ginsburg. 2014. “Beyond Presidentialism and Parliamentarism.” British Journal of Political Science 44(3): 515–544. Döring, Herbert. 1995. Parliaments and Majority Rule in Western Europe. Frankfurt: Campus Verlag. Döring, Herbert. 2001. “Parliamentary Agenda Control and Legislative Outcomes in Western Europe.” Legislative Studies Quarterly 26(1): 145–165. Državni zbor. 1991. “Constitution of the Republic of Slovenia.” Online: https:// www.dz-rs.si/wps/portal/en/Home/AboutNA/PoliticalSystem/Constitution/!ut/p/z1/04_Sj9CPykssy0xPLMnMz0vMAf Ijo8zivSy9Hb283Q0N3E3dLQwCQ7z9g7w8nNxDnUz0w8EKDHAARwP9KGL041EQhd94L_2o9Jz8JIhTHfOSjC3S9aOKUtNSi1KL9EqLgMIZJSUFxVaqBqoG5eXleun5-ek5qXrJ-bmqBti0ZOQXl-hHoKrUL8gNjajySQ0HAOpsxSw!/dz/d5/ L2dBISEvZ0FBIS9nQSEh/. Accessed: June 2014. Državni zbor. 2007. “Rules of Procedure of the National Assembly.” Online: https://www.dz-rs.si/wps/portal/en/Home/ODrzavnemZboru/PristojnostiInFunkcije/RulesoftheProcedureText/!ut/p/z1/04_Sj9CPykssy0xPLMnMz0vMAfIjo8zinfyCTD293Q0N3MOczAw8QwJcXE0tLIwsgk31w8EKvD39gy2dDB0N_C0N3Qw8Q43dTYx9QwzcfQ30oyjQbxAK0Y_HeIh-AxzAkUj9eBwYhd_4gtxQIHBUBAAf0K_t/dz/d5/L2dBISEvZ0FBIS9nQSEh/. Accessed: June 2014. Garritzmann, Julian L. 2017. “How Much Power Do Opposition Have? Comparing the Opportunity Structures of Parliamentary Opposition in 21 Democracies.” The Journal of Legislative Studies 23(1): 1–30. IDEA. 2014. “Institute for Democracy and Electoral Assistance. Direct Democracy Database.” Online: http://www.idea.int/elections/dd/search.cfm. Accessed: January 2014. Martin, Shane. 2011. “Electoral Institutions, the Personal Vote, and Legislative Organization.” Legislative Studies Quarterly 36(3): 339–361. Martin, Shane. 2018. Electoral Systems and Legislative Organization. In The Oxford Handbook of Electoral Systems, eds. Erik S. Herron, Robert J. Pekkanen, and Matthew S. Shugart. New York, NY: Oxford University Press: 303–319. Mattson, Ingvar. 1995. Private Members Initiatives and Amendments. In Parliaments and Majority Rule in Western Europe, ed. Herbert Döring. New York: St. Martins: 448–487. McCubbins, Mathew D. 2005. Legislative Process and the Mirroring Principle. In Handbook of New Institutional Economics, eds. Claude Menard and Mary M. Shirley. Dordrecht: Springer Verlag: 123–147. Müller, Wolfgang C. and Ulrich Sieberer. 2014. Procedures and Rules in Legislatures. In The Oxford Handbook of Legislative Studies, ed. Shane Martin,

Power of Opposition Players around the World  121 Thomas Saalfeld and Kaare W. Strøm. Oxford, New York: Oxford University Press: 312–331. Poslanecká sněmovna. 2006. “Rules of Procedure of the Chamber of Deputies.” Online: https://public.psp.cz/en/docs/laws/1995/90_index.html. Accessed: June 2014. Powell, G. Bingham. 2000. Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven & London: Yale University Press. Regan, Patrick and Dave Clark. 2014. “The Institutions and Elections Project.” Online: http://www2.binghamton.edu/political-science/institutions-and-elections-project.html. Accessed: January 2014. Shugart, Matthew S. 2006. Comparative Executive-Legislative Relations. In The Oxford Handbook of Political Institutions, eds. R.A.W. Rhodes, Sarah A. Binder and Bert A. Rockman. Oxford and New York: Oxford University Press: 344–365. Sieberer, Ulrich. 2011. “The Institutional Power of Western European Parliaments: A Multidimensional Analysis.” West European Politics 34(4): 731–754. Sieberer, Ulrich, Wolfgang C. Müller and Maiko Isabelle Heller. 2011. “Reforming the Rules of the Parliamentary Game: Measuring and Explaining Changes in Parliamentary Rules in Austria, Germany, and Switzerland, 1945–2010.” West European Politics 34(5): 948–975. Strøm, Kaare. 1990. Minority Government and Majority Rule. New York: Cambridge University Press. Svolik, Milan W. 2013. “Learning to Love Democracy: Electoral Accountability and the Success of Democracy.” American Journal of Political Science 57(3): 685–702. Taylor, Andrew J. 2006. “Size, Power, and Electoral Systems: Exogenous Determinants of Legislative Procedural Choice.” Legislative Studies Quarterly 31(3): 323–345. Wawro, Gregory J. and Eric Schickler. 2006. Filibuster. Obstruction and Law Making in the U.S. Senate. Princeton and Oxford: Princeton University Press. Wegmann, Simone. 2022. “Policy-Making Power of Opposition Players: A Comparative Institutional Perspective.” The Journal of Legislative Studies 28(1): 1–25. Zubek, Radoslaw. 2015. “Coalition Government and Committee Power.” West European Politics 38(5): 1020–1041.

Chapter 6

Policy-Making Power of Opposition Players and Democratic Consolidation

Democratic consolidation is inherently about the views and preferences of individuals. Research in this area has repeatedly highlighted the importance of individual political support as a crucial element for successful democracy (see, e.g. Claassen 2019; Welzel and Inglehart 2009; Diamond 1999; Lipset 1959). The central question in consolidation studies is why losers “(…) comply with the outcomes [of elections] and continue to participate rather than subvert democratic institutions” (Przeworski 1991: 15). However, the legislature has not figured prominently among the potential institutional determinants of such individual political support. But, if consolidation is inherently about citizen attitudes and opinions – and democratic consent of losers is especially important for democratic consolidation – focusing on the role of the institution that actually represents citizens (i.e. the legislature) seems imperative. The theoretical framework I presented in Chapter 2 suggests that the policy-making power of opposition players should influence both winners and losers of democratic elections. However, I expect strong policy-­making power of opposition players to have different effects on the diffuse and specific support of winners and losers. From a procedural perspective, more political equality (guaranteed by strong policy-making power of opposition players) may be associated with more democratic representation. Therefore, I expect strong policy-making power of opposition players to be linked to more diffuse support from both winners and losers alike. In contrast, from an outcome-oriented perspective, strong policy-making power of opposition players naturally gives them more influence in the decision-making process. This is beneficial for losers, but not for winners. Thus, I expect strong policy-making power of opposition players to be linked to more specific support from losers and less specific support from winners. The empirical analyses in this chapter are structured into three parts following a continuum from more diffuse to more specific support. In the first section, I look at the influence of different levels of policy-making power of opposition players on individual opinions about democracy DOI: 10.4324/9781003005360-8

Power of Opposition Players and Democratic Consolidation  123

as the best form of government. In the second section, I analyze the influence of the policy-making power of opposition players on individual satisfaction with democracy. In the third section, I present evidence of the relationship between policy-making power of opposition players and individual satisfaction with government performance. I conclude and present the overall results in the final section.

6.1  Democracy as the Best Form of Government The theoretical framework I presented in Chapter 2 suggests that more inclusive parliamentary procedures (measured through policy-making power of opposition players) increases political equality between winners and losers. More political equality, in turn, influences the level of diffuse support from both winners and losers alike. Based on Easton (1975) and Norris (2011) diffuse support refers to the satisfaction with the basic principles upon which a political regime is based. From the perspective of both winners and losers alike, more political equality (guaranteed by strong policy-making power of opposition players) may be associated with more democratic representation, thereby increasing the satisfaction with the basic principles of democracy (i.e. diffuse support). To understand this potential effect of opposition power on diffuse support of winners and losers, I first look at the link between policy-making power of opposition players and individual opinions about democracy as the best form of government. Due to data availability, this first analysis is based on the LAPOP survey (Vanderbilt University 2019) only.1 Overall, a vast majority of respondents think that democracy is the best form of government (87.6 percent), and less than 50 percent of respondents identify as electoral losers (41.7 percent).2 Policy-making power of opposition players shows a mean score of 0.60. Trinidad and Tobago grants the weakest policy-making power to opposition players, whereas Panama shows the strongest policy-making power granted to opposition players among the countries considered in the survey. Furthermore, the models in this chapter include several individual-level and country-level control variables. The individual-level control variables show that most of the respondents in the survey report their economic family situation (household income) to be good enough (52.6 percent). Whereas a vast majority of respondents is satisfied with their life (87.9 percent), only about one-third of the respondents indicate to be close to a political party (36.5 percent) and to be interested in politics (36.8 percent). Decentralization, which is the combination of strong federalism and bicameralism, shows a mode of 0 among the countries in the LAPOP (Vanderbilt University 2019). A closer look at the political regime shows that most countries are presidential regimes, whereas only two out of the ten countries in the sample are parliamentary. None of the countries

124  Power of Opposition Players and Consequences

in the LAPOP (Vanderbilt University 2019) are semi-presidential. Furthermore, only the two parliamentary regimes have majoritarian electoral formulas whereas all remaining countries use proportional or mixed systems. Due to this distribution of the two variables, political regime and PR electoral system are combined into an index of political system for a further robustness check. This index takes a value of 1 for all parliamentary systems or for countries with a non-PR electoral formula, and 0 otherwise. This coding allows for the identification of countries deviating from the predominant institutions of presidentialism and PR electoral formula in the region. Three additional country-level control variables are added in the analyses. The GDP/capita variable is recoded into a three-point scale grouping individuals living in countries with low, intermediate, and high GDP/capita. Corruption also varies considerably from 0.11 in Chile to 0.75 in Guatemala. The mean respondent lives in a country with a corruption score of 0.31 (best represented by Argentina with the same score). Finally, on average, respondents in the Latin American Public Opinion Project (Vanderbilt University 2019) live in a country with 29 years of democratic experience. Table 6.1 shows two random intercept probit models on democracy as the best form of government as binary dependent variable based on data from the LAPOP (Vanderbilt University 2019). The models include the individual-level variable loser and several individual-level control variables. According to the theoretical framework, I expect a positive relationship between the policy-making power of opposition players and diffuse support for both winners and losers. To test for these effects, both models include interactions between policy-making power of opposition players (PPOP) and losers as well as winners, respectively. Furthermore, model 2 includes an additional interaction with the democratic experience variable to model possible differences in the effects of strong policy-making power of opposition players in older and younger democracies. The cross-level interactions show the effect of the policy-making power of opposition players (PPOP) on diffuse support for winners and losers, respectively. Model 1 shows positive effects of policy-making power of opposition players for both winners and losers (in line with expectations). This would indicate that individuals – winners and losers alike – living in polities that grant strong policy-making power to opposition players are more likely to think that democracy is the best form of government, but the results are statistically not significant. However, when including democratic experience as further interaction in Model 2, results show a positive and statistically significant effect of policy-making power of opposition players for both winners and losers alike. Thus, it seems that the effect of strong policy-making power of opposition players does unfold only with a certain degree of democratic experience and is not something very young democracies benefit from.

Power of Opposition Players and Democratic Consolidation  125 Table 6.1  D emocracy as Best Form of Government Model 1 Individual level Loser Household income (good enough) Close to political party Interest in politics Satisfaction with life Country level Electoral formula (PR) Decentralization (strong fed. and bic.) Political regime (presidentialism) GDP/capita (intermediate) GDP/capita (high) Corruption Cross-level interaction Winner*PPOP Loser*PPOP Winner*PPOP*democratic experience

−0.15 (0.21) 0.09** (0.03) 0.09** (0.03) 0.02 (0.03) 0.05 (0.04)

0.17 −0.17 0.09** (0.03) 0.09** (0.03) 0.01 (0.03) 0.05 (0.04)

−0.14 (0.13) −0.06 (0.15) 0.07 (0.28) 0.08 (0.17) 0.23 (0.18) −0.22 (0.29)

−0.16** (0.06) 0.06 (0.07) 0.30** (0.10) −0.01 (0.07) 0.22** (0.08) −0.25+ (0.15)

0.39 (1.01) 0.49 (1.01)

Loser*PPOP*democratic experience Intercept Log. Likelihood N N groups

Model 2

0.90 (0.57) −4724.45 12,988 10

0.01*** (0.00) 0.02*** (0.00) 0.77*** (0.12) −4784.34 12,988 10

*** p < 0.001; ** p < 0.01; * p < 0.05; + p < 0.1 Random intercept probit model. Binary dependent variable: democracy is better than any other form of government, Latin America Public Opinion Project (Vanderbilt University 2019).

126  Power of Opposition Players and Consequences

A closer look at the individual-level and country-level control variables shows statistically significant results only for household income and being close to a political party in both models as well as PR electoral formula, presidential regimes, high GDP/capita, and corruption in model 2. Hence, individuals living in good economic conditions (both individually and on the aggregate level) are more likely to think that democracy is the best form of government. Similarly, individuals that report being close to a political party as well as those living in presidential regimes and under non-PR electoral formulas are more likely to believe that democracy is the best form of government. As presented above, the countries in the survey show a particular distribution of presidential regimes and PR electoral formula as the only two parliamentary regimes are also the only countries using a majoritarian electoral formula. Robustness checks using the political system variable (combining non-PR electoral formula and parliamentary systems) show results consistent with the ones presented in Table 6.1 as do models including only the PR electoral system and decentralization variables. To illustrate these results, Figure 6.1 shows predicted probabilities of thinking democracy is the best form of government based on results of Model 2 in Table 6.1. The figure shows probabilities dependent on the years of democratic experience for a setting with strong (score of 0.9) and a setting with weak (score of 0.2) policy-making power of opposition players for both winners (right panel) and losers (left panel). Control variables are held constant at their median or mean.3 I follow King et al. (2010) and use a simulation-based approach to illustrate results in which each line in Figure 6.1 represents the predicted probability based on a simulated set of model estimates. Hence, the density of lines represents the uncertainty associated with the estimated probabilities. In addition, 95 percent confidence intervals are shown in dashed lines. Overall, Figure 6.1 shows that for winners and losers alike, policy­making power of opposition players has a positive effect. Importantly, individuals living in countries with strong policy-making power of opposition players (black lines) have higher probabilities of thinking that democracy is the best form of government than individuals living in countries with weak policy-making power of opposition players (grey lines). But, Figure 6.1 also shows an important and positive effect of democratic experience. The probability of thinking that democracy is the best form of government considerably increases with the years of democratic experience in all settings. Overall, individuals show a rather high probability of thinking that democracy is the best form of government. Nevertheless, results show a much more pronounced effect of strong policy-making power of opposition players (black lines) than of weak policy-making power of opposition players (grey lines) over time, especially for losers (left panel of Figure 6.1).

