Coalition Governance in Western Europe 0198868480, 9780198868484

Citation preview

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Coalition Governance in Western Europe

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

COMPARATIVE POLITICS Comparative Politics is a series for researchers, teachers, and students of political science that deals with contemporary government and politics. Global in scope, books in the series are characterized by a stress on comparative analysis and strong methodological rigour. The series is published in association with the European Consortium for Political Research. For more information visit http://www.ecprnet.eu The series is edited by Susan Scarrow, John and Rebecca Moores Professor of Political Science, University of Houston, and Jonathan Slapin, Professor of Political Institutions and European Politics, University of Zurich.      Beyond Turnout How Compulsory Voting Shapes Citizens and Political Parties Shane P. Singh Party System Closure Party Alliances, Government Alternatives, and Democracy in Europe Fernando Casal Bértoa Zsolt Enyedi The New Party Challenge Changing Cycles of Party Birth and Death in Central Europe and Beyond Tim Haughton Kevin Deegan-Krause Multi-Level Democracy Integration and Independence Among Party Systems, Parties, and Voters in Seven Federal Systems Lori Thorlakson Citizen Support for Democratic and Autocratic Regimes Marlene Mauk Democratic Stability in an Age of Crisis Agnes Cornell Jørgen Møller Svend-Erik Skaaning Coalition Governance in Central Eastern Europe Edited by Torbjörn Bergman Gabriella Ilonszki Wolfgang C. Müller The Reshaping of West European Party Politics Agenda-Setting and Party Competition in Comparative Perspective Christoffer Green-Pedersen Parliaments in Time The Evolution of Legislative Democracy in Western Europe, 1866–2015 Michael Koß Inequality After the Transition Political Parties, Party Systems, and Social Policy in Southern and Postcommunist Europe Ekrem Karakoç Democracy and the Cartelization of Political Parties Richard S. Katz Peter Mair

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Coalition Governance in Western Europe Edited by TORBJÖRN BERGMAN, HANNA BÄCK, and JOHAN HELLSTRÖM

1

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2021 The moral rights of the authors have been asserted First Edition published in 2021 Impression: 1

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2021934832 ISBN 978–0–19–886848–4 DOI: 10.1093/oso/9780198868484.001.0001 Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A. Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Preface and Acknowledgement This volume presents our joint research on the coalition life cycle in Western Europe. We place particular emphasis on the under-researched coalition governance aspect of that cycle. The work emanates from a project for which Johan Hellström is the principal investigator: PAGED—Party Government in Europe Database (sponsored by Riksbankens Jubileumsfond, Dnr IN15-0306:1). We also gratefully acknowledge the support of Wolfgang C. Müller at the University of Vienna who hosted our first workshop. Hanna Bäck at Lund University hosted a second one. In both cases, the long and strenuous hours during which we jointly discussed indicators and definitions were balanced by pleasant company and gracious hosts. In this research, we have relied on our structured collaboration approach. We presented our contributors with a set of suggested variables and indicators but together we also discussed and scrutinized these variables so that they were defined in a dialogue between the editors and the contributors. This facilitated a common understanding and common use of important variables. From that point of departure, our contributors analysed coalition politics in their country, building not least on sources in their native language, discussions with informed colleagues, and interviews. We believe that the combination of systematic data collection and the skills of authors who know their cases well is a productive one. We include cabinets formed up to and including January 2019. At Umeå University, Jonas Lindahl provided skilled research assistance throughout the project. We are also grateful for the cooperation with the Comparative Politics series editors and the support of the team at Oxford University Press. We believe in transparency and in sharing data. One result of our efforts will be an open access data set, with codebooks and documentation, published online at the European Representative Democracy Data Archive (www.erdda.org). This volume is a product of colleagues working together. It made the work fun! Torbjörn Bergman—Hanna Bäck—Johan Hellström

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Contents List of Figures List of Tables List of Contributors

ix xi xv

1. Coalition Governance in Western Europe Torbjörn Bergman, Hanna Bäck, and Johan Hellström

1

2. The Three Stages of the Coalition Life Cycle Torbjörn Bergman, Hanna Bäck, and Johan Hellström

15

3. Austria: Phasing-Out Grand Coalition Government Wolfgang C. Müller

41

4. Belgium: From Highly Constrained and Complex Bargaining Settings to Paralysis? Lieven De Winter and Patrick Dumont 5. Denmark: How to Form and Govern Minority Coalitions Flemming J. Christiansen

81 124

6. Finland: Forming and Managing Ideologically Heterogeneous Oversized Coalitions Tapio Raunio

165

7. France: Electoral Necessity and Presidential Leadership Beyond Parties Isabelle Guinaudeau and Simon Persico

206

8. Germany: From Stable Coalition Camps to New Complexity Marc Debus, Holger Döring, and Alejandro Ecker 9. Greece: From Coalitions as a ‘State of Exception’ to the New Normal? Myrto Tsakatika

247

284

10. Iceland: Political Change and Coalition Politics Indridi H. Indridason and Gunnar Helgi Kristinsson

324

11. Ireland: Coalition Politics in a Fragmenting Party System Paul Mitchell

357

12. Italy: Continuous Change and Continuity in Change Francesco Zucchini and Andrea Pedrazzani

396

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

viii



13. The Netherlands: Old Solutions to New Problems Tom Louwerse and Arco Timmermans

448

14. Norway: Towards a More Permissive Coalitional Order Torill Stavenes and Kaare W. Strøm

482

15. Portugal: Left-Wing Single-Party Governments and Right-Wing Coalitions Patrícia Calca

517

16. Spain: Single-Party Majority and Minority Governments Bonnie N. Field

544

17. Sweden: The Rise and Fall of Bloc Politics Johan Hellström and Jonas Lindahl

574

18. The United Kingdom: When a Coalition Meets the Westminster Model, Who Wins? Nick Barlow and Tim Bale

611

19. Croatia: Strong Prime Ministers and Weak Coalitions Dario Nikić Čakar

640

20. Coalition Governance Patterns across Western Europe Torbjörn Bergman, Hanna Bäck, and Johan Hellström

680

Appendix: On Definition and Measurement Torbjörn Bergman, Hanna Bäck, Johan Hellström, and Jonas Lindahl

727

Index

749

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

List of Figures 1.1 The life cycle of governments

2

1.2 Effective number of parliamentary parties, 1945–2019

5

2.1 Potential impact of party system change on the coalition life cycle

31

20.1 Trend over the government formation process length in Western Europe 1945–2019

693

20.2 Cabinet formation duration for post-election cabinets in Western Europe 1945–2019

694

20.3 The relationship between parties’ seat contributions and their share of ministerial posts

698

20.4 Power concentration and power dispersion

700

20.5 Size of agreements

706

20.6 Length of coalition agreements in Western Europe

706

20.7 Content of coalition agreements

707

20.8 Share of early terminations in Western Europe, 1945–2019

709

20.9 Trend of cabinet duration Western Europe, 1945–2019

713

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

List of Tables 1.1 Institutions and their changes over time in Western Europe, 1945–2018/19 2.1 Government types in Western Europe, 1945–2018/19

7 17

3.1 Austrian cabinets since 1945

46

3.2 Government formation period in Austria, 1987–2018

49

3.3 Distribution of cabinet ministerships in Austrian coalitions, 1987–2018

57

3.4 Size and content of coalition agreements in Austria, 1945–2018

60

3.5 Coalition governance mechanisms in Austria, 1945–2018

61

3.6 Cabinet termination in Austria, 1987–2018

70

4.1a Belgian cabinets, 1946–2018

87

4.1b Electoral alliances and pre-electoral coalitions in Belgium, 1991–2019

92

4.2 Government formation period in Belgium, 1991–2018

94

4.3 Distribution of cabinet ministerships in Belgian coalitions, 1991–2018

102

4.4 Size and content of coalition agreements in Belgium, 1991–2018

105

4.5 Coalition governance mechanisms in Belgium, 1946–2018

108

4.6 Cabinet termination in Belgium, 1991–2018

116

5.1a Danish cabinets since 1945

132

5.1b Electoral alliances and pre-electoral coalitions in Denmark, 1988–2018

135

5.2 Government formation period in Denmark, 1988–2018

137

5.3 Distribution of cabinet ministerships in Danish coalitions, 1988–2018

141

5.4 Size and content of coalition agreements in Denmark, 1945–2018

144

5.5 Coalition governance mechanisms in Danish coalitions, 1945–2018

146

5.6 Cabinet termination in Denmark, 1988–2018

155

6.1 Finnish cabinets since 1945

170

6.2 Cabinet formation in Finland, 1987–2018

178

6.3 Distribution of cabinet ministerships in Finnish coalitions, 1987–2018

181

6.4 Size and content of coalition agreements in Finland, 1945–2018

185

6.5 Coalition governance mechanisms in Finnish coalitions, 1945–2018

190

6.6 Cabinet termination in Finland, 1987–2018

199

7.1a French cabinets since 1959

209

7.1b Electoral alliances and pre-electoral coalitions in France, 1988–2018

218

7.2 Cabinet formation in France, 1988–2018

221

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

xii

  

7.3 Distribution of cabinet ministerships in French coalitions, 1988–2018

224

7.4 Size and content of coalition agreements in France, 1959–2018

228

7.5 Coalition governance mechanisms in French coalitions, 1959–2018

233

7.6 Cabinet termination in France, 1988–2018

238

8.1 German cabinets since 1949

252

8.2 Cabinet formation in Germany, 1987–2018

257

8.3 Distribution of cabinet ministerships in German coalitions, 1987–2018

262

8.4 Size and content of coalition agreements in Germany, 1949–2018

264

8.5 Coalition governance mechanisms in German coalitions, 1949–2018

267

8.6 Cabinet termination in Germany, 1987–2018

273

9.1a Greek cabinets since 1977

287

9.1b Electoral alliances and pre-electoral coalitions in Greece, 1977–2018

289

9.2 Cabinet formation in Greece, 1977–2018

294

9.3 Distribution of cabinet ministerships in Greek coalitions, 1977–2018

298

9.4 Size and content of coalition agreements in Greece, 1977–2018

301

9.5 Coalition governance mechanisms in Greek coalitions, 1977–2018

303

9.6 Cabinet termination in Greece, 1977–2018

311

10.1a Icelandic cabinets since 1944

331

10.1b Electoral alliances and pre-electoral coalitions in Iceland, 1987–2018

333

10.2 Cabinet formation in Iceland, 1987–2018

335

10.3 Distribution of cabinet ministerships in Icelandic coalitions, 1987–2018

339

10.4 Size and content of coalition agreements in Iceland, 1944–2018

340

10.5 Coalition governance mechanisms in Icelandic coalitions, 1944–2018

344

10.6 Cabinet termination in Iceland, 1987–2018

350

11.1a Irish cabinets since 1944

363

11.1b Electoral alliances and pre-electoral coalitions in Ireland, 1989–2018

368

11.2 Cabinet formation in Ireland, 1989–2018

369

11.3 Distribution of cabinet ministerships in Irish coalitions, 1989–2018

374

11.4 Size and content of coalition agreements in Ireland, 1944–2018

375

11.5 Coalition governance mechanisms in Irish coalitions, 1944–2018

380

11.6 Cabinet termination in Ireland, 1989–2018

387

12.1a Italian cabinets, 1946–2018

400

12.1b Electoral alliances and pre-electoral coalitions in Italy, 1987–2018

405

12.2 Government formation period in Italy, 1987–2018

413

12.3 Distribution of cabinet ministerships in Italian coalitions, 1987–2018

420

12.4 Size and content of coalition agreements in Italy, 1946–2018

424

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

xiii

12.5 Coalition governance mechanisms in Italy, 1946–2018

427

12.6 Cabinet termination in Italy, 1987–2018

435

13.1a Dutch cabinets since 1946

453

13.1b Electoral alliances and pre-electoral coalitions in the Netherlands, 1946–2018

456

13.2 Cabinet formation in the Netherlands, 1989–2018

458

13.3 Distribution of cabinet ministerships in Dutch coalitions, 1989–2018

461

13.4 Size and content of coalition agreements in the Netherlands, 1946–2018

463

13.5 Coalition governance mechanisms in Dutch coalitions, 1946–2018

467

13.6 Cabinet termination in the Netherlands, 1989–2018

474

14.1a Norwegian cabinets since 1945

489

14.1b Electoral alliances and pre-electoral coalitions in Norway, 1989–2019

493

14.2 Government formation period in Norway, 1989–2019

494

14.3 Distribution of cabinet ministerships in Norwegian coalitions, 1989–2019

499

14.4 Size and content of coalition agreements in Norway, 1945–2019

501

14.5 Coalition governance mechanisms in Norwegian coalitions, 1945–2019

504

14.6 Cabinet termination in Norway, 1989–2019

511

15.1a Portuguese cabinets since 1976

521

15.1b Electoral alliances and pre-electoral coalitions in Portugal, 1987–2018

522

15.2 Cabinet formation in Portugal, 1987–2018

524

15.3 Distribution of cabinet ministerships in Portuguese coalitions, 1987–2018

527

15.4 Size and content of coalition agreements in Portugal, 1976–2018

531

15.5 Coalition governance mechanisms in Portuguese coalitions, 1944–2018

535

15.6 Cabinet termination in Portugal, 1987–2018

539

16.1a Spanish cabinets since 1977

550

16.1b Electoral alliances and pre-electoral coalitions in Spain, 1977–2018

556

16.2 Cabinet formation in Spain, 1977–2018

558

16.3 Cabinet termination in Spain, 1977–2018

567

17.1a Swedish cabinets since 1945

578

17.1b Electoral alliances and pre-electoral coalitions in Sweden, 1988–2019

584

17.2 Cabinet formation in Sweden, 1988–2019

586

17.3 Distribution of cabinet ministerships in Swedish coalitions, 1991–2019

594

17.4 Size and content of coalition agreements in Sweden, 1945–2019

596

17.5 Coalition governance mechanisms in Swedish coalition governments, 1945–2018

599

17.6 Cabinet termination in Sweden, 1988–2018

605

18.1a British cabinets since 1945

614

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

xiv

  

18.1b Electoral alliances and pre-electoral coalitions in the United Kingdom, 1945–2018

617

18.2 Cabinet formation in the United Kingdom, 1945–2018

618

18.3 Distribution of cabinet ministerships in British coalitions, 1945–2018

625

18.4 Size and content of coalition agreements in the United Kingdom, 1945–2018

627

18.5 Coalition governance mechanisms in British coalitions, 1945–2018

629

18.6 Cabinet termination in the United Kingdom, 1945–2018

634

19.1a Croatian cabinets since 1990

646

19.1b Electoral alliances and pre-electoral coalitions in Croatia, 1990–2018

651

19.2 Cabinet formation in Croatia, 1990–2018

653

19.3 Distribution of cabinet ministerships in Croatian coalitions, 1991–2018

657

19.4 Size and content of coalition agreements in Croatia, 1991–2018

659

19.5 Coalition governance mechanisms in Croatian coalition governments, 1991–2018

663

19.6 Cabinet termination in Croatia, 1990–2018

671

20.1 Party system indicators in Western Europe and Croatia

682

20.2 Institutional rules and conventions concerning cabinet formation

684

20.3 Prime ministerial powers

689

20.4 Cabinet formation

692

20.5 Parliamentary cabinets

695

20.6 Government participation of parties with median legislator and the largest parliamentary party in Western Europe and Croatia

697

20.7 Coalition governance

701

20.8 Coalition discipline in legislation and parliamentary behaviour

704

20.9 Policy agreements

708

20.10 Mechanisms of cabinet termination

711

20.11 Cabinet duration

714

20.12 Three models and sixteen coalition governance structures

717

20.13 Coalition governance models, empirical manifestations at the country level

719

A.1a Swedish cabinets since 1988

731

A.1b Electoral alliances and pre-electoral coalitions in Sweden, 1991–2019

734

A.2 Cabinet formation in Sweden, 1988–2019

736

A.3 Distribution of cabinet ministerships in Swedish coalitions, 1991–2019

738

A.4 Size and content of coalition agreements in Sweden, 1991–2019

740

A.5 Coalition governance mechanisms in Swedish coalition governments, 1988–2018

742

A.6 Cabinet termination in Sweden, 1988–2018

744

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

List of Contributors Tim Bale is Professor of Politics at Queen Mary University of London, UK. Nick Barlow is a PhD candidate in Politics at Queen Mary University of London, UK. Torbjörn Bergman is Professor of Political Science at Umeå University, Sweden. Hanna Bäck is Professor of Political Science at Lund University, Sweden. Patrícia Calca is a researcher at Iscte – Instituto Universitário de Lisboa, CIES-Iscte, Lisbon, Portugal and visiting fellow at the Center for Data and Methods, University of Konstanz, Germany. Flemming J. Christiansen is Associate Professor of Political Science at Roskilde University, Denmark. Marc Debus is Professor of Political Science at the University of Mannheim, Germany. Lieven De Winter is Professor of Political Science at the University of Louvain, Belgium. Patrick Dumont is Professor of Political Science at The Australian National University, Australia. Holger Döring is Associate Professor of Political Science at the University of Bremen, Germany. Alejandro Ecker Junior Professor for Politics and Communication in Ibero-America at Heidelberg University. Bonnie N. Field is Professor of Global Studies at Bentley University, USA. Isabelle Guinaudeau is a CNRS researcher at Centre Emile Durkheim, Sciences Po Bordeaux, France. Johan Hellström is Associate Professor of Political Science at Umeå University, Sweden. Indridi H. Indridason is Professor of Political Science at the University of California, Riverside, USA. Gunnar Helgi Kristinsson is Professor of Political Science at the University of Iceland, Reykjavik. Jonas Lindahl is a PhD candidate in Political Science at Umeå University, Sweden. Tom Louwerse is Associate Professor of Political Science at Leiden University, the Netherlands. Paul Mitchell is Associate Professor of Political Science at the London School of Economics and Political Science, UK.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

xvi

  

Wolfgang C. Müller is Professor of Democratic Governance at the Department of Government, University of Vienna, Austria. Dario Nikić Čakar is Assistant Professor of Political Science at Zagreb University, Croatia. Andrea Pedrazzani is Assistant Professor of Political Sociology at the University of Milan, Italy. Simon Persico is Professor of Political Science at Pacte, Sciences Po Grenoble, France. Tapio Raunio is Professor of Political Science at Tampere University, Finland. Torill Stavenes has a PhD in Political Science from the University of Exeter, UK. Kaare W. Strøm is Distinguished Professor of Political Science at the University of California, San Diego, USA. Arco Timmermans is Professor of Public Affairs at Leiden University, the Netherlands. Myrto Tsakatika is Senior Lecturer in Politics at the University of Glasgow, UK. Francesco Zucchini is Full Professor of Political Science at the University of Milan, Italy.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

Chapter 1 Coalition Governance in Western Europe Torbjörn Bergman, Hanna Bäck, and Johan Hellström

In parliamentary democracies, the partisan composition of the legislature is the basis for the cabinet (Strøm et al. 2003). At times, parliamentary elections produce a single majority party in parliament and the outcome of the cabinet formation process is largely a foregone conclusion. The UK used to be one example of this. Alternating majority parties controlled the cabinet. However, more recently, the UK has also been governed by a multiparty cabinet and by a government formed by a party that held only a minority of the seats in parliament. Rare as these events have been in the UK, coalitions of political parties govern most of Europe. Across Western Europe, there is considerable empirical variation in the patterns of coalition formation and dissolution. The study of this variation is a vibrant research field, where the largest subfield focuses on the formation of governments (e.g. Riker 1962; Martin and Stevenson 2001; Döring and Hellström 2013). But there is also a substantial literature focusing on the termination of cabinets (e.g. Warwick 1979; Saalfeld 2008; Schleiter and Morgan-Jones 2009; Bassi 2017). For quite some time, these two facets, the coming together and the breaking up of coalition governments, remained largely separate research traditions. In this book, we employ a dynamic coalition politics perspective— specifically, we build on the idea that what happens at the formation stage shapes what happens during the government’s tenure, which in turn influences its durability. This is what Strøm et al. (2008) label the ‘life cycle’ of coalition governments. Our ambition with this volume is three-fold. First, we build on the lessons from earlier studies of governments in both Western and Central Eastern Europe to deepen our understanding of the coalition life cycle. Second, we seek to capture how recent changes in the West European party systems influence the various stages of the coalition life cycle. Third, we are interested in particular in how coalition partners cooperate and make policy once a government has formed, aiming to contribute to the growing literature on the topic of coalition governance, which has, as pointed out by Laver (2012), become an intriguing and challenging topic in coalition politics research. Thus, we look at three phases in the life cycle. We begin with the formation process, for example studying the bargaining rounds, the duration of the formation period, and the total amount of time it takes to form a government. Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Coalition Governance in Western Europe In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

2

̈   ,  ̈  ,   ̈ 

Moreover, during the formation phase, the coalition partners typically design various mechanisms to facilitate cooperation, which are implemented at the governance phase. In the final phase, coalitions eventually end, and when they do, we want to know the reason behind their termination. The three phases of the coalition life cycle applied to cabinets—formation, governance, and termination— are relevant to all parliamentary systems in democratic Europe. In this introductory chapter, we briefly outline the different stages in the coalition life cycle. We then discuss the changing context for government formation in the West European countries, in particular how the party systems have changed over the past few decades. We also briefly present the institutional context in which coalition politics is played out. We conclude by giving a brief overview of the content of our volume.

The life cycle of governments in parliamentary democracies Heuristically, the life cycle of cabinets can be conceptualized as having three major phases. These are illustrated in Figure 1.1.¹ In our illustration, the coalition life cycle starts with the government formation stage.² Focusing on this stage, scholars • (Elections)

GOVERNMENT TERMINATION

GOVERNMENT FORMATION • Bargaining duration/delays • Government composition/size • Payoff allocation (e.g. ministerial portfolios)

• Cabinet duration • Termination types

GOVERNANCE • Coalition governance models • Policy-making of ministers • Control mechanisms (e.g. coalition agreements)

Figure 1.1 The life cycle of governments ¹ For simplicity, we here speak about the three main stages in the coalition cycle, even though the coalition life cycle could be further sub-divided into more stages. Strøm et al. (2008) identify eight different steps in the cycle that all correspond to an identifiable body of literature. ² According to our definition, we count a new government (or cabinet) every time there is (a) a change in party composition, (b) a change in prime minister, or (c) a general election (Müller and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

    

3

have asked why specific types of government form. Where early coalition theories predicted that minimal winning coalitions would form (e.g. Riker 1962), later work has aimed to explain why minority or oversized governments form (e.g. Strøm 1990; Volden and Carrubba 2004; Mitchell and Nyblade 2008; Eppner and Ganghof 2017; Thürk et al. 2020).³ One aspect of government formation which has received considerable interest among scholars is the question of why negotiations over government formation takes a long time in some countries, whereas they are very brief in others (e.g. De Winter and Dumont 2008; Golder 2010; Laver and Benoit 2015; Ecker and Meyer 2017). Other questions focusing on this particular phase relate to the payoffs that political parties negotiate over. Here, a large body of work has focused on the question of why some parties receive more portfolios than others do (e.g. Browne and Franklin 1973; Warwick and Druckman 2005; Cutler et al. 2016). Still others have been interested in parties’ portfolio preferences and in predicting ‘who gets what’ in terms of portfolios (e.g. Bäck et al. 2011; Ecker et al. 2015). The second phase of the coalition life cycle is the period when parties govern and make policy, the coalition governance stage. This phase has received less attention than the first and third phases, but the literature on this topic is growing. An overarching question that is asked is, ‘how do coalitions govern?’ In order to understand this, it is important to know how the coalition contracts negotiated during the formation phase are set up. This means that the distinction between the end of the formation phase and the beginning of the coalition governance phase is analytically useful, but also that it is sometimes difficult to distinguish between the stages. Theoretically, this is in line with research that incorporates actor expectations about the future stages already at the formation stage (Laver and Shepsle 1996; Diermeier 2006; Lupia and Strøm 2008). In the emerging governance literature, several different types of coalition governance model exist to describe and understand government decision-making. One, the ministerial government model, suggests that ministers have complete autonomy to direct policy in their department as they prefer (Laver and Shepsle 1996). Another, the so-called prime minister model, places government authority largely in the hands of the prime minister (PM). A third model emphasizes that coalition governance should be characterized as individual parties negotiating and monitoring each other. An extreme variant of this model is the veto players model

Strøm 2000). A new government may occur after an election, or as a ‘replacement’. A replacement cabinet forms when the previous cabinet ended without there being a new general election held (i.e. changes in the governing parties or of the PM). ³ A minimal winning coalition is one that cannot lose a member and still be winning, that is hold the majority of seats in parliament. A surplus coalition is one that can actually lose a member (the smallest) and still be winning. A minority cabinet is one that include parties that together control less than half (or exactly half) of all the members of parliament (Riker 1962).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

4

̈   ,  ̈  ,   ̈ 

(Tsebelis 2002), which assumes that each coalition partner (party) holds the authority to block initiatives from the other coalition partners. A more useful model for empirical studies nonetheless is a theoretical model in which political parties have the dual incentive to promote their own agenda and also to monitor and shadow what other political parties are pursuing. In this, the so-called coalition compromise model, partners try to constrain ministers from other parties by using various coalition governance mechanisms (e.g. Martin and Vanberg 2014). Studying the third phase of the coalition life cycle, the ‘death’ of governments, several scholars have focused on explaining the duration of cabinets, asking why some cabinets last longer than others (e.g. Saalfeld 2008; Schleiter and MorganJones 2009; Hellström and Walther 2017). In this field, some early explanations focused on so-called structural attributes and how they influence cabinet duration (e.g. Dodd 1976), whereas others have focused on the role of critical events (e.g. Browne et al. 1986; Laver and Shepsle 1998) or the strategic timing of early elections (e.g. Lupia and Strøm 1995; Kayser 2005; Schleiter and Tavits 2016). The three main phases of the coalition life cycle are clearly interlinked. The political environment within which governments form and operate also shapes the different phases. Since the main actors are political parties, changes in contemporary party systems can have an immediate impact on the various phases of the coalition life cycle. For example, changes in the number of parties represented in parliament, and the relationship among these parties, clearly influence this life cycle. In the following section, we discuss such potential changes of the party system context, before turning our attention to the related topic of political institutions.

The party system context of coalition governance in Western Europe The roles of political parties and the party system are well understood in the process of government formation and termination. When election outcomes result in a one-party winning majority the questions of government formation, governance, and termination are often a foregone conclusion. These issues are more relevant in minority situations, when there is no clear majority winner after parliamentary elections—situations that are becoming more common in Western Europe’s increasingly fragmented and polarized party systems. To illustrate the latter, Figure 1.2 shows the increasing fragmentation of party systems in Western Europe in terms of the effective number of parliamentary parties (Laakso and Taagepera 1979). According to this measure, many party systems in Western Europe have seen an increased fragmentation, which is illustrated by the fact that the average value has increased slightly during the post-war period.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

5

9 Belgium 8 7 6 5

Average for Western Europe

4 UK

3

75 19 80 19 85 19 90 19 95 20 00 20 05 20 10 20 15 20 20

70

19

65

19

60

55

19

19

19

50

19

45

2 19

Effective number of parliamentary parties

    

Year

Figure 1.2 Effective number of parliamentary parties, 1945–2019 Notes: The lines in the figure show the effective number of parliamentary parties, which equals where pi is party i’s share of the parliamentary seats. In Western Europe, and as seen in the figure, Belgium has the most fragmented party system and the UK the least. 1 ∑ni¼1 pi2 ,

The most prominent changes in the party systems since the 1990s have been the loss of electoral support for previously strong centrist parties and the formation and rise of new anti-establishment parties and populist parties of the extreme left and/or right. These parties, the weakening of the dominant left–right dimension of conflict, and the increased salience of new policy dimensions have led to higher uncertainty and complexity in bargaining over government and policy. Thus, much of the recent debate about the changing party systems has been about the rise of populist parties.⁴ In this respect, most observers agree that populist parties, on both the left and the right, are now a regular fixture in most Western European parliaments and have both served as support parties to government and held cabinet positions in some countries.⁵ Furthermore, when populist parties do not enter government, they can have a noticeable impact on the party system and the political agenda. In many countries, ⁴ Populism, despite longstanding scholarly attention, can conceptually be difficult to pin down (Mudde 2007). One noted general trait is anti-elitism. In our case, we note a particular scepticism against ‘established’ political parties. In this book, we follow the definition by Mudde (2004). However, determining which parties are ‘populist’ and which are close to being populist is still a contested issue (Mudde 2007), and we recognize that here. ⁵ In our countries, some of the parties that have been labelled as populist, and have either acted as formal support parties in parliament and/or have even held seats in the cabinet, include: FPÖ, BZÖ (Austria); VB (both Vlams Belang and Vlams Blok, Belgium); DFP (Denmark); Finns Party, Blue Reform (Finland); FN (France); AfD (Germany); ANEL, LAOS, SYRIZA (Greece); FI/PdL, LN, M5S (Italy); LPF, PVV (Netherlands); FrP (Norway); Podemos (Spain); ND, SD (Sweden); and UKIP (UK; only returned a single seat after the 2015 election, while returning no seats after the 2017 snap election).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

6

̈   ,  ̈  ,   ̈ 

having large populist parties influences the coalition building among the mainstream parties, effectively making it more difficult to form majorities. They can also have a considerable impact on government policies from a position outside of the cabinet; for example, the Danish People’s Party and Partij voor de Vrijheid (PVV) in the Netherlands were able to influence policy in a significant manner. In the UK, despite only holding a single seat in the House of Commons after the 2015 general election, the UK Independence Party (UKIP) was instrumental in setting the stage for the Brexit referendum, posing as an anti-European Union, antiestablishment challenger to the Conservative and Labour parties.⁶ In sum, rather than being a temporary success, populist parties have become a long-standing presence in many West European national parliaments. Their rise coincides with a decline in popular support from many of the established political parties, or even party families. In this study, we ask our contributors to detail the important changes in the party system and introduce the main newcomer parties in their countries during the last decades. We also ask them to analyse the impact that the newcomers and the other changes in the party system have had on coalition politics. To capture this, we sometimes detail developments since the Second World War (or the democratic breakthrough), but in order not to burden our case descriptions with too much detail, we detail the changes since the late 1980s or early 1990s, depending on the particular country context.⁷

The institutional context of coalition governance in Western Europe As will become clear in the following sections and in the country chapters of this volume, institutional characteristics clearly play an important role in much of the previous scholarly work on the various steps of the coalition life cycle. We therefore dedicate this section to describing some of the important institutions that are likely to influence government formation, governance, and cabinet termination. Early coalition theories, based on game theoretical accounts (e.g. von Neumann and Morgenstern 1953), did not consider the institutional characteristics that may influence which governments form in different countries. However, since the 1990s, most scholars have followed the advice by Strøm et al. (1994: 331), who argued that ‘if we can synthesize our recognition of institutional determinants with a theory of choice under those constraints, we should be able to make considerable progress in our understanding of party coalitions’. In Table 1.1, we present some of the institutional variations across the countries studied here. ⁶ See Ford and Goodwin (2014) for an analysis of the electoral threat posed by UKIP to both parties. ⁷ The full time series are available for all indicators at www.erdda.org.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

    

7

Table 1.1 Institutions and their changes over time in Western Europe, 1945–2018/19 Country

Austria Belgium

Year of first Positive Constructive Bicameral Semicabinet parliamentarism vote of no system presidential under rule confidence system

1945 1946 1995 Denmark 1945 1953 Finland 1945 2002 France 1959 Germany 1949 Greece 1977 1985 Iceland 1944 Ireland 1944 Italy 1946 1948 The Netherlands 1946 2012 Norway 1945 Portugal 1976 1983 Spain 1977 1979 Sweden 1945 1970 United 1945 Kingdom

No Yes No No No No Yes No Yes Yes Yes No Yes Yes Yes No Yes No No No No Yes No No No

No No Yes No No No No No Yes No No No No No No No No No No No No Yes No No No

Yes Yes Yes Yes No No No Yes Yes No No No Yes No Yes Yes Yes No No No Yes Yes Yes No Yes

No No No No No Yes No Yes No Yes No No No No No No No No Yes No No No No No No

Source: Project data (all tables and figures in this volume are based on project data unless otherwise specified).

A lot of work on government formation has focused on the role of investiture requirements, since such rules may ‘hinder the formation of minority governments [ . . . ] because it requires a prospective government to pass a formal vote in the legislature before it can take office’ (Martin and Stevenson 2001: 36). For example, Bergman (1995) argues that the fact that the Swedish Constitution prescribes a negative parliamentary system helps explain the regularity of minority governments in Sweden. In this type of system, to win, a proposal for a new government does not need majority support in the parliament, only that a majority does not actively oppose it. In contrast, majority support in a parliamentary vote is required in a positive parliamentary system. In Table 1.1, we can see that several of the countries included in this book, such as Germany, Greece, and Italy, apply a positive parliamentary rule. Other countries, such as Sweden, Norway, Denmark, and Portugal, apply negative parliamentary rules.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

8

̈   ,  ̈  ,   ̈ 

Several other institutional rules also matter for coalition formation and have been included as important predictors in previous literature. For example, a constructive vote of no confidence, used in Belgium (since 1995), Germany, and Spain (since 1979), has been suggested as an important determinant of cabinet duration. In countries with such a constructive rule, it is more difficult to bring down a government, since the parliament can only withdraw its confidence from a PM if there is majority support for a prospective successor. Institutional variables, such as semi-presidentialism and bicameralism, have been suggested to increase the chances of the formation of majority coalitions (and decrease the probability of minority coalitions). For instance, Druckman et al. (2005) conclude that having a majority in the upper chamber is positively related to the formation of a coalition. They also suggest that the common distinction between ‘strong’ and ‘weak’ upper chambers is of little importance. Upper chambers matter for coalition formation even if they possess few formal powers. In their book, Strøm et al. (2008) also use a broad definition and include all bicameral systems in Western Europe (including, e.g., the Netherlands and the UK) as opposed to only the strong ones, such as Italy. The definition we rely on here suggests that bicameralism exists when there are two legislative parliamentary chambers and the weaker chamber has at least a temporary suspensive veto. The logic is that the very existence of an upper house, even if it has only a limited suspensive veto for some forms of legislation, means that the lower chamber must consider it when forming coalitions. It has also been argued that semi-presidentialism affects both coalition formation (Elgie 1999; Mitchell and Nyblade 2008; Bergman et al. 2015) and cabinet duration (Saalfeld 2008). An often-used criterion to identify semi-presidential system is that there is an elected head of state (e.g. Elgie 1999). We use a stricter criterion based on two conditions: that the president is popularly elected and that the president can have a direct influence on cabinet formation (Strøm et al. 2008). Thus, we classify as semi-presidential Finland from 1945 to 1991, France and Greece before 1985, and Portugal before 1982. In addition, our contributors explore the role of the PM. We return to this in the final chapter. The PM and his or her powers to control the agenda, dismiss ministers, and instruct individual ministers on policy makes for an important institutional variation. Prime ministerial powers may clearly also influence cabinet duration; for example, if a PM has the power to unilaterally dissolve parliament, the risk of early elections should increase (see e.g. Saalfeld 2008). Such powers may also influence other stages of the coalition life cycle, for example the governance stage, where some scholars have suggested that strong PMs are better able to control agency loss (see e.g. Bäck et al. 2019). This feature clearly varies across, and to some extent within, countries in Western Europe, with strong PMs in, for example, the UK, Spain, and Germany and ‘weaker’ PMs in countries such as Belgium and the Netherlands.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

    

9

The importance of the institutional powers and the rules and procedures discussed above varies with the party system and the individuals that work within (and around) them, and thus we expect to find that the recent and ongoing changes in the party systems interact with institutions to shape the coalition life cycle.

Plan of the book Our starting point for each country analysis is the end of the Cold War and the fall of the Berlin Wall, that is around 1990. Of course, earlier periods still shape politics, but for these earlier periods we largely refer to other literature, in particular the sister volume edited by Müller and Strøm (2000).⁸ In the 1990s, after the fall of the wall, the European Union was formally put in place, and new or increasing concerns about migration, integration, and environmental problems have meant that the Western European party systems have changed substantially, with right-wing populists entering the parliamentary arena in most countries. Chapter 2 places our study in the context of the literature on coalition formation in parliamentary democracies. It highlights some of the most important contributions upon which we base our analysis. The chapter describes in depth the different stages of the coalition life cycle, discusses what research questions have been asked in the literature in relation to each stage, and reviews important research on the various stages, for example work on government formation, portfolio allocation, coalition agreements, coalition governance mechanisms, and the duration of cabinets. Following our research agenda, the main part of the book consists of country chapters, which are based on a uniform structure. Each country chapter has six thematic sections: (1) an introductory section that addresses the major changes in the party system that are occurring; (2) a section on the institutional background of coalition politics in the respective country, summarizing the most important institutions that shape coalition politics, in particular important changes in said institutions and the rules of the political game; (3) a section on government formation; (4) a section on governance; and (5) finally a section on cabinet termination. The country chapters end with a brief (6) conclusion that highlights the main features of the chapter and how party system change has influenced coalition governance. This section also discusses what conclusions can be drawn about each case in terms of the coalition governance model used in the country. This volume is based on a cross-national joint effort. In order to present comparable data, we have devoted time and effort to secure that we define and ⁸ In the country chapters, and the concluding chapter, we present some basic information on government features over the entire post-war (or democratic) period.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

10

̈   ,  ̈  ,   ̈ 

measure the important variables consistently across countries. Readers who are not already familiar with variables and measurements that are commonly used in research on political coalitions should find interest in knowing more precisely what definitions and measurement we use in the analysis of the coalition life cycle in Western Europe. For this purpose, we include an extensive appendix on definitions and measurement (see Appendix, this volume). To produce the measurements detailed in the appendix, we have relied on our ‘structured collaboration’ approach to identify the most appropriate and applicable cross-national criteria (cf. Bergman et al. 2005). Thus, our research design follows the one that was the basis for the Müller and Strøm (2000), Strøm et al. (2008), and Bergman et al. (2019) volumes. This means that most of our data have been produced by scholars that are genuine country experts, on parliamentary democracy and coalition politics in their respective countries. The authors of the country chapters in this volume have investigated the general comparative literature covering their countries and then examined (in the relevant national language and with knowledge of the national cultural and institutions) parliamentary rules of procedures, intra-party documents, official government records, archival material, and have conducted elite interviews (e.g. with previous or current ministers), to collect systematic information based on criteria discussed and defined at regular meetings of the research group. In these meetings, the country experts interacted repeatedly for the purpose of discussing, defining, and implementing cross-national operational indicators that were applicable across the set of 17 countries in the study. In this way we believe that we have developed systematic cross-national data that share common standards that have minimal measurement error and a high cross-national validity. Based on our definitions and measurements, the country chapters centre on six tables, each reflecting a major aspect of the coalition life cycle. The structure and content of these tables are presented in detail in the appendix to this volume. In the chapters, the authors use the information in the tables to analyse the formation and termination stages, but the most original empirical contribution is perhaps the coalition governance part in each chapter. In order to write these chapters, our contributors have examined the literature, media, and archival resources, and they have conducted elite interviews with representatives from political parties and ministerial bureaucracies. The interviews illuminate what goes on behind closed doors and are aimed at capturing how cabinet governance actually works and what processes are at play. After the in-depth study of the coalition cycle in each country covering cabinets up to 2019, we conclude with a comparative analysis of our findings and analyse the main patterns of coalition governance across the Western Europe. Here we focus on the important general trends in coalition governance and analyse their connection to ongoing party system change. We present the coalition life cycle in all of Western Europe, with the exception of Luxembourg and Switzerland (where

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

    

11

the latter is not a parliamentary democracy in our terms). Instead, we introduce coalition politics in Croatia—a country that was not included in the Bergman et al. (2019) analysis of Central Eastern Europe. Croatia has successfully transitioned to a stable parliamentary democracy and is now a member of the European Union, and while we do not refer to Croatia as a country in ‘Western’ Europe, including this parliamentary democracy can help shed light on the general aspect of contemporary coalition politics in Europe.

References Bassi, Anna (2017). ‘Policy Preferences in Coalition Formation and the Stability of Minority and Surplus Governments’. The Journal of Politics, 79(1): 250–268. Bäck, Hanna, Jan Teorell, and Staffan I. Lindberg (2019). ‘Cabinets, Prime Ministers, and Corruption: A Comparative Analysis of Parliamentary Governments in PostWar Europe’. Political Studies, 67: 149–70. Bäck, Hanna, Marc Debus, and Patrick Dumont (2011). ‘Who Gets What in Coalition Governments? Predictors of Portfolio Allocation in Parliamentary Democracies’. European Journal of Political Research, 50(4): 441–78. Bergman, Torbjörn (1995). ‘Constitutional Rules and Party Goals in Coalition Formation: An Analysis of Winning Minority Governments in Sweden’. PhD thesis, Department of Political Science, Umeå University. Bergman, Torbjörn, Gabriella Ilonszki, and Wolfgang C. Müller (eds) (2019). Coalition Governance in Central Eastern Europe. Oxford: Oxford University Press. Bergman, Torbjörn, Svante Ersson, and Johan Hellström (2015). ‘Government Formation and Breakdown in Western and Central Eastern Europe’. Comparative European Politics, 13(3): 345–75. Bergman, Torbjörn, Wolfgang C. Müller, and Kaare Strøm (2005). ‘Comparative Parliamentary Democracy: A Project Report’. European Political Science, 4(2): 42–54. Browne, Eric C., and Mark N. Franklin (1973). ‘Aspects of Coalition Payoffs in European Parliamentary Democracies’. American Political Science Review, 67(2): 453–69. Browne, Eric C., John P. Frendreis, and Dennis W. Gleiber (1986). ‘The Process of Cabinet Dissolution: An Exponential Model of Duration and Stability in Western Democracies’. American Journal of Political Science, 30(3): 628–50. Cutler, Josh, Scott De Marchi, Max Gallop, Florian M. Hollenbach, Michael Laver, and Matthias Orlowski (2016). ‘Cabinet Formation and Portfolio Distribution in European Multiparty Systems’. British Journal of Political Science, 46(1): 31–43. De Winter, Lieven, and Patrick Dumont (2008). ‘Uncertainty and Complexity in Cabinet Formation’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

12

̈   ,  ̈  ,   ̈ 

(eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 123–57. Diermeier, Daniel (2006). ‘Coalition Government’. In Barry Weingast and Donald Wittman (eds), Oxford Handbook of Political Economy. New York: Oxford University Press, 162–79. Dodd, Lawrence C. (1976). Coalitions in Parliamentary Government. Princeton, NJ: Princeton University Press. Döring, Holger, and Johan Hellström (2013). ‘Who Gets into Government? Coalition Formation in European Democracies’. West European Politics, 36(4): 683–703. Druckman, James N., Lanny W. Martin, and Michael F. Thies (2005). ‘Influence without Confidence: Upper Chambers and Government Formation’. Legislative Studies Quarterly, 30(4): 529–48. Ecker, Alejandro, Thomas M. Meyer, and Wolfgang C. Müller (2015). ‘The Distribution of Individual Cabinet Positions in Coalition Governments: A Sequential Approach’. European Journal of Political Research 54(4): 802–818. Ecker, Alejandro, and Thomas M. Meyer (2017). ‘Coalition Bargaining Duration in Multiparty Democracies’. British Journal of Political Science, 50(1): 1–20. Elgie, Robert (1999). ‘The Politics of Semi-presidentialism’. In Robert Elgie (ed.), Semipresidentialism in Europe. Oxford: Oxford University Press, 1–21. Eppner, Sebastian, and Steffen Ganghof (2017). ‘Institutional Veto Players and Cabinet Formation: The Veto Control Hypothesis Reconsidered’. European Journal of Political Research 56(1): 169–186. Ford, Robert, and Matthew Goodwin (2014). Revolt on the Right: Explaining Support for the Radical Right in Britain. Abingdon and New York: Routledge. Golder, Sona N. (2010). ‘Bargaining Delays in the Government Formation Process’. Comparative Political Studies, 43(1): 3–32. Hellström, Johan, and Daniel Walther (2017). ‘How Is Government Stability Affected by the State of the Economy? Payoff Structures, Government Type and Economic State’. Government and Opposition, 54(2): 280–308. Kayser, Mark A. (2005). ‘Who Surfs, Who Manipulates? The Determinants of Opportunistic Election Timing and Electorally Motivated Economic Intervention’. American Political Science Review, 99(1): 17–27. Laakso, Markku, and Rein Taagepera (1979). ‘ “Effective” Number of Parties: A Measure with Application to West Europe’. Comparative Political Studies, 12 (1): 3–27. Laver, Michael (2012). ‘Parliaments and Coalitions: The Role of Legislative Institutions in Multiparty Governance’. The Journal of Legislative Studies, 18(1): 113–114. Laver, Michael, and Kenneth A. Shepsle (1996). Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

    

13

Laver, Michael, and Kenneth A. Shepsle (1998). ‘Events, Equilibria, and Government Survival’. American Journal of Political Science, 42(1): 28–54. Laver, Michael, and Kenneth Benoit (2015). ‘The Basic Arithmetic of Legislative Decisions’. American Journal of Political Science, 59(2): 275–91. Lupia, Arthur, and Kaare Strøm (1995). ‘Coalition Termination and the Strategic Timing of Parliamentary Elections’. American Political Science Review, 89(3): 648–65. Lupia, Arthur, and Kaare Strøm (2008). ‘Bargaining, Transaction Costs, and Coalition Governance’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining. The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 51–84. Martin, Lanny W., and Georg Vanberg (2014). ‘Parties and Policymaking in Multiparty Governments: The Legislative Median, Ministerial Autonomy, and the Coalition Compromise’. American Journal of Political Science, 58(4): 979–96. Martin, Lanny W., and Randolph T. Stevenson (2001). ‘Government Formation in Parliamentary Democracies’. American Journal of Political Science, 45(1): 33–50. Mitchell, Paul, and Benjamin Nyblade (2008). ‘Government Formation and Cabinet Type in Parliamentary Democracies’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe, Oxford: Oxford University Press, 201–36. Mudde, Cas (2004). ‘The Populist Zeitgeist’. Government and Opposition, 39(4): 541–63. Mudde, Cas (2007). Populist Radical Right Parties in Europe. Cambridge: Cambridge University Press. Müller, Wolfgang C., and Kaare Strøm (eds) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Riker, William H. (1962). The Theory of Political Coalitions. New Haven: Yale University Press. Saalfeld, Thomas (2008). ‘Institutions, Chance and Choices: The Dynamics of Cabinet Survival in the Parliamentary Democracies of Western Europe (1945–1999)’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 327–68. Schleiter, Petra, and Edward Morgan-Jones (2009). ‘Constitutional Power and Competing Risks: Monarchs, Presidents, Prime Ministers, and the Termination of East and West European Cabinets’. American Political Science Review, 103(3): 496–512. Schleiter, Petra, and Margit Tavits (2016). ‘The Electoral Benefits of Opportunistic Election Timing’. The Journal of Politics, 78(3): 836–850. Strøm, Kaare (1990). Minority Government and Majority Rule. Cambridge: Cambridge University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

14

̈   ,  ̈  ,   ̈ 

Strøm, Kaare, Ian Budge, and Michael J. Laver (1994). ‘Constraints on Cabinet Formation in Parliamentary Democracies’. American Journal of Political Science, 38(2): 303–35. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2008). Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press. Thürk, Maria, Johan Hellström, and Holger Döring (2020). ‘Institutional Constraints on Cabinet Formation: Veto Points and Party System Dynamics’. European Journal of Political Research, 60(2): 295–316. Tsebelis, George (2002). Veto Players: How Political Institutions Work. Princeton, NJ: Princeton University Press. Volden, Craig, and Cliff J. Carrubba (2004). ‘The Formation of Oversized Coalitions in Parliamentary Democracies’. American Journal of Political Science, 48(3): 521–37. Von Neumann, John, and Oskar Morgenstern (1953). Theory of Games and Economic Behavior. Princeton, NJ: Princeton University Press. Warwick, Paul V. (1979). ‘The Durability of Coalition Governments in Parliamentary Democracies’. Comparative Political Studies, 11(4): 465–98. Druckman, James N., and Paul V. Warwick (2005). ‘The Missing Piece: Measuring Portfolio Salience in Western European Parliamentary Democracies’. European Journal of Political Research, 44(1): 17–42.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 2 The Three Stages of the Coalition Life Cycle Torbjörn Bergman, Hanna Bäck, and Johan Hellström

As mentioned in Chapter 1, we look at a series of phases in the life cycle. We begin with the formation process, the length of which often has served as a proxy for the uncertainty and complexity of the bargaining situation. In this process, and as the winning coalition forms, we study the negotiations, the bargaining rounds, and the duration of the formation period. During the formation phase, the coalition partners typically also design various mechanisms to facilitate cooperation between the partners. How this works is captured at the governance phase. In the final phase, coalitions eventually end and here we focus on the reasons of their termination. Relating to these different stages, previous research has focused on a number of interesting research questions. In this chapter, we present the main questions and answers that have been provided in the literature on these different stages. We start out by paying attention to the first stage of the coalition life cycle, that is the birth of governments, focusing on how governments form, continuing with the so-called governance stage, and ending with the termination of governments. Here, and in our country chapters, we focus on the main actors involved in the making and breaking of governments, the political parties. In their pursuit of multiple goals, coalition partners have the dual incentive to cooperate and to compete (Bergman et al. 2013). Moreover, in order to understand the governance part, which can last for years, it is important to study the outcome of the bargaining process between the political parties and how the government coalition was set up at the beginning. This is because how it was set up, and the way it worked during its tenure, also impacts on how long it can stay in power (its duration). This, in turn, can influence government performance and stability (Bergman et al. 2013). In this chapter, we also address the various types of government that have formed across Western Europe. In relation to this, we focus on coalitions, that is cabinets that form between two or more political parties as opposed to those that are formed by a single political party. This does not mean that the governments formed by a single party are uninteresting or that they do not have their governance problems. However, as we will show in more detail, most West European governments consist of multiple parties cooperating in coalitions.

Torbjörn Bergman, Hanna Bäck, and Johan Hellström, The Three Stages of the Coalition Life Cycle In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0002

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

16

̈   ,  ̈  ,   ̈ 

Coalition formation—the birth of governments What specific types of government? Table 2.1 gives an overview of the type of governments in the West European countries in this study. As seen in the table, about seven out of ten governments in Western Europe are coalitions. Coalitions have also been the focus of most of the existing literature on government formation. Modern classics in this field of study, such as Riker (1962), Axelrod (1970), and De Swaan (1973), explain coalition formation in terms of the motives of goal seeking actors. In this respect, they follow in the tradition of Downs (1957). However, while Downs (1957) explained party behaviour based on the assumption that office-seeking politicians maximize their share of the popular vote, Riker (1962) proposed that once it has reached a winning size, an office-seeking party does not have an incentive to increase the coalition any further. This is because increasing the coalition by including additional political parties will mean that each member gets a smaller share of the perks of winning. In contrast to Riker (1962), both Axelrod (1970) and De Swaan (1973) assume that policy pursuit is the main objective of political parties. De Swaan’s argument is that an actor will want to join the winning coalition that would implement a policy that is as close as possible to its own preferred policy position. Axelrod predicts that connected coalitions will form, that ‘consists of adjacent members’ in policy space (Axelrod 1970: 170). The introduction of a policy constraint challenged a crucial assumption in early coalition theory that all combinations of parties are feasible coalitions (Laver 1986: 34). Thus, the introduction of policy pursuit as an important goal for political parties makes the formation of some coalitions more likely than others. When, as in many countries in Western Europe, the left–right socio-economic dimension has been the main conflict dimension, the median legislator party on that conflict dimension has been in a privileged position. If the opposition is strictly divided along the line of one major policy conflict, the median party essentially controls both coalition formation and decision-making in the parliament (de Swaan 1973; Laver and Schofield 1990: 110–13). Of course, while it is true that most governments include the median legislator party (e.g. Laver and Budge 1992; Döring and Hellström 2013), many parliamentary votes do not fall that neatly in a strict one-dimensional space. In a multidimensional policy space, it might be unclear which party is the median legislator party or there might be different median legislator parties on different alternative dimensions. But in other systems and at other times, in a predominantly onedimensional space, the median party is strategically placed to block any alternative government and, simultaneously, the opposition parties are too divided to bypass

28 43 39 51 37 31 21 36 29 64 30 33 19 15 31 27 537

Austria Belgium Denmark Finland France Germany Greece Iceland Ireland Italy Netherlands Norway Portugal Spain Sweden United Kingdom Total

1 2 15 3 1 3 2 4 9 14 0 13 6 11 18 2 104

3.2 4.7 38.5 5.9 2.7 9.7 9.5 11.1 31.0 21.9 0 39.4 31.6 73.3 58.1 7.4 19.4

4 3 0 0 2 1 11 0 6 0 0 6 0 4 2 24 63

12.9 7.0 0 0 5.4 3.2 52.4 0 20.7 0 0 18.2 0 26.7 6.5 88.9 11.7

0 3 20 6 7 0 0 2 5 14 6 8 1 0 5 0 77

Number 0 7.0 51.3 11.8 18.9 0 0 5.6 17.2 21.9 20 24.2 5.3 0 16.1 0 14.3

%

Number

Number

%

Minority

Majority

Minority %

Coalition cabinets

Single-party cabinets

Note: Non-partisan/caretaker cabinets are excluded.

Number of cabinets

Country

Table 2.1 Government types in Western Europe, 1945–2018/19

21 17 4 10 6 22 3 26 7 5 14 6 9 0 6 1 157

Number 67.7 39.5 10.3 19.6 16.2 71.0 14.3 72.2 24.1 7.8 46.7 18.2 47.4 0 19.4 3.7 29.2

%

Minimal winning coalitions

2 18 0 32 21 5 5 4 2 31 10 0 3 0 0 0 133

Number

Oversized coalitions

6.5 41.9 0 62.7 56.8 16.1 23.8 11.1 6.9 48.4 33.3 0 15.8 0 0 0 24.8

% OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

18

̈   ,  ̈  ,   ̈ 

the party in the centre of the policy space. This is one reason why the median party is often included in the winning coalition.

Why do minority and surplus majority governments form? Coalitions usually form when no single party has a majority of its own. Yet, even political parties that control a majority of members of parliament (MPs) have been known to form cabinet coalitions. This has, for example, happened when party cohesion has been low or when the majority has wanted to secure a threshold for constitutional change. But typically, a majority party is also the only cabinet party. When cabinets control a majority of the MPs this legislative majority can, if it is disciplined and cohesive, be sure to win majority votes in the parliament. Minority governments on the other hand have to find other ways to reach the majority threshold in order to legislate and pass budgets. In addition, if the minority cabinet is a coalition, the party leaders have to worry about cohesion not only within the party but also within the ruling coalition and within the legislative majority. In his seminal work, Strøm (1990) explains minority governments from the perspective of the multiple goals of political parties. Political parties are assumed to have three goals; they are simultaneous office-, policy- and vote-seekers.¹ If a party has reason to believe that its voters are going to disapprove of a particular coalition, and also believe it can influence policy from its position in the parliament, this decreases the party’s desire to get into government. In such a case, a party might rather avoid joining a government coalition than entering one that it expects its voters to disapprove of. Yet, as mentioned, all governments, once in power, have to be concerned with passing legislation and other decisions in parliament via majority votes. For this purpose, they need the support of one or more parties in the legislature (Bale and Bergman 2006). Some minority governments find such support parties by making deals with various opposition parties. These are substantive minority governments. Other minority governments, so-called formal minority governments, rely on negotiated agreements with support parties to reach the majority threshold. However, to identify such arrangements, and not to risk conflating them with temporary policy agreements that are a normal part of parliamentary decisionmaking for any minority government, one has to apply fairly strict criteria. Consequently, formal minority governments have negotiated agreements with opposition parties that are (1) comprehensive, (2) explicit, and (3) binding in a

¹ One may also mention a fourth party goal that act as a constraining factor for the other goals, namely the goal to keep the party together, or in other words a party's internal cohesion and its organizational survival (Sjöblom 1968).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

19

longer term.² Substantive minority cabinets lack such agreements with explicit support parties.³ As discussed in the first chapter, institutions play an important role for the type of government that assumes office. In this context, Strøm (1990: 79) argues that the presence of a constitutional requirement of investiture can make the formation of minority government more difficult and, conversely, that less restrictive rules can promote the formation of minority governments (Strøm 1990: 110). Bergman (1993, 1995) further developed the argument that among investiture votes, the distinction between positive and negative parliamentarism is important. Under the former principle, an incoming cabinet (PM) must win a parliamentary majority vote to assume power. Under the latter, a cabinet does not have to win such a parliamentary majority vote, but rather that the parliament does not vote against the cabinet. This principle is associated with a higher frequency of minority governments. In addition, semi-presidentialism, bicameralism, and constructive vote of confidence have been suggested to increase the chances of the formation of majority coalitions, and thus decrease the likelihood of minority cabinets. The logic is similar for both semi-presidentialism and bicameralism as parties need to take into account, already during the coalition formation process, that the cabinet has enough legislative support to implement its policy agenda (Sjölin 1993; Lijphart 2012). Thus, parties have incentives to either include the party of the president in the coalition or, alternatively, build a large enough coalition with sufficient share of seats in parliament to counterbalance a presidential veto. In bicameral systems, as second chambers can block or at least delay legislative proposals, parties tend to build larger majorities to ensure a majority in both chambers (Sjölin 1993; Lijphart 2012; Thürk et al. 2020). Finally, several scholars (e.g. Strøm et al. 1994; Mitchell and Nyblade 2008) have argued that a constructive vote of confidence, that is that a successful non-confidence vote must simultaneously propose an alternative government, largely excludes the formation of minority cabinets. As seen in Table 2.1, among majority governments, almost a quarter are surplus, or oversized, coalitions—coalitions that include one or more parties above the majority threshold. A smaller, related literature focuses on why these governments form and Jungar (2000) summarizes the main three explanations. The first explanation has to do with parliamentary decision rules—surplus coalitions form when a decision in parliament requires a qualified majority and the cabinets parties want to secure a majority beyond the 50% + 1 seat threshold. The second reason is that if party cohesion is low in one or more of the government ² In a British and Irish context these are often called 'confidence and supply agreements of external support'. ³ The definition of 'formal' minority governments is based on Strøm (1990: 62, 95) who found that they are rare but have occurred in Denmark, Finland, France, Israel, and Italy. More recently, they have also occurred in Ireland, Sweden, and the United Kingdom.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

20

̈   ,  ̈  ,   ̈ 

parties, cabinets cannot be sure that the coalition partners can deliver all the votes in crucial decision-making situations in parliament, thus driving them to include additional coalition partners (also see Laver and Schofield 1990). A third explanation focuses on an argument saying that surplus coalitions are a reaction to a crisis, such as a national emergency, and the political parties feel compelled to forge a broad coalition to counteract and handle the situation (also see Baron and Diermeier 2001). In addition, Geys et al. (2006) argue that the existence of strong extreme parties alters the set of feasible coalitions, making the formation of surplus coalitions more likely (also see Volden and Carruba 2004). Thus, according to this argument we should find a lower number of minority cabinets and a higher number of surplus majority coalitions when strong extreme parties are represented in parliament. However, contrary to this expectation, Strøm (1990: 66) argues that a high seat share of extremist parties should lead to more minority cabinets if there is severe policy disagreement amongst the opposition parties, as these cannot agree on feasible policy alternatives. Thus, the potential impact of strong extreme parties on government formation is likely to be conditional on parliamentary polarization (also see Dodd 1976: 68–9; Thürk et al. 2020).

How long does the formation process take and why are there delays? There is large variation between countries in bargaining duration after parliamentary elections or, in other words, the time it takes to form a government after a parliamentary election. In some countries, such as the Netherlands, Belgium, and Austria, this is a lengthy process that takes two to three months, while it takes less than a week in countries such as Greece, France, and the United Kingdom. In order to explain the variation in formation time, Diermeier and van Roozendaal (1998) suggest that formation duration processes covary with measures of uncertainty; for example, post-electoral formation processes and those following a cabinet defeat in parliament are longer than other formation processes. Martin and Vanberg (2003) also study the duration of negotiations and show that there is significant variation across countries; for example, in Denmark and Sweden negotiations usually conclude after a couple of weeks, whereas forming a new government often takes several months in other countries, such as the Netherlands and Belgium. They set out to explain ‘bargaining delays’ and argue that the uncertainty measures suggested by Diermeier and van Roozendaal (1998) are incomplete. Therefore, Martin and Vanberg (2003) add some features that may influence the uncertainty over acceptable offers, and thereby cause delays, focusing on the number of parties and ideological differences between the parties.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

21

Similarly, De Winter and Dumont (2008) focus on the uncertainty and complexity in the formation process and perform a comprehensive analysis of features influencing the occurrence of inconclusive bargaining rounds and the duration of the formation process. Differences in ideological preferences among the bargaining actors are critical to explaining both the number of bargaining rounds and how long bargaining takes. De Winter and Dumont contribute to the literature by introducing important institutional features; for example, they show that supermajority requirements increase formation duration. In addition, PM powers seem to influence duration and the number of bargaining rounds, with strong PMs being able to more quickly conclude the formation process. Focusing on bargaining complexity, measured as the number of parliamentary parties and ideological polarization in the legislature, Golder (2010) finds that delays in forming governments mainly occurs only after elections and not in interelection periods. Using a somewhat different approach to measure the bargaining complexity, Laver and Benoit (2015) have also confirmed its importance for the duration of government formation processes. Furthermore, Ecker and Meyer (2015) analyse data from both Western Europe and from Central and Eastern Europe to test theories explaining delays in government formation and find that uncertainty leads to longer formation processes, as originally suggested by Diermeier and van Roozendaal (1998), but this effect is weaker in Central and Eastern Europe. They also find that bargaining uncertainty conditions the impact of complexity in the bargaining context in Western Europe (see also Golder 2010), whereas no such conditional effects are found in the Central and East European countries.

How are portfolios distributed in coalition governments? To what extent do political parties get what they want in terms of their ideological or programmatic stance when they enter a coalition government? Many scholars would agree that portfolio allocation is central to the government formation process, where the control of relevant ministries in the government is a crucial intervening link between party ideology and government policy output (see e.g. Budge and Keman 1990; Klingemann et al. 1994; Laver and Shepsle 1996). Portfolio allocation certainly is the most straightforward indicator of the payoffs parties bargain for when entering a coalition government, as this is the part of the formation process that is just as important for policy-pursuing as for officeseeking political actors (Bäck et al. 2011). Most previous work on portfolio allocation has focused on predicting how many portfolios each party gets. The proportionality prediction was originally derived from Gamson’s (1961: 376) idea that coalition partners will expect a share of payoffs that is ‘proportional to the amount of resources which they contribute

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

22

̈   ,  ̈  ,   ̈ 

to a coalition’. Drawing on Gamson, Browne and Franklin (1973: 457) argue: ‘The percentage share of ministries received by a party participating in a governing coalition and the percentage share of that party’s coalition seats will be proportional on a one-to-one basis.’ Evaluating this hypothesis on a comparative data set, they demonstrated a near perfect correlation between seat share and portfolio share. Several empirical tests have confirmed that the proportionality relationship is associated with an extremely high explained variance. Therefore, it has been described as one of the strongest relationships in the social sciences and has been dubbed ‘Gamson’s Law’. Warwick and Druckman (2001, 2006) have contributed to this literature by evaluating the idea that all portfolios are not valued equally. Warwick and Druckman (2006: 636) argue that to assume that all ministerial portfolios are equally important ‘is clearly a gross mischaracterization of reality in extant parliamentary systems’. Therefore, Druckman and Warwick (2005) measure portfolio salience by performing an expert survey in 14 European countries and find that the proportionality rule is supported even when portfolio payoffs are weighted by salience. Even though the proportionality relationship has been strongly supported, some deviations have been noted. The main deviation that was found already by Browne and Franklin (1973: 460) is that small parties seem to be receiving ‘ “bonus” ministries to be distributed above their proportional shares’, suggesting that there is a ‘small party bias’ in portfolio allocation. This has been confirmed by Ecker and Meyer (2019) who find that even if the proportionality rule is dominant, there tends to be variation depending of the bargaining power of the parties involved. Furthermore, Falcó-Gimeno and Indridason (2013: 221) argue that in uncertain and complex bargaining situations, ‘parties that otherwise would be in an advantageous bargaining position will have a difficult time exploiting their bargaining advantage’. When analysing portfolio allocation in Western Europe, they find that under such circumstances portfolio allocation will be highly proportional. Contrary to what is assumed in quantitative portfolio allocation research, parties may have specific portfolio preferences since ‘each party has a particular set of policy concerns, seeing control over a specific portfolio as an instrumental means of advancing these’ when parties are policy-seekers (Laver and Schofield 1990: 183). Given a certain degree of ministerial discretion, control over some ministries gives the party an advantage in implementing its preferred policies in the relevant sector (also see Bäck et al. 2011). Browne and Feste (1975: 533) conjectured along these lines, and they classified the preferences of parties belonging to one of eight party families. Budge and Keman (1990) performed a similar analysis, creating a ranking of general party policy interests that was supposed to indicate substantive preferences of parties for specific ministries. Relying on studies on party families, historical studies of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

23

traditional party support groups, and programmatic statements, Budge and Keman (1990: 95) distilled a ranking of portfolio types for five party families. Their propositions were evaluated on a comparative data set, showing that, for instance, when in cabinet Socialist parties get the Labour, Health, and Social Welfare posts more than 80 per cent of the time. More recently, Bäck et al. (2011) build on the work by scholars such as Budge and Keman (1990), presenting a hypothesis based on a policy saliency theory of portfolio allocation. This theory suggests that the more important a party considers a particular policy area, that is the more salient this policy area is in its electoral programme, the more likely it is that this party will try to get the ministerial post controlling this particular policy field. Using data drawn from the Comparative Manifestos Project (see e.g. Budge et al. 2001) and studying portfolio allocation in 12 Western European countries, Bäck et al. (2011) show that parties who stress certain policy issues in their electoral programme are indeed more likely to obtain the portfolio controlling these particular policy areas. However, the most important theoretical contribution about the qualitative aspect of portfolio allocation comes from Laver and Shepsle’s (1996) ‘portfolio allocation model’. In order to solve the ‘chaos problem’ of indefinite cycling of majority preferences between policy proposals when two or more dimensions of policy are important, they relied on the assumption that each policy dimension is governed by a particular cabinet portfolio and that the minister of a department has considerable discretion to act on his or her own. Hence, to be credible, a proposed policy position must correspond to the position of the party assigned to the portfolio that controls this specific dimension. A main implication of this model is that parties controlling the median on a specific dimension that is the jurisdiction of a ministerial post are more likely to obtain the portfolio, since median parties have positions that are preferred by a majority (Bäck et al. 2011). More recently, Ecker et al. (2015) argue that ‘coalition negotiations are sequential choice processes that begin with the allocation of those portfolios most important to the bargaining parties’. Analysing coalition governments in Western and Central Eastern Europe, they show that a sequential logic in the bargaining process results in better predictions than assuming mutual independence in the allocation of portfolios.

How are policy payoffs distributed in coalition governments? Portfolios are only one type of payoffs that parties bargain over when forming a government. The other main payoff is more directly related to the policies that parties bargain over. The question is, to what extent do parties get what they want in terms of their policy priorities? Political parties clearly pursue different policy goals and are forced to compromise on these objectives when forming a coalition.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

24

̈   ,  ̈  ,   ̈ 

While the literature on coalition governments has devoted considerable attention to explaining which parties are likely to form coalitions and to predicting the allocation of ministerial portfolios (see above), the allocation of policy payoffs has largely been neglected. Nonetheless, a few exceptions exist. For instance, Budge and Laver (1993) measure the distribution of policy payoffs by comparing the ideal points of coalition parties expressed in their election manifestos and the policy positions reflected in government declarations (also see Negri 2019). Building on this work, Warwick (2001) uses the same data to compare the left– right positions adopted by coalition governments in government declarations with the left–right ideal points of its constituent parties. He shows that the seat share of parties positively affects policy payoffs and that this effect is more pronounced once controlling for other factors such as formateur status. Däubler and Debus (2009) present an analysis of state-level coalition agreements and election manifestos in Germany and compare the policy positions of coalition governments and their constituent parties on the economic and the social policy dimension. They find that parties’ seat share has a positive effect on the policy payoffs they receive. More recently, Klüver and Bäck (2017) argue that the policy payoffs a coalition party receives depend on how salient particular policy issues are to the party, drawing on previous work by Bäck et al. (2011) on portfolio allocation. Klüver and Bäck show, using a new comparative data set generated by a comprehensive quantitative content analysis of coalition agreements in Western and Eastern Europe, that certain cabinet-level features, more specifically the ideological divisiveness and tangentiality of policy preferences among the partners, affect the policy payoff allocation in coalition governments. We will return to these findings in the final chapter, after having seen how the formation process plays out in our countries, particularly in the context of new challenges and changing party systems. We now turn to describing the literature on the second phase of the coalition life cycle, the so-called governance phase; that is, we now focus on the question of what happens after a government has formed and on questions about how coalitions govern and make policy.

Coalition governance—politics during the life of governments Three different coalition governance models Michael Laver (2012: 113) has succinctly summarized the state of the art in the literature on coalition politics, stressing the importance of coalition governance: Perhaps the most important substantive development in the field over the past decade or so has been a growing interest in coalition governance as opposed to coalition governments. If we want to understand the making and breaking of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

25

governments then we have to understand what happens in between these defining events. Politicians forming governments will anticipate the coalition governance that is implied. Coalition governments that fail mid-term do so because of some failure of coalition governance.

According to our definition, the coalition governance phase covers the period between formation and termination. From the same perspective, Bergman et al. (2019) analyse coalition governance in ten countries in Central Eastern Europe. As we mention in the introductory chapter, three models of coalition governance help structure their and our analysis. The first one, the ministerial government model, is based on a division of power between individual parties and their ministers (Laver and Shepsle 1990, 1996). The second, the coalition compromise model, is based on inter-party compromise between the government parties (Müller and Strøm 2000; Martin and Vanberg 2014). The third, the dominant Prime Minister model, captures when the cabinet is dominated the leading (senior) party and in particular the prime minister (Dunleavy and Rhodes 1990). In the influential works of Laver and Shepsle (1990, 1994, 1996), the ministerial government model (also known as the ministerial ‘policy dictator’ model) ‘emphasizes that the cabinet is not simply a collection of coalition partners, but instead a distribution of specific powers over policy formulation and implementation among those partners’ (Laver and Shepsle 1996: 282). Specifically, Laver and Shepsle (1996) theorized each cabinet minister to have (close to) full power over the government policies in his or her policy jurisdiction. As a result, coalition policy is a mixed bag of individual party policies. Like all parsimonious theories, it does not always correspond well with what has been observed empirically (see e.g. Laver and Shepsle 1994). While cabinet ‘ministers do indeed appear to be functioning as agents of their party rather than as independent actors in their own right’ (Laver and Shepsle 1994: 302), it turns out that coalition partners keep ‘tabs’ on each other through mechanisms such as coalition agreements and policy monitoring (e.g. via junior ministers or shadowing by other ministers). Thus, the dominant model in research on Western Europe, the coalition compromise model, stems from research on conflict management mechanisms within governing coalitions, that is from the coalition governance literature (Müller and Strøm 2000; Thies 2001; Strøm et al. 2008; Martin and Vanberg 2014). It is also sometimes referred to as the ‘collegial’ model (Barbieri and Vercesi 2013). The basic assumption is that the partners credibly commit to policies that in each policy domain are somewhere between the individual cabinet parties’ ideal policies. In the end, credible commitment is possible because most government policies necessitate legislation, the passing of which requires cabinet consent and a parliamentary majority and hence the votes of the coalition parties. Scholars working in this tradition are concerned with how coalition partners try to constrain ministers from other parties via coalition governance mechanisms

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

26

̈   ,  ̈  ,   ̈ 

(e.g. Andeweg and Timmermans 2008), one of which can be the ‘shadowing’ by a junior minister or committee chair from a different party (Thies 2001; Carroll and Cox 2012). Indeed, a main challenge to multiparty governments in Western Europe is to reach collective decisions in situations in which the individual incentives of each coalition party and the minister formally in charge of the policy area might differ from that of the majority in the cabinet. While the ministerial government model makes the simplifying assumption that ministers are unconstrained, the coalition compromise model is likely to underestimate the role of individual ministers. The question of ministers’ impact on policy-making has only recently gained some interest. For example, Becher (2010) finds that labour ministers do influence labour market policy, but only when the distance between veto players is small. Goodhart (2013) finds that monetary policy is more responsive to changes in cabinet leadership than finance ministerial change. Martin and Vanberg (2014) analyse the legislative history of bills and show that coalition policies reflect a compromise rather than the preferences of individual ministers. In a more recent article, Martin and Vanberg (2020) show that the impact of labour market ministers is stronger when ‘legislative policing institutions’ are weak. Although ministers through their office are likely to have considerable advantages to shape the policy process, these can be mitigated if coalitions resort to mechanisms of ex ante coordination, such as coalition contracts and policy agreements, various means of mutual control to reduce the ministers’ informational advantage, and coalition bodies to manage and resolve the conflicts that may arise in the process (Müller and Strøm 2000; Thies 2001; Strøm et al. 2008). As the creation and use of all these coalition governance mechanisms is costly to the government parties, their presence is seen as indicating that the coalition compromise model works. In what can be seen as an extreme variant of this way of understanding coalition politics, Tsebelis’s (1995, 1999, 2002) offers the so-called veto player theory. Veto player theory predicts that, ceteris paribus, policy stability will be the result of a greater number of veto players with greater ideological differences between them. Hence, the model presented by Tsebelis (2002) suggests that ‘each coalition party is a veto player that can maintain the status quo policy position against the demands of its coalition partner or partners’ (Strøm et al. 2010: 523). Here, any party is a veto player, regardless of the allocation of posts and no policy decisions are taken without the support of each partner. Neither an extreme coalition compromise model nor a strict ministerial government model is something that we actually expect to find. These are ideal types to which our countries or individual cabinets can be more or less similar. Moreover, Bergman et al. (2019) show that a third model, dominance by the prime minister, provides perhaps the best fair characterization of the main governance pattern in some of the parliamentary systems in Central Eastern

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

27

Europe (such as Hungary, Slovakia, and Poland). While differences exist in the role played by the prime minister between countries, single-party and coalition cabinets, and individuals—to name a few relevant conditions—in some constellations it has been so important to give rise to the notion of the dominant Prime Minister model (Dunleavy and Rhodes 1990; Müller 1994). This has emerged as one of the ideal types of governance in parliamentary systems (Dunleavy and Rhodes 1990; Müller 1994). Although this model originated from the analysis of single-party cabinets, coalition cabinets with a very imbalanced power structure may be closer to it than to any of the alternative models.

What is the role of policy agreements in coalition governance? Strøm and Müller (1999) with Müller and Strøm (2000) clearly set the agenda for the empirical study of coalition agreements introducing a comparative data set on such contracts in 13 Western European countries (1945–1996), showing that three different points are typically settled in coalition treaties: policy agreements, portfolio allocation, and procedural rules. This data set, which was later extended (Strøm et al. 2008), records whether a coalition government had drafted a formal or informal agreement, provides data on the length of agreements measured by the number of words they contain, and includes a measure for how comprehensively policies were negotiated, measured by the share of the content devoted to policies more generally. Some more recent work focuses on explaining the existence and the overall length of coalition agreements. For example, Indridason and Kristinsson (2013) argue that coalition agreements are a mechanism for avoiding ‘ministerial drift’. They suggest that in cases where the risk of such agency loss is severe, we should expect coalition agreements to be negotiated and to be more extensive. Based on an investigation of coalition agreements in Western Europe, Indridason and Kristinsson (2013: 840) conclude that the length of coalition agreements increases with factors ‘likely to exacerbate the moral hazard problem’, for example as the ideological diversity between partners increases. Also focusing on predicting the length of coalition agreements, Bowler et al. (2016) present a formal model aimed at understanding the trade-off between making comprehensive policy agreements and detailing conflict resolution mechanisms. They evaluate predictions drawn from this model using data on coalition agreements in the German Länder, and they show that when intra-coalition conflict is high, parties write shorter coalition contracts (see also Eichorst 2014). In addition, Klüver and Bäck (2019) argue that the attention parties give to various policy issues in coalition agreements can be seen as a control mechanism to limit ‘ministerial drift’. Using a comparative data set on coalition agreements in West and East European countries from 1945 until 2015, they show that

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

28

̈   ,  ̈  ,   ̈ 

ideological divisiveness among coalition partners positively affects issue attention, as coalition parties have stronger incentives to negotiate a detailed policy agenda that constrains their coalition partners on issues they care about. However, only relatively few studies have focused on evaluating the impact of coalition agreements on policy-making. For example, Moury and Timmermans (2013) study coalition agreements in Germany, Belgium, Italy, and the Netherlands, focusing on whether the agreements include deals over policy issues that the governing parties do not agree on (see also Timmermans 2006). Zubek and Klüver (2015) show that when issues are less divisive among coalition partners, they are also implemented more rapidly. Moury (2011) analyses how coalition agreements constrain the impact of ministers on policy decisions, following the so-called ‘pledge scholars’ approach to analysing policy-making. Bäck et al. (2017) focus on the role of coalition agreements in solving so-called ‘common pool problems’. Analysing spending behaviour in 17 Western European countries, they find support for a conditional hypothesis, suggesting that coalition agreements significantly reduce the negative effect of government fragmentation on spending in institutional contexts where prime ministerial power is low.

Coalition duration—the death of governments How long do cabinets last? The last phase of the coalition life cycle, the ‘death’ of governments, has received considerable scholarly attention and the main interest has been on explaining the duration of cabinets, or the question of why some cabinets last longer than others. In post-war Western Europe, more than half of all governments have ended prematurely; however, there is a considerable variation of cabinet stability both between and within countries. For example, only about one in five of the Italian governments have not ended prematurely, while at the other end less than one in five of the Swedish governments have ended before the next scheduled election. Early attempts to explain variations of the duration of cabinets focused on so called ‘structural attributes’, such as political systems, party systems, and government attributes (e.g. Taylor and Herman 1971; Dodd 1976). For example, as shown by the cabinet duration studies dating back to the 1970s, and empirically confirmed later in several later studies, a government’s likelihood of survival until the next scheduled election is dependent on its type. Majority governments are, for instance, on average more durable than minority governments, and among majority cabinets, minimal winning coalitions are more stable than surplus majority governments (see Grofman and Van Roozendaal 1997). Interestingly, Bergman et al. (2019: 562–3) suggest that this relationship might be a distinctly West European pattern. In the context of more fluid party memberships, in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

29

Central and Eastern Europe oversized (or surplus) governments tend to be more stable. Thus, in this early work cabinet duration was modelled as a function of the attributes of the government, parliament, and general party system, the same factors that allegedly caused the formation of cabinets. Thus, the characteristics of a particular government and parliamentary attributes at the time that it took office was also expected to explain the longevity of the same cabinet, without taking into account the events that actually caused its breakdown (Warwick 1994). Nonetheless, as argued by Saalfeld (2008: 333) some cabinet-specific structural attributes, and most evidently the cabinet formed, may not be ‘exogenous to the explanatory model’ as political parties may choose to form particular cabinets during coalition bargaining because they are on average more durable (also see Diermeier and Merlo 2000). In the 1980s, the focus shifted to the role of critical events and what Warwick termed the ‘survival debates’ (Warwick 1994). The debates were started by Browne et al. (1984, 1986), who argued that the vast majority of government terminations were driven by unpredictable or random external events such as deaths or health problems of prime ministers, economic crises, corruption scandals, or personal conflicts. However, Strøm et al. (1988) criticized their work on both empirical and theoretical grounds, arguing that often events that cause the downfall of governments are neither random nor exogenous as they were often affected by parties within cabinets or parliament. The random-events approach had significant impact on the research to follow, not least methodologically, by moving away from cross-sectional analysis to event-history analysis, also called ‘survival analysis’. Event-history models were further developed by King et al. (1990), who presented a unified statistical model of cabinet duration combining the theoretical arguments made by the structuralattributes approach with the stochastic element suggested by the advocates for random (critical) events (i.e. the Cox proportional hazard model, see Cox 1972). Nonetheless, the question of how to conceptualize and measure critical events was largely ignored by King et al. (1990) but was later further developed instead by Laver and Shepsle (1998) who argued that governments are continuously exposed to different potentially disrupting events but some cabinets are more likely to survive ‘critical events’ or ‘shocks’ than others.

Why do cabinets terminate due to early elections and what is the role of institutions? An important theoretical contribution was made by Lupia and Strøm (1995), who argued that ‘discretionary terminations’ should be divided into two sub-categories— non-electoral replacements and early elections—because party incentives for the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

30

̈   ,  ̈  ,   ̈ 

two termination types are theoretically different and also differ in their likelihood over the term of office. In this respect, early elections refer to governments resigning and calling elections earlier than the next scheduled elections. They may do this for various reasons, for example due to changes in public opinion. However, early elections can be a major gamble for a government and are unlikely early on during a period of office. Early election calling is most likely to occur when the end of the scheduled term of office is approaching, that is when the incumbents do not expect to derive much utility from their remaining time in office (Lupia and Strøm 1995; Diermeier and Stevenson 1999; Laver 2003). Even so, early elections may be called, with the timing designed to maximize incumbent parties’ electoral prospects in relation to changing conditions such as economic booms and busts (e.g. Strøm and Swindle 2002; Smith 2003; Kayser 2005). Thus, governing parties are likely to consider a dissolution at the most favourable time during the parliamentary term, and there may thus be an inverse relationship between a government’s remaining time in office and the likelihood of an early election, at least when the government has the constitutional power to dissolve the legislature. Government replacement is the second main type of premature termination. In extreme cases, when allowed by domestic institutional rules (Laver and Schofield 1990), a replacement occurs when one government is completely replaced by another without an intermediary election. More commonly, however, a replacement cabinet is created when one or more members leave, or others join, a coalition, changing its party composition. In more recent research, one distinctive feature is the increasing emphasis on political institutions and constitutional design. Although Strøm (1985) pointed to the role of institutions—for instance, that cabinets that form in countries with an investiture vote are more durable as it reduces the likelihood of minority cabinets—additional institutional factors have been investigated, such as bicameralism (e.g. Druckman and Thies 2002), semi-presidentialism (e.g. Schleiter and Morgan-Jones 2009; Fernandes and Magalhaes 2016), and the constitutionally mandated powers of the PM—all features that can impact on cabinet stability (e.g. Damgaard 2008; Schleiter and Morgan-Jones 2009).

Expectations and conclusions This chapter has explained the research origins of the questions that we ask in relation to each stage in the coalition life cycle and has described how various features impact on government formation, coalition governance, and cabinet duration. As should have become clear, several features related to the party system, and its changing nature, are expected to influence the different stages in the coalition life cycle. Figure 2.1 summarizes some of the expectations that we

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

31

have on how the party system influences the coalition life cycle in Western European democracies. For example, previous research has shown that a higher number of parties involved in coalition bargaining and increased ideological differences between the parties affect the uncertainty over acceptable offers, and thereby cause bargaining delay (Martin and Vanberg 2003). Scholars have also suggested that the type of cabinets that form will be affected by the polarization of the party system and the presence of ‘extremist’ parties—for example, minority governments may be more likely to form when polarization in parliament is increasing (see e.g. Strøm 1990). Hence, for the first stage of the coalition life cycle, we would expect longer bargaining duration and a higher frequency of minority governments, when bargaining complexity increases, for example due to an increased fragmentation. The second stage of the coalition life cycle has been studied to a lesser extent than the other two stages, so that there is less certainty about what to expect in terms of how this stage is affected by changes in the party system. Some expectations can be formed by considering the indirect effect of party system changes on policy-making via government formation. For example, if we see an increase in minority governments or coalitions including partners with more diverse policy preferences due to the presence of right-wing populist parties in parliament, the policy-making literature and arguments about veto players would suggest that such governments are less able to make policy reforms in an efficient manner, which may lead to lower ‘reform productivity’ (see e.g. Tsebelis 2002; Angelova et al. 2018). However, there may also be other consequences from a changing party system. Drawing on the previous literature on coalition governance, we expect that the use of specific control mechanisms may be influenced by changes in the party system, for example if the system becomes more polarized or extreme parties increase in size. Such effects may also be indirect, operating via the government formation stage; for example, if more diverse coalition governments form, the partners are Government formation: - Increased bargaining duration - More bargaining rounds - Minority governments more likely Party system dynamics: - Rise of right-wing populist parties - Rise of new policy dimensions - Increased polarization - Increased fragmentation

Coalition governance: - Difficulties in making policy reforms - Increased importance of policy agreements - Increased use of control mechanisms Government survival: - Shorter cabinet survival - Higher risk of conflict terminations - Higher risk of early elections

Figure 2.1 Potential impact of party system change on the coalition life cycle

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

32

̈   ,  ̈  ,   ̈ 

likely to make use of various control mechanisms, such as comprehensive coalition agreements, in order to control agency loss that occur when a distant party controls a specific department (Klüver and Bäck 2019). Other consequences may be more direct; for example, the use of ‘shadowing’ committee chairs (see Carroll and Cox 2012) may become more common when parties have higher incentives to control other parties in parliament, which may be the case when parties are operating in a highly polarized setting. Lastly, the third stage of the coalition life cycle may be influenced by changes in the party system, and here the previous literature has come relatively far in analysing the impact of such changes. For example, previous research has shown that the ideological range of the party system and the share of seats controlled by extremist parties affect cabinet survival (Saalfeld 2008). Laver and Schofield (1990: 157) suggest that polarization has a direct causal effect on cabinet durability, by increasing the complexity in the bargaining environment, leading to a ‘distribution of bargaining power that is far more susceptible to slight perturbations’. Another argument for why polarization might matter suggests an indirect causal relationship. Warwick (1994) argues that in the presence of strong extremist parties, parties that are more ideologically diverse have to form coalitions to reach majority support, and these types of cabinet are less durable. Overall, we expect increases in polarization and fragmentation to decrease cabinet survival and increase the likelihood of conflictual terminations and early elections. The following chapters, with the exception of the chapter on Croatia (Chapter 19), deal with coalition politics and governance in the countries of Western Europe. While the present chapter has provided the background to the different concepts used in the respective chapters, as the reader will notice, there is large variation both between and within countries over time on how coalition politics and governance work in practice. In the final chapter we return to these differences, summarize, and present cross-national data of coalitions politics and governance in Western Europe, and analyse potential common trends and patterns of change. However, we now turn to an in-depth look at the coalition life cycle, and its potential changes mainly over the last three decades, in each of our 17 countries. Thus, the purpose of the next 17 chapters is to lay out country features, describe the institutional constraints on coalitions politics, and discuss how recent developments have affected the various stages of the coalition life cycle. Each chapter provides a summary of the most important institutions that shape coalition politics and the rules of the political game and also addresses any major changes in the party system that have an effect on coalition politics, before turning the attention to cabinet life cycle itself, namely government formation, coalition governance, and cabinet termination for the cabinets under study. All in all, the data provided in the respective chapters provide a unique guide to contemporary European parliamentary democracies and help us deepen our understanding of the coalition politics and governance in contemporary Western Europe.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

33

References Andeweg, Rudy B., and Arco Timmermans (2008). ‘Conflict Management in Coalition Government’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 269–300. Angelova, Mariyana, Hanna Bäck, Wolfgang C. Müller, and Daniel Strobl (2018). ‘Veto Player Theory and Reform Making in Western Europe’. European Journal of Political Research, 57(2): 282–307. Axelrod, Robert (1970). Conflict of Interest. Chicago: Markham. Bäck, Hanna, Marc Debus, and Patrick Dumont (2011). ‘Who Gets What in Coalition Governments? Predictors of Portfolio Allocation in Parliamentary Democracies’. European Journal of Political Research, 50(4): 441–78. Bäck, Hanna, Wolfgang C. Müller, and Benjamin Nyblade (2017). ‘Multiparty Government and Economic Policy-Making: Coalition Agreements, Prime Ministerial Power and Spending in Western European Cabinets’. Public Choice, 170(1–2): 33–62. Bale, Tim, and Torbjörn Bergman (2006). ‘A Taste of Honey Is Worse Than None at All? Coping with the Generic Challenges of Support Party Status in Sweden and New Zealand’. Party Politics, 12(2): 189–202. Barbieri, Cristina, and Michelangelo Vercesi (2013). ‘The Cabinet: A Viable Definition and Its Composition in View of a Comparative Analysis’. Government and Opposition, 48(4): 526–47. Baron, David P., and Daniel Diermeier (2001). ‘Elections, Governments, and Parliaments in Proportional Representation Systems’. Quarterly Journal of Economics, 116(3): 933–67. Becher, Michael (2010). ‘Constraining Ministerial Power: The Impact of Veto Players on Labor Market Reforms in Industrial Democracies, 1973–2000’. Comparative Political Studies, 43(1): 33–60. Benoit, Kenneth, and Michael Laver (2006). Party Policy in Modern Democracies. London: Routledge. Bergman, Torbjörn (1993). ‘Formation Rules and Minority Governments’. European Journal of Political Research, 23(1): 55–66. Bergman, Torbjörn (1995). ‘Constitutional Rules and Party Goals in Coalition Formation: An Analysis of Winning Minority Governments in Sweden’. PhD thesis, Department of Political Science, Umeå University. Bergman, Torbjörn, and Kaare Strøm (eds) (2011). The Madisonian Turn: Political Parties and Parliamentary Democracy in Nordic Europe. Ann Arbor: University of Michigan Press. Bergman, Torbjörn, Alejandro Ecker, and Wolfgang C. Müller (2013). ‘How Parties Govern: Political Parties and the Internal Organization of Government’. In Wolfgang C. Müller and Hanne Marthe Narud (eds), Party Governance and Party Democracy. New York: Springer Science+Business Media.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

34

̈   ,  ̈  ,   ̈ 

Bergman, Torbjörn, Gabriella Ilonszki, and Wolfgang C. Müller (eds) (2019). Coalition Governance in Central Eastern Europe. Oxford: Oxford University Press. Bergman, Torbjörn, Svante Ersson, and Johan Hellström (2015). ‘Government Formation and Breakdown in Western and Central Eastern Europe’. Comparative European Politics, 13(3): 345–75. Bowler, Shaun, Thomas Bräuninger, Marc Debus, and Indridi H. Indridason (2016). ‘Let’s Just Agree to Disagree: Dispute Resolution Mechanisms in Coalition Agreements’. Journal of Politics, 78(4): 1264–78. Browne, Eric C., and Karen Ann Feste (1975). ‘Qualitative Dimensions of Coalition Payoffs: Evidence from European Party Governments, 1945–1970’. American Behavioral Scientist, 18(4): 530–56. Browne, Eric C., and Mark N. Franklin (1973). ‘Aspects of Coalition Payoffs in European Parliamentary Democracies’. American Political Science Review, 67(2): 453–69. Browne, Eric C., John P. Frendreis, and Dennis W. Gleiber (1984). ‘An “Events” Approach to the Problem of Cabinet Stability’. Comparative Political Studies, 17 (2): 167–97. Browne, Eric C., John P. Frendreis, and Dennis W. Gleiber (1986). ‘The Process of Cabinet Dissolution: An Exponential Model of Duration and Stability in Western Democracies’. American Journal of Political Science, 11(4): 628–50. Budge, Ian, and Hans Keman (1990). Parties and Democracy: Coalition Formation and Government Functioning in Twenty States. Oxford: Oxford University Press. Budge, Ian, and Michael Laver (1993). ‘The Policy Basis of Government Coalitions: A Comparative Investigation’. British Journal of Political Science, 23(4): 499–519. Budge, Ian, Hans-Dieter Klingemann, Andrea Volkens, Judith Bara, and Eric Tanenbaum, with Richard C. Fording, Derek J. Hearl, Hee Min Kim, Michal D. McDonald, and Silvia M. Mendes (2001). Mapping Policy Preferences: Estimates for Parties, Electors, and Governments 1945–1998. Oxford: Oxford University Press. Carroll, Royce, and Gary W. Cox (2012). ‘Shadowing Ministers: Monitoring Partners in Coalition Governments’. Comparative Political Studies, 45(2): 220–36. Chiba, Daina, Lanny W. Martin, and Randolph T. Stevenson (2015). ‘A Copula Approach to the Problem of Selection Bias in Models of Government Survival’. Political Analysis, 23(1): 42–58. Cox, David R. (1972). ‘Regression Models and Life-Tables’. Journal of the Royal Statistical Society. Series B (Methodological), 34(2): 187–220. Damgaard, Erik (2008). ‘Cabinet Termination in Western Europe’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 301–26. Däubler, Thomas, and Marc Debus (2009). ‘Government Formation and Policy Formulation in the German States’. Regional and Federal Studies, 19(1): 73–95.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

35

De Swaan, Abram (1973). Coalition Theories and Cabinet Formations. Amsterdam: Elsevier. De Winter, Lieven, and Patrick Dumont (2008). ‘Uncertainty and Complexity in Cabinet Formation’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 123–57. Diermeier, Daniel, and Antonio Merlo (2000). ‘Government Turnover in Parliamentary Democracies’. Journal of Economic Theory, 94(1): 46–79. Diermeier, Daniel, and Peter van Roozendaal (1998). ‘The Duration of Cabinet Formation Processes in Western Multi-party Democracies’. British Journal of Political Science, 28(4): 609–26. Diermeier, Daniel, and Randy T. Stevenson (1999). ‘Cabinet Survival and Competing Risks’. American Journal of Political Science, 43(4): 1051–68. Dodd, Lawrence C. (1976). Coalitions in Parliamentary Government. Princeton, NJ: Princeton University Press. Döring, Holger, and Johan Hellström (2013). ‘Who Gets into Government? Coalition Formation in European Democracies’. West European Politics, 36(4): 683–703. Downs, Anthony (1957). An Economic Theory of Democracy. New York: Harper & Row. Druckman, James N., and Michael F. Thies (2002). ‘The Importance of Concurrence: The Impact of Bicameralism on Government Formation and Duration’. American Journal of Political Science, 46(4): 760–71. Druckman, James N., and Paul V. Warwick (2005). ‘The Missing Piece: Measuring Portfolio Salience in Western European Parliamentary Democracies’. European Journal of Political Research, 44(1): 17–42. Druckman, James N., Lanny W. Martin, and Lanny W. Thies (2005). ‘Influence without Confidence: Upper Chambers and Government Formation’. Legislative Studies Quarterly, 30(4): 529–48. Dunleavy, Patrick, and R. A. W. Rhodes (1990). ‘Core Executive Studies in Britain’. Public Administration, 68(1): 3–28. Ecker, Alejandro, and Thomas M. Meyer (2015). ‘The Duration of Government Formation Processes in Europe’. Research and Politics 2(4). https://doi.org/ 10.1177/2053168015622796 Ecker, Alejandro, and Thomas. M. Meyer (2019). ‘Fairness and Qualitative Portfolio Allocation in Multiparty Governments’. Public Choice, 181(3–4): 309–30. https:// doi.org/10.1007/s11127-019-00658-8 Ecker, Alejandro, Thomas M. Meyer, and Wolfgang C. Müller (2015). ‘The Distribution of Individual Cabinet Positions in Coalition Governments: A Sequential Approach’. European Journal of Political Research, 54(4): 802–18. Eichorst, Jason (2014). ‘Explaining Variation in Coalition Agreements: The Electoral and Policy Motivations for Drafting Agreements’. European Journal of Political Research, 53(1): 98–115.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

36

̈   ,  ̈  ,   ̈ 

Elgie, Robert (1999). ‘The Politics of Semi-presidentialism’. In Robert Elgie (ed.), Semi-presidentialism in Europe. Oxford: Oxford University Press, 1–21. Falcó-Gimeno, Albert, and Indridi H. Indridason (2013). ‘Uncertainty, Complexity, and Gamson’s Law: Comparing Coalition Formation in Western Europe’. West European Politics, 36(1): 221–47. Fernandes, Jorge. M., and Pedro. C. Magalhaes (2016). ‘Government Survival in Semipresidential Regimes’. European Journal of Political Research, 55(1): 61–80. Gamson, William A. (1961). ‘A Theory of Coalition Formation’. American Sociological Review, 26(3): 373–82. Geys, Benny, Bruno Heyndels, and Jan Vermeir (2006). ‘Explaining the Formation of Minimal Coalitions: Anti-system Parties and Anti-pact Rules’. European Journal of Political Research, 45(6): 957–84. Golder, Sona N. (2010). ‘Bargaining Delays in the Government Formation Process’. Comparative Political Studies, 43(1): 3–32. Goodhart, Lucy (2013). ‘Who Decides? Coalition Governance and Ministerial Discretion’. Quarterly Journal of Political Science, 8(3): 205–37. Grofman, Bernard, and Peter Van Roozendaal (1997). ‘Modelling Cabinet Durability and Termination’. British Journal of Political Science, 27(3): 419–51. Hartmann, Sebastian (2015). Partisan Policy-making in Western Europe: How Ideology Influences the Content of Government Policies. Wiesbaden: Springer VS. Häusermann, Silja, Georg Picot, and Dominik Geering (2013). ‘Review Article: Rethinking Party Politics and the Welfare State—Recent Advances in the Literature’. British Journal of Political Science, 43(1): 221–40. Hays, Jude C., and Aya Kachi (2009). ‘Interdependent Duration Models in Political Science’. Paper presented at the Annual Meeting of the American Political Science Association, Toronto. Hellström, Johan, and Daniel Walther (2017). ‘How is Government Stability Affected by the State of the Economy? Payoff Structures, Government Type and Economic State’. Government and Opposition, 54(2): 280–308. Hibbs, Douglas A. (1977). ‘Political Parties and Macroeconomic Policy’. American Political Science Review, 71(4): 1467–87. Hibbs, Douglas A. (1987). The Political Economy of Industrial Democracies. Cambridge and London: Harvard University Press. Imbeau, Loius M., François Pétry, and Moktar Lamari (2001). ‘Left-Right Party Ideology and Government Policies: A Meta-analysis’. European Journal of Political Research, 40(1): 1–29. Indridason, Indridi H., and Gunnar H. Kristinsson, (2013). ‘Making Words Count: Coalition Agreements and Cabinet Management’. European Journal of Political Research, 52(6): 822–46. Jungar, Ann-Cathrine (2000). Surplus Majority Government. A Comparative Study of Italy and Finland. Uppsala: Acta Universitatis Upsaliensis.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

37

Kayser, Mark A. (2005). ‘Who Surfs, Who Manipulates? The Determinants of Opportunistic Election Timing and Electorally Motivated Economic Intervention’. American Political Science Review, 99(1): 17–27. Keman, Hans (2006). ‘Party Government Formation and Policy Preferences: An Encompassing Approach?’. In Judith Bara and Albert Weale (eds), Democratic Politics and Party Competition. London: Routledge, 33–55. King, Gary, James E. Alt, Nancy E. Burns, and Michael Laver (1990). ‘A Unified Model of Cabinet Dissolution in Parliamentary Democracies’. American Journal of Political Science, 34(3): 846–71. Klingemann, Hans-Dieter, Richard I. Hofferbert, and Ian Budge, with Hans Keman, Torbjörn Bergman, Francois Pétry, and Kaare Strøm (1994). Parties, Policies, and Democracy. Boulder: Westview Press. Klüver, Heike, and Hanna Bäck (2017). ‘Policy Payoff Allocation in Multiparty Cabinets: Evidence from a Quantitative Content Analysis of Coalition Agreements in Western and Eastern Europe’. Paper presented at the 7th Annual EPSA Conference. Klüver, Heike, and Hanna Bäck (2019). ‘Coalition Agreements, Issue Coverage and Cabinet Governance’. Comparative Political Studies, 52: 1995–2031. Laver, Michael (1986). ‘Between Theoretical Elegance and Political Reality: Deductive Models and Cabinet Coalitions in Europe’. In Geoffrey Pridham (ed.), Coalitional Behaviour in Theory and Practice: An Inductive Model for Western Europe. Cambridge: Cambridge University Press. Laver, Michael (2012). ‘Book Reviews. Parliaments and Coalitions: The Role of Legislative Institutions in Multiparty Governance edited by Lanny Martin and Georg Vanberg. Oxford: Oxford University Press, 2011’. The Journal of Legislative Studies, 18(1): 113–14. Laver, Michael, and Ian Budge (eds) (1992). Party Policy and Coalition Government. London: Macmillan. Laver, Michael, and Kenneth A. Shepsle (1990). ‘Coalitions and Cabinet Government’. American Political Science Review, 84, 873–90. Laver, Michael, and Kenneth A. Shepsle (1994). Cabinet Ministers and Parliamentary Government. Cambridge: Cambridge University Press. Laver, Michael, and Kenneth A. Shepsle (1996). Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Laver, Michael, and Kenneth A. Shepsle (1998). ‘Events, Equilibria, and Government Survival’. American Journal of Political Science, 42: 28–54. Laver, Michael, and Kenneth Benoit (2015). ‘The Basic Arithmetic of Legislative Decisions’. American Journal of Political Science, 59(2): 275–91. Laver, Michael, and Norman Schofield (1990). Multiparty Government. The Politics of Coalition in Europe. Oxford: Oxford University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

38

̈   ,  ̈  ,   ̈ 

Lijphart, Arend (2012). Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, 2nd edn. New Haven and London: Yale University Press. Lupia, Arthur, and Kaare Strøm (1995). ‘Coalition termination and the strategic timing of parliamentary elections’. American Political Science Review, 89(3): 648–65. Lupia, Arthur, and Kaare Strøm (2008). ‘Bargaining, Transaction Costs, and Coalition Governance’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining. The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 51–84. Martin, Lanny W., and Georg Vanberg (2003). ‘Wasting Time? The Impact of Ideology and Size on Delay in Coalition Formation’. British Journal of Political Science, 33(2): 323–32. Martin, Lanny W., and Georg Vanberg (2014). ‘Parties and Policymaking in Multiparty Governments: The Legislative Median, Ministerial Autonomy, and the Coalition Compromise’. American Journal of Political Science, 58(4): 979–96. Martin, Lanny W., and Georg Vanberg (2020). ‘Coalition Government, Legislative Institutions, and Public Policy in Parliamentary Democracies’. American Journal of Political Science, 64(2): 325–40. Martin, Lanny W., and Randolph T. Stevenson (2001). ‘Government Formation in Parliamentary Democracies’. American Journal of Political Science, 45(1): 33–50. Mitchell, Paul, and Benjamin Nyblade (2008). ‘Government Formation and Cabinet Type in Parliamentary Democracies’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: the democratic life cycle in Western Europe. Oxford: Oxford University Press, 201–36. Moury, Catherine (2011). ‘Coalition Agreement and Party Mandate: How Coalition Agreements Constrain the Ministers’. Party Politics, 17(3): 385–404. Moury, Catherine (2013). Coalition Government and Party Mandate: How Coalition Agreements Constrain Ministerial Action. London: Routledge. Moury, Catherine, and Arco Timmermans (2013). ‘Inter-party Conflict Management in Coalition Governments: Analyzing the Role of Coalition Agreements in Belgium, Germany, Italy and the Netherlands’. Politics and Governance, 1(2): 117–31. Mudde, Cas (2004). ‘The Populist Zeitgeist’. Government and Opposition, 39(4): 541–63. Müller, Wolfgang C., and Kaare Strøm (eds) (1999). Policy? Office?, or Votes? How Political Parties Make Hard Choices. Cambridge: Cambridge University Press. Müller, Wolfgang C., and Kaare Strøm (eds) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Negri, Fedra (2017). ‘Who Affects Government Declarations and Why? Contrasting the Left-Right Scale with the Welfare Dimension’. Government and Opposition, 54 (4): 607–36. https://doi.org/10.1017/gov.2017.29 Riker, William H. (1962). The Theory of Political Coalitions. New Haven: Yale University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

       

39

Saalfeld, Thomas (2008). ‘Institutions, Chance and Choices: The Dynamics of Cabinet Survival in the Parliamentary Democracies of Western Europe (1945–1999)’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 327–68. Sanders, David, and Valentine Herman (1977). ‘The Stability and Survival of Governments in Western Europe’. Acta Politica, 12(3): 346–77. Schleiter, Petra, and Edward Morgan-Jones (2009). ‘Constitutional Power and Competing Risks: Monarchs, Presidents, Prime Ministers, and the Termination of East and West European Cabinets’. American Political Science Review, 103(3): 496–512. Schleiter, Petra., and Margit Tavits (2016). ‘The Electoral Benefits of Opportunistic Election Timing’. The Journal of Politics, 78(3): 836–50. Sjöblom, Gunnar (1968). Party Strategies in a Multiparty System. Lund: Studentlitteratur. Sjölin, Mats (1993). Coalition Politics and Parliamentary Power. Lund: Lund University Press. Smith, Alastair (2003). ‘Election Timing in Majoritarian Parliaments’. British Journal of Political Science, 33(3): 397–418. Strøm, Kaare (1985). ‘Party Goals and Government Performance in Parliamentary Democracies’. American Political Science Review, 79(3): 738–54. Strøm, Kaare (1990). Minority Government and Majority Rule. Cambridge: Cambridge University Press. Strøm, Kaare, and Stephen M. Swindle (2002). ‘Strategic Parliamentary Dissolution’. American Political Science Review, 96(3): 575–91. Strøm, Kaare, and Wolfgang C. Müller (1999). ‘The Keys to Togetherness: Coalition Agreements in Parliamentary Democracies’. Journal of Legislative Studies, 5(3–4): 255–82. Strøm, Kaare, Eric C. Browne, John P. Frendreis, and Dennis W. Glieber (1988). ‘Contending Models of Cabinet Stability’. The American Political Science Review, 82(3): 923–41. Strøm, Kaare, Ian Budge, and Michael J. Laver (1994). ‘Constraints on Cabinet Formation in Parliamentary Democracies’. American Journal of Political Science, 38(2): 303–35. Strøm, Kaare, Wolfgang C. Müller, and Daniel Markham Smith (2010). ‘Parliamentary Control of Coalition Governments’. Annual Review of Political Science, 13: 517–35. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2008). Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

40

̈   ,  ̈  ,   ̈ 

Taylor, Michael, and Valentine Herman (1971). ‘Party Systems and Government Stability’. American Political Science Review, 65(1): 28–37. Thies, Michael F. (2001). ‘Keeping Tabs on Partners: The Logic of Delegation in Coalition Governments’. American Journal of Political Science, 45(3): 580–98. Thürk Maria, Johan Hellström, and Holger Döring (2020). ‘Institutional Constraints on Cabinet Formation: Veto Points and Party System Dynamics’. European Journal of Political Research, 60(2): 295–316. Timmermans, Arco (2006). ‘Standing Apart and Sitting Together: Enforcing Coalition Agreements in Multiparty Systems’. European Journal of Political Research, 45(2): 263–83. 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: Princeton University Press. Volden, Craig, and Cliff J. Carrubba (2004). ‘The Formation of Oversized Coalitions in Parliamentary Democracies’. American Journal of Political Science, 48(3): 521–37. Warwick, Paul V. (1994). Government Survival in Parliamentary Democracies. Cambridge: Cambridge University Press. Warwick, Paul V. (2001). ‘Coalition Policy in Parliamentary Democracies: Who Gets How Much and Why’. Comparative Political Studies, 34(10): 1212–36. Warwick, Paul V., and James. N. Druckman (2001). ‘Portfolio Salience and the Proportionality of Payoffs in Coalition Governments’. British Journal of Political Science, 31(4): 627–49. Warwick, Paul V., and James. N. Druckman (2006). ‘The Portfolio Allocation Paradox: An Investigation into the Nature of a Very Strong But Puzzling Relationship’. European Journal of Political Research, 45(4): 635–65. Zubek, Radoslaw, and Heike Klüver (2015). ‘Legislative Pledges and Coalition Government’. Party Politics, 21(4): 603–14.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 3 Austria: Phasing-Out Grand Coalition Government Wolfgang C. Müller

Grand coalition government of the Austrian People’s Party (ÖVP) and the Social Democratic Party (SPÖ) had been the hallmark of post-war Austria, the Second Republic. This chapter describes and analyses the decline and eventual demise of this mode of governing since 1987.¹ In terms of government composition, this period falls in four episodes. First was the cabinets of a 13-year period of ‘grand coalition’ of the SPÖ and ÖVP, beginning with Vranitzky II. It was followed after the 1999 election by a two-term government of the ÖVP and Freedom Party of Austria (FPÖ)/Alliance for the Future of Austria (BZÖ) (the BZÖ established itself as a break-away of the FPÖ ministers and most MPs in the government’s second term). In 2006, Austria returned to the ‘grand coalition’ formula although the term ‘grand’ now was only a historic reminiscence, given the greatly diminished parliamentary support these cabinets enjoyed. This type of government was eventually ‘phased-out’ in an acrimonious struggle towards early elections in 2017. ‘Grand coalition’ was replaced by a centre-right cabinet of the ÖVP and FPÖ (just one seat short of what the traditional grand coalition parties would have mastered between themselves). This coalition by and large worked smoothly, until a personal scandal of the FPÖ leaders (‘Ibizagate’) triggered its fall in 2019. The historic ‘grand coalition’ formula thus seems exhausted but alternatives involving the FPÖ have proved instable as in each case an FPÖ event either caused a sudden end of the government (intra-party revolts in 1986 and 2002, a personal scandal in 2019) or prevented the government to be continued after elections (the 2005 FPÖ split).

¹ This chapter largely builds on primary research. The sources used include official government and party documents, the news media coverage, the personal accounts of leading politicians (in particular Khol 2001, 2009, 2013; Fischer 2003; Vranitzky 2004; Schüssel 2009, 2010; Busek 2014; Mitterlehner 2019), accounts by historians (Kriechbaumer 2014, 2016) and journalists (Pelinka 2000, 2003; Steininger 2007), and, in particular, interviews with 57 coalition ‘insiders’ (often interviewed more than once) holding office in the cabinets Vranitzky II to Faymann III plus several participants in inconclusive coalition negotiations. The interviewees include 43 cabinet members (including 3 chancellors, 7 vicechancellors, and 14 parliamentary party group leaders or staff members). In terms of party functions, the group comprises 10 leaders of coalition parties and many high-ranking functionaries. Wolfgang C. Müller, Austria: Phasing-Out Grand Coalition Government In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0003

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

42

 . ̈  

The chapter shows how routine formateur-led processes of coalition formation resulting in anticipated outcomes contrast with a few instances that significantly deviate. Coalition governance is characterized by the use of an elaborated set of rules and instruments and great continuity over time and different types of coalition. Most coalitions terminated early because of inter-party conflict.

The institutional setting Austria’s constitutional framework is semi-presidential but basically works as a system of (negative) parliamentarism with a head of state who has a few important prerogatives (Müller 2003). These occasionally allow the president to exercise some influence on coalition politics and government policy. The president appoints the chancellor and at the chancellor’s proposal the other cabinet members plus the secretaries of state (junior ministers). Although the constitution contains no restriction whatsoever as to who should be invited to form a government and of which components it should consist, de facto the president has little room for manoeuvre as the new government must face parliament within a week and can be voted out of office at this occasion. Given ‘negative’ parliamentarism (cf. Bergman 1993) the burden of proof rests with the opposition. To unseat a government requires a quorum of at least 50 per cent and a majority of those members of parliament (MPs) who are present. While this would leave room for minority cabinets, two factors have worked against that option. One is Austria’s strict rule-of-law principles that require governments to make policy largely via legislation. Thus, to be effective, governments need permanent parliamentary support to get their bills passed. The other is the large executive advantage vis-à-vis the opposition, which generally has incentivized parties not to tolerate minority cabinets (Müller 2011). Given the government’s needs for permanent parliamentary support all presidents have refrained from appointing a cabinet opposed by a parliamentary majority. The process of building parliamentary majorities for support of a cabinet is structured by the convention that the president hands out a mandate to form a government. The practice has been that it is first given to the leader or ‘chancellor candidate’ of the largest parliamentary party (Welan 1997; Dickinger 1999; Müller 2006b). Yet several presidents in their campaigns have explicitly denied that there is an automatic in choosing the candidate of the largest party. In the observation period, in all but one instance the president’s formation mandate was successfully completed. In 1999 the formation attempt failed and the cabinet formed after freestyle bargaining and without having received a presidential mandate—indeed it was against the president’s explicit preference (see Section ‘Coalition Formation’). The president has more leeway when it comes to appointment to ministerial offices. While bound to the proposals of the designated chancellor, the president

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



43

may reject candidates unlimitedly. Indeed, President Klestil rejected one or two of the FPÖ nominations for ministerial positions. President Van der Bellen communicated semi-publicly that he would not appoint two of the most likely FPÖ ministerial contenders and insisted on splitting control over law enforcement— the Interior and Justice portfolios—among the coalition parties rather than tolerating a one-party (i.e. FPÖ) monopoly. In other instances, such presidential communication had been carried out strictly confidential and became known only much later. Once formed and sworn in the cabinet must present itself to parliament (the Nationalrat) within one week after assuming office. In so doing it does not go through an investiture vote. Yet the opposition can use this occasion to introduce a vote of no confidence. Only the ÖVP–FPÖ government had to face an unsuccessful vote of no confidence (initiated by the Greens) in 2000 when it first presented itself to parliament and only once an individual minister had to do so. Austrian rules of parliamentary procedure do not include a formal vote of confidence. Yet the controversial ÖVP–FPÖ coalition in 2000 found a different way to demonstrate the support it enjoyed: it made its government programme a parliamentary resolution that was then passed with the votes of the government parties. In any case, Austria’s strict rule-of-law principles require governments to make policy largely via legislation. Thus, to be effective, governments need permanent parliamentary support for their bills. Turning them into law formally requires a quorum of one-third of the MPs and a majority of the votes cast—de facto, however, a majority of all MPs. Legislative decision rules constitute some kind of incentive to form coalitions even more broadly based (Müller 1993). Over the years Austria has accumulated a large stock of laws requiring a two-thirds majority in the Nationalrat (Müller 2000). Yet the importance of this supermajority requirement should not be overestimated. For one thing, even the ‘grand coalition’ formula increasingly has ceased to provide two-thirds majorities. For another, most cabinets short of a supermajority have either managed to find two-thirds support with the help of opposition parties or they reduced their ambition levels and passed reforms exhausting the leeway simple majorities have within the constitutional constraints—to be corrected occasionally by the Constitutional Court. Despite the appointment of its members by politicians, this court acts following judicial reasoning and, to the chagrin of governments, has become much more activist and unpredictable since the 1990s (Schaden 2006). As a federal country, the Austrian parliament includes a second chamber, the Bundesrat, representing the Länder, composed of delegations from the Land diets. This may produce majorities different from those in the Nationalrat (if the government is not a grand coalition). Yet the Bundesrat is irrelevant for government survival and save a few exceptions can no more than delay legislation passed by the Nationalrat for a few weeks.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

44

 . ̈  

The parliament is elected by a relatively pure PR system with a four per cent threshold for five (four before 2006) years causing only minor distortions (Müller 1996, 2005). The electoral system does not create incentives for pre-electoral coalitions. Nevertheless, political parties have frequently made clear their coalition preferences before the elections as part of their campaign strategy, trying to exploit the popularity or unpopularity of specific potential coalitions in their potential constituency (Müller 2000). Since 1986, the SPÖ in most elections indicated that it would aim at forming a coalition with the ÖVP and explicitly excluded a coalition with the FPÖ from its options. In contrast, the ÖVP committed to ‘grand coalition’ government only in 1994 and before and after has refrained from excluding any of the options. The small Green and liberal parties typically have excluded only any coalition that would include the FPÖ from the range of their options.

The party system and the actors The ÖVP’s and SPÖ’s historic dominance in the government arena was rooted in the party system (Müller 2006a). At the beginning of this chapter’s observation period, it was mainly structured by the socio-economic left–right dimension (with SPÖ and ÖVP as the poles) but a sociocultural dimension was emerging in form of the Greens being elected to parliament for the first time in 1986. Under party leaders Jörg Haider and Heinz-Christian Strache the FPÖ occupied the opposite pole on the second dimension. The ÖVP represents a mix of Christian-democratic, conservative, and liberal ideas and champions representation of the interests of farmers, the business community, a large part of the public bureaucracy, and the upper echelons of white-collar employees; it maintains interlocking directorates with the interest organizations of these groups. The SPÖ, in turn, combines leftist ideas in the socio-economic realm with liberal ideas in other areas, which, however, are not always shared by their core constituency of blue- and white-collar workers (Wineroither and Kitschelt 2012). The main SPÖ constituency is organized in the trade unions and the Chamber of Labour, which strongly overlap with the party at all levels. The most significant change in the party system with direct impact on coalition politics was the long-term ascent of the FPÖ. After a short period of flirtation with liberalism modelled after the German FDP in the early 1980s the FPÖ turned to a strategy of populist protest, replaced its traditional panGerman nationalism with Austrian Chauvinism, and abandoned its traditional anti-clericalism by an appeal to conservative Christian values and its traditional pro-integrationist perspective on Europe by Euroscepticism (Luther 2003, 2006, 2008). In terms of specific issues driving its success, the FPÖ first focused on the perversions of the ‘grand coalition system’—‘black-red’ (ÖVP-SPÖ) abuses of power, clientelistic networks, and a self-serving ‘political class’—and later

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



45

championed a strong anti-immigration stance, also taking benefit from increasing Euroscepticism in the electorate. The sociocultural dimension, with the FPÖ and the Greens as the opposite poles, became increasingly important in the observation period. The main benefiter was the FPÖ, but its growth was not linear. Rather it consists of two spectacular ups and, in-between, one even more spectacular down (followed by another down falling just outside the observation period in 2019). The FPÖ’s electoral performance is closely tied to its role as either opposition (the ups) or government (the downs). Party system change at the parliamentary level is mainly driven by the parties’ electoral fortunes. However, the observation period is also characterized by increasing party fluidity in terms of MPs resigning from their parties, switching to existing parties, or establishing new ones. Most spectacular and consequential were the FPÖ splits, leading to the foundation of the Liberal Forum (LIF) (in 1993) and the BZÖ (in 2005). After Haider’s death the BZÖ gradually decomposed, which allowed Team Stronach to gain parliamentary representation even before it first contested an election. One of those events, the BZÖ establishment, directly impacted on coalition politics while the establishment of the LIF significantly changed the bargaining environment of the incumbent coalition. In a more aggregate way electoral and parliamentary developments are captured in the effective number of parties (ENP) measure in Table 3.1, which during the second phase of grand coalition rose from 2.63 (1986) to 3.41 (1999), then fell to 2.88 (2002), and rose again under the third grand coalition period to 4.56 (2016).

Coalition formation Cabinets formed In the observation period all cabinets were two-party majority coalitions. All coalitions conformed to the ‘minimal number of parties’ (Leiserson 1966) and ‘minimal winning’ (von Neumann and Morgenstern 1953) expectations (Table 3.1). Only two ÖVP–FPÖ cabinets were ‘minimum winning’ (Gamson 1961; Riker 1962)—Schüssel I and Kurz I. All cabinets included the party with the median legislator in the socio-economic policy dimension and all were connected on the left–right scale. Altogether four parties have held government office in the almost 34 years from early 1987 until mid-2019: the SPÖ, ÖVP, FPÖ, and the FPÖ-breakaway BZÖ. The ÖVP was the permanent party of government. The SPÖ was excluded for a total of eight years. The FPÖ served two stretches in government. The first one lasted five years but was interrupted by an election that shrunk the party to a third

Cabinet

Figl I Figl II Figl III Raab I Raab II Raab III Gorbach I Gorbach II Klaus I Klaus II Kreisky I Kreisky II Kreisky III Kreisky IV Sinowatz Vranitzky I Vranitzky II Vranitzky III Vranitzky IV Vranizky V Klima Schüssel I

Cabinet number

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

1945-12-20 1947-11-20 1949-11-08 1953-04-02 1956-06-29 1959-07-16 1961-04-11 1963-03-27 1964-04-02 1966-04-19 1970-04-21 1971-11-04 1975-10-28 1979-06-05 1983-05-24 1986-06-16 1987-01-21 1990-12-17 1994-11-29 1996-03-12 1997-01-15 2000-02-04

Date in

Table 3.1 Austrian cabinets since 1945

1999-10-03

1986-11-23 1990-10-07 1994-10-09 1995-12-17

1966-03-06 1970-03-01 1971-10-10 1975-10-05 1979-05-06 1983-04-24

1962-11-18

1949-10-09 1953-02-22 1956-05-13 1959-05-10

1945-11-25

ÖVP, SPÖ, KPÖ ÖVP, SPÖ ÖVP, SPÖ ÖVP, SPÖ ÖVP, SPÖ ÖVP, SPÖ ÖVP, SPÖ ÖVP, SPÖ ÖVP, SPÖ ÖVP SPÖ SPÖ SPÖ SPÖ SPÖ, FPÖ SPÖ, FPÖ SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP ÖVP, FPÖ

Election date Party composition of cabinet sur sur mwc mwc mwc mwc mwc mwc mwc maj min maj maj maj mwc mwc mwc mwc mwc mwc mwc mwc

100 97.6 87.3 89.1 94.5 95.2 95.2 95.2 95.2 51.5 49.1 50.8 50.8 51.9 55.7 55.7 85.8 76.5 63.9 67.8 67.8 56.8

Type Cabinet of seat cabinet strength (%) 165 165 165 165 165 165 165 165 165 165 183 183 183 183 183 183 183 183 183 183 183 138

3 3 4 4 4 3 3 3 3 3 3 3 3 3 3 3 4 4 5 5 5 4

2.09 2.09 2.54 2.47 2.22 2.20 2.20 2.20 2.20 2.14 2.15 2.21 2.21 2.22 2.26 2.26 2.63 2.99 3.73 3.48 3.48 3.41

ÖVP ÖVP ÖVP ÖVP ÖVP ÖVP ÖVP ÖVP ÖVP ÖVP ÖVP SPÖ SPÖ SPÖ ÖVP ÖVP ÖVP ÖVP ÖVP ÖVP ÖVP SPÖ

Formal Number of Number of ENP, Median support seats in parties in parliament party in parliament parliament first policy parties dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Schüssel II Gusenbauer Faymann I Faymann II Kern Kurz I

2003-02-28 2007-01-11 2008-12-02 2013-12-16 2016-05-17 2017-12-18 2017-10-15

2002-11-24 2006-10-01 2008-09-28 2013-09-29

ÖVP, FPÖ SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP ÖVP, FPÖ

mwc mwc mwc mwc mwc mwc

53 73.2 59 54.1 55.7 61.7

183 183 183 183 183 183

4 5 5 6 6 5

2.88 3.38 4.27 4.59 4.56 3.60

ÖVP FPÖ FPÖ FPÖ FPÖ FPÖ

Notes: For a list of parties, consult the chapter appendix. Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1945–1998 period. The subsequent period is based on Bakker et al. (2015) and Polk et al. (2017). First policy dimension is economic left–right. Cabinet types: maj = Single-party majority cabinet; min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition. Table legends and variables are further defined in the Appendix, this volume.

23 24 25 26 27 28

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

48

 . ̈  

of its previous strength and ended by the FPÖ leadership’s deserting for the foundation of the BZÖ in 2005. Initially, this did not affect the cabinet, which continued until the regular election in 2006. In a way it was more akin to a party’s name change than the replacement of one party by another given the full continuity in cabinet personnel and the viability of the rump-FPÖ was all but clear.² The second FPÖ participation lasted less than two years and ended with its withdrawal from the Kurz I cabinet after a scandal and coalition conflict.

The formation process Altogether there were ten episodes of coalition formation in the observation period (Table 3.2). Two more new cabinets (Klima, Kern) originated from interelection within-party replacements of the head of government. Of the coalitions, eight instances represent the typical formation story: after a round of information talks with the leaders of all parliamentary parties, the president designated the leader (or chancellor candidate) of the strongest parliamentary party chancellor. Typically, the parties then conducted one or more rounds of consultations with each other. This was flowed by the formateur, backed by a party executive decision of his own party, inviting another party to negotiate a joint government. The invitation then was accepted, again backed by a party executive decision. Then inter-party bargaining unfolded, typically involving a main negotiation arena, where each party was represented by five or six leaders and several specialized sub-groups with an equal number of policy specialists from both parties. The subgroups hammered out the coalition policy programme, considering the real challenges ahead (e.g. budget concerns) and trying to find compromises between conflicting party positions. The main group received interim reports and issued directives for further negotiations to the sub-groups. Eventually it agreed on unresolved issues and, with the help of a smaller editorial committee, ‘packaged’ the results of the sub-groups into the coalition agreement. The main committee also discussed and agreed on the principles of coalition governance and portfolio allocation. The final details of portfolio design and allocation between the parties, however, were typically left to the formateur and the coalition partner’s leader. Generally speaking, those negotiations conducted between the ÖVP and FPÖ were more smooth than those between the SPÖ and ÖVP, which occasionally arrived at a stalemate over some important policy with bargaining failure looming. Nevertheless, all these coalition formation attempts were successfully concluded by the initial party combinations.

² Ironically, the parliamentary party group, for reasons of state party funding, continued under the FPÖ label, uniting the vast majority of MPs who had joined the BZÖ and the two who had remained in the FPÖ. For all the reasons given this chapter does not count a new cabinet from 2005.

1987 1990 1994 1996 1997 2000

2003

2007 2008 2013 2016 2017

Vranitzky II Vranitzky III Vranitzky IV Vranizky V Klima Schüssel I

Schüssel II

Gusenbauer Faymann I Faymann II Kern Kurz I

0 0 0 0 0

2

0 0 0 0 0 2

Number of inconclusive bargaining rounds SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP ÖVP, FPÖ (1) SPÖ, ÖVP (2) SPÖ, ÖVP, FPÖ, GA ÖVP, FPÖ (1) ÖVP, SPÖ (2) ÖVP, GA SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP SPÖ, ÖVP ÖVP, FPÖ

Parties involved in the previous bargaining rounds

Note: Total bargaining duration coded for the 1990–2018 period.

Year in

Cabinet

Table 3.2 Government formation period in Austria, 1987–2018

44 55 42 59 1 7 35 6 4 51 6 87 33 59 1 53

Bargaining duration of individual formation attempt (in days)

102 65 78 0 58

96

59 71 51 86 0 124

Number of days required in government formation

87 33 59 1 65

87

44 55 42 59 1 46

Total bargaining duration

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

50

 . ̈  

The two remaining episodes—1999/2000 and 2002/2003—stand out by accounting for two inconclusive formation rounds each (Table 3.2) and demand separate consideration. They were also the most extraordinary and turbulent coalition formations of the entire post-war period. While on average it took 52 days to form a post-election coalition in the pre-1986 period, the mean formation required 78 days thereafter. The two instances with inconclusive formation attempts rank first and third among the coalition formation durations after 1986. What is particularly interesting is that a breakdown of the two formation situations with multiple attempts shows that the inconclusive rounds between the SPÖ and ÖVP consumed much more time than the ones (between the ÖVP and FPÖ) which followed and were successfully concluded. The time required for SPÖ–ÖVP negotiations, both successfully concluded and failed ones, underlines that considerable tensions between the parties existed since the ‘grand coalition’ formula had been revived in the mid-1980s. It may also reflect that grand coalitions bring together the parties with the strongest organizations and dense relations with a multitude of interest organizations that provide information and exercise influence during the negotiations. These instances of coalition formation hence lend credibility to Luebbert’s (1986:52) assertion that it is the negotiators’ intra-party processes that make coalition negotiations ‘so long, difficult and complex’. In contrast to the spectacular processes of 1999/2000 and 2002/2003, the third instance of coalition formation to be discussed here, the 2006 episode, from the information reported in Table 3.2 looks like a standard case. Yet a closer inspection reveals a formation that was on the razor’s edge and could probably only be concluded due a skilful presidential intervention.

From grand to centre-right coalition: the 1999/2000 episode The government formation following the 1999 elections stands out as the second longest in Austria (after 1962/63) and places itself among the most timeconsuming in post-war Europe. Despite the small number of parties involved, it produced the highest number of inconclusive bargaining rounds in Austria and one of the highest in post-war Europe.³ Before the 1999 elections ÖVP leader Wolfgang Schüssel had attempted to mobilize voters by committing his party to opposition in case it would fall behind the FPÖ. While this strategy probably had some positive electoral effect for the ÖVP (Plasser et al. 2000) it was not fully successful, with the party remaining 415 votes behind the FPÖ. Despite the ÖVP and FPÖ winning an equal number of seats, the ÖVP and the mass media generally interpreted the outcome as the ÖVP ³ See Burkert-Dottolo and Moser 2000; Khol 2001, 2009; Fischer 2003; Pelinka 2003; Schüssel 2009, 2010.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



51

being only the third-strongest party. The SPÖ, in turn, had made another precommitment: as in earlier campaigns since 1986, it had excluded a coalition with the FPÖ. Given the strengths of the individual parties, collectively these precommitments would not have allowed the formation of any majority coalition. The president tried to help the ÖVP to come back into the game by asking the leader of the strongest party, the SPÖ party chairman, to start ‘sounding out’ talks (Sondierungsgespräche) with all parliamentary parties and provided a list with 10 topics that should be explored in these talks. He made clear that the government he wanted was a stable one, that is majority-based, and would enjoy reputation domestically and abroad, what was generally read as support of the grand coalition of SPÖ and ÖVP while excluding the FPÖ. The ÖVP was not very cooperative in these talks with the SPÖ. It initiated parallel talks with the FPÖ and the Greens—several rounds with the FPÖ and just one with the Greens because the Greens were suspicious about the motivations of these talks. The ‘sounding out’ talks were ended by a strong statement of the president: ‘Opposition means opposition’, he announced, and not opposition visà-vis one particular party but coalition with another. He also talked about the danger of new elections. This made the ÖVP to give up its commitment to opposition. The ‘zero’ round of inter-party negotiations thus had been devoted to overcoming the ÖVP’s pre-commitment to opposition. The first ‘real’ bargaining round aimed at returning the grand coalition of the SPÖ and ÖVP. Being conducted according to the pattern described here, it resulted in a policy programme that largely carried the ÖVP imprint, including measures such as raising the early retirement age, making substantial cutbacks in the public sector (e.g. by cutting 9,000 civil servant jobs), and moving Austria closer to NATO. Then the negotiations began to run into troubles. The SPÖ party presidium made the cryptic decision of unanimously accepting that this policy package was the result of the coalition negotiations and that the party leadership would try to convince the social democratic trade unionists to support the deal. In the larger SPÖ party executive body the youth organizations (as usual) and the trade unionists (as the single post-war exception) voted against the package that however was accepted by a large majority. While the conflict between the party leader and the powerful trade union wing was truly exceptional for the highly disciplined SPÖ, the ÖVP party executive’s acceptance of the policy package with four votes against was not. Negotiations now turned to the departmental structure and the allocation of cabinet positions to parties. The ÖVP first demanded a grand reshuffle—basically the change of all line ministries from one party to the other—to signal innovation and at a minimum demanded one of the SPÖ’s core departments: Interior, Social Affairs, or Finance. Later on the ÖVP concentrated on Finance and made this a conditio sine qua non for entering the coalition. According to this proposal the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

52

 . ̈  

ÖVP would exchange the Ministry of Economics for Finance. Later the ÖVP made a final offer, to keep Economics but find an independent expert for Finance. Given the SPÖ internal conflict over the policy package, the ÖVP also issued a procedural demand, requiring that all SPÖ negotiators sign the coalition agreement. One of them was the leader of the social democratic trade unionists, Rudolf Nürnberger. Members of this intra-party group occupied about 15 seats in parliament, making them a group sizeable enough to tip the balance in parliamentary voting. Nürnberger, who was critical about the proposed policy of adding two years to the minimum retirement age, rejected the demand and rather offered his resignation. In the end his position was shared by most of the party leadership, leaving party leader Viktor Klima as the only one who still wanted to continue the negotiations with the ÖVP. Eventually the SPÖ had enough and decided to terminate the coalition negotiations with the ÖVP. Next President Klestil made a bold and ill-informed attempt to prevent the emerging coalition of the ÖVP and FPÖ, attempting to reinstall a grand coalition government of the SPÖ and ÖVP under leaders hand-picked by the president rather than the parties. This attempt occurred behind the scenes. After it had failed, the president encouraged the venturing of forming a viable SPÖ minority cabinet, reappointing Klima as formateur. Officially, the cabinet would be SPÖ-led and include independent experts. Unofficially, however, the other three parliamentary parties—the ÖVP, FPÖ, and Greens—were invited to nominate experts whom they would trust. It was indicated that many individual policies of the package hammered out with the ÖVP would remain on the government’s agenda. While the Greens indicated that they might support individual policies, this attempt was rejected by the ÖVP and FPÖ and therefore terminated. As a consequence, Klima informed the president that his mission of forming a cabinet had failed. The third bargaining round had the ÖVP and FPÖ as the actors, negotiating the formation of a coalition government without having received a formal mandate from the president. President Klestil tried to exercise pressure on ÖVP leader Schüssel to abstain from turning to the FPÖ. Once these attempts failed, rather than appointing a formateur, President Klestil waited until the party chairmen of the ÖVP and FPÖ reported the results of their negotiations and required to be appointed. While the historic evidence is not yet clear about the beginning of the ÖVP–FPÖ negotiations, it is clear that the parties had exchanged important signals in the sounding out talks and actually started their negotiations while Klima made his final attempts at forming a cabinet. Interestingly, the idea of a coalition including the FPÖ led to a wave of warnings and threats by leading politicians of other European countries, not only from the political left but also the centre-right, including the Spanish PM José María Aznar, the Belgian PM Guy Verhofstadt, and the French president Jacques Chirac; eventually with the coordination of all other 14 EU member states it was

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



53

announced that the taking office of this coalition would be followed by a number of bilateral diplomatic sanctions such as having no political-level contacts and not voting for Austrian candidates in international organizations. Whatever the motives may have been, this strategy was ill-conceived to achieve its stated purpose.⁴ For one thing the ÖVP–FPÖ negotiations were successfully concluded and the government sworn in. Moreover, the sanctions steered Austrian nationalism, caused a kind of rallying around the flag, crippled the parliamentary opposition, and made domestic life easier for the government (as did partly violent street demonstrations that undermined the legitimacy of the parliamentary opposition).⁵ While internationally the storm gathered over the emerging coalition and street protest began at home, the ÖVP and FPÖ had demonstrated honeymoon feelings to the public, making the coalition negotiations a well-advertised media show, with news-grabbing positive results each day. This clearly distinguished the coalition negotiations from those of the typical grand coalition formation. While the latter were largely shielded from the public and media news were often fed by internal critics of particular deals that had been tentatively agreed, the leaders of the emerging ÖVP–FPÖ coalition followed a proactive information strategy and managed to present an agreement on a new policy area after each of the official meetings of the negotiation teams. Only the budget was discussed over more than one of these meetings. The duration of these negotiations was also very short, in particular when considering that one of the parties was new to the game and had issued serious disagreement with the policies of the previous governments. The swiftness of ÖVP–FPÖ negotiations was helped by taking the policy programme from the SPÖ–ÖVP bargaining round as the starting basis. Although important changes were made the final deal left enough of the SPÖ–ÖVP deal intact to undermine the SPÖ’s opposition and ridicule this party’s government critique. Confronted with two parties willing to form a government and commanding a solid majority in parliament, President Klestil had to give in and to appoint the ÖVP–FPÖ cabinet. Yet Klestil did not fully capitulate. He influenced a preamble to the coalition agreement drafted by Schüssel that aimed at signalling to the outside world that the government parties were committed to the principles of democracy, the rule of law, and European integration that many international observers saw challenged by the FPÖ government participation. The president also rejected one or two of the proposed FPÖ cabinet members because of earlier verbal radicalism.

⁴ The ‘sanctions’ not only were naïve concerning the prevention of a particular coalition but also illconceived in their potential mid-term consequences (given unanimity in many EU decisions) and obviously it had not been considered that terminating them in a face-saving manner in case of failure would require considerable diplomatic effort. ⁵ The ‘sanctions’ are covered by several analyses, e.g. Busek and Schauer (2003).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

54

 . ̈  

Measured against the Austrian post-war record the result of this coalition formation process was unusual in form and content. It brought to office an untried party combination, indeed a combination that (in the form of predecessor parties) had dominated the inter-war period (Müller et al. 1995) and that long had been treated as kind of historically discredited, even more so given the FPÖ’s status as a kind of pariah party that had been labelled being ‘out of the constitutional spectrum’ by ÖVP parliamentary party group leader Andreas Khol (1995) just a few years earlier. It was also a cabinet that had some unique features concerning the formation process: it was the first one that had been formed without a prior mandate by the president, it excluded the strongest party, it gave the chancellorship to the leader of a party that considered itself only the thirdstrongest one, while this party had been pre-committed to opposition under the conditions created by the election. The formation of the ÖVP–FPÖ coalition has seen conflicting interpretations. One is that of a bargaining failure: in trying to extract more and more concessions from the SPÖ the ÖVP had publicly committed itself to its demands and caused the breakdown of the negotiations. The conflicting interpretation is that of a master plan followed through by ÖVP leader Schüssel. Accordingly, the ÖVP deliberately conducted the negotiations with the SPÖ in a way leading to their breakdown and paving the way to handing over the chancellorship to the ÖVP in a new coalition. In both scenarios, Schüssel’s survival as party leader was at stake and he most likely adhered to Luebbert’s (1986: 46) first principle (‘leaders are motivated above all by a desire to remain party leader’): clearly, returning the chancellorship to his party after 30 years and the promise of huge policy gains were more rewarding than assuring another term as junior partner in the declining grand coalition.

A pivotal party’s hard choice: the 2002/2003 episode The Schüssel I ÖVP–FPÖ coalition was brought down by intra-party revolt against the FPÖ leadership (criticized as being too docile in the coalition with the ÖVP). The early election that followed resulted in a triumphal victory for Chancellor Schüssel’s ÖVP (gaining more than any party before in a single election) and leaving his party in a pivotal position: being able to form three different two-party majority coalitions (with the SPÖ, FPÖ, and Greens) while no viable alternative coalition existed. As formateur, Schüssel first turned to the SPÖ. These negotiations fell over irreconcilable policy differences mostly over welfare state and budget issues. Schüssel next turned to the Greens. Going into talks with the ÖVP led to considerable scepticism among the Green MPs and party executive members. It turned out that the ÖVP idea was to allow for budget largess and policy

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



55

concessions in some of the policy areas most salient to the Greens (which were also likely to be handed over to Green ministers) and in return get the Greens’ support for the largely unaltered ÖVP line with regard to the bulk of policy areas. In the end the Green negotiators recognized that their party was not prepared to accept such a deal and took the exit option. The ÖVP then made another brief attempt in high-level talks with the SPÖ that came to nothing (Khol 2013). Eventually, the ÖVP turned to the humiliated FPÖ and renewed its coalition, however with a greatly diminished share of the FPÖ in terms of portfolio allocation and policy imprint. While Schüssel’s turn to the SPÖ had been mainly tactical (demonstrating to the public and his own party that the grand coalition could not yet be revived) he later declared that ÖVP–Greens would have been his favoured outcome (Schüssel 2009, 2010). Indeed, a coalition with a left-libertarian party would have whitewashed his reputation (of having brought to government office a right-wing populist party) and made the life of the SPÖ opposition much more difficult.

The unhappy return to grand coalition government: the 2006 episode While the 2002/2003 episode had offered the ÖVP embarrassment of the riches, things had changed for the worse for this party after the 2006 elections, which had deprived the ÖVP–BZÖ cabinet of its majority. For one thing this defeat reflected the failure of Haider’s gamble of splitting the FPÖ and establishing the BZÖ, as the rump-FPÖ survived and eventually turned out even slightly stronger than the BZÖ. For another thing the election also deprived the ÖVP of its plurality and narrowly returned the SPÖ as the strongest party. The stunned ÖVP leaders first interpreted this result as an error on part of the voters (with too many nonsocialist citizens taking the ÖVP lead as a given). Accordingly, the ÖVP first attempted to refuse coalition negotiations with SPÖ leader Alfred Gusenbauer, whom President Heinz Fischer had appointed formateur. Gusenbauer, in turn, aimed for getting appointed chancellor of a minority cabinet, aiming at solidifying the SPÖ’s support from a government role and paving the way to new coalition options (a strategy masterful applied by Gusenbauer’s hero Bruno Kreisky at the beginning of the SPÖ’s hegemony in the 1970s (Müller 2011)). Yet, like Klestil in 1999, President Fischer aimed for bringing to office a grand coalition government. In a subtle and skilful intervention he blocked both partisan strategies. Fischer removed the option of a new election under the sitting ÖVP–BZÖ government (by indicating that he might accept the cabinet’s resignation offer to honour the election outcome) and he refrained from appointing an SPÖ minority cabinet. In so doing he was banking on considerable scepticism in both the SPÖ and ÖVP concerning the high-risk strategies of their leaders and the fact that in the newly

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

56

 . ̈  

elected parliament no other majority coalition seemed feasible (given the deep rift between the FPÖ and BZÖ and the impossibility to combine any of the two with the Greens). Moreover, Gusenbauer’s overture received less than enthusiastic support from the potential support parties, in particular the Greens. Once the other options had been eliminated, coalition formation between the SPÖ and ÖVP followed the standard pattern as described earlier.

Portfolio allocation Before ministries can be allocated to the coalition partners, the coalition builder has to decide on three interrelated issues: the number of departments, the ministerial structure, and the number of positions in the ranks of minister and secretary of state. These decisions then need to be transformed into an amendment to the law regulating the set-up of federal ministries (Bundesministeriengesetz) and into the appointment acts of the president. While the number of ministries has seen minor variation (Table 3.3) their design has greatly changed between cabinets (Table 3.6). The frequent changes in the ministerial structure typically follows a political (rather than administrative) logic—to balance the parties’ portfolio payoffs and to strategically bundle jurisdictions for the purpose of addressing the government’s substantive goals (Sieberer et al. 2021). A few conventions and patterns of portfolio allocation got established over the post-war period, which have largely remained unchallenged. First, the position of chancellor goes to the government party with the larger number of parliamentary seats. Second, the second government party nominates the vice-chancellor. Third, the distribution of ministerial posts is roughly proportional to the parliamentary strengths of the coalition parties. To some extent, secretary of state posts and some reallocation of portfolio competences are used to balance the division of ministerial office spoils. If a coalition is maintained after an election, changes in the parliamentary strength of parties are taken into account. Fourth, each party gets those departments that correspond most closely to its constituency (cf. Budge and Keman 1990; Bäck et al. 2011; Ecker et al. 2015). Thus, whenever the respective parties are represented in government, the SPÖ gets social affairs and transportation (including the railways and, for a long time, also the postal services) and the ÖVP gets agriculture and trade/economy. Fifth, there is a tradition to avoid oneparty monopolies in broad branches of government. Thus, because of historical reasons (the civil war in 1934), the first-generation grand coalition divided the responsibility for the armed forces between the SPÖ (interior) and ÖVP (defence). This pattern was maintained in the second-generation grand coalitions, and in the third generation the two ministries were switched between the coalition parties. In both periods, the principle was also applied to the two educational departments (which were still one large ministry in the first-generation grand coalitions). In

8 SPÖ, 7 ÖVP, 1 Ind.

8 SPÖ, 6 ÖVP, 1 Ind.

7 SPÖ, 6 ÖVP, 1 Ind. 6 SPÖ, 6 ÖVP, 1 Ind. 6 ÖVP, 6 FPÖ 9 ÖVP, 3 FPÖ

1990

1994

1996 1997 2000 2003

2007 2008 2013 2016 2017

Gusenbauer Faymann I Faymann II Kern Kurz I

7 SPÖ, 7 ÖVP 7 SPÖ, 7 ÖVP 7 SPÖ, 7 ÖVP 7 SPÖ, 7 ÖVP 6 ÖVP (2 Ind.), 6 FPÖ (1 Ind.)

7 SPÖ, 7 ÖVP, 1 Ind.

1987

Vranitzky II Vranitzky III Vranitzky IV Vranizky V Klima Schüssel I Schüssel II

Number of ministers per party (in descending order)

Year in

Cabinet

14 14 14 14 14

14 13 12 12

15

16

15

Total number of ministers

1 ÖVP

1 ÖVP, 1 FPÖ 2 FPÖ, 1 ÖVP, 1 Ind. 1 SPÖ, 1 ÖVP 1 SPÖ 1 SPÖ

3 ÖVP

1 ÖVP

1 ÖVP

13 13 13 13 12

13 12 12 13

14

15

15

SPÖ SPÖ SPÖ SPÖ ÖVP

SPÖ SPÖ ÖVP ÖVP

SPÖ

SPÖ

SPÖ

ÖVP ÖVP ÖVP ÖVP ÖVP (Ind.)

SPÖ SPÖ FPÖ ÖVP

SPÖ

SPÖ

SPÖ

ÖVP ÖVP ÖVP ÖVP FPÖ (Ind.)

ÖVP ÖVP ÖVP ÖVP

ÖVP

ÖVP

ÖVP

1 Chancellor 2 Finance 3 Foreign Number Number of affairs of watchdog junior ministers ministries per party

Table 3.3 Distribution of cabinet ministerships in Austrian coalitions, 1987–2018

SPÖ SPÖ SPÖ SPÖ FPÖ

SPÖ SPÖ FPÖ FPÖ

SPÖ

SPÖ

SPÖ

4 Social Affairs

ÖVP ÖVP ÖVP ÖVP FPÖ

SPÖ SPÖ ÖVP ÖVP

SPÖ

SPÖ

SPÖ

5 Interior OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

58

 . ̈  

coalitions involving the FPÖ (or BZÖ), these principles were maintained; however, the ÖVP was willing to allow FPÖ control over both the police and the army in the Kurz cabinet. The parties’ bargaining power matters within these parameters. So, the pivotal ÖVP could claim the chancellorship when it had the same number of seats as the FPÖ. Each government formation is also an opportunity to change the departmental structure. While single-party governments used such changes to highlight new policy priorities (such as higher education, environmental protection, and health), coalitions have mostly used them to balance out the allocation of ministerial spoils between the parties. Accordingly, the jurisdictions for youth and family, but also women, consumer protection, and the arts, have been combined in various ways with others or had been made departments in their own right. Portfolio allocation always has been an important issue in coalition negotiations. In terms of the coalition negotiations it typically is the last one to be addressed and it is left to the party leaders. While it is true that the formal cabinet decision rules and the coalition governance mechanisms that are described here substantially constrain ministerial discretion, holding a portfolio is still a major advantage in policy-making. This is particularly true in terms of agenda control: it is practically impossible to bypass a minister with an initiative that falls in his or her domain. Despite all instruments of mutual control a minister is in a privileged position to influence policies for which he or she is formally responsible. Thus the struggle over portfolio allocation is clearly related to the coalition parties’ policymaking capacity. Individual ministerial posts often come with large appointment powers that are typically used to the minister’s party’s advantaged (Müller 2007; Ennser-Jedenastik 2014).

Coalition governance In terms of coalition governance the chapter faces four distinct periods: two with grand coalition cabinets (1987–1999 and 2007–2017) and two with ÖVP–FPÖ/ BZÖ cabinets (2000–2006 and 2017–2019⁶). Table 3.5 suggests a rather stable pattern of coalition governance institutions in the observation period and indeed in the entire post-war period (Müller 2000). Yet a closer look reveals important differences in how coalitions worked in practice. By and large the grand coalition governments reflected that they brought together the two traditional antagonists in the party system. It has been suggested that coalitions emphasize their own profile immediately after and shortly before elections but demonstrate unity in the bulk of the electoral period to govern effectively (Sagarzazu and Klüver 2017). ⁶ The coverage of the final period remains rather preliminary as the number and quality of primary sources that had been available does not nearly match those for any of the other periods.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



59

While such cycle effects are likely to exist also under grand coalition governments, these have maintained a high level of party differentiation throughout the electoral circle. This is not always due to their teams in government. Rather the vested interests within these parties closely monitor government policy-making and are keen not to give away concessions to the coalition partner (at least not until compensation is agreed). During the policy process intra-party actors often cry ‘alarm’ from the sideline and try to block change. Negotiating a deviation from the status quo hence often requires a cumbersome process and typically results in making incremental steps only. These patterns were established in the grand coalitions of the post-war years and seem to be reproducing themselves regardless of the greatly changed environment and the coalition leaders occasionally making resolutions to arrive at a more smooth and efficient conduct of government policy-making. In contrast, coalitions involving the SPÖ (just before this chapter’s observation period) or the ÖVP and the FPÖ were characterized by a stronger emphasis on coalition unity and commitment to a joint programme. The ÖVP–FPÖ coalitions tried hard to distance themselves from stalemate grand coalitions and build an image of effective reformers agreeing on the goals and means of their policies. Combining parties with little policy distance in many areas, these coalitions, in particular the Schüssel cabinets, made indeed considerable changes from the policy status quo, in particular in former SPÖ domains (Tálos 2006, 2019; Kriechbaumer and Schausberger 2013). Yet the FPÖ was unable to carry cabinet membership through as party splits, intra-party revolt, or personal scandal brought down the cabinets or prevented their renewal after elections. Returning to the contractual set-up of coalitions (Tables 3.4 and 3.5), all negotiated a formal, contract-like coalition agreement. All these were postelection and public, though the ÖVP–FPÖ coalition of 2000 confined to publishing the policy programme and kept the governance part private. Table 3.4 takes a closer look at the coalition agreements. All but one were worked out immediately after an election. The exception is the 2017 agreement between the SPÖ and ÖVP, which had been negotiated as an addition to the postelection agreement some time into the life of the Kern cabinet. By then both government parties had replaced the leaders who had negotiated the post-election agreement. Most significantly, the new chancellor, Kern, had presented his vision (‘Plan A’) of the new directions the government policy should take. Design and presentation of ‘Plan A’ could have been the SPÖ’s break-out from the coalition and paved the way to early elections. The SPÖ’s price for refraining from such rupture was the renegotiation of the coalition’s policy programme and incorporating ideas promoted in ‘Plan A’. Negotiating and settling new agenda items, of course, is everyday business in coalitions, as is the filling-in of details to conclude earlier policy agreements painted with a broader brush. Such deals typically result in specific government bills. In some instances, when the issue is complex, is

Year in

1949 1953 1956 1959 1961 1963 1964 1983 1986 1987 1990 1994 1996 1997 2000 2003 2007 2008 2013 2016 2017

Coalition

Figl III Raab I Raab II Raab III Gorbach I Gorbach II Klaus I Sinowatz Vranitzky I Vranitzky II Vranitzky III Vranitzky IV Vranizky V Klima Schüssel I Schüssel II Gusenbauer Faymann I Faymann II Kern Kurz I

1,600 700 1,700 1,200 1,200 4,400 4,400 2,900 2,900 13,900 23,300 11,700 11,200 11,200 29,313 11,729 29,394 50,871 28,644 8,654 58,834

Size

11 28 41 50 50 21 21 7 7 2 1 6 8 8 0 0 0.7 0.7 0.9 0 0.02

General rules (in %) 44 33 30 13 13 27 27 0 0 1 0 4 3 3 0 0 0 0 0 0 0

Policy-specific procedural rules (in %)

Table 3.4 Size and content of coalition agreements in Austria, 1945–2018

16 30 4 26 26 19 19 3 3 1 1 1 1 1 0 0 0 0 0 0 0

Distribution of offices (in %) 25 4 24 6 6 24 24 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of competences (in %) 5 4 0 6 6 9 9 90 90 96 98 88 87 87 96.5 99.4 98.0 97.9 97.2 96.0 97.3

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Figl I Figl II Figl III Raab I Raab II Raab III Gorbach I Gorbach II Klaus I Sinowatz Vranitzky I Vranitzky II Vranitzky III Vranitzky IV Vranizky V Klima Schüssel I Schüssel II

Coalition

1945 1947 1949 1953 1956 1959 1961 1963 1964 1983 1986 1987 1990 1994 1996 1997 2000 2003

N N POST POST POST POST POST POST POST POST POST POST POST POST POST POST POST POST

N/A N/A N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y

N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y

CoC, CaC CoC, CaC CoC, CaC Coc, CaC CoC, CaC CoC, CaC CoC, CaC CoC, CaC, IC CoC, CaC, IC CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC

All used

CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC

CoC CoC CoC CoC CoC CoC CoC IC IC IC IC IC IC IC IC IC IC IC

Most For common most serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Table 3.5 Coalition governance mechanisms in Austria, 1945–2018

Y Y Y Y Y Y Y Y Y Y N N Y Y Y Y Y Y

Personal Issues union excluded from agenda

All/Most All/Most All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All Most/All Most/All All/All All/All

Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y

Few Few Few Few Few Few Few Few Few Varied Varied Comp. Comp. Comp. Comp. Comp. Comp. Comp.

Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y

Continued

N N Y Y Y Y Y Y Y Y Y N N N N N N N

Junior NonCoalition Freedom of Policy discipline appointment agreement ministers cabinet positions in legislation/ other parl. behaviour

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2007 2008 2013 2016 2017

POST POST POST POST POST

Y Y Y Y Y

N Y Y Y N

CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC CoC, IC, CaC

All used

CaC CaC CaC CaC CaC

IC IC IC IC IC

Most For common most serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Y Y Y Y Y

Y N N N N

Personal Issues union excluded from agenda

All/All All/All All/All All/All All/All

Y Y Y Y Y

Comp. Comp. Comp. Comp. Comp.

Y Y Y Y Y

N Y N N N

Junior NonCoalition Freedom of Policy discipline appointment agreement ministers cabinet positions in legislation/ other parl. behaviour

Notes: During periods where the values for the variables remain identical, the first and last applicable cabinets are listed. The last applicable cabinet is right-justified in the Coalition column. Coalition agreement: POST = Post-election; N = No coalition agreement Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee; CoC = Coalition committee Coalition discipline: All = Discipline always expected; Most = Discipline expected except on explicitly exempted matters; Spec. = Discipline only expected on a few explicitly specified matters Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement; No = No explicit agreement

Gusenbauer Faymann I Faymann II Kern Kurz I

Coalition

Table 3.5 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



63

controversial, and requires several measures to be taken sequentially, an intermediate step in the form of a side letter or special agreement may be required. Most prominently, when the privatization of Austria’s largest bank, the Creditanstalt, brought the grand coalition (Vranitzky V) close to divorce, the cabinet parties eventually hammered out a comprehensive deal and formalized it in a contractual inter-party agreement (Vranitzky 2004: 367–72; Schüssel 2009: 39–51; Kriechbaumer 2014: 48–69). Such deals are specific, mostly behind the scenes, and in many instances hard to access (and therefore not included in the table). The 2017 deal is different in constituting an official renegotiation of the coalition contract before a clear ‘new deal vs. termination’ alternative and resulting in a public document, signed with all the ritual that accompanies postelection coalition negotiations. It thus most clearly resembles the ‘coalition renegotiating’ situation in the Lupia and Strøm (1995, 2008) model. With a mean length of ca. 27,700 words all coalition agreements since 1987 have been comprehensive compared to the earlier documents (mean length ca. 2,000 words). Since 1987 the documents tended to become longer, but the trend was not linear. What was new was the attempt to lay out the policy agenda of the incoming cabinet comprehensively and with some precision, listing a large number of legislative projects, fixing the direction of policy change, and occasionally providing details about the envisaged solutions. In contrast, first-generation grand coalition agreements almost exclusively were built on procedural rules and were confined to provide policy details only for those few issues that had brought down or immobilized the previous cabinet. The single SPÖ–FPÖ coalition agreement was somewhat more comprehensive on policy but very vague about most such items. Thus both of these early coalition types had largely rested on procedural rules of coalition governance. Yet such rules are neither absent nor less detailed in the post-1987 grand coalition contracts but simply take very small shares of the longish documents (indeed so small that they figure as 0 in Table 3.4 although they are present). These documents thus combine two modes of coalition governance that theoretically could be seen as alternatives (Müller and Strøm 2000; Bowler et al. 2016). The first ÖVP–FPÖ coalition stands out by having no procedural rules in the official coalition agreement. Nor did it mention the allocation of cabinet positions. Coalition architects wanted to present the new government as a ‘new style’ cooperation, based on mutual trust and common goals rather than mutual checks, strict rules, and the division of spoils. However, the truth is that the negotiations had resulted in a detailed contract about coalition governance rules modelled after but even stricter than the SPÖ–ÖVP agreements. It also included the full list of all ministerial and secretary of state positions. This document was signed by the party leaders—in case of the FPÖ the outgoing (Haider) and the incoming (RiessPasser) leaders—plus the leaders and the administrative heads of the parliamentary party groups. This deal was negotiated in 2000 but not formally renewed in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

64

 . ̈  

2002 when the second ÖVP–FPÖ government was formed after a great shift in the cabinet parties’ strengths. Compared to the SPÖ–ÖVP documents, the Schüssel I rules were more specific on inter-party coordination, and they contained commitments to regularly take stock on the implementation of the government’s policy programme and to closely monitor budgetary developments, and it established a tighter grip on the coalition parties’ parliamentary behaviour. Presumably, the ÖVP—having conceded the finance minister post to the FPÖ— had been behind the budget provisions, while the FPÖ is reported to have been behind the tightening of parliamentary coalition discipline to avoid ÖVP collusion with the opposition. The ÖVP–FPÖ 2017 agreement included the allocation of ministerial posts and procedural rules but regarding the latter it was less specific than its 2000 model. All but two grand coalition contracts of the entire post-war period included the election rule, committing the partners to holding elections in case of coalition breakdown (Table 3.5). The termination of the exceptional Gusenbauer and Kurz cabinets, nevertheless, led straight into elections. Thus coalition practice so far suggests that coalition termination inevitably leads to new elections. Personal union is a general feature of coalition governance in Austria. Typically, the party leaders take the highest executive position allocated to their parties (i.e. chancellor or vice-chancellor, the latter combined with a ministry). A change in party leadership typically leads to a simultaneous change in cabinet office. A partial exception is Franz Vranitzky, who was installed chancellor by his predecessor Fred Sinowatz. Vranitzky’s lack of a typical party career would have made the simultaneous taking over of the party leadership a challenge and Sinowatz’s loyal support made it not necessary neither. Things are different in case of the Schüssel cabinets. Personal union resulted from Haider’s resignation from the FPÖ leadership after the coalition deal had been forged. Given Haider’s damaged reputation it had been clear to him and his ÖVP partners at the outset of the coalition negotiations that he could not take government office himself. Resigning from party leadership (in favour of his long-term confidant and deputy Susanne Riess-Passer, who also took the vice-chancellorship) was an additional move intended to dampen the turmoil the government inclusion of the FPÖ had created. Yet Haider’s new role was conceived not that of a ‘simple party member’, as he had downplayed it, but of the party’s spiritual leader and would-be chancellor candidate in the next election. In reality, Haider soon turned out a major problem for the government as he was stirring up FPÖ dissent with coalition policy (when he considered it unpopular) and causing headlines that shed a bad light on the coalition. Haider was also key to the FPÖ revolt that brought down Schüssel I. He returned to party leadership in the BZÖ when it split from the FPÖ in 2005 (during Schüssel II) but remained out of government; hence there was no personal unity in the final phase of this cabinet.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



65

While all coalition agreements spared out issues on which the parties disagreed when there was no need to touching them, the 2008 SPÖ–ÖVP agreement explicitly excluded the calling of referenda (by majority vote in parliament) against the will of any of the partners. While phrased as a general procedural clause, it related back to the SPÖ’s pledge not to accept any change of the EU treaty without prior acceptance by referendum. The SPÖ’s making this pledge had been the reason for the ÖVP to terminate the Gusenbauer cabinet and seek early elections. The price the SPÖ had to pay for coalition renewal hence was to formally grant the ÖVP veto right on the use of referenda, an issue on which earlier coalition agreements had remained silent. Coalition discipline in parliamentary voting has been a general characteristic of all coalition types and cabinets (Table 3.5). Yet there are differences in detail. Until 2008 the SPÖ–ÖVP agreements were explicit only on legislation (and, after EU accession, also EU affairs) but the common understanding was that coalition discipline would relate to other important parliamentary decisions as well. Consequently, occasional deviations from that informal rule caused severe conflict in the coalition.⁷ Yet maintaining iron coalition discipline caused intra-party tensions in both cabinet parties, and the ÖVP in particular was pushy to find some valve for releasing intra-party pressure while the SPÖ, fearing to be outvoted by ÖVP–FPÖ alliances, was more reluctant. This led to including a ‘coalition-free area’ in the 1994 and 1996 agreements, which, however, was very narrowly construed. Coalition discipline was lifted only for very few issues, one of which (the parliamentary rules of procedure) anyway requires a two-thirds majority (which was unlikely or impossible outside the coalition) while for the others (fixing alcohol limits for motorists and LGBT affairs in the criminal code) relaxing coalition discipline was irrelevant as there was no political majority ready to change the status quo (provided party discipline was maintained).⁸ These coalition agreements listed further policy areas—education, traffic rules, industrial code, consumer protection, vocational training, waste-management—that were opened for majority formation outside the coalition but maintained a veto for the coalition partners. Eventually, none of these issues was decided outside the coalition framework. The 2008 and 2013 SPÖ–ÖVP coalition agreements formally extended coalition discipline to all parliamentary behaviour and this is also the rule adopted by the various ÖVP–FPÖ/BZÖ coalitions.

⁷ The ÖVP joining forces with the FPÖ to elect an ÖVP party employee president of the audit court in 1992, and both government parties outvoting their partners to establish parliamentary committees of investigation to look into scandals in their partner’s domain in 1989 (Fallend 2000). ⁸ The two issues were controversial between the SPÖ, Greens and Liberals (aiming for change of the status quo) on one side and the ÖVP and FPÖ (insisting on the status quo) on the other. Only extraordinary circumstances—a tragic car accident and a subsequent media campaign—eventually allowed for the relaxing of party discipline and the lowering of the alcohol limit to 0.05 per cent (Wolf 1998).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

66

 . ̈  

According to the constitution, the cabinet decides by unanimous consent, that is each cabinet member has veto right, thus conforming to Tsebelis’s model of coalition governance (Tsebelis 2002). Constitutionally, the role of the chancellor is confined to that of a coordinator (Adamovich 1973;Wineroither 2009). In addition, there are liberal rules concerning the cabinet agenda, with each minister being allowed to table proposals and no central agenda control. It is the office of party leader and, institutionally, the chancellor’s appointment and dismissal rights, which give the chancellor and vice-chancellor authority over their own teams in government.⁹ Such authority may vanish, however, when ministers have powerful intra-party groups behind them and the party leader is considered weakened and his days appear numbered. One important procedure to ensure mutual information of coalition partners is the official consultation process through which ministerial draft legislation is required to go. It provides the other ministries, other institutions, and a range of interest groups with the ministerial bills and invite them to comment (Fischer 1972). In the course of the post-war period this procedure has been central to provide structured input and generate societal consent for government legislation. Yet ministers remain free to decide how to react to the inputs. At a minimum, the consultation process puts ministerial policy ambitions and instruments to broad scrutiny and alerts the coalition partners about potential problems. Yet given that this process is public, coalition governments generally aim for the ministerial drafts entering the consultation procedure when they are already politically agreed between the partners. For that purpose, individual ministers are expected to politically coordinate their draft legislation with a representative of the coalition partner. Initially, these representatives held various offices; most were their party’s parliamentary spokesperson for the respective policy domain or cabinet members. In the Vranitzky and Klima cabinets gradually a system was established that Fernandes et al. (2016) have labelled one of ‘wary partners’. Accordingly, pairs of ministers, ideally from ‘neighbouring’ departments (such as economy/social affairs and labour), were formed and the ministers charged to politically coordinate their policies (i.e. get the coalition partner’s consent). The same system was also applied in the post-2006 grand coalition cabinets. Ministers wanting to be on the safe side, however, continued to talk to their partner’s parliamentary spokesperson in order to ensure their legislation a smooth passage through parliament. Yet coordination between the coalition partners is not always fully successful, leaving unresolved issues for the cabinet stage. There are also ministers who want to send a strong signal to their party or the general public and for that purpose enter their un-coordinated draft legislation into the formal ⁹ Chancellors have invariably followed the proposals of the coalition partner’s leader with regard to appointments to and dismissals from positions allocated to the other party, even if they personally disagreed (save a few vetoes in the appointment process mentioned earlier).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



67

consultation procedure and submit ministerial bills to cabinet deliberation without prior consent with the coalition partner. Such behaviour was more a problem in the grand coalitions than the ÖVP–FPÖ cabinets. All governments had a similar mechanism to filter out such cabinet proposals and resolve the coordination problem before the weekly cabinet meeting (Table 3.5). Key was a small informal coordination body (CoC) bringing together cabinet representatives—typically ministers charged with the coordination task and the heads of the personal cabinets of chancellor and vice-chancellor—and the leaders and directors (i.e. leading staff members) of the parliamentary party groups. In a first step this group screened the proposals, checked whether they were agreed between the partners, and identified problems that would prevent the cabinet’s unanimous consent. While some issues were resolved by the coordination group in exchange with the ministers in charge, others were delegated upwards to the weekly meeting of chancellor and vice-chancellor (the inner cabinet (IC) in Table 3.5), and still others were handled by agreeing to defer cabinet decision-making until a later date. The better a coalition machinery works, the less strain there is on the coordination system. A smoothly running coalition means a large share of fully agreed proposals being tabled for cabinet consideration, few proposals delegated upwards, and few items being deferred until later—with, at the same time, sufficient productivity in terms of the government’s legislative agenda. Smooth running also implies that conflictual issues are handled coalition-internal. The institutional design of the grand coalition cabinets at the beginning of the observation period and the ÖVP–FPÖ coalition taking office in 2000 included a second, higher-level type of coalition committee, involving some politicians outside the cabinet system. Such type of coalition committee had been the key institution making all kinds of decisions to be implemented by the government in the first period of grand coalition in the post-war years. However, the more recent SPÖ– ÖVP coalitions never activated this body from the coalition designers’ drawing board and eventually refrained from including it in more recent coalition designs. In contrast, such a body came to real life in the ÖVP–FPÖ 2000 coalition. Its main purpose was to directly involve the FPÖ’s strong man Haider (who held the post of Governor in Carinthia) in the coalition machinery. Substantively its tasks were to agree on challenging issues, resolve conflicts, and, in particular, strategize and fix the mid-term planning of government work. Yet Haider’s personal strategy of distancing himself from (unpopular) government decisions and publicly whooping the FPÖ team conflicted with him being involved in that committee. He therefore faded out of it, greatly diminishing the committee’s usefulness and hence its being convened. The high-level coalition committee thus largely was a ‘phantom’, as one of the participants had put it when interviewed. Of what is known, the short-lived Kurz cabinet employed an elaborated structure of a small coordination committee (CaC) meeting on a weekly basis, a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

68

 . ̈  

larger committee (CaC/CoC) for strategic planning including the party leaders and meeting less frequently, and a committee charged with monitoring the implementation of the coalition agreement (CoC). There was considerable personnel overlap between these bodies. All coalitions agreed on and practised freedom of appointment. Yet in a few instances (including SPÖ–ÖVP and ÖVP–FPÖ coalitions) one party was particularly unhappy about a nominated politician (as having a record as a stirrer against the coalition partner or having been involved in particular detrimental action against the coalition partner) and vetoed the other party’s choice. These, however, were rare exceptions from a rule rather than a general veto right or a kind of joint appointment of cabinet members. All cabinets also included a number of junior ministers, not nearly matching that of ministers. About half of them were from the same party as the minister and served in the party leader’s ministries with the goal to disburden the department head. Almost all remaining junior ministers served in departments held by the coalition partner. The goal was to allow for mutual control and to facilitate coordination. Accordingly, they were primarily appointed in key departments such as finance, economy, and transport and public economy. While non-cabinet positions had been included in most coalition agreements in the first two post-war decades this is no longer the case (Table 3.5). Nevertheless, government parties have found ways to ensure that appointments to key public sector positions—in the central bank, the nationalized industries, the railway company—followed a coalition (rather than a ministerial government) logic. Some deals, for instance regarding the nomination rights for Austria’s EU Commissioner, had been fixed in side letters to the coalition agreement. While party conflict typically is the most important coordination problem in coalition governments, it is not the only one. Another potential conflict is departmental—with ministries having different objectives and therefore sometimes conflicting opinions concerning government bills. Needless to say, that party and departmental conflict can reinforce each other. The traditional set-up of Austrian cabinet ministries includes several such in-built tensions between ministries such as those between environmental and agriculture and labour and economy, with the former typically taking more restrictive positions than the latter on environmental or labour protection. The ÖVP–FPÖ coalitions deliberately removed some such tensions by an unconventional portfolio design, uniting potentially conflicting jurisdictions in one department and hence internalizing such conflicts in one ministry and avoiding their party political reinforcement. Parliamentary institutions, in particular the committee system and written questions to ministers, provide additional opportunities for mutual control in coalition governments (Strøm et al. 2010; Martin and Vanberg 2011). Grand coalitions used these instruments more frequently for that purpose than ÖVP– FPÖ coalitions, which were more concerned about maintaining a united

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



69

government front (Müller and Fallend 2004; Müller and Jenny 2004). Parliamentary checks on what had passed through earlier stages of coalition coordination sometimes had been anticipated and served to making last refinements of compromises. Yet they also killed some important reform projects agreed at the cabinet level, such as pension reform under the Klima cabinet, and thus contributed to policy immobility and grand coalition decay.

Cabinet termination Twelve cabinets served and terminated during the observation period (Table 3.6). Five terminated at the regular end of the parliamentary term (which was extended from four to five years by a constitutional amendment in 2007). The last Vranitzky cabinet ended by the voluntary resignation of the chancellor, shortly after having led the SPÖ successfully in the 1995 election and after the celebration of his 10th anniversary in the office. In contrast, the other early terminations were conflictdriven. The three grand coalition cabinets Vranitzy IV, Gusenbauer, and Kern were terminated by the junior partner ÖVP in attempts to replace the SPÖ as the plurality party and enter a new formation situation with a flexible approach (i.e. being open to other coalition options than grand coalition government). In all instances the polls signalled that the ÖVP was under competitive pressure from the FPÖ, a threat that would disappear if the FPÖ were to become the ÖVP’s junior partner in government. As described earlier, only in the last instance the ÖVP’s strategizing was successful when the ÖVP under its new leader Sebastian Kurz won the election and brought to an end the third period of grand coalition government by forming a centre-right cabinet with the FPÖ. The triggering event for bringing down Vranitzky IV in 1995 was the disagreement over budgetary policy, with the ÖVP aiming for cuts in welfare spending and the SPÖ more willing to tolerate budgetary deficit. The disagreement over next year’s budget, however, was only the tip of the iceberg with regard to inter-party differences between the SPÖ and ÖVP in welfare and economic policies. These remained on the agenda even after the post-election compromise that paved the way for the last of the second-generation grand coalitions taking office in 1996 and constituted the policy basis for the non-renewal of grand coalition after the 1999 elections. In 2008, Chancellor Gusenbauer had already been toppled by his party as party leader when the ÖVP terminated the coalition and enforced early elections. The event triggering coalition termination was a conflict over Austria’s strategy in case of further EU extensions and treaty reforms. Together with the new SPÖ leader Werner Faymann, Gusenbauer had committed—in a public letter to the editor of the country’s largest (and EU-critical) tabloid, the Neue Kronenzeitung—to

Date in

1987-01-21 1990-12-17 1994-11-29 1996-03-12 1997-01-15 2000-02-04

2003-02-28

2007-01-11

2008-12-02 2013-12-16

Cabinet

Vranitzky II Vranitzky III Vranitzky IV Vranizky V Klima Schüssel I

Schüssel II

Gusenbauer

Faymann I Faymann II

2013-09-29 2016-05-09

2008-09-28

2006-10-01

1990-10-07 1994-10-09 1995-12-17 1997-01-15 1999-10-03 2002-11-24

Date out

100 49.2

46.1

96

96.6 100 27.2 23.8 100 76.6

Relative duration (%)

Table 3.6 Cabinet termination in Austria, 1987–2018

1 7b

4, 7b

4

1 1 4, 7a 2 1 4, 7b, 8

Mechanisms of cabinet termination

Terminal events

ÖVP, SPÖ

7b: ÖVP, FPÖ8: FPÖ

SPÖ, ÖVP

Budget

Parties (when Policy conflict between area(s) or within)

Faymann resigned under intra-party fire and attempts to replace him as party leader and chancellor.

Schüssel terminated coalition leading to dissolution of the legislature and early elections, stating his party could no longer effectively work with FPÖ due to internal strife within FPÖ. Early elections were held after premature dissolution of the legislature. The ruling coalition collapsed and early elections were held after ÖVP claimed they could no longer work with SPÖ.

Voluntary retirement of the chancellor.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2017-12-18

Kurz I

2019-05-27

2017-10-15 29.3

57.7 7b

4, 7a ÖVP, FPÖ

SPÖ, ÖVP

New ÖVP leader Kurz enforced early elections to bring home his lead in the polls. Conflict over removal of FPÖ minister of the interior in the context of resignation of FPÖ leader Vice-Chancellor Strache over personal scandal.

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event Note: Relative duration: share of constitutionally allowed duration for this cabinet.

2016-05-17

Kern

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

72

 . ̈  

holding referendums in such instances from then on. This was clearly fishing for the newspaper’s support and constituted a break with the grand coalition’s proEU tradition. More important, however, was the widely held opinion within the ÖVP that the 2006 election result, giving the SPÖ a small lead over the ÖVP and breaking the government’s parliamentary majority, had been an ‘error’ on the side of the voters (with many bourgeois voters taking ÖVP victory for granted and not participating in the election). The SPÖ’s move on the EU issue thus provided the ÖVP with an opportunity to correct that ‘error’. The SPÖ being in a transition period from one leader to another (who would have to fight the election without a bonus from the office of chancellor) was a favouring condition. The termination of the Kern cabinet in 2017 by the ÖVP’s newly elected leader Kurz was just the last stroke to terminate not only a cabinet but also the traditional grand coalition formula. To begin with, the government formed after the 2013 election was a coalition of the losers ( 2.4 per cent, 2.0 per cent). It was a ‘grand’ coalition only by historic precedent and formal definition (the two largest parties) but not by substance (holding 101 of 183 seats). As all third-generation grand coalition cabinets it lacked a mission—such as economic reconstruction and getting the occupational forces out of the country (grand coalition mark I) or achieving EU membership and public sector reform (grand coalition mark II)— that would have justified the cooperation in government of the two main contenders in the party system. As a consequence, the coalition was characterized by infighting, minimal compromises, and mediocre economic performance. Government attrition was obvious. Soon after government formation the polls showed the FPÖ ahead of the coalition parties, and ahead of the ÖVP by a large margin. The ÖVP reacted by forcing its leader Michael Spindelegger into resignation and enjoyed a short honeymoon with the voters under its new leader Reinhold Mitterlehner. It was the European refugee crisis that put this to an end. It exposed coalition stalemate (Ultsch et al. 2017), dramatically increased the salience of the FPÖ’s core issue of migration, and returned the FPÖ to leading in the polls (Plasser and Sommer 2018). In that context, the presidential election of 2016 resulted in a historical debacle for the candidates of the SPÖ and ÖVP (which had nominated all successful presidential candidates in the post-war period), being eliminated in the first round, with wide margins behind three other candidates. Both government parties reacted to their decline. The SPÖ forced its leader Faymann into resignation, replacing him by Christian Kern as party leader and chancellor. Under Kern the SPÖ recovered. His strategy included a timid openingup towards the FPÖ, a move, however, that met intra-party resistance and fell short of producing much effect for the SPÖ. Lacking strategic alternatives and faced with the ÖVP strategy to sit out the SPÖ momentum, Kern let pass the chance to call early elections and attempted at coalition revitalization in tandem with ÖVP leader and Vice-Chancellor Mitterlehner. However, Mitterlehner was soon forced into resignation by his own party. The ÖVP elected Sebastian Kurz

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



73

(‘our electoral ace’) to party leader, who immediately terminated the coalition. Although the election was somewhat retarded by the SPÖ, it returned the ÖVP as the strongest party, allowing Kurz to form a government with the FPÖ and thereby bringing to an end the third period of grand coalition government. The first ÖVP–FPÖ cabinet was also terminated by the ÖVP. Chancellor Schüssel reacted to intra-party revolt in the FPÖ over government policy (Heinisch 2003; Fallend 2004; Luther 2006, 2010). The revolt had been engineered by the party’s spiritual leader Haider and largely mobilized the party’s right. Haider had negotiated the coalition deal with the ÖVP but then had left it to his long-term ally Riess-Passer to represent the party as its leader and vice-chancellor. For one thing, given Haider’s controversial image, this had been a precondition for the coalition. However, it also had left Haider free to exercise pressure on the government and to position himself for the next election, a process that got out of control in the revolt. It led to the resignation of Riess-Passer from both of her positions. Two FPÖ ministers and FPÖ parliamentary party group leader Peter Westenthaler followed suite, causing a deep rift in the FPÖ electorate. Chancellor Schüssel decided to ride out the storm by calling elections and appealing to the FPÖ voters (also wooing them by recruiting the most-popular FPÖ minister for his team). This turned out to be highly successful. While the 2002 termination had been gradually ‘built up’ by growing tensions within the FPÖ, the 2019 one of the ÖVP–FPÖ coalition came out of the blue. Despite some eruptions the coalition seemed to be stable and working smoothly so that observers and coalition actors predicted a 10-year lifetime shortly before it came to a sudden end over the ‘Ibiza scandal’—the making public of a highly compromising secret video recording featuring FPÖ leaders Strache and Johann Gudenus.¹⁰ The FPÖ tried to manage the scandal by the two resigning from their offices. Yet Chancellor Kurz also demanded the resignation of FPÖ Minister of the Interior Herbert Kickl. As Kickl did not resign, Kurz removed him with the president’s consent. As a consequence all remaining FPÖ ministers resigned, which then triggered the call for new elections.¹¹

¹⁰ Strache and the FPÖ parliamentary party group leader Gudenus had been set up in Ibiza already in 2017 to be recorded furtively in a conversation with a person perceived as being related to a Russian oligarch. In this conversation Strache expressed his disdain for journalists and desire to undermine the freedom of the press, indicated his willingness to hand out government contracts according to a party political rationale and in exchange for support to the FPÖ, explained how party donations could be hidden before the audit court, and slandered political competitors. ¹¹ Although narrowly falling out of the book’s observation period, the episode cannot be fully understood without mentioning how it continued. Kurz replaced the FPÖ ministers with independents and aimed for remaining in office until the early elections (cabinet Kurz II). Yet three of the opposition parties—the SPÖ, FPÖ, and Pilz—were not willing to tolerate an ÖVP single-party government (with little camouflage by independents) and thereby grant Kurz incumbency advantage. The Kurz II cabinet was therefore removed with their votes in the post-war period’s first successful vote of no confidence, giving way to Austria’s first presidential cabinet.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

74

 . ̈  

Intra-party conflict was most obvious in the termination of the Schüssel I and Faymann II cabinets, and it was also present as a kind of ‘favourable condition’ in the calling of the early elections that ended the short-lived Gusenbauer cabinet. The absence of conflictual termination, however, is no guarantee that the existing coalition formula will be renewed after the election, as can be seen from what happened after the regular election at the end of the Klima cabinet, the last second-generation grand coalition government, which was replaced by the controversial ÖVP–FPÖ coalition.

Conclusion Post-war Austria has been governed mostly by grand coalition governments of the ÖVP and SPÖ and this also applies to this chapter’s observation period. While grand coalition in the post-war period was built to reconcile the civil war opponents, to rebuild the country after the Second World War, and regain sovereignty, the return to grand coalition from 1987 was motivated mainly by the elites’ general consensus on accessing the European Union and reforming the public sector. The achievement of the post-war goals had removed the rationale for grand coalition government and allowed the main parties to prioritize their office and partisan policy goals by forming single-party cabinets (in the 1960s and 1970s) and venturing into new coalitions with the FPÖ (beginning with the 1980s). When a grand coalition was reinstalled in 1996/97 EU accession and public sector reform stood out as national goals requiring broad consensus-building. In contrast to earlier grand coalitions, the one taking office in 2006 lacked a positive agenda. Furthermore, given its greatly reduced size, it was a ‘grand’ coalition mainly by historic precedent. It was formed, with the help of the president, because of the apparent lack of alternative majority cabinets. After a bumpy start the most recent grand coalition cabinets found their own rationale in managing the global financial crisis unfolding from 2008 with the help of the social partners. While this endeavour was quite successful it could not stop the further erosion of the grand coalition. The European migration crisis of 2015 revealed government stalemate and led to political turmoil and intra-party processes in the ÖVP and SPÖ, which eventually put an end to this type of government. Yet, similar to earlier episodes with coalitions involving the FPÖ, the resulting ÖVP–FPÖ coalition fell over FPÖ-induced problems, indicating that a stable alternative to SPÖ–ÖVP coalitions has not yet emerged. The chapter has highlighted that grand coalitions and ÖVP–FPÖ coalitions differed by the latter quite radically dismantling policy legacies and structures in former SPÖ domains (Tálos 2006, 2019). Moreover, the ÖVP–FPÖ coalitions seemed more committed to coalition unity, at least between the government teams and until the coalitions collapsed, while the grand coalitions displayed a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



75

considerable amount of conflict or party differentiation in policy-making. Yet such differences should not be overstated. For one thing, grand coalitions have their own record of substantive reforms, even when they came late and were dictated by empty coffers (as the restructuring of the public sector under post1987 grand coalitions). What is different, of course, is the ideological orientation of policy reforms. In contrast, despite attempts of the ÖVP–FPÖ cabinets to distance themselves from grand coalition policy-making, the communalities in coalition governance overwhelm. All coalitions were built on the understanding that the party leaders negotiating the cooperation need to be able to present to their party something they can sell as a ‘fair deal’ (Ecker and Meyer 2019). This is most obvious with portfolios— sometimes resulting in ‘creative’ portfolio design—but also extends to coalition policy. The coalition compromise (Martin and Vanberg 2014) as fixed in the coalition agreement comprises all policy areas. The parties’ bargaining power weighs in, but it should not be too obvious to observers. The other thing that weighs in in actual policy-making is portfolio allocation. The minister in charge clearly has a major advantage to shape government policy. However, Austria’s strict rule-of-law regime forces ministers to specify their policies in great detail in legislation that needs the coalition partner’s approval first in cabinet and later in parliament. All coalitions had established systems of coalition governance to identify partisan deviations from the coalition compromise and to enforce the coalition deal. This system is less effective when ministers want to maintain the status quo than when they want to change it. In sum, coalition governance in Austria thus places itself somewhere between the ‘coalition compromise’ model (where all government policy rests on substantive agreement between the government parties) and a model recently suggested by Dragu and Laver (2019), aptly named ‘constrained ministerial government’ rather than the original model of fully blown ‘ministerial government’ (Laver and Shepsle 1990, 1996).