Power of Opposition Players and Democratic Consolidation  127

Winners

0.5

0.5

0.6

0.7

0.8

0.9

Pr (Democracy is preferable) 0.6 0.7 0.8 0.9

1.0

1.0

Losers

0

10 20 30 40 50 60 70 Strong PPOP (0.9) Weak PPOP (0.2)

0

10 20 30 40 50 60 70

Democratic experience in years

Figure 6.1  Predicted Probabilities: Democracy as Best Form of Government

More specifically, losers (left panel of Figure 6.1) show a very similar probability of thinking that democracy is the best form of government if they live in young democracies regardless of whether they live in a country granting weak or strong policy-making power of opposition players: 64.1 percent compared to 64.6 percent, respectively. With more years of democratic experience, these probabilities rise considerably. However, they rise much more in countries with strong policy-making power of opposition players (dark lines). A loser living in a country with weak policy-making power of opposition players has a 72.6 percent probability of thinking that democracy is the best form of government (corresponding to an increase of 8.5 percentage points) if living in a country with 70 years of democratic experience, all else equal. This change, however, is not statistically significant. On the contrary, a loser living in a country with 70 years of democratic experience which grants strong policy-making power to opposition players has a probability of 92.8 percent (corresponding to an increase of 28.2 percentage points) of thinking that democracy is the best form of government, all else equal. The general pattern for winners is similar, however, the effect of policy­making power of opposition players is somewhat smaller. Again, winners in young democracies show similar probabilities of thinking that democracy is the best form of government, all else equal. With more democratic experience, this probability increases, but is only statistically significant

128  Power of Opposition Players and Consequences

for winners living in countries with strong policy-making power of opposition players (black lines, right-hand panel of Figure 6.1). A winner living in a country with only one year of democratic experience and strong policy-making power of opposition players shows a probability of 70.4 percent of thinking that democracy is the best form of government, all else equal. This probability increases to 90.6 percent for a winner living in a country with 70 years of democratic experience and strong policy-­making power of opposition players (corresponding to an increase of 20.2 percentage points). Figure 6.1 also shows a slightly positive effect for winners in countries with weak policy-making power of opposition players (grey lines) over time, however, this effect is not statistically significant. Consequently, individuals living in countries with strong policy­making power of opposition players and extensive democratic experience are significantly more likely to think that democracy is the best form of government than individuals living in countries with weak policy-­making power of opposition players, all else equal. Hence, not only democratic experience but also strong policy-making power of opposition players seems to increase diffuse support. However, the pattern is more pronounced for losers than for winners. As the left-hand panel in Figure 6.1 shows, the effect of policy-making power of opposition players becomes statistically significant for losers of democratic elections after 29 years of democratic experience. We see a similar effect for winners only after 45 years of democratic experience, respectively (right-hand panel in Figure 6.1). Based on these results, Figure 6.2 shows the change in the winner-loser gap for settings with weak and strong policy-making power of opposition players, respectively. Overall, Figure 6.2 shows a decrease in the winner-­ loser gap (i.e. a development closer to 0) with more democratic experience. This result illustrates the fact again that strong policy-making power of opposition players seems to unfold its effect only after a certain level of democratic experience. But, not only does democratic experience show effects, but the winner-loser gap is reversed in settings with strong policy-making power of opposition players (dashed lines), indicating that losers have a higher probability of expressing diffuse support than winners (indicated by a negative winner-loser gap). This pattern does not arise in countries with weak policy-making power of opposition players (continuous lines). However, none of these predictions are statistically significant. Hence, the results do not indicate the existence of a large winner-­loser gap in the first place. In sum, strong policy-making power of opposition players has a positive effect on diffuse support of both winners and losers alike. Yet, it seems to more positively influence losers than winners. This finding generally comports with the theoretical expectations presented in Chapter 2,

Winner−Loser Gap (Pr(winner)−Pr(loser)) −0.05 0.00 0.05 0.10

Power of Opposition Players and Democratic Consolidation  129

Weak PPOP (0.2) Strong PPOP (0.9)

0

10

20 30 40 50 Democratic experience in years

60

70

Figure 6.2  Winner-Loser Gap: Democracy as Best Form of Government

even though I did not expect a stronger effect for losers than for winners. Consequently, the data do point toward a faster reduction (i.e. after a shorter time of democratic experience) in the winner-loser gap in countries with strong policy-making power of opposition players. However, overall, the results do not point to a statistically significant gap in the level of political support between winners and losers in the first place. Rather, the analysis points to the existence of a gap of policy-making power of opposition players developing with more democratic experience. Losers and winners alike show similar probabilities of thinking that democracy is the best form of government in young democracies no matter whether they live in a polity granting weak or strong policy-making power to opposition players. This pattern changes with more democratic experience as both winners and losers are more likely to think that democracy is the best form of government when living in countries that grant strong policy-making power to opposition players. The pattern is more pronounced for losers, but does apply also to winners (see Figure 6.1). Thus, it seems that policy-making power of opposition players does contribute to more diffuse support of both winners and losers, alike. This result is in line with the theoretical expectations presented in Chapter 2.

6.2  Satisfaction with Democracy The analyses in the previous section confirm the theoretical expectation that strong policy-making power of opposition players is linked to more

130  Power of Opposition Players and Consequences

diffuse support of both winners and losers. However, the theoretical framework I presented in Chapter 2 suggests a different relationship regarding specific support. Based on Easton (1975) and Norris (2011), specific support refers to the satisfaction with a particular outcome of a given regime. From the perspective of both winners and losers alike, more political equality (guaranteed by strong policy-making power of opposition players) may be associated with more democratic representation, thereby increasing the satisfaction with the basic principles of democracy (i.e. diffuse support) but not necessarily increasing the satisfaction with the outcome of a democracy (i.e. specific support). Hence, from a more outcome-oriented perspective, stronger policy-making power of opposition players (i.e. more influence in the decision-making process) is beneficial to losers, but not necessarily to winners. Therefore, I expect strong policy-making power of opposition players to be linked to higher levels of specific support from losers, but less support from winners. To understand this potential effect of opposition power on specific support of winners and losers, I first look at the link between policy-making power of opposition players and individual satisfaction with democracy. As discussed in more detail in Chapter 4, it is not entirely clear whether satisfaction with democracy more closely measures diffuse or specific support. However, a brief look at the data revealed that the satisfaction with democracy variable shows higher correlations with the satisfaction with government performance variable than with the democracy is the best form of government variable, indicating a closer link to specific support (see Chapter 4 for more details). This second set of analyses is based on three different surveys: The LAPOP (Vanderbilt University 2019), the CSES (2019), and the ESS (2019).4 Overall, a majority of respondents indicate that they are satisfied with the way democracy works in their country. The policy-making power of opposition players shows mean scores between 0.6 (LAPOP) and 0.67 (ESS). The individual-level control variables in the LAPOP (Vanderbilt University 2019) show similar distributions to the ones based on the dependent variable of democracy as best form of government (see Table A.6.1). Most of the respondents report their economic family situation (household income) to be good enough (54.0 percent). Whereas a vast majority of respondents are satisfied with their life (88.0 percent), only about onethird of the respondents report being close to a political party (36.5 percent) and being interested in politics (39.0 percent). Descriptive statistics of the CSES (2019) and the ESS (2019) also show similar distributions. Only a minority of respondents fall into the category of low education but most of them show at least intermediate household income. Contrary to respondents in the LAPOP (Vanderbilt University 2019), however, a majority of respondents in the ESS (2019) identify as being close to a political party and are interested in politics.

Power of Opposition Players and Democratic Consolidation  131

Similarly, the country-level variables in the LAPOP (Vanderbilt University 2019) have similar distributions to the ones based on the dependent variable democracy as best form of government (see Table A.6.1). However, regime type (i.e. presidentialism) is strongly correlated with policy-making power of opposition players (0.61), corruption (0.58), and GDP/capita (0.67) leading to problems of multicollinearity. Therefore, this second set of analyses considers the combined variable of political system scoring 1 for all countries that are parliamentary and/or use a majoritarian electoral formula in the main analysis. Most individuals included in the CSES (2019) and the ESS (2019) live in centralized parliamentary regimes with PR electoral formula. GDP/capita also show distributions very similar to the first set of analyses. GDP/capita shows a median of two in all three surveys. On the contrary, corruption shows some significant variance between the three surveys (with a mean score of 0.13 (ESS), 0.26 (CSES), and 0.49 (LAPOP), respectively). Finally, on average, democracies are younger in the LAPOP survey (mean of 15.8 years) than in the CSES (mean of 63.9 years) and ESS (mean of 53.1 years). Table 6.2 shows two random intercept probit models on satisfaction with democracy as a binary dependent variable using data from all three surveys. According to the theoretical framework, I expect opposite effects of policy-making power of opposition players for losers and winners.5 Policy-making power of opposition players is theorized to be especially important for losers. Thus, losers should be more likely to be satisfied with democracy if living in a country with strong policy-making power of opposition players. The same is not necessarily true for winners. Here, I expect a negative relationship between winners and political support (i.e. satisfaction with democracy). Consequently, these expectations postulate that strong policy-making power of opposition players also leads to a reduction in the gap of political support between winners and losers. As the results in the previous section indicated for diffuse support, the effect of policy-making power of opposition players seems to unfold only after a certain level of democratic experience. To account for this effect, all models include interactions to show the effect of policy-making power of opposition players for losers and winners dependent on the level of democratic experience, respectively. In line with previous research, all models show that being a loser6 has a negative and statistically significant effect on the probability of being satisfied with democracy. Again, however, the cross-level interactions are of most importance to understand the effect of policy-making power of opposition players on specific support. Overall, results only partially support the theoretical expectations. Whereas results show mostly positive and statistically significant results for policy-making power of opposition players for losers (in line with expectations), results are mixed for winners. No consistent pattern showing that strong policy-making power of opposition

GDP/capita (intermediate) GDP/capita (high)

Political system (parl. and/or non-PR) Decentralization

Country level Political regime (semi-pres.) Political regime (presidential) PR electoral system

Satisfaction with life

Interest in politics

Close to party

Household income (intermediate) Household income (high/good enough) Low education

Individual level Loser

−0.33* ** (0.07) 0.18*** (0.02) 0.07* * (0.02)

−0.11*** (0.02) 0.23** * (0.02) 0.15*** (0.02) 0.63* ** (0.02)

−0.03 (0.14) 0.84* * (0.31) 0.02 (0.14)

−0.33 (0.21) 0.18* (0.08) 0.30*** (0.08)

−0.09*** (0.02) 0.18* ** (0.02)

0.17 (0.30) 0.04 (0.43) −0.16 (0.35)

0.56 (0.48) 0.05 (0.05) −0.46 (0.33)

ESS M1

−0.91*** (0.07) 0.09*** (0.02) 0.19*** (0.02)

CSES M1

Table 6.2  S atisfaction with Democracy

0.04 (0.09) −0.13 (0.09) −0.24** (0.07)

0.11+ (0.06)

0.14*** (0.02) 0.06* * (0.02) 0.05* (0.02) 0.35*** (0.03)

0.19*** (0.02)

−0.19 (0.10)

L APOP M1

0.56 (0.51) −0.04 (0.05) −1.46*** (0.41)

0.17 (0.32) −0.29 (0.46) −0.31 (0.35)

−0.09*** (0.02) 0.18*** (0.02)

−0.25*** (0.02) 0.09*** (0.02) 0.19*** (0.02)

CSES M2

−0.19 (0.20) 0.14 (0.07) 0.25* * (0.08)

0.01 (0.139 0.06 (0.44) 0.04 (0.13)

−0.10*** (0.02) 0.23*** (0.02) 0.15*** (0.02) 0.63*** (0.02)

−0.45*** (0.03) 0.18*** (0.02) 0.07* * (0.02)

ESS M2

−0.42 (0.81) −0.79 (0.73) −0.52 (0.71)

0.32 (0.60)

0.13*** (0.02) 0.05* * (0.02) 0.06* * (0.02) 0.35*** (0.03)

0.19*** (0.02)

−0.47*** (0.05)

L APOP M2

0.50 (0.29) −0.15* (0.07) −0.25 (0.21)

0.16 (0.19) −0.19 (0.27) −0.01 (0.21)

−0.10*** (0.02) 0.18*** (0.02)

−0.69*** (0.07) 0.09*** (0.02) 0.19*** (0.02)

CSES M3

−0.35 (0.21) 0.12 (0.08) 0.25* * (0.09)

−0.07 (0.14) 0.86* * (0.31) −0.00 (0.14)

−0.11* * * (0.02) 0.23* * * (0.02) 0.15* * * (0.02) 0.63* * * (0.02)

−0.29* * * (0.08) 0.18* * * (0.02) 0.07* * (0.02)

ESS M3

0.06 (0.08) −0.16* (0.08) −0.25* * * (0.07)

0.09 (0.06)

0.14* * * (0.02) 0.06* * (0.02) 0.05* (0.02) 0.35* * * (0.03)

0.18* * * (0.02)

−0.19 (0.10)

L APOP M3

−17431.83 30,721 18

−1.18** * (0.35)

1.75 (1.36)

−21266.41 37,141 26

−1.30* * (0.41) 0.28 (0.44) 0.29 (0.44)

−5.05* * * (0.38) −0.80 (1.78) 0.19 (1.77)

*** p < 0.001; ** p < 0.01; * p < 0.05; + p < 0.1

Log Likelihood N N countries

Winner *Democratic experience*PPOP Loser *Democratic experience*PPOP Winner *Old democracy*PPOP Loser *Old democracy*PPOP Winner *Young democracy*PPOP Loser *Young democracy*PPOP Intercept

Loser *PPOP

Winner *PPOP

Corruption

−14263.97 22 ,297 11

−0.45* (0.21)

−0.86*** (0.16) 1.43*** (0.35) 0.66* (0.33)

−21199.56 37,141 26

0.94* (0.37)

0.02* * * (0.01) 0.02* * * (0.01)

−4.64* * * (0.36)

−17410.56 30,721 18

−1.29*** (0.22)

0.01* (0.00) 0.01* * (0.00)

−0.77 (0.45)

−14147.09 22 ,297 11

3.62*** (0.76)

−0.11*** (0.01) −0.11*** (0.01)

−2.77*** (0.42)

−21021.56 37,141 26

−0.48 (0.48) −0.07 (0.47) −1.10* (0.48) 0.14 (0.48) 1.27* * (0.39)

−3.44*** (0.45)

17426.35 30,721 18

0.21 (0.45) 0.15 (0.45) 0.25 (0.45) 0.39 (0.45) −1.03* * (0.36)

−1.55* * * (0.43)

−14244.93 22 ,297 11

1.49* * * (0.33) 0.61* (0.31) 1.13* * (0.35) 0.66* (0.33) −0.43* (0.19)

−0.79* * * (0.15)

134  Power of Opposition Players and Consequences

players reduces specific support of winners exists. On the contrary, in some models (CSES M2, ESS M2 and LAPOP M1 and M3), the results show significant and positive effects for winners and losers alike. Therefore, either strong policy-making power of opposition players does not necessarily have a negative effect on specific support for winners as measured by satisfaction with democracy, or the question actually captures diffuse rather than specific support (explaining the positive effect for winners and losers in line with theoretical expectations about diffuse support). A closer look at the individual-level control variables shows statistically significant results mostly in line with expectations. Household income has a positive and statistically significant coefficient in all models. Hence, individuals indicating that their household income is intermediate or high are more likely than individuals with a low household income to also report being satisfied with democracy. Except for results based on LAPOP, respondents with low education are less likely to be satisfied with the way democracy works in their country. The data furthermore show statistically significant coefficients consistent with expectations for the interest in politics, close to party, and satisfaction with life variables. Individuals that report being interested in politics and close to a political party are more likely to be satisfied with democracy. Similarly, respondents that are more satisfied with their life are also more likely to indicate being satisfied with democracy. On the country level, Models ESS M1 and M3 show positive and statistically significant results for presidential regimes. Hence, individuals living in presidential regimes are more likely to be satisfied with democracy compared to individuals living in parliamentary regimes. Switzerland, however, is the only country in the ESS survey categorized as a presidential regime and it is also the country with the most democratic experience in the sample. Dropping Switzerland from the sample for robustness checks shows consistent results for all variables, most notably for the positive and statistically significant effect of policy-making power of opposition players for both winners and losers in Model 2. Moreover, the results show statistically significant coefficients for GDP/capita and corruption in all models. The coefficients for GDP/capita, however, do not point in a clear direction. Whereas respondents are more likely to be satisfied with democracy in countries with intermediate or high GDP/ capita based on the ESS survey, they are less likely to be satisfied based on results of the CSES and the LAPOP survey. Finally, the measure for corruption has a statistically significant coefficient in the expected direction in all models. Hence, more corruption relates to a lower probability of being satisfied with democracy. A further robustness check was conducted for the ESS (2019) based on the original variable measuring satisfaction with democracy without recoding. These linear models show consistent results with the ones in