References Adamovich, Ludwig (1973). ‘Die Koordinationskompetenz des Bundeskanzlers in verfassungsrechtlicher Sicht’. Juristische Blätter, 95: 234–42. Bäck, Hanna, Marc Debus, and Patrick Dumont (2011). ‘Who Gets What in Coalition Governments? Predictors of Portfolio Allocation in Parliamentary Democracies’. European Journal of Political Research, 50(4): 441–78. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010.’ Party Politics, 21(1): 143–52. Bergman, Torbjörn (1993). ‘Formation Rules and Minority Governments’. European Journal of Political Research, 23(1): 55–66.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

76

 . ̈  

Bowler, Shaun, Thomas Bräuninger, Marc Debus, and Indridi H. Indridason (2016). ‘Let’s Just Agree to Disagree: Dispute Resolution Mechanisms in Coalition Agreements’. Journal of Politics, 78(4): 1264–78. Budge, Ian, and Hans Keman (1990). Parties and Democracy: Coalition Formation and Government Functioning in Twenty States. Oxford: Oxford University Press. Burkert-Dottolo, Günther R., and Bernhard Moser (eds) (2000). Die Regierungsbildung in Österreich 1999/2000. Vienna: Politische Akademie der ÖVP. Busek, Erhard (2014). Lebensbilder. Vienna: Kremayr & Scheriau. Busek, Erhard, and Martin Schauer (eds) (2003). Eine europäische Erregung. Die ‘Sanktionen’ der Vierzehn gegen Österreich im Jahr 2000. Vienna: Böhlau. Dickinger, Christian (1999). Der Bundespräsident im politischen System Österreichs. Innsbruck: Studienverlag. Dragu, Tiberi, and Michael Laver (2019). ‘Coalition Governance with Incomplete Information’. Journal of Politics, 81(3): 923–36. Ecker, Alejandro, and Thomas M. Meyer (2019). ‘Fairness and Qualitative Portfolio Allocation in Multiparty Governments’. Public Choice, 181(3–4): 309–30. Ecker, Alejandro, Thomas M. Meyer, and Wolfgang C. Müller (2015). ‘The Distribution of Individual Cabinet Positions in Coalition Governments: A Sequential Approach’. European Journal of Political Research, 54(4): 802–18. Ennser-Jedenastik, Laurenz (2014). ‘The Politics of Patronage and Coalition. How Parties Allocate Managerial Positions in State-Owned Enterprises’. Political Studies, 62(2): 398–417. Fallend, Franz (2000). ‘Demokratische Kontrolle oder Inquisition?’. Österreichische Zeitschrift für Politikwissenschaft, 29(2): 177–200. Fallend, Franz (2004). ‘Are Right-Wing Populism and Government Participation Incompatible? The Case of the Freedom Party of Austria’. Representation, 40(2): 115–30. Fernandes, Jorge M., Florian Meinfelder, and Catherine Moury (2016). ‘Wary Partners: Strategic Portfolio Allocation and Coalition Governance in Parliamentary Democracies’. Comparative Political Studies, 49(9): 1270–300. Fischer, Heinz (1972). ‘Zur Praxis des Begutachtungsverfahrens im Prozeß der Bundesgesetzgebung’. Österreichische Zeitschrift für Politikwissenschaft, 1(1): 35–54. Fischer, Heinz (2003). Wende-Zeiten. Ein österreichischer Zwischenbefund. Vienna: Kremayr & Scheriau. Gamson, William A. (1961). ‘A Theory of Coalition Formation’. American Sociological Review, 26(3): 373–82. Heinisch, Reinhard C. (2003). ‘Success in Opposition – Failure in Government: Exploring the Performance of the Austrian Freedom Party and other European Right-wing Populist Parties in Public Office’. West European Politics, 26(3): 91–130. Khol, Andreas (1995). ‘Die FPÖ im Spannungsfeld von Ausgrenzung, Selbstausgrenzung, Verfassungsbogen und Regierungsfähigkeit’. Österreichisches Jahrbuch für Politik, 1995: 193–221.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



77

Khol, Andreas (2001). Die Wende ist geglückt. Der schwarz-blaue Marsch durch die Wüste Gobi. Vienna: Molden. Khol, Andreas (2009). ‘Veritas filia temporis – die Wahrheit ist eine Tochter der Zeit. Die Nationalratswahlen am 3. Oktober 1999 im Rückblick’. Österreichisches Jahrbuch für Politik, 2009: 379–98. Kohl, Andreas (2013). ‘Verspielte Chancen. Die Regierungsbildung 2003’, in Brigitte Bierlein, Heinrich Neisser, and Klaus Poier (eds.), Über die Pflicht, ein Optimist zu sein. Graz: Styria, 41–64. Kriechbaumer, Robert (2014). ‘ . . . ständiger Verdruss und viele Verletzungen.’ Die Regierung Klima/Schüssel und die Bildung der ÖVP-FPÖ-Regierung. Österreich 1997–2000. Vienna: Böhlau. Kriechbaumer, Robert (2016). ‘Es reicht!’ Die Regierung Gusenbauer—Molterer. Österreich 2007/2008. Vienna: Böhlau. Kriechbaumer, Robert, and Franz Schausberger (eds) (2013). Die umstrittene Wende. Österreich 2000–2006.Vienna: Böhlau. Laver, Michael, and Kenneth A. Shepsle (1990). ‘Coalitions and Cabinet Government’. American Political Science Review, 84(3): 873–90. Laver, Michael, and Kenneth A. Shepsle (1996). Making and Breaking Governments. Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Leiserson, Michael (1966). ‘Coalitions in Politics: A Theoretical and Empirical Study’. PhD dissertation: Yale University. Luebbert, Gregory M. (1986). Comparative Democracy. Policymaking and Governing Coalitions in Europe and Israel. New York: Columbia University Press. Lupia, Arthur, and Kaare Strøm (1995). ‘Coalition Termination and the Strategic Timing of Parliamentary Elections’. American Political Science Review, 89(3): 648–65. Lupia, Arthur, and Kaare Strøm (2008). ‘Bargaining, Transaction Costs, and Coalition Governance’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining. The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 51–84. Luther, Kurt R. (2003). ‘The FPÖ: From Populist Protest to Incumbency’. In Peter H. Merkel and Leonard Weinberg (eds), Right-Wing Extremism in the 21st Century. London: Frank Cass, 191–219. Luther, Kurt R. (2006). ‘Strategien und (Fehl-)Verhalten: Die Freiheitlichen in den Regierungen Schüssel I und II’. In Emmerich Tálos (ed.), Umbau in Schwarz-Blau (Orange). Eine Bilanz. Vienna: Lit Verlag, 19–37. Luther, Kurt R. (2008). ‘Electoral Strategies and Performance of Austrian Right-Wing Populism, 1986–2006’. In Günther Bischof and Fritz Plasser (eds), The Changing Austrian Voter. New Brunswick: Transaction Publishers, 104–22. Luther, Kurt R. (2010). ‘Governing with Right-Wing Populists and Managing the Consequences: Schüssel and the FPÖ’. In Günther Bischof and Fritz Plasser (eds), The Schüssel Era in Austria. Innsbruck: Innsbruck University Press, 79–103.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

78

 . ̈  

Martin, Lanny W., and Georg Vanberg (2011). Parliaments and Coalitions: The Role of Legislative Institutions in Multiparty Governance. Oxford: Oxford University Press. Martin, Lanny W., and Georg Vanberg (2014). ‘Parties and Policymaking in Multiparty Governments: The Legislative Median, Ministerial Autonomy, and the Coalition Compromise’. American Journal of Political Science, 58(4): 979–96. Mitterlehner, Reinhold (2019). Haltung. Flagge zeigen in Leben und Politik. Salzburg: Ecowin. Müller, Wolfgang C. (1993). ‘Executive–Legislative Relations in Austria: 1945–1992’. Legislative Studies Quarterly, 18(4): 467–94. Müller, Wolfgang C. (1996). ‘Wahlsysteme und Parteiensystem in Österreich, 1945–1995’. In Fritz Plasser, Peter A. Ulram, and Günther Ogris (eds), Wahlkampf und Wählerentscheidung. Analysen zur Nationalratswahl 1995. Vienna: Signum, 233–72. Müller, Wolfgang C. (2000). ‘Austria: Tight Coalitions and Stable Government’. In Wolfgang C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 86–125. Müller, Wolfgang C. (2003). ‘Austria: Imperfect Parliamentarism but Fully-Fledged Party Democracy’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press, 221–52. Müller, Wolfgang C. (2005). ‘Austria: A Complex Electoral System with Subtle Effects’. In Michael Gallagher and Paul Mitchell (eds), The Politics of Electoral Systems. Oxford: Oxford University Press, 397–416. Müller, Wolfgang C. (2006a). ‘Parteiensystem: Rahmenbedingungen, Format und Mechanik des Parteienwettbewerbs’. In Herbert Dachs, Peter Gerlich, Herbert Gottweis, Helmut Kramer, Volkmar Lauber, Wolfgang C. Müller, and Emmerich Tálos (eds), Politik in Österreich: Das Handbuch. Vienna: Manz, 279–304. Müller, Wolfgang C. (2006b). ‘Der Bundespräsident’. In Herbert Dachs, Peter Gerlich, Herbert Gottweis, Helmut Kramer, Volkmar Lauber, Wolfgang C. Müller, and Emmerich Tálos (eds), Politik in Österreich: Das Handbuch. Vienna: Manz, 188–200. Müller, Wolfgang C. (2007). ‘The Changing Role of the Austrian Civil Service: The Impact of Politicisation, Public Sector Reform, and Europeanisation’. In Edward C. Page and Vincent Wright (eds), From the Active to the Enabling State. The Changing Role of Top Officials in European Nations. Basingstoke: Palgrave Macmillan, 38–62. Müller, Wolfgang C. (2011). ‘Successful Failure: Ill-conceived Pre-commitments and Welcome Bargaining Failure Paving the Way to Minority Government in Austria’. In Rudy B. Andeweg, Lieven De Winter, and Patrick Dumont (eds), The Puzzles of Government Formation: Coalition Theory and Deviant Cases. London: Routledge, 88–111. Müller, Wolfgang C., and Franz Fallend (2004). ‘Changing Patterns of Party Competition in Austria: From Multipolar to Bipolar System’. West European Politics, 27(5): 801–35.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



79

Müller, Wolfgang C., and Marcelo Jenny (2004). ‘ “Business as usual” mit getauschten Rollen oder Konflikt- statt Konsensdemokratie?’. Österreichische Zeitschrift für Politikwissenschaft, 33(3): 307–24. Müller, Wolfgang C., Wilfried Philipp, and Barbara Steininger (1995). ‘Die Regierung’. In Emmerich Tálos, Herbert Dachs, Ernst Hanisch, and Anton Staudinger (eds), Handbuch des politischen Systems der Ersten Republik. Vienna: Manz, 72–89. Müller, Wolfgang C., and Kaare Strøm (eds.) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Pelinka, Peter (2000). Österreichs Kanzler. Vienna: Ueberreuter. Pelinka, Peter (2003). Wolfgang Schüssel. Eine politische Biographie. Wien: Ueberreuter. Plasser, Fritz, and Franz Sommer (2018). Wahlen im Schatten der Flüchtlingskrise. Vienna: Facultas. Plassser, Fritz, Peter A. Ulram, and Franz Sommer (2000). ‘Do Campaigns Matter? Massenmedien und Wahlentscheidung im Nationalratswahlkampf 1999’. In Fritz Plassser, Peter A. Ulram, and Franz Sommer (eds), Das österreichische Wahlverhalten. Vienna: Signum, 141–73. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of Antielitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Riker, William H. (1962). The Theory of Political Coalitions. New Haven: Yale University Press. Sagarzazu, Iñaki, and Heike Klüver (2017). ‘Coalition Governments and Party Competition: Political Communication Strategies of Coalition Parties’. Political Science Research and Methods, 5(2): 333–49. Schaden, Michael (2006). ‘Verfassungsgerichtsbarkeit’. In Herbert Dachs, Peter Gerlich, Herbert Gottweis, Helmut Kramer, Volkmar Lauber, Wolfgang C. Müller, and Emmerich Tálos (eds), Politik in Österreich. Das Handbuch. Vienna: Manz, 213–31. Schüssel, Wolfgang (2009). Offengelegt. Aufgezeichnet von Alexander Purger. Vienna: Ecowin. Schüssel, Wolfgang (2010). ‘Interview’. In Robert Kriechbaumer and Franz Schausberger (eds), Die umstrittene Wende. Österreich 2000–2006. Vienna: Böhlau, 799–843. Sieberer, Ulrich, Thomas M. Meyer, Hanna Bäck, Andrea Ceron, Albert FalcóGimeno, Isabelle Guinaudeau, Martin Ejnar Hansen, Kristoffer Kolltveit, Tom Louwerse, Wolfgang C. Müller, and Thomas Persson (2021). ‘The Political Dynamics of Portfolio Design in European Democracies’. British Journal of Political Science, 51(2): 772–87. Steininger, Gerhard (2007). Das Dritte Lager. Aufstieg nach dem Fall? Vienna: Edition Steinbauer.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

80

 . ̈  

Strøm, Kaare, Wolfgang C. Müller, and Torbörn Bergman (eds.) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Strøm, Kaare, Wolfgang C. Müller, and Daniel Markham Smith (2010). ‘Parliamentary Control of Coalition Governments’. Annual Review of Political Science, 13: 517–35. Tálos, Emmerich (ed.) (2006). Schwarz-blau. Eine Bilanz des ‘Neu-Regierens’. Vienna: Lit Verlag. Tálos, Emmerich (ed.) (2019). Die Schwarz-blaue Wende in Österreich. Eine Bilanz. Vienna: Lit Verlag. Tsebelis, George (2002). Veto Players. How Political Institutions Work. Princeton: Princeton University Press. Ultsch, Christian, Thomas Prior, and Rainer Nowak (2017). Flucht. Wie der Staat die Kontrolle verlor. Vienna: Molden. von Neumann, John, and Oskar Morgenstern (1953). Theory of Games and Economic Behavior. Princeton, NJ: Princeton University Press. Vranitzky, Franz (2004). Politische Erinnerungen. Vienna: Zsolnay. Welan, Manfried (1997). Das österreichische Staatsoberhaupt. Aufwertung oder Abwertung? Vienna: Verlag für Geschichte und Politik. Wineroither, David (2009). Kanzlermacht—Machtkanzler? Die Regierung Schüssel im historischen und internationalen Vergleich. Vienna: Lit Verlag. Wineroither, David, and Herbert Kitschelt (2012). ‘Die Entwicklung des Parteienwettbewerbs in Österreich im internationalen Vergleich’. In Ludger Helms and David M. Wineroither (eds), Die österreichische Demokratie im internationalen Vergleich. Baden-Baden: Nomos, 193–221. Wolf, Armin (1998). ‘Die 0,5-Promille-Regelung – ein mediengemachtes Gesetz. Sonderfall oder Prototyp?’. Österreichisches Jahrbuch für Politik, 1998: 197–222.

Appendix. List of political parties Abbreviation KPÖ GA SPÖ PILZ ÖVP FPÖ LF BZÖ FRANK NEOS

Name Communist Party of Austria (Kommunistische Partei Österreichs) The Greens—The Green Alternative (Die Grünen—Die Grüne Alternative) Social Democratic Party of Austria (Sozialdemokratische Partei Österreichs) Peter Pilz List (Liste Peter Pilz) Austrian People’s Party (Österreichische Volkspartei) Freedom Party of Austria (Freiheitliche Partei Österreichs) Liberal Forum (Liberales Forum) Alliance for the Future of Austria (Bündnis Zkunft Österreich) Team Stronach for Austria (Team Stronach für Österreich) NEOS—The New Austria and Liberal Forum (NEOS—Das Neue Österreich und Liberales Forum)

Note: Party names are given in English, followed by the party name in German in parentheses.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 4 Belgium: From Highly Constrained and Complex Bargaining Settings to Paralysis? Lieven De Winter and Patrick Dumont

The chapter on Belgium in the previous coalition volume (De Winter et al. 2000) covered the 32 cabinets formed from the short-lived 1946 Spaak I minority government to the Dehaene II government that formed in 1995. This chapter focuses on the subsequent nine cabinets, from the novel Verhofstadt I ‘Rainbow’ coalition (1999–2003) to the equally innovative Michel I ‘Swedish’ coalition (11 October 2014 to 9 December 2018).¹ The former excluded the median Christian Democrats from government for the first time since 1958 and the latter was the first one to contain only one French-speaking party. We compare these new cabinets with their two predecessors of the 1990s, thereby covering Belgian coalition governance in the period between the 1991 and the 2019 elections.² For the 1946–1999 period, we claimed that ‘Belgium undoubtedly has the most complex coalition bargaining system in Western Europe’ (De Winter et al. 1999: 300). This complexity was explained by the increasing party system fragmentation, triggered by a growing saliency of a multitude of cleavages and multilevel politics. However, we also claimed that institutional and behavioural rules served as constraints, reducing the range of possible formation outcomes and types of coalition governance in a highly unpredictable environment. Yet, since 1999, some of these constraints seem to have weakened or even vanished, leading to even more complex formation processes and uncertainty in governing, with even less predictable outcomes. ¹ The ‘rainbow’ coalition contained liberals, socialists, and Greens and was called ‘purple-green’ coalition (a mix of blue liberals and red socialists, with Greens added) in Flanders. In 2014 the coalition was nicknamed ‘kamikaze’ coalition due to the potentially suicidal situation of the MR, the only French-speaking party in the coalition. But the PM’s communication services managed to field the concept of ‘Swedish’ coalition (the blue of liberals, the yellow of the Flemish nationalists, and the crossshape of the Swedish flag to reflect the Christian Democrat participation). ² This chapter is based on nine interviews with senior politicians who collectively occupied the following positions: two federal PMs, seven federal vice-PMs, six party leaders, four regional ministerpresidents (PMs), eight parliamentary party group leaders, and two directors of party research office. It relies on the analysis of the 1999, 2003, 2008, 2010, and 2014 coalition agreements (and government declarations when there was a change of government without a new coalition agreement). In the 2013–2014 period all parliamentary party group leaders and their political secretaries in the Chamber within the project Policy Advice were also interviewed (De Winter and Wolfs 2017). Lieven De Winter and Patrick Dumont, Belgium: From Highly Constrained and Complex Bargaining Settings to Paralysis? In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0004

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

82

   

The institutional setting While the Chamber of Representatives and the Senate are still congruent in terms of party composition, in terms of competences the Chamber has clearly become predominant since the constitutional reforms that took effect in 1995. This asymmetry concerns various components relevant to coalition governance. First, the government making powers of the Senate have been cancelled: only the Chamber can invest or dissolve a government. Second, since this reform only the Chamber approves the budget and conducts interpellations of ministers. Third, all government bills are introduced in the Chamber (though until 2014 the Senate kept the right to discuss and amend them). Both assemblies remain equally competent regarding constitutional and other institutional reforms regarding regional and linguistic matters. The Senate has nearly totally lost its right to discuss and amend bills approved by the Chamber or to initiate bills. While the 1995 constitutional revision intended the Senate to evolve into a reflection chamber and a meeting place for resolving conflicts between the national and regional/community legislatures, in practice it has rather failed in both roles. Hence, over the course of the last two decades, several proposals have been made to revise the role and composition of the Senate, including its abolition (Sägesser and Istasse 2014). The 2014 constitutional reform reduced the Senate into a largely de-funded non-permanent body, without directly elected members, that should only meet in plenary ‘at least’ eight times a year.³ The rules for the elections of the Chamber have also changed. Currently, the 150 representatives are elected in 11 constituencies (the 10 provinces plus Brussels), with district magnitudes varying from 4 to 24 seats. The electoral reforms of 2003 introduced a five per cent threshold at the provincial level. This nearly doubled disproportionality between the parties’ votes and seats, with Gallagher’s index rising from 2.85 in 1999 to 5.16 in 2003—a level that remains comparatively low. The main change in electoral rules actually concerns the timing of federal and regional elections. Since the 1995 constitutional reform the regional assemblies are elected directly, with a fixed legislative term of five years, set at the same date as the European elections (starting from 1999). Hence, the regional elections of 2004 and

³ Between 1995 and 2014, 40 senators were directly elected in 2 large constituencies (25 in the Flemish one and 15 in the French-speaking one). The parliaments of subnational entities (regions and communities) designated 21 ‘community senators’, while another 10 were appointed (6 Flemish and 4 French-speakers) by their directly elected and community-designated peers. This particular seat allocation guaranteed a quasi-perfect proportional composition of the Senate in terms of the population size of the three regions and communities. Since 2014, 50 senators are designated by and among subnational-level parliaments. Ten more senators are appointed by their designated peers, proportionally to the parties’ strength in the federal Chamber of Representatives (Dandoy et al. 2015). Given these radical changes in role and composition, state subsidies to Senate parliamentary parties were cut by 75 per cent, while subsidies to Chamber groups increased accordingly (Maddens et al. 2019).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



83

2009 were desynchronized with the federal elections of 2003, 2007, and 2010. This made the building of vertically congruent coalitions very difficult, as the negotiations leading to the formation of governments at each level were not held synchronously any longer (Deschouwer 2009). In addition, these government formations were also based on electoral outcomes that partially started to follow each level’s own dynamics but also the shifts of the popularity of parties between elections held at the same level, resulting in increasingly different balances of power across parliaments. In fact, election results between the regional and federal levels often varied, not so much due to split-ticket voting but because voters took the opportunity to use them as thermometers of the overall appreciation of the competing parties based on their performance in both arenas.⁴ This (con)fusion between the two electoral arenas was further enhanced by the lack of distinction parties themselves made. For instance, candidate lists for the federal elections featured numerous incumbent regional members of parliament (MPs) and ministers (and vice versa) who had no intention to hop levels but just wanted to—or rather were asked to—back their party.⁵ Hence, during the period of desynchronization of regional and federal elections, most parties used an ‘all hands on deck’ strategy, with regional and federal MPs and ministers participating in both elections in order to maximize their party’s score. Further change came with the following major constitutional reform that stipulated that from 2014 onwards, federal elections would be organized jointly with the fixed-term regional and European ones and multilevel electoral candidacies became prohibited.⁶ This meant that parties could not continue to pursue their ‘all hands on desk’ strategy, where many of these ‘multilevel’ candidates did not intend to give up their regional ministerial portfolio for a ‘mere’ seat in the federal Chamber. Instead, they now run at one single level in order to support their party, usually by ‘pushing’ the list as its last candidate. The 1995 constitutional reform changed cabinet investiture and removal rules (De Winter 1995). A confidence vote is still not required by the constitution after the debate following the reading of the governmental declaration by the prime minister (PM) to the Chamber (only)—the cabinet is in any case sworn in before that by the king—but such a vote has established itself as constitutional convention. What has changed since 1995 is that it is only if the government (1) loses this vote by an absolute majority and (2) the Chamber proposes a successor to the PM ⁴ Split-ticket in synchronized elections have broadly remained below 20 per cent. Willocq and Kelbel (2018: 672) report about 17 per cent at the Belgian level, from about 11.9 per cent in Flanders to 21.6 per cent in Brussels and 18.4 per cent in Wallonia in the newly re-synchronized elections of 2014, levels that compare with those of 1999 (De Winter 2006). ⁵ At the June 2010 general elections, three quarters of the members of the Flemish parliament were candidates for the federal elections. ⁶ In case of parliamentary dissolution at the federal level, the newly elected parliament would only serve until the end of the mandated five years between the preceding joint election and the next.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

84

   

for appointment by the king within three days of the rejection of the motion of confidence that the incumbent government is forced to resign.⁷ This means that losing a confidence vote by a mere relative majority would in principle not affect the taking off of the government, moving the Belgian case into the negative parliamentarism form of investiture rule category (André et al. 2015). Conversely the 1995 reform introduced a constructive motion of no-confidence whereby the Chamber can remove the government and propose a successor to the PM. So far these changes have had no real consequences at the federal level: the key players interviewed have not yet considered relying on the less constraining character of the ‘investiture’ vote to form a minority government—the expectation is still that the vote of confidence will display visible absolute majority support of the executive being formed—and except for the single-party minority Spaak I in 1946 no post-war Belgian cabinet has been removed by a formal (no-) confidence vote.⁸ No change occurred regarding the role of the head of state and of the executive, the king. Whereas King Baudouin (1950–1992) tried and sometimes influenced government formation and maintenance, this was less often the case for his brother Albert II (1992–2013; succeeded by his son Philippe) who was nevertheless quite active during the 2007 and 2010 long formation crises. Within the executive, the formal powers of the PM relate mainly to agenda setting (De Winter and Dumont 2003). The PM remains a primus inter pares who depends on the consensus of his inner cabinet (Kerncabinet, comprised of all vice-PMs),⁹ has a limited role in designing his team, and can only dismiss individual ministers with the consent of their party leader.¹⁰ On the other hand internal cabinet rules and the existence of a powerful inner cabinet prevent policy dictatorship by individual ministers. Note that since 1995 ministers cannot simultaneously hold a seat in parliament (they are replaced and can take it back in case of resignations; Dumont et al. 2009). There have been a few informal changes to the formal functioning of the cabinet. First, given the drastic reduction of the number of ministers to maximally 15 (since 1995) and the limited number of junior ministers, in practice the latter started to also attend the meetings of the cabinet. For the same reason, the large standing cabinet committees such as the Ministerial Committee for Social and ⁷ In case no successor is presented (voted on by a motion of presentation with an absolute majority in favour), the king may dissolve the Chamber (chapter IV of the Standing Orders of the Chamber). ⁸ Several have resigned to pre-empt a no-confidence vote; in the period under study this only happened for Michel I in 2018. ⁹ Since Belgium only had male PMs until October 2019 (date outside the period covered by this chapter) and the nomination of the French-speaking liberal Sophie Wilmès, we use a male pronoun to refer to Belgian PMs. ¹⁰ In Belgian political parties, the common label for the ‘party leader’ is the party ‘president’ (président de parti, partijvoorzitter), who is actually the leader of the party’s Extra Parliamentary Organisation (Fiers 1998). This position is usually incompatible with the leadership of the Parliamentary Party Group (président de groupe, fractievoorzitter), as well as a party’s top office in the government (PM and ‘vice-PM’).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



85

Economic Coordination, which comprised about a dozen of ministers, have been abolished by the first Dehaene cabinet.¹¹ Finally, rules pertaining to the composition the Kerncabinet were adapted as well as the appointment of vice-PM positions. Due to the high number of coalition partners and the limited number of ministerial positions, the Verhofstadt I hexapartite decided not to give a viceprime ministership to each coalition partner. Despite not getting one, the Flemish Greens were represented in the Kerncabinet by their key minister. The PM party did not get a formal vice-PM either but also had a second representative in the Kerncabinet, allowing the PM to act as non-partisan ‘arbiter’. Since the Leterme I cabinet, this evolution has become formalized: each coalition partner, including the party of the PM, is granted a vice-PM and only they are permanent members of the inner cabinet.

The party system and its actors Party system change The Belgian party system is unique, as since 1978, there have been no more politywide parties but only French-speaking or Flemish parties, each seeking to represent their own linguistic community only. The current party system partially reflects the end of the nineteenth-century one, based on three party families. From then onwards, the Christian Democrats (now called CD&V and cdH), Socialists (sp.a and PS), and Liberals (Open VLD and MR) alternately shared government offices in different coalition combinations. Each split into two organizationally and programmatically independent Flemish and French-speaking parties (respectively in 1968, 1972, and 1978) due to increasing internal divisions between Flemings and French-speakers under the electoral pressure of the breakthrough of ethnoregionalist parties during the mid-1960s (among which are the FDF and the Volksunie who later gave birth to the now largest party in Belgium, the Flemish independentist N-VA).¹² At the end of the 1970s, the Flemish separatist far-right Vlaams Blok (now Vlaams Belang) and the Green parties (now Groen and Ecolo) emerged.¹³ Hence, the number of parties in parliament rose dramatically, from four in 1949 to fourteen in 1981. Therefore, since 1978, one cannot strictly speak of a single Belgian or federal party system anymore. Only ¹¹ In 2006 the Ministerial Committee for Information and Security was created. Since 2015 it is labelled National Security Council (chaired by the PM, comprising the ministers of interior, justice, defence, and foreign affairs). ¹² In 2015, the FDF changed its name and is henceforth called ‘DéFI’ (Démocrate Fédéraliste Indépendant). It will, however, still be called FDF in this book because the name change occurred after the formation of the last post-electoral government covered in this chapter. ¹³ These parties made their main electoral breakthrough in the period under study, at the 1991 ‘Black Sunday’ election for the VB and at the 1999 election for the Greens.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

86

   

in the Brussels-Capital Region constituency do the Flemish and French-speaking party systems overlap and all parties compete—at least potentially—for the same voters (De Winter and Dumont 2006). The ‘Belgian’ party system reached a (European) record fragmentation at the 1999 general elections,¹⁴ with 9.1 effective parties in the Chamber of Representatives. Table 4.1a indicates that also at the subsequent elections, fragmentation remained very high (the effective number of parties (ENP) varying between 7.0 and 8.4). The reduction of fragmentation in 2003 was partially due to the creation of electoral cartels between parties that were previously represented separately in parliament: the introduction of a five per cent threshold pushed several small parties to enter electoral cartels with mainstream parties in order to survive.

Cleavage system Although no new cleavages emerged since 1991 and party positions changed only modestly on the four dimensions of competition, coalition dynamics changed dramatically due to the declining size of the mainstream parties around which post-war coalitions were formed. In 1987 the three traditional party families still controlled 86 per cent of the seats in the Chamber, allowing the median party on the left–right divide, the Christian Democrats, to choose their coalition partner among the two other party families—this would be the case until 1999.¹⁵ By the 2014 elections, these three families had dropped to 65 per cent of the seats, gradually making the formation of these two-party family majority coalitions much harder and, with only 72 seats out of 150 in 2019, eventually impossible. All across Belgium the ‘big three’ have been losing to the Green parties; in Flanders they also endured the rise of the xenophobic and separatist VB (ever since the 1991 ‘Black Sunday’ election) and since 2010 of the independentist N-VA. The socio-economic left–right cleavage remains salient. It did not undergo significant changes in terms of the placement of mainstream parties but we however note a modest emergence of new competitors at the fringes of each side of the spectrum. This started with the flash success of a Flemish populist right-liberal party (Lijst De Decker) in 2007 with five MPs and the Frenchspeaking Parti Populaire (one seat in 2010 and 2014). On the opposite side the communist Parti du Travail Belge—Partij van de Arbeid elected two candidates in 2014 before making a genuine breakthrough in 2019 with twelve seats. ¹⁴ In 1999 the largest party won 14 per cent of the national vote, while the 10th party still obtained 5 per cent. ¹⁵ In Table 4.1a, we consider the CD&V as containing the median legislator based on the higher probability of finding him/her in the largest of the two Christian Democratic parties that are generally similarly located on the left–right continuum.

Cabinet

Spaak Van Acker III

Huysmans

Spaak II

Eyskens

Duvieusart Pholien Van Houtte Van Acker IV

Eyskens II Eyskens III

Lefevre

Harmel

Cabinet number since 1946

1 2

3

4

5

6 7 8 9

10 11

12

13

1965-07-27

1961-04-25

1958-06-23 1958-11-06

1950-06-08 1950-08-16 1952-01-15 1954-04-22

1949-08-11

1947-03-20

1946-08-03

1946-03-13 1946-03-31

Date in

Table 4.1a Belgian cabinets, 1946–2018

1965-05-23

1961-03-26

1958-06-01

1954-04-11

1950-06-04

1949-06-26

1946-02-17

Election date

PSB–BSP PSB–BSP, LP–PL, PCB–KPB PSB–BSP, LP–PL, PCB–KPB CVP–PSC, PSB–BSP CVP–PSC, LP–PL CVP–PSC CVP–PSC CVP–PSC PSB–BSP, LP–PL CVP–PSC CVP–PSC, LP–PL CVP–PSC, PSB–BSP CVP–PSC, PSB–BSP mwc (mwc)

mwc (mwc)

min (mwc) mwc (mwc)

mwc (mwc) mwc (mwc) mwc (mwc) mwc (mwc)

mwc (sur)

mwc (mwc)

mwc (mwc)

min (mwc) mwc (mwc)

Party Type of composition cabinet of cabinet (Senate in parentheses)

66.5 (71.9)

84.9 (88.6)

49.1 (52) 59 (61.7)

50.9 (52) 50.9 (52) 50.9 (52) 52.4 (53.7)

63.2 (66.3)

79.7 (82.6)

54 (50.3)

34.1 (32.9) 54 (50.3)

Cabinet strength in seats (%, Senate in parentheses)

212 (178)

212 (175)

212 (175) 212 (175)

212 (175) 212 (175) 212 (175) 212 (175)

212 (175)

202 (167)

202 (167)

202 (167) 202 (167)

Number of seats in parliament (Senate in parentheses)

7 (6)

5 (5)

5 (4) 5 (4)

4 (4) 4 (4) 4 (4) 5 (4)

4 (4)

4 (4)

4 (4)

4 (4) 4 (4)

Number of parties in parliament (Senate in parentheses)

3.59 (3.17)

2.69 (2.57)

2.48 (2.42) 2.48 (2.42)

2.49 (2.49) 2.49 (2.49) 2.49 (2.49) 2.66 (2.6)

2.75 (2.58)

2.91 (2.7)

2.91 (2.7)

2.91 (2.7) 2.91 (2.7)

ENP, parliament (Senate in parentheses)

CVP–PSC

CVP–PSC

CVP–PSC CVP–PSC

CVP–PSC CVP–PSC CVP–PSC CVP–PSC

CVP–PSC

CVP–PSC

CVP–PSC

CVP–PSC CVP–PSC

Continued

Formal Median support party in first policy parties dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Cabinet

Van den Boeynants I Eyskens IV

Eyskens V

Leburton

Tindemans

Tindemans II

Tindemans III

Tindemans IV

Van den Boeynants II

Martens I

Cabinet number since 1946

14

15

16

17

18

19

20

21

22

23

Table 4.1a Continued

1979-04-03

1978-10-20

1977-06-03

1977-03-06

1974-06-12

1974-04-25

1973-01-26

1972-01-21

1968-06-17

1966-03-19

Date in

1978-12-17

1977-04-17

1974-03-10

1971-11-07

1968-03-31

Election date

CVP–PSC, LP–PL CDV, CDH, PSB–BSP CDV, CDH, PSB–BSP PSB–BSP, CDV, CDH, VLD, MR CDV, CDH, VLD, MR CDV, CDH, VLD, MR, RW CDV, CDH, VLD, MR CDV, CDH, PSB–BSP, VU, FDF CDV, CDH, PSB–BSP, VU, FDF CDV, CDH, PS, SP, FDF 80.7 (81.2)

sur (sur)

71.2 (74.6)

81.1 (81.2)

sur (sur)

sur (sur)

49.5 (56.4)

54.2 (59.1)

48.1 (51.4)

76.4 (78.1)

60.4 (61.8)

60.4 (65.7)

59 (65.2)

Cabinet strength in seats (%, Senate in parentheses)

min (mwc)

mwc (sur)

min (mwc)

sur (sur)

sur (sur)

sur (sur)

mwc (mwc)

Party Type of composition cabinet of cabinet (Senate in parentheses)

212 (181)

212 (181)

212 (181)

212 (181)

212 (181)

212 (181)

212 (178)

212 (178)

212 (178)

212 (178)

Number of seats in parliament (Senate in parentheses)

12 (10)

9 (9)

9 (9)

9 (8)

9 (8)

9 (8)

9 (9)

9 (9)

8 (7)

7 (6)

Number of parties in parliament (Senate in parentheses)

6.92 (6.3)

5.24 (5.21)

5.24 (5.21)

5.81 (5.43)

5.8 (5.43)

5.8 (5.43)

5.85 (5.75)

5.85 (5.75)

4.97 (4.6)

3.59 (3.17)

ENP, parliament (Senate in parentheses)

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CVP–PSC

Formal Median support party in first policy parties dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Martens IV

M Eyskens

Martens V

Martens VI

Martens VII

Martens VIII

Dehaene I

Dehaene II

Verhofstadt I

Verhofstadt II

Verhofstadt III 2003-07-12

Verhofstadt IV*

Leterme I

26

27

28

29

30

31

32

33

34

35

36

37

38

2008-03-20

2007-12-21

2003-05-05

1999-07-12

1995-06-23

1992-03-07

1991-09-29

1988-05-09

1985-11-28

1981-12-17

1981-04-06

1980-10-22

1980-05-18

Martens III

25

1980-01-23

Martens II

24

2007-06-10

2003-05-18

1999-06-13

1995-05-21

1991-11-24

1987-12-13

1985-10-13

1981-11-08

CDV, CDH, PS, SP CDV, CDH, PS, SP, VLD, MR CDV, CDH, PS, SP CDV, CDH, PS, SP CDV, CDH, VLD, MR CDV, CDH, VLD, MR CDV, CDH, PS, SP, VU CDV, CDH, PS, SP CDV, CDH, PS, SP CDV, CDH, PS, SP VLD, MR, SP, PS, E, G VLD, MR, SP, PS, G VLD, SP, PS, MR VLD, CDV, PS, MR, CDH CDV, VLD, PS, MR, CDH 67.3 (66.2)

67.3 (66.2)

sur (sur) sur (sur)

64.7 (64.8)

55.3 (57.7)

62.7 (66.2)

54.7 (54.9)

56.6 (58.7)

63.2 (65.6)

70.8 (72.7)

54.2 (55.7)

53.3 (54.7)

66 (69.6)

mwc (sur)

mwc (sur)

sur (sur)

mwc (mwc)

mwc (mwc)

sur (sur)

sur (sur)

mwc (mwc)

mwc (mwc)

sur (sur)

66 (69.6)

83.5 (84.5)

sur (sur) sur (sur)

66 (69.6)

sur (sur)

150 (71)

150 (71)

150 (71)

150 (71)

150 (71)

150 (71)

212 (184)

212 (183)

212 (183)

212 (183)

212 (181)

212 (181)

212 (181)

212 (181)

212 (181)

11 (11)

11 (11)

10 (9)

11 (10)

11 (10)

11 (10)

12 (11)

11 (11)

11 (11)

12 (10)

14 (13)

12 (10)

12 (10)

12 (10)

12 (10)

8.01 (8.39)

8.01 (8.39)

7.12 (7.56)

9.17 (8.86)

9.17 (8.86)

8.23 (8.09)

8.48 (8.01)

7.26 (6.93)

7.26 (6.93)

7.07 (6.68)

7.95 (7.58)

6.9 (6.3)

6.9 (6.3)

6.9 (6.3)

6.92 (6.3)

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

CDV

Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Van Rompuy

Leterme II

Di Rupo

Michel I

Michel II

39

40

41

42

43

2018-12-09

2014-10-11

2011-12-06

2009-11-25

2008-12-30

Date in

2014-05-25

2010-06-13

Election date

CDV, VLD, MR, PS, CDH CDV, MR, VLD, PS, CDH PS, CDV, MR, SP, VLD, CDH MR, CDV, VLD, NVA MR, CDV, VLD min (min)

34.7 (33.3)

56.7 (56.7)

62 (63.4)

sur (sur) mwc (mwc)

62.7 (63.4)

62.7 (63.4)

Cabinet strength in seats (%, Senate in parentheses)

sur (sur)

sur (sur)

Party Type of composition cabinet of cabinet (Senate in parentheses)

150(52)

150 (60)

150 (71)

150 (71)

150 (71)

Number of seats in parliament (Senate in parentheses)

11 (9)

11 (10)

12 (10)

12 (12)

12 (12)

Number of parties in parliament (Senate in parentheses)

8.33 (7.61)

8.01 (7.86)

8.71 (7.96)

9.04 (9.08)

9.04 (9.08)

ENP, parliament (Senate in parentheses)

CDV

CDV

CDV

CDV

CDV

E, G**

Formal Median support party in first policy parties dimension

Table legends and variables are further defined in the Appendix, this volume.

Cabinet types: min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition.

First policy dimension is economic left–right.

Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), 1946–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015).

** The Green parties only supported the institutional reform part of the coalition agreement, not the full government policy agenda.

Notes: For a list of parties, consult the chapter appendix. * Limited policy remit

Cabinet

Cabinet number since 1946

Table 4.1a Continued OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



91

On the church–state divide, historically the first cleavage, only some moral issues such as abortion, euthanasia, and same-gender marriages still pit the Christian democrats (and the far right) against the ‘secular parties’ (Greens, Socialists, and Liberals). This makes it by now the least salient divide, with however two main qualifications. First, policy proximity on this dimension allowed for the formation in 1999 of the first cabinet that excluded the Christian Democrats from government since 1958. Made of secular parties only, the Verhofstadt I to III coalitions (1999 to 2007) did indeed carry a number of ethical reforms that had been previously vetoed by the Christian Democrats. Second, the actual contemporary version of the religious cleavage largely overlaps with the GAL–TAN divide. Standing on the GAL-pole are the Greens, who are among the most successful in Europe despite their uneven electoral performances.¹⁶ On the TAN-pole we find in Flanders the Vlaams Belang, which was originally born on the community cleavage (infra). It reached its peak in 2003 with 18 seats out of 150 and managed to stay above the 10-seat level until a severe defeat in 2014 elections (and a striking recovery to its best electoral level in 2019). In the Frenchspeaking landscape, we only find the Front National (one seat from 2003 to 2010). The cleavage that created most volatility within—and polarization between— the two party systems is the community divide, which pits Flemish interests (60 per cent of the population) against those of the French-speakers (40 per cent of the population).¹⁷ The issues on this divide and the large variety of ‘state reforms’ to appease their conflicting linguistic and increasingly socio-economic tensions were the main source of political instability since the late 1960s, and only severe economic crises could at times, and temporarily, put the ‘genie back into the bottle’. Within Flanders, the most ‘autonomist’ party is the separatist VB, then the independentist N-VA, while the other parties defend various degrees of (con)-federalism (calling for more devolution in some sectors but some recentralization in others). Throughout the period all French-speaking parties oppose further devolution and tend to ‘form a front’ against Flemish claims for economic and fiscal autonomy. In short, cleavages and party positioning within the two party systems have not changed much, but the battle for regional leadership, the polarization, and the drifting away of Flemish and French-speaking parties on other dimensions have become more intense since 1999. The leadership of the PS has been contested by the liberal MR on the French-speaking side (the latter overtaking the former in 2007). In the Flemish party system, larger electoral volatility and a fragmentation among mid-sized parties have resulted in four different parties becoming the ¹⁶ After their best result ever in 1999 and their participation in the regional and federal governments, the Flemish Greens lost all their nine seats in the Chamber in 2003 before regaining federal representation in 2007. ¹⁷ Post-electoral surveys indicate that vote shifting among individual voters between two successive elections is about 30 per cent (with a peak of 41 per cent in 2014; Dassonneville and Baudewyns 2014).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

92

   

leader (in seats) since 1999.¹⁸ Given that since 1995 the plurality winner in each region is the first formateur party of the regional government, this competition has had important consequences for the vertical congruence of coalitions, and both contributed to a paralysis at the federal level. The latter is the result of multiple exclusives within and across the two party systems, the Flemish one having evolved substantively on the future of the state and on the GAL–TAN divide (under the pressure of the VB) and the move to the socio-economic right of the N-VA in 2014. The French-speaking party system is still led by the socialist party, but it is increasingly under the pressure on its left by the Greens and the PTB.

Electoral alliances and pre-electoral coalitions (1999–2019) The main trigger for this volatility in Flanders was the 2001 split of the democratic Flemish-nationalist Volksunie into a small independentist Nieuw-Vlaamse Alliantie (N-VA) and even smaller post-nationalist left-liberal Spirit, which had to look for potential partners given the introduction of a five per cent electoral threshold in 2003. The formation of electoral cartels (see Table 4.1b) became a coalition game-changer in the 2003–2009 period. With the defunct VU’s 10 per cent of the votes potentially up for grabs, the Flemish Christian Democrats, socialists, and liberals (whose electoral scores started to converge in the 20–25 per cent range) fought for political and cabinet leadership both in Flanders and at the federal level. Hence, the socialist sp.a signed a successful cartel with Spirit for the 2003 elections. The Vlaams Kartel formed by Christian Democrats with the Table 4.1b Electoral alliances and pre-electoral coalitions in Belgium, 1991–2019 Election date

Constituent parties

Type

Types of pre-electoral commitment

2003-05-18 2007-06-10

SP, Spirit SP, Spirit

EA, PEC EA, PEC

CDV, NVA

EA, PEC

Written contract, joint press conference Written contract, joint press conference, Separate declarations, Other* Written contract, joint press conference

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other * Joint election congress

¹⁸ The Open VLD first overtook the CD&V in 1999; at the 2007 federal election the CD&V–N-VA cartel became the largest group, and since 2010 the N-VA dominates the Flemish landscape. Note that in the 2004 Flemish regional election, the VB (thus a fifth party) had the highest number of seats (but the CD&V–N-VA cartel had the largest parliamentary group). Moreover the sp.a (a sixth party) came a close second in votes and could have beaten the Open VLD at the 2003 federal election.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



93

N-VA performed even better, allowing the former to come back at the helm of regional (2004) and federal (2007) governments after a period in opposition. In both cases the formation of these cartels however radicalized the community programme of their mainstream host cartel partner. It eventually also led to the rise of the N-VA to the rank of the largest party both in Flanders and in Belgium since the 2010 elections, after having broken in 2008 the cartel that had originally saved it from disappearance. Hence, from being the largest and pivotal party in the 1958–1999 period, the Flemish Christian Democrats first lost their formateur party status in the 1999–2007 period, to gradually become a mere midsize player after 2010.

Government formation Table 4.2 indicates a tendency towards increasing formation duration time and bargaining attempts, the former most often taking more than 100 days after elections since 1999. These reflect not only the difficulty of forming coalition governments in an ever more fragmented and polarized bargaining environment but also the main players’ behaviour who decided to break away with previous informal formation rules. Those two factors were largely interdependent and led to the formation of new types of coalition, themselves bringing novel patterns of coalition governance. What most of the governments formed since 1999 have in common, and distinguish themselves from the previous period, is that the Christian Democrats were (with the exception of 2007–2010) not in charge of leading the formation process. Despite its unconventional character, the unconnected sixparty coalition of Verhofstadt I was formed in one stroke in only 29 days— ironically the shortest post-electoral formation since 1958. Although a simplified version (still unconnected on the socio-economic divide but composed of only four parties) of its predecessor, the Verhofstadt III coalition took longer to form, 55 days, but still without any failed bargaining attempt.¹⁹ Leaving aside technical, inter-election government formations (the Frenchspeaking Greens leaving cabinet in 2003 and two PM changes in 2008 and one in 2009 without coalition composition change) that either took zero days to a week (in the case of Van Rompuy taking over Leterme after the latter was

¹⁹ The liberal and socialist parties had won the 2003 general elections and now had a clear majority. They decided to continue to govern together although their party leaders knew that the electoral benefits of the risky unconnected, ‘contre-nature’ (as the PS vice-PM herself called it) formula had already been consumed, in part because the economic situation had changed since 1999 and the budget did not allow for pleasing their respective electorates anymore. In addition, all four parties had now become of about equal strength, triggering an internal competition for coalition leadership between the two party families, which had been largely absent in 1999.

Year in

1992

1995 1999 2003 2003 2007

Cabinet

Dehaene I

Dehaene II Verhofstadt I Verhofstadt II Verhofstadt III Verhofstadt IV

0 0 0 0 4

3

Number of inconclusive bargaining rounds

VLD, MR, CDV/ NVA, CDH, PS (1) MR, VLD, CDV/NVA, CDH (2) CDV/NVA, CDH, VLD, MR (3) CDV/NVA, CDH, VLD, MR (4) CDV/NVA, CDH, VLD, MR

CDV, CDH, PS, SP (1) VLD, MR, PS, SP, (E, G) (2) CDV, CDH, PS, SP (3) CDV, CDH, VLD, MR, PS, SP

Parties involved in the previous bargaining rounds

94

39

10

21

0 0 0 0 4

21

44

14 9

Bargaining duration of individual formation attempt (in days)

Table 4.2 Government formation period in Belgium, 1991–2018

33 29 0 55 194

104

Number of days required in government formation

26 27 0 52 168

88

Total bargaining duration

80 90 96 NA 97

118

Pro

1 1 0 NA 1

0

Abstention

61 47 49 NA 47

82

Contra

Result of investiture vote

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2008 2008 2009 2011

2014

Leterme I Van Rompuy Leterme II Di Rupo

Michel I

3

0 0 0 7 PS, SP, CDV, CDH, VLD, MR (1) PS, SP, CDV, CDH, E, G, NVA (2) PS+NVA, SP, CDV, CDH, E, G (3) NVA, PS, SP, CDV, CDH, E, G (4) SP, PS, CDV, CDH, E, G, NVA (5) MR, VLD, PS, CDV, CDH, NVA (6) PS, SP, CDV, CDH, E, G, MR, VLD, NVA (7) PS, SP, CDV, CDH, E, G, MR, VLD MR, VLD, CDV, NVA (1) NVA, MR, CDV, CDH (2) MR, VLD, NVA, CDV (3) MR+CDV, VLD, CDH, NVA 1

25

28

78

84

128

27

97

10

31

57

0 0 0 54

139

0 8 0 541

107

0 8 0 488

84

97 88 83 89

1

1 0 0 0

58

48 45 53 54

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

96

   

forced to resign),²⁰ we thus observe two rather short post-electoral formations with new (or rare) party compositions in the period under study. The more recent post-electoral formations were however much longer (they all took more than 100 days) and bargaining failure-ridden. On the podium we find probably the most ideologically compact cabinet of the period, the centre-right four-party connected coalition of Michel I (2014), which nevertheless took 139 days to form. Verhofstadt IV’s five-party coalition took no less than 194 days to form after the 2007 elections. Both had several inconclusive bargaining rounds although only five parties (or cartel) were around the table. Finally it took 541 days and seven inconclusive rounds (in total no less than nine parties participated in one or more of the bargaining attempts) to form Di Rupo’s six-party coalition. This world famous, record-breaking formation duration occurred after the Leterme I cabinet did not deliver on its promises to reform the state and solve some perennial community-related institutional issues (infra). One of the coalition partners (the Flemish liberals) decided to pull out the plug and early elections were held in June 2010. This resulted in having the N-VA, a party whose goal is the creation of an independent Flemish Republic, becoming the largest one in the Chamber, while on the French-speaking side, the socialist party regained its leadership position. Negotiations started between those plurality winners in each linguistic camp, under multiple (in-)formateurs and in various constellations (Christiandemocrats, liberals, socialists, N-VA). Eventually the Flemish nationalists did not take part in the Di Rupo coalition, judging in July 2011 that the state reform proposals of the French-speaking socialist formateur were unsatisfactory. As amending the constitution requires a two-thirds majority, the other parties turned to the Greens, who agreed to take part in the negotiations on the state reform and to support the government from the opposition ranks on this topic.²¹ This agreement among eight parties on the sixth state reform was signed on 12 October 2011 and the six-party cabinet was sworn in some two months later. In spite of its very difficult birth, the Di Rupo government completed its full legal term without much internal crises.

Unconnected coalitions The formation of the Verhofstadt I coalition is probably the most theoretically and empirically puzzling one in the period under study (Dumont 2011; Dumont et al.

²⁰ As also two other ministers of his party resigned, and also a new Speaker for the Chamber had to be found, it took a few days to fill these new positions while respecting proportionality and other informal rules. ²¹ In the period under consideration, support arrangements from outside the government only occurred in order to solve the problem of reaching two-thirds majorities for institutional reforms. Hence, in 2007–2008, some support was given by the N-VA for half a year in order to give a chance to further negotiations to form a two-thirds majority.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



97

2011). First, it broke away with the post-war pattern of the Christian Democrats choosing a coalition partner between socialist and liberals. For the first time in 41 years, the median Christian Democrats were not included, making the coalition unconnected on the left–right dimension. Second, for the first time the Greens, although partially a surplus party, entered a federal government.²² This government was a success, as liberals and socialists won the subsequent election and decided to continue their unconnected coalition.²³ Several factors reinforced the cohesion and success of this unprecedented heterogenous coalition: (1) the wish of the liberals and socialists to exclude the ‘natural’ government party family, the Christian Democrats, at both the federal and regional levels; (2) the temporary end of the eight years of austerity policies of the Dehaene I and II centre-left governments aimed at reaching the Maastricht convergence criteria; (3) their policy proximity on ethical issues (such as euthanasia and same-sex marriage); (4) their commitment to downplay any community/linguistic issues that could destabilize the coalition; and (5) the ambition of the ‘Verhofstadt generation’ (in all parties) to rejuvenate politics and provide a more positive image of a country (‘model state’) that had increasingly been seen at home and abroad as a ‘failed state’ in the preceding years. The government did bet on a compartmentalized rather than compromise-oriented logic of policy coherence in both its coalition agreement and portfolio allocation (Dumont 2011).

Horizontal and vertical symmetry Other informal composition rules were also regularly broken since 1999. First, coalitions broke with the basic formation custom of horizontal symmetry, a number of that is having both parties of a same party family together either in government or in opposition. From 2007 to 2011 four cabinets did not include the Flemish socialists while their French-speaking sister party was part of those coalitions. This asymmetrical coalition started with a short-lived government headed by the incumbent liberal PM but with a quite distinct party composition as his previous cabinet since the Christian Democrats were now back in business. Verhofstadt IV was actually a limited (14 members overall) and temporary (a fixed three-month duration) team by design, destined to ‘land’ a PM from the winning party of the 2007 election, the CD&V. The horizontal asymmetry was even more blatant in the ²² The inclusion of the Flemish Greens was unnecessary at the federal level but needed to avoid inviting the Christian Democrats in the simultaneous government formation at the Flemish regional level. ²³ The French-speaking Greens left the coalition a few weeks before the elections because the Council of Ministers refused to adopt their minister’s plan to ban night flights at the national airport—located in Flanders but causing sound and safety nuisance over Brussels—and got a serious electoral beating. Hence, the Verhofstadt II five-party government was very short-lived, preparing for already-called elections.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

98

   

2014 Michel government, which only contained one party on the French-speaking side, the liberal MR, for three parties on the Flemish side (the Christian Democrats were present without their French-speaking counterpart cdH). Second, most post-1999 coalitions broke with another basic formation tradition, that of vertical symmetry: as far as numerically possible, coalitions at the regional level tended to include the same parties as the ones in the national-level coalition. As the negotiators at both levels were often identical (same party leaders and party experts), they usually managed to link the formation of executives at different levels and force the entry of their party at all levels. While this was often a difficult norm to respect in the Brussels region for numeric reasons,²⁴ the symmetry between the federal coalition and the Flemish and Walloon regional executives has vanished, largely due to the decision of desynchronizing federal and regional elections, from 2003 on. Once again, the most blatant example was the constitution of Michel I in 2014 where, on the one hand, the MR was only in government at the federal level while, on the other hand, none of the parties involved in the French-speaking regional governments (PS, cdH and FDF), formed at the same time as the federal cabinet, was included in the latter.

A non-constrained formation: the Michel I cabinet (2014) The ‘mother of all elections’ of May 2014 (the regular federal elections coincided with the regional elections as well as with the European elections) produced an opportunity to restore some of the stabilizing rules of the past such as vertical symmetry. In addition, the outgoing coalition parties collectively gained one seat, making a continuation of the Di Rupo six-party coalition numerically possible. However, the morning after the election, the Flemish Christian Democrats decided that it would be unwise to exclude the largest party again, given that the N-VA had made significant further electoral progress (to the detriment of the farright VB rather than the incumbent Flemish coalition partners). As the N-VA had sworn not to govern with the PS, the leaders of the latter party interpreted the CD&V’s move as a preparation of a federal centre-right coalition (made of the NVA, Christian Democrats, and liberals), which could be enabled by the king’s nomination of the N-VA leader Bart De Wever as informateur. As a result, the PS as the largest French-speaking status promptly started the formation of regional governments in Wallonia and Brussels with the Christian Democrats (cdH), upon which the N-VA as the largest party started formation talks with the CD&V for the formation of the Flemish regional executive.

²⁴ The Brussels regional executives, in part due to the much weaker strength of all Flemish parties, knew this kind of coalition asymmetry already in the 1990s.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



99

Instead of a potential reinstatement of full vertical symmetry these precocious subnational government formations guaranteed a different outcome as the concurrent formation processes were driven by different leading parties that had no intention to govern together at the federal level.²⁵ In addition horizontal symmetry was not reached either, as the French-speaking Christian Democrats (who just formed a government with the PS at the subnational level) refused to join the federal coalition. Moreover, as a result, the liberal MR remained the only Frenchspeaking party at the bargaining table and eventually in government, with a mere 20 seats in the Chamber compared with a combined 65 for its three Flemish partners. Although it was rewarded for it, gaining the prime ministership and half of the ministerial positions, this under-representation challenged the democratic legitimacy of the MR’s leadership of a Flemish-nationalist-dominated government. Bringing back memories of the 1970s when the most influential decisionmakers of Belgian governments were not ministers but rather party leaders, this cabinet also had a ‘ghost cabinet leader’: the leader of the N-VA did not join the government and decided to remain the mayor of Antwerp.

The allocation of ministerial portfolios The 1995 constitutional reforms imposed a maximum size of 15 cabinet ministers and maintained the linguistic ‘parity rule’ that requires since 1970 the Council of Ministers to count an equal number of Dutch and French-speaking members. Several paths exist in order to reach a proportional allocation of governmental offices across coalition parties to compensate for the usually larger parliamentary base of the cabinet coming from the Flemish linguistic group of the Chamber.²⁶ First, for most of period since the adoption of the parity rule, PMs have been Flemish and were exempted from the count: since 1995 a typical cabinet would consist of one Flemish PM, seven Flemish ministers and seven French-speaking ministers.²⁷ Second, junior ministers (‘secretaries of state’ or more punctual issue-

²⁵ The N-VA and CD&V had first privileged a coalition only containing themselves at both the federal and Flemish levels. They however had to invite the Flemish liberals to reach a majority at the federal level when the cdH refused to participate. As a result, the Open VLD imposed itself in the Flemish government as well, but was only invited at the time of portfolio allocation (it therefore had no influence on the coalition agreement), making the federal and regional governments congruent in Flanders and relegating the socialists in opposition for the first time since 1988 (at the Flemish level). ²⁶ A consociational rule respected in many other institutional and informal decision-making arenas (see Andeweg et al. 2008; Deschouwer 2012). ²⁷ Note that the two most recent Belgian PMs have been French-speakers—after three decades of Flemish heads of government. Since the parity rule actually specifies that the PM can but does not have to be exempted from the linguistic count, in the 14-minister Michel I cabinet the PM counted as the seventh French-speaker minister. Both its predecessor (Di Rupo) and its successor (Michel II, after the N-VA left the government) were composed of an odd number of ministers however (13 ministers), in which case the PM is automatically exempted from the parity count.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

100

   

specific ‘government commissioners’) are often added to the mix. Since these are formally not members of the Council of Ministers, neither numeric nor linguistic rules apply to these positions, which can therefore be mostly filled by members of Flemish coalition partners. Finally, some non-governmental positions are included in the portfolio bargaining basket, such as the Speakers of the Chamber and the Senate and the Belgian EU Commissioner. A linguistic balance is implemented for the two Speakers,²⁸ but this does not account for the increasing asymmetry in powers and relevance of the respective houses. It is therefore not a surprise to see that French-speaking parties only had the Speaker of the Chamber position for four years in the 1999–2019 period. In the extreme case of the Michel I cabinet, where only one French-speaking party agreed to enter a government with the N-VA, the linguistic parity rule forced the appointment of seven French-speaking ministers overall despite the fact that this party only held 20 seats in the Chamber. In comparison the Flemish coalition parties which counted 14 (Open VLD), 18 (CD&V), and 33 (N-VA) seats in the Chamber respectively received, 2, 2, and 3 ministers. There was a compensation in the qualitative allocation of ministerial portfolios, the most prestigious generally going to Flemish parties while the MR had to take some remits that none of their partners wanted, in part because of the likelihood of difficult and unpopular reforms to be implemented. In addition, all 4 junior minister positions were given to the Flemish parties (respectively 1, 1, and 2). Since the functional distinction between ministers and secretaries of state gradually ceased to apply, this means that there were 7 French-speaking and 11 Flemish members around the cabinet’s Oval Table. The N-VA also got the Speaker of the Chamber and the CD&V got the EU Commissioner. Conversely, the Verhofstadt III coalition had the most equal-sized partners possible for a four-party coalition: the two largest parties relied on 25 seats each and the smaller ones on 24 and 23. In linguistic terms, the division was 49 seats (French-speaking parties) to 48 (Flemish parties). In terms of party families, the liberal family got 49 seats, the socialists 48. Proportionality was largely respected with altogether 7 liberal ministers (including the PM) and 8 socialist ones. When junior ministers were added there were 10 liberals and 11 socialists. In addition, the chairs of the Chamber and of the Senate were awarded to the liberals, while the socialist European Commissioner appointed in 1999 was completing his mandate in Brussels. Altogether, at the start of the Verhofstadt III government there were thus 12 liberals and 12 socialists in the top positions to be allocated. In 2004, a French-speaking liberal replaced the French-speaking socialist European Commissioner. As a result, the French-speaking liberal chairman of the Senate was replaced by a French-speaking socialist to keep the books straight.

²⁸ A rolling balance is also sought for the European Commission position.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



101

Given the constitutional parity rule, in practice party leaders first negotiate the number of ‘portfolio chips’ for each linguistic community. They then allocate these points to the specific parties according to their parliamentary strength and weight of the portfolios.²⁹ In the first round, each party leader then chooses his most preferred portfolio, starting with the leader of the strongest coalition party and then alternating between partners. Just like for the overall ‘points system’, the coalition partners may refer more or less explicitly to the D’Hondt method to settle the order in which each party expressed its preferred choice among the remaining portfolios.³⁰ Once all portfolios are distributed amongst the players, the party leaders evaluate the initial result, which can sometimes be less than optimal for all parties. In this case, new bilateral and multilateral negotiations occur, until a new equilibrium is reached. These secondary bargaining rounds sometimes require the augmentation of the number of portfolios or the shifting of certain competencies from one minister to another. When elections were synchronized, payoffs at the level of the regional and community executives could also be considered among parties of the same community, provided the coalitions were congruent. Hence, during the stage of portfolio distribution, many scenarios could be tried out before one arrived at a final solution. In terms of qualitative portfolio allocation, Table 4.3 indicates that that all three traditional party families had the opportunity to hold the PM position in the 1999–2019 period.³¹ The liberals, who had only held this position for six months in the entire twentieth century (1937–1938), have been most often at the helm since 1999. What is more, both the Flemish and the French-speaking liberals managed to appoint one each.³² Moreover, in 2011 the French-speaking socialists were able to take the biggest prize, for the first time since 1973, whilst the once dominant CD&V only headed three cabinets (with two PMs) in the short and chaotic 2008–2011 period.³³ Whereas they were totally absent from foreign affairs

²⁹ The quantitative portfolio allocation weighing system (PM = 3 points; ministers, assemblies’ speakers, and EC Commissioner = 2 points; junior ministers = 1 point) was intended as a way to counteract the linguistic parity rule effect and actually usually improved proportionality. It is still in use, but in a rather flexible way, allowing some give and take in the qualitative side of portfolio allocation. ³⁰ Depending on the strength and experience of coalition parties, some ‘first choices’ are not accepted by partners, as was the case for the French-speaking Greens who did bid for the social affairs department in 1999 but got denied this portfolio by the French-speaking socialists claiming it was ‘theirs’ (Dumont 2011). ³¹ Despite being the largest single party in Belgium since 2010, and being in government from 2014 to 2018, the independentist N-VA so far never provided the PM. ³² In 1999 and 2003 the liberals were the largest party family and the Open VLD was the largest party of the country. In 2014 the largest party family, the socialists, were kept out of government. The liberals were the second largest and actually the only party family fully represented in government; with 34 seats overall they were just larger than the N-VA on its own, and the French-speaking MR was this time the largest of the two components of the liberal family. ³³ In his two spells as PM combined, Yves Leterme actually spent less time leading a fully empowered cabinet than a caretaker one!

1992

1995

1999

Dehaene I

Dehaene II

Verhofstadt I

2008

2008

2009

2011

2014

2018

Leterme I

Van Rompuy

Leterme II

Di Rupo

Michel I

Michel II

Verhofstadt IV 2007

Verhofstadt III 2003

Verhofstadt II 2003

Year in

Cabinet

5 CDV, 5 PS, 3 SP, 2 CDH 5 CDV, 5 PS, 3 SP, 2 CDH 4 VLD, 3 PS, 3 MR, 3 SP, 1 G, 1 E 4 VLD, 4 PS, 3 MR, 3 SP, 1 G 4 VLD, 4 PS, 4 SP, 3 MR 4 CDV, 3 MR, 3 PS, 3 VLD, 1 CDH 4 CDV,4 VLD, 3 MR, 3 PS, 1 CDH 4 CDV,4 VLD, 3 MR, 3 PS, 1 CDH 5 CDV, 3 MR, 3 PS, 3 VLD, 1 CDH 3 PS, 3 MR, 2 CDV, 2 SP, 2 VLD, 1 CDH 7 MR, 3 NVA, 2 VLD, 2 CDV 7 MR, 3 VLD, 3 CDV

Number of ministers per party (in descending order)

13

14

13

15

15

15

14

15

15

15

15

15

Total number of ministers

1 CDH, 1 SP

1 CDH, 1 MR

1 CDH, 1 MR

1 CDH, 1 MR

1 VLD, 1 G

Number of watchdog junior ministers per party

14

16

17

17

17

17

17

17

17

17

16

17

Number of ministries

Table 4.3 Distribution of cabinet ministerships in Belgian coalitions, 1991–2018

MR

MR

PS

CDV

CDV

CDV

VLD

VLD

VLD

VLD

CDV

CDV

1 Prime minister

VLD

NVA

CDV

MR

MR

MR

MR

MR

MR

MR

CDH

CDH

2 Finance

MR

MR

MR

CDV

VLD

VLD

VLD

MR

MR

MR

SP

SP

3 Foreign affairs

VLD

VLD

PS

PS

PS

PS

PS

PS

SP

SP

PS

PS

4 Social affairs

CDV

NVA

CDH

VLD

VLD

VLD

VLD

VLD

MR

MR

SP

SP

5 Interior

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



103

since the Second World War, ever since 1999 the liberals have been in charge of it.³⁴ This is one of the most striking changes together with their taking over of the Christian democrats’ chasse gardée Agriculture portfolio, which further shows the different strength and role of the liberals—permanently in power in the 1999–2019 period—as they could at the same time keep most of their hold on finance, interior, or middle classes. Socialists have been kept out of interior, justice, finance, and budget most of the time, as well as from the newer immigration/integration ministerial remits created in 2008, but held on to social affairs and health when in government. Finally, new parties in government mostly managed to get the portfolios they hoped for, the Greens receiving in 1999 transports, environment, and energy as well as one of the social remits (health) while the independentist N-VA successfully claimed the basic sovereign authority functions of the state, finance, interior, and defence, as well as the immigration department in 2014.³⁵ As to the selection and management of ministerial personnel, the personal influence of the PM is severely restricted by the party leaders, at least since the 1960s. Each party leader is decisive in choosing who his/her cabinet members will be.³⁶ This means that if the new PM is also his own incumbent party leader, he will of course decide on his party’s ministerial personnel. Furthermore, the PM will only dismiss individual ministers with the consent of the contested minister’s party leader.³⁷ Yet, party leaders are evidently not entirely free in nominating those who they personally prefer. Former ministers will demand to be reappointed, and politicians representing strong constituency parties or intra-party factions will demand ministerial representation. Ministerial appointments have to fairly represent the provinces and constituency parties, as well as gender.³⁸ In addition, a party leader’s discretion depends strongly on his/her power within his party, as well as on his/her personal resources and background.

³⁴ Except for the five months of the fully empowered Leterme II cabinet, and its continuation as a caretaker. ³⁵ Given that there was a terrorist attack (in the Jewish museum in Brussels) the day before the 2014 election, both interior and defence were expected to be important portfolios. The N-VA had also been very active in the previous parliamentary term on those policy remits, including the voting of immigration laws by a floating majority of centre-right parties during the 541 days of the Leterme caretaker government. On top of these the N-VA also reaped other symbolic and relevant federal positions such as the Civil Service portfolio and the position of the Speaker of the only relevant federal assembly (the Chamber). ³⁶ The French-speaking Greens choose their ministers in a General Member Assembly (about 800 participants in 1999). ³⁷ Formally, the PM will formulate the minister’s resignation (and potentially his or her replacement by someone else) as a royal decree, to be signed by the monarch. ³⁸ Belgium uses gender quotas for the establishment of electoral lists since 1995 and gender parity since 2003. From the same year onwards, the federal Council of Ministers can no longer be unisex.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

104

   

Coalition agreements Whilst the allocation of portfolios typically takes place at the end of the negotiations, during a last long night among tired party leaders, the drafting of policy agreements is the lengthiest phase of government formation. It involves a wide number of negotiators spread over thematic working groups, with the formateur and party leaders being part of a central negotiating body. Coalition agreements are crucial in Belgian coalition governance and largely constrain ministerial discretion. Moury (2011) analyses the 1992 and 1999 governments and finds that close to 80 per cent of measures included in Belgian coalition agreements became cabinet decisions and that the majority of policies discussed in the Council of Ministers originated from this document. In the period under study those agreements kept on increasing in length, from a few dozens of pages in the 1990s to more than 200 pages for Michel I (Table 4.4). About all contents (on average more than 95 per cent) concern policy, the rest mainly specify procedural ‘rules of the game’ such as delegating authority or setting up new bodies to deal with specific issues. The dramatic increase of the formation duration of the post-electoral coalitions since 1999 is due to the increasing difficulties in finding a comprehensive policy compromise satisfying all coalition partners in times where the splitting of the country has become a ‘not desirable but not impossible’ outcome in case coalition negotiations between Flemish and French-speaking parties fail to produce the typical satisficing ‘Belgian compromise’. However, the saliency of the community issues—which are no longer only linguistic but increasingly pertain to economic and budgetary solidarity—is to some extent ‘behavioural’: bargaining parties can decide either to make these issues central in the policy programme, to fully discard them, or to only specify the procedures through which those matters have to be dealt with.³⁹ Verhofstadt I belongs in the latter category: none of the six participating parties wanted to invest lengths of time on these thorny issues. Therefore, the coalition agreement barely created an ‘intergovernmental and interparliamentary conference for institutional renewal’ as a permanent forum of concertation between state levels (with parity between Flemish and French-speaking members and dual chairmanship) to discuss, evaluate and reflect on a better functioning of the Belgian institutional

³⁹ For instance, the Dehaene I government (1992–1995) coalition agreement called for the establishment of ‘community to community’ dialogue, which led to the passing of a major constitutional reform. Note that both of the Dehaene coalition agreements constrained its components by requiring a consensus within the cabinet and in parliament for all matters that would not be specified in these—by Belgian standards—rather short documents (De Winter et al. 2000: 332). The policy areas at stake were mainly ethical issues for which the Flemish Christian Democrats wanted the status quo and avoid the constitution of alternative secular majorities in parliament.

1992 1995 1999 2003 2008 (date of Leterme I agreement) 2011 (date of Di Rupo government agreement) 2014 (year of Michel gvt agreement)

Dehaene I Dehaene II Verhofstadt I Verhofstadt III Leterme I

1 1 1 0.5 0 0

0

53,000²

57,100

General rules (in %)

7,500 17,350 14,800 26,500 14,900

Size

3

2.5

3 1 5 5 5

Policy-specific procedural rules (in %)¹

0

0

0 0 0 0 0

Distribution of offices (in %)

0.5

0

0 1 1.5 0.5 1

Distribution of competences (in %)

96.5

97.5

96 97 92.5 94 94

Policies (in %)

¹ ‘Usual’ procedures such as consultation or concertation with social partners or stakeholders, as well as concertation with subnational entities in case of shared/overlapping competences, are not coded as policy-specific procedural rules. ² About 40 per cent of the Di Rupo coalition agreement dealt with institutional reforms (including the BHV constituency issue) and was negotiated by eight parties rather than only the six parties of the coalition. Most of those reforms concern a transfer of competencies (and resources) from the federal to subnational levels. Following earlier coding (De Winter et al. 2000), those complex multilevel transfers were not coded as a new distribution of competences within the federal government. When policy areas/issues included in this part of the coalition agreement were left undecided but led to the specification of procedures, these were coded under policy specific procedural rules together with those that concern the second part of the coalition agreement (mainly socio-economic policies only committing the six parties of the government).

Notes:

Michel I

Di Rupo

Year in

Coalition

Table 4.4 Size and content of coalition agreements in Belgium, 1991–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

106

   

architecture.⁴⁰ Second, given its unprecedented composition, the Verhofstadt I coalition applied a different logic for drafting its policy agreement. Its predecessors were characterized by the (centrist) Christian Democratic technique of reaching a consensus on every policy issue to come to a balanced outcome for every policy area. As the rainbow coalition had to be formed quickly to avoid losing the momentum of excluding the Christian Democrats and because the dioxin crisis that developed just before the elections needed a quick response, the solution was to departmentalize the coalition agreement. It made sure that the fiscal and economic sections would be broadly dictated by the liberal manifestos, that the welfare state and employment policies mainly derived from socialist preferences, and that a green touch would be recognizable (for instance with the decision of a nuclear power phase-out), instead of trying to reach a difficult compromise on each of these policy areas and risking the breakdown of the formation attempt. The qualitative allocation of portfolios along the same lines further allowed the implementation of what could appear incoherent policies.⁴¹ This would allow each party family to make their policies identifiable by an electorate that could have been flabbergasted by the unconnected party composition of the coalition (Dumont 2011). As a result, by Belgian standards, both government formation duration and the coalition agreement were voluntarily short (less than a month, 14,800 words). The Verhofstadt III agreement was longer (26,500 words) and took more time to be drafted despite the fact that only four parties negotiated its terms. Without the Greens and only made of same-sized parties that had all progressed at the 2003 elections, the negotiators had to draft a balanced text making the difficult synthesis between left and right policies, a delicate mission that required more detailed guidelines for the future ministers. Once again, community issues were merely redirected to a specific concertation forum between federal and federated entities. In December 2007 the Verhofstadt IV interim government was not based on a coalition agreement but the declaration to the Chamber specified that state reform talks between eight parties (Octopus) would continue under the unsuccessful government formateur (and one of the two vice-PMs in charge of institutional reforms) Leterme ‘in parallel’ with negotiations to form a fullyfledged government. The Leterme I coalition agreement was as short as Verhofstadt IV (14,900 words) and only announced that it would present a second package of

⁴⁰ This ‘conference’ actually only had a limited input on the conclusion of a number of institutional agreements voted on during that term. Most agreements were instead initiated by the federal government and passed in parliament with the support of some opposition parties. ⁴¹ PM Verhofstadt fully assumed and justified those inconsistencies and the public clashes between partners that came with them as an ‘open culture of debate’ innovation compared to his predecessors’ mode of coalition functioning of keeping the lid on all intracabinet conflicts.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi



107

institutional reforms—without specifying them—a few months later. The following cabinets were still based on the same coalition agreement. Both Van Rompuy and Leterme II declarations to the Chamber only mentioned procedures to deal with the thorny Brussels-Halle-Vilvoorde (BHV) electoral and judiciary constituency issue. Contrary to the preceding three government agreements, a large part of Di Rupo’s was dedicated to state reform. The text started with the deal signed by eight parties (the six coalition partners and the two Green parties, two months before the cabinet was sworn in) on the sixth state reform that transferred another huge part of federal competences to the regional level. This section made for a good third of the whole coalition agreement. The rest of the text consisted of all the other policies agreed on by only the six government partners. Altogether the coalition agreement was 53,000 words long. Finally, the Michel I coalition formed in 2014 avoided community issues altogether in its coalition agreement. This was the core condition made by French-speaking liberal party leader Charles Michel to the dominant N-VA, in order to seal a deal that would leave the MR as the sole French-speaking representative in a coalition with the independentist party. The emphasis was instead placed on socio-economic centre-right policy cohesion (employment, pensions, finances) as well as on a more TAN-tone than its predecessors (justice, security, asylum, and immigration) allowed by this unprecedented coalition formula. The Flemish Christian Democrats left some social touches to a centre-right socioeconomic programme and managed to avoid mentions of advances in ethical issues in the agreement, conditioning any related future discussion to consensus among coalition parties. The fact that the three Flemish coalition partners compete for the same centre to centre-right electorate may explain why the text had to be so long (Swyngedouw et al. 2015), with each of them claiming to include some of their own subtle markers and trying to avoid leaving too much room for ministerial discretion. As a result, despite discarding community issues altogether, the Michel I coalition agreement was the longest ever drafted in Belgium’s political history (57,100 words).

Coalition governance Coalition governance in Belgium is very collective, yet hierarchical. The most prominent decision-makers are the PM and the vice-PMs of each party, solving most conflicts in an inner cabinet, the Kerncabinet (the cabinet ministériel restreint, IC in Table 4.5), whilst formal inter-ministerial cabinet committees (CaC in Table 4.5) have been abolished in 1993 and replaced by more flexible ad hoc ‘intercabinet working group meetings’.

Van Acker III Huysmans Spaak II Eyskens Van Acker IV Eyskens III Lefevre Harmel Van den Boeynants I Eyskens IV Eyskens V Leburton Tindemans Tindemans II Tindemans III Tindemans IV Van den Boeynants II Martens I Martens II Martens III Martens IV M Eyskens

Coalition

N N N N N IE POST POST IE

POST POST IE POST IE N POST N

POST N IE IE IE

1946 1946 1947 1949 1954 1958 1961 1965 1966

1968 1972 1973 1974 1974 1977 1977 1978

1979 1980 1980 1980 1981

y N/A y Y Y

y y y y y N/A y N/A

N/A N/A N/A N/A N/A n n y y

n n n N N

n n n n n n n n

N N n n n n n n n

IC, CaC, PS IC, CaC, PS IC, CaC, PS IC, CaC, PS IC, CaC, PS

IC, CaC, PS IC, CaC, PS IC, CaC, PS IC, CaC IC, CaC IC, CaC IC, CaC, PS IC, CaC, PS

CaC CaC CaC, PS CaC CaC, PS CaC, PS IC, CaC, PS IC, CaC, PS IC, CaC

All used

IC IC IC IC IC

IC IC IC CaC CaC CaC IC IC

CaC CaC CaC CaC CaC CaC IC IC CaC

Most common

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Table 4.5 Coalition governance mechanisms in Belgium, 1946–2018

PS PS PS PS PS

PS PS PS IC IC IC PS PS

CaC CaC PS CaC PS PS PS PS IC

For most serious conflicts

Personal union

Issues excluded from agenda

Most/No Most/No Most/No Most/No Most/No

Most/No Most/No Most/No Most/No Most/No Most/No Most/No Most/No

Most/No Most/No Most/No Most/No Most/No Most/No Most/No Most/No Most/No

y y y Y Y

y y y y y y y y

y y y y y y y y y

Comp. Comp. Comp. Comp. Comp.

Comp. Comp. Comp. Comp. Varied No Comp. Few

Few Few Few Few Few Varied Varied Varied Varied

y y y Y Y

y y y y y y y y

n n n n n y n y y

Y Y Y Y Y

Y Y Y Y Y Y Y Y

y y y y y y y y Y

Junior NonCoalition Freedom of Policy discipline appointment agreement ministers cabinet positions in legislation/ other parl. behaviour

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

2008 N

2009 N

2011 POST

2014 POST

Van Rompuy

Leterme II

Di Rupo

Michel I

Y

Y

N/A

N/A

Y

Y Y Y Y Y Y Y N/A Y N/A

N

N

N

N

N

N N

N N N N N N N

IC

IC

IC, PS

IC

IC, CaC IC, CaC IC, CaC, PS IC, CaC, PS IC, PS IC, PS IC, PS IC IC, PS transitional gvt IC, PS

IC

IC

IC, PS

IC

CaC CaC IC IC IC IC IC IC IC transitional gvt IC

IC

IC

PS

None

IC IC PS PS PS PS IC, BC, PS IC IC, PS transitional gvt IC

Most/No Most/No Most/No SP, PS Abortion Most/No SP None⁵¹ All/No SP None All/No G, PS, SP None All/No SP None All/No PS, SP None All/No transitional None No/No gvt CDV, VLD, None No/No PS CDV, VLD, None All/no PS CDV, VLD, None All/no PS CDV, VLD, None All/Spec. PS, MR NVA, Comm. All/No CDV, VLD Probl. Y

Y

Y

Y

Y

Y Y

Y Y Y Y Y Y Y

Comp.

Comp.

Few

Varied

Comp.

Comp. Varied

Comp. Comp. Comp. Few Comp. Comp. Comp.

N

Y

Y

Y

Y

Y Y Y Y N N Y N N N

Y

Y

Y

Y

Y

Y Y

Y Y Y Y Y Y Y

Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement; No = No explicit agreement

Coalition discipline: All = Discipline always expected; Most = Discipline expected except on explicitly exempted matters; Spec. = Discipline only expected on a few explicitly specified matters

Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee; CoC = Coalition committee

Coalition agreement: POST = Post-election; N = No coalition agreement

Notes: During periods where the values for the variables remain identical, the first and last applicable cabinets are listed. The last applicable cabinet is right-justified in the Coalition column.

2008 IE

Leterme I

POST POST POST POST POST POST POST N POST N

1981 1985 1988 1991 1992 1995 1999 2003 2003 2007

Martens V Martens VI Martens VII Martens VIII Dehaene I Dehaene II Verhofstadt I Verhofstadt II Verhofstadt III Verhofstadt IV

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

110

   

Coalition governance in the executive arena Since 2003, each coalition party has one vice-PM who is also in charge of a large department. Vice-PMs serve as the chef de file, the cabinet leader of their party’s ministers, and, with the other vice-PMs and the PM, constitute the Kerncabinet, which meets regularly (from several times a week to a few times a month) and resolves major conflicts between the coalition parties. Its final decisions are later on formally ratified by the full cabinet. For their cabinet leadership role, vice-PMs are equipped with a personal advisory staff for general policy (up to 70 staff members) whose main tasks are to follow the decision-making of the other ministers in the government and to safeguard the party’s interest. In addition to this advisory staff for general affairs, vice-PMs rely on a regular ministerial staff for each ministerial portfolio they hold; together they constitute the minister’s ‘strategic unit’, in the New Public Management parlance of the ‘Copernicus’ reform implemented in 2000. The latter was intended to drastically reduce the size of these personal staff units acting as a buffer between the minister and his civil service and which previously could count more than 200 personal collaborators for vice-PMs with one or two substantial portfolios. The Kern’s decisions are prepared by the General Policy Directors (the ‘DABs’, Directeurs Algemeen Beleid), that is the chefs de cabinet for General Policy of each vice-PM, who are meeting at least weekly for a four-hour meeting but often meet several times a week. Only the issues not solved by those ‘genies’ working in the shade of their vice-PMs are referred to the Kern for discussion rather than for mere ratification of pre-cooked decisions. The Kern thus takes all major contentious decisions, with maybe the partial exception of community issues. It is also the main arena for logrolling, sometimes leading to huge package deals across policy sectors, for instance when deviations to the coalition agreement have been denounced by a partner and compensations are called for or more generally when limited policy conflicts cannot be solved in a discrete manner. One of the crucial moments for logrolling is the preparation of the yearly general state budget and the bi-annual budget revisions.

The role of the PM and his staff in policy-making Although the cabinet is only made of 15 ministers, the latter come from an increasing number of different political parties, and thus veto players. In terms of staff resources, the PM, like his ministers, has always had his personal ‘strategic unit’ to serve him as government leader in party political terms (Brans et al. 2017).⁴² ⁴² In 2001, the old administrative Chancellery was transformed into the Federal Government Service ‘Chancellery & General Services’ (about 190 civil servants) to support, amongst others, the PM. It

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi



111

The PM has mainly agenda-setting powers. When ministers wish to add an item to the Council of Ministers agenda, they submit a request to the Council Secretary. The draft agenda is examined by the Council Secretary, purely from a budgetary, technical, and legal/administrative perspective. The draft agenda is then submitted to the PM for a more political examination. It is at that moment that the PM can decide to block or delay certain issues from being brought up for decision in the full cabinet on Friday. Apart from this negative agenda power, the PM has also positive agenda powers: he can insist that an individual minister take a policy initiative, especially if the matter was mentioned in the coalition agreement and other ministers insist on its timely implementation. In more exceptional circumstances the PM can suggest to the Kern that a minister be removed from some competencies to deal with thorny issues. The power of the PM is also linked to his central position in the government networks: the PM is the best-informed cabinet member. This is for example due to his role as compromise-maker; one of the most common ways at arriving at compromise over major issues are the ‘confessionals’: face-to-face bilateral conversations between the PM and a vice-PM. Based on these confessionals with each vice-PM, he can decide to formulate a compromise and submit it for discussion to the Kernkabinet. These rounds of confessionals with all vice-PMs can take weeks. When all finally agree, the compromise is sent to the Council of Ministers for simple ratification. Hence, the PM will never act without the full Kern’s consent. If the PM does not manage to create a consensus on an issue, the Kern can decide to put it on hold (‘into the freezer’), sometimes for the entire term of the government. Months can be used for delaying a compromise, by calling for the noncompulsory advice of for example the Council of State, the National Bank, or corporate actors.⁴³ If even the Kern does not manage to find a solution, the PM can decide to call a summit of all coalition party leaders (PS in Table 4.5). These summits are often called for solving community conflicts. Party summits can be a simple ad hoc meeting of PM and the party leaders, or they can take more complicated forms. For instance, in the Leterme II government, the cabinet decided to delegate the thorny BHV issue to former PM Jean-Luc Dehaene. The latter negotiated for five months with the leaders of the coalition parties plus those of the Green parties in the opposition, but to no avail (see Coalition duration and termination). In the Di Rupo government, the Comori (Comité de Mise en Oeuvre des Réformes Institutionnelles) was created to deal with the implementation of the actually allowed him to relocate a large part of his personal strategic staff in the context of the Copernicus reform. ⁴³ Such as various parliamentary constructions; mixed intergovernmental commissions (federal plus regional); delegation to ‘Group of Ten’, the main neo-corporatist body made of representatives of major trade unions and employers’ umbrella organizations, to a group of ‘Wise—former politicians—Men’ or external experts, etc.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

112

   

government’s institutional reforms. This quasi-permanent party summit included the PM, his two secretaries of state for institutional reform, and the six coalition party leaders as well as those of the Green parties. It met regularly for three years (December 2011–January 2014) transforming the state reform enshrined in the coalition agreement into a thousand pages of legislative bills. Some PMs are of course more proactive than others in the Kern and in full cabinet by proposing various solutions and personal views, while others adopt the role of ‘notary’, listen to all sides to a conflict, let the debate take place, and then sum up the discussion and suggest outlines of a possible compromise. Others evacuate conflicts in the Council altogether by asking the DABs to formulate decisions even on contentious matters. These differences in PM style depend on the latter’s personality but also on the delicate cohesion of the coalition and the degree of trust of the coalition partners in the PM. It also depends on the status of the PM within his party, as well as that of his vice-PMs in their own and their respective party’s strength in the coalition. In about every cabinet since 1999, roughly half of the party leaders that negotiated the formation did enter government, without clear patterns between parties. Note that regarding ‘personal union’, most party statutes require that when a party leader enters the cabinet, a new party leader is selected. However, as shown in Table 4.5, several party leaders were exempted from this rule.

The role of individual ministers in policy-making In addition to the iron straightjacket of the coalition policy agreement and the collective leadership of the Kern, there also exist a large number of formal rules in the ministers’ deontological code of behaviour that prevent Belgian ministers to become policy dictators. When forming a government, the new PM gives each member a collection of guidelines relating to ministerial ethics (Alen and Dujardin 1986: 532–4), the working of the government, and the working of the Council of Ministers. These rules stipulate that the Council of Ministers take decisions collectively and by consensus; it does not vote on them. The ministers debate the issue until they reach a consensus; they are then all equally and jointly accountable to the outside world for their decision. Discussions should only take place behind closed doors and demand utmost discretion. The PM ends the discussions when he formulates a consensus decision that the whole government can endorse. A member of the government cannot openly voice reservations about a decision that has been taken collectively. If no consensus is reached, the matter is put on hold and referred to a working group comprising representatives of the ministers’ policy-making bodies. Hence, if a single party or minister opposes a certain policy proposal of his colleague, the full cabinet may be forced to consider the matter and decide by consensus. Moreover, ministers may not make declarations about

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi



113

matters that do not belong to their colleagues’ competencies that could embarrass the latter, nor express a personal point of view if that would question the governmental agreement. Finally, at the beginning of a legislative year, each minister also has to present his ‘policy brief ’ (and departmental budget) announcing what he wants to realize that year. These briefs are scrutinized by all parties and the media for their conformity to the coalition agreement. Hence, Belgian ministers are not at all policy dictators in their respective jurisdictions (Moury 2013). In practice however, since the so-called ‘open debate culture’ introduced by Verhofstadt, these rules are increasingly violated due to new means (ICT) and needs (a context of growing electoral volatility) of individual and party communication. For instance, ministers increasingly announce policies in the media before they inform their colleagues and the PM (‘politics by announcement’), criticize openly the proposals of ministers of other parties, and, in the Michel government, sometimes openly contested the leadership of the PM by not fully informing him about their policy.

Hierarchy between cabinet members Next to the vice-PM positions the most important portfolios generally remain the core sovereign state functions ministries, such as finance, foreign affairs, and justice, as well as the main welfare state departments (social affairs and health). The minister for budget (which is a split-off from finance) can play a role as important as the one of vice-PM and usually is associated to the Kern for solutions on budget conflicts. Junior ministers have generally not played a watchdog role vis-à-vis their minister in the period under study. In all four-party coalitions (Dehaene I and II, Verhofstadt III, Michel I), cabinets had no watchdogs despite counting between one and six junior ministers, while the more exceptional and short-lived cabinets, Verhofstadt IV and Michel II, did not even appoint any of these. In the 2007 to 2010 five-party coalitions two out of seven junior ministers were adjunct to a full minister, a lower proportion than in the 1999 six-party coalition headed by Verhofstadt (two out of three) and Di Rupo (three out of six). Note that while some juniors are just mere additions to the portfolio holder with no specific attribution in their title and help address proportionality in office payoffs, some have identifiable minor competences, while others still may have highly important or visible ones. This was the case for Melchior Wathelet Jr (cdH), who was in charge of the Budget in the 2007–2010 governments,⁴⁴ whilst with several waves of ⁴⁴ Wathelet Jr was formally a watchdog for the PM but was most importantly, given the transversality of his remit, present at cabinet and Kern meetings, giving the small cdH a second representative in those bodies despite the party being only allowed to appoint one full minister.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

114

   

refugee crises in Europe, the recent state secretaries for Asylum and Migration have become highly visible—making them even more notorious and popular than most senior ministers.

Governance mechanisms in the parliamentary arena As expected in a multiparty partitocracy where cohesion at the governmental level is a crucial challenge, since the 1970s parliament has little impact on coalition governance in Belgium (De Winter and Dumont 2006; de Vet 2019). Every Thursday morning each of the coalition parties holds a parliamentary party group (PPG) meeting, which starts with an exposé of the PM or vice-PM explaining the current governmental agenda and discussions (De Winter and Dumont 2000; de Vet 2019). Even when the PPG leaders are not part of the informal summit of their party, they usually do play some role in coordinating detailed positions and amendments in parliament. The meetings of the PPG leaders of the coalition parties have become more frequent in the latest decade, but they mostly focus on technical or procedural adaptations (De Winter and Wolfs 2017).⁴⁵ Coalition parties hardly keep tabs on partners through the allocation of committee chairs: although the latter are allotted proportionally, the party identity of the chair does not seem to play an important watchdog role in coalition governance (Chiru and De Winter 2021) and parliamentary rules dictate that government initiatives always have agenda priority over private member bills. Majority MPs usually also refrain from embarrassing other parties’ ministers through aggressive questions in parliament. As expected these MPs not only ask less questions but are also more constrained in voicing their party manifesto’s priorities or currently mediatized issues than those who belong to opposition parties (Dandoy 2011; Vliegenhart et al. 2013). Note that sometimes coalition agreements may leave some room to ministers and majority MPs, which open avenues for parliamentary hearings on specific policy matters. Finally, long government formations have surely led the parliament to regain some autonomy since 2007. But the latter is however only temporary, confined to those periods where the absence of a fully empowered executive and no clear indications on what the next government’s partisan composition would look like, allowing for unconstrained (but still partydisciplined) parliamentary voting alliances (Van Aelst and Louwerse 2014).⁴⁶ ⁴⁵ Often those meetings are the consequence of governmental legislative projects made in a rush. Hence majority PPGs introduce amendments in order to technically improve governmental work. Although those technical interventions are thus considered useful, they can sometimes lead to further difficult discussions. ⁴⁶ Note that formally a caretaker government cannot make any major decisions and must only attend to ‘current and urgent matters’, using for instance the same budget as its predecessor (in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi



115

Cabinet duration and termination In comparison with the cabinet instability of the 1972–1981 period (13 governments in less than 10 years), Belgium mostly had long-lasting governments ever since. The main reason for this stability was initially the gigantic economic challenges to be met, which had to be tackled by dedicated ‘socio-economic’ governments through two devaluations, special executive powers, several austerity plans, and since 1991 major budgetary cuts in order to meet the Maastricht convergence norms for entering the euro. But cabinet duration remained high even under better economic times, with governments often completing their full theoretical legislative term. Table 4.6 summarizes the main and explicit reasons for cabinet termination. Contrary to the pre-1999 period, the cases for which governments fell short of 90 per cent of their expected duration ever since were overwhelmingly mere changes of PM and did not involve alterations of partisan composition. From 2007 to 2010 we indeed record a previously agreed-upon PM change from Verhofstadt IV to Leterme I, the forced resignation of Leterme and some his ministers due to the 2008 Fortis bank crisis, the replacement of his successor, acting Chamber Speaker Van Rompuy, who after one year as PM was nominated as the president of the European Council, and finally the replacement of the latter by Leterme having been cleared for his role in the Fortis debacle. That cabinet however ended when the issue of the splitting of the electoral and judiciary BHV constituency, which had emerged as early as in 1961, resurfaced as a result of institutional reforms in the 1999–2003 legislature but had not been solved ever since. Former PM Dehaene was asked to fix the issue as ‘Royal Assignment Holder’ but after his endless discussions with the leaders of the majority parties as well as the opposition Green parties, the Flemish liberals decided to leave those stalled negotiations but also the federal government on 22 April 2010, triggering elections a year before the end of the legislative term.⁴⁷ Conflicts between one coalition member and its partners over policy led to the more classical downfall of two governments since 1999. On both occasions the term was nearly over and the rest of the coalition continued until elections

monthly slices, the ‘provisional twelfths’). However, the extremely long caretaker Leterme II has led to more flexibility in this regard, as critical events may push the temporary government to take action, contingent upon parliamentary approval. That caretaker government decided to send F16 fighter planes to Libya, nominated a new National Bank governor, presented a budget for 2011 to the parliament, and had it voted on. ⁴⁷ In that legislative term this issue had already led to the release of several consociational instruments by the French-speaking subnational entities (conflicts of interest) and federal MPs (the alarm bell procedure, see Swenden 2005) against the potential unilateral vote by a Flemish majority in the federal parliament on the splitting of the constituency.

Date in

1992-03-07

1995-06-23 1999-07-12

2003-05-05 2003-07-12 2007-12-21

2008-03-20

2008-12-30

Cabinet

Dehaene I

Dehaene II Verhofstadt I

Verhofstadt II Verhofstadt III Verhofstadt IV

Leterme I

Van Rompuy

2009-11-25

2008-12-22

2003-05-18 2007-06-10 2008-03-20

1999-06-13 2003-05-05

1995-05-21

Date out

37

23.5

100 100 7.1

100 97.3

86.2

Relative duration (%)

Table 4.6 Cabinet termination in Belgium, 1991–2018

9

9

1 1 9

1 7a

4

Mechanisms of cabinet termination

13

Terminal events

E

Parties

PM, Finance, Justice

Transports

Policy area(s)

Expectation of good electoral result for outgoing coalition End of term The two ECOLO Party ministers left the government over disagreements about night flights over Brussels. End of term End of term Verhofstadt, who had headed a transitional cabinet since the July 2007 general election, resigned when a new government could be formed. Resignation due to a long-running scandal of government interference with justice in the sale of Fortis bank (which had just been bailed out by the government). The minister of justice stepped down and on the proposal of the PM the rest of the cabinet followed. Van Rompuy resigned after being elected president of the European Council.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

2011-12-06 2014-10-11

2018-12-09

Di Rupo Michel I

Michel II

2018-12-21

2014-05-25 2018-12-09

2010-04-26

7.1

100 90

27

9

1 7a

7a

NVA

VLD

Migration

Institutional reforms

VLD left the cabinet after accusing the Francophone cabinet members of blocking attempts to divide the BHV electoral/judicial district along linguistic lines. End of term NVA wanted to force PM (and cabinet) to renege on promise made to the United Nations that Belgium would ratify the Global Compact for Safe, Orderly and Regular Migration. PM Michel could not find support in parliament for his minority government. Socialists and Greens announced they would introduce a non-confidence motion but PM did not wait and offered his resignation to the king.

10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event

Terminal events

4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

Discretionary terminations

1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of PM

Technical terminology

2009-11-25

Leterme II

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

118

   

were held. In 2018 the departure of the N-VA,⁴⁸ the largest party of the coalition made the government lose its majority.⁴⁹ For several days PM Michel tried to find partners in the opposition who would support his accommodating government agenda but the introduction of a no-confidence motion by the socialist party family on 18 December led him to pre-empt such a vote by going to the king to tender the resignation of his cabinet on the same day. The king would eventually accept it on 21 December. Five months of a minority-based caretaker cabinet would ensue until the due May 2019 elections. The previous case for which the loss of a partner was followed by a short-lived cabinet made of remaining partners occurred in 2003, when the French-speaking Greens were pushed out of office a week before the scheduled elections.

Conclusion Although since 1999 there were no really new parties in the formation process, nor any new cleavage, the formation dynamics changed dramatically due to the declining size of the mainstream parties around which post-war coalitions were formed, the growing saliency and Flemish radicalization of the community cleavage, and the market leadership of the independentist N-VA, whose positions remain unacceptable for any French-speaking party. The two party systems also increasingly drifted apart in terms of the relative strength of left and GAL parties versus right and TAN ones. More important, party system fragmentation skyrocketed, confronting some (in)formateurs with up to ten office-seeking and coalitionable parties. This party system fragmentation and polarization along the linguistic divide reinforced and fine-tuned Belgium’s interpretation of the coalition compromise model, which was already solidly entrenched in the consociational norms and practices since the 1960s (De Winter et al. 2000). The four to six coalition partners keep tabs on each other through elaborated inter-party negotiation and compromise mechanisms: very long comprehensive and detailed coalition agreements drafted during record long formation with numerous bargaining failures. Coalition parties have at their disposal a wide arsenal of inter-party negotiation and compromise mechanisms as well as policy monitoring devices. The inner cabinet (Kern), composed of the PM and the vice-

⁴⁸ Its predecessor, the regionalist VU, also left the government in 1991 but at that time it was the smallest partner in a government relying on a two-thirds majority. ⁴⁹ All four parties had just lost ground at the October local elections. Given the VB’s politization of the UN World Pact on migrations to be ratified by the Belgian government and the refusal of, among others, the right-wing Austrian government to sign it, the N-VA turned the issue of the participation of the PM to a summit in Marrakech intended as a confirmation of participants’ commitment into a question of government survival.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi



119

PMs of each party, serves as the supreme gremium that takes binding cabinet decisions by consensus. The PM remains only a primus inter pares, whose influence lies in his capacity to pre-cook the Kern’s decisions. Hence, regardless of the comparative size of coalition partners, no policy decision is in principle taken without the support of each of them. In addition, coalition discipline in legislative matters is iron-strong and often spills over into other forms of parliamentary behaviour. It is not just that the Belgian formation puzzles have become ever more difficult to solve, as the record formation durations indicate. More dramatically, in times where the splitting of the country has become a ‘not desirable but not impossible’ outcome, the deadlock created by repeated formation failures of coalition negotiations due to unbridgeable divides between main Flemish and French-speaking parties could trigger the demise of Belgium, often nicknamed a ‘failed’ state (Higgins 2015; Devos 2016).

References Alen, André, and Jean Dujardin (1986). Caseboek Belgisch grondwettelijk recht. Gent: Story and Scientia. Andeweg, Rudy B., Lieven De Winter, and Wolfgang C. Müller (2008). ‘Parliamentary Opposition in Post-consociational Democracies: Austria, Belgium and the Netherlands’. Journal of Legislative Studies, 14(1–2): 77–111. André, Audrey, Sam Depauw, and Kris Deschouwer (2015). ‘Changing Investiture Rules in Belgium’. In Bjørn Erik Rasch, Shane Martin, and José Antonio Cheibub (eds), Parliaments and Government Formation: Unpacking Investiture Rules. Oxford: Oxford University Press, 49–66. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010.’ Party Politics, 21(1): 143–52. Brans, Marleen, Christian de Visscher, Athanassios Gouglas, and Sylke Jaspers (2017). ‘Political Control and Bureaucratic Expertise: Policy Analysis by Ministerial Cabinet Members’. In Marleen Brans and David Aubin (eds), Policy Analysis in Belgium. Bristol: Policy Press, 57–76. Chiru, Mihail, and Lieven De Winter (2021). ‘The Allocation of Committee Chairs and the Oversight of Coalition Cabinets in Belgium’. Government and Opposition (forthcoming). Dandoy, Régis (2011). ‘Parliamentary Questions in Belgium: Testing for Party Discipline’. The Journal of Legislative Studies, 17(3): 315-326. Dandoy Régis, Jérémy Dodeigne, Min Reuchamps, and Audrey Vandeleene (2015). ‘The New Belgian Senate. A (Dis)Continued Evolution of Federalism in Belgium?’. Representation, 51(3): 327–39.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

120

   

Dassonneville, Ruth, and Pierre Baudewyns (2014). ‘Élections de Mai 2014, Signe de Volatilité Extrême? Une Analyse Des Transferts de Voix Lors Des Élections Du 25 Mai 2014’. Courrier Hebdomadaire du CRISP, (2225): 10–19. Deschouwer, Kris (2009). ‘Coalition Formation and Congruence in a Multi-layered Setting: Belgium 1995–2008’. Regional and Federal Studies, 19(1): 13–35. Deschouwer, Kris (2012). The Politics of Belgium. Basingstoke: Palgrave Macmillan. de Vet, Benjamin (2019). Between Party and Parliament. The Roles of Parliamentary Group Leaders in ‘Partitocratic’ Belgium. PhD dissertation, University of Ghent. Devos, Carl (2016). De failed state. Antwerpen: Van Halewijck. De Winter, Lieven (1995). ‘The Role of Parliament in Government Formation and Resignation’. In Herbert Döring (ed.), Parliaments and Majority Rule in Western Europe. New York: St. Martin’s, 115–51. De Winter, Lieven (2006). ‘Multi-level Party Competition and Coordination in Belgium’. In Dan Hough and Charlie Jeffery (eds), Devolution and Electoral Politics. Manchester: Manchester University Press, 119–39. De Winter, Lieven, and Patrick Dumont (2000). ‘PPGs in Belgium: Subjects of Partitocratic Dominion’. In Knut Heidar and Ruud Koole (eds), Behind Closed Doors: Parliamentary Party Groups in European Democracies. London: Routledge: 106–29. De Winter, Lieven, and Patrick Dumont (2003). ‘Belgium: Delegation and Accountability under Partitocratic Rule’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press, 253–80. De Winter, Lieven, and Patrick Dumont (2006). ‘Do Belgian Parties Undermine the Democratic Chain of Delegation?’. West European Politics, 29(5): 957–76. De Winter, Lieven, and Wouter Wolfs (2017). ‘Policy Analysis in the Belgian Legislatures: The Marginal Role of a Structurally Weak Parliament in a Partitocracy with No Scientific and Political Tradition of Policy Analysis’. In Marleen Brans and David Aubin (eds), Policy Analysis in Belgium. Bristol: Policy Press, 129–50. De Winter, Lieven, Arco Timmermans, and Patrick Dumont (2000). ‘Belgium: On Government Agreements, Evangelists, Followers and Heretics’. In Kaare Strøm and Wolfgang C. Müller (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 300–55. Dumont, Patrick (2011). ‘The Belgian “Rainbow coalition” puzzle’. In Rudy B. Andeweg, Lieven De Winter, and Patrick Dumont (eds), Puzzles of Government Formation. Coalition Theory and Deviant Cases. London: Routledge, 183–207. Dumont, Patrick, Lieven De Winter, and Rudy Andeweg (2011). ‘From Coalition Theory to Coalition Puzzles’. In Rudy B. Andeweg, Lieven De Winter, and Patrick Dumont (eds), Puzzles of Government Formation. Coalition Theory and Deviant Cases. London: Routledge, 1–23.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi



121

Dumont, Patrick, Stefaan Fiers, and Régis Dandoy (2009). ‘Belgium: Ups and Downs of Ministerial Careers in a Partitocratic Federal State’. In Keith Dowding and Patrick Dumont (eds), The Selection of Ministers in Europe: Hiring and Firing. London: Routledge, 125–46. Fiers, Stefaan (1998). Partijvoorzitters in België, of ’Le parti c’est moi? Rolverwachtingen, rolpercepties, en rolgedrag van de voorzitters van de politieke partijen in België in de periode 1981–1996. Leuven: KUL—Afdeling Politologie. Higgins, Andrew (2015). ‘Belgium: World’s Wealthiest Failed State? Terrorism Response Puts Belgium in a Harsh Light’. New York Times, 24 November. Van Aelst, Peter and Tom Louwerse (2014). ‘Parliament without Government: The Belgian Parliament and the Government Formation Processes of 2007–2011’. West European Politics, 37(3): 475–496 Maddens, Bart, Jef Smulders, and Wouter Wolfs (2019). De prijs van politiek: Over de portefeuille van de partijen. Tielt: LannooCampus. Moury, Catherine (2011). ‘Coalition Agreement and Party Mandate: How Coalition Agreements Constrain the Ministers’. Party Politics, 17(3): 385–404. Moury, Catherine (2013). Coalition Government and Party Mandate: How Coalition Agreements Constrain Ministerial Action. London: Routledge. Müller, Wolfgang C., and Kaare Strøm (2000) (eds.) Coalition Governments in Western Europe. Oxford: Oxford University Press. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of Antielitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Sägesser, Caroline, and Cédric Istasse (2014). ‘Le Sénat et ses réformes successives’. Courrier hebdomadaire du CRISP, 14: 5–115. Swenden, Wilfried (2005). ‘What–If Anything–Can the European Union Learn from Belgian Federalism and Vice Versa?’. Regional and Federal Studies, 15(2): 187–204. Swyngedouw, Marc, Koenraad Abts, Sharon Baute, Jolien Galle, and Bart Meuleman (2015). Het communautaire in de verkiezingen van 25 mei 2014: analyse op basis van de postelectorale verkiezingsonderzoeken 1991–2014. Leuven: Onderzoeksverslag Centrum voor Sociologisch Onderzoek, Instituut voor Sociaal en Politiek Opinieonderzoek. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (2003) (eds.) Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Vliegenhart, Rens, Stefaan Walgrave and Brandon Zicha (2013). ‘How preferences, information and institutions interactively drive agenda-setting: Questions in the Belgian parliament, 1993–2000, European Journal of Political Research, 52: 390–418.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

122

   

Willocq, Simon, and Camille Kelbel (2018). ‘Splitting Votes, Splitting Hairs? Rationale for Split-Ticket Voting at the Federal, Regional, and European Elections of May 2014 in Belgium’. Publius: The Journal of Federalism, 48(4): 664–85.

Appendix. List of political parties Abbreviation PCB–KPB PTB-PvdA PSB–BSP SP

PS G

E FDF

RW CVP–PSC CDV

CDH

VU Spirit

Name Belgian Communist Party (Parti Communiste Belge–Belgische Communistische Partij) Workers’ Party of Belgium (Parti du Travail de Belgique-Partij van de Arbeid van België) Belgian Socialist Party (Parti Socialiste Belge–Belgische Socialistische Partij) Socialist Party Differently (sp.a: socialistische partij.anders, 2001–) Socialist Party (SP: Socialistische Partij, 1980–2001) Belgian Socialist Party (BSP: Belgische Socialistische Partij, 1978–1980) Socialist Party (PS: Parti Socialiste) Green (Groen, 2012–) Groen! (Green!, 2003–2012) AGALEV (Anders Gaan Leven, 1979–2003) Ecolo (Ecolo, Ecologistes Confédérés pour l’Organisation de Luttes Originales) DéFI (Démocrate Fédéraliste Indépendant, 2015–) Francophone Democratic Federalists (FDF: Fédéralistes Démocrates Francophones, 2010–2015) Democratic Front of Francophones (FDF: Front Démocratique des Francophones, 1964–2010) Walloon Rally (Rassemblement wallon) Christian Social Party–Christian People’s Party (Parti Social Chrétien–Christelijke Volkspartij) Christian Democratic and Flemish (CD&V: ChristenDemocratisch & Vlaams, 2001–) Christian People’s Party (CVP: Christelijke Volkspartij 1972–2001) Humanist Democratic Centre (cdH: centre démocrate Humaniste, 2002–) Christian Social Party (PSC: Parti Social Chrétien, 1972–2002) People’s Union (Volksunie, 1954–2001) Humanist (Spirit: Sociaal, Progressief, Internationaal, Regionalistisch, Integraal-democratisch en Toekomstgericht. 2002–2008; for a short time period Spirit was called SociaalLiberale Partij, 2008–2009 before merging with Groen!)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 1/7/2021, SPi

 NVA LP–PL

VLD

MR

FN VB UDRT

LDD

123

New Flemish Alliance (N-VA: Nieuw-Vlaamse Alliantie, 2001–) Party for Freedom and Progress (Parti de la Liberté et du Progrès–Partij voor Vrijheid en Vooruitgang, 1961–1992) Liberal Party (Parti libéral–Liberale Partij, 1846–1961) Flemish Liberals and Democrats (Open VLD: Open Vlaamse Liberalen en Democraten, 2007–) Flemish Liberals and Democrats (VLD: Vlaamse Liberalen en Democraten, 1992–2007) Reformist Movement (MR: Mouvement Réformateur, 2002–) Liberal Reform Party (PRL: Parti Réformateur Libéral, 1972–2002) National Front (FN: Front National) Flemish Interest (VB: Vlaams Belang, 2004–) Flemish Bloc (VB: Vlaams Blok 1978/79–2004) Democratic union for the Respect of Labour-Respect for Labour and Democracy (Union Démocratique pour le Respect du Travail) Libertarian, Direct, Democratic (Libertair, Direct, Democratisch, 2011–) List Dedecker (Lijst Dedecker, 2007–2011)

Note: Party acronyms are first given in abridged version as appearing in the tables, then their full name in English, followed in parentheses by the party full acronym as used in the text and name in the party’s native language. Bakker, Ryan, Catherine De Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Anna Vachudova. (2015). ‘Measuring Party Positions in Europe The Chapel Hill Expert Survey Trend File, 1999–2010’. Party Politics, 21(1): 143–152. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova and Marko Zilovic. (2017). ‘Explaining the Salience of Anti-elitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey data’. Research & Politics (January-March): 1–9.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 5 Denmark: How to Form and Govern Minority Coalitions Flemming J. Christiansen

At the surface, coalition building has not changed much in Denmark since the 1980s and 1990s (periods analysed in Damgaard 2000). The constitution from 1953 remains unamended, setting up a unicameral parliamentary democracy with proportional representation. The procedure for government formation has not changed, the electoral threshold is still at two per cent, party fragmentation remains high, and for more than a century no single party has held a majority on its own. Denmark maintains a record-high share of minority government among parliamentary democracies, most of them minority coalitions. Four old parties formed more than one century ago still dominate government formation and present the prime ministers. Coalition negotiations remain swift, the number of cabinet members is stable around 20, and unlike many other countries, Denmark still has no junior ministers. Danish governments remain ideologically coherent within their own ‘blue’ (centre-right) or ‘red’ (centre-left) bloc. With no overall majority in parliament for the governments, they deal with various forms of ‘minority governance’ that involve ad hoc bargaining with opposition parties, as external support parties (Damgaard 1969), and through ‘legislative accommodations’ that are long-lasting agreements about policies (Christiansen and Pedersen 2014). Below the surface, important changes have taken place. Liberal prime ministers have become more frequent than Social Democrats who held the positions for much of the twentieth century (Christiansen 2020). The Danish People’s Party has risen as a political actor influencing government formation and the majority building necessary for minority government to pass policies in parliament. Meanwhile, the Social Liberals no longer hold a pivotal role. The old parties have lost some electoral support, and two parties on the political wings, the Socialist People’s Party and Liberal Alliance, have achieved cabinet seats for the first time. The chapter provides new insights about coalition governance supported by five interviews with high-ranking former ministers. Coalition agreements feature much more prominently for coalition governance than before the 1990s (cf. Damgaard 2000: 261), and the government has developed much stronger, Flemming Juul Christiansen, Denmark: How to Form and Govern Minority Coalitions In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0005

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



125

hierarchical, and centralized arrangement for coordination and control in the form of inner cabinets that make most important cabinet decisions. From 1990 onwards, general elections were no longer called due to political crisis, and although there have been terminations due to disagreements among coalition parties, governments have remained in office for a longer time than previously. This means that minority governments, even in coalition, now show more stability than before 1990. The purpose of this chapter is to analyse and characterize these developments in Danish coalition formations and governance. It takes into account the institutional setting, the party system and its actors, government formation, coalition governance, and cabinet duration and termination. It includes five interviews with former cabinet ministers.

Institutional setting Danish parliamentary democracy developed from protracted political and constitutional struggles in the nineteenth century that involved the rise of nationalism transformed by defeat in war and divisions between rural and urban interests, between estate owners and farmers, and finally also between industrialists and workers. The party system came to reflect these divisions. The constitution of 1849 replaced absolutism and limited the power of the king by introducing a ‘Madisionian’ separation of powers in legislative, executive, and judicial branches (Damgaard 2011: 67). The king could now only make decisions countersigned by a minister. Yet, he was still free to choose his own ministers, and parliamentary government did not become a practice until 1901. Since the introduction of universal and equal suffrage to parliament in the constitutional revision of 1915, Denmark qualifies for the term ‘polyarchy’ in the sense of Dahl (1971). The most recent constitutional revision of 1953 abolished the upper chamber and codified parliamentary government (Skjæveland 2019). The unicameral Danish parliament, Folketinget, consists of 179 members, 175 elected in Denmark proper and 2 each on Greenland and the Faroe Islands—autonomous parts of the kingdom. The constitution guarantees a proportional electoral system. The electoral law puts the threshold at two per cent, which is comparatively low. Denmark provides an example of negative parliamentary government since a government could remain in office as long as no majority topples it. In the case of a passed motion of no confidence against the prime minister, he or she should either call an election or resign together with the other cabinet ministers. Only on three historical occasions did parliament pass a motion of no confidence (most recently in 1975), Twice (most recently in 1983), the governments resigned when its annual budget bill got defeated (Christiansen 2012). In contrast to Sweden (see Chapter 17, this volume), a vote of no confidence passes when a simple majority

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

126



supports it. In 1975, 86 votes against 85 passed the no confidence motion, with 5 neither for nor against (i.e, abstaining) and 3 absent (Olesen 2017: 200). The logic of parliamentary government described for the Danish case unifies executive and legislative powers more in the tradition of Westminster, and not Madisonian, democracy (Damgaard 2011). Yet, as noted in the introduction, the parties forming most Danish cabinets do not control a majority in parliament. Indeed, since 1971, only one majority government was in office. This was between 1993 and 1994 (Skjæveland 2003). Furthermore, no single party has by itself ever managed to hold a majority, neither in the unicameral system since 1953 nor in both chambers of the bicameral parliament prior to that year. In that respect, a certain separation of power between government and parliament remains in place. There are also other constraints on Danish parliamentary government (Damgaard 2011: 68). These include referendums and powers of courts and the European Union. The constitution mandates a referendum when handing over sovereignty to the European Union (unless there is five-sixths majority in the Folketing), and since 1972 eight such referendums have taken place. Parliament has also other various tools for control of the government. A right for minorities (one-third of parliament) to call a referendum was only used once historically (in 1963). Ultimately, parliament may indict a minister or a former minister for the High Court of the Realm for maladministration and breaking the law. This happens rarely, with only one recent conviction from 1995 (and a new case raised in 2021). The constitution also allows votes of lack of confidence for individual ministers. Such a vote has never passed but the threat has led to resignations or replacements of ministers in a number of cases. A counterbalance is the power of the prime minister who can call an early election almost any time, or threaten to do so (Becher and Christiansen 2015). The only limitation found in the constitution was added in 1953, namely that an incoming government could not call an election without having presented itself for parliament. This is to prevent a government from forming that does not respect the ‘reflex effect’ (explained later) and then call an election before parliament could act.¹ Yet, no investiture vote takes place in parliament when a government presents itself, although any member of parliament (MP) could raise a motion of no confidence. Under normal circumstances, a caretaker government (one that has actually resigned but remains in office while a new one is being formed) is also not entitled to call an election; it should have called the election in the first place rather than resigning. Constitutional scholars argue that if a government formation drags on, with no cabinet likely to form, it may fall within the necessary tasks of the caretaker government to call an election (Germer 1988: 32). Such an event has never taken place.

¹ Such a ‘state coup’ almost took place during the ‘Easter crisis’ in 1920 (Kaarsted 1968).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



127

The specific rules surrounding government formation and termination Denmark can establish a new government in two different formal constitutional manners. One is when a resignation or defeat of a government triggers a new formation procedure with the political parties giving advice to the head of state in ‘rounds’. Alternatively, the current government becomes a new government without resigning since the prime minister can argue that the change will not be met with a majority against it in parliament at its presentation. This could happen after an election or between elections. A new government formed like this can increase or decrease the number of parties in cabinet, and it can even change the prime minister. A more detailed account of these two major forms follows. When a government has been defeated in parliament but not called an election, or it has resigned—following an electoral defeat or resigning for a political reason—the constitution states that it continues in office until a new government takes over. Such a caretaker government ‘shall perform only what may be necessary to ensure the uninterrupted conduct of official business’ (cited from Art. 15 of the constitution). A government formation procedure then begins. According to leading Danish constitutional scholars, a ‘reflex principle’ exists from the constitution, meaning that no government may be formed if it could be expected to be defeated in the Folketing right at its presentation (Jensen 1989: 81, Christensen 2019: 91). The written constitution is almost silent about the particular procedure for government formation. Article 14 states that the king appoints the prime minister and the other ministers. This is no longer a personal power held by the hereditary head of state, since 1972, a queen, Margrethe II. First, the constitution states that the ministers have the responsibility for the conduct of a government (Art. 13), not the monarch personally. Government formation is no exception to this rule (Christensen 2019: 93–4). Germer (1988: 31), supporting himself on other scholars, considers older diverging views as ‘outdated’. This means that the outgoing prime minister has the responsibility for the government formation process, while the role of the head of state is symbolical. The new prime minister signs his or her own appointment and the dismissal of the previous prime minister. Actions by a prime minister—outgoing or new—need to respect the reflex principle to be constitutionally legal (Christensen 2019: 89). In 1975, there was discussion as to whether the caretaker prime minister in one of the early formation rounds interpreted advices more cautiously than he should have done (Olesen 2017: 203–4). The cabinet secretary—the legal advisor of the head of state, trained in constitutional law—could warn against possible transgressions. Yet it remains the sole responsibility of the prime minister to listen to this advice or not. In the final round of negotiations in 1975, the cabinet secretary told the leader of the negotiations to be on the safe side and ask the other parties beforehand about whether his proposed new government would be met with a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

128



majority against it. The incoming prime minister was ‘un-bending’, insisting that the other parties should themselves stand forward if they resisted, and consequently he formed Jørgensen II without following this advice (Olesen 2017: 207). On a number of occasions after 1901, the king still personally tried to influence government formation within the limitations of parliamentary government. Later, a constitutional practice—introduced for the first time in 1909 and regularly used since the 1920s—developed to find out whom to appoint prime minister out of the possible candidates. The monarch receives recommendations on the government formation from representatives of the political parties. The caretaker prime minister interprets the advice of such a ‘queen’s round’. Based on careful study of practice, Knudsen and Rasmussen (2012) finds that these rounds end with the appointment of a ‘negotiation leader’ with a mandate that could be more or less limited with regard to forming a government or to seeking information about the possibilities to form a government. This means that the distinction in the literature between an informateur—searching for a government but not seeking to form one him or herself—and the designation of a formateur attempting to form a government is often less clear in Danish practice (cf. Damgaard 2000: 241; Christensen 2019: 96–8). There could be more queen’s rounds if necessary. Kaarsted (1988) and Damgaard (2000) argue that if the advice unambiguously points to a majority government this should be preferred, and the same goes for a minority government with clear support from external parties. Otherwise, the two authors agree that the minority government most likely to survive should be appointed. In practice, this implies that the advice given by the minority representing the largest number of seats gets the chance to search for a government first. Only if this is without success, and another round shows a similar largest minority, the baton may go to the second largest minority, something that happened in the complicated government formation of 1975 (Kaarsted 1988: 109–11; Olesen 2017: 207). The arguments of Kaarsted (1988) and Damgaard (2000) must primarily concern the order in which negotiation leaders would get a chance. If a government proposed by a duly appointed negotiation leader respects the reflex effect, it would be constitutionally legal to appoint such a government, even if one could imagine a government based on more seats (cf. Christensen 2019: 92–3). Christensen (2019: 93) further argues that the caretaker prime minister during the government formation process should not question the assessments made by the appointed negotiation leader unless they give reason to ‘serious doubt’. Governments usually stay in office during a general election, and could remain so afterwards as long as there is no majority against it. In official terms (different from the ones applied in this volume), Denmark historically had the same government over elections (1982–1987 despite elections in 1984 as the most recent example, cf. Olesen 2018: 21–155). A government would usually resign soon after an election when it is clear that there is a majority against it. Otherwise,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



129

parliament would have to topple it when it assembles. It would be constitutionally legal for a government not to resign, and considered legitimate if there is doubt about whether it has a majority or not after an election. That happened after the election in 1975 when the incumbent government had won many seats but as described earlier got narrowly defeated at its presentation for parliament. After the elections of 1998, 2005, and 2007, ‘new’ governments formed consisting of the same group of parties, and without a queen’s round, yet after some days of bargaining, and usually also with some shifts in cabinet posts. This reflects what appears to be a revision of official constitutional-legal practice, meaning that a government in Danish politics (more in line with our project terms) is now considered new after an election. A government with a new constellation of parties would always be considered as a new government in legal terms even if it has the same prime minister as the previous government (Kaarsted 1988: 18). A new government may be appointed without a queen’s round in a number of situations with respect to the reflex effect as common denominator. One example is when cabinet parties decide to leave the government but this does not alter the parliamentary basis for the government in a manner that would question whether it has a majority in parliament. This happened in 2014 when the Socialist People’s Party resigned from office and in 1996 when the Centre Democrats did. In both cases, the outgoing parties still declared to support the survival of the government; new governments formed after a few days without a queen’s round. Following the election of 1990 that led the Social Liberals to resign from a coalition government with Conservatives and Liberals, it still declared not to topple a government of the two parties left, and such a government formed (Bille 1998: 97). Similarly, after the election in 1994— when the Christian Democrats did not get re-elected to parliament—three parties continued in office, now as a ‘new’ minority government. Otherwise, if the resignation of cabinet parties puts the survival of the new remnant government into question, it should resign or call an election (Kaarsted 1988: 17–18). When the Social Democratic–Liberal coalition government broke up in 1979, both parties supported that the prime minister called an election (Olesen 2017: 391). A prime minister can also increase the number of parties in cabinet as long as he or she has no reason to expect parliament to pass a motion against such a new government at its presentation. This happened in 2016 when the single-party Liberal government was enlarged with Liberal Alliance and the Conservative People’s Party as coalition partners and became a new three-party government. A prime minister in office may resign for reasons (personal, including health or death, or political) that do not affect the party constellation behind the government. In that case, a new prime minister supported by those same parties could be appointed without a queen’s round. The most recent example was in 2009 when Anders Fogh Rasmussen resigned to become general secretary of NATO, and the deputy of the Liberals, and minister of finance, Lars Løkke Rasmussen, became the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

130



new prime minister. Yet, when Schlüter resigned in 1993 after a political scandal and without calling an election, the Social Liberals declared that it would not support a new prime minister from one of the cabinet parties. Hence, such a person could not present him- or herself for parliament without risking defeat. Instead, this resignation triggered a ‘queen’s round’, ending in a completely different government. The Danish government formation processes establish certain rules for the political parties to organize their internal bargaining since a new prime minister could only result from such a process, and not appear as a deux ex machina. Furthermore, it requires of the political parties to give advice on what they want and what they do not want. If a party refrains from giving clear advice on these matters, it would be against the spirit of the system, adding an unintended uncertainty to the government formation (cf. Kaarsted 1988: 110). Yet, practice from the formation of Jørgensen II described earlier shows that such an action would not necessarily prevent the formation of a government. Rather, it challenges the other parties in parliament to speak up and clearly state that they would support a no confidence vote for a proposed government if they want to prevent its formation. This also applies when current governments change forms and become new governments without having resigned. Having to speak up against a government proposal, rather than to explicitly express support for it, allows for expedient government formations, as we shall see later.

The party system and the actors Party system change When the 1915 constitution introduced universal suffrage and proportional representation there were four parties in parliament: Liberals, Social Democrats, Conservatives, and Social Liberals that reflected the social cleavages, class and urban/rural, important at the time (Elklit 1984). These ‘old parties’ remain in parliament until this day, they remain some of the most important party actors, and no other party has ever held the position of prime minister. They represent traditional party families and ideologies although they have changed character from mass parties with high membership ratios to more professionalized organizations (Bille 1997). In accordance with the cleavage theory, the Danish party system for decades remained ‘frozen’ with only few new parties (Damgaard 1974). The comparatively low electoral threshold of two per cent makes it easy for new parties to enter parliament. In the 1973 ‘Earthquake Election’ the voters increased the number of parties from five to ten and marked the decline of all of the older parties (Pedersen 1987). This indicated more volatility and led to increased fragmentation of parliament. Since then the number of parties have remained at

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



131

a higher level, with between seven and eleven parties in parliament. Table 5.1a shows the effective number of parties (ENP) to be above five most of the time since 1973. Borre (1995) noted a rise of ‘new politics’ dimension in the Danish party system with focus on immaterial values. These may include immigration, law and order, and environment. This may make it meaningful to portray Danish voters and political parties in two dimensions: traditional left–right concerned with ‘socioeconomic’ items and a ‘new’ value-based politics (Stubager 2010). In particular after 2001, the new dimension affects the dynamic of the party system. For many decades of the twentieth century, the Social Liberals controlled the median position in the party system. More emphasis on immigration in particular enabled the parties right of centre to win over enough votes from the left wing to control the median in 2001–2011, and again in 2015–2019 (cf. GreenPedersen and Krogstrup 2008). In 2001, the Liberals overtook the position as the largest party in parliament, something the Social Democrats had been ever since 1929. Since the 2015 election, the relative size of the two parties reversed back again (Christiansen forthcoming 2020). A certain polarization has been taking place with the newer centrist parties, Centre Democrats and Christian People’s Party, not passing the threshold in 2001 and 2005 respectively.² At the same time, the parties on the wings of the party system strengthened. On the right wing, the nationalist and immigration-sceptical Danish People’s Party was formed in 1995 by a leading figure in the Progress Party that in turn had shocked the old party system by winning 17 per cent of the votes in 1973. The Danish People’s Party gained a record-high 21 per cent support level in the 2015 election. In 2007 the party New Alliance, now Liberal Alliance, got elected. It was formed earlier the same year by former MPs and members of European parliament of Social Liberals and Conservatives, later also joined by Liberal MPs. The party originally promoted a centrist profile but now has a markedly market liberal platform. On the left wing, the Red–Green Alliance formed in 1989 and has been represented in parliament since 1994. That alliance united the Danish Communist Party and Left Socialists, both previously elected to parliament, as well as other groups. Another party, the Alternative, was formed in 2013 by a former Social Liberal MP and minister of culture, and it managed to get elected in 2015. The party identifies as a Green party, something Denmark did not get in the 1980s. The Socialist People’s Party combines Socialist and Green positions to the left of the Social Democrats. Despite this polarization, all of the mentioned parties participate in passing policies in the Danish parliament.

² The latter party in 2003 took the name Christian Democrats. It has been on the ballot paper for every general election since then. The Centre Democrats dissolved in 2008.

1945-11-07 1947-11-13 1950-09-16 1950-10-30 1953-04-21 1953-09-30 1955-02-01 1957-05-28 1960-02-21 1960-11-18 1962-09-03 1964-09-26 1966-11-22 1968-02-22 1971-10-11 1972-10-05 1973-12-19 1975-02-13 1977-02-15 1978-08-30 1979-10-26 1981-12-30 1982-09-10

Kristensen Hedtoft I Hedtoft II Eriksen I Eriksen II Hedtoft III Hansen Hansen II Kampmann I Kampmann II Krag I Krag II Krag III Baunsgaard Krag IV Jorgensen I Hartling Jorgensen II Jorgensen III Jorgensen IV Jorgensen V Jorgensen VI Schlüter I

Schlüter II

Schlüter III

Schlüter IV

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

24

25

26

1988-06-03

1987-09-10

1984-01-10

Date in

Cabinet Cabinet number

min

min

min

min min min min min min min mwc mwc min min min min mwc min min min min min min min min min

38

39.1

43.6

25.5 38.7 39.7 39.7 39.7 41.9 41.9 52.5 52.5 49.2 49.2 43 39.1 54.7 39.7 39.7 12.3 30.2 36.9 49.2 38.5 33.5 36.9

179

179

179

149 150 151 151 151 179 179 179 179 179 179 179 179 179 179 179 179 179 179 179 179 179 179

8

9

9

6 6 6 6 6 6 6 6 6 6 6 6 6 6 5 5 10 10 11 11 10 9 9

5.41

5.35

5.14

4.53 3.53 3.96 3.96 3.82 3.69 3.69 3.86 3.9 3.66 3.66 3.6 4.07 4.35 4.04 4.04 7.01 5.53 5.24 5.26 4.91 5.56 5.56

Party Type Cabinet Number of Number of ENP, parties in parliament composition of strength seats in of cabinet cabinet in seats parliament parliament (%)

1945-10-30 V 1947-10-28 S 1950-09-05 S V, KF 1953-04-21 V, KF 1953-09-22 S S 1957-05-14 S, RV, RF S, RV, RF 1960-11-15 S, RV S, RV 1964-09-22 S 1966-11-22 S 1968-01-23 RV, KF, V 1971-09-21 S S 1973-12-04 V 1975-01-09 S 1977-02-15 S S, V 1979-10-23 S 1981-12-08 S KF, V, CD, KrF 1984-01-10 KF, V, CD, KrF 1987-09-08 KF, V, CD, KrF 1988-05-10 KF, V, RV

Election date

Table 5.1a Danish cabinets since 1945

RV

CD

CD

RV RV RV RV RV RV RV RV RV RV RV RV S RV RV RV RV RV CD CD CD CD CD

Median party in first policy dimension

RV

SF SF

SF

RV

Formal support parties

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Rasmussen II Rasmussen III Rasmussen IV Fogh Rasmussen I Fogh Rasmussen II Fogh Rasmussen III Løkke Rasmussen I ThorningSchmidt I ThorningSchmidt II Løkke Rasmussen II Løkke Rasmussen III

29 30 31 32

2016-11-28

2015-06-28

2014-02-03

2011-10-03

2009-04-05

2007-11-13

2005-02-08

1994-09-27 1996-12-30 1998-03-11 2001-11-27

1990-12-18 1993-01-25

V, LA, KF

2015-06-18 V

S, RV

2011-09-15 S, SF, RV

V, KF

2007-11-13 V, KF

2005-02-08 V, KF

1990-12-12 KF, V S, RV, CD, KrF 1994-09-21 S, RV, CD S, RV 1998-03-11 S, RV 2001-11-20 V, KF

min

min

min

min

min

min

min

min min min min

min min

29.6

19

35.8

43

36.3

36.3

39.7

42.5 39.7 39.7 40.8

33.5 49.7

179

179

179

179

179

179

179

179 179 179 179

179 179

9

9

8

8

8

8

7

8 8 10 8

8 8

6.08

5.99

5.68

5.85

5.44

5.48

5.08

4.51 4.51 4.75 4.6

4.36 4.45

DFP

DFP

DFP

DFP

DFP

DFP

DFP

RV RV CD KF

CD CD

SF

DFP

DFP

DFP

DFP

CD

Table legends and variables are further defined in the Appendix, this volume.

Notes: For a list of parties, please consult the chapter appendix. Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1945–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015). The first policy dimension is economic left-–right. Cabinet types: maj = Single-party majority cabinet; min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition.

39

38

37

36

35

34

33

Schlüter V Rasmussen I

27 28

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

134



Of relevance to coalition formation, Danish parties can be divided into ‘blocs’ left and right of the centre, over the recent decades labelled ‘red’ and ‘blue’ by media and now also by many politicians themselves. Parties within a bloc declares to support—or at least agrees to not topple—a prime minister from its own bloc, should it win a majority. Usually, the parties declare such an intention in public before an election so that voters would know or fearing that uncertainty may cost voters. Shifts between blocs are rare. Historically the Social Liberals have shifted blocs occasionally, but not since 1993. From time to time, there are parties that insist on not, or no longer, belonging to a bloc, like the Alternative in the 2019 election. Competition between the two blocs to win office remains sharp. At many elections—such as 1998, 2011, and 2015—there was only a tiny difference between the sizes of the two blocs. Yet, this does not make Denmark a Westminster-style democracy since cooperation among political parties across blocs about legislation is also at high levels—usually above 80 per cent of all legislation (Green-Pedersen and Thomsen 2005; Christiansen 2018).

Pre-electoral coalitions The Danish electoral law at the national level does not allow for formal electoral alliances between parties with their own label. Attempts by minor parties to present a joint ticket, while remaining separate entities,- have occurred but not succeeded to cross the electoral threshold of two per cent.³ It is somewhat more common for two or more parties to declare their intent to form a government together after the election; see Table 5.1b. For incumbent coalition governments this has been the case at each election since 1998. Among opposition parties, Social Democrats and Social Liberals that had been in office together prior to 2001 campaigned on forming a new coalition government in 2005, should the election result allow for that. In 2007, the two parties did not have the same understanding, and in 2009, the Social Democrats and the Socialist People’s Party declared their intent to form a government. They issued two joint general policy documents, supplemented between 2009 and 2011 with a number of joint policy positions. The two parties also began to vote together in parliament on most issues. This process took place to ‘prepare’ the Socialist People’s Party for taking office for the first time (Christiansen et al. 2014). When that eventually happened after 2011, Social Liberals also became part of the government and, as

³ The Red–Green Alliance constitutes a partial exception (Seeberg and Kölln forthcoming). Today, it is a party in its own right, much stronger than what remains of its constituent founders. Moreover, the party has operated with one coherent group in the Folketing.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



135

Table 5.1b Electoral alliances and pre-electoral coalitions in Denmark, 1988–2018 Election date Constituent parties Type Types of pre-electoral commitment 1988-05-10 1987-09-08 1988-05-10 1990-12-12 1994-09-21

V, KF, CD, KrF* V, KF, CD, KrF* V, KF, CD, KrF* V, KF, RV* S, CD, RV, KrF* V, KF

PEC PEC PEC PEC PEC PEC

1998-03-11

S, RV* V, KF S, RV* V, KF

PEC PEC PEC PEC

V, KF* V, KF* V, KF* S, SF S, RV*

PEC PEC PEC PEC PEC

2001-11-20 2005-02-08 2007-11-13 2011-09-15 2015-06-18

Joint press conference, separate declarations Joint press conference, separate declarations Joint press conference, separate declarations Separate declarations Separate declarations Written contract, joint press conference, separate declarations Joint press conference, separate declarations Separate declarations Separate declarations Written contract, joint press conference, separate declarations Joint press conference, separate declarations Joint press conference, separate declarations Joint press conference, separate declarations Written contract, joint press conference Joint press conference, separate declarations

* Incumbent government and constituent parties intending to continue in office after the election if possible Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other

part of the compromise to form a cabinet coalition, strongly amended the policies of the pre-electoral pact. In 2015 none of the opposition parties made pre-electoral coalitions.

Government formation The bargaining process The coalition signals in blocs declaring their support for prime ministerial candidates narrow the possibilities, and reduce the risk, of impasse after an election. Since 1990, governments resigned after the elections in 2001, 2011, and 2015 and without election in 1993. The formal government formation procedures in 2001 and 2011 only required one queen’s round. Bargaining takes place at Christiansborg or in private homes, yet in 2011 at a hotel, making it a semi-public event. The negotiations to extend government in 2016 took place at the prime minister’s official home, Marienborg. When a Danish government is to form, the formation process usually lasts no more than a few weeks, with an average since

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

136



1988 of 10 days and no longer than 24 days (see Table 5.2).⁴ This is quite swift. Most governments form after just one bargaining round, although historically there were up to four rounds (Damgaard 2000: 240). After the elections in 1990, 1994, 1998, 2005, and 2007, no ‘queen’s rounds’ took place. Few members of the party leaderships conduct the coalition bargaining. Besides clarifying the party composition, they primarily bargain over government policies. It takes precedence and more time than the bargaining over portfolio allocation for coalition governments that usually takes place only after an agreement over a joint policy programme. The interviewees of mostly former high-ranking ministers confirmed this pattern. Following a shift in government or a general election, the incoming prime minister often gives a speech in parliament presenting the plans of the government followed by a debate, although in 2015 (and again in 2019) it was only after several months in office. Sometimes this ends with a vote over a motion that expresses support for the government. Yet, there is no formal investiture vote. Historically (as in 1975), there were dramatic government formations with uncertain outcomes as to who would become prime minister (Kaarsted 1988). Yet, often, and in particularly since the 1990s, the two ‘blocs’ of parties would declare their support before an election for either a centre-right or centre-left prime minister. The distribution of seats between the two blocs following the election result then determines who becomes prime minister. A bargaining process will affect the party composition of the government. As an example, at the 2015 election the four parties in the ‘blue bloc’ achieved a slim majority of 90 seats over the 89 seats of the ‘red bloc’. The coalition signals by the blocs (cf. Debus 2009) made it likely that the Liberal leader would become the new prime minister despite heavy electoral losses for his party, which only became the third largest. At the first ‘queen’s round’, three of the parties duly pointed at Løkke Rasmussen without reservations but Liberal Alliance added the condition that he should form a majority government of all four parties. The caretaker prime minister (Thorning-Schmidt) then concluded that Løkke Rasmussen should be appointed to search for the formation of a majority government of those four parties (kongehuset.dk 2015a). Hence, his mandate was limited, assuming that Liberal Alliance would not allow a smaller government to form. After a few days, he returned concluding that such a majority government was not possible. In the second round, Løkke Rasmussen received an open mandate by all four parties to seek to form a government, and after further negotiations, he returned with the conclusion that he could form a single-party government. The caretaker prime minister agreed with this conclusion, and the queen then appointed Løkke Rasmussen as the new prime minister with his own countersignature (kongehuset.dk 2015b).

⁴ Excluding governments that remain in office after an election, which the Danish custom allows for.

Year in

1988

1990 1993 1994 1996 1998 2001 2005 2007 2009 2011 2014 2015

2016

Cabinet

Schlüter IV

Schlüter V Rasmussen I Rasmussen II Rasmussen III Rasmussen IV Fogh Rasmussen I Fogh Rasmussen II Fogh Rasmussen III Løkke Rasmussen I Thorning-Schmidt I Thorning-Schmidt II Løkke Rasmussen II

Løkke Rasmussen III

0

0 0 0 0 0 0 0 0 0 0 0 1

2

Number of inconclusive bargaining rounds

V (1) DFP, V, LA, KF

KF, V, RV (1) All parties (2) S, RV, V, KF

Parties involved in the previous bargaining rounds

Table 5.2 Government formation period in Denmark, 1988–2018

9

0 0 0 0 0 0 0 0 0 0 0 10

24

Bargaining duration of individual formation attempt (in days)

9

6 10 6 10 12 7 10 10 0 18 4 10

24

Number of days required in government formation

9

6 10 6 10 12 7 10 10 0 18 4 10

24

Total bargaining duration

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

138



Under the system described, a candidate for prime minister forms a government when supported by the ‘bloc’ that controls the median. The prime minister does not necessarily come from the median party. A smaller centrist party often held this position, as noted in Table 5.1a, without always joining the government. In particular, the Social Liberal Party held this position for many years. Until 2001, the decision of this party to support either the centre-left or the centre-right determined which side it would be (Green-Pedersen 2005). A party with support in single digits may be too small to form a single-party government (Skjæveland 2003) or to aim for the position of prime minister itself. Since 1993, and despite occasional reservations, the Social Liberals has been part of the ‘red bloc’, in the sense that it has pointed to a Social Democratic prime minister before each election. Furthermore, since 2001 the Social Liberals only rarely held the median position (2011–2015 and from 2019). This profoundly changed the dynamics of Danish government formation since the key to prime minister’s office now no longer lies in presenting policies attractive to the Social Liberals but rather in maximizing votes for the ‘bloc’ at the election time (Green-Pedersen 2005).

The composition and size of cabinets Denmark holds the world record for most minority governments (Damgaard 2011). After 1990—with increased party fractionalization and strengthened legislative committee system—only one government (1993–1994) controlled a majority.⁵ In a parliamentary system, this means that some parties sustain the survival of the government, supporting or at least not toppling it. In Denmark, this so-called ‘parliamentary basis’ usually consists of both cabinet and non-cabinet parties. The 2019 government formation presented the first Danish example of of ‘contract parliamentarism’ of the kind found in Sweden, with a written comprehensive support agreement agreeing on policy and/or other procedural arrangements (Bale and Bergman 2006). Prior to 2019 there are examples of support parties holding prestigious parliamentary positions such as Speaker or chair of the Finance Committee. The literature points at a number of general explanations for the existence of minority governments, and most of them are present in the Danish case: fractionalization, a strong committee system, a negative version of parliamentary government, and no investiture vote (Strøm 1990; Bergman 1993; cf. Rasch 2011). Yet, it does not have one large party just a few seats short of majority. For a long time minority governments had a bad reputation for instability in the

⁵ The government during 1993–1994 controlled the 90 seats needed for majority most of its time in office but in periods at the beginning and its end it controlled only 89 seats due to independents readmitted or leaving government party parliamentary groups (Skjæveland 2003). These independents still declared to support the government.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



139

comparative literature. In countries such as Germany and France they are still anathema. Yet lack of stability does not seem a major concern when Danish parties decide to enter minority governments. Minority governments may allow more flexibility in majority building for legislation. Before forming Løkke Rasmussen II, the designated prime minister publicly stated (widely cited in newspapers) that a four-party majority government would make it difficult to manoeuvre and reach out to the opposition. One may argue that such a government becomes more vulnerable to attack from the opposition that would dominate the public agenda (Christiansen and Seeberg 2016). Since no single party has held a majority in parliament under the proportional electoral system first introduced in its first version in 1915, the few peacetime majority governments were all coalitions, and they have all been of the minimal winning type and so no surplus governments (see Table 5.3). Of the minority governments, there are both single-party government and minority coalitions of between two and four parties. In particular, the Social Democrats and the Liberals—the major parties of the centre-left and for most of the time of the centre-right—appear to maintain a strong ambition for office and heading the government themselves, and this upholds competition between distinct blocs about forming the government. Despite this appetite for office, with one historical exception in peacetime (1978–1979), lasting for only 14 months (cf. Borchsenius 2004), the two parties have not formed a government together. Many minority coalition governments have controlled about 40 per cent of the seats in parliament. At this level, minority governments may find a balance between stability and flexibility. Most of the single-party governments were smaller than 40 per cent, Løkke Rasmussen II controlling less than 20 per cent of the seats. Minority coalitions are rare compated with other countries but in Denmark, they have become the most frequent type of government. They emerge when majority governments fail to form or while some parties inside a bloc do not join the government while others do. Particularly the parties on the wings of Danish politics are rare guests in government. In 2011, the Socialist People’s Party joined for the first time after more than 50 years of existence, and in 2016, Liberal Alliance became part of the government. Yet the Danish People’s Party and the Red–Green Alliance have not been part of a government. The issue of the European Union still divides mainstream and non-mainstream parties on both wings. Participation in the government requires EU-sceptic parties to accept and administrate many EU initiatives (Christiansen and Pedersen 2014).

Portfolio Allocation The Danish constitution states that ‘the king’ decides upon the number of ministers and their tasks. The prime minister exercises this prerogative, meaning

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

140



that parliament cannot intervene in appointments. In practice, the cabinet party leaders decide appointments. The five former ministers interviewed confirmed this pattern, adding that the prime minister has a right to veto but none of them recalled any cases, except one with a security concern (belonging to the prime minister’s own party). For a single-party government, the constitution may actually provide an accurate description inasmuch as the prime minister has always also been the party leader. Yet parties about to form a coalition government bargain over the distribution of portfolios among them. In the Danish case, there is an over-representation of small parties compared to the proportional distribution expected by Gamson’s law (cf. Skjæveland 2003). In Thorning-Schmidt I the Social Democrats only got 11 out of 23 positions, yet most of the positions held some political saliency (cf. Warwick and Druckman 2001). Today, it is quite common to split and merge ministries, and in particular portfolios, as part of the balance between coalition parties but historically (prior to the 1960s) such a practice was rare. With one historical exception (1968–1971), the largest party took the position of prime minister. Table 5.3 presents the historical record for five ministries that are considered particularly important, in Denmark and comparatively since 1988. In most cases—including those in the table, the latest exception was from 1982 until 1984—the party of the prime minister also gets the Ministry of Finance. Thereby, the leading party in government controls the two most important ministries for policy coordination and control. This is of course desirable for the party but it also reflects that the largest party in government has usually been much larger than the second largest. Mortensen (2014) and two former ministers from different parties mention in the interviews, that the leader of the Social Liberals wanted to become the minister of finance during the 2011 government bargaining but this was turned down by the Social Democrats. According to the leader of Liberal Alliance, and the newspaper reports at the time of government negotiations in 2016, the Liberals insisted on keeping the portfolios of prime minister and of finance but left other portfolios open. Interviews with former ministers confirm that finance and foreign affairs are permanently important portfolios. Justice also ranks high consistently. One interviewee, a former minister of justice, argued that finance and justice are important because they can stop initiatives of other ministries for financial or legal reasons. In a number of cases, the second largest party got the Ministry of Foreign Affairs. Occasionally, a junior party leader took this position for him or herself (1984–1993, 2010–2011, 2011–2012, and again 2016–2019) and in other cases left it to former party leaders (1993–2000, 2001–2010, 2012–2013). Despite its prestige and possible influence on major events, party leaders risk becoming detached from domestic matters from that position, according to interviews, newspaper reports, and memoirs from several former ministers of foreign affairs (Interviews; cf. Hartling 1981; Lykketoft 2019). Other leaders of the second largest

Year in

1988 1990 1993 1994 1996 1998 2001 2005 2007 2009 2011 2014 2016

Cabinet

Schlüter IV Schlüter V Nyrup Rasmussen I Nyrup Rasmussen II Nyrup Rasmussen III Nyrup Rasmussen IV Fogh Rasmussen I Fogh Rasmussen II Fogh Rasmussen III Løkke Rasmussen I Thorning-Schmidt I Thorning-Schmidt II Løkke Rasmussen III

9 KF, 7 V, 5 RV 10 KF, 9 V 15 S, 4 CD, 3 RV, 2 KrF 15 S, 3 RV, 2 CD 15 S, 4 RV 16 S, 4 RV 13 V, 5 KF 12 V, 7 KF 12 V, 7 KF 12 V, 7 KF 13 S, 5 RV, 5 SF 13 S, 7 RV 13 V, 6 LA, 3 KF

Number of ministers per party (in descending order) 21 19 24 20 19 20 18 19 19 19 23 20 22

Total number of ministers 21 22 24 20 20 20 20 21 21 21 20 20 19

Number of ministries

Table 5.3 Distribution of cabinet ministerships in Danish coalitions, 1988–2018

KF KF S S S S V V V V S S V

1 Prime minister KF KF S S S S V V V V S S V

2 Finance V V RV RV RV RV KF KF KF KF SF RV LA

3 Foreign affairs

RV V S S S S KF V V V S RV KF

4 Social affairs/ welfare

KF KF S S S S KF KF KF KF S S KF

5 Justice

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

142



party held the Ministry of Economy, sometimes in combination with business or the interior (1993–2001, 2001–2010, 2011–2015, 2016–2019). Some junior party leaders have held posts of ‘coordination’ without any specific portfolio (1993–1994 and two more instances before 1990). Others held more specific portfolios such as energy (1993–194), justice (2010–2011, 2016–2019), business (1994–1996 and 2012–2013), and social affairs (2013–2014). A former party leader and highly ranking minister stated in the interviews that it would be harder for party leaders in ministries without experience in coordination (such as energy or social affairs) to perform surveillance of other ministries and support their minister in tasks of coordination within cabinet (Smith 2015: 85–7). With one historical exception (the Centre Democrats 1982–1987), all party leaders took government office themselves (cf. Bille 1997: 379–86). The number of cabinet members has increased over time, reflecting the size and complexity of the public sector (Damgaard 2000). It reached its highest point to date with 24 in 1993. In 2016 there were 22 (Table 5.3), enlarged from 17 in the single-party government from 2015. There are no junior ministers in Denmark; that is, all ministers are full cabinet members. Occasionally, one ministry may have more than one minister, with one of them holding the more ‘senior’ title similar to the ministry itself. Yet there is no hierarchy in decision-making since ministers have singular responsibility for the portfolio they are assigned. In the Ministry of Foreign Affairs, this has been the case with the portfolios of development aid, foreign trade, Nordic cooperation, and Europe. One interviewed former minister of foreign affairs colloquially labelled these posts of other ministers as ‘junior’, and added that they result from bargaining over portfolios. Hence, in most cases, more ministers within one ministry come from different parties. This was also the case for the minister for the elderly appointed in 2016 who found herself inside the Ministry of Health, whereas the minister for Innovation also appointed in 2016 was inside the Ministry of Finance who has a minister from her own party. Sometimes, one cabinet member is the minister for separate ministries. Thus, since 2016 (also 2010–2011 and 2014–2015) the ministries for culture and for church have shared the same minister. Since 1994, the number of ministries has remained more stable (19–21) than the number of cabinet members. The distribution of portfolios in Danish coalition government confirm some ‘party family preferences’. Without exception, Social Democrats have held the Ministry of Labour, Liberals the Ministry of Agriculture, and, with only two historical exceptions, Conservatives the Ministry of Justice. This party has also shown preference for the business portfolio. Social Liberals have always either held education or research, and quite often culture. The interviews with former ministers confirm this pattern, with one former party leader stating, ‘It is important to get portfolios on issues where the party has a strong profile.’ The same former minister also added that some issue areas, such as energy and climate, have become more important over time.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



143

Coalition Agreements The coalition agreement of Nyrup Rasmussen I introduced coalition agreements in Denmark as a coherent document made publicly available (Knudsen 2000: 64; Christiansen and Pedersen 2014). Until then not every coalition had such a document and those that existed were often not published until later, if ever (Damgaard 2000: 244–6). The prevalence of minority governments was a possible reason (Damgaard 2000: 261). For minority coalitions, it would seem difficult first to negotiate a compromise among the cabinet parties and next to negotiate another compromise with opposition parties. Yet permanent support parties may support coalition agreements in return for major concessions in other arenas as part of a larger ‘logroll’ and that was very much the case for the governments between 2001 and 2011 (Christiansen and Pedersen 2014). Table 5.4 presents the record. We know the historical origin of the decision to have a coalition agreement in 1993 in some detail (Christiansen and Pedersen 2014; Lykketoft 2019; interviews with former leading ministers from the government 1993–1994). It was a condition raised by the then leader of the Social Liberals, Marianne Jelved, for taking part in any future government. Her party participated in Schlüter IV (1988–1990), and as parliamentary group leader she found it difficult to defend joint policies without knowing the overall goal. She had communicated this condition to one of the most prominent Social Democrats who had won support for the idea internally prior to the formation of a majority government in 1993. Since it had the power to pass the policies agreed upon, the content of the document became even more important (interviews with former leading ministers from the government 1993–1994). The documents prior to 1993 were of short length. See Table 5.4. The first document issued in 1993 and for the Nyrup Rasmussen cabinets I–IV until 2001 contained about 4,000–5,000 words, that is the length of a pamphlet. Later, the documents have increased very much in size, with Thorning-Schmidt I and II setting the current record of more than 25,000 words, that is the size of a book, and much longer than their Swedish counterparts although still shorter than the German ones. Consequently, the linguistic style of the text has become less punctual, yet it also covers more topics and thus become more exhaustive. The relative share of different types of content according to the categories in Table 5.4 has not changed much over time. Most of coalition agreements, about 90 per cent, concerns policy. Some of the policy content is quite specific initiatives while other content indicates a certain direction, but there are also pure descriptions, perceptions, prognosis, or wishes (cf. Christiansen and Pedersen 2014). The second most important category is procedural rules related to policy. It often includes scrutiny to prepare a policy decision in a certain direction. For most governments there are also a few general procedural remarks. They may concern how to make certain decisions among the cabinet parties themselves or with noncabinet parties.

Year in

1957 1993 1994 1996 1998 2001 2005 2007 2009 2011 2014 2016

Coalition

Hansen II Nyrup Rasmussen I Nyrup Rasmussen II Nyrup Rasmussen III Nyrup Rasmussen IV Fogh Rasmussen I Fogh Rasmussen II Fogh Rasmussen III Løkke Rasmussen I Thorning-Schmidt I Thorning-Schmidt II Løkke Rasmussen III

910 4,130 3,720 3,720 5,613 11,095 21,335 22,530 22,530 26,649 26,649 22,697

Size (words) 0 0 0 0 0 2 1 1 1 1 1 0

General rules (in %) 30 6 10 10 6 9 13 13 13 6 6 12

Policy-specific procedural rules (in %)

Table 5.4 Size and content of coalition agreements in Denmark, 1945–2018

0 0 0 0 0 0 0 0 0 0 0 0

Distribution of offices (in %) 0 0 0 0 0 1 1 0 0 0 0 0

Distribution of competences (in %)

70 94 90 90 93 88 85 86 86 93 93 88

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



145

According to the five interviewed ministers, the coalition agreement is the most important document, in which the parties commit themselves to joint policies. A former cabinet party leader describes it as ‘the bible’ (Bendtsen 2019: 229). The ministers agree that the government leadership, and the individual ministers, find it most important to get them implemented. One of them (a former party leader and top-ranking minister) stated that coalition agreements ‘do not preclude fight over its implementation’ and mentioned specific examples of reverted decisions. The coalition agreements do not preclude activity on other items but it does reflect what the government parties want to do, could agree upon when disagreeing, and prioritize. They may also promise to uphold the status quo. An important example was the statement in the Thorning-Schmidt I coalition agreement to continue the economic policies of its Liberal–Conservative predecessor. This was a requirement from the Social Liberals to join the government, in addition to skipping most policies from the pre-electoral pact described earlier (Mortensen 2014; Nielsen 2016). Minority governments cannot be certain to gets is coalition agreement and have to bargain with other parties, using one of two general strategies: compromise and logrolling (Christiansen and Pedersen 2014). In the former case the government gets support for the content of an item in the coalition agreement in an amended form. In the latter it gets support for an item as it stands in return for concession on other items. If cabinet parties anticipate problems, they could leave out such content from the coalition agreement altogether. For the opposition the content of the coalition agreement also receives strong attention and becomes focus of opposition (Christiansen and Pedersen 2014). For comparison, it is interesting to note that the single-party government Løkke Rasmussen II issued a government declaration somewhat shorter than for the coalition agreements before and after. One explanation could be that the government anticipated problems with getting a too comprehensive programme passed in parliament and that it did not want to exhibit this too much. One of the interviewed ministers from the Thorning-Schmidt I government stated that its coalition agreement may have been too detailed for a minority government. Furthermore, for an individual party in government, the document might not only serve a purpose of coordination between different parties but also help adjudicate among possible factions inside the party, factions that would otherwise object to some of the coalition proposals.

Coalition governance Table 5.5 demonstrates the use of coalition governance mechanism in Danish coalition governments.

1950 1953 1957 1960 1960 1962 1968 1978 1982 1984 1987 1988 1990 1993 1994 1996 1998 2001 2005 2007 2009 2011 2014 2016

N N POST POST N N POST IE N N N N N IE POST POST POST POST POST POST POST POST POST POST

n/a n/a y y n/a n/a n n n/a n/a n/a n/a n/a y y y y y y y y y y y

n n n n n n n n n n n n n n n n n n n n n n n n

IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O IC, CaC, Pca, O

All used

CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC

IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC

Y Y Y Y Y Y Y Y N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y

N N N N N N N N N N N N N N N N N N N N N N N N

Personal Issues union excluded from Most For most agenda common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All

y y y y y y y y y y y y y y y y y y y y y y y y

Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp.

– – – – – – – – – – – – – – – – – – – – – – – –

y y y y y y y y y y y y y y y y y y y y y y y y

Junior NonCoalition Freedom of Policy discipline appointment agreement ministers cabinet positions in legislation/ other parl. behaviour

Notes: Coalition agreement: IE = inter-election; POST = post-election Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee; Pca = Combination of cabinet members and parliamentarians; O = Other Coalition discipline: All = Discipline always expected Policy agreement: Comp. = Comprehensive policy agreement

Eriksen I Eriksen II Hansen II Kampmann I Kampmann II Krag I Baunsgaard Jorgensen IV Schlüter I Schlüter II Schlüter III Schlüter IV Schlüter V Nyrup Rasmussen I Nyrup Rasmussen II Nyrup Rasmussen III Nyrup Rasmussen IV Fogh Rasmussen I Fogh Rasmussen II Fogh Rasmussen III Løkke Rasmussen I Thorning-Schmidt I Thorning-Schmidt II Løkke Rasmussen III

Coalition

Table 5.5 Coalition governance mechanisms in Danish coalitions, 1945–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



147

The role of individual ministers in policy-making Since the introduction of cabinet government, each minister individually holds full executive power with formal responsibility for his or her portfolio. Unlike Sweden, the cabinet is not a collective decision-maker and the minister is the supreme administrative leader of the entire department. Misconduct in office violating the law on ministers’ responsibility could lead parliament to raise a case at the High Court of Realm. Formally, Denmark fits some of the important assumptions of the portfolio allocation model by Laver and Shepsle (1996) providing the minister with strong powers over a policy area. Previously, individual ministries were quite independent (Knudsen 2000: 31–2; Smith 2015: 85). Since then coalition agreements, and other forms of coordination and control described later has reduced the decision-making power of individual ministers. There was still much room left for individual ministers for policy formulation during the governments of Schlüter (1982–1993) who wanted to focus on ‘big lines’ and deliberately avoided interfering much in the affairs of individual ministers, except in cases of controversy (Schlüter 1998: 136; Olesen 2018: 54). The coalition agreements described earlier significantly reduce the actual discretionary powers of ministers. Today, most cabinet ministers, and their civil servants, will consider it a duty to fulfil the coalition agreement, although there may be some degrees of freedom in the process, and perhaps in particular in the implementation that follows. Furthermore, individual ministers may have to pass policies affecting them that result from legislative agreements that go across policy sectors, which is very often the case with the annual state budget agreement (Christiansen 2012). The policy role of ministers would be to supplement the collective and hierarchical decisions and to feed into these decisions. It would require a decision by the government top ministers to remove an item from the coalition agreement, and it could cause public embarrassment. That was the case during the Thorning-Schmidt I government concerning a road tax in Copenhagen and parental leave earmarked for fathers. These were well-known examples that were also referred to by the interviewed ministers.

Coalition governance in the executive arena At its origin (in 1848–1849), Danish cabinet government had seven ministries but with the expansion of the public sector into new domains and tasks, the number increased gradually. As pointed out earlier, the number is now around 20. Danish governments for a while attempted to solve problems with coordination horizontally (Christensen 1985; Knudsen 2000: 70) but now more hierarchical forms have

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

148



become important. The constitution only mentions one unit of coordination: the state council, a body with centuries-long historical roots, composed of the ministers as well as the monarch and the heir to the throne. This body meets about 12 to 13 times per year, usually on Wednesdays, and gives formal assent to measures, such as bills, proposed by a minister in the capacity of a member of the government. The five interviewed former ministers all stated that the state council is strictly ceremonial (cf. Knudsen 2000: 64–5). A more important ‘horizontal’ unit of coordination is the meeting of ministers in the cabinet as a whole. They often take place on Tuesdays about 40 times per year, which is less often than previously (Olsen 1999: 86; interviews with former ministers). Today, most decisions on behalf of the full government have already been prepared in advance in the coordinating committees. Although it does not happen often, a minister may voice dissent and have it in the records but none of the interviewed former ministers did remember or had heard of votes taking place. The prime minister leads the meeting, and concludes. The interviewed ministers agree that the Council of Ministers only has a very limited role in cabinet decisionmaking. Here, the central government committees are more important. Coordination and control with individual ministerial decision-making and spending prior to the 1970s has been described as ‘loose and generous’ (Olesen and Olesen 2018: 595). It led to an increased perception of planning and coordination problems, making it hard to control public spending and prevent contradictory policies (Knudsen 1995). During the 1970s a huge number of— horizontal—internal cabinet committees flourished with the purpose of overcoming various coordination problem (Christensen 1985; Olsen 1999: 95–100; Knudsen 2000: 68–78). During those years, the prime minister deliberately tried to prevent the Ministry of Finance from becoming too dominant (Olesen and Olesen 2018: 595). During the 1980s, the number of internal cabinet committees decreased. The Ministry of Finance began to take upon more of the policy coordination effort (Jensen 2003; Olesen 2018). In 1982, a coordination committee (K-udvalget) was put in place consisting of the leading ministers, soon described as ‘the command centre’ of the government (Olesen 2018: 54). It had seven members—with each of the four parties represented and with the heads of the department of the prime minister and the Ministry of Finance present (Olesen 2018). Later during the 1980s this committee may have lost some significance again (Olesen 2018). Yet, in general at the time, there was reduced horizontal coordination and somewhat expanded vertical coordination, allowing the Ministry of Finance to strengthen its role as coordinator through control of public expenditures and budgeting (Jensen 2003, 2008). The governments of Nyrup Rasmussen (1993–2001) reformed from their beginning executive coalition governance in addition to the introduction of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



149

coalition agreements. These put much more emphasis on hierarchy and coordination than previously (Lykketoft 2019; Knudsen 2000; interviews with two former leading ministers). The prime minister’s office got enhanced with the purpose of monitoring what happened in the ministries. In particular, the Ministry of Finance further strengthened its coordinating role. It monitored not only expenses but also more and more the efficiency of administrative practices. It also strengthened the committee for coordination (K-udvalget) and re-established the committee for economy (Ø-udvalget), founded in 1947 but having ceased to exist during 1990–1993 (Knudsen 2000: 74). These became important arenas for cabinet decision-making. There is often a partial overlap in their membership, with leading ministers, and often the party leaders, holding seats. Here, individual ministers would have to present and get approval for important measures (Knudsen 2000: 78). These events are important in the life of a minister and are compared with an examination. According to a former cabinet party leader, ministers during his time (2001–2008) literally waited outside the door until they were called in, and they had to be well prepared to have a chance to get a ‘green light’ (Bendtsen 2019: 257). Most proposals come to the committees in a short written form stating the purpose of the proposal. All five interviewed former ministers—of whom all had held a seat in one of the central cabinet committees but some of them were also at times ‘ordinary’, or line, ministers—stated that the economic committee solves most internal conflicts concerning domestic matters (cf. Knudsen 2000: 78). Its meetings are usually weekly, on Tuesdays, chaired by the minister of finance. It has members, ranging between three and seven, from leading economic ministries, but not only from them, representing all parties in a coalition government. The permanent secretaries gained access from 1993 to the economic committee, and since 2001 to the coordination committee, showing their increasing importance in internal cabinet decision-making (Jensen 2003). Here, cabinet parties and ministers bring forward disagreements and try to resolve them. The secretariat of civil servants assigned to these committees also help follow whether and how much of the coalition agreement and other government decisions have been fulfilled. At its meeting, the individual minister with a point on the agenda and his or her permanent secretary are also present. A ‘steering group’ controls access to the committee and consists of the heads of department of the ministers in the committee, that is of civil servants. Yet most of the interviewed former ministers point out that party leaders are allowed to bypass this system and put items on the agenda, making it possible for a line minister to ‘appeal’ to her or his party leader. The leading ministers have seats in the coordination committee headed by the prime minister. According to the interviewed former ministers, for most

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

150



governments the coordination committee meets less frequently than the economic committee does, usually in the week before the occurrence of an all-cabinet meeting. Other figures partly confirm this pattern but also an increased activity of the K-committee over time. Between 1993 and 2008, the Ø-committee met at an average of 35 times per year while the K-committee met 18 times during 1993–2001, increasing to 26 times during 2002–2008 (Fuglsang and Jensen 2010: 19–21). During 2011–2014 the two central committees met at about the same frequency according to Nielsen (2014). The coordination committee usually has up to seven members. It decides major issues concerning the government, including all bills. According to memoirs and interviewed former ministers of whom several used to be members of it, it deals with matters not solved in the economic committee and with matters related to foreign affairs (cf. Bendtsen 2019: 256). There is an attempt to avoid trivial matters (Bendtsen 2019). According to several of the interviewed former ministers, an individual minister who does not agree with the decision of the coordination committee will have to accept and follow it anyway, or ultimately resign. In both central government committees, holding the chair provides the prime minister and the minister of finance with very important roles in decision-making. They conclude on the decisions made. Yet, according to the interviewed ministers, including former party leaders, each party leader in government holds a veto over decision-making. The prime ministerial office monitors the implementation of government policy, including usually quite elaborate coalition agreements. The Ministry of Finance follows economic spending. In this structure, the relation between the prime minister and the minister of finance becomes very important (Knudsen 2000: 78). Although they usually come from the same party, the prime minister is higher in the party hierarchy but sometimes lacks experience (cf. Nielsen 2016: 227; interview with a leading former minister). During the governments of Anders Fogh Rasmussen (2001–2009), the role of the prime minister may have strengthened relative to the Ministry of Finance, and the junior partner in government. This was partly for political and personal reasons, with a prime minister able to dominate the work of the government and strong-hand decision-making. Furthermore, the economy was no longer in abyss, pointing at the economic conditions as a partial reason for the influence of this ministry (Jensen 2008). With the global financial crisis hitting Denmark around 2008, austerity and the pressure for economic coordination increased once more, and the Ministry of Finance became a central actor in passing and implementing a budget bill in 2012 that covered the entire public sector and introduced long-term planning. Hence, during the governments of Thorning-Schmidt (2011–2015), the minister of finance was once again a central actor. There was a lot of bargaining and conflict resolution inside the coordination committee. That was also the case for the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



151

somewhat ideologically diverse three-party Løkke Rasmussen III government (2016–2019).⁶ There are also a few other government committees: one for hiring of senior civil servants, increasing its functions over time since its establishment in 1977 (Nielsen 2017), and one for security matters. The Løkke Rasmussen III government (2016–2019) also had a committee for EU implementation and one for the ‘renewal’ of the public sector. As already mentioned, unlike the situation in many other countries, Danish governments have not developed ‘junior ministers’, that is members of the government assigned to a senior cabinet member with a hierarchical decisionmaking power. All ministers remain ‘equals’ in the sense that they are full cabinet members, and no other minister holds formal authority over the portfolio assigned to them. Why is that so? On the one hand, Danish ministers have extremely busy schedules and could need to be relieved (Knudsen 2000). On the other hand, junior ministers are also agents of control, in particular when they come from other parties, and the other mechanisms seem to work well in the Danish case. Furthermore, due to their minority status, Danish governments still need to negotiate its policies, and as we shall see later, the legislative agreements with opposition parties also provide governance mechanisms for the political parties involved.

Governance mechanisms in the parliamentary arena Danish cabinet members represent party groups, and ultimately they need the approval of their parliamentary groups to stay in office and get bills passed. Yet they could expect to get it in most cases. For most legislative terms government party groups have supported 100 per cent of the laws passed by the government (Christiansen 2018). According to the interviewed former ministers, there is a strong norm that MPs from cabinet parties support government initiatives when accepted by the party group in a meeting. This includes the content of the coalition agreement. Danish ministers are usually also members of the parliamentary party group. An even stronger norm requires of ministers who are also MPs to support the line of the government in votes—the logic is that they have given their assent in state council and not resigned—even if the party group should end up dissenting. In 2018, there was one very rare example of this distinction when a majority of the Liberal Alliance party group voted against a government-proposed

⁶ The government formed in 2019 has strenghtened the position of the prime minister’s office, relative to the ministry of Finance but this recent development goes beyond the data collection conducted for the purpose of this chapter.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

152



ban on face-covering clothing (or headgear) in public spaces while the ministers from the party voted for it. The party group leader acts as a liaison between the cabinet and party groups, keeping the latter in line with the government but also communicating when the cabinet is about to go further than it could find support for. When a government takes office, its party groups would have accepted it and the coalition agreement as well. Spokespersons of government parties take part in negotiating government policy together with the minister, and they monitor the coalition agreement on behalf of their party. The latter role is particularly important for the party spokesperson of other parties than the minister (interview with a former party leader and minister). The minister and the civil service would usually be ready to listen to government party spokespersons, provide them with information, and answer informal questions from them. It is very rare that government party members raise a formal parliamentary question to a minister. Likewise, according to the interviewed former ministers, government party members generally do not use the powers of legislative committees to control the ministers or to revise government bills. Hence, the formal channel through legislative committees to ‘shadow’ their own ministers does not play the role expected by Martin and Vanberg (2004) or Carroll and Cox (2012). Former ministers and MPs indicated in a public hearing in parliament that government parties prefer to maintain internal proceedings as much as possible rather than arenas where the opposition is also present and could take advantage of the vulnerability when a government shows internal division (Folketinget 2015). A few minority governments have had permanent support parties. They commit not only to survival in office but also to pass larger parts of the government’s programme. If the minority government thereby reaches a majority, they are ‘formal’ or a majority government in disguise (cf. Strøm 1990; Bale and Bergman 2006). That was very much the case between 2001 and 2011 with the Danish People’s Party supporting Liberal–Conservative cabinets. Back then, as a routine, most cabinet ministers informed or involved the spokespersons of its support party at early stages of legislation. Christiansen and Pedersen (2014) found that the Danish People’s Party supported the passing of most of the coalition agreement but got concession on other items. Minority governments need to bargain with non-cabinet parties in parliament to get policies passed. That includes the policies of the coalition agreement (Christiansen and Pedersen 2014). They find a parliament formally organized in party groups and legislative committees with features that support such bargaining. Every Danish parliamentary party group upholds strong party cohesion in final voting (Skjæveland 2001). This implies that each party in negotiations with the government credibly represents a voting weight equal to its number of seats in parliament. The party groups organize themselves with both a horizontal and a vertical division of labour. Horizontally, they have spokespersons that more or less

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



153

match the portfolios of the ministers. It means that they are able to represent their party in negotiations on specific areas. They also have a party group leadership that will help ensure that the party fulfils its obligations to the government and other parties when it has made a legislative bargain. Legislative bargaining between the government parties and other parties may result not only in an agreement to commit their party groups to pass one or more bills but also in more durable legislative agreements (forlig) (Christiansen 2008). These last for a specified number of years, or until cancelled or replaced. Yet, according to informal rules of legislative agreements, such a cancellation cannot have effect until after a general election. Furthermore, any political party that takes part in it holds veto power for changes in what the parties have agreed to (Christiansen 2008). After 2000 legislative agreements cover more than 30 per cent of government bills passed. At the end of the 2019 electoral term, about 120 such agreements were valid, and most of them with a broad group of parties preserving policies across the election (Finansministeriet 2019). Some legislative agreements cover multiple topics, and they are negotiated by the top leaderships of the parties whether in office or not. The annual negotiations over the state budget bill provides an important occasion for such bargaining that covers numerous topics beyond pure economics, a procedure that began in the 1980s. It developed further during the 1990s when the minister of finance excelled in putting together ‘patchwork agreements’ with different parties covering different topics within one larger agreement structure. According to Kristian Thulesen Dahl, who was finance spokesperson until 2012 and since then party leader for the Danish People’s Party, his party achieved much of its political influence as support party between 2001 and 2011 through these negotiations that also maintained an important political role for the minister of finance (Christiansen 2012). Ministers meet more or less regularly with the spokesperson of the parties within the agreement. As partners, they get easier access to information from the government and could ask questions through this channel. These features of coordination and control with non-cabinet parties resemble coalition governance between cabinet parties proper (Christiansen 2008). They provide opposition parties with a ‘quasi-governmental’ status as coalition partners on a policy issue (Klemmensen 2005). The history of legislative agreements dates back more than one century to the formative years of the party system (Pedersen 2011). They build on informal rules, so they are mentioned neither in the constitution nor in the parliamentary order. Most parties respect the norms; violations are rare and get sanctioned (Christiansen 2008). Arguably, the legislative agreements provide the most important precondition for making minority governance work in Denmark. The political parties know this system exists before they decide to enter a government or not and should be considered part of the institutional setting of coalition

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

154



formation and governance in Denmark in a broader sense. Most of the interviewed former ministers mentioned their importance.

Cabinet duration and termination The duration of cabinets The 14 Danish coalition cabinets between 1988 and 2016 were to remain in office for about 70 per cent, or a little more than two-thirds, of their possible duration. This level is higher compared with in particular the 1970s and 1980s (Damgaard 2000: 243). One possible explanation could be that parties on the political wings have become more included in legislative bargaining both because they have become less radical compared to the 1970s and 1980s and because the mainstream parties would let them in (Green-Pedersen 2005; Christiansen 2018). Some governments were also behind the opposition in the opinion polls and waited for the last moment (cf. Strøm and Swindle 2002).

The termination of cabinets The Folketing has a maximum term of four years. The prime minister is responsible for calling an election before the end of the term. This means that technically all elections have been ‘early elections’, although in 2019 it was just 12 days prior to the end of the term. Between 1971 and 1990, there were early elections almost every second year, most of them due to a conflict with parliament (cf. Becher and Christiansen 2015). Since then the electoral terms with one exception have ended a few months prior to the end of the electoral term (see Table 5.6) except in 2007 when the prime minister, Anders Fogh Rasmussen, called an early election 15 months prior to the end of the term. None of the elections after 1990 was called due to conflicts with the Folketing. Of the 14 cabinets terminated between 1988 and 2019, 9 ended with an election; see Table 5.6. The election of 2007 was constitutionally required for the revision of the Succession to the Throne Act to proceed but the election could have waited until the end of the ordinary term. Five times the prime minister called an election with six months or less remaining for the end of the electoral term (1994, 2001, 2011, 2015, and 2019). For two further elections (2005 and 2007) the timing seemed tactically motivated, and not from conflict with parliament. The 1988 and 1990 elections were due to a conflict between the government and parliament, something quite common in the 1970s and 1980s but not since then. Of the five terminations for reasons other than election, one was because the prime minister resigned to become general secretary of NATO (2009). The

Date in

1988-06-03

1990-12-18

1993-01-25

1994-09-27

1996-12-30

1998-03-11

2001-11-27

2005-02-08

Cabinet

Schlüter IV

Schlüter V

Rasmussen I

Rasmussen II

Rasmussen III

Rasmussen IV

Fogh Rasmussen I

Fogh Rasmussen II

2007-11-13

2005-02-08

2001-11-20

1998-03-11

1996-12-30

1994-09-21

1993-01-15

1990-12-12

Date out

69

80.4

92.4

69.2

56.7

88

52.2

64.2

4

4

1

4

7a

1

6

4

(11)

11

11

14

13

CD, S, RV

Relative Mechanisms Terminal Parties duration of cabinet events (when (%) termination conflict between or within)

Table 5.6 Cabinet termination in Denmark, 1988–2018

Finance, Economy, Taxation Justice

Policy area(s)

Continued

Elections six months before end of term An early election was called, four months before end of term, possibly to capitalize on a surge of popularity following Rasmussen’s support for the US invasion of Iraq. Early elections due to good polling. Elections 15 months before end of term, possibly to benefit from

Elections called to improve government’s bargaining position (economic policy) Prime minister decided to resign (possibly anticipating defeat in parliament because of ‘scandal’ (‘Tamilsagen’)) Elections three months before end of term. CD wanted to leave the coalition.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Date in

2007-11-13

2009-04-05

2011-10-03

Cabinet

Fogh Rasmussen III

Løkke Rasmussen I

Thorning-Schmidt I

Table 5.6 Continued

2014-02-03

2011-09-15

2009-04-05

Date out

59.2

92.8

34.8

8

1

9 11

14

SF

Relative Mechanisms Terminal Parties events (when duration of cabinet conflict termination (%) between or within)

Policy area(s)

government popularity due to strong economic performance and low unemployment. The election was also a constitutional precondition for passing an amendment to the Succession to the Throne Act introducing full primogeniture gender equality. Fogh Rasmussen resigned to take over as the head of NATO (nonpolitical). Early elections two months before end of term. The government had just reached an agreement with DFP on economic reforms, and polling improving for the government parties, although the government still went into the campaign behind in the polls. SF withdrew from the cabinet due to internal party instability over the sale of government shares in DONG Energy.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2016-11-28

2019-06-05

2015-06-28

Løkke Rasmussen II

Løkke Rasmussen III 2016-11-28

98.7

35.8

84.9

1

5

1

11

Note: Relative duration: share of constitutionally allowed duration for this cabinet.

Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

2015-06-18

Thorning-Schmidt II 2014-02-03

Early elections called, three months before end of term, possibly to capitalize on government popularity due to strong economic performance. Government enlargement as a result of bargaining over major economic reform had failed, and there was a threat from LA to topple the government over tax policies. The government enlargement may have prevented early election or government resignation. Election 12 days before end of term.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

158



remaining four involved an element of party conflict. One time a government got enlarged with more parties than previously (2016), and two times one party stepped out of the government after internal disagreements (1996 and 2014) but in both cases the party still supported the government. Finally, one government resigned without calling an election and got replaced by the opposition (1993).

Conclusion Danish governments form and govern in an institutional setting with a limited number of constitutionally binding rules, including most importantly the principle of negative parliamentary government, meaning that a government should resign or call an election if there is majority against it in parliament. Derived from that, a government cannot take office if there is a majority against it from the beginning. Government formations aim for expediency and for giving different government constellations a chance to form while respecting the principles of parliamentary government. They usually result in minority governments and over the recent decades primarily in coalitions. Governments form within ideological ‘blocs’ clear to the voters ahead of the election but not across blocs. The old and established mainstream parties still dominate government formation. Until 2001, the Social Liberal Party very often found itself in a pivotal position and shifted sides between the blocs from time to time. Yet since then the ‘blue bloc’ has been able to win office without the support of the Social Liberals, largely due to increased voter support to the populist Danish People’s Party. De jure, Danish governments follow ‘the Ministerial Government Model’ with full autonomy for each minister to decide, hence providing an influence over certain topics for each cabinet party. There are still topics left to decide for individual ministers. Yet, in particular since the 1990s, an increased hierarchical coordination in cases of disagreement took place at the top level of the government. Despite minority government, coalition agreements became more frequent from the 1990s onwards. Furthermore, inside the cabinets, a number of changes stressed hierarchical and centralized modes of coordination and monitoring. In particular, the economic and coordination committees now function as inner cabinets that work to resolve conflicts among the parties, the former dealing with the highest amounts of matters, and the latter dealing with the most conflictual matters. Most important and new decisions originate in the coalition agreement or pass through the central committees, providing party leaders with agenda setting and veto powers, meaning that Denmark is today best classified as following ‘the Coalition Compromise Model’. Yet, with powers of the Ministry of Finance, and of the prime minister, to monitor and control the individual ministries, the largest cabinet party, which usually controls both offices, has

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



159

elements of influence close to the ‘Dominant Prime Minister Model’ with the corollary that it has often been a duopoly of those ministers rather than the prime minister only (Rhodes and Salomonsen forthcoming). Governance and stability in the life of Danish minority governments also comes from coordination with opposition parties to build majorities in parliament. Here, legislative agreements are most important since they bind the parties over time and involve opposition parties as ‘quasi’-governmental coalition partners within their specific policy items. After turbulent decades of the 1970s and 1980s with many parties, frequent elections, and relatively short-lived governments, Danish cabinets became more stable since the 1990s. The number of parties in parliament remains high, and the governments are still of the minority type. Yet the governments have become more durable, there are few early elections, and few governments resigned for political reasons. This notwithstanding, if political conditions change, with more and stronger parties unwilling to compromise, as in the 1970s, there are only few formal institutions to prevent such a development.

References Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999-2010’. Party Politics, 21(1): 142–52. Bale, Tim, and Torbjörn Bergman (2006). ‘Captives No Longer, but Servants Still? Contract Parliamentarism and the New Minority Governance in Sweden and New Zealand’. Government and Opposition, 41(3): 422–49. Becher, Michael, and Flemming J. Christiansen (2015). ‘Dissolution Threats and Legislative Bargaining’. American Journal of Political Science, 59(3): 641–55. Bendtsen, Bendt (2019). Fra bondedreng til vicestatsminister. Erindringer. Odense: mellemgaard. Bergman, Torbjörn (1993). ‘Formation Rules and Minority Governments’. European Journal of Political Research, 23(1): 55–66. Bille, Lars (1997). Partier i forandring. En analyse af danske partiorganisationers udvikling 1960-1995. Odense: Odense Universitetsforlag. Bille, Lars (1998). Dansk partipolitik 1987–1998. Copenhagen: Jurist- og Økonomforbundets Forlag. Borchsenius, Søren (2004). ‘SV-regeringens forudsætninger og forløb’. Historisk Tidsskrift, 104(1): 122–61. Borre, Ole (1995). ‘Old and New Politics in Denmark’. Scandinavian Political Studies, 18(3): 187–205. Carroll, Royce, and Gary W. Cox (2012). ‘Shadowing Ministers: Monitoring Partners in Coalition Governments’. Comparative Political Studies, 45(2): 220–36.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

160



Christensen, Jens P. (2019). ‘Regeringsdannelse’. Ugeskrift for retsvæsen, 2019(12): 89–99. Christensen, Jørgen G. (1985). ‘In Search of Unity: Cabinet Committees in Denmark’. In Thomas T. Mackie and Brian W. Hogdwood (eds), Unlocking the Cabinet: Cabinet Structures in Comparative Perspective. London: Sage, 114–37. Christiansen, Flemming J. (2008). Politiske forlig i Folketinget. Partikonkurrence og samarbejde. Aarhus: Politica. Christiansen, Flemming J. (2012). ‘Raising the Stakes. Passing State Budgets in Scandinavia’. World Political Science, 8(1): 184–200. Christiansen, Flemming J. (2018). ‘Denmark: Strengthened Opposition, Yet High Levels of Cooperation’. In Elisabetta De Giorgi and Gabriella Illonszki (eds), Opposition Parties in European Legislature: Conflict or Consensus? London: Routledge, 17–34. Christiansen, Flemming J. (2020). ‘The Liberal Party: From Agrarian and Liberal to Centre-Right Catch-All’. In Peter M. Christiansen, Jørgen Elklit, and Peter Nedergaard (eds), The Oxford Handbook of Danish Politics. Oxford: Oxford University Press, 296–312. Christiansen, Flemming J., and Helene H. Pedersen (2014). ‘Minority Coalition Governance in Denmark’. Party Politics, 20(6): 940–9. Christiansen, Flemming J., and Henrik B. Seeberg (2016). ‘Cooperation between Counterparts in Parliament from an Agenda-Setting Perspective: Legislative Coalitions as a Trade of Criticism and Policy’. West European Politics, 39(6): 1160–80. Christiansen, Flemming J., Rasmus L. Nielsen, and Rasmus B. Pedersen (2014). ‘Friendship, Courting, and Engagement: Pre-Electoral Coalition Dynamics in Action’. Journal of Legislative Studies, 20(4): 413–29. Dahl, Robert A. (1971). Polyarchy. Participation and Opposition. New Haven: Yale University Press. Damgaard, Erik (1969). ‘The Parliamentary Basis of Danish Governments: The Patterns of Coalition Formation’. Scandinavian Political Studies. Yearbook, 4: 35–66. Damgaard, Erik (1974). ‘Stability and Chance in the Danish Party System over Half a Century’. Scandinavian Political Studies. Yearbook, 9: 103–25. Damgaard, Erik (2000). ‘Denmark: The Life and Death of Government Coalitions’. In Wolfgang C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 231–63. Damgaard, Erik (2011). ‘Change and Challenges of Danish Parliamentary Democracy’. In Torbjörn Bergman and Kaare Strøm (eds), Madisonian Turn. Political Parties and Parliamentary Democracy in Nordic Europe. Ann Arbor: University of Michigan Press, 67–111. Debus, Marc (2009). ‘Pre-electoral Commitments and Government Formation’. Public Choice, 138(1–2): 45–64.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



161

Elklit, Jørgen (1984). ‘Det klassiske danske partisystem bliver til’. In Jørgen Elklit and Ole Tonsgaard (eds), Valg og vælgeradfærd. Aarhus: Forlaget Politica, 21–38. Finansministeriet (2019). Oversigt over gældende forlig pr. 1. april 2019. ERU, alm. del, bilag 212. Retrieved from: https://www.ft.dk/samling/20181/almdel/ERU/bilag/212/ 2045907/index.htm (Accessed on 8 September 2019). Folketingets udvalg i Landstingssalen den 4. februar 2015. Copenhagen: Folketinget. Retrieved from: https://www.ft.dk/samling/20141/almdel/UFO/bilag/35/1520312/ index.htm (Accessed on 8 September 2019). Fuglsang, Niels, and Lotte Jensen (2010). ‘Magtens mekanismer – Regeringens udvalgsstruktur og koordineringspraksis under Nyrup og Fogh’. Samfundsøkonomen, (2): 17–23. Germer, Peter (1988). Statsforfatningsret I. Copenhagen: Jurist- og Økonomforbundets Forlag. Green-Pedersen, Christoffer (2005). ‘Folketinget under fortsat forandring’. Politica, 37 (4): 394–410. Green-Pedersen, Christoffer, and Jesper Krogstrup (2008). ‘Immigration as a Political Issue in Denmark and Sweden’. European Journal of Political Research, 47(5): 610–34. Green-Pedersen, Christoffer, and Lisbeth H. Thomsen (2005). ‘Bloc Politics vs. Broad Cooperation? The Functioning of Danish Minority Parliamentarism’. Journal of Legislative Studies, 11(2): 153–69. Hartling, Poul (1981). Politisk udspil. Erindringer 1964-1971. Copenhagen: Spektrum. Jensen, Jørgen Albæk (1989). Parlamentarismens statsretlige betydning. Det parlamentariske princips betydning for forholdet mellem Folketing og regering. Copenhagen: Jurist- og Økonomforbundet Forlag. Jensen, Lotte (2003). Den store koordinator. Finansministeriet som moderne styringsaktør. Copenhagen: Jurist- og Økonomforbundet Forlag. Jensen, Lotte (2008). Væk fra afgrunden. Finansministeriet som økonomisk styringsaktør. Odense: Syddansk Universitetsforlag. Kaarsted, Tage (1968). Påskekrisen 1920. Aarhus: Universitetsforlaget i Aarhus. Kaarsted, Tage (1988). Regeringen, vi aldrig fik. Regeringsdannelsen 1975 og dens baggrund. Odense: Odense Universitetsforlag. Klemmensen, Robert (2005). ‘Forlig i det danske folketing 1953-2001’. Politica, 37(4): 440–52. Knudsen, Tim (1995). Dansk statsbygning. Copenhagen: Jurist- og Økonomforbundet Forlag. Knudsen, Tim (2000). Regering og embedsmænd. Om magt og demokrati i staten. Århus: Systime. Knudsen, Tim, and Hanne Rasmussen (2012). ‘Den underlige kongesøger – om regeringsdannelsen’. Økonomi og Politik, 85(3): 75–84.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

162



Kongehuset.dk (2015a). Forhandlingsleder udpeget, 19 June. Retrieved from: http:// kongehuset.dk/menu/presse/meddelelser/forhandlingsleder-udpeget (Accessed on 8 September 2019). Kongehuset.dk (2015b). Forhandlinger gennemført - Regeringsdannelse, 27 June. Retrieved from: http://kongehuset.dk/menu/presse/meddelelser/forhandlingergennemfort-regeringsdannelse (Accessed on 17 January 2020). Laver, Michael, and Kenneth A. Shepsle (1996). Making and Breaking Governments. Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Lykketoft, Mogens (2019). Mogens Lykketofts erindringer. Holdning og handlinger. Copenhagen: Rosinante. Martin, Lanny W., and Georg Vanberg (2008). ‘Coalition Government and Political Communication’. Political Research Quarterly, 61(3): 502–16. Mortensen, Hans (2014). SF. Magtens pris. Copenhagen: People’s Press. E-book. Müller, Wolfgang C., and Kaare Strøm (eds) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Nielsen, Peter H. (2014). ‘Det utaknemmelige statsministerembede’. Unpublished master’s thesis, Copenhagen: Copenhagen University. Nielsen, Peter H. (2016). Fra kronprinsesse til statsministerkandidat. Historien om Mette Frederiksen og Socialdemokratiet siden 2001. Odense: mellemgård. Nielsen, Peter H. (2017). ‘Danske regerings ansættelsesudvalg 1977-2017’. Historisk Tidsskrift, 117(2): 363–98. Olesen, Niels W. (2018). Poul Schlüters tid 1982-1993. De danske ministerier. Copenhagen: Gads Forlag. Olesen, Niels W., and Thorsten B. Olesen (2018). ‘Anker Jørgensens og Poul Schlüters tider 1972-1993’. In Niels W. Olesen (ed.), Poul Schlüters tid 1982-1993. De danske ministerier. Copenhagen: Gads Forlag, 569–600. Olesen, Thorsten B. (2017). Anker Jørgensens tid 1972-1982. De danske ministerier. Copenhagen: Gads Forlag. Olsen, Søren-Ole (1999). ‘Regeringsarbejdet’. In Ole S. Andersen, Erik Damgaard, Claus Dethlefsen, Klavs Gravesen, Niels E. Holm, Henrik Jensen, Tim Knudsen, Sysette V. Kruse, Peer Lorenzen, Lars A. Rehof, Karen Siune, Palle Svensson, Ditlev Tamm, and Henrik Zahle (eds), Folketingets festskrift i anledning af Grundlovens 150 års jubilæum 5. juni 1999. Copenhagen: Folketingets Præsidium, 73–101. Pedersen, Helene H. (2011). ‘Etableringen af politiske forlig som parlamentarisk praksis’. Politica, 43(1): 48–67. Pedersen, Mogens N. (1987). ‘The Danish “Working Multiparty System”: Breakdown or Adaptation’. In Hans Daalder (ed.), Party Systems in Denmark, Austria, Switzerland, the Netherlands, and Belgium. London: Frances Pinter, 1–60. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



163

Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of AntiElitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’, Research & Politics, (January-March), 1–9. Rasch, Bjørn E. (2011). ‘Why Minority Governments? Executive-Legislative Relations in the Nordic Countries’. In Thomas Persson and Matti Wiberg (eds), Parliamentary Government in the Nordic Countries at a Crossroads. Coping with Challenges from Europeanisation and Presidentialisation. Stockholm: Santérus Academic Press, 41–61. Rhodes, Rod A. W., and Heidi H. Salomosen (forthcoming). ‘Doupoly, Court Politics, and the Danish Core Executive’, Public Administration. Schlüter, Poul (1998). Sikken et liv. Erindringer. Copenhagen: Aschehoug. Seeberg, Henrik B., and Ann-Kristin Kölln (2020). ‘Red-Green Alliance: Is it Red or Green?’. In Peter M. Christiansen, Jørgen Elklit, and Peter Nedergaard (eds) The Oxford Handbook of Danish Politics. Oxford: Oxford University Press, 329-46. Skjæveland, Asbjørn (2001). ‘Party Cohesion in the Danish Parliament’. Journal of Legislative Studies, 7(2): 35–56. Skjæveland, Asbjørn (2003). Government Formation in Denmark 1953-1998. Aarhus: Politica. Skjæveland, Asbjørn (2019). ‘Unicameralism in Denmark: Abolition of the Senate, Current Functioning and Debate’. In Nikolaj Bijleveld, Colin Grittner, David E. Smith, and Wybren Verstegen (eds), Reforming Senates. Upper Legislative Houses in North Atlantic Small Powers 1800-Present. Abingdon: Routledge, 225–38. Smith, Bo (2015). Embedsmanden i det moderne folkestyre. Copenhagen: Jurist- og Økonomforbundets Forlag. Strøm, Kaare (1990). Minority Government and Majority Rule. Cambridge: Cambridge University Press. Strøm, Kaare, Wolfgang C. Muller, and Torbörn Bergman (eds.) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Strøm, Kaare, and Stephen M. Swindle (2002). ‘Strategic Parliamentary Dissolution’. American Political Science Review, 96(3): 575–91. Stubager, Rune (2010). ‘The Development of the Education Cleavage: Denmark as a Critical Case’. West European Politics, 33(3): 505–33. Warwick, Paul V., and James N. Druckman (2001). ‘Portfolio Salience and the Proportionality of Payoffs in Coalition Governments’. British Journal of Political Science, 31(4): 627–49.

Appendix. List of political parties Abbreviation DKP FK

Name Communist Party of Denmark (Danmarks Kommunistiske Parti) Common Course (Fælles Kurs)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

164 EL VS SF S RV RF LC CD KrF KF V UN FrP DFP LA Alt

 Unity List (Enhedslisten) Left Socialists (Venstresocialisterne) Socialist People’s Party (Socialistisk Folkeparti) Social Democrats (Socialdemokraterne) Danish Social Liberal Party (Radikale Venstre) Justice Party of Denmark (Retsforbundet) Liberal Centre (Liberalt Centrum) Centre Democrats (Centrum-Demokraterne) Christian Democrats (Kristendemokraterne) Conservative People’s Party (Det Konservative Folkeparti) Liberal Party (Venstre) Independent Party (De Uafhængige) Progress Party (Fremskrittspartiet) Danish People’s Party (Danske Folkepartiet) Liberal Alliance (Liberal Alliance) The Alternative (Alternativet)

Note: Party names are given in English, followed by the party name in Danish in parentheses.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 6 Finland: Forming and Managing Ideologically Heterogeneous Oversized Coalitions Tapio Raunio

The Finnish political system has experienced quite a radical transformation since the late 1980s. The new constitution, which entered into force in 2000,¹ completed a period of far-reaching constitutional change that curtailed presidential powers and brought the Finnish regime closer to a classical form of parliamentary democracy. The president is today almost completely excluded from the policy process in domestic matters, and leadership by presidents has been replaced with leadership by strong majority governments. The strengthening of parliamentary democracy also means that political parties are in a much more central position. Political parties and their leaders form governments that are accountable to the Eduskunta, the unicameral national legislature, and not to the president as was the case until the 1980s. The prime minister (PM) is now the political leader of the country, but the PM is strongly constrained by the types of cabinet formed in Finland. Heading oversized, ideologically heterogeneous governments, PMs must strike a balance between active leadership and accommodating the preferences of the coalition partners. As essentially all of the politicians interviewed for this chapter underlined, the key to managing such broad coalitions is building and maintaining trust among the governing parties. Hence coalition governance in Finland certainly falls in the category of coalition compromise model outlined in Chapter 2 of this volume. Examining coalition governance in Finland, this chapter argues that due to the ideological heterogeneity of cabinets, PMs and governments emphasize the importance of ex ante mechanisms.² Government programmes have become ¹ The Constitution of Finland, 11 June 1999 (731/1999). An English version is found at https://www. finlex.fi/fi/laki/kaannokset/1999/en19990731.pdf, accessed 17 March 2021. ² In addition to the data compiled in this project, the chapter draws on various official documents such as the constitution, laws, government’s rules of procedure, coalition agreements, and 11 interviews with individuals who have served as both ministers and MPs and have considerable experience of government formation processes. Left Alliance, 14 March 2018; SDP, 13 March 2018; Green League, 12 March 2018; SDP, 10 March 2018; National Coalition, 29 March 2018; SDP, 18 March 2018; Green League, 25 April 2018; National Coalition, 15 March 2018; SDP, 9 March 2018; Centre Party, 13 March Tapio Raunio, Finland: Forming and Managing Ideologically Heterogeneous Oversized Coalitions In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0006

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

166



very detailed and there are both established written rules and informal conventions for cabinet decision-making and for solving conflicts inside the government. Coalition governance in Finland is also characterized by stability: the existing practices have remained basically unaltered at least since the mid-1990s, and the rise of populism has not changed how cabinets are formed or work. PMs, the coalition partners, and their parliamentary groups know and mainly respect the rules of the game, and this contributes to the survival of Finnish broad cabinets.

The institutional setting Before the constitutional reform in 2000, the comparative literature traditionally categorized the Finnish political system as semi-presidential, with the executive functions divided between an elected president and a cabinet that is accountable to the parliament. In fact, Finland is the oldest semi-presidential country in Europe, with the semi-presidential form of government adopted in 1919, two years after the country achieved independence. Under the old constitution, the president was recognized as the supreme executive power, particularly so in foreign affairs. For example, Duverger (1980) ranked Finland highest among the West European semi-presidential systems in terms of the formal powers of the head of state and second only to France with respect to the actual exercise of presidential power. The peak of presidential powers was reached during the long reign of President Urho Kekkonen (1956–1981), who made full use of his powers and arguably even overstepped the constitutional prerogatives of the presidency. During the Cold War, the balance between cabinet and president was therefore strongly in favour of the president until the constitutional reforms enacted from the late 1980s onwards, which were in part a response to the excesses of the Kekkonen era. The Constitution Act of 1919 was virtually silent on cabinet formation. In practice, the process was strongly dominated by the president from the 1950s to the late 1980s. After the outgoing cabinet had submitted its resignation, the president invited the Speaker of Eduskunta and the representatives of the parliamentary parties to bilateral discussions. The fragmented party system, with no clearly dominant party, strengthened the president’s hand in steering the negotiations. The president then appointed a formateur, but this person knew that the government needed the approval of the president. It was also common for the president to influence the selection of individual ministers. The process ended with the president appointing the new cabinet in the last full plenary meeting of

2018; and the Swedish People’s Party, 9 March 2018. Petra Kantola and Anna-Riikka Aarnio provided valuable research assistance. I am also grateful to Heikki Paloheimo for sharing his knowledge of Finnish governments.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



167

the resigning cabinet. The president also appointed caretaker cabinets consisting of civil servants: since 1945 Finland has had six caretaker governments, most recently the Liinamaa cabinet in 1975.³ The last case of presidential intervention in cabinet formation occurred in 1987 when President Mauno Koivisto overruled a coalition between the Centre Party and the National Coalition, indicating that a coalition between the National Coalition and the Social Democrats was preferable. The role of the president is now limited to formally appointing the PM and the cabinet chosen by Eduskunta.⁴ Hence cabinet formation is based on bargaining between political parties, with the understanding that the largest party (in terms of Eduskunta seats) will lead the negotiations. The Eduskunta then selects the PM and later votes on the entire cabinet through the investiture vote, which was first used in 1995 when the rainbow coalition headed by Paavo Lipponen (SDP) took office. Under the new constitution, the cabinet shall without delay submit its programme to the parliament in the form of a statement, which is then followed by a debate and a mandatory investiture vote. The decision rule is simple majority. There are no constitutional regulations about the number of ministers or how they are to be selected. Compared with the era of the old constitution, government formation is thus nowadays completely in the hands of the political parties (Karvonen 2016). The president can no longer dissolve the Eduskunta or force the government or an individual minister to resign.⁵ Until the 1990s the president alone had the right to dissolve the Eduskunta and order new elections, and he was not obliged to consult the cabinet or parliament before doing so. The president exercised this right four times during the post-war era (1953, 1962, 1971, and 1975). A constitutional amendment in 1991 altered the situation in favour of the cabinet, by requiring the explicit consent of the PM for parliamentary dissolution. Section 26 of the new constitution consolidated this practice: ‘The President of the Republic, in response to a reasoned proposal by the Prime Minister, and after ³ In addition, about two-thirds of the remaining post-war cabinets have included non-partisan ministers. ⁴ Section 61 of the constitution: ‘The Parliament elects the Prime Minister, who is thereafter appointed to the office by the President of the Republic. The President appoints the other Ministers in accordance with a proposal made by the Prime Minister. Before the Prime Minister is elected, the groups represented in the Parliament negotiate on the political program and composition of the Government. On the basis of the outcome of these negotiations, and after having heard the Speaker of the Parliament and the parliamentary groups, the President informs the Parliament of the nominee for Prime Minister. The nominee is elected Prime Minister if his or her election has been supported by more than half of the votes cast in an open vote in the Parliament. If the nominee does not receive the necessary majority, another nominee shall be put forward in accordance with the same procedure. If the second nominee fails to receive the support of more than half of the votes cast, the election of the Prime Minister shall be held in the Parliament by open vote. In this event, the person receiving the most votes is elected.’ ⁵ According to Section 64 of the constitution: ‘The President of the Republic grants, upon request, the resignation of the Government or a Minister. The President may also grant the resignation of a Minister on the proposal of the Prime Minister. The President shall in any event dismiss the Government or a Minister, if either no longer enjoys the confidence of Parliament, even if no request is made.’

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

168



having heard the parliamentary groups, and while the Parliament is in session, may order that extraordinary parliamentary elections shall be held. Thereafter, the Parliament shall decide the time when it concludes its work before the elections.’ Governments are thus now accountable to the Eduskunta and not to the president, as effectively was the case before. The president has only an ineffective delaying power in legislation, and even the appointment powers of the president have been drastically reduced. Overall, the president is almost completely excluded from the policy process in domestic matters. The government is responsible for European Union (EU) affairs, with foreign policy leadership shared between the president and the government. While jurisdictional disputes can emerge, a division of labour seems to have emerged: the government is responsible for those foreign and security policy issues handled through the EU, whereas the president focuses on bilateral ties with non-EU countries, especially those led by presidents, notably Russia. (Raunio and Sedelius 2020) Foreign and defence policy excluded, Finland is now thus effectively a parliamentary regime. The end of the Cold War removed the shadow of the Soviet Union from Finnish policy-making, but particularly through EU membership Finland has become much more involved in global and regional integration. A peculiar instrument of deferment rule also used to influence government formation and legislature–executive relations. It explained the propensity to form oversized coalitions and contributed to the practice of inclusive, consensual decision-making that reduced the gap between the government and opposition. Until 1987, one-third of the members of parliament (MPs) (67/200) could postpone the final adoption of an ordinary law over the next election, with the proposal adopted if a majority in the new parliament supported it. In 1987 the period of postponement was shortened to until the next annual parliamentary session, with the mechanism finally abolished in 1992. The rationale behind the deferment rule was that it would prevent tyranny by a simple parliamentary majority, offering in particular protection against potential radical socialist reforms (Forestiere 2008). Turning to other domestic constraints, Finland remains a strongly corporatist country. Corporatism was particularly prevalent in the 1970s and 1980s, and trade unions and employers’ confederations still wield considerable influence through collective wage bargaining and decision-making in a broad range of labour market issues. The changing nature of industrial relations is also related to changes in party-political cooperation. Until the 1970s the Finnish labour market was characterized by frequent strikes and work stoppages, but following the first comprehensive incomes policy agreement (1968) this conflictual style was gradually replaced with a more conciliatory approach to labour market issues. Such consensual practices spread also to the party system, with more pragmatic cooperation between parties of the right and the left. A broad consensus also emerged in favour of the welfare state. These changes contributed to the increased potential

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



169

for different kinds of coalition cabinets and also paved the way for the subsequent cabinets formed between the Social Democrats and the National Coalition. National referendums, which are consultative, have been used only twice: on the prohibition of alcohol in 1931 and on EU membership in 1994. A new constitutional amendment (2012) strengthened direct democracy by introducing the citizens’ initiative. At least 50,000 signatures are needed to submit an initiative for a new law to the Eduskunta. The mechanism has proved popular and it has facilitated debates about issues that might otherwise not rise to the political agenda—such as same-sex marriages, ban on fur farming, and the status of Swedish as a compulsory school subject. The popularity of the citizens’ initiative has also resulted in debates about how to combine electoral democracy and ‘between-elections’ democracy. To sum up, through the constitutional reforms outlined here, Finland has become a strongly government-driven polity. Finnish cabinets obviously operate in the context of EU membership and other international rules and commitments, but with the exception of corporatism, the cabinets do not face major domestic constraints. Presidential leadership has been replaced with leadership by strong oversized coalitions, which have ruled without much effective opposition since the 1980s. However, as argued in the next sections, the fragmented party system and the ensuing need to build and maintain heterogeneous coalitions act as important moderating factors on the powers of the PM (Nousiainen 2001; Raunio 2011; Karvonen 2014; Karvonen et al. 2016).

The party system and the actors In a comparative perspective, there are five intertwined features—high degree of party system fragmentation, increased weakness of the left, the strength of the Centre Party, waves of populist protest, and changing cleavage structure—that appear characteristic of the Finnish party system and which are also relevant for understanding government formation and coalition governance (Table 6.1a) (Paloheimo and Raunio 2008; Arter 2009; Bengtsson et al. 2014; Karvonen 2014).

A high degree of party system fragmentation Since the declaration of independence in 1917 no party has even come close to winning a majority of parliamentary seats (the post-Second World War high is 28.3 per cent won by SDP in the 1995 elections), and this fragmentation contributes to cooperation between the main parties. It also means that there is no party that would be decisively larger than its competitors. Forming majority cabinets is simply not possible unless the government has at least three parties.

Kekkonen III Kekkonen IV Tuomiojaabc Törngren Kekkonen V Fagerholm II

Fagerholm III 1957-05-17 Sukselainen I 1957-05-27

Sukselainen II Sukselainen III von Fieandtabc Kuuskoskiabc

6 7 8 9 10 11

12 13

14 15

16 17

Fagerholm Kekkonen I Kekkonen II

3 4 5

1957-11-29 1958-04-26

1957-07-02 1957-09-02

1951-09-20 1951-07-03 1953-07-09 1953-11-17 1954-05-05 1954-03-08 1954-10-20 1956-03-03

1948-07-29 1948-07-02 1950-03-17 1951-01-17

1946-03-26

Pekkala

2

Election date

1945-04-17 1945-03-18

Paasikivi IIIa

Date in

1

Cabinet Cabinet number

Table 6.1a Finnish cabinets since 1945

NN, SDP, SKDL, KESK, LKP SKDL, KESK, SDP, RKP SDP KESK, LKP, RKP KESK, SDP, LKP, RKP KESK, SDP, RKP KESK, RKP NN RKP, SDP, KESK KESK, SDP SDP, KESK, RKP, LKP SDP, KESK, RKP KESK, RKP, Suomen Kansanpuolue (LKP), Liberaalinen Kansanpuolue (LKP) KESK, LKP KESK, TPSL, LKP, SDP NN NN

Party composition of cabinet

non non

min min

sur min

sur min non sur mwc sur

min min sur

sur

sur

0 0

33 42.5

60 39.5

59 33 0 60 53.5 66.5

27 37.5 64.5

81

85.5

200 200

200 200

200 200

200 200 200 200 200 200

200 200 200

200

200

7 7

6 7

6 6

6 6 6 6 6 6

6 6 6

6

6

5.35 5.35

4.76 5.35

4.76 4.76

4.8 4.83 4.83 4.73 4.76 4.76

4.56 4.59 4.59

4.78

4.78

Type Cabinet Number of Number of ENP, parties in parliament of strength seats in cabinet in seats parliament parliament (%)

KESK KESK

KESK KESK

KESK KESK

KESK KESK KESK KESK KESK KESK

KESK KESK KESK

KESK

KESK

Formal Median support party in first policy parties dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Karjalainen II 1963-10-18

1968-03-22

1970-05-14 1970-03-16 1970-07-15

1971-10-29 1972-02-23 1972-01-03 1972-09-04

1975-06-13 1975-11-30 1975-09-22

Lehtoabc Virolainen

Paasio I

Koivisto I

Aura Iabc Karjalainen III Karjalainen IV Aura IIabc Paasio II Sorsa I

Liinamaaabc Miettunen II

Miettunen III 1976-09-29 Sorsa II 1977-05-15

22

23 24

25

26

27 28

33 34

35 36

30 31 32

29

20 21

Sukselainen IV Miettunen I Karjalainen I

19

1971-03-26

1966-05-27 1966-03-21

1963-12-18 1964-09-12

1961-07-14 1962-04-13 1962-02-05

1959-01-13

Fagerholm IV 1958-08-29 1958-07-07

18

KESK KESK, KOK, TPSL, RKP, LKP KESK, KOK, RKP, LKP NN KESK, KOK, RKP, LKP SDP, KESK, SKDL, TPSL SDP, KESK, SKDL, RKP, TPSL NN KESK, SDP, SKDL, LKP, RKP KESK, SDP, LKP, RKP NN SDP SDP, KESK, RKP, LKP NN KESK, SDP, SKDL, RKP, LKP KESK, LKP, RKP SDP, KESK, SKDL, LKP, RKP

SDP, KESK, KOK, RKP, LKP KESK, RKP

min sur

non sur

non min mwc

mwc

non sur

sur

sur

non mwc

mwc

min mwc

min

sur

29 76

0 75.5

0 27.5 53.5

54

0 72

82

76

0 56

55.5

24 56.5

31

66.5

200 200

200 200

200 200 200

200

200 200

200

200

200 200

200

200 200

200

200

8 8

8 8

8 8 8

8

8 8

8

8

7 7

7

7 7

7

7

5.32 5.32

5.56 5.32

5.57 5.52 5.56

5.57

5.57 5.57

5.42

5.42

5.23 5.23

5.2

5.52 5.2

5.52

5.49

KESK KESK

KESK KESK

KESK KESK KESK

KESK

KESK KESK

SDP

SDP

KESK KESK

KESK

SDP KESK

SDP

SDP

Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1978-03-02

Sorsa III

Koivisto II

Sorsa IV

Sorsa V

Sorsa VI

Holkeri I

Holkeri II Aho I

Aho II Lipponen I

Lipponen II

Lipponen III

Jäätteenmäki Vanhanen I

37

38

39

40

41

42

43 44

45 46

47

48

49 50

Election date

2003-04-17 2003-03-16 2003-06-24

2002-05-31

1999-04-15 1999-03-21

1994-06-28 1995-04-13 1995-03-19

1990-08-28 1991-04-26 1991-03-17

1987-04-30 1987-03-16

1983-05-06 1983-03-21

1982-12-31

1982-02-19

1979-05-26 1979-03-19

Date in

Cabinet Cabinet number

Table 6.1a Continued

SDP, KESK, SKDL, LKP SDP, KESK, SKDL, RKP SDP, KESK, SKDL, RKP SDP, KESK, RKP, LKP SDP, KESK, SMP, RKP KOK, SDP, RKP, SMP KOK, SDP, RKP KESK, KOK, RKP, SKL KESK, KOK, RKP SDP, KOK, RKP, SKDL, VIHR SDP, KOK, RKP, SKDL, VIHR SDP, KOK, RKP, SKDL KESK, SDP, RKP KESK, SDP, RKP

Party composition of cabinet

sur sur

sur

sur

mwc sur

sur sur

sur

sur

mwc

sur

sur

sur

58 58

64.5

69.5

53 72

60.5 57

65

61

51

66.5

66

71

200 200

200

200

200 200

200 200

200

200

200

200

200

200

8 8

8

8

9 8

8 9

8

8

8

8

8

8

4.93 4.93

5.15

5.15

5.25 4.89

4.87 5.25

4.87

5.37

5.26

5.26

5.22

5.32

Type Cabinet Number of Number of ENP, parties in parliament of strength seats in cabinet in seats parliament parliament (%)

KESK KESK

KESK

KESK

KESK KESK

KESK KESK

KESK

KESK

KESK

KESK

KESK

KESK

Formal Median support party in first policy parties dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Mari Kiviniemi Katainen I

Katainen II

Stubb I

Stubb II Sipilä I Sipilä II

52

54

55

56 57 58

2014-09-20 2015-05-29 2015-04-19 2017-06-13

2014-06-24

2014-03-25

2011-06-22 2011-04-17

2010-06-22

2007-04-19 2007-03-18

KESK, KOK, VIHR, RKP KESK, KOK, VIHR, RKP KOK, SDP, RKP, SKDL, VIHR, SKL KOK, SDP, RKP, VIHR, SKL KOK, SDP, RKP, VIHR, SKL KOK, SDP, RKP, SKL KESK, KOK, SMP KESK, KOK, SIN mwc mwc mwc

sur

sur

sur

sur

sur

51 62 53

56

55.5

62.5

62.5

62.5

200 200 200

200

200

200

200

200

8 8 9

8

8

8

8

8

5.81 5.84 6.46

5.81

5.81

5.83

5.13

5.13

VIHR KESK

VIHR

VIHR

VIHR

KESK

KESK

Notes: For a list of parties, consult the chapter appendix. a: Technocrat minister majority, b: Technocrat prime minister; c: Limited policy remit The number of parties in parliament does not include parties that have never held more than two seats when a cabinet has formed. Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm, Müller, and Bergman (2003), 1945–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015). The first policy dimension is economic left–right. Cabinet types: maj = Single-party majority cabinet; min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition; non = Non-partisan. Note: Table legends and variables are further defined in the Appendix, this volume.

53

Vanhanen II

51

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

174



In terms of party system development, the years after the Second World War can be roughly divided into two periods. First, until about 1970 the party system remained stable: class voting was high, electoral volatility was low, and practically no new parties entered the Eduskunta. As the class cleavage was crucial in the emergence of Finnish parties, it is not surprising that class dealignment has contributed to increasing electoral instability, in terms of both party system fragmentation and electoral volatility. However, despite the entry into the Eduskunta of new party families such as green, Christian, and populist parties, the party system has remained rather stable, with the three core parties—the Social Democrats, the liberal/agrarian Centre Party, and the conservative National Coalition—remaining dominant. The rise of the populist Finns Party has produced in the two latest Eduskunta elections (2011 and 2015) a situation where the party system has four about equally sized large parties. As a result, the Finnish party system is even more fragmented than previously.

An increased weakness of the parties on the left Whereas Social Democrats and the predecessor of Left Alliance, the Finnish People’s Democratic Union, won over 45 per cent of the vote between them in all but one election between 1945 and 1966 (when they won 48.3 per cent of the vote together), by 2015 the electoral strength of the left had decreased to 23.6 per cent. The prospect of a government consisting of only left-wing parties has not been realistic for several decades, and all cabinets formed after the 2003 elections have been led by centre-right parties. Social democracy has not been as strong in Finland as in the other Nordic countries, but SDP was the largest party in all Eduskunta elections held from 1907 to 1954, and since the 1966 elections it has finished first in all elections, apart from those held in 1991, 2003, 2007, 2011, and 2015. The Left Alliance, founded in 1990, was in government from 1995 to 2003 and again from 2011 to 2014, when after an uneasy three years it left the National Coalition-led cabinet due to differences over economic policy. The decline of the traditional left parties is to a certain extent compensated by the rise of the Green League, even though it refuses to be categorized as a left-wing party. It served in the government from 1995 to 2002, when it left the cabinet due to disagreements over nuclear energy, from 2007 to 2011, and from 2011 to 2014, when it again exited the cabinet over nuclear energy policy.

Persistent strength of the Centre Party The persistent strong support for the Centre Party, until 1965 the Agrarian Union, is a rare case of a survival of what is basically agrarian politics in the twenty-first

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



175

century. Aided by the lateness of the urbanization process and supported by a strong grassroots organizational network, the Centre remains by far the largest party in the overwhelming majority of rural municipalities. Located ideologically between the Social Democrats and the National Coalition on the left–right dimension, the Centre has formed coalitions with both left-wing and right-wing parties. The party has thus traditionally been a flexible bridge-builder in government formation.

Recurrent waves of populist protest The rise of populism has shaken party systems across Europe, and Finland is no exception. Already in 1970 and 1983 the populist Rural Party won a couple of spectacular victories in Eduskunta elections. Its successor, the Finns Party (originally known as the True Finns), achieved a major electoral breakthrough in the 2011 elections, winning 19.1 per cent of the votes, a staggering increase of 15 per cent from the 2007 elections and the largest ever increase in support achieved by a single party in Eduskunta elections. Many had predicted the Finns Party to fade away quickly, but in the 2015 elections it finished second with 38 seats and 17.7 per cent of the votes—a much better result than the polls had suggested, just like four years earlier. After the 2011 elections Timo Soini, the long-standing party chair, had opted to stay in the opposition, but in 2015 he guided his party to the centre-right cabinet that also included the Centre and National Coalition. There was clearly a demand for a party with a more critical view of European integration—and more broadly speaking for a party that would represent those sections of the citizenry with more traditional or socially conservative and nationalist preferences (Kestilä 2006). The party performed both in the 2011 and 2015 elections remarkably evenly across the country. According to surveys voters were drawn to the party mainly because they wanted to shake established patterns of power distribution and to change the direction of public policies, especially concerning immigration and European integration. Hence it is fair to claim that the phenomenal rise of the Finns Party is explained by both protest and issue voting (Borg 2012; Westinen 2014, Grönlund and Wass 2016). The future of the populists is currently difficult to predict. Having chaired the party since 1997, Soini decided to resign as the party leader, but in June 2017 the Finns Party split into two after the party congress had elected Member of the European Parliament Jussi Halla-aho as the new party chair. Halla-aho, who has been convicted in court for disrupting religious worship and of ethnic agitation, and the new party leadership look set to take the party economically further to the right whilst engaging in hard-line attacks on immigration and multiculturalism. Immediately following the election of Halla-aho, the more moderate or populist wing of the party left the Finns Party and established a new party, the Blue

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

176



Reform. The Centre and the National Coalition refused to work with Halla-aho-led Finns Party in the cabinet. This enabled Soini and his Blue Reform ministerial colleagues to remain in the government, but the future of Blue Reform is very uncertain.

A changing cleavage structure The dominant cleavage has traditionally been the left–right dimension, but since the early 1990s the sociocultural dimension, which is linked to the rural-urban/centre-periphery divide, has become the second main cleavage, partly because the European Union and the opening of borders have emerged on the political agenda (Westinen 2015). Parties tend to be internally divided in such issues, for example over European integration, immigration, or gay rights, and this has impacted coalition governance. Party positions on European integration have constituted an important element of government formation since the start of EU membership in 1995, and in 2011 the Finns Party’s decision to continue in the opposition was explained by the impossibility of joining a cabinet that was committed to further Eurozone bailout measures. In 2011 the parties forming the Katainen I ‘six pack’ government agreed not to introduce law proposals allowing same-sex marriages (as discussed later in the chapter), but following a citizens’ initiative, the Eduskunta voted in late 2014 in favour of same-sex marriages. Given the fragmented party system and the tradition of forming large multiparty cabinets, political parties and their leaders are engaged in an almost constant process of negotiation, and the art of making compromises and logrolls is an essential feature of daily politics. In order not to exclude themselves from cabinet formation negotiations, parties do not present voters with pre-election alliances nor do they make public statements ruling out sharing power with particular parties.⁶ This has so far applied also to working together with the Finns Party, but following the election of Halla-aho and the change in party image, at least the Greens have signalled their unwillingness to join a coalition that includes the Finns. Finnish parties are thus highly office-seeking in their behaviour. While partisan cooperation in multiparty governments and in the Eduskunta may enhance parties’ ability to defend the interests of their constituents, it simultaneously makes it harder for the voters to assess the performance of their

⁶ Parties can, however, form electoral alliances inside individual electoral districts. Within electoral alliances, the distribution of seats is determined by the plurality principle regardless of the total number of votes won by the respective parties forming the alliance. Smaller parties such as Christian Democrats have tended to enter electoral alliances with larger parties. These district-level alliances are typically presented to the voters as ‘technical’, not ideologically motivated, pacts.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



177

representatives, particularly when considering the lack of transparency that characterizes coalition government decision-making.

Government formation Bargaining and portfolio allocation Government formation processes have been relatively straightforward since the early 1990s (Table 6.2). The only real exception was the bargaining following the 2011 elections. The electoral triumph of the Finns Party complicated the situation, and the negotiations leading to the formation of the ideologically highly heterogeneous Katainen I cabinet were marked by significant confusion about government composition, including the possibility of the Finns entering the cabinet and of Katainen being replaced as the formateur.⁷ In other cases the election result has played a bigger role, not least through a party or parties suffering a major loss and thus preferring to stay in the opposition—for example the Centre following the 1995 and 2011 elections and SDP following the 1991, 2007, and 2015 elections. The negotiations centre around the government programme. In some instances, the leader of the largest Eduskunta party can start the process by publicly requesting answers from all parties regarding specific issues. For example, after the 2015 elections the future PM Juha Sipilä sent the parties a list of 15 questions, with most of them dealing with economy. This way the formateur can establish which parties could be potential coalition partners and thus worth starting actual government formation talks with.⁸ Even when such an initial round takes place, the formateur nonetheless normally has a basic idea of government composition immediately after the election, particularly if the election result has been favourable to the governing coalition or the PM’s party, as happened in the 1999 and 2007 elections. According to the interviews even as much as up to 95 or 98 per cent of the time is spent on policy issues. Typically a number of working groups are established to examine various topics (such as economy, EU and foreign affairs, environment

⁷ Essentially all Eduskunta parties were at various points included in the talks, with formation attempts led by Katainen complemented with informal negotiations between the other parties. The exact sequencing of events is difficult to establish with certainty, and this was also confirmed by several of the interviewees. ⁸ ‘Tässä ovat Sipilän 15 kysymystä eduskuntaryhmille’, Yle, 28 April 2015. Retrieved from: https:// yle.fi/uutiset/3-7960159, accessed 17 March 2021. Some parties may deliberately rule themselves out from government formation by not answering the questions at all or by providing replies that they know will not please the formateur. On the other hand, the future PM can perhaps use the (public) replies later to his advantage, for example by reminding the opposition of their post-election positions on key issues.

1999 0

2002 2003 2003 2007 2010 2011

Lipponen II

Lipponen III Jäätteenmäki Vanhanen I Vanhanen II Mari Kiviniemi Katainen I

SDP, KOK, SKDL, VIHR, RKP SDP, KOK, SKDL, VIHR, RKP 5

7

11 4

8

KOK, SDP, VIHR, RKP, SKL, SKDL

1994 0 1995 0

Aho II Lipponen I

KESK, KOK, RKP, SKL (1) KESK, KOK, RKP, SKL, VIHR

2

1990 0 1991 1

Holkeri II Aho I

KOK, SDP, SMP, RKP (1) SKDL, SDP, VIHR, KESK, SMP, SKL, RKP, KOK

KESK, SDP, RKP 12 KESK, SDP, RKP KESK, RKP, KOK, VIHR 12

1987 1

Holkeri I

Parties involved in Bargaining duration the previous bargaining of individual rounds formation attempt (in days)

0 0 0

Year Number of in inconclusive bargaining rounds

Cabinet

Table 6.2 Cabinet formation in Finland, 1987–2018

12

6 4

37

12

5

7

15

32

8 25

4 40

45

107 95 115 118 111 33

129

139

72 63 69 66 37

54

49

Result of investiture vote Number of days Total bargaining required in duration government formation Pro Abstention Contra

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2014 2014 2014 2015 0 2017 KESK, SMP, KOK

18

15

7

Note: The data on investiture votes concern the vote on the government programme.

Katainen II Stubb I Stubb II Sipilä I Sipilä II

(1) KOK, SDP, VIHR, RKP, SKL (2) KOK, SDP, VIHR, RKP, SKL, SKDL

40

18

114 13

99

72

79

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

180



and energy), and this increases the number of people per party present in the negotiations. Ministries and perhaps public sector agencies as well as various interest groups are also involved in the negotiations, either directly or through political parties (Paloheimo 2003: 228). Nonetheless, there is definitely an ‘inner core’ of persons, consisting primarily of party leaders and their closest aides. Bargaining over policy can be time-consuming and difficult, and given the importance of the government programme it is understandable that political parties take the negotiations very seriously. The remainder of the time parties negotiate about portfolio distribution and the ‘rules of the game’, with the latter category taking up just around one per cent of the talks. Parties agree on the distribution of portfolios between the coalition partners, with the understanding that the second largest coalition party gets to pick the second ministerial portfolio (Table 6.3). Since the late 1970s this has undoubtedly been the minister for finance, although in 2015 the Finns Party chose the post of the foreign minister. The post of the foreign minister has become less important over time, not least due to the internationalization of other policy areas, with for example the minister of interior responsible for representing Finland in EU meetings dealing with immigration and asylum-seekers. Questions related to law and order and immigration have overall become more salient, thus increasing the weight of the minister of interior. The prestige of the minister for social affairs is in turn explained by the Nordic-style welfare state regime and the large budget associated with the portfolio. Since the 1980s leaders of coalition parties have as a rule been ministers as well. Considering the prominent role of the Ministry of Finance as a kind of ‘super ministry’ (Murto 2014), the fact that the leader of the second largest coalition party is the finance minister emphasizes cooperation between the two largest cabinet parties (Paloheimo 2002: 213). Naturally the whole government and individual parties pay attention to gender balance and perhaps the representation of various regions in the cabinet, but only seldom do the names of potential individual ministers come up in the negotiations. Should that occur, it usually happens in the form of coalition partner(s) informing the leader of another coalition party that a certain politician from her party should not be included in the cabinet. Otherwise, coalition partners are responsible for choosing their own ministers, with the formateur having only limited opportunities to influence the choices made by the parties. The procedures are largely similar in all parties, with the parliamentary group, perhaps together with the party council/executive, taking the decisions. Naturally the party leaders are influential as well, but much depends on their support within the parties. In any case, neither the PM nor the leaders of other cabinet parties can reshuffle ministerial posts or replace ministers without legitimate reasons

Year in

1987 1990 1991 1994 1995

1999

2002

2003 2003 2007

2010

2011

2014

2014

2014 2015 2017

Cabinet

Holkeri I Holkeri II Aho I Aho II Lipponen I

Lipponen II

Lipponen III

Jäätteenmäki Vanhanen I Vanhanen II

Mari Kiviniemi Katainen I

Katainen II

Stubb I

Stubb II Sipilä I Sipilä II

8 SDP, 7 KOK, 2 RKP, 1 SMP 9 SDP, 6 KOK, 2 RKP 8 KESK, 6 KOK, 2 RKP, 1 SKL 8 KESK, 6 KOK, 2 RKP 7 SDP, 5 KOK, 2 RKP, 2 SKDL, 1 VIHR, 1 Ind. 6 SDP, 6 KOK, 2 RKP, 2 SKDL, 1 VIHR, 1 Ind. 7 SDP, 7 KOK, 2 RKP, 2 SKDL 8 KESK, 8 SDP, 2 RKP 8 KESK, 8 SDP, 2 RKP 8 KESK, 8 KOK, 2 RKP, 2 VIHR 8 KESK, 8 KOK, 2 RKP, 2 VIHR 6 KOK, 6 SDP, 2 RKP, 2 SKDL, 2 VIHR, 1 SKL 6 KOK, 6 SDP, 2 RKP, 2 VIHR, 1 SKL 6 KOK, 6 SDP, 2 RKP, 2 VIHR, 1 SKL 7 KOK, 7 SDP, 2 RKP, 1 SKL 6 KESK, 4 KOK, 4 SMP 7 KESK, 5 KOK, 5 SIN

Number of ministers per party (in descending order)

17 14 17

17

17

19

20

18 18 20

18

18

18 17 17 16 18

Total number of ministers

12 12 12

12

12

12

12

13 13 13

13

13

13 13 13 13 13

Number of ministries

Table 6.3 Distribution of cabinet ministerships in Finnish coalitions, 1987–2018

KOK KESK KEKS

KOK

KOK

KOK

KESK

KESK KESK KESK

SDP

SDP

KOK KOK KESK KESK SDP

1 Prime minister

SDP KOK KOK

SDP

SDP

SDP

KOK

SDP, RKP SDP, RKP KOK

KOK

KOK

SDP SDP KOK KOK KOK

2 Finance

SDP SMP SIN

SDP

SDP

SDP

KOK

SDP SDP KOK

SDP

SDP

SDP SDP KESK KESK SDP

3 Foreign affairs

KOK, SDP SMP, KESK SIN, KESK

KOK, SDP

KOK, SDP

KOK, SDP

KESK, KOK

SDP, KESK SDP, KESK KESK, KOK

KOK, RKP

KOK, RKP

KOK KOK KESK KESK SDP

4 Social affairs

SKL KOK KOK

SKL

SKL

SKL

KOK

SDP SDP KOK

KOK

KOK

SDP SDP KESK KESK RKP

5 Interior

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

182



(Paloheimo 2003: 236–7). In such cases, the party bodies are involved exactly as they are when ministers are selected to the new government.⁹ The parties do not negotiate about other, non-cabinet, positions. There is a general rule according to which the Speaker of Eduskunta has to represent a different party than the PM. Otherwise the expectation is that once the government has entered into office, various positions, such as heads of public sector agencies or the highest-ranking civil servants in the ministries, are appointed on the basis of a recommendation by the respective line minister. Several of the interviewees suggested that the mid-1990s were a turning point, with the formation of the Lipponen I cabinet witnessing much less agreements about such non-cabinet positions than before. Exceptions to this are certain individual appointments such as that of the Finnish EU Commissioner and perhaps other international positions, but even then the agreements between the parties may only be provisional and may involve only the largest two parties in the cabinet. Finally, coalition parties need hardly discuss the ‘rules of the game’ at all, either during the government formation process or right after the cabinet has entered into office. Typically the cabinet, or at least the PM together with the leaders of other coalition parties, briefly agree verbally about conflict management mechanisms and other such rules. Most of the interviewed politicians underlined that coalition governance mechanisms, from the key role of the government programme to ministerial committees and informal meetings of party leaders, are so institutionalized that everybody knows them. Continuity is reinforced by the inclusion in the new government of at least one of the parties from the outgoing coalition.

Composition and size of cabinets When compared with other European countries, Finnish cabinets are outliers in three respects: their parliamentary support, level of fragmentation, and ideological diversity (see Table 6.1a). Finland used to be characterized by short-lived and unstable governments living under the shadow of the president. As reported in the introductory chapters of this volume, among the West European countries, only Italy had more cabinets between 1945 and 2016 than Finland. Of the 48 cabinets (excluding caretaker governments) as high a share as 60.4 per cent were surplus majority coalitions, 18.8 per cent were minimal winning coalitions, 12.5 per cent were minority coalitions, and 8.3 per cent were single-party minority cabinets. The governments appointed after the era of President Kekkonen have basically

⁹ Examining decision-making in the National Coalition, the Social Democrats, and the Green League about joining governments after the 1983, 1987, 1991, 1995, and 2011 elections, Koskimaa (2016) detected a clear move towards a stronger influence of the party leaders.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



183

stayed in office for the whole four-year electoral period—a period Nousiainen (2006) has labelled ‘stable majority parliamentarism’. Since 1983, the broad, mainly oversized coalitions have controlled safe majorities in the Eduskunta. The centre-right cabinet led by Aho (1991–1994) had the narrowest majority with 57.5 per cent of the seats (although the National Coalition-led cabinet was down to 50.5 per cent in 2014 after the Left Alliance and the Greens had left the government), while the first rainbow coalition led by Lipponen in 1995–1999 controlled as many as 72.5 per cent of the seats. The Sipilä I coalition formed after the 2015 election controlled 62 per cent of seats, but following the split of the Finns Party it was down to 53 per cent of the seats. Not surprisingly, these broad coalitions have ruled without much effective parliamentary opposition. Particularly important has been the fragmented nature of the opposition. As the cabinets have, with the exception of the Sipilä governments formed after the 2015 elections and the centre-right cabinet of 1991–1995, brought together parties from both the left and the right, the opposition has been both numerically weak and ideologically fragmented. The overwhelming majority of Finnish governments have been cross-bloc coalitions, bringing together parties from the left and the right. The first ‘redochre’ coalition between Social Democrats and the Agrarian Union was formed in 1937, while the Holkeri I cabinet formed in 1987 was the first government based on cooperation between the National Coalition and the Social Democrats. Recent governments have as a rule included two of the three main parties, the Social Democrats, the Centre, and the National Coalition. The Swedish People’s Party has participated in most governments, including all cabinets formed between 1979 and 2015. The near-permanent government status of the party can be interpreted as a mechanism for protecting minority rights, but it is also explained by the centrist and flexible ideology of the party. Clearly the ‘rainbow’ coalitions led by Lipponen from 1995 to 2003 and the ‘six pack’ Katainen I cabinet (2011–2014) stand out in international comparison due to their strong ideological heterogeneity. Particularly the formation of the Lipponen I cabinet in 1995 was important. Bringing together five parties, including the most right-wing (National Coalition) and left-wing (Left Alliance) parties in the Eduskunta, it enabled PM Lipponen and his government to push through the necessary budgetary and economic measures needed to take Finland into the Economic and Monetary Union (EMU) (Jungar 2002). Several factors have contributed to the prevailing practice of forming oversized coalitions or at least governments that enjoy strong parliamentary majorities: the fragmented party system and the ensuing need to build workable coalitions with ‘safe’ majorities; the role of the Centre Party as a bridge-builder, forming coalitions with both parties to its left and its right; and the deferment rule that until the early 1990s allowed one-third of the MPs to postpone the adoption of laws. Furthermore during the Cold War the external factor of the Soviet Union played

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

184



a role. With the National Coalition excluded from government formation between 1966 and 1987 due to the need to anticipate reactions from Moscow, parties under the watchful eye of President Kekkonen formed broad and ideologically centrist cross-bloc coalitions. Here one needs to remember that already behind the formation of the first cross-bloc coalition before the Second World War was the goal of building broader societal consensus and marginalizing more radical ideological alternatives.¹⁰

Coalition agreements Considering the ideological and numerical breadth of Finnish cabinets, it is not surprising that the government programme has become such an important document.¹¹ The high number of parties forming the cabinet and the need to commit them and their party groups to established rules and policies primarily explain the length of the programmes. Ministers and their parties also know that ‘it is difficult to introduce new initiatives once the government has started to work’, and hence they have an interest in including a large number of items in the programme (Nousiainen 1996: 117). The programmes have become longer and more detailed over the decades (especially since the early 1980s), with the coalition partners investing a lot of resources in bargaining over the programme. There was a major leap at the turn of the millennium: while the programme of the Lipponen II cabinet from 1999 had 6,698 words, the governments appointed since then have drafted programmes in excess of 12,000 words. The programme of the Katainen I ‘six pack’ government, formed after the 2011 elections, had 90 pages and 26,654 words, while the ‘strategic’ programme of the Sipilä I government formed in 2015 had 17,709 words. The programmes focus 100 per cent on policy and do not say anything about portfolio allocation, appointment to non-cabinet positions, or about how the government will conduct its business or resolve its internal disputes (Table 6.4). Economic policy, including the role of corporatist actors, remains the most important issue, although the sociocultural dimension and especially attitudes ¹⁰ Formation of the Vanhanen II cabinet after the 2007 elections is a good example of how numerically large coalitions have become the dominant pattern. Immediately after the election result became clear, it seemed that the likeliest coalition alternative was a centre-right cabinet between the Centre, the National Coalition, and the Swedish People’s Party. However, Vanhanen announced that his new cabinet should control around 120 of the 200 seats. He justified this by referring to the need to ensure the smooth functioning of the cabinet. Soon afterwards Vanhanen declared that the government would bring together the Centre, the National Coalition, the Swedish People’s Party, and the Green League, commanding a comfortable majority in the Eduskunta with 126 seats (63 per cent). Jari Laurikko, ‘Vanhasen tavoite neljän puolueen ja 120 kansanedustajan hallitus’, Turun Sanomat, 20 March 2007. Retrieved from: http://www.ts.fi/uutiset/kotimaa/1074188456/Vanhasen+tavoite+neljan +puolueen+ja+120+kansanedustajan+hallitus, accessed 17 March 2021. ¹¹ The programmes are available at http://valtioneuvosto.fi/tietoa/historiaa/hallitusohjelmat.

Year in

1945 1946 1948 1950 1951 1951 1953 1954 1954 1956 1957 1957 1957 1957 1958 1959 1961 1962 1963 1964 1966 1968 1970 1971 1972 1972

Coalition

Paasikivi III Pekkala Fagerholm I Kekkonen I Kekkonen II Kekkonen III Kekkonen IV Törngren Kekkonen V Fagerholm II Fagerholm III Sukselainen I Sukselainen II Sukselainen III Fagerholm IV Sukselainen IV Miettunen I Karjalainen I Karjalainen II Virolainen Paasio I Koivisto I Karjalainen III Karjalainen IV Paasio II Sorsa I

448 418 571 248 268 413 224 354 561 225 225 242 242 751 1,415 391 158 1,103 1,103 404 777 841 1,723 1,723 783 1,936

Size

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

General rules (in %)

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Policy-specific procedural rules (in %)

Table 6.4 Size and content of coalition agreements in Finland, 1945–2018

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of offices (in %) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of competences (in %) 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100

Continued

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

204 2,222 512 512 1,118 1,025 1,025 1,788 2,861 2,861 2,697 2,697 4,541 6,698 6,698 12,211 12,061 15,399 1,067 26,654 26,654 1,916 1,916 17,709 17,709

1975 1976 1977 1978 1979 1982 1982 1983 1987 1990 1991 1994 1995 1999 2002 2003 2003 2007 2010

2011 2014 2014 2014 2015 2017

Miettunen II Miettunen III Sorsa II Sorsa III Koivisto II Sorsa IV Sorsa V Sorsa VI Holkeri I Holkeri II Aho I Aho II Lipponen I Lipponen II Lipponen III Jäätteenmäki Vanhanen I Vanhanen II Mari Kiviniemi Katainen I Katainen II Stubb I Stubb II Sipilä I Sipilä II

Size

Year in

Coalition

Table 6.4 Continued

0 0 0 0 0 0

0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0

General rules (in %)

0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Policy-specific procedural rules (in %)

0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of offices (in %)

0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of competences (in %)

100 100 100 100 100 100

100 100 100 100 100 100 100 99 100 100 100 100 100 100 100 100 100 100 100

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



187

towards the European Union have become more salient. The level of detail depends very much on the ability of the coalition partners to agree on common goals. If no such agreement is found, the programme can state more vaguely that the government will ‘look into’ certain issues (meaning that there is no commitment to necessarily do anything related to the matter). The programme can also explicitly bind the coalition partners to status quo—for example that Finland will not seek NATO membership or that certain social security benefits or taxes will not be raised or lowered, and in 2011 the six parties forming the Katainen I government agreed at the insistence of the Christian Democrats not to introduce law proposals allowing same-sex marriages (the coalition leaders at the same time agreed that individual MPs could be active in the issue, for example through tabling private members’ bills). Coalition partners can also make verbal, essentially non-public, agreements about various policy issues, but in such cases they are probably not as binding as when the issues are explicitly mentioned in the government programme. There are no other public documents about coalition governance. As stated earlier, the ‘rules of the game’ seem so institutionalized and accepted by the parties that they are not written down anywhere (although the constitution and other laws regulate more formal aspects of government work such as ministerial committees and plenaries). This applies to the entire post-Second World War period (see also Nousiainen 2000: 278–80). However, the government or at least the leaders of the coalition parties can reach informal, behind-the-scenes understandings about the ‘rules of the game’ or some policy issues.

Coalition governance The PM and the individual ministers Recent constitutional and political developments have undoubtedly strengthened the position of the PM. With the partial exception of the finance minister, the PM is the only person in the government whose jurisdiction covers all policy areas. According to Section 66 of the constitution, ‘The Prime Minister directs the activities of the Government and oversees the preparation and consideration of matters that come within the mandate of the Government.’ And, as stated earlier, this leadership applies now also to EU matters and to external relations where leadership is shared with the president. The PM’s office has risen in stature and in size in recent decades. It coordinates decision-making in the ministries, operates as a broker in the case of disputes within or between ministries, and monitors the implementation of the government programme. The PM thus is the key actor in resolving conflicts inside the cabinet, whereas under the old constitution this task was often carried out by the president. Increased cabinet duration has also provided

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

188



the PM with a more stable environment to exercise leadership. Furthermore, government communication is increasingly centralized to the PM’s office (Niemikari et al. 2019). The PM is thus stronger both in the whole national political system and inside the government (Paloheimo 2002, 2003, 2005, 2016). However, the bargaining involved in forming coalition cabinets and keeping them together act as significant constraints on the executive powers of the PM. It must be remembered that apart from ministers from her or his own party, the PM has little influence on the selection or sacking of ministers, with the coalition partners being responsible for such decisions. As essentially all interviewees strongly emphasized, Finnish PMs are first and foremost ‘managers’ of coalitions for whom building and maintaining trust is a prerequisite for policy success and cabinet survival. This is already evident in Eduskunta election campaigns where the parties seek to present their leaders as the most suitable next PM. This constrains party leaders from adopting strong political stances or engaging in confrontational discourse, privileging instead the quality of ‘statesmanship’ and the (perceived) ability to manage a coalition cabinet. As was discussed in the previous section, government formation is based on bargaining about policies between the coalition partners, with the formateur responsible for managing and overseeing the process. Similar logic guides the behaviour of PMs once the government has entered into office, as the PM needs to strike a balance between effective leadership and accommodating the preferences of the coalition parties. Leadership refers primarily to organizational matters— making sure that schedules are adhered to, that necessary decisions are taken, and that in general the ‘house is kept in order’. In terms of policy leadership, there was again broad consensus among the interviewees that the PM does not have much freedom of manoeuvre.¹² Apart from the government programme, the rather legalistic or formal nature of government work plays a role here: the ministerial committees and full plenaries are based on agendas prepared by civil servants and distributed beforehand to all participants, and it is not considered appropriate for the PM—who chairs all these meetings—to introduce new issues to the meetings or indeed to the broader government agenda without careful consultation with at least the leaders of the coalition parties. ‘Nothing should come as a surprise’, stated one of the interviewees. Moreover when interviewed by the media, Finnish PMs as well as line ministers by and large stick to the government agenda and do not express support for objectives not agreed upon by the cabinet. As a result, Finnish coalition governance operates very much along the lines of the coalition compromise model outlined in Chapter 2 of this volume.

¹² Sometimes the PM can exercise significant policy leadership in individual policy domains but again much depends on the cohesion of the cabinet. A good example was PM Lipponen, who had a specific interest in European affairs and played a major role in shaping national EU policy.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



189

Turning to individual ministers, their number has stayed fairly constant since the Second World War, but there has been a slight increase over the decades. The Vanhanen II government formed after the 2007 elections had an all-time high of 20 ministers. The Sipilä I cabinet appointed in 2015 had only 14 ministers, the lowest number since the 1950s and 5 less than in the cabinet appointed after the 2011 elections, but in the Sipilä II government there are 17 ministers. The number of ministries has also stayed about the same, with the current number being 12. Individual ministers have arguably become more autonomous actors in recent decades, and they wield stronger influence in their fields of competence than previously. This delegation of authority from the PM and the entire cabinet to the individual ministers has been necessitated by the gradually increasing workload of the cabinet since the 1960s and the resulting need to divide labour and delegate power to the line ministries. Two laws enacted in the mid-1990s specifically transferred authority to the individual ministries in matters not requiring decision-making by the plenary session of the cabinet. Since 1970 all ministers have had their own special political advisors, distinct from civil servants in the ministries, and since 2005 ministers may also have their own state secretaries. Both the special advisors and the state secretaries are political appointees whose terms coincide with those of the respective ministers. They are thus normally from the same party as the minister, and their job is to assist the minister in her or his duties. Most ministers have state secretaries, while the number of special advisors has varied both between cabinets and between individual ministers, with on average two to five advisors per minister. The basic rule is that ministers do not intervene in questions falling under the jurisdiction of their colleagues, thus respecting each other’s autonomy.¹³ Nevertheless, ministerial autonomy is strongly kept in check by the government programme and the agreements between the leaders of the coalition parties, at least in politically significant matters. In fact, it has been argued that Finnish line ministers have little autonomy compared to their European counterparts (Nousiainen 2000: 270; see also Nousiainen 1994).

Conflict management mechanisms There are various levels or forums for solving disputes between coalition parties (Table 6.5). The government programmes contain nothing about such conflict management mechanisms. But before discussing them individually, it is worth ¹³ For example, when discussing its ‘rules of the game’, the Vanhanen I cabinet agreed that ‘the fields of other ministers are not to be fiddled with’. Ilkka Ahtiainen, ‘Rkp otti aikalisän kansliapäällikkönimitykseen: Enestam ei halunnut erotuomariksi keskustan ja Sdp:n kiistaan’, Helsingin Sanomat, 5 September 2003. Retrieved from: https://www.hs.fi/kotimaa/art2000004165197.html, accessed 17 March 2021.

1945 1946 1948 1950 1951 1951 1953

1954 1954 1956 1957 1957

1957 1957

1958 1959 1961 1962 1963 1964

Törngren Kekkonen V Fagerholm II Fagerholm III Sukselainen I

Sukselainen II Sukselainen III

Fagerholm IV Sukselainen IV Miettunen I Karjalainen I Karjalainen II Virolainen

POST IE IE POST POST IE

IE IE

POST IE IE IE IE

PRE, POST IE POST IE IE POST IE

Y Y Y Y Y Y

Y Y

Y Y Y Y Y

Y Y Y Y Y Y Y

N N N N N N

N N

N N N N N

N N N N N N N

CaC CaC CaC CaC CaC IC, CaC CaC CaC CaC CaC IC, CaC IC IC, CaC CaC CaC CaC CaC CaC CaC

IC CaC

CaC CaC CaC CaC CaC

CaC CaC CaC CaC CaC CaC

CaC CaC

CaC -

IC CaC

CaC CaC CaC CaC CaC

CaC CaC CaC CaC CaC IC

All Most For most used common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Paasikivi III Pekkala Fagerholm I Kekkonen I Kekkonen II Kekkonen III Kekkonen IV

Coalition

Table 6.5 Coalition governance mechanisms in Finnish coalitions, 1945–2018

-

-

-

-

-

-

-

-

Personal Issues union excluded from agenda

Spec./Spec. Spec./Spec. Spec./Spec. Spec./Spec. Spec./Spec. Spec./Spec.

Spec./Spec. Spec./Spec.

Spec./Spec. Spec./Spec. Spec./Spec. Spec./Spec. Spec./Spec.

Spec./Spec. Spec./Spec. Spec./Spec. Spec./Spec. Spec./Spec. Spec./Spec. Spec./Spec.

Coalition discipline in legislation/ other parl. behaviour

y N y

y y

y y y y

N N N y y N

Varied Varied Few

Few Few

Few Varied Varied Few

Few Few Few Few Few Few

y y y y y y

y y

y y y y y

y y y y y y y

Freedom of Policy Junior Nonappointment agreement ministers cabinet (watchdogs) positions

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1966 1968 1970 1971 1972 1972

1975

1976

1977

1978

1979

1982

1982

1983

1987

Paasio I Koivisto I Karjalainen III Karjalainen IV Paasio II Sorsa I

Miettunen II

Miettunen III

Sorsa II

Sorsa III

Koivisto II

Sorsa IV

Sorsa V

Sorsa VI

Holkeri I

POST

POST

IE

IE

POST

IE

IE

IE

POST

POST IE IE IE POST IE

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y Y Y Y Y Y

N

N

N

N

N

N

N

N

N

N N N N N N

CaC CaC CaC CaC CaC, IC, Pca CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca CaC

CaC

CaC

CaC

Pca

Pca

Pca

Pca

Pca

Pca

CaC

CaC

Pca

Pca

Pca

CaC CaC CaC CaC Pca

CaC

CaC

Pca

CaC CaC CaC CaC CaC

N (SMP, SDP)

-

-

-

-

-

-

-

-

-

N

-

-

-

-

-

-

-

-

-

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Spec./Spec. Spec./Spec. Most/Most Most/Most Most/Most Most/Most

y

y

y

y

y

y

y

y

y

y y y y y

Comp.

Comp.

Comp.

Comp.

Comp.

Few

Few

Few

Comp.

Varied Varied Comp. Comp. Comp.

Continued

y

y

y

y

y

y

y

y

y

y y y y y y

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1990

1991

1994

1995

1999

2002

2003

2003

Aho I

Aho II

Lipponen I

Lipponen II

Lipponen III

Jäätteenmäki

Vanhanen I

IE

POST

POST

POST

POST

POST

POST

POST

Y

Y

Y

Y

Y

Y

Y

Y

N

N

N

N

N

N

N

N

IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca Pca

Pca

Pca

CaC

CaC

Pca

Pca

Pca

Pca

Pca

CaC

CaC

CaC

CaC

CaC

CaC

All Most For most used common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Holkeri II

Coalition

Table 6.5 Continued

N

N

N (SDP)

N (SDP)

Y

Y

N

N

N

N

N (VIHR) N

N (KOK) N

Y

Y

Personal Issues union excluded from agenda

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Coalition discipline in legislation/ other parl. behaviour

y

y

y

y

y

y

y

y

Comp.

Comp.

Comp.

Comp.

Comp.

Comp.

Comp.

Comp.

y

y

y

y

y

y

y

y

Freedom of Policy Junior Nonappointment agreement ministers cabinet (watchdogs) positions

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2010

2011

2014

2014

2014

2015

2017

Mari Kiviniemi

Katainen I

Katainen II

Stubb I

Stubb II

Sipilä I

Sipilä II

POST

POST

POST

IE

POST

POST

IE

POST

Y

Y

Y

Y

Y

Y

Y

Y

N

N

N

N

N

N

N

N

IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca IC, CaC, Pca CaC

Pca

Pca

Pca

CaC

CaC

Pca

Pca

Pca

Pca

Pca

CaC

CaC

CaC

CaC

CaC

Y

Y

Y

Y

Y

Y

Y

Y

N

N

N

N

N

N

N

N

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

Most/Most

y

y

y

y

y

y

y

y

Comp.

Comp.

Comp.

Comp.

Comp.

Comp.

Comp.

Comp.

y

y

y

y

y

y

y

y

Notes: Coalition agreement: IE = inter-election; POST = post-election Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee; Pca = Combination of cabinet members and parliamentarians; O = Other Coalition discipline: Most = Discipline expected, except on explicity exempted matters; Spec. = Discipline only expected on a few explicitly specified matters. Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement

2007

Vanhanen II

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

194



stating the obvious: not all disagreements take place between coalition partners. In fact, several of the interviewees underlined that most conflicts occur inside individual cabinet parties, with the understanding that the party in question and particularly its ministerial group or the party leader is responsible for solving the matter. When there is a conflict between coalition parties or typically between two line ministers, the special advisors of those ministers are normally the first actors entrusted with looking into the issue. In a lot of instances such behind-the-scenes discussions between the ministers and advisors are sufficient. There are no ‘watchdog’ junior ministers in Finland. Yet the governing parties operate a kind of a shadow minister or ‘overcoat’ (the word used by some of the interviewees was päällystakki in Finnish) system, whereby each minister is responsible for keeping an eye on her or his cabinet colleagues from the other parties. In case of disagreement between the responsible line minister and an individual cabinet party, the ‘overcoat’ can negotiate with the minister in charge of the issue. Such a system is more demanding for small coalition parties: for example, the Left Alliance had 2 ministers in the Katainen I cabinet, meaning that these 2 were responsible for monitoring the other 17 ministers.¹⁴ The next stage is the four statutory ministerial committees: Ministerial Committee on Foreign and Security Policy, Ministerial Finance Committee, Ministerial Committee on Economic Policy, and Ministerial Committee on European Union Affairs. The first two have existed throughout Finland’s independence, the third became a permanent body in 1977, while the last was established upon Finland joining the EU in 1995. All committees are chaired by the PM and bring together ministers from all coalition parties. If no agreement is found in the ministerial committee, a typical strategy is to continue the preparatory work under the guidance of the respective line minister, with the issue then reintroduced in the ministerial committee when a solution has been found. As shown earlier, one of the key tasks of the PM is indeed to ensure that matters are not delayed too much and that decisions are taken. Governments can also set up various ad hoc or issue-specific ministerial working groups. Their heyday was probably in the 1970s, and subsequent cabinets have on average established much less such committees. However, according to Paloheimo (2003: 233) the Lipponen II cabinet had ‘nine ministerial working groups for special policy areas’, whereas the Katainen I cabinet had over ten such working groups (Valtioneuvoston kanslia 2015: 32). The Sipilä II cabinet utilized nine working groups: on employment and competitiveness; knowledge and education; health and wellbeing; bioeconomy and clean solutions; reforming ¹⁴ This ‘overcoat’ system existed already before the 1990s, with Nousiainen (1994: 94) reporting that ‘it is habitual in large coalitions that a ministerial group assigns its members to monitor ministries led by other parties’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



195

operating practices; reforms; internal security and administration of justice; migration; and Russia.¹⁵ The minister responsible for the issue area in question normally chairs these ministerial groups and again all cabinet parties are usually represented in the working groups. Both the ministerial committees and the working groups perform important roles in conflict-solving, not least through their ‘watchdog’ function, with coalition parties having the opportunity to monitor developments inside individual ministries (Paloheimo 2016: 78). The full plenary of the government in turn is responsible for final, formal decisions, but its meetings, held on Thursday afternoons, are essentially without exception very short, with the plenary just rubber-stamping decisions taken in the ministerial committees or by individual ministers. The formal decision rule in the government plenary is majority, with the vote of the PM decisive if there is a tie.¹⁶ Yet all of the interviewees noted that there is a ‘culture of consensus’, with votes avoided and a desire to reach unanimity. This is also achieved most of the time, with each interviewed minister remembering that voting was resorted to only exceptionally in a very small number of cases per government. According to Nousiainen (2000: 281–2; see also Paloheimo 2002: 212) there was a change from the late 1970s onwards (Sorsa II cabinet), with the PMs starting to pay more attention to the internal solidarity of the government, with subsequent cabinets avoiding votes and aiming at unanimous decisions. If no agreement is reached, the PM can postpone the matter to forthcoming meetings. The PM could force votes to be held, but according to the interviewees this would almost certainly undermine trust among the coalition partners. There is also an informal veto-player rule in the sense that each coalition partner can request that decisions are not taken or that the issue is postponed until the next meeting. Given that Finnish cabinets are usually oversized coalitions, smaller parties or individual ministers from larger cabinet parties can sometimes vote against the majority or enter dissenting opinions into the minutes of the meetings as they will not prevent decision-making. However, at least since the early 1990s the most important decisions are often taken in discussions between the leaders of the coalition parties. The name of the meeting depends on the number of cabinet parties: in the Lipponen I and II governments they were referred to as ‘quintet’, in the Katainen I cabinet as ‘sextet’, and in a three-party cabinet as ‘trio’. These are informal gatherings without any actual written agendas or rules, with typically only the party leaders (and perhaps their advisors) present. All interviewees agreed that they are held ‘when the need arises’. Often short talks will suffice, but in case of serious conflicts, the meetings

¹⁵ https://valtioneuvosto.fi/en/sipila/ministerial-working-groups. ¹⁶ There are also presidential sessions of the government chaired by the president, the agenda of which covers those issues still in the competence of the head of state. In these sessions there is no voting and the president’s decisions do not have to follow the opinion of the government.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

196



can last long into the night or even consume much of the weekend. Party size clearly matters here, with several of the interviewed persons, both from smaller and larger parties, reporting that often the crucial discussions are held between the PM and the leader of the second largest cabinet party. On the one hand, these meetings ensure that also the smaller parties take part in decisions, but, on the other hand, a key role is performed by the leaders of the two biggest parties, especially in budgetary matters as the chair of the second largest party typically is the finance minister. As aptly summarized by Paloheimo (2005: 256): ‘These are the two “biggest hitters” in Finnish government, and any informal understandings they might reach can seriously restrict the bargaining power of smaller coalition parties.’ This applies also to other aspects of cabinet’s work, including bargaining over the government programme. Overall, the interviewees recognize the necessity and importance of these informal discussions, but at the same time accusations of a ‘closed inner circle’ are understandable. The governing parties are also in constant dialogue with their Eduskunta party groups. It is commonly accepted among the coalition partners that the government programme forms the backbone of the cabinet and that it is binding on all the parties. The government parties also monitor that their party groups support the programme. The cooperation rules between the governing parties’ parliamentary groups that have been in use since the early 1980s effectively prevent any disagreements or public conflicts between the government and the party groups (Wiberg 2011). Dissenting MPs can expect tough sanctions, including expulsion from their parliamentary group. The only exceptions are matters that are clearly ‘local’ by nature and certain questions of conscience. Should government bills encounter unexpected problems in the party groups or a committee of the Eduskunta, the government approaches their MPs and particularly the chairs of their parliamentary groups about the matter. Usually this exchange occurs inside individual governing parties between the party leader or another minister and the chair of the parliamentary group. In other instances, the committee chair and party spokespersons in the committee can be involved. Such exchanges are fairly routine, and the problems are more difficult to solve when the cabinet is not cohesive and/or the issue is not included in the government programme. If the issue at hand concerns the implementation of the cabinet programme, the PM or other ministers typically remind (on occasions publicly) the MPs about their duty to respect the programme. Moreover, even though there is variation between parties, the chair of the parliamentary group normally takes part in the meeting of the party’s ministerial group, and this again can be seen as a way to pre-empt potential conflicts between the cabinet and the legislature. These ministerial groups of cabinet parties convene essentially on a weekly basis and perform, when needed, an important role in resolving disputes inside the party and thus inside the entire cabinet.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



197

Previously Finnish governments had more time for discussions. This applies especially to the ‘evening school’ (iltakoulu), which has since the late 1930s brought together ministers and perhaps some other individuals, particularly leaders of the parliamentary groups of the cabinet parties, to sit down informally in the evenings to discuss various topical matters, mainly larger items on the cabinet agenda. The presence of the leaders of parliamentary groups of cabinet parties in these meetings facilitates the processing of government bills in the Eduskunta. The PM chairs the meetings. The importance of the ‘evening school’ has declined, with cabinets formed since the 1990s convening such sessions much more infrequently. The Sipilä cabinets formed after the 2015 elections have held ‘strategy meetings’ instead of evening school sessions that focus on the key priorities of the government (Virtanen et al. 2016: 10). Various reasons have been given for this decline: busier timetables of ministers, EU meetings and other international activities of the ministers, the increase in particularly younger (female) ministers that has resulted in the cabinet trying to work ‘normal hours’, the increased rigidity and formality of the sessions whereby they started to resemble official government meetings, and perhaps especially the fact that most issues have already been agreed in other forums, for example in the informal meetings of the coalition party leaders (Paloheimo 2002: 212; Tiili 2003).¹⁷ Naturally the PM can always organize additional ad hoc meetings with ministers to work out disagreements, for example immediately after the formal meetings of ministerial committees. Instead of informal discussions, recent Finnish governments have invested resources in improving coordination and strategic planning inside the cabinet and the entire executive branch, not least to encounter the decentralization of policy-making to the line ministries described earlier. Hence the governments appointed since 2003 have tried to improve horizontal coordination inside the cabinet, mainly through government’s intersectoral policy programmes (which were used from 2003 to 2011) and other coordination instruments such as various government strategy documents (Tiili 2008; Kekkonen and Raunio 2011; Virtanen et al. 2016). Another important tool is the ‘mid-term review’ session first utilized by the Vanhanen I cabinet, whereby the entire cabinet comes together halfway through the four-year electoral period to talk more freely about what the

¹⁷ The binding nature of the government programme and the resulting lack of discussion or genuinely collective decision-making inside the cabinet have attracted criticism from many leading politicians and civil servants (Tiili 2003). Paavo Väyrynen from the Centre Party, who served in the government during every decade from the 1970s until the 2010s, commented in an interview that ‘the most important change concerns the detailed nature of the government programme. The real bargaining occurs during the government formation process. When the cabinet begins its work on the basis of the programme, there is not much room for discussion’. Jaakko Hautamäki, ‘Paavo Väyrynen: Hallitus käy varsin vähän sisäisiä keskusteluja’, Helsingin Sanomat, 11 June 2007. Retrieved from https://www. hs.fi/kotimaa/art-2000004489903.html, accessed 17 March 2021.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

198



government has achieved, where it has failed, and whether its targets should be readjusted for the remaining two years.

Cabinet duration and termination The constitutional reforms impact on cabinet termination (Table 6.6). With the president and the Kremlin no longer intervening in government work, recent cabinets have either stayed in office for the whole four-year period or changes in government composition have been explained by disputes between the cabinet parties (as opposed to disputes between the government and the president). It was customary for the government to resign when a presidential election was held, but the last time this happened was in 1982. In fact, one can argue that under the old constitution, and particularly during the reign of President Kekkonen, governments were more accountable to the president than to the parliament. During the Cold War a crisis in relations with the Soviet Union brought the government down twice—in 1959 and in 1962. The last time a cabinet lost office after a vote of no-confidence in the Eduskunta was in 1958. Cabinet duration has increased quite substantially over the decades, especially since the early 1980s. Out of the PMs appointed to office immediately after the elections since 1983, Sorsa, Holkeri, Aho, and Lipponen (twice) have survived in office for the whole electoral term, while Vanhanen (who had already served as the PM from the summer of 2003 to the 2007 elections) and Katainen left their posts voluntarily, Katainen to become an EU Commissioner. If the PM resigns, the whole cabinet is dissolved. Hence the resignations of Vanhanen in 2010 and Katainen in 2014 and the appointments of Kiviniemi and Stubb (the respective new leaders of the Centre and the National Coalition) as the PM required both the resignation of the government and the appointment by the president of the new cabinet. A rare piece of major government drama occurred in 2003, when PM Jäätteenmäki had to resign in June after allegations concerning her use of secret foreign ministry documents during the election campaign. The rift occurred mainly between the two largest cabinet parties, the Centre and the Social Democrats. The same three coalition parties formed a new cabinet immediately after Jäätteenmäki had resigned (Arter 2006: 217–37). In addition, small coalition partners have left the governments: the Rural Party in 1990 over budgetary disagreements, the Christian Democrats in 1994 owing to the government’s proEU stance, the Green League in 2002 and in 2014 over disputes concerning nuclear energy, and the Left Alliance in 2014 over disagreements about economic policy. However, these defections, all of which took place the year before the next scheduled elections, did not threaten the survival of the surplus coalition cabinets. Finally, in June 2017 the government survived the split in the Finns Party when the Blue Reform continued in the cabinet.

2011-04-17 2014-03-25

2014-06-24

1994-06-28 1995-04-13 1999-04-15

2002-05-31 2003-04-17

2003-06-24 2007-04-19

Aho II Lipponen I Lipponen II

Lipponen III Jäätteenmäki

Vanhanen I Vanhanen II

Mari Kiviniemi 2010-06-22 Katainen I 2011-06-22

Katainen II

2014-03-25

2007-03-18 2010-06-18

1990-08-28 1991-04-26

Holkeri II Aho I

2003-03-16 2003-06-18

1995-03-19 1999-03-21 2002-05-31

1991-03-17 1994-06-20

1990-08-24

1987-04-30

Holkeri I

Date out

Date in

Cabinet

23.3

100 72.1

100 79.2

100 4.3

100 100 79.8

100 80.9

85.5

Relative duration (%)

Table 6.6 Cabinet termination in Finland, 1987–2018

9

1 7a

1 9

1 7b

1 1 7a

1 7a

7a

Mechanisms of cabinet termination

Terminal events

KOK, SDP, RKP, SKL, VIHR, SKDL

SDP, KESK

SDP, KOK, RKP, VIHR, SKDL

SKL, KESK

SMP, KOK EU membership

Conflict over budget

Comments

Continued

SKDL withdrew from the Foreign affairs, Social cabinet due to reductions in welfare provisions in the 2015 affairs budget proposal. Katainen withdrew as both party chairman of KOK and as PM in order to pursue an international career.

Vanhanen stepped down voluntarily.

Jäätteenmäki resigned following accusations of leaking classified information about the Iraq war.

Environment VIHR withdrew from the cabinet after a decision to build new nuclear power plants was made.

Foreign affairs

Finance, Social affairs

Policy Parties (when conflict between area(s) or within)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2014-06-24

2014-09-20 2015-05-29

Stubb I

Stubb II Sipilä I

2015-04-19 2017-06-13

2014-09-20

Date out

100 52.7

29.3

Relative duration (%)

1 8

7a

Mechanisms of cabinet termination

Terminal events

SMP

KOK, SDP, RKP, SKL, VIHR

Comments

Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event Note: Relative duration: share of constitutionally allowed duration for this cabinet.

SMP (The Finns) split into two (SMP and SIN) after Jussi Halla-aho was elected as chair of the party in June 2017. Cabinet members of the old SMP stayed in the cabinet but defected from SMP to SIN.

Environment VIHR withdrew from the coalition after a positive cabinet vote on allowing the construction of a new nuclear power plant in northern Finland.

Parties (when Policy conflict between area(s) or within)

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of PM Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

Date in

Cabinet

Table 6.6 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



201

Conclusions Coalition governance in Finland displays considerable stability. Many of the features discussed in this chapter were already present during the era of strong presidency, especially the large ideologically heterogeneous cabinets, the limited autonomy of both the PM and the line ministers, the ministerial committees, and the informal ‘evening school’ sessions. Nonetheless, the period of major constitutional reform from the late 1980s onwards and especially the 1990s can be seen as a watershed. Until then the cabinets were primarily short-lived and based largely on the premise that each coalition party was responsible for its own turf, resembling thus the ministerial government model outlined in Chapter 2 of this volume, whereas since the 1980s it is expected that cabinets will serve the whole four-year electoral term. As the government became the main executive and with the president no longer expected to intervene, the political parties simply needed to agree on rules about the formation and work of coalition cabinets.¹⁸ Government formation is led by the largest party in terms of Eduskunta seats, and the bargaining environment is definitely different from the Cold War era when the president dominated the process and also the Soviet Union cast its shadow over cabinet formation. But while such external factors have disappeared, the changing cleavage structure and the emergence of populism have complicated the situation. The left–right dimension continues to be the main axis of contestation in the party system, with questions about state finances and the welfare state still in a central role in government formation talks. Such questions also continue to be the ones producing most conflict among the coalition partners. At the same time the sociocultural dimension has emerged as the second cleavage, especially since the mid-1990s through EU membership and the broader internationalization of Finnish society. When putting together governments, the formateurs have needed particularly to take into account party positions regarding European integration. Yet the impact of populism should not be exaggerated. The Finns Party achieved an electoral breakthrough in the 2011 elections and first entered the government in 2015. The party effectively split into two camps in 2017, and in the run-up to the Eduskunta elections scheduled for spring 2019 some parties ruled out sharing power with the more hard-line version of the Finns Party—the first time in decades Finnish parties were making such pre-election promises. Finnish cabinets continue to be broad coalitions, in most cases including parties from both the left and the right. The fragmented party system, with no party winning more than around a quarter of the votes in the elections, contributes to

¹⁸ Elements of continuity and change can be best captured by comparing this chapter with earlier research on Finnish governments. Nousiainen (1994, 1996, 2000) covers the post-Second World War period until the mid-1990s, whereas Paloheimo (2002, 2003, 2005, 2016) extends the coverage up to the very first years of the twenty-first century.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

202



the formation of ideologically quite heterogeneous coalitions that contain at least three and, in most cases, more parties. This internal heterogeneity also in large part explains the importance attached to the government programme and to the ex ante coalition management mechanisms. The government programmes have become considerably longer and more detailed since the late 1990s and it is expected that the cabinet parties and their parliamentary groups respect them. Furthermore once the cabinet has entered into office, there are institutionalized procedures for solving disputes among coalition partners. Apart from ministerial committees, the most salient issues are discussed informally between the leaders of coalition parties. Such informal talks are held when needed and they are essential for cabinet survival. While over time comparisons are difficult to make, it appears that the ministerial committees and particularly the informal talks between leaders of coalition parties are more important as conflict management mechanisms than before the 1990s. The PM has emerged from the shadow of the president as the undisputed political leader of the country. Yet the PM and the individual line ministers are strongly constrained by the very procedures discussed in this chapter—the government programme and the firmly entrenched practices of cabinet decisionmaking. In line with the coalition compromise model (Chapter 2), PMs are obviously expected to provide leadership, but they must respect the established rules while paying close attention to the preferences of the coalition partners. Managing multiparty coalitions is first and foremost a matter of trust: trust between the coalition parties and trust between the PM and individual cabinet parties.

References Arter, David (2006). Democracy in Scandinavia: Consensual, Majoritarian or Mixed? Manchester: Manchester University Press. Arter, David (2009). ‘From a Contingent Party System to Party System Convergence? Mapping Party System Change in Postwar Finland’. Scandinavian Political Studies, 32(2): 221–39. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999-2010’. Party Politics, 21(1): 143–52. Bengtsson, Åsa, Kasper M. Hansen, Ólafur T. Hardarson, Hanne Marthe Narud, and Henrik Oscarsson (2014). The Nordic Voter: Myths of Exceptionalism. Colchester: ECPR Press. Borg, Sami (ed.) (2012). Muutosvaalit 2011. Helsinki: Oikeusministeriön selvityksiä ja ohjeita 16/2012.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



203

Duverger, Maurice (1980). ‘A New Political System Model: Semi-presidential Government’. European Journal of Political Research, 8(2): 165–87. Forestiere, Carolyn (2008). ‘New Institutionalism and Minority Protection in the National Legislatures of Finland and Denmark’. Scandinavian Political Studies, 31 (4): 448–68. Grönlund, Kimmo, and Hanna Wass (eds) (2016). Poliittisen osallistumisen eriytyminen – Eduskuntavaalitutkimus 2015. Helsinki: Oikeusministeriön julkaisu 28/2016. Jungar, Ann-Cathrine (2002). ‘A Case of Surplus Majority Government: The Finnish Rainbow Coalition’. Scandinavian Political Studies, 25(1): 57–83. Karvonen, Lauri (2014). Parties, Governments and Voters in Finland: Politics under Fundamental Societal Transformation. Colchester: ECPR Press. Karvonen, Lauri (2016). ‘No Definite Decline. The Power of Political Parties in Finland: A Focused Analysis’. In Lauri Karvonen, Heikki Paloheimo, and Tapio Raunio (eds), The Changing Balance of Political Power in Finland. Stockholm: Santérus Academic Press, 91–125. Karvonen, Lauri, Heikki Paloheimo, and Tapio Raunio (eds) (2016). The Changing Balance of Political Power in Finland. Stockholm: Santérus Academic Press. Kekkonen, Sirpa, and Tapio Raunio (2011). ‘Towards Stronger Political Steering: Program Management Reform in the Finnish Government’. In Carl Dahlström, B. Guy Peters, and Jon Pierre (eds), Steering from the Centre: Strengthening Political Control in Western Democracies. Toronto: University of Toronto Press, 241–60. Kestilä, Elina (2006). ‘Is There Demand for Radical Right Populism in the Finnish Electorate?’. Scandinavian Political Studies, 29(3): 169–91. Koskimaa, Vesa (2016). Towards Leadership Democracy: The Changing Balance of Power in Three Finnish Political Parties 1983–2012. Tampere: Tampere University Press, Acta Universitatis Tamperensis 2224. Müller, Wolfgang C., and Kaare Strøm (eds) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Murto, Eero (2014). Virkamiesvaltaa? Ministerien ja virkamiesten väliset valtasuhteet Suomessa viime vuosikymmenien aikana. Tampere: Tampere University Press. Niemikari, Risto, Tapio Raunio, and Tom Moring (2019). ‘Finland – Informal Interdependence and Occasional Clashes’. In Karl Magnus Johansson and Gunnar Nygren (eds), Close and Distant: Political Executive-Media Relations in Four Countries. University of Gothenburg: NORDICOM, 29–54. Nousiainen, Jaakko (1994). ‘Finland: Ministerial Autonomy, Constitutional Collectivism, and Party Oligarchy’. In Michael Laver and Kenneth A. Shepsle (eds), Cabinet Ministers and Parliamentary Government. Cambridge: Cambridge University Press, 88–105. Nousiainen, Jaakko (1996). ‘Finland: Operational Cabinet Autonomy in a PartyCentred System’. In Jean Blondel and Maurizio Cotta (eds), Party and Government: An Inquiry into the Relationship between Governments and Supporting Parties in Liberal Democracies. Basingstoke: Macmillan, 110–27.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

204



Nousiainen, Jaakko (2000). ‘Finland: The Consolidation of Parliamentary Governance’. In Wolfgang C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 264–99. Nousiainen, Jaakko (2001). ‘From Semi-presidentialism to Parliamentary Government: Political and Constitutional Developments in Finland’. Scandinavian Political Studies, 24(2): 95–109. Nousiainen, Jaakko (2006). ‘Suomalainen parlamentarismi’. In Eduskunnan muuttuva asema. Suomen eduskunta 100 vuotta, osa 2. Helsinki: Edita, 179–335. Paloheimo, Heikki (2002). ‘Pääministerin vallan kasvu Suomessa’. Politiikka, 44(3): 203–21. Paloheimo, Heikki (2003). ‘The Rising Power of the Prime Minister in Finland’. Scandinavian Political Studies, 26(3): 219–43. Paloheimo, Heikki (2005). ‘Finland: Let the Force Be with the Leader – But Who Is the Leader?’. In Thomas Poguntke and Paul Webb (eds), The Presidentialization of Politics: A Comparative Study of Modern Democracies. Oxford: Oxford University Press, 244–66. Paloheimo, Heikki (2016). ‘The Changing Balance of Power between President and Cabinet’. In Lauri Karvonen, Heikki Paloheimo, and Tapio Raunio (eds), The Changing Balance of Political Power in Finland. Stockholm: Santérus Academic Press, 57–90. Paloheimo, Heikki, and Tapio Raunio (eds) (2008). Suomen puolueet ja puoluejärjestelmä. Helsinki: WSOY. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of Antielitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Raunio, Tapio (2011). ‘Finland: Moving in the Opposite Direction’. In Torbjörn Bergman and Kaare Strøm (eds), The Madisonian Turn: Political Parties and Parliamentary Democracy in Nordic Europe. Ann Arbor: University of Michigan Press, 112–57. Raunio, Tapio, and Thomas Sedelius (2020). Semi-presidential Policy-Making in Europe: Executive Coordination and Political Leadership. Basingstoke: Palgrave Macmillan. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Tiili, Minna (2003). Ministerit strategisina johtajina: Tutkimushankkeen loppuraportti. Helsinki: Helsingin yliopiston yleisen valtio-opin laitos ja Valtiovarainministeriön hallinnon kehittämisosasto ja Edita. Tiili, Minna (2008). Ministers as Strategic Political Leaders? Strategic Political Steering after NPM Reforms in Finland. Helsinki: Acta Politica 34, Department of Political Science, University of Helsinki.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



205

Valtioneuvoston kanslia (VNK) (2015). Ministerin käsikirja. Helsinki: Hallituksen julkaisusarja 9/2015. Virtanen, Petri, Petri Uusikylä, Janne Jalava, Seppo Tiihonen, Lasse Laitinen, and Kirsi Noro (2016). Valtioneuvoston yhtenäisyys – kansainvälinen vertaileva tutkimus. Helsinki: Valtioneuvoston kanslia, Valtioneuvoston selvitys- ja tutkimustoiminnan julkaisusarja 49/2016. Westinen, Jussi (2014). ‘True Finns: A Shock for Stability? Testing the Persistence of Electoral Geography in Volatile Elections’. Scandinavian Political Studies, 37(2): 123–48. Westinen, Jussi (2015). Cleavages in Contemporary Finland: A Study on Party-Voter Ties. Åbo: Åbo Akademi University Press. Wiberg, Matti (2011). ‘Hallituspuolueiden eduskuntaryhmien yhteistoimintasäännöt’. Politiikka, 53(4): 321–2.

Appendix. List of political parties Abbreviation SKDL TPSL SDP VIHR KESK SIN SMP LKP SKL RKP KOK

Name Left Alliance (Vasemmistoliitto), 1990– Finnish People’s Democratic League (Suomen Kansan Demokraattinen Liitto), before 1990 Social Democratic Union of Workers and Smallholders (Työväen ja Pienviljelijäin Sosialidemokraattinen Liitto) Social Democratic Party of Finland (Suomen sosialidemokraattinen puolue) Green League (Vihreä liitto) Centre Party of Finland (Suomen Keskusta) Blue Reform (Sininen tulevaisuus) Finns Party (Perussuomalaiset), 1995– Finnish Rural Party (Suomen maaseudun puolue), before 1995 Liberal People’s Party (Liberaalinen kansanpuolue), 1965–2011 People’s Party of Finland (Suomen Kansanpuolue), 1951–1965 Christian Democrats (Kristillisdemokraatit, before 2001 Suomen Kristillinen Liitto) Swedish People’s Party of Finland (Suomen ruotsalainen kansanpuolue) National Coalition Party (Kansallinen Kokoomus)

Note: Party names are given in English, followed by the party name in Finnish in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 7 France: Electoral Necessity and Presidential Leadership Beyond Parties Isabelle Guinaudeau and Simon Persico

Scholars of the French polity have seldom studied coalition-making between political parties (Bué and Desage 2009).¹ In classic textbooks and handbooks dealing with French politics and institutions, the word ‘coalition’ hardly appears at all (e.g. Chevallier et al. 2004; Safran 2008; Elgie et al. 2016). The lack of interest might stem from the peculiar way coalitions are formed, maintained, and terminated in France, where coalition politics work differently from most other West European parliamentary democracies (Thiébault 2003). There are two main reasons for this specificity. First, due to the majoritarian two-round electoral system, parliamentary elections often result in a one-party parliamentary majority, which leaves little room for post-electoral coalition bargaining. Coalition agreements mostly take place prior to elections and mostly consist of pre-electoral deals in which the coalition’s senior party grants a few seats to its partner after both parties agree on a brief policy document. A party’s prospects to take part in a government coalition hence depends on its proximity to one of the big government parties. These have traditionally been the successive right-wing parties in the Gaullist tradition and the Socialist Party, and since 2017 Emmanuel Macron’s La République En Marche (LREM). In a highly bipolar party system, this means being close to those parties along the left–right division and within a realm of acceptability, which until now has excluded the far right. Moreover, a party must have some ‘political weight’, which can be reflected in its number of parliamentary seats, its performances at the last presidential or other second-order elections, or its past alliances. Second, in a semi-presidential regime where the executive enjoys increasing powers (Grossman and Sauger 2009), coalition members play a small role ¹ This chapter is based on the analysis of archive material and interviews. We have been able to retrieve several coalition agreements, data on election outcomes, and the composition of the assembly, taken in particular from the Journal Officiel, the government and National Assembly online archives, and the Comparative Political Data Set (Armingeon et al. 2019), press archives from Le Monde (using LexisNexis), and also used information gathered in Wikipedia. We also led 17 semi-structured interviews with (former) party leaders, ministers from all governments since 1981, members from some parliamentary majorities, and chief advisers. Two of our oral sources are cited by name; they have both agreed by written consent to this. Isabelle Guinaudeau and Simon Persico, France: Electoral Necessity and Presidential Leadership Beyond Parties In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0007

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



207

compared to the president (or the prime minister in times of cohabitation).² Cabinet formation and portfolio allocation rest in the discretionary power of the chief of the executive and no real (in)formal coordination or negotiation takes place. The chief of the executive also exclusively presides over government’s reshuffles, unless—a very rare case—a party decides to leave the coalition. Cabinet termination, as a result, mostly takes place following new national (presidential or legislative) elections. Noël Mamère, one of the leaders of the (left-ecologist) Green party and candidate in the 2002 presidential election, argued in our interview that French coalitions were nothing like ‘government coalitions in the Anglo-Saxon or German sense of the term’. Rather, they amounted to ‘electoral agreements linked to the electoral system constraining small parties to be in the hand of a mainstream party needing allies to show its openness during elections. [ . . . ] In order to foster the acceptance of the electoral agreement, junior parties usually obtain only one or two emblematic policy concessions’.³ In strict definitional terms, it makes sense to speak of coalition politics (and to include France in a comparative volume on government coalitions!) insofar as this political system features political parties that sometimes form coalition cabinets. However, as will be discussed in this chapter, French coalition politics do not imply joint structured agreements and stable coalition formation and governance mechanisms, as is common in pure parliamentary systems. It is rather based on bilateral and often informal agreements between one main party and one or several separate junior partners. More generally, given the power asymmetries between the main coalition party and its junior partners, France can be characterized as an extreme case of dominant Prime Minister model. The chief of the executive, that is the president in the absence of cohabitation, predominates (Dunleavy and Rhodes 1990; Laver and Shepsle 1996; Bergman et al. 2019; see also Chapter 2 in this volume). As we will see, he has very large autonomy in coalition building and in coalition governance, leaving single ministers (and to an even stronger extent junior coalition parties) virtually no leeway to shape government policies against his position. Since the beginning of this century, France has undergone major institutional and political transformations that have reinforced those dynamics. At the institutional level, the synchronization of presidential and legislative elections has increased the weight of the president in coalition bargaining and management. The party system has also evolved considerably. Both traditional governing parties—the social-democratic Parti Socialiste (PS) and the conservative Les Républicains (LR)—have suffered a severe decline. In parallel, two strong challengers have become forerunners: the far-right Rassemblement national (formerly Front national) and, in very recent times, the breakthrough of LREM, a catch-all ² Power sharing between the president and a parliamentary majority. ³ Interview with Noël Mamère, 26 May 2017, in Bègles, France.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

208

  

organization with liberal stances, gathering among others former members of the PS and LR around Macron after his 2017 presidential victory. These changes have increased the dimensionality of the political space but have left minor parties as powerless as before in terms of coalition bargaining.

The institutional setting In France, coalition building and coalition management take place within a series of constraints linked to several institutional features. The political system is semipresidential, with an increasingly strong president and asymmetrical bicameralism. Both presidential and legislative elections lead to the formation of a new government. The head of state (the president) has the exclusive power of nominating the prime minister (Art. 8, al. 1 of the constitution⁴). There is de jure no restriction to this power but there is a strong norm that the prime minister should stem from the party that has a parliamentary majority.⁵ If no party obtains an absolute majority—a rare case—the president approaches the leader from the party with the largest seat share. The French constitution foresees that the prime minister then proposes a list of ministers to the president who nominates them by presidential decrees, countersigned by the prime minister (Art. 8, al. 2). Note that except the references to the functions of prime minister and of minister of justice (Art. 65), the constitution does not provide any indication regarding the number of ministers and the allocation of competences. Over the first decades of the Fifth Republic, French politics were dominated by a coalition between centrist and right-wing parties (see Table 7.1a). Presidential and legislative elections were systematically desynchronized but came up with rather stable results with both centre-right parties getting a strong majority. Since 1981, by contrast, frequent political alternations between left and right have taken place both at the presidency and in government. In this context, Socialist President François Mitterrand dissolved the National Assembly after both the 1981 and 1988 elections, in order to obtain a parliamentary majority. Because of the shorter parliamentary term, he had to face parliamentary elections (in 1986 and in 1993), which his party (PS) lost, resulting in two phases of cohabitation. In 1995, the newly elected president Jacques Chirac already had a parliamentary ⁴ Following this Article, ‘the President of the Republic nominates the prime minister. He terminates his functions when the prime minister presents the government’s resignation. Following the proposition of the prime minister, he nominates the other members of government and terminates their functions’ (authors’ translation). ⁵ When a new legislative majority different from the presidential majority is elected, the president chooses a prime minister from this new majority, which leads to cohabitation. Note that, in contrast to countries with symmetrical bicameralism, government is only responsible to the National Assembly (lower house) and not to the Senate, so that bicameralism does not significantly affect French coalition politics. For a comparison between France and Italy in this respect, see di Virgilio et al. (2015).

Pompidou I

Pompidou II Pompidou III Pompidou IV Couve de Murville Chaban Delmas Messmer I

2

3 4 5 6

Messmer II and III Chirac I

Barre I

Barre II

Barre III Mauroy Ia Mauroy II and III

10

11

12

13 14 15

9

8

7

1959-01-20

Debré

1

1978-04-06 1981-05-22 1981-06-23

1977-04-01

1976-08-27

1974-06-08

1973-04-12

1972-07-06

1969-06-22

1962-12-06 1966-01-08 1967-04-07 1968-07-12

1962-04-15

Date in

Cabinet Cabinet number

Table 7.1a French cabinets since 1959

1981-06-21

1978-03-19

1973-03-11

1967-03-12 1968-06-30

1962-11-25

1958-11-30

UDR, RI, RCDS UDR, RI, RCDS RPR, RI, RCDS RPR, UDF PS, MRG PS, MRG, PC mwc min sur

sur

sur

sur

56.5 41 67

61.6

61.6

61.6

61.6

79.5

79.5

55.6 55.6 49.7 72.7

73.7

sur mwc mwc min sur

73.7

sur

UDR, RI, sur PDM UDR, RI, sur PDM UDR, RI, UC sur

UNR, CNIP, MRP, Radical Party UNR, CNIP, MRP UNR, RI UNR, RI UNR, RI UNR, RI

490 490 491

490

490

490

490

487

487

482 482 487 487

579

579

Number of Cabinet Type Election date Party strength in seats in composition of parliament cabinet seats (%) of cabinet

4 4 4

4

6

6

6

5

5

6 6 5 5

6

6

Number of parties in parliament

4.22 4.17 2.76

4.41

4.41

4.41

4.41

2.49

2.49

3.51 3.51 3.73 2.49

4.84

4.84

CD&CR CD&CR S

CL

CL

CL

CL

G

G

G G CL G

G

G

ENP, Median party* parliament in first policy dimension

Continued

Formal support parties

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1991-05-17 1992-04-04 1993-03-30 1995-05-18 1997-06-04

2002-05-07 2002-06-17

Cresson Beregovoy Balladur Juppé I and II Jospin

Raffarin Ia Raffarin II and III de Villepin Fillon Ia Fillon II Ayrault Ia

Ayrault II

Valls I Valls II Valls III

20 21 22 23 24

25 26

31

32 33 34

2014-04-09 2014-08-26 2016-02-11

2012-06-21

2005-06-02 2007-05-18 2007-06-19 2012-05-16

1984-07-23 1986-03-20 1988-05-13 1988-06-28

Fabius Chirac II Rocard Ia Rocard II

16 17 18 19

27 28 29 30

Date in

Cabinet Cabinet number

Table 7.1a Continued

2012-06-17

2007-06-17

2002-06-16

1997-06-01

1993-03-28

1988-06-12

1986-03-16

UMP UMP UMP, NC PS, Greens, PRG PS, Greens, PRG PS, PRG PS, PRG PS, PRG, Ecologist party

PS, MRG RPR, UDF PS, MRG PS, MRG, UDC PS, MRG, GE PS, MRG RPR, UDF RPR, UDF PS, PC, Greens, PRG, MDC UMP UMP

sur sur sur

sur

mwc sur sur min

min mwc

min min mwc mwc sur

sur mwc min sur

54.9 54.9 55.6

54.9

68.3 68.3 59.4 36.7

43.8 68.3

47.7 47.7 81.8 81.8 55.3

58.2 49.6 36.9 54.8

577 577 577

577

577 577 577 577

577 577

577 577 577 577 577

491 577 577 577

Number of Cabinet Type Election date Party strength in seats in composition of parliament cabinet seats (%) of cabinet

7 7 8

7

6 6 7 7

5 6

5 5 4 4 5

4 5 5 5

Number of parties in parliament

2.82 2.82 2.82

2.82

2.18 2.18 2.41 2.41

3.43 2.18

3.38 3.38 2.87 2.87 3.43

2.76 3.77 3.77 3.38

S S S

S

G G G G

S G

S S CD&CR CD&CR S

S CD&CR CD&CR S

ENP, Median party* parliament in first policy dimension

Formal support parties

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2017-05-17

Philippe Ia

Philippe II

36

37

2017-06-18

PS, PRG, Ecologist party LREM, MODEM, PRG LREM, MODEM, PRG sur

min

sur

62.7

55.6

577

577

577

9

8

8

3.08

2.63

2.82

LREM

S

S

Notes: For a list of parties, consult the chapter appendix. a = limited policy remit * refers to the party groupings (see below). Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1959–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015). The election date refers to the parliamentary election. The first policy dimension is economic left–right. The total number of seats in parliament is 579 during 1959–1962, 482 during 1962–1967, 487 during 1967–1973, 490 during 1973–1981, 491 during 1981–1986, and 577 from 1986. Cabinet types: min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition. Cabinet strength in seats was calculated based on the website on the National Assembly (http://www.assemblee-nationale.fr/elections/historique-2.asp) and the Wikipedia encyclopaedia, which delivers congruent information. Table legends and variables are further defined in the Appendix, this volume.

2017-06-21

2016-12-06

Cazeneuve

35

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

212

  

majority but provoked legislative elections in 1997, which his conservative party RPR and its allies lost, causing a third episode of cohabitation. Since 2002, the presidential term has been reduced to five years, which led to the alignment of presidential and legislative electoral cycles. In this context, electorally induced coalition building has become less frequent and more regular (it takes place every five years, following national elections). The 2002 reform also gives more weight to the presidential election, which takes place first: since 2002, legislative elections have granted a strong majority to the coalition of parties supporting the president. This alignment of presidential and legislative elections had important consequences on government, as it translated into the disappearance of cohabitations. The relative weight of the president and the prime minister in shaping coalition politics—especially through the appointment of ministers (Montay 2013; Sponchiado 2017)—used to depend on whether, or not, there was a cohabitation (Duverger 1978). When the prime minister and the president belong to the same party, the president is, in practice, free to choose the prime minister, and he will be present in coalition bargaining. By contrast, in times of cohabitation, the prime minister had the most say (Conley 2007).⁶ The reduction of the presidential term to five years has strengthened the role of the president, who is now the leader of coalition negotiations (Duverger 2004: 98–122; François 2011).⁷ Since governmental coalitions are responsible to the National Assembly, the two-ballot majority electoral system used for national elections crucially shapes coalition politics. France confirms Golder’s (2005) finding that disproportional systems encourage pre-electoral coalitions in party systems with a sufficiently large number of parties (see also Duverger 1954). Importantly, this system favours the biggest parties—who long remained the only ones having a chance to maintain themselves in the second round. At the presidential election, smaller parties are disadvantaged by the run-off voting system. At parliamentary elections, for which only candidates receiving the vote of at least 12.5 per cent of registered voters can run in the second round, smaller parties rely on pre-electoral alliances to access parliament. Mainstream parties accept not to run in a negotiated number of districts, in order to allow some smaller parties’ candidates to make it to the second round and have a chance to get elected. This offers main government parties strong bargaining power towards smaller parties while creating strong incentives for all

⁶ The president keeps nonetheless a veto. There are historical examples of presidents rejecting the prime minister’s choice and asking him or her to choose someone else. (The example of Mitterrand regarding the choice of the ministers of foreign affairs and defence is cited in Adler 2015; after Chirac became president in 1995, he also used his veto during the formation of the Jospin government in 1997. However, this veto was exerted regarding individual persons and not regarding coalition formation.) ⁷ The attempt to gather as many components as possible of the centrist and right-wing conservative parties in one single party, the Union pour un movement populaire (UMP), created in 2002 between the presidential and the legislative elections to provide President Chirac with a majority, can also be seen as a consequence of this new institutional logic (Haegel 2012). Even if this did not succeed in the long-term, it had implications on coalition building in the 2000s.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



213

parties to agree on electoral alliances (Golder 2006; see Andolfatto and Greffet 2008 for the example of the French Communist party). At the party level, it also deters dissidents from splitting, as a tiny party has no chance to achieve representation at the national level (Gallagher and Mitchell 2008). In sum, the majoritarian electoral system, associated to the synchronization of presidential and legislative elections, gives huge leverage to the two largest parties during the campaign and, after the election, to the de facto leader of the winning party, the president.

The party system and the actors The mainstream parties’ predominance that is induced by the institutions of the Fifth Republic is challenged by the growing fragmentation of the French party system. However, while this is important for representation and politics in general, this has had little impact, so far, on coalition politics in cabinets.

Party system change An overview of the evolution of the relative strength of parties can be gathered from the distribution of parliamentary seats in Table 7.1a. The first decades of the fifth Republic witnessed the progressive structuration of an initially highly fragmented party system inherited from the Fourth Republic around two main blocs (Evans 2003). The right-wing bloc remained predominant until 1981. Until 1974, this bloc was organized around the UNR (then UDR), a conservative party in direct filiation with Fifth Republic’s founder and first president Charles De Gaulle. Gaullists benefited from the support from most non-Gaullist (more liberal) rightwing members of parliament (MPs), gathered in Valery Giscard d’Estaing’s parliamentary group Independent Republicans.⁸ As discussed later, the left-wing bloc also progressively structured itself around the Socialist party (PS). Each bloc was relatively stable, but subject to internal competition (on the left, between Communists and Socialists; on the right, between centrist/moderate and conservative parties), so that Duverger (1985) spoke of a quadrille bipolaire. The year 1981 marks a turning point in French political history. After decades of exclusive predominance of centre-right coalitions, the Socialist candidate François Mitterrand won the presidential election and, after he dissolved the National Assembly, obtained an absolute parliamentary majority. The PS initially ⁸ D’Estaing later founded a political party of his own, the Union pour la démocratie française (UDF), gathering multiple centre-right non-Gaullist organizations. Between 1978 and 2002, the UDF acted as the successive Gaullist parties’ main coalition partner.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

214

  

formed a surplus coalition government with the Communists, who left the government in 1984 because of major divergences on macroeconomic policies. Since 1981, France has experienced frequent political alternations, with coalitions led by either the PS or the main conservative party (successively named RPR, UMP, and LR). The 1986 parliamentary elections led to the first experience of cohabitation between a Socialist president and an RPR–UDF coalition government led by Chirac. In 1988, François Mitterrand was re-elected and immediately dissolved the National Assembly, but the PS did not control a majority of seats and built a single-party minority government facing opposition both from the Communists and from centre-right parties. The 1993 legislative elections gave the RPR–UDF alliance a parliamentary majority, which resulted in a second cohabitation between Mitterrand and Prime Minister Edouard Balladur. The RPR candidate Chirac won the 1995 presidential election, but after he dissolved the Assembly in 1997, the alliance between the PS, Greens, and Communists obtained a parliamentary majority and formed a government under the leadership of Lionel Jospin, who remained prime minister until 2002 (the third and so far last cohabitation experience). During the 1990s, two transformations in the party system progressively put an end to the bipolar order that had structured French politics since the 1950s. The first noticeable evolution is the decline of the Parti Communiste (PC): while this party attracted close to 21 per cent of the votes in the first ballot of the 1978 legislative elections, it progressively declined to below 5 per cent in the 2002 and 2007 elections—and from 15.4 per cent in the 1981 presidential election to 1.9 per cent in 2007. This party was challenged, on the far left, by multiple parties (in particular the Trotskyist Lutte Ouvrière and Ligue Communiste Révolutionnaire) that achieved higher scores in the late 1990s and early 2000s. More recently, the creation of the Front de Gauche (now called France Insoumise), an electoral alliance led by the Socialist dissident Jean-Luc Mélenchon, has given a new dynamic to the far-left vote. However, this new party has consistently refused, so far, to enter any coalition with the PS. In parallel, the Green party has become more relevant in French coalition politics.⁹ After breakthroughs in the 1989 municipal and European elections, ecologists received more than 14 per cent of the votes at the 1992 regional elections.¹⁰ Given this strength and the concomitant decline of the Communists, the Green party imposed itself as a relevant coalition partner for the PS.¹¹ In 1997, ⁹ While the party was created in 1984, ecologist candidates and movements have run in French national elections since the 1974 presidential election, with only very limited scores and overall political impact. ¹⁰ There were initially two ecologist parties—the Greens and a less durable party, Génération Ecologie—that agreed on electoral alliances for the 1992 and 1993 elections. ¹¹ At the regional level, the ecologist parties’ spatial location along the left–right spectrum and the distribution of seats allowed them to be pivotal in coalition formations on both the left and the right (Brouard 1999). However, the emergence of the Greens took place above all at the cost of the PS and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



215

for the first time, the Greens joined the left electoral alliance under the leadership of the PS, which left 29 districts to Green candidates (while the Greens conceded 79 districts to the Socialists, see Boy 2002). This allowed the ecologists to be represented in the National Assembly for the first time and to enter government. Since then, they have negotiated electoral alliances with the PS for each parliamentary election, without success in 2002 and 2007. After the election of Socialist François Hollande for president in 2012, the Greens entered the PS-led coalition government. Transformations occurred also within the right-wing bloc, in particular with respect to the relations between centrists and conservative parties. These two political families have built a durable electoral alliance but have also challenged each other within the right-wing bloc. This relation was embodied in the 1970s by the competition between the centrist president Giscard d’Estaing and Chirac, who respectively founded new parties, the UDF and the RPR. Each of the two parties have most often presented their own candidate to the presidential election. This division was partially bridged after the 2002 presidential election with the creation of UMP, which aimed at gathering moderate right-wing parties around President Chirac (Haegel 2002; Sauger 2008). Most of the UDF office holders joined the UMP, even though the UDF continued to exist and managed to maintain its own parliamentary group. At the 2007 presidential election, its leader François Bayrou gathered 18 per cent of the votes, did not support the UMP candidate Nicolas Sarkozy, and founded a new party, the Mouvement Démocrate (MODEM), just after the election. This new party was supposed to break up France’s bipolarization and be independent from left and right,¹² but it has so far failed to conquer a pivotal position: only right-wing parties consider this party as a potential coalition partner. The rise of the radical right National Front (FN) is undoubtedly the most important evolution in French politics over the last three decades. Following the pioneering work by Grunberg and Schweisguth (1997), many scholars have observed that the French political space is no longer bipolar but structured in three blocs: left, right, and far right, embodied by the FN. This party made a breakthrough in the 1984 European elections, gathering close to 11 per cent of the votes, and obtained 35 parliamentary seats at the 1986 parliamentary elections, favoured by the proportional voting system adopted by the Socialist government (and immediately abolished by the next government). The FN has since attracted a growing number of votes in national, local, and European elections. The party often remains in the second ballot of two-round elections, which perturbs PC, which shows that this party was particularly attractive to left-leaning voters. On the influence of local-level coalitions on national coalition politics, see Bué and Desage 2009: 21–2. ¹² As a reaction, those in the UDF disapproving Bayrou’s line founded the Nouveau Centre as a centre party supporting the right-wing presidential majority (see section on Electoral alliances and preelectoral coalitions).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

216

  

significantly the functioning of the majoritarian logic of the Fifth Republic, in particular regarding the building of majorities at several levels. Following the 1986 regional elections, no absolute majority was found in several regions: in five of them, RPR–UDF presidents were elected thanks to the votes of FN regional councillors (and alliances with the FN were decided in further regions). The acceptability of alliances with the FN was hotly debated within the RPR over the next decades and there have been multiple examples of alliances at the local levels until the late 1990s. Since the 2000s, this type of alliance has no longer occurred. Most of the RPR (and, then, the UMP) leaders indeed rejected these alliances on ideological and moral grounds, even though it often implied losing the constituency election. The FN adopted, itself, a strong line criticizing the collusion between mainstream parties and did not send any signal of interest in coalitions with the UMP. Its participation to the second round of many local elections, with scores suggesting its capacity to win the election, regularly raises the question for Socialists and conservatives to mutually withdraw their candidate. In the 2002 presidential election, the FN candidate Jean-Marie Le Pen eliminated the Socialist candidate Lionel Jospin from the second round—Chirac was then elected with over 82 per cent of the vote. In 2007, the right-wing candidate Sarkozy strategically took up the FN’s issues—mostly immigration—which first proved successful in diminishing the FN electoral success (Martin 2010). However, the FN has since then achieved historically high scores at the 2014 municipal and European elections and again at the 2015 departmental and regional elections, before making it once again to the second round in the 2017 presidential election and reaching the first place at the 2019 European elections. The French political space is hence more two-dimensional than ever. French voters’ attitudes and party competition are not only structured by the traditional class cleavage (essentially linked to the conflict between neoliberalism and interventionism) but also by a second dimension related to globalization and international openness, focused on immigration, European integration, globalization, and moral issues (Bornschier and Lachat 2009; Stimson et al. 2010; Tiberj 2012; Gougou and Labouret 2013). Mainstream parties (and their traditional coalition partners) have appeared to be internally divided on those two conflict dimensions—the PS especially on economic issues, LR especially on cultural issues. Yet, until 2017, coalitions were based on ideological proximity on economic matters. Coalitions founded on issues of internationalization and morality had never been considered, although some discussions took place about the desirability of a left–right convergence of those who agree on liberal reforms and the principle of international openness. This changed following the 2017 presidential election. This election represents a landmark in party system change in France since it saw none of the candidates from the two main parties (PS and Gaullists) make it

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



217

into the second round.¹³ For the first time, four candidates—Emmanuel Macron, who had founded a new catch-all party LREM, Marine Le Pen (FN), François Fillon (LR), and Jean-Luc Mélenchon (FI)—came neck and neck in the first round. This suggests a political space divided into four even blocs embodying strictly distinct policy alternatives: Macron and LREM advocate liberal policies on economic and cultural matters, in sharp contrast to the FN, which opposes immigration as much as economic globalization. LR, under the leadership of François Fillon, campaigned for neoliberal reforms and moral conservatism, while Jean-Luc Mélenchon adopted traditional left-wing positions in favour of redistribution and government intervention in the economy (see Gougou and Persico 2017 for a correspondence analysis). Macron benefited from the ‘republican front’ uniting the majority of voters anxious to keep the FN out of power. Without surprise, he obtained a clear-cut victory in the run-off, appointed ministers leaving the PS and LR in government, and obtained a large parliamentary majority a month later (Elgie 2018). The 2019 European elections confirm the changes discussed, with the FN and LREM confirming their predominance and the Greens being able to attract a considerable share of votes, while traditional government parties did not reach 10 per cent of the votes.

Electoral alliances and pre-electoral coalitions In the French context, two forms of alliances exist in view of legislative elections. The most ambitious one, referred to as ‘pre-electoral coalition’ in Table 7.1b, involves a repartition of districts (mutual withdrawal between candidates from the same pre-electoral coalition for the first round). In such alliances, candidates might even run under the flagship of all parties forming the coalition (e.g. the PS and the Greens). This indicates the willingness to form a government coalition, even though this is not always formalized or explicitly communicated. The second form of electoral alliance (‘electoral alliances’ in Table 7.1b) is limited to mutual withdrawal for the second round. This reflects the belonging to the same ‘bloc’: without such withdrawal, the chances of the other bloc winning the election would indeed be much higher.¹⁴ In many cases, coalition agreements are not explicitly subject to public communication by the parties, and the actual figures of mutual withdrawals never appear in coalition agreements.

¹³ In contrast to previous elections, in which they usually attracted more than half of the votes in the first round of the presidential election, their cumulative score fell to only 26 per cent in 2017—with the PS gaining only 6.5 per cent of the vote (Gougou and Persico 2017). ¹⁴ Accordingly, mutual withdrawal in the first round is coded in Table 7.1b as an electoral alliance associated to a pre-electoral coalition, while national-level agreements on mutual withdrawal for the second round are coded as an electoral alliance only. Note that local-level agreements, which are very frequent, are not listed in the table.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

218

  

Table 7.1b Electoral alliances and pre-electoral coalitions in France, 1988–2018 Election date

Constituent parties

Type

Types of pre-electoral commitment

1988-06-12

PS, PC PS, MRG RPR, UDF PS, PC PS, PRG Greens, GE RPR, UDF PS, PC PS, PRG PS, Greens PS, MDC RPR, UDF PS, PRG RPR, UDF PS, PRG PS, MRC UMP, NC, PR PS, PRG PS, Greens PS, PRG PS, Greens LR(?), NC, PR

EA EA EA, PEC

Written contract

1993-03-28

1997-06-01

2002-06-16 2007-06-17 2012-06-17 2017-06-18

EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC EA EA EA EA EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC

Written contract Written contract Written contract Written contract Written contract Written contract Written contract Written contract Written contract Written contract Written contract

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other Sources: Thiébault (2003), own qualitative interviews with former ministers and party officials.

The first electoral alliance emerged within the left-wing bloc, when Socialists and Communists agreed on mutual withdrawal for the second round of the 1962 parliamentary elections. This agreement was renewed for the 1967 and 1968 elections. In 1973, a further step was taken with the adoption of a common governing programme (Le programme commun) and the mutual withdrawal of candidates from both parties before the first round. The Movement of Left Radicals, a splinter party from the centrist Radical Party, joined this alliance and, contrary to the PC, has since then never ceased to form pre-electoral alliances with the PS. As shown by Table 7.1b, the relationship between the PC and the PS has fluctuated. While the agreements between both parties in 1981, 1986, 1988, and 1993 were limited to a mutual withdrawal for the second round, due to programmatic divergences, both parties, along with the Greens and the Left Radicals, reached an agreement for mutual withdrawal for the first round in 1997. Since 1997, both parties have never reached any formal electoral agreement.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



219

The Greens have also not entered electoral alliances in 2002 and 2007 but have reached an agreement with the PS regarding office and policy in 2012, as did the Left Radicals. Following Hollande’s presidential victory, the PS built an oversized coalition with his two partners. Multiple policy conflicts opposed the PS not only to its left factions, embodied by ministers like Arnaud Montebourg or Benoît Hamon, but also to the ecologists. Green ministers Cécile Duflot and Pascal Canfin decided to end their government participation in March 2014, when Hollande appointed a new government following the majority’s defeat at the municipal elections. Green MPs and left PS factions (les frondeurs) became increasingly mobilized against government policies. In August 2014, this motivated the appointment of a new government designed to be more cohesive around Hollande’s moderate positions. In February 2016, at the occasion of a government reshuffle, three former leaders of the Green party—among others the Secretary General Emmanuelle Cosse—left the Greens to join The Ecologist Party, a splitter party from the Greens formed shortly before, and joined the PSled government. These deep divisions within the left-wing bloc explain part of the result of the 2017 elections. After the Socialist primary election appointed Benoît Hamon, a leading figure of the frondeurs, the PS reached a coalition agreement with the Greens—including, for the first time, the withdrawal of the Green candidate from the presidential election. Yet, Hamon’s social and environmental political line was not supported by the liberal factions of the PS, which turned to Macron, paving the way for Macron’s victory and for Hamon a disastrous score. Electoral alliances within the right-wing bloc also go back to the 1960s. Since 1967, Gaullist parties—de Gaulle’s UNR, the RPR, the UMP, and the LR—have often formed electoral alliances with liberal and/or more centrist parties. After 1978, electoral alliances and coalition agreements gathered the Gaullist party RPR and the centrist party UDF. This foundation of the UMP translated into the fact that governments formed under Chirac’s second presidential mandate (2002–2007) were no coalition governments but comprised only UMP ministers and a few non-partisan personalities. Under Sarkozy (2007–2012), governments reflected the surplus coalition between the UMP and the Nouveau Centre, founded in 2007, to gather centrists within the presidential majority (initially with the financial support of the UMP), while the Modem refused to enter rightwing coalitions. François Bayrou supported Macron in the 2017 campaign, quoting four blurred policy priorities to justify his support. After Macron’s presidential victory, his party LREM reached a pre-electoral agreement with the Modem, which implied no policy implications but was limited to a mutual withdrawal on most constituencies, with LREM supporting Modem candidates in 75 districts. As noted by Thiébault (2003), French electoral alliances are more the product of electoral necessity than of genuine affinities between the parties. If they want to enter government, junior coalition partners have no choice but to agree on

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

220

  

electoral alliances with one of the biggest parties.¹⁵ This asymmetry shapes French coalition politics. In particular, coalitions are not negotiated collectively by all political parties involved, as in many other countries, but take the form of bilateral agreements between the leading (mainstream) party and each junior partner.

Government formation As Bédéï formulates it, ‘In France, government composition goes down to a weird and unlikely formula including unpreparedness and chance’ (2015: 9–10). When thinking about French peculiar government formation,¹⁶ three distinctive features stand out. First, presidents—prime ministers in times of cohabitation—benefit from a great autonomy when selecting cabinet members.¹⁷ Second, pre-electoral negotiations and agreements, rather than actual electoral results, have a tremendous impact on the power relationships between coalition members. Third, political parties are weak: their electoral result only partially translates into a parliamentary balance of power and parliamentarians representing those parties enjoy a very large autonomy. All this means that the president’s role in cabinet formation should not be understated (Elgie and Grossman 2016) and that postelectoral negotiations among parties are reduced to the bare bones.

The bargaining process While government formation and bargaining depend on the last parliamentary election’s results in many other democracies, it starts much sooner in France and does not last very long after the elections. Following a presidential election, all presidents have appointed a new government, even when they did not have a parliamentary majority during the few weeks leading to the parliamentary elections. As appears in Table 7.2, bargaining duration is short on average—less than

¹⁵ Additional incentives derive from French rules of public financing, which represents 40 per cent of parties’ resources on average (François and Phelippeau 2018). Parties’ state subvention is calculated based on their share of parliamentary seats but also on the number of votes obtained in the first round of the last parliamentary elections—provided the party has received at least one per cent of the votes in at least 50 districts. From the point of view of a small party, having candidates running with the support of a mainstream party therefore brings, beyond the prospect of participating in a coalition government, considerable financial resources. ¹⁶ If France is a special case, similar features, including strong presidential leadership and fluid party politics, can be observed in other countries, for instance in Romania. ¹⁷ As Safran puts it, ‘the Constitution states that the President chooses the Prime Minister and then confirms the selection by the latter of his or her cabinet colleagues. In fact, under normal conditions – that is, if the majority of the National Assembly, and hence the prime minister, belongs to the same political party as the President – the president selects most, if not all, members of government’ (2008: 201).

0 0 0 0

0

1991 1992 1993 1995

1997

Cresson Beregovoy Balladur Juppé I and II Jospin

0 0 0 0 0 1

2005 2007 2007 2012 2012 2014

2014 2016 2016 2017 2017

Valls II Valls III Cazeneuve Philippe I Philippe II

0 0 0 0 0

0 0

2002 2002

Raffarin I Raffarin II and III de Villepin Fillon I Fillon II Ayrault I Ayrault II Valls I

1

1988

Rocard II

Number of inconclusive bargaining rounds

Year in

Cabinet

UMP UMP UMP, NC PS, Greens, PRG PS, Greens, PRG PS, PRG (1) PS, Greens, PRG PS, PRG PS, PRG, PE PS, PRG, PE LREM, MODEM, PRG LREM, MODEM, PRG

PS, PC, Greens, PRG, MDC UMP UMP

PS, MRG, UDC (1) PS, UDC PS, MRG, GE PS, MRG RPR, UDF RPR, UDF

Parties involved in the previous bargaining rounds

Table 7.2 Cabinet formation in France, 1988–2018

9 1

16 1

Bargaining duration of individual formation attempt (in days)

2 3 1 6 3 10 10 1 0 0 7 2

1 0

2

17 17 2 2 1 7

Number of days required in government formation

0 0 0 11 11 0 0 0 6 0 0 0

0 0

8

5 5 0 0 0 3

Total bargaining duration

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

222

  

two days—and governments are formed after a single round of negotiations (which are not even real negotiations as we will explain). Amongst the few exceptions, the Messmer II’s and III’s, Mauroy I’s, Rocard II’s, and Valls I’s government formations have necessitated more than five days of bargaining and/or two rounds of negotiations. In these cases, contrary to most other governments, the withdrawal of former coalition parties after a reshuffle or the hesitations of a potential coalition member led to some uncertainty regarding the exact perimeter of the coalitions. Even then, government formation never required more than 17 days. The reason for this rapid government formation process is to be found in the pre-electoral nature of coalitions. The power relationships between coalition members do not really depend on their latest electoral performance, because this performance is influenced by how political parties viewed this power balance during the pre-electoral negotiations. Indeed, the number of districts reserved to each coalition parties and hence the number of elected MPs depend on preelectoral considerations. Pre-electoral negotiations thus partly determine the composition of the majority in parliament and in cabinet, at least by clarifying which parties will take part in the coalition. When the pre-electoral agreement involves mutual withdrawals for the first round and a plan to govern as a coalition, two rounds of negotiations take place separately, dealing respectively with (1) the district repartition among parties and (2) programmatic agreements. The distribution of ministerial portfolios is not subject to interpartisan bargaining before the elections. Our qualitative interviews with (former) party leaders and ministers from all governments since the early 1980s have revealed that all parties do not grant the same importance to programmatic and office considerations. For example, Greens and Communists can be seen as more policy-seeking, while Left Radicals or centrist parties, which happen to be located closer to the PS or the UMP in terms of policy positions, tend to put more emphasis on office-related considerations. Mainstream parties are mostly office- and vote-seeking: they have strong incentives to build electoral alliances with as many small parties in their ideological bloc as possible to maximize their chances to obtain a majority of votes in many districts. How much are mainstream parties ready to pay for such coalition agreements? The final answer is hard to provide, given party officials’ reluctance to provide precise information on the content of electoral agreements. These agreements do indeed carry the negative image of partisan arrangements—they are seen as deals behind closed doors to defend each party’s interests. Yet the role of the main party in the coalition is tremendous. Symbolically, negotiation talks between left-wing parties tend to take place in the PS headquarters. More importantly, many negotiations have failed because the main party of the coalition (the PS or the Gaullist party) did not leave enough seats to its coalition partners. Even when an agreement is reached, those mainstream parties might not succeed in having it

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



223

respected everywhere. Withdrawn local candidates may run as independents, hence hindering the chances of the coalition candidates. For instance, the Green party’s current leader and former negotiator David Cormand estimates that, in 2012, some 20 out of 65 districts reserved to the Greens witnessed dissident candidacies from the PS. Nevertheless, the consequences of these electoral agreements—the actual electoral results—hardly affect cabinet composition. Had they managed to get 10 more MPs, the Greens would probably have had the same number of ministers in Ayrault’s government. Electoral outcomes are not shaping the composition of government through the definition of a parliamentary balance of power, as in a pure parliamentary system: they are not exogenous insofar as they strongly depend on coalition negotiations and related electoral agreements. In this context, the president and his party are very prevalent in forming the coalition.

The composition and size of cabinets Once the elections have passed, government formation in France is different from pure parliamentary systems. In the absence of cohabitation, the president leads the negotiations mostly together with the new prime minister. Discussions take place informally with individual officials from the parties in the coalition, who are discretionarily offered ministerial positions. While composing the government, presidents and prime ministers do not only consider coalition-related arguments but also try to represent ethnic, gender, territorial origin, and age (Bédéï 2015). Since no collective bargaining among parties really takes place, junior parties can hardly influence the number of ministers they will be granted, the nature of portfolios allocated, let alone the overall composition and size of cabinets. The bargaining process rather takes the form of a ‘take-it’ or ‘leave-it’ proposal by the main government party (in fact, the president or the prime minister), with little margins for manoeuvre for coalition partners. Our interviews suggest that they are not even in the position to suggest the names of party officials they would like to enter government. It has, however, happened that party leaders vetoed the proposed appointment of a personality for minister. As a result, the decision over the number of allocated portfolios rests on the president. Table 7.3 illustrates how the number of portfolios has evolved over time. While the number of ministries has remained rather stable—around 20—for most of the period, variations are mostly due to the president reshuffling the cabinet within the same term.¹⁸ Reshuffling allows the president or the prime

¹⁸ Reshuffling is a very common practice under the Fifth Republic. Presidents tend to use government reshuffles as a way of improving their popularity (Grossman 2009) even though this does not always prove effective.

1988 1991

1992 1993 1995 1997

2007 2012

2012

2014 2014 2016

Rocard II Cresson

Beregovoy Balladur Juppé Jospin

Fillon II Ayrault I

Ayrault II

Valls I Valls II Valls III

2017

22 PS, 2 MRG, 13 Ind. 23 PS, 2 MRG, 1 GE, 4 Ind. 22 PS, 3 MRG, 3 Ind. 14 RPR, 14 UDF, 1 Ind. 16 RPR, 12 UDF, 1 Ind. 14 PS, 2 PC, 1 Greens, 1 MDC, 1 PRG 17 UMP, 1 NC, 2 Ind. 30 PS, 2 PRG, 2 Greens, 1 Ind. 32 PS, 3 PRG, 2 Greens, 1 Ind. 15 PS, 2 PRG, 15 PS, 2 PRG 14 PS, 2 PRG, 1 PE, 1 Ind. 14 PS, 2 PRG, 1 PE, 1 Ind. 13 Ind., 2 LREM, 2 MODEM, 2 PRG 9 LREM, 8 Ind., 2 PRG, 1 MODEM

Number of ministers per party (in descending order)

20

19

18

17 17 18

38

20 35

28 29 29 19

37 30

Total number of ministers

4 LREM, 1 MODEM, 1 Ind.

1 PE

1 PE

1 G, 1 LGM*

2 RPR

3 Ind. 1 Ind.

Number of watchdog junior ministers per party

17

17

18

17 17 18

21

16 19

22 24 27 16

22 20

PS Ind.a LREM

Ind.a Ind.a

PS PS PS

PS

UMP PS

PS UDF UDF PS

PS PS

PS

PS PS PS

PS

UMP PS

PS RPR RPR PS

PS PS

1 Prime 2 Finance Number minister and of economy ministries

Ind.b

Ind.b

PS

PS PS PS

PS

Ind. PS

PS RPR UDF PS

PS PS

3 Foreign office

LREM

Ind.

PS

PS PS PS

PS

PS

PS Ind. RPR -

PS PS

4 Social affairs

Ind.b

Ind.b

PS

PS PS PS

PS

UMP PS

PS RPR RPR MDC

PS PS

5 Interior

Notes: * LGM, while having no cabinet-level ministers and thus not constituting a cabinet member by our coding criteria, had a single secretary of state (Secrétaires d’État). a Minister affiliated with LR b Minister affiliated with PS Ministers who are affiliated with a political party that does not support the government coalition are considered and counted as independent ministers.

Philippe II 2017

Philippe I

Cazeneuve 2016

Year in

Cabinet

Table 7.3 Distribution of cabinet ministerships in French coalitions, 1988–2018 OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



225

minister to increase the participation of party factions or junior parties that might otherwise be too vocal against the government. This is no surprise, then, that it most often leads to an increase in the number of ministries. This also leads to an increase in the number of government-level appointments. Most of the time, those new positions are junior positions (called Secrétaires d’Etat, state secretaries), which were not accounted for in Table 7.3, but they can also be full delegated minister positions.¹⁹ The latter are cabinet members. Apart from the inflation linked to reshuffles, the number of cabinet members also tends to be higher during surplus coalitions and in left-wing governments compared to their right-wing counterparts. This might derive from left-wing coalitions associating more parties on average and/or with the PS split into multiple rival factions seeking representation in government. Finally, the most recent evolutions in cabinet composition point towards a reduction of the number of cabinet members. Although all presidential candidates since Chirac had pledged to limit the size of the cabinet, this had not been fulfilled up until recently. The Valls, Cazeneuve, and Philippe cabinets’ size would seem to indicate that this is now the case. When assessing the composition of cabinet, two features stand out. First, as in many other democracies, the number of women has increased over time and has even reached 50 per cent in the Philippe II’s cabinet. Yet, women more often hold junior positions (secretaries of state) below cabinet status and only seldom take charge of the most important ministries such as the interior, justice, defence, foreign Affairs, or the economy (see Table 7.3 and Grossman and François 2013). Second, the number of ministers with no previous political background has remained quite low. The same goes for ‘independent’ ministers. The proportion of ministers who were members of parliament varies but rarely falls below 60 per cent (Grossman and François 2013).

The allocation of ministerial portfolios As mentioned previously, the constitution does not prescribe the number of ministers and the allocation of competences. Consequently, coalition bargaining might concern not only the parties involved in government but also the distribution of portfolios and competences. However, as for most other dimensions of coalition politics in France, the chief of the executive has an overwhelming power in that regard.

¹⁹ Although delegated ministers do not control any independent administrative structure, they were counted as ministers because they can be entitled to participate in the Council of Ministers, while secretaries of state do not (they assist in the meetings only when their competences are on the agenda).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

226

  

This explains the considerable variations in the number and content of portfolios over time. Listing the more or less permanent ministerial departments in France proved impossible since the ministerial organizational structure is extremely dependent on the actual minister’s nominations (Siné 2008). The perimeter of each ministry is instable and dependent on what the cabinet looks like, but the ministries listed in Table 7.3 belong to the most stable ones, with a relatively consistent denomination. Obviously, coalition parties aim at getting a high number of important ministerial portfolios and other non-ministerial positions in government (ministerial chief of staff, councillors, etc.) but their influence on the final decision is limited. The way the Greens entered government in the 2012 Ayrault I cabinet offers a good example of this. After signing a pre-electoral agreement in November 2011, they knew that they would enter government if Hollande won the presidential elections, but no negotiations took place before the first round of the presidential election, that is less than two weeks before the government was proclaimed. Even so, the negotiations were very informal and the first-rank Green party leaders ignored the portfolios their party would obtain at the end of the process. They were confident that the party would most probably get two members, one full minister and one secretary of state, and tried with some success to influence who these cabinet members would be and which portfolio they would receive. As a prominent figure of the Green party told us during an interview, there was a consensus amongst Green leaders to avoid the Ministry for the Environment. They thought obtaining another ministry would help the party gaining credibility, by enlarging their issue profile in the eyes of the voters. They managed to negotiate the allocation of the Ministry for Housing to Cécile Duflot. The other green cabinet member, Pascal Canfin, received the development portfolio under the supervision of a socialist minister for foreign affairs. Prime Minister J.-M. Ayrault informed the two future ministers and the party leadership about the content of their portfolio just hours before the government was announced. Portfolio allocation is thus not significantly influenced by junior coalition members. Sometimes, when they get a full ministership position, the prime minister’s or president’s party also gets a junior position, a state secretary, or a delegated minister, in the same ministry, hereby serving as a ‘watchdog’.²⁰ Conversely, junior coalition parties that obtain secretaries of state to a ministry controlled by the main coalition party are not in the position to act as ‘watchdogs’ and have no way of monitoring the action of their senior minister. They can rather be considered as a ‘minister’s pet’ since they have to conform to the president’s, the prime minister’s, or the senior minister’s wishes (Thies 2001). Other ²⁰ In Table 7.5, we do not record the actual practice of ‘watchdog’ junior ministers; instead, we code ‘no’. This is because such junior ministers are not included in formal coalition agreements. This is different from the coding for other countries, where the actual practice is recorded.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



227

considerations rule over the allocation of junior ministries. Some ministries are often attributed to ministers with no previous political experience—and hence representing no coalition party—but a strong professional competence. The Ministry for the Economy—with former CEOs—and the Ministry for Culture— with former authors, a former director, and a former publisher—are good examples of this. Macron and Philippe tried to go further in that direction: in the Philippe cabinet, 10 ministers emanate from the civil society and were selected based on their competence and field experience. This differs from the usual dynamics: as Huber and Martinez Gallardo (2004) have shown, France can be characterized as a country with low levels of ministers’ experience.

Coalition agreements An important part of the pre-electoral agreement rests on the number of mutual withdrawals allowing parties wishing to form a coalition to be represented in parliament. Our interviews pointed out that negotiators deal more with office— than with policy—considerations. Yet, the electoral agreement is rarely made public. Documents presented in Table 7.4 hence mostly limit themselves to policy agreements. This table shows how diverse those agreements can be. Most of them are very short while others are longer—without though reaching the length and the level of detail of coalition agreements in pure parliamentary systems. There is no straightforward explanation for these differences. Right-wing parties tend, but this is not always the case, to publish shorter documents. Moreover, 1997 agreements were very short, which is due to the fact that early legislative elections were not anticipated, and negotiations for the next elections had hardly started when Chirac dissolved the Assembly. Parties’ strategies regarding the content of those documents also differ greatly. The Greens have notably tried to gain concessions from the PS in terms of environmental policies. They negotiated the abandonment of the Superphenix nuclear reactor and of the digging of a new canal between the Rhin and Rhône rivers in 1997. The same goes with the decrease of nuclear energy production in 2012. However, they did not always achieve their goals: in 2011, the PS refused to abandon the contested airport project at Notre-Dame-des-Landes and the building of a new nuclear reactor at Flamanville, and the published coalition agreement highlighted those disagreements (Deront et al. 2018). Communists’ and Greens’ pressure also led to the decrease in weekly working hours by the Jospin government, as well as the extension of public medical insurance for immigrant and poorest citizens (couverture maladie universelle). The 1997 coalition agreement included those pledges. However, most policy agreements are laconic. The Left Radicals and the liberals (UDF) have never granted much significance to it; it is hard to recall major policy

1984 1986 1993 1995 1997

2012 2012

Ayrault I Ayrault II

c

b

a

1973

Messmer II and III Barre III Mauroy I Mauroy II and III Fabius Chirac II Balladur Juppé I and II Jospin

870 2,060 1,780 1,780 1,907a 2,069b 1,631c 11,332 11,332

870 870

Size

5.1 0 0 0 0 3,6 97,5 0 0

5.1 5.1

General rules (in %)

Coalition agreement between PS and the Greens Coalition agreement between PS and PC Coalition agreement between PS and MRG

1978 1981 1981

Year in

Coalition

0 0 0 0 0 0 0 0 0

0 0

Policy-specific procedural rules (in %)

Table 7.4 Size and content of coalition agreements in France, 1959–2018

0 0 0 0 0 0 0 0 0

0 0

Distribution of offices (in %)

0 0 0 0 0 0 0 0 0

0 0

Distribution of competences (in %)

94.9 100 100 100 100 96.4 2.5 100 100

94.9 94.9

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



229

pledges for which they really fought. Our unsuccessful attempt at collecting all coalition agreements gave us evidence of this lack of interest.²¹ In any case, those coalition agreements are not as central in the French context as in others.²² They are written well before the election and are not binding. For instance, the Jospin government stuck to its key commitments towards the Greens on environmental matters, but it did not respect the negotiated line on immigration, security, and hunting; neither did the participation of the PC in government prevent the government to privatize public companies in an unprecedented rhythm in French history (Cos 2018). The same goes for the closing of the symbolic Fessenheim nuclear plant, which socialists agreed upon in 2011 but was never achieved in the following five years. All party officials encountered for interviews pointed towards the insignificance of coalition agreements. With respect to the coalition between UMP and Nouveau Centre (2007–2012), Sarkozy’s cabinet director for instance explained to us: ‘I don’t remember how the “agreement” was formalized [ . . . ]. We agreed on amendments to the UMP’s initial legislative project. I then had to draft a note to N. Sarkozy to keep him posted. [ . . . ] We then translated this into public documents including online communication and leaflets. It is also possible that a synthesis meeting took place between N. Sarkozy, F. Fillon, their advisers and F. Sauvadet and Hervé Morin [leaders of the NC] to “record” the amendments and I suppose that in this meeting, we talked about everything except this government agreement about which nobody cared.’ Finally, the manifesto of the presidential winner, which is often written after the coalition agreement, gets way more attention in the campaign and serves as the real benchmark for pledge fulfilment (Guinaudeau and Persico 2018). In contrast, coalition agreements are hardly noticed and do not provide any significant resources for junior coalition members during the term.

Coalition governance French coalitions have no formal and hardly any informal coalition governance arrangements. The power of coalition parties once in government is highly asymmetrical. Junior coalition partners and their ministers have scarce means to encourage the implementation of the coalition agreement. The chief of the executive and the main party of the coalition, on the other hand, enjoy extensive leverage in policy-making and usually manage to get their way. Coalition ²¹ Three missing agreements still have to be located: (1) UDR–RI–CDP for the 1973 election, (2) RPR–UDF for the 1978 election, and (3) the agreement for the 1981 election (between the PS and the PC and/or between the PS and the Left Radicals. As regards the 2007 coalition, the former president’s chief of staff confirmed that the coalition agreement, if it has ever existed, was never made public. ²² See Bué 2009 for congruent observations with respect to local-level policy agreements.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

230

  

coordination is hence very informal and rests on the central role of the chief of the executive and of individual ministers’ or other key figures’ ability to defend their party’s interests. To do so, they may either use public communication to highlight disagreements and defend their positions in the public arena or exert internal pressure through inter-ministerial coordination meetings (cabinet committees) or informal discussions with the members of the executive (Weil 2004). Overall, most of the interviewees we met to write this chapter—former ministers, chiefs of staff, and party leaders—agree that coalition agreements do not constrain governmental action and that the discretion of individual ministers override coalition-related consideration.

The role of individual ministers in policy-making As for most other aspects of coalition politics, the chief of the executive (president or prime minister in times of cohabitation) enjoys an extensive power in policymaking, making France a clear case of dominant Prime Minister model (Dunleavy and Rhodes 1990; Laver and Shepsle 1996; Bergman et al. forthcoming). He does not have to (and rarely does) consult other parties over what should be done: exceptions can be found after great focusing events (e.g. the November 2015 terrorist attack, when President Hollande consulted all parties, thereby sending a signal of national unity). Apart from these rare moments, most of the difficulties, for the president, are about accommodating the personal sensibility of each individual minister. We speak of ‘personal sensibility’ since many ministers have a distant relationship with their parties. They are free to act independently rather than bound by their partisan affiliation. This explains why parties’ rank-and-files are often quite critical against their party’s governmental participation and against their own party’s ministers. As we will see later, however, being strongly supported within its own party is a resource a minister can use in case of disagreement with the chief of the executive’s decision. Policy divergences sometimes emerge between coalition partners, mainly with ministers from parties with strong programmatic claims such as the Communists and Greens. Both these parties have had a tradition of criticizing some of the decisions taken by a government in which they participate (Voynet 2003; Fiterman 2005; Duflot 2014). In contrast, other ministers from the left wing tend not make strong policy claims and instead comply with the president’s leadership. Centre-right ministers also tend to be submissive to the chief of the executive. Beside the action of ministers from other coalition parties, the chief of the executive controls that of the presidential party’s members. For example, many of the ministers challenging Hollande and his successive prime ministers after 2012 belonged to the PS (e.g. the industry minister Montebourg or education minister

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



231

Hamon). Those ministers represented a similar challenge to Hollande’s decisions as the Green housing minister Duflot, who was particularly vocal. The history of the Fifth Republic is filled with names of prominent ministers able to mark their differences with the government’s policy choices: Jean-LouisBorloo in the Fillon cabinet, Nicolas Sarkozy in the Raffarin cabinet, and Laurent Fabius in the Jospin cabinet are all good examples of such ministerial counter-powers. The role of individual ministers in policy-making hence depends on whether they agree, or not, with the chief of the executive. If they agree, they can have some leeway to implement consensual policy choices. If they do not, their ability to influence policy-making depends on their individual leverage—which, in turn, reflects their previous political weight (faction or junior party leaders, former ministers, etc.), their public popularity, the importance of their ministry (e.g. economy, interior), itself correlated to their protocolary (status) order. In case of strong disagreement, interviewees consistently cited the request to meet the president or the prime minister as the primary way to seek policy concessions. But all also agreed that this instrument can only be used with moderation. In this respect, a former Green minister explained that ministers have to define a little number of ‘red-lines’ on which to feel legitimate to ask for a meeting with the chief of the executive. He also emphasized that going public was a risky strategy open only to politicians with considerable political weight. This might explain why Macron and Philippe decided to form a government with secondary political figures, mostly unknown to the public and/or deprived of any strong partisan support. Without political weight, going public over a disagreement with the chief of the executive is politically risky. Delphine Batho, a former environment minister in the Ayrault cabinet, was for example fired immediately after she openly criticized budgetary decisions in 2013. Many ministers think they should never communicate their disagreements. Remaining might mean being less influential in policymaking, but cases of ministers winning their argument against the president or the prime minister are rare anyway. Jean-Pierre Chevènement, a former leader of the main left-wing faction of the PS and the founder of a splitter party, who resigned from both the Mauroy and Jospin cabinets, famously declared: ‘As a minister, either you shut up or you resign.’

Coalition governance in the executive arena No law has ever been passed against the chief of the executive’s will—or that of the president when there is no cohabitation. The president’s power of assembly dissolution makes him a key actor—through the (implicit) threat of dissolution—in coalition management (Elgie 1999). The situation has never occurred, but one can imagine that a president facing the adoption of a law he

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

232

  

does not want would consider dissolving parliament and calling new elections. This may be regarded as a form of informal veto power. Similarly, there is no de jure right for the president to initiate legislation but, de facto, the president is (in particular outside cohabitations) the initiator of many laws passed. He enjoys a large presidential cabinet with highly qualified councillors allocated to most policy areas and is therefore able to follow the legislative and governmental processes very closely. The prime minister is clearly not in full control of the governmental agenda. Even regarding the most formal institution, the Council of Ministers, his proposition has to be agreed on by the president (the prime minister proposes, the president decides). The president always presides the Council of Ministers in the Fifth Republic (i.e. also in times of cohabitation). He may use these meetings to get information about government policies and even to exert forms of veto. On some topics, the president may ask each minister to express his or her opinion. However, all interviews confirm that decisions are made not in the Council of Ministers but in previous informal negotiations among ministers under the leadership of the prime minister and/or the president. The Council of Ministers is a formal enumeration of decisions already made; press releases are ready before the actual meeting. Several other structures allow the prime minister to monitor governmental action: the government’s General secretary; the General Secretary for European Affairs; the General Commission for Prospective and Planning and the Government Information Service, which not only provides expertise and analyses but also organizes the public communication on government policies.²³ These structures are mainly composed of high-level civil servants but also possibly of political appointees and provide valuable information to the prime minister, who has the final say in inter-ministerial disputes. Table 7.5 shows that coalition members can rest upon few conflict management mechanisms in the executive arena. Cabinet committees (inter-ministerial meetings) are the most commonly used mechanism of the sort, even though they mostly try to find a compromise between two ministries over a policy dispute, regardless of the party affiliation of the ministers. Some cabinets have tried to use other ad hoc conflict management mechanisms, which, however informal and temporary they may be, can be compared to mechanisms in use in other democracies. For instance, under Edouard Balladur’s and Alain Juppé’s prime ministership, parties in government decided to meet on a regular basis in coalition committees. During the Jospin government, most prominent figures of the majority were invited to weekly ‘majority breakfasts’, where current affairs were

²³ An interview with a former director of this service revealed that it also works closely with the presidential cabinet.

Debré Pompidou I Pompidou II Pompidou III Pompidou IV Couve de Murville Chaban Delmas Messmer I Messmer II and III Chirac I Barre I Barre II Barre III Mauroy I Mauroy II and III Fabius Chirac II Rocard I Rocard II

Coalition

N N N N N N

N

N PRE

N PRE PRE PRE PRE, POST PRE, POST

PRE PRE N N

1959 1962 1962 1966 1967 1968

1969

1972 1973

1974 1976 1978 1978 1981 1981

1984 1986 1988 1988

Y Y N N

n/a Y Y Y Y Y

n/a Y

n/a

n/a n/a n/a n/a n/a n/a

N N N N

N N N N N N

N N

N

N N N N N N

CaC, PS, O CaC, CoC CaC, O CaC, O

CaC, O CaC, O CaC, O CaC, O CaC, PS, O CaC, PS, O

CaC, O CaC, O

CaC, O

CaC, O CaC, O CaC, O CaC, O CaC, O CaC, O

All used

CaC CaC CaC CaC

CaC CaC CaC CaC CaC CaC

CaC CaC

CaC

CaC CaC CaC CaC CaC CaC

O CoC O O

O O O O O O

O O

O

O O O O O O

Most For most common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Table 7.5 Coalition governance mechanisms in French coalitions, 1959–2018

N N N N

N N N N N N

N N

N

N N N N N N

N N N N

N N N N N N

N N

N

N N N N N N

Personal Issues union excluded from agenda

Spec./No Spec./No Spec./No Spec./No

Spec./No Spec./No Spec./No Spec./No Spec./No Spec./No

Spec./No Spec./No

Spec./No

Spec./No Spec./No Spec./No Spec./No Spec./No Spec./No

N N N N

N N N N N N

N N

N

N N N N N N

Varied Varied No No

No No No No Varied Varied

No No

No

No No No No No No

Coalition Freedom of Policy discipline in appointment agreement legislation/ other parl. behaviour

N N N N

N N N N N N

N N

N

N N N N N N

Continued

N N N N

N N N N N N

N N

N

N N N N N N

Junior Nonministers cabinet (watchdogs) positions in contracts

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2007 2007 2012 2012 2014 2014 2016 2016 2017 2017

Fillon I Fillon II Ayrault I Ayrault II Valls I Valls II Valls III Cazeneuve Philippe I Philippe II

N N Y Y N N N N N N

N N N N N N N N N N

N

N

N N N

CaC, O CaC, O CaC, CoC, 0 CaC, CoC, O CaC, CoC, O Cac, O Cac, O Cac, O Cac, O Cac, O Cac, O Cac, O Cac, O Cac, O Cac, O CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC

CaC

CaC

CaC CaC CaC

O O O O O O O O O O

PS

O

O O CoC

Most For most common serious conflicts

N N N N N N N N N N

N

N

N N N

N N N N N N N N N N

N

N

N N N

Personal Issues union excluded from agenda

Spec./No Spec./No Spec./No Spec./No Spec./No Spec./No Spec./No Spec./No Spec./No Spec./No

Spec./No

Spec./No

Spec./No Spec./No Spec./No

N N N N N N N N N N

N

N

N N N

Few Few Varied Varied Few Few Few Few No No

Varied

Varied

No No Varied

Coalition Freedom of Policy discipline in appointment agreement legislation/ other parl. behaviour

N N N N N N N N N N

N

N

N N N

N N N N N N N N N N

N

N

N N N

Junior Nonministers cabinet (watchdogs) positions in contracts

Notes: Coalition agreement: PRE = pre-election; POST = post-election; PRE, POST = Both pre- and post-election Conflict management mechanisms: CaC = Cabinet committee; PS = Party summit; O = Other (informal exchanges between minister or party head and prime minister) Coalition discipline: Spec. = Discipline only expected on a few explicitly specified matters (annual budgetary votes) Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement

POST POST PRE PRE PRE PRE PRE PRE PRE PRE

PRE, POST Y

1997

Jospin

Y

PRE

Juppé I and II 1995

N N Y

N N PRE

All used

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

1991 1992 1993

Cresson Beregovoy Balladur

Coalition

Table 7.5 Continued OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



235

discussed collectively. Yet, such initiatives rarely lasted the whole term, and they never profoundly influenced the actual decision-making process. Our interviews suggest that some conflicts within government are not settled at all. At the end of the day, the president—and his cabinet advisors—have the final say and the practice of asking for a meeting with one of them in case of strong policy disagreement prevents direct confrontation between ministers. A former minister of the interior explained in our interview that he strongly opposed the minister of planning and the environment on multiple issues. In particular, she asked for the closure of an incineration plant in his home constituency, which she judged not sufficiently safe. The préfet, representative of the state in this region, placed under the responsibility of the minister of the interior, did not comply with her demands. He ironically observed that he had to send the police to oversee public demonstrations backed by his cabinet colleague. Ministers mostly have to accept government policies more or less divergent from their own party line. This is a frequent reason for resigning—which both ministers involved in this episode respectively did.

Governance mechanisms in the parliamentary arena Coalition governance mechanisms in the parliamentary arena are also very informal. Most formal instruments relate to governmental stability and the defence of the chief of the executive’s policy positions. Junior coalition members can only hope to have their voice heard either during informal talks, within parliamentary committees, in the National Assembly, or in the press. The first encounter between a newly formed government and the parliament provides a good example of this asymmetric relationship. There is no investiture vote in the French Fifth Republic. However, prime ministers have the possibility to seek a vote of confidence following their discours de politique générale (general policy discourse) to the National Assembly. This is the first instrument through which the prime minister can coordinate his coalition in the parliamentary arena, and junior coalition members have no say regarding the content of that general policy speech. The prime minister can voluntarily ask for a confidence vote after having presented his or her general policy discourse—which is then equivalent to an investiture vote. This is not compulsory and only governments with a clear majority have done so. Parliamentary groups hardly influence coalition governance. Although group members select their leaders through a formal election, these leaders tend to support the interest of the chief of the executive and not necessarily those of their parties. As regards the election of the head of the main coalition party group, it is often just a formal validation of the president’s choice: all candidates supported by the president or the prime minister have won the nomination in the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

236

  

end. As regards junior coalition parties’ groups, there is also a tendency to select MPs that are rather favourable to the government (e.g. François De Rugy for the Greens in 2012 or Alain Bocquet for the Communists in 1997). Some group leaders have managed to take some autonomy towards the president or the prime minister. For example, Jean-François Copé, the UMP group leader under Sarkozy’s presidency, was a prominent figure of the Gaullist party before taking office and made sure that his and his supporters’ positions were defended. In France, coalition discipline mostly relates to party discipline and it is not mentioned in any coalition agreement. All interviewees made it clear that there is an implicit expectation (strong convention) regarding party discipline on important—if not all—policies. Parties in parliament can keep their MPs in line through diverse means: they can offer prestigious positions either within the National Assembly (e.g. the leadership of a parliamentary committee) or within the party (e.g. the charge of a policy domain), or guarantee the presence of potential defectors’ close acquaintances in future second-order elections. They may also not renominate potential defectors for the next parliamentary elections. A further relevant institutional feature is the existence of strong restrictive legislative procedures in the constitution of the Fifth Republic. In response to the trauma of cabinet instability under the Fourth Republic (Huber and MartinezGallardo 2004), linked to the lack of a stable parliamentary majority, the Fifth Republic was designed in order to rebalance power between executive and legislative powers. The imperative was to enhance the government’s policy-making capacity even in the absence of a reliable parliamentary majority (Huber 1996). In particular, the legislative procedures defined in Articles 44.3 and 49.3 of the constitution allow the government to pass legislation despite a policy conflict with coalition partners or backbenchers within their own party.²⁴ This is a factor of government duration in France—it notably explains that PS-led minority governments could effectively govern over five years (between 1988 and 1993). Those restrictive legislative procedures can offer unexpected instruments for junior coalition partners, in order to defend their positions. Huber (1996) shows that using Article 49.3 may also be a way to manage internal disagreement and allow coalition partners to show their disagreement without having to leave the majority. The 2008 constitutional revision has considerably restricted the usage of Article 49.3, which may now be used only once per parliamentary session (the only exception being budget laws, for which there is no limitation). This is likely to reinforce coalition partners’ (or factions’) bargaining power in case of policy conflict and/or to reduce government stability. ²⁴ Article 49.3 institutionalizes a confidence vote procedure. If invoked by the prime minister in the legislative process, the government’s bill is adopted unless a majority of legislators vote to censure and bring down the government. Article 44.3 defines the so-called package procedure (vote bloqué), in which the Assembly has to deliberate on a bill under exclusive consideration of the amendments proposed or accepted by the government (Huber 1996).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



237

Other governance mechanisms In contrast to pure parliamentary systems, there are overall very few (formal and informal) coalition governance mechanisms and no instance coordinating government coalitions across the partisan, the executive, and the legislative arenas. As already mentioned, coalition agreements focus on general policy orientations and do not state any conflict management mechanism. This is probably a reflection of the de facto predominance of the main government party, junior parties having virtually no other leverage, once elections are passed, than the threat to leave the coalition. In the absence of any reform of the majoritarian electoral system, there is no tendency towards a tighter monitoring of coalition agreements’ enforcement. Overall, political parties appear to be weak actors in government policy-making, as in other dominant Prime Minister models. Junior parties (but also minority factions within mainstream parties) are more vocal and try to influence the government’s policy line on the issues that matter most to them. Rank-and-files are often more critical about governmental participation and less prone to policy concessions, but this rarely had any impact. As already exemplified, small parties face a dilemma between participating in government and having to accept numerous policy compromises that are not negotiated in advance and remaining in the opposition in which the electoral system and the constitution provide them with only little resources.

Cabinet duration and termination The duration of cabinets After the spectacular instability of the Fourth Republic, the constitution of the Fifth Republic was meant to create the conditions for stable and effective governments, with a majoritarian electoral system and comparatively strong restrictive legislative procedures. These objectives were achieved, since governments appear as more stable, with an average duration of more than 20 months. This average is not fully representative of reality because it is biased by the practice that newly elected presidents immediately appoint a new (usually very similar) government that ends a few weeks later, after legislative elections have taken place (as discussed earlier): leaving these caretaker governments out, cabinets serve more than 700 days (i.e. 23 months) on average.

The termination of cabinets The main cause for cabinet termination, as illustrated in Table 7.6, is the occurrence of regular legislative or presidential elections. In one instance (in 1997),

Date in

1988-06-28

1991-05-17

1992-04-04 1993-03-30 1995-05-18

1997-06-04 2002-05-07 2002-06-17

2005-06-02 2007-05-18 2007-06-19 2012-05-16

Cabinet

Rocard II

Cresson

Beregovoy Balladur Juppé I and II

Jospin Raffarin I Raffarin II and III

de Villepin Fillon I Fillon II Ayrault I

2007-05-15 2007-06-18 2012-05-10 2012-06-18

2002-05-06 2002-06-17 2005-05-31

1993-03-29 1995-05-11 1997-06-02

1992-04-02

1991-05-15

Date out

99.7 91.2 100 100

99.9 97.6 60.5

99.4 99.9 71.3

47.1

60.6

Relative duration (%)

Table 7.6 Cabinet termination in France, 1988–2018

2 1 2 1

2 1 9

1 2 4

8, 9

8

Mechanisms of cabinet termination

10

Presidential election

Presidential election Legislative elections following dissolution by President Chirac Presidential election

Internal majority conflict (opposition between Prime Minister Rocard and President Mitterrand) Poor showing at regional elections; high unemployment; very unpopular prime minister + internal majority conflict

Comments

10

Economy

Policy area(s)

After losing the referendum on the EU constitutional treaty, Raffarin resigned. Presidential election

PS

PS

Parties (when conflict between or within)

10

10

10

10, 11, 13

14

Terminal events

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2014-04-09

2014-08-26

2016-02-11

2016-12-06 2017-05-17

Valls I

Valls II

Valls III

Cazeneuve Philippe I

2017-05-10 2017-06-19

2016-12-06

2016-02-11

2014-08-25

2014-03-31

100 100

66.4

54.3

12.

36.4

2 1

9

5

8

9

10

14

10, 11

PS

Economy

Note: Relative duration: share of constitutionally allowed duration for this cabinet.

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event

2012-06-21

Ayrault II

After PS suffered considerable losses in the 2014 municipal elections, President Hollande replaced Ayrault with Valls. Faced with growing internal contestation (after Minister Montebourg openly criticized his political line), Valls resigns. Several Green leaders decide to join government. They leave their party and found the ‘ecologist party’, resulting in a new party composition of government. Prime Minister Valls decides to be candidate to the 2017 presidential election and resigns. Presidential election

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

240

  

legislative elections followed the dissolution of the National Assembly by President Chirac. A second important cause is government reshuffling, used by the president to increase his popularity. See section on the composition and size of cabinets. Examples of this strategic reshuffling are the termination of the Mauroy government following massive mobilization against a bill on education policy in 1984, the replacement of prime minister Edith Cresson with Pierre Beregovoy in 1992, Jean-Pierre Raffarin’s resignation after losing the 2005 referendum on the European constitutional treaty, and the end of the Ayrault government in 2014, after the PS suffered considerable losses at the 2014 municipal elections. Overall, ministerial tenure usually ends due to collective reasons (80 per cent) rather than to ministers’ individual motivations or to conflict within coalitions. In keeping with this, French history has witnessed junior coalition partners leaving government against the background of strong policy disagreement. Note that even when this is not the apparent cause for government termination, reshuffling might open a window of opportunity for dissatisfied parties to leave the coalition. This occurred for example in 1984, when the PC decided not to join the Fabius government formed by the PS after the termination of the Mauroy cabinet. They officially justified this decision with respect to the austerity policies implemented, but there is evidence that electoral sanctions at intermediary elections were also an incentive. Overall, our observations suggest that policy conflict within coalitions becomes a more frequent cause of government crisis and (sometimes) terminations over time. Our interviews suggest that this is due to junior parties’ past frustrations with respect to their limited policy achievements when in government. Many of their representatives are now reluctant to accept compromises.²⁵

Conclusions As we have shown, French coalition politics are distinctive in Western Europe. There is a strongly imbalanced power structure in favour of the party of the president (or prime minister in times of cohabitation). France is therefore a particular and rather extreme case of dominant Prime Minister model. The constitution of the Fifth Republic creates strong incentives to build pre-electoral alliances, but the majoritarian electoral system and semi-presidentialism have led to an extreme predominance by the main parties, and in particular of the chief of the executive, over their junior partners. Institutional reforms have further reinforced the weight of the president. Small parties can exert leverage mainly before

²⁵ This idea that government coalitions are influenced by past interactions between parties is developed in Franklin and Mackie (1983).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



241

the elections, given their capacity to negotiate their support during the campaign and mutual withdrawals for the first or the second round of elections. Once the election is passed, coalition politics are strongly structured by presidential logics (outside cohabitations) and take place to a large extent beyond parties: individual MPs from the same party may join different parliamentary groups depending on their attitude towards the president. Given the absence of formal coalition governance arrangements, junior parties’ resources for influencing government policies are very much restricted to public communication, informal exchanges with the chief of the executive, and the threat to leave the coalition. Modalities of coalition governance provide the chief of the executive (i.e. the president since 2002) with strong leadership, extensive leverage in policymaking, and a final say in inter-ministerial disputes. He usually manages to get his way and only rarely consults parties in his coalition. Consequently, these parties will affect policy outcomes only with respect to very emblematic issues (usually one or two) and to the definition of ‘red lines’ likely to be a motivation for terminating their participation to government. Political parties are aware of this and do not expect coalition agreements to be implemented—they conceive them at best as a document allowing them to present their agreement as a political project and not as a mere electoral alliance. This situation nourishes considerable frustration within most junior coalition parties. This context explains recurrent attempts to build large movements gathered around the presidential candidate/president within each bloc. However, this has not succeeded in terms of durability and the party system is increasingly fragmented. Internal divisions and the emergence of a large far-right bloc puts coalitions under considerable pressure. The rise of the National Front, which has so far been excluded from the coalitionable parties, reduces the space for building coalitions, while the PS and the Republicans are respectively facing growing difficulties in building alliances and in managing internal dissent. This poses all the more a challenge, as the key tool for ensuring government survival in face of policy conflict, the particular French version of the confidence vote procedure, in many cases is not a viable option anymore.

References Adler, Laure (2015). François Mitterrand, Journée particulière. Paris: Flammarion. Andolfatto, Dominique, and Fabienne Greffet (2008). ‘La “semi-cartellisation” du parti communiste français’. In Yohann Aucante and Alexandre Dézé (eds), Les systèmes de partis dans les démocraties occidentales. Le modèle du parti cartel en question. Paris: Presses de Sciences Po, 321–46.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

242

  

Armingeon, Klaus, Virginia Wenger, Fiona Wiedemeier, Christian Isler, Laura Knöpfel, David Weisstanner, and Sarah Engler (2019). Comparative Political Data Set 1960–2017. Zurich: Institute of Political Science, University of Zurich. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010.’ Party Politics, 21(1): 143–52. Bédéï, Jean-Pierre (2015). Sur proposition du Premier Ministre . . . Dans les coulisses du remaniement. Paris: Editions de l’Archipel. Bergman, Torbjörn, Gabriella Ilonszki, and Wolfgang C. Müller (eds) (2019). Coalition Governance in Central Eastern Europe. Oxford: Oxford University Press. Bornschier, Simon, and Romain Lachat (2009). ‘The Evolution of the French Political Space and Party System’. West European Politics, 32(2): 360–83. Boy, Daniel (2002). ‘France’. Environmental Politics, 11(1): 63–77. Bué, Nicolas (2009). ‘Les accords de coalition dans une municipalité d’union de la gauche. Contribution à l’étude de la régulation des rapports coalitionnels’. Politix, 4 (88): 105–31. Bué, Nicolas, and Fabien Desage (2009). ‘Le “monde reel” des coalitions. L’étude des alliances partisanes de gouvernement à la croisée des méthodes’. Politix, 4(88): 7–37. Chevallier, Jean-Jacques, Guy Carcassone, and Olivier Duhamel (2004). La Ve République 1958–2004: histoire des institutions et des régimes politiques de la France. Paris: Armand Colin. Conley, Richard S. (2007). ‘Presidential Republics and Divided Government: Lawmaking and Executive Politics in the United States and France’. Political Science Quarterly, 122(2): 257–85. Cos, Rafael (2018). ‘Dénoncer le programme. Les logiques du désengagement électoral au révélateur des privatisations du gouvernement Jospin’. Revue française de science politique, 68(2): 239–63. Deront, Eva, Aurélien Evrard, and Simon Persico (2018). ‘Tenir une promesse électorale sans la mettre en œuvre. Le cas de la fermeture de Fessenheim’. Revue française de science politique, 68(2): 265–89. Di Virgilio, Aldo, Bernard Dolez, and Annie Laurent (2015). ‘Systèmes électoraux « complexes », coordination pré-électorale complexe. Une comparaison FranceItalie’. Revue internationale de politique comparée, 22(3): 343–370. Duflot, Cécile (2014). De l’intérieur. Voyage au pays de la désillusion. Paris: Fayard. Dunleavy, Patrick, and R. A. W. Rhodes (1990). ‘Core Executive Studies in Britain’. Public Administration, 68(1): 3–28. Duverger, Maurice (1954). Les partis politiques. Paris: Armand Colin. Duverger, Maurice (1978). Échec au Roi. Paris: Albin Michel. Duverger, Maurice (1985). Le système politique français. Paris: Presses universitaires de France.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



243

Duverger, Maurice (2004). Les constitutions de la France, 15th edn. Paris: Presses universitaires de France. Elgie, Robert (1999). ‘France’. In Robert Elgie (ed.), Semi-presidentialism in Europe. Oxford: Oxford University Press, 67–85. Elgie, Robert (2018). ‘The Election of Emmanuel Macron and the New French Party System: A Return to the Éternel Marais?’. Modern & Contemporary France, 26(1): 15–29. Elgie, Robert, and Emiliano Grossman (2016). ‘The Study of Executive Politics’. In Robert Elgie, Emiliano Grossman, and Amy G. Mazur (eds), The Oxford Handbook of French Politics. Oxford Handbooks. Oxford and New York: Oxford University Press, 677–92. Elgie, Robert, Emiliano Grossman, and Amy G. Mazur (eds) (2016). The Oxford Handbook of French Politics. Oxford Handbooks. Oxford and New York: Oxford University Press. Evans, Jocelyn A. (2003). The French Party System. Vol. 1. Manchester: Manchester University Press. Fiterman, Charles (2005). Profession de foi. Pour l’honneur de la politique. Paris: Seuil. François, Abel, and Eric Phelippeau (2018). ‘Financing of the French Political Parties’. In Jonathan Mendilow and Eric Phelippeau (eds), Handbook of Political Party Funding. Cheltenham and Northampton: Edward Elgar, 271–90. François, Bastien (2011). Le régime politique de la Ve République, 5th edn. Paris: La Découverte. Franklin, Mark N., and Thomas T. Mackie (1983). ‘Familiarity and Inertia in the Formation of Governing Coalitions in Parliamentary Democracies’. British Journal of Political Science, 13(3): 275–98. Gallagher, Michael, and Paul Mitchell (2008). The Politics of Electoral Systems. Oxford: Oxford University Press. Golder, Sona N. (2005). ‘Pre-electoral Coalitions in Comparative Perspective: A Test of Existing Hypotheses’. Electoral Studies, 24: 643–63. Golder, Sona N. (2006). ‘Pre-electoral Coalition Formation in Parliamentary Democracies’. British Journal of Political Science, 36(2): 193–212. Gougou, Florent, and Simon Labouret (2013). ‘The End of Tripartition?’. Revue française de science politique, 63(2): 279–302. Gougou, Florent, and Simon Persico (2017). ‘A New Party System in the Making? The 2017 French Presidential Election’. French Politics, 15(3): 303–21. Grossman, Emiliano (2009). ‘The President’s Choice? Government and Cabinet Turnover under the Fifth Republic’. West European Politics, 32(2): 268–86. Grossman, Emiliano, and Abel François (2013). ‘Who Are the Ministers of the Fifth Republic?’. French Politics, 11(3): 272–83. Grossman, Emiliano, and Nicolas Sauger (2009). ‘The End of Ambiguity? Presidents versus Parties or the Four Phases of the Fifth Republic’. West European Politics, 32 (2): 423–37.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

244

  

Grunberg, Gerard, and Etienne Schweisguth (1997). ‘Vers une tripartition de l'espace politique’. In Daniel Boy and Nonna Mayer (eds), L’électeur a ses raisons. Paris: Presses de Sciences Po, 339–362. Guinaudeau, Isabelle, and Simon Persico (2018). ‘Tenir promesse. Les conditions de la réalisation des programmes électoraux’. Revue française de science politique, 68(2): 215–37. Haegel, Florence (2002). ‘Faire l’union: la refondation des partis de droite après les élections de 2002’. Revue Française de Science Politique, 52(5–6): 561–76. Haegel, Florence (2012). Les droites en fusion. Transformations de l'UMP. Paris: Presses de Sciences Po. Huber, John D. (1996). Rationalizing Parliament. Legislative Institutions and Party Politics in France. New York: Cambridge University Press. Huber, John D., and Cecilia Martinez-Gallardo (2004). ‘Cabinet Instability and the Accumulation of Experience: The French Fourth and Fifth Republics in Comparative Perspective’. British Journal of Political Science, 34(1): 27–48. Laver, Michael, and Kenneth A. Shepsle (1996). Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Martin, Pierre (2010). ‘L’immigration, un piège pour la droite?’. Commentaire, 132: 1027–36. Montay, Benoît (2013). ‘Le pouvoir de nomination de l’Executif sous la Ve République’. Jus Politicum, 11. http://juspoliticum.com/article/Le-pouvoir-denomination-de-l-Executif-sous-la-Ve-Republique-817.html Müller, Wolfgang C., and Kaare Strøm (eds.) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of Antielitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Safran, William (2008). The French Polity. 7 ed. New York: Routledge. Sauger, Nicolas (2008). ‘L’UDF et la création de l’UMP’. In Yohann Aucante and Alexandre Dézé (eds), Les systèmes de partis dans les démocraties occidentales. Le modèle du parti cartel en question. Paris: Presses de Sciences Po, 347–367. Siné, Alexandre (2008). ‘Les contours ministériels vus de la LOLF’. In: Julien Meimon (ed), Les réorganisations administratives: Bilan et perspectives en France et en Europe. Paris: Institut de la gestion publique et du développement économique, 41–50. Sponchiado, Lucie (2017). La compétence de nomination du président de la Cinquième République. Paris: Dalloz. Stimson, James A., Vincent Tiberj, and Cyrille Thiébaut (2010). ‘Le mood, un nouvel instrument. Au service de l’analyse dynamique des opinions. Application aux

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



245

évolutions de la xénophobie en France (1990–2009)’. Revue française de science politique, 60(5): 901. Strøm, Kaare, Wolfgang C. Muller, and Torbörn Bergman (eds.) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Thiébault, Jean-Louis (2003). ‘France: Forming and Maintaining Government Coalitions in the Fifth Republic’. In Wolfgang. C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe (Comparative Politics). Oxford and New York: Oxford University Press, 498–528. Thies, Michael F. (2001). ‘Keeping Tabs on Partners: The Logic of Delegation in Coalition Governments’. American Journal of Political Science, 45(3): 580–98. Tiberj, Vincent (2012). ‘Two-Axis Politics: Values, Votes and Sociological Cleavages in France (1988-2007)’. Revue française de science politique (English Edition), 62(1): 67–103. Voynet, Dominique (2003). Voix off. Paris: Stock. Weil, Thierry (2004). ‘La machine Matignon’. Le journal de l’école de Paris du management, 45: 28–36.

Appendix. List of political parties Abbreviation Name C Communist Party (Parti Communiste, PC) Left Front (Front de Gauche, FDG) L Left Party (Parti de Gauche, PDG) Unbowed France (La France Insoumise, FI) V Europe Ecology—The Greens (Europe Écologie Les Verts, EELV), 2010– Greens (Les Verts), 1984–2010 OG Ecology Generation (Génération Ecologie, GE) Ecologist Party (Parti écologiste, PE) S Socialist Party (Parti Socialiste, PS), 1969– Democrat Movement (Mouvement des democrats, MD), 1974–???? Unified Socialist Party (Parti Socialiste Unifié, PSU), 1960–1990 French Section of the Workers’ International (Section Française de l’Internationale Ouvrière, SFIO), before 1969 R Radical Party (Parti radical) Movement of Left Radicals (Mouvement Radical de Gauche, MRG) Party of the Left Radicals (Parti Radical de Gauche, PRG) Movement of Citizens (Mouvement des Citoyens, MDC) Movement of Republicans and Citizens (Mouvement Républicain et Citoyen, MRC) LREM The Republic in Motion (La République en Marche) CD&CR Popular Republican Movement (Mouvement Républicain Populaire, MRP) Democratic Center (Centre Démocrate, CD) Progress and Modern Democracy (Progès et démocratie modern, PDM)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

246

  

Social Democrat Reformers (Réformateurs democrats sociaux, RDS) Center Untion (Union de Centre, UC) Reformists, Censtrists and Social Democrats (Réformateurs, centristes, et démocrates sociaux, RCDS) Center Democracy and Progress (Centre Démocratie et Progrès, CDP) Movement of Reformers (Mouvement Réformateur, MR) Centrist Union (Union Centriste, UC) Union of Centrist Democrats (Union des Démocrates centrists, UDC) Union for French Democracy (Union pour la Démocratie Française, UDF) Democratic Movement (Mouvement Démocrate, MODEM), 2007 CL National Center of Independent Professions and Peasants (Centre National des Indépendents et Paysans, CNIP) National Federation of Republicans (Fédération Internationale des Républicains Indépendents, FNRI) Independent Republicans (Républicains indépendants, RI) Liberal Democracy (Démocratie libérale, DL) New Centre (Nouveau Centre, NC) Union of Independents (Union des Indépendants, UDI) G Union for the New Republic (Union pour la Nouvelle République, UNR) Democratic Union for the 5th Republic (Union Démocratique pour la Ve République, UDV) Union of Democrats for the Republic (Union des Démocrates pour la République, UDR) Rally for the Republic (Rassemblement pour la République, RPR) Union for a Popular Movement (Union pour un Mouvement Populaire, UMP) The Republicans (Les Républicains, LR) UR Unity of the Republic (Untés de la République, UR) FN National Front (Front National, FN), from June 2018 National Rally (Rassemblement National, RN) RDA African Democratic Rally (Rassemblement démocratique africain, RDA) Note: Due to the functioning of the French party system both larger parliamentary groups—sometimes consisting of multiple parties simultaneously—and their constituent parties are presented. Parties that have been coded under a specific group are listed in reverse chronological order. Party names are presented in English, with the original French names in parentheses along withs its abbreviation.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

Chapter 8 Germany: From Stable Coalition Camps to New Complexity Marc Debus, Holger Döring, and Alejandro Ecker

In 1998, the first Schröder government ushered in a new era of coalition politics in Germany. It was the first complete government turnover in post-war Germany and foreshadowed substantial party system changes in the years to come. Previously, government formation and coalition governance had been very stable, both in the ‘two-and-a-half ’ party system which existed from the end of the 1950s until the mid of the 1980s, and during the 1990s in which the Green Party became a relevant actor in the coalition game, in particular at the state level (e.g. Saalfeld 2000; Decker and Jesse 2013; Bräuninger et al. 2020: 21–7). However, these aspects of the political process have significantly changed since the beginning of the 2000s. One reason for these changes is a dealignment process in the traditional party preferences of voters and therefrom resulting changes in the party system from four parliamentary parties—Christian Democrats (CDU/CSU), Social Democrats (SPD), the liberal Free Democratic Party (FDP), and the Greens—to a parliament with up to six party groups, which includes the socialist Left (Linke) party (until 2005 the Party of Democratic Socialism, PDS) and—since 2017—the antiimmigrant and right-wing populist Alternative for Germany (AfD). In addition, changes in the policy profiles of the parties and how potential coalitions are discussed during election campaigns resulted in a more complex coalition formation process. This chapter aims at presenting the characteristics of cabinets in Germany, in particular for the cabinets formed since the beginning of the twenty-first century. We cover two decades of coalition dynamics and an era that has led to significant changes in German politics in general and the German party system in particular. Since the 2017 Bundestag election, ‘pariah parties’ (van Spanje and de Graaf 2018) are represented in the Bundestag both on the left and on the right of the ideological spectrum, which makes the formation of a majority coalition between centre-left parties (SPD and Greens) or centre-right parties (CDU/CSU and FDP) less likely. We start out by giving a brief overview on the institutional setting in which parties in Germany operate and which influences the government formation process as well as the daily business of coalition governance. In addition, we outline recent Marc Debus, Holger Döring, and Alejandro Ecker, Germany: From stable coalition camps to New Complexity In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI:10.1093/oso/9780198868484.003.0008

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

248

  .

dynamics in the structure of the German party system that changed significantly since the 1990s. The Section ‘Conclusions’ summarizes the findings, considers if an overall trend in terms of changes in coalition governance in Germany exists, and discusses the impact of the parliamentary presence of a left-wing and a rightwing ‘pariah’ party—the Left and the AfD—for coalition politics in Germany in the future.

The institutional setting The outcomes of government formation processes and the structure of coalition governance in modern democracies depend on the interplay of a large number of factors. These factors have their origins in party-specific incentives like the expected payoffs in terms of offices and policies and incentives that emerge from institutionalist factors like the power of the head of state, the existence and competencies of a second chamber, and a multilevel structure of a political system. Additional factors are the pre-electoral commitments of parties, intra-party conflict, and the coalition preferences of voters and party supporters (for an overview see Laver and Schofield 1990; Strøm et al. 2008; Müller 2009; Andeweg et al. 2011; Clark et al. 2013: 465–524). The general findings also apply to the government formation process in Germany on all levels of the political system. That is, political parties and their representatives on the national, regional, and local levels try to maximize their office and policy payoffs and consider incentives or restrictions that emerge from the respective institutional setting (e.g. Downs 1998; Pappi et al. 2005; Debus and Gross 2016). In terms of the institutional structure that shapes government formation in particular and coalition politics in general, the multilevel structure of the German political system and, in this context, the role of the second chamber, the Bundesrat, in the legislative process play an important role, as do the corresponding parliamentary rules. In addition, the constitutionally provided competencies of the chancellor matter. They matter not only for ministerial appointments; the chancellor’s preferences are also decisive for the overall cabinet decision-making process and the formulation of future policies. Specifically, the prerogatives mentioned in Article 65 of the German constitution allow the chancellor to set the general guidelines of government policy, which strengthens her position vis-à-vis other cabinet members (Richtlinienkompetenz, see Saalfeld 2000; Niclauß 2004; Bäck et al. 2016). Overall, these institutional prerogatives are an elementary part of the ‘Dominant Prime Ministerial Model’ of coalition governance in Germany and is one of the three modes of government decision-making (see Section ‘Coalition governance’). Whether minority (coalition) governments are a viable alternative during government formation depends to a decisive degree on the existence of an

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



249

investiture vote for the prime minister (‘positive parliamentarism’; see also Bergman 1993). In Germany, a politician can only become chancellor when winning the support of the majority of members of the Bundestag in the first or second round of voting or—if no candidate receives a majority—by a plurality of votes in a final voting round and appointment by the head of state. This results in a strong incentive to form majority coalitions in order to secure a successful instalment of the (new) chancellor and the takeover of the cabinet and ministerial posts (Ganghof and Stecker 2016). Normally, the largest parliamentary party starts negotiations with all other parties represented in the Bundestag after the elections to find out which coalition is feasible in terms of expected policy and office payoffs. Some parties are a priori excluded from the coalition formation process, due to their ‘pariah’ status (see section ‘Electoral alliances and preelectoral coalitions’ for more details). This first negotiation period normally takes only a few days, since pre-electoral commitments of parties in terms of their preferred and rejected coalitions are the standard in Germany (Saalfeld 2000; Pappi et al. 2006; Debus 2007, 2009; Decker 2009, 2013; Linhart 2009; Best 2015). In case the majority situation in the Bundestag is less clear and no candidate for the office of the chancellor is likely to win a majority in the Bundestag after the two weeks following the first—failed—voting round, the president comes into play. If a candidate for the chancellor position wins the election in the first round of voting or in the following rounds by receiving a majority of votes, the president has to appoint the respective candidate as chancellor. If, however, the candidate wins only a plurality of votes in the final parliamentary vote, then the president can either appoint the candidate or call a new election for the Bundestag. While the latter never happened in German post-war history, it was a widely debated scenario after the 2017 Bundestag election, which resulted in an unclear majority situation, in particular after the negotiations between CDU/CSU, FDP, and Greens failed (Bräuninger et al. 2019). In such a situation, the president has a more active role, because she or he is—in particular in this situation but also in situations of gridlock when an incumbent coalition seemingly has no majority in the parliament anymore—negotiating the possibilities of forming a government with the representatives of the parliamentary parties. The multilevel structure of the German political system and the role of the second chamber—the Bundesrat—in the legislative process also shapes government formation not only at the subnational—that is the Länder—level but also at the federal level. There is a high degree of ‘shared rule’ (Hooghe et al. 2016) in the institutional configuration of the German political system, and state governments can initiate bills via the Bundesrat or stop bills that passed the Bundestag. While the share of bills that needed the consent of the Bundesrat was at around 60 per cent before the reform of German federalism, it was reduced after the implementation of the reform in 2006 to around 40 per cent (Stecker 2016). Thus, federal parties still take the majority situation in the Bundesrat into

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

250

  .

account when forming a coalition government. Specifically, law proposals initiated by the government and its majority in the Bundestag but without sufficient support in the second chamber will harm the government’s legislative record, which in turn could lead to defeats in upcoming state elections. The mutual impact of the majority situation in the first and second chambers for government formation at the federal and state levels has, however, decreased since the 1990s. Earlier, the main opposition party was able to block legislation in the Bundesrat, a strategy used by the SPD leader Lafontaine in the 1990s and by the Christian Democrats in the early 2000s. However, since the 2000s, an increasingly diverse partisan composition of state governments has weakened the dominance of the two major parties in the Bundesrat. In fact, their junior coalition partners at the state level are often in opposition at the federal level. As a result, the chances for installing a coalition government in a state (at the Länder level) that is fully congruent with either the partisan composition of the federal government or the opposition in the Bundestag decreased clearly. A final key institutional characteristic in Germany that affects multiparty government throughout the coalition life cycle is the constructive vote of no confidence (Article 67 of the German constitution). Borne out of the adverse historical experience of repeated government breakdown during the Weimar Republic, it stipulates that the current government can be defeated only by an alternative (coalition) government that is explicitly supported by a parliamentary majority. In contrast to a regular vote of no confidence, the alternative parliamentary majority needs to be willing to jointly and actively support a new chancellor. Such an institutional provision naturally constrains the number of viable government alternatives—effectively ruling out minority governments (Bergman 1993; Strøm et al. 1994)—while generally enhancing government stability (Diermeier et al. 2002). As such, it is an important factor contributing to the prevalent patterns of multiparty government in Germany.