Power of Opposition Players and Democratic Consolidation  135

Table 6.2. However, results are statistically only significant for losers in young democracies. For the LAPOP survey, alternative model specifications including only PR electoral formula and decentralization as control variables show consistent results with the ones presented here. To illustrate these results, I rely again on predicted probabilities of being satisfied with democracy. Following King et al. (2010), the figures show a simulation-based approach. Each line represents the predicted probability based on a simulated set of model estimates. Thus, the density of lines again represents the uncertainty associated with the estimated probabilities. In addition, 95 percent confidence intervals are shown in dashed lines. Figure 6.3 shows predicted probabilities of being satisfied with democracy based on data from the CSES (2019). Predicted probabilities are based on results of Model CSES M2 in Table 6.2. Again, the effect of policy-making power of opposition players on satisfaction with democracy is shown for winners (right panel) and losers (left panel) living in countries with weak policy-making power of opposition players (grey lines) and strong policy-making power of opposition players (black lines) dependent on the level of democratic experience.7 Results in Figure 6.3 show a clear positive effect of the policy-making power of opposition players on the level of satisfaction with democracy for both winners and losers. The probability of being satisfied with democracy is higher for individuals living in countries with strong policy-making power of opposition players (i.e. a score of 0.9, in black) than for individuals living in countries with weak policy-making power of opposition players (i.e. a score of 0.2, in grey). However, Figure 6.3 again shows an interesting effect of democratic experience. This analysis shows a positive effect of the years of democratic experience on the probability of being satisfied with democracy for winners and losers. Losers (left panel) in a new democracy show a rather low probability of being satisfied with democracy independent of the level of policy-making power granted to opposition players. The probability is slightly higher, but still low for winners (right panel). This probability significantly increases over time, but does so much faster for winners and losers living in countries with strong-policy-making power of opposition players (black lines) than for winners and losers living in countries granting only weak policy-making power to opposition players (grey lines). More specifically, whereas a loser living in a country with weak policy-making power of opposition players (grey lines) and only one year of democratic experience has a probability of only 30.8 percent of being satisfied with democracy, all else equal, this probability rises to 61.9 percent after 170 years of democratic experience (an increase of 31.1 percentage points). The situation is different for individuals living in countries with strong policy-making power of opposition players (black lines). Here, a loser living in a country with only one year of democratic

136  Power of Opposition Players and Consequences

Winner 1.0 0.8

0.8

0.4

0.6

0.6

0.2

0.4

0.0

0.2 0.0

Pr (Satisfaction with Democracy)

1.0

Loser

0

50 Strong PPOP (0.9) Weak PPOP (0.2)

100

150

0

50

100

150

Democratic experience in years

Figure 6.3  Predicted Probabilities: Satisfaction with Democracy. CSES

experience shows a probability of 31.9 percent of being satisfied with democracy, all else equal. This probability rises to 99.9 percent after 170 years of democratic experience (and reaches 99.0 percent after 126 years), all else equal, corresponding to an increase of 68.0 percentage points. Predicted probabilities for winners show extremely similar patterns. A winner living in a country with only weak policy-making power of opposition players and no democratic experience has a probability of 40.5 percent of being satisfied with democracy, all else equal. This probability increases to 72.1 percent after 170 years of democratic experience, all else equal (corresponding to an increase of 31.6 percentage points). The pattern is more pronounced again for strong policy-making power of opposition players (black lines). The effect of strong policy-making power of opposition players increases by 59.1 percentage points: from 40.8 percent for an individual living in a country without democratic experience to 99.9 percent after 170 years of democratic experience (and reaches 99.0 percent after 112 years), all else equal. Hence, as Figure 6.3 shows, despite the positive influence of the years of democratic experience in both settings (weak and strong policy-making power of opposition players), the effect of strong policy-making power of opposition players is much more striking compared to the effect of weak policy-making power of opposition players.

Power of Opposition Players and Democratic Consolidation  137

Winner−Loser Gap (Pr(winner)−Pr(loser)) −0.05 0.00 0.05 0.10 0.15

Consequently, it seems that strong policy-making power of opposition players unfolds its potential only with a certain degree of democratic experience. As Figure 6.3 shows, this effect happens at similar points in time for both winners and losers alike. Losers are significantly more likely to be satisfied with democracy if living in a country with strong policy-making power of opposition players compared to living in a country with weak policy-making power of opposition players after 47 years of democratic experience, all else equal. The same effect happens only three year later, after 50 years of democratic experience, for winners, all else equal. Figure 6.4 shows the winner-loser gap for both countries with weak (solid line) and with strong (dashed line) policy-making power of opposition players. A closer look at the winner-loser gap in countries with weak policy-making power of opposition players (solid line) shows a rather continuous pattern. Thus, in countries with weak policy-making power of opposition players, more democratic experience does not contribute to a reduction in the winner-loser gap. In contrast, the winner-loser gap in countries with strong policy-making power of opposition players is very similar in young democracies but starts to decline with more democratic experience. However, neither of these winner-loser gaps is statistically significant. Hence, the results seem to confirm a gap of policy-making power of opposition players rather than a strong winner-loser gap. Models using data from the ESS survey confirm these results. Figure 6.5 shows predicted probabilities of being satisfied with democracy based

weak PPOP (0.2) strong PPOP (0.9) 0

50 100 Democratic experience in years

150

Figure 6.4  Winner-Loser Gap: Satisfaction with Democracy. CSES

138  Power of Opposition Players and Consequences

Winner

0.0

0.0

0.2

0.4

0.6

0.8

Pr (Satisfaction with Democracy) 0.2 0.4 0.6 0.8

1.0

1.0

Loser

0

50 Strong PPOP (0.9) Weak PPOP (0.2)

100

150

0

50

100

150

Democratic experience in years

Figure 6.5  Predicted Probabilities: Satisfaction with Democracy. ESS

on data from the European Social Survey (ESS 2019).8 Predicted probabilities are based on results of Model ESS M2 in Table 6.2. Again, the effect of policy-making power of opposition players on satisfaction with democracy is shown for winners (right panel) and losers (left panel) living in countries with weak policy-making power of opposition players (grey lines) and strong policy-making power of opposition players (black lines) dependent on the level of democratic experience. Results in Figure 6.5 show a clear positive effect of strong policy-making power of opposition players on the level of satisfaction with democracy for both winners (right panel) and losers (left panel). These results also show a positive effect of the years of democratic experience on the probability of being satisfied with democracy especially for individuals living in countries with strong policy-making power of opposition players. More specifically, losers living in a country with only one year of democratic experience have a probability of 28.9 percent of being satisfied with democracy, regardless whether strong policy-making power is granted to opposition players, all else equal. But, this probability increases to 86.8 percent after 170 years of democratic experience if strong policy-making power is granted to opposition players and only to 42.5 percent if weak policy-making power is granted to opposition players, all else equal. The pattern, though similar, is less pronounced, for winners. For example, winners living in a country with only one year of democratic experience

Power of Opposition Players and Democratic Consolidation  139

have a probability of 45.8 percent of being satisfied with democracy, all else equal. This probability increases to 83.6 percent after 170 years of democratic experience when strong policy-making power is granted to opposition players, all else equal. Overall, Figure 6.5 shows that strong policy-making power of opposition players significantly increases the probability of being satisfied with democracy over time which is not the case for weak policy-making power of opposition players. Furthermore, Figure 6.6 shows the winner-loser gap for both countries with weak (left panel) and with strong (right panel) policy-making power of opposition players. As results show, there is a statistically significant winner-loser gap in countries with weak policy-making power of opposition players (left panel). The same is not the case in countries with strong policy-making power of opposition players (right panel). Here, results show a significant winner loser gap only in younger democracies. After 55 years of democratic experience, the winner-loser gap is not statistically significant anymore, meaning that winners are not significantly more likely to be satisfied with democracy than losers are. Overall, the pattern resembles the one based on the CSES data. On the one hand, in countries with weak policy-making power of opposition players, democratic experience does not contribute to a reduction in the winner-loser gap. On the other hand, strong policy-making power of opposition players does reduce the winner-loser gap (and ultimately even reverses it) with more democratic experience. Hence, strong policy-making power of

Strong PPOP

−0.4

0.0

0.2

0.4

0.6

0.8

Winner−Loser Gap (Pr(winner)−Pr(loser)) −0.4 0.0 0.2 0.4 0.6 0.8 1.0

1.0

Weak PPOP

0

50

100 150 0 50 Democratic experience in years

100

Figure 6.6  Winner-Loser Gap: Satisfaction with Democracy. ESS

150

140  Power of Opposition Players and Consequences

opposition players seems to not only reduce the winner-loser gap but also to increase the level of satisfaction with democracy in older democracies. Finally, Figure 6.7 shows predicted probabilities of being satisfied with democracy based on the LAPOP data.9 Overall, Figure 6.7 clearly shows that for a loser (grey lines) and a winner (dark lines) alike, policy-making power of opposition players has a positive effect on the probability of being satisfied with democracy. However, this effect is only statistically significant for winners (dark lines). More specifically, a winner living in an old democracy (dark lines in left panel of Figure 6.7) granting only weak policy-making power to opposition players has an 84.7 percent probability of being satisfied with democracy. This probability rises to 99.8 percent for the same individual living in a country with strong policy-making power of opposition players, all else equal. We see a similar pattern for winners in young democracies (dark lines in the right panel of Figure 6.7) that grant only weak policy-making power to opposition players, with a 75.2 percent probability of being satisfied with democracy. This probability rises to 96.5 percent for an individual living in a young democracy that grants strong policy-making power to opposition players, all else equal. This corresponds to an increase of 21.3 percentage points based on a change from weak to strong policy-making power of opposition players. Even though the effects are also positive for losers (grey lines), they do not reach statistical significance. Strong policy-making power of opposition players

Young democracies

0.0

0.0

0.2

0.4

0.6

0.8

Pr (Satisfaction with Democracy) 0.2 0.4 0.6 0.8

1.0

1.0

Old democracies

0.0

0.2

Winner Loser

0.4

0.6

0.8

1.0

0.0

0.2

0.4

0.6

0.8

1.0

Policy−Making Power of Opposition Players

Figure 6.7  Predicted Probabilities: Satisfaction with Democracy. LAPOP

Power of Opposition Players and Democratic Consolidation  141

leads to an increase of 22.7 percentage points in the probability of being satisfied with democracy (an increase from 48.9 percent to 71.6 percent) for losers in old democracies, all else equal. The effect is slightly more pronounced for losers in young democracies where strong policy-making power of opposition players leads to an increase in the probability of being satisfied with democracy of 24.0 percentage points (from 50.5 percent to 74.5 percent), all else equal. Hence, it seems that policy-making power of opposition players has similar effects independent of the level of democratic experience.10 Furthermore, Figure 6.7 nicely shows how strong policy-making power of opposition players can contribute to a reduction in the winner-loser gap both in young and old democracies. However, Figure 6.7 also shows that the winner-loser gap is smaller in young democracies (right panel) than in older ones (left panel). Winners and losers living in an old democracy without policy-making power of opposition players have a statistically significant gap in their probability of being satisfied with democracy of 35.8 percentage points. The gap between winners and losers living in an old democracy with strong policy-making power of opposition players is only 27.9 percentage points. A much stronger pattern emerges in young democracies (right panel): a significant winner-loser gap of 24.7 percentage points in countries with weak policy-making power of opposition players and of 22.0 percentage points in countries with strong policy-making power of opposition players. Most importantly, however, this latter winner-loser gap is not statistically significant. Hence, strong policy-making power of opposition players in young democracies seems to contribute to a situation where winners and losers are not significantly different in their probability of being satisfied with democracy. These results emphasize that a winner-loser gap might only develop over time as the results do not show a statistically significant winner-loser gap in young democracies granting at least intermediate policy-making power to opposition players. Similarly, however, strong policy-making power of opposition players might help prevent the development of a significant winner-loser gap, as results only show a considerable winner-­ loser gap in old democracies with very weak policy-making power of opposition players. In this case, strong policy-making power of opposition players does not directly reduce the winner-loser gap (as it is not automatically present in new democracies), but it may help prevent the development of a significant gap between winners and losers of democratic elections. In sum, the results here confirm a positive effect of the policy-making power of opposition players on the probability of being satisfied with democracy for both winners and losers alike. This finding is in line with expectations regarding losers but does not necessarily match those regarding winners. As presented in Chapter 2, from an outcome-oriented

142  Power of Opposition Players and Consequences

perspective, I expected winners to express less specific support if living in countries that grant strong policy-making power to opposition players. Results, however, indicate that this is not necessarily the case, or, that the question about satisfaction with democracy does not capture specific support but is more likely a measure of diffuse support. Interestingly, these data also show different effects of the policy-making power of opposition players dependent on the level of democratic experience. Several results show that strong policy-making power of opposition players only unfolds its potential to contribute to more satisfaction with democracy in countries with a certain level of democratic experience. For example, whereas both winners and losers have an increasing probability of being satisfied with democracy if living in countries with many years of democratic experience, the effect is much stronger in countries granting strong policy-making power to opposition players (see, for example, Figure 6.3). Despite the positive effect on both winners and losers, strong policy-making power of opposition players can still reduce the winner-­ loser gap. Combined with democratic experience, the results point to the effect that strong policy-making power of opposition players may actually reduce the gap in the levels of satisfaction between winners and losers.