The party system and the actors Party system change Germany was dominated by a two-and-a-half party system in the 1970s and 1980s. The two major parties (Christian Democrats and Social Democrats) jointly won about 90 per cent of the popular vote and the smaller Liberals were pivotal for cabinet formation. This led to predictable patterns of government formation. The Greens entered parliament first in 1983, but were considered a feasible coalition partner only at the state level in the 1980s. Only after the Social Democrats considered the Green Party as a coalition partner also at the federal level, a competition between two ‘coalition camps’ emerged in the 1990s: Christian

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



251

Democrats and Liberals on the one side and Social Democrats and Greens on the other side signalled directly or indirectly their common interest to form a coalition government and these two ‘camps’ or ‘blocs’ provided distinct platforms to the German electorate. However, these pre-electoral commitments were less explicit than the criteria used to record pre-electoral coalitions in comparative work (Golder 2006: 195; Chiru 2015: 11). The two-bloc system led to stable majorities for one of the two blocs until 2005: CDU/CSU and FDP won parliamentary majorities in 1990 and 1994, whereas SPD and Greens managed to win a majority of seats in the Bundestag in 1998 and 2002 despite the parliamentary presence of the PDS (see e.g. Decker and Jesse 2013: 13–17). The resulting pattern of cabinet formation over time is presented in Table 8.1 A significant change took place in the German party system during the 2000s, which affected coalition building dynamics (see also Decker and Jesse 2013: 17–22). A new party—The Left (Die Linke)—was formed by the post-communists from Eastern Germany, former members of the Social Democrats and splinter elements of the radical left. The party was established on the organizational basis of the post-communist PDS from East Germany (see the contributions in Spier et al. 2007). However, former Social Democrats from Western Germany who opposed the social and welfare policy reforms implemented by the second ‘red– green’ Schröder cabinet in the period between 2003 and 2005 were the crucial element that helped to consolidate the new party. This new party altered the twobloc system of electoral competition and has since then been a challenger in particular for the Social Democrats (see also Niedermayer 2007; Holtmann 2009). Because of the increasing strength of the new party on the left of the political spectrum, the dynamics of coalition bargaining in Germany changed significantly with the early election held in September 2005. The electoral campaign was fought within the old two-bloc dynamics. However, for the first time none of the two blocs was able to win a majority of seats in the Bundestag. Proksch and Slapin (2006) provide an analysis of this particular instance of cabinet formation and demonstrate on the basis of an election manifesto analysis that an SPD-led minority government was feasible from a policy perspective. However, the main actors were not willing to experiment with new cabinet types and a grand coalition between the Christian Democrats and Social Democrats was formed in 2005. At the institutional level, the Bundesrat, Germany’s second chamber, was seen as one major reason to form a grand coalition in order to avoid legislative gridlock (Kropp 2010). The structure of the party system that emerged in the 2005 election remained in place for a decade. The left was now split into three parties and the Social Democrats continuously lost electoral support. While they had won 38.5 per cent of the vote in 2002, seven years later and after serving as the junior partner in a grand coalition the party won the support of only 23 per cent of the voters. In combination with the policy positions of the parties that entered the Bundestag

Adenauer II

Adenauer III

Adenauer IV

Adenauer V Adenauer VI Adenauer VII Adenauer VIII Adenauer IX Erhard I Erhard II Erhard III Kiesinger Brandt I Brandt II Schmidt I Schmidt II Schmidt III Schmidt IV

2

3

4

5 6 7 8

9 10 11 12 13 14 15 16 17 18 19

1949-09-15

Adenauer I

1

1962-12-13 1963-10-16 1965-10-20 1966-10-28 1966-12-01 1969-10-21 1972-12-14 1974-05-16 1976-12-15 1980-11-05 1982-09-17

1957-10-22 1960-07-02 1961-11-07 1962-11-19

1956-02-25

1955-07-23

1953-10-09

Date in

Cabinet Cabinet number

1976-10-03 1980-10-05

1969-09-28 1972-11-19

1965-09-19

1961-09-17

1957-09-15

1953-09-06

1949-08-14

CDU/CSU, FDP CDU/CSU, FDP CDU/CSU, FDP CDU/CSU CDU/CSU, SPD SPD, FDP SPD, FDP SPD, FDP SPD, FDP SPD, FDP SPD

CDU/CSU, FDP, DP CDU/CSU, FDP, DP, GBBHE CDU/CSU, FDP, DP CDU/CSU, DP, DA-FVP CDU/CSU, DP CDU/CSU CDU/CSU, FDP CDU/CSU

Election date Party composition of cabinet

Table 8.1 German cabinets since 1949

mwc mwc mwc min mwc mwc mwc mwc mwc mwc min

sur mwc mwc min

sur

sur

sur

mwc

61.7 61.7 59.3 49.4 90.1 51.2 54.6 54.6 51 54.5 43.3

57.7 54.5 61.9 48.3

57.7

64.7

68.6

52

Type Cabinet of strength cabinet in seats (%)

499 499 496 496 496 496 496 496 496 497 497

497 497 499 499

487

487

487

402

3 3 3 3 3 3 3 3 3 3 3

4 4 3 3

7

6

6

9

2.52 2.52 2.38 2.38 2.38 2.24 2.34 2.34 2.31 2.44 2.47

2.39 2.38 2.51 2.52

2.7

2.67

2.77

3.99

FDP FDP FDP FDP FDP FDP FDP FDP FDP FDP FDP

CDU/CSU CDU/CSU FDP FDP

CDU/CSU

CDU/CSU

CDU/CSU

FDP

Number of Number of ENP, Median party seats in parties in parliament in first policy parliament parliament dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

Kohl V Kohl VI Schroeder I Schroeder II Merkel I Merkel II Merkel III Merkel IV

24 25 26 27 28 29 30 31

1991-01-17 1994-11-15 1998-10-27 2002-10-22 2005-11-22 2009-10-28 2013-12-17 2018-03-14

1982-10-01 1983-03-29 1987-03-11 1990-10-30 1990-12-02 1994-10-16 1998-09-27 2002-09-22 2005-09-18 2009-09-27 2013-09-22 2017-09-24

1983-03-06 1987-01-25

CDU/CSU, FDP CDU/CSU, FDP CDU/CSU, FDP CDU/CSU, FDP, DSU CDU/CSU, FDP CDU/CSU, FDP SPD, GR SPD, GR CDU/CSU, SPD CDU/CSU, FDP CDU/CSU, SPD CDU/CSU, SPD mwc mwc mwc mwc mwc mwc mwc mwc

mwc mwc mwc sur 60.1 50.7 51.6 50.7 73 53.4 79.9 56.3

56.1 55.8 54.1 54.6 662 672 669 603 614 622 631 709

497 498 497 663 5 5 5 5 5 5 4 6

3 4 4 5 2.65 2.91 2.9 2.8 3.44 3.97 2.8 4.64

2.47 2.51 2.8 2.93 FDP FDP SPD GR GR CDU/CSU GR CDU/CSU

FDP FDP FDP FDP

Notes: For a list of parties, consult the chapter appendix. Median parties are retrieved from Müller and Strøm (2000) and Strøm et al. (2003) for the 1949–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015). The first policy dimension is economic left–right. The number of parties in parliament does not include parties that have never held more than two seats when a cabinet has formed. Cabinet types: min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition. Table legends and variables are further defined in the Appendix, this volume.

Kohl I Kohl II Kohl III Kohl IV

20 21 22 23

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

254

  .

and the parties’ pre-electoral coalition statements (see section ‘Electoral alliances and pre-electoral coalitions’), the weakness of the SPD put the Christian Democrats in a strong bargaining position. In fact, the CDU/CSU controlled the median legislator on the economic left–right dimension. After winning almost 15 per cent of the votes in the 2009 Bundestag election, which made the majority for the then-formed Christian–Liberal coalition government possible (Niedermayer 2011; Vorländer 2011), the Free Democrats went into a serious electoral decline. They failed to implement tax reforms, which was a key promise in their 2009 election manifesto (Rixen 2014), lost parliamentary representation in most of the German state legislatures, and did not pass the five per cent threshold in the 2013 national election. After the 2013 election, a situation similar to that experienced in 2005 emerged and the Social Democrats entered government. A left-bloc coalition was not feasible due to the pariah status of the Left party. Yet, because of the weak electoral performance after the first Merkel cabinet, the SPD leadership initiated a referendum among party members, asking them if they support the coalition agreement that the SPD negotiated with CDU and CSU (Sturm 2014; Saalfeld and Zohlnhöfer 2015). This threat for the success of the coalition negotiations brought the Social Democrats into a very good bargaining situation, and a huge majority of 76 per cent of the party members were in favour of Germany’s third grand coalition in post-war history (Spier and von Alemann 2015: 65). More recently, a new party challenger has arisen on the right. Germany has been one of the few West European countries without a strong right-wing populist party, but this situation changed following the Euro crisis in 2010/2011. The AfD was founded 2013 in opposition to the monetary policy of Chancellor Merkel in the Eurozone (Arzheimer 2015). Originally, it focused on economic and European policy issues, but quickly added immigration issues to its profile and adopted nativist and nationalist positions after the ‘refugee crisis’ in 2015 (see e.g. Franzmann 2018; Bräuninger et al. 2019; Kortmann and Stecker 2019). The AfD almost gained parliamentary representation in the 2013 Bundestag elections by winning 4.7 per cent of the vote, entered three East German state parliaments in 2014, and won 7.1 per cent of the votes in the 2014 European Parliament election. After severe intra-party conflicts and shifting the focus to migration issues in course of the ‘refugee crisis’, the AfD won up to 24 per cent in state elections in 2016, secured parliamentary representation in the Bundestag by winning 12.6 per cent of the popular vote in September 2017 and become the third largest party. Following suit with most party systems in Western Europe, Germany also features a right-wing populist party which mobilizes the economic losers of globalization and emphasizes issues of national identity and cultural change (Kriesi et al. 2012). As a result, there are now two parties that are shed by the four mainstream parliamentary party groups in the Bundestag. In total, the Left and the AfD

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



255

control more than 20 per cent of the seats in the parliament elected in September 2017, but the two parties are not considered to be potential partners in a coalition government. These structural changes in the German party system have not led to new coalition formation strategies by the parties represented in parliament. The electoral success of the AfD further lowered the chances that a two-party coalition among centre-left parties (SPD and Greens) or centre-right parties (CDU/CSU and FDP) could win a parliamentary majority. Both camps struggle at the federal level to form coalitions that include at least one party from the other ideological camp. Recently, there have been some small and hesitant attempts to form coalitions such as those between Christian Democrats, Greens, and Liberals after the 2017 Bundestag election. While such ‘unusual’ coalitions already exist at the state level—for instance, the Free Democrats formed a coalition government with SPD and Greens in Rhineland-Palatinate in 2016, while in Saxony-Anhalt the Greens joined the incumbent coalition government of CDU and SPD in the same year—the German parties at the federal level seemingly have problems in forming ‘innovative’ coalitions. The default option at the federal level thus has become a coalition between Christian Democrats and Social Democrats, which by now controls only a narrow majority of the seats.

Electoral alliances and pre-electoral coalitions Pre-electoral commitments follow directly from the structure of the German party system and reflect recent changes. Again, a two-bloc logic was dominant from the 1980s until the mid-2000s. Pre-electoral commitments were made within each bloc with different levels of commitment. Up to the 2005 election, the minor parties made strong pre-electoral commitments to join only a coalition within the bloc, Liberals promised not to enter into coalitions with the Social Democrats/ Greens, and a coalition of the Greens and the Christian Democrats (and Liberals) was not yet feasible. In addition, there were strong public commitments not to enter into a coalition with the Left party. In 2005, 2013, and 2017/2018, this led to a situation in which only a coalition between the two main parties of the centre was feasible. Hence a very limited set of options was on the table, which has been striking from the perspective of coalition formation models. From a theoretical perspective, the Social Democrats were in a strong position in 2005 and 2013: the party was the second largest in parliament and close to the median position, with several parties to its left and right. Though the CDU had more seats in parliament, it was more isolated on the right of the policy space. However, a left coalition was not feasible in 2005 and 2013 due to conflicts following from the support of the Left by former Social Democrats and the fierce campaigning of the Left against the Social Democratic welfare reforms of the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

256

  .

2000s. At the federal level, no new coalition options were seriously considered. Only the level of commitment that political parties put into pre-electoral commitments changed. They still stated a preference for an in-bloc coalition, but put less emphasis on ruling out other options.

Government formation The bargaining process Government formation in Germany is structured by informal rules and procedures and can generally be characterized as a swift and effective process. As described in Table 8.2, there has been only a single failed attempt to form a multiparty government over the last two decades. In a similar vein, coalition negotiations have taken approximately four weeks on average, with a maximum of 36 days following the general elections in 2013—if we leave out the complex government formation process in 2017/2018. While the actual negotiations between parties thus usually take between three and five weeks, Table 8.2 suggests that there is considerably more variation if we look at the entire government formation period. In fact, the formal coalition negotiations in the German ‘freestyle bargaining system’ (Laver and Schofield 1990: 208) are regularly preceded by informal (mostly bilateral) exploratory talks between parties (Sondierungsgespräche), which are usually initiated by the strongest parliamentary party. Informal talks played a particularly important role after the 2005 and the 2013 elections and largely explain the prolonged government formation periods of 65 days and 86 days respectively. Both these elections resulted in a political stalemate as neither the centre-right (CDU/CSU and FDP) nor the centre-left (SPD and Greens) controlled a majority of seats in the Bundestag and a within-bloc coalition was not feasible. Consequently, parties had to look beyond established coalition options and, in both instances, the Christian Democrats and the Greens discussed the possibility of a joint government. Yet, these talks were substantially more advanced in 2013 as compared to 2005, in which CDU/CSU and Greens would have needed the additional support of the Liberal party to win a majority in the Bundestag. A further retarding factor for the formation process in 2005 was the refusal of the Social Democrats, in particular by the incumbent Chancellor Schröder, to recognize the joint parliamentary group of CDU and CSU as the largest party and their informal right to nominate the future chancellor in a coalition that includes the Christian Democrats (Jun 2007). While Schröder on election evening stated that his SPD would never elect CDU chair Angela Merkel as chancellor, Schröder’s decision to sidestep a position at the cabinet table a few days after the election was crucial for the successful subsequent coalition negotiations

Year in

1987 1990 1991 1994 1998 2002 2005 2009 2013 2018

Cabinet

Kohl III Kohl IV Kohl V Kohl VI Schroeder I Schroeder II Merkel I Merkel II Merkel III Merkel IV

0 0 0 0 0 0 0 0 0 1

Number of inconclusive bargaining rounds

CDU/CSU, SPD (1) CDU/CSU, FDP, GR

Parties involved in the previous bargaining rounds

Table 8.2 Cabinet formation in Germany, 1987–2018

33 0 42 16 20 22 26 22 55 32 33

Bargaining duration of individual formation attempt (in days) 45 0 46 30 30 30 65 31 86 141

Number of days required in government formation

33 0 42 16 20 22 26 22 55 113

Total bargaining duration

6 9 0 27 2 12 4 9 9

378 338 351 305 397 323 462 364

Abstention 253

Pro

257 333 287 292 202 285 150 315

225

Contra

Result of investiture vote OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

258

  .

between Christian Democrats and Social Democrats in 2005. Overall, these two episodes of government formation in 2005 and 2013 highlight the important differentiation between the length of the entire government formation process and the actual bargaining duration between parties (Diermeier and van Roozendaal 1998; Golder 2010). The comparatively short government formation periods in Germany in the time period until 2013 are largely the result of established patterns of coalition politics on the one side and the low rate of government turnover on the other. In fact, the formal coalition negotiations often start only a few days after the elections, and even in situations in which none of the established coalition options controls a legislative majority, at least one of the major parties is in government during the negotiations.¹ Another expediting factor is the coalition signals and pre-electoral commitments that parties send before the elections, in which they announce not only their preferred but also the excluded coalition options (Debus 2007, 2009). Prominent examples of such signals are the reciprocal statement of intent by the Christian Democrats and the Liberals to form a centre-right coalition before the 2005 elections and the unilateral rejection of a possible ‘traffic light’ coalition between SPD, Greens, and FDP by the Liberal party in the 2009 campaign (Schubert 2014). These signals considerably reduce the set of potential governments and shorten the phase of orientation immediately after the elections (Diermeier and van Roozendaal 1998; Ecker and Meyer 2020). Finally, the formal coalition negotiations and the drafting of the coalition agreement follow an established modus operandi, which ensures their timely conclusion. Here, issue-specific working groups discuss conflictual issues and formulate compromise solutions. Issues that cannot be resolved within these working groups are delegated to a larger round encompassing all negotiators to determine possible package deals across issue areas. All remaining conflictual issues are then brought to a small body generally comprised of the party leaders who ultimately resolve them. As all our interviewees consistently indicate, it is this small body that also discusses possible ministerial candidates and the quantitative and qualitative allocation of portfolios among the government parties. In a final step, the parties’ executive body, a special party congress or—in case of a high degree of internal party conflict over the next coalition government—a referendum among all party members ratifies the coalition agreement. The latter was, for example, the case within the SPD after the coalition negotiation talks in 2013 and 2018.

¹ In fact, the Schröder I cabinet was the first cabinet in the post-war period that experienced a complete change of government parties (see section ‘The composition and size of cabinets’).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



259

The government formation process of the Merkel IV cabinet marks the one main outlier and exception to this general pattern of swift government formation. It was only on 14 March 2018, almost half a year after the general elections in September 2017, that she passed the investiture vote and assumed office for the fourth consecutive time. With 172 days, the government formation process of 2017/2018 is more than twice as long as the second longest period of government formation in Germany since 1945. Naturally, the failed attempt to form a coalition between Christian Democrats, Liberals, and the Greens considerably contributed to this lengthy period and substantially increased the political uncertainty. For the first time in post-war Germany, new (early) parliamentary elections were considered a viable alternative. Both the considerable concessions by the CDU/CSU in terms of policy and ministerial positions and the rising political pressure on the SPD, not least by President Steinmeier (SPD), finally resulted in a coalition government formed between CDU/CSU and SPD.

The composition and size of cabinets According to the established empirical pattern in Germany since the 1960s, Table 8.1 indicates that all governments formed after 2000 are minimal winning coalitions. Furthermore, we observe that most of these recent cabinets follow a pattern of partial turnover in which one of the two government parties stays in office (Mershon 2002). In this context, the Schröder I government, which was formed in 1998, marks the main exception. It was the first German cabinet to include the Green party at the federal level about 10 years after such a red–green cabinet first formed at the state level (Börner 1985 cabinet in the State of Hesse). It was also the only cabinet in post-war Germany that experienced a full government turnover. The major difference we observe in government formation since the turn of the millennium is the increased frequency of grand coalitions formed by the two major German parties (see Table 8.1). A grand coalition was long considered as last resort reserved for exceptional situations by main political actors and commentators. Indeed, the CDU/CSU–SPD coalition led by Chancellor Kiesinger (CDU) between 1966 and 1969 was the only such government within six decades of post-war German democracy. Over the last decade, however, grand coalitions have become an established form of multiparty government and an option regularly preferred over alternative governments. Table 8.1 shows that three of the five most recent cabinets unite the two Christian democratic parties and the Social Democrats at the cabinet table. This increased frequency of grand coalitions is largely due to the interplay of two main developments. First, the fragmentation of the party system as a consequence

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

260

  .

of the gradual transformation from a two-and-a-half party system until the early 1980s to a more polarized multiparty system since 2005 with a decreasing vote share for the two major parties has exponentially increased the set of potential governments. The limited choice set of the 1960s to 1980s with a centre-right coalition between Christian Democrats and Liberals and a centre-left coalition between SPD and FDP has now been expanded and parties may choose from a much broader set of coalition governments. The downside of this development is that the established coalitions—since the 1990s between CDU/CSU and FDP on the one side and SPD and Greens on the other—often do not control a majority in the Bundestag since the 2005 parliamentary election. Second, the numerous incompatibilities and the presence of ‘pariah’ parties in parliament (Van Spanje and Van Der Brug 2007, 2009) have increased the complexity of government formation and thus reduced the number of politically viable coalition alternatives. Innovation in cabinet formation can be found at the state level (Müller 2018). Here, we observe many new coalition types that seemed impossible at the beginning of the 2000s and became feasible over the last decade. The Greens have been in several state-level coalitions with the Christian Democrats since such a coalition was formed for the first time in 2008 in the state of Hamburg. In the Eastern states, the Left became a viable coalition partner for Social Democrats and has even led a coalition in Thuringia since 2014. In a major West German state, North RhineWestphalia, a minority cabinet by Social Democrats and Greens was in office between 2010 and 2012 with the support of the Left. In 2019, Social Democrats, Greens, and the Left even formed the first coalition government in the West German state of Bremen. The Christian Democrats went into a short-lived coalition with a regional right-wing populist party between 2001 and 2004 in Hamburg and joined a coalition with the Greens as a junior partner in BadenWürttemberg in 2016. The experience won at the state level has informed the recent government formation at the federal level in 2017. It has not yet taken in the dynamics from the states. At the federal level, the paradoxical situation after the 2013 election nicely exemplifies the increased complexity of the bargaining situation. In this context, office-based theories would predict a coalition of the Social Democrats, the Greens, and the Left party, which jointly controlled 320 of the 631 seats in the Bundestag. Yet, although all three parties are part of the ‘left’ ideological spectrum, the substantial policy differences in particular between the SPD and the Left— especially in foreign and defence policy but also in economic and welfare policy issues—made such a government highly unlikely, and in fact the SPD considered the Left to be non-coalitionable at the federal level. Even though the three left parties thus controlled a majority of seats, the SPD preferred to join a coalition as the junior partner of the CDU/CSU led by Chancellor Merkel, as the results of the referendum among the SPD members on the coalition agreement between CDU, CSU, and SPD in 2013 indicate.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



261

The allocation of ministerial portfolios The quantitative allocation of ministerial portfolios in Germany is generally proportional to each party’s seat contribution to the government. In line with the general pattern in European parliamentary democracies, however, we observe a slight overcompensation of smaller government parties (Warwick and Druckman 2006). For instance, the Greens in 2002 controlled 3 (21 per cent) of the 14 cabinet positions in the Schröder II cabinet although contributing approximately 18 per cent of the seats in parliament. In a similar vein, the Liberals obtained 5 out of 15 cabinet positions (33 per cent) while controlling approximately 28 per cent of the cabinet seats in 2009. As in other European parliamentary democracies, ministerial portfolios in Germany vary substantially in (constitutional) importance and party-specific salience (Druckman and Warwick 2005; Druckman and Roberts 2008). The partisan distributions over five important portfolios are shown in Table 8.3. Specifically, government parties value cross-cutting portfolios such as finance, internal affairs, and justice as well as the head of the Federal Chancellery and particularly prestigious posts such as foreign affairs. Besides this shared understanding among government parties on the value of different ministries, parties also have particular preferences for specific ministries (Budge and Keman 1990; Bäck et al. 2011). The classic cases are the strong preferences of the Green party for the Ministry of Environment and that of the Social Democrats for the Ministry of Social Affairs. In addition, the Liberals have particularly strong ministerial preferences for the Ministry of Economic Affairs and the Ministry of Justice while the CSU aims to control the Ministry of Agriculture. The actual allocation of ministerial portfolios to government parties and individual ministers is a complex interaction between parties. Generally speaking, our interviewees suggest that parties follow the principle of alternation and choose ministries or subset of ministries sequentially based on party size (O’Leary et al. 2005; Ecker et al. 2015; Raabe and Linhart 2015). Within this framework, however, parties act strategically and initial proposals are often made with the sole intention to set the agenda for the subsequent round of negotiations.² Thus, while alternation is the general pattern of portfolio allocation, it is not a deterministic function of size and party preferences. The final stage of the nomination of ministerial candidates within parties is as complex as the distribution between government parties. Here, party leaders have to consider and

² The interview evidence indicates that party leaders anticipate other parties’ strong preferences for individual ministries and exploit those preferences to their advantage. Specifically, they will sometimes claim ministries that are of particular value to their coalition partners to then exchange them with (several of) their high-valued ministries.

Year in

1987 1990

1991 1994 1998 2002 2005 2009 2013 2018

Cabinet

Kohl III Kohl IV

Kohl V Kohl VI Schroeder I Schroeder II Merkel I Merkel II Merkel III Merkel IV

15 CDU/CSU, 4 FDP 19 CDU/CSU, 5 FDP, 1 DSU 15 CDU/CSU, 5 FDP 15 CDU/CSU, 3 FDP 13 SPD, 3 GR 11 SPD, 3 GR 8 CDU/CSU, 8 SPD 11 CDU/CSU, 5 FDP 10 CDU/CSU, 6 SPD 10 CDU/CSU, 6 SPD 20 18 16 14 16 16 16 16

19 25

2 CDU/CSU, 1 FDP 1 CDU/CSU

4 CDU/CSU, 1 FDP 2 FDP, 1 CDU/CSU 1 SPD, 1 GR 3 GR, 2 SPD

2 CDU/CSU, 2 FDP 3 CDU/CSU, 2 FDP

Number of Number of ministers Total per party (in number watchdog junior ministers per party descending order) of ministers

Table 8.3 Distribution of cabinet ministerships in German coalitions, 1987–2018

19 17 15 14 15 15 15 15

18 18 CDU/CSU CDU/CSU SPD SPD CDU/CSU CDU/CSU CDU/CSU CDU/CSU

CDU/CSU CDU/CSU

Number 1 Federal of chancellor ministries

CDU/CSU CDU/CSU SPD SPD SPD CDU/CSU CDU/CSU SPD

CDU/CSU CDU/CSU

2 Finance

FDP FDP GR GR SPD FDP SPD SPD

FDP FDP

3 Foreign affairs

SPD CDU/CSU SPD SPD

CDU/CSU CDU/CSU SPD

CDU/CSU CDU/CSU

4 Labour and social affairs

CDU/CSU CDU/CSU SPD SPD CDU/CSU CDU/CSU CDU/CSU CDU/CSU

CDU/CSU CDU/CSU

5 Interior

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



263

proportionally represent a series of personal traits and characteristics such as age, gender, intra-party group, and regional representation.

Coalition agreements The policy position of the government laid down in a coalition agreement can be interpreted as a joint understanding over how to distribute policy payoffs among the coalition partners (Däubler and Debus 2009). In this view, policies amount to benefits for the governing parties, comparable to manifest office payoffs, such as cabinet posts. Analyses of the full text of German coalition agreements—at both the federal and state levels—show that the coalition agreements cover the positions of the coalition parties, weighted by their intra-coalition seat strength (Debus 2008). All recent multiparty governments in Germany since 1998 are based on extensive coalition agreements. As in most other European parliamentary democracies, these agreements are generally considered as the most authoritative documents that constrain party behaviour in government (Müller and Strøm 2000). Our interviews confirm the importance of coalition agreements, in particular at the beginning of a new cabinet. They are the key aspect of the government formation process and fulfil both an internal and an external function. Internally, the general guidelines and the specific policy proposals are the main reference point for policy-making and possible renegotiations between parties during the legislative term. They are in fact a contract between parties in government. Externally, they communicate the government agenda to the general public and are meticulously analysed by the media, interest groups, and party factions alike. Coalition agreements have continued to grow considerably in length and coverage since the 2000s. As Table 8.4 suggests, their length has more than doubled from approximately 26,600 words in 2002 to approximately 62,000 words in 2013 and 2018. This substantial within-country variation and increasing length of agreements conform to established patterns among Western European democracies (Müller and Strøm 2000; Indridason and Kristinsson 2013). At the same time, their general structure has been largely constant in recent times. Specifically, coalition agreements in Germany revolve almost exclusively around government policies. Only a fraction of the coalition agreements, for example 2 out of 132 pages in the 2009 agreement, is concerned with codifying procedural rules and distributing government offices. These sections are often adopted and slightly adjusted from previous agreements. The substantial increase in length and coverage over the last two decades is a prevalent challenge for existing explanations of the length of agreements. Specifically, bargaining complexity as one key factor influencing how extensive

Year in

1961 1962 1963 1980 1982 1983 1987 1991 1994 1998 2002 2005 2009 2013 2018

Coalition

Adenauer VII Adenauer IX Erhard I Schmidt III Kohl I Kohl II Kohl III Kohl V Kohl VI Schroeder I Schroeder II Merkel I Merkel II Merkel III Merkel IV

1,837 513 513 1,199 3,350 2,341 7,153 15,322 10,580 16,536 26,614 44,038 41,115 62,009 61,524

Size 16.4 27.9 27.9 0 0 0 0 0 0 1.6 0.9 0.7 0.6 0.6 0.7

General rules (in %) 3.4 4.1 4.1 0 0 0 0 0 0 0 0 0.2 0 0 0.2

Policy-specific procedural rules (in %)

Table 8.4 Size and content of coalition agreements in Germany, 1949–2018

0 0 0 0 0 0 0 0 0 1.1 0.8 0.3 0.2 0.1 0.2

Distribution of offices (in %) 0 0 0 0 0 0 0 0 0 0 0.1 0.4 0.6 0.3 0

Distribution of competences (in %)

80.2 68 68 100 100 100 100 100 100 97.3 95.5 95.7 95.3 95.6 97.8

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



265

coalition agreements are has been largely constant since 2000 (Indridason and Kristinsson 2013). In a similar vein, policy conflict between government parties has not intensified to such an extent that would warrant such drastic increase in the comprehensiveness of policy agreements. Moreover, the iterated interactions between parties during this period would suggest that the increased trust on informal coordination between parties alleviates the necessity to codify the joint policy platform (Strøm and Müller 1999). One potential explanation for the increasing length of coalition agreements based on evidence from our qualitative interviews is their changing nature over time. Specifically, coalition parties agreed that the status quo prevails in all policy areas that were not explicitly mentioned. Coalition agreements were thus largely understood as policy programmes that outlined the main policy goals of the incoming government. This tacit agreement became increasingly difficult to uphold as agreements were more and more interpreted as impossibility contracts (Verhinderungsabkommen) which ruled out any changes of the status quo unless explicitly specified. This rigid interpretation limited the room for manoeuvre, in particular when responding to unforeseen events so that parties now also specify the current state of affairs in many policy areas. Indeed, more detailed data on the actual content of the coalition agreements suggest that the share of policy statements pledging to contain the status quo has increased from approximately 8 per cent to 16 per cent of all coalition pledges (Thomson et al. 2017).

Coalition governance The role of individual ministers in coalition governance The constitution provides three modes of government decision-making in Article 65 and thus reflects the three ideal types of coalition governance: the ‘Dominant Prime Minister Model’, the ‘Coalition Compromise Model’, and the ‘Ministerial Government Model’ (e.g. Müller-Rommel 1994; Bergman et al. 2019). Article 65 gives a prominent role to the chancellor (Richtlinienkompetenz), emphasizes that ministers have full responsibility for their respective ministries, and also highlights cabinet conflict resolution. In practice, the role of ministers has been dependent on particular cabinet dynamics but most importantly on the standing of ministers within their party, as our interviews suggested. Which of the three ideal types of coalition governance dominates in the German cabinet depends, however, on the type of the coalition on the one hand and the chancellor’s ‘style’ of governing on the other: while the government style of the first chancellor of Germany after the Second World War, Konrad Adenauer (CDU), became norm setting (Kanzlerdemokratie; see Niclauß 2004), other chancellors preferred the adoption of the ‘Coalition Compromise Model’. For instance,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

266

  .

such politicians were more likely to be appointed as ministers in the ‘red–green’ cabinet the closer their policy position was to the one of Chancellor Schröder, which points into the direction that the ‘Prime Minister Model’ was dominant in between 1998 and 2005 (Bäck et al. 2016). Yet, in the first cabinet led by Chancellor Merkel, politicians were more likely to become cabinet members the smaller their distance was to their own party, that is to CDU/CSU or SPD in the cabinet formed in 2005. In Germany, the portfolio allocation does already provide information about the relevance of ministries and the emphasis parties put on particular portfolios. Political parties have strong preferences for specific portfolios because controlling them makes it easier to implement a party’s policy positions to cater to the interests of its voters and supporters (Budge and Keman 1990; Bäck et al. 2011). For the coalitions we study, the Social Democrats, for instance, were always in control of the Ministry of Social Policy whereas the Greens had the Ministry of the Environment. In coalition bargaining, parties weight the importance and prestige of portfolios and their relevance for their policy platform (Ecker et al. 2015). In addition, the constitutionally provided competencies of some ministries make them an attractive target during the coalition negotiations. The Ministry of Finance is considered to be the most important cabinet portfolio because the minister of finance has the privilege to veto the budget. Table 8.5 presents the array of other governance mechanisms that are used. Potential candidates for the ministerial positions are mostly either senior and well-established party members with significant political experience or policy experts in a particular field. All ministers are expected to put a party’s policy platform into actual policies and are chosen by their ability to do so. There is more flexibility in the allocation of minor portfolios and other constraints are taken into account. The less important ministries are used to balance internal party constraints such as regional and gender requirements as well as to ensure an inclusion of relevant actors into cabinet.

Coalition governance in the executive arena Saalfeld (2000: 59) highlights that the coalition agreement of the Schröder I cabinet in 1998 is ‘unusually explicit’ with respect to the details of conflict management rules and deviates from previous coalition agreements. This high degree of formality remains for cabinets in the succeeding period we study here. However, they are different from the informal mechanisms that exist to regulate conflicts between coalition partners and that are more in line with dynamics Saalfeld (2000) describes for previous cabinets. There is a strong norm that no dissent will be taken into cabinet meetings. In practice, no controversial topics are put on the agenda for a cabinet meeting as our

1982 POST

1983 POST

1987 POST

Kohl I

Kohl II

Kohl III

Y

Y

Y

N/A N/A N/A N/A N/A N/A Y

N N N N N N POST

1965 1966 1969 1972 1974 1976 1980

N/A N/A N/A N/A N/A

Erhard II Kiesinger Brandt I Brandt II Schmidt I Schmidt II Schmidt III

N N N N N

N N N

1949 1953 1955 1956 1957

N

N

N

N N N N N N N

N N N

N N N N N

PS

Parl, PS, IC, Pca Parl, PS, IC, Pca Parl, PS, IC, Pca Parl, PS, IC, Pca PS

PS

PS

Pca Parl

Parl Pca Pca

Pca Pca Pca Pca Pca

Pca Pca, Parl

Parl Parl, Pca Parl, Pca

Pca, Parl Pca Pca Pca Pca

All used

PS

PS

PS

PS

Pca Pca

Parl Pca Pca

Pca Pca Pca Pca Pca

Y

Y

Y

N (CSU) N (CSU) N (CSU) N (CSU) N (CSU, DP) N (FDP) N (FDP) N (CDU, CSU) N(CSU) N(CDU) Y Y N (SPD) N (SPD) N (SPD)

-

-

-

-

-

-

Personal Issues union excluded from Most For most agenda common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Adenauer VII 1961 POST Adenauer IX 1962 IE Erhard I 1963 IE

Adenauer I Adenauer II Adenauer III Adenauer IV Adenauer V

Coalition

Table 8.5 Coalition governance mechanisms in German coalitions, 1949–2018

All/All

All/All

All/All

All/All Most/Most All/All All/All All/All All/All All/All

All/All All/All All/All

All/All All/All All/All All/All All/All

Coalition discipline in legislation/ other parl. behaviour

N

N

N

N N N N N N N

N N N

N N N N N

Comp.

Varied

Varied

Few Few Varied Varied Varied Varied Varied

Varied Few Few

Few Few Few Few Few

Y

Y

Y

Y Y Y Y Y Y

Continued

Y

Y

Y

Y Y Y Y Y Y Y

Y Y Y

Y Y Y Y Y

Freedom of Policy Junior Nonappointment agreement ministers cabinet (watchdogs) positions

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

1991 POST

1994 POST

1998 POST

2002 POST

2005 POST

2009 POST

2013 POST

Kohl V

Kohl VI

Schroeder I

Schroeder II

Merkel I

Merkel II

Merkel III

Y

Y

Y

Y

Y

Y

Y

N/A

N

N

N

N

N

N

N

N

Parl, PS, IC, Pca CoC, PS, IC, CaC Parl, PS, IC, Pca CoC, CaC, IC COC, CaC, IC CoC, CaC, Pca CoC, CaC, Pca CoC, CaC, Pca CaC

CaC

CaC

CaC

CaC

PS

PS

PS

Pca

Pca

Pca

IC

IC

PS

PS

PS

N (CSU)

N (SPD, CSU) N (CSU)

N (GR)

N (GR)

Y

N (FDP)

N

N

N

N

-

-

-

-

Most/All

Most/All

Most/All

Most/All

All/All

All/All

All/All

All/All

Coalition discipline in legislation/ other parl. behaviour

Y

Y

Y

Y

N

N

N

N

Comp.

Comp.

Comp.

Comp.

Comp.

Comp.

Comp.

Comp.

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Freedom of Policy Junior Nonappointment agreement ministers cabinet (watchdogs) positions

Notes: Coalition agreement: IE = inter-election; POST = post-election; N = no coalition agreement Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee; CoC = Coalition committee; Pca = Combination of cabinet members and parliamentarians; PS = Party summit; O = Other Coalition discipline: All = Discipline always expected; Most = Discipline expected on all matters except those explicitly exempted Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement

1990 N

All used

N (FDP)

Personal Issues union excluded from Most For most agenda common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Kohl IV

Coalition

Table 8.5 Continued OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



269

interviews highlight. The agenda for cabinet meetings is agreed upon before and will not be changed. Voting will only take place on issues where agreement has been found. There is some time left at the end of a cabinet meeting to talk about open issues. However, even this last part of a cabinet meeting is mainly used to discuss recent events and to coordinate cabinet activities. All major differences are dealt with outside the cabinet. The main responsibility in finding a cabinet agreement lies with the Federal Chancellery (Bundeskanzleramt) and its head, a junior minister. This signals the evidence for the prevalence of the ‘Dominant Prime Minister Model’, even in coalitions between two equally large parties—CDU/CSU and SPD—and during the chancellorship of Angela Merkel. Coordination between ministers and ministries takes place in the Federal Chancellery and issues of lower salience are dealt with at this level. The Chancellery has the key role in all regular cabinet coordination among ministers and ministries. More controversial issues are dealt with at other levels. At the parliamentary arena, party group leaders coordinate with ministers within a party group. Members of parliament with policy expertise in the respective issues are included to ensure coherent voting behaviour of parliamentary groups. There is also some coordination between party group leaders and policy experts of the coalition partners in parliament and outside the committee system. Junior ministers (Parlamentarische Staatssekretäre) recruited by the coalition partners have a low visibility and little importance in cabinet decision-making. These positions are an important mechanism to balance the interests and positions within coalition parties. At the appointment stage, junior minister positions are allocated to important senior party members. They are put into cabinet mainly to include all relevant actors and to integrate and satisfy potential opponents. Policy expertise is less relevant for these junior ministerial positions. Comparative work has shown that junior ministers are used as a control device in some in some parliamentary systems (Lipsmeyer and Pierce 2011). In Germany, there is no watchdog mechanism and junior ministers mainly share the party affiliation with the respective minister. Junior ministers may be put into ministries held by coalition partners only to satisfy particular personal constraints. For example, in the Schröder II government, the smaller coalition partner, the Greens, initially demanded a fourth ministry after increasing their relative seat share compared to the Social Democrats. At the end of the coalition formation, they were given more junior ministerial positions to meet these demands. These discussions about junior ministers are limited to the politically appointed junior ministers. Within the German government, the most senior public officials (Beamtete Staatssekretäre) are also appointed by the minister and can be dismissed at any point. They are recruited within the ministry and are carefully selected. These public administrators are important for their policy expertise and to ensure a frictionless delegation within a ministry. They are of less relevance for inner cabinet dynamics so that we focus on (regular) junior ministers.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

270

  .

All cabinets since 1998 have had long coalition agreements. Within these agreements, there is an explicit section about the coalition committee and conflict resolution mechanisms between coalition partners. These sections create the impression of a highly formalized conflict resolution within cabinets and a strong role of the coalition committee. However, less emphasis is put on these mechanisms in practice according to our interviews. Important issues are dealt with informally between the major actors within cabinet and by integrating key actors from the parliamentary groups. There are several regular and ad hoc meetings with these main actors to settle disagreements. Within the executive, major conflicts are dealt with at an individual level and the process is less formalized than the coalition agreement suggests (see section ‘Governance mechanisms with different types of actors’). Regular coordination within the executive takes place at the Chancellery and is led by its head.

Governance mechanisms in the parliamentary arena The parliamentary groups of the parties that form the coalition government coherently support all bills initiated by the cabinet. Coalition discipline in the Bundestag applies both to votes over legislative proposals and to all other votes in parliament. This holds for all coalition governments in the time span under study here and is explicitly formulated in the coalition agreement. Exceptions in which members of parliament (MPs) from governing parties are ‘allowed’ to deviate from the party line are moral issues that touch fundamental ethical values. For these particular issues, there is often bipartisan legislative activity in terms of introducing and voting on law proposals (e.g. Baumann et al. 2015; Euchner and Preidel 2018). If an MP from a government party deviates from the party line on a ‘normal’ issue, the respective MP will have a meeting with the parliamentary party group leader, so that—as one of our interviewees mentioned—the MP can bring forward the reasons for the vote. However, the respective MP will be reminded how important it is to stick with the party line in terms of voting with the parliamentary party leadership and, thereby, supporting the government. An important control mechanism of the parliament is—according to Carroll and Cox (2012)—the committee system and the role committee chairs can play in this context. They argue that coalition parties can use the committee system to shadow the ministers of their partners and find that the greater the policy disagreement between a minister’s party and its partners, the more likely the minister is to be shadowed. The parliamentary committees in the German Bundestag are considered to be rather strong (André et al. 2016). However, the degree of shadowing in committees in Germany is rather weak (Carroll and Cox 2012) and the powers of committee chairs are considered to be weak as well

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



271

(Sieberer and Höhmann 2017). None of the former ministers we interviewed pointed into the direction that committee chairs play an important role in terms of controlling the minister of the coalition partner.

Governance mechanisms with different types of actors All recent multiparty governments agreed on detailed procedural rules on the composition and the meeting schedule of a coalition committee. The coalition committee generally includes the chancellor and the vice-chancellor, the leaders of the parliamentary party groups, and (if not already included) the party leaders.³ It meets at the beginning of each session week in parliament, but at least once a month, and provides a forum to coordinate cabinet policies and to resolve possible conflicts within government. The coalition committee is an important body of inter-party coordination, but all interviewees consistently indicate that it plays only a subordinate role in day-to-day policy-making. It is not the key body to resolve the most serious conflicts in government. In all Merkel cabinets, the coalition committee is also not conveyed on a regular basis, contrary to parties’ initial agreement. Overall, there are thus substantial differences between what parties codify as their main arena of conflict resolution and the actual practice coalition governance, which relies on alternative (informal) bodies. Concerning the coordination of government policies and the resolution of the most common conflicts between government parties, the head of the Federal Chancellery plays a crucial role. The Chancellery’s different departments are aligned with the individual ministerial portfolios (Spiegelreferate) and provide the necessary expertise to coordinate policy proposals that cut across portfolios and to resolve any ensuing conflicts between ministries. Only policy proposals in which all responsible ministries signal their approval beforehand are brought to the cabinet table, where the cabinet then decides unanimously. The Chancellery’s different departments not only have a coordinative function, but also provide a strong instrument to oversee the ministries controlled by other government parties. In contrast to prevalent theories of intra-cabinet coordination, it is the Federal Chancellery rather than ‘watchdog’ junior ministers that constitute the main executive oversight mechanism in German multiparty governments (see also section ‘Coalition governance in the executive arena’). An important informal governance mechanism is the recurrent meetings of the chancellor with the vice-chancellor, selected ministers, and key members of the parliamentary party groups. In the two Schröder cabinets (1998–2005), Chancellor Schröder and Vice-Chancellor Fischer often tried to solve conflicts if ³ In this context, the CDU and the CSU constitute two separate parties so that they are each potentially represented by one party and one parliamentary group leader in the coalition committee.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

272

  .

no agreement was found at other levels. Conflict resolution of Chancellor Merkel since 2005 has been conducted by gathering key actors on Sunday evenings regularly. It is the key mechanism to resolve potentially destabilizing conflicts between the government parties and the main body to coordinate the immediate government response to major exogenous shocks. Examples hereof include the 1999 NATO intervention in Serbia and Montenegro, the onset of the financial crisis in 2008, the European sovereign debt crisis in 2009/2010, and the Fukushima nuclear disaster in 2011. The strong role that the chancellor has in these informal mechanisms of coalition governance indicates again her or his strong general position and points towards the ‘Dominant Prime Minister Model’ of coalition governance.

Cabinet duration and termination The duration of cabinets Multiparty governments in Germany can be generally characterized as comparatively stable. As the data in Table 8.6 suggest, the average time in office of German cabinets since 1945 is approximately 800 days or two-thirds of their maximum possible duration. This puts Germany on the ninth rank of all EU-27 countries (both in absolute and in relative terms). This tendency of stability becomes even more pronounced when focusing exclusively on cabinets since the 1990s. Multiparty governments in this more recent sub-period last for approximately 1150 days on average, putting Germany on the fourth rank in the EU-27 (again also in relative terms). One factor explaining this increased stability of multiparty governments is the profound structural change of party competition in Germany since the mid1980s described in Section ‘Party system and the actors’. Specifically, the Liberals were the pivotal party in the prevalent two-and-a-half party system (Ware 1996), as they could choose to coalesce either with the Christian Democrats to their right or with the Social Democrats to their left. It was exceptionally successful in exploiting this strong bargaining position as a ‘hinge’ half party (Siaroff 2003) and featured in all but four cabinets in German post-war history until 1998. At the same time, the changes in the FDP’s preferred coalition partner in 1969 and 1982 (from the Christian Democrats to the Social Democrats and back to the Christian Democrats) were a key driving factor of premature government termination for much of the post-war period in Germany (Saalfeld 2000: 75ff.). Turning to the record of individual cabinets since the 1990s in Table 8.6, we observe that relative cabinet duration ranges from 37.9 per cent of the legislative term for the fourth Kohl cabinet (1990) to complete terms in case of the Schröder

1987-03-11 1990-10-30 1991-01-17 1994-11-15 1998-10-27 2002-10-22

2005-11-22 2009-10-28 2013-12-17

Kohl III Kohl IV Kohl V Kohl VI Schroeder I Schroeder II

Merkel I Merkel II Merkel III

2009-09-27 2013-09-22 2017-10-24

1990-10-30 1990-12-02 1994-10-16 1998-09-27 2002-09-22 2005-09-18

Date out

100 100 100

93.9 37.9 98.8 98.7 100 74.4

Relative duration (%)

1 1 1

5 1 1 1 1 4, 6

Mechanisms of cabinet termination

Terminal events

Parties (when conflict between or within)

Policy area(s)

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event Note: Relative duration: share of constitutionally allowed duration for this cabinet.

Date in

Cabinet

Table 8.6 Cabinet termination in Germany, 1987–2018

The governing coalition intentionally lost a vote of no confidence in order to trigger early elections which proceeded after legal challenges were dismissed by the supreme court.

Coalition joined by DSU

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

274

  .

I cabinet and the first three Merkel cabinets.⁴ The cabinets headed by Kohl, Schröder, and Merkel between 1990 and 2017 jointly constitute an era of outstanding cabinet stability, even for German standards. In fact, Table 8.6 indicates that all but one cabinet were terminated by regular elections. Accordingly, the relative duration of these cabinets closely approximates or equals one.

The termination of cabinets In line with the exceptional empirical record of high government stability in Germany, all but one cabinet between 1990 and 2017 were terminated by regular parliamentary elections. The one major exception to this pattern is the premature termination of the second Schröder cabinet between the Social Democrats and the Greens via early parliamentary elections in 2005. The German constitution allows to dissolve the Bundestag prematurely only in a few exceptional circumstances. Schröder had to purposely lose a vote of confidence while still controlling a majority in the Bundestag, so that MPs from the coalition parties abstained from voting.⁵ This instance of ‘engineering’ a lost parliamentary vote for the third time after Brandt in 1972 and Kohl in 1983 (see also Saalfeld 2000: 75) stirred a controversial debate about its legitimacy. Finally, the German constitutional court ruled on the lawfulness of this mechanism to arrive at early elections.⁶ It was the second time after 1983 that the court ruled on the use of the confidence vote to dissolve the Bundestag by a government that controlled a majority in the parliament. A previously restrictive legal interpretation was weakened and the ability of a chancellor to dissolve parliament through a confidence vote by abstention was strengthened (Reutter 2006). A number of considerations among the Social Democrats culminated in the decision of Chancellor Schröder to seek early elections. The core electorate withdrew their support in German state elections, punishing Social Democrats for welfare retrenchments implemented in 2003. On average, the party lost six percentage points in 11 state elections between September 2002 and May 2005 (von Alemann and Spier 2008). Approval ratings further suffered from the ongoing recession since 2002. Following the substantial electoral losses at the state elections in North Rhine-Westphalia on 22nd May 2005, a traditional Social Democratic stronghold, support for structural reforms further dwindled within

⁴ The Kohl IV cabinet was a transitional government between the German unification in October 1990 and the election on 2 December 1990. It included one additional cabinet member from the EastGerman DSU. Consequently, the observed minimum duration is largely an artefact of the standardized criteria that constitute a ‘new’ cabinet. ⁵ These are defined in Article 63, Paragraph 1, and Article 68, Paragraph 1, of the German constitution. See also Section ‘Institutional setting’. ⁶ It was labelled a ‘dissolving-targeted’ vote of confidence (auflösungsgerichtete Vertrauensfrage).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



275

the Social Democratic party. At the same time, this lost state election further strengthened the ability of the Christian Democrats to veto government proposals in the second chamber. Overall, the SPD’s unilateral decision to seek a renewed electoral mandate exemplifies how adverse regional election results, incongruence in bicameral systems, and intra-party conflict may culminate in discretionary government termination. Although all other German cabinets since 2000 have proved to be particularly stable and served throughout the entire constitutional inter-election period, this does not necessarily imply that there were no exogenous shocks. In fact, a series of events had the potential to destabilize and terminate multiparty governments in Germany. Prominent examples hereof are the NATO-led military intervention in Serbia and Montenegro in 1999, the European sovereign debt crisis in 2009/2010, and the Fukushima nuclear disaster in 2011. These events shook the ideological foundations of coalition parties and resulted in fundamental policy shifts. The decision of the Schröder I cabinet to intervene in the Kosovo War in 1999, for instance, was preceded by a fierce internal debate in the Green party and marked a radical shift from the party’s general pacifist position. At the same time, the governance structure of German multiparty governments proved to be flexible enough to accommodate these exogenous shocks. Specifically, it allowed the government parties to coordinate their policies and to swiftly respond to these events.

Conclusions The largest differences between the post-war period until the 1990s and the last two decades refer to the structure of the German party system and the policy profile of the main parties. These changes have affected the process of coalition formation and made it more complex, but the general rules of the game in coalition governments have not altered. The increasing vote share for the Left and—since 2013—for the AfD decreased the chances for traditional centre-left or centre-right majority coalitions between either SPD and Greens or CDU/CSU and FDP. For the two decades we study, ‘grand coalitions’ between Christian and Social Democrats have become a default option and formed three times since 1998. The developments over the next years will show if the German parties at the federal level become more open for policy compromises, so that coalitions of three parliamentary party groups between CDU/CSU, FDP, and Greens, between SPD, FDP, and Greens, between Social Democrats, Greens, and the Left, or even between CDU/CSU, SPD, and Greens are more likely to form, as it is already the case at the state level. In addition, the programmatic development of the AfD will be decisive for a potential cooperation between this new right-wing populist

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

276

  .

party with CDU/CSU and FDP. These parties share some overlap in their policy positions on issues such as economic or immigration policy (Bräuninger et al. 2019, 2020). Currently, a coalition between moderate and populist right-wing parties is impossible because the AfD seems to be dominated by a far-right nationalist faction that adopts nativist and anti-Islam positions. If a moderate wing within the AfD won control over the party, a cooperation or even a formal coalition between CDU/CSU, FDP, and the AfD could become more likely. Such new forms of coalitions would most likely also affect patterns of coalition governance like portfolio allocation, the content and length of the coalition agreements, and the survival of multiparty cabinets. As our findings indicate, the length of coalition agreements increased and—as the 2017/2018 government formation process has shown—finding compromises within a more heterogeneous coalition is more complicated. If the German party system remains complex and consists of six parties represented in parliament, then coalition negotiations will take more time than before, coalition agreements are likely to become more complex and to cover more detailed procedural rules, and the risk for early cabinet termination might increase. Finally, intra-party conflicts and discussions within parties over the ‘right’ track in terms of choosing coalition parties might destabilize future coalition governments in Germany. The recent use of party members’ referenda by the Social Democrats and similar developments by the Greens point into this direction. All ministers we interviewed highlighted the importance of the coalition agreements as the ‘policy programme’ for the whole legislative period, to which the coalition parties refer in cases of policy conflicts within the government. Given also the increasing length of the coalition agreements, we can indeed consider them as the most important factor in the process of coalition governance for all cabinets we studied here, whereas coalition committees and the frequency of these meetings depend strongly on the way how the chancellor prefers to ‘direct’ the cabinet and the coalition. Future research could focus at the state level and processes of coalition governance there. Since there is more variety in terms of the partisan composition of coalition governments at the Länder level, we could learn how more ideologically complex coalitions on that sphere of the German political system organize their daily business (e.g., Bowler et al. 2016). The latter would help to understand patterns of coalition governance in Germany at the federal level because the shift towards a more polarized and fragmented party system will make more heterogeneous coalitions more likely in the near future.

References Andeweg, Ruby B., Lieven De Winter, and Patrick Dumont (2011). Puzzles of Government Formation: Coalition Theory and Deviant Cases. Abingdon: Routledge.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



277

André, Audrey, Sam Depauw, and Shane Martin (2016). ‘ “Trust Is Good, Control Is Better.” Multiparty Government and Legislative Organization’. Political Research Quarterly, 69(1): 108–20. Arzheimer, Kai (2015). ‘The AfD: Finally a Successful Right-Wing Populist Eurosceptic Party for Germany?’. West European Politics, 38(3): 535–56. Bäck, Hanna, Marc Debus, and Patrick Dumont (2011). ‘Who Gets What in Coalition Governments? Predictors of Portfolio Allocation in Parliamentary Democracies’. European Journal of Political Research, 50(4): 441–78. Bäck, Hanna, Marc Debus, and Wolfgang C. Müller (2016). ‘Intra-party Diversity and Ministerial Selection in Coalition Governments’. Public Choice, 166(3–4): 355–78. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Anna Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010’. Party Politics, 21(1): 143–52. Baumann, Markus, Marc Debus, and Jochen Müller (2015). ‘Personal Characteristics of MPs and Legislative Behavior in Moral Policymaking’. Legislative Studies Quarterly, 40(2): 179–210. Bergman, Torbjörn (1993). ‘Formation Rules and Minority Governments’. European Journal of Political Research, 23(1): 55–66. Bergman, Torbjörn, Gabriella Illonszki, and Wolfgang C. Müller (eds) (2019). Coalition Governance in Central Eastern Europe. Oxford: Oxford University Press. Best, Volker (2015). Koalitionssignale bei Landtagswahlen. Baden-Baden: Nomos. Bowler, Shaun, Thomas Bräuninger, Marc Debus, and Indridi Indridason. ‘Let’s just agree to disagree: Dispute resolution mechanisms in coalition agreements’. The Journal of Politics, 78(4): 1264–1278. Bräuninger, Thomas, Marc Debus, Jochen Müller, and Christian Stecker (2019). ‘Party Competition and Government Formation in Germany: Business as Usual or New Patterns?’. German Politics, 28(1): 80–100. Bräuninger, Thomas, Marc Debus, Jochen Müller, and Christian Stecker (2020). Parteienwettbewerb in den deutschen Bundesländern. 2nd edition. Wiesbaden: Springer. Budge, Ian, and Hans Keman (1990). Parties and Democracy: Coalition Formation and Government Functioning in Twenty States. Oxford: Oxford University Press. Carroll, Royce, and Gary W. Cox (2012). ‘Shadowing Ministers: Monitoring Partners in Coalition Governments’. Comparative Political Studies, 45(2): 220–36. Chiru, Mihail (2015). ‘Early Marriages Last Longer: Pre-electoral Coalitions and Government Survival in Europe’. Government and Opposition, 50(2): 165–88. Clark, William, Matt Golder, and Sona N. Golder (2013). Principles of Comparative Politics. Washington: CQ Press. Däubler, Thomas, and Marc Debus (2009). ‘Government Formation and Policy Formulation in the German States’. Regional and Federal Studies, 19(1): 73–95.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

278

  .

Debus, Marc (2007). Pre-electoral Alliances, Coalition Rejections, and Multiparty Governments. Baden-Baden: Nomos. Debus, Marc (2008). ‘Office and Policy Payoffs in Coalition Governments’. Party Politics, 14(5): 515–38. Debus, Marc (2009). ‘Pre-electoral Commitments and Government Formation’. Public Choice, 138(1–2): 45–64. Debus, Marc, and Martin Gross (2016). ‘Coalition Formation at the Local Level: Institutional Constraints, Party Policy Conflict, and Office-Seeking Political Parties’. Party Politics, 22(6): 835–46. Decker, Frank (2009). ‘Koalitionsaussagen der Parteien vor Wahlen. Eine Forschungsskizze im Kontext des deutschen Regierungssystems’. Zeitschrift für Parlamentsfragen, 40(2): 431–53. Decker, Frank (2013). ‘Koalitionssignale – ein von der Koalitionstheorie zu Unrecht vernachlässigter Faktor?’. In Frank Decker and Eckhard Jesse (eds), Die deutsche Koalitionsdemokratie vor der Bundestagswahl 2013. Baden-Baden: Nomos, 75–96. Decker, Frank, and Eckhard Jesse (2013). ‘Koalitionslandschaft im Wandel? Eine Einführung’. In Frank Decker and Eckhard Jesse (eds), Die deutsche Koalitionsdemokratie vor der Bundestagswahl 2013. Baden-Baden: Nomos, 9–34. Diermeier, Daniel, and Peter van Roozendaal (1998). ‘The Duration of Cabinet Formation Processes in Western Multi-party Democracies’. British Journal of Political Science, 28(4): 609–26. Diermeier, Daniel, Hülya Eraslan, and Antonio Merlo (2002). ‘Coalition Governments and Comparative Constitutional Design’. European Economic Review, 46(4–5): 893–907. Downs, William M. (1998). Coalition Government, Subnational Style: Multiparty Politics in Europe’s Regional Parliaments. Columbus: Ohio State University Press. Druckman, James N., and Andrew Roberts (2008). ‘Measuring Portfolio Salience in Eastern European Parliamentary Democracies’. European Journal of Political Research, 47(1): 101–34. Druckman, James N., and Paul V. Warwick (2005). ‘The Missing Piece: Measuring Portfolio Salience in Western European Parliamentary Democracies’. European Journal of Political Research, 44(1): 17–42. Ecker, Alejandro, and Thomas M. Meyer (2020). ‘Coalition Bargaining Duration in Multiparty Democracies’. British Journal of Political Science, 50(1): 261–80. Ecker, Alejandro, Thomas M. Meyer, and Wolfgang C. Müller (2015). ‘The Distribution of Individual Cabinet Positions in Coalition Governments: A Sequential Approach’. European Journal of Political Research, 54(4): 802–18. Euchner, Eva-Maria, and Caroline Preidel (2018). ‘Dropping the Curtain’. Politics and Religion, 11(2): 221–48. Franzmann, Simon T. (2018). ‘Von der EURO-Opposition zur KosmopolitismusOpposition’. In Lisa H. Anders, Henrik Scheller, and Thomas Tuntschew (eds),

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



279

Parteien und die Politisierung der Europäischen Union. Wiesbaden: Springer VS, 365–402. Ganghof, Steffen, and Christian Stecker (2016). ‘Investiture Rules in Germany’. In Bjorn Erik Rasch, Shane Martin, and Jose Antonio Cheibub (eds), Parliaments and Government Formation: Unpacking Investiture Rules. Oxford: Oxford University Press, 67–85. Golder, Sona N. (2006). ‘Pre-electoral Coalition Formation in Parliamentary Democracies’. British Journal of Political Science, 36(2): 193–212. Golder, Sona N. (2010). ‘Bargaining Delays in the Government Formation Process’. Comparative Political Studies, 43(1): 3–32. Holtmann, Everhard (2009). ‘Die politische Vorgeschichte der vorgezogenen Bundestagswahl’. In Oskar W. Gabriel, Bernhard Weßels, and Jürgen W. Falter (eds), Wahlen und Wähler. Wiesbaden: VS Verlag für Sozialwissenschaften, 15–39. Hooghe, Liesbet, Gary Marks, Arjan H. Schakel, Sandra C. Osterkatz, Sara Niedzwiecki, and Sarah Shair-Rosenfield (2016). Measuring Regional Authority. Volume I: A Postfunctionalist Theory of Governance. Oxford: Oxford University Press. Indridason, Indridi H., and Gunnar H. Kristinsson (2013). ‘Making Words Count: Coalition Agreements and Cabinet Management’. European Journal of Political Research, 52(6): 822–46. Jun, Uwe (2007). ‘Parteiensystem und Koalitionskonstellationen vor und nach der Bundestagswahl 2005’. In Frank Brettschneider, Oskar Niedermayer, and Bernhard Weßels (eds), Die Bundestagswahl 2005. Analysen des Wahlkampfes und der Wahlergebnisse. Wiesbaden: VS Verlag für Sozialwissenschaften, 491–515. Kortmann, Matthias, and Christian Stecker (2019). ‘Party Competition and Immigration and Integration Policies: A Comparative Analysis’. Comparative European Politics, 17(1): 72–91. Kriesi, Hanspeter, Edgar Grande, Martin Dolezal, Marc Helbling, Dominic Höglinger, Sven Hutter, and Bruno Wüest (2012). Political Conflict in Western Europe. New York: Cambridge University Press. Kropp, Sabine (2010). ‘The Ubiquity and Strategic Complexity of Grand Coalition in the German Federal System’. German Politics, 19(3–4): 286–311. Laver, Michael, and Norman Schofield (1990). Multiparty Government: The Politics of Coalition in Europe. Ann Arbor: University of Michigan Press. Linhart, Eric (2009). ‘A Rational Calculus of Voting Considering Coalition Signals: The German Bundestag Elections 2005 as an Example’. World Political Science Review, 5(1): 1–28. Lipsmeyer, Christine S., and Heather Nicole Pierce (2011). ‘The Eyes That Bind: Junior Ministers as Oversight Mechanisms in Coalition Governments’. Journal of Politics, 73(4): 1152–64. Mershon, Carol (2002). The Costs of Coalition. Stanford: Stanford University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

280

  .

Müller, Jochen (2009). ‘The Impact of the Socio-economic Context on the Länder Parties’ Policy Positions’. German Politics, 18(3): 365–84. Müller, Jochen (2018). ‘German Regional Elections: Patterns of Second-Order Voting’. Regional and Federal Studies, 28(3): 301–24. Müller, Wolfgang C., and Kaare Strøm (eds) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Müller-Rommel, Ferdinand (1994). ‘The Role of German Ministers in Cabinet Decision Making’. In Michael Laver and Kenneth Shepsle (eds), Cabinet Ministers and Parliamentary Government. Cambridge: Cambridge University Press, 150–68. Niclauß, Karlheinz (2004). Kanzlerdemokratie. Paderborn: Schöningh. Niedermayer, Oskar (2007). ‘Der Wahlkampf zur Bundestagswahl 2005: Parteistrategien und Kampagnenverlauf ’. In Frank Brettschneider, Oskar Niedermayer, and Bernhard Weßels (eds), Die Bundestagswahl 2005. Analysen des Wahlkampfes und der Wahlergebnisse. Wiesbaden: VS Verlag für Sozialwissenschaften, 21–42. Niedermayer, Oskar (2011). ‘Das deutsche Parteiensystem nach der Bundestagswahl 2009’. In Oskar Niedermayer (ed.), Die Parteien nach der Bundestagswahl 2009. Wiesbaden: VS Verlag für Sozialwissenschaften, 7–35. O’Leary, Brendan, Bernard Grofman, and Jørgen Elklit (2005). ‘Divisor Methods for Sequential Portfolio Allocation in Multi-party Executive Bodies: Evidence from Northern Ireland and Denmark’. American Journal of Political Science, 49(1): 198–211. Pappi, Franz U., Alexander Herzog, and Ralf Schmitt (2006). ‘Koalitionssignale und die Kombination von Erst- und Zweitstimme bei den Bundestagswahlen 1953 bis 2005’. Zeitschrift für Parlamentsfragen, 37(3): 493–513. Pappi, Franz U., Axel Becker, and Alexander Herzog (2005). ‘Regierungsbildung in Mehrebenensystemen: Zur Erklärung der Koalitionsbildung in den deutschen Bundesländern’. Politische Vierteljahresschrift, 46(3): 432–58. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of AntiElitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Proksch, Sven-Oliver, and Jonathan B. Slapin (2006). ‘Institutions and Coalition Formation: The German Election of 2005’. West European Politics, 29(3): 540–59. Raabe, Johannes, and Eric Linhart (2015). ‘Does Substance Matter? A Model of Qualitative Portfolio Allocation and Application to German State Governments between 1990 and 2010’. Party Politics, 21(3): 481–92. Reutter, Werner (2006). ‘Yet Another Coup d’État in Germany? Schröder’s Vote of Confidence and Parliamentary Government in Germany’. German Politics, 15(3): 302–17. Rixen, Thomas (2014). ‘Ambitious Goals, Deficient Output: Tax and Fiscal Policies of the Conservative–Liberal Government, 2009–13’. German Politics, 23(4): 304–21.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi



281

Saalfeld, Thomas (2000). ‘Members of Parliament and Governments in Western Europe: Agency Relations and Problems of Oversight’. European Journal of Political Research, 37(3): 353–76. Saalfeld, Thomas, and Reimut Zohlnhöfer (2015). ‘Von der Wunschkoalition zur Krisengemeinschaft?’. In Reimut Zohlnhöfer and Thomas Saalfeld (eds), Politik im Schatten der Krise. Wiesbaden: Springer VS, 9–24. Schubert, Thomas (2014). ‘Politikfloskeln oder Bündnissignale? Koalitionsaussagen zwischen Wahlkampfstrategie und Bündnispolitik’. In Eckhard Jesse and Roland Sturm (eds), Bilanz der Bundestagswahl—Voraussetzungen, Ergebnisse, Folgen. Baden-Baden: Nomos, 75–94. Siaroff, Alan (2003). ‘Two-and-a-Half-Party Systems and the Comparative Role of the “Half” ’. Party Politics, 9(3): 267–90. Sieberer, Ulrich, and Daniel Höhmann (2017). ‘Shadow Chairs as Monitoring Device? A Comparative Analysis of Committee Chair Powers in Western European Parliaments’. Journal of Legislative Studies, 23(3): 301–25. Spier, Tim, and Ulrich von Alemann (2015). ‘In ruhigerem Fahrwasser, aber ohne Land in Sicht? Die SPD nach der Bundestagswahl 2013’. In Oskar Niedermayer (ed.), Die Parteien nach der Bundestagswahl 2013. Wiesbaden: Springer VS, 49–69. Spier, Tim, Felix Butzlaff, Matthias Micus, and Franz Walter (2007). Die Linkspartei. Zeitgemäße Idee oder Bündnis ohne Zukunft? Wiesbaden: VS Verlag für Sozialwissenschaften. Stecker, Christian (2016). ‘The Effects of Federalism Reform on the Legislative Process in Germany’. Regional and Federal Studies, 26(5): 604–24. Strøm, Kaare, and Wolfgang C. Müller (1999). ‘Political Parties and Hard Choices’. In Wolfgang C. Müller and Kaare Strøm (eds), Policy, Office, or Votes? How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press, 1–35. Strøm, Kaare, Ian Budge, and Michael J. Laver (1994). ‘Constraints on Cabinet Formation in Parliamentary Democracies’. American Journal of Political Science, 38(2): 303–35. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2008). Cabinets and Coalition Bargaining: The Democractic Life Cycle in Western Europe. Oxford: Oxford University Press. Sturm, Roland (2014). ‘Die Regierungsbildung nach der Bundestagswahl 2013: lagerübergreifend und langwierig’. Zeitschrift für Parlamentsfragen, 45(1): 207–30. Thomson, Robert, Terry Royed, Elin Naurin, Joaquín Artés, Rory Costello, Laurenz Ennser-Jedenastik, Mark Ferguson, Petia Kostadinova, Catherine Moury, François Pétry, and Katrin Praprotnik (2017). ‘The Fulfillment of Parties’ Election Pledges: A Comparative Study on the Impact of Power Sharing’. American Journal of Political Science, 61(3): 527–42.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

282

  .

van Spanje, Joost, and Nan Dirk de Graaf (2018). ‘How Established Parties Reduce Other Parties’ Electoral Support: The Strategy of Parroting the Pariah’. West European Politics, 41(1): 1–27. Van Spanje, Joost, and Wouter Van Der Brug (2007). ‘The Party as Pariah: The Exclusion of Anti-immigration Parties and Its Effect on Their Ideological Positions’. West European Politics, 30(5): 1022–40. Van Spanje, Joost, and Wouter Van Der Brug (2009). ‘Being Intolerant of the Intolerant. The Exclusion of Western European Anti-immigration Parties and Its Consequences for Party Choice’. Acta Politica, 44(4): 353–84. Von Alemann, Ulrich, and Tim Spier (2008). ‘Doppelter Einsatz, halber Sieg? Die SPD und die Bundestagswahl 2005’. In Oskar Niedermayer (ed.), Die Parteien nach der Bundestagswahl 2005. Wiesbaden: Springer, 37–65. Vorländer, Hans (2011). ‘Als Phönix zurück in die Asche? Die FDP nach der Bundestagswahl 2009’. In Oskar Niedermayer (ed.), Die Parteien nach der Bundestagswahl 2009. Wiesbaden: VS Verlag für Sozialwissenschaften, 107–29. Ware, Alan (ed.) (1996). Political Parties and Party Systems. Oxford: Oxford University Press. Warwick, Paul V., and James N. Druckman (2006). ‘The Portfolio Allocation Paradox: An Investigation into the Nature of a Very Strong But Puzzling Relationship’. European Journal of Political Research, 45(4): 635–65.

Appendix. List of political parties Abbreviation KPD Linke

GR B90 SPD FDP CDU/CSU

DSU Z GB-BHE

Name Communist Party of Germany (Kommunistische Partei Deutschlands) The Left (Die Linke), 2007– The Left Party (Die Linkspartei), 2005–2007 Party of Democratic Socialism (Partei des Demokratischen Sozialismus), 1990–2005 Alliance 90/The Greens (Bündnis 90/Die Grünen), 1993– The Greens (Die Grünen), 1979–1993 Alliance 90 (Bündnis 90) Social Democratic Party of Germany (Sozialdemokratische Parti Deutschlands) Free Democratic Party (Freie Demokratische Partei) Christian Democratic Union of Germany & Christian Social Union in Bavaria (Christlich Demokratische Union Deutschlands & ChristlichSoziale Union in Bayern) German Social Union (Deutsche Soziale Union) German Centre Party (Deutsche Zentrumspartei) All-German Bloc/League of Expellees and Deprived of Rights (Gesamtdeutscher Block/Bund der Heimatvertriebenen und Entrechteten)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 28/6/2021, SPi

 DA-FVP DP WAV BP NR AfD

283

Free People’s Party (Freie Volkspartei) German Party (Deutsche Partei) Economic Reconstruction League (Wirtschaftliche Aufbau-Vereinigung) Bavarian Party (Bayernpartei) National Right (Nationale Rechte) Alternative for Germany (Alternative für Deutschland)

Note: Party names are given in English, followed by the party name in German in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order and with the period during which a specific party was coded or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 9 Greece: From Coalitions as a ‘State of Exception’ to the New Normal? Myrto Tsakatika

During the first decades since Greece became democratic in 1974, coalition governments were short-lived exceptions. Instead, the dominant pattern was single-party majority governments formed by either centre-left Panhellenic Socialist Movement (PASOK) or centre-right New Democracy (ND). A confrontational and bipartisan political culture reinforced overlapping historical cleavages that divided parties and polarized voters into two opposing camps (Moschonas 1994). Political elites have considered coalition governments less effective and desirable than single-party governments, paying only lip service to the need for more cooperation among parties.¹ The negative perceptions of coalition government in post-1974 Greece make sense if examined against the background of the successive, short-lived, and unstable coalitions of the Greek Civil War period (1947–1949) and those of the late 1960s that functioned as a preamble for the country’s slide towards dictatorship (1967–1974) (Linardatos 1977; Nicolakopoulos 2000). In this context, in the few instances when coalition government has occurred, it has been justified in terms of an exceptional situation, usually related to a national emergency, an economic crisis, or a series of political scandals, not to a ‘normal’ state of affairs. Public discourse on coalition government continues to reflect the ‘state of exception’. However, a greater acceptance of the practice of power-sharing developed in 2012, when in the aftermath of the economic crisis the Greek party system underwent radical transformation. The main characteristics of the change have been fragmentation and the emergence of new political actors. Coalition governments have since been formed regularly, experience with the practice has been accumulating in public life, and coalitions could be more frequent in the future. Drawing on the limited and (mostly) fairly recent experience of Greece with coalitions, the chapter will engage with the literature on the formation,

¹ The chapter is based on 14 interviews with individuals who were well placed to follow the inner working of Greek coalition cabinets. To protect these individuals, they are not named as individual sources. The author would like to acknowledge Georgios Katsambekis’ expert research assistance. Myrto Tsakatika, Greece: From Coalitions as a ‘State of Exception’ to the New Normal? In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0009

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



285

governance, and termination of coalition governments, which mainly refers to the West European experience but may also be helpful in analysing the Greek case. Focus will be on the life cycle of Greek coalition governments during the democratic period and particularly after 2012 when the economic crisis and domestic party system change began to impact the establishment of coalition government and its dynamics. A brief reflection will be offered on the relationship between the different phases of the Greek coalition ‘life cycle’.

The institutional setting Greece’s transition to democracy and the abolition of the monarchy by referendum in 1974 heralded an era of unprecedented constitutional and political stability in the country. The 1975 constitution (amended 1986, 2001, and 2008) established a unicameral parliamentary system of government with an indirectly elected president as Head of State. Initially the president was endowed with significant formal powers termed ‘superpowers’, such as the dissolution of parliament, the appointment of a head of government of his own choice in the absence of a clear parliamentary majority, and the right to call a referendum. The president also had what was referred to as ‘royal powers’. This included approving laws, granting amnesty, and convening the ‘Democracy Council’, following a semipresidential model inspired by France (Clogg 1987). The first amendment of the constitution in 1986 transferred the president’s ‘super powers’ to the prime minister (PM) and abolished his ‘royal powers’. The role of the president was now rendered largely symbolic and ceremonial and the Greek political system took a clear parliamentary turn. The Greek political system applies a principle of positive parliamentarism, whereby the PM, after forming his government and reading the ‘programmatic declarations’ (i.e. the main parameters of his government’s priorities and plans for the four years of their term in office) in parliament, requests a vote of investiture. When the investiture vote gives him a parliamentary majority of at least 151 of the 300 seats, the government’s term of office commences. There are three ways in which a Greek government can be terminated before the end of its four-year term, according to the Greek constitution. First, the government may lose their parliamentary majority after a vote of no-confidence (which in the Greek case is not a constructive vote of no confidence); second, the parliament may fail to elect a president of the Republic (requires a supermajority). In such cases the government resigns, parliament is dissolved, and new elections are called. The third way is when the PM decides to offer the government’s resignation and call new elections due to reasons of ‘national importance’. The latter is often used strategically by PMs to choose the timing of elections in ways that are most suitable for them and their parties (see Table 9.6).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

286



The party system and the actors The Greek party system can be neatly divided into two distinct phases in the post1974 democratic period. The first spans 1974 to 2012, while the second phase that began in 2012 is ongoing. The events that triggered party system change in 2012 are clearly associated with the social and economic repercussions of the 2008 global financial crisis whose impact on Greece was much deeper than on any of its South European neighbours. Fragmentation, extreme levels of electoral volatility, the rise of new parties, and the electoral success of populism on both the left and right are key aspects of the recent party system transformation. The party system in the early democratic period was marked by the establishment of a stable two-party system and the persistence of a weaker ‘third pole’ of the left (Pappas 2003; Lyrintzis 2005) following the legalization of the Communist Party in 1974. Two main parties, the centre-left PASOK and the centre-right ND formed single-party governments, often relying on stable parliamentary majorities of more than 160 of the 300 parliamentary seats and alternating in power from 1977 until 2012. Minor exceptions were the short-lived coalition governments of 1989–1990. The two major parties taken together gathered well above 80 per cent of the national vote throughout this period (Teperoglou and Tsatsanis 2014: 224). Meanwhile, the left’s electoral strength remained on aggregate at around 10–12 per cent, split between the Marxist, Eurosceptic Communist Party and a smaller Eurocommunist party of the moderate left, Synaspismos (SYN) (Kalyvas and Marantzidis 2003), both of which secured parliamentary representation. Only the elections during the 1989–1990 period were conducted under a relatively strict proportional representation system. All other elections post-1974 have been carried out under different versions of a system of ‘reinforced proportionality’, a system that was meant to have majoritarian effects. As it discouraged coalition government post-1974, the electoral system played a significant role in the consolidation of a two-party system with single-party governments. Table 9.1a provides information on the Greek cabinets since 1977. Voting takes place in 58 constituencies of unequal size (48 multiseat, 8 singleseat). Party lists are issued in each constituency and voters may select their member of parliament (MP) by marking their name on the ballot. Proportional principles apply in the translation of votes to seats with two important caveats: there is a three per cent national threshold for a party’s entry to parliament and a 50-seat bonus out of a total of 300 seats for the party winning a plurality of votes. Note that it is only parties and not electoral alliances or pre-electoral coalitions that are granted the 50-seat bonus. The effects of these provisions are to encourage government stability by privileging the survival of few, unified parties, capable of exercising single-party government and to discourage coalition government in general. They also provide few incentives for major parties to seek and smaller parties to agree to become junior coalition partners. As seen in Table 9.1b, there have only been three

1977-11-28 1980-05-10 1981-10-21 1985-06-05 1989-07-02 1989-10-12 1989-11-23

1990-04-11 1993-10-13 1996-01-22 1996-09-25 2000-04-13 2004-03-10 2007-09-19 2009-10-06 2011-11-16

2012-02-11 2012-05-16 2012-06-21

Karamanlis Rallis Papandreou I Papandreou II Tzannetakisc Grivasabc Zolotasbc

Mitsotakis Papandreou III Simitis I Simitis II Simitis III Karamanlis I Karamanlis II Papandreou Papademos Ibc

Papademos IIbc Pikramenosabc Samaras I

Samaras II Tsipras I

1 2 3 4 5 6 7

8 9 10 11 12 13 14 15 16

17 18 19

20 21

2013-06-22 2015-01-26

Date in

Cabinet Cabinet number

Table 9.1a Greek cabinets since 1977

2015-01-25

2012-05-06 2012-06-17

1996-09-22 2000-04-09 2004-03-07 2007-09-16 2009-10-04

1990-04-08 1993-10-10

1989-11-05

1981-10-18 1985-06-02 1989-06-18

1977-11-20

Election date

ND ND PASOK PASOK ND, SYN NN ND, PASOK, SYN ND PASOK PASOK PASOK PASOK ND ND PASOK PASOK, ND, LAOS PASOK, ND NN ND, PASOK, DIMAR ND, PASOK sur mwc

sur non sur

min maj maj maj maj maj maj maj sur

maj maj maj maj mwc non sur

52.3 54

64.3 0 59.7

50.3 56.7 55.3 54 52.7 55 50.7 53.3 84

57 57.7 57.3 53.7 57.7 0 99

300 300

300 300 300

300 300 300 300 300 300 300 300 300

300 300 300 300 300 300 300

7 7

5 7 7

6 4 4 5 4 4 5 5 5

7 7 3 4 5 5 5

3.94 3.09

3.5 4.83 3.76

2.37 2.17 2.25 2.36 2.21 2.19 2.62 2.59 2.87

2.35 2.31 2.09 2.15 2.4 2.4 2.32

Party Type Cabinet Number of Number of ENP, parties in parliament composition of strength seats in of cabinet cabinet in seats parliament parliament (%)

PASOK SYRIZA

PASOK

PASOK

ND PASOK PASOK PASOK PASOK ND ND PASOK PASOK

PASOK

ND ND PASOK PASOK PASOK

Continued

DIANA

Formal Median support party in first policy parties dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2019-01-16

2015-09-20

Election date

SYRIZA, ANEL SYRIZA

SYRIZA, ANEL NN

min

mwc

non

48.3

51.7

0

300

300

300

7

8

8

3.21

3.24

3.93

Party Type Cabinet Number of Number of ENP, parties in parliament composition of strength seats in of cabinet cabinet in seats parliament parliament (%)

SYRIZA

SYRIZA

Formal Median support party in first policy parties dimension

Notes: For a list of parties, consult the chapter appendix. a = Technocrat minister majority; b = Technocrat PM; c = Limited policy remit Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1977–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015). The first policy dimension is economic left–right. Cabinet types: min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition; non = non-partisan.

2019-01-16

Tsipras III

24

23

Thanou2015-08-28 Christophilouabc Tsipras II 2015-09-23

Date in

22

Cabinet Cabinet number

Table 9.1a Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



289

Table 9.1b Electoral alliances and pre-electoral coalitions in Greece, 1977–2018 Election date

Constituent parties

Type

Types of pre-electoral commitment

1989-06-18 1989-11-05 1990-04-08 2015-09-20

KKE, EAR KKE, EAR KKE, EAR SYRIZA, ANEL

EA, PEC EA, PEC EA, PEC PEC

Other

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other

electoral alliances and four cases of a pre-electoral coalition. Three of these precoalitions occurred during the ‘proportional’ period, 1989–1990. In terms of electoral party competition, the first democratic period was dominated by the continued salience of the deep left–right cleavage that was engrained from the time of the Greek Civil War (1947–1949) when communist guerrilla forces were defeated by the Greek army. In its aftermath, in the 1950s there was a period of communist repression by strong single-party governments of the right. Post-1974 it was PASOK rather than the radical left that succeeded in appealing to this deep division in Greek society, consolidating its position as the ‘anti-right’ pole of the party system. Without overshadowing the left–right division that remained central, the 1990s and 2000s saw the emergence of GAL–TAN as a secondary cleavage, which largely overlapped with left–right. Popular Orthodox Rally (LAOS), a new party of the nationalist and populist right that emerged to the right flank of ND building its appeal on national identity, Orthodox Christianity, traditional values, and a soft version of Euroscepticism as well as politicizing and opposing immigration, covered the TAN end of the new cleavage. On the other hand, SYN took over the main role of promoting left-liberal agendas such as LGBT rights, gender, the environment, and multiculturalism. Both parties achieved parliamentary representation but did not much exceed the three per cent threshold that the electoral system imposes for entry, or 10–13 of the 300 seats on average. The Communist Party of Greece maintained its electoral niche with 15 seats on average while continuing to stress materialist values rather than addressing the ‘new’ issues. In the late 1980s for PASOK and then again for both major parties later in the 2000s, high levels of corruption and clientelism (Featherstone 1990) haunted the parties. A corruption scandal in the late 1980, involving collusion between private banker and businessman Giorgos Koskotas and high-ranking officials of the second government of PASOK, led to a brief crisis of the party system and the short-lived experiment with proportional representation. An unprecedented coalition government of the two radical left parties and ND was formed under

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

290



ND senior politician Tzannis Tzannetakis, whose main purpose was to ensure that the judicial investigation of leading former PASOK ministers and the former PM Papandreou himself would take place. The stated political aim was to ‘purge’ the political system of PASOK’s scandals and restore public trust. However, the elections that took place after four months did not produce a single-party majority. In the meantime, the country’s economy had taken a significant downturn that brought the country to a state of emergency due to rising public deficit and inflation. Subsequent to the elections, an ‘ecumenical’ coalition government of national unity under the premiership of former Bank of Greece governor Xenofon Zolotas with ND, PASOK, and the left was formed whose limited mandate was to deal with the economic crisis (Verney 1990; Pridham and Verney 1991). This interval of extraordinary and purpose-built coalition governments was an exception that confirmed the two-party system with single-party government rule. However, there were three aspects of this brief spell of coalition government that with hindsight could be considered a general rehearsal for the radical changes to the party system to come in 2012. These were: the decline in public trust in the party system as a whole and in the two major parties in particular, a dire economic situation that called for broader political cooperation and consensus among political parties, and the participation of the radical left in a coalition government that expanded the range of ‘coalitionable’ political parties that could and would take on the responsibility of government. In the aftermath of the 2008 financial crisis, between 2010 and 2015 the Greek government received three multibillion ‘bailout’ loans from the European Union and the IMF in order to avoid bankruptcy as a result of its inability to access international financial markets to service its public debt. These loans were attached to strict conditionality (in Memoranda of Understanding, or MOUs) involving rapid and deep fiscal adjustment, extensive packages of public administration, tax, labour market and pension reform, and a wide-ranging programme of privatization. Greece entered a recession spiral that peaked in 2012, coupled with a steep rise in unemployment levels, wage and pension cuts, and public sector retrenchment. This ‘internal devaluation’ was not without political consequences, including the rise of mass and at times violent anti-austerity protest and the overhaul of the stable two-party system of the first democratic period, which was ushered in by the ‘earthquake elections’ of 2012. This phase witnessed the electoral demise of PASOK, which lost 75 per cent of its vote share and was relegated to minor party status, the downsizing of ND, which in the elections of May 2012 lost about half of its parliamentary seats, and the emergence of the radical left-populist Coalition of the Radical Left (SYRIZA), successor to moderate left-liberal SYN, as a major party with a claim to government power (Tsakatika 2016). The period is marked by high levels of party system fragmentation, the emergence of new parties across the political spectrum, and coalition government.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



291

While in the 1974–2012 period there were only three to five parties represented in the Greek parliament, after 2012 this number rose to seven to eight. This is reflected in the effective number of parties (ENP) in terms of seats, which jumped from 2.6 in 2009 to 3.9 in 2013 (Table 9.1a). Among the new parties that emerged after 2012, the one that stood out was the xenophobic, nativist, and hard Eurosceptic party of the extreme right, Golden Dawn, which achieved a surprising seven per cent of the vote and 21 seats in the Greek parliament in the May 2012 election (Ellinas 2013). Another newcomer was the Independent Greeks (ANEL), a right-wing split from the ND with a populist, soft Eurosceptic, and socially conservative agenda. Despite their small size they played a significant role in the new party system as they formed two government coalitions with SYRIZA between January 2015 and January 2019. After participating in the Papademos I government (PASOK, ND, LAOS) the LAOS party failed to achieve parliamentary representation, with its voters being absorbed largely by the other numerous contenders on the right and far right. Four parties also emerged in the centre and left over this period achieving parliamentary representation. DIMAR is a party of the democratic left that emerged as a result of a 2010 split by SYRIZA’s moderate and pro-European wing. After participating as a junior partner in the Samaras I government (ND, PASOK, DIMAR) the party lost its parliamentary representation and eventually merged with PASOK. In 2014, a new party of the liberal centre with a strong anti-corruption message, To Potami, was founded by maverick journalist Stavros Theodorakis incorporating politicians from liberal groups, DIMAR, and former PASOK modernizers, achieving parliamentary representation in the 2015 elections. Two more parties also made breakthroughs in 2015— Centre Union, a small party of the traditional centre with a long presence in Greek politics but unsuccessful in achieving parliamentary representation until 2015, and LAE, a party started by 25 SYRIZA MPs who opposed SYRIZA’s acceptance of the third MoU in the final months before the September 2015 elections. Coalition governments became the norm for a period, including two coalitions under ‘technocrat’ former Bank of Greece governor Loukas Papademos (Pastorella 2014), two coalitions at whose core were the former archenemies PASOK and ND, and two ‘populist’ government coalitions between radical left SYRIZA and the right-wing ANEL (Aslanidis and Rovira Kaltwasser 2016). These coalitions were made possible and justified on the grounds of the emergence of a new cleavage that came to divide the political system particularly after 2011, cutting across the left–right and GAL–TAN cleavages: the so-called ‘Memorandum–anti-Memorandum’ cleavage. The division was initially between parties that had lent their support to the ‘bailout’ loans and associated conditionality-related measures (PASOK, ND, LAOS, DIMAR) from those who claimed to be in favour of renegotiating the terms of these agreements and/or rupture with the European Union (SYRIZA, ANEL, KKE, GD) (Dinas and Rori 2013). Both the Papademos and Samaras governments were coalitions supported

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

292



by parties that accepted the MoU terms as necessary, potentially beneficial, and/or unavoidable and whose aim was to pull the country out of the difficult economic situation it found itself in. The SYRIZA–ANEL governments have also been explained in terms of the two parties being on the ‘anti-Memorandum’ side of the new division (Pappas 2014) despite being diametrically opposed in terms of their placement on the left–right and GAL–TAN axes (Andreadis 2015). However, the new ‘cleavage’ may well have a short life span, especially since the completion of the programme associated with the third (and final) MOU in August 2018. It has been argued that in the post-2012 period Greece is witnessing the emergence of a new bipolar party system with SYRIZA and ND as major parties that are nonetheless unable to form single-party government (Tsirbas 2015). In this new party system minor parties with parliamentary representation emerge as significant players in the coalition game (Kiapidou 2015), with the exception of Golden Dawn, which is not ‘coalitionable’, and the KKE, which refuses to participate in any government coalition. Yet, in 2019, ND succeeded in forming a single-party government. It remains to be seen whether coalition government will re-emerge or if the Greek party system will again revert to a two-party system involving alternation between single-party governments of SYRIZA and ND with minor parties returning to the margins. Party system fragmentation and the fact that no major party was able to form a single-party government between 2012 and 2019, despite the 50-seat bonus that ‘reinforced PR’ provides to the party that wins a plurality of the vote, weigh in favour of the prospect of more coalition governments. Recent changes to the electoral system also increase the odds of more coalition governments in the future, since the Tsipras II coalition government reintroduced proportional representation, meant to come into force in the national elections planned for 2023.

Government formation The bargaining process The Greek constitution explicitly renders the president of the Republic the custodian of government formation, assigning them to the role of informateur. Their contribution is by and large limited to ensuring the implementation of the very precise rules written into the Greek constitution. These rules describe in a fairly detailed way not only the temporal limits of bargaining duration and the maximum number of bargaining rounds but also the actors among which bargaining is to take place in each bargaining round. According to Article 37 of the constitution, immediately after the official election results are issued the president must call the leader of the largest (in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



293

terms of seats) party and give them the mandate to form a government. If the first party commands an absolute parliamentary majority (151/300 seats) its leader will be appointed PM by the president and should present themselves along with their chosen ministers before the parliament within 15 days to request a vote of confidence. Should the largest party not command an absolute majority but only a plurality of seats, its leader will be offered a mandate to explore possibilities of cooperation with other parties for the formation of a coalition government. Should the exploratory mandate fail, the leader of the largest party is obliged to return the mandate to the president within a maximum of three days. The president will then call upon the leader of the second largest party to attempt to form a collation government and if they also fail to do so in three days the president will offer the exploratory mandate to the third largest party. If after three days the third exploratory mandate fails, the president invites the leaders of all parliamentary parties and explores the possibility of a coalition government from the existing parliament or a minority government that will receive external support and thus achieve a parliamentary majority (Tsatsos 1993). If the fourth round of negotiations does not lead to fruition then the president tries to achieve the formation of a transitional government composed of all parliamentary parties with the aim of conducting new elections. Should this option also fail, the president offers the mandate to one of the presidents of the country’s three higher courts (the Supreme Administrative Court, the Supreme Civil and Criminal Court, the Court of Auditors) to form a caretaker cabinet ‘as widely accepted as possible to carry out elections’ and dissolves parliament.² The constitutional framework hence necessarily limits the number of bargaining rounds to four and the duration of each bargaining round to three days. It also effectively renders the leaders of parliamentary parties (their bargaining power increasing in line with the number of seats their parties occupy in parliament) key players in the bargaining process for coalition formation. In so doing, it minimizes uncertainty as political actors bargain within very precise institutional constraints. As can be surmised from Table 9.2, bargaining duration for the formation of a coalition government in Greece is short. Bargaining for the formation of the eight government coalitions of the democratic period has lasted on average 4.9 days. Two of these coalitions (Tzanetakis and Zolotas) have been agreed in the fourth round under the aegis of the president of the Republic after three failed bargaining ² Greece has a long tradition of caretaker governments. Caretaker governments are appointed to conduct elections at the same time as parliament is dissolved. Where there is no especially appointed caretaker government, the government of the day remains in the caretaker role after elections are called and the parliament is dissolved. In such occasions a handful of non-partisan ministers (senior academics, prominent personalities, or supreme court judges) are called in to replace the ministers in areas directly related to the conduct of elections, that is Ministry of Public Administration, Ministry of the Interior, etc. This practice is due to long-term lack of trust for the party/parties in government to play by the rules and conduct elections fairly (Pagoulatos, interview, cited in Pastorella 2014).

Year in

1977 1980 1981 1985 1989

1989

1989

1990 1993 1996 1996 2000 2004 2007 2009 2011 2012

Cabinet

Karamanlis Rallis Papandreou I Papandreou II Tzannetakis

Grivas

Zolotas

Mitsotakis Papandreou III Simitis I Simitis II Simitis III Karamanlis I Karamanlis II Papandreou Papademos I Papademos II

0 0 0 0 0 0 0 0 0 0

3

4

0 0 0 0 3

Number of inconclusive bargaining rounds

ND ND PASOK PASOK ND, SYN (1) SYN, ND, PASOK (2) PASOK, SYN (3) ND, SYN NN (1) ND, PASOK, SYN (2) SYN (3) PASOK, SYN (4) ND ND, PASOK, SYN (1) SYN, ND, PASOK, OIKEN (2) PASOK, ND, SYN (3) ND, PASOK, SYN, OIKEN ND PASOK PASOK PASOK PASOK ND ND PASOK PASOK, ND, LAOS PASOK, ND

Parties involved in the previous bargaining rounds

Table 9.2 Cabinet formation in Greece, 1977–2018

3 3 6 3 4 3 3 2 5 0

3 3

1 1 2 1 5 3

8 0 3 3 3 3 3 3

Bargaining duration of individual formation attempt (in days)

3 3 6 3 4 3 3 2 5 0

18

5

8 0 3 3 14

Number of days required in government formation

3 3 6 3 4 3 3 2 5 0

14

5

8 0 3 3 12

Total bargaining duration

152 170 166 161 157 165 160 155 255

292

171 180 172 161 174

Pro

0 0 3 0 0 0 0 0 0

2

0 1 13 0 1

Abstention

146 129 123 134 141 135 140 143 38

1

126 115 113 138 124

Contra

Result of investiture vote OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2012

2012 2013 2015 2015

2015 2019

Pikramenos

Samaras I Samaras II Tsipras I ThanouChristophilou

Tsipras II Tsipras III

0 0

0 0 0 4

4

(1) SYRIZA, ND, PASOK, TP, ANEL, LAE, KKE, GD (2) LAE, TP, ND, PASOK (3) ND, TP, PASOK, LAE SYRIZA SYRIZA, ANEL SYRIZA

NN (1) ND, SYRIZA, PASOK, DIMAR, ANEL, KKE, GD (2) PASOK, ND, SYRIZA (3) SYRIZA, ND, PASOK, DIMAR, ANEL (4) ND, SYRIZA, PASOK, DIMAR ND, PASOK, DIMAR ND, PASOK SYRIZA, ANEL NN

1 3 0

3

3

1

4 0 1 0

1

2

2

0 4

3 0

4 0 1 8

10

3 0

4 0 1 8

9

0 0

0

162

155 151

0

179

144 148

137

121

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

296



rounds. The remaining six (Papademos, Samaras, and Tsipras I and II) have been agreed during the first and only bargaining round. The three caretaker governments of the post-1974 period (Grivas, Pikrammenos, and ThanouChristophilou), all led by the presidents of the higher courts, were formed five, nine, and eight days respectively after the four bargaining rounds provided for by the constitution failed to produce a coalition government. All bargaining rounds were initiated either by the leader of one of the three largest political parties or the president of the Republic (Presidency website).

The composition and size of cabinets All eight post-1974 coalition governments in Greece have without exception included the largest party. Until the 2012 party system change the largest party was also the median party in both the left–right and GAL–TAN dimensions. In the multidimensional space that characterizes Greek politics, from 2012 coalitions have included the median party along the new MoU–anti-MOU dimension. Five out of the eight coalition governments have been surplus coalitions that commanded sizeable majorities in the Greek parliament, as could be seen in Table 9.1a. The surplus coalition governments of Zolotas, Papademos I and II, and Samaras I and II were without exception presented as ‘ecumenical’ or ‘governments of national unity’. They were formed with the aim of addressing a national crisis. In contrast, minimal winning coalitions have been rarer and are so far limited to three cases: the Tzannetakis government and the two more recent Tsipras governments. They also have been presented as governments of ‘national salvation’. An interesting pattern emerges concerning the composition of these coalitions. Surplus coalitions were connected coalitions, always involving the participation of PASOK and ND and in some cases also including a third, minor partner. The Zolotas government included ND, PASOK, and SYN (a coalition of the left), collaborating under a ‘technocrat’ PM, a well-respected economist and former governor of the Bank of Greece. Likewise, the first Papademos coalition brought together PASOK, ND, and LAOS (a party of the populist right) under the leadership of another former governor of the Bank of Greece and ECB vicepresident. The first Samaras government included ND, PASOK, and DIMAR (a party of the moderate left). Papademos II and Samaras II were coalitions of PASOK and ND that decided to continue their collaboration after the departure of LAOS and of DIMAR. All these connected surplus coalition governments had the task of addressing a severe economic crisis. In contrast, all three minimal winning coalitions have not been connected coalitions, but rather, they have been governments of the radical left with the right (either the centre-right or the nationalist right). The Tzannetakis

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



297

government was the product of collaboration between SYN, a coalition of the Communist Party of Greece and its Eurocommunist ally (EAR), and ND, the main party of the centre-right. Likewise, the two more recent Tsipras governments have been collaborations between SYRIZA and the Independent Greeks (ANEL), a party of the nationalist right. All three of these governments have presented themselves as coalitions against corruption, which they associate with the centre ground of politics. In 1989 the Tzannetakis government promised to ‘purge’ the political system of PASOK’s wrongdoings, while in 2015 the Tsipras government pronounced that its components, the ‘new parties’, had come together to address the root causes of the country’s decline into disrepute by allying against the ‘old parties’ (PASOK, ND, and their allies), that is those parties that had first brought the country to the brink with their corrupt practices before signing the MoU that was detrimental for the survival of its people. Like the Tzannetakis government before it, the Tsipras government also promised to ‘clean up’ the political system from the clientelistic networks, patronage, and corrupt practices of the ‘old parties’.

The allocation of ministerial portfolios Governments in Greece tend to be large, comprising of 20 ministries and 24 ministers on average. Coalition governments are larger than average with respect to the number of ministers they involve (Table 9.3). This can be attributed to the extensive use of alternate ministers. These are ministers who are members of cabinet with full voting rights but are not heads of their ministries. Portfolio allocation to specific ministers, alternate ministers, and junior ministers is a function that is formally the sole responsibility of the PM whether there is a single-party or a coalition government. In practice, where there are more than one government partners the responsibility falls collectively to the party leaders. According to interview evidence, in those coalition governments where the PM was not the leader of the larger party, mainly the Tzannetakis, Zolotas, and Papademos governments, party leaders first met, negotiated, and reached a consensual position on the person of the PM (Interviews 2, 3, 5, 12). Subsequently, the leaders agreed on the persons of each of the ministers, alternate ministers, and junior ministers by consensus. Each party would have the freedom to propose a candidate of their choice. In the case of the Papademos government, the ‘technocrat’ PM also proposed a handful of his own personal choices for ministerial posts to the party leaders’ forum (Interview 5). Proposals generally become accepted by the other party leaders and no vetoes are posed. This is not to say that opinions are not expressed in some cases on other parties’ candidates, such as for instance during the relevant discussion in the Samaras II government. It was commonly understood that an objection could not

1989 20 ND, 4 SYN 1989 10 ND, 10 PASOK, 3 SYN, 5 Ind. 2011 19 PASOK, 3 ND, 2 LAOS, 2 Ind. 2012 17 PASOK, 4 ND, 3 Ind. 2012 21 ND, 2 PASOK, 2 DIMAR 2013 22 ND, 6 PASOK, 1 Ind. 2015 32 SYRIZA, 3 ANEL 2015 32 SYRIZA, 2 ANEL

Tzannetakis Zolotas

34

35

29

25

25

28

24 28

Total number of ministers 24 22

3 ANEL, 1 SYRIZA 3 ANEL 22

15

4 ND, 3 PASOK, 2 21 DIMAR 3 ND, 3 PASOK 23

Ind.

PASOK

PASOK

PASOK

ND ND

SYRIZA SYRIZA

SYRIZA SYRIZA

ND

ND

Ind.

Ind.

ND1 Ind.

SYRIZA

SYRIZA

PASOK

ND

ND

ND

ND1 ND

SYRIZA

SYRIZA

ND

ND

PASOK

PASOK

ND Ind.

Number 1 Prime 2 Finance 3 Foreign 4 Health, of minister affairs welfare, and ministries social insurance

4 PASOK, 2 ND, 2 20 LAOS 4 PASOK, 2 ND 20

Number of watchdog junior ministers per party

SYRIZA

SYRIZA

ND

ND

Ind.

Ind.

SYN PASOK

5 Interior

Note: Party labels followed by a number (e.g. ND1) identifies individual ministers holding multiple portfolios simultaneously. The identification numbers are on a per-party and per-cabinet basis.

Tsipras II

Tsipras I

Samaras II

Samaras I

Papademos II

Papademos I

Year Number of in ministers per party (in descending order)

Cabinet

Table 9.3 Distribution of cabinet ministerships in Greek coalitions, 1977–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



299

be raised for ‘aesthetic’ reasons, that is ‘if the candidate had been around for too long, if they were too populist, etc.’, but could only be put forward if a particular choice ‘raised an issue concerning the character of the coalition’ (Interview 13). The PM, in liaison with the (other) party leaders, has the final say over the placement of the selected appointees in particular posts. In the case of Tsipras I where there was asymmetry between the major partner SYRIZA (35.5 per cent) and the minor partner ANEL (3.7 per cent) the process was slightly different. In that case, the two leaders met immediately after the January 2015 election and within 15 minutes had agreed to form a coalition in which ANEL was to receive one ministry (Defence) and two or three other posts at alternate or junior ministerial level. The remaining posts were decided upon by the PM (Interview 11). In most cases the distribution has followed the proportionality principle although there were variations in the degree to which this was strictly adhered to across governments. Some governments, such as Zolotas and Samaras I, were quite insistent on proportionality (Interviews 8, 12). On the other hand, in some cases the proportionality principle was respected less; for example, in the Papademos governments, the second largest partner (ND) was largely underrepresented. The particular circumstances in which Papademos I was formed help explain the under-representation. In particular, the resignation of PM Papandreou under pressure from the European Union, the proposition of Papademos as a choice of PM that the European Union and lenders would find acceptable, and the forced participation of ND in this coalition government were seen as ways to demonstrate political consensus to the outside world. ND only dispatched few high-profile members of its leadership group (though not its leader) to demonstrate its symbolic support for this government rather than trying to pull its full weight in the process of portfolio allocation (Interview 5). In Tsipras I and II the minor partner, ANEL, was over-represented (Interview 1). This was more a voluntary concession of the larger partner than a demand coming from the minor partner (Interviews 4, 11). Minor partners would in most cases ‘ask for’ specific ministries. For instance, in the Tzannetakis government SYN asked for the Ministry of Justice (Interview 3). In the Papademos governments ND and LAOS asked for few, specific ministries that were not directly related to the negotiation or implementation of the second MoU (Interview 5). In Samaras I DIMAR expressed a preference for the Ministry of Justice, while PASOK asked to keep the Ministry of the Environment. Names and posts entered the process of negotiation at the same time (Interview 13). In the Tsipras I government, the minor partner ANEL was offered the Ministry of Defence (Interview 11). This was, for among other reasons, because SYRIZA, a radical left party, would have had difficulty in finding an appropriate person from its own ranks to place in that particular ministry, traditionally led by the right, whereas the leader of ANEL would have no problem ‘fitting in’. The choice was hence beneficial to both partners (Interview 14).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

300



There is some limited evidence of policy saliency in portfolio allocation. In particular, parties on the left and centre left (PASOK, SYN, SYRIZA, DIMAR) have been consistently in charge of justice, agriculture, industry, and the environment while parties on the right and centre-right (ND, LAOS, ANEL) have consistently preferred defence and foreign policy portfolios. However, social welfare and labour, portfolios that are expected to be priorities by parties on the left, are not more commonly occupied by these parties. Likewise, as mentioned, agriculture, a portfolio that cross-nationally would be preferred by right-wing parties, has consistently been the responsibility of the left. Table 9.3 illustrates the distribution of the five cross-nationally defined core ministerial positions. It demonstrates the predominance of the right-wing parties in the foreign policy portfolio as well as the non-dominance of centre-left parties in social welfare. Allocation of political posts below cabinet and junior ministers presents particular interest in the Greek case due to the extensive range of positions potentially available for patronage at different levels. These range from ministry secretary generals to the directors of the organizations of the broader public sector supervised by the ministries to the directors of hospitals and other public institutions. Significant turnover has taken place in such posts when single-party governments have alternated in power. Appointments to the positions of ministry secretary generals tend to be the prerogative of the minister, whichever party he or she belongs to in a coalition. Below that level, coalition governments have varied both in terms of how extensively they have engaged in making such appointments (or preferred to simply leave incumbents in post) and in terms of how they have engaged in making these appointments. The Papademos (I and II), Zolotas, Samaras II, and Tsipras II governments did not engage in extensive appointments. In the Papademos case this was because the government would be short-lived, and no point was seen in disrupting the continuity of the administration, especially in a turbulent period when the country was negotiating the second MoU (Interview 7). In the three other cases, appointments did not take place to any significant extent because they had already been made by their predecessor governments shortly before. Hence the three coalition governments that engaged in extensive appointments below the level of ministries were the Tzannetakis, Samaras I, and Tsipras I governments. In the case of Samaras I proportionality was important. A strict ‘algorithm’ was followed (4-2-1), meaning that for every four appointments made by the first party (ND) there would be two appointments made by the second party (PASOK) and one by the third (DIMAR). This included 60-–0 hospital directors. The parties vetted these candidates and approved them by consensus. The lower the level of appointment, the more intense were the negotiations among the parties (Interview 8). In both Tzannetakis and Tsipras I a similar mechanism of appointment was used. A committee of two high-ranking party cadres, each of which acted as a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



301

liaison between their party and the government, was tasked with vetting candidates. There had to be consensus among the partners for all appointments. Ministers proposed most candidates and the proposed had to be agreed upon unanimously. Nevertheless, the proposed candidates were largely respected. No vetoes were exercised but in some cases one of the partners presented additional candidates to the committee in order for them to be considered for particular posts. In the case of the Tzannetakis government the main principle for selection was ‘party loyalties and personal acquaintances rather than suitability’ (Interview 1). In the Tsipras I case, the committee selected candidates but also provided political direction to the appointees. The committee had a stable structure and membership, although it was informal and only met when there were issues to resolve. The body did not resolve differences/conflicts; it operated more as a preventive mechanism and as a channel of communication among parties. ‘For example, ANEL wanted to place someone in the Organisation of the Thessaloniki Port, this body would make that decision. It prevented such minor issues escalating and become political’ (Interview 14).

Coalition agreements Few coalition agreements have been made public in post-1974 Greece. The ones that have are presented in Table 9.4. The first was the tripartite agreement on economic policy signed by centre-right ND, centrist PASOK, and left SYN in the context of the 1990 Zolotas government by three senior, moderate, and widely respected representatives of the three parties (Souflias, Gennimatas, Dragasakis). The second was the agreement signed by centre-right ND, centrist PASOK, and centre-left DIMAR, who governed after the 2012 June elections (Samaras I), as well as its updated and more extended version released after DIMAR left the government in 2013 (Samaras II). Coalition agreements have been brief, with the Zolotas agreement at 987 words, Samaras I 2,185 words, and Samaras II 4,054 words. In all other cases the content and core aims of coalition agreements in Greece were made public during parliamentary declarations. The statements can Table 9.4 Size and content of coalition agreements in Greece, 1977–2018 Coalition

Year Size General Policy-specific Distribution Distribution Policies (in %) of of offices procedural in rules competences (in %) (in %) rules (in %) (in %)

Zolotas 1989 987 0 Samaras I 2012 2,185 10.9 Samaras II 2013 4,054 8.1

0 0 0

0 0 0

0 0 0

96.5 83.9 89.3

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

302



in a way be regarded as equivalents to coalition agreements. However, these parliamentary declarations are not directly comparable to the three coalition agreement documents because their main purpose is not to formalize and make public the content of the informal agreement that has taken place between coalition partners but rather to state the government’s general programme for the next term. They are therefore not included in Table 9.4. In terms of their content, the limited coalition agreements available (much like the ‘programmatic declarations’) focus on policy and only to a limited extent the main aims of the government, as can be seen in Table 9.4. The Zolotas agreement, entitled ‘framework of principles’ (πλαίσιο αρχών), focuses specifically on key policies regarding the economy. It does not include any references to general rules, procedural rules, distributions of offices, or distribution of competence. The Samaras I coalition agreement is comprehensive. It also stresses that there are other policy areas on which progress and coordination is needed, in which the aim is to proceed with consensus. General rules refer to the context and character of the coalition: that the government is based on the collaboration of three parties and that it aims to take ‘full responsibilities’ and exhaust its four-year term in office. ‘Dealing with the crisis, opening up the way to growth and re-negotiating the Loan Agreement (Memorandum of Understanding) without risking the country’s place within Europe and the Eurozone’ was the central aim of the government. While specific procedural rules are not mentioned the agreement makes reference to principles of cooperation, for example that the cabinet will operate collectively and transparently, following the programmatic agreements reached among constituent parties and following meritocratic criteria in administration. It also stresses that flexibility will be sought, so that disagreements would not impede on the work of the government nor undermine the consensus. The revised Samaras II document is similar but more thorough and detailed in describing the policy aims and targets, which largely accounts for its greater length. Neither of the two Samaras documents make reference to specific procedural rules and distribution of offices or of competence.

Coalition governance The coalition governance mechanisms that are used in Greek coalitions are presented in Table 9.5. The Greek political system can be characterized as ‘prime-ministerial’, even though formally speaking it is the cabinet (ministerial council) that decides on all major issues of general concern, operating under the principle of collective responsibility. However, as we shall see in this section, this assessment also needs some qualification when referring to coalition governments.

2011 N

2012 N

2012 POST

2013 POST

2015 N

2015 N

Papademos I

Papademos II

Samaras I

Samaras II

Tsipras I

Tsipras II

N/A

N/A

Y

Y

N/A

N/A

N/A Y

N

N

N

N

N

N

N N PS, CaC, CoC PS, CaC, CoC PS, CaC, CoC, Pca PS, CaC, CoC, Pca PS, CaC, CoC, Pca PS, CaC, CoC, Pca PS

PS

CoC

CoC

CoC

CoC

PS, CaC PS PS, CaC PS

PS

PS

PS

PS

PS

PS

PS PS

Y

Y

N (ND, SYN) N (ND, PASOK, SYN) N (ND, PASOK, LAOS) N (ND, PASOK) N (PASOK, DIMAR) Y

Personal union

N

N

N

N

N

N

N N

Issues excluded from agenda

No/No

No/No

Spec./No

Spec./No

Spec./No

Spec./No

No/No No/No

N

N

N

N

N

N

N N

Varied

Varied

Comp.

Comp.

Few

Few

Few Few

Y

Y

Y

Y

Y

Y

N N

Y

Y

Y

Y

Y

Y

Y Y

Coalition Freedom of Policy Junior Nondiscipline appointment agreement ministers cabinet in (watchdogs) positions legislation/ other parl. behaviour

Notes: Coalition agreement: POST = post-election; N = no coalition agreement Conflict management mechanisms: CaC = Cabinet committee; CoC = Coalition committee; Pca = Combination of cabinet members and parliamentarians; PS = Party summit Coalition discipline: Spec. = Discipline only expected on a few explicitly specified matters; No = Coalition discipline not always expected Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement

1989 N 1989 POST

All used Most For most common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Tzannetakis Zolotas

Coalition

Table 9.5 Coalition governance mechanisms in Greek coalitions, 1977–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

304



The cabinet consists of the PM (also called the president of the government), the vice-president/s of the government, ministers, alternate ministers, and ministers without portfolio. Alternate ministers hold specific portfolios within ministries. Junior ministers are not part of the cabinet. A number of cabinet subcommittees are convened regularly or on an ad hoc basis, all of which are chaired and convened by the PM. The Government Committee, chaired by the PM and comprised of the most senior cabinet ministers (among which would normally be counted the ministers for the interior, foreign affairs, public administration, defence, economy; see also Table 9.3) and some of the PM’s close associates, stands out among these configurations. Its task is to take the necessary measures, monitor and oversee the implementation of government policy within the framework of the directions given by the PM and the cabinet, and deliberate and take decisions on all important matters put to it by the PM. Other important cabinet configurations are the Government Council for Foreign Affairs and Defence, the Committee on Institutions and the Economic and Social Policy Committees. The PM is authorized to secure the unity of the government including resolving disagreements between ministers, providing direction for and coordination of all government policy, and overseeing the implementation of the laws. The cabinet is meant to decide on the grounds of consensus, which is articulated by the PM. If explicit disagreement emerges a vote may be called, which is won by absolute majority. If no view gains an absolute majority the PM will try to elicit a consensus, but if the vote results in a tie, the PM has the deciding vote (PD 69/ 2005). In practice the PM is ‘primus solus’ rather than ‘primus inter pares’ in relation to his cabinet (Koutsoukis 1994). The PM determines the cabinet agenda, presides over all collective government organs, has the sole responsibility for appointing and dismissing cabinet members and junior ministers, and reconfigures ministerial departments and the composition of collective government organs at will. The PM is directly accountable to parliament and will answer questions during ‘The PM’s hour’. However, the PM’s dominance is consolidated in the legislative arena by Greek parties’ hierarchical nature as well as the fact that the PM often is the substantive leader of his party, the leader of the party’s parliamentary group, and the final arbiter of his party’s electoral lists. Hence the PM has several means at his disposal to ensure cabinet dominance, party discipline, and a stable parliamentary majority.

The role of individual ministers in policy-making Individual ministers enjoy high levels of policy-making autonomy within the bounds of their competencies, enjoying a status of ‘small PMs’. Ministers set the policy agenda within the remit of their portfolio; they develop and propose policy

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



305

to the cabinet and are responsible for policy implementation with the help of their staff and the civil service. Hence the policy of their ministry will normally reflect the policy principles of their party as well as have their own personal ‘stamp’. Ministers often use their positions as springboards for re-election (if running for parliament, which is the most common practice) or as important ‘stepping stones’ if their ambition is party leadership. Since ministers are usually also elected MPs their personal political clout often derives from their popular legitimacy as seen in the number of preference votes they receive on election ballots, the significance of their constituency, and their position in their party and (in some cases) their policy expertise or previous ministerial record. The parliament has very limited scope for constraining a minister’s autonomy given high levels of party discipline, weak oversight mechanisms, and restrictive legislation that protects individual ministers from impeachment. It is from other sources that limitations to ministerial autonomy arise. First, ministers are autonomous, provided their decisions do not interfere with the remit of other ministries or require coordination with other ministries because of their decisions’ financial or staffing implications or because of the interdepartmental nature of those decisions (requiring the cooperation of two or more ministries) (Koutsoukis 1994: 275–6). Second, while they are meant to have the support of the civil service under their jurisdiction for the implementation of their policy there may be constraints posed by policy experts with an agenda or politically hostile civil servants. Yet the minister has the capacity to overcome such constraints by assembling a personal office of their own collaborators and, importantly, by appointing a ministry secretary general of their own choice (Interviews 3, 6). Third, a minister’s autonomy can to a limited extent be restricted by the appointment of an alternate minister or junior minister/s from the same party in the case of single-party government or from other parties in the case of coalition governments. Alternate ministers and junior ministers are more often than not chosen with the rationale of accommodating different tendencies within a single-party government or coalition parties in the case of coalition governments. Their role can be to guarantee proportionate representation or, in some cases, provide special skills or knowledge in a particulate policy area. Sometimes however they are placed with a ‘watchdog’ rationale, both in cases of different tendencies within a single-party government or to keep an eye on the minister of a different party. With the exception of the Tzannetakis coalition government (ND–SYN), which did not engage in ‘watchdog’ practices (Table 9.5), all other coalition governments since 1974 did so to different extents. It would seem that there is a pattern whereby PASOK seems to want to keep a watchful eye on ministries where the right dominates, in particular foreign policy, defence, and the interior, all ministries relevant to national security and the armed forces traditionally politically closer to the right. There is also a second pattern whereby minor coalition partners use

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

306



‘watchdogs’ in order to remain closely informed on policy developments in areas especially important to their programmes or constituencies. Based on information from interviews it appears that the aim of the coalition partners, who are not in charge of a particular ministry, is to provide channels of information and coordination among parties rather than limit ministerial autonomy (Interviews 5, 9). Finally, especially in single-party governments a minister’s policy autonomy can be restricted by the PM and his ‘inner cabinet’ whereby the PM makes key decisions on collective government policy based on a small group of trusted ministers and associates operating as a quasi-think-tank that presents him with policy options and assists him in making decisions through deliberation on those options (Koutsoukis 1994). If the PM and his inner cabinet make decisions that encroach on a minister’s territory this may lead to a minister’s plans being modified or even overruled (Interview 6). One might expect that the power of the PM over his ministers’ policy choices brings Greece closer to the PM dominance model of coalition governance. However, in coalition governments a slightly different dynamic seems to be at work. On the whole, in coalition governments the PM is more ‘discrete’ when it comes to intervening in a minister’s work compared to single-party governments. He has to take into account the wishes of the party leader whose party the minister is representing (Interview 13). However, interview evidence points to the fact that when the PM is also the leader of the largest party they may still try to intervene to avert a minister’s policy going through when the matter conflicts with their political agenda, when the minister belongs to a party with a different agenda. One example mentioned was the intervention of the leader of ND to block the law on judicial codes that was being prepared by the Ministry of Justice (at the time held by a minister of SYN) in the Tzannetakis government; the outcome was not the side-lining of the competent minister but it certainly was the modification of the planned law (Interview 6). Another example is the intervention of ND, the largest party, which proposed significant additions to the proposed law on immigration prepared during the Samaras I government. The minister of justice, appointed by DIMAR, a minor partner, was asked to include clauses that treated the punishment of criminal offenses differently when these were conducted by nationals and migrants without papers. The minister, supported by his party leader, refused to follow the wishes of the minister of justice (Interview 6). In cases where the PM was an independent and not the leader of one of the coalition parties their ability to intervene and dominate a policy argument was limited. In the case of the Tzannetakis government for instance, where the PM was a senior ND figure, it was the leader of the party, Mitsotakis, that made key policy decisions, which Tzannetakis carried out (Interview 6). Zolotas initially thought he might be able to influence economic policy given his expertise and proposed ‘27 measures for the management of the crisis’ to the economic ministries. These were rejected by the political leaders (Interview 12). Likewise, Papademos’ attempts to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



307

take policy initiatives in the areas of higher education (Interview 5) and employment policy (Interview 7) were shunned by the political leaders. It would seem that neither the ‘ministerial government model’ nor the ‘coalition compromise model’ in their pure forms accurately describe the dynamics of ministerial autonomy in Greek coalition government. Rather, PMs continue to have a central role, but one that is less clear-cut than in single-party governments. While individual ministers are strong and fairly autonomous in policy-making, they are not entirely immune to external limitations, particularly those coming from the other parties, especially the strongest partner if they also hold the premiership. Yet is seems that parties intervene in each other’s policy domain directly through their leaders rather than through ‘watchdogs’. While there is some evidence of ‘watchdog’ practices in coalition governments these are not applied thoroughly or across the board among coalition partners. ‘Watchdogs’ normally apply to a small subset of policy areas where coalition partners wish mainly to be kept informed of policy developments in a policy area where they have little influence or in an area especially important to their programme or constituents. Let us examine more closely the governance mechanisms in three specific areas: inside the executive arena, in the parliamentary arena, and when these arenas are mixed.

Governance mechanisms in the executive arena Cabinet Committees (CaC) can be a locus for resolving policy specific disagreements among (typically) ministers that belong to different coalition parties, mainly when there are issues of joint responsibility. This however happens to the same extent in single party governments where CaC are used to resolve issues among ministers of the same party. It is not specifically used in coalition cabinets. In coalition governments some disputes are addressed within the internal context of ministries where the political leadership is shared between different parties, for example where a SYRIZA minister co-exists with an ANEL junior minister (Interview 14). It is also quite common that the PM, a minister of state, or the vice-president plays a mediating role between the ministers in conflict. PM Zolotas, for example, would sometimes meet with the competent minister/s to work things out for less politically central or more specialized problems (Interview 1). Tsipras would also hold a meeting with the competent minister and the other ministers involved from both parties himself, although this only happened a few times (on the refugee issue and the salaries of military personnel) (Interview 11). In the Papademos governments likewise, for specific policy issues, the PM or the minister of state met the competent ministers (Interview 10). In the Samaras governments the minister of State would sometimes meet with the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

308



ministers who had disagreements to mediate. In the Tsipras governments the vicepresident (Interview 1) or minister of state for government coordination would sometimes mediate between the ministers.

Governance mechanisms in the parliamentary arena In Greek coalition governments no coalition discipline is enforced but rather each party group is responsible for its own MPs. With regard to parliamentary questions towards ministers there is neither party nor coalition discipline. Loose and in most cases ad hoc and informal mechanisms of coordination of the parliamentary groups of the coalition parties and the government are in operation. There are no joint meetings of the parliamentary groups or any institutionalized meetings among parliamentary group leaders or any regular meetings among parliamentary group leaders and members of the cabinet. There have been, nonetheless, informal meetings between parliamentary group representatives and ministers or between parliamentary group leaders and the secretary general of the government (Interview 8), cases where ministers appeared before the parliamentary groups of other coalition partners (Interview 14) as well as cases where an issue upon which coalition parties were not entirely in agreement was discussed in a standing committee. These may all be considered instances of embryonic combined coalition forums for cabinet members and parliamentarians (Pca in Table 9.5).

Governance mechanisms with different types of actors (mixed) As can be seen clearly in Table 9.5, the party summit (PS) comprising ad hoc, informal, and frequent meetings of party leaders (who may or may not be cabinet members) is the key mechanism used in order to resolve the most significant conflicts among coalition partners. The party summit will often also deal with issues that arise among the partners over policy specific issues that cannot be resolved at lower levels. In such cases the minister/s involved will also be invited to attend. These mechanisms may in some cases be described as Coalition Committees (CoC) when specific ministers and officials relevant to particular high salience issues are regularly invited to contribute. Such was the case in the Papademos and Samaras governments for instance when the negotiation and implementation of the second MoU was the central concern of the government. In the Tzannetakis government the main mechanism to coordinate and resolve differences/conflicts among government partners was the meeting of party leaders (and a few trusted advisers) with the PM (Interviews 2, 3). This informal forum met regularly, sometimes two to three times a week. Where there was disagreement on a specific issue, the minister responsible was invited and there was

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



309

discussion. Two examples: the law on judicial codes and the law on rehabilitating the victims of (both sides of) the civil war. The first disagreement was resolved in favour of the view of the minister and against the view of the largest party, ND. The second issue was resolved consensually. These meetings with the participation of the competent ministers took place when issues arose, ad hoc, more sparsely than the leaders’ meetings, as often as every two weeks (Interview 3). In the Zolotas government likewise, the three political leaders and their close associates would make key decisions and resolve conflict on all major issues, often discussing with the PM just before or during the meetings (Interview 12). They met frequently, every week (Interview 1). It was clear the leaders’ forum was the centre of real power and this trumped any collective decision taken by the cabinet, the result being that cabinet ministers often felt like mere ‘implementers’ of leaders’ decisions. This was also due to the limited timespan and scope of this government (Interview 1). A serious disagreement surfaced when a decision had to be made regarding assigning a major digitization contract for the modernization of the national telecoms company (OTE) to Greek provider Intracom. The cabinet was against assigning the contract but the informal leaders’ forum decided in favour and got its way (Interviews 2, 12). In the Papademos governments the main political disagreements were discussed and resolved ad hoc by the political party leaders and the PM. Meetings took place often, at crucial times on a weekly basis, especially when major issues such as the Greek negotiating position on the second MoU were at stake (Interviews 5, 10). The PM had to step in sometimes, acting as a mediator between the political leaders. One example was a disagreement between PASOK and ND on the question of where to implement cuts in order to cover the fiscal gap, with PASOK arguing for cuts in auxiliary pensions and ND arguing for cuts in the public investment programme. This was ultimately resolved by the PM. At a more day-to-day level their main associates, that is the minister of state on behalf of the PM and the directors of the offices of the political leaders, were in close collaboration throughout the period and dealt with the main issues and resolved conflicts on specific issues as and when those arose (Interview 10). In the two Samaras governments the political leaders and the minister of state (who was a senior ND cadre close to PM Samaras) met to both coordinate the main policies of the government, which were mainly related to the second MoU, and to manage/resolve conflict among the partners. This mechanism was ad hoc and informal. It was called regularly, sometimes with a specific agenda and sometimes without one. One-on-one meetings between the minister of state and individual political leaders would sometimes substitute the meeting of the four (Interview 9). In the first Samaras government the meeting of the political leaders and the ministers in the economic ministries as well as the governor of the Bank of Greece were very common given that the priority was the MoU (Interview 3). Leaders met on a weekly basis (Interview 13). When there were issues this forum

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

310



met with the relevant ministers (Interview 8). Where there was serious disagreement for non-MOU issues the matter was avoided. One example was the vote for Greeks abroad (Interview 9)—there was agreement that there would be no constitutional reform (Interview 13). In the case of symbolic issues that pertained to left–right differences the political parties sometimes agreed to disagree, for example anti-racism law and citizenship (Interview 13). When a serious political disagreement arose in the Tsipras I and II governments it was discussed by the two party leaders, who were in regular contact, daily or every other day. Discussions took place informally and ad hoc when need arose on particular issues. It was well known that there were some blatant political differences in matters such as the separation of church and state, LGBT rights, citizenship law, and foreign policy. On a few potentially heavily controversial issues where SYRIZA would have liked to intervene, such as church–state relations, they did not want to generate problems with ANEL (Interview 1). On the other hand, SYRIZA did bring forward other such issues such as laws on social rights and relations with FYR Macedonia, which were not supported by ANEL in parliament and which, in the case of the latter, caused the demise of the Tsipras II government. Given the central role of political leaders in Greek coalition governments, an important point raised in some of the interviews is the significance of personal relationships. One interviewee stressed the importance of the person of the PM in the Tzannetakis government in building trust among the coalition partners. Before going into politics with ND Tzannetakis was a navy officer who had good personal relations with participants in the resistance movement against the 1967–1974 military dictatorship, including politicians from the left (Interview 2). The friendly personal relations between PM Alexis Tsipras and Panos Kammenos were also found to be significant in keeping together the two unusual coalition governments between SYRIZA and ANEL (Interview 4). Interviewees emphasized this link and its importance in smoothing out disagreements and difficulties in the coalition’s lifespan, noting that ‘they are on the phone every day’ (Interview 11).

Coalition termination The duration of cabinets Coalition governments have been more short-lived than single-party governments in post-1974 Greece. As can be seen in Table 9.6 most single-party governments in the 1980s, 1990s, and 2000s reached near completion of their four-year term. The longevity of the first governments of Andreas Papandreou, the Mitsotakis government, Simitis III, and Karamanlis I is indicative. On the contrary, the duration of coalition governments has on average been brief, hardly exceeding a

Date in

1977-11-28

1980-05-10

1981-10-21

Cabinet

Karamanlis

Rallis

Papandreou I

1985-06-02

1981-10-18

1980-05-10

Date out

90.5

92.8

61.2

Relative duration (%)

Table 9.6 Cabinet termination in Greece, 1977–2018

4, 9

4, 9

2

Mechanisms of cabinet termination

11

11

Terminal Parties (when events conflict between or within)

Policy area(s)

Continued

Karamanlis resigned in order to take up the office of the president of the Republic. The Rallis government submitted its resignation to the president of the Republic. The reasons stated for the resignation were trivial (the proximity of the four-year cut-off point with the festive season, the damage done to the economy by the prolonged pre-electoral period, etc.) The political reasons were the loss of public popularity of the government, whose PM Rallis was seen as a substitute for the previous PM (Karamanlis), infighting in the government majority, and the rapidly increasing popularity of PASOK, which was on its way to win the next election by a landslide. The Papandreou government submitted its resignation to the president of the Republic, putting forward reasons of ‘national importance’, in particular the need to take forward the amendment of the Greek constitution and significant developments in the Cyprus issue.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Date in

1985-06-05 1989-07-02

1989-10-12

1989-11-23

Cabinet

Papandreou II Tzannetakis

Grivas

Zolotas

Table 9.6 Continued

1990-04-08

1989-11-05

1989-06-18 1989-10-07

Date out

9.4

1.8

100 6.7

Relative duration (%)

2, 7b

4

1 4, 9

Mechanisms of cabinet termination

ND, PASOK, SYN

Terminal Parties (when events conflict between or within)

Policy area(s)

Early elections were held after the Tzannetakis government had fulfilled its mandate. The later involved the referral to trial in a Special Court of the former PM Andreas Papandreou and a small number of his former ministers for allegations of corruption. Grivas’ caretaker government had a limited mandate: to conduct early elections. It resigned immediately after completing its task. Early elections were held after parliament had failed to elect a new president of the Republic after three rounds of voting (constitutional provision). The three government parties failed to agree on a common candidate for the presidency

(According to the Greek constitution the resignation of a government followed by legislative elections is possible when a matter of ‘grave national importance’ requires a fresh mandate from the people.)

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1996-01-16

1996-09-22

Papandreou III 1993-10-13

Simitis I

1996-01-22

1993-10-10

1990-04-11

Mitsotakis

38.9

56.6

87.7

4, 9

9

4, 8

11

11

ND

Continued

The Mitsotakis government submitted its resignation to the president of the Republic, putting forward reasons of ‘national importance’ , in particular developments in former Yugoslavia, significant developments in the Cyprus issue, and the Greek presidency of the European Union. However, this took place against the background of internal infighting in Mitsotakis' own governing party (his foreign minister, Antonis Samaras, had left and founded a new party Politiki Anoixi) and the threatened loss of his very slim parliamentary majority. Andreas Papandreou resigned due to falling terminally ill. The Simitis I government submitted its resignation to the president of the Republic, putting forward reasons of ‘national importance’ , in particular developments in the European Union, the Balkans, Turkey, and Cyprus. The background was the need to receive a fresh parliamentary mandate, given that Simitis had been the PM select of the governing PASOK (replacing Andreas Papandreou) and had won an investiture vote but his government did not carry the legitimation of a popular election.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Date in

1996-09-25

2000-04-13

2004-03-10

2007-09-19

2009-10-06

Cabinet

Simitis II

Simitis III

Karamanlis I

Karamanlis II

Papandreou

Table 9.6 Continued

2011-11-11

2009-10-04

2007-09-16

2004-03-07

2000-04-09

Date out

52.5

51.2

88.1

97.7

88.6

Relative duration (%)

9

4, 9

4, 9

4, 9

4, 9

Mechanisms of cabinet termination

11

11

11

11

11

Terminal Parties (when events conflict between or within)

Policy area(s)

The Simitis II government submitted its resignation to the president of the Republic, putting forward reasons of ‘national importance’ , in particular the country’s admission to the Eurozone. The Simitis III government submitted its resignation to the president of the Republic, putting forward reasons of ‘national importance’. The Karamanlis I government submitted its resignation to the president of the Republic, putting forward reasons of ‘national importance’. The Karamanlis II government submitted its resignation to the president of the Republic, putting forward reasons of ‘national importance’, in particular, educational and economic reforms as well as the drafting of the 2008 budget. Georgios Papandreou resigned after his government narrowly survived a vote of no confidence on 5 November. Papandreou had been heavily criticized for announcing a referendum on a new, second, bailout package without informing other cabinet members or EU officials.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2011-11-16

2012-02-11

2012-05-16

2012-06-21

2013-06-22

2015-01-26

Papademos I

Papademos II

Pikramenos

Samaras I

Samaras II

Tsipras I

2015-08-20

2015-01-25

2013-06-22

2012-06-17

2012-05-06

2012-02-10

14.1

53.3

25

100

14.2

12.6

4, 8

2

7a

4

4, 9

7a

11

SYRIZA

ND, PASOK, DIMAR

ND, PASOK, LAOS

Continued

LAOS withdrew from the coalition in advance of a crucial vote in parliament regarding a package of austerity measures associated with the second bailout package. LAOS refused to back the government measures in the parliament. Pre-planned early elections were held after the Papademos government had fulfilled its mandate, that is to complete negotiations for the second bailout package and pass through parliament the associated austerity measures. Pikramenos caretaker government had a limited mandate: to conduct early elections. It resigned immediately after completing its task. Communications DIMAR withdrew from the coalition after the government decided to abolish the Hellenic Broadcasting Corporation and fire its employees. Early elections were held after parliament had failed to elect a new president of the Republic following three rounds of voting (according to the Greek constitution this automatically leads to legislative elections). Economy The Tsipras government lost its parliamentary majority after a third of SYRIZA members failed to support a

Economy

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Date in

2015-08-28

2015-09-23

Cabinet

ThanouChristophilou

Tsipras II

Table 9.6 Continued

2019-01-13

2015-09-20

Date out

82.8

100

Relative duration (%)

7a

4

Mechanisms of cabinet termination

12

SYRIZA, ANEL

Terminal Parties (when events conflict between or within)

Foreign policy

Policy area(s)

government proposal to implement further austerity measures as a precondition to receiving a third bailout package from the European Union and IMF (passing with opposition support on 16 July and 13 August respectively). The dissenting MPs formed a new party and parliamentary group, Popular Unity (LAE). Thanou-Christophilou’s caretaker government had a limited mandate: to conduct early elections. It resigned immediately after completing its task. The minor government partner ANEL withdrew from the coalition after a prolonged disagreement over the government’s forging of the Prespes Agreement. This was a bilateral agreement forged between Greece and the Republic of North Macedonia to settle a historical dispute over the name of the latter. ANEL contested the agreement on patriotic lines. It was unable to vote down the Prespes

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Note: Relative duration: share of constitutionally allowed duration for this cabinet.

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of PM Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event

Agreement in parliament while staying in government since it considered this a core issue of principle for its party. It hence left the government on these grounds.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

318



year. A change can be detected in the post-2012 period with nearly two years for Samaras II and nearly three and a half years for Tsipras II. These may well be the first stable coalition governments under the new party system. In contrast to previous coalition governments, the duration and remit of the Tsipras and Samaras governments were not pre-determined but open to completion of the four-year term. That said, there was a clear drop in relative cabinet duration in the 2010s, which reflects the higher number of coalition governments after 2012.

The termination of cabinets During the first democratic period (1974–2011), only one cabinet formally lasted a full term, Papandreou II. All other governments were terminated early, although several of the single-party governments of this period could be argued to have reached near full term. The most common reason for early cabinet termination has been a strategic decision by the PM when they believed it would be beneficial for them and their parties at that stage of the political cycle. Formally speaking, the constitution allows a PM to hand in the resignation of his government to the president of the Republic and call early elections for reasons of ‘national importance’. This provision has been used by nearly all PMs that headed the governments of this period. The circumstances they appealed to varied, but in most cases made reference to a controversial matter of foreign policy, security, or economic policy that requires a fresh mandate from the people. As can be seen in Table 9.6, matters related to the country’s relation to the European Union are one of the most common justifications. The Mitsotakis government resigned in October 1993 so that the next government could run the Greek Presidency of the European Union cohesively, and the second Simitis government resigned in April 2000 so that the next government could handle the country’s accession to the Eurozone. Likewise, developments in the Cyprus issue and regional developments in the Balkans, considered vital to Greek foreign policy, have been appealed to at least twice by PMs Mitsotakis and Simitis I when they resigned from government. Another common reason has been crucial economic policy choices or the need to implement proposed reforms, policy related or constitutional, that require legitimation from the people. The Karamanlis II government for instance resigned so that it could receive a fresh mandate to implement controversial education and economic reforms. Rallis, Simitis (I), and George Papandreou handed in their resignations and triggered early elections because their own position was weak. Rallis and Simitis had succeeded the previous PM (Karamanlis and Andreas Papandreou respectively) mid-term and felt they needed to strengthen their mandate by requesting the vote of the people, while George Papandreou resigned despite narrowly surviving a vote of no confidence because his position was no longer tenable

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



319

after his personal handling of the political crisis of 2011 that led to the instalment of the first Papademos government. Two PMs have resigned as a result of their personal inability to continue with their duties: Karamanlis resigned so that he could take on the office of the president of the Republic and Andreas Papandreou brought his third government to an end because of his ill health. Finally, the Zolotas government resigned because parliament failed to muster the supermajority required to appoint a president of the Republic, which is a trigger for the dissolution of parliament and early elections. In the post-2012 period there is no single predominant reason or clear pattern for cabinet termination, that is, as was the appeal to reasons of ‘national importance’ for the 1974–2012 period. However, one thing that can be noted is that this has been a time when internal policy disagreement among coalition partners has been more frequent. This has led to the departure of a coalition partner twice, when LAOS left the Papademos I government over economic policy disagreement and DIMAR left the Samaras I government over its handling of the national broadcaster, ERT. Policy disagreement has also been common in the Samaras II and Tsipras II governments (in the latter case SYRIZA and ANEL have publicly disagreed over LGBT rights, citizenship law, church–state relations, and foreign policy) but this only led to cabinet termination in the case of Tsipras II. Finally, the three caretaker governments (Grivas, Pikrammenos, and ThanouChristophilou) and four special purpose governments (Tzannetakis, Zolotas, and Papademos I and II) since 1974 were terminated early when their agreed task was (perceived to be) completed. With regard to the governments led by ‘technocrats’ Zolotas and Papademos it is further well known that they were undermined by their component parties as those parties positioned themselves for upcoming elections, a fact that sped up the end of their term (Interviews 5, 7, 12).

Conclusion The coalition ‘life cycle’ in Greece is characterized by a brief coalition formation phase, which is more likely to result in surplus connected coalitions and more often than not does not involve formal coalition agreements. Moreover, it is characterized by the continued dominance of the PM, but also by individual ministers as regards policy-making and of party leaders concerning government coordination and conflict resolution during the governance phase. More recently, there are low levels of coalition cabinet duration and no particular pattern in coalition termination. The main effect of the economic crisis on the coalition life cycle has been first and foremost that coalition government itself, perhaps temporarily, has become the norm. The coalition formation phase remains unchanged, as does the governance phase, which remains party leader dominated, ad hoc, informal, and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

320



poorly institutionalized. The main change so far concerns the duration of coalition governments, which is much lower than that recorded for single-party governments. The rise of new actors, some of which have exhibited populist characteristics, have not affected any aspect of the coalition life cycle. Coalitions of mainstream parties have given their place to ‘populist’ coalitions. Parties that enter government with a populist agenda ultimately used similar ways of forming, governing, and terminating coalitions to parties considered non-populist. Despite the continuities in the composition and operation of coalition governments over the democratic period there does seem to be a significant change after 2012: while all coalition governments before 2012 were set up to perform a specific task and had a pre-determined temporal duration that corresponded to the completion of the task, both the Samaras and the Tsipras governments declared themselves governments with a broad remit and unlimited temporal duration. Despite being confined by the imperative of implementing MoUs and/or mitigating their social consequences, the fact that they were open time horizon governments means that they were expected to deal with the management of the country fully and that they accepted that responsibility. If party system fragmentation remains a feature of Greek politics there may well be incentives for parties to develop more elaborate coalition governance practices with positive knock-on effects on coalition duration. In 1991, in their study of the first coalition governments of the democratic period, Pridham and Verney (1991: 50) wrote that parties entered these collaborations without any experience and were not used to the ‘continual intra-party compromise’ required. Since 2012, experience of coalition government has been accumulating and significant learning has taken place in terms of working collaboratively across parties. Interviewees from these last four coalition governments have stressed that in many respects working in a coalition has been ‘an undoubtedly successful and positive experience’ (Interview 13), that their collaboration with politicians from other parties at the level of policy has been ‘respectful and without fault’ (Interview 8), and that ‘it is often more political’ (as opposed to tribal and/or personal) to work in coalitions (Interview 13) than it is to participate in single-party governments. This accumulated positive experience of collaboration may well lead to a significant shift in the perception of coalition government among political elites in Greece and eventually establish coalition government as the ‘new normal’.

References Andreadis, Ioannis (2015). ‘The Ideological Foundations of the Greek Coalition Government’. In Roman Gerodimos (ed.), First Thoughts on the 25 January 2015 Election in Greece. PSA Greek Politics Specialist Group Pamphlet No. 4, 44–5.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



321

Aslanidis, Paris, and Cristóbal Rovira Kaltwasser (2016). ‘Dealing with Populists in Government: The SYRIZA-ANEL Coalition in Greece’. Democratization, 23(6): 1077–91. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010.’ Party Politics, 21(1): 143–52. Clogg, Richard (1987). Parties and Elections in Greece. The Search for Legitimacy. Durham, NC: Duke University Press. Dinas, Elias, and Lamprini Rori (2013). ‘The 2012 Greek Parliamentary Elections: Fear and Loathing in the Polls’. West European Politics, 36(1): 270–82. Ellinas, Antonis (2013). ‘The Rise of Golden Dawn: The New Face of the Far Right in Greece’. South European Society and Politics, 18(4): 543–65. Featherstone, Kevin (1990). ‘The “Party-State” in Greece and the Fall of Papandreou’. West European Politics, 13(1): 101–15. Kalyvas, Stathis and Nikos Marantzidis (2003). ‘The Two Paths of the Greek Communism’. In Joan Botella and Luis Ramiro (eds), The Crisis of Communism and Party Change: The Evolution of West European Communist and PostCommunist Parties. Barcelona: ICPS, 15–32. Kiapidou, Nikoleta (2015). ‘A New Greek Party System Is Here to Stay’. In R. Gerodimos (ed), First Thoughts on the 25 January 2015 Election in Greece. PSA Greek Politics Specialist Group Pamphlet No. 4, 21. Koutsoukis, Kleomenis (1994). ‘Cabinet Decision Making in the Hellenic Republic, 1974-1992’. In Michael Laver and Kenneth A. Shepsle (eds), Cabinet Ministers and Parliamentary Government. Cambridge: Cambridge University Press, 270–82. Linardatos, Spyros (1977). From the Civil War to the Dictatorship. Volume A, 1949–1952. Athens: Papazisis. Lyrintzis, Christos (2005). ‘The Changing Party System: Stable Democracy, Contested Modernisation’. West European Politics, 28(2): 242–59. Moschonas, Gerasimos (1994). ‘The right - anti-right cleavage in the post-dictatorship period’. In Nikos Demertzis (ed), Contemporary Greek Political Culture. Athens: Odysseas, 159–216. Müller, Wolfgang C., and Kaare Strøm (eds.) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Nicolakopoulos, Elias (2000). Stunted Democracy. Parties and Elections 1946-1967. Athens: Patakis. Pappas, Takis (2003). ‘The Transformation of the Greek Party System since 1951’. West European Politics, 26(2): 90–114. Pappas, Takis (2014). Populism and Crisis Politics in Greece. Basingstoke and Houndmills: Palgrave Macmillan.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

322



Pastorella, Giulia (2014). ‘All Technocratic Governments Are Equal, but Some Are More Equal than Others: The Peculiarities of the Greek Case’. Paper presented at the ECPR General Conference University of Glasgow, September 2014. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of Antielitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Pridham, Geoffrey, and Susannah Verney (1991). ‘The Coalitions of 1989-90 in Greece: Inter-party Relations and Democratic Consolidation’. West European Politics, 14(4): 42–69. Strøm, Kaare, Wolfgang C. Müller, and Torbörn Bergman (eds.) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Teperoglou, Eftichia, and Emmanouil Tsatsanis (2014). ‘Dealignment, De-legitimation and the Implosion of the Two-Party System in Greece: The Earthquake Election of 6 May 2012’. Journal of Elections, Public Opinion and Parties, 24(2): 222–42. Tsakatika, Myrto (2016). ‘SYRIZA’s Electoral Rise in Greece: Protest, Trust and the Art of Political Manipulation’. South European Society and Politics, 21(4): 519–40. Tsatsos, Dimitris (1993). Constitutional Law, Volume B. Organisation and Functioning of the Polity, 2nd edition. Athens: Sakkoulas. Tsirbas, Yannis (2015). ‘The January 2015 Parliamentary Election in Greece: Government Change, Partial Punishment and Hesitant Stabilisation’. South European Society and Politics, 21(4): 407–26. Verney, Susannah (1990). ‘Between Coalition and One-Party Government: The Greek Elections of November 1989 and April 1990’. West European Politics, 19(4): 131–38.

Appendix. List of political parties Abbreviation KKE KKE-es SPAD LAE OIKEN SYRIZA SYN DIKKI PASOK EK

Name Communist Party of Greece (Kommonounistiko Komma Hellados) Communist Party of Greece-Interior (Kommounistiko Komma ElladasEsoterikou) Alliance of Progressive and Left Wing Forces (Symmachia Proodeftikon kai Aristeron Dynameon) Popular Unity (Laiki Enotita) Federation of Ecological and Alternative Organizations (Omospondia Oikologikon kai Enallarktokon Organoseon) Coalition of the Radical Left (Synaspismós Rizospastikís Aristerás) Coalition of the Left and Progress (Synaspismos tis aristeras kai tis proodou) Democratic Social Movement (Dimokratiko Koinoniko Kinima) Panhellenic Socialist Movement (Panhellinio Sosialistiko Kinema) Union of Centrists (Enosi Kentroon)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 EDIK TP KN DIANA ND POLAN ANEL IM LAOS EP XA

323

Union of Democratic Center (Enosi Dimokratikou Kentrou) The River (To Potami) New Liberal Party (Neofileleftheroi) Democratic Renewal (Dimokratiki Ananeosi) New Democracy (Nea Democratia) Political Spring (Politiki Anoixi) Independent Greeks (Anexartitoi Ellines) Independent Muslims Popular Orthodox Rally (Laïkós Orthódoxos Synagermós) National Front (Ethniki Parataxi) Popular Association – Golden Dawn (Laïkós Sýndesmos – Chrysí Avgí)

Note: Party names are given in English, followed by the party name in Greek in parentheses.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 10 Iceland: Political Change and Coalition Politics Indridi H. Indridason and Gunnar Helgi Kristinsson

Prior to the economic crash of 2008, coalition governance in Iceland was probably more influenced by ministerial government than in most other parliamentary systems. Ministerial government, according to Laver and Shepsle (1990, 1996), is the method of coalition governance that grants each coalition party full autonomy over the portfolios it holds so that the initial allocation of ministries effectively contains the policy bargain on which the coalition is based, provided that ministers follow their party’s programme. While this model tends to be viewed as a weak form of coalition governance in the literature, given its susceptibility to principal– agent problems, it may function to the satisfaction of coalition partners, given certain conditions. These include ideological proximity among coalition partners and a powerful parliament as an arena for continuous bargaining among coalition partners (Indridason and Kristinsson 2018). Ministerial government in Iceland was characterized by relatively little coordination at the cabinet level, including short coalition agreements, lack of a ministerial hierarchy, little use of collective cabinet decisionmaking, and very little watchdog control of ministers. The 2008 economic crash and its aftermath called for greater cabinet coordination and, thus, placed ministerial government under pressure. The most significant of these pressures came in the form of growing distrust of politicians that led to electoral volatility and fragmentation of the party system. This, in turn, made the task of forming strong coalitions and keeping effective control over government and parliament more difficult than before. An additional source of uncertainty was the ambition of president Grímsson (1996–2016) to expand the role of the presidency and intervene in government business. Subsequent governments have responded to these challenges by attempting to strengthen cabinet management and improving coordination at the sub-cabinet level. Coalition agreements have become longer than before, and watchdog mechanisms are increasingly used in the parliamentary committees. Nonetheless, more coordinated forms of coalition governance remain a challenge in many respects. Many politicians in the established power bloc (Independence Party, IP, and Progressive Party, PP) remain unconvinced of the desirability of greater coordination at the cabinet level and may look for opportunities to strengthen ministerial Indridi H. Indridason, and Gunnar Helgi Kristinsson, Iceland: Political Change and Coalition Politics In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0010

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

325

autonomy anew. A complete U-turn, however, back to the golden days of ministerial autonomy, may be difficult to achieve, given increasing difficulties in forming ideologically cohesive coalitions with only two or three partners and comfortable majorities in parliament.

The institutional setting The parliamentary system that took shape in Iceland during the twentieth century was influenced partly by the Danish constitution and partly by the political forces emerging during and after the secession from Denmark. This included negative parliamentarism where the president (in 1944) took over the role of the king in appointing government formateurs but kept his distance from politics in other respects. With single-party majorities being absent since the 1920s, political leaders maintained a strong preference for minimum winning coalitions, usually containing two or three parties, while minority governments were rare and mainly provisional. Although the president was popularly elected, presidential candidates were usually not affiliated with the political parties and adhered to the principles of parliamentary government, while party leaders were in control of parliament and government. The constitution that was adopted in 1944 was in most respects a recycled version of the Danish constitution except for a new section on the presidency. Although the form of government changed from monarchy to semipresidentialism, the intentions of most political leaders were probably for politics to remain much the same. The president selects a formateur after consulting with the party leaders. There is no informateur although the president is free to seek expert advice if that was called for. The norms concerning selection of formateur are fairly simple, although implementing them can be complicated. The norm is that the president should select as formateur the party leader most likely to be able to form a majority government and often the role of formateur goes first to the leader of the largest party. But that need not be the case if an alternative majority coalition that excludes the largest party is feasible. In 2013, when the IP and PP tied as the largest parliamentary party, President Grímsson handed the formateurship to the PP leader, stating that the PP’s electoral victory had been more convincing despite the PP’s smaller vote share. When no obvious majority is in sight, the president may either decide on freefor-all negotiations or continue appointing new formateurs in the hope of a breakthrough. When party leaders feel they are getting nowhere, they usually turn in their mandate to form a government to the president. Presidents have been known to threaten politicians with the prospect of a non-parliamentary cabinet (i.e. not based on the parties in parliament) but no such cabinet has been formed

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

326

 .     

since 1942–1944 (e.g. Jóhannesson 2005). The president has not appointed a formateur authorized to form a minority government except with a very limited mandate, since 1949. Minority governments are considered too weak to compete with parliament and have primarily served as caretaker governments or been established in the process of undertaking constitutional reform (which requires two votes in parliament with an election in the interim). No radical revisions of the constitution have been undertaken, despite promises to that effect, and constitutional evolution in the post-war era became very much a gradual affair concerning things such as electoral rules, human rights provisions, and the organization of the legislature. The bicameral (or even tricameral if the United House is counted as a separate chamber) Alþingi (Icelandic parliament) became a unicameral legislature in 1991.¹ Renewed interest in a more thorough revision of the constitution emerged after President Grímsson, in 2004, took the unprecedented step of refusing to countersign legislation. During his presidency (1996–2016), Grímsson showed clear signs of wanting to expand the political role of the president and became the first president to activate a clause in the constitution that grants the president the right to refer new legislation to a binding referendum. The president’s action was highly controversial—with the controversy centring in large part on a clause in the constitution that states that the cabinet exercises the powers of the president. The government’s decision to withdraw the legislation without holding a referendum may have been intended to undermine the president’s claim to being able to exercise this power independently but, in effect, it merely served to acknowledge the president’s power. President Grímsson went on to refer legislation to a referendum another two times (in 2010 and 2011); in both instances the legislation concerned an agreement regarding repayment of deposits in the United Kingdom and the Netherlands covered by deposit insurance that the British and the Dutch authorities had taken on after the collapse of the Icelandic banks. Voters rejected the agreements in both instances. The importance of the president expanding his political role by reinterpreting ambiguous constitutional clauses for coalition governance is fairly obvious (Protsyk 2005). Whereas previously government coalitions could form without paying heed to the political preferences of the president, the ‘expanded’ powers of the president imply that the president now has an effective legislative veto, provided he can rally voters against the legislation. In the presence of a politically assertive president, the political parties may want to take the president’s preferences into account in forming coalitions—in particular if there are issues on the horizon on which voters are divided.

¹ Indridason (2005) discusses the implication of the move from a bicameral to unicameral legislature for government formation. See also Druckman et al. (2005) for government formation in bicameral systems.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

327

Whether this actually does, or will, impact coalition formation remains an open question—so far, there have been no indications of the parties seeking to build coalitions to include the president. Moreover, it is not clear whether future presidents will subscribe to the model of a political presidency that Grímsson adopted. The question about what the role of the president should be was somewhat salient in the last presidential elections (in 2012, 2016, and 2020) with some candidates leaving open the possibility of exercising the presidential power while others vowed to revert to the traditional, ceremonial, role of former presidents. Overall, the elections were focused far more on the personal qualities of the candidates rather than any political agenda that the candidates might have. With Grímsson leaving office in 2016, Iceland’s transition to a true semipresidential system appears to have been stalled. The newly elected president, Jóhannesson, advocates popular initiatives as a method of introducing referenda rather than presidential intervention but wants to keep the presidential option open in case of unforeseen and exceptional circumstances. Grímsson’s interpretation of the president’s constitutional powers in 2004 placed constitutional reform on the agenda. Those efforts appeared to have fizzled out by 2008 when the economic collapse brought constitutional reform back on the agenda. A constitutional council produced a draft of a new constitution that has been at the centre of demands for a comprehensive institutional reform after the economic crash. The Social Democratic Alliance (SDA)–Left Green (LG) coalition government, however, opted not to press on with constitutional reform ahead of the 2013 election. The constitution reform issue had little impact on the outcome of the 2013 election although it was prominent in the campaigns of some of the new parties that failed to gain representation. In 2016, however, it played a significant role in the Pirate Party’s populist platform—as a prime example of the established political elites paying no heed to the wishes of the people as expressed in a consultative referendum on the (failed) attempt at revising the constitution in 2012. The Pirate Party was, however, far from being the only party advocating constitutional reform—all left-wing parties, though divisions existed within them, expressed a willingness to move forward, while the parties on the right took a more conservative stance, favouring a more limited reform than that produced by the constitutional council.

The party system and the actors Party system change The year of the financial meltdown in Iceland, 2008, marks a neat dividing line in the political landscape. The economic crash altered the political discourse and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

328

 .     

affected public confidence in politicians and political institutions. Overnight, the tone of politics changed from drowsy neoliberal euphoria to one of anger and moral righteousness. Instead of privatization, growth, and further expansion of Icelandic businesses at home and abroad came a number of thorny issues relating to austerity and how to deal with international organizations and claimants on the estates of the fallen banks that the government had taken over. There was also a strong demand for settling accounts with the culprits and to introduce a new and improved political order. Party politics during the first half of the twentieth century were shaped by a fairly simple cleavage structure where left–right differences played a primary role, while the urban–rural divide provided an important second dimension. After the Second World War, foreign policy added an important third dimension. Thus, the conservative IP represented the urban middle classes and part of the farming communities while the Social Democrats (SDP) and the left socialists (Communist Party, CP; United Socialist Party, USP; People’s Alliance, PA) represented the urban working class. The PP represented the rural areas and sparsely populated regions with flexible preferences concerning left–right governments. Despite declining electoral strength, its centrist location on the left–right dimension has often given it a pivotal position when it comes to coalition formation. The left socialists were strongly opposed to Icelandic participation in Western security cooperation while opposition to market integration in Europe had a broader appeal, not least in the rural areas and periphery. Iceland hesitatingly joined the European Economic Area in 1994 while none of the parties advocated full membership of the European Union at the time. Iceland entered the twenty-first century with five parties in parliament, namely the IP, PP, SDA (an amalgamation of the left-wing forces of the 1990s), the LeftGreens (LG, a new left-socialist green party representing similar policies as the PA had done) and the Liberal Party (LP) (a small splinter group from the IP with different views on fisheries policies). The IP-led coalition of the IP and PP, formed in 1995, was relatively secure although opinion polls indicated a potential threat from the SDA. The coalition’s majority, nonetheless, survived the 2003 election, but after further losses by the PP in 2007, the IP decided to change partners and formed a coalition with the SDA. If we count the SDA and LG as direct descendants of the established left-wing forces, the established party system showed little signs that frozen cleavages were ‘thawing’ until after the crash. Since 1971 the occasional fifth or sixth party has gained representation but, even though traditional class voting has probably been on the decline since the 1950s, the established parties continued to poll a combined share of around 90 per cent in most parliamentary elections. The crash, however, not only shook confidence in the system profoundly but also set large groups of voters afloat (Önnudóttir et al. 2017). Signs of stronger challenges to the established order may be observed, for example in an increasing number of parties

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

329

contesting elections, greater fluctuations in party support, and more support for new parties. The first elections following the crash, however, indicated a strong swing to the left among voters, that is to those of the established parties considered least accountable for the banking expansion leading to the first ever clear socialist majority in parliament (Indridason 2014). In addition, a left-wing populist movement claiming to represent the ‘pots and pans revolution’ obtained seven per cent of votes. Amidst lofty expectations and rather far-reaching promises, the realities of crisis management and austerity, however, quickly undermined the popularity of the left government, which even experienced great difficulties in keeping its own members of parliament (MPs) in line. In fact, by the end of 2010, it was rapidly losing control over its own forces in parliament. The left government’s lack of popularity and its loss of credibility in the highly controversial Icesave issue (the settling of foreign claims following the fall of the Landsbankinn) led to a major swing in the election of 2013, partly to the PP (which, in addition to having opposed compromises on the Icesave issue all along, made extensive promises to relieve household debts) and partly to new political forces (Indridason et al. 2017). A record number of 15 parties contested the 2013 election. The combined vote share of new parties was 25 per cent, although only two of them obtained seats in Alþingi (Bright Future, BF, and Pirates). Both can be regarded as protest parties, emphasizing improved and more democratic procedures. Large vote fluctuations continued in the parliamentary elections of 2016 and 2017, with the SDA entering a deep crisis while new parties such as the Pirates, the People’s Party, and Reform managed to establish themselves in parliament. Bright Future, on the other hand, lost all its seats in 2017. Although the economic crash in 2008 put the traditional four-party system to the test, what was perhaps most remarkable was how well the four-party system survived the test in the first two elections following the crash. New parties won only a limited number of seats and the reins of government remained in the hands of the established parties. This was in part due to the 2009 election being an early election that allowed the new political forces limited opportunity to organize, campaign, and coordinate (in particular, electorally) and the 2013 election was dominated by economic issues on which the traditional parties could claim ownership (Grittersova et al. 2016). This was no longer the case in the 2016 election—another early election held against the backdrop of the Panama papers scandal that saw prime minister (PM) Gunnlaugsson leave office but also in more stable economic conditions. With the economy having less salience, the seeds of the political dissatisfaction sown by the economic crash finally came to upend the party system. The biggest beneficiary was the Pirate Party that polled at close to 40 per cent half a year ahead of the election although it won a more modest 14.5 per cent in the election, which was still enough to more than triple its representation in Alþingi (from 3 to 10 seats). At the same time, two of the established parties were decimated—the PP and the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

330

 .     

SDA lost more than half their vote share from 2013. Between the two parties, the SDA was undeniably the bigger loser as the PP had won big in 2013 while the 2013 election had also been a disaster for the SDA. Between 1999 and 2009 the SDA had hovered around 30 per cent of the vote—in 2016 it won a measly 5.7 per cent. The other notable change in the party system was the emergence of the Reform Party, which aimed at siphoning off votes from the IP by offering a pro-European, centre-right platform. The Reform Party won 10.5 per cent of the vote, which is the second-best result for a party contesting an election for the first time (i.e. excluding parties formed in a merger of existing parties). Seven parties were thus represented in Alþingi after the 2016 election. Unlike in many of the other European parliamentary systems, no significant right-wing populist party has emerged in Iceland. Although some of the parties have occasionally flirted with right-wing populism (e.g. with regard to immigration), the elite consensus to avoid conflict on this front seems to have held for the most part. Voter demand for such policies also remains small, judging by the 0.2 per cent of votes obtained by the Icelandic National Front in the parliamentary election of 2016. We can only speculate on why this is the case. For one thing, Iceland is not a member of the European Union and has maintained fairly restrictive immigration policies towards people outside the European Economic Area. For another, there is no structural unemployment, despite the crash, and foreign labourers are generally not seen as competing with the domestic work force for highly valued jobs. In the absence of single-party majorities, there is a strong preference among Icelandic politicians for minimal winning coalitions, preferably with a strong majority in parliament. Table 10.1a illustrates the historical pattern.² No threeway coalition in Iceland has lasted for an entire term and the experience of political instability during the 1970s and 1980s has fostered the belief that coalitions involving many parties and small majorities should be avoided. Accordingly, all governments between 1991 and 2016 were two-party coalitions. Indeed, the period from 1991 to 2008 may be regarded as one of unusual stability in many respects with stable two-party governments, all of which lasted a full term until the ‘pots-and-pan’ revolution brought down the coalition of the IP and SDA with a great deal of noise in January 2009. Coalition formation in 1991–2016 followed a relatively simple pattern. With one exception, only minimal winning and ideologically connected coalitions were formed. In all instances, two-party coalitions formed with relative ease in about two to four weeks. Sigurðardóttir I, however, was a caretaker SDA–LG minority government formed in 2009 after the pots and pans revolution with neutrality

² Sources of data are Indridason (2005) and Önnudóttir and Harðarson (2018). Retrieved from: https://www.stjornarradid.is/rikisstjorn/sogulegt-efni/rikisstjornartal (Accessed on 12 May 2018) and https://hagstofa.is/talnaefni/ibuar/kosningar/ (Accessed on 12 May 2018).

1944-10-21 1946-06-30 1947-02-04 1949-12-06 1950-03-14 1953-09-11 1956-07-24 1958-12-23 1959-06-28 1959-11-20 1963-06-09 1963-11-14 1967-06-11 1970-07-10 1971-07-14 1974-08-28 1978-09-01 1979-10-15 1980-02-08 1983-05-26 1987-07-08 1988-09-28 1989-09-10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Thors II Thors III Stefansson Thors IV Steinthorsson Thors V Jonasson III Jonsson I Jonsson II Thors VI Thors VII Benediktsson I Benediktsson II Hafstein Johannesson I Hallgrimsson Johannesson II Gröndal Thoroddsen Hermansson I Palsson Hermansson II Hermansson III

Date in

Cabinet Cabinet number

1979-12-03 1983-04-23 1987-04-25

1971-06-13 1974-06-30 1978-06-25

1967-06-11

1959-06-28 1959-10-26 1963-06-09

1953-06-28 1956-06-24

1949-10-24

1946-06-30

IP, SP, SDP IP, SP, SDP SDP, PP, IP IP PP, IP IP, PP PP, PA, SDP SDP SDP IP, SDP IP, SDP IP, SDP IP, SDP IP, SDP PP, PA, ULL IP, PP PP, PA, SDP SDP IP, PA, PP PP, IP IP, SDP, PP PP, PA, SDP

sur sur sur min mwc mwc mwc min min mwc mwc mwc mwc mwc mwc mwc mwc min sur mwc mwc min mwc

71.2 75 80.8 36.5 69.2 71.2 63.5 15.4 11.5 55 53.3 53.3 53.3 53.3 53.3 70 66.7 23.3 81.7 61.7 65.1 49.2 60.3

52 52 52 52 52 52 52 52 52 60 60 60 60 60 60 60 60 60 60 60 63 63 63

4 4 4 4 4 5 4 4 4 4 4 4 4 4 5 5 4 4 5 6 7 7 7

3.49 3.61 3.61 3.47 3.47 3.44 3.48 3.48 3.2 3.44 3.33 3.33 3.48 3.48 3.85 3.38 3.85 3.85 3.78 4.06 5.34 5.34 5.34

Election date Party Type Cabinet Number of Number of ENP, parties in parliament composition of strength seats in of cabinet cabinet in seats parliament parliament (%)

Table 10.1a Icelandic cabinets since 1944

PP PP PP PP PP PP PP PP PP PP PP PP PP PP PP PP PP PP PP PP PP PP PP

Median party in first policy dimension

Continued

Formal support parties

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1991-04-30 1995-04-23 1999-05-28 2003-05-23 2004-09-15 2006-06-15 2007-05-24 2009-02-01 2009-05-10 2013-05-23 2016-04-07 2017-01-11 2017-11-30

24 25 26 27 28 29 30 31 32 33 34 35 36 2016-10-29 2017-10-28

2009-04-25 2013-04-27

2007-05-12

1991-04-20 1995-04-08 1999-05-08 2003-05-10

PP, PA, SDP, CP IP, SDP IP, PP IP, PP IP, PP PP, IP IP, PP IP, SDA SDA, LG SDA, LG PP, IP PP, IP IP, RP, BF LG, IP, PP mwc mwc mwc mwc mwc mwc mwc min mwc mwc mwc mwc mwc

57.1 63.5 60.3 54 54 54 68.3 42.9 54 60.3 60.3 50.8 55.6

63 63 63 63 63 63 63 63 63 63 63 63 63

5 6 5 5 5 5 5 5 5 6 6 7 8

3.78 3.95 3.45 3.71 3.71 3.71 3.62 3.62 4.18 4.42 4.42 5.09 6.54

Election date Party Type Cabinet Number of Number of ENP, parties in parliament composition of strength seats in of cabinet cabinet in seats parliament parliament (%)

PP PP PP PP PP PP PP PP PP PP PP PP PP

Median party in first policy dimension

PP

Formal support parties

Notes: For a list of parties, consult the chapter appendix. a = Technocrat minister majority; b = Technocrat PM; c = Caretaker government The first policy dimension is economic left–right. Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1945–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015). Cabinet types: min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition Table legends and variables are further defined in the Appendix, this volume.

Oddsson I Oddsson II Oddsson III Oddsson IV Asgrimsson Haarde I Haarde II Sigurdardottir I Sigurdardottir II Gunnlaugsson I Johannsson I Benediktsson I Jakobsdottir

Date in

Cabinet Cabinet number

Table 10.1a Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

333

from the PP until a new election could be held. The coalition of the same parties formed after the election in the spring of 2009 started out as a majority coalition but lost its majority towards the end of 2012. The fragmentation of the party system in the parliamentary elections of 2016 and 2017 can be seen as a belated effect of the crash in 2008. With seven parties represented in parliament and no possibility of a two-party coalition, a prolonged period of coalition negotiations followed. After negotiations following both elections, three-party coalitions that were not ideologically connected were formed.³ Especially noteworthy was the coalition formed in November 2017 containing the parties furthest to the left (LG) and right (IP) as well as the centre (PP) with LG leader Jakobsdóttir as the PM.

Electoral alliances and pre-electoral coalitions As shown in Table 10.1b, pre-electoral coalitions and electoral alliances are rare in Iceland and the coalition game a relatively open-ended one.⁴ The main rationale of electoral alliances is usually to prevent the waste of votes in the case of small parties while pre-electoral coalitions aim at presenting credible government alternatives (Golder 2006). The larger parties in Iceland, however, may see the electoral system as protection against new challengers and hence are not likely to encourage minor parties at their fringe through electoral alliances. Most often, electoral alliances have been a prelude to a full merger of the constituent parties. This was the case, for example, with the SDA electoral alliance in 1999, which subsequently became a full-fledged political party, aiming to be a left-wing counterweight to the IP on the right. Table 10.1b Electoral alliances and pre-electoral coalitions in Iceland, 1987–2018 Election date

Constituent parties

Type

1991-04-20 1999-05-08

NP, HP SDP, WM, PA, THPM

EA EA

Types of pre-electoral commitment

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other

³ One may, however, question whether the IP–R–BF coalition should be considered not to be a connected coalition as the relative positions of BF and the PP on the left–right dimension are not completely clear. ⁴ Information on electoral alliances comes from Kristinsson (2000).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

334

 .     

Regarding pre-electoral coalitions, the Icelandic political parties rarely state clearly their intentions with regard to post-electoral government formation. The concern with pre-electoral coalitions is that they may influence the perceptions of voters on the policy positioning of their parties and potentially alienate those furthest from the ‘pact’ mean position.⁵ This may be a problem for both large and small parties. From the point of view of small parties, moreover, participation in a pre-electoral coalition may weaken their bargaining position during negotiations after the election. A smaller party with a free hand can be pivotal in coalition formation and obtain a very strong bargaining position. Hence, the absence of pre-electoral coalitions has contributed to strong overcompensation of smaller parties in terms of government portfolios. In most cases, cabinet positions are divided equally in two-party coalitions, and smaller parties are overcompensated in larger coalitions.

Government formation The bargaining process The government formation process in the period under consideration was in most cases a relatively simple affair, as may be seen in Table 10.2. While in some cases it was uncertain which parties would end up in government in the immediate aftermath of the election (e.g. 1991, 1995, and 2007), the party leaders formed coalitions relatively quickly and decisively. In other cases, government formation was relatively predictable (1999, 2003, 2009, and 2013) and some cabinet changes were primarily formalities as one PM took over from another without renegotiation of the coalition agreement (e.g. Ásgrímsson I). Following the more complex coalition formation in 1987 and 1988, the relative simplicity of government formation in 1991–2016 was greatly facilitated by the possibility of forming majority coalitions with only two parties. As the party system became more fragmented in 2016 and 2017, two-party coalitions were no longer an option. The parties, therefore, had to weigh the relative costs of three negative factors, that is the need to compromise on policy, a greater number of partners, and narrow parliamentary majorities. This proved difficult for the parties, as can be seen by the larger number of bargaining rounds and the more drawn-out government formation processes following these elections. After the election of 2016, the IP opted for a coalition with the smallest possible number of parties and a modest policy compromise but a parliamentary majority of only one. After the election the following year, the LG, IP, and PP opted for the smallest

⁵ See, however, Gschwend et al. (2017).

Year in

1987

1988

1989 1991

1995

1999 2003 2004 2006 2007 2009 2009 2013 2016

Cabinet

Palsson

Hermansson II

Hermansson III Oddsson I

Oddsson II

Oddsson III Oddsson IV Asgrimsson Haarde I Haarde II Sigurdardottir I Sigurdardottir II Gunnlaugsson I Johannsson I

0 0 0 0 0 0 0 0 0

1

0 0

3

5

Number of inconclusive bargaining rounds IP, SDP, PP (1) PP, ? (2) IP, SDP, WA (3) SDP, PA, WA* (4) PP, CP* (5) SDP, IP, PP PP, PA, SDP (1) PP, SDP, PA (2) PP, SDP, PA, CP (3) IP, ? PP, PA, SDP, CP IP, SDP (1) PP, SDP, PA IP, PP (1) IP, SDP IP, PP IP, PP IP-PP PP, IP IP, PP IP, SDA SDA, LG SDA, LG PP, IP PP, IP

Parties involved in the previous bargaining rounds

Table 10.2 Cabinet formation in Iceland, 1987–2018

6 6 14 22 0

27 2 7 1 1 0 5 3 5 8 17 10

7 3 14

Bargaining duration of individual formation attempt (in days)

20 13 0 0 12 6 15 26 0

15

0 10

11

74

Number of days required in government formation

Continued

6 6 14 22 0

17 10

(13?)

0 (8?)

Total bargaining duration

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2017

2017

Benediktsson I

Jakobsdottir

1

5

Number of inconclusive bargaining rounds IP, RP, BF (1) IP, RP, BF (2) LG, PP, SDA, RP, BF (3) IP, LG (4) RP, PP, BF, SDA (5) PP, RP, BF, SDA, LG LG, IP, PP (1) LG, SDA, BF, Pir

Parties involved in the previous bargaining rounds 25 14 8 3 2 11 21 7

Bargaining duration of individual formation attempt (in days)

33

73

Number of days required in government formation

(28?)

69

Total bargaining duration

Note: Negotiations marked * were informal and occurred without a change of formateur. A question mark indicates that no information is available about which parties were involved in the negotiations. Duration is difficult to establish but perhaps no more than a day might be a fair estimate.

Year in

Cabinet

Table 10.2 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

337

number of parties and a much larger parliamentary majority while accepting a substantial policy compromise as the coalition covered the whole left–right spectrum, from the party furthest to the left (LG) to the one furthest to the right (IP). Coalition bargaining, according to our interviews, is usually a highly elitist affair with the party leaders at centre stage. The deputy leaders sometimes play a role and the personal assistants of the party leaders seem to be involved to an increasing degree as well. Others are rarely involved. The only document produced during the negotiations is the coalition agreement and while the relevant party organs must formally ratify the outcome of the negotiations, the decision whether or not to form the government is effectively that of the party leader. The only known instance of a more extensive negotiation process, reaching beyond the top leadership level, is the preparation of the coalition agreement after the election of 2009. These negotiations appear to have been based on broad consultations involving a relatively large number of people, including experts, and can be attributed to the highly unusual context of the election being held in the middle of an economic crisis. It also appears that the coalition parties had ambition to build their cooperation on a more comprehensive coalition agreement that reflected a break with the traditional ministerial form of government.

The allocation of ministerial portfolios The number of ministries was set by law between 1970 and 2011, as the Ministry Act listed the ministries in existence and limited the number of ministers to one minister per ministry. This set an upper limit to the number of ministers although to begin with it was quite common for ministers to oversee more than one portfolio. Strong demand for ministerial positions created pressure to enlarge the cabinet, which grew from seven prior to 1971 to twelve in the 2000s. Making cabinet positions more easily available provides party leaders with rewards to parliamentarians for good behaviour, which is likely to have contributed to party cohesion (see e.g. Kristinsson 2011; Indridason and Bowler 2014). The allocation of portfolios to parties in two-party coalitions since 1991 almost always resulted in an even number of portfolios for each party. The two deviations were related to special agreements, one that handed the post of PM to the smaller partner for part of the term (i.e. the PP in 2005) and another that compensated the party not receiving the PM post part of the term (the IP in 2013).⁶ This reflects the tendency—mentioned earlier—to overcompensate minor partners in terms of portfolio allocations in Iceland, which is consistent with earlier history as well. ⁶ In the beginning the IP had five ministerial posts compared to the PP’s four. At the end of December 2014, however, the number of PP ministers was increased to five.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

338

 .     

Formally, the allocation of portfolios is discussed only at the end of the negotiation process when policy differences have been settled. This seems to indicate that policy comes before portfolio. Things may not be this simple, however. In many cases the preferences of the parties for specific portfolios are known to participants beforehand, for example on account of connections to specific interest groups or election pledges. Left-wing parties are more likely to obtain welfare and service-related ministries while those farther to the right are more concerned with the economy (Bäck et al. 2011). Education and culture are in particular demand among the parties furthest to the right and the left. Personal considerations may also play a role. Table 10.3 illustrates the partisan distribution on the five cross-nationally important ministries.⁷ The posts of PM and to a lesser extent minister of finance are, moreover, highly valued by all parties and as a rule no coalition partner expects to get both. The post of minister of foreign affairs used to be included among the top-ranking ministerial posts but seems less valued in recent years. However, all three posts have not fallen to the same party in a coalition government since 1971–1974. In two-party coalitions, the party of the PM usually gets either the finance or the foreign affairs portfolio as well, the only exception being in 2005 when the larger partner (IP) was compensated for conceding the PM post with both the other posts. Usually the largest party in the coalition gets the post of PM, but the smaller parties have on occasions been able to leverage their strategic position into obtaining the post as the PP did in 2004 and the LG in 2017. Hence, the matrix of portfolio allocations may not be as enigmatic as it is often portrayed to be. While the formal allocation of portfolios is left to the end, the parties gradually develop a picture of what the end result is likely to be, and this may shape the negotiations over policy.

Coalition agreements Coalition agreements in Iceland (cf. Table 10.4⁸) tend to be relatively short by international standards. Usually titled ‘policy declarations’, they deal mainly with major points of policy in broad terms and seldom with procedural matters. Their brevity reflects the general preference of many politicians for dealing with political issues on an ongoing basis, where parliament plays a significant role, as well as the strong emphasis on ministerial government and ministerial autonomy. Coalition

⁷ Data obtained from https://www.stjornarradid.is/rikisstjorn/sogulegt-efni/rikisstjornartal/ (Accessed on 12 May 2018). ⁸ Information on coalition agreements is from Indridason (2005) and authors’ analysis of subsequent coalition agreements.

Year in

1987 1988 1989 1991 1995 1999 2003 2004 2006 2007 2009 2009 2013 2016 2017 2017

Cabinet

Palsson Hermansson II Hermansson III Oddsson I Oddsson II Oddsson III Oddsson IV Asgrimsson Haarde I Haarde II Sigurdardottir I Sigurdardottir II Gunnlaugsson I Johannsson Benediktsson Jakobsdottir

4 IP, 4 PP, 3 SDP 3 PP, 3 PA, 3 SDP 4 CP, 3 PP, 3 PA, 3 SDP 5 IP, 5 SDP 5 IP, 5 PP 6 IP, 6 PP 6 IP, 6 PP 7 IP, 5 PP 6 IP, 6 PP 6 IP, 6 SDA 4 SDA, 4 LG, 2 Ind. 5 SDA, 5 LG, 2 Ind. 5 IP, 4 PP 5 PP, 5 IP 6 IP, 3 RP, 2 BF 4 IP, 3 LG, 3 PP

Number of ministers per party (in descending order) 11 9 11 10 10 12 12 12 12 12 10 12 9 10 11 11

Total number of ministers 13 14 15 15 15 15 13 13 13 12 10 12 10 10 11 11

Number of ministries

Table 10.3 Distribution of cabinet ministerships in Icelandic coalitions, 1987–2018

IP PP PP IP IP IP IP PP IP IP SDA SDA PP PP IP LG

1 Prime minister SDP PA PA IP IP IP IP IP IP IP LG LG IP IP RP IP

2 Finance

PP SDP SDP SDP PP PP PP IP PP SDA SDA SDA PP PP IP IP

3 Foreign affairs

SDP SDP SDP SDP PP PP PP PP PP SDA SDA SDA PP PP RP PP

4 Social affairs

SDP PP CP IP IP IP IP IP IP IP Ind. Ind. IP IP IP IP

5 Justice OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Year in

1944 1947 1950 1953 1956 1959 1967 1971 1974 1978 1980 1983 1987 1988 1989 1991 1995 1999 2003 2004 2006 2007 2009 2009 2013 2016 2017 2017

Coalition

Thors II Stefansson Steinthorsson Thors V Jonasson III Thors VI Benediktsson II Johannesson I Hallgrimsson Johannesson II Thoroddsen Hermansson I Palsson Hermansson II Hermansson III Oddsson I Oddsson II Oddsson III Oddsson IV Asgrimsson Haarde I Haarde II Sigurdardottir I Sigurdardottir II Gunnlaugsson I Johannsson I Benediktsson Jakobsdottir

1,145 950 600 571 697 319 2,200 2,300 942 1,750 2,750 1,700 4,900 3,850 2,700 750 1,800 2,100 1,786 1,786 1,786 2,393 1,101 6974 3907 3907 2955 6128

Size

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

General rules (in %) 2 2 3 6 4 0 0 0 0 2 1 0 0 0 0 0 0 1 0 0 0 0 0 2 0 0 1.5 0.2

Policy-specific procedural rules (in %)

Table 10.4 Size and content of coalition agreements in Iceland, 1944–2018

0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 3 0 0 0 0 0 0 0 0 0 0

Distribution of offices (in %) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of competences (in %) 98 98 97 94 96 100 100 100 100 98 99 100 100 100 99 100 100 96 100 100 100 100 100 98 100 100 98.5 99.8

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

341

agreements are thus recognized as incomplete contracts by the coalition partners that need fleshing out on an informal basis during the lifespan of the coalition. Following the economic crash in 2008, however, coalition agreements have become increasingly elaborate, which may indicate willingness to move away from ministerial government to a more collective norm. However, in line with Bowler et al. (2016), when there is either substantial policy disagreement between coalition partners or uncertainty with regard to the type of compromise that may be reached, the policy declarations sometimes designate a procedure to be followed rather than deciding on a specific option. This may involve prioritizing an issue area, setting up an investigation, or improving policy-relevant procedures. In the case of Iceland’s application for membership of the European Union in 2009, the policy declaration stated that following an eventual agreement on membership the issue should be put to a referendum where the right of the coalition partners to advocate distinct positions should be respected.

Coalition governance The role of individual ministers in coalition governance In terms of the three models of governance discussed in the introduction of this volume, Iceland resembles the ministerial government model more than the coalition compromise or the dominant PM model. Icelandic cabinets grant ministers a substantial degree of autonomy within their portfolios. The minister, according to Icelandic law and practice, is the highest executive within his or her portfolio and is neither subordinate to the PM nor the cabinet. Ministers consider themselves to have a relatively large degree of autonomy in conducting matters according to their own will within their portfolio regarding administrative matters, policy priorities, and even in making public statements relevant to public policy. Only with regard to introducing government bills in Alþingi do they require the formal approval of the cabinet. Ministerial autonomy sometimes causes complications with regard to accountability. Although the ministers may formally be within their rights to act in a particular manner, they owe their position to the legislative majority that may not always be happy with their actions. The practice of ministerial government has been strongly shaped by the IP and PP and, according to our interviews, still has considerable backing in these parties. The economic crash, however, and subsequent investigations into its causes (e.g. by the Special Investigation Commission) suggested that the lack of coordination within government was one of the factors responsible. This, and subsequent developments, has put advocates of ministerial government on the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

342

 .     

defensive to a certain extent. In the Ministry Act of 2011, an attempt was made to clarify the role of cabinet meetings. According to the Act, the duty to inform the cabinet applies to government bills, other matters that require presidential confirmation (e.g. government-initiated parliamentary resolutions), ‘important governmental matters’ (e.g. regulations, important policy declarations, or declarations containing budgetary commitments), and information that may be seen as relevant to the cabinet as a whole. The duty to inform the cabinet, however, does not formally infringe on the autonomous decision-making powers of the minister in most cases. Thus, over the past decade, there have been attempts to edge coalition governance closer to the coalition compromise model, but it remains to be seen how effect and permanent those changes have been. The Icelandic PM has relatively few formal powers. He or she chairs cabinet meetings and, if ministerial jurisdictions collide, the PM adjudicates. Formally the PM selects all government ministers but in effect this is recognized to be entirely the decision of each coalition party. The PM can initiate an early election, which calls for a presidential countersignature, and most of our interviewees believed that formal agreements on consent by coalition partners had not been made. It is, however, extremely rare for PMs to use this power, and Gunnlaugsson’s attempt in 2016 to use it for political advantage, in fact, failed. The most important of the PM’s informal powers for running the cabinet probably derives from her role as party leader and her ability to set the agenda and interpret the results of cabinet meetings. Different PMs, however, may wield their informal power with varying amount of skill and their personality may also play a role in how effectively they employ those powers. Overall, there is little to suggest that the dominant prime minister model of coalition governance has any purchase in Iceland. The constitutional division of labour between the PM and the president has been the source of some confusion. Most of the tasks of the president de facto belong to the PM in accordance with paragraph 13 of the constitutions, which states that ministers execute the powers of the president. According to the Ministry Act of 2011, the PM has a special duty to clarify ministerial jurisdictions when necessary and to coordinate the actions of ministers. The PM can appoint cabinet committees with cabinet approval and chairs their meetings or appoints another minister to that role. The only instance in Icelandic history where an impeachment case has been brought against a minister—that is in the case against former PM Haarde following the economic crash—saw the former PM charged with negligence in the face of the impending financial crisis and neglecting to call cabinet meetings on important governmental issues. Haarde was acquitted of the former charge since it was not proven that he could have done anything to avert the crisis but convicted of the latter.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

343

Governance mechanisms in the executive arena Table 10.5 outlines the combination of coalition governance mechanisms that are used.⁹ The main mechanism for coordination in the Icelandic cabinet is the cabinet meeting, which usually takes place once or twice per week. A more formal arena, the Council of State, which the president also attends, takes place only two or three times per year and serves an entirely ceremonial function. The cabinet meeting, by contrast, is often the venue for strategizing about policy and political tactics and its proceedings are confidential. The PM sets the agenda for cabinet meetings and individual ministers can present issues (e.g. legislative proposals) that they wish to discuss. The agenda is, however, often set fairly late, and documents are only distributed shortly before the meeting, often leaving ministers limited opportunities to prepare for issues presented by others. They are, however, assisted by ministerial assistants as well as public officials from the ministries and, after the economic crash, increasing efforts have been made to improve coordination below the cabinet level. The main tasks of officials, however, are focused on their own ministries while ministerial assistants tend to focus on political networking and publicity more than substantive policy issues (Kristinsson 2016). Junior ministers do not exist in the Icelandic system, neither as ministers’ pets nor watchdogs (Thies 2001). Ministerial assistants are not formally part of the ministerial hierarchy, but being placed close to the minister, their influence can in some cases be significant. In an effort to strengthen ministerial control, the number of ministerial assistants was increased in the Ministry Act of 2011, while at the same time, the number of ministries was reduced, in an attempt to simplify inter-ministerial coordination. The ministerial assistants appear to play a relatively small role in inter-party coalition governance, according to our interviews, with the possible exception of the PM’s assistant. Formally there exists no inner cabinet in Iceland and the full cabinet plays a central role in government decision-making. The personal relationship between the party leaders, however, is generally recognized as playing an important role in how smoothly the government leadership functions.

Governance mechanisms in the parliamentary arena Once government bills have passed the cabinet stage, they are presented in the parliamentary groups of the coalition parties before being formally submitted in parliament. Although bills are introduced in each of the coalition parties’ parliamentary groups, the minister is present only in the meeting of his or her own party

⁹ See Indridason (2005). Additional information obtained in interviews with former ministers.

Thors II Thors III Stefansson Steinthorsson Thors V Jonasson III Thors VI Thors VII Benediktsson I Benediktsson II Hafstein Johannesson I Hallgrimsson Johannesson II Thoroddsen Hermansson I Palsson Hermansson II Hermansson III Oddsson I Oddsson II Oddsson III

Coalition

1944 1946 1947 1950 1953 1956 1959 1963 1963 1967 1970 1971 1974 1978 1980 1983 1987 1988 1989 1991 1995 1999

POST N POST POST POST POST POST N N POST N POST POST POST POST POST POST POST POST POST POST POST

Y N/A Y Y Y Y Y N/A N/A Y N/A Y Y Y Y Y Y Y Y Y Y Y

N N/A N N N N N N/A N/A N N/A N N N N N N N N N N N

IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC

IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC

IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC

All Most For most used common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Table 10.5 Coalition governance mechanisms in Icelandic coalitions, 1944–2018

-

-

Personal Issues union excluded from agenda

Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most

Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y

Varied Varied Varied Few Few Few Few Few Few Varied Varied Comp. Varied Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp. Comp.

N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A

N N N N N N N N N N N

N

N N N N N

N

Coalition Freedom of Policy Junior Nondiscipline appointment agreement ministers cabinet in (watchdogs) positions legislation/ other parl. behaviour

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2013

2016

2017

2017

Gunnlaugsson I

Johannsson I

Benediktsson

Jakobsdottir

POST

POST

POST

POST

POST POST POST POST POST POST

Y

Y

Y

Y

Y Y Y Y Y Y

N

N

N

N

N N N N N N

IC CoC IC IC IC, CaC IC, CaC

IC, CaC, CoC CoC IC, CaC, CoC CoC, IC, CaC, CaC CoC N/A N/A

IC CoC IC IC IC IC, CaC CoC

N/A

IC, CaC

IC, CaC

IC, CaC

IC, CaC IC, CaC IC, CaC IC, CaC IC IC, CaC

Y

Y

Y

Y

Y Y Y Y Y Y

N

N**

N

N

N N N N N N

All/All

All/All

All/All

All/All

All/All All/All All/All All/All All/All Most*/All

Y

Y

Y

Y

Y Y Y Y Y Y

Comp.

Comp.

Comp.

Comp.

Comp. Comp. Comp. Comp. Varied Comp.

N/A

N/A

N/A

N/A

N/A N/A N/A N/A N/A N/A

N/A

Y

N/A

N/A

N/A N/A N/A N/A N/A N/A

Notes: * MPs allowed to vote according to their own conviction on membership, following negotiations of a draft treaty ** Matters regarding the European Union would not be brought up until the third year Coalition agreement: POST = post-election; N = no coalition agreement Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee; CoC = Coalition committee Coalition discipline: All = Discipline always expected; Most = Discipline expected except on explicitly exempted matters Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement

2003 2004 2006 2007 2009 2009

Oddsson IV Asgrimsson Haarde I Haarde II Sigurdardottir I Sigurdardottir II

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

346

 .     

and the coalition partner’s parliamentary groups get less of an opportunity to have an impact at this stage. A legislative bill that meets with strong opposition in the parliamentary groups may need further preparation by the minister or the cabinet. The main opportunity for ordinary MPs to influence legislation is through the parliamentary committees. The parliamentary committees play a central role in parliamentary scrutiny of legislative bills and they intervene to a much greater extent than is customary in the neighbouring Nordic states. During 2003–2009, committee chairs were the ministers’ co-partisans in about 83 per cent of cases and were, thus, likely to cooperate with ministers on the introduction of amendments. Nonetheless, the preparation of legislation in Iceland takes place to a relatively great extent at the parliamentary stage where various experts, the opposition, and stakeholders are often consulted for the first time. Instead of strengthening the pre-submission stage of the policy-making process, the response to the economic crash was primarily to strengthen the role of the parliamentary committees—partly through re-organization intended to break up the match between committee and ministry subject areas and partly through greater use of watchdog committee chairs (Indridason and Kristinsson 2018). Amendments of government bills in parliament have increased after the crash. Party cohesion in the Icelandic Alþingi is generally high and comparable to the other Nordic countries (Kristinsson 2011). Contrary to expectations in some of the literature (Hazan and Rahat 2010), a relatively inclusive and decentralized process of candidate nominations in Iceland has not resulted in weaker party cohesion (Indridason and Kristinsson 2015). Party cohesion tends to be greater among government parties and when party leaders play a greater role in distributing positions of influence within the parliamentary group. After the crash, however, party cohesion appears to have declined somewhat, reflecting a more uncertain political situation and weakening of the established parties. Alþingi is in many ways an unusually influential assembly for a parliamentary system with relatively little limitations on the rights of members to speak, introduce bills, propose amendments, put questions to ministers and ask for reports, or raise issues outside the agenda. The permanent parliamentary committees are also very powerful and a strong influence on legislation. The traditionally strong position of the Alþingi may be among the reasons for the emphasis on majority governments. Minority governments enjoy few privileges against such a strong assembly. Weak control over the parliamentary agenda resulted in government bills constituting only 57 per cent of legislative bills in 2000–2015, the remainder being private members’ bills (38 per cent) and committee bills (5 per cent). The government, as a rule, has a majority on all permanent committees. Bills are referred to the permanent committees between hearings but in many cases the committees fail to report the bill back to the house, which means it dies (although it can be re-introduced in the following session). Most private members’ bills

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

347

never make it out of the committee stage while around 25 per cent of government bills share that fate. An equally direct committee intervention in government legislation is the amount of amendments made to bills. Considering the extent of amendments made to government bills from the original bill to the final text of legislation as percentage word change in 2003–2014, we found an average of about 20 per cent change (Indridason and Kristinsson 2018). The most influential counterpart to the minister in parliament is the committee chair, who can exert strong influence on the committee agenda. For the minister it is important to be on good terms with the committee chair, which is more likely when the minister and chair are co-partisans. Prior to the crash, co-partisan committee chairs were the rule rather than the exception, that is in 73 per cent of the parliamentary committees in the IP–PP coalition of 2003–2007 and 93 per cent in the IP–SDA coalition of 2007–2009. After the crash, however, ministers lost some of their influence over the committees as the share of co-partisan committee chairs fell to 33 per cent in the SDA–LG coalitions of 2009–2013 and 31 per cent in the PP–IP coalition formed in 2013. Other actors that the ministers may have to consider for the safe passage of bills through parliament include the government parties’ parliamentary groups, stakeholders (through their access to the parliamentary committees), and the opposition. Government bills are introduced in the government majority’s parliamentary groups before being presented in parliament and if they meet with a hostile reception this may create difficulties, especially if government majorities are slim. Stakeholder pressure in Iceland is heavily concentrated on the parliamentary committees and heavy pressure may reduce the chances of a bill passing or lead to greater amending of the bill. The opposition generally has little influence on legislation except towards the end of the session when it can use filibusters to force the government to negotiate about the agenda. This presents a dilemma for ministers, who may be tempted to introduce bills late in order to escape major changes but may risk failing to get them through when time pressures mount.

Other governance mechanisms The Icelandic cabinet meets once or twice a week while parliament is in session. It has no direct administrative authority and relies entirely on individual ministers for administrative control. This means that ministers are formally and effectively highly autonomous concerning such matters as the issuing of regulations, the internal organization of their ministries or major agencies, high-level public appointments, and various other administrative actions, including the allocation of various quotas, grants, and subsidies. Although the Icelandic cabinet spends more time in formal meetings than cabinets in many neighbouring states, often

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

348

 .     

one to three hours per week, these meetings seem to focus to a substantial extent on strategic issues and the presentation of draft legislation that require cabinet approval in order to be introduced as government bills in parliament. Preparations for discussions of draft legislation, however, are not elaborate, although at the beginning of each session in parliament the government issues a list of issues to be presented in the coming months. Rules on cabinet procedures suggest that draft legislation must be presented to the PM’s office at least two weeks in advance of formal presentation in the cabinet but according to a report on the work of the cabinet and ministries, published in 2010, there are in practice no limits to how late draft legislation can be introduced and in practice many bills receive scant attention by the cabinet (Samhent stjórnsýsla 2010: 38). The Icelandic cabinet never votes on anything. Decision-making is by consensus, interpreted by the PM. In some cases, the PM can decide that an issue is not sufficiently prepared for a decision and remove it from the agenda. Such issues can be referred back to the responsible minister or they can enter an informal process of private discussions between those with an interest in the issue. In some instances, the party leaders get involved to settle disagreements and sometimes even members of the parliamentary groups. Ministerial assistants also meet regularly as do the permanent secretaries. In addition to the PM, the minister of finance plays an important interministerial role and as those two roles are often occupied by leaders of the coalition parties, they collectively wield considerable influence. The minister of foreign affairs can also be in an influential position, although less so—it seems—in recent coalitions. The use of more or less permanent coalition committees began in the 1990s and increased after the turn of the century. Coalition committees include the PM and often other party leaders as well as some of the other ministers and generally aim at increasing coordination in the cabinet and across ministries. At present there are four such committees, one on government finances, one on economic issues and the re-organization of the financial sector, one on equal rights, and one on general coordination across ministries. The use of coalition committees may in some cases reduce ministerial autonomy although some ministers welcome the opportunity of gaining access to party and government leaders (Samhent stjórnsýsla 2010: 43). Coalition agreements are not legally binding documents and individual ministers may in some cases be tempted to try to thwart government policy: the LG minister for agriculture and fisheries even voted against, and without resigning as minister, an application for membership of the European Union in July 2009,¹⁰ despite the coalition agreement addressing the issue of EU membership. In a different context, at a later date, however, he was forced to step down. ¹⁰ https://www.althingi.is/thingstorf/thingmalin/atkvaedagreidsla/?nnafnak=41080 (Accessed on 15 March 2021).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

349

Although examples of shirking can be found, the general view is that coalition agreements are binding for coalition partners and their ministers. The problem is not so much that ministers fail to deliver on their parties’ promises as the vagueness of many policy statements that give a large room for interpretation. Given the strong tradition of ministerial autonomy, moreover, the scope for ministers to prioritize according to personal or partisan preferences is considerable. Prior to the economic crash there do not appear to have been any instance of the coalition parties systematically overseeing the implementation of the coalition agreement. In a juridical sense, the PM may have had some supervisory duties but in actual situations it was primarily the role of the coalition partners to keep their own ministers loyal to the coalition pact. Following the crash, however, the PM’s office appears to have taken on a more active role in this respect.

Cabinet duration and termination The duration of cabinets Conventional wisdom among Icelandic politicians is that two-party coalitions with large majorities are more stable than larger coalitions with narrow majorities. No coalition with three parties has ever survived a full term (although one with four parties formed in the second year of the 1987–1991 electoral term managed to sit to the end of the term). The small size of the Alþingi (only 63 members during the period in Table 10.6¹¹) makes it likely that a handful of MPs can obtain a pivotal position. Two-party coalitions have generally been more stable and, prior to the economic crash, the last time a two-party coalition collapsed mid-term was in 1956. After the crash, however, things have been less certain. Only one coalition has sat for a full term after the crash (at the time of writing)—the SDA–LG coalition in 2009–2013—but with a very shaky majority, and regular political upheavals, it ended up as a minority government during its last year in office. Of the subsequent coalitions, the first lasted three years and the next less than a year. The crash seems to have made it more difficult to meet the conditions for maintaining stable and effective coalitions. The reasons for this have partly to do with a more unstable political situation characterized by less trust in politicians, greater volatility, and greater fragmentation of the party system (that makes, e.g., the formation of connected two-party coalitions infeasible). Partly, however, they may have to do with an influx of politicians who question the old ways of doing politics. The renewal rate in parliament has reached ¹¹ Information on duration obtained rikisstjornartal/ (Accessed on 12 May 2018).

from

www.stjornarradid.is/rikisstjorn/sogulegt-efni/

2004-09-15 2006-06-15 65, 8

2006-06-15 2007-05-12 100 2007-05-24 2009-01-26 42, 3

2009-02-01 2009-04-25 10

Asgrimsson

Haarde I Haarde II

Sigurdardottir I

37, 2 100 100 100 100 33, 2

1988-09-28 1989-09-10 1991-04-30 1995-04-23 1999-05-28 2003-05-23

Hermansson II Hermansson III Oddsson I Oddsson II Oddsson III Oddsson IV

1989-09-10 1991-04-20 1995-04-08 1999-05-08 2003-05-10 2004-09-15

1987-07-08 1988-09-28 32, 4

4

1 4, 7b

9

5 1 1 1 1 9

7a, 7b

10

IP, SDA

7a: IP, SDP, PP7b: IP, SDP, PP

Relative Mechanisms Terminal Parties duration of cabinet events (when (%) termination conflict between or within)

Palsson

Date out

Date in

Cabinet

Table 10.6 Cabinet termination in Iceland, 1987–2018

Finance

Policy area(s)

The coalition fell after the SDA demanded a change in leadership of the coalition, from the IP to the SDA. Sigurdardottir had led a caretaker government until the previously scheduled early elections were held in April 2009.

Oddsson was replaced by Foreign Minister Asgrimsson as part of a 2003 coalition deal. Asgrimsson resigned after his party performed poorly in the May 2004 local elections.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2016-04-07 2016-10-30 54, 2

2017-01-11 2017-10-29 16, 7

Johannsson I

Benediktsson I

4, 7a

4

1 9 11, 14

IP, BF

Gunnlaugsson stepped down as PM after he had been implicated in the Panama Papers scandal. An early election had been promised in relation to Gunnlaugsson’s resignation although no date was mentioned. The coalition collapsed in September after BF grew unhappy with an IP minister and ended the coalition.

Note: Relative duration: share of constitutionally allowed duration for this cabinet.

Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event 14: Personal event

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of PM

2009-05-10 2013-04-27 100 2013-05-23 2016-04-06 72

Sigurdardottir II Gunnlaugsson I

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

352

 .     

unprecedented heights both through greater turnover among MPs within the established political parties and through the emergence of new parties. Thus, the average share of newly elected MPs in the twelve-year period 1995–2007 was 30.2 per cent compared to 39.7 per cent in the eight-year period 2009–2017, indicating a rather sharp drop in parliamentary experience in Alþingi. All the new parties entering parliament since 2009 have placed reform of political practices on the agenda and called for greater accountability in politics. Some of these parties, however, are seen by the old parties as risky coalition partners, lacking the experience to run an effective coalition—which in turn reduces the options for a viable coalition. With the economic crash of 2008 Iceland moved from a period of relative government stability to one of uncertainty and instability. As may be seen in Table 10.6, no coalition broke down in the 1991–2008 period as a result of losing its majority in parliament or policy disagreements. All completed their legislative terms, except for the IP–SDA coalition formed in 2007. The IP–SDA coalition collapsed in January 2009 amidst massive public protests and a rank-and-file revolt in the SDA against continuing cooperation with the IP, which many on the left blamed for the crash. It was replaced by the first socialist coalition in Icelandic political history, which was then removed from power through a resounding defeat in the elections of 2013, where the combined vote shares of the two partners fell from 51.5 per cent to 23.8 per cent. The PP–IP coalition that took over had a solid backing in parliament but soon began losing heavily in the opinion polls. The revelation in April 2016 that PM Gunnlaugsson had been associated with an offshore business company, holding a share he later sold to his wife for the price of 1 US dollar, led to loud demands for his resignation. The company in question, Wintris Inc., was among the claimants to the estates of some of the fallen Icelandic banks. Gunnlaugsson met the president to discuss the position and, according to President Grímsson, wished to call an election, which Grímsson claims to have denied. No direct evidence, however, supports this. It seems clear, however, that the PM had lost the support of his parliamentary group at that time and he was replaced by the PP’s deputy leader Jóhannsson as PM (who later also replaced Gunnlaugsson as party leader) until a new election was called in October. The issue of Gunnlaugsson’s resignation raised interesting questions concerning the PM’s right to dissolve parliament and call a new election. According to mainstream jurisprudence, the PM calls an election and has the right to dissolve parliament but doing so requires the countersignature of the president. Thorarensen (2015: 428) points out that there are no valid arguments for the president to refuse such a request but should he or she nonetheless do so the dissolution would fail to take effect. The president, however, can never dissolve parliament on his own initiative. Grímsson’s claim to have refused

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

353

Gunnlaugsson’s wish to call an election is most convincingly interpreted in the context of Grímsson’s rather unusual approach to the role of president and attempt to expand the powers of the office. The coalition of the IP–RP–BF, which formed following the elections of 2016, collapsed in September 2017 when BF decided to leave the government. While the immediate reasons given had to do with confidential information passed on by the minister of justice (IP) to the PM (IP) concerning the latter’s father, it seems evident that a number of other issues—all involving the controversial minister of justice—had stretched the willingness of BF to accept responsibility for the actions of the minister to a breaking point. Cases in point include her actions regarding immigrants and her appointments to the new Court of Appeal, which deviated significantly from the recommendations of an evaluation committee.

Conclusions After nearly two decades of relatively stable coalition governance in Iceland, the pots and pan revolution in 2009 marked a turning point, introducing a greater degree of instability. In the first place, it appears to have set a large number of voters afloat leading to greater electoral volatility and party system fragmentation. Secondly, it affected party cohesion significantly and made coalition governance more difficult than before. Thirdly, it led to greater difficulties in forming and maintaining the stable two-party type of coalitions, which characterized the period 1991–2009. Coalition governance since the crash in 2008 has been characterized by more instability than before. No two-party coalition in this period has completed a full term except Sigurðardóttir II—but her cabinet became a minority coalition a little over three years into its term. Two coalitions broke down under pressure (2009 and 2016). This period saw two minority cabinets—and the second of these was the first minority cabinet since the 1940s whose role was not that of a caretaker government. Faced with increasing difficulties of maintaining stable coalitions, governments have emphasized coalition management and coordination to a greater degree than before, thus indicating a shift away from the ministerial government model to incorporate elements more closely associated with the coalition compromise model. This is apparent in more elaborate coalition agreements, greater emphasis on coalition leadership, and inter-ministerial coordination as well as the increasing use of watchdog chairs in parliamentary committees. Moves away from ministerial government to a more coordinated system are contested, however, and may meet with strong opposition. With the traditional two-party minimal winning coalition model weakening, they may, nonetheless, be forced to coordinate their actions better than during the heyday of ministerial government. Not

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

354

 .     

creating the necessary consensus or coordination on administrative or legislative issues may create difficulties for weak governments and individual ministers in a way that would have been much less likely prior to the economic crash.

References Bäck, Hanna, Marc Debus, and Patrick Dumont (2011). ‘Who Gets What in Coalition Governments? Predictors of Portfolio Allocation in Parliamentary Democracies’. European Journal of Political Research, 50(4): 441–78. Bakker, Ryan, Catherine De Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Anna Vachudova. ‘Measuring party positions in Europe: The Chapel Hill expert survey trend file, 1999–2010.’ Party Politics 21, no. 1 (2015): 143–152. Bowler, Shaun, Indridi H. Indridason, Thomas Bräuninger, and Marc Debus (2016). ‘Let’s Just Agree to Disagree: Dispute Resolution Mechanisms in Coalition Agreements’. Journal of Politics, 78(4): 1264–78. Druckman, James N., Lanny W. Martin, and Michael F. Thies (2005). ‘Influence without Confidence: Upper Chambers and Government Formation’. Legislative Studies Quarterly, 30(4): 529–48. Golder, Sona Nadenichek (2006). The Logic of Pre-electoral Coalition Formation. Ohio: Ohio State University Press. Gschwend, Thomas, Michael F. Meffert, and Lukas F. Stoetzer (2017). ‘Weighting Parties and Coalitions: How Coalition Signals Influence Voting Behavior’. The Journal of Politics, 79(2): 642–55. Grittersova, Jana, Indriði H. Indriðason, Ricardo Crespo, and Christina Gregory (2016). ‘Austerity and Niche Parties: The Electoral Consequences of Fiscal Reform’. Electoral Studies, 42: 276–89. Hazan, Reuven Y., and Gideon Rahat (2010). Democracy within Parties. Candidate Selection Methods and Their Political Consequences. Oxford: Oxford University Press. Indridason, Indridi (2014). ‘The Collapse: Economic Considerations in Vote Choice in Iceland’. Journal of Elections, Public Opinion & Parties, 24(2): 134–59. Indridason, Indridi, and Gunnar H. Kristinsson (2015). ‘Primary Consequences. The Effects of Candidate Selection through Party Primaries in Iceland’. Party Politics, 21: 565–76. Indridason, Indridi, and Gunnar H. Kristinsson (2018). ‘The Role of Parliament under Ministerial Government’. Icelandic Review of Politics and Administration, 14(1): 149–66. Indridason, Indridi H., and Shaun Bowler (2014). ‘Determinants of Cabinet Size’. European Journal of Political Research, 53(2): 381–403.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

:     

355

Indridason, Indridi, Eva H. Önnudóttir, Hulda Thorisdottir, and Ólafur Þ. Harðarson (2017). ‘Re-electing the Culprits of the Crisis? Elections in the Aftermath of a Recession’. Scandinavian Political Studies, 40(1): 28–60. Jóhannesson, Guðni Th. (2005). Völundarhús valdsins. Reykjavík: Mál og menning. Kristinsson, Gunnar H. (2000). ‘The Icelandic Parliamentary Election of 1999’. West European Politics, 23(1): 187–92. Kristinsson, Gunnar H. (2011). ‘Party Cohesion in the Icelandic Althingi’. Icelandic Review of Politics and Administration, 7(2): 227–48. Retrieved from: http://www. irpa.is/article/view/1137/pdf_227. Kristinsson, Gunnar H. (2016). ‘Specialists, Spinners and Networkers: Political Appointees in Iceland’. Acta Politica, 51(4): 413–32. Laver, Michael, and Kenneth A. Shepsle (1990). ‘Coalitions and Cabinet Government’. American Political Science Review, 84(3): 873–90. Laver, Michael, and Kenneth A. Shepsle (1996). Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Müller, Wolfgang C., and Kaare Strom, eds. (2003). Coalition Governments in Western Europe. Oxford: Oxford University Press. Önnudóttir, Eva H., and Ólafur Þ. Harðarson (2018). ‘Political Cleavages, Party Voter Linkages and the Impact of Voters’ Socio-economic Status on Vote-Choice in Iceland, 1983-2016/17’. Icelandic Review of Politics & Administration, 14(1): 101–30. Önnudóttir, Eva H., Hermann Schmitt, and Ólafur Þ. Harðarson (2017). ‘Critical Election in the Wake of an Economic and Political Crisis: Realignment of Icelandic Party Voters?’. Scandinavian Political Studies, 40(2): 157–81. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, et al. “Explaining the Salience of Anti-Elitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data.” Research & Politics, (January 2017). Protsyk, Oleh (2005). ‘Prime Ministers’ Identity in Semi-presidential Regimes: Constitutional Norms and Cabinet Formation Outcomes’. European Journal of Political Research, 44(5): 721–48. Samhent stjórnsýsla (Joined up government) (2010). Reykjavík: Prime Minister’s Office. Retrieved from: https://www.forsaetisraduneyti.is/frettir/nr/4514. Strøm, Kaare; Müller, Wolfgang C. and Bergman, Torbjörn eds. (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Thies, Michael F. (2001). ‘Keeping Tabs on Partners: The Logic of Delegation in Coalition Governments’. American Journal of Political Science, 45(3): 580–98. Thorarensen, Björg (2015). Stjórnskipunarréttur. Undirstöður og handhafar ríkisvalds. Reykjavik: Bókaútgáfan Codex.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

356

 .     

Appendix. List of political parties Abbreviation SP LG PA WA NPP ULL THPM SDA ASD SDP AESI PP LP CP IP CM BF P RP CentP PeP

Name People’s Unity Party—Socialist Party or United Socialist Party (Sameiningarflokkur alþýðu—Sósíalistaflokkurinn) Left-Green Movement (Vinstrihreyfingin—grænt framboð) People’s Alliance (Alþýðubandalagið) Women’s List (Samtök um kvennalista) National Preservation Party (Þjóðvarnarflokkurinn) Union of Liberals and Leftists (Samtök frjálslyndra og vinstrimanna) Thjodvaki—People’s Movement (Þjóðvaki—Hreyfing fólksins) Social Democratic Alliance (Samfylkingin jafnaðarmannaflokkur Íslands) Alliance of Social Democrats (Bandalag jafnaðarmanna) Social Democratic Party (Alþýðuflokkurinn) Association for Equality and Social Justice (Samtök um jafnrétti og félagshyggju) Progressive Party (Framsóknarflokkurinn) Liberal Party (Frjálslyndi flokkurinn) Citizens’ Party (Borgaraflokkur) Independence Party (Sjálfstæðisflokkurinn) Citizens’ Movement (Borgarahreyfingin) Bright Future (Björt framtíð) Pirate Party (Pírataflokkurinn) Reform Party (Viðreisn) Centre Party (Miðflokkurinn) People’s Party (Flokkur fólksins)

Note: Party names are given in English, followed by the party name in the party’s native Icelandic in parentheses.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 11 Ireland: Coalition Politics in a Fragmenting Party System Paul Mitchell

Ireland is now unequivocally a country whose political system is dominated by coalition governance. Before 1989 Ireland’s pattern of government formation was one in which a single dominant party, Fianna Fáil, governed for most of the twentieth century, sometimes rotating with a coalition of (most of) ‘the rest’. Fianna Fáil dominance is now a fading memory and since 1989 all governments in Ireland have been coalitions.¹ The chapter analyses the evolution of coalition governance in Ireland. It considers the institutional context and party system and then sequentially examines life cycle of coalitions: formation bargaining, governance, and termination. There is a greater than usual focus on the governance phase with particular attention to the increasingly important (and lengthy) coalition agreements, which set out the initial policy bargain. Politicians quickly became aware that the coalition contract is not automatically implemented and have sought to develop a whole range of monitoring and enforcement devices, in an attempt to attenuate both agency and partisan losses.

The institutional setting The new state of Ireland, which was founded in 1921 following a revolutionary secession from the United Kingdom, adopted a set of governing institutions that are in many respects quite similar to the Westminster model. The state was conceived as a parliamentary democracy with many features such as the organization of parliament, procedures for lawmaking, and the administrative structure clearly inherited from Westminster.

¹ With the partial exception of two governments that formed after the 2016 election, which were single-party minority governments (with three independent cabinet ministers) and crucially an external support arrangement with the main opposition party. Following the 2020 election a ‘normal’ threeparty coalition was formed, but with Fianna Fáil and Fine Gael governing together for the first time ever in a formal coalition. Paul Mitchell, Ireland: Coalition Politics in a Fragmenting Party System In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0011

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

358



Nevertheless, Irish political institutions contain some important institutional differences that are highly relevant in shaping how the country is governed. Probably the most important institutional divergence from the United Kingdom was the adoption of limited government by a written constitution (first in 1922 and superseded by the current 1937 constitution—Bunreacht na hÉireann) protected by a Supreme Court and developed by judicial interpretation and popular referendums. All changes to the constitution require the consent of the people via a referendum.² Proposals to amend the constitution must be initiated in the Dáil (the most important lower house of parliament). The wording of the referendum is determined by the government and then the people decide the outcome by a majority of valid votes cast. The frequency of referendums has been increasing: while only two occurred before 1968, since then (1968–2019) forty more have taken place, averaging slightly less than one a year.³ A second very important innovation with a clear relevance to the party system and pattern of governance was the adoption of proportional representation by means of the single transferable vote (STV) for all parliamentary elections.⁴ Ireland’s candidate-centred electoral system can be expected to influence the incentives and loyalties of deputies (TDs). In most proportional representation systems each voter’s principal decision is to choose between rival party lists. Although many countries’ list systems have some elements of intra-party preference voting, electors nevertheless usually vote first for the chosen party, and their vote may help elect an individual whom they oppose. The central feature of Ireland’s electoral system is that the electorate votes directly for individual candidates in multimember districts. There was (and still is) support amongst the electorate for a pattern of small district constituency representation similar to the Westminster plurality system, but without the latter’s pronounced disproportionality at national level. Two aspects of STV are particularly important. First, it is a preferential electoral system in which voters have the opportunity to rank individual candidates in constituencies with a small district magnitude (since 1947 between three and five seats). This typically means that only two or three of a major party’s candidates have much chance of being elected and that the voters alone decide which of the party’s candidates are successful. A frequent result is intense intra-party competition because STV puts candidates of the same party in

² But citizens themselves cannot initiate referendums. ³ Since the turn of the century (2000–2019) there have been 19 referendums. ⁴ The electoral system of PR-STV is protected by the constitution and can thus only be changed by referendum. Fianna Fáil governments have on two occasions proposed replacing STV with single member plurality (in 1959 and 1968), but lost both referendum and on the second attempt decisively (61 per cent against the proposed change). More recently the Fine Gael–Labour government (which formed in 2011) established a convention on the constitution (comprised of 66 randomly selected citizens, 33 politicians, and an independent chair)—the convention voted overwhelmingly (79 per cent) against changing the electoral system. The government accepted this recommendation (see Farrell et al. 2017 on the convention).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



359

competition with each other. They often attempt to differentiate themselves from their party running mates through intense ‘constituency service’.⁵ They are right to be worried: it has been calculated that between 1922 and 1997, 34 per cent of all TDs who suffered defeat at an election lost their seat not to a rival party’s candidate but to one of their running mates (Gallagher 2000; Gallagher and Komito 2018). The second relevant feature of STV is that preference voting is not limited to an intra-party choice: voters can (and do) vote across party lines. There are strong incentives for candidates (and parties) to try and attract lower preferences from partisans of other parties, and this clearly has important consequences for campaigning and for legislative and executive coalition building. Two other institutional features should be mentioned more briefly. Although Ireland is bicameral, the Dáil (the lower house) is by far the stronger. Bills come before the Seanad (Senate) but, at most, it can delay them for 90 days. The Fine Gael–Labour coalition wanted to abolish the upper house, and a referendum on abolition was held in October 2013. In something of a surprise (and on a low 39 per cent turnout) the people voted against abolishing the Senate by 52–48 per cent. Despite the fact that Ireland has a directly elected president (on a seven-year term) he or she plays no role in government formation. There is, however, a constitutional clause relating to the powers of legislative dissolution, which could influence government composition if invoked by the president. In normal circumstances, when a Taoiseach requests that the president dissolve the Dáil in order to hold an election, the president is required to do so (Art. 13.2.1). However, according to Article 13.2.2 the president ‘may in his absolute discretion refuse to dissolve Dáil Éireann on the advice of a Taoiseach who has ceased to retain the support of Dáil Éireann’. A Taoiseach who has just lost a no-confidence motion would clearly justify invoking this clause, but there is some debate about whether there are other circumstances in which a president could refuse to dissolve—for example, if it was obvious that the government’s legislative base had diminished but there has not actually been a formal vote. This question was partially answered ‘when President Mary Robinson “let it be known” that if the Taoiseach Albert Reynolds, leading a rump Fianna Fáil minority government in November 1994 following the collapse of his coalition with Labour, had asked her to dissolve the Dáil, she would have invoked Article 13.2.2 and refused to do so’ (Gallagher, 2018: 166). With the signal received, Reynolds resigned, and an alternative coalition took office for the first time without an election. While this power has not yet been explicitly used, a president could encourage coalition renegotiation during an inter-electoral period as a possible alternative to the traditional resort to an ⁵ This electoral incentive towards candidate differentiation does not of course necessitate that intense constituency service be the method of competition. Candidates could compete as effective legislators or ministers or take up distinctive policy positions.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

360



early election. Given an increasingly fragmented party system and hence a higher incidence of minority situations in parliament (every parliament since 1979), this power might become more relevant.

The party system The title of the Irish chapter in Müller and Strøm (2000) was ‘Ireland: From Single-Party to Coalition Rule’, and it reflected a dramatic transformation in the bargaining environment from a predominant party system,⁶ famously characterized by Peter Mair (1987, 1993) as ‘Fianna Fáil versus the rest’, to a less constrained context in which a wider range of coalitions might be feasible. For most of the twentieth century a Fianna Fáil single-party government was the default outcome of the government formation process, though many of these cabinets were ‘large’ minority administrations. Since Fianna Fáil first entered government in 1932 up until it entered its first executive coalition in 1989 the party enjoyed an average first preference vote of 46.4 per cent (19 elections from 1932 to 1987). Given that the electoral system is not perfectly proportional and typically provides a ‘seat bonus’ to the largest party, Fianna Fáil in the same period (1932–1987) on average won 49.6 per cent of the seats with the result that it could typically form either a majority government or a large and fairly secure minority government on its own. Fianna Fáil was ideologically opposed to even the principle of coalition⁷, though no doubt this stance also had strategic underpinnings. Thus, at elections before 1989 there were really only two possible outcomes of the government formation process that typically followed a general election. If Fianna Fáil had a majority of seats it always formed a single-party government which it managed to do on eight occasions.⁸ If it did not have a majority of seats, then the availability of an alternative to a Fianna Fáil minority cabinet depended on the current attitudes of the opposition parties to working together. In the same period of 1932 to 1987 there were eleven minority situations following general elections: five of these resulted in coalitions and six in Fianna Fáil minority governments. The two main opposition parties Fine Gael (usually the second largest party) and the Labour Party (usually the third party) faced heavy strategic constraints given Fianna Fáil’s size combined with its refusal to cooperate with ⁶ For Sartori (1976: 174) a predominant party system is one in which a single-party democratically wins a majority of seats on a regular basis, but this ‘threshold can be lowered to the point at which minority single-party governments remain a standing and efficient practice’, as is often the case in Norway, Sweden, and Denmark. Sartori (1976: 175) judged that ‘the Irish system has been predominant between 1933–1948 and 1957–1973)’. ⁷ The Irish Times observed on 3 November 1947 that ‘Mr de Valera’s abhorrence of coalitions is proverbial’, quoted in MacDermott (1998: 67). ⁸ 1933, 1937, 1938, 1944, 1957, 1965, 1969, and 1977. In 1937 and 1965 Fianna Fáil won exactly half of the seats in the Dáil.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



361

any other party. Their reactions to this predicament oscillated over the decades between pursuing mutually exclusive strategies in the hope of achieving sufficient electoral growth to enable them to lead a government and then subsequently resorting to a mutually dependent coalition as the only way of removing Fianna Fáil once it becomes clear that the long hoped for seat gains have not materialized in sufficient quantities.⁹ Patterns of government formation often appear to be on an endless repeat cycle. Fianna Fáil governs most of the time either by virtue of obtaining a majority or by default when Labour and Fine Gael pursue divergent strategies in order to forge stridently independent identities untainted by coalition compromises. Eventually, they reconsider their lengthy tenure on the opposition benches and form a coalition that usually results in electoral losses, hence necessitating lengthy periods out of government to recuperate.¹⁰ The Labour Party, always the second party in any coalition, has always suffered electoral costs, most recently severely, from any time spent in coalition.¹¹ Given that Fianna Fáil would always cast a large shadow over the process, the extra problem for the other parties was that ‘the coalition alternative’ would require almost all of the other parties to combine against the dominant party, irrespective of whatever policy differences, personal dislikes, or divergent strategies existed among the other parties. Overall, during the 57-year period from when Fianna Fáil formed its first government in 1932 until it entered its first coalition in 1989, Fianna Fáil governed on its own for 41.6 years (73 per cent of the entire duration), whereas coalitions of ‘the rest’ governed for only 15.2 years (27 per cent of the duration). The bargaining environment permanently changed in 1989 when Fianna Fáil broke the habit of a lifetime and entered its first coalition with the Progressive Democrats. Why Fianna Fáil abandoned its total opposition to coalitions is a complex issue (see chapters in Gallagher and Sinnott 1990; Marsh and Mitchell 1999). But be that as it may a new era of coalition politics began in 1989, so that there were nine successive coalition cabinets (1989–2016) interrupted only by the rather unusual formation of a ‘small’ single-party minority government in 2016 (discussed in the next section). Back in 1989 and reacting to Fianna Fáil’s new coalitionable status Mair (1990: 213–14) rather presciently predicted that ‘ironically . . . far from weakening Fianna Fáil, this may well have the effect of cementing its hold on office even more securely than before’. The point of course is that as by ⁹ For example, Labour pursued an anti-coalition strategy in the mid-1960s and famously declared at its 1967 ‘new republic’ conference that ‘the Seventies will be Socialist’, though they did not always define what ‘socialist’ meant (Gallagher 1982: 69). ¹⁰ Of the 17 countries included in Strøm et al. (2010), cabinet parties in Ireland suffered the second worst ‘adverse incumbency effect’, losing on average 4.47 per cent, compared to a loss of an average of 2.59 per cent for all of the countries included (see Narud and Valen 2010: 379). ¹¹ The only partial exception was the 1951 election in which Labour managed to increase its vote by 0.1 per cent following the first coalition government (if the vote for National Labour, now remerged with Labour, is added to Labour’s 1948 vote).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

362



far still the largest party and with a fairly non-ideological position in the centre of the party system, Fianna Fáil would be a member of most potential winning coalitions. From 1989 to 2017 Fianna Fáil led eight of the twelve cabinets that formed, and all of them were coalitions.¹² To put this another way, during 1989–2017 Fianna Fáil governed in coalition for 18.8 years (69 per cent of the entire duration), and Fine Gael-led governments for 8.7 years (31 per cent of the duration; calculated from Table 11.6). One final way of looking at it is to observe that it took almost the economic and financial collapse of the country to remove Fianna Fáil from government in 2011! The party system has been changing and most recently more dramatically than normal. Three parties have dominated and been ever present since the first election in 1922. While the fortunes of Fianna Fáil, Fine Gael, and Labour (usually in that order of size) have oscillated over the decades (with some notable challenges from new parties in the 1940s and 1950s that no longer exist), we can see from Table 11.1a that they have generally dominated the party system and hence also government formation. In the post-war period until 1987 the combined twoparty vote (of Fianna Fáil + Fine Gael) was generally over 80 per cent. It fell sharply in 1987 (when another new party that no longer exists, the Progressive Democrats, took 11.8 per cent at its first election), but the two-party vote has never been above 74 per cent since 1987. The three-party vote was mostly above 90 per cent from the mid-1960s to mid-1980s, but this has also fallen away sharply. Fianna Fáil last won an overall majority of votes and seats in 1977, but since then all parliaments have been minority situations. Fianna Fáil, which used to regularly win 45 per cent plus, slipped gradually to averaging nearer 40 per cent in the 1990s and 2000s, no doubt encouraging its participation in coalition politics. Table 11.1a lists Irish election results in seats and is arranged in a general left– right order. Cross-checking Laver’s (1994) expert survey with the 2014 CHES data we can approximate the ordinal placement of currently existing parties as follows (from left to right):¹³ Anti-Austerity Alliance–People Before Profit —Sinn Féin— Green Party—Social Democrats—Labour— Fianna Fáil — Fine Gael. There is some doubt, however, that any of these surveys, which are snapshots since the 1990s, can accurately reflect the ordinal placement of the parties for the entire post-war period. For example, using manifesto data Mair argued that Fine Gael can be placed to the left of Fianna Fail for the period 1965–1977 (Mair 1986: 463–5). Part of the problem here is that there are no really large policy differences between Fine Gael and Fianna Fáil. Nevertheless, as an approximation, the above ¹² Except the very short-lived Cowan II, which only lasted 33 days. ¹³ Older parties, for example, Clann na Talmhan and Clann na Poblachta are simply our best judgement since they are not included in any of the expert surveys. A party that was important for about 20 years (and was a member of six coalition cabinets) was the Progressive Democrats. This was the most ‘free-market’ party but also socially relatively liberal.

1944-06-09 1948-02-18

de Valera VI Costello I

de Valera VII Costello II de Valera VIII Lemass I Lemass II Lemass III Lynch I Lynch II Cosgrave I Lynch III Haughey I FitzGerald I Haughey II FitzGerald II Haughey III Haughey IV Reynolds I Reynolds II Bruton Ahern I

1 2

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

1951-06-13 1954-06-02 1957-03-20 1959-06-23 1961-10-11 1965-04-21 1966-11-10 1969-07-02 1973-03-14 1977-07-05 1979-12-11 1981-06-30 1982-03-09 1982-12-14 1987-03-10 1989-07-12 1992-02-11 1993-01-12 1994-12-15 1997-06-26

Date in

Cabinet Cabinet number

Table 11.1a Irish cabinets since 1944

1997-06-06

1992-11-25

1981-06-11 1982-02-18 1982-11-24 1987-02-17 1989-06-15

1969-06-18 1973-02-28 1977-06-16

1961-10-04 1965-04-07

1951-05-30 1954-05-18 1957-03-05

1944-05-30 1948-02-04

FF FG, Lab, CnT, CnP, NL FF FG, Lab, CnT FF FF FF FF FF FF FG, Lab FF FF FG, Lab FF FG, Lab FF FF, PD FF, PD FF, Lab FG, Lab, DL FF, PD min mwc maj Maj Min Min Min Maj mwc Maj maj min min mwc min min min mwc mwc min

maj min 47 50.3 53 53 48.6 50 50 52 50.7 56.8 56.1 48.2 48.8 51.8 48.8 50 50 60.8 50.6 48.8

55 45.6 147 147 147 147 144 144 144 144 144 144 148 166 166 166 166 166 166 166 166 166

138 147 5 5 6* 5* 5 4 4 3 3 3 3 4* 4 4 5 5 5 5 5 6

5 6 3.26 3.26 2.72 2.72 2.78 2.61 2.61 2.46 2.59 2.36 2.36 2.62 2.53 2.52 2.89 2.94 2.94 3.46 3.63 3.00

2.83 3.66

Election date Party Type Cabinet Number of Number of ENP, parties in parliament composition of strength seats in of cabinet cabinet in seats parliament parliament (%)

CnT CnT FF FF FF FF FF FF Lab FF FF Lab Lab Lab Lab Lab Lab Lab Lab FF

FF FF

Continued

CnP

Formal Median support party in first policy parties dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2002-06-06 2007-06-14 2008-05-07 2011-01-23 2011-03-09 2016-05-06 2017-06-14

23 24 25 26 27 28 29 2011-02-25 2016-02-26

2002-05-17 2007-05-24

FF, PD FF, GR, PD FF, GR, PD FF FG, Lab FG FG

mwc sur sur min mwc min min

53.6 51.8 51.8 46.9 68 37.3 36.1

166 166 166 166 166 158 158

6 6 6 5 5 7 7

3.38 3.03 3.03 3.03 3.52 4.93 4.93

Cabinet Number of Number of ENP, Type Election date Party parties in parliament strength seats in composition of cabinet in seats parliament parliament of cabinet (%) FF FF FF FF FG FF FF

FF FF

Formal Median support party in first policy parties dimension

Notes: For a list of parties, consult the chapter appendix. Median parties are retrieved from Müller and Strøm (2000) for the 1977–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015). The first policy dimension is economic left–right. The number of parties only includes parties that have won more than two seats in parliament at least once during the period of observation. *The number of parties includes the then abstentionist Sinn Féin in 1957 and 1959, and the abstentionist Anti-H Block in 1981. Cabinet types: min = maj = Single-party majority cabinet; Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition. Table legends and variables are further defined in the Appendix, this volume.

Ahern II Ahern III Cowen I Cowen II Kenny I Kenny II Varadkar I

Date in

Cabinet Cabinet number

Table 11.1a Continued OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



365

ordinal left–right placement seems reasonable.¹⁴ On this dimension, Fianna Fáil, as normally the far largest party, is the median party for most of the history of the state (they lost this position to Fine Gael in the ‘earthquake election’ in 2011). We should note that there is an increasing problem in confidently identifying the median party on the first and second dimensions due to the greatly increased number of elected independents. There have always been a significant number of TDs belonging to no party, quite high numbers in the early years of state, fewer in the 1960s–1980s, but a noticeable growth since 2002. This reached a new peak in 2016: 23 TDs belong to no party at all, that is almost 15 per cent of the parliament. Since there is no reliable way of locating each of these in the relevant policy spaces they have been excluded from all estimates.¹⁵ We cannot complete discussion of the party system and recent context without a fuller consideration of the implications of the financial crisis. Without exaggeration in 2011 Ireland experienced a truly earthquake election (see Gallagher and Marsh 2011 for the full story). At the previous election in 2007 things looked pretty much like business as usual. Fianna Fáil won 41.6 per cent of the first preference votes and 78 seats (46.7 per cent). Six votes short of an overall majority Fianna Fáil decided to form a three-party coalition with the Green Party (six seats) and the Progressive Democrats (two seats). Thus, Bertie Ahern formed his third coalition government and looked forward to a happy retirement from politics in 2011 when he would be 60 years old (Murphy 2011: 9). Instead, Ahern was overwhelmed by a bizarre series of personal and party finance scandals so that he was forced into early retirement and resigned in May 2008 (he was later expelled from Fianna Fáil). The same coalition continued with Brian Cowan as the new Fianna Fáil leader and Taoiseach. Cowan’s first setback was the loss of the referendum on the Lisbon Treaty (53 per cent to 47 per cent), but this was nothing compared to what was about to follow. The crisis struck in September 2008. There had been about six months of serious turbulence on the stock market following the collapse of the US investment bank Bear Sterns. The Irish banks were heavily reliant on borrowed money from international banks and this credit flow had begun to dry up.

¹⁴ It is however far from clear that Irish voters think in terms of a left–right dimension. Analysing data from the 2011 Irish National Election Study, Gail McElroy (2017: 80) concludes: ‘despite almost three years of economic crisis, left and right appear to have no more substantive meaning for Irish voters in 2011 than in 2002.’ But see also chapters in Marsh et al. (2018). ¹⁵ In Mitchell (2000), a Northern Ireland policy scale was listed as the second dimension of party competition, even though in Laver’s survey (1998) this ‘second dimension’ was actually the most salient dimension. This may have reflected the fast-moving events of that decade with two IRA ceasefires and the eventual negotiation of the Belfast Agreement in 1998. From the more recent Chapel Hill data (2014 survey) the two leading candidates for an Irish second policy dimension (that meaningfully differentiates the parties) would be GAL–TAN and/or Nationalism. On GAL–TAN the parties are ordered (from liberal to traditional): Anti-Austerity Alliance–People Before Profit—Green Party—Labour— Sinn Féin—Fine Gael— Fianna Fáil. On Nationalism (from cosmopolitan to nationalist): Green Party—Anti-Austerity Alliance–People Before Profit—Labour— Fine Gael—Fianna Fáil—Sinn Féin.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

366



Murphy (2011: 16; 2016) notes: ‘as the Irish property market began to crash spectacularly in the autumn of 2008, a concomitant collapse of the banks’ liquidity ensued due to the enormous sums loaned by all the main banks, particularly Anglo-Irish Bank, to property developers.’ There was a serious threat that the collapse of one of the main Irish banks would trigger the collapse of all the others and hence undermine the entire solvency of the state. On the night of 29 September 2008 an incorporeal cabinet meeting took place with ministers contacted by phone between 3 a.m. and 5 a.m. (Murphy 2011: 15).¹⁶ The decision was taken to ‘guarantee the deposits, loans, obligations and liabilities of the six Irish banks, a total sum of Euro 440 billion, more than twice the country’s gross national product’ (Murphy 2011: 15). Ultimately Ireland could ill afford this bank bailout: GDP dropped by 7 per cent in 2009, unemployment rose sharply, and public sector pay was cut. Although the government tried hard to resist, it eventually had to accept a 85-billion-euro bailout of the country from the European Union and IMF on 28 November 2010. The Green Party announced that it would only stay in government to pass the December 2010 budget, but then wanted an election in January 2011. The first opinion poll of 2011 estimated support for Fianna Fáil at 14 per cent, prompting after some more theatrics the resignation of Brian Cowan. Fianna Fáil would contest the 2011 election under a new leader, Micheál Martin. Electoral change in 2011 was ‘truly seismic. Fianna Fáil suffered a negative tsunami of votes that has few parallels among governing parties anywhere’ (Gallagher 2011: 139). Fianna Fáil plummeted from 41.6 per cent in 2007 to 17.4 per cent in 2011, losing 58 of their 77 seats. The Green Party was entirely wiped out. By contrast Fine Gael (36.1 per cent) and Labour (19.4) had their best elections in many years and Sinn Féin rose to 9.9 per cent. The combined twoparty vote (of Fianna Fáil + Fine Gael) fell to only 53.5 per cent. There is no mystery as to what happened—it was a classic case of ruthless electoral accountability. The opinion polls show a clear pattern of two sharp drops in support for Fianna Fáil—the first after the Irish government’s bailouts of the Irish banks in September 2008 and the second two years later when Ireland withdraws from the bond market and has to be bailed out by the European Union and IMF. The ‘outcome was in essence a classic example of the voters exacting punishment for what was widely seen as a succession of bad economic decisions. Fianna Fáil (and the Greens) destroyed the economy, and the voters went some way to destroying those parties in revenge’ (Marsh and Cunningham 2011: 172–4). Some years of deep austerity and competent government by Fine Gael and Labour between 2011–2016 helped facilitate a remarkable turnaround in Ireland’s macroeconomic position, so that by the time of the next election in 2016 Ireland ¹⁶ John Gormley, the leader of the Green Party, could not be contacted by phone—he was woken by a garda (police) at his home asking him to please ring the Taoiseach (Leahy 2009: 333).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



367

had the fastest growing economy in the European Union (Murphy 2016a: 2). Despite many objective economic achievements, sustained austerity is nearly always electorally dangerous for incumbent governments. Fine Gael was reduced to 25.5 per cent (-10.6 and the loss of 27 seats) and Labour was almost destroyed, scoring only 6.6 per cent (-12.8; Labour lost 30 of its 37 seats in the Dáil). In 2016 the three-party vote (of Fianna Fáil + Fine Gael + Labour) was only 56 per cent, meaning that ‘the others’ now comprised 46 per cent of the vote. The others include 23 TDs for Sinn Fein, 6 for Anti-Austerity Alliance–People Before Profit, 3 for a new party, the Social Democrats, and an astonishing 23 TDs who belong to no party at all. The party system fragmented with some sharp rises in the effective number of parties (ENP) and disproportionality.

Government formation The bargaining context: coalition formation before and after elections Certain electoral systems, especially those that encourage transferring votes from one party to another, provide extra incentives to engage in coalition formation before as well as after the election. The STV, by encouraging voters to rank-order candidates (and thus parties), rewards cooperative electoral strategies. High transfer rates (of lower preference votes) among parties engaged in pre-electoral coalitions can make all the difference to the distribution of a fairly small number of seats. Given that the Dáil is often delicately balanced and governments rarely enjoy large majorities (if they have one at all), transfers can make or break a prospective coalition.¹⁷ Thus, there is a distinction between pre-ballot coalitions (transfer pacts under STV) designed to maximize the seat share of a particular set of cooperative parties and the negotiation of an executive coalition once the results are in (Mitchell 1999). Table 11.1b illustrates the empirical record. Clearly these are analytically distinct arenas, and parties may cooperate with different parties in each stage, though they risk incurring credibility costs if they do so. This is what happened in 1992: the Labour Party fought an independent campaign with no electoral alliances. It was rewarded with its then best ever election result partly based on such a pro-change anti-Fianna Fáil message. Labour then dismayed many of its new voters by forming an ill-fated coalition immediately after the election with Fianna Fáil. This was seen as such a breach of trust that Labour was heavily punished at the next election.

¹⁷ Of the 29 cabinets listed in Table 11.1, 9 had no majority. A further 4 cabinets controlled exactly 50 per cent of the seats. Of the remaining 16 cabinets only 5 had an initial legislative strength of 55 per cent or above. Thus most Irish cabinets have had very small majorities, or no majority at all.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

368



Table 11.1b Electoral alliances and pre-electoral coalitions in Ireland, 1989–2018 Election date

Constituent parties

Type

Types of pre-electoral commitment

1989-06-15 1997-06-06

FG, PD FG, Lab, WP-DL

PEC EA, PEC

2007-05-24 2016-02-26

FF, PD FG, Lab FG, Lab

PEC EA, PEC EA, PEC

Written policy statement Written policy statement, Joint press conference Written policy statement Joint press conference Joint press conference

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC). Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other.

Alternatively, a party may cooperate with different partners in the ‘before and after’ phases of coalition building. An example occurred in 1989. During the campaign, Fine Gael and the Progressive Democrats issued a joint statement claiming that they were a viable alternative government to Fianna Fáil. The statement covers areas of policy agreement (‘An Agreed Agenda for Action’). Thus, they formed a ‘pre-electoral coalition’ but rather undermined the success of this venture by failing to also form an ‘electoral alliance’ (under STV this would manifest itself as an explicit pact between parties in the pre-electoral coalition to recommend that they exchange terminal transfers with the favoured partner). Neither leader issued a national direction on transfers but left it to local candidates to decide for themselves what was the best way of defeating Fianna Fáil. After the election the hoped for Fine Gael–Progressive Democrats coalition was not numerically viable (only 61 seats combined when 83 was needed for a bare majority). Instead the Progressive Democrats changed partners and formed a government with Fianna Fáil (83 seats). On other occasions, for example in 2016, the incumbent coalition (Kenny I between Fine Gael and Labour) formed a pre-electoral coalition and an electoral alliance but still failed to form after the election, in this case due to the ‘voluntary’ withdrawal of Labour following its electoral meltdown.

The bargaining process Coalition building in Ireland is a fairly unstructured process in which the party leaders examine their bargaining weights and then try to explore what might be feasible. There are no recognition rules and no one is institutionally designated to lead or chair the negotiations. Table 11.2 displays the historical record since 1989. For most of the history of the state there was never a choice of more than two alternative governments. As mentioned, if Fianna Fáil had a majority (or near majority) it would form a single-party government; if not, the existence of an

1989

1992 1993

1994

1997 2002 2007

2008 2011 2011 2016

2017

Haughey IV

Reynolds I Reynolds II

Bruton

Ahern I Ahern II Ahern III

Cowen I Cowen II Kenny I Kenny II

Varadkar I

0

0 0 0 3

0 0 1

1

0 1

2

Number of inconclusive bargaining rounds

FG, Inds (1) FG, FF* (2) FG, Inds, GP, Lab (3) FG, Inds

FF, GP, PD, Inds (1) FF, GP, PD

FF, Lab (1) FG, Lab, WP-DL, PD, FF FG, Lab, WPDL (1) FF, Lab, FG, WP-DL

FF, PD (1) FF, PD (2) FF, PD

Parties involved in the previous bargaining rounds

12 1

22 4 9

5

9

7 2 5

Bargaining duration of individual formation attempt (in days)

1

1 1 12 70

20 20 21

28

1 48

27

Number of days required in government formation

1

1 1 9 47

8 9

-

1 -

14

Total bargaining duration

6 47

21 49 28 25

117 59 52 51 57 57

2

3 5 1

6

3 3

2

Abstention

88

85 93 89

85

84 102

84

Pro

94 50

27 49 77 81

76

78 68 76

74

78 60

79

Contra

Result of investiture vote

Notes: * Fianna Fáil was involved in some of the bargaining rounds but did not seriously consider joining. The investiture vote concerns the votes on the nomination of a Taoiseach. Abstentions are not recorded and are an estimate of the total assembly size with the Chair’s vote, pro votes and contra votes subtracted.

Year in

Cabinet

Table 11.2 Cabinet formation in Ireland, 1989–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

370



alternative would depend on the willingness (or not) of Fine Gael and Labour to coalesce. This facilitated speedy government formation so that all governments (before 1993) were put together within the ‘official gap’ between the dissolution of parliament for the purposes of elections and the first scheduled meeting of the new Dáil. De Winter and Dumont (2010: 130) list the average formation durations for 17 countries. The average for 16 countries (i.e. excluding Ireland) is 23 days. Ireland averages 15.7 days. Of course this average figure for the entire 1944–2007 period disguises a more recent evolution to some more lengthy duration periods that reflects the more complicated bargaining environment as a wider range of options became possible. Thus Reynolds II took 48 days to form, Bruton I 28 days, and more recently Kenny II took 70 days. Patterns of government formation have closely corresponded to what Laver and Schofield (1990) described as ‘free style bargaining between elites’. There are no size or composition requirements, formateurs or informateurs (see Strøm et al. 1994). There is an investiture requirement so that a proposed Taoiseach must win a plurality vote of those voting in the legislature (Art. 13.1.1–2). It is important to note that a proposed candidate does not need an absolute majority of TDs but only a majority of those who vote. Thus, parties and/or independent TDs can help to facilitate the formation of a particular proposed government by deciding to abstain on the investiture vote. While this has always been true the importance of this detail was highlighted much more dramatically in 2016 when Enda Kenny won an investiture vote 59 to 49 (with 49 abstentions). Once the incoming Taoiseach is approved by the Dáil and subsequently appointed by the president, he (so far always a he) must return to the Dáil and subject his list of ministers to a second vote. The cabinet is invested collectively in that the Dáil must vote to accept or reject the nomination of cabinet ministers en bloc (Martin 2015).¹⁸

Composition and Size of Cabinets Since the modern coalition era began in 1989 all governments have been coalitions.¹⁹ Of the twelve cabinets formed between 1989 and 2017 four have been minimal wining coalitions, two have been surplus majority cabinets, and six have been minority cabinets. The average parliamentary strength of these twelve cabinets was just 50.3 per cent. Government’s with large and seemingly secure

¹⁸ The Taoiseach is not required to indicate portfolio allocations, which means that reshuffles are possible without a further legislative vote. But new additions to the cabinet must be approved by the Dáil (Martin 2015). ¹⁹ With the exception of the two Fine Gael-led cabinets that have formed since the 2016 election. These are technically single-party minority cabinets that each contain three independent cabinet ministers. Each was sustained by an external ‘confidence and supply’ arrangement with Fianna Fáil.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



371

majorities can happen;²⁰ for example Reynolds II in 1993 controlled 61 per cent of seats in the Dáil, and Kenny I in 2011 was supported by 68 per cent, but these are very much the exception rather than the rule. Government formation in Ireland has often been facilitated by the availability of non-party independent TDs who might be persuaded to support a prospective government in the investiture vote and might even be willing to offer ongoing legislative support in exchange usually for spending commitments in their constituencies. Independent MPs are not common in European politics,²¹ so this is clearly an interesting and unusual feature of coalition politics in Ireland. The presence of sometimes quite large numbers of independents increases the ability of minority cabinets, single-party or coalitions, to pass their key legislation. The number of independents was high in the early decades of the state, much lower in the 1960s and 1970s, but has been on the rise in recent elections. Ten or more independent TDs have been elected in each of the last four general elections, with 15 in 2011 (out of 166) and an astonishing 23 (out of 158) in 2016. Thus, currently about 15 per cent of TDs in Ireland do not belong to any party.²² For example, after the 1997 election Bertie Ahern formed the first of his three successive cabinets by putting together a minority coalition with the Progressive Democrats. The coalition with 81 votes in a 166-seat parliament was thus two or three votes short of a working majority. In 1997 six non-party independent TDs were elected so that Ahern was able to negotiate legislative support with three of them in return for mostly pork-barrel spending commitments in their respective constituencies (Mitchell 1999: 256–9). Thus the minority coalition had extra legislative support and won the investiture vote by 85 votes to 78. The 1997–2002 Fianna Fáil—Progressive Democrat coalition (Ahern I) lasted a full five-year term (1786 days)—quite stable for a minority two-party coalition propped up by support from three independents. In 2002 the economy was booming and the available data support the idea that ‘a feel good factor favoured the incumbent government in general and Fianna Fáil in particular’ (Garry et al. 2003: 140). Fianna Fáil won 81 seats (+ 4) on an increased vote, while the Progressive Democrats doubled its seats to 8 (on a slightly reduced vote). Given this arithmetic, combined with greater fragmentation among the opposition forces, Ahern chose to rebuild the outgoing coalition with the Progressive Democrats, rather than govern as a large Fianna Fáil minority government with negotiated legislative support from a few independents (Mitchell 2003). Ahern II also lasted a full five years. ²⁰ Despite its large majority Reynolds II fell apart in coalition conflict after only 674 days of a possible five-year term. ²¹ Weeks (2016: 209) calculates that of the 18 EU democracies that currently permit independent candidates for lower house election, there were just 19 independents elected at the first set of elections held in the 2010s. These were 1 in the United Kingdom, 3 in Lithuania, and the remaining 15 in Ireland. ²² For much more on the independent TD phenomenon see Weeks (2014, 2015, 2016).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

372



The actual process of government formation can be illustrated with a short account of what happened after the 2011 election. Some things about this formation were ‘typical’ of how Irish coalition governments are put together and some things were not, the changes due largely to the country’s unprecedented financial crisis. The outgoing lead governing party, Fianna Fáil, had collapsed in the earthquake election. Fine Gael achieved its highest ever number of seats (76), just six short of a bare majority. Labour also achieved a record seat total: 37. Given that Fine Gael would not contemplate a coalition with Sinn Fein (14 seats), Fine Gael had two main options. Either it forms a hopefully stable coalition with Labour or it negotiates an external support arrangement with some of the independents and/or small parties (of which there were 19). However, given the extreme economic crisis and the tough austerity measures that the incoming government would have to implement, neither Fine Gael nor Labour seriously contemplated refusing to govern together. Election day was 25 February 2011 (Friday). The new Dail was due to meet on 9 March—and it is during this approximately two-week period that most Irish coalitions are negotiated. The two party leaders, Enda Kenny of Fine Gael and Eamon Gilmore of Labour, first met on Sunday evening for 80 minutes (O’Malley 2011: 270; see also Leahy 2013). Among other things they discussed the portfolio split (with Labour pitching for 9–6 but eventually having to settle for 10–5) and who would get the department of finance. For the most part the party leaders do not directly take part in the detailed policy negotiations but are kept informed and serve almost like a political appeals court for matters that the negotiators cannot resolve. Each party appointed a negotiating team of three TDs, and being selected for this role is a good predictor of securing a cabinet position. The Fine Gael team was led by Michael Noonan (who would become the finance minister) and included Alan Shatter and Phil Hogan. The Labour team was led by Brendan Howlin (who would become the minister for public enterprise) and included Joan Burton and Pat Rabbitte. All six of the negotiators would become cabinet ministers. In an innovation, prompted by the financial crisis, both parties’ chief economic advisors (not elected TDs) were included as the fourth member of their respective teams. Negotiations commenced on Tuesday, 1 March, at Government Buildings. But rather than immediately getting down to inter-party negotiations (e.g. identifying the ‘red lines’ in each party’s election manifestoes) the first two days moved into seminar mode, whereby all of the negotiators had to listen to successively dire warnings and summaries by the country’s top economic, civil service, financial, and banking officials. Once the negotiations returned to inter-party issues, they compromised on most of the big economic questions. For example, Fine Gael had wanted to target reducing the deficit to 3 per cent of national income by 2014, while Labour favoured the slower pace of achieving this target by 2016. They split the difference on this and many other items, agreeing on 2015 for the deficit reduction. One other unusual aspect of this formation was that it was subject to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



373

external, indeed even international, veto-players: the agreed deal had to be within the parameters set by the EU/IMF bailout (O’Malley 2011: 272).

Portfolio allocation of cabinet ministers Portfolio allocations are decided during the coalition negotiations after the policy programme has been agreed. Having said that party leaders will have a good idea how many cabinet positions they are likely to get since Ireland like many countries follows an approximately proportional norm. Browne and Franklin (1973) first pointed out what they called a ‘relative weakness effect’ whereby small parties tend to get a little more than their proportional share. Ireland cannot indulge in portfolio inflation (at least at cabinet level) because the maximum number of cabinet ministers is restricted to 15 by the constitution. Thus the set of 15 ministries are a fixed prize that has to be divided in a zero-sum manner among the coalition parties. Most allocations are fairly proportional, but it is striking how frequently smaller parties are able to successfully bargain and gain themselves one ‘extra’ cabinet position. For example, since 1973 the Labour Party has taken part in six coalition governments and has been allocated an ‘extra’ cabinet seat in four of them. Another way to look at it is that in seven of these ten coalition governments the largest party has had to give up one (and in one occasion two) cabinet ministers that it was proportionally entitled to (Table 11.3). There is clearly also some relationship between the core policy concerns of some parties and the portfolios that they secure during negotiations. We can see this most clearly with the case of the Labour Party, the only smaller party that has been in enough coalitions for us to fairly reliably discern a pattern. Labour has taken part in eight coalition governments between 1948 and 2016. It is striking that in six of these eight governments Labour has filled the social welfare cabinet portfolio (and on one occasion when it did not the social welfare cabinet minister was from Democratic Left, a party that subsequently merged with Labour). Labour (1948–2016) has had a total of 36 cabinet ministers. Two-thirds of them have been from just six departments (social welfare six times, health four, industry four, employment four, education three, and foreign affairs three times).

Coalition agreements All coalition governments in Ireland have had a coalition agreement. In almost all cases these have been post-electoral written coalition agreements (see Table 11.4). Irish parties mostly prefer to conclude negotiations after elections in order to emphasize their independent appeals during the election campaign and hence maximize their first preference votes. The two exceptions are 1973 when Fine Gael

Year in

1989 1992 1993 1994

1997 2002 2007 2008

2011

Cabinet

Haughey IV Reynolds I Reynolds II Bruton

Ahern I Ahern II Ahern III Cowen I

Kenny I

13 FF, 2 PD 13 FF, 2 PD 10 FF, 5 Lab 8 FG, 6 Lab, 1 WP-DL 14 FF, 1 PD 13 FF, 2 PD 14 FF, 1 PD 12 FF, 2 GR, 1 Ind. 10 FG, 5 Lab

Number of ministers per party (in descending order)

15

15 15 15 15

15 15 15 15

Total number of ministers

6 Lab, 4 FG

1 FF, 1 PD 2 FF, 1 PD 7 FF, 4 Lab 4 FG, 4 Lab, 3 WP-DL 2 FF, 2 PD 4 FF, 2 PD 6 FF, 1 GR 3 FF, 1 GR

Number of watchdog junior ministers per party

16

15 15 15 15

17 16 17 17

Number of ministries

Table 11.3 Distribution of cabinet ministerships in Irish coalitions, 1989–2018

FG

FF FF FF FF

FF FF FF FG

1 Taoiseach

FG, Lab

FF FF FF FF

FF FF FF Lab

2 Finance

Lab

FF FF FF FF

FF FF Lab Lab

3 Foreign affairs

Lab

FF FF FF FF

FF FF FF WP-DL

4 Social welfare

FG

FF FF FF FF

FF FF FF FG

5 Justice

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Year in

1948 1954 1973 1981 1982 1989 1992 1993 1994 1997 2002 2007 2008 2011

Coalition

Costello I Costello II Cosgrave I FitzGerald I FitzGerald II Haughey IV Reynolds I Reynolds II Bruton Ahern I Ahern II Ahern III Cowen I Kenny I

119 750 1,206 9,751 5,627 4,817 4,817 18,593 13,338 7,351 14,222 33,000 33,000 23,000

Size

0 0 0 0 0 0 0 0 0 0 0 0 0 0

General rules (in %) 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Policy-specific procedural rules (in %)

Table 11.4 Size and content of coalition agreements in Ireland, 1944–2018

0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of offices (in %) 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of competences (in %)

100 100 100 100 100 100 100 100 100 100 100 100 100 100

Policies (in %) OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

376



and the Labour Party agreed to a pre-electoral manifesto and 1994 when for the first and so far only time one coalition replaced another without an election and hence wrote its own agreement. These written and published documents are not coalition ‘agreements’ in the wider sense, but rather they are coalition policy documents. Thus they do not discuss general rules of coalition behaviour, procedural rules, or the distribution of offices or competences (see Table 11.4). They are all about the initial policy bargain on which the coalition is founded. These coalition agreements have evolved over the decades and have become much more detailed and hence more lengthy. Of the coalition agreements analysed across 12 European countries (Müller and Strøm 2010: 173) the average length was 7,511 words, and in Ireland 10,161. While the lengthening of these agreements does not follow a strictly unilinear pattern, the overall trend is clearly towards longer and more comprehensive documents. At one extreme consider the first-ever ‘coalition agreement’. On 17 February 1948 Fine Gael leader Richard Mulcahy issued a statement to the press on behalf of Fine Gael, Labour, National Labour, Clann na Poblacta, Clann na Talmhan, and the independents supporting the prospective coalition. The five parties had agreed on ten points of policy. The agreement amounted to 119 words (National Archives of Ireland, S10719 A, D/T memo, 31.3.48, reproduced in McCullagh 1998). The agreement contained some specific pledges like cutting taxes on beer and cinema tickets, improving the treatment of tuberculosis (here the government was very successful), and pledges of a vaguer variety, like reducing ‘the cost of living’. John A. Costello in his speech accepting his appointment as Taoiseach of Ireland’s first coalition government said: The various Parties who have formed this Government have sought to find, and have found, numbers of points on which they can completely agree. This Government has been formed on the basis of full agreement on all those points. Any points on which we have not agreed have been left in abeyance. (Dáil Éireann Debates, 18 February 1948, vol. 110, col. 77)

The second coalition led by the same Taoiseach formed in 1954 on a similar basis. Having learnt from some of the difficulties of policy implementation during the first coalition, the Labour Party wanted to drive a harder bargain and agree a somewhat more detailed policy agreement. Labour said that it would only take part if the coalition was publicly committed to an agreed policy programme ‘in broad conformity with Labour policy’. Furthermore, Labour in 1952 decided that a special delegate conference would have to approve any proposal to enter government, helping ensure the inclusion of enough ‘Labour inspired’ policy content. A historian of the Labour Party noted: ‘Labour had learned from experience that if the devil is in the detail, it was best to summon these demons and deal with them at the outset’ (Puirseil 2007: 175). Following the election result,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



377

McCullagh (2010) reports that intensive negotiations between Fine Gael and Labour followed on a coalition agreement, with a sequence of position papers being exchanged. For example, on social welfare Fine Gael’s proposals were vague (promising to ‘improve social welfare’)—Labour responded with much more specific policy demands. Fine Gael needed Labour to have any chance of forming a government, so in this key area for the smaller party, all of Labour’s proposals were included in the draft government programme (JACP P190/551). On 31 May 1954 Fine Gael and Labour held simultaneous press conferences to announce a coalition programme of ‘12 objects’ (previously passed by a Labour Special Delegate Conference). The 1954 coalition agreement was 750 words, very short by contemporary standards, but three times longer than the previous agreement in 1948. Coalition agreements as we currently know them expanded in length in the 1980s (to 5,000–10,000 words), in the 1990s (to 10,000–20,000 words), and since then to 20,000–33,000 words. Interestingly, the governments that formed after the 2016 election (Kenny II; and then in 2017 Varadkar I continuing with the same agreement) had the longest ‘coalition agreement’ ever though under our project rules it is not technically a coalition—because there is only one ‘party’ in the cabinet. Kenny II and Varadkar I are Fine Gael-led governments with three independent cabinet ministers (see more later). The ‘Programme for a Partnership Government’ was an agreement between Fine Gael, the Independent Alliance,²³ and a number of other independents. It is composed of 156 pages, and 42,182 words.

Government formation during and after the financial crisis Until 2016 all Irish governments either had a majority or had been ‘large’ minority governments close to a legislative majority and often facilitated by explicit or implicit external support deals with a number of independent deputies. But in 2016 the party system fragmented to its highest ever level of 4.93 (ENP), so that the two largest parties were no longer very large. Fine Gael won 32 per cent of the seats (from only 25 per cent of the votes), while Fianna Fail had a small recovery to 28 per cent of the seats (from 24 per cent of the votes). In the absence of even a single large party as a pole of coalescence government formation was less certain than ever. In other countries, it might reasonably be expected that these two would form a coalition. Such a coalition would be a clear minimal winning coalition with 93 seats in a down-sized Dáil in which the absolute majority threshold was ²³ The Independent Alliance did not register as a political party. The group was established by five independent TDs and two Senators. ‘In a group charter, it outlined ten principles and priorities but did not devise a constitution or rule book’ (Reidy 2016: 66). The Alliance’s ‘leader’ joined the cabinet.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

378



now 79.²⁴ Fianna Fáil and Fine Gael had always refused to govern together,²⁵ and this was a major bargaining constraint, as increasingly does the refusal of the other parties to coalesce with Sinn Féin. In 2016 it took 70 days to form a government, a national record for Ireland, although far short of a European record. As mentioned earlier, an important feature of the institutional rules is that to win an investiture vote a proposed candidate for Taoiseach only needs a simple majority of those who choose to vote, not an overall majority of all TDs. Thus, while an absolute majority would require 79 votes (if everybody votes) if either Fianna Fáil or Fine Gael abstain on the investiture vote then Kenny could be elected with 58 votes and Micheál Martin could be elected with 54 votes. This means that Kenny needed the support of at least eight more TDs and Martin at least eleven more (O’Malley 2016: 262). Thus, both Fine Gael and Fianna Fáil opened up talks with some of the smaller parties and independents to see who could attract the most support and to try to avoid an immediate second election. The already scheduled first meeting of the Dáil was held on 10 March, at which a Fianna Fáil member was elected Ceann Comhairle (Speaker) for the first time by secret ballot. Fine Gael leader Kenny was proposed for Taoiseach and defeated, with 94 against and 57 for. The leaders of Fianna Fáil, Sinn Féin, and the AntiAusterity Alliance–People Before Profit were then proposed and all defeated by larger margins. Having lost the investiture vote Kenny was then required to tender his resignation to the president, though of course he and the outgoing government continued to serve in a caretaker role. The Dáil met on two further occasions and failed to elect anybody as Taoiseach. Labour, heavily damaged in the election, effectively ruled itself out. After the second failed meeting of the Dáil, Kenny’s office issued a statement saying that he had made an offer of ‘full partnership government’ to Fianna Fáil with an equal number of cabinet positions for each party. Many in Fianna Fáil saw this as an insincere trap, that Fine Gael knew Fianna Fáil would refuse (it would need to pass a Fianna Fáil special party conference, which was unlikely)—they believed it was designed to shift the blame to Fianna Fáil if an early election was required. Negotiations between Fine Gael and Fianna Fáil finally opened on 11 April and then switched venues a week later to Dublin’s Trinity College. Finally, on 3 May both parties signed an 1,800-word agreement—not a Programme for Government—but a ‘confidence and supply’ arrangement by which Fianna Fáil would facilitate the election of Fine Gael’s Kenny, in return for various policy concessions (see O’Malley 2016 for full details). Negotiations with independents continued so that finally on 6 May Kenny was elected Taoiseach by 59 votes for ²⁴ At the 2020 election the Dáil was increased from 158 to 160 seats. ²⁵ Labour has long fantasized about the merger of Fianna Fail and Fine Gael, leading to a left–right realignment of the party system. But as Gallagher (1982: 168) observed Fianna Fail and Fine Gael are each ‘steeped in [their] own traditions, and for many members, opposing the other party at every turn is the essence of political activity; a merger would take all the fun out of politics’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



379

and 49 against, with Fianna Fáil abstaining. Kenny was supported by five votes from the Independent Alliance (which had agreed to ministerial appointments with him) and by positive votes from four other independents. The cabinet that formed contained twelve Fine Gael ministers and three independents (a further three independents became junior ministers). Thus, the 2016 cabinet with only 59 votes out of 158 was a government with the smallest legislative coalition in the history of the state.

Coalition governance There are no real ‘official rules’ of coalition politics. The practice of coalition governance (see Tables 11.4 and 11.5) has evolved from the early coalitions that made essentially ad hoc minor adaptations to the policy and decision-making procedures that prevailed during earlier single-party governments to the more recent practice of the coalition parties adjusting the policy process and informal norms of behaviour, in order to facilitate better coalition governance²⁶ For example, during one of the early coalitions (Fine Gael–Labour, 1973–1977) the ministers chose not to act as party blocs within the cabinet. Indeed, the ministers of each party did not even sit together in party teams: ‘the decision not to do so was taken at the start, as a token of the two parties’ willingness to operate amicably’ (Gallagher 1982: 219). A lot has changed since then. There have been at least three significant changes since the 1980s that have moved Ireland closer to the ‘coalition compromise model’ (see Chapter 2, this volume), whereby parties simultaneously attempt to promote their own policies and scrutinize the activity of their coalition partners, via the evolution of a range of monitoring and credible commitment devices. First, coalition agreements became much more detailed and were taken more seriously as guidelines for the policy priorities of the government. Second, and connected with this, much more explicit methods of monitoring the implementation of the government programme were developed. The exact practices have varied somewhat by cabinet, but beginning in the 1990s there was a much greater use of ministerial ‘special advisors’ and even ‘partnership programme managers’,²⁷ whose essential job was to look after the partisan interests of the minister and his or her coalition party, as well as facilitating better cooperation with the relevant part of the civil service. In addition, and more recently, some coalition governments are publishing annual ²⁶ Given the earlier normative prejudice against even the concept of coalition governments (constantly encouraged by Fianna Fail) some of the early coalitions almost pretended that they were not coalitions. The first two coalitions avoided the word ‘coalition’ calling themselves ‘inter-party’ governments. ²⁷ In September 2020 it was announced that at least 10 of the junior ministers would get their own special advisors.

1948 1954 1973 1981 1982 1989 1992 1993 1994 1997 2002 2007 2008 2011

POST POST PRE POST POST POST POST POST IE POST POST POST POST POST

Y Y Y Y Y Y Y Y Y Y Y Y Y Y

N N N N N N N N N N N N N N

Parl, PS Parl, PS Parl, PS Parl, PS Parl, PS Parl, PS Parl, PS Parl, PS Parl, PS Parl, PS Parl, PS Parl Parl, PS CoC, Parl, PS

All used

PS PS PS PS PS PS PS CoC CoC PS PS PS PS CoC

Parl Parl Parl Parl Parl Parl Parl Parl Parl Parl Parl Parl PS CoC, Parl

Y Y Y Y Y Y Y Y Y Y Y Y Y Y

N N N N N N N N N N N N N N

Personal Issues union excluded from Most For most agenda common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

All/All All/All Most/All All/All Most/All All/All All/All All/All All/All All/All All/All All/All All/All All/All

Y Y Y Y Y Y Y Y Y Y 1 1 1 1

No No Few Varied Varied Varied Varied Varied Varied Varied Varied Comp. Comp. Comp.

Y Y Y Y Y Y Y Y Y Y Y Y Y Y

Y Y Y Y Y Y Y Y Y Y Y Y Y Y

Coalition Freedom of Policy Junior Nondiscipline appointment agreement ministers cabinet in (watchdogs) positions legislation/ other parl. behaviour

Notes: Coalition agreement: PRE = pre-election; IE = inter-election; POST = post-election; N = no coalition agreement Conflict management mechanisms: CoC = Coalition committee; Parl = Parliamentary leaders; PS = Party summit Coalition discipline: All = Coalition discipline is expected on all matters; Most = Coalition discipline is expected on all matters except those explicitly exempted Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement

Costello I Costello II Cosgrave I FitzGerald I FitzGerald II Haughey IV Reynolds I Reynolds II Bruton Ahern I Ahern II Ahern III Cowen I Kenny I

Coalition

Table 11.5 Coalition governance mechanisms in Irish coalitions, 1944–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



381

monitoring reports on the coalition programme, some of which are longer than the original coalition agreement. Third, there has been starting in the early 1980s but escalating in the 2000s a significant growth in the numbers of ministers of state (Ireland has one level of junior minister). It was mentioned earlier that portfolio inflation is not possible in the cabinet,²⁸ but this is not true at the junior ministerial level.

The role of individual ministers in policy-making The style of cabinet decision-making is structured by Article 28.4.2 of the constitution: ‘The Government shall meet and act as a collective authority, and shall be collectively responsible for the Departments of State administered by the members of the Government.’ The cabinet acts as a clearing house for most major decisions and the doctrine of collective cabinet responsibility ‘normally denies ministers the right to record private dissent, let alone public opposition, to cabinet decisions’ (Farrell 1993: 174). The first coalition government (1948–1951) was unusual in this respect in that it had quite frequent breaches of collective responsibility and a number of instances of ministers trying to interfere in each other’s’ jurisdictions. In response to a question Taoiseach Costello said that in the inter-party government ‘it was considered permissible for a Minister in an individual or party capacity to give public expression to views which might not necessarily be those of the government as such’ (quoted in McCullagh 1998: 53). The opposition made much political capital from these publicly aired divisions between ministers and coalition parties, so that during the 1954 election campaign (which would lead to the second coalition government, 1954–1957) Costello declared that collective cabinet responsibility would be observed the same way as in a single-party government (McCullagh 2010). While of course there have been occasional breaches of this doctrine in general it is upheld.

²⁸ Even here though creative politicians can try and find a way around this problem. In 1994 Democratic Left only managed to secure one cabinet post, which was what they were proportionally entitled to. They wanted a second representative in cabinet, so one of their ministers of state was designated as a ‘super junior’ minister with the right to attend cabinet but without a vote (the argument being that the lack of a vote would avoid contravening the constitution). This way the sole Democratic Left cabinet minister had a party colleague at cabinet, and the cabinet does not vote anyway. Indeed, in practical terms the main substantive manner in which the ‘super junior’ is not a full cabinet minister is not really to do with the lack of a vote and more to do with the fact that the ‘super junior’ will have an immediate ‘boss’, the cabinet minister of the ministry in which he or she is a junior. This practice of appointing a super junior was repeated by some subsequent governments, for example to give the Progressive Democrats a second presence in cabinet. Furthermore the cabinet formed in 2016 (not technically a ‘coalition’) appointed two ‘super junior’ ministers, one from Fine Gael and one independent. In 2017 three were appointed. In addition the government chief whip (a junior minister) automatically attends the cabinet but is not usually referred to as a super junior.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

382



The Taoiseach is the boss both formally and in practice. The Taoiseach’s ministerial nominees will be appointed by the president provided that they have been previously approved by Dáil Eireann (Art. 13.1.2). Clearly, the Taoiseach’s power over ministerial selection is amended during coalition governments, so that the leaders of the other governing parties pick their own ministers.²⁹ Similarly the Taoiseach can fire any member of the government ‘for reasons which to him seem sufficient’ (Art. 28.9.4) though again in practice a Taoiseach cannot fire a minister of another party without that party leader’s consent. There is a far-reaching division of labour in government and this gives decision-making a departmental structure in which each cabinet minister has considerable agenda powers to determine which policy options are brought to the cabinet for a formal decision. However, while energetic and determined ministers no doubt have considerable scope to shape the policies of their departments, this autonomy should not be exaggerated. Particularly during coalitions party leaders cannot afford to allow full ministerial discretion since they would then lose all policy input in jurisdictions in which they did not control the minister. Parties have strong incentives to underwrite the credibility of their key policies—and attempt to commit their coalition partners to them— by negotiating their inclusion in the coalition policy document and subsequently policing these promises by a variety of devices designed to monitor progress towards the party’s goals (Mitchell 1999). Nevertheless, most policy-making does take place within departments. Ministers have clear agenda-setting powers—they decide which policy options emerge from their department for decision elsewhere. It is virtually unheard of for a policy proposal to emerge from one department that is clearly the jurisdiction of another department. O’Malley and Martin (2018: 258) quote former Taoiseach John Bruton as saying: ‘most ministers spend 90 per cent of their time immersed in their own departments, with only 10 per cent concerned with the rest of the government’ (John Bruton, Irish Times, 1 November 2014).

Coalition governance in the executive arena There is a Cabinet Handbook that sets out, at least in theory, the rules and procedures. The cabinet normally meets once a week at 10 a.m. on Tuesday mornings when the Dáil is in session and each Wednesday when it is not. The agenda for each government meeting is confidentially available online to ministers, their private secretaries, secretaries general and other approved high-level

²⁹ Technically, the nominees are recommended to the Dáil as members of the government rather than to particular departments. This underlines the Taoiseach’s power to alter assignments or jurisdictions at will.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



383

users. The documents relevant to each agenda item are also available online. The main agenda is finalized on the previous Friday morning. The supplementary agenda is finalized at 4.00 p.m. on the day preceding the meeting. O’Malley and Martin (2018: 253) report that while the cabinet is still an important forum a large amount of the policy discussions and potential conflicts are discussed in advance of the cabinet in other places. The cabinet’s job is to take a decision on the broad principle of a proposal and examine the ‘political implications’ of a proposal, that is to try and ensure that the proposal will not lead to bad publicity or worse a public revolt against the policy. ‘There are a number of reasons for the move away from detailed policy formation and debate at cabinet. One is that less contentious items are cleared in advance by the team of special advisors, especially those of the Taoiseach and Tánaiste. Another is the expansion of the cabinet committee system’ (O’Malley and Martin 2018: 253). Historically, cabinet committees were not used very often before the 1990s. There are, however, some examples from the distant past. The first coalition, partly perhaps in an effort to hold the five parties together, used cabinet committees quite extensively. There were 6 in 1948, 9 in 1949, and 16 in 1950. In 1949 a cabinet committee of four members (Costello (PM), Norton (the Labour leader), McGilligan (minister for finance) and O’Higgins) was set up to examine all ‘outstanding estimates’. This became known as the ‘Estimates Committee’ and became powerful, much like an inner cabinet. McCullagh (1998: 63–4) comments: ‘the Estimates Committee had become extremely powerful—and in doing so usurped some of the functions of the Minister for Finance.’ Cabinet committees, when they exist, are set up by each administration and end with that government (Table 11.5). A more recent example is the 2011–2016 coalition between Fine Gael and the Labour Party—the government that had to deal with the aftermath of the financial crisis. Given the scale of the emergency this coalition had to ‘work’. During the negotiations on the Programme for Government they decided on two institutional innovations. Firstly, the Department of Finance would be split into Finance and a new Department of Public Expenditure, politically allowing both coalition parties to have a finance ministry. Secondly, an inner cabinet was set up—the Economic Management Council (EMC). The EMC was made up of four cabinet ministers, the Taoiseach, the Tánaiste, and the two finance ministers. This was understandably attractive to Labour. While they were outnumbered two-to-one in the full cabinet—it would be fifty-fifty in the inner cabinet. Meetings of the EMC would contain these four cabinet ministers, plus their key policy advisors and top civil servants. The EMC met weekly. Initially, the idea was that it would meet at 8 a.m. on the day of cabinet meetings (which started in this government at 10.15 on Tuesdays). This was found to be too crowded an agenda since the Labour and Fine Gael ministers held separate meetings at 9 a.m. Many EMC meetings are reported to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

384



have gone on for several hours so the scheduling had to be changed. They were moved to Wednesday afternoons—and thus they became the start of the weekly political schedule (Leahy 2013: 111). Leahy (2013: 101) comments: the EMC ‘would turn out to be the engine of government, the forum in which almost all the most important decisions of the new administration would be made, deciding all economic and budgetary policy’. He also reports, not surprisingly, that the EMC was ‘a source of constant unhappiness’ among those cabinet ministers that were excluded (Leahy 2013: 112). Junior ministers (in Ireland ‘ministers of state’) are used to achieve portfolio balance between the coalition partners and to stress party interests. There has been a substantial growth in the number of junior ministers. In 1948 only three were appointed, and generally throughout the 1950s to 1970s there were typically five to seven in each government. The number of junior ministers rose sharply in the 1980s—with typically around 15 in each government. The number peaked in 2007 and 2008 when 20 were appointed. Their numbers were reduced in 2009 and 2011 from 20 to 15—this was explicitly presented as part of cost-cutting measures in response to the financial crisis. This retrenchment may well prove to be temporary given the inherent attraction to party leaders of distributing patronage to their parliamentary parties. In 2016 and 2017, 18 and 19 junior ministers respectively were appointed. There is some evidence that the distribution of junior ministers is intended to allow each party to monitor what their coalition partner is doing in the cabinet portfolios that they hold. There seems to be some positive evidence of this watchdog role in the Irish data. Clearly the evidence varies by cabinet, principally it seems according to the sheer size of junior coalition partners. For example, very little oversight will be possible if the junior coalition partner is very small—for example the Progressive Democrats in 1997 had only two junior ministers, and in 2007 it got zero junior ministers. However, when there is a closer balance of strength between the coalition parties, there appears to be more evidence of ‘divided portfolios’. When, for example, Labour is relatively strong (after the 1992 and 2011 elections for example), 10 of the (15) ministries have a cabinet minister and junior minister from different parties, suggesting that in some cabinets they do partially play a watchdog role. However, this should probably not be exaggerated: junior ministers in Ireland have specific policy roles within the department and the junior minister can be overruled by the department’s minister.

Implementing and monitoring coalition agreements It is clear that the coalition programme is central to the work of the government: senior civil servants increasingly treat the relevant section as a plan of work for

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



385

their department. The coalition programme plays a key role in the work of the cabinet and the relations between the parties. One former Taoiseach, Albert Reynolds, has been reported as saying about an issue that was causing division in cabinet: ‘if it’s in the programme, then it’s a deal; if not it’s up for discussion’ (private interview conducted by Eoin O’Malley; see O’Malley and Martin 2018: 246). Costello et al. (2016) report some very interesting findings concerning the 2011–2016 coalition government (Kenny I). First, inclusion of a policy item (a pledge) in the Programme for Government ‘has a very significant impact on the likelihood of pledge fulfilment: 78 per cent of all election pledges that featured in the Programme for Government were fulfilled at least in part, compared to 46 per cent of other pledges’ (Costello et al. 2016: 37). Second, they report the marginal effects of variables (like controlling the relevant ministry, being in the programme for government, consensus with government party, etc.) on the probability of pledge fulfilment. Controlling for other factors, they found that ‘the likelihood of [pledge] fulfilment increases by 12 per cent if the party goes on to hold the ministry relevant to that pledge’ (37). But they found that the inclusion of the pledge in the Programme for Government has an even bigger effect than controlling the relevant ministry. Again controlling for other factors, if the item was in the coalition agreement it was 29 per cent more likely to be fulfilled than if it was not. This is quite strong evidence that the increasingly detailed coalition agreements are a very important commitment device during the life cycle of coalition governments. While coalition agreements have been increasingly monitored especially since the innovations of the 1992–1997 coalition government, involving the much greater use of special advisors and programme managers (Mitchell 2000), more recently governments have been publishing annual monitoring reports on the coalition programme. These are very detailed annual reports, some much longer than the original coalition agreement. For example, the 2011 Programme for Government, which was the coalition policy document on which the 2011–2016 government was based, totalled 23,172 words. Five annual monitoring reports were published during the life of this government, ranging in size from 15,793 to 43,774 words (averaging 30,000).

Coalition duration and termination Considering all of the cabinets (1944–2017) the average duration in Ireland is 971 days, about two years and eight months. This is perhaps shorter than expected but is partly an artefact of how we count cabinets. For example, there are four occasions on which a Taoiseach ‘voluntarily’ retired to facilitate succession in a fairly dignified manner (De Valera, Lemass, Lynch, and Kenny). The governments that they led were not brought down by conflict. The three cabinets that followed

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

386



the first three of these changes of leader lasted on average for a further 609 days.³⁰ Rather than these ‘terminations’ being about insurmountable conflict they are more like changing the team captain at half time. To the best of my knowledge this does not happen a lot in football but it does periodically in Irish politics! In addition, being PM in Ireland seems to be a fairly healthy career choice: to date in the history of the state no PM has ever died in office (one of the other technical reasons for termination in Table 11.6). The third technical reason for termination—a regularly scheduled parliamentary election—has only been the principal cause of termination on three fairly recent occasions (in 1997, 2002, and 2016). A key reason for this of course is that Ireland is a system of government in which the PM has dissolution powers. The attempted strategic timing of elections is (at least believed) to be an important resource for the PM’s party and few PM’s have wanted to risk being ‘boxed-in’ (diminishing the other parties’ uncertainty) by governing into the last few months of the maximum term, especially since it is a five-year term. Governments in Ireland have been terminated by combinations of almost the full range of discretionary mechanisms listed in Table 11.6 (except that there has never been a voluntary enlargement of the coalition). To balance this rather rosy picture of the life of an Irish Taoiseach (given earlier) we should note that it is possible to get kicked out. For example, the resignations of Haughey in favour of Reynolds and Reynolds in favour Ahern were due to severe coalition and intraparty conflicts. In the first case, Reynolds replacing Haughey saved the coalition; in the second it did not. The resignation of Ahern (much lauded for his good work alongside Tony Blair in relation to Northern Ireland) was one of the low points in recent Irish politics. With Fianna Fáil back in government in 2007 in a surplus majority three-party coalition many observers wondered if the party would ever be removed from its dominance of the coalition era. Certainly, none predicted that by the next election in 2011 the Progressive Democrats would no longer exist,³¹ the Greens would lose all of their seats, and Fianna Fáil would be punch drunk from losing nearly three quarters of its parliamentary party. The 2007–2008 cabinet (Ahern III) was ended by the forced resignation of the Taoiseach caused by bad publicity in relation to a complicated series of personal financial scandals for which Ahern appeared to have no plausible answers. Ahern was subsequently expelled from Fianna Fáil. In ³⁰ We cannot include Kenny’s resignation since the replacement Vardakar I had not yet been terminated at the time of writing. ³¹ The Progressive Democrats was nearly wiped out at the 2007 election, holding only two of their eight seats. It joined the three-party coalition with Mary Harney as its sole cabinet minister. In November 2008 the Progressive Democrats held a special delegate conference and voted by 201 votes to 161 to dissolve the party (Murphy 2016b: 131–2). Never before has a party in government in Ireland decided to dissolve itself. After the Progressive Democrats’ dissolution—its only minister stayed on in the cabinet as an independent for another two years. By convention this is not regarded as a new cabinet. But it does seem a grey area if one of the ‘parties’ in the coalition no longer exists!

Date in

1989-07-12

1992-02-11

1993-01-12

1994-12-15

1997-06-26 2002-06-06 2007-06-14

2008-05-07

Cabinet

Haughey IV

Reynolds I

Reynolds II

Bruton

Ahern I Ahern II Ahern III

Cowen I

2011-01-23

2002-05-17 2007-05-24 2008-05-06

1997-06-06

1994-11-17

1992-11-25

1992-02-11

Date out

67.1

100 100 18.1

84

37.9

33.7

52.5

Relative duration (%)

Table 11.6 Cabinet termination in Ireland, 1989–2018

7a, 8

1 1 9

4

7b

4, 6, 7b

7a, 7b

Mechanisms of cabinet termination

11, 13

14

14

Terminal events

7a: FF, GR8: FF

FF, Lab

FF, PD

FF, PD

Parties (when conflict between or within)

Justice

Policy area(s)

Continued

Ahern resigned after having become involved in a corruption case. Opinion poll—FF at 14 per cent led to leadership conflict in FF. Martin votes against Cowen on a confidence motion, and has to resign. The Greens left the cabinet citing difficulties working with FF, in particular because of its

Intense factionalism within FF; PM forced to resign by an ultimatum from coalition partner PD’s withdrawal from government after sharp personal conflict with PM Conflict between PM and Labour leader; PM resigns but Labour still leaves the government No conflict: a popular threeparty government holds an early election and loses.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2011-01-23

2011-03-09 2016-05-06

Cowen II

Kenny I Kenny II

2016-02-26 2017-06-14

2011-02-25

Date out

99.5 22.1

6.8

Relative duration (%)

Terminal events

Parties (when conflict between or within)

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event Note: Relative duration: share of constitutionally allowed duration for this cabinet.

4 2

4

Mechanisms of cabinet termination

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of PM

Date in

Cabinet

Table 11.6 Continued Policy area(s)

PM voluntarily resigns (retires) to allow succession.

leadership situation, where Cowen had resigned as leader of FF but wished to remain as PM—a proposal the Greens refused to accept. Cowen requested that the president dissolve the Dáil on 1 February, with early elections to be held on 25 February. Between the Green’s withdrawal from the cabinet in January and the request to dissolve the Dáil, the government managed to pass an expedited budget through the Dáil.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



389

general though we can see that Irish cabinets are often ended by ‘early’ elections and by conflicts within and between the respective coalition partners. In terms of measures of ‘relative duration’ three summary findings can be mentioned briefly. First, if we look at average relative duration across the decades there is very little variation to observe. From the 1940s to the 2010s all decades, except one, have average relative durations ranging only from 0.61 to 0.69 (the sole exception is the 1980s—0.42—largely because there were five general elections). Second, there is no difference (at all!) between the average relative durations of single-party and coalition cabinets (single-party—0.59; coalitions—0.59). Third, as expected there is a significant difference in average relative duration between minority and majority cabinets (0.49 and 0.68 respectively). Finally, serving in government in Ireland is usually electorally costly. Of the eight parties that have served in government no party has on average benefitted electorally from governing. Of course there are occasional examples of a party electorally gaining votes after service but overwhelmingly the evidence points to a negative incumbency effect. Of the three parties that have most often been in government, Fianna Fáil’s average loss is 2.8 per cent, Fine Gael’s is 3.3 per cent, and Labour’s is 4.4 per cent. The average loss of ‘the government’ (that is all governments) is 6 per cent, with coalitions doing worse (8.9 per cent) than singleparty governments (2.8 per cent).³² And of course almost all governments since 1989 have been coalitions. Clearly, parties are willing to pay the electoral price for the opportunity to govern.

Conclusion While other parties have come and gone the only three ever-present parties, Fianna Fáil, Fine Gael, and Labour (normally in that order of magnitude), so dominated Irish politics in the twentieth century that the party system was often loosely described as a ‘two and a half ’ party system. Irish politics decisively changed in 1989 when Fianna Fáil changed tactics and entered its first ever coalition. Before this, single-party Fianna Fáil governments were the default outcome of the government formation process. Before 1989 there were 5 coalition governments and 12 single-party governments (1932–1987). Since 1989 there has been 30 years of coalition governments. The only partial exceptions to this are the cabinets formed in 2016 and 2017 (both formed according to the same formula: single-party Fine Gael, with three independents in cabinet and an external legislative support arrangement with the main opposition party). These exceptions have largely been justified by the economic crisis, the associated electoral collapse

³² Authors calculations.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

390



of Fianna Fáil, and more recently by the existential, economic, and political threat of BREXIT. There are signs of party system fragmentation. The ENP (votes) averaged 3 in the earlier period (1932–1987) and rose to an average of 4.4 in the period since 1989. Associated with this there have been significant increases in the levels of electoral system disproportionality: the average was 3.7 (1932–1987) and this has risen to 5.8 (1989–2016). These summary indices reflect the declining dominance of the three ever-present parties. In the same earlier period, the combined vote of Fianna Fáil + Fine Gael + Labour averaged 88 per cent. The average share of these parties has declined to 75 per cent (1989–2016). Indeed, in the parliament elected in 2016, they only account for 56 per cent of the vote. Another way of looking at this fragmentation is that of the 158 seats allocated in 2016, Fianna Fáil + Fine Gael + Labour control 101 seats, while other parties and independents control 57. While it is possible that this greater level of fragmentation could be reversed, multiple opinion polls in recent years suggest that this is unlikely any time soon. With 7 parties and 23 independents or others groups in parliament after the 2016 election the number of coalition permutations are greater than they have ever been. There were two remaining behavioural constraints: the refusal (so far) of Fianna Fáil and Fine Gael to share an executive coalition and the refusal (so far) of all parties to strike a coalition deal with Sinn Féin, the third largest party.³³ Irish politicians have in the last 30 years learnt how to make coalitions work. It is no longer a case of forming a coalition but pretending as if it is a ‘normal’ singleparty government. These days incumbents will have spent time and a lot of energy in negotiating a detailed Programme for Government, a document that all regard as crucial for the subsequent life of the government. While not quite a holy grail, it codifies the initial coalition policy bargain and is a document that can be subsequently invoked as evidence that there has been too much agency loss and drift away from a party’s key policies. Recent governments have explicitly examined performance by producing annual monitoring reports to determine how much of the coalition programme has been implemented. These documents are sometimes more detailed than the initial coalition agreement. Similarly, parties in coalition have increasingly taken steps to coordinate policy and prevent excessive conflict through the use of cabinet committees and regular meetings of special advisors to pre-empt trouble. Meetings of the latter have become very important: ‘the growth of special advisors has reduced the influence of senior civil servants and shifted the

³³ Following the February 2020 election Fianna Fáil, Fine Gael, and Sinn Féin controlled respectively 23 per cent, 21 per cent, and 23 per cent of the seats. Fianna Fáil and Fine Gael at least for now continue to rule out Sinn Féin as a coalition partner. But after about four months of negotiations, the single biggest bargaining constraint has been eclipsed when on 27 June 2020 Fianna Fáil and Fine Gael agreed to govern together in a three-party coalition with the Greens. The office of Taoiseach (PM) is to ‘rotate’ between the leaders of Fianna Fáil and Fine Gael at the half-way point of the five-year maximum term.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



391

balance of power back in the direction of the minister’ (O’Malley and Martin 2018: 262). It seems likely that a probable area of coalition turbulence in the near future may not so much be in the realm of coalition governance but rather of coalition formation and this directly connects to party system fragmentation and the declining electoral strength of the two largest parties. At the 2020 election the respective sizes of the three larger parties (in rounded first preference votes) were Sinn Féin (25), Fianna Fáil (22) Fine Gael (21), and all the other parties have less than 10 per cent. The other parties claim they would like Fianna Fáil and Fine Gael to merge. But they are unlikely to do so. Hostilities will resume at the next election. They are currently governing together (with unhappy opinion poll results for Fianna Fáil). But if they both try to avoid governing together in the future and also refuse to accept Sinn Féin as a coalition partner, then it is hard to see an easy path to stable majority coalition governments, as was the case in 2016. Coalition formation bargaining may well become more challenging.

References Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010.’ Party Politics, 21(1): 143–52. Browne, Eric and Mark Franklin (1973). ‘Aspects of Coalition Payoffs in European Parliamentary Democracies’. American Political Science Review, 67(2): 453–69. Costello, Rory, Paul O’Neill, and Robert Thomson (2016). ‘The Fulfilment of Election Pledges by the Outgoing Government’. In Michael Gallagher and Michael Marsh (eds), How Ireland Voted 2016: The Election that Nobody Won. London: Palgrave Macmillan, 27–46. De Winter, Lieven, and Patrick Dumont (2010). ‘Uncertainty and Complexity in Cabinet Formation’. In Kaare Strøm, Wolfgang Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 123–57. Farrell, Brian (1993). ‘The Government’. In John Coakley and Michael Gallagher (eds), Politics in the Republic of Ireland, 2nd edn. Limerick: PSAI Press and Folens, 167–89. Farrell, David, Clodagh Harris, and Jane Suiter (2017). ‘Bringing People into the Heart of Constitutional Design: The Irish Constitutional Convention of 2012–14’. In Xenophon Contiades and Alkmene Fotiadou (eds), Participatory Constitutional Change: The People as Amenders of the Constitution. London: Routledge, 120–35. Gallagher, Michael (1982). The Irish Labour Party in Transition, 1957–82. Manchester: Manchester University Press. Gallagher, Michael (2000). ‘The (Relatively) Victorious Incumbent under PR-STV: Legislative Turnover in Ireland and Malta’. In Shaun Bowler and Bernard Grofman

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

392



(eds), Elections in Australia, Ireland and Malta under the Single Transferable Vote. Ann Arbor: University of Michigan Press, 81–113. Gallagher, Michael (2011). ‘Ireland’s Earthquake Election: Analysis of the Results’. In Michael Gallagherand Michael Marsh (eds), How Ireland Voted 2011: The Full Story of Ireland’s Earthquake Election. London: Palgrave Macmillan, 139–71. Gallagher, Michael (2018). ‘The Oireachtas: President and Parliament’. In John Coakley and Michael Gallagher (eds), Politics in the Republic of Ireland, 6th edn. London: Routledge, 164–90. Gallagher, Michael, and Lee Komito (2018). ‘The Constituency Role of Dáil Deputies’. In John Coakley and Michael Gallagher (eds), Politics in the Republic of Ireland, 6th edn. London: Routledge, 191–215. Gallagher, Michael, and Michael Marsh (eds) (2011). How Ireland Voted 2011: The Full Story of Ireland’s Earthquake Election. London: Palgrave Macmillan. Gallagher, Michael, and Michael Marsh (eds) (2016). How Ireland Voted 2016: The Election that Nobody Won. London: Palgrave Macmillan. Gallagher, Michael, and Richard Sinnott (1990). How Ireland Voted 1989. CSIE, University College Galway. Garry, John, Fiachra Kennedy, Michael Marsh, and Richard Sinnott (2003). ‘What Decided the Election?’. In Michael Gallagher, Michael Marsh and Paul Mitchell (eds), How Ireland Voted 2002. London: Palgrave Macmillan, 119–42. Laver, Michael (1994). ‘Party Policy and Cabinet Portfolios in Ireland 1992: Results from an Expert Survey’. Irish Political Studies, 9(1): 157–64. Laver, Michael (1998). ‘Party Policy in Ireland in 1997: Results from an Expert Survey’. Irish Political Studies, 13(1): 159–71. Laver, Michael, and Audrey Arkins (1990). ‘Coalition and Fianna Fáil’. In Michael Gallagher and Richard Sinnott (eds), How Ireland Voted 1989. CSIE, University College Galway, 192–207. Laver, Michael, and Norman Schofield (1990). Multiparty Government: The Politics of Coalition in Europe. New York: Oxford University Press. Leahy, Pat (2009). Showtime: The Inside Story of Fianna Fáil in Power. London: Penguin. Leahy, Pat (2013). The Price of Power: Inside Ireland’s Crisis Coalition. London: Penguin. MacDermott, Eithne (1998). Clann na Poblachta. Cork: Cork University Press. Mair, Peter (1986). ‘Locating Irish Parties on a Left-Right Dimension: An Empirical Enquiry’. Political Studies, 34: 456–65. Mair, Peter (1987). The Changing Irish Party System. London: Pinter. Mair, Peter (1990). ‘The Irish Party System into the 1990s’. In Michael Gallagherand Richard Sinnott (eds), How Ireland Voted 1989. CSIE, University College Galway, 208–20.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



393

Mair, Peter (1993). ‘Fianna Fáil, Labour and the Irish Party System’. In Michael Gallagher and Michael Laver (eds), How Ireland Voted 1992. PSAI Press and Folens, 162–73. Marsh, Michael, and Kevin Cunningham (2011). ‘A Positive Choice, or Anyone but Fianna Fáil’. In Michael Gallagherand Michael Marsh (eds), How Ireland Voted 2011: The Full Story of Ireland’s Earthquake Election. London: Palgrave Macmillan, 172–204. Marsh, Michael, and Paul Mitchell (1999). ‘Office, Votes and Then Policy: Hard Choices for Political Parties in the Republic of Ireland’. In Wolfgang Müller and Kaare Strøm (eds), Policy, Office or Votes. Cambridge: Cambridge University Press, 36–62. Marsh, Michael, David Farrell and Theresa Reidy (eds) (2018). The Post-crisis Irish Voter: Voting Behaviour in the Irish 2016 General Election. Manchester: Manchester University Press. Martin, Shane (2015). ‘Government Formation in Ireland’. In Bjorn Erik Rasch, Shane Martin and Jose Antonio Cheibub (eds), Parliaments and Government Formation: Unpacking Investiture Rules. Oxford: Oxford University Press. McCullagh, David (1998). A Makeshift Majority: The First Inter-Party Government, 1948–51. Dublin: IPA. McCullagh, David (2010). John A. Costello: The Reluctant Taoiseach. Dublin: Gill and MacMillan. McElroy, Gail (2017). ‘Party Competition in Ireland: The Emergence of a Left-Right Dimension?’. In Michael Marsh, David Farrell and Gail McElroy (eds), A Conservative Revolution: Electoral Change in Twenty-First-Century Ireland. Oxford: Oxford University Press, 61–82. Mitchell, Paul (1999). ‘Government Formation: A Tale of Two Coalitions’. In Michael Marshand Paul Mitchell (eds), How Ireland Voted 1997. Boulder, CO: Westview Press, 243–63. Mitchell, Paul (2000). ‘Ireland: From Single-Party to Coalition Rule’. In Wolfgang Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 126–57. Mitchell, Paul (2003). ‘Government Formation in 2002: “You Can Have Any Kind of Government as Long as It’s Fianna Fáil” ’. In Michael Gallagher, Michael Marsh and Paul Mitchell (eds), How Ireland Voted 2002. London: Palgrave Macmillan, 214–29. Müller, Wolfgang, and Kaare Strøm (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Müller, Wolfgang, and Kaare Strøm (2010). ‘Coalition Agreements and Cabinet Governance’. In Kaare Strøm, Wolfgang Müller and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 159–99.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

394



Murphy, Gary (2011). ‘The Background to the Election’. In Michael Gallagherand Michael Marsh (eds), How Ireland Voted 2011: The Full Story of Ireland’s Earthquake Election. London: Palgrave Macmillan, 1–28. Murphy, Gary (2016a). ‘The Background to the Election’. In Michael Gallagherand Michael Marsh (eds), How Ireland Voted 2016: The Election that Nobody Won. London: Palgrave Macmillan, 1–26. Murphy, Gary (2016b). Electoral Competition in Ireland since 1987. Manchester: Manchester University Press. Narud, Hanne Marthe and Henry Valen (2010). ‘Coalition Membership and Electoral Performance’. In Kaare Strøm, Wolfgang Müller and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 369–402. O’Malley, Eoin (2008). ‘Government Formation in 2007’. In Michael Gallagherand Michael Marsh (eds), How Ireland Voted 2007. London: Palgrave Macmillan, 205–17. O’Malley, Eoin (2011). ‘Government Formation in 2011’. In Michael Gallagherand Michael Marsh (eds), How Ireland Voted 2011: The Full Story of Ireland’s Earthquake Election. London: Palgrave Macmillan, 264–82. O’Malley, Eoin (2016). ‘70 Days: Government Formation in 2016’. In Michael Gallagherand Michael Marsh (eds), How Ireland Voted 2016: The Election that Nobody Won. London: Palgrave Macmillan, 255–76. O’Malley, Eoin, and Shane Martin (2018). ‘The Government and the Taoiseach’. In John Coakley and Michael Gallagher (eds), Politics in the Republic of Ireland, 6th edn. London: Routledge, 243–69. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of Antielitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Puirseil, Niamh (2007). The Irish Labour Party, 1922–73. Dublin: University College Dublin Press. Reidy, Theresa (2016). ‘Candidate Selection and the Illusion of Grass-Roots Democracy’. In Michael Gallagherand Michael Marsh (eds), How Ireland Voted 2016: The Election that Nobody Won. London: Palgrave Macmillan, 47–73. Sartori, Giovanni (1976). Parties and Party Systems: A Framework for Analysis. Cambridge: Cambridge University Press. Strøm, Kaare, Ian Budge, and Michael Laver (1994). ‘Constraints on Cabinet Formation in Parliamentary Democracies’. American Journal of Political Science, 38(2): 303–35. Strøm, Kaare, Wolfgang C. Muller, and Torbörn Bergman (eds.) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



395

Strøm, Kaare, Wolfgang Müller, and Torbjörn Bergman (2010). Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press. Weeks, Liam (2014). ‘Crashing the Party: Does STV help independents?’. Party Politics, 20: 604–16. Weeks, Liam (2015). ‘Why Are There Independents in Ireland?’. Government and Opposition, 51(4): 580–604. Weeks, Liam (2016). ‘Independents and the Election: Party Crashers’. In Michael Gallagherand Michael Marsh (eds), How Ireland Voted 2016: The Election that Nobody Won. London: Palgrave Macmillan, 207–26.

Appendix. List of political parties Abbreviation S-PBP

WP-DL SF GR SD Lab NL CnP CnT FF FG PD

Name Solidarity—People Before Profit, 2017– Anti-Austerity Alliance–People Before Profit, 2015–2017 United Left Alliance, 2010–2013 Democratic Left (Daonlathas Clé), 1992–1999 Workers’ Party (Páirtí na nOibrithe), 1970– Sinn Féin (Ourselves) Green Party (Comhaontas Glas) Social Democrats (Na Daonlathaigh Shóisialta) Labour Party (Páirtí an Lucht Oibre) National Labour Party (Páirtí Náisiúnta an Lucht Oibre) Clann na Poblachta (Party of the Republic) Clann na Talmhan (Party of the Land) Fianna Fáil (Soldiers of Ireland) Fine Gael (Irish Race) Progressive Democrats (An Páirtí Daonlathach)

Note: Party names are given in Irish or English, with a translation into the other in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 12 Italy: Continuous Change and Continuity in Change Francesco Zucchini and Andrea Pedrazzani

Few democracies in the world have experienced so many transformations in the electoral rules and party system as has Italy since the early 1990s. The study of the Italian case is then an excellent opportunity to investigate if and how these changes impact on the role of the government in the decisionmaking process, on government formation and termination, and on the governance stage. The first section of this chapter illustrates the institutional framework and the institutional practices characterizing the life cycle of government coalitions in Italy. Since the late twentieth century, these practices have changed much more than the formal rules. They favoured, temporary and de facto, the government in the decision-making process and the prime minister (PM) vis-à-vis the other ministers. The subsequent section summarizes the evolution of the Italian party system and traces it back to the interaction between electoral rules and electoral preferences. The next sections show how the evolution of the Italian party system shaped cabinet formation, duration, and termination, according to the post-electoral or interelectoral nature of the governments. While the party competition dynamics between 1994 and 2013 induced the political parties to build electoral alliances and pre-electoral coalitions, the persisting high level of internal fragmentation made Italian governments very unstable compared to the governments in many other European democracies. As discussed in Section ‘Coalition governance’, the instruments of intra-coalitional conflict resolution used in Italy have typically been quite informal and mostly based upon decision-making bodies at least partially external to the executive. Starting from the 1990s, however, the increased overlapping between government leadership and party leadership makes these mechanisms more internal to the government arena than in the past. Recent political developments from 2013 onwards—especially after the 2018 general elections—leave conclusions very open about the consolidation of the changes in the government arena.

Francesco Zucchini and Andrea Pedrazzani, Italy: Continuous Change and Continuity in Change In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0012

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



397

The institutional setting In parliamentary democracies, cabinets differ from each other depending on the ways in which a cabinet interacts with parliament and on the internal cabinet structure, particularly the power relationship between the PM and the other ministers. In this respect, in spite of the important changes in the party system and electoral rules that have taken place since the early 1990s (see Section ‘The party system and the actors’), the Italian political system has not experienced major institutional changes after the Second World War. Two ambitious attempts of constitutional reform—one by the centre-right government coalition in 2005 and one by the centre-left coalition in 2016—were stopped by popular vote in constitutional referenda.¹ Transformations have instead been incremental, informal, and driven by the evolution of party competition dynamics. In order to take office, the Italian government must be explicitly supported by an investiture vote carried by a majority of voting members of parliament (MPs), separately both in the Chamber of Deputies and in the Senate.² The ‘double’ confidence vote, together with the perfectly equal legislative prerogatives of the two chambers and a navette system, which means that a bill is passed from one house to the other until both agree, makes the Italian parliament the most symmetric bicameral system in the realm of European parliamentary democracies. No bill, including those sponsored by the government, can get passed unless the same text is approved in both chambers (Zucchini 2008, 2013). Notwithstanding the positive parliamentarism, minority governments have been quite frequent (25 per cent) in Italy. Most of the time, they were based upon an explicit agreement with some parties outside the government. In the legislative arena, the Italian government does not enjoy strong formal prerogatives (Döring 1995). Nevertheless, in practice, Italian governments—above all after the collapse of the so-called ‘First Republic’ in 1994 and the shift to a ‘Second Republic’—have been able to increase their de facto agenda-setting power by circumventing ordinary procedures and by relying upon special procedures and legislative ‘ploys’, such as law decrees and delegation laws (Zucchini 2011).

¹ In September 2020, after the end of the period covered in this volume, a popular referendum approved a constitutional reform reducing the number of members of the Chamber of Deputies (from 630 to 400) and the number of elected members of the Senate (from 315 to 200). This reform did not alter the prerogatives of the two legislative branches, nor it affected the relationship between parliament and government. ² Until recently, the criteria to identify the voting members had differed between the Chamber and the Senate. In the Chamber, those MPs who abstain are not considered as voting members. Nevertheless they are counted in order to reach the legal threshold that allows to qualify a legislative session as valid. In the Senate, on the contrary, up to 2018 the MPs who abstained had been considered as voting members. However, at the start of the XVIII Legislature in 2018 the Senate adopted the same counting rules of the Chamber.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

398

  

With respect to the cabinet internal structure, no formal rule assigns to the PM strong prerogatives vis-à-vis the other ministers. Most of the PM’s authority during the Second Republic derives from an implicit electoral mandate. The names of ministers are proposed by the appointed PM and approved by the president of the Republic (Art. 92 of Italian constitution). Once ministers are in office, they cannot be dismissed by the PM. This prerogative belongs instead to the parliament. Since 1984 in the Senate, and few years later in the Chamber, an interpretation of the Standing Orders allows for a motion of no confidence against a single minister. Although Italian opposition parties have repeatedly tried to get rid of a minister through this procedure, only in one case (in 1995), during Lamberto Dini’s caretaker government, was a similar motion approved. The attempts to regulate by law the number of ministers and ministerial portfolios have been often circumvented. The overall number of ministers (and of viceministers and junior ministers) has usually been proportional to the level of fragmentation and heterogeneity of the governing coalition. The noticeable growth of the premiership’s department (presidenza del consiglio) and the creation of the Ministry of Economy and Finance have probably been the most important and enduring changes in the internal structure of the cabinet. Before 2001, three different portfolios (budget, finance, and treasury) had jurisdiction over the various aspects of government economic policy. The increasing Europeanization of economic policies after the euro’s birth as well as the necessity to control Italy’s huge public debt by limiting the financial requests coming from the other ministers led to a centralization of the main competences in economic policy under one minister in a unified structure. Italian ministers of economy are now typically independent technocrats who enjoy a good reputation from international economic organizations and who rely upon a direct relationship of trust with the PM.

The party system and the actors After the collapse of the First Republic (1948–1993), the Italian party system experienced a fundamental change, one that after more than 25 years is still ongoing. The First Republic was characterized by a ‘pivotal party system’ (Keman 1994; Strøm 2003) and had a very proportional electoral system. The presence of anti-system political parties—the neo-fascist Italian Social Movement (MSI) and the Italian Communist Party (PCI)—at the extreme sides of the ideological spectrum prevented real government alternation. Pro-system parties— Italian Liberal Party (PLI), Italian Republican Party (PRI), Italian Socialist Democratic Party (PSDI), and from 1963 Italian Socialist Party (PSI)—were sometimes ideologically distant from each other but coalesced to form governments (Sartori 1976). All the governments were centred upon the Christian

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



399

Democracy (DC), the largest party, which until 1992 never won less than 30 per cent of votes. This resulted in heterogeneous and bickering post-electoral multiparty coalitions in support of short-lived and usually ineffective cabinets (Cotta and Verzichelli 2007). The main elements of this era persisted for about 45 years. Things abruptly changed with the fall of the Berlin Wall, the collapse of the Soviet Union, and the disappearance of the communism–anticommunism cleavage as a fundamental dimension of the party competition (Bardi 2007). Symbolically, in 1991 the PCI changed its name to Democratic Party of the Left (PDS). At the same time, the popular support for the traditional parties dramatically declined. The financial resources for distributive and particularistic policies, policies that had sustained the ruling parties’ support, were decreasing. This was because of the high public debt and the requirements imposed by the new European treaties (Cotta 1996). Meanwhile, judicial prosecutions on grounds of misuse of public money removed the residual legitimacy of the traditional government parties. During the XI Legislature (1992–1994), a parliament literally decimated by the judicial inquiries approved a complex mixed electoral system. This was an attempt to mitigate and harmonize the effects of a referendum on the electoral rules held in 1993 that had resulted in a call for a more majoritarian electoral system. The new system assigned 75 per cent of seats by plurality in single-member districts, while the rest of the seats were allocated by a proportional system. The proportionality element was engineered to partially compensate the parties that had been defeated in the district competition.

An unstable bipolar and alternational party system In the general elections of 1994, the first ones under the new electoral system, the Italian political landscape was completely reshaped. Newcomer parties such as Go Italy (FI), and parties previously excluded from the government arena, received a large proportion of the votes that were once collected by the old centre and centreleft parties (Morlino 1996). As Table 12.1a shows, cabinets were now formed by new parties and new constellations. The new electoral rules also favoured the formation of alliances and broad pre-electoral coalitions, a phenomenon almost unknown during the First Republic (Table 12.1b). From 1994 up to 2013, the Italian party system was alternational and bipolar. Elections were contested by two alternative coalitions that secured most of the votes. Centre-left and centre-right coalitions prevailed over each other at any new election. The Italian political system, which had never experienced complete government alternation for 45 years, had complete government alternation six times in less than two decades. All these coalitions were rather fragmented and heterogeneous (Zucchini 2013). Until 2006 no party overcame the threshold of 30 per cent of seats in parliament,

20

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

3 4

2

1

Date in

Election date

60.8 (55.7) 55.1 (47) 44.4 (45.9) 44.4 (45.9) 44.4 (45.9) 50 (48.8) 50 (48.6) 44.4 (45.5) 48.7 (50.2) 45.8 (47.8) 45.8 (48.6) 45.8 (48.6) 49.8 (50.8) 41.3 (41.3) 57 (60.6) 42.1 (42.5)

min

47.5 63.4 (52.8)

76.6

80.8

Cabinet strength in seats (%, Senate in parentheses)

sur sur min min min min min min min min min min min min sur

min sur

sur

sur

Party Type of composition of cabinet cabinet (Senate in parentheses)

1946-07-14 1946-06-02 DC, PCI–PDS– DS, PSI, PRI De Gasperi III 1947-02-02 DC, PCI–PDS– DS, PSI De Gasperi IV 1947-06-01 DC, PLI De Gasperi V 1948-05-24 1948-04-18 DC, PSDI, PLI, PRI De Gasperi VI 1950-01-28 DC, PSDI, PRI De Gasperi VII 1951-07-26 DC, PRI De Gasperi VIII 1953-07-16 1953-06-07 DC Pellac 1953-08-17 DC Fanfani I 1954-01-19 DC Scelba 1954-02-10 DC, PSDI, PLI Segni I 1955-07-06 DC, PSDI, PLI Zoli 1957-05-20 DC Fanfani II 1958-07-02 1958-05-25 DC, PSDI Segni II 1959-02-16 DC Tambroni 1960-03-26 DC Fanfani III 1960-07-27 DC Fanfani IV 1962-02-22 DC, PSDI, PRI Leone Ic 1963-06-22 1963-04-28 DC Moro I 1963-12-05 DC, PSI, PRI, PSDI 1968-06-25 1968-05-19 DC Leone IIc

De Gasperi II

Cabinet Cabinet number

Table 12.1a Italian cabinets, 1946–2018

630 (322)

574 (336) 574 (336) 590 (242) 590 (242) 590 (242) 590 (242) 590 (243) 590 (242) 596 (253) 596 (253) 596 (249) 596 (249) 596 (248) 630 (320) 630 (320)

556 574 (343)

556

556

Number of seats in parliament (Senate in parentheses)

10 (10)

9 (8) 9 (8) 9 (8) 9 (8) 9 (8) 9 (8) 9 (8) 9 (8) 9 (8) 9 (8) 9 (8) 9 (8) 9 (8) 9 (9) 9 (9)

6 9 (8)

6

6

Number of parties in parliament (Senate in parentheses)

3.64 (3.77)

2.87 (3.84) 2.87 (3.88) 3.54 (3.6) 3.54 (3.6) 3.54 (3.6) 3.54 (3.6) 3.54 (3.6) 3.54 (3.59) 3.44 (3.25) 3.44 (3.28) 3.44 (3.19) 3.44 (3.19) 3.44 (3.15) 3.74 (3.75) 3.74 (3.75)

4.30 2.87 (3.99)

4.30

4.30

ENP, parliament (Senate in parentheses)

DC

DC DC DC DC DC DC DC DC DC DC DC DC DC DC DC

DC DC

DC

DC

Formal Median support party in first policy parties dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1969-08-06 1970-03-28

1970-08-06

1972-02-18 1972-06-26 1972-05-07 1973-07-08

1983-08-04 1983-06-26

1987-04-18 1987-07-29 1987-06-14

Rumor IIc Rumor III

Colombo

Andreotti Ic Andreotti II Rumor IV

Rumor V Moro II Moro III Andreotti III Andreotti IVc Cossiga I Cossiga II Forlani

Spadolini

Fanfani V

Craxi

Fanfani VIc Goria

De Mita

22 23

24

25 26 27

28 29 30 31 32 33 34 35

36

37

38

39 40

41

1988-04-13

1982-12-01

1981-06-28

1974-03-15 1974-11-23 1976-02-12 1976-07-30 1976-06-21 1979-03-21 1979-08-05 1979-06-03 1980-04-04 1980-10-18

1968-12-13

Rumor I

21

DC, PSI, PSDI, PRI DC DC, PSI, PSDI, PRI DC, PSI, PSDI, PRI DC DC, PSDI, PLI DC, PSI, PSDI, PRI DC, PSI, PSDI DC, PRI DC DC DC, PSDI, PRI DC, PSDI, PLI DC, PSI, PRI DC, PSI, PRI, PSDI PRI, DC, PSI, PSDI, PLI DC, PSI, PSDI, PLI PSI, DC, PRI, PSDI, PLI DC, NN DC, PSI, PRI, PSDI, PLI DC, PSI, PRI, PSDI, PLI sur (sur)

min sur (sur)

sur

sur

sur

sur min min min min min sur sur

min min sur

sur

min sur

sur

59.8 (59.1)

35.9 (37.3) 59.8 (59)

58.1 (57.8)

55.9 (56.5)

58.4 (59.3)

56.5 (57.1) 44.4 (43.8) 42.1 (42.2) 41.6 (42.2) 46.2 (47.5) 46.2 (46.6) 53.8 (55.6) 57 (58.7)

42.1 (42.1) 50 (49.4) 58.9 (58.7)

57.9 (57.6)

42.1 (42.5) 57.9 (57.6)

57.9 (57.5)

630 (323)

630 (324) 630 (324)

630 (322)

630 (322)

630 (322)

630 (322) 630 (322) 630 (322) 630 (322) 630 (322) 630 (322) 630 (322) 630 (322)

630 (321) 630 (322) 630 (322)

630 (321)

630 (322) 630 (321)

630 (322)

12 (12)

11 (10) 12 (12)

11 (10)

11 (10)

11 (10)

8 (9) 8 (9) 8 (9) 10 (7) 10 (8) 11 (10) 11 (10) 11 (10)

10 (11) 8 (10) 8 (9)

10 (11)

10 (11) 10 (11)

10 (10)

4.35 (4.01)

4.33 (4.11) 4.36 (4.02)

4.33 (4.12)

3.58 (3.49)

3.58 (3.51)

3.57 (3.76) 3.57 (3.74) 3.57 (3.74) 3.16 (3.45) 3.16 (3.45) 3.58 (3.51) 3.58 (3.51) 3.58 (3.51)

3.64 (3.93) 3.57 (3.9) 3.57 (3.76)

3.64 (3.87)

3.64 (3.89) 3.64 (3.85)

3.64 (3.77)

PSDI

DC PSDI

DC

DC

DC

DC DC DC DC DC DC DC DC

DC DC DC

DC

DC DC

DC

Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1989-07-23

Andreotti V

Andreotti VI

Amato I

Ciampib

Berlusconi I

Diniab

Prodi I

D'Alema I

D'Alema II

42

43

44

45

46

47

48

49

50

Election date

1996-05-18 1996-04-21 PCI–PDS–DS, min (min) PPI, RI, Verdi 1998-10-21 PCI–PDS–DS, sur (sur) PPI, UDR– UDEUR, RI, PdCI, Verdi, SDI 1999-12-22 PCI–PDS–DS, min (sur) PPI, Dem., UDR–UDEUR, PdCI, RI, Verdi

630 (325) 630 (325)

630 (324)

52.2 (55.1)

48.6 (53.7)

629 (325)

630 (326)

630 (325)

630 (325)

630 (322)

630 (322)

Number of seats in parliament (Senate in parentheses)

44.3 (48)

-

57 (46.9)

sur (min) non (non)

51.9 (52.6)

52.5 (52.6)

56.7 (55.9)

58.9 (59.3)

Cabinet strength in seats (%, Senate in parentheses)

mwc (sur)

mwc (sur)

sur (sur)

sur (sur)

Type of Party composition of cabinet (Senate in cabinet parentheses)

DC, PSI, PRI, PSDI, PLI 1991-04-13 DC, PSI, PSDI, PLI 1992-06-28 1992-04-06 PSI, DC, PLI, PSDI 1993-04-29 DC, PSI, PLI, PSDI 1994-05-11 1994-03-28 FI–PdL, AN, LN, CCD– CDU–UDC 1995-01-17 NN

Date in

Cabinet Cabinet number

Table 12.1a Continued

16 (16)

14 (14)

11 (11)

14 (14)

14 (12)

13 (12)

13 (12)

12 (12)

12 (12)

Number of parties in parliament (Senate in parentheses)

7.09 (6.22)

6.69 (6.43)

6.21 (6.22)

7.18 (8.05)

7.07 (7.24)

5.86 (5.03)

5.71 (5.03)

4.42 (4.22)

4.35 (3.93)

ENP, parliament (Senate in parentheses)

PPI

RI

RI

LN

LN

DC

DC

PSDI

PSDI

Rep., SVP, (VA)

PCI–PDS– DS, LN, PPI, Verdi, PS, AD, SVP, (VA) Ex Left, Rete

PRI

SVP, (VA)

Formal Median support party in first policy parties dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Amato II

Berlusconi II

Berlusconi III

Prodi II

Prodi III

Berlusconi IV Berlusconi V Berlusconi VI Berlusconi VII

Montiab

Letta I

51

52

53

54

55

56 57 58 59

60

61

54 (45.7)

54.1 (44.7)

sur (min)

sur (min)

non (non)

69.4 (69.9)

-

53.2 (53.4) 52.4 (53.1) 46.7 (49.8) 50 (52)

55.1 (54.1)

sur (sur)

mwc (mwc) mwc (mwc) min (min) min (mwc)

56.1 (53.6)

49.5 (54)

sur (mwc)

min (sur)

2013-04-28 2013-02-25 PD, FI–PdL, SC, sur (sur) CCD–CDU– UDC, PR

PCI–PDS–DS, PPI, Dem., PdCI, Verdi, UDR– UDEUR, RI, SDI 2001-06-11 2001-05-13 FI–PdL, AN, LN, CCD– CDU–UDC 2005-04-23 FI–PdL, AN, CCD–CDU– UDC, LN, NPSI, Rep. 2006-05-17 2006-04-10 PD, Ex Left, RnP, IdV, Verdi, PdCI, UDR– UDEUR 2007-05-16 PD, Ex Left, RnP, SD, IdV, Verdi, PdCI, UDR–UDEUR 2008-05-08 2008-04-14 FI–PdL, LN 2010-07-30 FI–PdL, LN, FLI 2010-11-17 FI–PdL, LN 2011-03-23 FI–PdL, LN, PTCN 2011-11-16 NN

2000-04-26

630 (319)

630 (322)

630 (322) 630 (322) 630 (321) 630 (321)

630 (322)

630 (322)

619 (320)

619 (323)

630 (324)

12 (10)

15 (11)

7 (7) 12 (9) 12 (9) 11 (10)

17 (15)

15 (13)

15 (14)

13 (13)

16 (16)

3.59 (4.22)

4.21 (3.6)

3.09 (2.85) 3.75 (3.24) 3.8 (3.26) 3.93 (3.44)

5.93 (6.28)

5.32 (5.35)

5.73 (6.85)

5.49 (6.44)

7.1 (6.42)

PD

FLI

FI–PdL FLI FLI FLI

IdV

IdV

AN

AN

PPI

Continued

FI–PdL, PD, CCD–CDU– UDC, FLI, PT–CN, IdV, CD, MpA CD, SVP, Soc., MAIE

MpA MpA NS Rep.

SVP

SVP

NPSI, Rep.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2013-11-15

Letta II

Renzi I

Renzi II

Gentiloni Silveri 2016-12-12

Conte Ib

62

63

64

65

66

52.5 (43.9)

sur (min) 629 (320)

630 (321)

630 (320)

630 (320)

630 (321)

Number of seats in parliament (Senate in parentheses)

13 (10)

17 (17)

14 (13)

14 (12)

14 (12)

Number of parties in parliament (Senate in parentheses)

4.43 (4.47)

3.77 (5.68)

3.54 (5.01)

3.77 (5.24)

3.77 (5.14)

ENP, parliament (Senate in parentheses)

M5S

PD

PD

PD

PD

MAIE

SC, CD, SVP, Soc.

SC, PpI, Soc., SVP, CD

PpI, CD, SVP, Soc.

CD, SVP, Soc., MAIE

Formal Median support party in first policy parties dimension

Table legends and variables are further defined in the Appendix, this volume.

Formal support parties have been coded for the 1987–2018 period. Only formal support parties in the Chamber of Deputies are reported.

Cabinet types: min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition; non = Non-partisan.

The number of parties in parliament does not include parties that have never held more than two seats when a cabinet has formed. First policy dimension is economic left–right.

Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1945–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015).

a = Technocrat minister majority; b = Technocrat PM; c = Limited policy remit

54.3 (46.9)

sur (min)

mwc (mwc) 55.0 (52.2)

56.5 (46.6)

58.6 (48.9)

Cabinet strength in seats (%, Senate in parentheses)

sur (min)

sur (min)

Party Type of composition of cabinet cabinet (Senate in parentheses)

PD, NCD, SC, PpI, CCD– CDU–UDC, PR PD, NCD, SC, CCD–CDU– UDC PD, NCD, CCD–CDU– UDC PD, NCD, CCD–CDU– UDC 2018-03-04 M5S, LN

Election date

Notes: For a list of parties, consult the chapter appendix.

2018-06-01

2015-02-18

2014-02-22

Date in

Cabinet Cabinet number

Table 12.1a Continued OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



405

Table 12.1b Electoral alliances and pre-electoral coalitions in Italy, 1987–2018 Election date Constituent parties

Type

Types of pre-electoral commitment

1994-03-28

Ex Left, Rete, Verdi, PCI–PDS–DS, PSI, AD PPI, PS

EA

Written contract

EA

FI–PdL, LN, CCD–CDU–UDC, PR FI–PdL, AN, CCD–CDU–UDC PCI–PDS–DS, Verdi, PPI, RI PCI–PDS–DS, Verdi, PPI, RI, Ex Left

EA EA EA, PEC EA

FI–PdL, CCD–CDU–UDC, AN PCI–PDS–DS, PPI, Dem., UDR– UDEUR, RI, Verdi, SDI, PdCI, SVP, Movement of European Republicans (MRE), Sardinian Action Party (PSdAZ) FI–PdL, CCD–CDU–UDC, LN, AN, NPSI, Rep. PCI–PDS–DS, DeL, IdV, PdCI, Verdi, UDR–UDEUR, Ex Left, RnP, SVP FI–PdL, CCD–CDU–UDC, LN, AN, NPSI, DCA, Rep. PD, IdV

EA, PEC EA, PEC

Separate declarations Written contract Written contract Other* Separate declarations Other* Other*

EA, PEC

Other*

EA, PEC

Other*

EA, PEC

Other*

EA, PEC

FI–PdL, LN, MpA Ex left, PdCI, Verdi, SD PD, SEL–SI, CD, SVP FI–PdL, LN, FdI, La Destra (the Right), GS, MpA SC, CCD–CDU–UDC, FLI, MAIE Ex Left, PdCI, Verdi, IdV PD, PR, Italy Europe Together (IEI), Lorenzin’s Civic Popular list (CPL), SVP LN, FI–PdL, FdI, NcI, CCD–CDU– UDC

EA, PEC EA, PEC EA, PEC EA, PEC

Separate declarations Other* Other* Written contract Other*

1996-04-21

2001-05-13

2006-04-10

2008-04-14

2013-02-25

2018-03-04

EA, PEC EA, PEC EA, PEC

Other* Other* Separate declarations

EA, PEC

Other*

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other *Joint electoral manifesto.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

406

  

and until 2008 no coalition was formed by less than four parliamentary parties (Table 12.1a). This also meant that government agendas and government’s initiatives looked like a patchwork of compromises and exchanges. Party fragmentation was temporarily reduced in the 2008 election, when just seven parties were able to enter the parliament and a centre-right government led by Silvio Berlusconi involved only two coalition members. However, during the term the number of parliamentary party groups again began to rise in both chambers. Unlike in the First Republic, in the Second Republic the party system has been quite unstable. Since the mid-1990s, most Italian parties merged with other parties, suffered from party split, or even disappeared. Even the Northern League (LN), which in 2013 was the only party running with the same label held in the 1994 elections, renamed itself ‘the League’ in 2018, to mark the party’s transformation into a nationalist and radical right force under Matteo Salvini’s leadership. Moreover, the dimensions of the political competition have changed. Although left–right, interpreted mainly in economic terms, has been by and large the dominant dimension of competition, new issues (formerly decentralization, more recently pro/anti-European Union and immigration) have been strategically emphasized by some key actors (Di Virgilio et al. 2015; Giannetti et al. 2017, 2018). The fragility of the party system is intertwined with the instability of the electoral rules. For instance, both the electoral rules approved in 1993 and in 2005 produced unintended consequences that required remedies.³ In addition, short-run electoral interests and institutional preferences drove the centre-right parties in 2005 and the Constitutional Court in 2014 and 2017 to promote changes that would have led to the adoption of a predominantly proportional system in 2017.⁴ The unstable party system collapsed with the 2013 elections. The Great Recession played a fundamental role in explaining the dramatic change in the voters’ preferences. From 2008 on, a sequence of stagnation and recession years

³ The political players learned how to manipulate the farraginous mixed electoral system approved in 1993. In 2001, this brought to the formation of a Chamber of Deputies ‘unconstitutionally’ composed of only 619 members instead of 630. A second example concerns the electoral law enacted in a hurry by the centre-right coalition in late 2005, just four months before the 2006 elections. To meet the constitutional requirement (Art. 54) that the Senate must be elected on a regional basis, the new electoral law replicated at the regional level the prize in seats that, in the election of the Chamber, was provided for the winning coalition at the national level. This created the premises for a very incongruent bicameralism, a problematic feature as Italian governments must receive an investiture vote in both chambers. The Prodi II cabinet, formed after the 2006 election, was the first (but not last) victim of the discrepancy between the two parliamentary branches in terms of the coalition supporting the government. ⁴ In 2014 the Court declared as unconstitutional the majority bonus of the 2005 electoral law. In 2017 it declared as unconstitutional the majority run-off system introduced with the electoral reform sponsored by Matteo Renzi’s government and that had never been actually applied. By cancelling these ‘majoritarian’ features, the Court’s sentences induced the parties to agree on the proportional electoral rules adopted in 2017.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



407

hit a country whose economic growth was already weak and whose high public debt left no room for expansive manoeuvres of fiscal policy. Starting from 2010, Berlusconi’s reputation and his government’s parliamentary support also eroded. The divergences between Berlusconi’s party and its main coalition partner (LN) over the government’s policy priorities were exacerbated by lack of economic growth and by the restrictive fiscal policy carried out by the minister of economy. The autumn of 2011 was marked by European authorities’ reiterated requests for ‘structural’ reforms, as well as by heavy losses in the Italian stock market and an unrestrainable widening of the spread between Italian and German bonds. In December, during the parliamentary budgetary session, the already narrow legislative majority supporting the government completely disappeared. The president of the Republic nominated the economist Mario Monti, former European commissioner, to head a technical caretaker government—the second one in the Italian history—charged to make the unpopular reforms suggested by European authorities. In spite of the absence of partisan ministers, the composition of the parliamentary majority—Democratic Party (PD), People of Freedom (PdL), and Union of the Centre (UDC)—gave a clear political connotation to the Monti’s cabinet. The extreme wings of the previous government coalitions—LN, the radical left, and later Italy of Values (IdV)—now opposed (inside and outside the parliament) the government and its policies, including the crucial umpteenth restructuring of the pension system. The issues of European integration and the Italian membership in the Euro zone were also becoming much more salient in the party competition than in the past. In the 2013 elections, the above-mentioned opposition parties did not take much advantage of the unpopularity of Monti’s policies and of the very critical conditions of the labour market. A new and innovative political entrepreneur did it. Beppe Grillo—blogger, comedian, and in 2009 co-founder with the web guru Gianroberto Casaleggio of the Five Star Movement (M5S)—channelled the anger and despair of many voters by harshly criticizing Monti’s government and all the mainstream parties. The M5S does not have a clear ideological profile. The call for a direct and digital democracy makes it populist and anti-system. The elected representatives are considered to be at most spokesmen of the decisions made on the web by Movement supporters on issues that a centralized leadership sets in agenda. Originally, the priorities of the M5S resembled those of the European environmental and left libertarian movements in the 1980s. Gradually, valence issues such as moral integrity and fight against corruption have prevailed, together with the polemic against the political establishment. Ambiguous positions on the immigration policies and a clear anti-EU attitude made the Movement’s electoral platform attractive also for rightist voters (Bordignon and Ceccarini 2013; Pedrazzani and Pinto 2015; Tronconi 2018).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

408

  

The birth of a third pole and the (provisional?) end of an alternational party system In the 2013 general elections, the centre-left headed by the PD, the centre-right led by the PdL, and the M5S obtained between 30 per cent and 25 per cent of votes each. Bipolarism had now vanished (Chiaramonte and Emanuele 2014). Although the electoral rules granted a majority of seats to the centre-left coalition in the Chamber, no coalition (or party) attained a majority of seats in the Senate. A government (Letta I) was formed not only irrespective of, but even against, the alliances built during the electoral campaign. The new executive included the PD, Berlusconi’s PdL, and a centrist alliance led by Monti. A large coalition of ‘pro-system’ parties was made. The parties’ positions regarding Italian membership in the European Union and the Euro club turned out to be a crucial criterion for a role in the government. However, as a consequence of the judicial vicissitudes of Berlusconi, his party soon quit the government majority. During the 2018 electoral campaign, Berlusconi’s FI and the League created once again an electoral alliance despite their different positions on EU-related issues. A more homogeneous electoral cartel was built around the centre-left proEuropean PD, while the M5S ran alone. Once again, none of these three poles was able to win a majority of the popular vote, and a coalition cabinet was formed by the M5S and the League after the latter decided to break up (only at the national level) the centre-right electoral alliance. The current situation is a party system in which no large party is close to another large party on both the two main dimensions of party competition— economic left–right and European integration. Unlike during the First Republic, in the current party system no political axis is considered to be the most important by all political forces. In this multidimensional political space without a ‘core’ party, uncertainty and instability reign supreme.

The appearance of electoral alliances Alongside parties, electoral alliances have been the major political actors during the Italian Second Republic.⁵ From one election to the next, alliances were reshaped in response to the internal evolution of the parties and the changes in the electoral rules, as illustrated in Table 12.1b. In 1994, a left-wing alliance was ⁵ During the First Republic, the creation of electoral alliances was episodic. In 1948, PCI and PSI formed the Popular Democratic Front, while PLI and the rightist populist Front of the Common Man constituted the National Bloc. In 1953, DC, PSDI, PRI, and PLI stipulated an electoral agreement to obtain the majority prize set up by the electoral law that only in that year was in force. In 1958, the Radicals and PRI ran in a joint list. In 1968 and 1972, PCI and PSIUP formed an electoral cartel for the Senate election.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



409

built by the heirs of the PCI—the PDS and the smaller Communist Refoundation (RC), with orthodox communist views—together with other minor left and centre-left parties as well as the remains of the PSI. A centre alliance was composed of the moderate Italian Popular Party (PPI), that is the main DC’s successor, and the Segni Pact, which gathered many of the promoters of the electoral referendums of 1991 and 1993. In a bid to build an anti-left coalition and fill the political vacuum generated by the collapse of DC and PSI, Berlusconi devised two electoral cartels. In northern and most central Italy, the Freedom Pole included FI, the Christian Democratic Centre (CCD)—a conservative fragment of the DC—, the Radicals, and the regional/separatist LN. In the southern part of the country, the Good Government Pole comprised FI, the CCD, the Radicals, and the post-fascist National Alliance (AN). All these four alliances were nothing more than purely electoral aggregations, as they did not entail any commitment to form a joint government after the election (Di Virgilio 2006).

The consolidation of pre-electoral coalitions After 1994, the poor performance of the centrist pole pushed its members to either the centre-left camp (PPI) or the centre-right camp (the United Christian Democrats or CDU, which was a splinter of the PPI). A centre alliance would not have been formed any more until 2013. Two truly pre-electoral coalitions formed with an explicit PM candidate. Despite the changes occurred over time, each of the two main alliances has always been characterized by a core set of parties. The centre-left coalition has developed around a stable agreement between the main party of the left (the PDS, then relabelled Left Democrats or DS) and the centre-left heirs of the DC (the PPI, then the Daisy). This bloc has coalesced (or concluded pacts of mutual withdrawal of candidates) with the extreme left and other minor lists. The centre-right coalition has hinged on the FI–AN dyad and has involved the LN (except for the 1996 election) and the ex-DC segments known under the label of CCD, CDU, and later UDC (up until 2006). The median positions in the Italian parliament changed depending on the electorally winning coalition, but very often they did not correspond to the position of the main coalition party (Table 12.1a). During the XIII Legislature (1996–2001), Italian Renewal, a party founded by Dini, and later the PPI were median on the economic left–right dimension. In the following legislature (2001–2006), National Alliance was median. In the XV Legislature (2006–2008), a small corruptionbashing party led by the former prosecutor Antonio Di Pietro, the IdV, was median in the Chamber, while the UDC was median in the Senate on the left–right.⁶ ⁶ Median parties may differ between the Chamber and the Senate because of the possible incongruent seat distributions created by the electoral system.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

410

  

In 2006, under the electoral system approved the year before, no party ran outside the two coalitions assembled by Romano Prodi and Berlusconi. Alongside the parties included in the previous election, the centre-right camp incorporated also a small new Catholic party called Christian Democracy for the Autonomies (DCA). The centre-left bloc was even more heterogeneous: its members were the leftist DS, the centre-left Daisy, the Greens, two communist parties (RC and PdCI), the IdV, centrist UDEUR (Union of Democrats for Europe, which had left the Daisy during the XIV Legislature), SVP (the German-speaking South Tyrol Popular Party) and the Rose in the Fist (RnP), a new party originating from the merger between the socialist SDI and the Radicals. In 2007 the DS and the Daisy merged into a single party (PD), prompting the most leftist wing of the former to create a new party labelled Democratic Left (SD). The following year, the fusion between FI and AN gave birth to the conservative PdL. Pre-electoral coalitions were unusually small-sized in 2008. Taking a defeat for granted, the PD refused any agreement with left forces and decided to involve only IdV in the centre-left coalition. In turn, Berlusconi built a coalition with just the LN and the Movement for the Autonomies (MpA), a moderate party whose electoral base was in southern Italy. The UDC took an autonomous position as a centrist party in between the two opposing blocs, while a joint list encompassing the Greens and the most leftist parties (the Left, the Rainbow) did not overcome the threshold to win parliamentary seats. After the elections the number of parliamentary parties considerably increased. Two new parties emerged as splinters of the PdL: the centrist Future and Freedom (FLI), which left Berlusconi’s government in late 2010, and the very right-wing Brothers of Italy (FdI), which formed towards the end of the legislative term. In the Chamber, FLI occupied the median position on the economic left–right up until the end of the legislature (Table 12.1a), while the IdV was median in the emerging European integration dimension. After FLI’s exit from the governing coalition, in either chamber a very composite parliamentary group was assembled to support Berlusconi’s cabinet. The group—named People and Territory in the Chamber and National Cohesion in the Senate—gathered MPs from UDC, IdV, and even PD. Meantime, a small centrist party called Democratic Centre (CD) was created by a former UDC leader, and a bunch of other tiny parties formed.

Pre-electoral coalitions at the end of bipolar competition In the 2013 elections the centre-right coalition, led for the sixth time by Berlusconi, comprised PdL, LN, FdI, and MpA. The centre-left camp included PD, CD, SVP, and Left Ecology Freedom (SEL), a small leftist libertarian party. Three new political actors ran for the elections: the M5S, a centrist coalition (with Monti’s party SC, the UDC, and FLI), and an extreme left joint list labelled Civil

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



411

Revolution that eventually failed to obtain legislative seats. Such a configuration proved not to be stable in parliament. None of these pre-electoral coalitions actually became a post-electoral government. Enrico Letta’s cabinet was initially supported by PD, PdL, SC, and UDC.⁷ In November 2013 Berlusconi withdrew the PdL’s support for Letta’s cabinet and relabelled his party Go Italy, while a PdL’s segment remained loyal to the executive after naming itself New Democratic Centre (NCD). Under Renzi’s leadership, from 2014 to 2017, the PD’s shift towards more centrist positions on the economic left–right caused some minor fissions in the party. Nevertheless, thanks to the electoral bonus obtained in 2013, the PD controlled the median legislator along all dimensions in the Chamber. During the legislative term, the size of the parliamentary groups of M5S and SC gradually shrank, and a plethora of small parties was created in both houses. Again, five years later, the configuration of electoral alliances did not survive after the elections. Two pre-electoral coalitions contested the general elections of March 2018. Besides FI, the League, and the extreme right FdI, the centre-right bloc included a centrist joint list between UDC and a new party called Us with Italy (NcI). The centre-left coalition headed by the PD involved several minor lists, among which was the radical Euro-enthusiastic +Europe. The election was also contested by a number of single-party lists like the M5S and a left-wing PD splinter named Free and Equal (LeU). In the end, a breakup of the centre-right electoral coalition allowed the formation of the Conte I cabinet in June 2018. This was a coalition between the M5S, which in both chambers of the new parliament was median in the most important dimensions (economic left–right, GAL–TAN, European integration), and the League, which for the first time surpassed Berlusconi’s party in terms of votes.

Government formation In the shift from the First to the Second Republic, the formal procedures and the custom regulating the process of government formation remained the same. The process is initiated by the president of the Republic, who nominates a formateur after holding consultations with the leaders of the parliamentary party groups and the former presidents. The formateur then starts a bargaining round involving those parties that seem available to enter a new government or to provide external support. If an agreement is reached, the formateur reports to the head of state, who gives her or him an official mandate to form a cabinet. The appointed PM then takes the oath of office and presents the government’s programme to the parliament, where the government has to pass an investiture vote. If during the ⁷ Although being assigned the Ministry of Foreign Affairs, the Radicals (PR) had no seats in the parliament.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

412

  

bargaining round one of the involved parties drops out or the formateur relinquishes her or his mandate, a new bargaining round starts. What was new in the Second Republic, especially from 1996 onwards, was the presence of pre-electoral coalitions, each with a candidate for the prime ministership and a joint electoral manifesto. This implied that the candidate of the victorious coalition was naturally designated as the formateur and that the bargaining over the formation of a new government typically involved just the members of the pre-electoral coalition. This apparently simple picture was complicated by some important political and institutional factors. As already pointed out, one of these was the relatively high number of coalition members characterizing the multiparty executives that formed after each election. Another critical element is related to the bicameral nature of the Italian parliament, as the government has to retain the confidence in both the Chamber and the Senate. Given the increasing incongruence between the two houses, above all after the 2005 electoral reform (Zucchini 2008, 2013; Pedrazzani and Zucchini 2020), several cabinets lacked a majority in one of the two parliamentary branches. In addition, none of the cabinets formed after the election managed to last for the entire length of the constitutional term, and several inter-electoral executives were sworn in thanks to the support of a parliamentary coalition that was at least in part different from the winning pre-electoral coalition. During the inter-electoral periods of crisis and government replacement, the president of the Republic was often able (and sometimes forced) to play a major role in the government formation process (Tebaldi 2014).

The bargaining process While in the First Republic failed bargaining attempts were frequent before the formation of new cabinets (Verzichelli and Cotta 2000), in the period from 1994 to the 2018 elections there were only four (very short) inconclusive attempts. As reported in Table 12.2, the first of these was in 1996, when after Dini’s resignation the head of the state appointed Antonio Maccanico to form again a caretaker government. However, most of the parties in parliament preferred early elections to a new cabinet. The second instance of a failed formation attempt occurred in 1998, when the RC withdrew its external support from the first Prodi’s cabinet. After the PM’s resignation in October 1998, the president of the Republic nominated again Prodi as a formateur, whose attempt failed as the new centrist UDR refused to support a centre-left government with Prodi as PM. After less than a week, the UDR backed instead a new cabinet led by the Left Democrat Massimo D’Alema. A third failed bargaining attempt occurred 10 years later, following the fall of Prodi III cabinet (2007–2008). In January 2008, the head of the state commissioned Franco Marini,

1992 0

1993 0

1994 0

1995 0 1996 1

Amato I

Ciampi

Berlusconi I

Dini Prodi I

5 2

14 (1) PCI–PDS–DS, Verdi, Rete, CCD–CDU–UDC, Ex Left, PPI, LN, FI–PdL, AN, SVP, Sinistra Democratica, Laburisti, Lega Italiana Federalista, I Democratici, Federalisti e Liberaldemocratici, Comunisti Unitari

PPI, PCI–PDS–DS, RI, Verdi

14

3

11

26 8

(1) DC, PSI, PSDI, PLI, PRI

1989 1

Andreotti V

Andreotti VI 1991 0

1988 0

De Mita 11

1987 0

Goria

Bargaining duration of individual formation attempt (in days)

DC, PSI, PSDI, PLI, PRI

Year Number of Parties involved in the previous in inconclusive bargaining rounds bargaining rounds

Cabinet

Table 12.2 Government formation period in Italy, 1987–2018

26 132

118

7

84

15

65

33

91

5 16

14

3

11

8

37

302 (191) 270 (2) 322 (173) 0 (1)

366 (159) 0 (2)

309 (162) 60 (50)

330 (173) 2 (0)

339 (177) 0 (0)

371 (187) 3 (1)

366 (177) 2 (1)

Continued

207 (104) 280 (140) 185 (36) 245 (153) 39 (17) 299 (139)

237 (123) 215 (143) 200 (117)

Abstention Contra

371 (183) 0 (0)

Pro

Result of investiture vote Number of Total bargaining (Senate result in parentheses) days required in duration government formation

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2 2

Berlusconi II 2001 0

2005 0

2006 0

Berlusconi III Prodi II

2007 0 2008 1

3

2000 0

Amato II

Prodi III Berlusconi IV

6

1999 0

(1) PD, SD, Ex Left, IdV, Verdi, PdCI, RnP, UDR–UDEUR, FI– PdL, AN, CCD–CDU–UDC, NPSI, Rep., MpA, DCA, SVP

FI–PdL, LN 6

1 2

3

3

6

D'Alema II

PCI–PDS–DS, PPI, Verdi, RI, SDI, UDR–UDEUR, PdCI (1) PPI, PCI–PDS–DS, Verdi, RI, SDI, UDR–UDEUR, PdCI

1998 1

D'Alema I

Bargaining duration of individual formation attempt (in days)

Year Number of Parties involved in the previous in inconclusive bargaining rounds bargaining rounds

Cabinet

Table 12.2 Continued

1 105

37

3

29

7

4

12

1 8

2

2

3

6

3

9

335 (173) 1 (2)

344 (165) 0 (0)

334 (170) 2 (0)

351 (175) 1 (5)

319 (179) 5 (2)

310 (177) 18 (4)

275 (137)

287 (100) 298 (112) 261 (133) 240 (117) 268 (155)

281 (116)

Abstention Contra

333 (188) 3 (1)

Pro

Result of investiture vote Number of Total bargaining (Senate result in parentheses) days required in duration government formation

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1 8

7 1 6 1 2 2 3 4 19

M5S, LN (1) M5S, LN, FI–PdL, FdI (2) M5S, PD (3) M5S, LN

2013 0 2014 0

2015 0 2016 0

2018 3

Letta II Renzi I

Renzi II Gentiloni Silveri Conte I

Note: Total bargaining duration coded for the 1989–2018 period.

4 128

4 5

PD, FI–PdL, SC, CCD–CDU– UDC, PR (1) PD, M5S

2011 0 2013 1

89

1 5

1

1

2011 0

1 1

1 1

2010 0 2010 0

Berlusconi V Berlusconi VI Berlusconi VII Monti Letta I

28

1 2

1 6

4 12

1

1 1

350 (171) 35 (25)

368 (169) 0 (0)

378 (169) 1 (0)

556 (281) 0 (0) 453 (233) 17 (18)

105 (99) 236 (117)

220 (139)

61 (25) 153 (59) OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

416

  

at that time the Speaker of the Senate, to assemble a temporary coalition in order to approve a new electoral law. After meeting all the leaders of the parliamentary party groups (except for the LN, which refused to negotiate), Marini gave up due to the unwillingness of centrist and right-wing parties. Similar to what had happened after Maccanico’s failure, Marini’s unsuccessful attempt led to early elections. The fourth case of inconclusive bargaining was before the Letta I cabinet took office in April 2013. After the election of February 2013, President Giorgio Napolitano designated the PD leader Pier Luigi Bersani to form a new cabinet. As Bersani’s attempt to involve the M5S proved unfruitful, the whole process was postponed because the term of the president was expiring. Once re-elected, Napolitano nominated Letta, who at the time was deputy secretary of the PD. The last cabinet covered in this chapter was formed in June 2018, after three unsuccessful attempts. During three months characterized by many twists and turns by the main Italian parties, President Sergio Mattarella designated the presidents of both the Chamber and Senate as informateur, and appointed Giuseppe Conte as PM on two occasions, the latter of which led to the birth of his government.⁸ On average, the length of successful formation attempts was shorter than that of failed attempts. This holds both for post-electoral cabinets, whose bargaining rounds lasted slightly more than four days on average, and for inter-electoral cabinets, with negotiations of less than three days. Let us note, however, that a two-week bargaining round was needed to form Berlusconi I government (1994–1995), a result related to the merger of two partially overlapping electoral cartels. As far as the Second Republic is concerned, the overall length of the process of government formation does not seem substantially affected by the (often short) duration of the bargaining rounds. Given our counting rules, the total number of days required for forming a new government was particularly high for those cabinets that took office following an early election: Berlusconi I (1994–1995), Prodi I (1996–1998), Berlusconi IV (2008–2010), and Letta I (2013). The birth of these governments required about three or four months, as the resignation of the previous PM and the unfeasibility of an inter-electoral cabinet prompted the head of state to call early elections. Therefore, the duration of the crisis preceding the formation of these governments also includes the weeks of the electoral campaign and the days a new parliament needs to be convened and elect its internal bodies. Apart from the Conte I executive, the time needed for forming a new government was indeed much shorter for post-electoral cabinets that came to power after the natural end of the previous legislature: Berlusconi II (2001–2005) and Prodi II (2006–2007). By the same token, no more than two weeks were required to form ⁸ Between the two Conte’s attempts, Mattarella also appointed (inconclusively) the economist Carlo Cottarelli to form a technocratic government, which would have asked for support in parliament.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



417

inter-electoral governments, with the only (partial) exception of Dini’s technocratic cabinet.⁹

The composition and size of cabinets Whereas all the governments formed during the First Republic were centred upon the pivotal role of the DC (albeit according to various coalitional formulas, as shown by Verzichelli and Cotta 2000), the shift to the Second Republic brought about a competition for government between a centre-right and a centre-left bloc that actually alternated to power for about two decades. In particular, the following phases can be identified (see Table 12.1a): -

Centre-right coalition (1994–1995) Technocratic cabinet (1995–1996) Centre-left coalition (1996–2001) Centre-right coalition (2001–2006) Centre-left coalition (2006–2008) Centre-right coalition (2008–2011) Technocratic cabinet (2011–2013) Grand coalition (2013) Centre-left coalition (2013–2018) Yellow-green coalition (2018–2019)

In addition to government alternation, another phenomenon of the Second Republic has been the appointment of technocratic cabinets supported by large majorities in parliament, specifically designated to cope with critical economic circumstances.¹⁰ All the governments in the 1994–2018 period were supported by legislative coalitions that were quite fragmented and quarrelsome. However, some differences between centre-right and centre-left coalitions are worth noting. In all the centre-right cabinets, the leader of the largest party (Berlusconi) was appointed as the PM. The same did not hold for the centre-left cabinets: in fact, Prodi was not affiliated with any of the coalition parties when he headed his first cabinet in 1996, and Giuliano Amato was an independent politician when he became PM in 2000. ⁹ According to the coding rules of this book, when a change in the set of parties holding ministerships occurs a new government is coded even in the absence of a formal investiture vote. This is why our numbering of Berlusconi’s, Prodi’s, Letta’s, and Renzi’s cabinets deviates from Italian conventions. In these cases, we consider both the duration of the bargaining attempt and the time required in government formation to be equal to one day. ¹⁰ Although a cabinet led by a technocratic PM had been previously appointed (Ciampi government in 1993), the executives headed by Dini and Monti were entirely formed by non-partisan ministers (McDonnell and Valbruzzi 2014).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

418

  

Moreover, centre-right governments were less fragmented in terms of the number of coalition members than centre-left governments: excluding external supporting parties, we count two to six coalition parties for the former, and three to eight coalition parties for the latter. In addition, centre-right cabinets enjoyed a more stable support in parliament. Four of the seven Berlusconi’s governments controlled a majority in both houses of the Italian parliament. Only one of the ten centre-left governments did the same: D’Alema I (1998–1999). This brings attention to an aspect that has become increasingly paramount in the process of government formation during the Second Republic, at least up to the 2018 elections: the growing incongruence between the partisan composition of the Chamber and the Senate. While almost perfectly congruent throughout the First Republic, from 1994 onwards the two legislative branches have diverged more and more in terms of the distribution of party seats. This was mainly due to the impact of the electoral reforms of 1993 and, above all, of 2005. Some technicalities in the mixed member majoritarian system introduced in 1993 did not guarantee bicameral congruence (Bartolini and D’Alimonte 1994). The PR adopted in 2005 assigned a seat bonus on a national basis in the Chamber, and on a regional basis in the Senate (Chiaramonte and Di Virgilio 2006). Out of the nineteen partisan cabinets of the Second Republic (from Berlusconi I to Conte I), only eight had the same status in both legislative houses: three oversized coalitions, three minimal winning coalitions, and two minority coalitions (Table 12.1a). Moreover, just one-third of the partisan cabinets of the Second Republic controlled a majority of seats in both houses, while more than half were supported by a legislative majority in one house but not in the other. In the latter group, seven cabinets enjoyed a majority in the Chamber but not in the Senate (Berlusconi I, Prodi II and III, Letta II, Renzi I and II, Gentiloni Silveri); vice versa, three cabinets held a majority in the Senate but not in the Chamber (D’Alema II, Amato II, Berlusconi VII). All the centre-left governments that formed after the approval of the 2005 electoral rules were unable to assemble a majority in the Senate despite relying on a surplus coalition in the Chamber. The Conte I cabinet—also dubbed ‘yellow-green’ coalition—is the last government covered in this chapter and marks an important difference with all the partisan cabinets of the Second Republic. It is no coincidence that the party leaders of the two coalition partners defined the new cabinet as the ‘government of change’ (Pedrazzani 2018; Conti et al. 2020). Indeed, the two government parties share strong anti-establishment and anti-EU stances and emphasize the divide between ‘the elite’ and ‘the people’. Moreover, the M5S promotes the passage to a direct (and digital) democracy while the League advocates nationalistic and xenophobic stances. These positions and issues were previously at the margins of political debate and electoral competition. At the same time, in 2018 the two political forces that had alternated in government almost uninterruptedly since 1994—centre-right FI/PdL and centre-left PD and its forerunners—were

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



419

both at the opposition. Finally, the (predominantly proportional) electoral rules in force during the 2018 elections meant increased congruence in the partisan composition of the two parliamentary houses, which allowed the Conte I cabinet to be supported by a minimal winning coalition in both the Chamber and the Senate (Chiaramonte and D’Alimonte 2018). Generally speaking, cabinets of the Second Republic differ from those of the First Republic not only because of their partisan composition but also with regard to their size and ministerial structure. Italian cabinets reached their maximum size during the 1987–1992 legislature, which was characterized by the coalition formula known as pentapartito (five-party coalition). Since then, both the number of ministries and the overall number of cabinet members have gradually decreased, especially after 2001, when a single Ministry of Economy and Finance was created and a number of other ministries were merged (transports with public works, industry with external trade, education with university and research). However, there was again a noticeable government enlargement under the fragmented Prodi II and III cabinets, when some ministries were split into two (education/university and research, labour/social solidarity) as a way to please the numerous partners in the coalition. During the XVI Legislature, the number of ministries finally became 14—that is one-third less than the ministries of the late First Republic. Likewise, since the start of the 1990s the overall number of cabinet members—which includes the vice-PMs and ministers without portfolio—has decreased from 30 to 19.

The allocation of ministerial portfolios The coalitional nature of Italian governments constrains the PM’s formal right to propose ministers to the president of the Republic. Usually the leadership of each coalition party proposes the nominees for the ministerial portfolios the party has obtained. In this regard, the prerogatives of the PM have increased after 1994 because, differently from the First Republic, during the Second Republic the PM has often been the leader of the largest party in the coalition (Verzichelli 2009). The technocrats Dini and Monti, leading governments without party delegations, probably enjoyed the greatest discretion in the selection of ministers even though, in those cases, the choices were strongly influenced also by the president of the Republic. Shifting from the First to the Second Republic, the portfolio allocation patterns have changed with respect to the proportionality norm and the partisanship of ministers. Table 12.3 introduces the portfolio distribution over five crossnationally important ministries. Most of the time, the distribution of ministerial positions has followed by and large a proportionality norm (Gamson’s law). Throughout the First Republic

1987 15 DC, 8 PSI, 3 PRI, 2 PSDI, 1 PLI, 1 Ind. 1988 16 DC, 9 PSI, 3 PRI, 2 PSDI, 1 PLI, 1 Ind. 1989 15 DC, 9 PSI, 3 PRI, 3 PSDI, 2 PLI 1991 16 DC, 10 PSI, 2 PSDI, 2 PLI 1992 7 PSI, 13 DC, 2 PLI, 2 PSDI, 1 Ind. 1993 8 DC, 4 PSI, 1 PLI, 1 PSDI, 10 Ind.

Goria

Berlusconi 1994 11 FI–PdL, 5 AN, I 5 LN, 2 CCD– CDU–UDC, 3 Ind. Prodi I 1996 9 PCI–PDS–DS, 5 PPI, 3 RI, 1 Verdi, 3 Ind. D'Alema I 1998 8 PCI–PDS–DS, 6 PPI, 3 UDR– UDEUR, 2 RI, 2 PdCI, 2 Verdi, 1 SDI, 3 Ind. D'Alema II 1999 9 PCI–PDS–DS, 5 PPI, 4 Dem., 2 UDR–UDEUR, 2 PdCI, 2 Verdi, 1 RI 1 Ind.

Ciampi

Andreotti V Andreotti VI Amato

De Mita

Year Number of in ministers per party (in descending order)

Cabinet

26

27

21

26

24

25

30

32

32

30

Total number of ministers 15 PSI, 12 DC, 4 PLI, 3 PRI, 3 PSDI 15 PSI, 12 DC, 4 PLI, 4 PRI, 4 PSDI 15 PSI, 12 DC, 4 PLI, 4 PRI, 4 PSDI 16 PSI, 9 DC, 5 PLI, 5 PSDI 5 DC, 3 PLI, 3 PSDI, 10 PSI 10 PSI, 9 DC, 4 PSDI, 3 PLI, 1 PRI, 1 Ind. 12 AN, 10 FI–PdL, 10 LN, 1 CCD– CDU–UDC 13 PCI–PDS–DS, 6 PPI, 3 RI, 3 Verdi, 3 Ind. 13 PCI–PDS–DS, 11 PPI, 8 UDR– UDEUR, 5 RI, 3 Verdi, 3 PdCI, 1 SDI, 3 Ind. 11 PCI–PDS–DS, 11 PPI, 7 Dem., 5 RI, 4 UDR–UDEUR, 3 Verdi, 3 PdCI, 1 VA, 4 Ind.

Number of watchdog junior ministers per party

PCI– PCI–PDS– PDS–DS DS

PCI– PCI–PDS– PDS–DS DS

20

19

PCI–PDS– DS

PPI

Ind.

Ind.

DC

PSI

PSI

DC

DC

21

FI–PdL

Ind.

22 20

PSI

DC

DC

DC

DC

23

23

23

22

22

RI

RI

RI

FI–PdL

DC

DC

PSI

PSI

DC

DC

PCI– PDS–DS

LN

DC

DC

DC

DC

DC

DC

PCI– Dem. PDS–DS

PCI– PPI PDS–DS

CCD– CDU– UDC RI

PSI

DC

DC

DC

PSI

PSI

Number 1 Prime 2 Economy 3 Foreign 4 Labour 5 Interior of minister and affairs ministries finance*

Table 12.3 Distribution of cabinet ministerships in Italian coalitions, 1987–2018 OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Berlusconi 2005 11 FI–PdL, 6 AN, 3 III CCD–CDU–UDC, 3 LN, 1 NPSI, 1 Rep., 1 Ind. Prodi II 2006 19 PD, 1 Ex Left, 1 RnP, 1 IdV, 1 Verdi, 1 PdCI, 1 UDR– UDEUR, 1 Ind. Prodi III 2007 18 PD, 1 Ex Left, 1 RnP, 1 SD, 1 IdV, 1 Verdi, 1 PdCI, 1 UDR–UDEUR, 1 Ind. Berlusconi 2008 18 FI–PdL, 4 LN IV Berlusconi 2010 19 FI–PdL, 3 LN, 1 V FLI Berlusconi 2010 20 FI–PdL, 3 LN VI Berlusconi 2011 19 FI–PdL, 3 LN, 1 VII PT–CN Letta I 2013 10 PD, 5 FI–PdL, 2 SC, 1 CCD–CDU– UDC, 1 PR, 3 Ind. 22

23

23

23

22

26

26

26

25 2000 7 PCI–PDS–DS, 4 PPI, 3 Dem., 2 PdCI, 2 Verdi, 2 UDR– UDEUR, 1 RI, 1 SDI, 3 Ind. Berlusconi 2001 13 FI–PdL, 5 AN, 3 25 II LN, 2 CCD–CDU– UDC, 2 Ind.

Amato II

7 PD, 7 Ex Left, 4 RnP, 2 IdV, 2 PdCI, 2 Verdi, 1 UDR– UDEUR, 6 Ind. 8 PD, 7 Ex Left, 4 RnP, 3 SD, 2 IdV, 2 PdCI, 2 Verdi, 1 UDR–UDEUR, 6 Ind. 3 FI–PdL, 3 LN, 2 MpA, 1 Ind. 4 FLI, 3 LN, 2 MpA, 1 FI–PdL, 1 Ind. 3 LN, 1 FI–PdL, 1 NS, 1 Ind. 3 LN, 1 FI–PdL, 1 PT–CN, 1 Ind. 10 FI–PdL, 9 PD, 3 SC, 1 CCD–CDU– UDC, 1 GS, 7 Ind. 14

14

14

14

13

PD

FI–PdL

FI–PdL

FI–PdL

FI–PdL

PD

19

Ind.

FI–PdL

Ind.

FI–PdL

FI–PdL

FI–PdL

FI–PdL

Ind.

Ind.

FI–PdL

FI–PdL

PD

SDI

Ind.

19

13 PCI–PDS–DS, 11 19 PPI, 6 Dem., 5 UDR–UDEUR, 3 RI, 2 PdCI, 2 SDI, 2 Verdi, 3 Ind. 12 AN, 9 FI–PdL, 6 15 CCD–CDU–UDC, 5 LN, 1 NPSI, 1 Rep., 1 Ind. 12 FI–PdL, 11 AN, 9 15 CCD–CDU–UDC, 7 LN, 2 NPSI, 1 Rep.

PR

FI–PdL

FI–PdL

FI–PdL

FI–PdL

PD

PD

AN

Ind.

RI

Ind.

FI–PdL

FI–PdL

FI–PdL

FI–PdL

PD

PD

LN

LN

Continued

FI–PdL

LN

LN

LN

LN

PD

PD

FI–PdL

FI–PdL

PCI– Dem. PDS–DS

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2013 9 PD, 5 NCD, 1 SC, 1 PpI, 1 CCD– CDU–UDC, 1 PR, 3 Ind. 2014 10 PD, 3 NCD, 1 SC, 1 CCD–CDU–UDC, 2 Ind. 2015 10 PD, 4 NCD, 1 CCD–CDU–UDC, 2 Ind. 2016 13 PD, 3 NCD, 1 CCD–CDU–UDC, 2 Ind. 2018 9 M5S, 6 LN, 4 Ind.

Letta II

19

19

17

17

8 NCD, 7 PD, 2 CD, 2 PpI, 1 SC, 1 Soc., 4 Ind. 8 NCD, 7 PD, 3 CD, 1 SC, 1 PpI, 1 Soc., 4 Ind. 10 NCD, 6 PD, 3 CD, 1 SC, 1 Soc., 4 Ind. 13 LN, 9 M5S, 1 MAIE, 1 Ind. PD

14 Ind.

PD

14

14

PD

PD

14

9 PD, 6 NCD, 2 SC, 14 1 CCD–CDU–UDC, 1 PpI, 7 Ind.

21

Ind.

Ind.

Ind.

Ind.

Ind.

Ind.

NCD

PD

PD

PR

M5S

PD

PD

PD

Ind.

LN

PD

NCD

NCD

NCD

Number 1 Prime 2 Economy 3 Foreign 4 Labour 5 Interior of minister and affairs ministries finance*

Number of watchdog junior ministers per party

Total number of ministers

Note: *The Ministry of Finance and the Ministry of Economy merged in 2001, becoming the Ministry of Economy and Finance. Prior to 2001, the column refers to the Ministry of Finance.

Conte I

Gentiloni Silveri

Renzi II

Renzi I

Year Number of in ministers per party (in descending order)

Cabinet

Table 12.3 Continued OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



423

deviations from Gamson’s law usually occurred when parties entering cabinet offices for the first time received a disproportionally high share of ministries (Marradi 1982). In contrast, during the Second Republic it can be noted that, within the same legislative term, the degree of proportionality was particularly high when cabinets formed immediately after elections and lower for interelectoral cabinets (Cotta and Marangoni 2015). This is because, after the fall of a post-electoral cabinet, the—often small—parties that offered their support to a new cabinet received some office rewards in return. Non-partisan ministers were rather exceptional—albeit not totally absent— until 1993, when a technocrat-led cabinet (Ciampi) was appointed to deal with the negative economic juncture. During the Second Republic, non-partisan ministers increased in number (Table 12.3).¹¹ Two fully technocratic executives took office and in the partisan cabinets a number of non-partisan figures were appointed to some key ministries. In as many as two-thirds of these cabinets, from Berlusconi I to Conte I, the Ministry of Economy was given to a technocrat. It can also be noted that several times non-partisan policy experts have been appointed to the ministries of justice, labour, and health. Such a dynamic is common to a number of European countries (Costa Pinto et al. 2018) and is undoubtedly related with the increasing complexity and ‘Europeanization’ of policy-making in several policy areas, first of all in the economic and financial domain. National-level policy-makers need more and more an enhanced profile of competences to deal with the technical and specific inputs coming from supranational authorities (Verzichelli and Cotta 2018). In contrast, other ‘traditional’, prestigious, and powerful portfolios such as the interior, defence, and foreign affairs have almost always been assigned to partisan ministers—usually, high-ranking and long-tenured government parties’ members. There are almost only partisan appointees also to the ministerial posts without portfolio as the latter are mostly created to increase the spoils to be given to the government parties.

Coalition agreements During the whole First Republic, formal coalition agreements were the exception rather than the rule. After negotiating with the members of the coalition, the designate PM used to combine a bunch of programmatic guidelines in a rather generic and vague text that he then illustrated in parliament before asking an

¹¹ Starting from the 1990s, Italian governments also present a growing share of ministers without any political background, that is without any previous experience in the parliament or in a party (Verzichelli 2014; Verzichelli and Cotta 2018). In this regard, let us recall that the yellow-green government featured as PM a law professor with no previous political experience.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

424

  

Table 12.4 Size and content of coalition agreements in Italy, 1946–2018 Coalition

Year Size in

Distribution Distribution Policies General Policy(in %) of offices of rules specific competences (in %) procedural (in %) (in %) rules (in %)

Moro I Prodi I Berlusconi II Prodi II Berlusconi IV Conte I

1963 1996 2001 2006 2008 2018

38.1 0 0 0 0 5.2

3,680 41,521 15,168 85,140 3,619 17,431

23 0 0 0 0 0

0 0 0 0 0 0

7.5 0 0 0 0 1.3

31.3 100 95.8 100 100 93.5

investiture vote (Verzichelli and Cotta 2000).¹² As illustrated in Table 12.4, coalition agreements appeared during the Second Republic. However, at least until the 2018 coalition contract between the M5S and the League, they had been quite different from the documents that are found in other European democracies (Timmermans 2003, 2006; Müller and Strøm 2008; Moury 2013). The coalition agreements reported in Table 12.4 for the governments Prodi I and II and Berlusconi II and IV were just the joint electoral programme issued by the winning coalition before the election. In fact, none of the many inter-electoral cabinets of the 1994–2019 period was based on a coalition treaty. Moreover, the above-mentioned four coalition agreements were much less formalized than the coalitional pacts subscribed in Belgium, Germany, and the Netherlands. They did not set up mechanisms for handling possible conflicts within the coalition, nor did they provide formal resolution procedures for bringing the government to an end in case of failure to implement its platform. On the whole, they did not seem very effective as tools for constraining ministerial discretion (Conti 2015). They were all almost entirely devoted to policy issues but their length and scope differed from one coalition to the other. As shown in Table 12.4, the joint documents agreed upon by the centre-left coalitions (The Olive tree in 1996 and The Union in 2006) were much longer than the electoral manifestos of the centre-right coalitions (House of freedom in 2001 and the coalition formed by PdL, LN, and MpA in 2008). The former reflected an attempt to stake out a compromise among the various parties participating in the coalition, while the agreements of the centre-right coalitions were more centred upon a set

¹² The only exception to this pattern was the coalition document agreed on in 1963 before cabinet Moro I, although not being properly formalized as a coalition treaty. Another partial exception was a programmatic document signed in 1977 by DC, PCI, PSI, PRI, PSDI, and PLI at the beginning of the ‘national solidarity’ phase. This agreement cannot be properly considered as a coalition pact as the participating parties supported the single-party government Andreotti III formed by the DC alone.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



425

of electoral promises (Cotta and Marangoni 2015). The higher heterogeneity of the centre-left coalitions made less dense the agreements’ content in terms of programmatic pledges (Marangoni 2013). For the first time in Italian republican history, in 2018 the M5S–League cabinet relied upon a formal post-electoral contract signed by the leaders of the two coalition parties. This treaty was a public document formalized after the election. It enlisted governments’ policy priorities and established a ‘conciliation committee’ as the primary instrument for solving conflicts that could arise between coalition partners.

Coalition governance The role of individual ministers in policy-making As mentioned earlier, during the First Republic the cabinet’s programmes, presented by the PM in parliament before the investiture vote, were too vague to serve as an effective constraint on ministers’ behaviour. Moreover, the coalition agreements, which are elsewhere one of the main instruments for restraining ex ante minister’s discretion, have been rare in Italy, even during the Second Republic. Accordingly, Italian ministers have often been described as highly autonomous (Criscitiello 1994). Nevertheless, the cabinet decisionmaking in Italy has never adhered to a ministerial government model—that is, Michael Laver and Kenneth A. Shepsle’s (1996) full ministerial discretion model. A set of institutional mechanisms (see subsections ‘Coalition governance in the executive arena’, ‘Mixed and external governance mechanisms’, and ‘Governance mechanisms in the parliamentary arena’) help to prevent ‘ministerial drift’ and mitigate disagreements among the ministers and among the coalition parties they represent. The limitations to the ministers’ margins of manoeuvre have not implied a strengthening of PM’s role, which remained relatively weak vis-à-vis line ministers throughout the First Republic. In that period, Italian PMs did not enjoy sufficient organizational resources and popular legitimacy and the frequent party summits and coalition committees (see subsection ‘Mixed and external governance mechanisms’) moved the decisionmaking at least partially outside the cabinet. Things started changing in 1988 with the enactment of Law 400, which gave the PM a proper administrative structure and increased the financial resources at her or his disposal.¹³ A further and more important reinforcement of the PM’s role derived from the electoral rules of the Second Republic: at least for

¹³ Reforms during the 1990s formally defined the primacy of the PM in relation to line ministers.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

426

  

post-election cabinets, the systems approved in 1994 and 2005 aligned, de facto, the leader of the winning coalition with the appointed PM. Furthermore the numerous appointments of technocrats to the Ministry of Economy or to other prominent portfolios signal a similar strengthening, as these non-partisan actors are usually selected on the basis of a strong trust relationship with the PM (Cotta and Marangoni 2015). Still, as Maurizio Cotta and Francesco Marangoni (2015: 159) noticed, the recent dynamics favouring the Italian PM have contained the ministerial discretion only to a limited extent. Other factors were and are at work. The range of coalition governance mechanisms employed in Italy is summarized in Table 12.5.

Coalition governance in the executive arena The type of coalition governance that dominates in Italy is the coalition compromise model and it is based upon a set of institutional tools both at the executive level and at the legislative level. Among the former, junior ministers (JMs) are for sure of primary importance. From the first executive of the Italian Republic (De Gasperi II, 1946–1947) to the first cabinet led by Conte (2018–2019), the distribution of government jobs has followed a proportionality norm not only with regard to ministerial posts but also concerning junior ministerial positions. As highlighted by Carol Mershon (2002), Italian JMs have often been appointed in such a way as to compensate for possible imbalances in the distribution of ministerial portfolios among coalition parties. Even under Monti’s full technocratic government, non-partisan JMs were selected following the traditional dynamics of the spoil system, whereby each of the parties supporting the cabinet received a proportional share of ‘trustworthy’ JMs (Giannetti 2013). At the same time, Italian JMs serve as instruments of mutual monitoring among governing parties, as each minister is typically ‘shadowed’ by one or more ‘watchdog’ JMs coming from a different party than the minister. This especially holds for the most salient portfolios, that is where parties that do not control the ministry have the most to lose from abdication of control to their coalition partners (Thies 2001; Giannetti and Laver 2005). A particular type of junior ministerial post is the office of the state undersecretary to the presidency of the Council of Ministers (sottosegretario di stato alla presidenza del consiglio dei ministri). The politicians who hold this position are usually chosen by the PM himself. Especially during the Second Republic, the task of this state undersecretary has been to help the PM in coordinating ministers.¹⁴

¹⁴ A remarkably powerful undersecretary to the presidency was Gianni Letta (uncle of the PD politician Enrico Letta, PM in the 2013–2014 period), who was charged with coordination tasks under all cabinets headed by Berlusconi.

1999 N

2000 N

2001 PRE 2005 N 2006 PRE

D'Alema II

Amato II

Berlusconi II Berlusconi III Prodi II

Y

Y

N N N

N

N

CoC, PS CoC, PS CoC, PS CoC, PS, Pca CoC, PS, Pca CoC, PS, Pca CaC, CoC CaC, CoC PS, Pca

N N PRE N N N

CoC, PS CoC, PS

N N

Y

CoC, PS CoC, PS

N N

CoC PS PS PS

PS PS

PS PS

CoC CoC

CoC

CaC CaC PS, Pca

CoC CoC PS

CoC, Pca PS

CoC, Pca PS

CoC PS CoC CoC, Pca

PS PS

PS PS

CoC CoC

CoC

Y N N

N

N

N N N N

Y N

N N

Y N

N

Y Y Y

Y

Y

Y Y

CoC, PS CoC, PS

N

IE N

N

CoC

N

N

CoC

N

1946 1955 1958 1962 1963 1968 1974 1979 1979 1982 1983 1987 1992 1993 1994 1996 1998

De Gasperi II Segni I Fanfani II Fanfani IV Moro I Rumor I Moro II Andreotti IV Cossiga I Fanfani V Craxi Goria Amato I Ciampi Berlusconi I Prodi I D'Alema I

CoC

Personal Issues union excluded from All used Most For most agenda common serious conflicts

CoC

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Coalition

Table 12.5 Coalition governance mechanisms in Italy, 1946–2018

All/All All/All All/All

All/All

All/All

Most/Spec Most/Spec All/All All/All

Most/Spec Most/Spec

Most/Spec Most/Spec

Most/Spec Most/Spec

Most/Spec

Most/Spec

Coalition discipline in legislation/ other parl. behaviour

Y Y N

N

N

Y Y N N

Y Y

Y Y

Y Y

Y

Y

Few Few Comp.

Few

Few

Few Few Comp. Few

Few Few

Varied Few

Varied Varied

Varied

Few

Y Y N

N

N

N Y N N

Y Y

Y Y

Y Y

Y

Y

Continued

N N N

N

N

N Y N N

Y Y

Y Y

Y Y

Y

Y

Freedom of Policy Junior Nonappointment agreement ministers cabinet (watchdogs) positions

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Y

N

IC, PS

IC, PS

IC, PS

PS CoC CoC CoC CoC Parl Parl PS PS PS Y

N Y N Y Y N N N N N Y

Y Y Y Y Y N N N N N All/None

All/All All/All All/All All/All All/All All/All All/All All/All All/All All/All

Coalition discipline in legislation/ other parl. behaviour

Y

N Y Y Y Y N N N N N

Comp.

Comp. Few Few Few Few Varied Varied Varied Varied Varied

N

N Y Y Y Y Y Y N N N

Y

N N N N N N N N N N

Freedom of Policy Junior Nonappointment agreement ministers cabinet (watchdogs) positions

Coalition discipline: All = Discipline always expected; Most = Discipline expected except on explicitly exempted matters; Spec. = Discipline only expected on a few explicitly specified matters Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement; No = No explicit agreement

Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee; CoC = Coalition committee; Pca = Combination of cabinet members and parliamentarians; Parl = Parliamentary leaders; PS = Party summit

Coalition agreement: IE = inter-election; PRE = pre-election; POST = post-election; N = No coalition agreement

Notes: Up until 1994, during periods where the values for the variables remain identical, the first and last applicable cabinets are listed. The last applicable cabinet is right-justified in the Coalition column.

2018 POST

PS, Pca CaC CaC CaC CaC Parl Parl PS PS Pca

PS, Pca CaC, CoC CaC, CoC CaC, CoC CaC, CoC IC, Parl IC, Parl PS PS PS, Pca

N N N N N N N N N N

2007 2008 2010 2010 2011 2013 2013 2014 2015 2016

Prodi III Berlusconi IV Berlusconi V Berlusconi VI Berlusconi VII Letta I Letta II Renzi I Renzi II Gentiloni Silveri Conte I

Y

Personal Issues union excluded from All used Most For most agenda common serious conflicts

N PRE N N N N N N N N

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Coalition

Table 12.5 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



429

This function is usually carried out in the informal meetings—known as preconsiglio—that take place before the official meeting of the cabinet. During such informal meetings, the agenda of the Council of Ministers is decided and some preliminary agreement among all cabinet members is sought. Often, the preconsiglio chooses to postpone the discussion of the most divisive issues to avoid open conflicts in the Council of Ministers. In 2001, for the first time a minister ‘for the Implementation of the government programme’ was established for the purpose of ensuring the realization of government electoral pledges. However, this minister without portfolio has been concerned above all with communicating to the public the main achievements of the government (Verzichelli 2014: 193), without performing any real coordination task of the other ministers’ activities. Besides the appointment of JMs, delegation to ministers in Italy has been handled also through a set of governance mechanisms that are common in Western European democracies (Andeweg and Timmermans 2008). Generally speaking, Italian cabinets have made a very limited use of conflict resolution instruments at the purely executive level, which differentiates Italy from countries such as Austria, Belgium, and the Nordic democracies. These tools, which are also known as ‘internal arenas’ as they recognize only ministers as legitimate participants, typically include inner cabinets and cabinet committees. As reported in Table 12.5, for almost none of the governments from 1946 to 2018, inner cabinets represented the most frequently used mechanism or the option chosen to deal with the most serious conflicts. An inner cabinet, with the participation of the PM and the coalition parties’ leading ministers, was indeed set up under Bettino Craxi’s prime ministership in the 1980s. However, inner cabinets have never been formalized. Almost the same holds in the case of cabinet committees, which are a subset of ministers deciding upon the government’s joint initiatives in specific policy domains. During the 1960s, some inter-ministerial committees (comitati interministeriali) were created to coordinate the activity of the Italian cabinet in specific policy areas. The most powerful of these was the CIPE (Interministerial Committee for Economic Planning). Although facilitating the coordination among a small number of ministers, Italian inter-ministerial committees were largely ineffective in solving conflicts emerging in the cabinet as a whole (Hine 1993; Criscitiello 1994). During the Second Republic cabinet committees have been used as the most common method for managing intra-coalition divergence by the centre-right executives, and inner cabinets were employed under the Conte I cabinet. Another peculiarity the M5S–League coalition shared with most of Berlusconi’s governments was the presence in the cabinet of all the leaders of the government parties. Before Conte’s and Berlusconi’s governments, in the past a ‘personal union’ was found only in Aldo Moro’s first government (1963–1968) and in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

430

  

Craxi’s executive (1983–1987). Indeed, even the leader of the largest coalition party has only rarely been a member of the cabinet in the entire post-war period.¹⁵

Mixed and external governance mechanisms In Italy, the attempts to resolve conflicts within the cabinet have often taken place in ‘mixed’ decisional arenas—where both ministers and actors who do not belong to the cabinet participate—and even in arenas that are entirely ‘external’ to the executive. In other words, the deepest divisions emerging in the Italian cabinets have typically been settled in meetings where external figures such as party leaders played a primary role (Cotta and Isernia 1996). Intra-coalition disputes were often triggered by one or more governing parties with the purpose of renegotiating the allocation of portfolios through a reshuffle (rimpasto) or adjusting the policy agenda of the cabinet (Verzichelli and Cotta 2000). A huge part of the conflicts inside Italian cabinets have involved two or more ministers but not the PM. As reported by Jaakko Nousiainen (1993), in the 1970–1987 period the Italian PM took part in just 8 per cent of all intra-coalition disputes. During the Second Republic, this percentage rose up to 25 per cent, indicating a greater exposure of the PM as well as his moderately increased role in the cabinet decision-making (Marangoni and Vercesi 2015). As Table 12.5 illustrates, during the First Republic two types of mechanism turned out to be particularly crucial: coalition committees and party summits. Coalition committees, also known as ‘majority summits’ (vertici), were semi-informal meetings of mixed nature, where the party leaders, the PM, and possibly other ministers discussed the government’s agenda and handled divergences within the cabinet. According to Annarita Criscitiello (1994), the impact of coalition committees in Italy was ambivalent. For sure, majority summits strengthened the executive by providing an instrument for increasing coordination and reaching decisions. At the same time, however, they limited the cabinet’s autonomy from parties by enabling party leaders to actively participate in the decision-making process. While majority summits represented the most prominent instrument of conflict resolution for Italian cabinets up to the late 1970s, decisions concerning the management of intra-coalition disagreement shifted further apart from the cabinet thenceforward. During the 1980s and early 1990s, party summits—which bring together the leaders of the coalition parties—became the most common method for dealing with intra-coalition divergence as well as the one used to confront the most severe conflicts.