6.3  Satisfaction with Government Performance The statistical models in the previous section only partially confirm the theoretical expectations I presented in Chapter 2. Whereas results are in line with expectations for losers (i.e. a positive effect of policy-making power of opposition players), they do not confirm expectations for winners (i.e. a negative effect of policy-making power of opposition players). But, these results may also be due to the fact that satisfaction with democracy captures diffuse rather than specific support. Therefore, to further test the theoretical expectations regarding the effect of the policy-making power of opposition players on specific support, in this final empirical section, I present results for satisfaction with government performance which clearly captures specific support. Two different surveys provide information on the dependent variable satisfaction with government performance: the CSES (2019) and the ESS (2019).11 Again, overall a majority of respondents of the CSES survey are satisfied with the government performance. This is not the case for respondents in the ESS survey where the majority of respondents are not satisfied with the government performance. The policy-making power of opposition players has mean scores of 0.71 (ESS) and 0.68 (CSES). The individual-level and country-level control variables in both surveys again have similar distributions to the previous analyses. Descriptive statistics show that the majority of respondents do not have low education

Power of Opposition Players and Democratic Consolidation  143

but do have an intermediate level of household income. Furthermore, a majority of respondents feel close to a political party. Respondents in the ESS survey are mostly interested in politics and satisfied with their life. On the country level, most individuals live in decentralized parliamentary regimes with PR electoral formula. GDP/capita and corruption were recoded into binary variables for the ESS survey scoring 1 for values above the median and 0 otherwise. Finally, democratic experience shows a mean of 48.4 years for the CSES survey and 53.2 years for the ESS survey. However, in the CSES survey, democratic experience is strongly correlated with regime type (0.77), decentralization (0.95), and GDP/capita (0.76). Furthermore, regime type and decentralization are strongly correlated (0.78) leading to problems of multicollinearity. Switzerland is the only country in the sample categorized as a presidential regime and the only decentralized country. Furthermore, only two countries in the sample (Hungary and Romania) use a non-PR electoral formula. Therefore, I include the variable political system combining PR electoral formula, parliamentary and semi-presidential regimes and decentralization in the main analysis. Alternative model specifications including only PR electoral formula and political regime as control variables show mostly consistent results with the ones presented here as do models excluding Switzerland from the sample. Table 6.3 presents random intercept probit models on satisfaction with government performance as binary dependent variable based on data from both surveys. According to the theoretical framework, I expect opposite effects of policy-making power of opposition players for losers and winners. When evaluating specific outcomes, strong policy-making power of opposition players is theorized to be especially important for losers. Hence, losers should be more likely to be satisfied with government performance if living in a country with strong policy-making power of opposition players, whereas winners should be less satisfied in these settings. The results in the previous sections confirmed an effect of the policy-making power of opposition players dependent on the level of democratic experience. To test for these different effects, Table 6.3 also includes additional interactions with the democratic experience variable to account for possible differences in the effect of the policy-making power of opposition players in younger and older democracies. As in the previous analyses, these cross-level interactions are the most important for understanding the effect that the policy-making power of opposition players has on specific support. Overall, results in Model CSES M1 (first column of Table 6.3) confirm the theoretical expectations whereas Model ESS M1 shows positive effects of policy-making power of opposition players for both winners and losers alike. However, a closer look at the second set of models – including democratic experience – shows positive effects of policy-making power of opposition players for

144  Power of Opposition Players and Consequences Table 6.3  S atisfaction with Government CSES M 1 Individual level Loser Household income (intermediate) Household income (high) Low education Close to political party Interest in politics

−5.02*** (0.20) 0.01 (0.04) 0.04 (0.03) 0.07+ (0.04) −0.04 (0.03)

Satisfaction with life Country level Political regime (semi-pres.) Political regime (presidential) PR electoral system Decentralization Political system (1) Political system (2) GDP/capita (intermediate) GDP/capita (high) Corruption Cross-level interaction Winner*PPOP Loser*PPOP Winner*democratic experience*PPO Loser*democratic experience*PPOP Intercept Log. Likelihood N N countries

−0.86* (0.34) −0.77* (0.34) 0.46 (0.31) −0.20 (0.35) −0.26 (0.69)

ESS M1

−0.84*** (0.08) 0.05** (0.02) 0.00 (0.02) 0.24*** (0.02) 0.17*** (0.02) 0.67*** (0.02) −0.18 (0.15) 0.69* (0.35) 0.11 (0.16) −0.94*** (0.23)

0.57*** (0.05) −0.51*** (0.05)

−4.32*** (0.80) 2.83*** (0.80)

1.69*** (0.51) 2.04*** (0.51)

3.85*** (0.68) −4950.80 8,576 12

−1.95*** (0.37) −16961.22 30,846 18

CSES M2

0.11** (0.04) 0.04 (0.04) 0.04 (0.03) 0.10** (0.03) −0.03 (0.03)

0.10 (0.41) 0.11 (0.37) 0.38 (0.25) −0.53 (0.41) 0.49 (0.62)

0.02 * (0.01) 0.01 + (0.01) −0.52 (0.57) −5321.91 8,576 12

ESS M2

−0.78*** (0.03) 0.06*** (0.02) 0.02 (0.02) 0.25*** (0.02) 0.18*** (0.02) 0.67*** (0.02) −0.17 (0.62) −4.75** (1.54) 0.01 (0.63) −1.03 (0.92)

0.77*** (0.05) −0.44*** (0.05)

0.06 *** (0.01) 0.06 *** (0.01) −2.38*** (0.59) −16877.87 30,846 18

*** p < 0.001; ** p < 0.01; * p < 0.05; + p < 0.1. Random intercept probit model. Binary dependent variable: satisfaction with government. Political system: Combined score of PR electoral formula, parliamentary or semi-presidential regime, and centralization (weak federalism and bicameralism). GDP/capita and corruption ESS: dummy variable with score of 1 for higher than median.

Power of Opposition Players and Democratic Consolidation  145

both groups. These results hold also based on models using the original variable of satisfaction with the government performance from the ESS survey as well as based on robustness checks dropping Switzerland (as the only presidential regime) from the sample. Again, it seems that strong policy-making power of opposition players only has an effect in conjunction with democratic experience. Taking a closer look at the control variables we see statistically significant results for low education based on the CSES survey. Hence, individuals with low education are more likely to be satisfied with the government performance compared to individuals with high education. Results based on the ESS survey further show statistically significant results in the expected direction for all other individual-level control variables. Accordingly, individuals with a high household income who are close to a political party, show interest in politics, and are satisfied with their life are more likely to be satisfied with the government performance. On the country level, results show statistically significant effects for high GDP/capita and corruption for the ESS survey. This indicates that individuals living in countries with higher GDP/capita and lower corruption are more likely to be satisfied with the government performance. To illustrate these findings, I rely again on predicted probabilities of being satisfied with the government performance. Figure 6.8 shows predicted probabilities of being satisfied with the government performance based on model CSES M2 in Table 6.3. The figure presents results for both weak (right panels) and strong (left panels) policy-making power of opposition players for both winners (lower panels) and losers (upper panels) dependent on the years of democratic experience.12 Again, following King et al. (2010), the figure shows a simulation-based approach where each line represents the predicted probability based on a simulated set of model estimates. The density of lines represents the uncertainty associated with the predicted probability. In addition, 95 percent confidence intervals are shown in dashed lines. As reflected in Table 6.3, the figure shows a positive influence of the policy-making power of opposition players. The pattern is similar to that for satisfaction with democracy, however, much less clear in its effect. The probability of being satisfied with the government performance is higher for individuals living in countries with strong policy-making power of opposition players (left panels) than for individuals living in countries with weak policy-making power of opposition players (right panels). This finding holds for both, winners and losers alike. However, the results again show an increased effect of strong-policy-making power of opposition players with increasing democratic experience. More specifically, a loser living in a new democracy with only very weak policy-making power of opposition players has a probability of 57.3 percent of being satisfied with the government performance, all else

146  Power of Opposition Players and Consequences

equal. This probability is very similar for an individual living in a new democracy granting strong policy-making power to opposition players (58.9 percent). However, as Figure 6.8 shows, the positive effect that develops with more democratic experience is slightly more pronounced for strong policy-making power of opposition players (left panels) than for weak policy-making power of opposition players (right panels). A loser living in an old democracy with strong policy-making power of opposition players will have a probability of 98.6 percent of being satisfied with the government performance (compared to 58.9 percent in a new democracy), all else equal. This amounts to an increase of 39.7 percentage points. The effect is similar for winners living in countries with strong policy-making power of opposition players (lower panels in Figure 6.8). Here, an individual living in a new democracy granting strong policy-making power to opposition players has a probability of 53.9 percent of being satisfied with the government performance, all else equal. This probability rises to 99.3 percent for winners living in old democracies, all else equal. This amounts to an increase of 45.4 percentage points which is statistically significant at the 0.05 level. The results are also positive, but much weaker for both winners and losers living in democracies with only weak policy-making power of opposition players. This result is not in line with the theoretical expectation that strong-policy-making power of opposition players has a negative effect on the specific support

0.4 0.0

0.0

0.4

Pr (Satisfaction with Government)

0.8

Loser: weak PPOP

0.8

Loser: strong PPOP

0

50

100

150

0

100

150

0.4 0.0

0.0

0.4

0.8

Winner: weak PPOP

0.8

Winner strong PPOP

50

0

50

100

150

0

50

100

Democratic experience in years

Figure 6.8  Satisfaction with Government Performance. CSES

150

Power of Opposition Players and Democratic Consolidation  147

Winner−Loser Gap (Pr(winner)−Pr(loser)) −0.10 −0.05 0.00 0.05

of winners. On the contrary, strong policy-making power of opposition players seems to positively influence winners and the level of specific support they express. Figure 6.9 shows the winner-loser gap of satisfaction with government performance based on these results. Results show a negative winner-loser gap in young democracies regardless of the level of policy-making power of opposition players, indicating that it is actually losers that are more satisfied with the government performance than winners. Although this effect is reversed in countries with weak policy-making power of opposition players after 48 years of democratic experience, the gap remains negative in countries with strong policy-making power of opposition players. Hence, strong policy-making power of opposition players contributes to a situation in which losers are more likely to be satisfied with the government performance than winners. However, these effects are not statistically significant. Figure 6.10 shows predicted probabilities of being satisfied with the government performance based on data from the ESS (2019).13 Predicted probabilities are based on results of Model ESS M2 in Table 6.3. Again, the effect of policy-making power of opposition players on satisfaction with government performance is shown for winners (right panel) and losers (left panel) living in countries with weak policy-making power of opposition players (grey lines) and strong policy-making power of opposition players (black lines) dependent on the level of democratic experience.

weak PPOP (0.2) strong PPOP (0.9)

0

50 100 Democratic experience in years

150

Figure 6.9  W inner-Loser Gap: Satisfaction with Government Performance. CSES

148  Power of Opposition Players and Consequences

Winner

0.8 0.6

0.8

0.4

0.6

0.2

0.4

0.0

0.2 0.0

Pr (Satisfaction with Government)

1.0

1.0

Loser

0

50

Strong PPOP (0.9) Weak PPOP (0.2)

100

150

0

50

100

150

Democratic experience in years

Figure 6.10  Satisfaction with Government Performance. ESS

Results in Figure 6.10 show a clear positive effect of strong policy-making power of opposition players on the level of satisfaction with democracy for both winners (right panel) and losers (left panel). Even though Figure 6.10 also shows a positive effect of the years of democratic experience on the probability of individuals being satisfied with government performance in countries with both strong and weak policy-making power of opposition players, the effect of strong policy-making power of opposition players is clearly more pronounced. More specifically, a loser living in a country with only one year of democratic experience has a probability of less than 5 percent of being satisfied with government performance, regardless of whether strong policy-making power is granted to opposition players, all else equal. But, for losers this probability increases to 99 percent after 85 years of democratic experience if strong policy-making power is granted to opposition players and only to 34.9 percent after 170 years of democratic experience if weak policy-making power is granted to opposition players, all else equal. The pattern is similar, but less pronounced, for winners. For example, winners living in countries with strong policy-making power of opposition players reach a probability of 99 percent of being satisfied with government performance after 77 years of democratic experience, all else equal. In contrast, winners living in a country granting weak policy-making power to opposition players reach a probability of only

Power of Opposition Players and Democratic Consolidation  149

Winner−Loser Gap (Pr(winner)−Pr(loser)) −0.1 0.0 0.1 0.2 0.3

58.8 percent of being satisfied with government performance after 170 years of democratic experience, all else equal. Finally, Figure 6.11 shows the winner-loser gap of satisfaction with the government performance based on results in Figure 6.10. Results show a very similar gap in new democracies for countries that grant weak and strong policy-making power to opposition players. The gap between winners and losers increases with more democratic experience if opposition players are granted only weak policy-making power (solid line). The results for countries with strong policy-making power of opposition players show a different picture. Here, the gap in the probability of being satisfied with the government performance between winners and losers increases as new democracies gain more experience and decreases again after 37 years of democratic experience until reaching a level close to 0. However, none of these results are statistically significant. In sum, models based on both surveys point in the same direction. The analyses show a positive effect of policy-making power of opposition players on the satisfaction with government performance for both winners and losers. This result is only partially in line with expectations. I expected a positive effect of strong policy-making power of opposition players on specific support of losers, which is confirmed in both surveys. However, I also expected a negative effect of the policy-making power of opposition players on specific support of winners. The results in this section do not entirely confirm this pattern. Rather, strong policy-making power of opposition players has a significant and positive effect on the

weak PPOP (0.2) strong PPOP (0.9) 0

50 100 Democratic experience in years

150

Figure 6.11  W  inner-Loser Gap: Satisfaction with Government Performance. ESS

150  Power of Opposition Players and Consequences

probability of winners of being satisfied with the government performance. Again, the analyses highlight the importance of democratic experience. The policy-making power of opposition players seems to influence the level of specific support to different degrees depending on democratic experience. The difference between weak and strong policy-making power of opposition players becomes especially clear in countries with more democratic experience. This also influences the development of the winner-loser gap. Whereas results show a reduction of the winner-loser gap in countries granting strong policy-making power to opposition players over time, this is not the case in countries granting weak policy-making power to opposition players.

6.4 Conclusion In this chapter, I have analyzed the relationship between the policy-making power of opposition players and democratic consolidation as measured by support for democracy at the individual level. To present as detailed a picture as possible, both specific support (measured by satisfaction with the government performance and satisfaction with democracy) as well as diffuse support (measured by the opinion of democracy as the best form of government) were considered in these analyses. Based on the theoretical argument I presented in Chapter 2, I expected the policy-­making power of opposition players to have different consequences for winners and losers. For losers, strong policy-making power of opposition players should positively influence both diffuse and specific support. For winners, however, strong policy-making power of opposition players should only positively influence diffuse support but have a negative effect on specific support. The results mostly confirm these expectations. The statistical models show that strong policy-making power of opposition players has a positive and statistically significant effect on the level of both diffuse and specific support of losers. Hence, in general, losers living in countries with strong policy-making power of opposition players are more likely to express political support than losers living in countries with weak policy-making power of opposition players. These results confirm the theoretical expectations about the influence of the policy-making power of opposition players for losers. Furthermore, results also confirmed a positive effect of strong policy-making power of opposition players on diffuse support of winners. This result is also in line with the theoretical expectations presented in Chapter 2. Thus, according to the theoretical framework, results presented in this chapter provide initial evidence that, by granting more political equality, strong policy-making power of opposition players is associated with more democratic consolidation as measured by diffuse support. This effect is present for both winners and

Power of Opposition Players and Democratic Consolidation  151

losers, alike. Strong policy-making power of opposition players seems to be beneficial to losers by granting more influence in the decision-making process. By doing so, it contributes to more democratic consolidation among losers as they are more likely to express higher specific support. However, the results also highlight some important and interesting deviations from the theoretical expectations. The statistical models show a positive effect of the policy-making power of opposition players on specific support of winners. As measured by satisfaction with democracy, results for winners are mixed. While results based on the CSES and ESS surveys show positive effects for winners, a more fine-grained analysis based on CSES data shows a negative effect for winners in young democracies. Measuring specific support with satisfaction with the government performance again shows a negative effect of the policy-making power of opposition players for winners in the CSES survey. This effect, however, is also positive when considering a possible interaction with democratic experience. Consequently, most models point to a positive effect of strong policy-making power of opposition players on specific support for winners. However, there are some results pointing in the opposite direction (confirming the theoretical expectations). Several possible explanations may contribute to situating these results with reference to the existing literature. First, the findings regarding the support of winners are most contradictory regarding the satisfaction with democracy variable. These results may actually be seen as confirming the critique repeatedly raised in the literature about using satisfaction with democracy to measure political support. Several authors have highlighted that it is unclear whether satisfaction with democracy actually measures diffuse or specific support (see, e.g. Norris 2011; Linde and Ekman 2003; Canache, Mondak and Seligson 2001). Norris’ (2011: 24) conceptualization of political support places the evaluation of regime performance (i.e. satisfaction with democracy) on the intermediate level between diffuse and specific support. Correlations between satisfaction with democracy and diffuse support (i.e. democracy as the best form of government) and specific support (i.e. satisfaction with government performance) indicated a closer relation to specific rather than diffuse support (see Chapter 4 for details). However, the results indicating a positive effect of the policy-making power of opposition players for the probability of winners being satisfied with democracy are more in line with the theoretical argument about diffuse support. The results presented in this chapter may therefore be interpreted as confirming the problems associated with the use of the satisfaction with democracy variable. Second, a possible explanation for the positive influence of the policy-­making power of opposition players for winners may lie in the measure of the policy-making power of opposition players itself. Based on the logic of the policy influence differential,14 the proposed measure

152  Power of Opposition Players and Consequences

of the policy-making power of opposition players assesses the power of opposition players compared to that of the government (Strøm 1990: 42). Consequently, strong policy-making power of opposition players may indicate more power not only for opposition players but also for government actors.15 This increase in power for both government and opposition players may then lead to an increase of political support of both winners and losers alike. A final explanation can be linked to one robust and interesting finding of this chapter. The results highlight a relatively consistent pattern of the influence of democratic experience on the effect of the policy-making power of opposition players on democratic consolidation. The statistical models show that the potential of strong policy-making power of opposition players only unfolds with a certain degree of democratic experience but is not present in very young democracies. A possible explanation for this result may lie not only in the level of policy-making power of opposition players but also in the correlation with democratic experience and past communist regimes in European countries. Several studies have pointed to the effect of communist socialization on political attitudes. Tóka (1995: 356, 366), for example, links the difference in the level of political support between Eastern and Western Europe to the economic performance of the two regions. Similarly, Pop-Eleches and Tucker (2014, 2012, 2011) analyze the effect of legacies on political values in countries with past communist regime experience. The authors highlight that “(…) economic conditions and institutional arrangements play a role in explaining why trust in political parties is lower in postcommunist countries than elsewhere” (Pop-Eleches and Tucker 2011: 396–397). Furthermore, these results may point toward a learning mechanism and the fact that institutions unfold their effect only after some time. The results presented in this chapter have important implications for the consolidation of democracies around the world. The findings in this chapter point to the importance of legislative organization in the policy-making process and thereby identify legislatures as a crucial institution for successful democratic consolidation. The institutionalization of opposition power as measured by the policy-making power of opposition players may positively influence political support of both winners and losers and, furthermore, contribute to a smaller winner-loser gap. Strong policy-making power of opposition players seems to be especially important in countries with more democratic experience. In more established democracies, strong policy-making power of opposition players may help to minimize the risks of a decline in political support of democracy and the rise of anti-democratic forces. Through its limited effect on political support, weak policy-making power of opposition players may lead citizens to consider less democratic alternatives of government. In turn, high

Power of Opposition Players and Democratic Consolidation  153

levels of political support, facilitated by strong policy-making power of opposition players, is essential also in these more established regimes. As the results in this chapter show, the policy-making power of opposition players should be among the elements considered when thinking about institutional design in newly established democracies. Even if its influence only manifests itself over time, focusing on the institutionalization of opposition power may be an important mechanism to foster democratic consolidation. Therefore, opposition power and its positive effects on democratic consolidation ought to be an important aspect to consider for scholars and policy-makers alike. Institutionalizing opposition power may contribute to political support for democracy which is essential for every successful democratic regime.

Notes 1 Table A.6.1 shows the descriptive statistics of the variables in this first analysis. 2 However, as presented in Chapter 4, the question in the Latin American Public Opinion Project (Vanderbilt University 2019) is different from the questions in the other two surveys: VB20: Vote Intention in the Next Presidential Election. More specifically, for which candidate would you vote if the presidential elections were held this week? (Vanderbilt University 2019). 3 Predicted probabilities are based on an individual with the following scores: Household income=0; Close to party=0; Interest in politics=0; satisfaction with life=1; political regime =presidential; electoral system= PR; Decentralization=0; GDP/capita=intermediate, corruption=0.31. 4 Table A.6.2 shows the descriptive statistics of the variables in this second set of analyses. The summary statistics are shown for each survey separately as the analyses presented in the following sections are based on the individual surveys and not on a combined dataset. This decision is based on the different availability of control variables in the different surveys. 5 As discussed in the literature and presented in Chapter 4, the question of satisfaction with democracy lies at the intersection between diffuse and specific support. However, based on correlations, it seems to relate more closely to specific than diffuse support in the surveys considered here. 6 In the context of the Latin America Public Opinion Project (Vanderbilt University 2019), reporting that one would not vote for the current presidential/ government party if elections were held this week. 7 Predicted probabilities are shown for an individual with the following scores: household income = intermediate, low education = 0, close to party = 1, centralized parliamentary regime with PR electoral formula, GDP/capita = 2, corruption = 0.24 (mean). 8 Predicted probabilities are shown for an individual with the following scores: household income = intermediate, low education = 0, close to party = 1, interest in politics = 1, satisfaction with life=1, centralized parliamentary regime with PR electoral formula, GDP/capita = 2, corruption = 0.13 (mean). 9 Results are shown for an individual with the following characteristics: Household income=good enough, low education = 0, close to party=0, interest in

154  Power of Opposition Players and Consequences

10

11

12 13

14 15

politics=0, satisfaction with life=1, Political system=0 (presidential); Decentralization=0; GDP/capita=intermediate, corruption=0.49. However, as Table A.6.2 shows, countries included in the LAPOP survey show a smaller range of democratic experience compared to the CSES and ESS surveys. Whereas in the former the average individual lives in a country with 32 years of democratic experience, in the latter the average democratic experience amounts to 63.9 (CSES) and 53.1 (ESS) years, respectively. Similarly, the maximum number of years considerably differs between the LAPOP and the other surveys: 58 years (LAPOP) compared to 164 years (CSES) and 165 years (ESS), respectively. Table A.6.3 shows the descriptive statistics for this last set of analyses. Again, summary statistics are shown for each survey separately as the analyses presented in the following sections are based on the individual surveys. This decision is based on the different availability of control variables. Predicted probabilities are shown for an individual with the following characteristics: Household income = 1, low education = 0, close to party = 1, political system = 1, GDP/capita = high, corruption = 0.24. Predicted probabilities are shown for an individual with the following scores: household income = intermediate, low education = 0, close to party = 1, interest in politics = 1, satisfaction with life=1, centralized parliamentary regime with PR electoral formula, GDP/capita = 2, corruption = 0.13 (mean). See Chapter 3 for a more detailed discussion. As it is, for example, the case with fewer restrictions to introduce bills; as such rules not only grant more power to opposition MPs but also to those from government parties.

Bibliography Canache, Damary, Jeffery J. Mondak and Mitchell A. Seligson. 2001. “Meaning and Measurement in Cross-National Research on Satisfaction with Democracy.” Public Opinion Quarterly 65: 506–528. Claassen, Christopher. 2019. “Does Public Support Help Democracy Survive?” American Journal of Political Science 64(1): 118–134. CSES. 2019. “The Comparative Study of Electoral Systems. CSES Modules 1 to 4 Full Release. dataset. December 15, 2015 version. DOI Module 1: 10.7804/cses. module1.201512-15; DOI Module 2: 10.7804/cses.module2.2015-12-15; DOI Module 3:10.7804/cses.module3.2015-12-15; DOI Module 4: 0.7804/cses.module4.2015–0320. CSES Module 5 First Advance Release. dataset. May 21, 2019 version. doi:10.7804/cses.module5.2019-05-21.” Online: www.cses.org. Diamond, Larry J. 1999. Developing Democracy: Toward Consolidation. Baltimore, MD: Johns Hopkins University Press. Easton, David. 1975. “A Re-Assessment of the Concept of Political Support.” British Journal of Political Science 5(4): 435–457. ESS. 2019. “European Social Survey. Multilevel Data.” Online: http://www. europeansocialsurvey.org/. Accessed: October 2019. King, Gary, Michael Tomz and Jason Wittenberg. 2010. “Making the Most of Statistical Analyses: Improving Interpretation and Presentation.” American Journal of Political Science 44(2): 347–361.

Power of Opposition Players and Democratic Consolidation  155 Linde, Jonas and Joakim Ekman. 2003. “Satisfaction with Democracy: A Note on a Frequently Used Indicator in Comparative Politics.” European Journal of Political Research 42(3): 391–408. Lipset, Seymour Martin. 1959. “Some Social Requisites of Democracy Economic Development and Legitimacy.” American Political Science Review 53: 69–105. Norris, Pippa. 2011. Democratic Deficit. Critical Citizens Revisited. New York: Cambridge University Press. Pop-Eleches, Grigore and Joshua A. Tucker. 2011. “Communism’s Shadow: Postcommunist Legacies, Values, and Behavior.” Comparative Politics 43(4): 379–408. Pop-Eleches, Grigore and Joshua A. Tucker. 2012. “Post-Communist Legacies and Political Behavior and Attitudes.” Demokratizatsiya: The Journal of Post-Soviet Democratization 20(2): 157–166. Pop-Eleches, Grigore and Joshua A. Tucker. 2014. “Communist Socialization and Postcommunist Economic and Political Attitudes.” Electoral Studies 33: 77–89. Przeworski, Adam. 1991. Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America. Cambridge: Cambridge University Press. Riker, William H. 1983. Political Theory and the Art of Heresthetics. In Political Science: The State of the Discipline, ed. Ada W. Finifter. Washington, DC: American Political Science Association Washington: 47–67. Strøm, Kaare. 1990. “A Behavioral Theory of Competitive Political Parties.” American Journal of Political Science 34: 565–598. Tóka, Gábor. 1995. Political Support in East-Central Europe. In Citizens and the State, eds. Hans-Dieter Klingemann and Dieter Fuchs. Oxford and New York: Oxford University Press: 354–381. Vanderbilt University. 2019. “Latin American Public Opinion Project (LAPOP).” Online: http://www.vanderbilt.edu/lapop/. Accessed: October 2019. Welzel, Christian and Ronald F. Inglehart. 2009. Political Culture, Mass Beliefs, and Value Change. In Democratization, eds. Ronald F. Inglehart, Christian W. Haerpfer, Patrick Bernhagen and Christian Welzel. Oxford and New York: Oxford University Press: 126–144.

Chapter 7

Why Should We Care About Opposition Power? Conclusion and Implications

This book examines the role of opposition power in democratic consolidation, thereby addressing two gaps in current research. First, I argue that research analyzing the influence of institutions on democratic consolidation neglects the role of individual citizens. This is extremely surprising as the dimension of the individual is repeatedly mentioned in the definitions of democratic consolidation. Diamond (1994: 15), for example, has emphasized that “(…) consolidation is the process by which democracy becomes so broadly and profoundly legitimate among its citizens that it is very unlikely to break down.” Furthermore, Przeworski (1991: 15) has highlighted that the central question in consolidation studies is why losers “(…) comply with the outcomes [of elections] and continue to participate rather than subvert democratic institutions.” Second, I argue that the research on the institutional determinants of democratic consolidation has largely focused on different procedures of government formation and neglected the legislature as the central body representing citizens and especially losers of democratic elections. However, if democratic consolidation is inherently about citizens’ attitudes and opinions – as definitions of consolidation suggest – focusing on the role of the legislature as the institution representing these citizens seems imperative. Finally, I show that not only have democracy studies paid little attention to the role of the legislature for democratic consolidation, but legislative studies have also paid little attention to the influence of legislative characteristics on extra-legislative outcomes. Therefore, the purpose of this book is to bridge these current gaps and to contribute part of the answer to the general question of the determinants of successful democratic consolidation. To fulfil its purpose, the goals of this book are to understand the importance of legislatures in democratic consolidation, by using the insights generated by legislative studies, to preserve the important distinction between winners and losers, and to analyze the question of consolidation from the perspective of individuals. In the first section of this concluding chapter, I present the main results and discuss the theoretical and empirical contributions of DOI: 10.4324/9781003005360-9

Conclusion and Implications  157

this book. In the second section, I address the limitations of the present analyses and suggest future avenues for research on opposition roles in legislatures, democratic consolidation, and individual political support. I conclude this chapter with a last thought.

7.1  Main Findings To analyze the influence of parliamentary procedures on democratic consolidation, I built upon several different research areas. First, research on the institutional design of democracies has emphasized the importance of elections and legislatures for the successful consolidation of democracies. However, the literature dealing with the question of democratic consolidation has largely focused on government formation processes. To expand this research, this book proposes a shift in focus from the government formation process to the (mostly neglected) legislative institutions. Second, research on political support has highlighted the importance of individual political support for the survival and success of democratic regimes. Similarly, research in this area has also identified a large number of explanatory factors for the varying levels of support for democracy. However, it has not considered the specific power of different groups, mainly winners and losers, in great detail. Building on this literature, this book looks at the rights of the representatives of the electoral losers as realized in the policy-making power of opposition players. The book mobilizes the vast literature in legislative studies focusing on legislative organization to propose a measure of the policy-making power of opposition players. Finally, this book mobilizes approaches of the democratization literature and the coalition formation literature to connect the research on democratic consolidation and individual political support. By combining these research traditions, I propose a novel way of looking at the consolidation of democratic regimes. First and foremost, I bring legislatures into focus as the main body representing both winners and losers of democratic elections. Even though a large research tradition exists on legislative organization, the consolidation literature has largely neglected these achievements of legislative studies. Building on these insights from the legislative studies literature, I adapt the logic of rational choice approaches of the democratization literature and the literature on coalition formation to a post-transition setting. This individual-based argument relies on the costs and benefits of different legislative organizations in the policy-making process for winners and losers. Winners and losers will express the most political support in a setting that generates the most benefits for them. Therefore, from the perspective of losers, the inclusiveness of parliamentary procedures decides the degree of inequality between them and winners of the political game (i.e. the power

158  Power of Opposition Players and Consequences

of the government). Thus, losers should express a higher level of both diffuse and specific political support in settings with inclusive parliamentary procedures. The same is not necessarily true for winners. From an outcome-oriented perspective, I expected more political equality to lead winners to express lower levels of specific support. Whereas, from a procedure-­oriented perspective, winners should still express higher levels of diffuse support in a setting with inclusive parliamentary procedures (i.e. more power granted to opposition players). To test these hypotheses, this book presents novel data on the policy-­ making power of opposition players. The measure used to assess the power of opposition players focuses on one of the core functions of legislatures: policy-making and checking majority and executive power. This measure combines a total of seven variables1 that assess the power granted to opposition players during the policy-making process. The data collection was based on official documents of national parliaments and governments (i.e. national constitutions and standing orders of legislative chambers). Some coding was also based on existing data2 and website information of national parliaments. The data presented in this book is a first attempt to assess the policy-making power of opposition players across a wide range of countries. The measure of policy-making power of opposition players contributes to the research area in legislative studies looking at the organization of legislatures more generally and presents one of the few accounts of measuring legislative organization from the perspective of oppositions. Thereby, the presented measure goes beyond classical regime characteristics, such as parliamentary and presidential regimes or federal and unitary states and allows for a comparison of legislative organization across a wider range of countries. Based on the theoretical argument that the consent of losers is of crucial importance for the survival and stability of every democratic regime, the presented measure of the policy-making power of opposition players assesses the power granted to the representatives of these losers. The measure directly relates to the power distribution in the legislative chamber between winners and losers. It thereby allows for a more direct link between the potential influence of institutional settings on the political support of losers than traditional regime classifications propose. Research presented in this book also fits into an emerging research agenda in legislative studies analyzing extra-legislative effects of parliamentary rules. Descriptive results presented in Chapter 5 show that countries vary considerably in the degree of influence they grant to opposition players. Hence, the results of the descriptive analysis confirm the findings in the legislative studies literature – that legislative organization varies

Conclusion and Implications  159

significantly between countries. Most importantly, descriptive results show considerable variance, not only across countries but also across regime types, with regards not only to the separate dimensions (i.e. initiation power, debate power, and veto power) but also to the overall index of the policy-making power of opposition players. The results show a slight tendency of semi-presidential regimes to grant strong policy-making power to opposition players more often than presidential and parliamentary regimes. Whereas presidential regimes show a cluster of policy-making power of opposition players ranging from 0.5 to 0.8, parliamentary regimes show a wide range of power granted to opposition players. This significant variance regarding the policy-making power of opposition players is important as it suggests that looking more closely at how legislatures are organized and how they include opposition players in the policy-making process may be a promising approach to explain the different levels of democratic consolidation. Therefore, the potential role and influence of opposition players may not be fully understood when considering only the process of government formation. Hence, research that is interested in the role of institutional characteristics and its link to individual political support and democratic consolidation ought to also consider the actual rights granted to the representatives of not only winners (i.e. the government) but also of losers (i.e. the opposition). This new measure of the policy-making power of opposition players marks a first step in moving beyond classical definitions of regime types and to looking more closely at the specific rights of opposition players and their influence on democratic consolidation. Results on the influence of policy-making power of opposition players on individual political support for democracy (Chapter 6) show a significant link between the two. Results presented in Chapter 6 are based on different surveys (ESS, LAPOP, and CSES) and several questions on individual political support. To present a picture as detailed as possible, both specific support measured by satisfaction with the government performance and satisfaction with democracy as well as diffuse support measured by the opinion of democracy as the best form of government were considered. Based on the theoretical argument I presented in Chapter 2, I expected the policy-making power of opposition players to have different consequences for winners and losers. For losers, I expected a positive relationship between strong policy-making power of opposition players and both diffuse and specific support. For winners, however, I expected a positive relationship between strong policy-making power of opposition players and diffuse support but a negative effect on specific support. The results presented in Chapter 6 mostly confirm these expectations. Strong policy-making power of opposition players has a positive effect on the level of both diffuse and specific support of losers. Similarly, results also