¹⁵ Remarkable exceptions to this pattern are Alcide De Gasperi’s cabinets during the ‘centrist coalition’ formula (1948–1953), Fanfani II and IV cabinets in the stages preparing for the ‘centre-left coalition’ of the 1960s, De Mita cabinet (1988–1989) in the late ‘five-party coalition’ phase, and more recently the centre-left governments led by Renzi (2014–2016).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



431

During the Second Republic, there was a greater use of decision arenas less open to politicians outside the executive. As previously noted, Berlusconi’s cabinets usually relied on cabinet committees (an arena purely internal to the executive), which they combined with coalition committees (mixed) for the most serious disputes. At the same time, however, centre-left governments of the Second Republic revived the patterns of First Republic cabinets by adopting external (party summits) or mixed arenas (coalition committees). Coalition committees—typically including the PM and other ministers as well as some party leaders outside the cabinet—are then a still crucial conflict resolution mechanism for Italian executives (Cotta and Marangoni 2015; Marangoni and Vercesi 2015). However, it should be added that, in the latest period, when the overlap between government leadership and party leadership has been larger than during the First Republic, party summits and coalition committees have in practice been more internal to the government arena than in the past. The Conte I government fits into this account as intra-cabinet disagreements, rather than being settled in the conciliation committee mentioned in the coalition treaty, were reconciled in party summits involving the leaders of the two coalition parties—who were also ministers and deputy PMs—or by inner cabinets where the PM also participated.

Governance mechanisms in the parliamentary arena Despite the renowned ‘centrality’ of the Italian parliament (Cotta 1994), almost no Italian cabinet has relied on governance mechanisms entirely made up of parliamentary leaders. The only exceptions were the two governments headed by Letta between 2013 and 2014, where external decisional arenas composed of parliamentary leaders were employed. Attendants of such informal meetings were the heads of the coalition parties’ parliamentary groups. Nevertheless, the legislative arena is a crucial venue where Italian coalition parties try to resolve the government’s internal conflicts and make joint policy. The Italian parliament has traditionally been depicted as institutionally strong, especially due to the outstanding prerogatives of its permanent committees (Lees and Shaw 1979; Strøm 1990; Della Sala 1993; Mattson and Strøm 1995; Lijphart 2012). In the course of Italy’s First Republic, lawmaking was extremely consensual: a major part of legislation was directly adopted behind the closed doors of the committees, and those bills that were voted on the floor usually received very high percentages of favourable votes (Di Palma 1977).¹⁶ Under the governments led by the DC, the cohesiveness of the ¹⁶ Italian standing committees can give final approval to legislation without referring it to the assembly on the whole. In the Chamber, such a power can be withheld by the government, by onetenth of all representatives, or by one-fifth of the committee members. Similar rules exist in the Senate.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

432

  

legislative majority was not really assured, although discipline in the parliament was somehow stronger during the initial phase of each coalition cycle (Verzichelli and Cotta 2000). A certain level of consensus seems to have persisted also in the early stage of the Second Republic (Giuliani 2008; De Giorgi 2016) but, compared to the past, the lawmaking process in Italy has experienced important changes. First, the government has ostensibly increased its agenda-setting powers in the legislative process. The shift from a pivotal to an alternational party system that occurred in the early 1990s has prompted Italian governments to be more effective in lawmaking (Zucchini 2013) and to make a special effort to implement their electoral pledges (Marangoni 2013). This is why Italian cabinets of the Second Republic are based not just on some agreed-on policy compromise but also on an explicit understanding that there will be coalition discipline in parliamentary votes on legislation and in other types of parliamentary behaviour (Table 12.5).¹⁷ Remarkably, the agenda-setting prerogatives of the Italian government have been strengthened without any change in the formal rules. During the Second Republic, cabinets have extensively—and in many cases improperly—drawn upon a number of tactics to circumvent the ordinary legislative procedures: law decrees, delegation laws, maxi-amendments, and requests of a parliamentary confidence vote on cabinet-sponsored bills (Vassallo 1994; Capano and Giuliani 2001; Zucchini 2011; De Micheli 2014). Second, whereas during the First Republic parliamentary institutions hosted continuous negotiations that often also involved opposition parties, in more recent times the Italian parliament has become more a bargaining arena for the members of the governing coalition. As Lanny W. Martin and Georg Vanberg (2004, 2005, 2011) pointed out, under multiparty government the implementation of the coalition compromise is potentially undermined by ministerial drift: each minister may attempt to use his or her discretion and informational advantage in drafting legislation to move policy in directions favoured by his or her own party, but that can be damaging with respect to the interests of other coalition members. The legislative review—namely changing government bills through amendments—is one of the most effective means for containing ministerial drift. Some recent analyses of the Italian legislative process have indeed shown that bills drafted by ministers, once introduced in the legislature, become subjected to a bargaining process among the coalition members. As a result,

Throughout the First Republic, parties outside the government only rarely opposed a decision to endow a committee with direct legislative powers, thus showing a clearly cooperative attitude towards the executive. ¹⁷ Apart from the 2013–2018 period, the Italian governments of the Second Republic adopted to some degree a conflict-avoidance strategy. Both the members of centre-right coalitions and those of centre-left coalitions agreed at the onset not to change the status quo in certain policy issues and hence excluded them from the government’s agenda.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



433

government bills are more extensively amended in parliament compared to the past. The amount of modifications made, as well as the overall duration of the legislative process, is found to depend on the ideological conflicts within the coalition, rather than the divide between government and opposition. Furthermore, during the First Republic, the parliamentary modifications extended the consensus beyond the government majority, while during the Second Republic they have ended up to narrow it. All in all, voting behaviour on government legislation during the Second Republic more clearly reflects the divide between government and opposition (Pedrazzani and Zucchini 2013; Pedrazzani 2017). In the Italian parliament, patterns of managed delegation are realized not only by means of legislative review but also through the allocation of a set of agendasetting offices in the legislature that are sometimes known as important ‘mega seats’ (Carroll et al. 2006).¹⁸ In this regard, it is worth mentioning the key role of the Speakers of the two houses. The reforms of the parliamentary Standing Orders that were enacted during the 1990s strengthened the prerogatives of the two presidents in setting the legislative agenda (De Micheli and Verzichelli 2004). As a result, in the alternational era the Speakers of the Chamber and Senate have changed their role compared to the past, shifting from a neutral position towards a role as agents of the coalition supporting the government. Another powerful legislative-level instrument for containing ministerial drift are the chairs of parliamentary committees (also mega seats). In countries with multiparty executives like Italy, committee chairs tend to be distributed among coalition parties in such a way as to monitor each other. Each cabinet minister is often ‘shadowed’ in his or her policy jurisdiction by a committee chair belonging to one of the other parties in the coalition (Kim and Loewenberg 2005; Carroll and Cox 2012).

Cabinet duration and termination The duration of cabinets If the Italian First Republic stood out for the extremely short life span of its governments (on average less than one year), the mean cabinet duration throughout the Second Republic has not been much longer: only 13 months in the 1994–2017 period (Curini and Pinto 2017). However, in the decade dominated by Berlusconi governments (2000–2009), governments’ relative duration with respect to their maximum potential duration was above 50 per cent—that is the highest value from the birth of the Republic. Berlusconi’s extra-political resources, his personal charisma, and his full control of the main coalition party (FI/PdL) ¹⁸ At the same time, there is no clear evidence that parliamentary questions are used by coalition members in such a way as to check one another (Russo and Wiberg 2010; Russo 2011).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

434

  

made centre-right cabinets more stable over time than any other. The centre-left governments’ high instability reflects the great ideological heterogeneity of their parliamentary coalitions. The decade from 2010 onwards is characterized by a relative duration around 25 per cent, similar to the short government duration during the 1980s and 1990s (Table 12.6). None of the cabinets formed by the winning pre-electoral coalition lasted for the entire length of the constitutional term, and several inter-electoral executives were sworn in. In just one case—the second Prodi’s government, here split into Prodi II and III—the fall of a cabinet that had resulted victorious in the elections led to early elections. Moreover, in all the legislatures of the Second Republic, the coalition supporting the government changed from one cabinet to the next one. The only partial exception was the XIV Legislature, during which two minor parties shifted from providing external support for Berlusconi II to direct participation in Berlusconi III cabinet.

The termination of cabinets During the First Republic, most crises leading to cabinet termination had an ‘extra-parliamentary’ nature, as the cabinet resigned without a parliamentary vote of no confidence. Several First Republic executives were pushed to end because some important government-sponsored piece of legislation failed to be approved in the legislature, without the PM and his government being openly defeated by a vote of no confidence.¹⁹ In contrast, a vote of no confidence took place twice during the Second Republic, in Prodi’s first and second (here III) cabinets. Table 12.6 provides a detailed overview of the various causes of termination of Italian governments (see also Cotta and Marangoni 2015). In the Second Republic, just three cabinets (Amato II, Berlusconi III, and Gentiloni Silveri) ended for technical reasons—that is because regular parliamentary elections were held. All the other governments fell due to discretionary reasons—most of the time, as a consequence of conflicts of policy or personal nature among coalition parties. In many instances, in the presence of serious intra-coalition tensions an external shock accelerated cabinet termination or provided a pretext for the government’s fall. In a couple of cases (D’Alema II and Berlusconi II), negative results of the ruling parties in the regional elections exacerbated pre-existing intra-coalition divisions. The crisis of the last Prodi’s cabinet was instead triggered by the resignation of the UDEUR’s leader and minister of justice, whose wife had been investigated for corruption. However, the cohesion of Prodi’s coalition had been ¹⁹ However, some governments—the most recent of which was Fanfani VI in 1987—failed to obtain the parliament’s confidence at the time of their investiture vote.

1991-03-29

1992-04-06 1993-04-22

1994-01-13

1988-04-13

1989-07-23

1994-05-11

1995-01-17

1996-05-18

Andreotti V

Andreotti VI 1991-04-13 Amato I 1992-06-28

1993-04-29

De Mita

Ciampi

Berlusconi I

Dini

Prodi I

1998-10-09

1996-01-11

1994-12-22

1989-05-19

1988-03-11

1987-07-29

Goria

Date out

Date in

Cabinet

46.8

22.4

12.2

17.2

72.4 16.4

54.6

25.2

12.2

Relative duration (%)

Table 12.6 Cabinet termination in Italy, 1987–2018

6, 7a

4

7a, 7b

4

1 9

7b, 8

7b

6, 7a

Mechanisms of cabinet termination

10, 11

Terminal events

Ulivo, Ex Left

(7a) AN, LN (7b) FI–PdL, LN

(7a) DC, PRI, PSI, (8) DC

PSI, DC

DC, PSI

Parties (when conflict between or within)

3, 8, 16

1

22

1, 8

Policy area(s)

Continued

Voluntary resignation of the PM after the results of referenda (which meant, impossibility to put a majority together on different issues). But an important cause was also the dramatic decline in legitimacy of the political class (‘mani pulite’ investigation etc.). Cabinet formed only to implement electoral and economic reforms before a new early election. The conflict was mainly personal (between the LN leader, who withdrew his party’s ministers from cabinet) and Berlusconi. There was also a conflict between LN and AN about the federal reform. Temporary cabinet to complete some reforms before the early elections. After the decision of RC (Ex Left) to revoke the support for the cabinet, the PM asked a confidence vote which failed for one vote. During previous months there were policy conflicts

Defeat of the financial bill (already delayed out of the financial year). A number of criticisms about different policies between DC and PSI. Extra-parliamentary crisis determined at the end of the PSI congress. Primary reason of crisis: conflict within the DC about TV regulation. After the replacement of some ministers a number of personal conflicts among parties and ministers.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2000-04-19

2001-05-13 2005-04-20

2006-04-10

2007-05-15

1999-12-22

Amato II 2000-04-26 Berlusconi II 2001-06-11

2005-04-23

2006-05-17

2007-05-16

D'Alema II

Berlusconi III Prodi II

Prodi III

2008-01-24

1999-12-18

1998-10-21

D'Alema I

Date out

Date in

Cabinet

Table 12.6 Continued

16.9

19.5

77.5

89 75.5

21.4

43.1

Relative duration (%)

4, 6

8

1

1 7a

7b

7b

Mechanisms of cabinet termination

14

10

10

Terminal events

FI–PdL, LN, AN, CCD– CDU–UDC

PCI–PDS–DS, SDI

Parties (when conflict between or within)

2, federalism

Policy area(s)

A minister who previously was member of DS left the parliamentary group of Ulivo-PD and became member of a new parliamentary group (SD). The government lost a vote of confidence in the Senate on 24 January 2008, prompting Prodi’s resignation. The president of the Republic hence dissolved parliament to make way for early elections. The crisis was triggered by the justice

between the government coalition (Ulivo) and Ex Left. SDI withdrew from the cabinet, arguing that D’Alema was too uncharismatic to win the next general election. D’Alema resigned after heavy losses in the regional elections. Analysts attributed the defeat of the ruling coalition to internal strife within the coalition. Following a negative result in the regional elections for the governing coalition, UDC withdrew from the coalition after Berlusconi’s refusal to amend the government programme or call early elections. AN also threatened to withdraw. Both also criticized Berlusconi for the influence he gave LN within the coalition, whose policy positions UDC and AN considered as one of the reasons for the electoral defeat. Berlusconi finally resigned on 20 April 2005.

Comments OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2013-04-28

2013-11-15

2014-02-22

2015-02-18

Letta I

Letta II

Renzi I

Renzi II

2011-11-12

2011-03-23

2016-12-07

2015-02-17

2014-02-14

2013-11-14

2012-12-21

2011-03-22

2010-11-17

Berlusconi VI Berlusconi VII

2011-11-16

2010-11-16

Berlusconi V 2010-07-30

Monti

2010-07-29

2008-05-08

Berlusconi IV

56.1

23.5

5.6

10.9

68.7

28.5

13.2

10.3

43.4

9

8

8

7b, 8

4, 9

9

5

7b

8

10

11, 14

SC

PD

(7b) PD, FI– PdL, SC, CCD–CDU– UDC, PR (8) FI–PdL

FI–PdL, FLI

FI–PdL

Continued

Berlusconi resigned amidst pressure from LN, from the finance minister, and from the president of the Republic. Monti resigned after passing the 2013 budget, having lost the external support of PdL the month prior. After the government refused to offer immunity from prosecution to Berlusconi, PdL withdrew from the government. PdL also suffered a split in the process, as a number of its deputies and senators and all of its ministers formed the NCD and remained in the government. Letta resigned after losing a vote of no confidence undertaken by the PD party executive. The only minister affiliated with SC left the SC parliamentary group to become member of the PD. Renzi resigned after the defeat in the constitutional referendum of 4 December 2016.

minister, who went under investigation for corruption. A minister, who was previously member of PdL, left the PdL parliamentary group to get affiliated to another new parliamentary group (FLI). The FLI parliamentary group withdrew its support to the government. A new party (PT-CN) entered the government. OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2019-08-20

2018-06-01

24.6

87.8

Relative duration (%)

7a, 7b

1

Mechanisms of cabinet termination

Terminal events

M5S, LN

Parties (when conflict between or within)

Comments

Transports The results of the European elections of May 2019 made clear that the League (LN) had surpassed the M5S in terms of popularity. This prompted the League’s leader Salvini to claim greater room for manoeuvre within the government. In early August, the government coalition appeared split in parliament on the construction of the TurinLyon high-speed railway. PM Conte resigned after a parliamentary motion of no-confidence was tabled (and then withdrew) by the League.

Policy area(s)

Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event Note: Relative duration: share of constitutionally allowed duration for this cabinet.

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of PM

2018-03-04

2016-12-12

Gentiloni Silveri Conte I

Date out

Date in

Cabinet

Table 12.6 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



439

undermined by policy conflicts since the birth of the government. Severe tensions within the executive had already emerged about the government’s foreign policy, the approval of the budget, and an impending referendum on the electoral law. As already discussed, a mix of factors led Berlusconi to resign in November 2011 (here Berlusconi VII). Notably, the fall of Letta’s government (here Letta II) is the only case—in the Second Republic—in which the PM had to resign because of a decision of his party. Driven by the new secretary Renzi, in February 2014 the PD national directorate approved a motion urging a new government. Less than three years later, the Renzi government (here Renzi II) fell due to the PM’s decision to resign after a defeat in the 2016 institutional referendum. Finally, in August 2019 the Conte I cabinet fell because of tensions triggered by the League, which had surpassed the coalition partner M5S in the European elections held three months earlier.

Conclusions The evolution of government coalitions and cabinet governance in Italy is complex and challenges any attempt to forecast future developments. After the fall of the First Republic, as the effects of new party competition dynamics and electoral rules on the government arena began to settle, competition and rules further changed. Instability seems to reign supreme, especially since new parties and forces gained strength in the 2013 and 2018 elections. At the same time, under the surface of the continuous transformation some ancient features, just slightly attenuated, of the First Republic remain. What emerges is a disharmonic and certainly provisional picture of the Italian coalition life cycle, characterized on the one hand by few changes in the formal rules and on the other hand by practices that are very sensitive to political developments. For about 25 years during the Second Republic, government alternation has almost been the rule. Simultaneously, party fragmentation of the governments has often been greater than was the case during the First Republic. Competing preelectoral coalitions have been built, but no detailed agreement has ruled the life of the governments that have emerged from these. The explicit electoral mandate for the leader of the winning coalition has simplified bargaining for the birth of postelection governments, but this has not prevented the governments from collapsing frequently. Inter-election formations have often lacked a clear and obvious link with electoral results. To complicate this picture, the cabinets formed after the last two general elections (2013 and 2018) could be installed only after the breakdown of pre-electoral coalitions. Italian governments have strengthened in the legislative process, but they still have a short life span and, up to 2018, had been victims of a symmetrical and incongruent bicameralism. Among the instruments of conflict resolution, at least

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

440

  

those used against the most serious divisions continue to be largely external to the government arena. While ministers have lost some of their former discretionary authority, PMs have not necessarily benefited from this loss. To add complexity and uncertainty, all these characteristics—here roughly summarized—have not appeared with the same intensity in all of the governments of the Second Republic. Berlusconi’s governments—perhaps because of their greater ideological compactness and even more likely because of the PM’s political and extra-political resources—lasted longer and relied more often on internal government mechanisms to resolve conflicts. Still, traces of the past reappear from time to time, as no political and institutional equilibrium seems able to prevail.

References Andeweg, Rudy B., and Arco Timmermans (2008). ‘Conflict Management in Coalition Government’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 269–300. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Anna Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999-2010’. Party Politics, 21(1): 143–52. Bardi, Luciano (2007). ‘Electoral Change and Its Impact on the Party System in Italy’. West European Politics, 30(4): 711–32. Bartolini, Stefano, and Roberto D’Alimonte (1994). ‘La competizione maggioritaria: le origini elettorali del parlamento diviso’. Rivista Italiana di Scienza Politica, 24(3): 631–86. Bordignon, Fabio, and Luigi Ceccarini (2013). ‘Five Stars and a Cricket. Beppe Grillo Shakes Italian Politics’. South European Society and Politics, 18(4): 427–49. Capano, Giliberto, and Marco Giuliani (2001). ‘Governing without Surviving? An Italian Paradox: Law-Making in Italy, 1987-2001’. Journal of Legislative Studies, 7 (4): 13–36. Carroll, Royce, and Gary W. Cox (2012). ‘Shadowing Ministers: Monitoring Partners in Coalition Governments’. Comparative Political Studies, 45(2): 220–36. Carroll, Royce, Gary W. Cox, and Mónica Pachón (2006). ‘How Parties Create Electoral Democracy, Chapter 2’. Legislative Studies Quarterly, 31(2): 153–74. Chiaramonte, Alessandro, and Aldo Di Virgilio (2006). ‘Da una riforma elettorale all’altra: partiti, coalizioni e processi di apprendimento’. Rivista Italiana di Scienza Politica, 36(3): 363–91.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



441

Chiaramonte, Alessandro, and Roberto D’Alimonte (2018). ‘The New Italian Electoral System and Its Effects on Strategic Coordination and Disproportionality’. Italian Political Science, 13(1): 8–18. Chiaramonte, Alessandro, and Vincenzo Emanuele (2014). ‘Bipolarismo addio? Il sistema partitico tra cambiamento e de-istituzionalizzazione’. In Alessandro Chiaramonte and Lorenzo De Sio (eds), Terremoto elettorale. Le elezioni politiche del 2013. Bologna: Il Mulino, 233–62. Conti, Nicolò (2015). ‘From Words to Facts: The Implementation of the Government Agreement’. In Nicolò Conti and Francesco Marangoni (eds), The Challenge of Coalition Government: The Italian Case. Abingdon: Routledge, 58–74. Conti, Nicolò, Andrea Pedrazzani, and Federico Russo (2020). ‘Policy Polarisation in Italy: The Short and Conflictual Life of the “Government of Change” (2018-2019)”. South European Society and Politics, DOI: 10.1080/13608746.2020.1840110. Costa Pinto, António, Maurizio Cotta, and Pedro Tavares de Almeida (eds) (2018). Technocratic Ministers and Political Leadership in European Democracies. Cham: Palgrave Macmillan. Cotta, Maurizio (1994). ‘The Rise and Fall of the “Centrality” of the Italian Parliament: Transformations of the Executive-Legislative Subsystem after the Second World War’. In Samuel C. Patterson and Gary W. Copeland (eds), Parliaments in the Modern World: Changing Institutions. Ann Arbor: University of Michigan Press, 59–84. Cotta, Maurizio (1996). ‘La crisi del governo di partito all’italiana’. In Maurizio Cotta and Pierangelo Isernia (eds), Il gigante dai piedi di argilla: la crisi del regime partitocratico in Italia. Bologna: Il Mulino, 11–52. Cotta, Maurizio, and Francesco Marangoni (2015). Il governo. Bologna: Il Mulino. Cotta, Maurizio, and Luca Verzichelli (2007). Political Institutions in Italy. Oxford: Oxford University Press. Cotta, Maurizio, and Pierangelo Isernia (eds) (1996). Il gigante dai piedi di argilla: la crisi del regime partitocratico in Italia. Bologna: Il Mulino. Criscitiello, Annarita (1994). ‘The Political Role of Cabinet Ministers in Italy’. In Michael Laver and Kenneth A. Shepsle (eds), Cabinet Ministers and Parliamentary Government. Cambridge: Cambridge University Press, 187–200. Curini, Luigi, and Luca Pinto (2017). L’arte di fare (e disfare) i governi. Milano: Egea. De Giorgi, Elisabetta (2016). L’opposizione parlamentare in Italia. Dall’antiberlusconismo all’antipolitica. Roma: Carocci. De Micheli, Chiara (2014). Parlamento e governo in Italia. Partiti, procedure e capacità decisionale (1948–2013). Milano: Franco Angeli. De Micheli, Chiara, and Luca Verzichelli (2004). Il Parlamento. Bologna: Il Mulino. Della Sala, Vincent (1993). ‘The Permanent Committees of the Italian Chamber of Deputies: Parliament at Work?’. Legislative Studies Quarterly, 18(2): 157–83. Di Palma, Giuseppe (1977). Surviving without Governing: The Italian Parties in Parliament. Berkeley: University of California Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

442

  

Di Virgilio, Aldo (2006). ‘Dal cambiamento dei partiti all’evoluzione del sistema partitico’. In Leonardo Morlino and Marco Tarchi (eds), Partiti e caso italiano. Bologna: Il Mulino, 173–206. Di Virgilio, Aldo, Daniela Giannetti, Andrea Pedrazzani, and Luca Pinto (2015). ‘Party Competition in the 2013 Italian Elections: Evidence from an Expert Survey’. Government and Opposition, 50(1): 65–89. Döring, Herbert (ed.) (1995). Parliaments and Majority Rule in Western Europe. New York: St. Martin’s. Giannetti, Daniela (2013). ‘Mario Monti’s Technocratic Government’. In Aldo Di Virgilio and Claudio Radaelli (eds), Italian Politics 2012. Technocrats in Office. New York and Oxford: Berghahn, 133–52. Giannetti, Daniela, and Michael Laver (2005). ‘Policy Positions and Jobs in the Government’. European Journal of Political Research, 44(1): 91–120. Giannetti, Daniela, Andrea Pedrazzani, and Luca Pinto (2017). ‘Party System Change in Italy: Politicising the EU and the Rise of Eccentric Parties’. South European Society and Politics, 22(1): 21–42. Giannetti, Daniela, Andrea Pedrazzani, and Luca Pinto (2018). ‘The Rising Importance of Non-economic Policy Dimensions and the Formation of the Conte Government in Italy’. Italian Political Science, 13(2): 27–44. Giuliani, Marco (2008). ‘Patterns of Consensual Law-Making in the Italian Parliament’. South European Society and Politics, 13(1): 61–85. Hine, David (1993). Governing Italy: The Politics of Bargained Pluralism. Oxford: Clarendon Press. Keman, Hans (1994). ‘The Search for the Centre: Pivot Parties in West European Party Systems’. West European Politics, 17(4): 124–48. Kim, Don-Hung, and Gerhard Loewenberg (2005). ‘The Role of Parliamentary Committees in Coalition Governments: Keeping Tabs on Coalition Partners in the German Bundestag’. Comparative Political Studies, 38(9): 1104–29. Laver, Michael, and Kenneth A. Shepsle (1996). Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Lees, John D., and Malcolm Shaw (eds) (1979). Committees in Legislatures: A Comparative Analysis. Durham, NC: Duke University Press. Lijphart Arend (2012). Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, 2nd edn. New Haven: Yale University Press. Marangoni, Francesco (2013). Provare a governare, cercando di sopravvivere. Esecutivi e attività legislativa nella Seconda Repubblica. Pisa: Pisa University Press. Marangoni, Francesco, and Michelangelo Vercesi (2015). ‘The Government and Its Hard Decisions: How Conflict Is Managed within the Coalition’. In Nicolò Conti and Francesco Marangoni (eds), The Challenge of Coalition Government: The Italian Case. London: Routledge, 17–35.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



443

Marradi, Alberto (1982). ‘Italy: From “Centrism” to Crisis of the Center-Left Coalitions’. In Eric C. Browne and John Dreijmanis (eds), Government Coalitions in Western Democracies. New York: Longman, 33–70. Martin, Lanny W., and Georg Vanberg (2004). ‘Policing the Bargain: Coalition Government and Parliamentary Scrutiny’. American Journal of Political Science, 48(1): 13–27. Martin, Lanny W., and Georg Vanberg (2005). ‘Coalition Policymaking and Legislative Review’. American Political Science Review, 99(1): 93–106. Martin, Lanny W., and Georg Vanberg (2011). Parliaments and Coalitions. The Role of Legislative Institutions in Multiparty Governance. Oxford: Oxford University Press. Mattson, Ingvar, and Kaare Strøm (1995). ‘Parliamentary Committees’. In Herbert Döring (ed.), Parliaments and Majority Rule in Western Europe. New York: St. Martin’s, 249–307. McDonnell, Duncan, and Marco Valbruzzi (2014). ‘Defining and Classifying Technocrat-Led and Technocratic Governments’. European Journal of Political Research, 53(4): 654–71. Mershon, Carol (2002). The Costs of Coalition. Stanford: Stanford University Press. Morlino, Leonardo (1996). ‘Crisis of Parties and Change of Party System in Italy’. Party Politics, 2(1): 5–30. Moury, Catherine (2013). Coalition Government and Party Mandate. How Coalition Agreements Constrain Ministerial Action. London: Routledge. Müller, Wolfgang C., and Kaare Strøm (eds) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Müller, Wolfgang C., and Kaare Strøm (2008). ‘Coalition Agreements and Cabinet Governance’, in Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 159–99. Nousiainen, Jaakko (1993). ‘Decision-Making, Policy Content and Conflict Resolution in Western European Cabinets’. In Jean Blondel and Ferdinand Müller-Rommel (eds), Governing Together. The Extent and Limits of Joint Decision-Making in Western European Cabinets. London: Palgrave Macmillan, 259–83. Pedrazzani, Andrea (2017). Fare le leggi nella Seconda Repubblica. Come cambia il Parlamento. Milano: Egea. Pedrazzani, Andrea (ed.) (2018). ‘Special Issue: “Governo del Cambiamento”? Italian Politics under the Yellow-Green Government’. Italian Political Science, 13(2): 1–114. Pedrazzani, Andrea, and Luca Pinto (2015). ‘The Electoral Base: The “Political Revolution” in Evolution’. In Filippo Tronconi (ed.), Beppe Grillo’s Five Star Movement. Organization, Communication and Ideology. Farnham: Ashgate, 75–98. Pedrazzani, Andrea, and Francesco Zucchini (2013). ‘Horses and Hippos: Why Italian Government Bills Change in the Legislative Arena, 1987–2006’. European Journal of Political Research, 52(5): 687–714.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

444

  

Pedrazzani, Andrea, and Francesco Zucchini (2020). ‘ “Useless Approvals”. Italian Bicameralism and Its Decisional Capacity’. Journal of Legislative Studies, 26(4): 578–605. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of AntiElitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research and Politics, 4(1): 1–9. Russo, Federico (2011). ‘The Constituency as a Focus of Representation: Studying the Italian Case through the Analysis of Parliamentary Questions’. Journal of Legislative Studies, 17(3): 290–301. Russo, Federico, and Matti Wiberg (2010). ‘Parliamentary Questioning in 17 European Parliaments: Some Steps towards Comparison’. Journal of Legislative Studies, 16(2): 215–32. Sartori, Giovanni (1976). Parties and Party Systems: A Framework for Analysis. Cambridge: Cambridge University Press. Strøm, Kaare (1990). Minority Governments and Majority Rule. Cambridge: Cambridge University Press. Strøm, Kaare (2003). ‘Parliamentary Democracy and Delegation’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press, 55–106. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Tebaldi, Mauro (2014). ‘From Notary to Ruler: The Role of the President of the Republic during the Italian Crisis (2010-14)’. South European Society and Politics, 19(4): 561–81. Thies, Michael F. (2001). ‘Keeping Tabs on Partners: The Logic of Delegation in Coalition Governments’. American Journal of Political Science, 45(3): 580–98. Timmermans, Arco (2003). High Politics in the Low Countries. An Empirical Study of Coalition Agreements in Belgium and the Netherlands. Aldershot: Ashgate. Timmermans, Arco (2006). ‘Standing Apart and Sitting Together: Enforcing Coalition Agreements in Multiparty Systems’. European Journal of Political Research 45(2): 263–83. Tronconi, Filippo (2018). ‘The Italian Five Star Movement during the Crisis: Towards Normalisation?’ South European Society and Politics, 23(1): 163–80. Vassallo, Salvatore (1994). Il governo di partito in Italia. Bologna: Il Mulino. Verzichelli, Luca (2009). ‘Italy: The Difficult Road towards a More Effective Process of Ministerial Selection’. In Keith Dowding and Patrick Dumont (eds), The Selection of Ministers in Europe: Hiring and Firing. London: Routledge, 79–100. Verzichelli, Luca (2014). ‘C’era una volta la partitocrazia. La dinamica del governo di partito nel ventennio berlusconiano’. In Marco Almagisti, Luca Lanzalaco, and Luca Verzichelli (eds), La transizione politica italiana. Da Tangentopoli a oggi. Roma: Carocci, 171–96.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



445

Verzichelli, Luca, and Maurizio Cotta (2000). ‘Italy: From “Constrained” Coalitions to Alternating Governments?’. In Wolfgang C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 433–97. Verzichelli, Luca, and Maurizio Cotta (2018). ‘Shades of Technocracy: The Variable Use of Non-Partisan Ministers in Italy’. In António Costa Pinto, Maurizio Cotta, and Pedro Tavares de Almeida (eds), Technocratic Ministers and Political Leadership in European Democracies. Cham: Palgrave Macmillan, 77–110. Zucchini, Francesco (2008). ‘Dividing Parliament? Italian Bicameralism in the Legislative Process (1987–2006)’. South European Society and Politics, 13(1): 11–34. Zucchini, Francesco (2011). ‘Government Alternation and Legislative Agenda Setting’. European Journal of Political Research, 50(6): 749–74. Zucchini, Francesco (2013). La repubblica dei veti. Un’analisi spaziale del mutamento legislativo in Italia. Milano: Egea.

Appendix. List of political parties Abbreviation Ex Left

PdCI PCI–PDS–DS

Indep. Left Verdi Rete PR

PSI SDI PSDI DC PPI RI UDR– UDEUR

Name Communist Refoundation (Rifondazione Comunista), 1991–2008 Party of Proletarian Unity for Communism (Partito di Unità Proletaria per il Comunismo), 1979–1987 Proletarian Democracy (Democrazia Proletaria), 1976–1991 Party of Italian Communists (Partito dei Comunisti Italiani) Democrats of the Left (Democratici di Sinistra), 1998–2006 Democratic Party of the Left (Partito Democratico della Sinistra), 1991–1998. Includes deputies and senators from Social Christians and Labour Federation 1992–1994. Italian Communist Party (Partito Comunista Italiano), 1946–1991 Independents of the Left (Indipendenti di sinistra) Federation of the Greens (Federazione dei verdi) The Net: movement for democracy (La Rete: movimento per la democrazia) More Europe (+Europa), 2018– Italian Radicals (Radicali Italiani), 2001–2014. Present only in the cabinet. Pannella List (Lista Pannella), 1994–1996 Radical Party (Partito Radicale), 1976–1994 Italian Socialist Party (Partito Socialista Italiano) Italian Social Democrats (Socialisti Democratici Italiani) Italian Socialist Democratic Party (Partito Socialista Democratico Italiano) Christian Democracy (Democrazia Cristiana) Italian Popular Party (Partito Popolare Italiano) Italian Renewal (Rinnovamento Italiano) Union of Democrats for Europe (Unione Democratici per l’Europa), 1999–2008 Democratic Union for the Republic (Unione Democratica per la Repubblica), 1998–1999

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

446

  

CCD–CDU– UDC

PRI LN PLI FI–PdL AN MSI SVP NPSI DeL IdV RnP PD MpA M5S SC SEL–SI

FdI CD

NCD FLI PT-CN PpI AD Dem. PS FeL LIF Rep.

Union of Christian and Centre Democrats/Union of the Centre (Unione dei Democratici Cristiani e di Centro/Unione di Centro), 2002–2018 United Christian Democrats (Cristiani Democratici Uniti), 1995–2002 Christian Democratic Centre (Centro Cristiano Democratico), 1994–2002 Italian Republican Party (Partito Repubblicano Italiano) League (Lega), 2018– Northern League (Lega Nord), 1992–2018 Partito Liberale Italiano (Italian Liberal Party) Go Italy (Forza Italia), 1994–2008, 2013– People of Freedom (Popolo della Libertà), 2008–2013 National Alliance (Alleanza Nazionale) Italian Social Movement - National Right (Movimento Sociale Italiano Destra Nazionale) South Tyrolean People’s Party (Südtiroler Volkspartei) New Italian Socialist Party (Nuovo Partito Socialista Italiano) Democracy is Freedom - The Daisy (Democrazia è Libertà - La Margherita) Italy of Values (Italia dei Valori) Rose in the Fist (Rosa nel Pugno) Democratic Party (Partito Democratico), 2007– Olive tree (L’Ulivo), 2006–2007 Movement for Autonomies (Movimento per le Autonomie) 5 Star Movement (Movimento 5 Stelle) Civics and Innovators (Civici e Innovatori), 2016–2018 Civic Choice (Scelta Civica), 2013–2016 Italian Left (Sinistra Italiana), 2016–2018 Possible (Possibile), 2015–2018 Left Ecology Freedom (Sinistra Ecologica Libertà), 2013–2016 Brothers of Italy (Fratelli d’Italia) Democratic Centre (Centro Democratico), 2013–2018 Alliance for Italy - Democratic Centre (Alleanza per l’Italia - Centro Democratico), 2010–2013 Popular Alternative (Alternativa Popolare), 2018– New Centre-Right (Nuovo Centrodestra), 2013–2018 Future and Liberty for Italy (Futuro e Libertà per l’Italia) People and Territory - National Cohesion (Popolo e Territorio - Coesione Nazionale) For Italy; Populars for Italy (Per l’Italia; Popolari per l’Italia) Democratic Alliance (Alleanza Democratica) The Democrats, Dem. (I Democratici, Dem.) Segni Pact (Patto Segni) Federalists and Liberaldemocrats (Federalisti e Liberaldemocratici) Italian Federalist League (Lega Italiana Federalista) Republicans, Actionists (Repubblicani, Azionisti), 2009–2013

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



CR AeF DCA

SD MAIE

NS FareItalia Lib.

Soc. GS IDeA

Fare Mod. CeR LeU NcI Other Life senators

447

Republicans, Liberals, and Reformers (Repubblicani, Liberai e Riformatori), 2007–2008 Republicans (Repubblicani), 2001–2006 Federalists, Liberal democrats, and Republicans (Federalisti, Liberaldemocratici e Repubblicani), 1999–2001 Reformer Centre (Centro Riformatore) Autonomists and Federalists (Autonomisti e Federalisti) Christian Democracy for Autonomies (Democrazia Cristiana per le Autonomie), 2006–2008 Democratic Ecologists (Ecologisti Democratici), 2005–2006 Democratic Left (Sinistra Democratica) Associative Movement Italians Abroad (Movimento Associativo Italiani all’Estero), 2018– Civic Choice - Liberalpopular Alliance Autonomies - Associative Movement Italians Abroad (Scelta Civica - Alleanza Liberalpopolare Autonomie - Movimento Associativo Italiani all’Estero), 2013–2018 Liberaldemocrats - Associative Movement Italians Abroad (Liberaldemocratici - Movimento Associativo Italiani all’Estero), 2009–2013 We South (Noi Sud) Doing Italy (Fare Italia) Small liberal parties: Free Italy - Liberals for Italy - Italian Liberal Party (Piccoli partiti liberali: Italia Libera - Liberali per l’Italia - Partito Liberale Italiano) Socialist Party (Partito Socialista) Great South (Grande Sud) Identity and Action (Identità e Azione) South American Union Italian Emigrants - Identity and Action (Unione Sudamerica Emigrati Italiani - Identità e Azione) To do! (Fare!) Moderates (Moderati) Conservatives and Reformers (Conservatori e Riformisti) Free and Equal (Liberi e Uguali) Us with Italy (Noi con l’Italia) Other minor (< 3 seats) parties. Always includes the deputy/senator who represents the autonomous region Val d’Aosta. Life senators who are not affiliated with any of the existing party groups (only Senate)

Note: Party names are given in English, followed by the party name in Italian in parentheses, with the exception of SVP, which is in German. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 13 The Netherlands: Old Solutions to New Problems Tom Louwerse and Arco Timmermans

New lines of conflict, political parties, and the fragmentation of the political landscape have contributed to cabinet volatility in the Netherlands. Politicians have tried to deal with those problems using mainly ‘old’ consensus-politics solutions. When the second Rutte cabinet faced a minority in the Senate in 2012, it chose to strike deals with opposition parties to ensure passage of major legislation. The prime minister did not like this, however: ‘I have such a [big, red] ear from calling [the opposition leaders].’ For his next cabinet, he preferred to go back to a coalition with a stable majority, as has been the norm in Dutch politics since the Second World War. Coalition politics in the Netherlands is rooted in a system of consociational democracy that allowed for a great deal of inclusiveness in governance (Lijphart 1968). While the old system of ‘pillarized’ political parties catering to a specific socio-economic or religious constituency has faded, the way governance is conducted still largely follows the same unwritten rules (Andeweg and Irwin 2014). This has resulted in long government formation processes in which policy is negotiated in great detail. The increasing complexity of the political landscape has made this formation process even more difficult in recent years and it has also required extensive efforts to deal with conflicts within the coalition. This chapter deals with the way in which politicians have tried to fit these old solutions to new problems. We first describe the institutional setting with a specific emphasis on changes since the turn of the century. We then describe the party system and political parties: how immigration and integration have become an important line of conflict and have given rise to several new political parties. We subsequently assess how these developments have affected the formation, functioning, duration, and termination of Dutch governments.¹ ¹ This chapter is based on various sources, including the (detailed) reports from (in)formateurs that are available online for all government formation processes since 2002 (via https://www. kabinetsformatie2017.nl/documenten). For the 2017 formation all documents related to the cabinet formation are available; for previous government formations all reports by (in)formateurs are available. The recent Dutch-language volume by Van Baalen and Van Kessel (2016) on Dutch cabinet formations also provides background information from interviews with key players. Moreover, newspaper sources Tom Louwerse and Arco Timmermans, The Netherlands: Old Solutions to New Problems In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0013

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

449

The institutional setting Government coalitions have been the norm in Dutch politics since its inception. Even before the introduction of a system of proportional representation (1917), no political party normally commanded a parliamentary majority. From a purely institutional perspective, coalition formation was not too complex. While there were and still are two chambers in parliament, the second chamber (Eerste Kamer) is clearly less important when it comes to questions of government inauguration or survival. Government formation is the remit of the party leaders in the lower house and the question of confidence is (almost) never made explicit in the Senate. Moreover, the government until recently usually commanded a majority in both chambers. Still, when the Rutte II government did not command a Senate majority, it believed that the Senate would vote for bills based on their merits. That did not quite work out as expected: opposition parties in the Senate demanded concessions when supporting government bills, which lead to a practice of ensuring opposition support for major bills. Thus, while the Dutch Senate does not play a big role in government formation, its veto power on bills results in the need for government parties to take the party political composition of the Senate into account. This worked out relatively well during the Rutte II government, which was the first to complete its full term since 1998. While majority coalitions have been the norm in Dutch politics, until 2012 there was no requirement that this majority support be demonstrated through an investiture vote. This has recently changed, however, when parliament decided it wanted to coordinate government formation itself instead of this process being led by the head of state (the king). In implementing this change, the lower house (Tweede Kamer) included in its Standing Orders the provision that (in)formateurs be appointed by the lower house of parliament. This can be regarded as a form of proactive investiture (Rasch et al. 2015). One should, however, note that the entire basis for the procedure is the Tweede Kamer’s own Standing Orders. There is no constitutional or legal requirement that the formateur ought to be appointed by parliament. In practice, the appointment of (in)formateurs has been supported by a broad coalition of parties. Thus while strictly a positive investment requirement, it seems not to have led to a much more contested or politicized inauguration process than in the past. After all, the main reason for the change in procedure was to take this power away from the head of state, not to create a positive investiture rule. Still, the effects of this procedural change have yet to be grasped in full. An even more recent, and temporary, institutional change was the introduction of an advisory corrective referendum in 2015 (Jacobs et al. 2016). While this institution does not directly affect government formation, it might impact on offer detailed contemporary accounts of these cabinet formation processes. Therefore, we have not conducted our own interviews for this chapter.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

450

  

how governments function, their stability, and eventually the type of government formed. The 2015 legislation opened up the possibility to request a non-binding referendum on a bill that has passed parliament. This requires 10,000 signatures for the initial request and another 300,000 signatures (in a six-week period) once this initial threshold has been met. Since the availability of this instrument, two referendums have been held: one on the EU Association Treaty with Ukraine (2016) and the Law on Intelligence and Security Services (2018). Both referendums led to a defeat of the government’s legislation. The need for the government to deal with negative referendum outcomes might have impacted future governance. Every successful bill could be subject to a referendum and once it has been held the result would at least have been ‘politically binding’. Due to the unease with particularly the 2016 referendum, the Rutte III coalition parties abolished the advisory corrective referendum in 2018. The government itself is characterized by a relative equality among ministers. The prime minister is characterized as primus inter pares (the first under equals). The prime minister’s position was strengthened somewhat in 2007 when a change in the Standing Order of the Council of Ministers was implemented, which (somewhat) strengthens the agenda-setting power of the prime minister. Otherwise, the formal position of the prime minister has hardly changed since 1945. Fiers and Krouwel (2005) have argued that his authority has increased particularly in terms of policy coordination. As party leader of (usually) the largest party, the prime minister has gained a stronger position within his own party but not towards ministers from other parties (Andeweg and Irwin 2014: 160). Dutch coalitions are a balancing act of the parties involved; if a serious political or policy conflict arises, all parties have to be involved in finding some sort of compromise.

The party system and the actors Party system change Before 1967, Dutch elections were essentially censuses. Voting behaviour followed the patterns of pillarization along religious and class lines (Andeweg and Irwin 2014). Only few seats changed party hands between elections: after the 1948 elections only 4 out of 100 parliamentary seats were occupied by a different party than before the voters had their say.² While volatility did increase somewhat in the 1970s and 1980s to around 10 per cent of parliamentary seats changing hands, the watershed election was in 1994 when 22.7 per cent of parliamentary seats changed party hands. Electoral volatility has remained high ever since, with a ² As the electoral system is very proportional, electoral volatility in terms of votes essentially follows the same pattern. We focus on seats here, as coalition government is our main objective.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

451

minimum of 15.3 per cent of seats changing hands. The most volatile election was in 2002 when 30.7 per cent of the parliamentary seats was won by a different party. These levels are not only high compared to the 1950s and 1960s but also in international comparative terms (Mair 2008). The increased levels of volatility go hand in hand with an increase of the effective number of parliamentary parties (EPP). The pattern is, however, not identical due to the position of the Christian democrats. Before 1977 there were three separate Christian democratic parties, which resulted in a relatively high EPP of about 4.5 in the 1950s and 1960s. This increased to around 6.4 in the 1970s because of the rise of several new parties. After the 1977 merger of the Christian democrats, the EPP declined substantially to 3.7. It remained low throughout the 1980s. The volatile 1994 election increased the EPP to 5.4. Ever since, the level has remained between about 4.7 and a high of 8.1 in the 2017 elections. Another source of fragmentation has been the relatively high number of splitoff parties that have formed in parliament during a parliamentary term. In the 2012 parliament the number of parliamentary parties increased from 11 in 2012 to 17 just before the 2017 elections, due to members of parliament (MPs) leaving their parliamentary party. This resulted in the Rutte II government losing its parliamentary majority at the end of its term. The increase of the number of parties mirrors the decline of the ‘big three’: Christian Democratic Appeal (CDA), Labour Party (PvdA), and the People’s Party for Freedom and Democracy (VVD). In the 1986 election, these three parties won 133 out of 150 seats (89 per cent), which had declined to only 82 seats (55 per cent) in 2010. New parties have emerged right, left, and centre on the political spectrum. The very low electoral threshold of one seat (0.67 per cent) means that it is not too difficult for new parties to gain parliamentary representation and since 2002 seven parties have done so: the populist parties Liveable Netherlands (LN), List Pim Fortuyn (LPF), Party for Freedom (PVV), and Forum for Democracy (FvD) as well as the special interest parties the Party for the Animals (PvdD), 50PLUS (elderly), and Denk (migrants). While LN and LPF turned out to be short-lived, the others look likely to be around for at least some time. These completely new parties are not the only ones to challenge the ‘big three’. On the left, the Socialist Party (SP) and Green Left (GL), in the centre Democrats 66 (D66) and the Christian Union (CU), and on the right the Reformed Political Party (SGP) have been alternatives to the incumbent parties. D66 and SP in particular have done so successfully at times, winning more than 20 seats at least once since 1994. The electoral success of new parties is related to the restructuring of the political competition. Pellikaan et al. (2003) argue that the 2002 elections brought about a change in the second dimension of political conflict. For many years, Dutch politics had arguably been structured by a socio-economic left–right dimension as well as a religious dimension. The latter was replaced by a cultural dimension in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

452

  

2002, according to Pellikaan et al. (2003). The cultural dimension relates to issues of immigration and the integration of migrants. Initially the LPF was the main owner of this new issue, a position that was later taken over by the PVV. The exact meaning of the second dimension can be discussed. Whereas Pellikaan et al. (2003) call it a cultural dimension, Kriesi et al. (2006) talk about the winners and losers of globalization. Parties like LPF and PVV represent those who do not stand to gain from open borders and supranational institutions and therefore generally oppose immigration and European integration. It should be stressed that the cultural dimension deviates somewhat from the GAL–TAN dimension that is often used in the literature. Parties like the PVV are by far the most extreme on the cultural dimension, but they are more moderate in terms of GAL–TAN because of their support for social liberal values, such as gay marriage (and vice versa for the Orthodox Protestant SGP party).³ Issues relating to the cultural dimension have remained relatively important since 2002. The LPF disappeared quickly as a result of internal conflict after its leader was assassinated just before the 2002 elections. Its position in the electoral landscape was filled by the PVV, which takes a somewhat ambiguous position on the economic left–right dimension but unambiguously opposes immigration and European integration. The party has become more radical over time in this respect. Initially Wilders argued that he wanted to combat the ‘excesses of Islam’; more recently the party opposes Islam altogether because it regards it as an extremist political ideology, not a religion. Whereas in 2006 the PVV argued that European cooperation should be ‘mainly economic’, it currently advocates a Dutch withdrawal from the Union. More recently, the Forum for Democracy (FvD) has presented itself as a new radical right-wing populist challenger, gaining two seats in the 2017 elections. Partly in response to these right-wing nativist parties, migrants have organized themselves into a new political party as well (Denk), which entered parliament for the first time in 2017 with three seats. The entry of new political parties was not limited to the cultural dimension. In 2006, the Party for the Animals (PvdD) entered parliament for the first time and has seen continued representation since. The 50PLUS party, which is aimed mostly towards elderly voters, won its first parliamentary seats in 2012. Both parties managed to gain seats in the elections of 2017, signalling that their success is not short term. Within the volatile and fragmented Dutch political landscape, these ‘single issue’ parties seem to be able to attract a small but sufficiently large electorate. Table 13.1a introduces the resulting historical cabinet record. It shows that after two cabinets led by a PvdA prime minister (1994–2002), the following eight years witnessed five CDA-led cabinets. First with the new right-wing LPF and VVD, ³ The correlation between Dutch party positions on GAL–TAN and immigration policy is 0.72 in the CHES 2010 data, while social lifestyle and GAL–TAN correlate more strongly (r = 0.96).

1946-07-03 1946-05-17 KVP, PvdA 1948-08-07 1948-07-07 PvdA, KVP, CHU, VVD 1951-03-15 PvdA, KVP, CHU, VVD 1952-09-02 1952-06-25 PvdA, KVP, CHU, ARP 1956-10-13 1956-06-13 PvdA, KVP, CHU, ARP 1958-12-22 KVP, CHU, ARP 1959-05-19 1959-03-12 KVP, CHU, ARP, VVD 1963-07-24 1963-05-15 KVP, CHU, ARP, VVD 1965-04-14 KVP, PvdA, ARP 1966-11-22 ARP, KVP 1967-04-05 1967-02-15 KVP, ARP, CHU, VVD 1971-07-06 1971-04-28 ARP, KVP, CHU, VVD, DS70 1972-08-09 ARP, KVP, CHU, VVD 1973-05-11 1972-11-29 PvdA, PPR, D66, KVP, ARP 1977-12-19 1977-05-25 CDA, VVD 1981-09-11 1981-05-26 CDA, PvdA, D66

Beel I Drees I

Drees II

Drees III

Drees IV

Beel IIa

De Quay

Marijnen

Cals

Zijlstraa De Jong

Biesheuvel I

Biesheuvel IIa

Den Uyl

Van Agt I Van Agt II

3

4

5

6

7

8

9

10 11

12

13

14

15 16

51.3 72.7

64.7

sur mwc sur

49.3

54.7

mwc min

42 57.3

70.7

61.3

62.7

51.3

84.7

81

76

61 76

min mwc

sur

sur

sur

mwc

sur

sur

sur

mwc sur

150 150

150

150

150

150 150

150

150

150

150

150

100

100

100 100

11 10

14

14

14

10 11

10

10

8

7

7

8

8

7 8

3.7 4.29

6.42

6.4

6.4

4.51 5.71

4.51

4.51

4.15

4.07

4.07

4.65

4.68

4.47 4.68

Party Type Cabinet Number of Number of ENP, parties in parliament composition of strength seats in of cabinet cabinet in seats parliament parliament (%)

1 2

Election date

Year in

Cabinet Cabinet number

Table 13.1a Dutch cabinets since 1946

CDA CDA

CDA

KVP

KVP

KVP KVP

KVP

KVP

KVP

KVP

KVP

KVP

KVP

ARP KVP

Median party in first policy dimension

Continued

Formal support parties

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1982-05-29 1982-11-04 1986-07-14 1989-11-07 1994-08-22

1982-09-08 1986-05-21 1989-09-06 1994-05-03

Election date

34.7 52.7 50.7

31.3

min min mwc mwc

53.3

47.3

52

62

64.7

43.3 54 54 68.7 61.3

mwc

min

mwc

mwc

sur

min mwc mwc mwc mwc

150 150 150

150

150

150

150

150

150

150 150 150 150 150

10 11 13

10

10

9

9

10

9

10 12 9 9 12

6.74 5.69 8.12

5.54

5.54

4.74

4.74

5.79

4.81

4.29 4.01 3.49 3.75 5.42

Party Type Cabinet Number of Number of ENP, parties in parliament composition of strength seats in of cabinet cabinet in seats parliament parliament (%)

CDA, D66 CDA, VVD CDA, VVD CDA, PvdA PvdA, D66, VVD Kok II 1998-08-03 1998-05-06 PvdA, D66, VVD Balkenende I 2002-07-22 2002-05-15 CDA, LPF, VVD Balkenende II 2003-05-27 2003-01-22 CDA, VVD, D66 Balkenende 2006-07-07 CDA, VVD IIIa Balkenende 2007-02-22 2006-11-22 CDA, PvdA, IV CU 2010-02-23 CDA, CU Balkenende Va Rutte I 2010-10-14 2010-06-09 VVD, CDA Rutte II 2012-11-05 2012-09-12 VVD, PvdA Rutte III 2017-10-26 2017-03-15 VVD, CDA, D66, CU

Van Agt IIIa Lubbers I Lubbers II Lubbers III Kok I

Year in

PVV PVV D66

PVV

D66

CDA

CDA

CDA

D66, CDA

CDA CDA CDA CDA CDA

Median party in first policy dimension

PVV

Formal support parties

Notes: For a list of parties, consult the chapter appendix. a = Limited policy remit Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1946–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015). The first policy dimension is economic left–right. The number of parties only includes parties that have held more than two seats in parliament when a cabinet is formed during the period of observation. Cabinet types: min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition Table legends and variables are further defined in the Appendix, this volume.

28 29 30

27

26

25

24

23

22

17 18 19 20 21

Cabinet Cabinet number

Table 13.1a Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

455

later with VVD and D66 (followed by a short CDA–VVD caretaker) and finally with PvdA and CU (which ended in a CDA–CU caretaker situation). From 2010 onwards, the VVD won most seats in national elections, which resulted in a number of Rutte cabinets. The first was a minority cabinet with CDA, formally supported by the radical right PVV, followed by a majority cabinet with PvdA, and finally, since 2017, a four-party coalition of VVD, CDA, D66, and CU. Despite the increasing importance of cultural issues in Dutch politics, socioeconomic issues have remained of central importance to electoral and parliamentary politics. Before the financial crisis of 2008, CDA and PvdA failed to overcome their socio-economic differences to form a government in 2003 and succeeded only after a lengthy formation process in 2006–2007. During and after the economic crisis that started in 2008, socio-economic issues took centre stage in electoral and parliamentary politics. An attempt to form a ‘purple-plus’ coalition of VVD, PvdA, D66, and Green Left failed in 2010 for similar reasons. There is an argument to be made that the left–right dimension is changing in nature. While the classical ‘state versus market’ distinction remains one aspect of the socioeconomic divide, the willingness to reform seems to be a new dimension related to these questions, especially since the financial crisis (Louwerse 2010; Otjes 2015). On the left, we can distinguish between those parties (and voters) that generally favour reforms, such as increasing the pension age and labour market reforms (PvdA, Green Left), and those that do not (SP). Similarly, on the right there are parties that are more willing to reform (VVD, CDA) than others (PVV). The ‘reform’ dimension to socio-economic policy is correlated with pro- and antiEuropean policy positions. As a result of this partial redefinition of socioeconomic divide, the median position on the socio-economic divide is now occupied by the PVV, which takes left-wing positions on some issues (notably health care) and right-wing positions on other issues (taxation).

Electoral alliances and pre-electoral coalitions Despite the large degree of fragmentation, coalition formation before elections through electoral alliances and pre-electoral coalitions has not happened on a regular scale. Only in 1971 and 1972 did left-wing parties (PvdA, D66, PPR) form a pre-electoral coalition with a common manifesto and a joint candidate for prime minister. Table 13.1b records the event. In 1972 this coalition did form the core of the new government, but as it failed to win an absolute majority it resulted in very complicated coalition negotiations and, eventually, a coalition of the left with two Christian democratic parties. More recently, there has been discussion about cooperation (particularly) on the left, but this never materialized. There have been so-called ‘connected lists’ (lijstverbindingen) between parties, particularly CU and SGP, as well as between left-wing

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

456

  

Table 13.1b Electoral alliances and pre-electoral coalitions in the Netherlands, 1946–2018 Election date

Constituent parties

Type

Types of pre-electoral commitment

1971-04-28 1972-11-29

PvdA, D66, PPR PvdA, D66, PPR

PEC PEC

Joint press conference, Other Joint press conference, Other

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other.

parties. This now-abolished mechanism does not amount to an electoral alliance or pre-electoral coalition, however, and mainly serves to increase the chances of winning a ‘remainder seat’ under the D’Hondt electoral formula.

Government formation The bargaining process Coalition formation in the Netherlands is a slow process. In comparative terms, Dutch formations are among the longest in Europe, particularly for post-election cabinets (De Winter and Dumont 2008). After elections, parties take their time, to find the right combination of parties and subsequently agree on a government programme. The procedure is only very lightly regulated (the Standing Orders of the lower house of parliament include some regulations) and therefore changes somewhat from year to year. In the most recent years, however, the general pattern is as follows. On the day after the election, the leaders of the newly elected parties are invited by the Speaker of the lower house of parliament for an informal consultation. Before 2012, the head of state (king) consulted with party leaders and advisors. Usually this results in the appointment of a scout (verkenner) or informateur. Customarily, this person belongs to the largest party but is somewhat removed from party politics to gain the trust of other parties. (S)he is tasked with consulting with individual party leaders to explore which parties might be part of a new government coalition. The scouts usually advise on the coalition composition that should be explored first. Their report is debated by the lower house, which then appoints one or more informateurs to lead these negotiations. When negotiations fail a new informateur is appointed to re-explore options or to start negotiations on a new cabinet. If successful, these negotiations would result in a coalition agreement. The lower house of parliament would then appoint a formateur, usually the prospective prime minister, who is tasked with staffing the cabinet. The division of work between the informateur and formateur that has

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

457

arisen is thus roughly between gathering information and leading negotiations on the coalition agreement, while the formateur speaks with prospective (junior) ministers. Finally, the head of state appoints the new ministers. Table 13.2 presents the empirical record since the late 1980s. There is significant variation in the duration of the bargaining process. Since the Cals cabinet (1966–1967) there has been an informal rule that government resignations have to be followed by new elections. Therefore, bargaining on governments that are not formed immediately after an election only concerns ‘caretaker governments’ that can be formed relatively quickly. When the PvdA withdrew from the fourth Balkenende cabinet, the formation of Balkenende V (CDA and CU) was even instantaneous. There was no official process of government formation: the remaining ministers just continued in office and a few junior ministers were promoted. As for regular cabinets formed immediately after elections, the procedure takes much more time. In the last 40 years, even when there was only one (successful) bargaining round, bargaining has taken between 53 and 90 days. When there were multiple (failing) bargaining rounds, the whole procedure has taken more than 200 days in two instances (Van Agt I and Rutte III). The average total bargaining duration for non-caretaker cabinets formed since the Second World War was 87 days; this increases to 104 if we limit ourselves to the last 40 years. Most of this time is taken up by negotiations on the coalition agreement. The exploration phase usually does not take more than one to two weeks and the formation phase also is concluded within a matter of weeks at most. Reflecting the importance (and length) of the coalition agreements, parties take their time to iron out many details. While coalition agreements are not set in stone (Timmermans 2006), inclusion of electoral commitments in the agreement is one of the best ways to ensure their fulfilment (Thomson 1999: 206). In order to speed up the detailed negotiations, most recent coalition formation processes have relied on ‘side tables’ in which specific areas of policy were discussed by experts from each parliamentary party group. The leaders (and recently, the secondants they bring along) at the main table refer specific or technical issues to their specialists, who can iron out details and compromises that are then finalized at the main table. An additional advantage is that the parliamentary party group is more involved in the negotiation, which the leaders hope increases party unity down the line. When multiple bargaining rounds are necessary, there are often changes of informateur and/or the parties involved in the negotiations. The formation of the Rutte I cabinet in 2010, for example, saw no less than seven failed bargaining rounds. The last four bargaining rounds, including the successful one, all concerned the same parties (VVD, PVV, and CDA). First, informateur Lubbers explored the option of a minority cabinet with PVV support, then informateur Opstelten started negotiations on a government programme. These talks collapsed after internal turmoil within the CDA. This was resolved a few days later, but the

Year in

1989

1994

1998 2002 2003

2006 2007 2010 2010

2012 2017

Cabinet

Lubbers III

Kok I

Kok II Balkenende I Balkenende II

Balkenende III Balkenende IV Balkenende V Rutte I

Rutte II Rutte III

0 2

0 0 0 7

0 0 1

2

1

Number of inconclusive bargaining rounds

CDA, PvdA (1) CDA, PvdA, D66 PvdA, D66, VVD (1) PvdA, VVD, D66 (2) PvdA, CDA, VVD, D66 PvdA, D66, VVD CDA, LPF, VVD CDA, VVD, D66 (1) CDA, PvdA CDA, VVD CDA, PvdA, CU CDA, CU VVD, CDA (1) VVD, PVV, CDA (2) VVD, PvdA, CDA, D66, GL (3) VVD, PvdA, D66, GL (4) VVD, PvdA, D66, GL (5) VVD, CDA (6) VVD, CDA (7) VVD, CDA VVD, PvdA VVD, CDA, D66, CU (1) VVD, CDA, D66, GL (2) VVD, CDA, D66, GL

Parties involved in the previous bargaining rounds

Table 13.2 Cabinet formation in the Netherlands, 1989–2018

11 17 14 32 7 54 121 49 14

88 67 43 79 7 90 1 32 6 9

48 14 48 52 9

Bargaining duration of individual formation attempt (in days)

196 226

8 92 0 127

90 68 223

111

189

Number of days required in government formation

54 213

7 90 1 125

88 67 124

109

61

Total bargaining duration

118 111

Pro

-

Abstention

32 39

Contra

Result of investiture vote

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

459

queen appointed a third informateur (notably former vice-president of the Council of State Tjeenk Willink, a member of the PvdA⁴) to assess the constitutional situation first (‘a constitutional penalty round’), before informateur Opstelten was allowed to finalize negotiations and hand over to formateur Rutte. Therefore, in the more complex bargaining situations there are usually more (sets of) informateurs than there are viable bargaining options. There is a lot of backward and forward before everything is finalized.

The composition and size of cabinets The first section of this chapter alludes to the norm of majority coalition governance in the Netherlands. All governments since 1945 have been coalitions and, save for one recent exception, all (non-caretaker) cabinets since 1945 consisted of ministers from parties that together commanded a majority in the lower house of parliament. Moreover, in the years of ‘pillarization’ (before 1967, roughly speaking) many government coalitions were surplus majority governments, which contained parties that were not strictly necessary in order to obtain a parliamentary majority. This conforms to the patterns of consensus democracy: broad power-sharing coalitions. Since the 1970s, however, the number of surplus majority cabinets has declined considerably and most governments are minimal winning coalitions. This change can be partly attributed to the merger of three Christian democratic parties into the CDA in 1980. Beforehand, these three parties (increasingly) worked together, even though regularly one of these parties was outside of the government coalition. Another reason to find an increase in minimal wining coalitions is the growing fragmentation of the party landscape. This has made it numerically impossible to form two-party (or in some cases even three-party) coalitions. Because four- or five-party coalitions are considered overly complex, especially after the CDA merger, the only realistic option is often to form a minimal winning coalition. In recent years the only exception is the Kok II cabinet, which was a continuation of Kok I in terms of the three parties involved. While D66 was not needed anymore to win a majority, because of its ideological position in between the other two parties involved (PvdA and VVD) as well as the feeling that ‘D66 belonged in the coalition’, the party was included in Kok II anyway. A third reason for an increase in minimal winning coalitions is the electoral cost of ruling, which seems to have increased in the last decades (Müller and Louwerse 2020). As a result parties, especially smaller, potential ‘surplus’ ⁴ The Council of State is one of the High Councils of State and has two functions. One is an advisory function, carried out by the Advisory Division, while the other is a judicial function as the highest administrative court, carried out by the Administrative Jurisdiction Division. The two divisions are separate (Council of State n.d.). Before 2012, as one of the advisors of the head of state, the vicepresident of the Council of State would present his views on the government formation to the queen.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

460

  

parties, will think twice before entering a coalition in which their votes are not required. If their votes are not necessary, this limits their bargaining position and therefore potential policy payoffs that may offset future electoral losses. An alternative to the minimal winning coalition that has been considered more seriously in recent years is the option of a minority government (Strøm 1990). In 2010, VVD and CDA agreed to support and provide ministers for the Rutte I cabinet. These parties jointly controlled only 52 out of 150 parliamentary seats. The radical right-wing populist party PVV agreed to provide support from parliament; it committed to parts of the coalition agreement as well as ‘confidence and supply’. Earlier attempts to include the PVV in a regular majority coalition had been met with scepticism from the other two parties involved, but this arrangement, which had worked successfully in countries like Denmark, was acceptable to all three parties. While this government may thus be considered a minority government, it operated as a majority coalition, which was also apparent in terms of parliamentary behaviour of government and opposition parties (Otjes and Louwerse 2014). Despite its nominal status as a minimal winning coalition, the subsequent Rutte II cabinet came closer to substantive minority cabinet status due to its lack of majority support in the Senate. While it would have been unlikely for the Senate to support a motion of no confidence, opposition parties were reluctant to support major government legislation. Eventually, the government chose to strike (ad hoc) deals with opposition parties. In exchange for policy concessions on the bill concerned, opposition parties pledged to support the bill in parliament. The deals were made with the party leaders in the lower house, but the real target was obtaining support in the Senate. Therefore, in many ways Rutte II behaved as a substantive minority cabinet by looking for opposition support on an ad hoc basis. In terms of the party composition of government the traditional parties continue to dominate the field: CDA, PvdA, VVD, and D66. At least two out of these four have been a part of every (non-caretaker) coalition since the CDA was formed. Two new parties have joined coalitions since 1990: LPF and the CU. After coming second in the 2002 elections the LPF was included in the formation of the first Balkenende cabinet. Due to internal conflicts, however, it also contributed greatly to its (very) early demise and the party soon disintegrated. The CU seems to have been more successful and has participated in two cabinets (Balkenende IV and Rutte III). As a moderate party on most major issues, it can find common ground with parties on the left and right of the political spectrum.

The allocation of ministerial portfolios The allocation of ministerial portfolios occurs in the final stages of the government formation process when the formateur is in charge. The end result with regard to five cross-nationally important ministries is presented in Table 13.3.

Year in

1989 1994

1998

2002

2003

2006 2007

2010 2010 2012 2017

Cabinet

Lubbers III Kok I

Kok II

Balkenende I

Balkenende II

Balkenende III Balkenende IV

Balkenende V Rutte I Rutte II Rutte III

7 CDA, 7 PvdA 5 PvdA, 5 VVD, 4 D66 6 PvdA, 6 VVD, 3 D66 6 CDA, 4 LPF, 4 VVD 8 CDA, 6 VVD, 2 D66 9 CDA, 7 VVD 8 CDA, 6 PvdA, 2 CU 9 CDA, 3 CU 6 CDA, 6 VVD 7 VVD, 6 PvdA 6 VVD, 4 CDA, 4 D66, 2 CU

Number of ministers per party (in descending order)

12 12 13 16

16 16

16

14

15

14 14

Total number of ministers

2 CDA 4 CDA, 2 VVD 3 PvdA, 2 VVD 2 CDA, 2 D66, 2 VVD, 1 CU

3 CDA, 2 PvdA 4 PvdA, 3 D66, 3 VVD 6 PvdA, 4 VVD, 2 D66 5 LPF, 4 CDA, 4 VVD 5 CDA, 4 VVD, 1 D66 3 CDA, 3 VVD 5 PvdA, 4 CDA

Number of watchdog junior ministers per party

Table 13.3 Distribution of cabinet ministerships in Dutch coalitions, 1989–2018

13 11 11 12

13 13

13

13

13

13 13

Number of ministries

CDA VVD VVD VVD

CDA CDA

CDA

CDA

PvdA

CDA PvdA

1 Prime minister

CDA CDA PvdA CDA

VVD PvdA

VVD

VVD

VVD

PvdA VVD

2 Finance

CDA VVD PvdA VVD

CDA CDA

CDA

CDA

VVD

CDA D66

3 Foreign affairs

CDA VVD PvdA D66

CDA CDA

CDA

CDA

PvdA

CDA PvdA

4 Social affairs

CDA VVD VVD CDA

CDA CDA

CDA

CDA

VVD

CDA D66

5 Justice

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

462

  

Informally, before the end stage, parties will have already discussed the total number of ministers, including both departmental and non-departmental ministers, as well as the division of portfolios. This will, however, only be finalized when parties agree on everything and are convinced that they have suitable candidates for the posts allocated to them. The number of ministers for each party is roughly proportional to seat share, with the division of portfolios approximated relatively well by a sequential logic of portfolio allocation (Ecker et al. 2015). Some posts are valued relatively highly by all parties, especially the finance ministry, which since 1990 has almost always been held by the second largest party. Party issue saliency is also important for the allocation of ministerial posts, although arguably less so than in other countries due to the existence of elaborate coalition agreements (Bäck 2011). Perhaps the most prominent example of catering to party ownership of issues is the creation of so-called ‘non-departmental’ ministers. For example, the antiimmigrant LPF got a non-departmental minister on immigration and integration (2002), the CU was given a minister for ‘youth and family’ (2007), and the propolitical reform D66 party obtained a ministerial post on ‘political-administrative reform’ (2003). The use of this instrument has increased in recent years; expanding beyond the usual non-departmental minister for development cooperation, the Rutte III government has no less than 4 non-departmental ministers out of a total of 16 cabinet ministers.

Coalition agreements Since the early 1960s, coalition agreements are almost always made when a new government is formed after parliamentary elections. There is no habit of preannouncing or committing to any policy intentions between parties prior to the start of the government formation process. Mostly, parties keep their positions on major topics quite open; party leaders or spokespersons usually reveal very little until the negotiations have already advanced and it is possible for the new partners to share common viewpoints. Table 13.4 shows that coalition agreements vary more in size, the number of words, than in the composition of these documents when considering the proportions of policy relative to other kinds of statements, such as on policy specific procedures, ministerial competencies, and offices. The longest agreement to date is the one produced by a four-party coalition in October 2017, with a length of some 40,500 words. There is no clear pattern when relating the party composition of coalitions to the size of agreements. Governments including CDA and VVD (1982, 1986, 2002, 2003, 2010, 2017) and governments with the PvdA on board (1981, 1989, 1994, 1998, 2007, 2012) produced agreements of similar length.

Year in

1963 1965 1967 1971 1977 1981 1982 1986 1989 1994 1998 2002 2003 2007 2010 2012 2017

Coalition

Marijnen Cals De Jong Biesheuvel I Van Agt I Van Agt II Lubbers I Lubbers II Lubbers III Kok I Kok II Balkenende I Balkenende II Balkenende IV Rutte I Rutte II Rutte III

3,350 3,600 3,100 6,100 7,900 15,900 20,300 15,500 28,450 16,250 36,000 16,500 8,600 16,000 16,600 27,000 40,500

Size 16 5 7 2 0 1 1 0 0 0 0 0 0 0 0 0 0

General rules (in %) 29 0 0 5 2 1 2 2 2 1 1 1 2 2 2 1 1

Policy-specific procedural rules (in %) 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of offices (in %)

Table 13.4 Size and content of coalition agreements in the Netherlands, 1946–2018

0 0 0 0 0 4 9 9 1 2 1 1 1 1 1 1 1

Distribution of competences (in %)

52 95 93 93 98 94 88 89 97 97 98 98 97 97 97 98 98

Policies (in %) OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

464

  

With size, the agenda capacity for expressed intentions also varies from one cabinet to the next, so coalition agreements can have a narrow or wider scope. While the electoral programmes of participating parties are a basis, there is no straightforward mechanism or rule according to which statements in coalition agreements mirror them. Sometimes pledges are visible, at other times a deal is included that cannot be traced back to any electoral programme. In this sense, government formation is a venue not only for prospective coalition parties to set the agenda but also for all kinds of interest organizations and groups to try and influence what is included in the coalition agreement. Compared to other types of policy agenda, coalition agreements made in the ‘institutionalized extrainstitutional arena’ of government formation (Peterson et al. 1983) are unbounded by formal limits. One practice peculiar to making coalition agreements in the Netherlands is that since the 1980s the policy intentions contained in it are assessed by the national Bureau for Economic Policy Analysis (Centraal Planbureau, CPB), the public organization that also assesses the electoral programmes of political parties. These assessments are meant to rationalize intentions on economic effects and in this way set boundaries to the plans that new governments can put on the agenda. In reality, parties involved in this process make their expenditure choices to a large extent on the basis of political rationality and estimated electoral effect (Bolhuis 2018). Coalition agreements may contain many items, often more than what is central to the formation of governments, and also more than what the aforementioned Bureau of Economic Policy Analysis considers for economic effect assessment. Statements in coalition agreements can be symbolic or substantive, opaque or detailed, draw up a shiny future or be written in terms of realizable output. Table 13.4 shows that the proportions of components of coalition agreements are quite stable in the past 30 years. The text is almost entirely about policy statements, part of which are clear-cut intentions and other parts indications of direction for the course of action. But agreements never are completely unambiguous; sometimes formulas are included that are meant to reduce the inflammation risk of controversial issues, but the subsequent interpretation of them gives rise to confusion (Timmermans 1996). With the size the policy scope of the agreement also varies. While main fields of public policy such as the economy, social security, health, education, and international affairs are always present, coalitions may include different ranges of policy topics. A systematic content analysis of coalition agreements done within the Comparative Agendas Project shows that these documents vary in scope between some 60 and 120 different policy topics (Timmermans and Breeman 2017).⁵ This ⁵ The maximum scope is some 240 policy topics, clustered in 20 main topic categories. This is a standardized international coding scheme used in the Comparative Agendas Project. See www. comparativeagendas.net where data from the Netherlands Policy Agendas Project are available.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

465

scope correlates with the length of agreements. Apart from habitual elements that make coalition agreements look like quasi-official government programmes, including some mission statement (‘building bridges’ in 2012, ‘confidence in the future’ in 2017), it is the scope that expresses both the ambitions of the coalition and the alleged functions of the coalition agreement, to set the agenda and keep manifest or potential controversy under control. The long-term trend in the policy scope is curvilinear: after a rise in the 1960s to the 1990s, it went down in the 2000s and started to rise again in the 2010s. This trend deviates from the development of governmental policy agendas more generally, which after a ‘great issue expansion’ until the 1980s started to narrow down (Baumgartner and Jones 2015; Timmermans and Breeman 2017). Within the scope of the agenda, not all issues are equal. The top five priority themes usually take some 50 per cent of the total agenda space in the coalition agreement. These are typically ‘guns and butter’ topics (Jennings et al. 2011): the economy, international affairs and domestic security, the structure and organization of government, and domains of ongoing policy reform, such as labour market, social security, and health care. A final observation about the content of coalition agreements is that they seem to have become more open to signals from other policy agendas in the parliamentary system. While agreements emerge from a relatively closed environment meant to cement the coalition internally, the parties taking office show more sensitivity to the parliamentary arena and the public arena. Reasons for this are that majority support is more conditional and even uncertain to obtain in both legislative chambers and that electoral volatility requires parties to constantly monitor the public environment where attention to issues may cascade to problems that must be addressed even if not initially acknowledged. Oral questions in the Second Chamber, for example, long followed the major themes in the coalition agreement as opposition parties tried to find vulnerable parts of the coalition. But since the turn of the century, oral questions have also become a predictor of what themes rise and fall in prominence in the next coalition agreement (Timmermans and Breeman 2010).

Coalition governance The political transaction costs that coalition parties make in producing joint policy agreements are expected to facilitate consensus and stability. The challenge for coalition governments is to guard the priorities set in the joint policy agreement and also react to pressures for updating the policy agenda as conditions change and focus events may make a previously lower key topic into a matter of urgency. The more turbulent public environment requires constant monitoring by a coalition government in order not to put its legitimacy at stake. But with two to four parties in office, this is a difficult balancing act.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

466

  

In this sense, cabinet governance in the Netherlands closely follows the coalition compromise model (see Chapter 2, this volume). Coalition agreements are important documents that individual ministers have to adhere to. While deviations are possible (and not infrequent), these always have to be coordinated with all government parties. One (very) partial exception, at least at the coalition formation stage, was the Rutte II cabinet. During the government formation, the parties explicitly chose to ‘logroll’ on some issues rather than to compromise. Still, the agreed position was included in the coalition agreement, but it potentially provided the party that initiated the policy somewhat more leeway in terms of its implementation. In practice, coordination between the government parties was still the norm. While coalition agreements are never exhaustive and events or developments can require modification and policy adaptation, the compromise model in the Netherlands signifies a wide practice. This becomes more pronounced as the number of coalition partners increases due to declining vote shares of parties that are most ‘coalitionable’. Governments cannot table an infinite number of bills in a legislative year, and even if they might wish to do this for political reasons, they face constraints of administrative preparation and formal legislative calendars. There is no formal maximum in legislative production in the Netherlands, but in practice the number of bills placed on the legislative agenda has a limit of some two hundred per year. With capacity limits, choices must be made. Cabinets also vary in their legislative ambition level. The Rutte II cabinet proposed some 30 per cent more bills in its first year than the following Rutte III cabinet in its first year (Korteweg 2018). The government formation table is a typical arena of exchange and compromise. But when a specific agreement has to be implemented it usually is disconnected from the other matters that were on the government formation table. It requires patience and loyalty to prevent one coalition partner from undermining the equilibrium of the agenda. For example, shortly after the Rutte II took office in 2012, a PvdA intention in the agreement to make health insurance premiums income-dependent was to be turned into a law by the VVD minister in charge of the health portfolio. But then her party rank and file rebelled, and the plan caused a major problem and put the coalition to an early test of survival. To reduce the risk of conflict, coalition agreements not only contain new ambitions and joint intentions to initiate new policy but often also mention explicitly whether policies are to continue and stay the same or even mention issues that are not to be addressed, as is indicated in Table 13.5. Maintenance of the status quo in policy is a much more often used explication of agreement to not initiate new policies than the joint declaration that an issue is supposed to be left aside. Policy continuity also is not necessarily an appeasement strategy—the coalition partners can simply prefer a policy programme to remain intact, and this is more likely in case the party holding the portfolio or the whole coalition was the initiator of it during the previous term in office. Explicit negative agenda

1946 N 1948 N

1951 N

1952 N

1956 N

1958 N

1959 N

1963 POST

1965 IE 1966 N 1967 POST

1971 POST 1972 N

1973 N

Drees II

Drees III

Drees IV

Beel II

De Quay

Marijnen

Cals Zijlstra De Jong

Biesheuvel I Biesheuvel II

Den Uyl

N/A

Y N/A

N N/A Y

Y

N/A

N/A

N/A

N/A

N/A

N/A N/A

Y

Y Y

N N Y

N

N

N

N

N

N

N N

IC IC IC

Parl

Parl

Parl

Parl

Parl

Parl

Parl Parl

Parl, IC, CoC IC

CoC

CoC CoC

Parl Parl Parl

Parl

Parl

Parl

Parl

Parl

Parl

Parl Parl

Most For most common serious conflicts

Parl, IC, CoC IC Parl, IC, CoC IC

Parl, IC Parl, IC Parl, IC

Parl

Parl

Parl

Parl

Parl

Parl

Parl Parl

All used

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Beel I Drees I

Coalition

Table 13.5 Coalition governance mechanisms in Dutch coalitions, 1946–2018

N (KVP) N (KVP, VVD, CHU) N (KVP, VVD, CHU) N (KVP, ARP, CHU) N (KVP, ARP, CHU) N (KVP, CHU) N (KVP, VVD, CHU, ARP) N (KVP, CHU) N (KVP) N (KVP) N (KVP, CHU, VVD, ARP) N (KVP) N (KVP, VVD, CHU) N (KVP, ARP, D66, PPR)

Personal union

Yes

Yes N/A

No N/A No

No

N/A

N/A

N/A

N/A

N/A

N/A N/A

Issues excluded from agenda

Most/No

Most/No Most/No

Most/No Most/No Most/No

Most/No

Most/No

Most/No

Most/No

Most/No

Most/No

Most/No Most/No

Y

Y Y

Y Y Y

Y

Y

Y

Y

Y

Y

Y Y

Comp.

Varied

Varied

Varied

Few

Few

Few

Few

Few

Few Few

Y

Y Y

Y Y Y

Y

Y

Y

Y

Y

Y

N N

Continued

N

N N

N N N

N

N

N

N

N

N

N N

Coalition Freedom of Policy Junior Nondiscipline appointment agreement ministers cabinet in (watchdogs) positions legislation/ other parl. behaviour

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2003 2006 2007 2010 2010 2012 2017

Balkenende II Balkenende III Balkenende IV Balkenende V Rutte I Rutte II Rutte III

Y N/A Y N/A Y Y Y

Y

Y Y N/A Y Y Y Y Y

Y Y Y Y Y Y Y

Y

Y Y Y Y Y Y Y Y

IC IC IC IC IC IC IC IC

Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC

IC IC IC IC IC IC CoC

Parl, IC, CoC IC

Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC Parl, IC, CoC

CoC CoC CoC CoC CoC CoC CoC

CoC

CoC CoC CoC CoC CoC CoC CoC CoC

For most Most common serious conflicts Y Y Y N (VVD) N (VVD) Y N (VVD) N (VVD, D66) N (LPF, VVD) N (VVD) N (VVD) Y Y Y N (PvdA) N (CDA, D66, CU)

Personal union

Yes N/A Yes N/A Yes Yes Yes

Yes

Yes Yes N/A Yes Yes Yes Yes Yes

Issues excluded from agenda

Most/No Most/No Most/No Most/No Most/No Most/No Most/No

Most/No

Most/No Most/No Most/No Most/No Most/No Most/No Most/No Most/No

Y Y Y Y Y Y Y

Y

Y Y Y Y Y Y Y Y

Comp. Comp. Comp.

Comp.

Comp.

Comp.

Comp. Comp. Comp. Comp. Comp.

Comp. Comp.

Y Y Y Y Y Y Y

Y

Y Y Y Y Y Y Y Y

N N N N N N N

N

N N N N N N N N

Coalition Freedom of Policy Junior Nondiscipline appointment agreement ministers cabinet in (watchdogs) positions legislation/ other parl. behaviour

Notes: Coalition agreement: POST = post-election; N = no coalition agreement Conflict management mechanisms: IC = Inner cabinet; CoC = Coalition committee; Parl = Parliamentary leaders Coalition discipline: Most = Coalition discipline is expected on all matters except those explicitly exempted; No = Coalition discipline not always expected Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement

POST N POST N POST POST POST

2002 POST

Balkenende I

POST POST N POST POST POST POST POST

1977 1981 1982 1982 1986 1989 1994 1998

All used

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Van Agt I Van Agt II Van Agt III Lubbers I Lubbers II Lubbers III Kok I Kok II

Coalition

Table 13.5 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

469

setting, mentioning issues that shall remain away from the governmental agenda, points to disagreement, and while this may be a procedural arrangement to avoid trouble, the fact that it also is a pointer to underlying disagreement makes coalitions cautious in listing them. For all these reasons, institutional and political, coalition governance during the term in office means that priorities and payoffs to the partners must be balanced. Carrying out the coalition agreement thus is anything but a routine mechanistic enterprise, and requires very close monitoring of signals from within the coalition and from the wider environment. Looking at successive years a coalition government is in office, we see the political agenda moving in which priorities are reset. Policy agenda change for coalition government need not be a survival risk, as long as such change happens in concert. Thus we see governments addressing different sets of political problems and policy intentions over the course of the term, a display of the way in which attention shifts to different parts of the joint policy agenda set at the beginning. When comparing the legislative agenda of the years in office to the coalition agreement negotiated at the beginning, most of the governments between 1963 and 2012 show some cyclic pattern in which they allocate their attention first to some clear priorities and then in the second and third year in their term in office move back to the wider range of topics included in the coalition agreement. The last year in office then again moves away from the broad agenda and shows concentration of some topics (Timmermans and Breeman 2014). This apparent sequencing of attention to policy issues suggests that coalition governments begin with the topics considered most important to move into the policy-making pipeline, then address more issues, and finally pay attention to issues that were left somewhat ignored or that play a part in timing towards the elections. This pattern in coalition governments in the Netherlands looks dissimilar from the model of the political business cycle, in which a single party in office strategically times its ‘harvest season’ for policy success towards the next elections. The ongoing process of agenda setting is thus not a case of automated government. Furthermore, to secure all this happens as peacefully and productively as possible, there are informal rules and structures for coalition governance.

The role of individual ministers in policy-making One aspect of this coalition governance is the space given to individual ministers. While individual ministers enjoy leeway in carrying out the tasks related to their portfolio and their role as party prominent in the cabinet, they are in no sense policy ‘dictators’ (Laver and Shepsle 1996). The principles of collegiality and collective responsibility also underline the importance of the rule of nonintervention into the business of other ministers. Cabinet ministers thus must

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

470

  

stay on their own turf, even though it is sometimes hard to determine where one jurisdiction ends and another begins. Many problems and matters of government policy do not behave according to formal boundaries drawn around a government department. When formal border lines are hard to draw and follow, experiential learning should help preventing that ministers compete openly over who is the first responsible for an issue to address it. Departmental reshuffles made in government formation may also help in this ministerial orientation (Sieberer et al. 2017). In November 2017 for example, the Rutte III government placed the first responsibility for climate policy in the hands of the minister of economics affairs—as was expressed by the new departmental name economic affairs and climate policy. In the cabinet in office between 2012 and 2017, climate policy was the first responsibility of the minister of infrastructure and environment. Increased media attention and the rapid rise of issues in the public arena make it imperative that ministers carefully balance claiming jurisdiction over an issue and avoidance of blame when the issue is believed to be ill-addressed. This exposure and risk of reputation damage may contribute to the ongoing practice of adherence to the coalition agreement. The prime ministerial position gained importance in representing the country in international arenas, but for domestic coalition governance to be successful, the prime minister must still pay systematic attention to securing the internal harmony and the credibility of the government. Prime minister Mark Rutte, who has run three successive cabinets since 2010, has a gavel when formally presiding meetings of the Council of Ministers, but he never uses it (Niemantsverdriet et al. 2016). The task of preserving internal peace becomes more pressing as government coalitions contain more and smaller parties and have less certainty of stable majority support in parliament. For this reason, the prime minister is also no exception to the rule that policy intentions included in the coalition agreement must be guarded, and in case plans need reconsideration or new issues intrude, the prime minister has special responsibility for establishing agreement about the policy response. This is mostly a matter of personal competence and coalition management skills, and much less of institutional privilege since Dutch prime ministers lack formal equipment for it.

Coalition governance in the executive arena The external pressures on governments imply that internal governance mechanisms are becoming ever more important. If persuasive communication skills of a prime minister is an advantage, this is not sufficient for managing the major coalition business and even less if matters become tense. The informal cabinet committees that since long are used to streamline and smooth cabinet decisionmaking are re-operationalized at the beginning of each new government. A variable factor in this is the number of vice-prime ministers, as each coalition

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

471

party normally provides one. Thus the Rutte III government taking office in November 2017 counted three vice-prime ministers. If informal cabinet committees deal with policy domains and in this sense have an interdepartmental rationale, the vice-prime ministers act in their role mostly when matters require a more party political nature. Junior ministers’ portfolios are positions for taking a part of the larger ministerial package and specializing on it. They are also created to establish a portfolio balance between the coalition partners and to express priorities in the policy agenda of the new government. Such priorities usually are explicated to allow individual coalition parties to be visible on themes that do not incur real political risk to the coalition as a whole. Thus junior minister positions have been created for promoting women emancipation and for family matters. The most common use of junior ministers however is to reduce the minister’s workload; they are usually responsible for a part of the minister’s portfolio. It is also common practice to allocate such positions across the coalition; more often than not, junior ministers are from a different party than their minister. Junior ministers are, however, not necessarily ‘watchdogs’ acting as their own party’s agents in the other party’s department, but they rather function as bidirectional messengers to fine tune coalition policy between them (Timmermans and Andeweg 2000: 380).

Governance mechanisms in the parliamentary arena The leaders of parliamentary groups face the double task of keeping up a party profile and harmonize matters important to the continuation and survival of the coalition. Electoral uncertainty and an increasingly harsh culture of accountability in the public arena have pushed the parliamentary face of these members of the party leadership more to the foreground and reduced the tolerance for ‘lip service’ in unconditionally supporting the government. Party leaders in parliament that move too close towards a position of identification with the government are vulnerable in their role in parliament. During the Rutte III cabinet not only the parliamentary leaders but also area specialists in charge of specific policy themes pushed the leeway for opinionated messages to a maximum (Korteweg 2018). The principle of dualism, a clear separation of responsibilities between government and supporting parties in parliament, is also reinforced by the sometimes very narrow majority or even the absence of a majority for the coalition in the Senate, the other legislative chamber where support parties must be found to secure that legislative proposals actually receive final political endorsement. Party profiling by coalition partners produced more risk of internal controversy. In weekly meetings between the vicechairs of the coalition parties in the Tweede Kamer, voting on matters on the parliamentary agenda is discussed and coordinated in order to contain the risk of escalation (Korteweg 2018).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

472

  

The Torentjesoverleg (named after the small octagonal tower that houses the prime minister’s office) is an essential informal venue where more agenda items are being scheduled for discussion between the top of the cabinet and the top of the parliamentary groups. In the early 2000s the idea came up to refresh Dutch politics (a so-called ‘new politics’) and abolish such mysterious and nontransparent venues as the Torentjesoverleg. But the rejuvenation idea foundered quickly and the prime ministerial tower was unlocked again for coalition summits. What does seem to have changed, however, is the timing of these summits. Parties are more keen to avoid press coverage of parliamentary party leaders entering the prime minister’s office as not to create too strong an image of collusion. While the weekly press conference of the prime minister is on Fridays, the consultation of party and cabinet leaders has moved away from the prime minister’s office and to Monday mornings—when everyone else also is busy beginning the work week. Interviews with ministers in the Rutte II cabinet (2012–2017) reveal that this weekly summit was crucial to the survival of the two-party coalition that was almost permanently in a process of political negotiation (Niemantsverdriet et al. 2016). Moreover, more specific mechanisms of similar type of composition are used for policy streamlining: the Rutte III cabinet organized ‘cockpit talks’ involving the cabinet and the parliamentary leaders and area specialists to monitor progress in elaborating a national climate agreement (Meeus 2018). Partly as a result of the fragmentation of parliamentary politics and the fact that the Rutte III government includes no less than four parties, these consultations between parliamentary party leaders and the prime minister and vice- prime ministers have become more central in the last years. Even government ministers sometimes take a back seat, referred to by Koole (2018) as a process of ‘governmentalisation’.

Governance mechanisms with different types of actors (mixed) With a less stable parliamentary majority and the need to secure a majority case by case even on important matters in the Senate, coalition governance also has come to involve a wider politics of agreements with societal stakeholders with major influence on sectors of the economy and society. This extension of the politics of agreements resembles the practice of neo-corporatism but it is more closely connected to the governance of the coalition itself. It not only is a reinvigoration of a policy-making style but also serves the stability of the government itself. In the period since 2010 a housing agreement, a pensions agreement, an energy agreement, a health care agreement, and a climate agreement were negotiated, all meant to provide a wide enough support basis for government policy that it would be possible to move intentions beyond mere plans and set a course of reforms.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

473

Cabinet duration and termination The duration of cabinets Dutch governments often end prematurely. In the 2000s, none of the governments formed by Prime Minister Balkenende achieved the maximum possible duration: they ended before the next constitutionally mandated election. Historically, there have been ups and downs in terms of government duration. In the 1940s, both governments ended prematurely, resulting in a relative duration of just under 60 per cent of the maximum possible duration. Table 13.6 displays the relative duration as a share of the maximum possible duration, although it underestimates cabinet duration somewhat because it includes caretaker (transitional) governments, which are almost never intended to last until the next constitutionally mandated elections but rather to organize early elections. Governments in the 1950s were much more stable than the two cabinets in the 1940s, achieving about 75 per cent relative duration, which goes up to almost 90 per cent if we exclude the caretaker cabinet Beel II. In the 1960s this declined to 71 per cent and it remained similarly at 74 per cent in the 1970s (excluding caretakers). While the 1980s started with the short-lived Van Agt II cabinet, the subsequent Lubbers cabinets were much more stable, achieving 70 per cent relative duration (excluding the Van Agt III caretaker government). The 1990s were the most stable decade, with both cabinets of Kok (almost) fulfilling their full term. The break with the ‘unstable 2000s’ was significant. The average relative duration of the three non-caretaker Balkenende cabinets was 57 per cent, the lowest since the 1940s. More recently, Rutte II’s completion of its full term means that relative duration in the 2010s has gone up to 76 per cent. The fact that Rutte II managed to complete its full term in office is not mere coincidence. Aware of the instability of the Balkenende cabinets and after an early demise of his first cabinet, Prime Minister Rutte was very much focused on a quick government formation and providing stable government. His coalition partner’s (PvdA) counterpart agreed with this and the formation of Rutte II was indeed among the quickest in recent times. The Rutte II cabinet was the first since 1998 (Kok I) to complete its full term. The fact that both government parties started to lose support in opinion polls fairly quickly into the government’s term will also have contributed to the government’s stability: neither party would stand to gain much from early elections. Perhaps Rutte II was the stable exception in unstable times. The mostly unstable governments that were formed since the ‘2002 Fortuyn’ revolution seem to be connected to the fragmentation of parliament and the subsequent difficulty in forming a government. After the 2017 election, there was only one ‘large’ party with over 15 per cent of the vote, while five ‘mid-sized’ parties obtained between 9 per cent and 12 per cent. Coalitions require more

84.5

69.3 80.1

2002-07-22 2002-10-16

2003-05-27 2006-06-30

Balkenende I

Balkenende II

Balkenende III 2006-07-07 2006-11-22

Balkenende IV 2007-02-22 2010-02-23

6.2

1989-11-07 1994-05-03 100 1994-08-22 1998-05-05 100 1998-08-03 2002-04-16 98.0

Lubbers III Kok I Kok II

7a

4

7a

8

1 1 9

Comments

The cabinet resigned after a report released by the Dutch Institute for War Documentation, from which Kok drew the conclusion that Dutch UN troops had failed to prevent the fall of Srebrenica in 1995. The government remained in office in a caretaker capacity until the May elections, however. LPF The government resigned after infighting in LPF paralysed the ruling coalition. VVD, D66 Immigration D66 withdrew from the coalition over policy controversial statements made by Minister for Immigration and Integration Rita Verdonk of VVD. The government then survived a vote of no confidence initiated by D66, but resigned anyway and called early elections. Caretaker cabinet to organize early elections. CDA, Foreign PvdA left the ruling coalition over policy PvdA policy disagreements on the Dutch troop

Foreign policy

Policy Relative Mechanisms Terminal Parties area(s) events (when duration of cabinet conflict termination (%) between or within)

Date in

Cabinet

Date out

Table 13.6 Cabinet termination in the Netherlands, 1989–2018 OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2010-10-14 2012-04-23

2012-11-05 2017-03-15 100

Rutte I

Rutte II

1

4, 9

4

Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event Note: Relative duration: share of constitutionally allowed duration for this cabinet.

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister

52.4

2010-02-23 2010-06-09 39

Balkenende V

presence in Afghanistan. Early elections were called for 9 June. Caretaker cabinet to organize early elections. The government resigned on 23 April 2012 after it lost the external support of PVV, on which it relied for majority status in the lower house. Early elections were called for 12 September 2012.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

476

  

parties to win a majority and those majorities are more often than not small. Combining this with a relatively high number of politicians ‘crossing the floor’, that is leaving their parliamentary party group (usually to start their own party), it means that governing becomes even more of a balancing act than in the past.

The termination of cabinets Table 13.6 also records the main reasons for termination. The reasons for a premature end to many cabinets are mostly related to political conflict within the coalition. In recent times, all (non-caretaker) Balkenende cabinets resigned due to (policy) conflict between or within coalition parties. Rutte I resigned when the support party PVV left negotiations on financial reforms. This is a continuation of the reasons for cabinet termination over the whole post-Second World War period (see Table 13.6). Some cabinets end due to early parliamentary elections, but these mostly concern caretaker cabinets that were installed to organize those elections. Cabinets in regular elections have become a relatively rare phenomenon in the last 20 years, due to high levels of early terminations. Since the Cals cabinet in the 1960s, the unwritten rule has been that cabinet resignations have to be followed by early elections. Governments are generally not defeated in parliament. The most recent case of a parliamentary vote leading to a cabinet crisis was in 1999, when the government failed to obtain two-thirds majority support in the Senate for the introduction of a referendum. The government’s resignation was, however, revoked after the parties found a compromise that allowed them to continue their coalition. Because the cabinet’s resignation was not officially accepted, we do not treat this and similar cases as early terminations (and a new cabinet being formed). The only case where a parliamentary vote directly led to the resignation of a cabinet (that was not later revoked) is the infamous ‘Night of Schmelzer’ in 1966, when the leader of the KVP introduced a motion that was interpreted by the prime minister as a nonconfidence motion. The motion was subsequently adopted, which led to the government’s resignation. While individual ministers have lost confidence motions, no cabinets have done so since 1966.

Conclusion While substantial changes in the party system and electoral context presented new challenges for coalition governance in the Netherlands, Dutch politicians have mainly looked to old solutions to address these new problems. Majority coalition formation is still the norm and the (unintended) introduction of an investiture vote for the formateur has further cemented this norm. Electoral volatility and the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

477

effective number of parties have increased over the last 30 years. Immigration and integration have established themselves as a new line of conflict. Despite the rise of new parties on the right, centre, and left, governance is still dominated by the ‘traditional’ three-plus-one (CDA, VVD, PvdA, and D66). Since 1994, only two new parties have participated in government. Where something new has been tried, most notably the Rutte I supported minority cabinet, this was not evaluated very positively by the actors involved. While part of the negative assessment was due to the early demise of this cabinet and the policy conflict between PVV on the one hand and VVD and CDA on the other, the dependency on a ‘third party’ was also seen as something that should be avoided. Fragmentation and party instability have impacted on cabinet governance. The start of the twenty-first century was an era of cabinet instability. Parties leaders have tried to address this by insisting on party unity so that they could deliver the votes for compromises that were agreed to in the coalition. These demands of loyalty and discipline in some cases had the effect of alienating some government MPs, which led to floor-crossing and thus a further erosion of parliamentary support. Party leaders therefore have to balance between appeasing their MPs and voters as well as avoiding (serious) conflict with other government parties. Coordination within the coalition and conflict resolution mechanisms have therefore become even more important for the survival of cabinets. The Netherlands can clearly be described in terms of the Coalition Compromise Model, in which the coalition agreement as well as coordination and conflict management mechanisms are of central importance. Balance has to be maintained in several arenas. In the parliamentary arena the coalition party leaders regularly meet to avoid unnecessary conflicts. In government, ministers who formally have a relatively strong autonomy have to take into account the views of their parliamentary party group and the sensitivities of coalition partners and those of other ministers. At the very top, the prime minister and leaders of the other parties have to see whether the coalition agreement is upheld or if ad hoc changes are necessary. While the prime minister’s formal powers are quite limited, his leadership of the largest party and his central role in policy coordination at home and in Europe strengthen his position. Whereas electoral politics and political campaigns have become more adversarial in style over the last 20 years, coalition governance still seems to be guided by the politics of accommodation. In that sense, our analysis confirms earlier accounts of Dutch (coalition) politics by Lijphart (1989) and Timmermans and Andeweg (2000). At the same time, the changes in volatility, fragmentation, and the party system are much more extensive now than they were 20 years ago. It needs to be seen whether old solutions keep working in this significantly changed context.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

478

  

References Andeweg, Rudy, and Galen Irwin (2014). Governance and Politics of the Netherlands, 4th edn. Basingstoke: Palgrave Macmillan. Bäck, Hanna, Marc Debus, and Patrick Dumont (2011). ‘Who Gets What in Coalition Governments? Predictors of Portfolio Allocation in Parliamentary Democracies’. European Journal of Political Research, 50(4): 441–78. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Anna Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999-2010’. Party Politics, 21(1): 143–52. Baumgartner, Frank, and Bryan Jones (2015). The Politics of Information: Problem Definition and the Course of Public Policy in America. Chicago and London: University of Chicago Press. Bolhuis, W. (2018). ‘Van Woord tot Akkoord. Een analyse van de partijkeuzes in CPB -doorrekeningen van verkiezingsprogramma’s en regeerakkoorden, 1986–2017’. Leiden University. Council of State (n.d.). ‘The Council of State’. Retrieved from: https://www. raadvanstate.nl/talen/artikel/ (Accessed on 19 December 2019). De Winter, Lieven, and Patrick Dumont (2008). ‘Uncertainty and Complexity in Cabinet Formation’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 123–57. Ecker, Alejandro, Thomas M. Meyer, and Wolfgang C. Müller (2015). ‘The Distribution of Individual Cabinet Positions in Coalition Governments: A Sequential Approach’. European Journal of Political Research, 54(4): 802–18. Fiers, Stefaan, and André Krouwel (2005). ‘The Low Countries: From “Prime Minister” to President-Minister’. In Thomas Poguntke and Paul Webb (eds), The Presidentialization of Politics: A Comparative Study of Modern Democracies. Oxford: Oxford University Press, 128–58. Jacobs, Kristof, Marijn van Klingeren, Henk van der Kolk, Tom van der Meer, and Eefje Steenvoorden (2016). Het Oekraïne-referendum: Nationaal Referendum Onderzoek 2016. Den Haag: Ministerie van Buitenlandse Zaken. Retrieved from: http://repository.ubn.ru.nl/handle/2066/166055 Jennings, Will, Shaun Bevan, Arco Timmermans, Gerard Breeman, Sylvain Brouard, Laura Chaques-Bonafont, Christoffer Green-Pedersen, Peter John, Peter Mortensen, and Anna Palau (2011). ‘Effects of the Core Functions of Government on the Diversity of Executive Agendas’. Comparative Political Studies, 44(8): 1001–30. Koole, Ruud (2018). ‘Gouvernementalisering: De veranderende verhouding tussen regering en parlement in Nederland’. Tijdschrift Voor Constitutioneel Recht, 9 (October): 317–41.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

479

Korteweg, Ariejan (2018). ‘Na negen maanden is het “praatkabinet” Rutte III nog steeds een groot raadsel’. De Volkskrant. 6 July. Retrieved from: https://www. volkskrant.nl/nieuws-achtergrond/na-negen-maanden-is-het-praatkabinet-rutte-iiinog-steeds-een-groot-raadsel~bd9ccc7c/ Kriesi, Hanspeter, Edgar Grande, Romain Lachat, Martin Dolezal, Simon Bornschier, and Timotheos Frey (2006). ‘Globalization and the Transformation of the National Political Space: Six European Countries Compared’. European Journal of Political Research, 45(6): 921–56. Laver, Michael, and Kenneth A. Shepsle (1996). Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Lijphart, Arend (1968). The Politics of Accommodation. Pluralism and Democracy in the Netherlands. Berkeley: University of California Press. Lijphart, Arend (1989). ‘From the Politics of Accommodation to Adversarial Politics in the Netherlands: A Reassessment’. In Hans Daalder and Galen A. Irwin (eds), Politics in the Netherlands: How Much Change? London: Cass, 139–58. Louwerse, Tom (2010). De CPB-ruimte: een grafische presentatie van de doorrekening. Retrieved from: http://blog.tomlouwerse.nl/2010/05/de-cpb-ruimte-eengrafische-presentatie.html (Accessed on 28 October 2016). Mair, Peter (2008). ‘Electoral Volatility and the Dutch Party System: A Comparative Perspective’. Acta Politica, 43(2–3): 235–53. https://doi.org/10.1057/ap.2008.1 Meeus, Tom-Jan (2018). ‘Cockpitoverleg, Urgenda, Rutte versus Hoekstra:de klimaatblues in de coalitie’. NRC Handelsblad. 8 December. Retrieved from: https://www. nrc.nl/nieuws/2018/12/08/cockpitoverleg-urgenda-rutte-versus-hoekstrade-klimaat blues-in-de-coalitie-a3059955 Müller, Wolfgang C. and Kaare Strøm (eds.) (2020). Coalition Governments in Western Europe. Oxford: Oxford University Press. Müller, Stefan and Tom Louwerse (2020). ‘The Electoral Cycle Effect in Parliamentary Democracies’. Political Science Research and Methods, 8(4): 795–802. Niemantsverdriet, Thijs, Philip de Witt Wijnen, and Annemarie Kas (2016). ‘Het geheim van Rutte II: Vallen? Dat nooit’. NRC Handelsblad. 24 December. Retrieved from: https://www.nrc.nl/nieuws/2016/12/23/dan-ga-je-maar-naar-dekoning-5881783-a1538226 Otjes, Simon (2015). ‘How the Eurozone Crisis Reshaped the National Economic Policy Space: The Netherlands 2006–2012’. Acta Politica, 51: 1–25. Otjes, Simon and Tom Louwerse (2014). ‘A Special Majority Cabinet? Supported Minority Governance and Parliamentary Behavior in the Netherlands’. World Political Science, 10(2): 343–63. Pellikaan, Huib, Tom van der Meer, and Sarah de Lange (2003). ‘The Road from a Depoliticized to a Centrifugal Democracy’. Acta Politica, 38(1): 23–49. Peterson, Robert, Martine de Ridder, John Hobbs, and Fletcher McClellan (1983). ‘Government Formation and Policy Formulation: Patterns in Belgium and the Netherlands’. Res Publica, 25: 49–82.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

480

  

Rasch, Bjørn Erik, Shane Martin, and José Antonio Cheibub (2015). ‘Investiture Rules and Government Formation’. In Bjørn Erik Rasch, Shane Martin, and José Antonio Cheibub (eds), Parliaments and Government Formation: Unpacking Investiture Rules. Oxford: Oxford University Press, 3–26. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of AntiElitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research and Politics, 4(1): 1–9. Sieberer, Ulrich, Thomas Meyer, Andrea Ceron, Albert Falcó-Gimeno, Isabelle Guinaudeau, Martin Hansen, Wolfgang Müller, and Thomas Persson (2017). ‘The Politics of Portfolio Design in European Democracies’. Paper Presented at the 2017 ECPR General Conference, Oslo, 6–9 September. Strøm, Kaare (1990). Minority Government and Majority Rule. Cambridge: Cambridge University Press. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds.) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Thomson, Robert (1999). The Party Mandate. Election Pledges and Government Actions in the Netherlands, 1986–1998. Amsterdam: Thela Thesis. RUG/UU/KUN: ICS. Timmermans, Arco (1996). High Politics in the Low Countries: Functions and Effects of Coalition Policy Agreements in Belgium and the Netherlands. Florence: European University Institute. Timmermans, Arco (2006). ‘Standing Apart and Sitting Together: Enforcing Coalition Agreements in Multiparty Systems’. European Journal of Political Research, 45(2): 263–83. Timmermans, Arco, and Gerard Breeman (2010). ‘Politieke waarheid en dynamiek van de agenda in coalitiekabinetten’. In Carla van Baalen, Willem Breedveld, Marij Leenders, Johan van Merriënboer, Jan Ramakers, and Jouke Turpijn (eds), Jaarboek Parlementaire Geschiedenis: Waarheidsvinding en waarheidsbeleving. Amsterdam: Boom, 47–62. Timmermans, Arco, and Gerard Breeman (2014). ‘The Policy Agenda in Multiparty Government: Coalition Governments and Legislative Activity in the Netherlands’. In Christoffer Green-Pedersen and Stefaan Walgrave (eds), Agenda Setting, Policies, and Political Systems: A Comparative Approach. Chicago: University of Chicago Press, 87–104. Timmermans, Arco, and Gerard Breeman (2017). ‘Information Processing in Coalition Governance: How Political Reset the Policy Agenda when in Office Together’. Paper Presented at the 2017 ECPR Joint Sessions of Workshops, Nottingham, UK. Timmermans, Arco, and Rudy B. Andeweg (2000). ‘Coalition Cabinets in the Netherlands: Still the Politics of Accommodation?’. In Wolfgang C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford and New York: Oxford University Press, 356–98. Van Baalen, Carla, and Alexander van Kessel (eds) (2016). Kabinetsformaties 1977–2012. Amsterdam: Boom.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 

481

Appendix. List of political parties Abbreviation SP CPN PSP GL EVP PPR PvdD PvdA DENK 50PLUS AOV Unie 55+ CU D66 ARP KVP CDA CHU DS70 NMP CP-CD GPV RPF SGP RKPN VVD BP-RVP LPF PVV

Name Socialist Party (Socialistische Partij) Communist Party Netherlands (Communistische Partij Nederland) Pacifist-Socialist Party (Pacifistisch-Socialistische Partij) GreenLeft (GroenLinks) Evangelical Peoples Party (Evangelische Volkspartij) Political Party of Radicals (Politieke Partij Radicalen) Party for the Animals (Partij voor de Dieren) Labour Party (Partij van de Arbeid) Denk 50PLUS General Elderly Alliance (Algemeen Ouderen Verbond) Union 55+ (Unie 55+) ChristianUnion (ChristenUnie) Democrats 66 (Democraten 66) Anti-Revolutionary Party (Anti-Revolutionaire Partij) Catholic People’s Party (Katholieke Volkspartij) Christian Democratic Appeal (Christen Democratisch Appèl) Christian-Historical Union (Christelijk-Historische Unie) Democratic Socialists ‘70 (Democratisch Socialisten ‘70) New Middle Party (Nieuwe Midden Partij) Centre Democrats (Centrum Democraten), 1984–2002 Centre Party (Centrumpartij), 1982–1986 Reformed Political Union (Gereformeerd Politiek Verbond) Reformatory Political Federation (Reformatorische Politieke Federatie) Reformed Political Party (Staatkundig Gereformeerde Partij) Roman Catholic Party of the Netherlands (Rooms Katholieke Partij Nederland) People’s Party for Freedom and Democracy (Volkspartij voor Vrijheid en Democratie) Right-wing People’s Party (Rechtse Volkspartij), 1981 Farmers’ Party (Boerenpartij), 1958–1981 Pim Fortuyn List (Lijst Pim Fortuyn) Party for Freedom (Partij voor de Vrijheid)

Note: Party names are given in English, followed by the party name in Dutch in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 14 Norway: Towards a More Permissive Coalitional Order Torill Stavenes and Kaare W. Strøm

Norwegian coalition governance has over time been characterized by consensus, compromise, and collegiality. Consensus has meant a preference and search for broad agreements among coalition parties, and sometimes also including support parties. Compromise has meant a willingness among coalition parties to negotiate across policy areas and ministries and accept common platforms and constraints. Collegiality has meant intra-cabinet negotiations on the basis of equal standing, with less deference to rank and hierarchy than in many other cultures. These characteristics of Norwegian coalition governance have necessitated extensive face-to-face interaction. In fact, it has been argued that in order to be a cabinet minister in Norway, one must love meetings. In consequence, Norwegian coalition governance thus adheres to the Coalition Compromise Model, characterized by inter-party compromise and negotiation, rather than the Prime Minister Model, characterized by deference to the head of government, or the Ministerial Government Model, featuring extensive deference to individual cabinet members within their respective jurisdictions. Within this overarching feature of compromise rather than command or dispersion of authority, central aspects of Norwegian coalition politics have come into flux since the 1990s. In short, Norwegian parliamentarism has become increasingly permissive or flexible. Permissiveness in this context has several features. One is that the coalitional options have expanded and become more negotiable. An especially notable manifestation of this development is that parties that were previously considered non-coalitionable have become accepted members of the ‘club’. A second development is that portfolio allocation has become much less predictably based on partisan issue ownership. Moreover, small but pivotal parties have been able to reap much greater and numerically disproportional rewards, compared to coalitions of the past. Finally, there has been continuous experimentation with new or modified governance mechanisms. Most notably, Kåre Willoch’s coalition cabinet in 1983 introduced an inner cabinet, the so-called sub-committee, to deal with serious policy conflicts. In subsequent governments, this sub-committee at times, for example in Jens Stoltenberg’s second and third cabinets (2005–2013), became so powerful that it was criticized Torill Stavenes and Kaare W. Strøm, Norway: Towards a More Permissive Coalitional Order In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0014

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



483

as undemocratic. Moreover, there has been an increased tendency in coalitions after the 1990s to make so-called cross-partisan junior minister appointments, that is representatives from a party that is not holding the cabinet ministership. This chapter will explore the politics of coalition compromise in Norway and especially these trends towards coalitional permissiveness.¹

The institutional setting The constitutional chain of governance in Norway is comparatively simple. Under the constitution, all legislative authority is exercised by the people through parliament (Art. 49): ‘The people exercise the Legislative Power through the Storting’ (Norwegian Constitution 1814). Although national direct democracy has been employed on six occasions since national independence in 1905, it fell into disuse from the 1920s to the 1970s. Like all major Western European states except Switzerland, Norway is a parliamentary democracy, albeit a peculiar one. Norway conforms to Westminster parliamentarism in the sense that members of parliament (the Storting) are the only national agents elected by the people. In its heyday in the 1950s, Norway also came very close to the broader Westminster model of a singular chain of democratic delegation, running through parliament and controlled by political parties (see Strøm 2000). More recent trends, however, have in many ways distanced Norway from that model. Norwegian democracy was born in a spirit of nationalism, egalitarianism, and Madisonianism. The Norwegian constitution, dating back to 17 May 1814, is currently the second oldest codified constitution in the world, after the US constitution of 1787. The 1814 constitution was brought about by the Napoleonic Wars, which brought to an end the 400-year era of Danish supremacy. The Norwegian constitution was thus part of the nationalist movement that shaped much of Europe’s nineteenth century and that had roots in the French Revolution of 1789. Nationalism meant a demand for national sovereignty, which was eventually tempered into a dual monarchy with Sweden, under a Swedish king but with extensive domestic autonomy. Nationalism also meant that to the framers of the constitution, and to many politicians in the nineteenth-century Norway, the Storting (parliament) would occupy a very privileged place in the ¹ A previous version of this chapter was presented at a conference at Lund University, Sweden, 17–18 May 2018. We thank the organizers for their contributions to the data collection on which this chapter builds. Our thanks also to Knut Heidar, Kristoffer Kolltveit, Paul Mitchell, and other conference participants for their insightful comments and suggestions. We base the chapter on nine interviews with party representatives. For Bondevik II cabinet: the Liberal Party, 14 September 2017; the Christian People’s Party, 15 September 2017; the Conservative Party, 12 September 2017; and the Conservative Party, 15 September 2017. For Stoltenberg II and III: the Centre Party, 14 September 2017; the Labour Party, 06 October 2017; and the Socialist Left, 20 September 2017. For Solberg I: the Progress Party, 21 September 2017 and the Conservative Party, 22 September 2017.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

484

  ø

regime they built. Coming out of absolute monarchy under a Danish king and adapting to the prospect of a dual monarchy with an ambitious Swedish head of state, the Storting represented not only ordinary citizens but also the national community of Norway. Thus, empowering the Storting both promoted the interests of ordinary citizens (or at least those with voting rights) and erected a bulwark against Swedish domination.² A second strand of the constitutional spirit was egalitarianism, driven in large part by the examples of the American and especially the French Revolution. By international standards, the 1814 constitution was generous (though less so than some of the constitutions on which it was modelled) in extending voting rights to about 30–40 per cent of all men aged 25 or above. Throughout the nineteenth century, this franchise was extended, so that Norway in 1898 became the first Nordic country to extend voting rights to all adult men (and some categories of women), followed by universal adult suffrage in 1913. Egalitarianism also meant a phasing out of the (very small) Norwegian nobility and a rejection of the legislative models of the other Nordic countries, which at that time provided for separate chambers for the various social estates. Finally, egalitarianism, and the strong homogeneity of Norwegian society, precluded any form of federalism in the new state. Finally, the 1814 constitution was Madisonian in its commitment to separation of powers and limited government. James Madison was president of the United States when Norway got its 1814 constitution, and although he did not personally influence its drafting, Madison’s ideas and the example of the American Revolution certainly did. The 1814 constitution was a separation-of-powers document, with a multitude of checks and balances and no provisions for political parties to play any significant role. The constitutional provisions concerning the powers and operations of the Storting were effectively a carbon copy of the provisions that Madison had written into the American constitution. Yet, the Norwegian document departed from Madison’s legacy in rejecting religious liberties, federalism, and bicameralism. The latter feature became the subject of one of the most divisive debates of the five to six weeks during which the Constitutional Assembly deliberated. The quasi-bicameral solution that was finally adopted is a good illustration of the fact that compromise does not necessarily guarantee ‘the best of both worlds’. In fact, with the advent of organized political parties, the Storting quickly became effectively unicameral. Norwegian democracy did not always remain Madisonian. For about a century and a half after 1814, Norway moved ever closer to a Westminsterian form of ² A less attractive reflection of nationalism in the 1814 constitution was its rejection of religious and ethnic pluralism. Some of the constitutional provisions that would subsequently become most controversial were its denial of access to Jewish individuals and to members of Catholic orders. The ban on Jewish immigration was lifted in 1851, whereas remarkably the prohibition against Catholic orders remained in effect as late as the 1950s.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



485

government. The initial battle lines were drawn between a parliament elected by national voters (many of them small farmers) and a cabinet appointed by the dual monarchy king. The constitutional pre-eminence of the Storting became a rallying cry for the constitutional reformers of the 1880s. ‘All power in this assembly!’ was the slogan of Liberal leader Johan Sverdrup. And Sverdrup succeeded in his aims. The right of the parliamentary majority to dismiss the cabinet was established in the constitutional crisis of 1884, when the Liberal Party gained control of parliament and impeached the members of the incumbent Conservative cabinet. Although the parliamentary principle remained somewhat contested until Norwegian independence in 1905, there has since been a clear and unambiguous convention that individual ministers or cabinets must resign if they lose confidence votes in the Storting (Andenæs 1981; Hansen and Mo 1994: 118; Nordby 2000). Norway thus became the first Nordic country to adopt parliamentary democracy. It had predictable consequences. Sustained since the 1880s by strong and cohesive political parties, the cabinet increasingly came to dominate national policy-making, especially in the decades following the Second World War. Thus in 1959, at the height of his party’s predominance, Labour Party parliamentary leader Nils Hønsvald proclaimed that parliamentary control of the executive had effectively been transferred from the floor of parliament to the internal organs of his party. Party government had eclipsed parliamentary deliberation. Around 1960 Norway had thus reached a constitutional form that, apart from some party fragmentation on the centre-right caused by proportional representation, looked decidedly Westminsterian. Since then, however, Norway has moved back towards a more Madisonian separation-of-powers system. A series of weaker minority governments have given rise to parliamentary reassertion and to more flexible and permissive coalition formulas and parliamentary practices. The Norwegian party system has fragmented, and the individual parties have atrophied as mass membership organizations. In addition, a heightened assertiveness on the part of the judiciary has helped contain parliamentary power. Furthermore, two hotly contested EU membership referendums in 1972 and 1994 have firmly established the role of direct democracy in critical political decisions. In this chapter, we will dwell on the implications for coalition bargaining in particular. The place to look for these important changes has rarely been in the formal constitution itself. In recent years, the Norwegian constitution had become an increasingly archaic and often ambiguous document, with major discrepancies between formal constitutional provisions and prevailing political practice. Thus, ministerial responsibility to parliament has a long history in Norway, but for a long time had only a tenuous place in the formal constitution. In the run-up to the constitutional bicentennial of 2014, however, parliament launched a project to revise the constitution and update its formal provisions to conform to conventions that had evolved since the 1880s. Thus, it was not until 2007 that the constitution

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

486

  ø

was amended to require cabinet members to resign in the event of a parliamentary vote of no confidence. Ministerial responsibility thus became a part of the constitution through the new Article 15 (Constitutional Amendment No. 364). At the same time, the old and unsuitable division of parliament into an upper quasi-chamber (Lagtinget) and a lower one (Odelstinget) was abolished (Constitutional Amendment No. 365).³ In addition, changes were introduced to the Court of Impeachment, which played a critical role in launching parliamentarism in the 1880s but had been of little political consequence since the 1920s. One reform proposal that was debated but ultimately rejected would have introduced a mechanism of parliamentary dissolution. Norway thus remains the only parliamentary system in Europe that has no provision for the early dissolution of its national legislature (or any of its subnational assemblies, for that matter). This lack of a parliamentary ‘safety valve’ and the consequent lack of an option to consult the voters rather than form an alternative government have affected coalition bargaining in Norway in systematic ways.

The party system and the actors Party system change For a long time, parliamentary and cabinet politics in Norway has been party politics. Independents and non-partisans have been rare indeed among members of parliament and virtually non-existent in the cabinet. Norwegian parties date back to the 1880s and the divisive constitutional debates that gave rise to the organization of Liberals versus Conservatives. Liberals represented the egalitarianism and nationalism of the constitution, Conservatives its Madisonianism. Bipartism quickly faded, however, as the Norwegian Labour Party was founded in 1889 and rapidly grew from a minor to a dominant party between the early 1900s and the 1920s. With the adoption of Proportional Representation after the First World War, an agrarian, later centre, party gained parliamentary representation, along with a splinter Labour Party faction that became the Norwegian Communist Party (NKP). The NKP rapidly lost support from the 1950s on and eventually merged into the Socialist Left Party in the mid-1970s. A Christian People’s Party emerged on the national scene after the Second World War—it had in the 1930s operated only in western Norway, where Christian communities were strong. In 1961, a left-wing splinter group broke off from the Labour Party and formed the Socialist People’s Party, which in 1975 merged with other left-leaning (and anti-European Union) forces to form the

³ These two amendments took effect after the 2009 general election.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



487

Socialist Left. In 1973, a populist right-wing party formed that eventually became the Progress Party.⁴ The same year saw the formation of a small extreme left Marxist-Leninist party, currently known as the Red Party. Finally, a Green (Environmental) party (MDG) was founded as early as 1988 but did not gain parliamentary representation until 2013. The most dramatic party system changes in the 2000s have been the rise and subsequent decline of the two most prominent ‘flank’ parties: the Progress Party and the Socialist Left. From its founding in 1973 until the late 1980s, the Progress Party was a small party, hovering on the brink of parliamentary extinction. It then made a breakthrough in the 1989 election and continued to grow rapidly, albeit with the occasional setback, until it became Norway’s second largest party in 2005 and thus eclipsed the Conservatives as the largest centreright party. Its electoral support has since subsided but remains north of 15 per cent. The party’s growth had much to do with popular scepticism towards immigration and towards politicians and establishment elites. It is thus a fairly typical populist right party, though a comparatively moderate one with a significant libertarian following. The Socialist Left had much greater success in its early phase in the 1970s but then experienced fluctuating fortunes until the Labour Party began its decline in the late 1990s and early 2000s. Long considered fringe parties with little governing potential, both parties have entered coalition governments in the 2000s: the Socialist Left from 2005 to 2013 and the Progress Party from 2013 to 2020. Despite its gradual fragmentation (for details, see Table 14.1a⁵), from the 1930s until the early 1970s Norway had one of the most stable party systems in Western Europe. It was defined around six dimensions of political cleavage rooted in economic, geographical, and cultural circumstances (see Rokkan 1967, 1970; Valen and Rokkan 1974). The policy dimensions of Norwegian politics have traditionally been shaped largely by these dominant social cleavages. The socio‐economic class cleavage is expressed through the left–right policy dimension, whereas territorial and sectoral cleavages are reflected in the centre–periphery and the urban–rural policy dimensions respectively. Cultural cleavages have given rise to a moral–religious dimension related to the establishment and political influence of the Lutheran church and also to matters such as education, abortion, and regulations of the sale and consumption of alcohol. The major division, however, has traditionally been between socialists (or centre-left) and non-socialists (or centre-right) located along a left–right axis.

⁴ The Progress Party was originally Anders Lange’s Party, named after its founder. ⁵ Compiled with the help of Ray and Narud (2000), Svåsand et al. (1997: 96), and Stortinget.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

488

  ø

While different studies of Norwegian policy dimensions yield somewhat different spatial representations, they leave no doubt that the dominant dimension of political contention over the past century has been this left–right axis. From left to right,⁶ the parties as of March 2021 are: the Red Party (Rødt, R), the Socialist Left (Sosialistisk Venstreparti, SV), the Greens (Miljøpartiet de grønne, MDG), the Labour Party (Arbeiderpartiet, A or DNA), the Centre Party (Senterpartiet, SP), the Christian People’s Party (Kristelig Folkeparti, KRF), the Liberals (Venstre, V), the Conservative Party (Høyre, H), and the Progress Party (Fremskrittspartiet, FRP).⁷ In the 1990s the predominance of the left–right dimension lessened somewhat, but it rose again after the turn of the millennium (Aardal and Valen 1995; Narud 1996, 2007). Table 14.1a shows that the Labour Party controlled the median position on the left–right dimension until 1981. As the non-socialists (or, more recently, the centre-right) gradually gained ground at the expense of Labour, the Christian People’s Party came to control the median position from 1981 to 2017, after which the median position was shared between the Christian People’s Party and the Centre Party. Two other important policy dimensions in Norwegian politics are strongly correlated: the urban–rural dimension and the centre–periphery axis. The policy positions of the various parties tend to be highly consistent across these two dimensions, as the parties that most favour rural interests are also the strongest defenders of the peripheries, and vice versa. On the urban–rural dimension, the ordering of the parties’ policy positions differs considerably from the left–right axis.⁸ The Progress Party has traditionally been the most pro-urban party and the Centre Party the most pro-rural one. The Conservatives are fairly close to the Progress Party, whereas the Christian People’s Party tends to be pro-rural. The other parties hold more centrist positions on this dimension (see e.g. Valen and Narud 2007a; Narud and Rasch 2007). Thus, the traditional allies in the centre-right camp, who are fairly close neighbours on the left–right axis, are highly polarized along the urban–rural dimension. After 2000, the

⁶ Rankings are based on values on the economic left–right dimension (the ranking is the same when using the values from the ideological left–right dimension) from the Chapel Hill Expert Survey (Bakker et al. 2015; Polk et al. 2017). The Red Party is not included in the ranking itself, but its position as the most far-left party in the Storting today is uncontested. ⁷ Two smaller and largely regional parties have gained parliamentary representation in recent decades: Folkeaksjonen Fremtid for Finnmark [Popular Movement for the Future of Finnmark] (1989–93) and Tverrpolitiske Folkevalgte, from which a group split and formed Kystpartiet [the Coastal Party] (1997–2005). ⁸ The urban–rural (cum centre–periphery) dimension has on several occasions influenced coalition politics as well as party competition. For this reason, we define it as the second most important dimension. The third most important dimension may be the moral–religious axis, on which the Christian People’s Party is at one pole and the parties of the left–right extremes (the Socialist Left and the Progress Party) at the other.

Cabinet

Gerhardsen II Gerhardsen III Torp I Torp II Gerhardsen IV Gerhardsen V Gerhardsen VI Lyng Gerhardsen VII Borten I Borten II Bratteli I Korvald Bratteli II Nordli I Nordli II Brundtland I Willoch I Willoch II Willoch III Brundtland II Syse Brundtland III Brundtland IV Jagland

Cabinet number

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

1945-11-05 1949-10-10 1951-11-19 1953-10-12 1955-01-22 1957-10-07 1961-09-11 1963-08-28 1963-09-25 1965-10-12 1969-09-08 1971-03-13 1972-10-18 1973-10-16 1976-01-15 1977-09-12 1981-02-04 1981-10-14 1983-06-08 1985-09-09 1986-05-09 1989-10-16 1990-11-03 1993-09-13 1996-10-25

Date in

1993-09-13

1989-09-11

1985-09-09

1981-09-14

1977-09-12

1973-09-10

1965-09-13 1969-09-08

1957-10-07 1961-09-11

1953-10-12

1945-10-08 1949-10-10

Election date

Table 14.1a Norwegian cabinets since 1945

A A A A A A A H, SP, V, KRF A SP, H, V, KRF SP, H, V, KRF A KRF, SP, V A A A A H H, KRF, SP H, KRF, SP A H, KRF, SP A A A

Party composition of cabinet maj maj maj maj maj maj min min min mwc mwc min min min min min min min mwc min min min min min min

Type of cabinet 50.7 56.7 56.7 51.3 51.3 52.0 49.3 49.3 49.3 53.3 50.7 49.3 31.3 40.0 40.0 49.0 49.0 34.2 51.0 49.7 45.2 37.6 38.2 40.6 40.6

Cabinet strength in seats (%) 150 150 150 150 150 150 150 150 150 150 150 150 150 155 155 155 155 155 155 157 157 165 165 165 165

Number of seats in parliament 6 5 5 6 6 6 6 6 6 6 5 5 5 8 8 6 6 7 7 6 6 6 6 8 8

Number of parties in parliament 3.18 2.67 2.67 3.09 3.09 2.99 3.22 3.22 3.22 3.51 3.18 3.18 3.18 4.14 4.14 2.97 2.97 3.18 3.18 3.09 3.09 4.23 4.23 4.04 4.04

ENP, parliament

A A A A A A A A A V V V V A A A A A KRF KRF KRF KRF KRF KRF KRF

Median party in first policy dimension

Continued

Formal support parties

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Bondevik I Stoltenberg I Bondevik II Stoltenberg II Stoltenberg III Solberg I Solberg II Solberg III

26 27 28 29 30 31 32 33

1997-10-17 2000-03-17 2001-10-19 2005-10-17 2009-09-14 2013-10-16 2018-01-17 2019-01-22

Date in

2001-09-10 2005-09-12 2009-09-14 2013-09-09 2017-09-11 2019-01-22

1997-09-15

Election date

KRF, SP, V A KRF, H, V A, SF-SV, SP A, SF-SV, SP H, FRP H, FRP, V H, FRP, V, KRF

Party composition of cabinet min min min mwc mwc min min mwc

Type of cabinet 25.5 39.4 37.6 51.5 50.9 45.6 47.3 52.1

Cabinet strength in seats (%) 165 165 165 169 169 169 169 169

Number of seats in parliament 7 7 7 7 7 7 7 7

Number of parties in parliament 4.36 4.36 5.35 4.56 4.07 4.39 4.95 4.95

ENP, parliament

KRF KRF KRF A A V KRF KRF

Median party in first policy dimension

KRF, V

Formal support parties

Notes: For a list of parties, consult the chapter appendix. Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1945–1998 period. The subsequent period is based on data from Ray and Narud (2000), Benoit and Laver (2006), Bakker et al. (2015), and Polk et al. (2017). The first policy dimension is economic left–right. Formal support parties have been coded for the 1990–2018 period. Cabinet types: maj = Single-party majority cabinet; min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition. Table legends and variables are further defined in the Appendix, this volume.

Cabinet

Cabinet number

Table 14.1a Continued OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



491

Christian People’s Party has consistently controlled the median legislator on the urban–rural dimension, except for in 2013 when the Labour Party captured the spot and in 2018 when the Christian People’s Party shared the median party position with the Conservatives.⁹ This dimension is therefore particularly inimical to centre-right cooperation. This secondary dimension has been mobilized most dramatically during the EC/EU campaigns in the early 1970s and again in the 1990s. On these occasions, the left–right dimension declined in significance, and the established patterns of left–right opposition weakened. Indeed, the saliency of the urban–rural dimension has led to the termination of two coalitions (Narud 1995) and inhibited the formation of others. It has also caused great internal strain on several parties, notably the Labour Party and the Liberals. The nearly identical results of the EC/ EU campaigns of the 1970s and the 1990s indicate that such conflicts may be latent but far from negligible. Instead, they reflect entrenched conflict dimensions that are likely to constrain future coalition bargaining, especially on the centreright, and which may pose serious future challenges to several major parties and especially the Labour Party. The urban–rural policy dimension can be powerfully mobilized even when no EU issue is on the agenda, as in 2017 when the Centre Party almost doubled its vote share, largely due to rural resistance to the incumbent government’s plans to merge many small Norwegian municipalities and several of the 19 provinces.

Electoral alliances and pre-electoral coalitions The post-war Norwegian party system has, in Laver and Schofield’s (1990) terms, changed from a unipolar, through a bipolar, to a multipolar format and is possibly now shifting back towards bipolarity. The unipolar phase reflects the Labour Party’s predominance between 1945 and 1961, during which time the party won a parliamentary majority in every election. The 1961 emergence of the Socialist People’s Party eroded Labour’s support and began a transition to a competitive two-bloc format. The dominant left–right issue dimension separated Norwegian parties into a socialist or leftist bloc—consisting of the Labour Party, the Socialist Left (and previously the Socialist People’s Party), and the Communists—and a non-socialist or bourgeois one, consisting of Conservatives, Liberals, the

⁹ Rankings post-2013 are based on values from the urban–rural dimension from the Chapel Hill Expert Survey (Polk et al. 2017). Rankings for the cabinets between 2000 and 2001 (Stoltenberg I and Bondevik II) are based on Ray and Narud (2000) decentralization scale, while we use the decentralization scale of Benoit and Laver (2006) for Stoltenberg’s second and third cabinets. Note that the actual distance (in values) between the parties on the urban-rural dimension is very small in the Chapel Hill data, which helps to explain the placement of the Conservatives, traditionally a pro-urban party, in the median position in 2018.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

492

  ø

Christian People’s Party, until the early 2000s the Centre Party, and since 1973 the Progress Party. Until the 1990s especially, the socialist and non‐socialist blocs were extremely evenly balanced, and the party system could best be described as an alternational two-bloc system. Most elections resulted in a close balance between the two blocs, and the result was frequent alternation between them, such as in 1963 (twice), 1965, 1971, 1972, 1973, 1981, 1986, 1989, 1990, 1997, 2000, 2001, 2005, and 2013. Since then, the non-socialists (or, in more recent terms, the centre-right) have gradually gained ground at the expense of the socialists (or centre-left). At the same time, however, the centre-right bloc has become less cohesive, and the Centre Party has abandoned the centre-right in favour of the centre-left. The bloc patterns have thus changed and gradually become more flexible, as Table 14.1a outlines. The three-party centrist coalition government formed in 1997 represented an important step away from the two-bloc format of the past. In 2001, a massive loss for the incumbent Labour Party paved the way for a minority government of the Conservatives with the Christians and the Liberals, while their former ally, the Centre Party, stayed in opposition. The 2005 election once again brought about massive changes in party support and a new coalition formula. The Labour Party regained some of its former strength, even though the result was the second worst in its post-war history. Two of the coalition parties, the Conservatives and the Christians, suffered severe losses, whereas the Progress Party increased its vote share by almost eight percentage points. The resulting centre-left coalition of Labour, the Centre Party, and the Socialist Left meant a departure from Labour’s long-standing anti-coalition stance (Valen and Narud 2007b; Aardal 2007). Pre-electoral coalitions and alliances have historically been a tacit, rather than explicit, feature of the Norwegian coalitional order. Between the early 1960s and the early 1990s, the battle lines were clearly drawn between the Labour Party and a bourgeois (centre-right) coalition. When the latter form of cabinet faced elections as incumbents, in 1969 and 1985, they left no records of formal pre-electoral negotiations. Nevertheless, in both cases the option of continuing in office in the event of a joint majority was always presumed and was in fact realized. In more recent years, pre-electoral negotiations and agreements have become more formalized, and, as shown in Table 14.1b¹⁰, in three consecutive elections (2005, 2009, and 2013) the centre-left coalition of Labour, the Centre Party, and the Socialist Left campaigned on the basis of such precommitments.

¹⁰ Compiled with the help of news reports, Arbeiderpartiet (2009), and Sandberg (2007).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



493

Table 14.1b Electoral alliances and pre-electoral coalitions in Norway, 1989–2019 Election date

Constituent parties

Type

Types of pre-electoral commitment

2005-09-12 2009-09-14 2013-09-09

A, SP, SF-SV A, SP, SF-SV A, SP, SF-SV

PEC PEC PEC

Other Other Other

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other

Government formation The bargaining process The Norwegian cabinet is still officially known as the King’s Council, and the constitution formally gives the monarch broad discretion to appoint its members. In practice, however, the king has exerted no influence on the composition of any cabinet since 1928.¹¹ When he formally calls upon someone to form a new government, the king always follows the advice of the leaders of the parliamentary parties. Because of the absence of formal rules and mechanisms, Norwegian government formation is best described as ‘freestyle bargaining’ among the parliamentary parties. The use of informateurs has no codified place in the Norwegian constitution and has occurred only once since 1945.¹² In practice, the choice of a prime minister (PM) has rarely been difficult. As Table 14.2¹³ shows, the initial attempt at cabinet formation has rarely been inconclusive. Due to the long-term bloc format, party cohesiveness, and the transparency of Norwegian coalition politics, bargaining has rarely been protracted or complicated. Still, difficulties have arisen: one example is the bargaining that eventually led to the creation of Kjell Magne Bondevik’s second government in 2001, when negotiations initially broke down as the parties could not agree on the prime ministership. After a break in the negotiations, however, the coalition parties ¹¹ In 1928, King Haakon VII, against the advice of the outgoing PM, called upon the Labour Party leader to form the first socialist government in Norwegian history (Björnberg 1939). The government proved short-lived but the king’s behaviour did much to solidify his support among Norwegian social democrats. ¹² In 1971, when a bourgeois majority coalition had just broken down over the EC issue, Storting president Bernt Ingvaldsen, a Conservative, was asked to explore the feasibility of another non-socialist coalition. Ingvaldsen’s role was purely that of an informateur. ¹³ Compiled with the help of news reports, negotiation protocols from 1989, 1997, and 2001. There is no indication that there was a new negotiation period when the (Willoch III) government continued after the 1985 elections. The reason is that the state archive only has the 1983 negotiation protocols, indicating that there were none in 1985 (as the state archive has made negotiation protocols in existence after 1971 public).

1989 1990 1993 1996 1997 2000 2001 2005 2009 2013

2018

2019

Syse Brundtland III Brundtland IV Jagland Bondevik I Stoltenberg I Bondevik II Stoltenberg II Stoltenberg III Solberg I

Solberg II

Solberg III

0

1

0 0 0 0 0 0 0 0 0 1

Number of inconclusive bargaining rounds

H, FRP (1) H, FRP, V, KRF H, FRP, V (1) H, FRP, V, KRF H, FRP, V, KRF

Parties involved in the previous bargaining rounds 22 0 0 0 21 0 8 18 10 7 15 13 7 0 (21)*

Bargaining duration of individual formation attempt (in days)

0

128

35 0 0 2 32 8 39 35 0 37

Number of days required in government formation

0 (21)*

116

22 0 0 0 21 0 8 18 10 22

Total bargaining duration

Note: *The parties negotiated for 21 days prior to the coalition enlargement. However, according to our coding rules the government formation period is zero days.

Year in

Cabinet

Table 14.2 Government formation period in Norway, 1989–2019

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



495

agreed that the Christian People’s Party would get the PM, and the Conservatives, as the largest party, would get both the Ministry of Finance and the Ministry of Foreign Affairs as ‘compensation’. This opened the door for formal bargaining (Conservative Party 12 September 2017; Liberal Party 14 September 2017; NTB 2001a, 2001b). Inconclusive bargaining rounds have happened on four occasions only, each time in an attempt to form a broad non-socialist coalition. Yet, notably this has happened after both of the two most recent parliamentary elections (2013 and 2017). Conservative PM Erna Solberg had a strong wish, both in 2013 and in 2017, to unite her own party in a four-party coalition with the Progress Party, the Christian People’s Party, and the Liberals. In 2013 the two centrist parties pulled out simultaneously after the initial talks but signed a comprehensive support agreement with the coalition partners. The bargaining process following the 2017 election was by far the longest bargaining process in history, lasting 128 days before the new Solberg cabinet could finally take office. Once again, a fourparty cabinet, in accordance with the PM’s wishes, proved impossible, mainly because the leader of the Christian People’s Party, Knut Arild Hareide, had promised ahead of the election that his party would not govern with the Progress Party. No other bargaining process has lasted more than about a month. Early in 2019, however, Solberg finally succeeded in uniting the Conservatives, the Liberals, the Progress Party, and the Christian People’s Party, in a majority, centre-right coalition. The backdrop was a leadership change in the Christian People’s Party in the autumn of 2018, which was sparked by the incumbent leader Knut Arild Hareide’s proposal to seek governmental power—for the very first time in the Christian People’s Party’s history—with the centre-left. The proposal was highly contested internally, as many party officers, among them the party’s two deputy leaders, believed the party should rather seek inclusion in Solberg’s incumbent centre-right government. In a dramatic vote in an extraordinary party congress, Hareide’s ‘red’ side lost to the ‘blue’ side, effectively paving the way for the party’s entry into Solberg’s coalition. According to our interviews, the bargaining processes in Norwegian cabinets after 2000 have nearly exclusively dealt with policy, and in comparison, minimal time has been spent on portfolio allocation, the distribution of other positions, or governance mechanisms (Conservative Party 12 September 2017; Liberal Party 14 September 2017; Centre Party 14 September 2017; Labour Party 06 October 2018; Progress Party 21 September 2017; Conservative Party 21 September 2017). When the red–green coalition partners negotiated the Soria Moria agreement in 2005, for instance, only the party leaders were involved in portfolio allocation negotiations (Centre Party 14 September 2017; Labour Party 06 October 2018), while in the Sundvolden negotiations in 2013 the portfolio allocations were discussed mainly after the policy document had been agreed upon (Progress Party 21 September 2017; Conservative Party 22 September 2017).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

496

  ø

The composition and size of cabinets The post-Second World War period has seen 18 Norwegian single-party administrations and 14 coalitions (see also Table 14.1a). Twelve cabinets have included representatives of parties that have collectively controlled a majority of the seats in the Storting, while 21 have been minority cabinets, including many of the non-socialist coalitions. Non-socialist cabinets have tended to be coalitions, and ten of the twelve have included at least three parties. There have been three types of non-socialist coalition: broad centreright governments, centre-right governments excluding the Centre Party, and two centrist coalitions consisting of the Christian People’s Party, the Centre Party, and the Liberals. Both of the latter (Korvald and Bondevik I) formed under circumstances in which dissent over European integration precluded coalescence between Conservatives and the other non-socialist parties, and specifically the Centre Party. Given the frequency of minority governments, most Norwegian cabinets have needed to attend carefully to their relationships with the formal opposition parties (Strøm 1990). The simple calculus of parliamentary decision-making has required most of these governments to seek outside support for their most important decisions. Contrary to Denmark, where formal—though often substantively limited—accommodations have been common, majority-building in Norway has been mostly ad hoc. Labour minority governments in particular have felt relatively secure in playing their left socialist opposition against the ‘bourgeois’ parties in order to reach acceptable agreements. Only two governments, both led by Conservative PMs (Willoch I and Solberg I), have had stable and prenegotiated support from parties outside the cabinet. Prior to its entry into the cabinet in 2013, the Progress Party had often been an unpredictable ad hoc partner for centre-right governments. The often uneasy relationship between the Progress Party and the centrist parties was further exacerbated by the latter’s moral (or moralistic) opposition to many of the populist appeals of the Progress Party. Until 2005, all Norwegian cabinets were either socialist or non-socialist. In fact, for a long time the Norwegian Labour Party was the only major social democratic party in Western Europe never to have entered a cabinet coalition with any bourgeois party. Labour also eschewed coalitions with any party to its left. Thus, a socialist government meant a cabinet of Labour alone and therefore between 1961 and 2005 also a minority cabinet. However, Labour’s massive 2001 electoral losses provoked a change in the party’s position. In the general election of 2005, Labour joined the Centre Party and the Socialist Left in a pre-electoral alliance. This alliance formalized into the first majority cabinet since 1985, as the three parties gained a parliamentary (though not a popular) majority and formed a ‘red–green’ coalition.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



497

Non-socialist minority cabinets have built majority support through a number of different mechanisms. One of the most critical occasions for such majoritybuilding is the annual budget process, which typically goes on more or less continuously through the fall session, with budget agreements often reached under severe time pressure just before the Christmas recess. From the early 1970s to 2005, budget negotiations were particularly challenging and high profile. Especially in the early part of this period, minority governments would often solve their budget problems through the so-called ‘slalom’ method, which meant darting from one legislative coalition to another on different budget issues. This form of budget-making often meant little overall coordination and impeded fiscal discipline. To avoid these problems, the Storting imposed on itself a series of procedural reforms aimed at streamlining the budgetary process and imposing a specific budget ceiling. While these reforms have met with some success, they have (predictably) not solved such underlying problems as the inherent parliamentary weakness of many minority governments or weakness of will on the part of politicians. While there is variation in budget coalitions (for an overview, see Narud and Valen 2007: 223), one interesting regularity is that governments typically prefer to negotiate their budgetary support towards the political right. Thus, Labour governments have much more often reached budget agreements with the centrist parties (especially the Christian People’s Party) than with the Left Socialists. Nonsocialists have tended not to reciprocate this favour. Centrist coalitions have tended to coalesce with the Conservatives, and centre-right coalitions (including the Conservatives) with the Progress Party. Deviations from this regularity appear to have been driven more by the government’s need to maintain credibility (‘we are open to cooperation with any opposition party’) than by the true inclinations of the incumbents.

The allocation of ministerial portfolios In partisan terms, Norwegian portfolio allocation was long highly predictable, conforming to the impressive Gamsonian regularity first described by Eric Browne and Mark Franklin (1973) as meeting the ‘interocular trauma test’ (hitting you between the eyes). The size of a coalition party’s parliamentary delegation was almost perfectly correlated with its number of cabinet members. Since the centreright Syse coalition of 1989–1990, however, this correlation has been much less perfect. The main deviation has been caused by smaller parties ‘punching well above their weight’. Thus, in 2001 the Liberals negotiated three cabinet seats in the Bondevik II coalition despite electing only two members of parliament. According to interviews, Liberal Party leader Lars Sponheim demanded at least three ministers, which he believed was the minimum necessary to enable the party to do a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

498

  ø

proper job in the cabinet and be an effective coalition partner (Liberal Party 14 September 2017). In subsequent coalitions, portfolio distribution has continued to favour the smaller parties above what would be expected under Gamson’s Law. The prime ministership and the Ministries of Finance and Foreign Affairs— particularly the first two—have (unsurprisingly) been perceived as the most prestigious portfolios in Norwegian cabinets (Socialist Left Party 20 September 2017; Liberal Party 14 September 2017; Conservative Party 12 September 2017; Centre Party 14 September 2017; Progress Party 21 September 2017; Conservative Party 22 September 2017). Interestingly, as Table 14.3 reports,¹⁴ in Stoltenberg’s second cabinet (2005) as well as in Solberg’s first, second, and third cabinets (2013, 2018, 2019), the Ministry of Finance was allocated to the party leader of the fringe party in the coalition: the Socialist Left and the Progress Party respectively. One Norwegian researcher has commented: ‘To the prime ministers Stoltenberg and Solberg this was a possibility to place the somewhat more rebellious little sisters in a post that would discipline them’ (Rongved 2018). The process of distributing ministerial portfolios is, according to interviewees, complex as different considerations clash: the allocation needs to take into account party strength, each party’s portfolio preferences, gender distribution, and regional representation. Concerning the latter, one interviewee states that it impossible not to include a cabinet member from the northernmost region of Norway (Conservative Party 12 September 2017). The negotiators have also needed to balance the distribution of the perceived ‘heavy’ ministries such as foreign affairs, education, or health, as well as the ‘light’ ministries, such as culture. Finally, the PM often wishes to have some of his or her close confidants in the cabinet (Conservative Party 15 September 2017). Generally, the allocation of ministerial portfolios has been overseen by the PM, who has had veto power over these appointments. Interviews confirm that both Bondevik (in his second cabinet) and Stoltenberg used this veto power to block certain ministerial appointments (Labour Party 06 October 2017; Conservative Party 15 September 2017). An aspect of portfolio allocation that has caused particular political interest has been the cabinet’s gender composition. Since Gro Harlem Brundtland’s second cabinet in 1986, Norwegian cabinets have been characterized by (and at least initially famous for) their virtual gender parity, and a reshuffle in 2007 resulted in the first female-majority cabinet. In the 1980s, this high proportion of women coexisted with rather ‘traditional’ appointment patterns, with female cabinet members assigned to portfolios such as family affairs, social affairs, development assistance, and increasingly justice. In recent years, this tendency has been reversed, with female cabinet members being appointed in such traditionally male posts as foreign affairs, finance, and defence. ¹⁴ The number of ministries and cabinet members reported in the table have been retrieved from www.regjeringen.no. See also Müller and Strøm (2000).

1989 1997 2001 2005 2009 2013 2018 2019

Syse Bondevik I Bondevik II Stoltenberg II Stoltenberg III Solberg I Solberg II Solberg III

9 H, 5 KRF, 5 SP 9 KRF, 6 SP, 4 V 10 H, 6 KRF, 3 V 10 A, 5 SF-SV, 4 SP 12 A, 5 SF-SV, 4 SP 11 H, 7 FRP 10 H, 7 FRP, 3 V 9 H, 7 FRP, 3 V, 3 KRF

Number of ministers per party (in descending order) 19 19 19 19 21 18 20 22

Total number of ministers 1 KRF, 1 H 2 SP, 2 V, 1 KRF 1 H, 1 KRF, 1 V 3 SP, 1 A, 1 SF-SV 4 SF-SV, 2 SP, 1 A 8 FRP, 3 H 3 H, 3 FRP, 3 V 3 H, 2 FRP, 2 V, 2 KRF

Number of watchdog junior ministers per party 19 17 20 17 18 18 16 16

Number of ministries H KRF KRF A A H H H

1 Prime minister

H SP H SF-SV A FRP FRP FRP

2 Finance

KRF KRF H A A H H H

3 Foreign affairs

H SP H A A FRP H H

4 Labour/ Social affairs

H KRF V A A FRP FRP FRP

5 Justice

Note: During Bondevik II, the number of ministries increased from 17 to 18, when the Ministry of Health was established on 1 January 2002. During Stoltenberg III, the number of ministries decreased to 16, as 2 ministries were dissolved. During Solberg I, a new Ministry for Immigration was established (Dec 2015), increasing the number of ministries to 19.

Year in

Cabinet

Table 14.3 Distribution of cabinet ministerships in Norwegian coalitions, 1989–2019

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

500

  ø

Alongside this development, the parliamentary seniority of incoming cabinet members has increased, and more appointees have had major prior appointments within parliament. Just as notably, the proportion of cabinet members with experience from a high-ranking office in their party’s youth organization has increased. If anything, Norwegian cabinet ministers have become more partisan, and the extent of parliamentary vetting has seemingly also increased. These background factors appear to have been prioritized at the expense of experience in subnational executive offices (e.g. mayors) and professional expertise.

Coalition agreements Since the first post-war coalition in 1963, Norwegian multiparty governments have taken office and governed on the basis of a coalition agreement, occasionally also based on additional agreements with support parties, such as was the case with Solberg I in 2013. The size of coalition agreements has grown steadily, from about 3,000 words under Lyng in 1963 to typically more than 20,000 words in the coalitions that have formed since 2000. As Table 14.4 shows,¹⁵ the most comprehensive coalition agreement to date is the one of Solberg’s third cabinet (2019), comprising about 40,000 words. Coalition agreements are almost exclusively policy documents; they normally contain very few procedural rules and hardly any commitments concerning the allocation of cabinet portfolios or other government posts. The latter fact, of course, does not mean that coalition payoffs do not matter to the political parties, but rather that such matters are handled in more informal ways.

Coalition governance The role of individual ministers in policy-making As mentioned at the outset, coalition governance in Norway is characterized by compromise, consensus, and collegiality and a lack of anything approaching jurisdictional dictatorship for individual ministers. Norwegian coalition governance, both presently and in the past, thus adheres to the coalition compromise model. Constitutionally speaking, the cabinet is a collegial organ of at least seven members (in addition to the PM) (Constitution, Art. 12), in which the PM does not have an elevated position compared to his ministers.¹⁶ Overall, the formal

¹⁵ Compiled based on coalition agreements. ¹⁶ In its operation, the Norwegian government has to abide by the constitution and the Rules for Cabinet (instruks for Regjeringen 1909), first passed in 1909 (Berggrav 1994).

Year in

1963 1965 1969 1972 1983 1985 1989 1997 2001 2005 2009 2013 2018 2019

Coalition

Lyng Borten I Borten II Korvald Willoch II Willoch III Syse Bondevik I Bondevik II Stoltenberg II Stoltenberg III Solberg I Solberg II Solberg III

2,941 4,456 5,996 2,919 25,137 31,138 6,650 20,240 15,721 28,111 22,274 25,467 28,399 40,059

Size 0 0 0 0 0 0 0 0 0 0 0 0.9 0 0

General rules (in %) 0 0 0 0 3.5 0 0.2 0 0 0.04 0.1 0 0.04 0.03

Policy-specific procedural rules (in %)

Table 14.4 Size and content of coalition agreements in Norway, 1945–2019

0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of offices (in %) 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of competences (in %)

100 100 100 100 96.5 100 99.8 100 96.7 99.3 98.5 98.6 99.6 99.7

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

502

  ø

powers of Norwegian government as laid out in the 1814 constitution have been described as comparatively weak (Chubb and Peterson 1989; see also Christensen and Lægreid 2002). The relevant regulations permit individual cabinets fairly wide latitude in designing their internal operations. Yet, most Norwegian governments since 1945 have functioned in similar ways (Berggrav 1994). The PM is the head of the Council of State (the formal title of the cabinet) and can, according to the Rules for Cabinet, demand any kind of information that he wishes from his cabinet colleagues and their civil servants (Prime Minister’s Office 1969). Yet, the Norwegian PM scores low on power compared to her opposite numbers in many other countries. Specifically, the Norwegian PM ‘only’ has the right to appoint ministers and set the agenda for cabinet conference meetings. The collegial nature of the cabinet also extends to the ministers themselves, as none of them enjoy full autonomy in their own areas of responsibility. Moreover, they are unable to opt out of deliberation in policy areas outside their own areas of expertise (Skjeie 2001: 157). There are two forms of cabinet meetings, the weekly (every Friday) and formalized Council of State meetings in which the king presides and the cabinet conferences, without the presence of the monarch. The latter are held once or twice a week, depending on the cabinet. Whereas Solberg’s first government met just once a week, cabinet conferences were held twice a week in both Bondevik II’s and Stoltenberg’s second and third cabinets. Cabinet conferences are not regulated in the constitution and their decisions are not official, though they are binding for members of the cabinet (Prime Minister’s Office 2018: 5). The rules regarding which matters have to be discussed in the cabinet conferences are vague. According to the guidelines (Prime Minister’s Office 2018: 8), such matters include ‘issues that have significant financial and administrative implications, issues that are politically difficult and cases where there is disagreement among ministers’. The individual ministers are responsible for preparing so-called R-notes, which are short documents (two to three pages) to be distributed to other concerned ministers. The Ministry of Finance is always a concerned ministry (Prime Minister’s Office 2018). According to some of the interviewees (Liberal Party 14 September 2017; Centre Party 14 September 2017), the comments from the concerned ministries determine how conflictual an issue is, which in the case of Bondevik II and Stoltenberg II and III would determine whether the cabinet conference, or some other conflict resolution mechanism, would be responsible for finding a solution.

Coalition governance in the executive arena The collegial nature of the Norwegian cabinets—which has been stable over time—means that the ministers are occupied in cabinet meetings several times a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



503

week: ‘To be a minister is to love the Meeting’, writes Skjeie (2001: 159). Interviewees state that in addition to the cabinet conference, there are also cabinet lunches (prior to the cabinet conferences), in addition to meetings in the cabinet sub-committee (for relevant members) and meetings in various ministry committees (Skjeie 2001; Prime Minister’s Office 2018; Progress Party 21 September 2017; Liberal Party 14 September 2017; Centre Party 14 September 2017; Conservative Party 12 September 2017). Ministers report that they spend around two-thirds of their time on specific matters related to their own ministry (Christensen and Lægreid 2002: 70; Statskonsult 2003). Respondents emphasize these different meetings, the extensive contact both between ministries and between senior and junior ministers during the process of policy development generally, and R-notes specifically, as core ways to avoid intra-coalition conflict (Conservative Party 12 September 1017; Progress Party 21 September 2017; Conservative Party 22 September 2017; Liberal Party 14 September 2017). More specifically, in Bondevik II, if there was indication (prior to a cabinet conference) that a matter was conflictual, the junior ministers in the PM’s office would contact the relevant ministers to try to solve the problem before it would reach the cabinet conference (Christian People’s Party 15 September 2017). Another mechanism was to solve issues in the cabinet lunch meeting (Liberal Party 14 September 2017), where questions could be asked and answered informally. These procedures were also common in the Stoltenberg and Solberg cabinets that succeeded Bondevik II (Centre Party 14 September 2017; Progress Party 22 September 2018). Should conflicts arise, cabinets in office after 1990 have had similar mechanisms to deal with them. Yet, their relative importance has varied in interesting ways. In the sister volume of this book, Narud and Strøm (2000: 178–9) observed that the most common form of conflict mechanism in Norwegian cabinets had been the cabinet conference itself, while the different forms of inner cabinet committees had dealt with the most serious policy conflict. In 1983, when PM Kåre Willoch formed a coalition cabinet composed of his own Conservative Party, the Christian People’s Party, and the Centre Party, he simultaneously created an inner cabinet comprising the three party leaders (including himself) and the minister of finance. From then on, all coalition cabinets have adopted a similar organ with a similar composition, which has been denominated the subcommittee, to deal with the more serious policy conflicts. There is broad evidence that from 1990 onwards, the cabinet conference and the cabinet sub-committee have been the two key arenas of conflict resolution, as reflected in Table 14.5.¹⁷ As important as the sub-committee has become, however, it is constitutionally a discretionary and flexible institution. Thus, no Norwegian coalition agreements

¹⁷ Compiled with the help of interviews and Müller and Strøm (2000).

1963 1965 1969 1972 1983 1985 1989 1997 2001 2005 2009 2013

Lyng Borten I Borten II Korvald Willoch II Willoch III Syse Bondevik I Bondevik II Stoltenberg II Stoltenberg III Solberg I

Y Y Y Y Y Y Y Y Y Y Y Y

N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A IC, CaC IC, CaC Pca, CaC IC, CaC IC, CaC IC, CaC IC, CaC IC, CaC IC, CaC IC, CaC IC, CaC

CaC CaC CaC CaC CaC CaC CaC CaC CaC CaC Cac

IC IC Pca IC IC IC IC IC IC IC IC

N N N N N N Y Y Y Y Y Y Y Y Y Y

Personal Issues union excluded from All used Most For most agenda common serious conflicts

Agreement Election Conflict management public rule mechanisms

All/Most All/Most All/Most All/Most Most/Most Most/Most Most/Most All/Most Most/All Most/Most Most/Most Most/All

Y Y Y Y N N Y N N N N N

Few Comp. Comp. Few Comp. Comp. Varied Comp. Comp. Comp. Comp. Comp.

Y Y Y Y N N Y N Y Y Y Y

N Y Y N N Y Y Y N Y Y Y

Coalition Freedom of Policy Junior Nondiscipline appointment agreement ministers cabinet in (watchdogs) positions legislation/ other parl. behaviour

Notes: Coalition agreement: IE = inter-election; PRE = pre-election; POST = post-election; PRE, POST = pre- and post-election Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee; Pca = Combination of cabinet members and parliamentarians Coalition discipline: All = Discipline always expected; Most = Discipline expected except on explicitly exempted matters Policy agreement: Few = Policy agreement on a few selected policies; Varied = Policy agreement on a non-comprehensive variety of policies; Comp. = Comprehensive policy agreement

IE PRE, POST PRE, POST POST PRE, POST PRE, POST PRE, POST PRE, POST POST POST POST POST

Year Coalition in agreement

Coalition

Table 14.5 Coalition governance mechanisms in Norwegian coalitions, 1945–2019

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



505

have specified any particular conflict resolution mechanisms, as these documents have almost without exceptions been about policy (see Table 14.4). Interestingly, there is evidence that the cabinet’s sub-committee has had a more prominent role in some coalition cabinets compared to others (see also Christensen and Lægreid 2002: 74; Kolltveit 2012, 2013). The increased use of the sub-committee was first detected in Bondevik’s first coalition cabinet (1997), a practice that was continued in his second coalition (2001). Then its use decidedly escalated under Stoltenberg’s two coalitions (2005–2013), while present PM Solberg (since 2013) has barely used this conflict resolution mechanism at all. In Bondevik’s first government, the frequency of meetings in the cabinet’s subcommittee increased, and the organ also decided on more controversial matters than before (Kolltveit 2014: 274). This led Christensen and Lægreid (2002: 74) to argue that the use of the sub-committee in Bondevik I contrasted with the consensual, collegial decision-making that had previously characterized Norwegian coalition cabinets. The development continued in Bondevik II.¹⁸ One minister from the Conservative Party (interview 12 September 2017) states that conflictual matters were passed on to the sub-committee, in addition to matters that were seen as so important that they would be discussed there before a cabinet conference. This view is supported by other interviewees (Christian People’s Party 15 September 2017; Conservative Party 15 September 2017). The increasing use of the sub-committee sparked internal criticism in both of Bondevik’s cabinets, as some members were unhappy that some matters were already decided before the cabinet conference. Some observers believed this development was undemocratic (Conservative Party 15 September 2017; Christensen and Lægreid 2002; Kolltveit 2013). The reason for this development can be linked to the composition of the cabinets, as the presence of the Conservative Party—instead of the Centre Party—made Bondevik II more conflictual than Bondevik I (Kolltveit 2014: 275). This explanation also resonates with the fact that in Bondevik II some issues were by mutual agreement excluded from the agenda (see Table 14.5). Specifically, the government’s Sem declaration stated that the EEA agreement would govern Norway’s relationship with the European Union, effectively precluding any EU debate from the agenda, and that the public broadcaster NRK would still be financed through a TV license fee imposed on all viewers. Stoltenberg’s red–green coalitions (2005–2013), comprising the Labour Party, the Socialist Left, and the Centre Party, were more conflictual than Bondevik II. This is evident from Table 14.5, as issues that were excluded from the agenda now included Norway’s continued membership in NATO and its nonmembership in the European Union. The conflicts between the coalition parties ¹⁸ In Bondevik II the sub-committee consisted of the three party leaders, the prime minister, and the finance minister.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

506

  ø

were, according to Kolltveit (2014), strengthened by the fact that the coalition parties got portfolios that strongly aligned with their prime policy preferences (thus, the Socialist Left got the minister of environment, for instance) and that it was a majority cabinet, which increased expectations of effectiveness both inside and outside the cabinet (Kolltveit 2014: 276). All in all, the conflictual nature of the coalition significantly shaped how conflicts were solved and turned the subcommittee into an ‘over-committee’, in the words of a former centrally located Labour Party source (interview 06 October 2017): ( . . . ) the ministers did not manage to reach a consensus, and then over time it was just very easy to leave the issue to the sub-committee. And there were endless meetings, endless meetings.

More specifically, under Stoltenberg the sub-committee took over as the main decision-making organ (Kolltveit 2014: 275). This was ‘both to relieve the cabinet conference and a way to conclude finally following a cabinet conference’ (Centre Party 14 September 2017), which led to criticism in the news media (Spence 2008; Clemet 2010). The sub-committee comprised the three party leaders, who were then involved in a very wide range of issues. This led to serious wear and tear on their cooperative efforts, and not least on the patience of PM Stoltenberg (Labour Party 06 October 2017). To compensate for this, a new conflict resolution mechanism was born in 2009 with the appointment of Karl Eirik Schjøtt-Pedersen as minister without portfolio at the PMs Office.¹⁹ The new arena for conflict resolution was named ‘Seminar Room 3’, after the room in the PM’s office in which the meetings were held. Meeting room 3 conferences included the relevant ministers and Schjøtt-Pedersen, in addition to (non-political) representatives from the PMs Office. Sometimes junior ministers and bureaucrats would sit in.²⁰ So, where previous cabinets had had just two main organs of conflict resolution, the red–green coalition now had three: the cabinet conference, meeting room 3, and the sub-committee. The latter kept its position as the conflict resolution mechanism for the most serious conflicts, while meeting room 3 became the most common arena for conflict management, for example to solve issues that required time, which was a scarcity in the cabinet conference (Socialist Left Party 20 September 2017; Centre Party 14 September 2017; Labour Party 06 October 2017). When Solberg and her Conservative Party won the parliamentary elections in 2013, the way to resolve cabinet conflicts reverted to what it had been prior to

¹⁹ Schjøtt Pedersen was a close confidant of the PM, and referred to by some as ‘Stoltenberg’s cerebellum’ (Boe Hornburg 2012). ²⁰ Note that over time, the sub-committee was formalized and thus had an agenda and notes that were prepared in advance of meetings (Kolltveit 2014: 276).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



507

Bondevik I. According to interviewees, the sub-committee was only summoned a few times during her first term and most issues were dealt with in the cabinet conference (Conservative Party 15 September 2017, Conservative Party 22 September 2017, Progress Party 21 September 2017). That said, throughout the first term (2013–2017), the two cabinet party leaders, Solberg and Finance Minister Siv Jensen from the Progress Party, consistently met prior to cabinet conferences to talk all the matters through and discuss how they could be solved (Conservative Party 22 September 2017; Progress Party 21 September 2017). So, while the role of the sub-committee became less prominent, another type of inner cabinet arose to ease deliberations in the ‘decision-making machine’, as the cabinet conference has been called (Conservative Party 22 September 2017). Similar to the development in conflict resolution mechanisms, the period after 1990 has also witnessed an interesting shift with regard to the appointment of junior ministers. Specifically, in nearly every cabinet since the 1990s, there has been an increasing number of junior ministers from a party that is not holding the cabinet ministership, so-called cross-partisan appointees (CPAs). This also means that the overall number of junior ministers has increased.²¹ While Bondevik II had CPAs only in the PM’s office and in the Ministry of Finance, Stoltenberg had them in 5 ministries between 2005 and 2009 and Solberg had 12 in her first government (2013–2017). Interestingly, however, Solberg had CPAs in only 6 and 5 ministries in her second and third cabinets respectively. Analysts have argued that CPAs can have four different roles, depending on whether these junior ministers are expected to perform monitoring functions and/or give cross-partisan advice to his or her minister: watchdog, liaison officer, regular political appointee, and coalition advisor (Askim et al. 2018).²² The interviews show that across all cabinets from Bondevik I through Solberg I, the CPAs have taken on all of these roles in their work. More specifically, interviewees see the rationale for CPA junior ministers as easing inter-party cooperation and coordination (Progress Party 22 September 2017; Centre Party 14 September 2017; Liberal Party 14 September 2017; Conservative Party 12 September 2017). Providing cross-partisan advice has consequently been one of the core rationales for the increasing number of CPAs. Fewer interviewees refer to the monitoring, or watchdog role of junior ministers, but some do. A representative from Solberg’s first cabinet states that it has been important for the party to have CPAs to monitor and make sure that the profile of the party

²¹ Before 1990 Norwegian cabinets generally included a fairly limited number of political appointees and generally each ministry had just one junior minister (Narud and Strøm 2000: 183). ²² A watchdog minister is not expected to give cross-partisan advice, for example provide information about his or her own party to the minister, but to monitor the minister and report back to his or her party superiors. A liaison officer is expected to both monitor and to give cross-partisan advice, while a coalition advisor provides cross-partisan advice but does not monitor. Finally, a regular political appointee neither provides cross-partisan advice nor monitors (Askim et al. 2018: 4).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

508

  ø

comes through (Progress Party 21 September 2017). These results are in line with those of Askim et al. (2018) who find that a majority of Norwegian CPAs can be described as coalition liaison officers, that is as ministers that combine the watchdog role with that of providing cross-partisan advice.

Governance mechanisms in the parliamentary arena There is broad agreement across interviewees from both minority and majority cabinets after 1990 that communication and collaboration with the Storting is of paramount importance for the cabinet. The cabinets have organized their formal contacts in similar ways: every week the individual ministers will meet with their own party’s members of parliament (MPs) in the party group meeting in parliament. Additionally, ministers meet with all the coalition’s representatives from their respective parliamentary committee, in so-called ‘fraction meetings’ (fraksjonsmøte): ‘These were important meetings, if you were not there, you made a blunder’ (Conservative Party 12 September 2017). Not surprisingly, for the ministers serving in minority cabinets, nurturing relationships with other parties’ representatives in the relevant committee is reportedly also a special priority. Such contact is often informal and can take the form of grabbing a cup of coffee or meeting over dinner (Christian People’s Party 15 September 2017; Conservative Party 15 September 2017). The cabinet’s relationship with the Storting was particular both in Stoltenberg’s two majority cabinets (2005–2013) and in Solberg’s minority cabinet (2013–2017), but for very different reasons. As Stoltenberg was heading a majority coalition, the locus of power shifted from the Storting to the cabinet. Interestingly, throughout the eight years that it was in power, the coalition’s way of consulting its MPs changed, partly because some MPs from the governing parties publicly opposed proposals from their own cabinet (e.g. see NRK 2005). Thus, instead of consulting the parliamentary party groups after the cabinet had started its decision-making processes, the group was consulted earlier²³—and this trend was even ‘institutionalised with own documents presenting the party factions’ views’ (Kolltveit 2014: 277; see also Conservative Party 15 September 2017). Under Solberg I, contact with the Storting was different for another reason. A formalized agreement existed between the government and its two support parties in parliament: the Christian People’s Party and the Liberal Party. The agreement was the result of initial cabinet formation talks following the 2013 parliamentary elections between the Liberal Party, the Christian People’s Party,

²³ Progress Party representatives have since 2013 been criticized for publicly opposing their own cabinet’s proposals (see e.g. Gitmark and Prestegård 2014). However, this has not altered the executive– legislative relations as in Stoltenberg’s cabinet.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



509

the Conservative Party, and the Progress Party. When the talks stranded and only the Conservatives and Progress Party decided to form a cabinet, the four parties still decided to cooperate in parliament, which resulted in the so-called Cooperation Agreement (Samarbeidsavtalen). Among other provisions, the agreement states that the PM is to invite the cooperating parties’ leaders to meetings about important matters and that the parliamentary party leaders are to meet regularly to discuss pending matters (Samarbeidsavtalen 2013). Besides the group and faction meetings, the four parties had additional contact arenas in order to secure the stability of this cooperation. Even though coalition agreements include no formal or explicit rules regulating voting over legislative proposals in parliament, there is a strong expectation that the MPs vote in line with their cabinet, as is shown in Table 14.5. Moreover, party cohesion is very strong in the Norwegian case, which has been related to the parties’ strict control over ballot access and state funding (Narud and Strøm 2000: 180).²⁴ Exceptions from coalition discipline happen in matters of great personal commitment, such as abortion policies, or in matters of geographical (regional/ local) importance. Adherence to the coalition line is also common in other parliamentary behaviour, for example in the oral question hour: ‘Generally there was an agreement that you would not surprise each other with questions in the Storting ( . . . )’ (Christian People’s Party 15 September 2017). According to the interviewees, if they had questions or required information about an issue, parliamentarians were encouraged to use back channels and contact their ministers and junior ministers personally (Conservative Party 12 September 2017; Centre Party 14 September 2017; Conservative Party 22 September 2017). As Table 14.5 shows, the discipline in parliamentary behaviour was absolute in Bondevik’s second and Solberg’s first cabinet. In Stoltenberg’s two coalition cabinets, however, the interviews indicate that such discipline was less rigid (Labour Party 06 October 2017) and that it did happen that MPs from one of the governing parties would decide to take a divergent stand on an issue in parliament (Socialist Left Party 20 September 2017).

Governance mechanisms with different types of actors (mixed) Reflecting the coalition compromise model, Norwegian coalition cabinets have generally made decisions in a collegial fashion in cabinet conferences or in the sub-committee. Other coalition committees have played a very limited role in decision-making (Statskonsult 2003: 81–4; Kolltveit 2013: 179–80), but as ministers often participate in committees such as the Standing Committee on Security

²⁴ Note that some state funding is also allocated to the parties’ subnational branches.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

510

  ø

or the Standing Committee on Research,²⁵ such committee participation matters for the individual minister’s time management (Skjeie 2001: 160). Still, decisions in such committees ‘have not been politically binding on the cabinet, which has meant that they have had limited importance’ (Kolltveit 2013: 179–80).

Cabinet duration and termination Norwegian cabinets have tended to be durable, in the sense that the majority have served out their maximum constitutional term, which for post-election cabinet is four years. As Table 14.6 outlines, only two of ten cabinets formed since 1990 have failed to do so: Brundtland IV in 1996 because of the PM’s voluntary retirement and Bondevik I in 2000 on a failed confidence motion introduced by the PM himself. Overwhelmingly, the main cause of termination for Norwegian cabinets has been regular parliamentary elections. Recall that all parliamentary elections in Norway are regular, since there is no provision for parliamentary dissolution and since there are no special elections for vacant seats due to the list-PR electoral system. While these facts point to a great and growing state of cabinet stability, this stability does not mean that incumbent governments lead sheltered lives. Since the 1970s it has become common in Norwegian journalism and political parlance to refer to governments suffering ‘wear and tear’ (slitasje). While this term is rarely defined with any precision, loss of popular support and more specifically anticipated electoral losses are surely a major part of this concern. The general recognition of this phenomenon has developed along with, and no doubt partly as a result of, a growing tendency for incumbent parties to suffer at the polls. Overall, Norway since 1945 exhibits a moderate adverse incumbency effect: a slight tendency for governing parties to lose vote shares when they face the voters. The mean net loss during 1945–2013 amounts to 2.0 per cent of the total national poll (Grindheim et al. 2017: 120), which is not significantly worse than for most comparable countries. Yet, there are differences across parties and cabinet types, and the scariest aspect (for party leaders) is the temporal trend. Coalition governments have tended to suffer larger losses than single-party administrations, and minority cabinets have done relatively better than those with a parliamentary majority. But by far the most striking pattern is what has happened over time. From the 1940s until the 1970s, governing parties were actually more likely to gain vote shares than to lose. Thus, Labour gained vote shares in its first three post-war ²⁵ The composition of the Standing Committee on Security may vary slightly from cabinet to cabinet but usually comprises the PM, the minister of foreign affairs, and the ministers of defence, justice, and finance.

2000-03-17 2001-10-19 2005-10-17 2009-09-14 2013-10-16 2018-01-17

Stoltenberg I Bondevik II Stoltenberg II Stoltenberg III Solberg I Solberg II

100 100 100 100 100 27.4

1 1 1 1 1 5

1 2 1 6

7

12

SP, H

Note: Relative duration: share of constitutionally allowed duration for this cabinet.

Voluntary retirement

EC membership

Comments

KRF became part of the coalition, thus forming Solberg III.

Environment Bondevik resigned after losing a vote of no confidence in Stortinget.

Foreign affairs

Policy Parties (when conflict between area(s) or within)

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister

2001-09-10 2005-09-12 2009-09-14 2013-09-09 2017-09-11 2019-01-22

100 77.6 100 61.4

1990-11-03 1993-09-13 1996-10-25 1997-10-17

Brundtland III Brundtland IV Jagland Bondevik I

1993-09-13 1996-10-23 1997-09-15 2000-03-09

1989-10-16 1990-11-03 26.8

Syse

Relative Mechanisms Terminal events duration (%) of cabinet termination

Date in

Cabinet

Date out

Table 14.6 Cabinet termination in Norway, 1989–2019

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

512

  ø

elections. Since that time, however, the prospects for governing parties have gotten progressively worse, and for every decade from the 1950s to the 2000s, the average vote share lost by incumbents grew monotonically. In the two elections in the 2010s, the average losses have declined slightly compared to the previous decades, but it is too early to know whether this constitutes any reversal of the downward trend. Thus, in the elections of 2001, 2005, and 2013, the governing parties suffered massive losses. In 2009 and 2017 the incumbent losses were much smaller, but even in the latter year every one of the four parties supporting Solberg I registered a net loss of vote shares.

Conclusions Norway is not the epitome of parliamentary democracy. In fact, the 1814 constitution was drawn up to establish a very different kind of regime, a separation-ofpowers system. Only gradually and informally was the constitution transformed into a parliamentary one. Central parliamentary principles, such as the cabinet’s obligation to inform the Storting and to resign after a vote of no confidence, were not included in the constitution until 2007. Moreover, the constitution still lacks some of the typical features of parliamentarism, such as parliamentary dissolution powers. Nonetheless, delegation of authority seems to follow the parliamentary pattern fairly closely. This is largely due to two interrelated factors: the strength of party government and a strong reliance on ex ante controls over agents. The seemingly incomplete and informal nature of Norwegian parliamentarism does not translate into a lack of stable governance or a growing susceptibility to political crises. On the contrary, Norwegian cabinets have since the early 1990s become increasingly stable, in the sense that eight out of ten cabinets formed since 1990 have served out their maximum term. The political leadership of the major parties has also since 2005 succeeded in incorporating two growing ‘fringe’ parties (the Socialist Left and the Progress Party) that had previously been considered non-coalitionable. As new parties have thus been incorporated and the incidence of coalitions has increased, new mechanisms of governance have evolved and existing institutions (such as junior ministers) have been strengthened. Most notably, this chapter has documented the increased use of the so-called subcommittee to solve serious conflicts within Bondevik I and II and Stoltenberg II and III and the increased tendency to appoint cross-partisan junior ministers from the 1990s onwards. Finally, patterns of portfolio allocation have certainly been changed so as to favour the smaller parties and deviate more clearly from Gamson’s well-established regularity. Despite these changes, the practice of Norwegian cabinets to make decisions in the weekly (or bi-weekly) cabinet conferences and the fact that Norwegian ministers lack full autonomy in their

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



513

daily operation have remained consistent over time. Thus, Norwegian coalition governance has been a stable expression of the coalition compromise model. The only development likely to disturb the Norwegian political establishment is the growing tendency for governing parties to be severely punished at the polls. Despite unprecedented prosperity, political stability, and political leaders most of whom have avoided even a whiff of political scandal, Norwegian voters have grown increasingly critical and fickle in their partisan attachments. While these trends have so far generated no systemic crisis, it remains to be seen to what extent Norwegian politicians will be able to respond to this challenge.

References Aardal, Bernt (2007). Norske velgere. En studie av stortingsvalget i 2005. Oslo: Damm forlag. Aardal, Bernt, and Henry Valen (1995). Konflikt og opinion. Oslo: NKS-forlag. Andenæs, Johs (1981). Statsforfatningen i Norge. Oslo: Tanum-Nordli. Arbeiderpartiet (2009). ‘Vedtaksprotokoll fra Det norske arbeiderpartis 62. ordinære landsmøte18.-21. april 2009 i Oslo Kongressenter, Folkets hus’. Congress minutes. Retrieved from: https://www.arbark.no/eldok/DNA2009_54.pdf (Accessed on 12 October 2018). Askim, Jostein, Rune Karlsen, and Kristoffer Kolltveit (2018). ‘The Spy Who Loved Me? Cross-partisans in the Core Executive’. Public Administration, 96(2): 243–58. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999-2010’. Party Politics, 21(1): 143–52. Benoit, Kenneth, and Michael Laver (2006). Party Policy in Modern Democracies. London: Routledge. Berggrav, Dag (1994). Slik styres Norge. Oslo: Schibsted. Bergman et al. (2003). Democratic Delegation and Accountability: Cross-national Patterns. Oxford: SMB, D&A, 109–220. Björnberg, Arne (1939). Parlamentarismens utveckling i Norge efter 1905. Uppsala: Almqvist & Wiksell. Boe Hornburg, Thomas (2012). ‘Bakmaktens mann’. Aftenposten. 13 August. Retrieved from: https://www.aftenposten.no/meninger/i/Ard3E/Bakmaktens-mann (Accessed on 26 April 2018). Browne, Eric C., and Mark Franklin (1973). ‘Aspects of Coalition Payoffs in European Parliamentary Democracies’. American Political Science Review, 67(2): 453–69. Christensen, Tom, and Per Lægreid (2002). Reformer og lederskap. Omstilling i den utøvende makt. Oslo: Universitetsforlaget.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

514

  ø

Chubb, John E., and Paul E. Peterson (1989). Can the Government Govern? Washington, DC: The Brookings Institution. Clemet, Kristin (2010). ‘Hvem styrer Norge?’. Aftenposten, 30 November. Forhandlingene mellom Kristelig Folkeparti, Senterpartiet og Venstre om dannelse av regjering (1997). ‘Minutes from Coalition Negotiations’. Retrieved from: https:// www.digitalarkivet.no/source/39783 (Accessed on 12 October 2018). Gitmark, Hannah, and Sofie Prestegård (2014). ‘Fremskrittspartiets to ansikter’. Dagsavisen. 21 February. Retrieved from: https://www.dagsavisen.no/innenriks/fremskrittspartietsto-ansikter-1.281493 (Accessed on 26 April 2018). Grindheim, Jan Erik, Knut Heidar, and Kaare Strøm (2017). Norsk politikk. Oslo: Universitetsforlaget. Hansen, Guttorm, and Erik Mo (1994). Om Stortingets arbeidsordning. Oslo: Stortinget. Instruks for regjeringen (1909). Retrieved from: https://lovdata.no/dokument/NL/lov/ 1909-03-23 (Accessed on 24 April 2018). Kolltveit, Kristoffer (2012). ‘Maktkonsentrasjon og politisk kontekst’. Norsk Statsvitenskapelig Tidsskrift, 28(1):1–24. Kolltveit, Kristoffer (2013). ‘Concentration of Decision-Making Power: Investigating the Role of the Norwegian Cabinet Subcommittee’. World Political Science, 9: 173–95. Kolltveit, Kristoffer (2014). ‘Concentration of Power in Cabinets: Exploring the Importance of the Party Political Context’. Acta Politica, 49: 266–85. Laver, Michael J., and Norman Schofield (1990). Multiparty Government: The Politics of Coalition in Europe. Oxford: Oxford University Press. Müller, Wolfgang C., and Kaare Strøm (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Narud, Hanne Marthe (1995). ‘Coalition Termination in Norway: Models and Cases’. Scandinavian Political Studies, 18(1): 1–24. Narud, Hanne Marthe (1996). ‘Party Policies and Government Accountability’. Party Politics, 2(4): 479–506. Narud, Hanne Marthe (2007). ‘Fra mindretallsregjering til flertallsregjering’. In Bernt Aardal (ed.), Norske velgere. En studie av stortingsvalget i 2005. Oslo: Damm forlag, 253–79. Narud, Hanne Marthe, and Bjørn Erik Rasch (2007). ‘Ideologisk avstander og regjeringskoalisjoner – om vilkårene for samarbeid til venstre og høyre’. In Per Kristen Mydske, Dag Harald Claes, and Amund Lie (eds), Nyliberalisme—ideer og politisk virkelighet. Oslo: Universitetsforlaget, 228–43. Narud, Hanne Marthe, and Henry Valen (2007). Demokrati og ansvar. Politisk representasjon i et flerpartisystem. Oslo: Damm.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



515

Narud, Hanne Marthe, and Kaare Strøm (2000). ‘Norway. A Fragile Coalitional Order’. In Wolfgang C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 158–91. Nordby, Trond (2000). I politikkens sentrum: Variasjoner i Stortingets makt 1814–2000. Oslo: Universitetsforlaget. Norwegian Constitution (1814). Retrieved from: https://lovdata.no/dokument/NL/lov/ 1814-05-17-nn#KAPITTEL_2 (Accessed on 24 April 2018). NRK (2005). ‘Demonstrerte mot egen regjering’. 9 November. Retrieved from: https:// www.nrk.no/norge/demonstrerte-mot-egen-regjering-1.517392 (Accessed 30 April 2018). NTB (2001a). ‘Bondevik statsminister om vel 14 dager’. 1 October. NTB (2001b). ‘Tirsdag starter regjeringsforhandlingene’. 1 October. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of AntiElitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Prime Minister’s Office [Statsministerens kontor] (1969). Instruks for regjeringen. Oslo: Prime Minister’s Office. Prime Minister’s Office (2018). Retrieved from: https://www.regjeringen.no/contentassets/ b2dd39c6c22e4d32b508e0a4d82ba914/no/pdfs/b-0504-b_om-rkonferanser.pdf (Accessed on 26 April 2018). Ray, Leonard, and Hanne Marthe Narud (2000). ‘Mapping the Norwegian Political Space: Some Findings from an Expert Survey’. Party Politics, 6(2): 225–39. Rokkan, Stein (1967). ‘Geography, Religion and Social Class: Cross-cutting Cleavages in Norwegian Politics’. In Seymor Martin Lipset and Stein Rokkan (eds), Party Systems and Voter Alignments. New York: Free Press, 367–444. Rokkan, Stein (1970). Citizens, Elections, Parties. New York: David McKay. Rongved, Gjermund (2018). ‘Sigbjørn Jensen’. Morgenbladet. 25 May. Retrieved from: https://morgenbladet.no/ideer/2018/05/sigbjorn-jensen (Accessed on 29 May 2018). Samarbeidsavtalen [The Cooperation Agreement] (2013). ‘Avtale mellom Venstre, Kristelig Folkeparti, Fremskrittspartiet og Høyre’. Retrieved from: https://www. aftenposten.no/norge/i/G1zbq/Her-er-samarbeidsavtalen-for-Hoyre_-FrP_-KrF-ogVenstre (Accessed on 29 May 2018). Sandberg, Nina C. Foss (2007). ‘Strategier i valgkamp: Arbeiderpartiet og Høyres strategier ved stortingsvalget i 2005 - var de vellykket?’. MA thesis. Oslo: Department of Political Science, University of Oslo. Skjeie, Hege (2001). ‘Inne i “beslutningsmaskinen”. Regjeringen som kollegium’. In Bent Sofus Tranøy and Øyvind Østerud (eds), Den fragmenterte staten: Reformer, makt og styring. Oslo: Gyldendal, 157–89. Spence, Thomas (2008). ‘Det rødgrønne Overhuset’. Aftenposten. 8 October.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

516

  ø

Statskonsult (2003). I Kongens navn. Den norske regjeringen i europeisk perspektiv. Oslo: Statskonsult. Strøm, Kaare (1990). Minority Government and Majority Rule. New York: Cambridge University Press. Strøm, Kaare (2000). ‘Delegation and Accountability in Parliamentary Democracies’. European Journal of Political Research, 37(3): 261–89. Strøm, Kaare, Wolfgang C. Muller, and Torbörn Bergman (eds.) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Svåsand, Lars, Kaare Strøm, and Bjørn E. Rasch (1997). ‘Party Organization’. In Kaare Strøm and Lars Svåsand (eds), Challenges to Political Parties: The Case of Norway. Ann Arbor: University of Michigan Press, 91–123. Valen, Henry, and Hanne Marthe Narud (2007a). ‘The Conditional Party Mandate. A Model for the Study of Political Representation’. European Journal of Political Research, 46(3): 293–318. Valen, Henry, and Hanne Marthe Narud (2007b). ‘The Storting Election in Norway, September 2005’. Electoral Studies, 26(1): 219–23. Valen, Henry, and Stein Rokkan (1974). ‘Norway: Conflict Structure and Mass Politics in a European Periphery’. In Richard Rose (ed.), Electoral Behavior: A Comparative Handbook. New York: Free Press, 315–70.