160  Power of Opposition Players and Consequences

show a positive effect of strong policy-making power of opposition players on the diffuse support of winners. These results presented in Chapter 6 provide initial evidence that strong policy-making power of opposition players is linked to more democratic consolidation. Strong policy-making power of opposition players positively influences democratic consolidation as measured by diffuse support. This result holds for both winners and losers of democratic elections alike. Moreover, strong policy-making power of opposition players seems to also contribute to more democratic consolidation as measured by specific support of losers. However, the results show a relatively consistent pattern of the influence of democratic experience on the effect of policy-making power of opposition players on democratic consolidation. The potential of strong policy-making power of opposition players only unfolds with a certain degree of democratic experience and is not present in very young democracies. Institutionalizing strong policy-making power of opposition players does not seem to help consolidate very young democracies. However, institutionalized opposition power does seem to contribute to higher levels of democratic consolidation (i.e. political support) in more established democracies. This result is especially important regarding the question of democratic backsliding and increasing dissatisfaction with democracy in consolidated democratic regimes. As the effects of strong policy-making power of opposition players is something that only manifests with a certain degree of democratic experience, it may help prevent these potentially dangerous developments for the survival of democracy. The results presented in this book have important implications for the consolidation of democracies around the world. First, the results highlight the importance of legislative organization in the policy-­making process. Legislatures seem to be a crucial institution for the successful democratic consolidation. Even though this has been highlighted in some contributions (see, e.g. Barkan 2009; Fish 2006), the legislature has not figured prominently among the institutions linked to successful democratic consolidation. But the variance in policy-making power of opposition players in the descriptive chapter shows that neglecting the specific organization of legislatures shows an incomplete picture of regime types. Second, the institutionalization of opposition power as measured by the policy-making power of opposition players may positively influence political support of both winners and losers. Therefore, strong policy-­ making power of opposition players may be among the institutional characteristics contributing to democratic consolidation. However, strong policy-making power of opposition players seems to be especially important in countries with more democratic experience. Even if strong policy-making power of opposition players may not be among the institutions helping young democracies to consolidate in the first place, it

Conclusion and Implications  161

can have important consequences in established democracies. Thus, for scholars and policy-makers alike, the legislature and especially the role of opposition players and the level of power granted to them ought to be a chief concern in institutional design. Through its positive influence on political support, both diffuse and specific, strong policy-making power of opposition players may be one of the factors preventing strong anti-system forces in democracies (which, in the extreme, also endanger democratic consolidation and stability). Hence, in more established democracies, strong policy-making power of opposition players may contribute to minimizing the risks of a decline in political support of democracy. Through its effect on political support, weak policy-making power of opposition players may lead citizens to consider less democratic government alternatives. In turn, high levels of political support, facilitated through strong policy-making power of opposition players, is essential in these more established regimes. In sum, even though I cannot finally identify causal effects based on the underlying research design, results show strong support for policy-­ making power of opposition players ought to be among the elements considered when thinking about institutional design in newly established democracies. Even if its influence only manifests over time, focusing on the institutionalization of opposition power may be an important mechanism in fostering democratic consolidation. Therefore, opposition power and its positive effects on democratic consolidation should be an important aspect for scholars and policy-makers to consider.

7.2  T he Study of Institutions, Oppositions, and Democratic Consolidation – Limitations and Future Research Although I have tried to expand the research on the question of democratic consolidation and provide new insights into extra-legislative effects of legislative organization, this book only marks a starting point. It is an appeal to combine results of various research traditions to advance the research on the question of successful democratic consolidation. Considerable limitations remain that highlight several avenues for future research. The theoretical argument about the influence of the policy-making power of opposition players on individual opinions lies on the assumption that citizens somehow know what the specific procedures in their polity are. As discussed in Chapter 2, this might be more obvious for some of the institutional settings, such as the possibility of citizen initiatives, but less so for others, such as the organization of parliamentary committees. The assumption in this book is that the legislative behavior of MPs and political parties is mediated by mass media and that political

162  Power of Opposition Players and Consequences

actors themselves have an incentive to communicate their legislative activities to their potential voters. Therefore, even if they do not know the procedural details, citizens can be aware of the extent to which opposition players are able to influence the policy-making process. Research increasingly confirms this assumption and suggests that the activity of political actors does reach voters (see, e.g. Tuttnauer and Wegmann 2022; Fortunato and Stevenson 2021; Huber et al. 2022; McKay 2020; Bouteca et al. 2019; Bowler 2010; Carey 2009). As parliamentary procedures determine the possibility of action by representatives upon which voters evaluate their MPs, parliamentary rules and procedures may also be likely to influence how individuals evaluate democracy. Nevertheless, the link between the specific parliamentary procedures and the knowledge of these procedures relies on an assumption that was not specifically tested in this book. Future research should therefore look more closely into this relationship and analyze how political actors communicate with their constituents and to what extent citizens are aware of specific opportunities for different actors to influence the policy-making process. Moreover, in this book I have provided detailed data on the potential influence of opposition players in the policy-making process for a total of 55 countries.3 However, the data presented allow only for a short period of time-series-cross-sectional analysis. This constitutes an important shortcoming as far as the detail of possible analyses and, especially, the identification of the causal link between the policy-making power of opposition players and democratic consolidation are concerned. Due to the wide range of years in which national constitutions and standing orders came into force, the data on the policy-making power of opposition players presented here lacks complete time series information and is not able to catch changes in parliamentary rules. Therefore, future research efforts should consider a more extended time dimension. Such future research efforts would be able to also examine changes in the policy-making power of opposition players over time. Such a dataset would allow for further and more detailed analyses of the impact of the policy-making power of opposition players on individual political support. As more recent contributions to legislative studies have emphasized, parliamentary rules change over time (see, e.g. Sieberer and Höhmann 2021; Sieberer, Müller and Heller 2011). Ignoring these changes in institutions and assuming institutional stability is very likely to lead to biased results. In this book, I have tried to take this shortcoming into consideration by only considering survey waves that were conducted after the coming into force of the standing order used to measure the policy-making power of opposition players for each country. I therefore believe that the design of the analysis still allows for the conclusion that the policy-­ making power of opposition players and democratic consolidation are linked. Nevertheless, clearly identifying a causal effect and addressing

Conclusion and Implications  163

the question of endogeneity appropriately would require the inclusion of the so-called constitutional moment (see, e.g. Fish 2006), i.e. the analysis of the evolution of policy-making power of opposition players from the first democratic standing orders and constitutions and corresponding surveys on political opinion over time. Furthermore, more detailed data on the policy-making power of opposition players would certainly allow subsequent questions to be addressed. For example, the availability of time-variant information on the policy-making power of opposition players would allow for the inclusion of the question regarding the impact of changes to the policy-making power of opposition players on political support in a specific country. Such an analysis would contribute considerably to the understanding of extra-legislative effects of parliamentary rules and further allow for an identification of the causal link between the level of policy-making power of opposition players and individual political support. Also, with time-variant data, it would be possible to assess the role of changes in government parties and coalition formation on political support. This indicates a further avenue for research linking different research traditions. There is a vast and rich research tradition on questions of government formation that could be linked to questions on legislative organization and democratic consolidation. A further shortcoming of the data is its focus on parliamentary organization of first chambers only. Even though the analyses in Chapter 6 control for the effect of strong bicameral systems, second chambers may show different opportunities for opposition players than first chambers. Therefore, future data collection efforts should also apply the operationalization of the policy-making power of opposition players, presented here, to second chambers. For bicameral systems, a combined score of the policy-making power of opposition players for the two chambers would present a more detailed picture of the overall power of opposition players in the policy-making process. Similarly, separate measures of the policy-making power of opposition players for both chambers advance the literature on bicameralism and the specific power granted to the two chambers (see, e.g. the literature on symmetrical bicameral systems, Lijphart 1999). Finally, in this book I tried to bring legislatures into focus in the democratization literature by building a bridge between the research on legislative organization and that on individual political behavior and rational choice approaches of democratization. Research presented in this book speaks to the emerging research in legislative studies analyzing extra-legislative effects of parliamentary rules. But, a detailed measure of the policy-making power of opposition players may also contribute considerably to the research on legislative organization more broadly. For example, as Martin (2014: 361) has put it: a “(…) significant factor

164  Power of Opposition Players and Consequences

preventing construction of comparative theories relates to a lack of data to test any such theories”. Very recent contributions to the field have presented such comparative data and further pointed to the importance of broad comparative analyses for our understanding of the origin and effects of legislative procedures (see, e.g. Sieberer and Höhmann 2021; Zubek 2021; Wegmann 2022; André, Depauw and Martin 2016; Hug, Wegmann and Wüest 2015; Garritzmann 2017). Hence the measure of policy-making power of opposition players allows for more detailed comparative research in the field of legislative studies that still largely relies on case studies or comparisons of a small number of cases. The level of policy-making power granted to opposition players may, for example, also influence the strategic behavior of political actors. The behavior of parties and coalition formation may be dependent upon the level of power granted to opposition players. Or, the level of the policy-making power of opposition players may also influence legislative activities of MPs or explain country specific government structures.

7.3  Last Thoughts Research presented in this book shows that institutionalizing opposition power may contribute to political support for democracy which is essential for every successful democratic regime. It thereby focuses on two neglected aspects in the extant literature: the role of the legislature for democratic consolidation and the importance of opposition players (i.e. losers of democratic elections). Results show that analyzing extra-legislative effects of parliamentary rules and giving these neglected aspects a more important role in research on democratic consolidation is important. This book marks a point of departure. I hope that the theoretical framework and the results presented here can be a starting point for future research and show that – even if challenging – it is worthwhile to combine different research areas to find answers to the most challenging problems of our time.

Notes 1 Bill introduction, agenda setting, committee structure, committee procedures, amendments, executive power, and referendums. 2 Hug, Wegmann and Wüest (2015); Elkins, Ginsburgh and Melton (2014); IDEA (2014); Regan and Clark (2014); Fish and Kroenig (2009). 3 Partial information on policy-making power of opposition players is available for a total of 74 countries.

Bibliography André, Audrey and Sam Depauw and Shane Martin. 2016. “‘Trust is Good, Control Is Better’ Multiparty Government and Legislative Organization.” Political Research Quarterly 69(1): 108–120.

Conclusion and Implications  165 Barkan, Joel D. 2009. African Legislatures and the “Third Wave” of Democratization. In Legislative Power in Emerging African Democracies, ed. Joel D. Barkan. Boulder, CO: Lynne Rienner Publishers: 1–32. Bouteca, Nicolas, Jef Smulders, Bart Maddens, Carl Devos and Bram Wauters. 2019. “‘A Fair Day’s Wage for a Fair Day’s Work’? Exploring the Connection between the Parliamentary Work of MPs and their Electoral Support.” The Journal of Legislative Studies 25(1): 44–65. Bowler, Shaun. 2010. “Private Members’ Bills in the UK Parliament: Is There an ‘Electoral Connection’?” The Journal of Legislative Studies 16(4): 476–494. Carey, John M. 2009. Legislative Voting and Accountability. Cambridge: Cambridge University Press. CSES. 2019. “The Comparative Study of Electoral Systems. CSES Modules 1 to 4 Full Release. dataset. December 15, 2015 version. DOI Module 1: 10.7804/cses. module1.201512-15; DOI Module 2: 10.7804/cses.module2.2015-12-15; DOI Module 3: 10.7804/cses.module3.2015-12-15; DOI Module 4: 0.7804/cses.module4.2015–0320. CSES Module 5 First Advance Release. dataset. May 21, 2019 version. doi:10.7804/cses.module5.2019-05-21.” Online: www.cses.org. Diamond, Larry J. 1994. “Toward Democratic Consolidation.” Journal of Democracy 5(3): 4–17. Elkins, Zachary, Tom Ginsburgh and James Melton. 2014. “Characteristics of National Constitutions, Version 2.0.” Online: http://www.comparativeconstitutionsproject.org. Accessed: June 2013. ESS. 2019. “European Social Survey. Multilevel Data.” Online: http://www.europeansocialsurvey.org/. Accessed: October 2019. Fish, Steven M. 2006. “Stronger Legislatures, Stronger Democracies.” Journal of Democracy 1: 5–19. Fish, Steven M. and Matthew Kroenig. 2009. The Handbook of National Legislatures: A Global Survey. Cambridge: Cambridge University Press. Fortunato, David and Randolph T. Stevenson. 2021. “Party Government and Political Information.” Legislative Studies Quarterly 46(2): 251–295. Garritzmann, Julian. 2017. “How Much Power Do Oppositions Have? Comparing the Opportunity Structure of Oppositions in 21 Democracies.” The Journal of Legislative Studies 23(1): 1–30. Huber, Lena Maria, Anita Bodlos, Elisabeth Graf and Thomas M. Meyer. 2022. “Disseminating Legislative Debates: How Legislators Communicate the Parliamentary Agenda.” Party Politics 28(2): 365–376. Hug, Simon, Simone Wegmann and Reto Wüest. 2015. “Parliamentary Voting Procedures in Comparison.” West European Politics 38(5): 940–968. IDEA. 2014. “Institute for Democracy and Electoral Assistance. Direct Democracy Database.” Online: http://www.idea.int/elections/dd/search.cfm. Accessed: January 2014. Lijphart, Arend. 1999. Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. New Haven, CT and London: Yale University Press. Marshall, Monty G., Keith Jaggers and Ted Robert Gurr. 2018. “Polity IV Project, Political Regime Characteristics and Transitions, 1800–2018.” Online: http://www.systemicpeace.org/inscrdata.html. Accessed: October 2019. Martin, Shane. 2014. Committees. In The Oxford Handbook of Legislative Studies, eds. Shane Martin, Thomas Saalfeld and Kaare W. Strøm. Oxford and New York: Oxford University Press: 352–368.

166  Power of Opposition Players and Consequences McKay, Lawrence. 2020. “Does Constituency Focus Improve Attitudes to MPs? A Test for the UK.” The Journal of Legislative Studies 26(1): 1–26. Przeworski, Adam. 1991. Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America. Cambridge: Cambridge University Press. Regan, Patrick and Dave Clark. 2014. “The Institutions and Elections Project.” Online: https://havardhegre.net/iaep/ Accessed: January 2014. Sieberer, Ulrich and Daniel Höhmann. 2021. “Do Party System Parameters Explain Difference in Legislative Organization? Fragmentation, Polarization, and the Density of Regulation in European Parliaments, 1045–2009.” Party Politics: 1–14. Sieberer, Ulrich, Wolfgang C. Müller and Maiko Isabelle Heller. 2011. “Reforming the Rules of the Parliamentary Game: Measuring and Explaining Changes in Parliamentary Rules in Austria, Germany, and Switzerland, 1945–2010.” West European Politics 34(5): 948–975. Tuttnauer, Or and Simone Wegmann. 2022. “Voting for Votes: Opposition Parties’ Legislative Activity and Electoral Outcomes.” American Political Science Review: 1–18. Vanderbilt University. 2019. “Latin American Public Opinion Project (LAPOP).” Online: http://www.vanderbilt.edu/lapop/. Accessed: October 2019. Wegmann, Simone. 2022. “Policy-Making Power of Opposition Players: A Comparative Institutional Perspective.” The Journal of Legislative Studies 28(1): 1–25. Zubek, Radoslaw. 2021. “Committee Strength in Parliamentary Democracies: A New Index.” European Journal of Political Research 69(4): 1018–1031.