Appendix. List of political parties Abbreviation NKP

R SF-SV A V KRF SP DLF H FRP

Name Communist Party (Norges Kommunistiske Parti). Merged with other political forces to form the Red Party in 2007. Norwegian Communist Party (Norske kommunistiske parti) Red Party (Rødt) (since 2007; RV Red Electoral Alliance 1973–2007) Socialist Left Party (Sosialistisk venstreparti), from 1975 Socialist People’s Party (Sosialistisk folkeparti), 1961–1975 Labour Party (Det norske arbeiderpartiet) Liberal Party (Venstre) Christian People’s Party (Kristelig folkeparti) Centre Party (Senterpartiet) Liberal People’s Party (Det liberale folkeparti) Conservative Party (Høyre) Progress Party (Fremskrittspartiet)

Note: Party names are given in English, followed by the party name in Norwegian in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 15 Portugal: Left-Wing Single-Party Governments and Right-Wing Coalitions Patrícia Calca

Portugal was hit hard by the economic and financial crisis in 2010, after the global crisis of 2008. When a country is hit by an external shock, executives are expected to respond and adjust their political positions quickly—governments simply cannot wait and see (Calca and Gross 2019). This is what happened in Portugal. After international intervention and the European Union (EU)-and International Monetary Fund (IMF)-led rescue plan for Portugal’s sovereign debt crisis, Portugal became something of a ‘good student’ adhering to international recommendations, and several governments, both during and after the crisis, played an important role in this. In this respect, the crisis constitutes an important context for the understanding of governments’ actions, government formation, and coalition dynamics. Not at least as the last Portuguese coalition government during the observation period was in office while the impact of the crisis was at its peak, which heavily conditioned the politics of the country. Although the study of coalition governments is a recurrent theme in political science, research on Portuguese governments is, nevertheless, somewhat rare. For instance, data and important information about previous Portuguese coalitions (1975–1995) have been made available by Magone (2000). There are also a few more recent studies, for example Calca (2015a, 2015b, 2021) that analyses government legislative behaviour (1982–2009) in interaction with the Portuguese parliament and the president. Calca (2015b, 2021) shows that the status of the parliamentary majority defines the likelihood of issuing certain types of legislation. More specifically, she shows under which conditions a government is more likely to issue a government bill and an executive law. This chapter¹ draws upon the previous literature on the government and the parliament (Calca 2015b, 2019b, 2021) but also provides new data and interviews with major political actors concerning coalition governance and politics from 2000 to 2015.

¹ I thank Inês Carneiro for the committed and helpful research assistance regarding the interviews. I would also like to thank the former members of coalition governments in Portugal who agreed in contributing with their valuable testimonies for this chapter, as well as the PSD secretariat. Patrícia Calca, Portugal: Left-Wing Single-Party Governments and Right-Wing Coalitions In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0015

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

518



The institutional setting Portugal is a unicameral system in which the parliament (Assembleia da República) is composed of 230 deputies elected for a term of four years. The members of parliament (MPs) are directly elected in 22 (18 continental districts, two in the autonomous regions, one of Portuguese living in other European countries, and one of Portuguese living in the rest of the world) multimember districts with a variable district magnitude (i.e., a varying number of seats per district). In the closed list electoral system, the D’Hondt proportional method is used to calculate the number of elected MPs. The president of the republic is elected in a separate election. As the president is directly elected and that she nominates the prime minister, several scholars have considered the Portuguese system as a semi-presidential system. However, because of the way the state functions, especially in terms of the decision-making framework and legislative dynamics, it should be defined as a parliamentary system, with a strong executive power headed by the party (or parties) in office and with strong prime ministerial powers (Calca 2015b, 2021). At inauguration, before the government becomes a fully working body, it has to seek parliamentary approval for its programme. If parliament rejects the document the government cannot continue in office. Until parliamentary approval of the government programme, the executive can only deal with urgent and managerial/administrative issues. An example from 2015 is telling. The Portugal Ahead (Portugal à Frente, PàF) coalition of the Social Democratic Party (Partido Social Democrata, PSD)² and the Party of the Democratic Social Centre (Partido do Centro Democrático e Social, CDS)³ won a plurality of the votes, not the absolute majority, and was appointed into government by the president. That is, President Cavaco Silva invested Passos Coelho as the next prime minister. However, the Passos Coelho II government was short-lived. Following the nomination of Passos Coelho the parties on the left disagreed and stated that they would collaborate to form a leftist government. Thus, the government faced a left-leaning majority in parliament. Later a motion of rejection from the Socialist Party Partido Socialista (PS)⁴ received a parliament majority (123 against 107) and rejected the government programme, causing the very short tenure of the Passos Coelho II

² The PSD is acknowledged as liberal, especially in regard to economic aspects, but also conservative (Calca 2019b; Jalali 2015), occupying the centre-right. It is a member of the European People’s Party (https://www.epp.eu/parties-and-partners/) and of the Centrist Democrat International (https://idccdi.com/en/). ³ The CDS is presented as a Christian democratic and conservative party (https://www.cds.pt/). Like the PSD, it is a member of the European People’s Party (https://www.epp.eu/parties-and-partners/) and also belongs to the International Democrat Union (https://www.idu.org/members/). ⁴ The PS is acknowledged as social democrat (Jalali 2015), occupying the centre-left. It is a member of the Progressive Alliance of Socialists and Democrats (https://www.socialistsanddemocrats.eu/) because it belongs to the Party of European Socialists (https://www.pes.eu/en/).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



519

government. Afterwards the president had to excuse the government and invest Costa (PS) as the new prime minister. This type of opposition to a presidential nomination for government had occurred earlier. For example, in 2002, at the time of the formation of the Barroso’s government, there were three separate motions. Although none of the three motions was successful they still show the conflictual environment amongst the parties with parliamentary representation.

The party system and the actors Party system change In 1999, Portugal witnessed significant changes in the party system, with an additional party gaining seats in the parliament: the Left Bloc (Bloco de Esquerda, BE).⁵ Although not especially significant at the time, since BE won only two seats, this new party would change the dynamics of the party system, as BE continued to gain seats in future elections. BE’s number of seats would more than double in the next election—from three seats in 2002 to eight in 2005, and sixteen in the 2009 elections. In 2011, the number of seats for the party was reduced to eight, but in 2015 the party reached its highest number of MPs (19). In contrast, the Unitary Democratic Coalition (Coligação Democrática Unitária, CDU), formally a pre-electoral coalition between the Portuguese Communist Party⁶ (Partido Comunista Português, PCP), and the Ecologist Party⁷ (Partido Ecologista “Os Verdes”, PEV), had a more stable share of seats over the years. This is even more noticeable if compared it with the number of seats of the other left-aligned parties (BE and PS). The communists received a lower number of MPs (10) in the 2002 elections but recovered in the next elections. However, the vote share of the CDU is currently smaller than in former times. This is especially evident when comparing the period of democratic consolidation, before the fall of the Berlin Wall, and today. The PSD has been in office the most of all Portuguese parties—however, since the 2000s, as coalition governments and where two of three cabinets took office after the socialists prime ministers resignations. On the other side, PSD has had ⁵ The BE (https://www.bloco.org/) is a party resulting from the association of several Portuguese leftist movements, a founding member of the European Anti-Capitalist Left and of the Party of the European Left, and a member of the European United Left, now named The Left in the European Parliament (https://www.guengl.eu/). ⁶ The PCP (http://www.pcp.pt/) is a communist party and a member of the European United Left (https://www.guengl.eu/). It is represented in the Portuguese parliament in pre-electoral coalition (CDU) with PEV. ⁷ The PEV (http://www.osverdes.pt/) is a green and socialist party, being a member of the European Greens and a founder of the European Federation of Green Parties.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

520



more stable majorities then PS in the 2000s when it has governed in coalitions with the CDS, but this has raised distinctive challenges on its own. The PSD typically does well in the Portuguese elections and usually has the highest or the second-highest vote share of all parliamentary parties as the PS. Nonetheless, in the 2005 election the PSD had relatively the lowest number of MPs (75) in contrast to its main opponent, the PS, which in that year gained an absolute majority in parliament (121). The most right-wing party in the Portuguese political system (before the last elections of 2019), the CDS, has seen its number of mandates increase, in general, with the exception of the 2005 election. Nevertheless, the number of mandates is inferior (approximately by half) to what it used to be in the 1970s and the first half of the 1980s. Finally, the PàF, the coalition between the PSD and CDS, had a high vote share in the 2015 elections (102 plus five PSD members that did not run under the PSD label in the election) but still received fewer mandates than the sum for both parties in the previous elections in 2011 (132). Thus, it was not enough to guarantee a majority against the combined number of mandates of the BE, the PCP, the PEV and the PS. The conflict dimensions in Portugal are mainly left–right, but in many ways PSD and PS go hand in hand on the measures that they propose and take. However, during the financial crisis there were larger differences in the way they wanted to address it (economic liberalism), but the policy differences were not profound. The median party on the second policy dimension, presented in Table 15.1a, is the PS or the PSD. More specifically, when the PS was in office, the party was also the median on the second policy dimension, and the same occurred when the PSD was the government party. In 1999, when BE entered parliament, the left gained an additional member and diversity. From that point onwards, there were three left-wing and two right-wing parliamentary parties. Technically they are perhaps four left-wing parties, but given both parties’ strategies and behaviour, in this respect, I count PCP and PEV as belonging to one pre-electoral coalition—CDU. A novelty in the 2015 elections was the entry of a new party into parliament with one MP—The party, People, Animals, and Nature (Pessoas, Animais e Natureza, PAN),⁸ a party that wants to avoid being presented as either a left or right party.⁹ In 2015, while voting for the 2016 budget, the sole MP of PAN abstained; however, he would vote in favour of the PS budget in the following years together with PCP, PEV, and BE.

⁸ The PAN (https://www.pan.com.pt/) is a political party with a centre-left alignment. The party defends animal rights, environment, humanism, and pacifism. ⁹ As described in the appendix chapter of this volume, parties including only one MP are not included in the calculation of the effective number of parties.

1976-07-23 1978-01-23 1978-08-29 1978-11-22 1979-07-31 1980-01-03 1980-10-05 1981-01-09 1981-09-04 1983-06-09 1985-11-06 1987-08-17 1991-10-31 1995-10-28 1999-10-25 2002-04-06 2004-07-17 2005-03-12 2009-10-26 2011-06-21 2015-10-30 2015-11-26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 2005-02-20 2009-09-27 2011-06-05 2015-10-04

1983-04-25 1985-10-06 1987-07-19 1991-10-06 1995-10-01 1999-10-10 2002-03-17

1979-12-02 1980-10-05

1976-04-25

PS PS, CDS NN NN NN PSD, CDS, PPM PSD, CDS, PPM PSD, CDS, PPM PSD, CDS, PPM PS, PSD PSD PSD PSD PS PS PSD, CDS PSD, CDS PS PS PSD, CDS PSD, CDS PS

Election date Party composition of cabinet min mwc non non non mwc sur sur sur mwc min mwc mwc min min mwc mwc maj min mwc min min

40.7 56.7 N/A N/A N/A 51.2 53.6 53.6 53.6 70.4 35.2 59.2 58.7 48.7 50 51.7 51.7 52.6 42.2 57.4 46.5 37.4

Cabinet Type strength of cabinet in seats (%) 263 263 263 263 263 250 250 250 250 250 250 250 230 230 230 230 230 230 230 230 230 230

5 5 5 5 5 6 7 9 9 4 5 5 5 4 5 5 5 5 5 5 5 5

3.43 3.43 3.43 3.43 3.43 3.9 4.28 4.23 4.23 3.35 4.18 2.36 2.23 2.55 2.61 2.57 2.57 2.57 3.14 2.94 2.87 2.87

PS PS PS PS PS PSD PSD PSD PSD PS PRD PSD PSD PSD PS PSD PSD PS PS PSD PS PS

BE, PCP, PEV

Formal Median Number of Number of ENP, parties in parliament party in first support seats in parties policy parliament parliament dimension

Notes: For a list of parties, consult the chapter appendix. NN = independent Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1977–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015). The first policy dimension is economic left–right. The number of parties only includes parties that have won more than two seats in parliament during the period of observation. Cabinet types: min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition; sur = Surplus majority coalition. Table legends and variables are further defined in Appendix, this volume.

Soares I Soares II Nobre de Costa Mota Pinto Pintassilgo Sá Carneiro I Sá Carneiro II Balsemão I Balsemão II Soares III Cavaco Silva I Cavaco Silva II Cavaco Silva III Guterres I Guterres II Durão Barroso Santana Lopes Socrates I Socrates II Passos Coelho I Passos Coelho II Costa I

Date in

Cabinet Cabinet number

Table 15.1a Portuguese cabinets since 1976

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

522



Table 15.1b Electoral alliances and pre-electoral coalitions in Portugal, 1987–2018 Election date

Constituent parties

Type

1987-07-19 1991-10-06 1995-10-01

PCP, PEV PCP, PEV PCP, PEV PPM, MPT PCP, PEV PCP, PEV BE, UDP* PCP, PEV PSD, MPT, PPM PCP, PEV MPT, P.H. PCP, PEV PCP, PEV CDS, PPM** PTP, MAS PSD, CDS

EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC PEC EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC EA, PEC EA EA, PEC

1999-10-10 2002-03-17 2005-02-20 2009-09-27 2011-06-05 2015-10-04

Notes: * Only in Madeira ** Only in the Azores Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC)

Electoral alliances and pre-electoral coalitions As illustrated by Table 15.1b, electoral alliances and pre-electoral coalitions are common in Portuguese politics. Typically, during the electoral campaigns leading up to this, associations of parties have formed electoral alliances and pre-electoral coalitions (but not necessarily gained office). From the early 2000s onwards this kind of agreement was quite common. For example, the CDU (PCP and PEV) has been a pre-electoral coalition since the end of the 1980s, more specifically since 1987, and this agreement continues until today. There have been less frequent pre-electoral coalitions among other parties. In addition, they usually ended up not winning seats,¹⁰ with the exception of the 2015 preelectoral coalition PàF, which became the winner of the election (at least temporally). Typically, coalitions between PSD and CDS were agreed after elections rather than before. That is, both parties frequently have formed a post-electoral coalition

¹⁰ Of the several pre-electoral coalitions with no MPs elected there were cases of the coalitions between the BE and the Popular Democratic Union (União Democrática Popular, UDP) for Madeira (2002); Ecology and Humanism Front (Frente Ecologia e Humanismo, FEH) constituted by the Earth Party /Earth Movement Party (Partido da Terra/Movimento Partido da Terra, MPT) and the Humanist Party (Partido Humanista, PH) in 2009; the CDS and the People’s Monarchist Party (Partido Popular Monárquico, PPM) in Alliance Azores (Aliança Açores), 2015; and, AGIR, that is a coalition of the Portuguese Labour Party (Partido Trabalhista Português, PTP) (alone in the list to Madeira) and the Socialist Alternative Movement (Movimento Alternativa Socialista, MAS).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



523

with a parliamentary agreement. This happened for the elections of 2002, 2005, and 2011 and resulted in four PSD and CDS coalition governments. As previously mentioned, in 2015 PS became a single-party government without parliamentary majority. The party was not in an electoral alliance or with a pre- or post-electoral coalition, but it had the post-parliamentary support of the BE and CDU.

Government formation The bargaining process Since 1999, the Portuguese parliament has contained five parties. The two parties placed at the centre-right and -left, PSD and PS, do have some distinctive characteristics but also many similarities. In spite of their similarities, PSD and PS have chosen different patterns of government formation. For instance, PS was always the single party in government and PSD was always in a coalition with the CDS during the 2000s. In the Portuguese case there is no official or even informal information available regarding the separate bargaining processes or on the duration of these processes. This may be because coalitions for government formation are still rare or because the parties simply do not want to disclose the details of the negotiations per se. However, as stated in several of the interviews conducted for this chapter, it may be of importance to consider that most of these approaches are made informally. These informalities, in particular at the initial moments, make it hard to systematize and access information, which is also typically not publicly available. This also explains why there are some missing data in Table 15.2. Since the government of Guterres II, the average number of days for government formation is 25 days (or 27 days if we consider the period between 1987 and 2015). Nonetheless, the government of Santana Lopes is included in this calculation and at that time there were no intervening parliamentary elections, hence the time for that government formation was zero days.¹¹ A closer and deeper look into the time it takes to form a cabinet brings to surface some aspects worth mentioning. In the case of the Santana Lopes cabinet, the government formation process was swift since it was a continuation of the cabinet headed by Durão Barroso without any new elections. After the 2011 election, the first government formation by Passos Coelho took an unusually short time, less than three weeks (i.e., 16 days after the election). However, our cross-national definition of the days required for the government formation

¹¹ The president of the republic considered that the majority in parliament was still representative of the voters’ preferences. Thus, just the prime minister and the government changed but not the distribution of mandates in parliament.

1987 1991 1995 1999 2002 2004 2005 2009 2011 2015 2015

Cavaco Silva II Cavaco Silva III Guterres I Guterres II Durão Barroso Santana Lopes Socrates I Socrates II Passos Coelho I Passos Coelho II Costa I

PSD PSD PS PS PSD, CDS PSD, CDS PS PS PSD, CDS PSD, CDS PS

Parties involved in Bargaining duration the previous of individual bargaining rounds formation attempt (in days) 29 25 27 24 20 0 20 29 82 26 16

Number of days required in government formation

30 15 110

122 107

146 54 134 65 119 0

107 1 122 1

Pro Abstention Contra

Total Result of investiture vote bargaining duration

Note: The investiture vote concerns both motions of rejection on the government programme put forward by the opposition and votes of confidence put forward by the government concerning the same. Pro and contra votes concern whether the vote is supportive of the government (i.e., a no vote on a motion of rejection or a yes vote on a motion of confidence) or against (i.e., a yes vote on a motion of rejection or a no vote on a motion of confidence).

0 0 0 0 0 0 0 0 0 0 0

Year Number of in inconclusive bargaining rounds

Cabinet

Table 15.2 Cabinet formation in Portugal, 1987–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



525

begins when the previous government resigns or at the general election, whichever comes first. This is in contrast with how official records define a bargaining period. Counting our way, 82 days passed before the Passos Coelho I government could be formed (i.e., the bargaining period begins with the resignation of the Socrates II cabinet and ends with the formation of Passos Coelho’s first cabinet). Nonetheless, it is probably accurate to say that the government formation crisis in 2015 took more than a month. A total of 42 days (26 plus 16) passed before Portugal had a fully working government. In that period, two separate governments were formed, Passos Coelho II and the 2015 Costa cabinet. The period it took to form a government in 2015 is extensive compared to the average number of days for other government formations (not counting the total formation period for the Passos Coelho I cabinet). The timing for the coalition negotiations was a topic addressed in the interviews. On that basis, it is safe to state, in spite of the extended government formation period, that the actual negotiations for Passos Coelho I were short and the Programme of the Government was written in one week. Since ministers had already been named, the future ministers were asked about which issues they considered relevant and the information was then incorporated into the final government programme document. For this specific circumstance, but also for other coalition governments, the coalition partners already had well-known and pre-defined portfolio allocation preferences. Consequently, according to the interviewees, most discussions have been easy. Yet, there are also important examples of conflict between negotiating parties. For example, for the Passos Coelho I government, no proposals were made regarding electoral legislation, the privatization of the national television (RTP), the reduction of pensions, and respecting the election for the president of the parliament— because the CDS, the other party of the coalition, did not want to jeopardize the parliamentary vote support for Fernando Nobre, the leading PSD candidate for the presidency of the parliament. Although this resulted in non-action, Fernando Nobre was not elected. Some years earlier, during the Durão Barroso government, a hot topic in the coalition agenda was that one of the parties did not want to act on the law of pregnancy termination (abortion) and the coalition did not take action. Likewise, in the written agreements, especially for Passos Coelho I, the CDS managed to impose several questions that were important to them from an electoral perspective.

The composition and size of cabinets Portugal, after its initial moves towards democratization in the late 1970s and early 1980s, has been headed by governments leaning either towards the right or

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

526



left. From 2000 to 2015 just one single-party government had an absolute majority; all other cabinets were minority governments with external parliamentary support or majority coalitions. Since the beginning of the 2000s, Portugal has had eight governments—four single-party governments (three of which were minority governments) and four majority coalitions. The coalition governments comprised two parties: the PSD, which is centre-right, and the CDS, which at the time embodied the right. Since the 2000s, apart from the centre-right coalitions, the other party in government was the centre-left PS. Typically, the PSD and PS have rotated in power. The 2015 electoral results meant a change in the traditional patterns. Probably influenced by the economic and financial crisis and the external intervention by the Troika,¹² the elections reflected a shift back to the left of politics. A government headed by the PS was elected. Against all odds and traditional norms, it was the socialists, with only the second-highest vote share, that formed the government. The novelty was due to the support from other parties on the left. Thus, the BE together with PCP and PEV (CDU) supported a new left PS government without formally joining it. This four-party alliance was mainly made possible by BE’s electoral successes but also by the fact that voters supporting these leftist parties were discontent with the previous policies of the PSD and CDS coalition government. As a consequence the parties were more open to an agreement, in spite of their previous differences.

The allocation of ministerial portfolios Among the four coalition governments that formed in Portugal since the 2000s, the distribution of cabinet ministerships is quite diverse. Size-wise, the government of Santana Lopes with 20 ministers was the government with the highest number of ministers (Table 15.3). In contrast, the government of Passos Coelho I took office with 12 ministries, which makes it the government with the lowest number of ministers. When considering the distribution of ministers per coalition partner, we see what appears to be an improvement of the negotiation skills of the smaller partner, CDS. In 2002 CDS held two out of the 18 ministries. This increased to three out of the 12 with Passos Coelho I. If accounting for the total number of independent ministers, in 2011 CDS gained 25 per cent of the ministerial positions. Second, if not accounting for independent ministers, CDS gained 37.5 per cent of the ministerial positions. Either way, this relatively small party achieved to gain a high share of ministerial posts. ¹² Composed of the representatives of the European Commission (EC), International Monetary Fund (IMF), and the Central European Bank (CEB).

Year in

2002 2004

2011

2015

Cabinet

Durão Barroso Santana Lopes

Passos Coelho I

Passos Coelho II

16 PSD, 2 CDS 12 PSD, 3 CDS, 5 Ind. 5 PSD, 3 CDS, 4 Ind. 10 PSD, 4 CDS, 3 Ind.

Number of ministers per party (in descending order)

17

12

18 20

Total number of ministers

Number of watchdog junior ministers per party

16

14

18 20

Number of ministries

PSD

PSD

PSD PSD

1 Prime minister

Table 15.3 Distribution of cabinet ministerships in Portuguese coalitions, 1987–2018

PSD

Ind.

PSD CDS

2 Finance

PSD

CDS

PSD Ind.

3 Foreign affairs

CDS

CDS

CDS PSD

4 Social affairs

PSD

PSD

PSD Ind.

5 Interior

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

528



Independent ministers in Portuguese coalition governments represent an increasingly common reality. In the 2002 government, none of the ministers were independent, but in the next three coalition governments this changed. In 2004, five out of the 20 ministers were independents (20 per cent). In 2011, four out of the 12 ministers were independents (33.3 per cent). Thus, during the crisis, one out of three ministers did not have any party affiliation. Finally, in 2015 three out of the 17 ministers were not members of a political party (17.7 per cent). Regarding the allocation of ministerial portfolios in coalitions, it is important to reflect on the distinctive perspectives that diverse contextual settings generate. Hence, there are some pertinent points to consider. For example, in all coalition cases (the PSD and CDS) the prime minister represented the PSD, perhaps not unexpectedly since the party received the largest share of the popular vote. Out of the four coalition governments considered in this chapter, there were three different prime ministers. Just two were party leaders at the time of the legislative elections: Durão Barroso and Passos Coelho. The first was substituted without new elections and the latter was re-elected in 2015. In contrast, Passos Coelho II was an executive government without full powers and did not reach the two-week mark. Three ministers from the PSD have been in charge of the portfolio for finance. During the rule of the Santana Lopes government a minister representing the CDS headed the department. Finally, during the years of Passos Coelho I an independent headed the ministry. In the middle of the mandate of Passos Coelho I government, the leader of the CDS, Paulo Portas, took office as a deputy prime minister. The Ministries of Foreign Affairs, Environment, and Justice has merely once been headed by the minor coalition party (CDS). There are also ministries that when the party is in government are always allocated to the PSD such as the Ministries of Planning and Coordination, of Parliamentary Affairs, and of Industry and Energy. The Ministry of Social Affairs is usually the domain of the CDS. The party has held the office three times, by opposition of one time by the PSD. The Ministry of Education has once been held by the PSD; else it has been headed by independents. The Sea Ministry only existed for two terms of office and was during those years controlled by the CDS. Some observations emerged from the interviews regarding the allocation of ministerial portfolios. For example, it was easier to allocate the portfolios in 2011 than in 2002, as in the early 2000s the parties were less used to working together and lack of trust was an issue. One of the interviewees stated that in the coalition under the Durão Barroso government, the definition of portfolios was essentially decided by the PSD. This was agreed with the CDS leader Paulo Portas and later with Santana Lopes.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



529

Considering the governments of Passos Coelho I and II, the agreement between partners was negotiated by portfolio and not by ministers’ names. Apart from the Finance Ministry, which was an exigency of the PSD, both party leaders and ministers had the freedom to choose with whom to work. The broader restriction was that the allocation of portfolios should respect the division among parties previously agreed on. But for instance in the government of Passos Coelho I, Paulo Portas demonstrated a major interest in being the minister of foreign affairs. Moreover, agriculture and social security were secured by the CDS because of its particular interest in these policy areas. Especially in times of crises, finance and social security are very important ministries, and as a result the coalition decided to divide them between the two parties. Before, the CDS, while negotiating the coalition, pressured Durão Barroso to gain control of portfolios in sovereign areas such as defence, justice, and internal affairs. On the other hand, the PSD, specifically the finance minister, Manuela Ferreira Leite, wanted all the members of her team to belong to her party: the PSD. But, in general, there was an equilibrium between ministers and junior ministers from both parties. Where there were ministers of a given party, there should also be junior ministers from the other. For this government, the negotiation over portfolios was smooth. The CDS wanted to be in office; because the party had not been in a government since 1982, no CDS members except Bagão Félix had executive experience. On the contrary, the team of PSD had a great deal of experience. The Passos Coelho I government was headed by someone who did not have office experience, contrary to the leader of the coalition partner CDS (Paulo Portas). One of the interviewees believes that the CDS leader firstly regarded who he wanted in government and then negotiated the portfolios that were in his and his party’s interest. According to some interviewees, the actions of Paulo Portas were central to the governments of Passos Coelho I and II. Owing to Passos Coelho and his relative lack of experience, the smaller party in the coalition managed to run several relevant portfolios that are believed to have had important political outcomes. The specific cases were fiscal affairs, internal affairs, agriculture, and employment, as well as the direct contact with the Misericórdias.¹³ Other areas included public administration, private schools, social solidarity, and foreign affairs. Thus, Paulo Portas was an efficient leader who had central players in several influential ministries, for instance, placing the tax lawyer Paulo Núncio as junior minister of fiscal issues. The CDS also showed its strength by possessing

¹³ The Portuguese Misericórdias are religious associations—charities—with network characteristics. These charities have benevolent and compassionate aims and a strong geographical and cultural implementation as they are socially and financially relevant for Portuguese society. More information can be found at https://www.ump.pt/.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

530



veto power over certain ministries and decisions. A main area that the CDS leader was very interested in controlling was the Portuguese Trade and Investment Agency (AICEP). The agency was initially under the influence of the Ministry of Economics, but this shifted to the Foreign Affairs Ministry, which Paulo Portas controlled. During the crisis of 2013 the leader of the CDS asked for resignation. In response, the prime minister addressed the CDS concerns while maintaining an equilibrium in the portfolio allocation among the two parties. The strategy worked, and the coalition government lasted its full term in office.

Coalition agreements The size and content of the coalition agreements in Portugal differ from the West European average in being rather short and unspecific. Since the late 1990s, the coalition agreements have not, besides the main and general policy areas, contained any information on procedural rules, office, and competence distribution nor specific policies. Nevertheless, there is some information that can be gathered from the general size of the coalition agreements, and this is related to their number of words. For the four coalition agreements that have been written in the 2000s, the average number of words was 1,575, but there is some variation per agreement. Since the 2002 Durão Barroso cabinet, which had one of the shortest agreements of all times, the number of words in the coalition agreements slightly increased (from 1,468 words to 2,167 for Passos Coelho I). Then there was a distinct drop in length for the next cabinet. This drop, for the Passos Coelho II in 2015, may be due to a change to a calmer era in the general state of affairs, after the most intense parts of the political crisis (Table 15.4). As mentioned by several interviewees, in general the coalition agreements have been quite unspecific, being rather general and short. This seems to be more common when the two involved parties had divergences about specific points. For instance, in the negotiations of Passos Coelho I, some interviewees raised the point that the CDS had very strong and clear positions about precise topics. Because of this, there was an extra effort, especially from the PSD, to conciliate both parties’ positions on these issues. In this case it was not that the partners had an initial disagreement; conversely, they are quite close on aspects like the reinforcement of state authority, justice, or bank supervision, but it was relevant to both partners to state their political limits regarding these issues. However, this seems to have been very different a few years before: in the Durão Barroso government negotiations were almost non-existent, mainly because of the lack of experience of the CDS. In contrast to the PSD, the CDS did not have experience of how the government functioned.

Year in

1978 1980 1980 1981 1981 1983 2002 2004 2011 2015

Coalition

Soares II Sá Carneiro I Sá Carneiro II Balsemão I Balsemão II Soares III Durão Barroso Santana Lopes Passos Coelho I Passos Coelho II

4,476 34,300 34,300 34,300 2,461 2,461 1,468 1,905 2,167 758

Size 39 0 0 0 10.4 10.4 0 0 0 0

General rules (in %) 0.6 0 0 0 14.2 14.2 0 0 0 0

Policy-specific procedural rules (in %)

Table 15.4 Size and content of coalition agreements in Portugal, 1976–2018

0 0 0 0 4.5 4.5 0 0 0 0

Distribution of offices (in %) 0 0 0 0 0 0 0 0 0 0

Distribution of competences (in %)

60.4 100 100 100 70.9 70.9 100 100 100 100

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

532



According to several of the interviewees, out of all the coalition agreements, the Durão Barroso agreement, the longest one, was the easiest to write. This may have been because it was a stable time in the country’s history. During times without pressing economic problems, the government programme was what guided the coalition. The programme reflected the coalition’s own intentions and was accepted by parliament. In the case of the Passos Coelho I government, there were other aspects that some interviewees underlined, namely that some of the information on the country was not publicly available at the time for the preparation of the agreement and the government programme. Some proposals could not be implemented because they were planned without full knowledge of the situation and given the needed economic measures as well as to decisions made by the Constitutional Court. It seems like a few of these non-expected situations obliged the government to adjust and change some previously defined policy choices. During the interviews, several argued that the coalition agreement is a good indicator for the future behaviour of the governing parties. At the same time, the notion that the formal government programme is probably seen as the most important document was highlighted several times. Nevertheless, there were cases where disagreement among partners was strong and they could not agree on a common direction, for instance on issues such as the Electoral Law, the privatization of the Portuguese television (RTP) and the vote for the president of the parliament. These issues were then left out of the coalition agreement.

Coalition governance The role of individual ministers in policy-making The information on the role of the individual ministers in the four coalitions was mainly obtained from interview questions regarding the role of each government cabinet. More specifically, for the Portuguese case, the coordination of executive policy-making is conducted by the cabinet, the Council of Ministers (CM). Several interviewees declared that the agreement and coordination among members of government is made in the council. Usually, the way to achieve agreements is by discussing the topics and not voting on issues in the CM. The prime minister would normally consider several opinions before deciding. When a difficult question arose in a meeting, even if it was not on the agenda, there would be time to discuss it. If in the CM an agreement on a issue could not be reached, the decision would be put on hold until there was a greater prospect of agreement. As a last resort, to avoid major problems within the coalition, a given decision would not be advanced if it was too controversial. At times meetings took place only among the involved ministers, in order to achieve a common position.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



533

In the Passos Coelho I government the minister of the presidency of the CM—a minister in charge of the proceedings of the CM, a position not used by all cabinets—was Luís Marques Guedes, who coordinated the agenda with issues brought by several ministers. One of the interviewees stated that in some governments, when the questions were very technical, they were first discussed in meetings with junior ministers. Thus, the junior ministers’ meetings functioned as a place for discussing political issues before passing them on, after the technical details had been sorted out, to be actioned by the CM. The junior ministers’ meetings were open-ended, sometimes lasting nine to 10 hours—all for the sake of achieving an agreement. There have also been many examples of coordinated position-taking. The Durão Barroso government made sure to coordinate ministers, junior ministers, and MPs before debates. Later, the Passos Coelho I cabinet held regular meetings among all ministers, parliamentary meetings, and conferences as well as visits, dinners, and other social happenings. Several interviewees motioned that personal relationships were relevant and even vital in the maintenance of the coalitions, as one of the interviewees stated: ‘coalition agreements and government programmes do not substitute personal relationships.’ During the Passos Coelho I government, there was a high degree of formality and systematization for the strategies taken by the coalition. Here, the junior ministers’ meetings, which according to an interviewee defined and controlled the government functioning, were a good example of more formal ways of coordinating efforts. There was, nevertheless, a constant and less formal dialogue between the two party leaders and the coordination teams of each of the ministries. Yet, complications existed. One interviewee stated, regarding the Passos Coelho I government, there were four distinct ‘government’ sections: 1) a government of the PSD; 2) a government of the CDS; 3) a government of Troika; and 4) a government that existed in the first two years (2011–2013) with external people who had great influence on the prime minister like the economists and university Professors António Borges and Braga de Macedo as well as certain ministers. There were also regular meetings between Paulo Portas and the ministers of the CDS, which did not include PSD members. These statements indicate division within the coalition, or, if one prefers, the creation of sub-groups acting alongside the government. The interviews did not reveal how these ‘moving parts’ relate to each other and which groups had higher internal hierarchical power than others, which nonetheless seems to have been the case. When looking at the Portuguese example concerning the specific roles of ministers, junior ministers, and the prime minister in the policy-making stages it is hard not to conclude that the main coalition government model applicable to this case is the Dominant Prime Minister Model. In the Portuguese version of the model, the one who controls all the domains of the governance process is the prime minister and, more broadly, the dominant party.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

534



However, there are some variations among cabinets. In interviews it was stated that during the years of Passos Coelho I, a great many conversations and coordination took place among representatives of either PSD or CDS. Yet, this level of coordination and negotiation is not as constant or equilibrated as one would expect in the Coalition Compromise Model, which entails negotiations of the entire governmental action as well as extensive control mechanisms between the parties involved. Moreover, the Ministerial Government Model is also not applicable to the Portuguese example. Effectively there is no division of power between the parties in government but a differentiated share of power where the main party and thus the prime minister have a higher share of responsibilities and prerogatives.

Governance mechanisms in the parliamentary arena The conflict management mechanisms used in each of the governmental coalitions were different. Especially between the first, second, and third coalition governments the ways to deal with conflicts were related to the specific characteristics of each government. Table 15.5 illustrates this variation. When Durão Barroso headed the coalition as prime minister there was an inner cabinet of ministers that managed government issues. Santana Lopes also had an inner cabinet, as did the two governments headed by Passos Coelho. The two cabinets also had a party summit, which included the smaller partner of the coalition, the CDS. According to the interviewees, the governments’ way of solving disagreements was to let the inner cabinet deal with conflicts. In most coalitions party discipline has been standard. There is an understanding that the MPs should vote the same way in areas where there was no enforced and explicit party discipline. The advantages are seen to be twofold: it is both for the good of the coalition and for the good of the party. However, from the interviews it is possible to conclude that party discipline was not enforced by strict monitoring. None of the interviewees declared this as necessary. Coalition governments had a good relationship between the members of government, parliamentary group leaders, and MPs. In addition, in the PSD there are several aspects, which are exempted from discipline of vote. On some issues, related to societal matters or when the party does not have a common and defined position, the MPs have greater freedom to vote according to their own preferences. Concerning the Passos Coelho government, some interviewees underlined that voting discipline was actually called upon in situations where the coalition agreement was invoked. On important issues like the national health system reforms, party discipline had to be used as they came to have a big impact on overall government performance. All issues that generally and specifically dealt with

2015 PRE

Passos Coelho II

Y

Y Y

Y Y Y Y Y Y Y

N

N N

N N N N N N N

CoC, PS CaC, PS CaC, PS CaC, PS CaC, PS CoC, PS IC, CaC, Pca CaC IC, CaC, PS (CDS) IC, CaC, PS (CDS) IC

IC IC

CoC CaC CaC CaC CaC CoC IC

IC

IC IC

PS PS PS PS PS PS IC

Y

Y Y

Y

Y

Y Y

Y

Y Y Y Y Y Y Y

Most/Most Y

Most/Most Y Most/Most Y

Spec./Spec. Most/Spec. Most/Spec. Most/Spec. Most/Spec. Most/Spec. Most/Most

Comp.

Comp. Comp.

Comp. Comp. Comp. Comp. Comp. Comp. Comp.

Y

Y Y

N N N N N Y Y

-

-

N N N N N N -

Coalition Freedom of Policy Junior Nondiscipline appointment agreement ministers cabinet in (watchdogs) positions legislation/ other parl. behaviour

Notes: Coalition agreement: PRE = pre-election; IE = inter-election; POST = post-election Conflict management mechanisms: CaC = Cabinet committee; CoC = Coalition committee; Parl = Parliamentary leaders; PS = Party summit Coalition discipline: Most = Coalition discipline is expected on all matters except those explicitly exempted Policy agreement: Comp. = Comprehensive policy agreement

2004 IE 2011 POST

Santana Lopes Passos Coelho I

IE PRE PRE PRE PRE POST POST

1978 1980 1980 1981 1981 1983 2002

All used

Personal Issues union excluded from For most Most agenda common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Soares II Sá Carneiro I Sá Carneiro II Balsemão I Balsemão II Soares III Durão Barroso

Coalition

Table 15.5 Coalition governance mechanisms in Portuguese coalitions, 1944–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

536



finance, health, work, or territorial aspects were usually considered to fall under strict voting discipline. Nevertheless, the coordination among MPs of both parties was declared as straightforward; therefore, usually there was no need to advocate party discipline. One of the main difficulties was that MPs from autonomous regions (i.e., Madeira and Azores) usually had somewhat different incentives behind their actions at the national level (Calca and Ruel, 2018). Actually, this reality has already been emphasized in other research, for instance Ruel (2017, 2020), who argues that the MPs elected by the regions have strong bargaining powers, especially because their votes are highly relevant to the approval (or not) of the national budget. That is, these MPs have a stronghold on what are the calculus of voting totals given the close seat distribution between the left and right and the dynamics of the system until then.

Governance mechanisms with different types of actors During the last few years of the observation period, there were three coalitions in Portugal and a pre- and post-electoral coalition that did not become a government coalition, at least not a fully invested one. In 2002 and 2011 the PSD and CDS had a post-electoral agreement to form a government with members of both parties, but with the 2015 elections, PSD and CDS went to elections with PàF, a pre-electoral agreement. The electoral results were positive to the coalition as they gained the higher number of seats but without a parliamentary majority. The majority of mandates was held by the centre-left and left parties (BE, PCP, PEV, and PS) that agreed to go against a right-wing government. The government did not survive even two weeks and the president had to finally, and against his first action, to invested Costa from PS as the prime minister. In regard to coding (see Chapter 3, this volume), a party is considered a parliamentary one if it has at least three MPs with seats. Two of the three party lists represented in parliament (BE, PAN, and PEV) in the last few years are not listed as separate party groups. However, since 2002 BE has gained three MPs. In the last elections in 2015, another party gained one MP, the PAN, but for our purposes and based on our definition only a total number of five parties can be considered stable in this party system for the elections of 2015. The conflict resolution mechanisms between the coalition members relied on the CM for this last government. Likewise, in coordination meetings between both party leaders, among parliamentary leaders, and in meetings with the minister of parliamentary affairs, the parties used their own conversations and negotiation processes as conflict resolution mechanisms. However, what most of the interviewees underlined were the personal relationships of trust among government

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



537

members for both the PSD and CDS. That is, there was a continuous attempt to mitigate conflict through one-on-one conversations and there were attempts to create consensual behaviour. Durão Barroso and Passos Coelho I governments had political coordination meetings every Monday morning. Ministers were invited to the meetings if the content of their portfolio matched the topic of the day. Existing conflicts were reduced in prolonged discussions. Specific rules were pre-established, which turned out to be helpful in solving problems. When the issues were very socially salient, the CDS would be less insistent on public position taking, perhaps because it was the smaller party in the coalition. Some of the conflicts, for instance with the Durão Barroso government, originated from a lack of trust among PSD members regarding Paulo Portas. However, Durão Barroso and Paulo Portas maintained a good personal relation and, one salient case, Universidade Moderna did not affect the coalition too deeply.¹⁴ Conversely, according to one of the interviewees, the personal relationship between Passos Coelho and Paulo Portas was never as good as the one between Durão Barroso and Paulo Portas. Relationships among other members of the two parties and even within government were also strained at some point. The interviewees stated that there was no Plan B if the coalition fell apart. Nonetheless, there was a general notion that in the end the MPs would have to decide the future of the coalition. The top-level negotiation was also important. In the middle of the term, the Passos Coelho I coalition almost collapsed; there were conflicts because of the allocation of structural funds. The collapse was avoided because the prime minister was determined to find a solution with the leader of the CDS. It is known that there were also crises before 2013 regarding the rectification budget; early elections were a distinct possibility. In the end, this did not happen; the president decided against them after consulting the PS.

Cabinet duration and termination The duration of cabinets Since the mid-2000s, instability seems to be the norm in the Portuguese political arena. The average maximum duration for governments is around 1189 days. However, the effective number of days in office, since 2000, is higher, around 1263 days. There are differences between the duration of governments, comparing the maximum and minimum number of days in office that can prove significant. The average government duration is almost half of the total maximum tenure ¹⁴ This situation addresses a criminal case about a private university and involved past and current members of the government.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

538



(approximately 645 days) and this is higher when dealing with data collected after the year 2000 (close to 812 days in office). Taking into consideration the latter number it should be underlined that, especially in the last government for which we have coded the termination date, the duration was only 11 days. Differences can be detected when comparing governmental duration in the 1980s and 1990s, but the spectrum somehow changed in the 2000s, with two prime ministers (Guterres II and Sócrates II) asking for resignation before the constitutionally mandated term had ended and Durrão Barroso leaving government and being replaced by Santana Lopes. Finally, following the elections of 2015 the government was in office for 11 days, with very limited powers (Passos Coelho II). The short number of days in office has an impact on the average calculus. The government that stayed longest in office since the 2000s was Sócrates I, 1160 days, and the executive with fewest days in office was Passos Coelho II, with 11. In terms of the relative duration of these cabinets it is evident that since the 2000s only two governments served the full constitutionally mandated maximum term, that is, they had a 100 per cent relative duration, Sócrates I and Passos Coelho I. On very distinct levels we find Durão Barroso with more than a half of the percentage (57.8 per cent), then a bit higher than the 30 per cent mark appears Santana Lopes (35.9 per cent) and Sócrates II (36.4 per cent), and finally the lowest Passos Coelho II with 0.8 per cent. Context, institutional dynamics, and additional events (that may be unexpected and exogenous) influence the duration of coalition governments in Portugal. The institutional context, a minority or majority government, and the stability of a coalition seem to be main aspects in this calculus.

The termination of cabinets The reasons for early termination of a governmental mandate in Portugal are diverse and do not show a clear pattern. Nonetheless, as Table 15.6 reveals, a closer look at these features does illustrate some interesting patterns, for instance, the general results in second-tier elections—local elections—or a political crisis immediately following the local elections of December 2001. In the latter case, there was a minority government in office, a government with known difficulties in bargaining at the parliamentary level. The governance context for this executive was quite particular and the result of elections where 115 MPs for PS, which was exactly half of the mandates, making the party dependent on the goodwill of at least one oppositional parliamentarian. The reasons for cabinet termination in the last few years have been distributed almost evenly. There are three cases of an early parliamentary election that ended the Guterres II, the Santana Lopes, and the Sócrates II cabinets; however, the reasons were different to other cases. One concerned the prime minister resigning

2015-10-04 2015-11-26

2019-10-06

2004-07-17

2005-03-12 2009-10-26

Santana Lopes

Socrates I Socrates II

Passos Coelho I 2011-06-21 Passos Coelho II 2015-10-30

Costa I

2004-12-13

2004-07-17

99.4

100 0.8

100 36.4

35.9

57.8

100 100 99.4 89.7%

1

1 6

1 6

4

9

1 1 1 4

11

10

Relative Mechanisms Terminal Parties (when duration of cabinet events conflict (%) termination between or within)

Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event Note: Relative duration: share of constitutionally allowed duration for this cabinet.

The government’s programme was defeated in a motion of rejection introduced by PS, and were thus the PàF (the PSD and the CDS) coalition was exonerated by the president.

Socrates resigned after the government lost an important vote in parliament on new austerity measures.

After PS suffered a crushing defeat in the municipal elections held in late 2001, Guterres resigned as prime minister and leader of PS, and early elections were called by President Sampaio. Barroso stepped down in order to take office as the president of the European Commission. The legislature was dissolved and early elections were called after Santana Lopes had publicly expressed his wish to see it dissolved two weeks earlier. A lack of popular support was one of the reasons.

Policy Comments area(s)

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister

2015-11-26

2009-09-27 2011-03-31

2002-04-06

Durão Barroso

1991-10-06 1995-10-01 1999-10-01 2001-12-17

1987-08-17 1991-10-31 1995-10-28 1999-10-25

Cavaco Silva II Cavaco Silva III Guterres I Guterres II

Date out

Date in

Cabinet

Table 15.6 Cabinet termination in Portugal, 1987–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

540



after bad local election results, the other, a presidential dismissal of the government, and the last one, a prime minister resignation after a parliamentary rejection of a plan to deal with the financial crisis. The termination of the Santana Lopes government has very specific characteristics as it implied presidential action. The situation was partly due to events that can be traced back to the Durão Barroso government. In 2004 Durão Barroso decided to accept appointment to the presidency of the European Commission, which led to office Santana Lopes (the vice deputy of the PSD), but the government would be dismissed after the president Jorge Sampaio decided to dissolve the parliament. It seems that bad executive decisions and a lack of popular support for the cabinet were reasons presented by the president to dissolve the parliament. The results of the 2005 elections saw the PS regain office with an absolute majority. In this first mandate, Sócrates would serve the full four years and be reelected in 2009. Conversely, during the second term, the PS was in a minority government and thus in a more fragile situation, needing to negotiate much more than before. The bargaining seems to have been unsuccessful, and in 2011, Sócrates resigned. The reason given was a non-agreement in parliament with the aim of approving special measures to avoid a Troika intervention in the country. The 2011 elections brought the country back to a centre-right/right government consisting of the PSD and CDS. There are some interesting differences to the previous cases of the Santana Lopes (a coalition government) and Sócrates II governments (minority singleparty government). In the first case the parliament was dissolved by the president and thus the legislature and the government ended, and in the second case, the prime minister made a decision to resign after the failure of parliament to approve the non-intervention of Troika via a rescue plan. The Passos Coelho II government failed to get approval of its programme of government. As a result, according to the Portuguese Constitution, the cabinet had to be dismissed and the term in office became short. The programme of government was defeated in the approval of a motion of rejection introduced by the PS, forcing the government to resign. Technically the government had to resign and the president present this dismissing as the Constitution indicates (Constituição da República Portuguesa, Art. 133, g)). The motion of rejection implies that the government cannot be in office, but an official dismissal can only be made by the president as it is the president that, from the legislative electoral results and with the parliament indication, invites the party that won the elections to form government. The other two cases (Santana Lopes and Sócrates II governments) met new elections at their termination. Once more, the outcome of these elections was very much the result of the will to vote on the prime minister, even when the legislative elections concerned mean to vote for members of parliament and not for the government or the prime minister.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



541

Conclusions From 2000 to 2015 Portugal had eight governments, four single-party and four coalition governments. The coalitions in Portugal during this period were formed by the same two parties (the PSD and the CDS), and some actors were the same in several governments. The country would later experience a turn to a quasicoalition towards the left; however, examining this would go beyond the scope (and observation period) of this chapter. During the period that this chapter focuses on, there was one party that gained the status of parliamentary party group and even managed to win 19 MPs in 2015, which is a big increase over time (BE). Furthermore, on the left we have the CDU, a pre-electoral coalition between the communists and the greens, with fewer MPs but also with a stable electoral return over the years. Portugal has not had too many stable governments in the 2000s. There have been several reasons for early government dissolution. The inner dynamics of Portuguese governments is closer to a black box than crystal clear. Most information, at this level, gathered for this chapter could only be accessed by conducting interviews. It was possible to conclude that the coordination of executive policy-making was done at the presidency of the CM, but with some variations either in the earlier process or in the post-CM process, meaning it was mainly dependent on the prime minister’s personality or whom he trusted for advice. Thus, precisely because of the predominance of the prime minister’s powers and decision-making abilities, Portugal falls under the Dominant Prime Minister Model of governance rather than under the other two—Coalition Compromise Model and Ministerial Government Model. This chapter helps to shed light into the processes and dynamics inherent to governments, especially coalition governments, in Portugal. However, abundant work still has to be done on this topic in a country where polarization in parliament seems to be increasing, something that was made explicit in the recent electoral results of October 2019.

References Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Anna Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010’. Party Politics, 21(1): 143–52. Calca, Patrícia (2015a). ‘Government’s Decision: A Theoretical and Empirical Study on Legislative Initiative as Strategic Behaviour’. PhD Dissertation. Lisbon: University of Lisbon. Calca, Patrícia (2015b). ‘Introducing a New Database on Legislative Behaviour in Portugal’. Working Paper – Mannheimer Zentrum für Europäische Sozialforschung (MZES), 160: 1–22.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

542



Calca, Patrícia (2019a). ‘Responsividade e Inovação Legislativa em Portugal: Dos Programas Eleitorais à União Europeia’. In André Corrêa d’Almeida (ed.), Reforma do Sistema parlamentar em Portugal: Análises e Instrumentos para um Diálogo Urgente. Parede: Principia, 211–58. Calca, Patrícia (2019b). ‘O Partido Social Democrata: Os Programas Eleitorais e a União Europeia’. In Alice Cunha (ed.), Os Partidos Políticos Portugueses e a União Europeia. Coimbra: Almedina, 52–91. Calca, Patrícia (2021). Executive-Legislative Relations in Parliamentary Systems: Policymaking and Legislative Processes. London & NewYork: Palgrave Macmillan. Calca, Patrícia, and Martin Gross (2019). ‘To Adapt or to Disregard? Parties’ Reactions to External Shocks’. West European Politics, 42(3): 545–72. Calca, Patrícia, and Teresa Ruel (2018). ‘From Tip to Toe: Evidence from National and Subnational Parliaments in Europe’, EPSA Conference Paper. Freire, André, Marco Lisi, Ioannis Andreadis, and José Manuel Leite Viegas (2014). ‘Political Representation in Bailed-Out Southern Europe: Greece and Portugal Compared’. Southern European Society and Politics, 19(4): 413–33. Jalali, Carlos (2015). Partidos e Democracia em Portugal 1974–2005. Lisboa: Imprensa de Ciências Sociais. Magone, José M. (2000). ‘Portugal: The Rationale of Democratic Regime Building’. In Wolfgang C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 529–58. Müller, Wolfgang C. and Kaare Strøm (eds.) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of AntiElitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research and Politics, 4(1): 1–9. Ruel, Teresa (2017). As Regiões Autónomas (Açores e Madeira) nos Debates Parlamentares da Assembleia da República 1975–2015. Lisboa: Assembleia da República Portuguesa. Ruel, Teresa (2020). Political alternation in the Azores, Madeira and the Canary Islands. London & NewYork: Palgrave Macmillan. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press.

Appendix. List of political parties Abbreviation UDP BE

Name Popular Democratic Union (União Democrática Popular) Left Bloc (Bloco de Esquerda)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



543

MDP-CDE

Portuguese Democratic Movement-Democratic Electoral Commissions (Movimento Democrático Português-Comissões Democráticas Eleitorais) PCP Portuguese Communist Party (Partido Comunista Português) ASDI Independent Social Democratic Action (Acção Social Democrata Independente) UEDS Left-wing Union for the Socialist Democracy (União da Esquerda para a Democracia Socialista) PS Socialist Party (Partido Socialista) PRD Democratic Renewal Party (Partido Renovador Democrático) PSD Social Democratic Party (Partido Social Democrata) PSN National Solidarity Party (Partido da Solidariedade Nacional) PPM People’s Monarchist Party (Partido Popular Monárquico) CDS CDS-People’s Party (CDS-Partido Popular), 1992– Democratic and Social Centre (Centro Democrático e Social), 1975–1992 PáF Portugal Ahead (Portugal à Frente), comprised of PSD and CDS Note: Party names are given in English, followed by the party name in Portuguese in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 16 Spain: Single-Party Majority and Minority Governments Bonnie N. Field

Spain, along with Greece and Portugal, re-democratized in the 1970s. It formally joined the European integration project in 1986, and it has been in the Eurozone since 1999. Spain stands out in comparative Western European perspective. Between its transition to democracy and 2019, it did not have a coalition government.¹ Instead, governments were either single-party minority or single-party majority ones (Field 2016a). Of the Western European countries covered in this volume, it is the only one without a coalition government during the period under study. The United Kingdom is similarly resistant to coalitions; yet, even it had a coalition government between 2010 and 2015. Even after the transformation of the party system in 2015—in which the dominance of the Socialist Party (Spanish Socialist Workers’ Party/PSOE) and the Popular Party (PP) ended and two significant newcomers, Podemos (We can) and Cuidadanos (Citizens/Cs), entered parliament—the pattern of single-party governments continued. When parliament failed to select a government, new elections in 2016 produced a PP minority government. Parliament, led by the PSOE, brought that government down in a constructive vote of no confidence in 2018, and the PSOE formed another single-party minority government. The April 2019 elections produced even greater fragmentation with the radical right party, Vox, entering parliament for the first time. Again, parliament failed to form a government, and new elections occurred in November 2019. It was not until January 2020 that Spain got its first coalition government—a minority one of the PSOE and Podemos.²

¹ Nonetheless, coalition governments were the norm during Spain’s Second Republic democracy (1931–1936/1939). ² The coalition falls outside of the observation period; the formation of this cabinet is intertwined with the transformation of Spain’s party system. The strong bargaining position of its governing parties, with large pluralities in parliament, disappeared. While newcomers, particularly Cs and Podemos, were initially willing to accept support party roles, Podemos subsequently sought cabinet posts. Nonetheless, the PSOE’s resistance to a coalition was strong. After the April elections in 2019, it preferred a single-party minority government, supported externally by Podemos and smaller parties. A coalition agreement only succeeded after Podemos refused to support the PSOE’s candidate for PM and new elections were held in November of 2019. Bonnie N. Field, Spain: Single-Party Majority and Minority Governments In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0016

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



545

This chapter reviews the institutions relevant to the life cycle of governments, the parliamentary party system and changes therein, and the record of government formation, governability, and government termination through 2018. In comparative perspective, Spain’s governments before 2015 generally formed rather easily, governed without great difficulty, and were quite stable. In contrast, the party system change in 2015 led to severe difficulties of government formation and, thus far, governability.

The institutional setting The literature suggests that a variety of institutions affect the life cycle of governments. However, the ones typically highlighted do not convincingly explain the prevalence of single-party minority governments in Spain. Nonetheless, they help explain governability and government duration. Spain has a parliamentary system with a constitutional monarch as its head of state. In broad strokes, its institutions do not impose a high hurdle for government formation and strengthen governments once in office.

Bicameralism Research suggests that bicameralism affects government formation—favouring coalitions (Druckman et al. 2005), even with weak upper chambers—and duration (Druckman and Thies 2002). In particular, governments with upper chamber majorities last longer than those that do not. Spain’s parliament, las Cortes Generales, is bicameral. However, bicameralism did not correlate with coalitions and in general has not affected governability significantly. The lower house, Congress of Deputies, is substantially more powerful than the Senate.³ The Congress of Deputies invests the prime minister (PM) and can censure the government. Ultimately, the Congress can pass the legislation it prefers over opposition from the Senate. Because the Senate is weak, the government normally is most concerned with its standing in the lower house. Nonetheless, the government cannot completely ignore the Senate. The Senate can approve amendments and vetoes, in effect giving it delay power. However, the Congress must approve Senate amendments, and Congress can lift a Senate veto with an absolute majority vote within the first two months and with a simple

³ This discussion refers to the period after the approval of the 1978 constitution. Before the 1978 constitution, the Senate was more powerful in the lawmaking and constitution-making processes than it was subsequently.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

546



majority thereafter. The Senate also has certain powers that the Congress of Deputies does not, such as its ability to suspend regional (Art. 155 of the constitution) and local governments (Art. 61 of the Local Regime Law) (Harguindéguy et al. 2017: 533–4).⁴ While exceptional, the minority government of Mariano Rajoy attained Senate approval to suspend Catalonia’s regional autonomy in response to its independence declaration in 2017. Normally, the governing party has a stronger position in the Senate than it does in the Congress of Deputies, in part due to the more majoritarian nature of the Senate’s electoral system.⁵ The governing party has often been the largest party in the Senate during minority governments; however, not always. During the Socialist governments of José Luis Rodríguez Zapatero, the PP held more seats in the Senate, though short of a majority. Furthermore, during the PSOE government of González between 1993 and 1996, the PSOE’s representation slipped behind the PP in the Senate.⁶ During the Sánchez minority government (2018–2019), the opposition controlled the Senate. Thus far, political parties in the Senate have not used the full potential for obstruction that the institutional setup permits.

Positive parliamentarism: investiture of the PM Spain has positive parliamentarism (Bergman 1993), meaning that parliament must explicitly select the PM in an investiture vote (see the introductory chapters of this volume), as do about half of the countries covered in this volume. Yet, positive parliamentarism has not stymied minority governments, as scholars suggest it does elsewhere. This may be in part because, in Spain, parliament can elect the PM with less than an absolute majority vote. Government investiture is regulated in the 1978 constitution and in the Standing Orders of the Congress of Deputies (Sections 170–172).⁷ According to the constitution (Art. 99), following a round of consultations with the leaders of the political parties in parliament, the monarch nominates a candidate for PM

⁴ An English version of the 1978 constitution is available at: http://www.congreso.es/constitucion/ ficheros/c78/cons_ingl.pdf (last accessed 18 March 2021). ⁵ There are 208 directly elected senators. In the overwhelming majority of districts, Spain’s peninsular provinces, they are elected through a limited vote system in which voters can cast three votes to fill four seats per district. The candidates who receive the most votes win. (Spain’s island and autonomous city districts elect between one and three senators.) Additionally, Spain’s regional parliaments indirectly elect a variable number of senators (currently 58). ⁶ Based on data from the Senate’s website, available at http://www.senado.es/web/com posicionorganizacion/gruposparlamentarios/gruposparlamentariosdesde1977/index.html (last accessed 29 March 2018). ⁷ An English version of the Standing Orders is available at: https://www.congreso.es/webpublica/ ficherosportal/standing_orders_02.pdf (last accessed 18 March 2021).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



547

through the Speaker of the Congress.⁸ The prime ministerial candidate then presents their programme to the Congress, which it debates.⁹ Afterwards, the candidate is subject to a formal investiture vote.¹⁰ If the prime ministerial candidate (and parliament only votes on the PM candidate), receives the support of an absolute majority of the total number of deputies (50 per cent + 1 yes votes), the monarch appoints them. If an absolute majority is not attained, a simple majority of more yes than no votes suffices in a second vote 48 hours later. The investiture vote is a public, roll-call vote (Standing Orders Art. 85.2). If this hurdle is not met, the monarch can nominate (or renominate) successive candidates according to the aforementioned provisions. Each candidate begins (again) at round one. If parliament does not elect a prime minister within two months after the first investiture vote, the head of state dissolves parliament and new elections are triggered.¹¹ Once parliament elects the head of government, the PM then selects the government ministers without parliament intervening (Constitution, Art. 100). The monarch formally appoints them. The government consists of the PM, deputy PMs, if any, and the ministers.¹²

Constructive vote of no confidence and confidence votes Spain adopted a constructive vote of no confidence in its 1978 constitution (Art. 113), setting a high threshold for parliament to remove the government.¹³ This helps account for the comparatively long duration of Spain’s governments, at least until 2015. However, it has not prevented minority governments, as some research suggests it would (Strøm et al. 1994). Amongst the Western European cases in this volume, only Belgium, Germany, and Spain have such a high threshold for government censure. At least one-tenth of the members of parliament (MPs) in the Congress must propose the motion that also includes a candidate for PM. To pass, an absolute majority of the MPs in

⁸ Prior to the approval of the 1978 constitution, a formal investiture vote was not required. The monarch appointed Adolfo Suárez as the PM in 1977. ⁹ To date, Spain has not had a women prime ministerial candidate or PM. ¹⁰ Most constitutional jurists consider that the investiture vote is on the PM candidate and their programme (Vintró Castells 2007: 302). ¹¹ The constitution did not regulate what should happen if no investiture vote takes place (e.g. no candidate agrees to stand for investiture, which was a possibility after the 2016 elections). If no investiture vote occurs, the two-month timetable for new elections would not be triggered. ¹² Secretary of state is the rank immediately below minister. While not part of the cabinet (Consejo de Ministros), secretaries of state may be called to attend cabinet meetings to report on areas of their jurisdiction (Real-Dato and Rodríguez Teruel 2016: 499). ¹³ Law 51/1977 (14 November 1977) regulated votes of no confidence and confidence motions until the approval of the 1978 constitution.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

548



the Congress must vote to support it. Moreover, parliament cannot censure individual ministers. The nomination and removal of ministers is the PM’s exclusive authority (Art. 100). The PM may also call a confidence vote on their programme or on a policy statement. Confidence is considered granted with a simple majority in favour (Art. 122). The government must resign if it loses the confidence of parliament, after general elections, or on the resignation or death of the PM (Art. 101). The government remains in office as a caretaker cabinet until the next government forms. Thus, while parliament can select a government without absolute majority support, it is very difficult for parliament to bring it down.

PM’s cabinet powers Spain’s PM has strong powers, comparable to in Germany and the United Kingdom (see the introductory chapters of this volume).¹⁴ As mentioned previously, the PM in Spain selects and removes cabinet members. Parliament cannot censure an individual minister; it can only censure the entire government in a constructive vote of no confidence. This means individual ministers owe their positions directly to the PM. Article 98 of the constitution gives the PM formal steering or coordination rights vis-à-vis cabinet ministers. Since 1977, the PM has set the jurisdiction of ministries using executive decree authority (real decreto). Parliament does not need to pass a law. However, this process was not explicitly regulated in the constitution but was formalized in the 1997 Government Law (Rodríguez Teruel 2011: 57). The same 1997 law formalized the PM’s control of the cabinet agenda (Art. 2g, 18.3). Nonetheless, the PM’s control is not complete. The General Committee of Secretaries of State and Sub-secretaries, which a deputy PM or minister of the presidency chairs, is important in setting the agenda for cabinet meetings. The committee assists in handling the large volume of matters that require cabinet approval. A 1983 law first regulated the committee, though its origin stretches back to 1976 (Parrado 2017: 103). The committee, referred to as the ‘consejillo’ or small cabinet, largely determines which matters appear on the cabinet’s agenda. It places items on a ‘green’ or ‘red’ list. Matters on the green list are those that the committee unanimously approved and are not likely to require further debate in cabinet. Matters on the red list likely do. The constitution specifies which decisions cabinet needs to debate or approve, but it does not stipulate a precise decision rule. In practice, cabinet makes decisions by consensus, which the PM

¹⁴ The rules discussed here encompass constitutional rules, procedural rules, and conventions.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



549

defines. Votes typically are not taken and the cabinet’s deliberations are secret. The head of state typically does not attend cabinet meetings (Parrado 2017: 102). Finally, the PM has important resources to monitor ministers. The size of the PM’s office, which includes political appointees and civil servants, has increased significantly over the democratic period (Parrado 2017: 79). A minister of the presidency monitors and coordinates the actions of ministers. PMs have appointed one or more deputy PMs, who have a monitoring and coordinating role. The PM can, and typically does, create Delegated Committees of the Government on particular matters, such as economic affairs, that can include and monitor various ministries and tiers of government, for example at the regional level.

The multilevel state While it is not typically included as an institution associated with the life cycle of governments, Spain’s multilevel state, with 17 powerful regional governments, is essential for understanding government formation, governability, and government termination. This is because parties engage in multilevel bargaining to form and sustain governments (Colomer and Martínez 1995; Reniu i Vilamala 2002; Falcó-Gimeno and Verge 2013; Field 2016a). The dynamics of party politics and governing at the regional level affect national-level governments, particularly during periods of minority governments, and vice versa. Moreover the political dynamics at the regional level can change during the national government’s tenure, affecting its ability to govern.

The party system and the actors During the 40 years between the first democratic elections in 1977 and the end of 2018, Spain had 15 governments. In addition to the distinct lack of coalitions, the prevalence of minority governments also makes Spain stand out (Table 16.1a). Excluding the Rajoy II caretaker cabinet in 2015, 10 of the 14 (71 per cent) cabinets were minority ones. This places Spain among the countries with the highest share of minority governments, along with Denmark, Norway, and Sweden. However, in contrast, the Scandinavian countries have all formed minority and majority coalitions. Spain’s major parties have all governed in minority. The Union of the Democratic Centre (UCD), which was the main centre-right party during the early years of democracy, did so between 1977 and 1982. Since then, the centre-left PSOE and the conservative PP governed with majority and minority governments. Notably, the PSOE has not governed with a majority since the 1986 González II cabinet.

1977-07-05 1979-04-06 1981-02-27 1982-12-03 1986-07-26 1989-12-07 1993-07-14 1996-05-06

Suárez II Suárez III Calvo-Sotelo González I González II González III González IV Aznar I

Aznar II Zapatero I Zapatero II Rajoy I Rajoy IIa Rajoy III Sánchez I

1 2 3 4 5 6 7 8

9 10 11 12 13 14 15

2000-03-12 2004-03-14 2008-03-09 2011-11-20 2015-12-20 2016-06-26

1982-10-28 1986-06-22 1989-10-29 1993-06-06 1996-03-03

1977-06-15 1979-03-01

PP PSOE PSOE PP PP PP PSOE

UCD UCD UCD PSOE PSOE PSOE PSOE PP maj min min maj min min min

min min min maj maj min* min min 52.3 46.9 48.3 53.1 34 38.3 24

47.4 48 47.1 57.7 52.6 50 45.4 44.6

Election date Party Type Cabinet composition of strength of cabinet cabinet in seats (%)

350 350 350 350 350 350 350

350 350 350 350 350 350 350 350 8 8 9 11 10 9 9

8 9 9 9 9 9 8 9 2.48 2.5 2.34 2.62 4.33 3.98 3.98

2.89 2.81 2.9 2.35 2.9 2.85 2.67 2.72 PP PSOE PSOE PP PNV Cs Cs

Cs, CC

CC CiU, PNV CiU, PNV, CC CC CC

Formal Number of Number of ENP, Median seats in parties in parliament party in first support parties parliament parliament policy dimension

Notes: For a list of parties, consult the chapter appendix. * De facto a majority government due to HB not taking its seats, except for when an absolute majority of votes is required (e.g. on organic laws) a = Limited policy remit Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1977–1998 period. Median parties for the subsequent period are based on data from Polk et al. (2017) and Bakker et al. (2015). The first policy dimension is economic left–right. The number of parties only includes parties that have held more than two seats in parliament when a cabinet has formed. Cabinet types: maj = Single-party majority cabinet; min = Minority cabinet (both single-party and coalition cabinets). Table legends and variables are further defined in the Appendix, this volume. Source: Ministry of the Interior, Spain.

2000-04-28 2004-04-18 2008-04-14 2011-12-22 2015-12-20 2016-11-04 2018-06-06

Date in

Cabinet Cabinet number

Table 16.1a Spanish cabinets since 1977 OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



551

In elections for the Congress of Deputies, voters elect representatives via a closed list, proportional representation electoral system format. However, several traits distort proportionality in practice, including the allocation of a guaranteed minimum number of seats per district (independent of population) and many low magnitude districts.¹⁵ In general, the system has favoured larger parties, harmed smaller parties whose support is geographically dispersed, and has not punished smaller parties whose support is geographically concentrated. The two primary dimensions of the party system are the left–right and centre– periphery (Alonso 2012; Field and Hamann 2015). In the Spanish case, it is best to consider the left–right dimension in general terms, that is both in terms of socioeconomic policy related to the role of the state in the economy and redistribution and in liberal–conservative (or libertarian–authoritarian) terms regarding individual morality, religion, civil liberties, and so forth. These dimensions have largely overlapped one another in the national party system, and therefore do not clearly represent distinct dimensions. The centre–periphery cleavage is also prominent in this multinational country. Using Alonso’s conceptualization, the centre–periphery dimension is about territory. It encompasses diverse institutional, cultural, and fiscal issues, such as the ‘political status of the peripheral territory inside the state’; the ‘protection and preservation of the peripheral group’s cultural distinctiveness and identity’; ‘the way in which the power of revenue and expenditure should be distributed between the central state and peripheral administrations’ (Alonso 2012: 25). This cleavage manifests in distinct party types, regional and national or statewide parties, and in distinct policy positions. While regional parties tend to be more pro-periphery than the statewide parties, there is variation within each party type. Regional parties also span the left–right ideological spectrum. To measure party placement, the collective project uses the Chapel Hill Expert Survey and its placements on the left–right economic and centralization–decentralization questions.

Party system change Spain has experienced significant changes in its party system since redemocratization in the mid-1970s. Focusing on the post-1990 period, Spain’s party system changed from one in which two main parties predominated to one with four and then five significant national parties. Other smaller parties were and are relevant. Yet, unlike many Western European countries, Spain did not have a relevant right populist or radical right party in parliament until 2019. It is useful to distinguish the period between 1993 and 2015 from the post-2015 period. ¹⁵ Two of the 350 members are elected in single-member districts.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

552



1993–2015: During this period, competition between the two main parties, the centre-left PSOE and conservative PP, was stiff, and the parties regularly alternated in government. There were two majority governments of the PP and four minority governments (three PSOE and one PP). Additionally, regional parties gained prominence as support parties for minority governments. This is a period of comparatively stable governments, both in cross-national terms and compared to subsequent governments. Amongst the regional parties, the centre-right parties were the most relevant until about 2000, when leftist regional parties grew in importance. The centreright includes the Catalan Convergence and Union (CiU) (between ten and sixteen seats), the Basque Nationalist Party (PNV) (between five and eight seats) and the Canary Islands Coalition (CC) (between one and four seats). Taking moderate nationalist positions, these parties allied with both the PSOE minority government of Felipe González (1993–1996) and the PP minority government of José María Aznar (1996–2000) (Field 2016a). Beginning in 2000, leftist regional parties won more seats than they had previously, including the Galician Nationalist Bloc (BNG) (with two or three seats) and the Catalan Republican Left (ERC). ERC had nearly consistently held one seat in the Congress, which increased to eight in 2004 (though it declined to three in the next two elections). These leftist parties (and others) at times allied with the PSOE minority governments of Zapatero (2004–2011) (Field 2016a). Notably, ERC supports Catalonia’s independence from Spain, but at this time it did not actively take steps to accomplish it, as it would later. BNG, while more radical in its origins, did not openly support Galician independence (GómezReino Cachafeiro 2011: 142–3). 2015–2019: In the 2015 parliamentary elections, four significant national parties entered parliament, along with a variety of regional parties. Podemos (We can) and Ciudadanos (Citizens) gained a strong foothold, alongside the PSOE and PP, and the effective number of parties increased to 4.33, from 2.62 in the prior election. It also meant that the votes of regional parties were no longer sufficient to support a minority government, as they had been in the past. Complicating governability even more, some Catalan parties actively pushed for secession. Catalan parties had previously been essential for governability in Spain. The roots of the transformation lay in the economic crisis that hit Spain beginning, mildly, in 2008 and accelerated thereafter, and to the Zapatero II (PSOE) government’s implementation of austerity measures in 2010 (Field and Botti 2013b; Miley 2016), about which many PSOE voters felt betrayed. It was the worst economic crisis that Spain had experienced in the previous 80 years (Field and Botti 2013a: 4). In this context, PM Zapatero called early elections for November 2011, in which he did not stand as a candidate. The voters severely punished the PSOE, which plummeted from 169 to 110 seats (29 per cent of the vote). The conservative PP benefitted, winning an

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



553

absolute majority of seats and 45 per cent of the vote. Many smaller parties also increased their votes, producing greater fragmentation among the opposition parties. Nonetheless, a party capable of capitalizing on societal discontent with the main parties had yet to emerge. This occurred during the PP majority government of Rajoy I (2011–2015), which intensified austerity measures and structural reforms. Simultaneously, the PP faced a series of severe corruption scandals. Other parties, including CiU in Catalonia and the PSOE at the regional level, and state institutions also faced corruption charges and scandals, leading to the widespread public perception that corruption was endemic. Signs of change appeared in the 2014 European parliament elections, with the rise of the radical left Podemos, formed just months before, which won eight per cent of the vote amidst extraordinary levels of vote switching (Cordero and Montero 2015). The May 2015 local and regional elections confirmed Podemos’ appeal but also the rise of the centre-right Ciudadanos (Gomez Fortes and Urquizu 2015; Rodon and Hierro 2016). Ciudadanos started as a regional party in Catalonia, opposed to Catalan nationalism. It began to expand throughout Spain in 2013 (Rodríguez Teruel and Barrio 2016). The 2015 parliamentary elections bought the new parties into parliament (Medina and Correa 2016). Podemos and its allies won 69 seats (20 per cent); Ciudadanos won 40 seats (11 per cent). The PSOE once again declined, this time to only 90 seats (26 per cent). Voters also punished the PP, which declined from 185 to 123 seats (35 per cent). Because parliament could not agree on a new government, elections occurred again in June 2016, which produced broadly similar results in terms of seats: PP (39 per cent), PSOE (24 per cent), Podemos (20 per cent), and Ciudadanos (9 per cent). Research suggests that both economic and political disaffection, the latter related to a crisis of representation and corruption, led voters to support new parties (Orriols and Cordero 2016; Bosch and Durán 2017; Vidal 2018). The newcomers campaigned as outsiders, intent on changing politics in Spain, including ‘political renewal, democratic regeneration and the fight against political corruption’ (Rodríguez Teruel and Barrio 2016: 597). Therefore, the insider– outsider distinction was important. However, arguably, the insider–outsider cleavage lost importance in the years since, and the newcomers clearly slotted into the left–right and centre–periphery dimensions of party competition. Podemos’ populist messaging initially attempted to eschew the terms left–right, deemed antiquated. Nonetheless, the Spanish public in 2019 placed it on the far left, with an average placement of 2.3 on a 10-point left–right spectrum, compared to the PSOE’s average of 4.1 in opinion polls.¹⁶ Ciudadanos also tried to avoid

¹⁶ Data from CIS poll, March 2019. Available at: www.cis.es (last accessed 18 March 2021).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

554



ideological labels as part of a catch-all strategy. It claimed to draw on European traditions of liberalism, democracy, and socialism (Rodríguez Teruel and Barrio 2016). Polls of citizens placed it at 6.9 on the left–right spectrum, compared to 8.1 for PP. Cs more clearly placed itself on the political right in recent years. There are important differences across the new and the old parties along the centre–periphery dimension. Ciudadanos formed in 2006 to oppose Catalan nationalism and nation-building policies in Catalonia—instead it supported Spanish identity and the Castilian language (Rodríguez Teruel and Barrio 2016). The Chapel Hill Expert Survey of party positioning in 2014 placed Ciudadanos at 7.0 on a 10-point decentralization–centralization scale.¹⁷ This compares to 2.6 for Podemos, 4.0 for PSOE, and 7.5 for PP. Nonetheless, Ciudadanos, particularly when the Catalan crisis heated up in 2017 with a failed independence declaration, adopted a harder line stance towards peripheral nationalism than the PP under Rajoy, and stressed Spanish nationalism. Podemos, on the other hand, developed, in part, on the basis of regional-level alliances with left-wing regional organizations in Catalonia, Valencia, and Galicia. It adopted and stressed a multinational view of Spain (Rodríguez Teruel et al. 2016). Party leaders, weighing in on the Catalan independence debate, expressed their desire for Catalonia to remain part of Spain. Yet, the party also supported a referendum on Catalonia’s independence, which the PP, PSOE, and Ciudadanos opposed. In 2015, under the influence of its Catalan MPs, Podemos made the holding of a referendum on Catalonia’s relationship with Spain a condition for it to support a PSOE-led government (Rodríguez Teruel et al. 2016)—a nonstarter. These changes coincided with a radicalization of territorial demands in Catalonia (Gillespie and Gray 2015). Moreover, Catalan parties are often a lynchpin of governability in Spain. CiU split in 2015 because one of the component parties in the two-party federation, Democratic Convergence of Catalonia (CDC), moved from a moderate nationalist stance to the active pursuit of secession. CiU had been the largest regional party and typically the third largest party in parliament. It had also been a critical support party for minority governments. CDC renamed itself Partit Demòcrata Europeu Català (Catalan European Democratic Party) (PDeCAT) in 2016. The name change responded to corruption scandals that tarnished the CDC brand and accommodated the party’s new profile. Jointly with ERC, in 2012, it began to actively pursue holding a referendum on Catalonia’s relationship with Spain, which the PP governments and the constitutional court blocked (Barrio and Field 2017b, 2017a). The Catalan government, led by the two secessionist parties, backed a referendum on independence on 1 October 2017, in defiance of a court order. On the basis of the referendum, the

¹⁷ Available at: http://chesdata.eu/ (last accessed 18 March 2021).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



555

pro-independence forces declared independence on 27 October, which led the Spanish government to dismiss the Catalan government, disband parliament, and temporarily suspend the region’s autonomy. Notably, Spain had not had an electorally viable radical right party, unlike in much of Western Europe (Alonso and Rovira Kaltwasser 2015). However, a new far right party, Vox, won seats in the regional elections in the region of Andalusia in 2018 and became the kingmaker of a PP–Cs minority coalition government. In the April 2019 national parliamentary elections, Vox won 24 seats in parliament and over 10 per cent of the vote. It became the third largest party in Spain’s parliament in the November 2019 elections.

Electoral alliances and pre-electoral coalitions Electoral alliances and pre-electoral coalitions have not been important for understanding governments in Spain. Between the 1990s and 2015 (see Table 16.1b), pre-election alliances formed between small parties, such as between the leftist IU and left-green ICV, and among the Canary Islands’ parties (CC, NC, and PNC). The PP also formed pre-election alliances with small regional parties, such as the Aragonese PAR and Navarre UPN. Between 2015 and 2018, electoral alliances among regional parties and between one of the large national parties and a small regional party (e.g. PP + UPN and PSOE + NC) continued. Moreover, Podemos formed electoral alliances more than other national parties. In 2015 and 2016, it presented candidates in alliance with distinct regional forces across the country—with en Comú (in common or jointly) in Catalonia, en Marea (tides or en masse) in Galicia and Compromís (commitment) in Valencia. The MPs therefore did not belong to a unified party. For example, in 2015, only 40 per cent of Podemos’ MPs elected in Catalonia, Galicia, and Valencia were members of the party (Rodríguez Teruel et al. 2016: 68). After the governing board of the Congress rejected the formation of separate parliamentary groups representing the distinct components of the alliance, most agreed to form a single parliamentary group. Compromís opted to join the mixed parliamentary group, which is an amalgam of parties that do not meet the requirements to form their own group.¹⁸ Also, in the 2016 elections, Podemos and IU reached an agreement to form an electoral alliance—Unidos Podemos (Together We Can). The election manifesto contained their agreement on ‘50 steps to govern together’. Following the elections, they formed a joint parliamentary group with en Comú and en Marea.¹⁹

¹⁸ All parliamentary group information is taken from the Congress of Deputies’ website. Available at: http://www.congreso.es (last accessed 18 March 2021). ¹⁹ This alliance was reprised in 2019, as Unidas Podemos.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

556



Table 16.1b Electoral alliances and pre-electoral coalitions in Spain, 1977–2018 Election date

Constituent parties

Type

1977-06-15

UDC, Centre Català PSP, US AP, ACL, PDP, others AP, PDP, PAR, UV, UPN, others AP, PDP, PL PA, UPN PP, UPN PP, PAR IU, ICV CC, NC IU, ICV CC, PNC PP, UPN IU, ICV, CHA, others CC, NC, PNC PP, UPN PP, PAR CC, PNC PP, UPN PP, PAR PP, FORO Podemos, En Comú (ICV), CPV, En Marea, others PSOE, NC CC, PNC PP, UPN PP, PAR PP, FORO Podemos, En Comú (ICV), CPV, En Marea, others Podemos, IU, EQUO, others PSOE, NC

EA EA EA, PEC EA, PEC EA, PEC EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA

1979-03-01 1982-10-28 1986-06-22 1989-10-29 1996-03-03 2004-03-14 2008-03-09

2011-11-20

2015-12-20

2016-06-26

Types of pre-electoral commitment

Other* Other* Other*

EA EA EA EA EA EA EA, PEC EA

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other *Joint ‘governing programme’/election manifesto

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



557

Government formation Prior to 2015, governments had typically formed rather easily.²⁰ Between 1977 and 2015, the party with the most seats in parliament formed the government. All governments except the Calvo-Sotelo government in 1981 formed immediately after a parliamentary election. On average, it took 37 days to form a new government, and it did not take minority governments longer than majority ones to form (see Table 16.2). This changed dramatically in 2015, with the change of the party system. In a series of firsts since Spain’s re-democratization, the leader of the party that won the election, Rajoy of the PP, did not accept the formateur role because he did not think he would be able to form a government. The leader of the second largest party, Pedro Sánchez of the PSOE, accepted but failed. New elections had to be held in 2016. Meanwhile, the caretaker government of Rajoy II lasted 189 days. Government formation after the 2016 elections was not easy either. While the political actors avoided having to face the public in the polls yet again, government formation took 131 days. The PP minority government of Rajoy III was the weakest cabinet in terms of seats in parliament to date, at 134 of 350 seats (38.3 per cent). That government did not last long; parliament brought it down in a constructive vote of no confidence that simultaneously elected PM Pedro Sánchez of the PSOE in 2018, which was also short-lived.

Bargaining over coalition formation Bargaining over forming a coalition government has been rare in Spain. Until 2019, the only time that bargaining occurred, publicly, over the possibility of forming a coalition was after the 1993 parliamentary elections.²¹ The PSOE, in government since 1982, won the election with 159 seats in the Congress of Deputies, but fell 17 seats short of an absolute majority. Under the leadership of Felipe González, it entered into separate negotiations with the Catalan CiU (17 seats) and the Basque PNV (5 seats). Each party internally debated entering a coalition and decided that it was not in their interest to assume the costs of governing (Aguilera de Prat 2001a, 2001b; Reniu i Vilamala 2002; Field 2016a: 52, 152–4), potentially costing them support in their regions. CiU discarded entering

²⁰ The investiture of Leopoldo Calvo-Sotelo in 1981 was notable for the coup attempt, known as 23F. In the midst of the second-round investiture debate, coup instigators in the Civil Guard and Lieutenant Colonel Antonio Tejero took over the in-session Congress of Deputies. After the coup attempt failed, MPs returned two days later to elect Calvo-Sotelo, this time with an absolute majority that he had not received in the first vote prior to the coup attempt. ²¹ While outside the period of analysis in this book, bargaining over coalition formation occurred in 2019/2020.

1977 1979 1981 1982 1986 1989 1993

1996 2000 2004 2008 2011 2015 2016 2018

Suárez II Suárez III Calvo-Sotelo González I González II González III González IV

Aznar I Aznar II Zapatero I Zapatero II Rajoy I Rajoy II Rajoy III Sánchez I

0 0 0 0 0 0 0 0

0 0 0 0 0 0 2

Number of inconclusive bargaining rounds UCD UCD UCD PSOE PSOE PSOE PSOE (1) PSOE, PNV (2) PSOE, CiU, PNV PP PP PSOE PSOE PP PP PP PSOE

Parties involved in the previous bargaining rounds

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 17 18

Bargaining duration of individual formation attempt (in days)

64 47 35 36 32 0 131 0

20 36 29 36 34 39 38

Number of days required in government formation

0 0 0 0 0 0 0 0

0 0 0 0 0 0 35

Total bargaining duration

181 202 183 169 187 ** 170 180

* 183 186 207 184 167 181

Pro

166 148 148 158 149 111 169

68 1

148 158 116 144 155 165

Contra

1 0 19 23 14

8 0 21 6 6 1

Abstention

Result of investiture vote***

Notes: * No investiture vote was required in 1977. ** No successful investiture vote; caretaker cabinet *** Investiture vote for Calvo-Sotelo, Zapatero II, and Rajoy III are second-round votes. Investiture vote for Sánchez I is also the constructive no confidence vote against the Rajoy III government.

Year in

Cabinet

Table 16.2 Cabinet formation in Spain, 1977–2018 OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



559

a coalition first; PNV followed. Nonetheless, both opted to provide external support to the González IV minority government. While it does not reach the standard for formal, public bargaining used in this volume, Suárez of the UCD in the early years of democracy and later Aznar of the PP in 1996 and 2000 privately put forward the notion of CiU taking portfolios (Field 2016a: 52, 152–4, 229fn15). Furthermore, after the 2015 election, Podemos advanced the idea of forming a PSOE + Podemos coalition government (Simón 2016: 502–4). However, the PSOE never publicly recognized any coalition negotiations. Bargaining centred on forging a deal to attain the votes necessary to invest the PSOE candidate Pedro Sánchez, which ultimately failed.

Bargaining over support for (single-party) minority governments However, what Table 16.2 does not capture is inter-party bargaining about providing support for a single-party minority government. This occurs regularly in Spain. It has taken place typically between the formateur party and the smaller regionally based parties. Bargaining occurs over policy, regional autonomy, and investments. It also often entails bargaining over exchanging political support across state levels so that regional parties can govern or govern more comfortably in their regions in exchange for giving support to a minority government from parliament at the national level (Heller 2002; Reniu i Vilamala 2002; Field 2016a). At times, the bargains have been formalized in support agreements, while at other times support has been informal or ad hoc. In the 1989 elections, the PSOE won 175 seats, exactly half of the seats in the Congress. Nonetheless, at the time the chamber first met there were only 332 MPs because the results of elections in a few provinces were annulled. Therefore, the number of votes required for an absolute majority, normally 176, was set at 167 (Vintró Castells 2007: 304–05). At the time of investiture, the PSOE needed one additional vote to elect Felipe González in a first-round investiture vote. The PSOE negotiated to get the support of the one deputy from the Canary Islands Independent Groups (AIC), in exchange for employment and investment projects in the Canary Islands (Field 2016a: 53).²² The government took 39 days to form. The formation of the 1993–1996 PSOE minority government and the 1996–2000 PP minority government involved bargaining to attain the parliamentary support of regionally based nationalist parties on the centre-right (Field 2016a: 98–111). In 1993, the Catalan CiU and the Basque PNV voted in favour of González’s investiture. In 1996, CiU, PNV, and CC voted in favour of Aznar’s investiture. However, there were important differences between the formation

²² AIC later became the Canary Islands Coalition.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

560



processes. In the case of the González IV (1993–1996) minority government, PNV and CiU only publicly committed to investiture agreements and not to agreements for the entire legislature (pacto de legislatura), though informally CiU’s support amounted to a legislature agreement. In contrast, CiU and CC signed formal documents that were considered legislature agreements with the PP in 1996 and gave the PP an absolute majority in parliament. PNV only publicly committed to support Aznar’s investiture. The González IV minority government took 38 days to form compared to 64 days for the Aznar I minority government. The more prolonged government formation process was due to the greater difficulty of reaching agreement between the PP, representing the Spanish nationalism pole, and the regional nationalist parties. Moreover, the PP had spent the prior legislature criticizing the agreements between the González IV government and the regional nationalist parties, whose support it now needed. In contrast, the Socialists under the leadership of Zapatero did not negotiate formalized agreements with potential support parties when it won the elections short of an absolute majority in 2004 and 2008 (Field 2016a: 111–18). Yet neither government had difficulty forming. In 2004, the PSOE benefitted from good will among most of the left and left regional parties, which assured Zapatero’s investiture. Zapatero was invested in a first-round vote with the positive votes of this own party, along with those of IU, ERC, BNG, and CHA, while CC, CiU, PNV, EA, and Na-Bai abstained. The PP deputies voted no. Yet the positive investiture votes did not entail a commitment of ongoing support, except in the case of CC. In 2008, Zapatero did not even attempt to assure a first-round investiture. Instead, parliament elected him in a second-round relative majority vote of only the members of his own party. This was the first time since 1981 that parliament selected a PM in a second-round vote. Nonetheless, Zapatero’s investiture was never in question. The governments took 35 (Zapatero I) and 36 (Zapatero II) days to form. With the new party system that emerged in 2015, Spain, for the first time, faced serious challenges forming a government. This is related to the transformation of the party system with the rise of Podemos and Cuidadanos (Simón 2016) and changes in the Catalan parties (Field 2016b), as well as contextual factors. For the first time, bargaining amongst the larger statewide parties was required. In 2015, the PP won the largest number of seats, 119—but the fewest seats of any lead party since the transition. Parliamentary arithmetic and political divisions were tricky. An alliance of the right and centre-right parties, including newcomer Ciudadanos and the centre-right regional nationalist parties, added up to 178 votes, 2 over an absolute majority. However, in contrast to the previously moderate Catalan CiU, CDC embraced independence. CDC held eight seats. This complicated an alliance on the right—not only because of CDC’s stance but also because Ciudadanos, a party strongly opposed to Catalan nationalism, refused to support a minority government that also relied on secessionist parties.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



561

On the left, the hurdles were also high. The PSOE, while it held onto its position as the lead party of the left, faced a serious challenge from newcomer Podemos. The Socialists won 90 seats, while Podemos won 65. Collaboration between the two was risky for the Socialists who feared it would bolster Podemos’ stature. For Podemos, collaboration with insider PSOE could dampen its anti-establishment credentials. Even if they were able to agree, their joint votes were not sufficient to elect a candidate. The incumbent PM Rajoy turned down the monarch’s offer to put him forth as the PM candidate. Rajoy and the PP did not have the allies necessary to form a government. This occurred in the context of declining public support for the party and the corruption charges engulfing the party—in some cases implicating Rajoy himself. Pedro Sánchez, PSOE party leader, did accept the king’s charge to try to form a government. The PSOE signed an agreement with Ciudadanos, which assured its support for Sánchez’s investiture. Sánchez also reached an agreement with the Canary CC, which had one deputy, to abstain in a second-round investiture vote. However, Sánchez could not secure either the affirmative support or the abstention of the MPs of Podemos or the PP, one of which was necessary for him to be elected. While there was much talk of cooperation between PSOE and Podemos, ultimately competition between the two parties for leadership of the left, as well as the policy distance between Podemos and Ciudadanos, prevented any such deal. Sánchez lost the investiture vote. New elections were called to take place in June 2016. One of the big questions was whether Podemos would overtake the PSOE; it did not. The elections produced broadly similar results—though they strengthened the PP. In this context, Rajoy attempted to form a government and signed a support agreement with Ciudadanos, assuring its support for Rajoy in the investiture vote, and with the Canary CC, which had one seat in the Congress. Yet this was not enough to win office in the parliamentary investiture votes that took place in August and September 2016. PSOE leader Sánchez had refused to have the PSOE deputies abstain to facilitate the formation of a Rajoy-led PP minority government. With the PSOE severely divided on how to deal with forming a government, the party forced Sánchez’s resignation. Under new leadership, the party decided the PSOE deputies would abstain in a second investiture attempt by Rajoy in late October 2016. Most MPs followed orders, yet several broke party discipline to vote no—in this normally highly disciplined party. Rajoy was invested with a simple majority of 170 votes, 6 short of an absolute majority. Bargaining over the formation of the Sánchez I government in 2018 was also anomalous. Sánchez had fought his way back in a leadership battle. After a court ruling that the PP had benefitted from an illegal kickback scheme in the early 2000s, Sánchez rallied disparate parliamentary actors to remove the Rajoy

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

562



government in a constructive vote no confidence and simultaneously invest him as the PM. The alliance included Podemos and multiple regionally based parties, including the Catalan and Basque secessionist parties.

Cabinets and ministerial portfolios Because Spain regularly formed single-party governments, either of the UCD, PSOE, or PP, the allocation of ministerial portfolios has not been as complex as in other Western European countries. Additionally, Spain only had one caretaker cabinet prior to 2019—as defined in the collective volume—Rajoy III between 2015 and 2016. Since re-democratization, it has never had a technocratic PM or cabinet. Because of the lack of coalitions, this chapter does not contain a table on portfolios. Nonetheless, it is useful to discuss some details of ministerial portfolios. Between 1977 and 2018, Spain’s governments had 16 ministries on average, ranging from a low of 14 to a high of 19 during the Suárez III government at the outset of democracy. On average, governments have had 17 cabinet members. In the earlier years of democracy, deputy PMs did not simultaneously serve as ministers, though since 1996 it has been the norm. While the rules still allow for the appointment of ministers without a portfolio, Spain has not had one since the Aznar II government in 2000. As in other Western European countries, the PM’s office is very important, and as mentioned previously, the size of the PM’s office has increased significantly over the democratic period (Parrado 2017: 79). In addition, since the transition, there has been a ‘presidency’ minister or equivalent (the PM is referred to as the president of the government) with its own portfolio. Between 1996 and 2019, this portfolio was held by the only deputy PM or the first deputy PM responsible for political affairs, giving it paramount political importance. The minister of the presidency monitors and coordinates the actions of ministers. While more systematic research needs to be done on the relative importance of portfolios in Spain (and across parties and time), the political importance of the ministry of economy and/or finance is apparent, and at times equally or more important than the presidency portfolio. It is noteworthy that some governments have had separate finance (hacienda) and economic (economía) portfolios, while others have combined them. Several governments have also appointed a second deputy PM focusing on economic matters that simultaneously occupies the economy or economy and finance post, including PMs Aznar and Zapatero. Moreover, similar to other Western European countries, the ministries of interior and foreign affairs are considered among the most important; however, interior in the Spanish context is likely a more important post than foreign affairs. Ministries responsible for social policies may not be as important as in other political contexts due to the devolution of social policy competencies to the autonomous communities.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



563

Governing in minority Between 1977 and 2019, Spain did not have a coalition government. Nonetheless, it is important to analyse how minority governments govern, and their relationship with the political parties that remain outside of the cabinet.²³ It is useful to distinguish between formal and substantive minority governments (Strøm 1990). Formal minority governments are those that rely on agreements with at least one support party that does not take cabinet positions and, if their parliamentary seats are counted, the government has the support of an absolute majority of MPs. I define a formal support party as one that (1) has reached an explicit agreement to, at least, support the investiture of the PM and (2) votes in favour of the PM’s investiture. In this conceptualization, it does not necessarily commit the party to ongoing support. Substantive minority governments either do not rely on support agreements or the party (parties) that enters such agreements do not add up to a majority. Excluding the peculiar González III government (1989–1993) and the Rajoy caretaker cabinet in 2015, only the González IV (1993–1996) and Aznar I (1996–2000) governments were formal minority governments. Both relied on support agreements with centre-right regionally based parties. These agreements were made explicit through the media or in the investiture debate, or culminated in written documents. The remaining six were substantive minority governments. Occasionally parties have negotiated a written policy agreement. For example, in 1996, the PP under José María Aznar and the Catalan CiU negotiated an ‘investiture and governance agreement’, the so-called Majestic Pact. In 2016, the PP and Cs entered an agreement titled ‘150 commitments to improve Spain’ that undergirded the Rajoy II government. The first document was less extensive than the second. Measured in words, that of PP and Cs was four times longer than the Majestic Pact. Nonetheless, the Rajoy II government was a substantive minority government. The governing party’s MPs along with those of its support parties (Cs and CC) did not amount to a majority. The government thus needed to negotiate with other parties in parliament, which often included the Basque PNV. Other minority governments have used a strategy of shifting alliances, termed ‘variable geometry’ or ‘geometría variable’ in Spain. This occurred, in particular, during the Socialist governments of PM Zapatero (2004–2011). During both governments, the governing party occupied the central policy position on the left–right and centre–periphery policy dimensions. Thus, it was often in a position to negotiate with distinct allies.

²³ The analysis in this section draws on Field (2016a) and the dozens of interviews carried out for that book. Since Spain had no (formal) coalition cabinets in the observation period, some tables that are present in the other chapters do not appear in this chapter.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

564



The Sánchez I government (2018–2019) did not count on a written agreement with support parties. Since the investiture occurred on the basis of the vote of no confidence against the Rajoy government, due to its constructive nature, it is difficult to identify formal support parties. Nonetheless, Podemos clearly expressed a willingness to support a PSOE-led government, if the motion succeeded. In practice, the short-lived government engaged in ongoing negotiations with leftist (mainly Podemos) and regionally based parties. Negotiations with support parties and other potential allies take place at various levels. Interviews with ministers from the governments before 2015 indicated that negotiations normally take place between the governing party’s spokesperson in parliament and the relevant representative of the potential ally in parliament. If negotiations stall or problems arise, or the subject is of critical importance to the government, a minister may engage in direct negotiations with leaders of support parties. In the most important negotiations, the PM (or a designee) may negotiate directly with the leader of the support party. Often, this has been the leader of a regional party who may also be the head of a regional government. When agreements are reached, whether to support the government as a whole or a particular piece of legislation, including budgets, monitoring typically takes place in parliament, through parliamentary debates and questions. Parties have also formed inter-party committees to monitor the progress of agreements. Support parties also track the fulfilment of agreements from their party headquarters.

The governing capacity of minority governments It is also important to consider the governing capacity of minority governments.²⁴ My research on governments in Spain between 1982 and 2015 finds that minority governments exhibited strong governing capacity, equivalent to majority ones (Field 2016a). For example, minority governments approved 88 per cent of their bills and lasted 3.51 years, compared to 89 per cent and 3.56 years for majority ones. They also tended to fulfil their election pledges (Artés 2013). However, in recent years, minority governments in Spain have not worked well. The PP government of Rajoy III, formed after the 2016 elections, lasted only 575 days (1.58 years). At that time, it was the briefest government since the transition. The Socialist government of Sánchez I, formed in 2018, lasted only 326 days (.89 years). In addition, both governments had difficulty carrying out their political agendas and approving budgets, which contributed to Sánchez calling early elections in 2019.

²⁴ This section reproduces analyses from (Field 2015, 2016b, 2019).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



565

To help account for governing capacity, here I apply an analytical framework, developed in Field (2016a), that is based on three factors: political institutions, the government’s partisan bargaining position, and the reconcilability of party goals. 1982–2015: In this period, Spain’s minority governments were effective in part because of parliamentary institutions that strengthen the government, partisan bargaining circumstances favourable to the government, and the alliances between the statewide governing party and the regionally based parties. The design of Spain’s parliamentary institutions tips the balance more in favour of the government than do the parliamentary institutions in most European countries. The constructive vote of no confidence, strong governmental agendasetting capabilities, and low effective voting thresholds for decision-making in parliament all strengthen the government’s bargaining position, making governing in minority easier. With their specificities, these minority governments also had strong bargaining positions because the governing party (always only one) controlled a large number of seats. Moreover, the governing party occupied a central policy position in parliament, allowing it to shift allies, and/or it encountered potential support parties in Congress that needed political reinforcement to be able to govern, or govern more easily, at the regional level. The latter relates to whether regional parties governed in minority or in coalition with the statewide party governing Spain. The objectives of the parties were reconcilable because the government (of the PSOE or PP) was able to ally with moderate regionally based parties from Catalonia, the Basque Country, the Canary Islands, or elsewhere. These parties had little to no interest in governing Spain but were willing to exchange their support for concessions on public policies, regional investment, decentralization, and/or political support at the regional level. 2015–2019: These three factors also help us understand why the Rajoy III and Sánchez I minority governments did not work well. While the governments still counted on some institutional advantages, they were in weaker bargaining positions and reconciling party goals was more difficult. This is in part related to the changes of the party system in 2015. These minority governments had fewer seats of their own in parliament— Rajoy’s PP held 39 per cent of the seats in the Congress of Deputies and Sánchez’s Socialists held 24 per cent, compared with between 45 per cent and 48 per cent for the previous minority governments. Measured by the Chapel Hill Expert Survey, the PP under Rajoy III did not occupy the central position in parliament on either of the two main axes of party competition—the left–right and centre–periphery cleavages. The PSOE under Sánchez I only occupied it on the centre–periphery dimension. In terms of reconciling party goals, some regionally based parties, such as the PNV and the CC, continued to provide support for minority governments.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

566



However, polarization surrounding the political status of Catalonia and the Catalan nationalist parties’ move to secessionism made getting and governing with their support far more difficult. However, support from regionally based parties would not have been enough either, as it had been before 2015. The post-2016 minority governments also had to seek support from the new parties, Ciudadanos or Podemos, who were strong electoral competitors with interest in displacing the dominant parties from power or overtaking them electorally. While Spain’s relevant formal institutions did not change, the Sánchez I 2018 government and its allies never gained control of the parliament’s governing board, which has agenda-setting power. Instead, there was a rightist majority of the PP and Ciudadanos, established after the 2016 elections. This allowed the political right to stymie the Sánchez government’s parliamentary agenda. In part as a result, the Sánchez government passed many laws through executive decree. My prior work showed that decree power helped minority governments advance their agendas. Minority governments issued an average of 17 decrees per year in that study. The Rajoy III minority government issued 30 in little more than a year and a half (for an annualized average of 19). The Sánchez I government issued 35 decrees in less than a year in office.

Cabinet duration and termination The duration of cabinets Spain is among the Western European countries with the most durable governments. Excluding the Rajoy caretaker government, governments lasted 75 per cent of their maximum possible term—i.e. their relative duration—between 1977 and 2019 (see Table 16.3).²⁵ However, there are important differences during this time span. In the early post-transition years, when democracy and political parties had yet to consolidate, relative government duration averaged 54.5 per cent. Between 1982 and 2015, relative government duration averaged 90 per cent. Moreover, there was little difference between the average duration of minority (88 per cent) and majority (92 per cent) governments. Since the party system changed in 2015, relative government duration decreased to an average of 42 per cent (excluding the Rajoy caretaker government). ²⁵ Per the 1978 constitution, the lifespan of parliament is four years from its election (Art. 68.4), and elections must occur between 30 and 60 days from the end of its mandate (Art. 68.6). The 1985 election law further stipulated that parliament is automatically dissolved 25 days before the end of its term, unless early parliamentary elections are called. The next day, the elections are officially called. Elections are then held 54 days after the official call. In practice, the government has typically timed the call for elections so that they occur on a Sunday.

1996-05-06 2000-04-28 2004-04-18 2008-04-14

2011-12-22 2015-12-20 100 2015-12-20 2016-06-26 100 2016-11-04 2018-06-02 42.3

2018-06-06 2019-04-28 41.8

Aznar I Aznar II Zapatero I Zapatero II

Rajoy I Rajoy II Rajoy III

Sánchez I

98.5 98.1 97.6 90.4

4, 6

1 2 6

1 1 1 4, 9

8, 9 4, 8 1 4 4, 8 4, 6, 8

4 UCD UCD

11

11, 13

11, 12, 13 11, 13 PSOE 11, 12 PSOE

12 10

Inability to form a government triggered new elections. National Court’s verdict found that the PP participated in extensive corruption in the so-called Gürtel case on 24 May 2018. The government’s budget proposal was defeated in parliament.

Early elections precipitated by a lack of public support due to Spain’s economic crisis.

General strike; end of European presidency Corruption scandals The government’s budget proposal was defeated in parliament because a support party refused to support it. Corruption and abuse of authority scandals. End of European presidency

Early elections were called after the new democratic constitution was approved. (12) Rumours of a possible military coup.

Notes: Relative duration: share of constitutionally allowed duration for this cabinet. Rajoy II cabinet is a caretaker cabinet.

Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of PM

2000-03-12 2004-03-14 2008-03-09 2011-11-20

44.7 76.8 87.3 81.7 87.9 66.3

1979-04-06 1981-02-27 1982-12-03 1986-07-26 1989-12-07 1993-07-14

Suárez III Calvo-Sotelo González I González II González III González IV

1981-01-29 1982-10-28 1986-06-22 1989-10-29 1993-06-06 1996-03-03

1977-07-05 1979-03-01 41.9

Suárez II

Policy Comments Relative Mechanisms Terminal Parties (when events conflict between area(s) duration of cabinet or within) termination (%)

Date in

Cabinet

Date out

Table 16.3 Cabinet termination in Spain, 1977–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

568



The termination of cabinets Cabinets in Spain have typically ended with regular or early parliamentary elections (see Table 16.3). Using the project’s decision rule that defines an early election as one occurring prior to the point at which there is 10 per cent of the maximum term remaining, eight cabinets ended with early parliamentary elections. However, using a qualitative approach, González I was not viewed as calling early elections due to a particular event or governance challenges. On the other hand, the Zapatero II government, while lasting 90 per cent of its term, was viewed to have called early elections in 2011 due to political problems related to the economic crisis. Between 1977 and 1982 governments were short-lived. The first (minority) government in post-transition Spain, under the leadership of PM Adolfo Suárez, lasted approximately two years. However, its short tenure was not due to a political crisis. The PM called early elections following the approval of Spain’s new democratic constitution. However, the Suárez II minority cabinet ended in 1981 when the PM resigned. The exact motivations are still debated. There were clear divisions in the ruling party, and there is also speculation that sabre-rattling from the military may have influenced his decision. PM Calvo-Sotelo called early parliamentary elections in 1982 due to intra-party conflict. The majority governments of PM González lasted longer, but not their full terms. González I was not viewed as calling early elections due to a particular event or governance challenges. In 1989, PM González (II) called early elections in the face of a general strike and at the end of Spain’s European presidency. While the government was down in the polls, there was little risk of the PSOE losing to the main opposition party. PM González (III) also called elections a bit early in June 1993. This was in the context of declining public support, after more than a decade in government, and rising corruption scandals. The results produced a minority situation in parliament; yet, the PSOE held on to its position as the lead party. In 1996, PM González (IV) called elections primarily in response to a parliamentary defeat. In the context of growing corruption scandals surrounding the governing party and accusations of the abuse of state power in the fight against terrorism, CiU decided it could no longer support the González minority government. CiU refused to approve the government’s budget, and CiU’s leader, Jordi Pujol, negotiated the end of the government with González. Both agreed that CiU’s support would end with Spain’s presidency of the European Union (Field 2016a: 102–3). Subsequently, the Aznar I minority government, the Aznar II majority government, and the Zapatero I minority government served their full terms. While the Zapatero I government (2004–2008) lasted nearly the entire constitutional term, PM Zapatero called elections a few months early, in 2011, which ended his second

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



569

government (2008–2011). This occurred in the context of increasing public frustration with the economic crisis and the government’s response to it. The minority government did not end due to the withdrawal of parliamentary support or because of a clear indication that it could not find the parliamentary allies to govern (Field 2016a: 119). The elections brought to power a PP absolute majority government under Rajoy, which lasted the full term (2011–2015). The Rajoy II caretaker cabinet is anomalous. It ended in 2016 because parliament was unable to form a government, which triggered new elections. The two subsequent minority governments terminated because of government defeats in parliament. The Rajoy III government was the first to be brought down in a vote of no confidence since the transition to democracy. A court ruling that found the PP had participated in extensive corruption in the early 2000s triggered the motion. In contrast, the Sánchez I minority government terminated when it called early elections after parliament rejected its budget bill.

Conclusions During the period covered in this volume, Spain stood out because of its distinct lack of coalitions and the prevalence of minority governments. During the early democratic period (1977–1982), the minority governments easily formed yet were comparatively brief. The 1977 Suárez government was able to govern effectively, while subsequent governments faced greater challenges. Between the consolidation of democracy in 1982 and 2015, minority and majority governments formed rather easily, governed without great difficulty, and lasted—though, of course, with some variation. In contrast, between 2015 and 2019, Spain experienced greater difficulty forming governments and problems of governability. In 2020, a coalition government formed in Spain for the first time since redemocratization; yet the prevalence of minority governments continues as it is a leftist minority coalition of the PSOE and Unidas Podemos, providing the opportunity to study coalition governance in Spain. The Spanish case also presents lessons for the comparative literature on minority governments and coalitions. Many of the institutions that scholars suggest encourage coalition governments have not done so in Spain. Scholars of Spanish politics have demonstrated the need to incorporate the decentralized state institutions in order to understand the life cycle of governments. Multilevel dynamic affect government formation, governability, and government duration in Spain. The decentralized state and the party system with strong left–right and a centre– periphery dimensions of competition structure the multilevel dynamics. Spain also illustrates the need to better understand how minority governments govern and the mechanisms that facilitate or complicate governability during minority governments. Minority governments need to manage relationships

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

570



with support parties and other potential allies. In addition, minority coalitions need to manage inter-party relations among parties in cabinet. With greater political fragmentation in Europe, minority governments are likely to become even more common. We have much to learn about their life cycle.

References Aguilera de Prat, Cesáreo R. (2001a). ‘Convergència i Unió ante los pactos de gobernabilidad de 1993 y 1996’. Sistema, 165: 99–129. Aguilera de Prat, Cesáreo R. (2001b). ‘Los socialistas ante los pactos de gobernabilidad de 1993 y 1996’. Revista de Estudios Políticos (Nueva Época), 111: 9–43. Alonso, Sonia (2012). Challenging the State: Devolution and the Battle for Partisan Credibility. Oxford: Oxford University Press. Alonso, Sonia, and Cristóbal Rovira Kaltwasser (2015). ‘Spain: No Country for the Populist Radical Right?’. South European Society and Politics, 20(1): 21–45. Artés, Joaquín (2013). ‘Do Spanish Politicians Keep Their Promises?’. Party Politics, 19(1): 143–58. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010.’ Party Politics, 21(1): 143–52. Barrio, Astrid, and Bonnie N. Field (2017a). ‘Is Spain Francoland? No. Has the Catalan Crisis Damaged Spanish Democracy? Yes.’. Monkey Cage. The Washington Post. 22 December. Barrio, Astrid, and Bonnie N. Field (2017b). ‘Spain Is Trying to Block Catalonia’s Independence Referendum Tomorrow. What’s Going On?’. Monkey Cage. The Washington Post. 30 September. Bergman, Torbjörn (1993). ‘Formation Rules and Minority Governments’. European Journal of Political Research, 23(1): 55–66. Bosch, Agusti, and Iván M. Durán (2017). ‘How Does Economic Crisis Impel Emerging Parties on the Road to Elections? The Case of the Spanish Podemos and Ciudadanos’. Party Politics, 25(2): 257-67. Colomer, Josep M., and Florencio Martínez (1995). ‘The Paradox of Coalition Trading’. Journal of Theoretical Politics, 7(1): 41–63. Cordero, Guillermo, and José Ramón Montero (2015). ‘Against Bipartyism, Towards Dealignment? The 2014 European Election in Spain’. South European Society and Politics, 20(3): 357–79. Druckman, James N., and Michael F. Thies (2002). ‘The Importance of Concurrence: The Impact of Bicameralism on Government Formation and Duration’. American Journal of Political Science, 46(4): 760–71.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



571

Druckman, James N., Lanny W. Martin, and Michael F. Thies (2005). ‘Influence without Confidence: Upper Chambers and Government Formation’. Legislative Studies Quarterly, 30(4): 529–48. Falcó-Gimeno, Albert, and Tània Verge (2013). ‘Coalition Trading in Spain: Explaining State-wide Parties’ Government Formation Strategies at the Regional Level’. Regional and Federal Studies, 23(4): 387–405. Field, Bonnie N. (2015). ‘How the Next Spanish Government Might Be Formed after Spain’s Election’. LSE European Politics and Policy (EUROPP) Blog. Field, Bonnie N. (2016a). Why Minority Governments Work: Multilevel Territorial Politics in Spain. New York: Palgrave Macmillan. Field, Bonnie N. (2016b). ‘Will the New Rajoy Minority Government in Spain Work?’. LSE European Politics and Policy (EUROPP) Blog. Field, Bonnie N. (2019). ‘Do Spain’s Minority Governments Work?’. LSE European Politics and Policy (EUROPP) Blog. Field, Bonnie N., and Alfonso Botti (2013a). ‘Introduction: Political Change in Spain, from Zapatero to Rajoy’. In Bonnie N. Field and Alfonso Botti (eds), Politics and Society in Contemporary Spain: From Zapatero to Rajoy. New York: Palgrave Macmillan, 1–19. Field, Bonnie N., and Alfonso Botti (eds) (2013b). Politics and Society in Contemporary Spain: From Zapatero to Rajoy. New York: Palgrave Macmillan. Field, Bonnie N., and Kerstin Hamann (2015). ‘Framing Legislative Bills in Parliament: Regional-Nationalist Parties’ Strategies in Spain’s Multinational Democracy’. Party Politics, 21(6): 900–11. Gillespie, Richard, and Caroline Gray (eds) (2015). Contesting Spain? The Dynamics of Nationalist Movements in Catalonia and the Basque Country. London and New York: Routledge. Gomez Fortes, Braulio, and Ignacio Urquizu (2015). ‘Political Corruption and the End of Two-Party System after the May 1015 Spanish Regional Elections’. Regional and Federal Studies, 25(4): 379–89. Gómez-Reino Cachafeiro, Margarita (2011). ‘The Bloque Nacionalista Galego’. In Anwen Elias and Filippo Tronconi (eds), From Protest to Power: Autonomist Parties and the Challenges of Representation. Vienna: Braumüller, 125–46. Harguindéguy, Jean-Baptiste, Xavier Coller, and Alistair Cole (2017). ‘Why Is the Spanish Upper Chamber So Difficult to Reform’. Parliamentary Affairs, 70(3): 530–47. Heller, William B. (2002). ‘Regional Parties and National Politics in Europe: Spain’s Estado de las Autonomías, 1993 to 2000’. Comparative Political Studies, 35(6): 657–85. Medina, Iván, and Patricia Correa (2016). ‘The 2015 Spanish Elections: The Times They Are A’ Changing’. Regional and Federal Studies, 26(3): 407–17. Miley, Thomas Jeffrey (2016). ‘Austerity Politics and Constitutional Crisis in Spain’. European Politics and Society, 18(2): 263–83.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

572



Müller, Wolfgang C., and Kaare Strøm (eds.) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Orriols, Lluis, and Guillermo Cordero (2016). ‘The Breakdown of the Spanish TwoParty System: The Upsurge of Podemos and Cuidadanos in the 2015 General Election’. South European Society and Politics, 21(4): 469–92. Parrado, Salvador (2017). ‘El gobierno y el diseño organizativo de la administración central’. In José Antonio Olmeda, Salvador Parrado, and César Colino (eds), Las Administraciones Públicas en España. Valencia: Tirant lo Blanch, 77–113. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of Antielitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Real-Dato, José, and Juan Rodríguez Teruel (2016). ‘Politicians, Experts or Both? Democratic Delegation and Junior Ministers in Spain’. Acta Politica, 51: 492–516. Reniu i Vilamala, Josep María (2002). La formación de gobiernos minoritarios en España, 1977–1996. Madrid: Centro de Investigaciones Sociológicas. Rodon, Toni, and María José Hierro (2016). ‘Podemos and Ciudadanos Shake up the Spanish Party System: The 2015 Local and Regional Elections’. South European Society and Politics, 21(3): 339–57. Rodríguez Teruel, Juan (2011). Los ministros de la España democrática: Reclutamiento político y carrera ministerial de Suárez a Zapatero (1976–2010). Madrid: Centro de Estudios Políticos y Constitucionales. Rodríguez Teruel, Juan, and Astrid Barrio (2016). ‘Going National: Ciudadanos from Catalonia to Spain’. South European Society and Politics, 21(4): 587–607. Rodríguez Teruel, Juan, Astrid Barrio, and Oscar Barberà (2016). ‘Fast and Furious: Podemos’ Quest for Power in Multi-level Spain’. South European Society and Politics, 21(4): 561–85. Simón, Pablo (2016). ‘The Challenges of the New Spanish Multipartism: Government Formation Failure and the 2016 General Election’. South European Society and Politics, 21(4): 493–517. Strøm, Kaare (1990). Minority Government and Majority Rule. Cambridge: Cambridge University Press. Strøm, Kaare, Ian Budge, and Michael J. Laver (1994). ‘Constraints on Cabinet Formation in Parliamentary Democracies’. American Journal of Political Science, 38(2): 303–35. Strøm, Kaare, Wolfgang C. Muller, and Torbörn Bergman(eds.) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Vidal, Guillem (2018). ‘Challenging Business as Usual? The Rise of New Parties in Spain in Times of Crisis’. West European Politics, 41(2): 261–86. Vintró Castells, Joan (2007). La investidura parlamentaria del Gobierno: perspectiva comparada y Constitución Española. Madrid: Congreso de los Diputados.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



573

Appendix. List of political parties Abbreviation HB Amaiur Podemos CPV IU

BNG ERC PSOE

PSP PSA PNV CC UPyD Cs CiU

CDS UCD PP PDP

Name United People (Herri Batasuna) Amaiur We can (Podemos) Commitment Coalition (Coalició Compromis) Commitment for the Valencian Country (Compromís pel País Valencià) United Left (Izquerda Unida), 1986– Communist Party of Spain (Partido Comunista de España), 1977–1986; also includes United Socialist Party of Catalonia (Partit Socialista Unificat de Catalunya) Galician Nationalist Bloc (Bloque Nacionalista Gallego) Republican Left of Catalonia (Esquerra Republicana de Catalunya) Spanish Socialist Workers’ Party/Socialist Party (Partido Socialista Obrero Español); also includes Socialist Party of Catalonia (Partit dels Socialistes de Catalunya) Popular Socialist Party (Partido Socialista Popular) Socialist Party of Andalusia-Andalusian Party (Partido Socialista de Andalucía-Partido Andaluz) Basque Nationalist Party (Partido Nacionalista Vasco) Canary Islands Coalition (Coalición Canaria) Union Progress and Democracy (Unión Progreso y Democracia) Citizens (Ciudadanos) Catalan European Democratic Party (Partit Demòcrata Europeu Català), 2016– Democratic Convergence of Catalonia (Convergènca Democràtica de Catalunya), 2016 Democracy and Freedom (Democràcia i Llibertat), 2015–2016 Convergence and Union (Convergència i Unió), 1979–2015 Democratic and Social Centre (Centro Democrático y Social) Union of the Democratic Centre (Unión de Centro Democrático) Popular Party (Partido Popular), 1989– Popular Alliance (Alianza Popular), 1977–1989 Popular Democratic Party (Partido Demócrata Popular)

Note: Party names are given in English, followed by the party name in the party’s native language (or other languages, for multilingual parties or polities) in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 17 Sweden: The Rise and Fall of Bloc Politics Johan Hellström and Jonas Lindahl

Historically, Swedish coalition politics has been characterized by a nearunidimensional political conflict along the left–right economic divide and minority government rule. While the tendency towards minority government and the institutional factors underlying this remain intact, developments in the party system following the 2010 Riksdag elections have complicated matters thoroughly.¹ In the 2000s both of the two traditional political blocs of the Swedish party system, the socialist (or centre-left) bloc and the non-socialist (or centreright) bloc moved towards greater within-bloc cooperation, entrenching the bipolar system further and presenting the voters with something of a two-party system. However, the entry of the populist radical right party Sweden Democrats into parliament in 2010 upended this entrenchment, as neither political bloc could realistically expect to win a majority of the parliamentary seats. Attempts to isolate the Sweden Democrats from direct influence over cabinet formation and policy has resulted in a number of compromises with one overarching goal: allowing minority governments to rule.

The institutional setting The foundation of the current system of government was introduced in the Swedish constitution of 1809. Typical of its era, it was based on Madisonian system of separation of powers between the king and the parliament. Formally, the prime minister was appointed by the king, as were the individual cabinet ministers (at the suggestion of the prime minister). However, the political system gradually developed into a parliamentary system of government and the king ¹ Some analyses in the chapter are based on interviews with junior ministers and cabinet members: Junior minister A (2017), interview conducted on 29 September 2017; Cabinet member A (2017), interview conducted on 5 October 2017; Junior minister B (2017), interview conducted by telephone on 28 August 2017 and 3 October 2017; Cabinet member B (2017), interview conducted on 22 August 2017; Junior minister C (2017), interview conducted on 30 August 2017; Junior minister D (2017), Interview conducted on 29 September 2017. Johan Hellström and Jonas Lindahl, Sweden: The Rise and Fall of Bloc Politics In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0017

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



575

eventually lost most of his political powers when parliamentarism was firmly established in 1917. Before 1970s Sweden was an example of a strongly bicameral system of parliament, comprised of the directly elected lower chamber and the indirectly elected upper chamber. Thus, on most legislative matters, separate majorities were needed in both chambers to pass legislation. However, as part of the major revisions of the constitution in the late 1960s and 1970s the upper chamber was abolished in 1971. Later with the implementation of the Instrument of Government of 1974 in 1975, parliamentarism also became the formal form of government.² The latest constitutional reform of 2010 did not alter the basic features of Swedish institutional system. Its peculiarities, such as negative parliamentarism and the non-involvement of the head of state in appointing the prime minister (and therefore government), a task left to the Speaker of the Riksdag (parliament), remain intact.

Electoral system Elections to the unicameral parliament—the Riksdag—are held every four years on a fixed election date. The election specifically concerns political parties, with voters casting their ballot for the party lists. Beginning with the 1998 election, voters can also choose a specific candidate on a party’s list. Nonetheless, even after the introduction of the (partially) open-list system, the political parties as a whole, and not their candidates, are at the centre in election campaigns.³ Of the parliament’s 349 seats, 310 are fixed seats apportioned to the parties using a modified Saint-Laguë method, based on the results in the electoral districts (29 in the 2018 election), and the remaining 39 seats are adjustment seats apportioned on the basis of the party’s performance for the entire country. Although the electoral system results in highly proportional election outcomes in a comparative perspective (Lijphart 2012: 149–52), the inclusion of a four per cent threshold for a party to be eligible for any seats introduces a minor degree of disproportionality. Several of the minor parties in the Riksdag, such as the Christian Democrats, the Green Party, and the Liberals, have frequently found themselves close to or below the threshold in opinion polls. Although, the most recent changes in the electoral laws, in effect during the 2018 election, have further increased the degree of

² The constitutional basis for the division of power and the principles for the political system is stated in one of the four fundamental laws compromising the Basic laws of Sweden (i.e. the Swedish constitution), namely the Instrument of Government (regeringsformen). ³ There is a threshold of five per cent of the candidate’s party’s votes in an electoral district for these ‘personal’ votes to count, at which point they bypass the candidate order on the party’s list. Even though the threshold has been lowered from the previous eight per cent there were only twelve candidates that were elected into parliament on a ‘personal vote’ in the 2014 election. In the 2018 election the number dwindled to merely five candidates.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

576

ö  

proportionality in an already highly proportional electoral system, the electoral threshold remains unchanged.

Parliamentary procedures Before the constitutional reforms in 1974, the king was involved in the government formation process as an informatuer, a role that the Speaker of the Riksdag took over in 1975. The formal process involves the Speaker consulting with the leaders of the parties represented in the Riksdag on possible candidates for prime minister—in practice who can build a sustainable government, either as a coalition or as a single party, with sufficient support in the Riksdag to pass the investiture vote. In the government formation process the Speaker has little involvement in the actual bargaining process however, the details of which we will discuss in further in this chapter. The investiture vote—formally a vote on the candidate for prime minister— implemented with the 1974 Instrument of Government and first taking effect after the 1976 election is different from those in most other European countries. The voting rule is negatively formulated, in that an absolute majority of the Riksdag (i.e. at least 175 members of parliament, MPs) is needed to vote against a suggested prime minister candidate for the investiture to fail. Thus, the Swedish negative parliamentary system differs from most other European countries in that a government must only be tolerated by the parliament and does not need to have the support of a majority of its members. This form of institutionalized negative parliamentarism is one contributing factor to the high share of minority governments in Sweden (Bergman 1993, 1995). Consequently, a given candidate does not need to be approved by a majority of all MPs but must at the very least be tolerated to enter office and form a government. This rule originates from, and also applies to, motions of no-confidence in the prime minister or individual ministers, in that a majority of all Riksdag members must vote in favour of the motion of noconfidence for it to succeed. A recent change to the investiture rule involves when it applies. In its original formulation, only non-incumbent prime minister candidates proposed by the Speaker of the Riksdag needed to pass an investiture vote. Incumbent prime ministers could simply remain in office until they resigned or were removed through a motion of no-confidence. Following a 2010 amendment to the Instrument of Government, incumbent prime ministers need to subject themselves to a mandatory vote of confidence unless the prime minister resigns after a parliamentary election. Such a vote follows the same voting rules as for an investiture vote for prime minister. The first such vote took place following the 2018 election, when incumbent prime minister Stefan Löfven was predictably voted out of office.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



577

The party system and the actors Party system change As Table 17.1a illustrates, for much of the post-Second World War period, Sweden’s party system has remained remarkably stable, including only five parliamentary parties, with splits and divisions being nearly non-existent (Berglund and Lindström 1978: 16–18; Bergman 2000). The five-party system was comprised of a communist party (the Left Party Communists, later the Left Party), a social democratic party (the Social Democrats), an agrarian party (the Centre Party), a liberal party (the People’s Party, and later the Liberals), and a conservative party (the Moderates, originally the Right Party). Beginning in the late 1980s, the Riksdag saw the entry of three new parties, one only briefly. In 1988, the Green Party entered the Riksdag, exiting in 1991 after failing to pass the four per cent threshold, before getting into parliament again in 1994. In 1991, the Christian Democrats and populist radical right New Democracy entered the Riksdag.⁴ While the latter party proved to be short-lived, failing to gain any seats in any subsequent election, both the Green Party and the Christian Democrats have maintained their parliamentary presence. In the parliamentary election of 2010 the most recent newcomer, the populist radical right Sweden Democrats was able to secure representation in the parliament and won 20 of the 349 seats. Alongside the entry of new parties, the old parties have also experienced shifts in their ideological positions. The Social Democrats, as left-of-centre parties elsewhere, in the 1990s embraced Third Way politics. Early in the same decade, the Left Party Communists rebranded to just the Left Party, alongside abandoning its ties to the disintegrating Soviet Union and transitioning to democratic socialism. The Moderates had already rebranded themselves as a liberal-conservative party when they abandoned its Right Party moniker in 1969. The Centre Party, while still with focusing on issues relating to rural communities has shifted to promoting more market-liberal policies. Similarly, the Liberals took a less social-liberal and more market-liberal tack in the latter half of the 1990s. The Swedish five-party system model could broadly be divided into two ‘blocs’:⁵ a ‘socialist’ (or left to centre-left bloc) and a ‘bourgeois’ (or centre-right to right bloc),⁶ ⁴ While the majority of the parties are presented here with a direct translation of their Swedish party name, we have opted to follow Swedish parlance in the case of the Moderates (formally the Moderate Unity Party) and the Social Democrats (formally the Social Democratic Workers’ Party). We also refer to what is formally the Environmental Party, the Greens, as the Green Party, in line with what environmentalist parties are commonly referred to in English. ⁵ See Green-Pedersen and Hoffman Thomsen (2005) for an in-depth discussion on what exactly constitutes bloc politics. ⁶ The use of bourgeois, including by the parties themselves, is not a sign of an underlying classtheoretical perspective and is also generally used with no pejorative connotations. The bourgeois bloc will also be referred to interchangeably as the non-socialist or centre-right bloc. The socialist bloc will also be referred to as the centre-left, red, or red-green bloc.

1945-07-31 1946-10-11 1948-09-19 1951-10-01 1952-09-21 1956-09-16 1957-10-31 1958-06-01 1960-09-18 1964-09-20 1968-09-15 1969-10-14 1970-09-20 1973-09-16 1976-10-08 1978-10-18 1979-10-12 1981-05-22 1982-10-08 1985-09-15 1986-03-14 1988-09-18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Hansson Erlander I Erlander II Erlander III Erlander IV Erlander V Erlander VI Erlander VII Erlander VIII Erlander IX Erlander X Palme I Palme II Palme III Fälldin I Ullsten Fälldin II Fälldin III Palme IV Palme V Carlsson I Carlsson II

Date in

Cabinet Cabinet number

1988-09-18

1982-09-19 1985-09-15

1979-09-16

1970-09-20 1973-09-16 1976-09-19

1958-06-01 1960-09-18 1964-09-20 1968-09-15

1952-09-21 1956-09-16

1948-09-19

Election date

Table 17.1a Swedish cabinets since 1945

S S S S, C S, C S, C S S S S S S S S C, L, M L C, L, M C, L S S S S

min min min mwc mwc mwc min min min min mwc mwc min min mwc min mwc min min min min min

50 (55.3) 50 (55.3) 48.7 (56) 61.7 (70) 59.1 (69.3) 54.1 (69.3) 45.9 (52.7) 48.1 (52.3) 49.1 (51) 48.5 (51.7) 53.7 (52.3) 53.7 (52.3) 46.6 44.6 51.6 11.2 50.1 29.2 47.6 45.6 45.6 44.7

230 (150) 230 (150) 230 (150) 230 (150) 230 (150) 231 (150) 231 (150) 231 (151) 232 (151) 233 (151) 233 (151) 233 (151) 350 350 349 349 349 349 349 349 349 349

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 6

Number of Number of Party Type Cabinet parties in composition of strength (%, seats in parliament parliament of cabinet cabinet upper (upper house in parentheses) house in parentheses) 3.14 (2.67) 3.14 (2.67) 3.06 (2.68) 3.06 (2.68) 3.09 (2.67) 3.18 (2.83) 3.18 (2.83) 3.17 (2.85) 3.11 (2.93) 3.18 (2.92) 2.81 (2.87) 2.81 (2.87) 3.32 3.35 3.45 3.45 3.48 3.48 3.13 3.39 3.39 3.67

ENP, parliament (first chamber in parentheses)

S S S S S, C C C S S S S S S S, C C C C S S S S S

Formal Median party in the support first policy parties dimension (upper house in parentheses)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Bildt Carlsson III Persson I Persson II Persson III Reinfeldt I Reinfeldt II Löfven I Löfven II

1991-10-04 1994-10-07 1996-03-22 1998-09-20 2002-09-15 2006-10-06 2010-09-19 2014-10-03 2019-01-21 1998-09-20 2002-09-15 2006-09-17 2010-09-19 2014-09-14 2018-09-09

1991-09-15 1994-09-18

M, C, KD, L S S S S M, C, KD, L M, L, C, KD S, MP S, MP

min min min min min mwc min min Min

48.7 46.1 46.1 37.5 41.3 51 49.6 39.5 33.2

349 349 349 349 349 349 349 349 349

7 7 7 7 7 7 8 8 8

4.19 3.50 3.50 4.29 4.23 4.15 4.54 4.99 5.63

C S S MP S KD SD SD SD C, L

V, MP V, MP

Notes: For a list of parties, consult the chapter appendix. Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1945–1998 period. Median parties for the subsequent period are based on data from Polk et al. (2017) and Bakker et al. (2015). The first policy dimension is economic left–right. The number of parties only includes parties that have held more than two seats in parliament when a cabinet has formed. Cabinet types: maj = Single-party majority cabinet; min = Minority cabinet (both single-party and coalition cabinets); mwc = minimal winning coalition. Table legends and variables are further defined in the Appendix, this volume.

23 24 25 26 27 28 29 30 31

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

580

ö  

until the 2010 arrival in parliament of the Sweden Democrats. Two of the newer entries to the party system, the Green Party and the Christian Democrats, initially maintained that they belonged to neither bloc. Furthermore, during 1995–1998, the Centre Party, one of the bourgeois parties, supported the Social Democratic minority government (in an explicit agreement on the budget). Later nonetheless, in 1998, the Social Democratic minority government struck a formal support agreement with the Left Party and the Green Party, the two other parties in what has later been referred to as the red-green bloc, which remained in effect throughout the government’s tenure. The same parties reached a new support agreement after the 2002 election, again acting as support parties for a Social Democratic minority government. The two-bloc character of the Swedish party system seemed to be further entrenched in 2006 when the four bourgeois parties—the Moderates, the Centre Party, the Liberals, and the Christian Democrats—formed a pre-electoral coalition ahead of the 2006 election—the Alliance for Sweden (Aylott and Bolin 2007). The remaining three Riksdag parties—the Social Democrats, the Left Party, and the Green Party— responded in kind for the 2010 election, resulting in all seven Riksdag parties belonging to one of two different pre-electoral coalitions. The recent pre-electoral coalitions reflect the increasingly fragmented parliament and party system and particularly the continued decline of the oncedominant Social Democratic Party. While remaining the largest party in parliament—barely, in the case of the 2010–2014 Riksdag—the Social Democrats controlled less than a third of the seats in the Riksdag following the 2010, 2014, and 2018 elections. While the decline of the major centre-left party is less stark than in many other Western European countries (see Chapters 7, 9, and 13 in this volume), it is nevertheless readily discernible. The major right-wing party, the Moderates, have experienced a parabolic growth and then decline following their dismal performance in the 2002 election. In 2010, they nearly became the largest party in the Riksdag with 30.1 per cent of the vote (compared to the 30.7 per cent of the Social Democrats). In the 2018 election they received a far more modest 19.8 per cent of the vote, more in line with their historical performance beginning in the 1980s. By far the most remarkable development, however, is the rise of the radical right Sweden Democrats. In the 2014 election they more than doubled their vote share compared to the 2010 election, becoming the third largest party in parliament with 12.9 per cent of the vote. The trend continued in the 2018 election, albeit to a lesser extent, with the party winning 17.5 per cent of the vote. This nevertheless meant that the party was within striking distance of becoming the second largest party in parliament. Following their entry into parliament, the other parties in the Riksdag placed the Sweden Democrats behind a cordon sanitaire due to their historical ties with Swedish white power and neo-Nazi groups, limiting their access to direct policy influence (Van Spanje and Van der Brug 2009; Loxbo 2014). Nevertheless, this did not prevent the party from indirectly influencing the policy agenda, especially as the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



581

Sweden Democrats as a hinge party, positioned between two equally strong political blocs where neither bloc has a parliamentary majority, became decisive for many votes in parliament. In terms of direct influence, they were however explicitly excluded on immigration policy—the most salient policy for the party. In 2011, with the explicit purpose of excluding the Sweden Democrats from influence, the Alliance government reached an agreement on immigration policy with the Green Party. Later during the autumn of 2015 a new agreement on immigration policy between six of the Swedish parliament’s eight parties was made to deal with the refugee crisis, excluding the Swedish Democrats (who wanted to see even more restrictive policies) as well as the Left party (who wanted less restrictive policies). Alongside the entry of new parties, the dimensions of political conflict have gradually changed as well. Sweden has for long been, and to a large extent still is dominated by an economic left–right dimension. Nonetheless, there has existed a persistent growth vs. ecology dimension—variously expressed over time— complementing it (Bergman 2000). With regard to the economic left–right dimension there has been a remarkable convergence within the bourgeois bloc in the 2000s. Both the Centre Party and the Liberals have moved considerably toward the right, particularly the Centre Party after the initiative to form the preelectoral coalition Alliance for Sweden was launched in 2004, while the Moderates moved towards the centre of politics after Fredrik Reinfeldt was elected party leader in 2003. An underlying factor was the Alliance’s formation ahead of the 2006 election, requiring the parties to agree on a common economic programme before presenting themselves as a coherent government alternative compared to the incumbent Social Democratic government. The Alliance also effectively neutralized one of the Social Democrats’ go-to strategies for remaining in office after elections: raising the ‘government question’ (i.e. which bloc or party most fit to rule) and pointing to dissent within the bourgeois bloc as evidence for the bourgeois parties being a poor choice. Simultaneously, cohesion within the Alliance also prevented the Social Democrats from peeling off the two ‘middle’ parties, the Centre Party and the Liberals, with promises of policy concessions (Aylott and Bolin 2007). This was a strategy that they had used earlier. The Centre Party was in a coalition with the Social Democrats between 1951 and 1957 and provided external support to the Social Democrats between 1995 and 1998. Similarly, the Liberals have worked with the Social Democratic government on, for example, reforming the tax system in the 1990s. Regarding the second political dimension of party conflict, we see a gradual shift towards immigration and integration, followed by a broader, but related, focus on cultural, or GAL–TAN, issues ahead of and following the 2018 election. While the two major Swedish parties, the Social Democrats and the Moderates, for decades had a shared view on immigration issues, in 2002 the Liberals tried to politicize the issue by pitching a language test in Swedish for would-be immigrants

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

582

ö  

as part of its election campaign. The strategy was successful and became a more important issue for the vote choice than in 1991, the year when Sweden saw its first populist radical right party in parliament (Holmberg and Oscarsson 2004: 173). The Liberal party also received its greatest share of the vote—13.4 per cent— in almost 20 years, nearly tripling its performance compared to the previous election. Nevertheless, despite such digressions from the status quo, the Swedish parties were primarily orientated towards the more permissive extreme on immigration. For example, in the Chapel Hill Expert Surveys (Bakker et al. 2015), which begins to include an immigration policy scale in the 2006 survey, the party with the most restrictive position in the 2006 survey—the Liberals—is not far off from the centre of the scale. Similarly, in 2010 the Moderates were the most restrictive on immigration of the ‘old’ parties in the Riksdag, just a bit off from the centre of the scale. In other words, there was a vacant position towards more restrictive immigration policies in the Swedish party system. Arguably, the Sweden Democrats—with the main policy goal of drastically lowering immigration, in particular from countries outside of Western Europe—have managed to leverage this vacancy in the policy space into seats in parliament (Loxbo 2014) while also gradually becoming the issue owner of the immigration policy issue (Berg and Oscarsson 2015 27–30; Martinsson 2017: 405–6). However, whether immigration or alternate dimensions, such as integration or GAL–TAN, do provide more accurate descriptions of the Swedish party political space is far from a settled question. In the run-up to the 2018 election, questions of law and order, such as harsher sentences for convicted criminals, have become more pronounced and have drastically increased in the public consciousness, as well as in the parties’ public communication (Aylott and Bolin 2019). Similarly, several party leaders made frequent references to ‘Swedish values’ in the public debate.⁷ Both of these factors, reflecting more traditionalist (‘Swedish values’) and authoritarian (law and order) positions, are captured by the GAL–TAN dimension. Nevertheless, immigration (often bundled with integration) has been the most salient of these issues (Andersson et al. 2017; Polk et al. 2017) and, as we will discuss further, has gradually emerged as an important dimension for understanding recent developments within the political blocs, both among voters (Oscarsson 2017: 422–3) and the party system in general. Furthermore, according to the Chapel Hill Expert Survey expert surveys party positions on immigration are highly, and increasingly so, correlated with positions on GAL–TAN.⁸ Developments before and after the 2018 election have included an increased attention to GAL–TAN issues more broadly, however. Moreover, the five parties ⁷ This was particularly evident during several party leaders’ speeches at the annual politics week in Almedalen in 2016. ⁸ For Swedish parties the correlation between immigration and GAL–TAN positions has increased for each successive survey. In the data from the 2006 survey, when positions on immigration is first introduced, the (Pearson’s) correlation coefficient is 0.643, rising to 0.873 in 2010 and 0.956 in 2017.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



583

clearly opposed to the Sweden Democrats—the Social Democrats, Green Party, Left Party, Centre Party, and Liberals—have all made it abundantly clear that they perceive the Sweden Democrats as a threat to the liberal-democratic order, rather than only or primarily opposing them on immigration policy (e.g. Riksdag minutes 2018/19: 16, and the second paragraph of the so-called 73-point programme, discussed later). Further developments ahead of the 2018 election have also set up new conditions for the Swedish parties and party system, ultimately affecting future government formation. In late 2015, following the European refugee crisis, where Sweden was one of the top recipients of asylum applications per capita (only after Hungary), the government implemented a new, temporary immigration law— which meets the bare minimum required by EU law—in an effort to lessen the number of refugees entering Sweden. The Social Democrats have also increasingly framed policy proposals—even if the proposals themselves are not always new per se—in terms of conflict between natives and non-natives, most clearly in relation to immigrant labour. The Moderates, especially after electing Ulf Kristersson as party leader in 2017, have also focused more extensively on issues of immigration, integration, and law and order (Aylott and Bolin 2019). In the most recent Chapel Hill Expert Survey available—the 2017 flash survey (Polk et al. 2017)—both the Social Democrats and the Moderates can now be positioned on the more restrictive half of the immigration policy divide, along with the Christian Democrats and— albeit not to the same extreme—Sweden Democrats, while the Liberals are closer to the centre of the scale, and the Left Party, Green Party, and Centre Party can be found within the more permissive half of the scale. The fracture within the Alliance became especially evident in April 2018, when the Centre Party stated it would vote in favour of a heavily criticized government proposal (including by the Centre Party!) concerning continued residency for certain unaccompanied refugee children. The Moderates, in turn, declared publicly that they would not participate in a government that implements immigration policy in line with the Centre Party’s stances in the policy area. Furthermore, at the vote three Liberals and three Christian Democrats also broke ranks with their parties and voted in favour of the proposal. The long-standing trend of the government (either within the government coalition or through external support arrangements) almost always including the median party in the left–right economic dimension was broken already in 2010, when the Sweden Democrats entered parliament. To date there have been seven cases where the median party on the economic dimension was not included in the government coalition:⁹ when the Centre Party left the government coalition

⁹ Arguably, there have actually been eight cases. After the 1958 election, the socialist bloc controlled 116 seats while the non-socialist bloc controlled 115 seats. However, the Speaker of the Riksdag, who did not have voting rights in parliament until an amendment of the Riksdag Act taking effect in 1961, was a Social Democrat, resulting in two median parties.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

584

ö  

in 1957, during the so-called lottery Riksdag after the 1973 election when each bloc controlled 175 seats each resulting in two median parties, during the Ullsten cabinet in 1978–1979, and after the 2010, 2014, and 2018 elections when the Sweden Democrats were the median party on that dimension. While, as should be clear, the median party is not always included in the governing coalition, the three latest cabinets are the only cases so far where the median party on the economic left–right dimension has not been included in several cabinets sequentially.

Electoral alliances and pre-electoral coalitions While electoral alliances are not allowed since 1952,¹⁰ pre-electoral coalitions, albeit not frequent, have at times had a marked impact on the politics of government formation (Table 17.1b). As already noted, ahead of the 2006 election, the four bourgeois parties presented themselves to the voters as a united government option: the Alliance for Sweden, later simply referred to as the Alliance. They had reached a detailed policy agreement and had issued a joint electoral manifesto (the Moderates, the largest of the parties, issued no manifesto of its own, running solely on the Alliance’s manifesto). The competing red-greens in the 2010 election, comprised of the Left Party, Green Party, and Social Democrats, proved to be a shorter-lived affair, only surviving a single election, with all three parties opting to Table 17.1b Electoral alliances and pre-electoral coalitions in Sweden, 1988–2018 Election date

Constituent parties

Type

Types of pre-electoral commitment

1991-09-15 2006-09-17 2010-09-19

M, L M, C, L, KD M, C, L, KD S, MP, V M, C, L, KD S, MP M, C, L, KD

PEC PEC PEC PEC PEC PEC PEC

Written contract, joint press conference Written contract, joint press conference, other* Written contract, joint press conference, other* Joint press conference, other Written contract, joint press conference, other* Joint press conference Joint press conference

2014-09-14 2018-09-09

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other * Joint electoral manifesto

¹⁰ Nonetheless, one exception was the collaboration between the Centre Party and the Christian Democrats in the 1985 election. In the end, only a single Christian Democrat was elected through this arrangement by technically running as a Centre Party’ candidate.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



585

contest the 2014 election on their own. Nevertheless, the Green Party and Social Democrats made a clear commitment that they would govern together if they managed to emerge as the biggest of the political blocs. The Social Democrats were reluctant to even mention the Left Party as a possible coalition partner and constantly dodged the question from both the media and the Left Party itself. Although the Left Party itself prepared to be able to appoint ministers for a future coalition government, including the Left Party in such an arrangement was largely out of the question for the Social Democrats, as they had concluded that including the Left Party in the previous pre-electoral coalition in 2010 was one reason for its failure (Blombäck 2015; Hinnfors 2015).

Government formation The bargaining process The Swedish government formation process has always been one of the shortest in Europe (being shorter only in Greece and Malta). The government formation remained rapid until 2018 (Table 17.2), never having taken longer than 19 days to complete before 2018, and on average only taking 6.1 days.¹¹ The bargaining process typically begins following the results of the election being known and continue until the government declaration is read by the prime minister in the Riksdag, which outlines the main policy goals of the government for the coming year. The normal process before that involves several steps—in case of coalition cabinets, bargaining over the general policy programme, distribution of portfolios, and the appointment of junior ministers and political appointees in the different ministries (Bergman 2003). The selection of who gets the first opportunity to form a government, ‘formateur’, follows largely the general pattern in most parliamentary democracies, where the size of a party and ideological centrality has been identified as important (Warwick 1996; Diermeier and Merlo 2004; Bäck and Dumont 2008). Thus, the largest party and the most centrally located party, or the median legislative party, will have a better chance of forming a viable government and thus becoming the formateur. In most cases the largest party is also the median party, which was usually the case in Sweden until the 2010s. However, the Speaker in Sweden does not foremost appoint the party leader from the largest party to formateur but rather the party that is likely to be able to form a cabinet that will not be opposed by a parliamentary majority, which are not always the same. It used to be that the

¹¹ However, the low average reflects the fact that many cabinets continued in office, without any bargaining between parties taking place. Moreover, after the amendment of the Swedish constitution in 2010 (and its adoption in 2011) this option is removed.

1988 1991 1994

1996 1998 2002 2006 2010 2014

2019

Carlsson II Bildt Carlsson III

Persson I Persson II Persson III Reinfeldt I Reinfeldt II Löfven I

Löfven II

4

0 0 0a 0 0 1

0 0 1

Number of inconclusive bargaining rounds

S M, C, L, KD S (1) L, S S S S M, C, L, KD M, C, L, KD S, MP (1) S, MP, V S, MPb (1) M, C, L, KD (2) M, KD (3) S, MPb (4) M, KD

Parties involved in the previous bargaining rounds

16 0 1 0 0 0 17 0 17 1 24 35 9 12 16

Bargaining duration of individual formation attempt (in days)

134

4 0 0 19 0 19

0 19 19

Number of days required in government formation

80

0 0 0 17 0 17

16 1

Total bargaining duration

154

132

153

0

175

115

154

23 130

Abstention

178

163 180

Pro

77

49

169

0

147 26

Contra

Result of investiture vote

Notes: Before 2011 an incumbent prime minister could remain in office after an election with no requirement to resign or submit to an investiture vote. a MP held parallel negotiations with on the one hand S and L, and on the other hand C and KD. However, due to our coding rules, this does not constitute as inconclusive bargaining round due to S instead opting to continue in office (see discussion on investiture rules). b In addition to S and MP, C and L were also party to negotiations on future support and collaboration on the budget.

Year in

Cabinet

Table 17.2 Cabinet formation in Sweden, 1988–2019

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



587

government formation was a process of ‘freestyle bargaining’ among the party leaders without much direct participation by the Speaker (Bergman 2000). Increasingly, and especially since the 2011 constitutional reform of a vote on the incumbent prime minister, the Speaker plays a greater role in appointing a formateur among the party leaders and attempting to steer the process. Between the period of 1945 and 1975 (when the new Instrument of Government came into effect) it was common that the Social Democrats, singlehandedly or when in coalition with the Centre Party (1951–1957), simply remained in office after an election (Bergman 2000). In addition, the Social Democrats have always been the largest party in parliament and the largest party of the centre-left bloc, and with the strict left–right alignment in party competition (which only changed recently as mentioned earlier), the party has almost always controlled the median legislator, both on the first and second dimensions. These circumstances have given the party a favourable position of being appointed the formateur by the Speaker. Nevertheless, there are some cases where the median position on the most important dimension of conflict was held by other parties, e.g. in the period 1956–1958, 1976–1981, and 1991–1994 when the Centre Party held this position. Later, in 1998 and again during 2006–2018 other parties (Green Party, Christian Democrats, and Sweden Democrats) held the median position. The latter period also coincides with a shift away from the party of the prime minister designate holding the median position on the first dimension, previously only occurring in 1978 and 1998. Beginning in 2010, we also find that the median party on the first dimension—the Sweden Democrats— is not included at all in the government, either as a member or as external support, which is a break with the previous periods. Similarly, with some other radical right parties (de Lange 2007), the Sweden Democrats occupy a centrist position on the economic left–right dimension, veering slightly towards the right more recently. The Sweden Democrats’ exclusion from any coalition negotiations should however, as already discussed, be considered in light of their extreme position on immigration, integration, and GAL–TAN issues more generally. In the post-2000 period, there have been five instances of inter-party bargaining over cabinet membership. The first two of these were after the 2002 election, where the Green party demanded ministerial posts to support a continued Social democratic lead cabinet, while the Left Party chose to enter the negotiations without such requests. The incumbent Social Democratic prime minister Göran Persson quickly dismissed a coalition cabinet, and the Liberal Party initiated parallel talks about a centre-right coalition government including the Green Party, with the Moderates as a support party. However, Persson opted to remain in office instead of resigning, surviving a motion of no confidence from the bourgeois parties after the Green Party abstained. For this reason, no government formation period has been coded in Table 17.2.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

588

ö  

The second case occurred after the 2006 election, when the Alliance emerged as the clear victors, controlling a majority of the seats in parliament. During the customary 19-day period between the election and the reading of the government declaration the parties further specified their policy positions to prepare for the inclusion in the government declaration and also on the distribution of portfolios. The Alliance contested the 2010 election with a joint election manifesto as well, which served as the basis for the subsequent government declaration, albeit this time the Alliance did not control a majority of seats after the Sweden Democrats had entered parliament. Nonetheless, Reinfeldt could remain in office without his tolerance by the Riksdag being tested as no party advanced a motion of no confidence. Following the 2014 election, the situation in parliament was more uncertain than expected, in large part due to the electoral support for the Green Party, by now more or less a member of the centre-left bloc, having been overestimated and the support for the Sweden Democrats, not a member of either bloc, being underestimated (Aylott and Bolin 2015: 735). Neither of the two traditional blocs had majority support on its own. Prime Minister Reinfeldt, despite some fears from the left-of-centre bloc, made good on his promise to resign if the Alliance did not emerge as the largest bloc in parliament. After the Social Democrat party leader Stefan Löfven was appointed formateur by the Speaker of the Riksdag after the election, he was quick to dismiss the Left Party as a possible coalition partner, informing the Left Party’s leader of this at their very first postelection meeting, mirroring the situation between the Social Democrats and the Liberals after the 1994 election. The Social Democrats and the Green Party initiated four parallel negotiation processes prior to the declaration of government: negotiations on the government agreement, which was later publicized (and focused almost exclusively on policy, see Table 17.4); negotiations on the budget; producing a document outlining how work should be organized and coordinated, both between the parties and with other actors, such as the parliamentary party groups; and the distribution of portfolios and the appointments of ministers (Junior minister A, interview 2017; Cabinet member B, interview 2017). In typical Swedish fashion, the negotiations were finished by the time the government declaration was read by Prime Minister Löfven in the Riksdag. The formation of the Löfven I cabinet is the first government coalition including the Social Democrats since the coalition between the party and the Centre Party (then Farmer’s League) fell apart in 1957. But the cabinet formation also reflects the development towards a more complex bargaining environment for the oncedominant Social Democratic Party. With the Green Party and the Left Party both requesting or even demanding cabinet representation at various points, keeping the parties out of the cabinet but still expecting their continued support seems less feasible, especially considering the declining electoral support for the Social Democrats.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



589

The situation after the 2018 election did not simplify matters of coalition bargaining for the parties. Despite the election outcome being more or less in line with the polls before the election the following bargaining process was remarkably static, especially as none of the parties were initially prepared to budge from their pre-election promises. The Alliance parties were set to form an Alliance government and, of crucial importance for the two liberal parties, without support from the Sweden Democrats. In addition, the Social Democrats stated that they were not going to accept a government formed by the smaller of the two traditional blocs. The Alliance, now on a one-seat deficit compared to the red-green bloc, sought to reframe the debate away from blocs towards government alternatives, thereby excluding the Left Party from the calculus, which ultimately proved to be a fruitless attempt (Aylott and Bolin 2019). The position of the Centre Party in particular has been central, refusing to give any influence to the Sweden Democrats while simultaneously being able to deny any pure right-ofcentre government its majority by defecting. Additionally, the Centre Party had two important criteria for joining or supporting a government that it had staked out during the election campaign. It would not join or support a government that necessitated negotiations with the Sweden Democrats, and it would not join or support a government headed by the Social Democrats. In practice, this meant that there was no realistic government option that was majority-tolerated by parliament. After two failed investiture attempts—one for Moderate leader Ulf Kristersson in November and caretaker prime minister Stefan Löfven on December—the Speaker set out a fixed deadline for the two remaining investiture attempts, which would be held in January the following year. The positioning of the Centre Party had again been a complicating factor. They, along with the Liberals, had voted against both suggested prime ministers. In the case of their Alliance colleague Kristersson both parties argued that his proposed cabinet—a minority cabinet consisting of the Moderates and between one and three additional Alliance parties, in practice only the Christian Democrats—would give too great of an influence to the Sweden Democrats (Riksdag minutes 2018/19: 16). An Alliance government without support from either the Left Party or the Social Democrats would always be at risk of the Sweden Democrats voting in favour of the Social Democrats’ budget proposal in any budget vote, thereby causing the government’s budget proposal to fall.¹² With Löfven, the Centre Party backed out from the negotiations in early December, arguing that the Social Democrats had

¹² If there are more than two competing budget proposals parliament sequentially eliminates them through pairwise provisional votes, beginning with the ones with the weakest support, until only two remain for the final vote. Parliamentary practice has been to abstain on any votes that do not include a party’s own proposal (either on its own or through a joint proposal with other parties) to facilitate minority government. The proposal that wins in the final vote is adopted as the state budget. After the 2018 election, the Social Democrats and the Sweden Democrats together held 162 seats, while the Alliance parties only held 143 seats.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

590

ö  

not been forthcoming enough on compromising with some of the Centre Party’s demands, particularly concerning linearization of the labour market. Negotiations resumed in late December/early January and finally resulted in a new government being sworn in on 21 January 2019 consisting of the Social Democrats and the Green Party, with external support from the Centre Party and the Liberals. The compromise underlying this successful government formation included hefty policy concessions from in particular the Social Democrats to the two liberal parties. The many policies included working towards deregulation of the housing market, reforming the Employment Protection Act in an employer-friendly direction, tax deductions on certain household services, and lowering marginal tax rates on high-income earners—that is policies that the Social Democratic Party had actively campaigned against during the election campaign. The agreement included cooperation on the budget between the four parties but also limited cooperation on other policy matters. The involved parties also committed themselves to not cooperate with any parties outside of the agreement on such matters, with the stated intent of limiting the influence of the Left Party. The Left Party initially responded that they would vote against Löfven based on the proposed socalled ‘73-point programme’ but agreed to abstain on the investiture vote after negotiations with Löfven (and a two-day delay of the investiture vote). Before 2018, all investiture votes had been successful. The closest votes prior was after the 1979 election, when the left-of-centre bloc held 174 seats and the centre-right bloc, which formed the two subsequent cabinets, held 175 seats. The first prime minister designate to be tried and fail to pass the investiture vote was Moderate Ulf Kristersson in 2018, with 154 votes in favour (Moderates, Sweden Democrats, and Christian Democrats) and 195 votes against (Social Democrats, Centre Party, Left Party, Liberals, and Green Party), followed by Stefan Löfven as the second case a few weeks later. There are two clear patterns regarding the investiture votes: the socialist parties, with a single exception in 1978, have always voted against any non-socialist prime pinisterial candidate. The non-socialist parties (not including the Sweden Democrats), on the other hand, have usually abstained on any socialist—which in practice has meant Social Democratic—prime ministerial candidate.¹³ From the parliamentary minutes (Riksdagens protokoll) it is evident that non-socialist parties usually interpret a ‘No’ vote as a statement that an alternative government formation would be possible. When no such alternative option exists, they instead abstain (Riksdag Minutes 1982/83: 5, 1994/95: 5, 1995/96: 73, 2014/15: 5). The latter trend was clearly broken when Löfven was voted out of office after the 2018 election and again when he was denied in his first government formation attempt afterwards, where all right-of-centre parties, including the Sweden Democrats,

¹³ Two exceptions are in 1990 and 1994, where not all non-socialist parties abstained in the vote.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



591

voted against Löfven despite there being no realistic alternative government preferred by a larger plurality.

The composition and size of cabinets Sweden has a long tradition of minority governments (as seen in Table 17.1a). Of the 31 cabinets that have formed after the Second World War, 23 have been minority cabinets, and the majority of these—15 to be exact—have been singleparty governments headed by the Social Democrats. The remaining 8 majority cabinets include 6 minimal-winning coalitions and 2 single-party majority cabinets. There have been no surplus majority coalitions and no non-partisan cabinets. Before 1971 the tendency for minority governments could be largely be explained by the fact that the Social Democratic Party controlled a majority of the legislators in the upper chamber.¹⁴ Although, an overall majority between both chambers would be sufficient to pass a budget, as budget resolutions were passed by a joint vote, other legislation needed separate majorities in both chambers. This effectively removed the possibility for a non-socialist government to win both the joint votes on the budget and both the separate votes for other legislation (Bergman 2000: 199–201). However, after the abolishment of the upper chamber in 1971 minority governments continued to be a frequent outcome of elections mainly because of, as previously mentioned, the negatively formulated investiture rule (Bergman 1993, 1995). But as argued by Strøm (1990: 70–3) it is also likely that the strong committee system in the Riksdag, allowing the opposition the possibility to exert some influence over policy outcomes without government participation, also contributes to the formation of minority cabinets. While minority government appears to be the continuing norm in Sweden, the 2006 government being an exception (and with a narrow majority of only four seats), there have nevertheless been marked changes in how minority governments ensure their continued survival. In 1998, the Social Democrats entered into a formal support agreement with the Left Party and the Green Party to remain in office. The cooperation continued after the 2002 election, with an even more extensive confidence and supply agreement than the previous term. Bale and Bergman (2006: 424) have referred to this form of external government support arrangements, based on comprehensive and long-term written policy contracts, as ‘contract parliamentarism’. In both the case of the support arrangement between the Centre Party and the Social Democrats between 1995 and 1997 and the later arrangements between the Social Democrats and the Left Party and Green Party, ¹⁴ Legislators in the first chamber were indirectly elected on the basis of local and regional elections, with only a portion of the legislators being elected each year, creating a time-lag between shifts in popular opinion, as expressed through the general elections, and the composition of the first chamber.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

592

ö  

staff from the support parties were placed in various government ministries to monitor the agreement. Another clear trend with regard to composition is that all cabinets formed after the 1974 Instrument of Government was adopted have followed bloc patterns with no cabinet including both left-wing and right-wing parties. Nevertheless, minority governments have frequently reached across the aisle to gain support for their policies, most notably the support arrangement the Social Democrats had with the Centre Party in 1995–1998, which mainly focused on ‘sanitizing’ public finances and on defence (Bergman 2000: 218). The Alliance government also negotiated with the Green Party on immigration policy in 2011, partly to marginalize the influence of the Sweden Democrats (or the ‘xenophobic forces’ as Prime Minister Reinfeldt put it). The most recent bout of cross-bloc collaboration, and one of the most extensive, is the budget negotiations between the Social Democrats, Green Party, Centre Party, and Liberals ahead of the formation of the Löfven II cabinet in 2019. However, the particulars of the cabinet that formed after the 2014 election can be considered a further example of the strong tendency in Swedish politics to sustain the possibility of minority government rule. The minority coalition consisting of the Social Democrats and Green Party was, based on the number of seats, the fourth smallest cabinet in post-war Sweden. Even when accounting for the external support of the Left Party—who remained negotiating partners to the government concerning the state budget for the entire term, starting in 2014—the government’s parliamentary support still does not amount to what Strøm (1990: 94–6) has referred to as a formal minority government—that is a minority government that including its support parties controls a parliamentary majority. Indeed, the government almost became the shortest-lived in post-war Swedish political history after the Sweden Democrats, against parliamentary praxis, voted in favour of the 2014 fall budget prepared by the Alliance parties, which therefore passed instead of the government’s proposed budget. While Prime Minister Löfven immediately announced that an ‘extra’ election would be held as early as possible (see the discussion on the rules governing this Swedish peculiarity in the section on cabinet duration), this outcome was avoided after an agreement was reached by the government parties and the Alliance parties in December of 2014. The core of the so-called December agreement (in Swedish rather ominously abbreviated as the Swedish word for ‘die’, DÖ) was that whichever bloc emerged as the largest in parliament would be allowed to form a government and to have its budget pass (Bäck and Hellström 2015: 272). In practice, this meant that the smaller bloc would not present a joint budget proposal, with each party from the opposition bloc instead preparing individual proposals. The agreement was a means to ensure the ability for minority governments to function, both in forming and in passing and thus governing on their own state budget, while simultaneously minimizing the influence of the Sweden Democrats in particular on budget

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



593

proceedings.¹⁵ While the agreement was supposed to last the entire 2014–2018 parliamentary term, as well as the subsequent 2018–2022 term, it formally died rather unceremoniously in October 2015, after the Christian Democrats’ members had voted against continuing to observe the agreement at the annual party conference, effectively rendering the agreement null and void. In practice, however, the agreement was upheld through the entire term, as is evident from the voting records on subsequent budgets. It was not until after the 2018 election, however, that the practical maintenance of the December agreement ceased, when the Moderates and Christian Democrats prepared a joint budget proposal for the 2018 fall budget vote. The budget proposal from the two opposition parties (and not the budget proposed by the caretaker government) passed with support from the Sweden Democrats.

The allocation of ministerial portfolios For both the two Reinfeldt cabinets and the Löfven cabinets, the junior partners in the coalition were slightly favoured in terms of the numerical distribution of portfolios. According to our interviewees, the Moderates and the Social Democrats lay claim to more important portfolios, especially prime minister and finance but also, for example, foreign affairs, interior, and justice (Table 17.3) The junior partners primarily obtained portfolios that were aligned with their parties’ profiles, such as the environment portfolio for the Green Party and the Centre Party, education for the Green Party and the Liberals, and social affairs for the Christian Democrats. The parties were given free rein on who to appoint for different portfolios, although formally the prime minister could veto any suggested minister (Junior minister A, interview 2017; Cabinet member A, interview 2017; Cabinet member B, interview 2017; Junior minister C, interview 2017; Junior minister D, interview 2017). Worth noting is that in both Reinfeldt cabinets and the Löfven cabinets, the prime minister and the minister of finance were drawn from the same party. While this is not unexpected with regard to the Löfven cabinets—during the three coalition cabinets under Prime Minister Erlander in the 1950s the Social Democrats likewise held both portfolios—it is a break for non-socialist coalition cabinets. In the Fälldin cabinets in the 1970s and 1980s, as well as the Bildt cabinet during 1991–1994, the position of minister of finance belonged to a different party than the prime minister’s, specifically the Liberals. The Liberals’ previous claim to the Finance Ministry falls in line with the expectations of their party family’s preference for this type of portfolio (e.g. Budge

¹⁵ The Left Party was also not a party to the December agreement, but was a negotiating partner with the Löfven I government on the state budget.

1991

2006 2010 2014 2019

Bildt

Reinfeldt I Reinfeldt II Löfven I Löfven II

9 (M), 4 (L), 4 (C), 3 (KD), 1 (ind.) 11 (M), 5 (C), 4 (L), 3 (KD) 13 (M), 4 (C), 4 (L), 3 (KD) 18 (S), 6 (MP) 18 (S), 5 (MP)

Number of ministers per party (in descending order)

23 24 24 23

21

Total number of ministers

13 12 11 12

14

Number of ministries

Note: There have been no watchdog junior ministers in Sweden during the presented period.

Year in

Cabinet

M M S S

M

1 Prime minister

Table 17.3 Distribution of cabinet ministerships in Swedish coalitions, 1991–2019

M M S S

L

2 Finance

M M S S

M

3 Foreign affairs

M M S S

C

4 Labour and equality affairs

M M S S

M

5 Justice

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



595

and Keman 1990). More recently, however, the Liberals have increasingly focused their efforts on education as their main policy issue. It is also common for there to be multiple ministers within the same ministry, and for recent cabinets, often with ministers from different coalition parties. In line with Bergman’s (2000: 217–18) earlier assessment, ministers without portfolio are continually assigned specific policy areas within their ministries. For example, the Ministry of Education in the Löfven I cabinet includes three ministers, the minister for education, who is the head of the ministry, as well as the minister for secondary education and youths and the minister for higher education and research. The latter two ministers are responsible for preparing these specific policy areas within the wider education portfolio. Multiparty ministries are present in all twenty-first-century coalition cabinets. The so-called ‘entry force’, which comprised of members from all the Alliance parties, advised against the practice of multiparty ministries based on the bad experiences from the Bildt cabinet in the 1990s (Junior minister B, interview 2017). Despite this, in the Reinfeldt I cabinet starting in 2016 there were three, and four in the Reinfeldt II cabinet. In Löfven I cabinet there were five multiparty ministries and three in the Löfven II cabinet.

Coalition agreements The Social Democrats–Centre Party coalitions during the 1950s in part set the format for future Swedish coalition agreements, in that they were (almost) entirely focused on policy (Table 17.4). This trend continued with both the bourgeois cabinets formed in the 1970s and 1980s, the Bildt cabinet in 1991, and the four coalition cabinets formed in the twenty-first century. It also appears as if the general trend identified by Bergman (2000: 214) of the coalition agreements becoming longer continued with the Reinfeldt cabinets, but then quickly abated with the Löfven cabinet. However, despite the continued tendency to almost exclusively focus on policy, there have been a few new developments regarding coalition agreements in Sweden for the coalition cabinets formed in the twentyfirst century, which we will return to later when we describe the agreements reached for the Reinfeldt and Löfven cabinets. The most central policy document in common for Swedish governments after 1975 is the government declaration (regeringsförklaringen). Read by the prime minister in one of the first sessions of the newly convened parliament in the autumn, it outlines the main policy goals of the government for the following year. The coalition partners negotiate extensively prior to the declaration being read, and our interviewees for both the Reinfeldt cabinets and the Löfven cabinets clearly indicate that the government declaration is a very important agreement for the coalition government.

1,100 1,100 1,900 2,900 3,500 1,400 5,200 6,421 6,262 3,772a 4,961b 3,683

Erlander III Erlander IV Erlander V Fälldin I Fälldin II Fälldin III Bildt Reinfeldt I Reinfeldt II Löfven I

0 0 0 0 0 0 0 0 0 0 0 0

General rules (in %)

0 0 2 8 4 0 4 0 0 0 0 0

Policy-specific procedural rules (in %)

Notes: a 2014 Government declaration b 2014 policy agreement between the Social Democrats and the Green Party

Löfven II

Size

Coalition

Table 17.4 Size and content of coalition agreements in Sweden, 1945–2019

0 0 0 0 0 0 0 0 0 0 0 0

Distribution of offices (in %) 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of competences (in %)

100 100 98 92 96 100 96 97.5 92.3 95.8 98 90.4

Policies (in %)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



597

There are several novelties concerning the coalition agreements for the most recent coalition cabinets in Sweden. First, the Alliance had an already agreed upon policy programme before entering office in 2006, in the form of their joint electoral manifesto. For almost two years before the election, the coalition partners had prepared and coordinated their policies. Since the inception of the Alliance, various task groups comprised of various members, such as MPs, other party members, civil servants, and others, had begun negotiations to reach a common policy platform for the parties, which was later discussed at the party leader level. The joint policy platform was then developed into the Alliance’s election manifesto, which served as the main policy agreement between the parties before the government declaration was read. To help facilitate the cooperation between the parties when in government, including the preparation of guidelines on coordination, the so-called ‘entry force’, comprised of members from each constituent party who worked closely with their respective party leaders, was assembled (Cabinet member A, interview 2017; Junior minister C, interview 2017, Junior minister D, interview 2017).¹⁶ The Alliance’s electoral manifestos, which grew longer and more detailed with each election (the 2014 manifesto weighed in at over 25,000 words, compared to the roughly 11,000-word-long 2006 manifesto), served as the basis for the subsequent government declarations. According to our interviewees, the most recent public-facing documents were also considered to supersede earlier documents if there were discrepancies between them (Cabinet member A, interview 2017; Junior minister C, interview 2017; Junior minister D, interview 2017). Similar to previous government declarations, 2006 and 2010 government declarations were almost entirely devoted to policy. They were also so far the longest government declarations, at over 6,000 words each. The Löfven government declaration also almost entirely focused on policy, but it was considerably shorter, at just under 4,000 words. The Social Democrats and Green Party however also reached a separate government agreement, as discussed earlier, which served as the basis for the government declaration and which our interviewees indicated was as, if not more, important than the government declaration, and similarly focused on policy (Junior minister A, interview 2017; Cabinet member B, interview 2017).

¹⁶ Prior to the Alliance’s election victory in 2006, the only previous case after the Second World War where an incoming government had a developed programme prior to the election was in 1991. However, the agreement reached between the Moderates and Liberals was focused on financial and economic policy and only included the aforementioned two parties out of the four that subsequently joined the cabinet. The programme therefore had to be renegotiated by all four coalition parties before it could be finalized (Bergman 2000: 209).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

598

ö  

Coalition governance The role of individual ministers in policy-making The formal rules underlying coalition governance in Sweden have by and large remained unchanged. Swedish ministers are still constitutionally prohibited from exerting direct ministerial control of state agencies, and executive decisionmaking remains a collegial affair, as a formal expression of the entire cabinet. Nevertheless, individual ministers are only loosely constrained by the constitution. The formal powers of the prime minister have likewise remained largely unchanged. While the prime minister appoints and dismisses ministers at their discretion, they cannot unilaterally decide on the jurisdictions of the various ministries, which is instead determined by the cabinet collectively through decree. Thus, as also indicated by Table 17.5, the decisions of the government can be seen as a common expression of the views of the members of the government under shared responsibility (Bergman 2000, 2003). In the Swedish constitution a cabinet meeting should compromise of at least five ministers along with the prime minister. Moreover, normally at a government meeting the head of department is the rapporteur for the matters belonging to his or her department (Instrument of Government Chapter 7, Sections 3–5). Before a reform in the end of 1990s the government offices consisted of 15 independent agencies, but the reform merged the ministries to form one joint agency headed by the prime minister. Although, this reform potentially could strengthen the prime minister’s powers, in practice it had little effect and business continued as usual (Ds 2003: 44; Erlandsson 2007). It is also evident in various memoranda on the rules of procedure for the government offices,¹⁷ both before and after the reform, that ministerial autonomy has always been limited as all government matters must be decided by the government as a joint collective. In coalition governments, as indicated be the interviewees, the room for ministerial autonomy was even more limited as the general rule for decision-making was always consensus among all coalition partners. Nonetheless, holding a ministerial portfolio still gives a party a great deal of influence in that the party that controlled a ministry had greater opportunity to set the agenda on the specific issues that belonged to that particular ministry. Although any party could raise an issue to the agenda on issues that concerned a ministry they did not control, the minister could stop the proposal if he or she did not like it, giving the minister a kind of veto power.

¹⁷ See, for instance, Förordning (1996: 1515) med instruktion för Regeringskansliet, RKF 2008: 17, SB PM 2012: 1.

1951 1952 1956 1976 1979 1981 1991 2006 2010 2014

IE Pre Post Post Post IE Post Pre, Post Pre, Post Post

Y Y Y Y Y Y Y Y Y Y

N N N N N N N N N N

IC IC IC IC, CaC IC, CaC IC, CaC IC, CaC IC, CaC IC, CaC IC, CaC

All used

IC IC IC CaC CaC CaC CaC CaC CaC CaC

IC IC IC IC IC IC IC IC IC IC

Most For most common serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Y Y Y Y Y Y Y Y Y Y N N N

Personal Issues union excluded from agenda

Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most Most/Most All/All All/All All/All

Y Y Y Y Y Y Y Y Y Y

No No No No No No No Comp. Comp. Comp.

Y Y Y Y Y Y Y Y Y Y

Y Y Y Y Y Y Y Y Y Y

Coalition Freedom of Policy Junior Nondiscipline appointment agreement ministers cabinet in positions legislation/ other parl. behaviour

Notes: Coalition agreement: PRE = pre-election; Pre, Post = pre- and post-election; IE = inter-election; POST = post-election Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee Coalition discipline: All = Coalition discipline is always expected; Most = Coalition discipline is expected on all matters except those explicitly exempted Policy agreement: Comp. = Comprehensive policy agreement

Erlander III Erlander IV Erlander V Fälldin I Fälldin II Fälldin III Bildt Reinfeldt I Reinfeldt II Löfven I

Coalition

Table 17.5 Coalition governance mechanisms in Swedish coalition governments, 1945–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

600

ö  

Coalition governance in the executive arena Coordination within the cabinet has undergone several transformations during the various periods of Swedish coalition governance under study here, as is noted in Table 17.5. In the coalition governments in the 1950s, disputes within the cabinet were handled by the party leaders and possibly other important ministers, comprising an informal inner cabinet. While the practice of using an inner cabinet has persisted, it has primarily dealt with more serious issues for later coalition cabinets. Conflicts have instead been intended to be resolved at levels lower than the party leaders, with two primary models attempted, comprising what we refer to as cabinet committees (CaC). The first was implemented during the Fälldin cabinets in the 1970s and 1980s. A staff of political appointees were provided to each party leader and were tasked with the coordination of the cabinet, as well as solving lower-level disputes before escalating them to the party leaders (i.e. the inner cabinet) (Bergman 2000: 212). A second system was implemented by the Bildt cabinet 1991–1994. Instead of each party leader being assigned a staff of political appointees, a coordination office was instated at the prime minister’s office, and each coalition party assigned a junior minister to the office, serving as an intermediary between the ministry level and the inner cabinet level both on coordination and conflict resolution (Bergman 2000: 212–13). The practice used in the Bildt cabinet was largely retained in the Reinfeldt cabinets 2006–2014. A central coordination office, located in the prime minister’s office, was tasked with coordination and staffed with junior ministers from the four coalition parties. In addition, a ‘coordination document’ was adopted directly after the election. The document set up rules and procedures for how initiatives by the ministers should be handled and receive support from all coalition partners, what types of decisions that should be up to party leader negotiations, and what types of decisions that could be resolved without the party leaders’ active involvement (Cabinet member A, interview 2017; Junior minister D, interview 2017). When the Löfven cabinet was formed, the procedures and the practice of having a coordination office to handle low-level issues in the cabinet were also carried over, in part due to the short time period in which the government needed to get up and running (Junior minister A, interview 2017; Cabinet member B, interview 2017). In addition to the coordination office, the Löfven I cabinet also included a minister for coordination after a cabinet reshuffle in 2016, placed at the prime minister’s office, who is tasked with assisting the prime minister in coordinating between the different ministries. While having a minister for coordination is not a new phenomenon (as the first one was already appointed in 1940s), the Löfven I cabinet is the first occurrence where a minister for coordination has been included in a coalition cabinet. In practice, the coordination procedure during the Reinfeldt cabinets, as well as the previous Bildt cabinet, was the following: the initial policy proposal was first

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



601

worked out between the involved ministries, and often including the Ministry of Finance, followed by a preparation in the coordination office, to finally be followed by the final negotiation by the involved ministries. The junior ministers for each party, either at the involved ministries or at the coordination office, or both, then contact the senior committee members in the Riksdag belonging to their parties (in the committee where the specific issue should be handled). If necessary, the matter can then be completed or moved on to the inner cabinet if there were any unresolved issues that needed additional bargaining (Junior minister B, interview 2017; Junior minister C, interview 2017; Junior minister D, interview 2017). The nomination and recruitment of junior ministers (statssekreterare), as with ministers, have been left to the discretion of the individual parties and ministers for the entire period, although, as is the case with ministers, there has been a principle of candidates potentially being excluded based on non-tolerance from other coalition parties or the prime minister. It has been rare, although not completely without precedence, for suggested junior ministers to be opposed or disregarded. In the coalition cabinets in the 1950s, a junior minister from the Social Democrats was placed in the Ministry for Education, and one suggested junior minister in the first Fälldin cabinet was vetoed by the prime minister (Bergman 2000: 211–12). As with previous cabinets, ministers, with backing from their parties, have largely been given free rein to select their own junior ministers in the Reinfeldt and Löfven cabinets. In the latter, potential junior ministers were also screened by the coordination office (Junior minister A, interview 2017; Cabinet member A, interview 2017; Junior minister B, interview 2017; Cabinet member B, interview 2017; Junior minister C, interview 2017; Junior minister D, interview 2017). Although, junior watchdog ministers—that is junior ministers who are attached to a minister from a different party—do not exist in modern coalition cabinets in Sweden,¹⁸ coalition partners are still able to ‘keep tabs’ on each other’s ministers. This is especially true in the multiparty ministries where ministers (with full voting rights) can act as watchdogs. Regardless, there is not much room for ‘ministerial drift’ as the coordination procedure between coalition partners (since the Bildt cabinet in the 1990s) gives the parties transparency and influence over policy proposals originating from the individual ministries. To summarize, when it comes to coalition governance, Swedish coalition governments from the 1990s onwards undoubtedly fit well into one of the ideal types of coalition governance, namely the coalition compromise model (Bergman

¹⁸ There has only been one watchdog junior minister in Sweden, namely in the coalition government between the Social Democratic Party and the Centre Party (then named the Farmers’ League) between 1951 and 1957, where the Social Democratic Party had a watchdog junior minister in the ministry of education.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

602

ö  

et al. 2019). The coalition partners voluntarily bind themselves to specific policy commitments within the context of coalition agreements, which are primarily hashed out in concert with coalition negotiations, most clearly in the intense work on the government declaration presented by the prime minister in parliament prior to being sworn in. The development of a coordination office during the Bildt cabinet, which has persisted during the Reinfeldt and Löfven cabinets, serves as a first means of combating policy drift between ministries. There are also institutional safeguards to prevent ministers from directly controlling government agencies or unilaterally enacting policy. All government policy is formally collegial in nature.

Governance mechanisms in the parliamentary arena Tendencies towards increasingly formalized manners of coordination is not strictly limited to the executive arena but includes the parliamentary arena as well. As described by Bergman (2000: 213), two novel features with regard to executive–parliamentary relations between and within the coalition parties introduced with the Bildt cabinet were the inclusion of the parties’ parliamentary group leaders in the informal weekly meetings where coalition policy was discussed and the parliamentary party groups meeting with each other, and sometimes also with cabinet ministers, to discuss policy and exchange information. As with the coordination and conflict resolution practices within the cabinet, these practices have likewise continued in both the Reinfeldt and the Löfven cabinets. Compared to previous cabinets, the expectations of coalition discipline among parliamentarians belonging to the coalition parties has, if anything, increased. According to our interviewees both in the Reinfeldt cabinets and the Löfven I cabinet, strict discipline was expected of all parliamentarians (Junior minister A, interview 2017; Cabinet member A, interview 2017; Junior minister B, interview 2017; Cabinet member B, interview 2017; Junior minister C, interview 2017; Junior minister D, interview 2017). While there were certain issues, such as samesex marriage, where one of the coalition partners in the Alliance government had a different position compared to the others, the general course of action was that there were no issues excluded from the agenda.¹⁹ MPs defecting from their party and government line—or threatening to do so publicly—are nonetheless not unheard of. The most clear-cut example is when several MPs from the Alliance government voted against a government bill to extend the surveillance capabilities of the National Defence Radio Establishment in 2008.²⁰

¹⁹ This particular issue was resolved through voting on a legislative proposal made by parliament rather than through a government-initiated process. ²⁰ The adoption of the bill was postponed for over a year, and several measures to safeguard personal integrity were included as concessions to critics.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



603

Cabinet duration and termination The duration of cabinets In Sweden it is relatively rare for a government to resign before the next regular election, and as of the turn of the millennium all cabinets formed have served their full term. The last time a Swedish government resigned and was replaced prematurely was the second Fälldin cabinet in 1981. The Carlsson government also resigned in 1990, but as there were no realistic alternatives to form a government, Carlsson was reinvested as PM and reformed his cabinet (so in practice it did not lead to any change in government). From 1945 until 2018, only 5 out of 30 (or about 17 per cent) of the cabinets have ended prematurely (excluding technical termination, for instance, deaths of prime ministers or constitutional reasons). This makes Sweden the country with the fewest number of discretionary terminations in Western Europe. Moreover, apart from Norway where the constitution does not permit early dissolutions, Sweden stands out by having only one cabinet that ended through an early election (the Erlander VI cabinet in 1958) since the introduction of universal suffrage. There are two main reasons for the few discretionary cabinet terminations in Sweden. Firstly, as previously mentioned, the institutional rules and inter-party competition traditionally have helped to maintain minority cabinets. Secondly, another important reason concerns how dissolution of the Riksdag and early elections is regulated in the Swedish constitution, which compared to other West European countries make the incentives for incumbents to call for early elections significantly lower. In some countries with more or less flexible election dates (e.g. Denmark, Ireland, Austria, Belgium, and the United Kingdom up to the 2010 parliamentary elections), incumbent parties can strategically engineer their own departure to maximize electoral support or to circumvent electoral loss (Strøm and Swindle 2002; Kayser 2005, 2006; Schleiter and Tavits 2016). In addition, in countries where strategic exploitation of dissolution powers is limited, an early election starts a new term of office. Thus, theoretically, the incentives and tendency for cabinets to announce early dissolutions are assumed to increase over time, that is the closer to the regular election date or maximum term of office. The reason for this is that governments usually manage to get through much of their policies in the beginning and middle of their term of office, and the benefits of remaining in office therefore reduces over time, making opportunistic or strategic early elections a viable option in some cases (e.g. Lupia and Strøm 1995). However, the Swedish constitution states that early elections are only first allowed to be called when three months have passed after preceding regular election, they cannot be called by caretaker governments and, most importantly, an early election does not commence a new term of office (Instrument of Government,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

604

ö  

Chapter 3, Section 11). In other words, a government formed after an early election only serves the remaining term of office, and the incentives for the prime minister to announcing an early election decreases over time. Prime ministers threatening to call early elections is nonetheless rather common. For instance, Prime Minister Göran Persson threatened to call an early election if the Green Party would not support a Social Democratic minority government to form after the parliamentary election of 2002. More recently, and as we have discussed in some detail earlier, Prime Minister Löfven stated that he would call an early election after the government’s budget proposal failed to pass in late 2014, which was averted through the December agreement. Löfven has also made less explicit threats of holding an early election as recently as 2017, stating that he would not rule out calling an early election if the Alliance parties were to vote against the 2017 spring budget. In the 2018–2019 government formation, the Speaker of the Riksdag also invoked the threat of an early election to entice the parties to be more forthcoming to each other in the negotiation process by setting a final date for the fourth and constitutionally last possible vote on a candidate for prime minister.