Appendix

Appendix to Chapter 3 Figure A.3.1 shows the correlations between the index of the policy-­ making power of opposition players resulting from standardized variables as discussed and alternative aggregation rules. Results in black show correlations between the additive index and an index based on factor analysis. Results in grey triangles show correlations between the additive index and an index based on both standardized variables and stages. Finally results in grey circles show correlations between the additive index and an index based on the sum of variable scores. As correlation scores in parenthesis show, all four indices show similar results of policy-making power of opposition players. strong .8 .6 .4 .2 weak weak .2 .4 .6 .8 strong Policy-making power of opposition players (PPOP) - weighted variables PPOP - factor (0.96) PPOP - weighted variables and stages (0.99) PPOP - sum of variables (0.98)

Figure A .3.1  P olicy-Making Power of Opposition Parties – Aggregation Rules

168 Appendix

Appendix to Chapter 4 Table A.4.1 shows an overview of all country-level variables and datasets used in the analyses in Chapter 6. The first column states the variable as used in the analyses, and the second column indicates the dataset that provided this variable. The third column gives detailed information about the original variable and coding in the dataset, and the fourth column presents any recoding decisions of the variable as included in the analyses. Variables are presented in alphabetical order. Table A.4.2 presents all individual-level variables and datasets used in the analyses in Chapter 6. Variables are presented in alphabetical order.

Appendix to Chapter 5 Table A.5.1 shows an overview of all of the countries included in the sample and details about information available on initiation, debate, and veto power. Furthermore, the table shows the final scores for the overall index of the policy-making power of opposition players (PPOP) for all countries with complete information. Table A.5.2 shows correlations between all seven variables of the index of the policy-making power of opposition players.

Appendix to Chapter 6 Table A.6.1 shows the descriptive statistics of the variables in the first set of analyses of the relationship between the policy-making power of opposition players and individual opinions about democracy as the preferable form of government. As this dependent variable is only available in the Latin American Public Opinion Project (Vanderbilt University 2016), Table A.6.1 shows summary statistics of this survey only. Table A.6.2 shows the descriptive statistics of the data used in the analyses of the influence of the policy-making power of opposition players on satisfaction with democracy. These analyses consider three different surveys (CSES 2019; ESS 2019; Vanderbilt University 2019). The summary statistics are shown for each survey separately as the analyses presented in the following sections are based on the individual surveys and not on a combined dataset. This decision is based on the different availability of control variables in the different surveys. Table A.6.3 shows the descriptive statistics for the last set of analyses of the influence of the policy-making power of opposition players on the satisfaction with the government performance. Information on the dependent variable is available in the Comparative Study of Electoral Systems (CSES 2019) and the European Social Survey (ESS 2019). Hence, Table A.6.3 shows summary statistics of these two surveys.

4.96.1 vdem_corr: How pervasive is political corruption? Clarification: The directionality of the V-Dem corruption index runs from less corrupt to more corrupt (unlike the other V-Dem variables that generally run from less democratic to more democratic situation). The corruption index includes measures of six distinct types of corruption that cover both different areas and levels of the polity realm, distinguishing between executive, legislative and judicial corruption. Within the executive realm, the measures also distinguish between corruption mostly pertaining to bribery and corruption due to embezzlement. Finally, they differentiate between corruption in the highest echelons of the executive (at the level of the rulers/cabinet) on the one hand, and in the public sector at large on the other. The measures thus tap into several distinguished types of corruption: both ’petty’ and ’grand’; both bribery and theft; both corruption aimed and influencing law making and that affecting implementation. Aggregation: The index is arrived at by taking the average of (a) public sector corruption index; (b) executive corruption index; (c) the indicator for legislative corruption; and (d) the indicator for judicial corruption. In other words, these four different government spheres are weighted equally in the resulting index. V-Dem replace missing values for countries with no legislature by only taking the average of (a), (b) and (d).

Corruption

Coppedge et al. (2017), provided by Dahlberg et al. (2019)

Original variable and coding in dataset

Variable in analyses Dataset

Table A .4.1  O verview Datasets – Country Variables

(Continued)

The variable ranges from 0 to 1. No recoding

Recoding for analyses

GDP/capita

Electoral system

Democratic experience

Decentralization

Original variable and coding in dataset

Recoding for analyses

Unitarism: Recoded into binary Five-point scale (0 –2). Average of Nonfederalism and variable: Nonbicameralism. Score of 1 for all A) Nonfederalism is coded as: (0) Federal (elective regional countries scoring legislatures plus conditional recognition of subnational authority); 0 or 0.5 on the (1) Semifederal (where there are elective legislatures at the Unitarism variable, regional level but in which constitutional sovereignty is reserved 0 otherwise. to the national government); (2) Non-federal. (B) Nonbicameralism is coded as: (0) Strong bicameral (upper house has some effective veto power; the two houses are incongruent); (1) Weak bicameral (upper house has some effective veto power, though not necessarily a formal veto, the two houses are congruent); (2) Unicameral (no upper house or weak upper house). Marshall, Jaggers and DEMOC Number of Gurr (2018a) Institutionalized Democracy: Democracy indicator of the additive consecutive years Polity IV democracy scale. as democracy (score of 6 or more). Recoded into binary Bormann and Golder 4.43.10 gol_est: variable: (2013) provided Electoral System Type-3 classes Score of 1 for by Dahlberg et al. This is a categorical variable that takes on one of three values proportional, score (2019) indicating the basic type of electoral system used in the elections. of 0 otherwise. Majoritarian Proportional Mixed World Bank (2016), wdi_ gdppcpppcon: No recodings provided by GDP per capita based on purchasing power parity (PPP). PPP GDP Dahlberg et al. is gross domestic product converted to international dollars using (2019) purchasing power parity rates.

Gerring et al. (2005)

Variable in analyses Dataset

Own coding (see Chapter 3)

Cheibub, Gandhi and Vreeland (2010), provided by Dahlberg et al. (2019)

PPOP

Regime type

Regime/gol_inst: Sixfold regime classification at the end of the given year: Parliamentary democracy Mixed (semi-pres.) democracy Presidential democracy Civilian dictatorship Military dictatorship Royal dictatorship The data for this variable come from Cheibub, Gandhi and Vreeland (2010) (available until 2008), Dahlberg et al. (2019) updated the information through 2011.

PPOP: Policy-making power of opposition players

The variable ranges from 0 to 1. No recoding No recoding

ESS (2019)

Close to party

Vanderbilt University (2019)

Vanderbilt University (2019)

Democracy as best form of government

Economic situation

CSES (2019)

Vanderbilt University (2019)

Dataset

Variable in analyses

Recoding for analyses

1 2 3 4

Good enough and can save Good enough, with no major problems Not enough, and are stretched Not enough, and having a hard time

1 We need a strong leader who does not have to be elected 2 Electoral democracy is the best q10d: Family economic situation. The salary that you receive and total household income…

Recoded into binary variable: Score of 1 for all individuals responding “good enough” (categories 1 and 2), score of 0 otherwise.

Recoded into binary variable: Score of 1 for all individuals answering yes, score of 0 otherwise. Recoded into binary variable: Score of 1 for all individuals answering yes, score of 0 otherwise. A3004/B3028/C3020_1/D3018_1/e3024_1: Recoded into binary variable: Do you usually think of yourself as close to any Score of 1 for all individuals particular party? answering yes, score of 0 (1) Yes; (2) No otherwise. aut1: Recoded into binary variable: There are people who say that we need a strong leader Score of 1 for all individuals who does not have to be elected by the vote of the responding electoral democracy is people. Others say that although things may not the best, score of 0 otherwise. work, electoral democracy, or the popular vote, is always best. What do you think?

clsprty: Is there a particular political party you feel closer to than all the other parties? (1) Yes; (2) No vb10: Do you currently identify with a political party? (1) Yes; (2) No

Original variable and coding in dataset

Table A .4.2  O verview Datasets – Individual Variables

Household income

Government performance

CSES (2019)

ESS (2019)

ESS (2019)

CSES (2019)

1 2 3 4 5

Lowest household income quintile Second household income quintile Third household income quintile Fourth household income quintile Highest household income quintile

A2012/B2020/C2020/D2020/e2010 Household income quintile appropriate to the respondent.

 1 1st decile  2 2nd decile  3 3rd decile  4 4th decile  5 5th decile  6 6th decile  7 7th decile  8 8th decile  9 9th decile 10  10th decile

B3011/C3006: Now thinking about the performance of the government in [capital]/president, in general, how good or bad a job do you think the government/ president in [capital] has done over the past [number of years between the previous and the present election OR change in govt.] years. Has it/he/she done a very good job? A good job? A bad job? A very bad job? stfgov: Now thinking about the [country] government, how satisfied are you with the way it is doing its job? Scores from 0 (extremely dissatisfied) to 10 (extremely satisfied) hinctnta: Household’s total net income.

(Continued)

1 Lowest and second household income quintile 2 Third household income quintile 3 Fourth and highest household income quintile

Recoded into three-point scale:

1 1st to 3rd decile 2 4th to 6th decile 3 7th to 10th decile

Recoded into three-point scale:

Recoded into binary variable: Score for 1 for all individuals responding 6 or more, score of 0 otherwise.

Recoded into binary variable: Score of 1 for all individuals responding very good job or good job, score of 0 otherwise.

Loser

Life satisfaction

ESS (2019)

Interest in politics

Vanderbilt University (2019)

ESS (2019)

Vanderbilt University (2019)

ESS (2019)

Vanderbilt University (2019)

Dataset

Variable in analyses

Very Satisfied Somewhat Satisfied Somewhat Dissatisfied Very Dissatisfied

Recoded into binary variable: Score of 1 for all individuals responding very satisfied and somewhat satisfied, score of 0 otherwise.

Recoded into binary variable: Score of 1 for all individuals answering categories 1 or 2, score of 0 otherwise. Recoded into binary variable: Score of 1 for all individuals answering categories 1 or 2, score of 0 otherwise. Recoded into binary variable: Score of 1 for all individuals with a score of 6 or more, score of 0 otherwise.

Recoding for analyses

B12(Country): Coded 1 for all individuals that did Party voted for in last national election? (Country) not vote for a government party Döring and Manow (2018): in the current election, score of 0 Identification of government parties: election_date, otherwise. start _date, party_name, cabinet _party VB20: Coded 1 for all individuals Vote Intention in the Next Presidential Election. More that would not vote for the specifically, for which candidate would you vote if the presidential party, score of 0 presidential elections were held this week? otherwise.

1 2 3 4

stflife: All things considered, how satisfied are you with your life as a whole nowadays? Please answer using this card, where 0 means extremely dissatisfied and 10 means extremely satisfied. ls3: To begin, in general, how satisfied are you with your life?

polintr: How interested would you say you are in politics? (1) Very interested; (2) Quite interested; (3) Hardly interested; (4) Not at all interested pol1: How much interest do you have in politics? (1) A Lot; (2) Some; (3) A Little; (4) None

Original variable and coding in dataset

Low education

CSES (2019)

ESS (2019)

CSES (2019)

1 None 2 Incomplete primary 3 Primary completed 4 Incomplete secondary 5 Secondary completed 6 Post-secondary trade/ Vocational sec. 7 University undergraduate degree incompl. 8 University undergraduate degree compl.

C3023_1/C3023_2: This variable reports the respondent’s vote choice for president. C3023_LH_DC: Vote choice lower house – district candidate. C3023_LH_ PL: Vote choice lower house – party list. C5013_ A to C5013_I: Identification of government parties. Number of portfolios (Government composition) after the election – Party A to I. eisced: Highest level of education. ES -ISCED I , less than lower secondary ES -ISCED II, lower secondary ES -ISCED IIIb, lower-tier upper secondary ES -ISCED IIIa, upper-tier upper secondary ES -ISCED IV, advanced vocational ES -ISCED V1, lower tertiary education ES -ISCED V2, higher tertiary education A2003/B2003/C2003/D2003/e2003: Education of the respondent.

(Continued)

Recoded into binary variable: Score of 1 for all individuals with incomplete secondary or lower, score of 0 otherwise.

Recoded into binary variable: Score of 1 for all individuals with lower tier upper secondary or lower, score of 0 otherwise.

Coded 1 for all individuals that did not vote for a government party, score of 0 otherwise.

ESS (2019)

Satisfaction with democracy

CSES (2019)

Vanderbilt University (2019)

Dataset

Variable in analyses

A3001/B3012/C3019/D3017/e3023: On the whole, are you very satisfied, fairly satisfied, not very satisfied, or not at all satisfied with the way democracy works in [COUNTRY]?

stfdem: And on the whole, how satisfied are you with the way democracy works in [country]? Scores from 0 (extremely dissatisfied) to 10 (extremely satisfied) pn4: In general, would you say that you are very satisfied, satisfied, dissatisfied or very dissatisfied with the way democracy works in (country)?

Original variable and coding in dataset

Recoded into binary variable: Score of 1 for all individuals responding very satisfied and fairly satisfied, score of 0 otherwise. Recoded into binary variable: Score of 1 for all individuals responding very satisfied and fairly satisfied, score of 0 otherwise.

Recoded into binary variable: Score of 1 for all individuals with 7 or more, score of 0 otherwise.

Recoding for analyses

Appendix 177 Table A .5.1  S ample Information Information available for: Country Argentina Australia Bangladesh Belgium Benin Bolivia Brazil Bulgaria Burundi Canada Cape Verde Chile Colombia Costa Rica Croatia Cyprus Czech Republic Denmark Dominican Rep. Ecuador El Salvador Estonia Finland Georgia Germany Greece Guatemala Guinea-Bissau Honduras Hungary India Ireland Israel Italy Jamaica Japan Kenya Korea South Latvia Liberia Lithuania Macedonia Malawi Mauritius Mexico New Zealand Niger Norway Pakistan

Initiation  Debate x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

x x x x x x x x x x x x

x x x

x x x x x x x x x x x x x x x x x x x x x

Veto

PPOP

x x x x x x x x x x x x x x x x x x

0.557 0.421 0.183 0.469 0.498 0.693 0.467 0.688 0.469 0.431

x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

0.593 0.617

0.645 0.717

0.729 0.538 0.824 0.486 0.681 0.795 0.288 0.550 0.521 0.871 0.562 0.633 0.514 0.652 0.824 0.724 0.557 0.350 0.626 0.724 0.450 0.295 (Continued)

178 Appendix Information available for: Country

Initiation Debate

Panama Paraguay Peru Philippines Poland Portugal Romania Senegal Serbia Sierra Leone Slovak Republic Slovenia Solomon Islands Spain Sri Lanka Sweden Switzerland Thailand Trinidad and T. Turkey United Kingdom United States Uruguay Venezuela

x x x x x x x x x x x x x x x x x

x

x x x x x x

x x x x

Veto

PPOP

x x x x x x x x x x x x x x

0.776

x x x x x x x x x

0.621 0.757

x x x x x x x x x x x x x x

0.510 0.605 0.729 0.674 0.757 0.724 0.464 0.710 0.924 0.498 0.398

0.374 0.564 0.507 0.650

Table A .5.2  C orrelations of Variables Bill intro. Bill introduction Agenda setting Amendments Committee structure Committee procedures Executive power Referendums

Agenda Setting

Amend.

Committee structure

Committee procedures

E xecutive power

Ref.

1.00 0.35 0.15 0.42

1.00 0.36 0.22

1.00 0.06

1.00

0.07

−0.00

−0.05

0.29

1.00

0.07

0.08

0.12

0.15

0.07

1.00

0.42

0.31

0.09

0.14

0.16

−0.03

1.00

LAPOP LAPOP

LAPOP LAPOP LAPOP LAPOP LAPOP LAPOP LAPOP LAPOP LAPOP LAPOP LAPOP

Control variables Household income Close to political party Interest in politics Satisfaction with life Political system PR electoral system Political regime Decentralization GDP/capita Corruption Democratic experience

LAPOP

Dependent variable Democracy as the best form of gov.

Independent variables Loser PPOP

Survey

Variable

12,988 12,988 12,988 12,988 12,988 12,988 12,988 12,988 12,988 12,988 12,988

12,988 12,988

12,988

N

Table A .6.1  D escriptive Statistics: Democracy is Preferable

0.526 0.365 0.368 0.879 0.294 0.706 1.827 0.312 1.902 0.312 29.018

0.417 0.596

0.876

Mean

0.499 0.482 0.482 0.326 0.456 0.456 0.563 0.463 0.703 0.463 13.504

0.493 0.085

0.330

Std. Dev.

0

1

1 0 0 1

1

1

Mode

2

1

0

Median

1 0.710

1

Max

0 1 0 1 0 1 0 1 0 1 0 1 0 2 0 1 1 3 0.113 0.751 12 56

0 0.402

0

Min

Appendix 179

Close to political party

Household income

Control variables Low education

Loser

Independent variables PPOP

Dependent variable Satisfaction with dem.

Variable

LAPOP CSES ESS LAPOP CSES ESS LAPOP CSES ESS

LAPOP CSES ESS LAPOP CSES ESS

LAPOP CSES ESS

Survey

22,297 37,241 30,721 22,297 37,241 30,721 22,297 37,241 30,721

22,297 37,241 30,721 22,297 37,241 30,721

22,297 37,241 30,721

N

0.386 0.479 0.380 0.540 1.026 2.192 0.365 0.541 0.588

0.600 0.645 0.675 0.517 0.432 0.449

0.535 0.549 0.388

Mean

Table A .6.2  D escriptive Statistics: Satisfaction with Democracy

0.487 0.499 0.485 0.498 0.889 0.796 0.482 0.498 0.492

0.111 0.141 0.145 0.499 0.495 0.497

0.499 0.498 0.487

Std. Dev.

0 1 1

0 0 0 1

1 0 0

1 1 0

Mode

1 2

Median

0 0 0 0 0 1 0 0 0

0.402 0.426 0.426 0 0 0

0 0 0

Min

1 1 1 1 2 3 1 1 1

0.805 0.924 0.924 1 1 1

1 1 1

Max

180 Appendix

Democratic experience

Corruption

Political system GDP/capita

Decentralization

PR electoral formula

Political regime

Satisfaction with life

Interest in politics

LAPOP ESS LAPOP ESS CSES ESS CSES ESS LAPOP CSES ESS LAPOP LAPOP CSES ESS LAPOP CSES ESS LAPOP CSES ESS

22,297 30,721 22,297 30,721 37,241 30,721 37,241 30,721 22,297 37,241 30,721 22,297 22,297 37,241 30,721 22,297 37,241 30,721 21,121 23,741 30,933

0.390 0.565 0.880 0.793 0.904 0.501 0.671 0.745 0.234 0.432 0.273 1.234 2.082 2.051 2.064 0.488 0.232 0.130 32.168 63.935 53.105

0.487 0.496 0.326 0.405 0.887 0.609 0.470 0.436 0.423 0.495 0.273 0.423 0.769 0.859 0.827 0.203 0.216 0.140 15.816 56.378 40.088 1 1 0 0 0

0 1 1 1

1 2 2 2

1 1

0 1 0 1 0 1 0 1 0 2 0 2 0 1 0 1 0 1 0 1 0 1 1 2 1 3 1 3 1 3 0.082 0.820 0.006 0.704 0.006 0.513 12 58 9 164 13 165

Appendix 181

Interest in politics Satisfaction with life

Close to political party

Household income

Control variables Low education

PPOP

Independent variables Loser

Dependent variable Gov. performance

Variable

CSES ESS CSES ESS CSES ESS ESS ESS

CSES ESS CSES ESS

CSES ESS

Survey

8,576 30,846 8,576 30,846 8,576 30,846 30,846 30,846

8,576 30,846 8,576 30,846

8,576 30,846

N

0.360 0.383 1.074 0.428 0.633 0.586 0.562 0.791

0.388 0.449 0.714 0.675

0.598 0.4.416

Mean

0.480 0.486 0.891 0.495 0.482 0.493 0.496 0.407

0.487 0.497 0.118 0.145

0.490 2.489

Std. Dev.

Table A .6.3  D escriptive Statistics: Satisfaction with Government Performance

1 1 1 1

0 0

0 0

1

Mode

1 0

Median

0 0 0 0 0 0 0 0

0 0 0.426 0.426

0 0

Min

1 1 2 1 1 1 1 1

1 1 0.924 0.924

1 10

Max

182 Appendix

Democratic experience

Corruption

Political regime PR electoral formula Decentralization Political system GDP/capita

ESS ESS ESS CSES CSES ESS CSES ESS CSES ESS

30,846 30,846 30,846 8,576 8,576 30,846 8,576 30,846 8,576 30,846

0.502 0.746 0.271 2.131 1.906 0.465 0.238 0.486 48.379 53.188

0.609 0.435 0.444 0.781 0.825 0.499 0.216 0.499 50.965 41.189 1 0 2 2 0

1

0 2 0 1 0 1 1 3 1 3 0 1 0.019 0.704 0.0 1 13 160 18 169

Appendix 183

Index

Note: Bold page numbers refer to tables; italic page numbers refer to figures and page numbers followed by “n” denote endnotes. Acemoglu, Daron 36–37, 41, 42n6, 42n8 agenda setting 50–51, 102–104, 164n1; collective actor 60n11; by regime type 103 agenda setting power 30, 51, 52–53, 102–103, 116 amendments 16, 50, 104, 105, 164n1; debate power 53–54 Anderson, Christopher J. 7, 79 Argentina 88, 90n18, 124 Bangladesh 60n17, 90n17, 97 Beckman, Lugvig 5 Belgium 60n9, 86, 88, 90n18 Beliaev, Mikhail V. 9 Beramendi, Pablo 78 bicameralism 76–77, 78, 123, 163 bill introduction 51–52, 101–102 Blais, André 33 Blondel, Jean 47 Bolivia 86, 88, 90n18 Bormann, Nils-Christian 80, 170 Bowler, Shaun 79 Canache, Damary 73 Carey, John M. 13, 14, 19n8, 38, 42n7, 50 Cheibub, José Antonio 9, 10, 171 Chile 88, 90n18; Camara de Diputados in 114; distribution of policy-making power of opposition players 114; Lower House Camara de Diputados 114–116; Proyecto de Elites Latinoamericanas project 42n7

citizens: attitudes and opinions 122, 156; consent to democracy 7; and democracies 3–4; role of 5–8; and winners and losers 29, 37–40 Clarke, Harold D. 73 Colombia 90n18; and level of democracy 86–87; Proyecto de Elites Latinoamericanas project 42n7 committee procedures 54–55, 106–107 committee structure 54, 105, 106 Comparative Constitutions Project (Elkins, Ginsburgh and Melton) 58 Comparative Study of Electoral Systems (CSES) 18, 69–71, 74–75, 80, 83–84, 87, 89n8, 130–131, 134–135, 139, 142–145, 151 consensus democracies 4, 31–32, 77–78 consolidation 5; defined 6; and democratic stability 6; political support 71–75 constitutional moment 163 control variables 75–84; corruption 81; decentralization 76–78; democratic experience 81–82; electoral system 79–80; GDP/capita 81; individual background 82–83; individual opinions and interests 83–84; political system 80–81; regime type 79 Coppedge, Michael 81, 169 corruption 81, 124, 126, 131, 134, 143, 145 Cox, Gary W. 53 Czech Republic 17, 95, 111–112; distribution of policy-making

186 Index power of opposition players 116; Lower House Poslanecka Snemovna 116–118 Dahl, Robert A. 47 debate power 53–55; amendments 53–54; committee procedures 54–55; committee structure 54 decentralization 76–78; bicameralism 76–77; federalism 77–78; measure of 78 democracies: as the best form of government 123–129; and citizens 3–4; consensus 4, 31–32, 77–78; and democratic consolidation 30; majoritarian 31; satisfaction with 129–142; stable 6–7, 9, 33; Western 47 democratic consolidation 3–18, 122–153; and democracy 30; future research 161–164; limitations 161– 164; and parliamentary procedures 33–40; study of 161–164; the winnerloser gap 34 democratic experience 81–82 democratic stability 6; and consolidation 6; defined 6 democratization 5; economic approaches to 41; initiated process of 3; legislative development and 19n6; literature 4; research on 10–11, 28; three distinct phases of 18n1 dependent variable 71–75; consolidation 71–75; political support 71–75 Diamond, Larry J. 4, 6, 7, 30, 156 diffuse and specific support 17–18, 30, 37, 39–41, 71–75, 89, 122–124, 128–131, 150–151, 153n5, 158–161 Direct Democracy Database 19n9, 58 dissatisfaction 3–4, 14, 160 Dominican Republic 86, 90n7; Proyecto de Elites Latinoamericanas project 42n7 Donovan, Todd 79 Döring, Herbert 52 Döring, Holger 70 Dutt, Nitish 73 Easton, David 17, 18, 71, 123, 130 Eckstein, Harry 7 Ecuador 86; Proyecto de Elites Latinoamericanas project 42n7

Ekman, Joakim 72 electoral system 79–80 Elgie, Robert 8, 10 Elkins, Zachary 10 El Salvador 90n18; Proyecto de Elites Latinoamericanas project 42n7 empirical strategy 67–69 European Social Survey (ESS) 18, 69–71, 74–75, 83–84, 87, 130–131, 134, 137–139, 142–145, 147, 151, 154n10, 168 executive power 56, 107–108, 108 federalism 76, 77–78, 89n9, 90n12, 123 Fish, M. Steven 19 Fish, Steven M. 31 Fortunato, David 12 Gabriel, Oscar W. 33 Gandhi, Jennifer 171 Gasiorowski, Mark J. 9 GDP/capita 81, 89, 124, 131, 134, 143, 145 Gerring, John 78 Gilligan, Thomas W. 11 Ginsburg, Tom 10 Golder, Matt 80, 170 government: democracy as the best form of 123–129; satisfaction with performance of 142–150 governmental organization 31–32 Guatemala 60n17, 88, 90n16, 118, 124; Proyecto de Elites Latinoamericanas project 42n7 Gurr, Ted Robert 58, 85–86, 170 Holden, Matthew 13 Honduras 90n18; Proyecto de Elites Latinoamericanas project 42n7 Hug, Simon 56 independent variable 70–71; policymaking power of opposition players 70–71; winners and losers 70–71 individual: background 82–83; opinions and interests 83–84 Inglehart, Ronald F. 7 initiation power 51–53; agenda setting power 52–53; bill introduction 51–52 institutionalized opposition power 160 institutions 3–18; future research 161–164; limitations 161–164; study of 161–164

Index 187 Institutions and Elections Project (Regan and Clark) 58 Israel 88, 90n18 Jaggers, Keith 58, 85–86, 170 Jamaica 86, 88, 90n18, 96 Journal of Democracy 19 Karl, Terry Lynn 7 King, Gary 126, 135, 145 Krehbiel, Keith 11 Kroenig, Matthew 31 Kronberg, Allan 73 Latin American Public Opinion Project (LAPOP) 69–71, 74–75, 83–84, 87, 123–124, 130–131, 134–135, 140, 154n10, 159 legislative organization 10–15, 17, 28–29, 49, 57, 68, 88, 95, 157; extra-legislative effects of 158, 161; measuring 158; policy-making process 152, 160; research on 163 legislatures: and governmental organization 31–32; importance of 31–32; role of 8–13 Lijphart, Arend 9, 77–78 Linde, Jonas 72 Linz, Juan J. 4, 8 Lipset, Seymour Martin 7 losers: and citizens 29, 37–40; importance of 32–33 Mainwaring, Scott 9, 18n3, 56 majoritarian democracies 31 Malawi 60n17, 90n17 Manow, Philipp 70 Marshall, Monty G. 58, 85–86, 170 Martin, Shane 163 Mattson, Ingvar 51, 60n12, 103 McKay, Lawrence 38 Mexico 88, 90n16; Proyecto de Elites Latinoamericanas project 42n7 Mishler, William 73 Mondak, Jeffery J. 73 Money, Jeannette 77 Morin-Chassé, Alexandre 33 Nadeau, Richard 33 Nicaragua 90n18; Proyecto de Elites Latinoamericanas project 42n7 Norris, Pippa 72–74, 89n7, 123, 130

O’Donnell, Guillermo 6 opposition power 156–164; and democratic consolidation 161– 164; future research 161–164; institutionalized 160; and institutions 161–164; limitations 161–164; main findings 157–161; measuring 47–59; method of data collection 58; and oppositions 161–164; policy-making power of opposition players 48–58 oppositions: future research 161–164; limitations 161–164; role of 3–18; study of 161–164 Pakistan 60n17, 86, 90n17 Panama 88, 90n18, 123 parliamentarism 31; characteristic of 9; constitutional systems of 31 Parliamentary Power Index (Fish and Kroenig) 31, 58 parliamentary procedures: democracy and democratic consolidation 30; and democratic consolidation 33–40; governmental organization 31–32; legislatures, importance of 31–32; moderating role of 35; political support 29–30; role of 28–41; winners and losers 29, 32–33 Peru 86, 88, 90n17; Proyecto de Elites Latinoamericanas project 42n7 policy influence differential 16, 49–50, 55–56, 59, 60n7, 151 policy-making power of opposition players 48–58, 49, 58, 95–119; agenda setting 102–104; amendments 104, 105; bill introduction 101–102; Chile 114–116; committee procedures 106–107; committee structure 105, 106; Czech Republic 116–118; debate power 53–55; and democratic consolidation 122–153; by democratic experience 100; executive power 107–108, 108; initiation power 51–53; Lower House Camara de Diputados 114–116; Lower House Drzavni Zbor 112–114; Lower House Poslanecka Snemovna 116–118; measuring 57–58; referendums 108–110, 109; Slovenia 112–114; veto power 55–57; winners and losers 70–71 political support 29–30 political system 80–81

188 Index Pop-Eleches, Grigore 152 Power, Timothy J. 9 presidentialism 31; characteristic of 9; constitutional systems of 31 presidentialism-parliamentarism dichotomy 10 presidentialism-parliamentarism distinction 9–10 Pridham, Geoffrey 5 Przeworski, Adam 7, 9, 10, 156 referendums 56–57, 108–110, 109 Robinson, James 36–37, 41, 42n6, 42n8 Rose, Richard 73 Saalfeld, Thomas 38 sample 84–88 satisfaction: with democracy 129–142; with government performance 18, 74–75, 89, 130, 142–150 Schmitter, Philippe C. 7 Seligson, Mitchell A. 73 Shugart, Matthew Søberg 9, 18n3, 56 Sierra Leone 60n17, 90n16 Singh, Shane P. 33, 79 Skach, Cindy 8 Slovenia 17, 95, 97, 117–118; Lower House Drzavni Zbor 112–114

stable democracy 6–7, 9, 33 Stepan, Alfred 4, 8 Strøm, Kaare 16, 39, 41, 42n6, 49, 54, 59, 60n12, 105 Svolik, Milan W. 81 Taylor, Andrew J. 51 Tóka, Gábor 152 Trinidad and Tobago 88, 90n18, 96, 123 Tsebelis, George 31, 56, 77 Tucker, Joshua A. 152 Turkey 90n18; and level of democracy 86–87 Tuttnauer, Or 38 Vanhanen, Tatu 5 Venezuela 86; Proyecto de Elites Latinoamericanas project 42n7 veto power 55–57; executive power 56; referendums 56–57 Vreeland, James 171 Wegmann, Simone 38, 82, 110–111 Welzel, Christian 7 Western democracies 47 winners 29; and citizens 29, 37–40; importance of 32–33