The termination of cabinets Cabinet terminations have, as Table 17.6 shows, since end of the 1980s been a rather uneventful affair in Sweden. Over two-thirds of cabinets—twenty-one of thirty—have terminated because of a regular election. The strong position of the Social Democrats for much of the post-Second World War period goes a long way to explain the predominance of this kind of termination. Of the twenty-one aforementioned cabinets, thirteen are single-party Social Democratic cabinets, while the remaining eight are comprised of seven coalition cabinets and one single-party cabinet. Of the nine cabinets that have not terminated as a result of a regular election, four are the result of the incumbent prime minister either resigning (two cases) or dying while in office (two cases), three cases of inter-party conflict within the governing coalition, one early election, and one case of coalition enlargement. Within the context of the historically strong position of the Social Democrats, the codification of negative parliamentarism with the 1974 Instrument of Government, and the comparatively weak incentives for calling an early election, the distribution of termination mechanisms among Swedish cabinets is not unsurprising. If we limit our enquiry to cabinets formed after the last election in the 1980s— which would coincide with the 1988 election and a time period of 30 years—the image is even more telling, as is clear from Table 17.6. Except for Carlsson’s resignation as prime minister in 1995, all other cabinets have terminated as the result of a regular election. As discussed, the disincentives to holding early

1988-09-18 1991-10-04 1994-10-07

1996-03-22 1998-09-20 2002-09-15 2006-10-06 2010-09-19 2014-10-03

Carlsson II Bildt Carlsson III

Persson I Persson II Persson III Reinfeldt I Reinfeldt II Löfven I

1998-09-20 2002-09-15 2006-09-17 2010-09-19 2014-09-14 2018-09-09

1991-09-15 1994-09-18 1996-03-18

Date out

100 100 100 100 100 100

100 100 36.6

Relative duration (%)

1 1 1 1 1 1

1 1 9

Mechanisms of cabinet termination

14

Terminal events

Parties

Policy area(s)

Prime Minister Carlsson voluntarily resigned as leader of the Social Democrats and as prime minister.

Comments

Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister

Date in

Cabinet

Table 17.6 Cabinet termination in Sweden, 1988–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

606

ö  

elections, along with the internal cohesion of the coalition partners in the Reinfeldt and Löfven cabinets, have certainly contributed to the stability of the respective cabinets.

Conclusions The possibility of forming majority governments often seems like a dim prospect in Swedish politics. During the first decade of the new millennium, with an increasingly formalized or even institutionalized forms of support arrangements within the red-green bloc, and the convergence of the bourgeois parties in the Alliance, Sweden almost had a bipolar, two-bloc party system. But this quickly changed when the Sweden Democrats entered parliament in 2010, not the least due to their rapid increase of seats in parliament and their blackmail potential as a hinge party. The development of the party system back into a strict bipolar, twobloc party system seems less and less likely, in particular as the bargaining process after the 2018 election resulted in the first government based on cross-bloc budget negotiations in decades. Certainly, with a changing party system, the coalition lifecycle is also likely to change. The initial attempts to isolate the Sweden Democrats through removing the issue of immigration policy from the political agenda in 2011 appear to have been mixed. While the Sweden Democrats have been excluded from direct policy influence if we limit ourselves to the national level, the same cannot be said of their indirect influence, with immigration policy and sociocultural issues—the bread and butter of radical right parties—becoming increasingly salient issues in the political debate as well as parliamentary politics. The rush towards more restrictive immigration policies, primarily undertaken by the Social Democrats and the Moderates, complicates matters further within the previous two-bloc system. In this regard, it is likely that two main questions determine the future of coalition governance in Sweden. Firstly, whether the current arrangement between the red-green bloc and parts of the bourgeois bloc will continue, whereby a two-and-a-half bloc system is maintained and the Sweden Democrats are still excluded from direct policy influence, or if the party system develops in a new direction, with more blocs, for example the current red-green bloc (or a red bloc), a liberal bloc comprised of the Centre Party and the Liberals (or a liberal-green bloc also including the Green Party), and a conservative bloc comprised of the Moderates, Christian Democrats, and Sweden Democrats. Heated exchanges between the former Alliance parties after the 2018–2019 government formation points towards the latter being a more likely scenario than previously. Secondly, whether the cordon sanitaire by the other parties in the Riksdag will survive or if the Sweden Democrats will be invited to support a right-wing formal minority government. Both the Moderates and the Christian Democrats appear to have

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



607

abandoned the stalwart opposition to the Sweden Democrats that was maintained during the Reinfeldt II cabinet, which is especially evident if we include the municipal level of politics. An increasingly heated debate between the bourgeois parties on relying on the Sweden Democrats’ support after the 2018 election has made the ideological differences within the bourgeois bloc clearer than during the Alliance’s years of policy convergence. Regardless, minority cabinets will be the rule, not the exception, even in the years to come.

References Andersson, Ulrika, Jonas Ohlsson, Henrik Oscarsson, and Marie Oskarson (2017). ‘Larmar och gör sig till’. In Ulrika Andersson, Jonas Ohlsson, Henrik Oscarsson, and Marie Oskarson (eds), Larmar och gör sig till. Göteborg: SOM-institutet vid Göteborgs universitet, 11–35. Aylott, Nicholas, and Niklas Bolin (2007). ‘Towards a Two-Party System? The Swedish Parliamentary Election of September 2006’. West European Politics, 30(3): 621–33. Aylott, Nicholas, and Niklas Bolin (2015). ‘Polarising Pluralism: The Swedish Parliamentary Election of September 2014’. West European Politics, 38(3): 730–40. Aylott, Nicholas, and Niklas Bolin (2019). ‘A Party System in Flux: The Swedish Parliamentary Election of September 2018’. West European Politics, 42(7): 1504–15. https://doi.org/10.1080/01402382.2019.1583885. Bäck, Hanna, and Johan Hellström (2015). ‘Efter valet 2014: Regeringsbildningen och det inställda extra valet’. Statsvetenskaplig tidskrift, 117(2): 261–78. Bäck, Hanna, and Patrick Dumont (2008). ‘Making the First Move. A Two-Stage Analysis of the Role of Formateurs in Parliamentary Government Formation’. Public Choice, 135(3–4): 353–73. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010’. Party Politics, 21(1): 143–52. Bale, Tim, and Torbjörn Bergman (2006). ‘Captives No Longer, but Servants Still? Contract Parliamentarism and the New Minority Governance in Sweden and New Zealand’. Government and Opposition, 41(3): 422–49. Berg, Lind, and Henrik Oscarsson (2015). Supervalåret 2014. Stockholm: Statistics Sweden. Berglund, Sten, and Ulf Lindström (1978). The Scandinavian Party System(s). Lund: Studentlitteratur. Bergman, Torbjörn (1993). ‘Formation Rules and Minority Governments’. European Journal of Political Research, 23(1): 55–6.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

608

ö  

Bergman, Torbjörn (1995). ‘Constitutional Rules and Party Goals in Coalition Formation: An Analysis of Winning Minority Governments in Sweden’. PhD thesis, Department of Political Science, Umeå University. Bergman, Torbjörn (2000). ‘Sweden: When Minority Cabinets Are the Rule and Majority Coalitions the Exception’. In Wolfgang C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 192–230. Bergman, Torbjörn (2003). ‘Sweden: From Separation of Power to Parliamentary Supremacy—and Back Again?’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press, 594–619. Bergman, Torbjörn, Gabriella Ilonszki, and Wolfgang C. Müller (eds) (2019). Coalition Governance in Central Eastern Europe. Oxford: Oxford University Press. Blombäck, Sofie (2015). ‘Vänsterpartiet: Pådrivare eller pragmatiskt stödparti?’. Statsvetenskaplig tidskrift, 117(2): 219–30. Budge, Ian, and Hans Keman (1990). Parties and Democracy: Coalition Formation and Government Functioning in Twenty States. Oxford: Oxford University Press. De Lange, Sarah L. (2007). ‘A New Winning Formula? The Programmatic Appeal of the Radical Right’. Party Politics, 13(4): 411–35. Diermeier, Daniel, and Antonio Merlo (2004). ‘An Empirical Investigation of Coalitional Bargaining Procedures’. Journal of Public Economics, 88(3–4): 783–97. Ds (2003). Regeringskansliet en myndighet. En analys av effekter och konsekvenser, Stockholm: Regeringskansliet, Förvaltningsavdelningen. Erlandsson, Magnus (2007). ‘Striderna i Rosenbad: Om trettio års försök att förändra regeringskansliet’. Doctoral dissertation, Statsvetenskapliga institutionen. Green-Pedersen, Christoffer, and Lisbeth Hoffman Thomsen (2005). ‘Bloc Politics vs. Broad Cooperation? The Functioning of Danish Minority Parliamentarism’. The Journal of Legislative Studies, 11(2): 153–69. Hinnfors, Jonas (2015). ‘Socialdemokraterna: Från klar vaghet till vag klarhet’. Statsvetenskaplig tidskrift, 117(2): 137–52. Holmberg, Sören, and Henrik Oscarsson (2004). Svenskt väljarbeteende. Redogörelse för 2002 års valundersökning. Stockholm and Örebro: SCB. Kayser, Mark Andreas (2005). ‘Who Surfs, Who Manipulates? The Determinants of Opportunistic Election Timing and Electorally Motivated Economic Intervention’. American Political Science Review, 99(1): 17–27. Kayser, Mark Andreas (2006). ‘Trade and the Timing of Elections’. British Journal of Political Science, 36(03): 437–57. Lijphart, Arend (2012). Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. Yale: Yale University Press. Loxbo, Karl (2014). ‘Voters’ Perceptions of Policy Convergence and the Short-Term Opportunities of Anti-immigrant Parties: Examples from Sweden’. Scandinavian Political Studies, 37(3): 239–62.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



609

Lupio, Arthur, and Kaare Strøm (1995). ‘Coalition Termination and the Strategic Timing of Parliamentary Elections’. American Political Science Review, 89(3): 648–65. Martinsson, Johan (2017). ‘Vilka partier äger de viktigaste sakfrågorna?’. In Ulrika Andersson, Jonas Ohlsson, Henrik Oscarsson, and Marie Oskarson (eds), Larmar och gör sig till. Göteborg: SOM-institutet vid Göteborgs universitet, 401–409. Müller, Wolfgang C., and Kaare Strøm (eds) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Oscarsson, Henrik (2017). Det svenska partisystemet i förändring, in Ulrika Andersson, Jonas Ohlsson, Henrik Oscarsson, and Marie Oskarson (eds), Larmar och gör sig till. Göteborg: SOM-institutet vid Göteborgs universitet, 411–27. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of Antielitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Schleiter, Petra, and Margit Tavits (2016). ‘The Electoral Benefits of Opportunistic Election Timing’. Journal of Politics, 78(3): 836–50. Strøm, Kaare (1990). Minority Government and Majority Rule. Cambridge: Cambridge University Press. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Strøm, Kaare, and Stephen M. Swindle (2002). ‘Strategic Parliamentary Dissolution’. American Political Science Review, 96(3): 575–91. Van Spanje, Joost, and Wouter van der Brug (2009). ‘Being Intolerant of the Intolerant. The Exclusion of Western European Anti-immigration Parties and Its Consequences for Party Choice’. Acta Politica, 44(4): 353–84. Warwick, Paul V. (1996). ‘Coalition Government Membership in West European Parliamentary Democracies’. British Journal of Political Science, 26(04): 471–99.

Appendix. List of political parties Abbreviation V

MP S L C

Name Left Party (Vänsterpartiet), 1990– Left Party Communists (Vänstperartiet kommunisterna), 1967–1990 Communist Party of Sweden (Sveriges kommunistiska parti), 1921–1967 Green Party (Miljöpartiet de gröna) Social Democratic Party (Socialdemokratiska arbetarepartiet) Liberals (Liberalerna), 2015– People’s Party (Folkpartiet liberalerna), 1934–2015 Centre Party (Centerpartiet), 1959– Agrarian League (Bondeförbundet), 1913–1959

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

610 M

KD NyD SD

ö   Moderate Unity Party (Moderata samlingspartiet), 1969– Right Party (Högerpartiet), 1952–1969 Right Party (Högerns riksorganisation), 1938–1952 Christian Democrats (Kristdemokraterna), 1995– Christian Democrats (Kristdemokratiska samhällspartiet), 1987–1995 New Democracy (Ny demokrati) Sweden Democrats (Sverigedemokraterna)

Note: Party names are given in English, followed by the party name in Swedish in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 18 The United Kingdom: When a Coalition Meets the Westminster Model, Who Wins? Nick Barlow and Tim Bale

Unique amongst the countries in this book, the United Kingdom has only had a single coalition government in the post-Second World War period, and only three times in since 1945 have general elections failed to deliver a parliamentary majority to either the centre-right Conservative Party or the centre-left Labour Party. In this chapter, we look at how a system that developed to support a single party of government adapted to deal with two parties in power, in so doing asking how much the system changed to support the advent of a coalition government in 2010 and how much the parties adapted themselves in order to retain familiar governmental institutions.¹ Not surprisingly, given it represented such an aberration from the United Kingdom’s traditional pattern of single-party government, many aspects of the 2010 coalition between the Conservatives and the more centrist Liberal Democrats have been covered in great detail by academics (a non-exhaustive list would include Fox 2010; Bale 2011, 2012; Debus 2011; Lees 2011; Quinn et al. 2011; Evans 2012; Lakin 2013; Beech 2015; Seldon and Finn 2015; Yong and Bale 2016), journalists (Gerard 2011; d’Ancona 2013; Crace 2014), and politicians (Laws 2010, 2016; Wilson 2010; Adonis 2013; Clegg 2016). Indeed, it is striking just how many of the key players in it have been willing to go on the record about its inner workings—partly because one effect of the coalition and its political aftermath (most notably, the Brexit referendum) has been to eject many of the leading figures in its negotiation and implementation out of active politics, while even those still involved have not remained tight-lipped about it. It is this literature that provides us with a good deal of the background for this chapter.

¹ Some results in the chapter are based on an interview conducted with senior Liberal Democrat adviser by Tim Bale, 16 October 2018. Nick Barlow and Tim Bale, The United Kingdom: When a Coalition Meets the Westminster Model, Who Wins? In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

612

  

The institutional setting The United Kingdom does not possess a single written constitution but it does still possess a monarchy, in whose name power is formally exercised. Power in the United Kingdom remains highly centralized and administered in a top-down manner through the Crown-in-Parliament: the government of the United Kingdom is appointed by and acts on behalf of the monarch and the Crown rather than the people. So, on the conclusion of an election, the new prime minister is not appointed to their role by a vote in parliament or the voice of the people, but by being summoned to Buckingham Palace where the monarch requests that they form a government on his or her behalf.² While theoretically merely primus inter pares, the prime minister then has very significant powers of appointment to cabinet and junior ministerial positions without having to seek parliamentary approval for any of their choices. The one hurdle a prime minister is expected to clear—but not required to immediately prove that they can—is to demonstrate that they command a majority in the House of Commons, an achievement normally recognized by the government passing the first Queen’s Speech (laying out the administration’s legislative programme) of a parliament. This is a system that has co-evolved with a party system based around two large catch-all parties that are expected to alternate in government with one gaining a majority as the other loses it. Initially, the two principal parties were the Conservatives and Liberals, but following the fall of the last purely Liberal government during the First World War, the Liberals were rapidly replaced as the main alternative to the Conservatives by the Labour Party. While other parties (including the Liberals) existed and won seats in parliament, Britain was principally a two-party system and, with only one exception (February 1974), every general election from 1945 to 2005 gave one of these two parties a majority in parliament. Both parties were organized around a principle of strong personalized leadership, and it would be the leader of the victorious party who was summoned to the palace to form a government post-election. This principle of strong leadership extended to those occasions where a prime minister was replaced during their term in office: their chosen successor would, once they had been selected by their party, be appointed as prime minister without any need for a parliamentary vote and on the assumption that they possessed their predecessor’s mandate and ability to guarantee a majority in the Commons. One final key point to note here is that the post-election transfer of power in the United Kingdom is very swift. General elections take place on Thursdays with votes being counted overnight, so by the time Friday starts it is usually clear who ² This important distinction—that it is the monarch requesting that a government be formed, not the new prime minister requesting to form a government—is notably dramatized in the 2006 film The Queen.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

613

has won and lost the election, with the victor visiting Buckingham Palace and being back at Downing Street as a newly appointed prime minister by 9 a.m. (while their predecessor is leaving by the back door having visited the monarch to resign their position). There is little to no tradition of a lengthy transfer of power or of caretaker administrations,³ and the assumption is that a new prime minister and government should be able to commence work immediately after the election.

The party system The United Kingdom is predominantly a two-party system though one with occasionally significant, though not large, third parties. Two-partism is strengthened and institutionalized by Britain’s use of the single-member plurality (‘first past the post’) electoral system that means that, even while the effective number of parties in terms of votes may be high, the effective number in parliament will be much lower. In recent years, the effective number of electoral parties in Britain has been almost 4 (Wager 2017: 123–4), but the effective number of parliamentary parties has rarely exceeded 2.5 (see Table 18.1a). By strengthening the electoral prospects of large parties with efficiently distributed (geographically concentrated rather than thinly spread) votes this system helped to ensure two-partism at the parliamentary level. Traditionally, the principal cleavage in British politics has been a left–right, owner–worker affair, as captured by Peter Pulzer’s (1967: 98) famous description in the late 1960s: ‘Class is the basis of British party politics; all else is embellishment and detail.’ While other issues have been of importance at particular elections, such as questions of the relationship with Europe, GAL–TAN-related issues, political scandal, and foreign policy, none have been so strong an influence on a nationwide basis that they have had (at least until recently) the same kind of shaping effect on party competition in the long term. Table 18.1a illustrates that competition has generally focused on that between the Conservatives and Labour, with every single prime minister since 1922 coming from one or the other. The Conservatives have spent the longest time in office, with Labour in opposition, but alternation between the two parties has been fairly common (see Quinn 2013 for one interpretation of this). Outside of the two main parties, the Liberals and their successor parties (the Social Democratic Party–Liberal Alliance in 1983–1987 and the Liberal Democrats from 1992) have generally played the role of a centrist third party in a two-party system, finding that, even when they achieved breakthroughs in ³ Churchill led a caretaker administration in 1945 between the end of the wartime coalition and the general election, but his Conservatives still held a parliamentary majority alone during that period.

Cabinet

Attlee I Attlee II Churchill II Eden I Eden II Macmillan I Macmillan II Douglas-Home Wilson I Wilson II Heath Wilson III Wilson IV Callaghan Thatcher I Thatcher II Thatcher III Major I Major II Blair I Blair II Blair III Brown

Cabinet number

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

1945-07-26 1950-02-24 1951-10-26 1955-04-06 1955-05-27 1957-01-10 1959-10-09 1963-10-19 1964-10-16 1966-04-01 1970-06-19 1974-03-04 1974-10-11 1976-04-05 1979-05-04 1983-06-10 1987-06-11 1990-11-28 1992-04-10 1997-05-02 2001-06-07 2005-05-05 2007-06-27

Date in

Table 18.1a British cabinets since 1945

1992-04-09 1997-05-01 2001-06-07 2005-05-05

1979-05-03 1983-06-09 1987-06-11

1964-10-15 1966-03-31 1970-06-18 1974-02-28 1974-10-10

1959-10-08

1955-05-26

1945-07-05 1950-02-23 1951-10-25

Election date Lab Lab Con Con Con Con Con Con Lab Lab Con Lab Lab Lab Con Con Con Con Con Lab Lab Lab Lab

Party composition of cabinet maj maj maj maj maj maj maj maj maj maj maj min maj maj maj maj maj maj maj maj maj maj maj

61.4 50.4 51.4 51.5 54.6 53.3 57.9 57.3 50.3 57.6 52.4 47.4 50.2 50.1 53.4 61.1 57.8 57.4 51.6 63.4 62.5 55 54.8

7 4 7 4 3 3 3 3 3 4 6 7 7 7 7 8 8 8 7 7 7 7 7

Number of Cabinet Type strength in parties in of parliament cabinet seats (%) 2.05 2.08 2.05 2.06 2.03 2.09 1.99 2.01 2.06 2.03 2.07 2.25 2.25 2.26 2.15 2.09 2.17 2.2 2.27 2.13 2.17 2.46 2.48

Lab Lab Con Con Con Con Con Con Lab Lab Con Lab Lab Lab Con Con Con Con Con Lab Lab Lab Lab

ENP, Median party parliament in first policy dimension

Formal support parties

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Cameron I Cameron II May I May II

2010-05-11 2015-05-07 2016-07-13 2017-06-08 2017-06-08

2010-05-06 2015-05-07

Con, Lib Con Con Con

mwc maj maj min

55.8 50.8 50.8 48.8

7 8 8 7

2.58 2.55 2.55 2.48

Lib Con Con Con DUP

Cabinet types: maj = Single-party majority cabinet; min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition.

The number of parties only includes parties that have won more than two seats in parliament during the period of observation.

The total number of seats in parliament is 640 between 1945 and 1950, 625 between 1950 and 1955, 630 between 1955 and 1974, 635 between 1974 and 1983, 650 between 1983 and 1992 and from 2010, 651 between 1992 and 1997, 659 between 1997 and 2005, and 646 between 2005 and 2010.

The first policy dimension is economic left–right, except for the May I cabinet when it is the position on the European Union.

Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1945–1998 period. The subsequent period is based on Polk et al. (2017) and Bakker et al. (2015).

For a list of party abbreviations and parties, consult the chapter appendix.

Notes:

24 25 26 27

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

616

  

popular support, they rarely transferred into winning large numbers of seats. Even when the party was more efficient at winning seats between 1997 and 2010, it found that vote shares of over 20 per cent were still winning it less than 10 per cent of the parliamentary seats. Officially, the party was seeking to win power outright, but it was implicitly accepted within and without the party that its main aim was to bring about a situation in which it would be required to be part of a coalition government, through which it could push political reforms (and particularly electoral reform) that would allow it to be fairly represented. The electoral system was also a major hurdle for the other UK-wide party of note: the United Kingdom Independence Party (UKIP), established after the Maastricht Treaty to campaign for Britain’s withdrawal from the European Union. For most of the next 20 years, its support grew steadily, albeit from a low base, and it had particular success at elections to the European parliament, thanks to the proportional electoral system used for those, coming third in 2004, second in 2009, and first in 2014. However, it was unable to turn this success into seats in the Westminster parliament—winning 14 per cent at the 2015 election only resulted in it winning one solitary seat. UKIP has generally been seen as a right-wing party, with debate about it centring on whether it is merely a rightwing challenger to the Conservatives, advocating ‘traditional’ conservatism, or a genuine populist radical right party (Ford and Goodwin 2014). Finally, there are also other party systems involved in UK politics within the constituent nations, particularly in Northern Ireland and Scotland. Northern Ireland has a completely separate party system, based on the unionist–nationalist cleavage, and while some of its parties have had links with the UK-wide parties (particularly the Ulster Unionists with the Conservatives and Social Democratic and Labour Party with Labour) the members of parliament (MPs) elected from there have tended to be counted separately from the other parties. In Scotland, the creation of the Scottish parliament in 1999 was intended to stymie the growth of nationalist sentiment, but it has risen instead, culminating in the Scottish National Party (SNP) becoming the largest party in the Scottish parliament from 2007 and having a majority of Scottish MPs in 2015 (56 out of 59), 2017 (38 out of 59), and 2019 (47 out of 59) reaping the benefits of being a strong party with a divided opposition under plurality voting. The party system and electoral system have reinforced each other to ensure that elections have returned majorities for one or other of the main two parties, despite challenges from outside the two-party system. The three occasions since 1945 when neither main party has won a majority have been as much about the closeness of the battle between the two main parties as they have been about the effect of the third parties. February 1974 saw a dramatic rise in the Liberal vote from the 1970 election, though it only delivered them eight more seats, with the more crucial factor being a mere 0.7 per cent gap between Labour and the Conservatives. In 2010, the total number of seats won by either Labour or

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

617

Table 18.1b Electoral alliances and pre-electoral coalitions in the United Kingdom, 1945–2018 Election date

Constituent parties

Type

Types of pre-electoral commitment

1983-06-09

Lib, SDP

EA, PEC

1987-06-11

Lib, SDP

EA, PEC

Written contract, Joint press conference, Separate declarations, Other* Written contract, Joint press conference, Separate declarations, Other*

Notes: * One joint ‘Alliance’ candidate supported by both parties in most (1983) or all (1987) Great Britain constituencies. Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other

the Conservatives was higher than in 2005, but the Conservatives were unable to turn their lead in votes into sufficient seats. In 2017, both parties gained significant numbers of votes on 2015, with the total number of third-party seats falling, but the closeness of their respective vote shares meant neither achieved a majority of seats. Electoral alliances and pre-electoral coalitions have not been a major feature of the UK electoral system, with the only full-blown electoral alliance of note being that between the Liberals and the Social Democratic Party (a centrist splinter from Labour) in 1983 and 1987, which led to the two parties merging into the Liberal Democrats. Aside from that, the broad catch-all nature and wide electoral appeal of the major UK parties, and the weight of that two-party tradition, has made electoral alliances seem unnecessary, even impossible (Wager 2017 and Table 18.1b).

Government formation As we have noted, the British political system is based around the idea of there being a single party in government, with a rapid transition from one government to its successor facilitated by there being no need for negotiation between parties to form a government. Table 18.2 demonstrates this overall pattern. The three occasions where this did not happen (February 1974, 2010, and 2017) led to three different outcomes (an unsupported minority government, a majority coalition, and a minority government granted a legislative majority by a supply and confidence deal respectively). There have also been two occasions (the Callaghan administration after 1977 and the Major administration after 1996) when a government that previously had a parliamentary majority lost it because of

Year in

1945 1950 1951

1955 1955 1957 1959 1963 1964 1966 1970 1974

1974 1976 1979 1983 1987 1990 1992 1997

Cabinet

Attlee I Attlee II Churchill II

Eden I Eden II Macmillan I Macmillan II Douglas-Home Wilson I Wilson II Heath Wilson III

Wilson IV Callaghan Thatcher I Thatcher II Thatcher III Major I Major II Blair I

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 1

0 0 1

Number of inconclusive bargaining rounds

Lab (1) Con, Lib

Con (1) Con, Lib

Parties involved in the previous bargaining rounds

Table 18.2 Cabinet formation in the United Kingdom, 1945–2018

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 1

0 0 1

Bargaining duration of individual formation attempt (in days)

1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 4

1 1 3

Number of days required in government formation

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 3

0 0 3

Total bargaining duration

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2001 2005 2007 2010

2015 2016 2017

Blair II Blair III Brown Cameron I

Cameron II May I May II

0 0 1

0 0 0 1

Con (1) Con, DUP

Con, Lib (1) Lab, Lib

0 0 0 5 3 0 0 17 3 0 0 1

1 1 1 5 0 0 17

0 0 0 5

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

620

  

defections and by-election defeats. While the bulk of this chapter will focus on the 2010 election and its aftermath, it is worth taking a brief look at those other occasions so as to provide contrast with, and illumination of, the decisions that led to coalition in 2010 and to suggest why it was seen as a break from previous experience. The February 1974 election produced a somewhat anomalous result: while the Conservatives (the incumbent government) won the most votes, Labour won the most seats—something that had not happened since 1951, when the Conservatives had been the beneficiaries of this quirk of the electoral system. Immediately following the election, Edward Heath, the incumbent Conservative prime minister invited Jeremy Thorpe, leader of the Liberals for talks at Downing Street about the possibility of a deal between the two parties. When Thorpe rejected Heath’s proposal, Heath resigned and Wilson then formed a minority government, which remained in place for eight months until he called a second general election in October that saw Labour returned with a majority. By 1977, Wilson had resigned, replaced by James Callaghan and the Labour government had lost the majority it had won in October 1974. Facing the possibility of losing a no confidence vote in the Commons, which would have ended the government and triggered a general election, Callaghan negotiated a one-year confidence and supply deal (the ‘Lib-Lab Pact’) with the new Liberal leader, David Steel, in which the Liberals agreed to maintain the government in power in return for very limited policy concessions—an arrangement that fell some way short of a full coalition. The deal was not renewed after one year of operation, and Callaghan’s government was eventually defeated in a parliamentary confidence vote tabled by the Conservatives under Margaret Thatcher, who went on to win the 1979 election and begin her long reign in office. By 1996, the Conservative government of Thatcher’s successor, John Major, lost its parliamentary majority and faced heavy defeat in the event of a confidence vote leading to a general election taking place. In order to buy the government some more time in the hope things would turn around electorally, an informal confidence and supply deal was agreed with the Ulster Unionist Party—a party that had been historically close to the Conservatives—but this deal was not publicly acknowledged in the same way as the Lib-Lab Pact. Fast forward to 2017 and the election held in June that year resulted in the Conservatives losing the majority they had won in 2015 following five years of coalition with the Liberal Democrats. However, no other party or combination of parties looked likely to be able to put together either a coalition or a deal that would defeat them. As the sitting prime minister, Theresa May was able to first obtain a request from the Queen to form a government and then to negotiate a confidence and supply deal with the Democratic Unionist Party. While short of a formal coalition (which appears to have been discussed briefly by the parties but

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

621

swiftly discarded⁴), it did mark the first time a modern British government would be maintained in office from the start of its term by the parliamentary support of another party in a formal published agreement. These examples—recorded in Tables 18.1b and 18.2—reveal some important points about the processes behind government formation in the United Kingdom that will have a bearing on the processes that created the Conservative–Liberal Democrat coalition of 2010. First, they demonstrate the power of incumbency: the sitting prime minister would seem to be able to remain in while making deals (the United Kingdom does not have any real tradition of caretaker leaders, nor of formateurs and informateurs). Second, it shows that parliamentary support can be and has been agreed upon in both formal and informal ways to support governments, while even governments with majorities have made informal deals with minor parties to ensure majorities on tricky votes. Third, they remind us that these situations are very rare occurrences, and that while there had been much theoretical discussion of parliamentary coalitions before 2010, there was little in the way of practical experience in how to negotiate one nor any agreement on a set of rules by which such negotiations would work.

The bargaining process The nature of UK election declarations—happening constituency by constituency throughout the night and into the next day—meant that the possibility of any coalition negotiations was not clear until well into Friday, 7 May. First, the parties had to be sure of what the result was—both for themselves and the other parties— before being able to decide what their strategy was, having grown accustomed to a system where this sort of decisions were not required of them at all, let alone on the day after the election. As time rolled on, it eventually became clear that the final result would be inconclusive: 307 Conservative MPs, 258 Labour MPs, 57 Liberal Democrat MPs, and 28 others. Some pre-election preparations had taken place to acclimatize the parties for this eventuality, particularly within the Liberal Democrats and Conservatives, while the cabinet secretary (a top civil servant), Gus O’Donnell, had issued some guidelines in February 2010 about the role of the civil service in facilitating negotiations and the position of the government during them. Critically, Liberal Democrat leader Nick Clegg had previously stated that in the event of a hung parliament after an election, his party would speak first to the party that had won the most seats and votes, which now implied the Conservatives (and luckily for Clegg, he did not face a situation like former Liberal leader Jeremy Thorpe in 1974 ⁴ Media reports around 9 June talk of the possibility of a coalition, but announcements the following day only talk of confidence and supply deals.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

622

  

where one party won the most seats and the other the most votes). This position had been developed by Clegg and his team during the election campaign for political reasons—so that they would have a response to ‘who would you back in a hung parliament?’ questions—not constitutional ones, as Clegg did not pay attention to anything recommended in the existing Cabinet Manual or guidelines. Unusually, this meant David Cameron, leader of the Conservatives in opposition since 2005, had the initiative rather than Gordon Brown, the incumbent Labour prime minister. In his first statement after the result was clear, Cameron effectively began the bargaining process by not declaring himself the victor of the election and insisting on becoming prime minister (despite some expectations that he would do this; see Adonis 2013) but by signalling that he wanted to make a ‘big, open and comprehensive offer’ to the Liberal Democrats. Hours later, the first formal meeting between the two parties began, and substantive talks between the Liberal Democrats and Conservatives continued over the weekend. These talks took place in government buildings, with limited assistance from the civil service (the participants opted not to take on all the assistance that was offered to them (Laws 2010)) and consisted of four senior politicians from each party, although not including either party leader directly. Alongside the Conservative–Liberal Democrat negotiations, a parallel process was going on, albeit at a slower, more halting pace, between Labour and the Liberal Democrats, with much of it taking place at a more informal, leadership level than through a formal negotiation process. Brown and Clegg spoke at several points throughout the weekend, both in person and by phone (Laws 2010; Adonis 2013). However, any formal process between the two parties did not begin until Monday, after Brown had conceded Clegg’s insistence that any coalition between the two parties would have to be under a new Labour prime minister rather than Brown himself, meaning the talks took place under the assumption that any deal they negotiated would also have to factor in the outcome of a Labour Party leadership election. What we can see from the many accounts of this negotiating process is that it was as much a political process as it was a technical one, with all sides bargaining in a situation where they possessed incomplete information, not least as to how any victory might be declared in the negotiation ‘game’. Both Conservative (Wilson 2010) and Labour (Adonis 2013) negotiators were sceptical of the motivation of the Liberal Democrats, each at points being sure that they were in a sham negotiation that was only there to increase the third party’s strength in negotiation with their rival, while the Liberal Democrats were unsure if either party’s negotiating team adequately represented the beliefs and interests of their wider party about the desirability of entering a coalition (Laws 2010, and interview). Outside of the negotiation process, other political actors were also attempting to influence proceedings, ranging from newspapers demanding Brown’s

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

623

immediate resignation (calling him a ‘squatter in Number 10’ when in fact he was only performing his constitutional duty), through dire warnings of financial market collapse should a deal not be reached within days, to marches and demonstrations calling for electoral and political reform that demanded that the Liberal Democrats not surrender that prize in negotiations. The Conservative–Liberal Democrat negotiations were the most productive of the two parallel processes with the politicians of both sides finding common ground and leading on the process of creating the outline of the coalition agreement that would be presented to both parties. This was principally a politically driven process, with initial work from the Conservative Oliver Letwin and Liberal Democrat Danny Alexander being supported by their political advisers with only limited contributions from civil servants. While Labour–Liberal Democrat negotiations had found some areas of potential concord, they never reached this productive stage of producing more than a rough draft of a potential agreement. This is perhaps a result of the comparative level of preparation the two sides had done before the election with the Conservative document on potential shared policies described to us as ‘being like a (government) white paper’ while the Labour equivalent was ‘a side and a half of A4’ (Interview). These political pressures were part of the reason the negotiation process was short, especially in comparison with the length of negotiation seen in other countries. The mix of Brown’s relative unpopularity, the impracticality of the alternatives to a Liberal Democrat–Conservative coalition, and public and media expectation of a rapid resolution of the election led to Brown announcing his resignation as prime minister on Tuesday, 11 May, with Cameron being summoned to Buckingham Palace and announcing, on his return to Downing Street, that his new government would be a coalition between the Conservatives and Liberal Democrats. Despite this announcement, the outline coalition agreement was not approved by the Liberal Democrat parliamentary groups and federal executive until later that night. The process of coalition negotiation in the United Kingdom in 2010 was a clear example of the expectations of a system established on the assumption of singleparty majority government having to deal with a situation where such a government was clearly impossible. These institutional expectations of a quick resolution and speedy appointment of a new prime minister were challenged by the need for negotiations between parties to establish a stable government, but still created a situation where those negotiations were essentially complete and a new government in place within a week of the election. Clearly the fact that an agreement between them would produce a minimum winning coalition was important (Bale 2011), but the decision by the Conservatives and Liberal Democrats to plan (separately) before the election for this situation undoubtedly facilitated the formation of the coalition between them, as the speed at which events were decided advantaged parties with firmly established positions. Furthermore, crucial

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

624

  

to this process was Cameron’s decision to not ‘claim victory’ and demand to be appointed prime minister immediately after the election but instead to allow space for the negotiations with the Liberal Democrats to proceed in a spirit of exploration rather than trying to bounce them into a coalition—a move that could easily have provoked the effective ‘kingmaker’ into trying harder than it eventually did to come to an agreement with Labour.

The composition and size of cabinets The British cabinet has stayed at approximately the same size since the war with usually just over 20 full members of the cabinet representing a slightly lower number of ministries (see Table 18.3). This is because not all members of the cabinet represent a ministry: it is common practice for the Leaders of the Commons and the Lords (the parliamentarians charged with managing the business of the lower and upper houses), the Government Chief Whip, and other ministers without portfolio to be given seats in the cabinet, as well as the Treasury normally having two members of the cabinet (the chancellor and their deputy, the chief secretary). All members of the cabinet are political rather than technocratic appointments, with the majority usually MPs, though Lords are occasionally appointed to cabinet positions (and more often to junior roles). The roles and names of the ministries are generally consistent, but occasionally changed by the government either to react to changing times or to signal political priorities. The coalition government of 2010–2015 made no major changes to the size and composition of the cabinet in technical terms, the only significant change being that it was the first cabinet since 1945 to include members of two different political parties.

The allocation of ministerial portfolios Ministerial portfolios are usually in the gift of the prime minister, acting as the representative of the Crown whom secretaries of state are technically representatives of. Appointments do not require the confirmation of parliament and ministers can be required to resign or else dismissed by the prime minister without recourse to parliament. As well as making appointments at the start of the parliamentary term, prime ministers will often conduct ‘reshuffles’ of the cabinet during the parliamentary term where ministers who are underperforming and/or politically troublesome (or past their ‘sell-by date’) can be removed and others promoted within or to the cabinet.

23

Year Number of ministers per Total number of in party (in descending ministers order)

Cameron I 2010 18 Con, 5 Lib

Cabinet

7 Lib, 3 Con

Number of watchdog junior ministers per party

Table 18.3 Distribution of cabinet ministerships in British coalitions, 1945–2018

18

Con

Con

Con

Con

Con

5 4 2 3 Number 1 Prime Finance Foreign Work and Interior of affairs pensions ministries minister

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

626

  

While these formal rules remained in place for the coalition, an additional informal step was put in place where Cameron would consult with Clegg, the deputy prime minister, on appointments and Clegg would be responsible for choosing who filled the positions within government allocated to the Liberal Democrats. Table 18.3 shows the coalition distribution over important ministries. Allocation of portfolios was based on the proportion of MPs each party had, rather than the share of the vote they had received. This meant that the Liberal Democrats had five members of the Cabinet, with a similar share of junior ministries. Clegg took the decision that the party should be involved with the full workings of government, so junior ministers were predominantly placed—as watchdogs perhaps—in the ministries where the Liberal Democrats were not represented at the cabinet level. The Liberal Democrats held the same five cabinet positions throughout the life of the coalition—deputy prime minister (with a responsibility for political and constitutional reform but no direct ministerial role), chief secretary to the Treasury, business secretary, energy and climate change secretary, and Scottish secretary. Extraordinarily, at least in comparison with their counterparts in most continental countries, the junior coalition partner made no bid for any of the biggest jobs in the cabinet (namely foreign secretary, home secretary, and chancellor of the exchequer) nor for big-spending departments like health, education, or work and pensions (which handles social security and welfare) (see Table 18.3). This is perhaps reflective of Clegg’s approach to the coalition as a whole. Beginning with the negotiations, his approach was to first seek agreement on policy and relegate issues of jobs and personnel to a later stage. Certainly, sources indicate that ‘policy, then jobs’ was Clegg’s intent in negotiations, which perhaps provides another explanation for his choosing the Conservative option. While they were willing to work on a joint policy platform (and had prepared for that eventuality), Labour’s approach was based on an assumption that the Liberal Democrats would generally agree with their policy platform and the main focus would be the allocation of positions.

Coalition agreements The full coalition agreement was made up of two documents: the shorter Coalition agreement for stability and reform, which principally covered procedural matters, and the longer The coalition: our programme for government,⁵ which focused mainly on the policy priorities of the new government. The separate publication of the two documents is important in showing the nature

⁵ Both are available at https://www.gov.uk/government/publications/the-coalition-documentation.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

627

Table 18.4 Size and content of coalition agreements in the United Kingdom, 1945–2018 Coalition

Year Size in

Distribution Distribution Policies General Policy(in %) of offices of rules specific competences (in %) procedural (in %) (in %) rules (in %)

Cameron I 2010 1,295* 67.5 13,682** 0.7

0 1.1

22.7 0

0 0

3.1 74.2

Notes: * ‘Agreement for stability and reform’ ** ‘The coalition: our programme for government’

of the coalition and its incongruity within the British system. The coalition needed to set out its policy priorities and show the electorate what the new government would be doing; it had to also establish the rules by which it would be operating, in order to show how it would adapt (and adapt to) the established customs of British government. Table 18.4 shows the relative emphasis on each part (policy versus portfolio and procedure), in terms of share of each agreement. Although the creation of a coalition government was a radical step in British politics, there was little radicalism in the Agreement for stability and reform. Instead, it merely sought to adapt the existing system of parliamentary and cabinet government to work within the framework of there being two parties in government rather than one. So, the document stressed that the two parties would work as a single government and that collective cabinet responsibility—an integral part of the so-called Westminster system of government—would remain in force. It also maintained the top-down distribution of power, merely amending it to acknowledge that the prime minister would consult with the deputy prime minister over certain matters rather than granting any additional powers to the cabinet or parliament. The one major change brought in through the agreement was the decision to announce a fixed-term parliament, with the power of dissolution taken out of the prime minister’s hands. This would be implemented as the Fixed Term Parliaments Act (FTPA). While most of the rules within this document were followed by the government, one fell by the wayside. A proposed ‘Coalition Committee’ that would review and resolve disputes as well as set further policy only met a handful of times and was superseded by the informal ‘Quad’ meeting of two senior figures from each coalition partner. The second document—the Programme for government—was important in fitting the coalition in with the norms of British government, as the new administration, by its very nature, was clearly unable to follow the convention that it had a single manifesto on which it had won the election. Instead, an approach setting

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

628

  

out the jointly agreed policies of the two parties in a single document was adopted, and it was on this document that the government’s programme for the next five years would be based (see Quinn et al. 2011 for an interpretation of how much each party was represented in this document). One consequence of this approach which in the end went untested, however, was the so-called ‘Salisbury Convention’ by which the House of Lords would not use its powers to block legislation that was included in the manifesto of the governing party. This is a major, yet self-imposed, limitation on the power of the legislature to limit the government’s ability to legislate its agenda: while a majority government commands a majority in the Commons, it does not necessarily enjoy one in the Lords and, indeed, no government has done so since the Lords was reformed in 1999. With no explicit manifesto to refer to, the coalition government could perhaps have been challenged as a result, but, perhaps fortunately (for the government if not for political scientists and constitutional historians!), such a challenge never arose.

Coalition governance Although the coalition agreements and formal rules were written to create a government in which little would supposedly be different between the coalition and other UK governments, the realities of two parties having to co-exist in government led to many differences in the process of government. Table 18.5 identifies the main practices in the terms established for the cross-national project that we share with the other country chapters.

The role of individual ministers in policy-making If anything, the coalition reduced the influence that ministers—even secretaries of state (i.e. senior rather than junior ministers)—could have in making policy. In previous governments, a minister would have possessed sufficient freedom to develop policies on their own and present them to the cabinet, if not as a fait accompli then as a serious set of proposals. The checks and balances brought in by the coalition, however, meant that this was no longer the case. For instance, in the majority of departments (and all of those responsible for major policy development in public services), there would be at least one Liberal Democrat junior minister whose support any Conservative secretary of state would require in order to put forward any wider policy proposals (and vice versa for those departments with a Liberal Democrat in charge). Thus, any individual freedom was severely curtailed by the need to engage in a wider process of involvement within their own department before they could even take their policy outside it. This led to departments being mostly concerned with implementing the policies agreed in

N

IC, IC CaC

IC

Y

N

Most/All

Y

Comp.

Y

Policy agreement: Comp. = Comprehensive policy agreement

Y

Junior NonCoalition Freedom of Policy discipline in appointment agreement ministers cabinet (watchdogs) positions legislation/ other parl. behaviour

Coalition discipline: All = Coalition discipline is always expected; Most = Coalition discipline is expected on all matters except those explicitly exempted

Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee

Coalition agreement: POST = post-election

Notes:

Y

All Most For used common most serious conflicts

Year Coalition Agreement Election Conflict management Personal Issues in agreement public rule mechanisms union excluded from agenda

Cameron I 2010 POST

Coalition

Table 18.5 Coalition governance mechanisms in British coalitions, 1945–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

630

  

the coalition agreement. One notable, but arguably disastrous, exception to this was health where Andrew Lansley, the coalition’s first health secretary, was able to introduce major reforms of the National Health Service in a way that had not been agreed at a wider political level because he had been able to persuade his Liberal Democrat minister of state, Paul Burstow, to support his vision and work with him on developing the policy rather than bringing in the wider policy structures of the government and coalition (d’Ancona 2013). It is perhaps not coincidental that, while the government eventually chose to soldier on with a version of the reforms proposed by Lansley and Burstow, both ministers were reshuffled out of their positions at Cameron’s first reshuffle in 2012.

Coalition governance in the executive arena The appointment of junior ministers—the United Kingdom appoints two levels of junior minister, a minister of state directly underneath the secretary of state and an undersecretary of state at a level below them—was used to extend coalition governance throughout the government. Eschewing the idea of the party taking responsibility for distinct policy areas, Clegg and the Liberal Democrats instead chose a policy of ‘owning the coalition’ and having involvement across the full range of government policy and activity, linked to Clegg’s role as the deputy prime minister. Both parties were also able to appoint special advisers (often known acronymically as SpAds) to departments—full-time, albeit temporary, employees of the department but who are appointed directly by (and work directly to) a minister and are therefore not bound by civil service rules on impartiality. In some cases, SpAds were used to monitor work going on throughout departments to ensure that it was in line with coalition policy but also to keep each party represented in policy development. Negotiations on policy, particularly in the later periods of the coalition when the main areas of the initial agreement had been delivered, would often involve special advisers both alongside and separately from ministers. The coalition stuck with the tradition of much work being done by cabinet committees, which would bring together a select group of ministers, usually chaired and led by a secretary of state and, in certain more high-profile cases, by the prime minister or deputy prime minister. These committees—containing representatives of both parties—would develop policy on issues that cut across departments and report back to the cabinet. Each sub-committee would have a chair from one party and deputy chair from the other in order to ensure both had involvement in the direction of the committees though this often led to individual Liberal Democrat ministers having to sit on many more committees (with the resultant increase in workload) than their Conservative colleagues (d’Ancona 2013: 37).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

631

Sitting over these levels of party working at the executive level was the Quad—the ad hoc committee of Cameron (Conservative prime minister), Clegg (Liberal Democrat deputy prime minister), Osborne (Conservative chancellor) and Alexander (Liberal Democrat chief secretary to the Treasury), who replaced the planned formal coalition committee with more members outlined in the coalition agreement. This group became effectively the inner cabinet, responsible for setting the policy direction of the government, particularly after the policies of the coalition agreement had been implemented. Disputes and arguments at lower levels would be escalated to the Quad for decisions, and even though this forum was not included as such in the coalition agreement, it effectively became the main location for dispute resolution between the two parties.

Government mechanisms in the parliamentary arena One interesting aspect of the 2010 coalition agreement was that, as well as promising wider political change, it also promised many reforms to the way parliament worked. Both parties had stressed in their manifestos the importance of cleaning up parliament following the MPs’ expenses scandal of 2009, and the initial package of reforms promised would have led to a major change in the way parliament worked (Cowley 2015). However, despite this promise of a brave new world in parliamentary procedure, the coalition worked in a similar way to previous governments, only now with a whipping operation that had to operate across two parties rather than one. Throughout the coalition, the government chief whip (with a seat at the cabinet table) was a Conservative and the deputy chief whip a Liberal Democrat. This was mirrored in the appointment of a Conservative leader of the House to manage its business in parliament along with a Liberal Democrat deputy. The coalition agreement pledged both parties to supporting the government’s agenda in full except for a few opt-outs on policies, with Liberal Democrat MPs allowed to abstain on proposals for university tuition fees, nuclear power, and a change in tax arrangements to favour married couples. In practice, while the second two held, on the issue of tuition fees, the political backlash to the Liberal Democrats’ decision to abandon a high-profile pledge to abolish tuition fees from the 2010 election led to the party splitting three ways in the Commons vote on the measure, with some voting for, some abstaining, and others voting against. Crucial, however, to all of these cases was that merely allowing abstention on issues did not threaten the government’s ability to command a parliamentary majority. With Liberal Democrats not voting (or cancelling themselves out, as they did on the tuition fees vote) the 306 Conservative MPs were still more than the combined total for all other parties in parliament.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

632

  

Governance mechanisms with different types of actors The coalition agreement set out the broad nature of the way governance would be handled, following the traditional pattern of British government with decisions made at the top with the collective responsibility of the cabinet and then party whips ensuring it passed through parliament. The major concessions from the traditional model of governance were the announcement of a Coalition Committee, superseded by the Quad, to handle disputes and two parallel whipping operations, one for each party. As Table 18.4 showed, the parties devoted much more attention in the agreement to setting out the government’s policies than they did to the rules by which the government would be run. Accounts of the coalition negotiations suggest too that policy discussions were of much more interest to the negotiators than process ones. This policy focus is perhaps the key to understanding coalition governance for the 2010–2015 government alongside the fact that national-level coalitions and minority governments have been very rare in the United Kingdom. Clegg’s position of the Liberal Democrats ‘owning the coalition’, choosing to be involved across all policy areas, and presenting coalition policy as more important than party policy meant that coalition governance was a process of behind-the-scenes policy negotiation with the Quad at the heart of the process. In this, we see again the idea that it was the parties who adapted to the accepted practices of UK government—control from the centre based around the prime minister’s office— rather than the processes of government and governance adapting to the new situation.

Cabinet duration and termination In discussing the duration of cabinets, and particularly the coalition, we must remember that the coalition does mark a major change in the duration and termination of British governments: after all, it oversaw the introduction of the FTPA replacing the rules described earlier, established by the Septennial Act and Parliament Act, which only gave a maximum duration to governments of five years, but did not prevent their easy termination at the whim of the prime minister, who could personally decide to hold a fresh election by requesting (and almost certainly getting) a dissolution from the monarch.

The duration of cabinets Since the Second World War, governments of the United Kingdom have varied in length from the eight months of the 1974 Wilson minority government

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

633

to the several that have run for all or almost all of their five-year term, with the Conservative–Liberal Democrat coalition in this latter group. As it is merely one case, coupled with the confounding factor of the FTPA, it is near impossible to extrapolate from this one example so as to produce a general rule about the relative stability of coalition compared to single-party government in the United Kingdom—especially as the nature of the other governments that have lasted the full five years has been quite varied, from the large majorities enjoyed by the Attlee (1945–1950), Macmillan/Douglas-Home (1959–1964), Thatcher/Major (1987–1992), and Blair/Brown (2005–2010) administrations to the vanishing majorities of the Wilson/Callaghan (1974–1979) and Major (1992–1997) administrations. Table 18.6 lists the main reasons for cabinet termination. Two points are particularly worth noting about the coalition government’s duration. First is that it remained a coalition government throughout the five years of its existence, with neither party seeking to withdraw from it and leave the other as a caretaker in the run-up to the election. The Liberal Democrats were the only party who could realistically have done this, and at times it was assumed or believed they might in order to gain some supposedly useful independence and freedom to criticize the government in the run-up to the 2015 election. Yet they chose not to take that path. Linked to this, in spite of opposition from within their parties, both Cameron and Clegg remained party leaders for the duration of the coalition with neither facing a formal challenge to their leadership. Clegg came closest to being removed following poor local and European election results in 2014, but none of his colleagues proved willing to try to unseat him. The same went for Cameron, although things might have been different had he not bowed in early 2013 to backbench and media pressure to concede a referendum on the UK membership of the European Union. Given that the coalition had—at least in part and at least initially—been built on a widely reported personal bond between the two men, both of whom came from similarly privileged backgrounds, it is questionable if it could have continued had one of them been removed. Secondly, it is worth asking whether the FTPA kept the coalition in place or if the two parties would have continued to hang together in its absence? Had Cameron had the power to call an election at a time of his choosing before May 2015, were there any circumstances in which he might reasonably have done so? In fact, it seems unlikely, given the consistent (if ultimately deceptive) opinion poll lead enjoyed by the Labour Party from the beginning of 2011, that he would have taken a risk on an early election in any case. So while the new rules may have helped keep the show on the road when things got tough (as they did, for instance, when the Liberal Democrats, already sore at the way Cameron campaigned against electoral reform in the referendum on the issue held in 2011, pulled the plug on boundary changes after Conservative MPs effectively scuppered Lords reform in 2012), the difficulty of calling an early election did not perhaps determine events as much as some had expected.

Date in

1945-07-26 1950-02-24 1951-10-26 1955-04-06 1955-05-27

1957-01-10 1959-10-09

1963-10-19 1964-10-16

1966-04-01 1970-06-19

1974-03-04

1974-10-11

1976-04-05 1979-05-04 1983-06-10 1987-06-11

Cabinet

Attlee I Attlee II Churchill II Eden I Eden II

Macmillan I Macmillan II

Douglas-Home Wilson I

Wilson II Heath

Wilson III

Wilson IV

Callaghan Thatcher I Thatcher II Thatcher III

1979-05-03 1983-06-09 1987-06-11 1990-11-28

1976-04-05

1974-10-10

1970-06-18 1974-02-28

1964-10-15 1966-03-31

1959-10-08 1963-10-19

1950-02-23 1951-10-25 1955-04-05 1955-05-26 1957-01-09

Date out

87.5 82 78.8 69.3

29.7

12.1

84.3 73.9

100 31.9

81.4 80.5

91.6 33.4 68.8 8.8 32.5

Relative duration (%)

4, 6 4 4 9, 10

9

4

4 4

1 4

4 9

1 4 9 4 9

Mechanisms of cabinet termination

Table 18.6 Cabinet termination in the United Kingdom, 1945–2018

10

13

13

12

12, 14

13 14

Terminal events

Parties (when conflict between or within)

Policy area(s)

Thatcher resigned after failing to win party leadership election in first round.

Heath called election to determine ‘who runs Britain?’. Minority government seeking working majority Wilson resigned for personal reasons.

Early election was called to gain a working majority.

Macmillan resigned because of Profumo scandal.

Eden resigned following Suez Crisis, citing personal reasons.

Churchill resigned because of age.

Comments

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

2001-06-07

2005-05-05

2007-06-27 2010-05-11 2015-05-07

2016-07-13

Blair II

Blair III

Brown Cameron I Cameron II

May I

2017-06-08

2010-05-06 2015-05-07 2016-07-13

2007-06-27

2005-05-05

1992-04-09 1997-05-01 2001-06-07

23.7

100 100 23.8

42.9

78.2

88.8 100 82

4

1 1 9

9

4

1 1 4

11

11

11

Note: Relative duration: share of constitutionally allowed duration for this cabinet.

Terminal events 10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event; 14: Personal event

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament; 7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties; 9: Other voluntary reason

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister

1990-11-28 1992-04-10 1997-05-02

Major I Major II Blair I

Cameron resigned after the leave option won in the 2016 EU referendum. Early election was called to gain mandate for negotiating Brexit.

Early election was called due to strong polling. Early election was called because of demands arising from EU presidency and G8 summit in later parts of 2005. Blair voluntarily resigned in May and in June Brown’s government took over.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

636

  

The termination of cabinets Owing to all this, the Conservative–Liberal Democrat coalition of 2010–2015 was that rare breed of British government that gets to die peacefully in its bed at the end of its term rather than its life being cut short either by crisis or else by opportunistic pursuit of election victory. It remains an open question as to whether or not this decision to legislate for and then stick to a five-year term strengthened the government, particularly during some of its more difficult periods of operation (the ‘Omnishambles’ budget and the clashes over political reform in 2012 noted earlier) in that it may have forced the two parties to find a resolution to their differences and a way forward, rather than simply walking away from the government altogether.

Conclusions The United Kingdom is an outlier (perhaps all the more so post-Brexit) from other European countries in numerous ways—in this instance through it only having had one coalition government since 1945. This makes the coalition itself an outlier within an outlier. But rather than focus simply on outlining and explaining its sheer peculiarity, we decided to ask, too, whether it had a transformative effect on British politics, both in the institutional arena and the party-electoral one. In terms of institutions, while the Liberal Democrats did have many radical ideas about reforming the way they worked, circumstances meant that very few of those ideas were instituted: the voting system was not changed, the House of Lords was not reformed, and the powers of parliament in relation to the executive were not fundamentally altered. In other ways, the coalition found itself fitting in with the established forms of institutions, attempting to mimic the form of a singleparty government to keep the Westminster model on track instead of challenging it with ideas about how a multiparty government might work. The one major institutional change of the coalition government, the FTPA, may, however, have caused more long-run change than was originally, though both sides of the argument for this were tested in 2017, two years after the end of the coalition. The supposed power of the FTPA to prevent an early election was proven illusory when Labour sided with the Conservatives in voting to dissolve parliament early. On the other hand, following that election, Theresa May, having lost her majority, was able to use the FTPA to keep her new government in power and damp down expectations of a further election that year. The FPTA did not, though, prevent another early election in 2019. Yet, leaving aside its relatively light impact on the rules of government, the 2010–2015 coalition has arguably played a more fundamental role in shaking up the wider political culture of the United Kingdom. Even before Brexit, the general

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

637

election result of 2015 was highly anomalous, with both main parties doing poorly, the Liberal Democrats crashing to a near-fifty-year low in terms of both seats and votes and UKIP and the SNP experiencing incredible surges of support—a surge which, in the case of the latter, was also rewarded in Westminster seats as well as votes. This post-coalition reshaping of the party landscape is still ongoing, with the Liberal Democrats unable to recover support or shed the stigma of coalition while the Conservatives, post-Brexit, have recently gone from strength to strength under Boris Johnson. While the end of the coalition may have appeared to have restored normal politics and the return of single-party government in 2015, any brief calm was upended by the Brexit referendum. As such, while the coalition may have adapted itself to fit into the established British system of government, that system itself may be under such strains that it ends up changing anyway.

References Adonis, Andrew (2013). 5 Days in May: The Coalition and Beyond. London: Biteback. Bakker, Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Anna Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999-2010’. Party Politics, 21(1): 143–52. Bale, Tim (2011). ‘I don’t Agree with Nick: Retrodicting the Conservative-Liberal Democrat Coalition’. Political Quarterly, 82(2): 244–50. https://doi.org/10.1111/j. 1467-923X.2011.02187.x. Bale, Tim (2012). ‘The Black Widow Effect: Why Britain’s Conservative-Liberal Democrat Coalition Might Have an Unhappy Ending’. Parliamentary Affairs, 65 (2): 323–37. https://doi.org/10.1093/pa/gsr033. Beech, Matt (ed.) (2015). The Conservative-Liberal Coalition: Examining the CameronClegg Government. Basingstoke: Palgrave Macmillan. Clegg, Nick (2016). Politics: Between the Extremes. London: Bodley Head. Cowley, Philip (2015). ‘The Coalition and Parliament’. In Anthony Seldon and Mike Finn (eds), The Coalition Effect, 2010-2015. Cambridge: Cambridge University Press, 136–56. Crace, John (2014). I Never Promised You a Rose Garden: An Insider’s Guide to Modern Politics, the Coalition and the General Election. London: Bantam Press. d’Ancona, Matthew (2013). In it Together: The Inside Story of the Coalition Government. London: Viking. Debus, Marc (2011). ‘Portfolio Allocation and Policy Compromises: How and Why the Conservatives and the Liberal Democrats Formed a Coalition Government’. Political Quarterly, 82(2): 293–304. https://doi.org/10.1111/j.1467-923X.2011. 02191.x

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

638

  

Evans, Stephen (2012). ‘Reluctant Coalitionists: The Conservative Party and the Establishment of the Coalition Government in May 2010’. Political Quarterly, 83 (3): 478–86. https://doi.org/10.1111/j.1467-923X.2012.02342.x Ford, Robert, and Matthew Goodwin (2014). Revolt on the Right: Explaining Support for the Radical Right in Britain. Abingdon and New York: Routledge. Fox, Ruth (2010). ‘Five Days in May: A New Political Order Emerges’. Parliamentary Affairs, 63(4): 607–22. https://doi.org/10.1093/pa/gsq030 Gerard, Jasper (2011). The Clegg Coup: Britain’s First Coalition Government since Lloyd George. London: Gibson Square. Lakin, Matthew (2013). ‘The Ideology of the Coalition: More “Muscular” than “Liberal”?’. British Politics, 8(4): 76–490. https://doi.org/10.1057/bp.2013.9 Laws, David (2010). 22 Days in May: The Birth of the Lib Dem-Conservative Coalition. London: Biteback. Laws, David (2016). Coalition: The Inside Story of the Conservative-Liberal Democrat Coalition Government. London: Biteback. Lees, Charles (2011). ‘How Unusual Is the United Kingdom Coalition (and What Are the Chances of It Happening Again)?’. Political Quarterly, 82(2): 279–92. https://doi. org/10.1111/j.1467-923X.2011.02192.x Müller, W. C., & Strom, K. (Eds.). (2000). Coalition governments in western Europe. Oxford: Oxford University Press. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of AntiElitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research and Politics, 4(1): 1–9. Pulzer, Peter (1967). Political Representation and Elections in Britain. London: Allen and Unwin. Quinn, Thomas (2013). ‘From Two-Partism to Alternating Predominance: The Changing UK Party System, 1950–2010’. Political Studies, 61(2): 378–400. https:// doi.org/10.1111/j.1467-9248.2012.00966.x Quinn, Thomas, Judith Bara, and John Bartle (2011). ‘The UK Coalition Agreement of 2010: Who Won?’. Journal of Elections, Public Opinion & Parties, 21(2): 295–312. https://doi.org/10.1080/17457289.2011.562610 Seldon, Anthony, and Mike Finn (eds) (2015). The Coalition Effect, 2010-2015. Cambridge: Cambridge University Press. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Wager, Alan (2017). ‘Friends with Benefits: A Temporal Comparison of Electoral Pact Negotiations in the British Context’. British Politics, 12: 115–33. https://doi.org/10. 1057/bp.2015.41

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

  

639

Wilson, Rob (2010). 5 Days to Power: The Journey to Coalition Britain. London: Biteback. Yong, Ben, and Tim Bale (2016). ‘Britain’s Experience of Coalition Government: Continuity and Change’. In Richard Heffernan, Colin Hay, Meg Russell, and Philip Cowley (eds), Developments in British Politics 10. 20–38. London, Palgrave.

Appendix. List of political parties Abbreviation Comm ILP SNP PC CW Lab Lib SDP Con UKIP IN U

Name Communist Party of Great Britain Independent Labour Party Scottish National Party The Party of Wales (Plaid Cymru) Common Wealth Labour Party SDP-Liberal Alliance, 1983–1987 Liberal Party, until 1979 Social Democratic Party Conservative Party United Kingdom Independence Party Irish nationalist parties (principally Social Democratic and Labour Party and Sinn Féin) Ulsten Unionist parties (including Democratic Unionist Party)

Note: Party names are given in English, and in the case of The Party of Wales, the party name in Welsh is listed in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 19 Croatia: Strong Prime Ministers and Weak Coalitions Dario Nikić Čakar

Since the introduction of democracy in 1990, the Croatian political system has experienced several transformations, starting with the illiberal democracy and semi-authoritarian rule of President Tuđman during the 1990s, then making a leap to a democratic consolidation at the turn of the new millennium. Major changes occurred in 1999/2000, when the mixed-member proportional electoral system was replaced with proportional representation prior to the parliamentary election in 2000. After that election, the new parliamentary majority adopted constitutional changes that effectively transformed the semi-presidential system into a functional parliamentary democracy. As a result of these institutional changes, a new pattern of executive politics was established, with coalition governments as a norm (Kasapović 2005). However, despite almost two decades of coalition politics in Croatia, there are no clear trends of coalition formation and coalition governance. These vary across coalitions and governments, with major challenges to the survivability of coalitions coming from internal conflict and the lack of established mechanisms to resolve these conflicting issues. The post-2000 period witnessed the strengthening of parliamentary democracy, in which the role of the president was significantly marginalized while prime ministers took over the leading role in executive coalition governance. Prime ministerial dominance in the executive is rooted in the nature of Croatian political parties, as they are highly centralized and ‘presidentialized’ organizations (Nikić Čakar and Čular 2016; Nikić Čakar 2019). Because of this, party leaders are usually the strongest actors in the coalition formation process, determining the content and the structure of negotiated coalitions. Over the last 20 years, two leading parties, the Croatian Democratic Union (HDZ) and the Social Democratic Party (SDP), have dominated coalition politics in Croatia, as their party leaders were prime ministers in every coalition government, except for a short period after the 2015 parliamentary election, when an independent technocrat, Tihomir Orešković, was acting as the prime minister. Orešković’s coalition cabinet was composed of HDZ and the Bridge of Independent Lists (Most) ministers. The cabinet was formed as a direct result of major changes in the Croatian party system, since the rise of the anti-establishment party, Most, and a Eurosceptic  Dario Nikic Cakar, Croatia: Strong Prime Ministers and Weak Coalitions In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0019

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



641

populist party, Human Shield (Živi zid), over the previous few years had altered the structure of party competition and voter behaviour. Furthermore, the electoral success of these new parties also deeply affected the government formation process and changed the conventional rules that had been established over the years. In the 2015 parliamentary election, for the first time, the traditional leftwing and right-wing political blocs failed to secure a sufficient parliamentary majority to form a coalition government by themselves, so they had to start negotiations with Most. Most, in turn, was acting very unpredictably, due to its complete inexperience in national politics. This chapter argues that political parties and party leaders, as the main actors of coalition politics in Croatia, (un)intentionally fail to use institutional memory in coalition governance. This means that they do not build coalition governments on the lessons from their predecessors or even on their own previous experience in coalition-making, particularly when it comes to the questions of coalition bargaining and the implementation of conflict resolution mechanisms within coalition agreements. This notion is especially evident in several features of Croatian coalition politics: the very general and rather brief coalition agreements without written rules on cabinet decision-making and on how to resolve internal conflicts; an informal and personalized way of handling conflicts between coalition parties; the dominant position of the prime minister and limited ministerial autonomy, and the policy and personnel conflicts between coalition parties as the main reason for cabinet termination.¹

The institutional setting Adopted in December 1990, the Croatian constitution at the time set up the institutional structure of a new democratic state as a semi-presidential system, with the president as the chief executive, in control of both foreign and domestic policy. President Franjo Tuđman’s (1990–1999) broad charismatic appeal and the strong concentration of power in the hands of the president, backed by absolute parliamentary majorities, were the supporting pillars of the HDZ’s predominant party rule during the 1990s (Čular 2000; Fisher 2006). The directly legitimized president was invested with constitutional powers to appoint and dismiss the prime minister and cabinet ministers, issue decrees, ¹ This chapter is based on in-depth interviews with two former prime ministers and 20 former or current deputy prime ministers and ministers active in all coalition governments since 1991. Furthermore, evidence is also drawn from the analysis of several pre-electoral agreements, six coalition agreements and a number of additional coalition documents, such as protocols and government programmes. In addition, the archive research was conducted in relation to the journal articles on the coalition governance published in two Croatian daily newspapers Večernji list and Vjesnik. The archives of Croatian government and parliament also served as valuable sources of information for the purposes of this chapter.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

642

´ ˇ 

initiate referendums, and dissolve parliament, although with the prime minister’s countersignature. Tuđman’s supreme dominance was further supported by his unquestioned standing within the party, as the constitution allowed him to simultaneously rule the country and the party (Boban 2017). In the period from 1990 to 2000, the HDZ was in control of absolute parliamentary majorities, giving them an opportunity to form single-party governments, with President Tuđman solely reshuffling prime ministers on a regular basis and deciding on the composition of the cabinet. The only deviation from the HDZ’s absolute dominance in the 1990s was the war cabinet, the Government of Democratic Unity, under Prime Minister Franjo Gregurić, which was established as a broad coalition in the summer of 1991 and consisted of virtually all relevant parties at that time. This coalition government was born in the wake of the war for national independence and was established for the purposes of organizing and coordinating defensive activities (Gregurić 1998, 2017). Although the HDZ alone was in control of an absolute majority in parliament, President Tuđman accepted the idea of including all parliamentary parties in the government of national unity, believing that all decisions concerning war activities would have much higher legitimacy. The necessity of forming the government was also recognized by the opposition leaders, as it was unanimously approved in parliament, but it lasted only for a year and ended with the early parliamentary and presidential elections in 1992, organized under the new constitutional and electoral provisions. By the end of the 1990s, the HDZ’s legitimacy and electoral support started to deteriorate under the pressure of the economic and social crisis (Lamza Posavec 2000; Zakošek 2002). In order to prevent a likely defeat, the HDZ’s ruling majority changed the electoral rules and instituted proportional representation, but with Tuđman’s death in December 1999, the HDZ was defeated in both parliamentary and presidential elections. The new Račan I government initiated comprehensive constitutional reforms to dismantle Tuđman’s authoritarian legacy embedded in the semi-presidential structure and, with the new constitutional rules adopted in November 2000, the presidential office was stripped of most of its previous powers in an attempt to establish a system with a balance of power between the cabinet and parliament (Boban 2016; Nikić Čakar 2019).² The centre of executive power moved away from the presidency to the cabinet, with the prime minister holding the central position in executive politics. Although the president is still directly elected, the presidency has been diminished to a symbolic and ceremonial role and left with only very limited prerogative powers in defence and foreign policy.

² At that point, all relevant political actors agreed that the constitution needed to be changed in order to consolidate democracy in Croatia, including the newly elected president Mesić and the HDZ as the largest party in parliament.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



643

The constitutional reform in the autumn of 2000 was directed not only towards reducing presidential prerogatives but also towards broadening the scope of human rights provisions, particularly in terms of gender equality and positive discrimination in favour of those citizens from minority ethnic groups, giving them special voting rights when electing their parliamentary representatives. New constitutional provisions also strengthened direct democracy by introducing citizen-initiated referenda, with the signatures of at least 10 per cent of registered voters needed to call a referendum. In March 2001, another constitutional change was initiated, this time to abolish the second chamber of the Sabor (the Croatian parliament). The abolished chamber, the Županijski dom (House of Counties), was the weaker chamber anyway and had a very limited constitutional role, both in legislative process and in government formation. In contrast to the period of the 1990s when the president regularly exercised his constitutional powers in the government formation process by appointing and dismissing prime ministers and cabinet ministers at his own will, the post-2000 period was quite different. The importance of the president in forming coalition governments is now entirely marginalized and the process is firmly in control of political parties and parliament. The formal process of government formation starts with the president consulting with the leaders of political parties represented in the Sabor and then the president entrusts the mandate to form a government to the individual who is in control of a majority of seats in parliament (formateur). Up to now, presidents have always entrusted the leader of the largest party with the mandate,³ and all formateurs have managed to form a government in the first round, without the need to change the formateur or even to call an early election. The next stage of the government formation process takes place in the Sabor where the formateur presents the government programme and the candidates for ministerial positions, followed by an investiture vote on the government as a whole, where a majority vote of all members of parliament is required for a government to take office (positive parliamentarism). On four occasions since 2000, coalition parties that won the support of the majority of members of the Sabor technically formed minority governments, although they also enjoyed the confidence of other (in)formal support parties that stayed outside the government (Račan III, Sanader I, Orešković, and Plenković I). As mentioned earlier, after the constitutional change in 2000, the office of the prime minister took over the dominant role in shaping coalition governance in Croatia. The prime minister has the constitutional right to appoint and dismiss cabinet ministers, although it should be taken into account that the appointment

³ The only exception to this norm was Orešković, who was invited by the HDZ and Most to take over the government after the 2015 parliamentary elections. Prior to taking office, he did not have any party or political background in Croatia and was the first person outside the conventional political arena to be appointed as prime minister.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

644

´ ˇ 

follows a successful investiture vote and is done with the formal countersignature of the Speaker of the parliament. However, we should bear in mind the fact that an investiture is voted on the government as a whole, and not on the individual ministers, which makes them less autonomous in relation to the prime minister. In the case of a cabinet reshuffle, the prime minister presents new ministerial candidates in the parliamentary committees, which is followed by an investiture vote on each individual candidate in the plenary session. Furthermore, when assessing power relations between the prime minister and ministers in the internal cabinet structure, it should be stressed that the Law on the Government empowers the prime minister to steer the tasks of individual ministers, as Article 11 states: ‘The prime minister can provide members of the Government with specific instructions for work, special assignments or tasks and authorise them to implement and execute specific projects.’ Besides this, a prime minister’s dominance also stems from the formal right to fully control the agenda for cabinet meetings, on the one hand, and from the administrative and personnel resources at his/her disposal, on the other hand.

The party system and the actors Party system change The party system in Croatia has gone through two main stages of development, with the predominant party rule established during the 1990s and the moderate pluralism that characterized the period after the 2000 elections (Čular 2001; Zakošek 2008). As already mentioned in the introduction, the change from one type of party system to another came about as a result of two factors—electoral reform and the introduction of proportional representation in 1999, on the one hand, and the downgrading of HDZ’s electoral support and the loss of its predominant position, on the other. Several scholars argue that the Croatian party system after the 2000 parliamentary election could be described ‘as based on two multiparty blocs with relatively high potential for intra-bloc volatility (including abstentions) but relatively low inter-bloc volatility’ (Henjak et al. 2013: 456). Ever since the critical juncture in 2000, the party system has featured high levels of stability, with the main parties HDZ and SDP advancing vote-seeking strategies, trying to approach the median voter, while smaller centrist parties were employing office-seeking strategies and changing coalition camps from one election to the next, participating in centre-left and centre-right governments (Nikić Čakar 2019). However, in the last few years these patterns of stability were put under serious pressure with the electoral success of new challenger and populist parties, Most and Živi zid.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



645

There have been three sets of electoral rules introduced in Croatia since the introduction of democracy in 1990. In the wake of the first multiparty elections in 1990, the ruling League of Communists of Croatia (SKH) decided to introduce a French-style two-round electoral system, believing that it would give them leverage over opposing democratic parties, particularly over the HDZ, which was enjoying enormous public support. Contrary to their expectations, the HDZ won the elections and very soon realized that electoral rules could be used as a powerful weapon in building a predominant rule and so it engaged in electoral engineering, which eventually produced the mixed-member proportional system introduced in 1992 (Nikić Čakar 2019: 27). In 1999, the HDZ’s parliamentary majority once again changed electoral rules by introducing proportional representation, and this new set of rules has had a long-lasting impact on party competition and government formation. The proportional representation system was further upgraded with the introduction of preferential voting in the wake of the 2015 parliamentary elections. In the very short initial phase of party system development in Croatia, party competition was determined by the centre–periphery cleavage, manifested as a dilemma between Croatian independence and the preservation of the Yugoslav federation. This traditional cleavage was coupled with much more dominant ideological–cultural and historical–identity cleavages, which were structured around religious, historical, and ethnic issues (Čular and Gregurić 2007; Henjak et al. 2013). These frozen cleavages made the structure of party competition predominantly one-dimensional, with parties competing mostly over ideological and historical issues, with the socio-economic cleavage having a very marginalized influence (Zakošek 2002; Henjak 2018). However, it is important to stress that, when identifying median parties along those different policy dimensions, there is a high degree of overlap identified between the economic and cultural dimensions, with a general left–right policy divide (see Table 19.1a). Starting from these societal factors, in the first multiparty elections, held under the two-round majority electoral system, the HDZ won 69 per cent of seats in the new parliament, SKH won 25 per cent, while the Coalition of People’s Accord (KNS), as the third actor, suffered badly from the disproportional effects of the electoral system; it won more than 15 per cent of votes yet only 3.8 per cent of seats. When the Croatian war of independence started in 1991, the parties quickly came together around the programme of the government of national unity, but this consensus-oriented politics did not last long. The HDZ seized the moment to push forward a nationalist right-wing ideology and, after the early parliamentary election in 1992, established a predominant party rule, left unchallenged by a weak opposition (Kasapović 2001; Zakošek 2008). The prolonged state of war and occupied parts of national territory forced the parties to adjust their electoral strategies to meet such extraordinary circumstances. Opposition

Date in

Election date

Sanader II

Sanader III

14

HDZ

2008-01-13 2007-11-25 HDZ, HSS, HSLS, SDSS

2006-02-09

93.8

sur

mwc

min

min

min

sur

50.3

41.7

44.4

49.0

58.3

61.6 61.6 59.1 60.9

93.8

sur

maj maj maj sur

68.8 68.8 95.0

maj maj sur

Type Cabinet Party strength composition of of cabinet in seats cabinet (in %)

1990-05-31 1990-05-07 HDZ 1990-08-24 HDZ 1991-07-17 HDZ, SDP, SDSH, HSLS, HNS, HKDU, SSH, HDS Greguric II 1992-02-20 HDZ, SDP, SDSH, HNS, HKDU, HDS, SSH Grecuric III 1992-05-14 HDZ, SDP, SDSH, HKDU, HDS, SSH Sarinic 1992-08-12 1992-08-02 HDZ Valentic 1993-04-03 HDZ Matesa 1995-11-07 1995-10-29 HDZ Racan I 2000-01-27 2000-01-03 SDP, HSLS, HNS, HSS, IDS, LS Racan II 2001-06-04 SDP, HSLS, HNS, HSS, LS Racan III 2002-07-30 SDP, Libra, HNS, HSS, LS Sanader I 2003-12-23 2003-11-23 HDZ, DC

Mesic Manolic Greguric I

13

12

11

10

6 7 8 9

5

4

1 2 3

Cabinet Cabinet number

Table 19.1a Croatian cabinets since 1990

153

151

151

151

151

138 138 127 151

80

80

80 80 80

10

12

12

11

10

8 8 9 10

5

5

5 5 5

3.07

4.14

3.90

5.14

4.83

2.48 2.48 2.66 4.58

1.90

1.90

1.90 1.90 1.90

HDZ

HDZ

HDZ

HDZ

HDZ

HDZ HDZ HDZ HDZ

HDZ

HDZ

HDZ HDZ HDZ

HSLS, SDSS, HSU HSLS, SDSS, HSU HSU

Formal Median Number of Number of ENP, support parties in parliament party in seats in parties the first parliament parliament policy dimension

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Plenkovic II 2017-04-27

Plenkovic III

20

21

Technocrat prime minister

HDZ, HNS

HDZ min

min

mwc min min

min

mwc

40.4

37.1

51.0 43.0 45.7

49.0

50.3

151

151

151 151 151

153

153

12

12

11 14 12

10

10

4.41

4.50

3.85 4.73 4.45

3.07

3.07

HDZ

HDZ

SDP Most HDZ

HDZ

HDZ

Table legends and variables are further defined in the measurement and operationalization Appendix, this volume.

Cabinet types: maj = Single-party majority cabinet; min = Minority cabinet (both single-party and coalition cabinets); mwc = Minimal winning coalition.

The first policy dimension is economic left–right. The number of parties only includes parties that have held more than two seats in parliament when a cabinet has formed.

Median parties are based on data from Polk et al. (2017) and Bakker et al. (2015).

For a list of parties, consult the chapter appendix.

a

Notes:

Milanovic Oreskovica Plenkovic I

17 18 19

2017-06-09

Kosor II

16

HDZ, HSS, HSLS, SDSS 2010-07-10 HDZ, HSS, SDSS 2011-12-23 2011-12-04 SDP, HNS, IDS 2016-01-22 2015-11-23 NN, HDZ, Most 2016-10-19 2016-09-11 HDZ, Most

2009-07-06

Kosor I

15

BM365 SDSS, HSLS, HDS SDSS, HSLS, HDS SDSS, HSLS, HDS

HSU

HSU

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

648

´ ˇ 

parties were pushed to offer programmes in which they placed more emphasis on the issues of national independence, sovereignty, and the building of a homogeneous national community. Moreover, the ruling party introduced the mixed-member proportional electoral system before the 1992 parliamentary election, which was supposed to be an institutional guarantor of the HDZ’s prolonged electoral dominance (Kasapović 2014). Its subsequent reforms were only fine-tuning in order to maintain the ruling party’s electoral performance and to secure its absolute majority in parliament. As indicated in Table 19.1a, during the 1990s, the HDZ repeatedly confirmed its superior electoral performance, but other parties also managed to carve out electoral niches for themselves. This was particularly the case with the SDP, as it managed to consolidate and gradually gain increased voter support. By the end of the 1990s, when state-building was successfully completed, the HDZ’s predominance started to deteriorate, while Tuđman’s death in December 1999 left the HDZ with devastating intra-party conflicts and a leadership crisis (Nikić Čakar 2013; Čular and Nikić Čakar 2019). A broad centre-left coalition of opposition parties, with the backbone alliance formed between the SDP and the Social Liberals (HSLS), defeated the HDZ in January 2000 and won the parliamentary and presidential elections. These events were a key milestone that marked a change from the predominant party rule to the multiparty system of moderate pluralism. Some scholars described these elections as critical (Kasapović 2011) or even revolutionary (Bunce and Wolchik 2006), since the elections marked a radical shift in the way the centre-left coalition government conducted external and internal policies and in the elimination of the semipresidential system. After just one year in office, the centre-left government faced a strong rightwing challenge in the form of mass protests organized by war veteran associations and massively supported by the HDZ. These events posed a serious threat to the stability of the government, and soon deep ideological cracks emerged between the SDP and the HSLS over the question of national sovereignty and the Hague Tribunal. The final split came in March 2002 when the HSLS left the government and immediately splintered, while the progovernment ministers and members of parliament (MPs) formed the new Party of Liberal Democrats (Libra). The post-Tuđman HDZ also faced a series of splintering and intra-party disruptions. However, the party was ideologically repackaged in 2002 when the new, moderate leader, Ivo Sanader, crushed the extreme nationalist faction within the party. As a prize, the HDZ won the 2003 parliamentary elections and formed a coalition government with Democratic Centre (DC), while the government also enjoyed the parliamentary support of several smaller parties, including HSLS and the Independent Democratic Serb Party (SDSS). If the war and struggle for national independence were key drivers of party competition during the 1990s, then the Europeanization of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



649

Croatian politics was the most important factor in the consolidation period in the 2000s. Over the next few years the party system’s development was characterized by significant centripetal tendencies, which brought a gradual reduction in party system polarization (Henjak et al. 2013). The HDZ confirmed its electoral strength in the 2007 parliamentary elections by securing a second consecutive term in office. However, in the following years, the HDZ governments faced a serious legitimacy crisis, propelled not only by a deep economic recession but primarily by numerous corruption affairs. The highest-ranking government officials, ministers, and other prominent members of the HDZ were involved in large-scale corruption, including Sanader himself as the mastermind who planned the entire corruption scheme. In March 2011, the HDZ’s governing coalition faced a series of anti-government protests that grew out of an anti-establishment, populist movement. Later that year, the HDZ lost the election to the centre-left coalition led by the SDP, despite its last-minute electoral strategy directed towards mobilization of its core supporters. A long-lasting recession and economic downfall, coupled with omnipresent political corruption, left a deep imprint on voters’ behaviour, as their disaffection with political institutions and the quality of democracy in Croatia started to grow. The climax of structural and substantial changes in the Croatian party system came in 2015 and 2016 with ‘earthquake’ elections, when the new challenger parties, Most and Živi zid, secured a significant share of votes and seats. The Most could be fairly described as an anti-establishment reformist party, which attracted many disaffected voters with its anti-corruption programme focused on structural reforms in public administration, the judiciary, and the interior (Grbeša and Šalaj 2018). On the other hand, Živi zid was created by youth activists as a social movement at the height of anti-government protests in 2011 and, over the next few years, evolved into a populist, Eurosceptic party, advocating radical change to the monetary policy of the central bank (Nikić Čakar and Raos 2016). Parliamentary elections in 2015 and 2016 also brought more polarization and fragmentation in the party system, as indicated in Table 19.1a. After both elections, Most even took the leading role in the government formation process, negotiating coalition agreements with the HDZ. However, both coalition governments formed between the new, anti-establishment, party Most and the old, mainstream, party HDZ were underperforming in terms of decision-making and ended after several months. Despite this negative outcome, the electoral success of the new parties seriously undermined the long-lasting bipolarity and ideologically determined structure of party competition, as those who voted for Most and Živi zid voiced their protest against the old mainstream parties, based on non-ideological issues of party corruption and disaffection with the malfunctioning democracy (Henjak 2018).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

650

´ ˇ 

Electoral alliances and pre-electoral coalitions Electoral alliances and pre-electoral coalitions have a long tradition in Croatia, starting with the first multiparty election in 1990, when the KNS, an umbrella organization of loosely bonded moderate nationalist and liberal parties, was formed as the first pre-electoral coalition. During the 1990s, smaller opposition parties also formed electoral alliances, trying to improve their electoral fortunes.⁴ However, in the same period, the opposition remained quite fragmented, with the two leading opposition parties, SDP and HSLS, focused on building their organizational structures. By the end of the 1990s, opposition party leaders realized that only a united opposition could bring the HDZ’s dominant rule to an end. For that purpose, in the summer of 1999, they formed two electoral alliances, one between SDP and HSLS and the other between HSS (Croatian Peasant Party), HNS (Croatian People’s Party), IDS (Istrian Democratic Assembly), and LS (Liberal Party), with a joint pre-electoral coalition between all six parties dedicated to form the coalition government after the elections (see Table 19.1b). After the centre-left coalition fell apart and ended its term, only the smaller coalition partners IDS, LS, and Libra decided to form, once again, a pre-electoral coalition with the Social Democrats for the 2003 parliamentary election, while the larger parties opted to fight the election on their own (HSS) or in an electoral alliance with a new coalition partner (HSLS–DC). Electoral politics of the SDP in 2000s also witnessed the absence of pre-electoral coalitions, which was particularly evident in the 2007 election, as the Social Democrats, with new party leader Milanović, decided to fight the election alone. When the party failed to win enough seats to form a coalition government with its traditional left-wing and liberal partners, there were some critical voices, who blamed Milanović for not forming a pre-electoral coalition, as it was considered that they could have won the election in this way. They did not make the same mistake twice; so, prior to the 2011 election, the SDP formed a pre-electoral coalition with HNS, IDS, and HSU (Croatian Party of Pensioners), which was known as Kukuriku koalicija.⁵ This pattern of left-wing pre-electoral coalition formation was also repeated in the 2015 and 2016 elections, when the SDP made coalition agreements with the ‘usual suspects’, primarily HNS and HSU. The 2015 elections were somewhat specific as, for the first time, the HDZ also engaged in forming a pre-electoral coalition by making arrangements with HSS and HSLS, but also with some other smaller right-wing parties, which would not have achieved any significant success in the election otherwise. ⁴ This was particularly evident in the case of the Istrian Democratic Assembly (IDS), a regionalist party that championed the programme of decentralization and regional autonomy, against the prevailing centralizing tendencies of the HDZ. ⁵ The coalition got its name after a restaurant in Istria, where party leaders made initial arrangements on coalition formation.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



651

Table 19.1b Electoral alliances and pre-electoral coalitions in Croatia, 1990–2018 Election date

Constituent parties

Type

Types of pre-electoral commitment

1990-05-07

HKDU, HDS, HSLS, SDH, HSS IDS, RDS, DA

EA, PEC

HSS, IDS, HNS, HKDU, SBHS SDP, HSLS

EA, PEC EA, PEC

HSS, IDS, LS, HNS

EA, PEC

HSP, HKDU HSLS, DC

EA EA

SDP, IDS, Libra, LS

EA, PEC

2007-11-25

HSLS, HSS, PGS

EA

2011-12-04

SDP, HNS, IDS, HSU

EA, PEC

HDZ, HGS

EA

HDZ, DC

EA

HDZ, HSS, HSP AS, HSLS, BUZ, HRAST, HDS, ZDS SDP, HNS, HSU, HL-SR, AHSS, ZS SDP, HNS, HSS, HSU

EA, PEC

EA, PEC

HDZ, HSLS, HDS ŽZ, PH, AM

EA EA

Written contract, joint press conference, other* Joint press conference, separate declarations Written contract, joint press conference, other* Written contract, joint press conference, other* Written contract, joint press conference Joint press conference Written contract, joint press conference, other* Written contract, joint press conference, separate declarations Written contract, joint press conference, other* Written contract, joint press conference, other* Written contract, joint press conference Written contract, joint press conference Written contract, joint press conference, other* Written contract, joint press conference, other* Written contract, joint press conference, other* Separate declarations Joint press conference

1992-08-02 1995-10-29 2000-01-03

2003-11-23

2015-11-23

2016-09-11

EA

EA, PEC

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other * Joint electoral manifesto

Government formation The bargaining process At first sight, the government formation process in Croatia has been a rather simple task for those involved, particularly when taking into consideration the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

652

´ ˇ 

almost complete absence of inconclusive bargaining rounds, meaning that parties successfully end every bargaining process with a government being formed (see Table 19.2). While this comes as no surprise for the period of 1990s, when singleparty governments were the norm, without the need for the bargaining process, it is interesting to observe a similar pattern in the period after 2000. The only exception to this informal rule happened after the parliamentary elections in 2015, when the newly created party Most, as the kingmaker in the government formation process, initiated simultaneous bargaining rounds with both left-wing and right-wing pre-electoral coalitions, with the ultimate goal of making a tripartite grand coalition consisting of HDZ, Most, and SDP. In the end, the Most leadership broke off negotiations with the left-wing coalition and finalized the bargaining process with the HDZ-led coalition, selecting independent technocrat Orešković as formateur. As seen in Table 19.2, the formation of the Orešković cabinet, compared to other coalition negotiations, was the longest bargaining process, as it was well above the average of 27 days for all coalition cabinets since 2000. When taking into consideration the entire government formation period, we can also see that there are considerable variations among the coalition cabinets, ranging from 19 days (Milanović) to 125 days (Plenković I). When observing these data, it could also be tentatively concluded that the existence of pre-electoral coalition agreements significantly shortens the time needed for negotiating the formation of a government. It takes on average almost 14 days to form a government when there is a pre-electoral agreement (Račan I and Milanović), while in the case of ‘freestyle bargaining’ after the election, it takes on average almost 38 days for a coalition to be negotiated (Sanader I, Sanader III, Orešković, and Plenković I). The formal coalition negotiations usually start a couple of days after the election, particularly in cases, though not exclusively, when there is no preelectoral coalition agreement, with party leaders of the prospective coalition taking the leading role. The initiative for opening the bargaining process comes from the leader of the largest party most likely to form the government. Up to now, it was always relatively easy for leaders of the HDZ and SDP to negotiate coalition agreements, as the smaller parties were primarily focused on gaining offices, leaving the policy aspects of coalition agreements rather marginalized. The behaviour of the alliance of HSS–HSLS in 2008 and Most after the 2015 and 2016 elections was somewhat of an exception to this pattern, as during the coalition bargaining, they first insisted on policy issues being negotiated and later in the negotiation process demanded control of specific ministries. Although coalition bargaining is a highly elitist process, with the most important conflicting issues being resolved between party leaders, it also usually involves party working groups, which negotiate specific policy areas. These include prominent party members and potential cabinet ministers, but are almost always controlled by the party leadership.

Year in

1990 1990 1991

1992

1992

1992 1993 1995 2000 2001 2002 2003 2006 2008

Cabinet

Mesic Manolic Greguric I

Greguric II

Greguric III

Sarinic Valentic Matesa Racan I Racan II Racan III Sanader I Sanader II Sanader III

0 0 0 0 0 0 0 0 0

0

0

0 0 0

0

0

0 0 0

Bargaining duration of individual formation attempt (in days)

0 0 0 SDP, HSLS, HSS, HNS, IDS, LS 23 SDP, HSLS, HNS, HSS, LS 0 SDP, Libra, HNS, HSS, LS 0 HDZ, DC 27 0 HDZ, HSLS, SDSS, HSS 36

HDZ, SDP, SDSH, HSLS, HNS, HKDU, HDS, SSH HDZ, SDP, SDSH, HNS, HKDU, HDS, SSH HDZ, SDP, SDSH, HKDU, HDS, SSH

Number of Parties involved in the inconclusive previous bargaining rounds bargaining rounds

Table 19.2 Cabinet formation in Croatia, 1990–2018

10 5 9 24 0 25 30 0 49

0

0

0 0 0

0 0 0 23 0 0 27 0 36

0

0

0 0 0

Total Number of days required bargaining in government duration formation

Continued

62

82

0

47 29

84 1 88 14

* * *

7 19 5 1

* * *

97 0 83 15 77 39 122 1

* * *

Pro Abstention Contra

Result of investiture vote

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

0 0 0

2009 2010 2011 2016

Kosor I Kosor II Milanovic Oreskovic

Plenkovic I 2016 Plenkovic II 2017 Plenkovic III 2017

HDZ, HSS, HSLS, SDSS 0 HDZ, HSS, SDSS 0 SDP, HNS, IDS, HSU 4 NN, HDZ, Most 64 (1) SDP, Most, HNS, HSU, HL- 35 SR HDZ, Most 24 0 HDZ, HNS 1

Bargaining duration of individual formation attempt (in days)

125 0 0

5 0 19 75 24 0 1

0 0 0 64

Total Number of days required bargaining in government duration formation

3 0

78

46

45

28 61

89 12 83 5 91

45

0

83

Pro Abstention Contra

Result of investiture vote

* Votes were not counted, as the president of the parliament, according to the number of the hands raised, just assessed that there were enough votes.

NN = Independent

Notes:

0 0 0 1

Year in

Cabinet

Number of Parties involved in the inconclusive previous bargaining rounds bargaining rounds

Table 19.2 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



655

The composition and size of cabinets According to the description of the Croatian party system previously outlined, it is somewhat to be expected to observe two contrasting periods when assessing cabinet types and cabinet composition. As a direct outcome of the HDZ’s electoral superiority during the 1990s, all the governments formed under President Tuđman were single-party governments, which enjoyed rather comfortable parliamentary majorities. The only exceptions to the single-party rule were the shortlived Gregurić I-III governments, which were composed of all relevant political parties at that time, working as some sort of all-inclusive grand coalition. The turn of the millennium witnessed a comprehensive constitutional and electoral reconstruction that resulted in coalition politics and minority situations in parliament as the norm. The effects of these changes were primarily visible in the composition of Račan’s cabinets, who started his prime ministership by leading the surplus sixmember coalition cabinet (Račan I) and ended his term with a minority coalition cabinet (Račan III), as a consequence of strong personal and policy conflicts between SDP and HSLS in his second cabinet. Interestingly, Table 19.1a reports that, since the turn of the millennium, there has been a strong tendency towards minority governments, with special external support arrangements being made in order for minority governments to secure the backing of parliamentary majorities. This became a common pattern of the government formation process, starting with the Sanader I government, when the HDZ failed to negotiate a coalition agreement with HSLS and HSS while forming the governing coalition only with the rather small and uninfluential DC. However, Sanader managed to secure formal support arrangements with HSLS, SDSS, and HSU, as those parties decided to support the governing coalition on a contractual basis without being represented in the cabinet, while the HSS party leader publicly announced that his party would support the governing coalition in the investiture vote and would continue to do so during the term with no strings attached. These contractual and voluntary support arrangements have become commonplace in the government formation process, even in cases when there was no real need for such support, as in the Sanader III, Kosor I, and Milanović governments, as those cabinets were already minimal winning coalitions. Despite this, smaller parties, such as HSU and SDSS, which advocate the particular interests of pensioners and the Serbian ethnic minority in Croatia respectively, decided to join with parliamentary majorities to secure special concessions for their electorate. The number of ministries has varied from one cabinet to another, with no linear trends either in increasing or reducing the number. It is interesting to note, however, that, after the breakpoint in 2000, prime ministers have preferred somewhat larger cabinets, with the exception of Prime Minister Sanader (and later his successor Kosor) who insisted on a smaller cabinet, in line with the government’s proclaimed policy of administrative efficiency. During the 1990s, in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

656

´ ˇ 

comparison to the period after 2000, there were many more cabinet members who did not hold any portfolio, no matter that they were ministers or deputy prime ministers. Furthermore, the fluctuation of cabinet members was on a much larger scale during the 1990s than afterwards, providing more evidence of how different institutional settings (semi-presidentialism vs parliamentarism) affect governmental processes (see Table 19.3).

The allocation of ministerial portfolios The allocation of ministerial portfolios in Croatia seems to follow similar trends to those in other European democracies, particularly in terms of overcompensation of junior coalition partners (Warwick and Druckman 2006). As they are primarily employing office-seeking negotiation strategies, smaller parties usually tend to get the maximum benefit of their bargaining power, in terms of both the targeted number of portfolios and the preferences for particular portfolios. For instance, the most overcompensated party was DC in 2003, as its one parliamentary seat (0.66 per cent of total parliamentary seats) was awarded with the Ministry of Justice (6.67 per cent of portfolios in the Sanader I cabinet), as this ministry suited the professional profile of the DC party leader very well. Another striking example of an overcompensated party is Most in 2015, since it managed to negotiate seven cabinet portfolios (33.33 per cent) based on less than 10 per cent of parliamentary seats (see Table 19.3). The bargaining over portfolios represents one of the central stages in the government formation process, so it is important to assess different policy or office motives behind coalition parties’ preferences for a specific cabinet portfolio (Manow and Zorn 2004; Bäck et al. 2011). Cabinet portfolios in Croatia have different status roles in the government formation process. As a rule of thumb, the classical state ministries of finance, interior, administration, foreign affairs, justice, and defence, including prime minister, are usually in the control of the largest coalition party. This is illustrated in Table 19.3. On the other hand, the portfolio preferences of smaller coalition parties can be explained with both policy-seeking and office-seeking behaviour. For instance, the HSS usually has strong preferences for the agriculture portfolio, which is in line with the party’s agrarian profile. In a similar vein, interviewees stressed that, in the 2015 and 2016 coalition negotiations, Most insisted on taking over the ministries according to its policy profile of anti-corruption and reformist party, primarily justice, interior, public administration, and economy. When assessing office-seeking strategies, it is interesting to observe that junior coalition parties generally ask for portfolios that perfectly fit the professional profile or personal ambitions of their leading figures. Evidence extracted from the interviews suggests that this kind of behaviour can be identified in the bargaining processes for the Račan I and III, Sanader I, and Milanović

1991

1992

Greguric I

Greguric II

2000

2001

2002

2003 2008

2009

2010

2011 2016 2016 2017

Racan I

Racan II

Racan III

Sanader I Sanader III

Kosor I

Kosor II

Milanovic Oreskovic Plenkovic I Plenkovic III

Greguric III 1992

Year in

Cabinet

16 SDP, 4 HNS, 1 IDS 1 Ind., 15 HDZ, 7 Most 17 HDZ, 4 Most 19 HDZ, 2 HNS

12 SDP, 5 HSLS, 3 HSS, 2 HNS, 1 LS 15 SDP, 4 HSS, 2 HNS, 2 Libra, 1 LS 14 HDZ, 1 DC 14 HDZ, 2 HSS, 1 HSLS, 1 SDSS 16 HDZ, 2 HSS, 1 HSLS, 1 SDSS 15 HDZ, 2 HSS, 1 SDSS

17 HDZ, 2 SDSH, 2 SDP, 1 HDS, 1 HKDU, 1 HNS, 1 HSLS, 1 SSH, 1 Ind. 22 HDZ, 2 SDP, 1 HDS, 1 HKDU, 1 HNS, 1 SDSH, 1 SSH, 1 Ind. 26 HDZ, 2 SDP, 1 HDS, 1 HKDU, 1 SDSH, 1 SSH, 1 Ind. 11 SDP, 5 HSLS, 3 HSS, 2 HNS, 1 IDS, 1 LS

22 23 21 21

18

20

15 18

24

23

23

5 HDZ, 3 Most

2 HDZ, 1 HSS, 1 SDSS 2 HDZ, 1 HSS, 1 SDSS 21 21 20 21

17

17

8 SDP, 2 HNS, 2 20 HSLS, 1 HSS, 1 IDS, 2 Ind. 8 SDP, 2 HSLS, 1 20 HNS, 1 HSS, 1 Ind. 8 SDP, 2, Libra, 1 20 HNS, 1 HSS, 1 Ind. 15 2 HDZ 16

21

21

30 33

17

27

Number of Number of Number of ministers per Total ministries watchdog junior number party (in descending ministers per party of order) ministers

Table 19.3 Distribution of cabinet ministerships in Croatian coalitions, 1991–2018

SDP Ind. HDZ HDZ

HDZ

HDZ

HDZ HDZ

SDP

SDP

SDP

HDZ

HDZ

HDZ

1 Prime minister

SDP HDZ HDZ HDZ

HDZ

HDZ

HDZ HDZ

SDP

SDP

SDP

HDZ

HDZ

HDZ

HNS HDZ HDZ HDZ

HDZ

HDZ

HDZ HDZ

SDP

SDP

SDP

HDZ

HDZ

HDZ

2 3 Finance Foreign affairs

SDP HDZ HDZ HDZ

SDP

SDP

SDP

HDZ

HDZ

HDZ

SDP Most Most HDZ

HDZ

HDZ

DC HDZ

SDP

SDP

SDP

HDZ

HNS

HNS

5 4 Labour and Justice equality affairs

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

658

´ ˇ 

cabinets, with the HNS as a prime example of an office-seeking party. More recently, however, the HNS has increasingly focused its efforts on education as its main policy issue and nominated an independent for this portfolio in the Plenković III cabinet.

Coalition agreements At first sight, the analysis of coalition agreements in Croatia reveals that they represent a constituent part of the government formation process, as they have appeared in every coalition government so far, usually featuring the policy dimension of negotiated coalitions (see Table 19.4). Nevertheless, when taking into consideration the size and the content of the agreements, we can see significant within-country variations, with no clear linear trends observed. Contrary to evidence from other European democracies, Table 19.4 shows that coalition agreements in Croatia are not growing longer and more detailed with every new election. They tend to be rather short in general, with the shortest one signed in 1991 when forming the Gregurić I cabinet (340 words). On the other hand, the longest coalition agreement to date is the one signed between HDZ, HSS, and HSLS in 2008, at almost 27,000 words, covering almost all the issues of coalition politics inside and outside the cabinet.⁶ It is also interesting to mention two specificities related to coalition agreements in Croatia. The first one is the case of the somewhat longer agreement of the Plenković I cabinet in 2016, primarily because it was not a customarily negotiated and signed agreement but actually the government programme presented in parliament at the investiture vote.⁷ The second observation relates to the Račan and Orešković cabinets, as Table 19.4 reports that, in those cases, there were two documents that constituted the outcome of the inter-party bargaining process. In both cases, the most authoritative document was the coalition agreement negotiated and signed by party leaders, while the second document was the policy agreement, in the form of either the government programme (Račan I) or the negotiated policy paper (Orešković). On both occasions, these documents were explicitly mentioned in the original coalition agreements as supplements. Content analysis of the coalition agreements shows that coalition parties are putting much more emphasis on government policies. Even in the cases of the

⁶ This agreement was also supplemented with the annex, which regulates the distribution of cabinet portfolios and the patronage over jobs in the public sector, but unfortunately it left out of author’s reach. ⁷ Interviewees pointed out that the coalition parties HDZ and Most did not sign a standardized agreement, as they came to the conclusion that the negotiated policies would be elaborated in the government programme anyway, while other issues, such as portfolio allocations were discussed and negotiated only in the form of verbal agreements.

1991 1992 1992 2000

2001

2002

2008 2009 2010 2011 2016

2016 2017

Greguric I Greguric II Greguric III Racan I

Racan II

Racan III

Sanader III Kosor I Kosor II Milanovic Oreskovic

Plenkovic I Plenkovic III

340 340 340 1,235 12,714* 1,235 12,714* 1,235 12,714* 26,999 26,999 26,999 674 2,345 43,751* 11,623* 828

Size 0 0 0 53.2 0 53.2 0 53.2 0 3.7 3.7 3.7 100 31.2 0 0 0

General rules (in %) 0 0 0 25.7 0 25.7 0 25.7 0 0 0 0 0 0 0 0 0

Policy-specific procedural rules (in %)

Note: *Government programme presented to the parliament prior to the investiture vote.

Year in

Coalition

Table 19.4 Size and content of coalition agreements in Croatia, 1991–2018

0 0 0 0 0 0 0 0 0 0.9 0.9 0.9 0 25.8 0 0 0

Distribution of offices (in %) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Distribution of competences (in %)

100 100 100 0 100 0 100 0 100 91.5 91.5 91.5 0 38.2 100 100 100

Policies (in %) OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

660

´ ˇ 

Račan I and Orešković cabinets, where the focus of coalition agreements was more directed towards general and policy-specific procedural rules, there were also supplemental documents devoted entirely to policy issues. The only exception to this rule is the pre-electoral coalition agreement of Milanović’s four-member coalition government, as its content is completely focused on general rules related to candidacy lists. Since the coalition agreements are mostly designed as government policy programmes, they do not say almost anything about how coalition governments will resolve potential internal disputes and conflicts. Even in the cases where conflict resolution mechanisms are specifically outlined in the coalition agreements, such as in the Račan I, Sanader III, and Orešković governments, coalition governance practice, at least according to the interview evidence, shows that the rules of the game are usually rather informal and dependent on the personal understandings of party leaders and their behind-the-scenes wheeling and dealing.

Coalition governance The role of individual ministers in policy-making Ministerial autonomy in Croatian cabinets is significantly constrained by several factors, primarily by the dominant role of their parties, prime ministerial powers, and the lack of constitutional and legal provisions of their roles and authority within the government. Starting with the latter, the Croatian constitution and the Law on the Government are quite silent on ministerial responsibilities and authority. The constitution only mentions ministers as being members of the government who are collectively and individually accountable to parliament, but it does not say anything about their institutionalized scope of work. In a similar vein, the Law on the Government is no more specific about the issue of ministerial autonomy than the constitution, giving no directions for the role of individual ministers in the policy-making process. However, both documents provide institutional frameworks for a prime minister to impose their executive leadership, having very clear intentions in defining the Croatian version of the ‘chancellor principle’ (Niclauss 2000; Helms 2005). As previously mentioned, the prime minister enjoys prerogatives in appointing and dismissing ministers and these powers give him/her a great deal of leverage over other cabinet members, since ministers’ political careers are dependent on the prime minister’s will. Besides this, the prime minister has the power of giving directions to the ministers, making them personally accountable for delivering special assignments. When those formal rules are taken into consideration within the reality of coalition governance, then we can notice how patterns of behaviour vary across governments. Interviewees pointed out that there is a trade-off between a prime

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



661

minister’s dominance and ministerial autonomy. On rare occasions when prime ministers are weak and unable to exercise their powers, then the role of individual ministers is much greater and, vice versa, when strong prime ministers impose their will on the cabinet, then ministerial autonomy is significantly limited. As an illustration, interviewees identified Račan, Kosor, and Orešković as being, for various reasons, rather weak prime ministers, as they allowed significant ministerial autonomy in defining priorities and directing policy-making within individual ministries. This was especially true in the case of Račan’s cabinets, where ministers were behaving like policy dictators within their portfolios (Laver and Shepsle 1996) while, at the same time, Račan tolerated ministerial disobedience and intra-cabinet disunity. On the other hand, the interview evidence also indicates the opposite examples of Sanader, Milanović, and Plenković, who are perceived by interviewees as rather strong prime ministers, which significantly contributes to the notion of prime ministerial government (Dunleavy and Rhodes 1990) as the dominant model of governance in Croatia. These prime ministers were able to make decisions across all policy areas, wanted to be involved in (or at least informed about) the policy-making process of every cabinet portfolio, and served as the ultimate authority for resolving conflicts within the cabinet and between coalition parties. Coalition parties involved in the bargaining process over cabinet formation can set limitations not only on the autonomy of their own ministers but also on the prime minister’s manoeuvring space. Coalition parties expect their ministers to follow party policy guidelines and to implement party programmes within their portfolios. Moreover, contrary to the formal prime ministerial powers in appointing and dismissing cabinet members, interviewees suggest that coalition parties are completely autonomous in selecting ministerial candidates, making prime ministers somewhat more constrained in that sense. Essentially all interviewees emphasize this characteristic of the coalition bargaining process, except for the somewhat peculiar case of the Orešković government.⁸ When it comes to the question of dismissing ministers, prime ministers are usually not inclined to make that move, because it usually endangers the coalition’s stability. However, on two occasions, in the Sanader III and Plenković I cabinets, prime ministers unilaterally decided to dismiss ministers; this decision in the case of Plenković I also brought the coalition government to an end.

⁸ In line with his previous experience in the business firm employment practices, Orešković organized interviews with the parties’ ministerial candidates, as there were usually several people nominated for each ministry, making the final choice based on his business instincts. However, in reality, coalition parties HDZ and Most were actually in control of the process, as they managed to get their preferred candidates appointed as ministers.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

662

´ ˇ 

Coalition governance in the executive arena Coordination mechanisms of the coalition governance in the executive arena mostly work through the cabinet committees and the inner cabinet. However, it should be noted at this point that the inner cabinet and cabinet committees in Croatia are not conflict management mechanisms specifically designed by coalition parties to coordinate their mutual relations within a specific coalition. Instead, they are formal instruments, legally imposed on every government as an intra-cabinet coordination mechanism, which means that their structures and functional roles are prescribed by the Law on the Government and the government’s Rules of Procedure. For instance, Article 16 of the Law sets the functions of the inner cabinet in the scope of ‘proposing the implementation of the government’s policy, monitoring the implementation of the government’s programme, discussing certain matters within the jurisdiction of the government and coordinating the work of the members of the government in performing their assigned tasks’ (emphasis added). Table 19.5 shows that the inner cabinet, together with the party summits, is used as the most common conflict management mechanism, since almost all interviewees identified it as the focal point in intra-cabinet decision-making. The first step in the process of coalition coordination over policies takes place in the cabinet committees, which are policy-specific and permanent coordinating bodies consisting of ministers, junior ministers, and senior civil servants. At this point, policy proposals are broadly discussed in larger groups, narrowing the initial differences between coalition parties. The next step takes place in the inner cabinet, formally consisting of the prime minister and deputy prime ministers, with other ministers and coalition party leaders who are outside the government only occasionally invited to the meetings by the prime minister. The inner cabinet works as the ultimate filter, with all major differences over policy negotiated and compromised between coalition partners. This is also the final instance prior to the cabinet meeting where consensus is reached over policy proposals; otherwise they are rejected or returned for reconsideration. A cabinet meeting is the final stage of the coalition governance in the executive arena. All interviewees agree that inter-party disagreements are almost always excluded from the meetings, given that they commonly serve the purpose of projecting symbolic unity and harmony between coalition parties. Although the formal voting takes place in the cabinet meeting on every decision, with the majority of all cabinet members required, interviews reveal that prime ministers strongly insist on unanimity. There are only a handful of open disputes and conflicts between ministers taking place in cabinet meetings, with the most common during the Račan I–III cabinets and the most notorious one in the Plenković I cabinet. That case recently raised a lot of controversy over the prime minister’s power to dismiss cabinet members, as Prime Minister Plenković made

2001

2002

Racan II

Racan III

2009 POST

2010 POST

Kosor II

IE IE IE PRE, POST PRE, POST PRE, POST POST POST

Kosor I

Sanader I 2003 Sanader III 2008

1991 1992 1992 2000

Y

Y

N Y

N

N

Y Y Y N

N

N

N N

N

N

N N N N

IC IC IC IC, CaC, PS, Parl IC, CaC, PS, Parl IC, CaC, PS, Parl IC* IC, CaC, CoC, PS IC, CaC, CoC, PS IC, CaC, CoC, PS

All used

IC* IC, CaC, PS IC, CaC, PS IC, CaC, PS

IC, PS

IC, PS

IC IC IC IC, PS

PS

PS

IC* PS

PS

PS

IC IC IC PS

Most For common most serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Greguric I Greguric II Greguric III Racan I

Coalition

N

N

Y N

N

N

N N N N

N

N

N N

N

N

N N N N

Personal Issues union excluded from agenda

Table 19.5 Coalition governance mechanisms in Croatian coalition governments, 1991–2018

All/All

All/All

All/All All/All

All/All

All/All

No/No No/No No/No All/All

Y

Y

Y Y

Y

Y

Y Y Y Y

Comp.

Comp.

No Comp.

Comp.

Comp.

Few Few Few Comp.

Y

Y

N Y

N

N

N N N N

Continued

Y

Y

N Y

N

N

N N N N

Junior NonCoalition Freedom of Policy discipline appointment agreement ministers cabinet (watchdogs) positions in legislation/ other parl. behaviour OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Y

Plenkovic III

Y

N

N

N

IC, CaC, PS, Pca IC, CaC, PS, CoC IC, CaC, PS IC, CoC PS

IC, PS

CoC

CoC

IC, CaC PS

PS

IC

Most For common most serious conflicts

N

N

Y

N

N

N

N

N

Personal Issues union excluded from agenda

All/All

No/No

No/No

All/All

Varied

Varied

Comp.

Comp.

N

Y

N

N

Policy agreement: Few = On few selected policies; Varied = On a variety of issues, but not comprehensive; Comp. = Comprehensive policy agreement

Coalition discipline: All = Coalition discipline is always expected; No = Coalition discipline not expected

Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee; CoC = Coalition committee; Parl = Parliamentary leaders; PS = Party summit

Y

Y

N

Y

N

Y

N

N

Junior NonCoalition Freedom of Policy discipline appointment agreement ministers cabinet (watchdogs) positions in legislation/ other parl. behaviour

* Corresponding to an informal inner cabinet, not the legally defined inner cabinet Coalition agreement: PRE = pre-election; Pre, Post = pre- and post-election; IE = inter-election; POST = post-election

Notes:

N

Plenkovic I 2016 POST

2017 IE

Y

2016 POST

Oreskovic

N

2011 PRE

All used

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Milanovic

Coalition

Table 19.5 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



665

the decision on his own, by referring to the constitution and the Law on the Government, which are not that explicit in defining the procedure and the prerogatives. Junior ministers in Croatian governments have a low political profile and very little significance for cabinet decision-making. Since the appointment of junior ministers is almost exclusively under the control of each coalition party, they prefer to use these positions to reward or to promote less prominent party officials. Ministers are primarily in charge of determining the scope of work of junior ministers, and their de facto status within the ministry is dependent on their personal relations with the minister. Although watchdog junior ministers are common in Croatian coalition governments (Račan I–III, Sanader III, Kosor I and II, Orešković), according to the interviewees, they are in most cases marginalized and almost without any influence, unable to perform the task of overview of ministers from other coalition parties. Table 19.5 indirectly confirms these observations, by making it clear that coalition parties during the bargaining phase pay almost no interest to making arrangements on the distribution of junior ministers.

Governance mechanisms in the parliamentary arena When coalition parties negotiate over government formation, there is almost always some kind of explicit or implicit understanding that there will be a coalition discipline in parliamentary votes on legislative proposals. Table 19.5 shows that all coalition parties in Croatia, except for the short-lived Gregurić I–III governments, which were specific in themselves, had agreed, at least in principle, that they would maintain discipline among their members of the Sabor. However, notwithstanding these negotiated rules, the interviews show that some coalition governments had a very hard time keeping their MPs in line with agreed coalition policy. Parliamentary rebellions and disobedience were especially frequent during Račan’s coalition governments, which can be partially explained by the spillover effect of the disrupted relations between coalition partners in the government. Interestingly, the six-member coalition initially included in the coalition agreement the rules on establishing a coordination mechanism between coalition parties in parliament (see Table 19.5). This coalition mechanism was supposed to coordinate the work of the coalition party groups in the Sabor, with leaders of parliamentary party groups as the main actors. However, interviewees suggest that, despite these formally established rules, coalition governance in the parliamentary arena was seriously undermined by parliamentarians who were behaving rather badly. In an attempt to establish the balance between executive and legislature power, members of the parliamentary majority were obstructing

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

666

´ ˇ 

parliamentary procedures by heavily amending proposed bills coming from their own ministers. The largest coalition party, SDP, was particularly burdened by the lack of discipline among its parliamentary representatives. Interviewees argue that personal frustrations and unfulfilled ambitions were behind this behaviour, as some of the MPs were left out when ministries were allocated in the government formation process. The parliamentary committee system does not play an important role in the coalition governance in the parliamentary arena, as none of the interviewed former ministers even mentioned that committee chairs can shadow cabinet ministers or exercise any other controlling mechanism over coalition cabinet members. On the other hand, the leaders of parliamentary party groups are identified as central figures in the parliamentary coordination mechanisms, since they are usually in charge of maintaining discipline and building cohesion within the parliamentary majority. For instance, the role of parliamentary leaders was particularly pronounced in the cases of the Sanader III (and consequently Kosor I and II) and Milanović cabinets, given that they served as a coordinating link between the government and the parliament. According to the coordination rules set up between coalition parties in these governments, ministers had to present legislative proposals either at a joint meeting of all members of the parliamentary majority or at individual meetings with each coalition parliamentary group. Parliamentarians were, in that way, directly informed about the controversial aspects of a proposed bill by the minister in charge of a specific policy portfolio and had an opportunity to resolve all contesting issues before the bill was put in front of the whole parliament for a vote.

Governance mechanisms with different types of actors (mixed) Given that the conflict resolution mechanisms in the executive arena and the inner cabinet are primarily enforced by law, it comes as no surprise that coalition committees, as specifically designed conflict management mechanisms, are somewhat rare. Table 19.5 shows that coalition committees were used as coordination and conflict resolution mechanisms in the Sanader III, Kosor I and II, Orešković, and Plenković III cabinets but, in all those cases, except for the latter, they played a very limited role in coalition decision-making. Coalition parties in Croatia are not usually inclined to design this type of conflict management mechanisms, partially due to the fact that they are negatively perceived in the public as some sort of parapolitical decision-making bodies, outside the conventional political arenas and out of the public eye. A coalition committee in the Sanader III cabinet was set up for the purposes of coordinating and harmonizing intra-coalition relations between HDZ, HSS, and HSLS, and it was the first coalition mechanism of this kind in Croatian politics.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



667

According to the interviewees, this was typically a permanent committee with relatively stable membership, made on the basis of similar coalition committees in West European countries. It consisted of a mixture of cabinet members, parliamentarians, and party leaders, with 10 to 15 members attending the committee meetings. One deputy prime minister from the HDZ acted as the trustee of the coalition agreement and had the task of chief coordinator in charge of convening meetings. Although the committee was regularly convened and members discussed various aspects of coalition governance, it was uninfluential and unable to deliver on its purpose of framing coalition policy and handling contested issues between coalition parties. Interviewees also suggest that the committee was completely overshadowed by the party summit as another conflict resolution mechanism, with a handful of leaders of coalition parties as the main actors. A similar thing also happened with the coalition committee set up to improve coordination in the Orešković coalition government. Formally entitled Council for Cooperation, it consisted of up to 14 members (seven from each party), working as a mix of parliamentary party leaders and less prominent party people. The council was supposed to oversee coalition coordination and make appointments of junior ministers, but in the end, it was too weak and marginalized to make any difference within a coalition structure that was characterized by personal animosities and constant clashes between two deputy prime ministers as the leading figures. On the other hand, the Plenković III coalition government is a somewhat better example of a functional coalition committee. It consisted of party leaders (including inner party leaderships) and parliamentary leaders, meeting regularly and has been rather successful in resolving all conflicting issues coming from inside or outside the coalition government. At this point it is important to address the issue of party summits as the most attractive conflict management mechanisms, since they are identified by interviewees as the most important external coordination mechanism in Croatian coalition governance. Table 19.5 indicates that party summits are used to manage the most serious conflicts that arise between coalition parties but, at the same time, are also regularly engaged as conflict resolution bodies in everyday coalition decision-making. When assessing the structure and the nature of these conflict management bodies, it should be stressed that they bring together only the top leaders of the coalition parties and can be set up either as permanent or ad hoc mechanisms for resolving intra-coalition disputes. By their nature, party summits are mostly characterized by a significant level of informality and personalization in making decisions, with very limited possibilities of imposing public scrutiny or intra-party controlling mechanisms upon the party leaders and their autonomy in framing the coalition governance. There is a general consensus among interviewees that almost all crucial aspects of coalition governance in Croatia are determined in the narrow circle of coalition party leaders. This pattern of coalition behaviour was first established in the Račan

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

668

´ ˇ 

I coalition government, when six party leaders regularly met to discuss and resolve all contested issues. However, according to the interviews, this governance arena functioned without any established rules of conduct and mostly relied on personal relations between party leaders, which was also the main source of terminal disruptions between them. Personal and informal ways of managing inter-party coalition relations were raised to a higher level of technical sophistication in the Sanader I and III cabinets, but also in the Milanović cabinet. In these cases, prime ministers preferred to handle coalition affairs in bilateral informal meetings with other coalition party leaders, usually discussing coalition politics and making initial policy decisions over private lunches or in some other unconventional context. Similar patterns of informality were also repeated in the Orešković and Plenković I governments. In the former case, two party leaders of the HDZ and Most, who were also acting as deputy prime ministers, were resolving all conflicting issues in bilateral, behind-the-scenes meetings, while in the latter case, the game was played very much in the same way, only with the two leaders playing different roles, this time as the prime minister and the Speaker of the parliament. Finally, coalition parties do not invest much effort in detailing conflict management mechanisms in coalition agreements. There are only three agreements that included specific conflict management mechanisms; the Račan I, Sanader III, and Orešković governments. In the first two cases, party summits were formally set up in coalition agreements as the most dominant instrument of conflict resolution between coalition parties. In the case of the short-lived Orešković coalition government, the coalition agreement negotiated between the HDZ and Most included the coalition committee as the ideal mechanism. However, as mentioned earlier, during the coalition life cycle this mechanism proved to be ineffective and was completely replaced with bilateral negotiations between two party leaders.

Cabinet duration and termination The duration of cabinets Coalition cabinets in Croatia can be evaluated not only in terms of the quality of the coalition governance and the way political parties handle their intra-coalition affairs but also with the somewhat ‘objective’ criteria of cabinet duration and the overall survival rate of prime ministers. When taking these two criteria into consideration, we notice a couple of more or less contradictory trends that need to be discussed, that is the relatively short cabinet duration and the very high survival rate of prime ministers. The average duration of all Croatian cabinets since 1990 is rather short. The average lifetime of Croatian cabinets is approximately 475 days, which is around

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



669

40 per cent of their average maximum possible duration. Duration records of individual cabinets show even more diversity, with the shortest living cabinet being that of Plenković II, which lasted for only 43 days. However, we should bear in mind that Plenković II served only as an interim minority cabinet until the moment the new coalition government was negotiated between HDZ and HNS. The longest serving cabinets were the HDZ majority government under Prime Minister Mateša in the mid-1990s and the left-wing coalition government led by Prime Minister Milanović, as both governments endured for almost full constitutional terms. These data point to a certain level of instability of Croatian governments, with the notion that there is no significant difference in the cabinet duration between the period of the 1990s, when single-party governments were the norm, and the period after 2000, when only coalition governments were formed. The main factor explaining cabinet instability in the 1990s is the behaviour of the president, as the fates of prime ministers and cabinets were entirely in his hands. Frequent changes of cabinets after constitutional reform in 2000 can be explained by the behaviour of coalition parties and their leaders, since coalition governments were generally too weak and, without established mechanisms, left at the whim of party leaders and their fragile personal relations. Although the turnover rate of coalition cabinets has been relatively high, as indicated by the data on cabinet durability, prime ministers have been much more successful in surviving this unstable coalition context. It is interesting to note that Prime Ministers Račan, Sanader, and Plenković managed to survive in office for the almost maximum possible duration of their first cabinets, although their initially negotiated coalition arrangements did not last that long. Of these three, Račan had the best chance to survive as prime minister for the whole parliamentary term, as he started his prime ministership with a surplus majority coalition and could afford to lose some of the coalition partners along the way, and even outlive a dramatic split with the largest coalition partner, HSLS, in 2002. On the other hand, Sanader and Plenković were in an unenviable situation right from the start, as they stepped into the prime minister’s office in 2003 and 2016 respectively, by leading minority governments that enjoyed a very thin parliamentary support of several smaller parties, including parliamentary representatives of national minorities in Croatia. Despite the confines of coalition dynamics, they were able to endure the pressure and hold on to the prime ministerial position, mostly due to their personal leadership capacities and the inter-party dealmaking.

The termination of cabinets While the above-mentioned evidence points to a certain level of instability of coalition governments in Croatia, it is also important to dig deeper into the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

670

´ ˇ 

reasons for cabinet terminations. Table 19.6 shows that the HDZ’s single-party governments in the 1990s were terminated for various reasons, with only the Mesić and Mateša cabinets being ended for technical reasons: Mesić resigned as prime minister due to his election as Croatia’s member of the federal collective state presidency of Yugoslavia, while Mateša’s government was terminated by a regular election. All other single-party governments in the same period faced a premature end, mostly due to being affected by some kind of terminal event, such as national security issues in the cases of the Manolić and Valentić cabinets or the popular resentment that brought down the Šarinić cabinet amidst a number of financial scandals and poor economic performance. However, the real driving force behind cabinet terminations in the 1990s was a powerful president, who imposed strong control over both party and cabinet. He was acting as an overall manager of all political developments in Croatia at that time, making prime ministers and other cabinet members merely his political puppets in the executive arena. In comparison to the previous period, the circumstances of cabinet termination have changed dramatically after the constitutional reform in 2000. Table 19.6 reveals that most coalition cabinets ended their life cycle abruptly, mostly due to disagreements between the coalition partners arising from both policy and personal conflicts. Only a handful of cabinets (Račan III, Sanader II, Kosor II, and Milanović) were terminated for technical reasons, namely because regular parliamentary elections were held. Furthermore, only in the case of the Sanader III government was the main reason for cabinet termination the resignation of the ‘chief executive’, though he left the office and politics without revealing the reasons behind his decision.⁹ All other coalition governments in the last two decades were terminated for discretionary reasons, primarily due to intracoalition conflicts that made the coalitions fall apart. It is also useful to take a look at the individual cases, as it is revealed that disagreements over policy led to the fall of the Račan I, Račan II, Kosor I, Orešković, and Plenković I governments, while the Sanader I cabinet was terminated over the constant personal conflicts between the two leaders of the coalition parties. However, at the same time, it is important to note that the absence of functional conflict resolution mechanisms contributed significantly to cabinets being terminated prematurely, since the coalition parties were left to deal with their internal conflicting issues without the established mechanisms that would preserve them for longer.

⁹ It was later discovered that Sanader had built a large-scale corruption scheme that involved the highest-ranking government officials, ministers, and other prominent members of the HDZ (Nikić Čakar 2013).

1990-08-24 1991-07-17 22.9

1991-07-17 1992-02-20 19.7

Manolic

Greguric I

1992-08-12 1993-03-29 14.8

9.9

Greguric III 1992-05-14 1992-08-02

Sarinic

9.4

Greguric II 1992-02-20 1992-05-14

5.6

1990-05-31 1990-08-24

Relative duration (%)

Mesic

Date out

Date in

Cabinet

9

4

7a

7a

9

2

Mechanisms of cabinet termination

Table 19.6 Cabinet termination in Croatia, 1990–2018

11, 13

12

10

HNS

HSLS

Continued

Mesic was elected as Croatia’s member of the federal collective state presidency of Yugoslavia. With the outbreak of the Yugoslavian civil war, president Tudjman dismissed Manolic as PM and appointed Greguric to head a government with the mandate to form the new government. HSLS withdrew from the cabinet following the disagreement over the peace agreement and the inability of the cabinet to influence presidential decisions. HNS withdrew from the cabinet following the passage of a controversial piece of legislation according special status to Serbs living in the Knin and Glina districts. On 24 June 1992, President Tudjman called the first elections, both presidential and parliamentary, under Croatia’s new constitution, for August 2. The stated reason for Sarinic’s resignation was the recent election to the upper house of the Croatian legislature, amidst popular resentment due to a

Terminal Parties Policy Comments events involved area(s)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1993-04-03 1995-10-29 71.5

1995-11-07 2000-01-03 98.4

2000-01-27 2001-06-04 32.4

2001-06-04 2002-07-05 38.3

Matesa

Racan I

Racan II

Relative duration (%)

Valentic

Date out

Date in

Cabinet

Table 19.6 Continued

7a

7a

1

4

Mechanisms of cabinet termination

12

SDP, HSLS

IDS

number of financial scandals and a poor economic climate. Terminated by early elections called by the government attempting to capitalize on important military victories over Serbian forces in May. Terminated by regular election. The election was originally scheduled for 22 December 1999 but was postponed due to President Tudjman’s poor health. The Istrian Democratic League (IDS) withdrew from the government, including its only minister, following continued policy differences with the other members of the coalition. The final catalyst was IDS planning to give the Italian language legal status in Istria and make the county officially bilingual. Racan resigned on 5 July, following policy differences within the ruling coalition. The Croatian Social Liberal Party (HSLS), one of the coalition members, did not support the ratification of an agreement with Slovenia on the use and funding of a nuclear power plant, as it required Croatia to dispose of nuclear waste.

Terminal Parties Policy Comments events involved area(s)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

9

7a

Sanader III 2008-01-13 2009-07-01 35.2

2009-07-06 2010-07-10 37.7

2010-07-10 2011-12-04 83.9 2011-12-23 2015-11-08 93.2 2016-01-22 2016-06-16 9.8

Kosor I

Kosor II Milanovic Oreskovic

1 1 6, 7a, 7b

1

2006-02-09 2007-11-25 88.3

Sanader II

1 7b

2002-07-30 2003-11-23 78.6 2003-12-23 2006-02-09 51.3

Racan III Sanader I

HDZ, Most

HDZ, HSLS

HDZ, DC

Continued

Oreskovic was successfully removed by a motion of no confidence initiated by HDZ. The previous day Deputy PM and HDZ leader Tomislav Karamarko had resigned amidst charges of conflicts of interest, averting a motion of no

DC’s sole minister, Vesna Škare Ožbolt, was dismissed from the government on 9 February 2006, over continued criticism from her against the government. Parliament voted to dissolve itself on 12 October, in order to hold elections due by November. Elections were held on 25 November. Sanader unexpectedly resigned as prime minister on 1 July 2009, announcing his withdrawal from politics as well. Effectively no reason was given for his resignation. HSLS withdrew from the government over policy differences and supposed allegations from HDZ that HSLS was a disloyal coalition partner.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Mechanisms of cabinet termination

HDZ, Most

Note: Relative duration: share of constitutionally allowed duration for this cabinet.

14: Personal event Parties involved = between or within which parties.

10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event;

Terminal events

9: Other voluntary reason

7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties;

4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament;

Discretionary terminations

confidence from SDP that was crucially supported by MOST. The coalition collapsed due to disagreements between the coalition partners arising during the Agrokor crisis in April 2017. HDZ entered into a coalition agreement with HNS.

Terminal Parties Policy Comments events involved area(s)

Technical terminations 1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of prime minister

3.2

5

Relative duration (%)

Plenkovic II 2017-04-27 2017-06-09

Date out

7a, 7b

Date in

Plenkovic I 2016-10-19 2017-04-27 12.5

Cabinet

Table 19.6 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



675

Conclusions As discussed in this chapter, it can be concluded that coalition governance in Croatia comes closest to the prime minister model, which is a rather similar governance pattern to those in several other countries in Central and Eastern Europe, namely Poland, Slovakia, and Hungary (Bergman et al. 2019). Generally speaking, all three stages of coalition governance in Croatia are heavily dominated by top party leaders and particularly prime ministers, thus creating the patterns of informal and personalized coalition decision-making. The prime ministerial dominance is reflected in weak coalition arrangements, with very limited coordination established between coalition parties and the lack of broader conflict resolution mechanisms. Although legal provisions pave the way for resolving intra-coalition conflicts in the inner cabinet as a formal government body, this chapter shows that party summits work rather well as a sort of informal and unstructured channel used in both the most common and the most serious conflicts. However, the coalition governance overdependence on an informal relationship between the prime minister and other coalition party leaders makes coalition cabinets considerably fragile and unstable, which can be seen in the frequent cabinet turnover that comes as a direct consequence of internal policy and personal conflicts. Prime ministers are the predominant actors in the executive arena, supported by constitutional and legal provisions and unchallenged by relatively weak ministers. The sources of their executive power lie both in the formal authority of the prime minister’s office and in their undisputed intra-party standing. The latter is also important for the survival of coalition governments in the parliamentary arena, as coalition parties usually insist on maintaining coalition discipline and cohesion in the parliament. On the other hand, parliamentarians have very limited opportunity to exercise any controlling mechanism over coalition cabinets; therefore in some cases they resort to disobedience in order to express their dissatisfaction with coalition policy. Furthermore, prime ministers are not constrained by other, specifically designed, conflict management mechanisms, particularly coalition committees, as they have proved to be inefficient and completely overshadowed by individualized ways of handling coalition governance. Finally, coalition governance in Croatia is faced with strong challenges, emanating from an increasingly unstable party system that has become distorted with the rise of new anti-establishment and populist parties. These recent developments are making the government formation process a rather difficult task, as was the case in the coalition negotiations on the government formation in 2015. The unexpected results of the most recent parliamentary elections in 2015 and 2016 have shaped a new political landscape in Croatia and destabilized two-bloc party system, putting the new challenger party Most into a pivotal position. However, in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

676

´ ˇ 

both instances, coalition governments with Most as the junior partner did not last for long and were terminated for reasons previously inconceivable in the Croatian politics: the Orešković government was brought down by a vote of no confidence in the parliament, initiated by the HDZ as senior coalition partner, while the Plenković I government collapsed due to policy disagreement at the moment when the prime minister unilaterally decided to dismiss Most’s ministers during a cabinet meeting (Nikić Čakar and Krašovec 2021). Since Most proved to be a rather unreliable coalition partner, behaving like the ‘opposition within the government’, traditional mainstream parties moved forward by experimenting with new coalition formation formulae, which entailed the breaking of traditional ideological barriers between parties in the coalition government formation.

Appendix. List of political parties Abbreviation SKH-SDP

Name League of Communists of Croatia-Party of Democratic Reform (Savez komunista Hrvatske-Stranka demokratskih promjena) SDSH Social Democrats of Croatia (Socijaldemokrati Hrvatske), 1992–1994 Social Democratic Party of Croatia (Socijaldemokratska stranka Hrvatske), 1990–1992 SNS Serbian People’s Party (Srpska narodne stranka) HLSR Croatian Labourists—Labour Party (Hrvatski laburisti-stranka rada) ŽZ Human Shield (Živi zid) IDS Istrian Democratic Assembly (Istrian Democratic Assembly) SDP Social Democratic Party of Croatia (Socijaldemokratska partija Hrvatske), 1993– Party of Democratic Reform (Stranka demokratskih promjena), 1990–1993 SDSS Independent Democratic Serb Party (Samostalna demokratska srpska stranka) HNS Croatian People’s Party—Liberal Democrats (Hrvatska narodna stranka— liberalni demokrati), 2005– Croatian People’s Party (Hrvatska narodna stranka) LIBRA LIBRA—Party of Liberal Democrats (LIBRA—Stranka liberalnih demokrata) LS Liberal Party (Liberalna stranka) HSLS Croatian Social Liberal Union (Hrvatski socijalno-liberalni savez) HSU Croatian Party of Pensioners (Hrvatska stranka umirovljenika) HSS Croatian Peasant Party (Hrvatska seljačka stranka) Most Bridge of Independent Lists (Most nezavisnih lista) DC Democratic Centre (Demokratski centar) HDZ Croatian Democratic Union (Hrvatska demokratska zajednica) HKDU Croatian Christian Democratic Union (Hrvatska krscanska demokratska unija), 1992– Croatian Christian Democratic Party (Hrvatska kršćanska demokratska stranka), 1990–1992 HDS Croatian Democratic Party (Hrvatska demokratska stranke)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 HDSSB HSP-AS HSP

677

Croatian Democratic Alliance of Slavonia and Baranja (Hrvatski demokratski savez Slavonije I Baranje) Croatian Party of Rights dr. Ante Starčević (Hrvatska stranka prava dr. Ante Starčević) Croatian Party of Rights (Hrvatska stranka prava)

Note: Party names are given in English, followed by the party name in Croatian in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use.

References Bäck, Hanna, Marc Debus, and Patrick Dumont (2011). ‘Who Gets What in Coalition Governments? Predictors of Portfolio Allocation in Parliamentary Democracies’. European Journal of Political Research, 50(4): 441–78. Bergman, Torbjörn, Gabriella Ilonszki, and Wolfgang C. Müller (eds) (2019). Coalition Governance in Central Eastern Europe. Oxford: Oxford University Press. Boban, Davor (2016). ‘Promjene u sustavu vlasti i konsolidacija demokracije: usporedba uloge predsjednika države u Hrvatskoj i Slovačkoj’. Studia lexicographica: časopis za leksikografiju i enciklopedistiku, 10(1): 151–72. Boban, Davor (2017). ‘Dispozicijska i relacijska obilježja hrvatskog sustava vlasti: političke posljedice mogućeg povećanja ustavnih ovlasti predsjednika Republike’. In Tvrtko Jakovina (ed.), Dvadeset pet godina hrvatske neovisnosti—kako dalje? Zagreb: Centar za demokraciju i pravo Miko Tripalo, 47–66. Bunce, Valerie J., and Sharon L. Wolchik (2006). ‘International Diffusion and Postcommunist Electoral Revolutions’. Communist and Post-Communist Studies, 39: 283–304. Čular, Goran (2000). ‘Political Development in Croatia 1990–2000: Fast Transition – Postponed Consolidation’. Politička misao, 37(5): 30–46. Čular, Goran (2001). ‘Vrste stranačke kompeticije i razvoj stranačkog sustava’. In Mirjana Kasapović (ed.), Hrvatska politika 1990–2000. Zagreb: Fakultet političkih znanosti, 121–46. Čular, Goran, and Dario Nikić Čakar (2019). ‘Institutionalisation of a Charismatic Movement Party: The Case of Croatian Democratic Union’. In Robert Harmel and Lars G. Svåsand (eds), Institutionalisation of Political Parties. London and New York: ECPR Press/Rowman & Littlefield, 171–92. Čular, Goran, and Ivan Gregurić (2007). ‘How Cleavage Politics Survives Despite Everything: The Case of Croatia’. Paper presented at the ECPR Joint Sessions of Workshops, Helsinki.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

678

´ ˇ 

Dunleavy, Patrick, and R. A. W. Rhodes (1990). ‘Core Executive Studies in Britain’. Public Administration, 68(1): 3–28. Fisher, Sharon (2006). Political Change in Post-Communist Slovakia and Croatia. Basingstoke: Palgrave Macmillan. Grbeša, Marijana, and Berto Šalaj (2018). Dobar, loš ili zao? Populizam u Hrvatskoj. Zagreb: Tim Press. Gregurić, Franjo (1998). Vlada demokratskog jedinstva Hrvatske. Zagreb: Naklada Zadro. Gregurić, Franjo (2017). Vlada demokratskog jedinstva—Istina o Vukovaru i škola demokracije u Hrvatskoj. Zagreb: Školska knjiga. Helms, Ludger (2005). Presidents, Prime Ministers and Chancellors. Basingstoke: Palgrave Macmillan. Henjak, Andrija (2018). ‘Lojalnost, glas ili izlazak: izborna participacija i potpora novim strankama u Hrvatskoj’. Anali Hrvatskog politološkog društva, 14(1): 79–103. Henjak, Andrija, Nenad Zakošek, and Goran Čular (2013). ‘Croatia’. In Stein Berglund, Joakim Ekman, Kevin Deegan-Krause, and Terje Knutsen (eds), The Handbook of Political Change in Eastern Europe. Cheltenham: Edward Elgar, 443–80. Kasapović, Mirjana (2001). ‘Demokratska konsolidacija i izborna politika u Hrvatskoj 1990–2000’. In Mirjana Kasapović (ed.), Hrvatska politika 1990–2000. Zagreb: Fakultet političkih znanosti, 15–40. Kasapović, Mirjana (2005). ‘Koalicijske vlade u Hrvatskoj: prva iskustva u komparativnoj perspektivi’. In Goran Čular (ed.), Izbori i konsolidacija demokracije u Hrvatskoj. Zagreb: Fakultet političkih znanosti, 181–209. Kasapović Mirjana (2011). ‘Drugi kritični izbori u Hrvatskoj – slom jednog modela vladanja’. Političke analize, 2(8): 3–9. Kasapović, Mirjana (2014). Kombinirani izborni sustavi u Europi 1945-2014. Zagreb: Plejada. Lamza Posavec, Vesna (2000). ‘Što je prethodilo neuspjehu HDZ-a na izborima 2000: rezultati istraživanja javnog mnijenja u razdoblju od 1991. do 1999. Godine’. Društvena istraživanja, 9(4–5): 433–71. Laver, Michael J., and Kenneth A. Shepsle (1996). Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Manow, Philip, and Hendrik Zorn (2004). Office versus Policy Motives in Portfolio Allocation: The Case of Junior Ministers. MPIfG Discussion Paper 04/9. Köln: MaxPlanck-Institut für Gesellschaftsforschung. Niclauss, Karlheinz (2000). ‘The Federal Government: Variations of Chancellor Dominance’. In Ludger Helms (ed.), Institutions and Institutional Change in the Federal Republic of Germany. London: Palgrave Macmillan, 65–83.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi



679

Nikić Čakar, Dario (2013). Prezidencijalizacija političkih stranaka. Zagreb: Fakultet političkih znanosti. Nikić Čakar, Dario (2019). ‘The Presidentialisation of Political Parties in Croatia: Institutional Change Matters’. In Gianluca Passarelli (ed.), The Presidentialisation of Political Parties in the Western Balkans. London: Palgrave Macmillan, 23–47. Nikić Čakar, Dario, and Goran Čular (2016). ‘Organizational Structures of Political Parties in Croatia’. In Katarzyna Sobolewska-Myślik, Beata Kosowska-Gąstoł, and Piotr Borowiec (eds), Organizational Structures of Political Parties in Central and Eastern European Countries. Krakow: Jagiellonian University Press, 109–32. Nikić Čakar, Dario, and Alenka Krašovec (2021). ‘Coping with the New Party Challenge: Patterns of Prime Ministerial Survival in Croatia and Slovenia'. East European Politics (Online first). Nikić Čakar, Dario, and Višeslav Raos (2016). ‘Croatia’. European Journal of Political Research Political Data Yearbook, 55(1): 50–8. Warwick, Paul V., and James N. Druckman (2006). ‘The Portfolio Allocation Paradox: An Investigation into the Nature of a Very Strong but Puzzling Relationship’. European Journal of Political Research, 45(4): 635–65. Zakošek, Nenad (2002). Politički sustav Hrvatske. Zagreb: Fakultet političkih znanosti. Zakošek, Nenad (2008). ‘Democratization, State-building and War: The Cases of Serbia and Croatia’. Democratization, 15(3): 588–610.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Chapter 20 Coalition Governance Patterns across Western Europe Torbjörn Bergman, Hanna Bäck, and Johan Hellström

The ambition of this book has been to deepen our understanding of the coalition life cycle in Western Europe, building on previous studies focusing on this topic. In 17 chapters, country experts have provided updated information and analysed the formation of governments, the design and use of various coalition governance mechanisms, and the duration and termination of cabinets up until 2019. The country experts have, through careful analysis of various sources, including interviews with elite actors, not only provided updated information on the various stages in the coalition life cycle but also produced new knowledge on how coalition partners make policy once a coalition has formed, thereby contributing to the growing literature on coalition governance. Before turning to a comparative overview of the life cycle of coalition governments, we briefly describe the party systems in Western Europe and the institutional rules that guide government formation and termination in the countries covered in this volume. We here show that important changes in the party systems have occurred in the past decades; for example, most countries have experienced an increased party system fragmentation and a significant rise of populist parties, mainly so called radical right-wing parties. Such party system changes are likely to influence various stages of the coalition life cycle, which we discuss in this concluding chapter.

The party systems To understand coalition politics, the party systems and their changes over time are of importance. In particular, the number of political parties in a political system, their relative size, and the main dimensions of competition among them are critical, as they to a large extent determine the complexity of intra-party interactions and bargaining among political parties. In coalition theory as well as in comparative research, party conflicts are usually conceptualized in terms of the left–right policy dimension. While there are, in most party systems, other dimensions structuring party competition, none is consistently as salient as the left–right Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Coalition Governance Patterns across Western Europe In: Coalition Governance in Western Europe. Edited by: Torbjörn Bergman, Hanna Bäck, and Johan Hellström, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/oso/9780198868484.003.0020

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

681

policy dimension. In Table 20.1, which is based on the respective country chapters, we show the placement of the political parties on various conflict dimensions including the economic left–right policy dimension. In most countries there is normally more than one relevant dimension of party conflict, and in some cases, the dimensionality also varies over time. According to our country experts, in a few countries and at certain time periods, party competition has been structured over as many as three policy dimensions (in Belgium, Greece, Italy, Sweden, and the United Kingdom). More commonly, the economic left–right policy dimension together with potentially one additional policy dimension is enough to provide a good description of the type of policy issues that structure party competition in a country. In addition, as mentioned in the introductory chapters, most countries have seen an increasing fragmentation of their party systems over time, with the loss of electoral support of centrist parties and the entry of new parties. Moreover, in Table 20.1, we can see that on average, the effective number of parliamentary parties (ENP) has increased over time in Western Europe and this trend is evident for most countries. When looking at the ENP measure over time, and comparing the columns for two periods (1945–1989 and 1990–2019), we can see that all countries, except France and Portugal, have seen an increase in party system fractionalization—often because of the electoral success of anti-establishment parties and populist parties of the extreme left and/or right. Here, we will discuss the importance of party system fractionalization and polarization for coalition formation and cabinet termination, but we will next turn to the institutional context and its variation over countries and time. Thus, the next section summarizes the institutional context in which coalition politics takes place and provides an overview of the most important institutional changes that have taken place over time.

Institutional rules and conventions The rules surrounding government formation and termination Returning to the institutions that we introduced in Chapter 1, Table 20.2 summarizes the major changes in the Western European institutions of parliamentary democracy during post-World War II Europe. While early coalition theorists did not pay much attention to institutions, from the late 1980s and onwards this began to change. In Chapter 2, we mentioned that Strøm (1990: 79) argues that the presence of a constitutional requirement of investiture can make the formation of minority government more difficult. In that context, Bergman (1993) suggests that negative parliamentarism is important. In contrast to early coalition theory, which assumed that winning meant having the support of an absolute majority in

Economic left–right

Economic left–right (1945–1998)Sociocultural (1999–2006) Communities (2007–2019) Economic left–right Economic left–right Economic left–right Economic left–right Economic left–right

Economic left–right

Economic left–right Economic left–right Economic left–right

Economic left–right

Economic left–right Economic left–right

Belgium

Croatia Denmark Finland France Germany

Greece

Iceland Ireland Italy

Netherlands

Norway Portugal

Urban–rural Northern Ireland Pro/anti friendly relations with USSR (1947–1993)Sociocultural (1994–2019) Religion (1946–2001) Immigration/ integration (2002–2019) Urban–rural

Foreign policy (1949–1989) Sociocultural (1990–2019) Sociocultural

Sociocultural/centre–periphery

Sociocultural

Religion (1945–1985) Sociocultural (1989–2019) Community conflict (1946–1998, 2007) Economic left–right (1999–2006, 2008–2019)

Relation with European Union (2011–2019)

Memorandum of Understanding (2011–2015)

Church/state (1946–1991) Sociocultural (1991–1998, 2007–2019)

3.25 3.56

4.64

3.79 2.76 3.65

4.48 2.68

5.71

4.17 3.53 5.06

3.00

3.37 5.14 5.39 2.82 3.23

– 4.51 5.16 3.60 2.54 2.29

8.35

3.65

4.69

2.25

1990–2019

1944–1989

Third dimension

First dimension

Second dimension

Effective number of parties

Number of dimensions

Austria

Country

Table 20.1 Party system indicators in Western Europe and Croatia

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Economic left–right (1945–2015, 2017–2019) Relation to Europe (2016–2017)

United Kingdom

Means

Economic left–right Economic left–right

Spain Sweden

Centre–periphery Environment–growth (1945–2001) Immigration (2002–2005, 2010–2014) Multiculturalism (2006–2010) Sociocultural (2014–2019) Sociocultural (1970–1973, 2005–2016) Relation to Europe (1997–2004, 2017–2019) Sociocultural (2001–2004) Relation to Europe (2005–2016)

Immigration (1991–1993)

3.69

2.10

2.78 3.23

4.19

2.39

3.07 4.34

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Year of first cabinet under rule

1945 1946 1995 1990 1991 1993 2000 2001 1945 1953 1945 1995 2002 1959 1949 1977 1985 1944 1944 1948 1988 1946 2012 1945 1976

Country

Austria Belgium Belgium Croatia Croatia Croatia Croatia Croatia Denmark Denmark Finland Finland Finland France Germany Greece Greece Iceland Ireland Italy Italy Netherlands Netherlands Norway Portugal

no yes no yes yes yes yes yes no no no yes yes no yes yes yes no yes yes yes no yes no no

Positive parliamentarism

no no yes no no no no no no no no no no no yes no no no no no no no no no no

Constructive vote of no confidence yes yes yes no no yes yes no yes no no no no yes yes no no no yes yes yes yes yes no no

Bicameral system

Table 20.2 Institutional rules and conventions concerning cabinet formation

no no no no yes yes no no no no yes yes no yes no yes no no no no no no no no yes

Semipresidential system no yes yes yes yes yes yes yes no no no yes yes no yes yes yes no yes yes yes no yes no yes

Investiture vote required 5 2 4 1 1 1 1 1 5 5 5 2 2 5 1, 1, 3 1 1 5 2 2 2 5 2 5 4

Investiture vote decision rule

no yes yes no no

no no

no no

yes no

Cabinet responsible to upper chamber

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1983 1977 1979 1945 1970 1976 1945

no no yes no no no no

no no yes no no no no

no yes yes yes no no yes

no no no no no no no

yes no yes no no yes no

4 5 1, 2 5 5 4 5 no

yes no yes

Note: Investiture decision rules: 1: Majority support, 50 per cent + 1 in support of government; 2: Plurality support; 3: Plurality support for different options; 4: No majority against; 5: No investiture vote. When there are multiple decision rules reported, this reflects a change in the decision rule as the voting rounds progress. That is, e.g. 1, 1, 3 for Germany, indicates that in the first two rounds, a 50 per cent + 1 vote in support of the government is required, while in the third round, the decision rule is changed to one requiring plurality support.

Portugal Spain Spain Sweden Sweden Sweden United Kingdom

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

686

  .

an investiture vote, in negative parliamentary systems, an incoming cabinet (prime minister, PM) does not have to win such a parliamentary majority vote. It is enough that a parliamentary majority does not vote against the cabinet. In such systems, there is a higher frequency of minority governments. It has also been suggested that other institutional arrangements, such as semipresidentialism, bicameralism, and a constructive vote of confidence may have important effects on coalition formation. The logic is similar for both semipresidentialism and bicameralism, as parties may need to take into account, already during the coalition formation process, that the cabinet has enough legislative support to implement its policy agenda. That is, at the government formation stage, political parties have an incentive to build a large enough coalition with sufficient share of seats in parliament to secure parliamentary success (Sjölin 1993; Lijphart 2012; Thürk et al. 2020). Finally, several scholars (e.g. Strøm et al. 1994; Mitchell and Nyblade 2008) have argued that a constructive vote of confidence largely excludes the formation of minority cabinets. However, it has also been shown that when the sample is large, and includes both Western and Central Eastern Europe, it is difficult to establish that there is a consistent, statistical, and cross-national impact of institutions (Bergman et al. 2015). Rather the impact of institutions is conditional on party competition and country-specific configurations of institutions. Thus, institutions are important, but their specific impact, or lack of impact, can vary between different contexts (see e.g. Thürk et al. 2020). Table 20.2 describes which countries have positive parliamentarism, bicameralism, semi-presidentialism, and a constructive vote of no confidence, and also if there have been any major changes. After the Second World War, or the beginning of the current democratic regime, it is clear that even if institutional change of the basic structures of each representative democracy is relatively rare, there have been some important changes (see the country chapters for more detail and sources). For example, Finland in the late 1990s and early 2000, first informally and then constitutionally, moved from negative to positive parliamentarism. The Netherlands moved in the same direction with the cabinet that formed in 2012. Belgium changed its government formation rules in the mid-1990s (e.g. André et al. 2015). Before, the constitutional convention was that the parliament voted on the incoming cabinet’s government declaration, resembling a system of positive parliamentarism. However, the incoming cabinet already held constitutional powers when it first was appointed by the king, which arguably could make Belgium a case of negative parliamentarism. With the reforms that went into effect in 1995, the rule is now that for the cabinet to be rejected by the parliament, an absolute majority must appoint an alternative PM. This suggests that the Belgian system can be classified as a case of negative parliamentarism; that is, there is no need for the cabinet to have active majority support at inauguration.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

687

The details of the investiture vote can in contested situations be important. For example, both Portugal and Sweden have investiture votes where the proposed cabinet is only rejected if a majority actively votes against the proposal. Abstentions count in favour of the proposed cabinet. Most countries instead require majority support for the proposed cabinet/PM. A few, Germany and Spain, have different decisions rules for different rounds of voting, with the highest threshold being required in the first round. In Austria, Denmark, France, Norway, and the United Kingdom, a new PM and cabinet are formally appointed by the head of state and no parliamentary vote is required. Constitutional reforms that are fairly rare, but do occur, include a move away from a two-chamber (bicameral) system to a unicameral one. This happened in Denmark in 1953, Sweden in 1970, and Croatia in 2001. Belgium kept its bicameral system but decided that the cabinet would from 1995 not be held accountable to the upper chamber. As for semipresidentialism, Croatia adopted semi-presidentialism during the first decade after democratization, only to abandon it again at the turn of the millennium. At about the same time, Finland decided to move from having a strong president in a semi-presidential system to establishing a parliamentary regime with limited and largely symbolic powers for the president. Portugal too, a bit earlier, moved in the same direction. In sum, the basic institutions and procedural rules of West European parliamentary democracy tend to be stable over time. This lack of variation makes it difficult to attribute changes in governance patterns to changes in institutions. However, there have been some important changes. In Western Europe, there have been some occasions of a move away from negative parliamentarism to positive majority support, away from bicameralism, and away from semi-presidentialism. In specific countries, and specific political systems, changes in these institutions can have significant effect on the coalition life cycle.

Prime ministerial powers Let us now turn to the role and power of the PM. It is likely that factors such as ideological heterogeneity can incite in coalition partners a perceived need to monitor each other, both in cabinet and in their legislative behaviour (Carroll and Cox 2012; Zubek 2015). An increased number of parties in the cabinet is also likely to make cabinet matters more complex. The powers of the PM is another important factor that can influence how the cabinet is set up and managed. Strøm et al. (2008) found that the PMs with more institutional powers were associated with fewer inconclusive bargain rounds, fewer surplus coalitions, and fewer

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

688

  .

cabinet members and that internal conflict management arenas were less likely to be used. Table 20.3 updates and outlines the variation in prime ministerial powers (PM Powers) relative to other ministers and the cabinet as a whole. First, we can note that how cabinets reach decisions, for example by majority or consensus defined by the PM, is established by formal rules in a majority of our countries. However, there are a few countries without formal rules governing cabinet decision-making procedures; here the cabinet can decide on its own procedures. In these countries (Belgium, Spain, Sweden, and the United Kingdom), decisions are typically made by consensus. However, this consensus is in the end defined by the PM. That is, rather than the cabinet voting on a proposal, the discussions end with the PM summarizing the discussion in favour of a particular proposal. To give a summary measure of PM Powers, we present seven indicators as a simple additive index, following Strøm et al. (2003). The indicators are based on the following questions: (1) does the PM appoint the cabinet ministers; (2) can he or she dismiss these ministers; (3) does the parliamentary accountability run through the PM, meaning that parliament can unseat ministers only by unseating the full cabinet (or the PM); (4) does the PM determine the jurisdiction of other ministers; (5) is there a formal steering mechanism; and (6) does the PM have full control over the cabinet agenda. When the PM also has a (7) bureaucratic structure in his or her own office to monitor other ministers and ministries, the PM has a full set of PM Powers. Strøm et al. (2003: 183–94) found the institutionally most powerful PM positions in Spain, Germany, and the United Kingdom, followed by the Portuguese, French, and Irish PMs. At the other end of this spectrum, the PMs in Austria, the Netherlands, and Italy clustered as weak PMs. In our updated ranking, PMs in Germany and the United Kingdom still have more institutional powers than their colleagues in most other countries. Portugal also belongs to this category with institutionally very powerful PMs. In these countries, the continued political survival of individual ministers relies on the PM and not directly on parliamentary support for the minister. In addition, these PMs determine the jurisdiction of ministers and have steering mechanisms and full control over the cabinet agenda. They also have an office staff or some form of structure that allows them to monitor individual ministers. Looking at the current systems, they are in our data shadowed by the PMs in Greece and Spain. At the other end of the scale, we find the PMs in Finland and the Netherlands, who each hold only one of the formal powers. In general, the PM formal powers are fairly stable over time. However, if there is a trend, it is in the direction of increasing powers for the PM. Croatia is a prime example of this and Italy is another. Notably, as Strøm et al. (2003, 2008) stress, the actual power of the PM is also dependent on relations to other actors, the party system, and the structure of the parliamentary democracy (see also Doyle 2020).

2 2 2 2 3 3 3 1 1

1 3 2 2 3 3 2 2

yes no no yes yes yes no no yes yes yes yes yes yes yes yes yes yes

2 2 3 4 2 2 3 2 3 2 2 2 2 2 2 2 2 2

yes no no yes yes yes no no yes yes yes yes yes yes yes yes no no

2 2 3 1 2 2 3 2 3 2 2 2 2 2 2 2

0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 2 2

1945 1946 1990 2001 1945 1964 1945 1999 1959 1949 1977 2013 2015 1944 2016 1944 1948 1988

Austria Belgium Croatia Croatia Denmark Denmark Finland Finland France Germany Greece Greece Greece Iceland Iceland Ireland Italy Italy

yes no yes yes yes yes yes yes no yes yes yes yes yes yes yes yes yes

Year Formal Type of Appoint Appoint Dismiss Dismiss Ministers’ of cabinet actual ministers ministers: ministers ministers: parliamentary change decision decision description description accountability rule rule of rules of rules

Country

Table 20.3 Prime ministerial powers

no no no no yes yes no no yes yes yes yes yes yes yes yes no no

Determine jurisdiction of ministries

yes yes yes yes no no no no yes yes yes yes yes no no yes no yes

Steering or coordination rights vis-àvis cabinet ministers

no yes no yes yes yes no no no yes yes no yes no no yes yes yes

Full control over agenda for cabinet meeting

2 1, 2

no yes

2 1, 2 1, 2 2 2 1, 2, 3 1, 2, 3 1, 2, 3

1, 2

2 1, 2

Continued

4 3 1 5 4 5 1 1 6 7 6 5 6 3 3 5 3 5

Regular PM bureaucratic Powers structure in PM’s office designed to monitor departmental affairs: description of structure (not mutually exclusive categories) yes yes no yes no yes yes yes yes yes yes yes yes no no

Regular bureaucratic structure in PM’s office designed to monitor departmental affairs

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

no no yes yes yes yes yes yes yes 2 1 2 2 2 1 2

no no yes yes yes yes yes yes yes 2 1 2 2 2 1 1

0 0 0 1 2 0 0 0 1

no no no yes yes no no no yes

Determine jurisdiction of ministries

no no no yes yes no no no yes

Steering or coordination rights vis-àvis cabinet ministers

no yes yes yes no no no no yes

Full control over agenda for cabinet meeting

yes yes no yes yes no yes yes yes

Regular bureaucratic structure in PM’s office designed to monitor departmental affairs

1 1 1

1, 2, 3, 4

2 2

Regular bureaucratic structure in PM’s office designed to monitor departmental affairs: 1 = Personal staff (political appointment); 2 = Civil service staff; 3 = PM occupies specific (PM) portfolio; 4 = Other

PM right to appoint/dismiss ministers: description of rules: 1 = Yes, PM alone; 2 = Yes, through formal act carried out by head of state; 3 = No, head of state has discretionary powers; 4 = Other

Type of actual decision rule: 1 = Unanimity; 2 = Majority; 3 = Consensus defined by PM; 4 = Consensus defined by head of state.

1 2 3 7 6 2 3 3 7

Regular PM bureaucratic Powers structure in PM’s office designed to monitor departmental affairs: description of structure (not mutually exclusive categories)

Notes: Included above are the characteristics for the first cabinet in each country, followed by any additional changes to the characteristics in subsequent years.

2 2 3 1 3 3 3 3 3

1946 2007 1945 1976 1977 1945 1964 1976 1945

Netherlands Netherlands Norway Portugal Spain Sweden Sweden Sweden United Kingdom

yes yes yes yes no no no no no

Year Formal Type of Appoint Appoint Dismiss Dismiss Ministers’ of cabinet actual ministers ministers: ministers ministers: parliamentary change decision decision description description accountability rule rule of rules of rules

Country

Table 20.3 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

691

The German chancellor and the UK PM both have similar formal powers, but the German leader is typically engaged in a coalition and often with another large party. The UK PM has typically been in charge of a one-party majority government. This provides for different relations with the cabinet ministers. Later, we return to discussing PM Powers in relation to the broader governance patterns of each country. Now, we turn to the first stage in the coalition life cycle, focusing on the ‘birth’, or formation, of governments.

Coalition bargaining and formation As mentioned in the introduction and in the appendix, we count the beginning of a new cabinet according to official criteria (sworn-in, passed parliamentary inauguration vote, date of the general election, etc.) and the end of the cabinet when the cabinet resigned or the date of a general election, whichever comes first. This also determines the length of the government formation process. The length of government formation process after an election is the number of days from the election until a new government is appointed (if the current government never resigns or is removed, the ‘new’ cabinet has a government formation duration of zero days). Similarly, the length of government formation processes that occur between elections is simply the elapsed time from when the previous government resigned (or was removed) until the next government took office. Table 20.4 shows two related aspects of bargaining delays in government formation processes, the number of inconclusive bargaining rounds and the number of days required for government formation. When comparing the two time periods in the table, 1945–1989 and 1990–2019, for most countries there is no major difference in the number of failed attempts (i.e. the number of inconclusive bargaining rounds). On average, however, the number of failed attempts before the final, successful one has fallen slightly in Western Europe. At the same time, it seems that the actual time it takes to form a new government is more protracted than before in a majority of the countries (in 10 of 17 countries). This is also reflected in the result that the average time it takes to form a government has increased. However, there are no large differences between the two time periods. To give a more detailed description of changes over time, Figure 20.1 shows the average government formation duration over the entire post-war period up to 2019. On average, it has only taken about 25 days to form a government in Western Europe over the whole period (34 days after parliamentary elections and 24 days between elections). But, as the figure shows, in the 1950s, a typical government formation was over in just over 25–30 days, and the time spent on bargaining over government gradually increased until the 1970s when it stabilized in just under 40 days. It was not until the late-1990s and early 2002 that an upward trend took over in the sense that government formation took longer—currently

Austria Belgium Croatia Denmark Finland France Germany Greece Iceland Ireland Italy Netherlands Norway Portugal Spain Sweden United Kingdom Means excl. Croatia Means incl. Croatia

Country

0 0 – 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 7 – 3 6 1 2 4 6 2 3 6 1 1 0 2 1

0.3 1.3 – 0.6 1.3 0.2 0.1 1.4 1.7 0.2 0.6 2.0 0.1 0.1 0.0 0.2 0.1 0.7

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 7 1 1 2 1 1 4 5 3 6 7 1 0 2 4 1

Max

Mean 0.4 1.2 0.0 0.1 0.2 0.1 0.1 0.5 0.5 0.5 0.4 1.2 0.2 0.0 0.2 0.7 0.2 0.4 0.4

0 1 – 0 0 0 0 0 0 0 3 10 0 0 20 0 1

Min 176 203 – 35 79 16 73 18 117 27 127 272 36 168 39 26 4

Max 43.0 43.7 – 8.9 26.0 4.9 18.4 7.3 28.6 13.4 37.2 77.5 9.0 48.0 32.3 5.2 1.0 26.2

Mean

Min

Mean

Min

Max

1945–1989

1990–2019

1945–1989

0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0

Min

124 589 125 18 66 9 141 10 73 70 128 225 128 110 131 134 5

Max

1990–2019

Number of days required for government formation

Number of inconclusive bargaining rounds

Table 20.4 Cabinet formation

67.2 88.5 17.9 4.4 15.7 2.7 51.0 3.6 17.2 20.0 33.9 116.8 26.0 26.9 42.6 23.8 1.2 32 30.7

Mean

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

693

Average number of days

60

50

40

30

20 1940

1950

1960

1970

1980

1990

2000

2010

2020

Year

Figure 20.1 Trend over the government formation process length in Western Europe 1945–2019 Notes: The figure shows the trend for the average number of days required for government formation in Western Europe between 1945 and 2019. The line is a so-called loess curve and shows the underlying trend in the data.

government formation takes around 55 days on average. From the 1990s onwards, there is also a disproportionately large proportion (about 60 per cent) of the government formations that have taken more than 100 days. As discussed in some of the country chapters, many of the most protracted government formations occurred in the last two decades—most recently, for example, after the German and Dutch elections in 2017 and after the Swedish election in 2018. After the Swedish election, it took 134 days and after the German election, it took 141 days before the same governments that were in office before the elections re-formed. In the Netherlands, it took even longer, 225 days. However, there is a lot of variation within countries in the duration of the bargaining processes, but also important systematic differences between countries. Figure 20.2 shows how long it takes to form governments after parliamentary elections on average in the different countries covered in this book. The figure shows that, on average, it has taken the longest time to form a government in the Netherlands, Belgium, and Austria, while it has taken the shortest time in Greece, France, and the United Kingdom. How long it takes to form governments is directly related to the type of governments that form. In this respect, the electoral system largely determines whether elections result in majority situations, where a single party controls a majority of seats in parliament, or minority situations, where two or more parties are required to reach a majority. In Western Europe, less than 15 per cent of the parliamentary elections has resulted in a single-party majority winner. This is

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

694

  . 88

86

Average number of days

80

60

61 49 42

40

42

40 31 27

20

24

21

21 12

5

4

1

N et

he rla Be nds lg i Au um str Fi ia nl an d Ita l S y G pai er n m a Ic ny ela Cr nd o Po atia rtu N gal or w Ire ay lan Sw d D e de en n m a Gr rk Un ee ite F ce d ra Ki nc ng e do m

0

9

Figure 20.2 Cabinet formation duration for post-election cabinets in Western Europe 1945–2019 Notes: The figure shows the average number of days required for the cabinet formation after parliamentary elections.

mainly because all countries, except the United Kingdom and France, use some form of proportional representation system. Regardless, coalitions are common in France, and the United Kingdom had its first regular coalition government in 2010. Table 20.5 provides an overview of party-based cabinets covered in this volume and also shows the variation of the different types of cabinet between countries and divided into two columns for two periods (1945–1989 and 1990–2019). In the table, we separate between five cabinet types, depending on the number of parties and whether the cabinet controls a legislative majority. In this regard, coalitions are the most common government type and constitute almost 70 per cent of the cabinets in Western Europe. In more detail, the most common form of government is minimal winning coalitions (MWCs), which make up about 28 per cent of all post-war cabinets, followed by oversized coalitions (24 per cent) and singleparty minority cabinets (18 per cent). About one-third of all post-war cabinets are minority cabinets (i.e. including minority coalitions). Here, the Scandinavian countries stand out in having most minority cabinets, with Denmark (90 per cent) at the top, followed by Sweden (74 per cent), Spain (73 per cent), Norway (64 per cent), Ireland (48 per cent), Italy (44 per cent), and Portugal (37 per cent). The countries not yet mentioned have a share between 7 and 22 per cent minority governments (except for Croatia, where the share constitutes 38 per cent of all cabinets).

555

28 43 21 39 51 37 31 21 36 29 64 30 33 19 15 31 27

Minority

Coalition cabinets (%) Minimal winning coalitions

Oversized coalitions

23.1

5.9 6.7 – 53.8 8.6 0 13.6 0 17.4 33.3 33.3 0 40.9 22.2 66.7 63.6 5.9

12.7

13.0

0 0 9.5 7.7 0 5.6 0 6.7 0 27.3 0 0 36.4 40 77.8 44.4 10 13.5

23.5 10 – 0 0 0 4.5 66.7 0 33.3 0 0 27.3 11.1 33.3 9.1 94.1

12.7

11.5

0 0 23.8 0 0 16.7 0 53.3 0 0 0 0 0 20 22.2 0 80 13.2

0 6.7 – 34.6 17.1 21.1 0 0 4.3 27.8 21.4 15 18.2 0 0 4.5 0

19.5

18.5

0 7.7 28.6 92.3 0 22.2 0 0 7.7 18.2 22.7 30 36.4 10 0 44.4 0 24.9

58.8 40 – 11.5 17.1 15.8 63.6 16.7 60.9 5.6 0 40 13.6 33.3 0 22.7 0

31.7

33.5

100 38.5 14.3 0 25 11.1 88.9 13.3 92.3 36.4 22.7 60 27.3 30 0 11.1 10 25.4

11.8 36.7 – 0 57.1 63.2 18.2 16.7 17.4 0 45.2 45 0 33.3 0 0 0

23.5

23.5

0 53.8 23.8 0 75 44.4 11.1 26.7 0 18.2 54.5 10 0 0 0 0 0

1945–1989 1990–2019 1945–1989 1990–2019 1945–1989 1990–2019 1945–1989 1990–2019 1945–1989 1990–2019

Number of Single-party cabinets (%) cabinets Minority Majority

Notes: Non-partisan cabinets are excluded.

Austria Belgium Croatia Denmark Finland France Germany Greece Iceland Ireland Italy Netherlands Norway Portugal Spain Sweden United Kingdom Total Means excl. Croatia Means incl. Croatia

Country

Table 20.5 Parliamentary cabinets

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

696

  .

Comparing the two time periods, two changes are particularly apparent, the decreasing share of single-party minority cabinets (from about 23 to 13 per cent) and the more frequent occurrence of MWCs (from about 25 to 34 per cent). Looking at the individual countries, the former change has mainly contributed to the fact that in all but five countries (i.e. France, Greece, Portugal, Spain, and the United Kingdom), single-party minority cabinets have become less common. The latter change is a trend covering several countries. In Austria, all cabinets from 1990 have been MWCs, followed by Iceland (92 per cent), Germany (89 per cent), and the Netherlands (60 per cent). However, MWCs have also become more common in Finland, Ireland, Italy, Norway, and the United Kingdom (although the latter only refers to one cabinet). Two of the most robust predictions on coalition formation is that parties will seek to form minimal connected winning coalitions when elections do not produce a single-party majority winner (De Swaan 1973). Moreover, under the assumption that coalition formation takes place in a one-dimensional political space, we could expect the median legislator party to have a high likelihood of being a cabinet member (De Swaan 1973; Laver and Schofield 1990: 110–13). Thus, political parties that are located at the centre on the most important ideological dimension have significantly greater opportunity to become a cabinet member than more ideologically extreme parties. In the post-war period, the median legislator party has been included in about three quarters of all cabinets. Table 20.6 shows that the median legislator party on the left–right policy dimension is less often in government than before (from 81 to 69 per cent), most likely due to the weakening of the left–right dimension in many party systems. In addition, in many countries there is also a strong practice that the head of state appoints the party leader of the largest party to form a government, that is that party leader becomes the formateur in the first formation attempt (e.g. Bäck and Dumont 2008; Döring and Hellström 2013). This is also reflected in the fact that the largest party is in government in about eight of ten cabinets in Western Europe since 1945, and it has become even more decisive in coalition formation over time (from 82 to 89 per cent). However, it should be noted that the median legislator party is also the largest party in about two-thirds of all cases. That the median legislator party and the largest party are important in government formations is apparent when considering that it is very uncommon that the government does not include one or the other (this has only happened in nine per cent of the cases).

Portfolio allocation and distribution of power We now turn to analysing how parties allocate ministerial posts in coalition governments in Western Europe. As described in Chapter 2 of this volume,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

697

Table 20.6 Government participation of parties with median legislator and the largest parliamentary party in Western Europe and Croatia Country

Austria Belgium Croatia Denmark Finland France Germany Greece Iceland Ireland Italy Netherlands Norway Portugal Spain Sweden United Kingdom Total Means excl. Croatia Means incl. Croatia

Number of party-based cabinets

28 43 21 39 51 37 31 21 36 29 64 30 33 19 15 31 27

Median party (first dimension) in cabinet (%)

Largest party in cabinet (%)

1945–1989

1990–2019

1945–1989

1990–2019

82.4 86.7 – 38.5 85.7 89.5 81.8 83.3 52.2 72.2 100 95 81.8 88.9 33.3 90.9 100

54.5 76.9 85.7 30.8 62.5 83.3 77.8 100 46.2 72.7 86.4 50 45.5 90 44.4 55.6 100

100 76.7 – 65.4 77.1 89.5 81.8 100 65.2 72.2 100 75 63.6 100 100 81.8 100

90.9 84.6 95.2 76.9 100 83.3 100 100 92.3 90.9 95.5 100 54.5 80 88.9 66.7 100

81.1

69

82

88.5

555 70.6

89.1

most previous work focusing on the distribution of ministerial posts has focused on predicting how many portfolios each party gets. Already in the 1970s, Browne and Franklin (1973) showed, using comparative data on Western European governments, that there is a very high correlation between the share of seats a governing party brings to the government’s total number of parliamentary seats and the share of ministerial posts the party obtains. This relationship has been dubbed ‘Gamson’s Law’ due to its high explained variance and has been replicated in more recent studies (e.g. Warwick and Druckman 2006). The main deviation from a ‘perfect one-to-one relationship’ between a party’s seat contribution and its share of ministerial posts that has been found in the literature is a ‘small-party-bias’—smaller parties in a coalition often receive more than a proportional share of the ministerial payoffs and larger parties receive less than what would be expected following a proportionality rule (e.g. Warwick and Druckman 2006).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

698

  . 100

Share of ministers (in percent)

80

60

40

20

20

40

60

80

100

Share of cabinet seats (in percent) OLS

Reference

Figure 20.3 The relationship between parties’ seat contributions and their share of ministerial posts Notes: The figure shows the relationship between share of ministers (in per cent) and share of cabinet seats (in per cent). Independent (nonpartisan) minsters are excluded from the calculation. The solid line is the estimated linear regression line and the dashed line a 45 degree reference line. The regression line is Y = 5.26 (4.62 ≤ β ≤ 5.90) + 0.834X (0.819 ≤ β ≤ 0.850), with 95 % confidence intervals within parentheses; R-squared: 0.905; n = 1157.

In Figure 20.3, we present the relationship between parties’ seat contributions and their share of ministerial posts using the data collected by our country experts, that is illustrating the proportionality relationship when analysing cabinets over the post-war period up until 2019. The figure shows that there is a positive and significant relationship between parties’ seat contributions and their portfolio payoffs—larger parties receive a large share of ministerial posts whereas smaller parties receive small shares, and the relationship between the two features is clearly a strong one in terms of explained variance (the R-square is equal to about 90 per cent). In addition, there are few larger deviations from the regression line. The fact that the regression line is flatter than the one-to-one line, as shown in the figure, illustrates that there is a ‘small-party-bias’, as shown in the previous literature—smaller coalition partners receive disproportionally more ministerial

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

699

portfolios than their partners. For example, in the coalition government between the Social Democrats and the Greens that formed after the 2014 election in Sweden, the Green party, with a seat contribution of 18 per cent, clearly the junior partner, obtained about 25 per cent of the ministerial posts in the cabinet. Hence, the Greens gained a substantially larger share of ministerial posts than a proportionality rule would have warranted. Why do larger parties seem to give up more of their ministerial share when forming a coalition with a smaller party? A number of potential explanations to this small-party-bias have been presented in the literature. One such suggestion is based on the work by Warwick and Druckman (2006), who argue and show that all ministerial portfolios are clearly not equally important. We also know that larger parties are often appointing important cabinet members, most importantly the PM (see e.g. Bäck and Dumont 2008). Are the smaller parties compensated with more portfolios because the larger parties hold the most important posts? This is a plausible explanation, but weighting portfolios by their importance does not seem to remove the small-party-bias since the regression slope is still flatter than the one-to-one line when focusing on predicting weighted portfolio shares (see Warwick and Druckman 2006; Bäck and Carroll 2020). Lastly, before turning to the second stage of the coalition life cycle, we will briefly discuss how power is distributed between partners in multiparty cabinets in Western Europe, focusing on the fact that in some cases the largest party gains control over most or all of the important portfolios in cabinet. In Figure 20.4, we present information on whether the same party controls the post as PM, as minister of foreign affairs, and as minister of finance, which are the most important portfolios in most countries in Western Europe (see Warwick and Druckman 2001). We here see that in about half of the Western European coalition cabinets formed during the post-war period, the party controlling the PM post also controlled the post of finance minister (51.5 per cent) or the post of minister of foreign affairs (46.9 per cent). In 25 per cent of the cases, one dominant party controlled all three of these important posts. This may suggest that large formateur parties typically gain control over the most important posts in cabinet in many of the coalition governments formed in Western Europe. However, Figure 20.4 also shows that there is variation across countries in terms of how concentrated power is in the cabinet. For example, in Croatia, Ireland, and Sweden, it is relatively common that one party controls all three of the important posts, whereas in other countries, such as Belgium, Denmark, and Finland, this is very rarely the case. These patterns most likely reflect the type of government that forms in these countries, where power concentration in terms of ministerial portfolios is more likely when a large dominant party, for example, a Social Democratic party, forms a coalition with one or more smaller parties, who do not gain control over any important posts.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

700

  . 100

100

85.71

75

71.43 63.64

Percent

57.58

50 37.5

25

24

21.74 21.43

24.93 20 18.52 16.67 9.38 8.33 7.89 4.17

Un ite d

Ki ng do Cr m oa t Ire ia lan Sw d ed e Fr n an c Gr e ee ce Ita Au ly str N ia or w Po ay rtu G ga e N rm l et an he y rla n Ic ds ela D nd en m Be ark lg iu m To Fin ta lan l( d M ea n)

0

Figure 20.4 Power concentration and power dispersion Notes: The figure show the share of cabinets within each country where the same party holds the Prime minister, Finance and Foreign affairs offices.

Coalition governance mechanisms A growing literature focusing on the second phase in the coalition life cycle, the so-called coalition governance phase, is concerned with how coalition partners try to constrain ministers from other parties via various mechanisms (e.g. Strøm et al. 2008). These mechanisms are set in place by political parties in multiparty cabinets to deal with the main challenge of reaching collective decisions in situations in which the individual incentives of each coalition party and the minister formally in charge of the policy area might differ from that of the majority in the cabinet. Putting it in principal–agent terms, the PM or the cabinet as a whole use such mechanisms to avoid that there is ‘agency loss’ when power in cabinet is delegated to individual line ministers. In Table 20.7, we present a summary of the results from the empirical investigations performed by our country experts in terms of what coalition governance mechanisms are used in their country. Coalition agreements will be discussed more in-depth in following section, but as part of giving an overview of various coalition governance mechanisms in Western European governments, we briefly present information of the occurrence of coalition agreements before turning to other control mechanisms.

23 38 14 24 48 33 27 8 32 14 50 30 14 10 11 1

Austria Belgium Croatia Denmark Finland France Germany Greece Iceland Ireland Italy Netherlands Norway Portugal Sweden United Kingdom

21 26 14 15 48 21 15 3 27 14 6 17 12 10 10 1

Coalition agreement 19 23 0 13 48 13 12 3 27 14 5 16 12 10 10 1

Agreement public 19 0 0 0 0 0 0 0 0 0 0 20 0 0 0 0

Election rule CaC, CoC, IC IC, CaC, PS IC, CaC, PS CaC, IC, O CaC, IC, Pca CaC, O, CoC Pca, Parl, IC CaC, PS, CoC IC, CoC, CaC Parl, PS, CoC CoC, PS, CaC Parl, IC, CoC CaC, IC, Pca CaC, PS, IC IC, CaC, Pca CaC, IC

Coalition management mechanisms (three most common) 20 25 7 0 0 22 15 7 0 14 50 30 9 10 3 1

Cabinets with watchdog junior ministers

O Other: this coding option has only been used with extra caution and scarcity and only if no other mechanism applied at all.

Other:

PS Party summit: typically ad hoc, consisting of one or several leaders for each coalition party, and some but not all attendants may be cabinet members

Parl Parliamentary leaders (heads of the coalition parties’ parliamentary groups)

External:

Pca Combination of cabinet members and parliamentarians

CoC Coalition committee: typically permanent with relatively stable membership, consisting of party leaders but not limited to cabinet members

CaC Cabinet committee(s): typically issue-specific and may include cabinet ministers, junior ministers, and/or civil servants Mixed:

IC Inner cabinet: a subset of cabinet ministers that is not issue-specific and is stable over time

Internal:

Conflict management mechanisms:

Number of coalition cabinets

Country

Table 20.7 Coalition governance

10 37 4 24 48 0 26 8 0 14 32 0 8 0 10 1

Noncabinet positions

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

702

  .

In some countries, written agreements are the norm when a coalition government forms; for example, all of the Finnish coalition cabinets are based on a written coalition contract. This is also the case for all of the coalition governments formed in Ireland and Portugal and almost all of the multiparty cabinets formed in Iceland, Norway, and Sweden. In total, 69 per cent of all the coalition governments formed during the post-war period in Western Europe have been based on a written coalition contract. Among those coalition agreements, most (68 per cent) were publicly available. This means that a large share of the multiparty governments in Western Europe make use of contracts that are in the public domain, making it likely that parties stick to these deals even when power has been delegated to individual ministers, as they can be held accountable for them by the voters. However, there are also countries where written agreements are less common; for example, in Italy, only 12 per cent (6 out of 50) of the coalitions that have formed are based on such contracts. In some countries, there is significant cross-cabinet variation; for example, in Germany and Denmark, about 60 per cent of the coalitions made use of written contracts as an ex ante control mechanism (56 per cent and 63 per cent respectively). As noted by several authors, coalition contracts sometimes contain specific mechanisms or rules for conflict resolution (see e.g. Timmermans 2003; Müller et al. 2008; Bowler et al. 2016). In the overview table of coalition governance mechanisms (Table 20.7), we provide information on how many coalitions include a so-called election rule. Such a rule entails a commitment of the parties forming a coalition government to call new elections in case the coalition breaks down. Only a handful of countries make use of this type of control mechanism, more specifically, all or almost all of the coalition cabinets formed in Austria and the Netherlands employ such a rule. Multiparty governments make use of a number of so-called conflict management mechanisms, and in Table 20.7, we provide an overview of the most commonly used such mechanisms in the Western European countries. Here, we are focusing on what, in the principal–agent literature, has been called ‘institutional checks’, aimed at controlling agency loss. As described by (Müller and Meyer 2010: 1075), these checks typically consist of ‘more or less permanent bodies with representatives from all coalition parties, including cabinet members, party leaders, and parliamentary leaders in various combinations’. Some coalitions use several such institutional mechanisms to control ministers, whereas others do not use any but instead rely on other control mechanisms. The most commonly used conflict management mechanism is a cabinet committee—such a committee is typically issue-specific and may for example include cabinet ministers, junior ministers, and/or civil servants. In 13 of the 16 countries (81 per cent) analysed in this volume, such cabinet committees are among the most common mechanisms used. The second most common institutional check

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

703

used in the Western European coalition governments is the so-called inner cabinet, which is a body including a subset of cabinet ministers, which is stable over time (and is non-issue-specific). Inner cabinets are among the most common mechanisms used in 12 of the 16 countries (75 per cent) analysed here. Another relatively common institutional check is the establishment of a so-called coalition committee, which is typically a permanent body with relatively stable membership, consisting of party leaders but not limited to cabinet members. In 7 of the 16 countries (44 per cent), this type of check is among the most commonly used mechanisms. Let us now turn to another control mechanism that was empirically investigated early in the literature on coalition governance, the use of so-called watchdog junior ministers. For example, Thies (2001) argued that parties in coalition governments appoint junior ministers coming from a different party than the head of the department to act as ‘watchdogs’ who are able to and have incentives to monitor the minister and provide information to their own party leaders, thereby reducing the minister’s informational advantage vis-á-vis the rest of the cabinet. In Table 20.7, we provide updated information on how many post-war cabinets in each country appoint junior ministers. In a number of countries, in Austria, Belgium, Greece, Ireland, Italy, The Netherlands, and Portugal, all or almost all coalition governments appoint junior ministers. In some countries, more specifically, in Denmark, Finland, and Iceland, no junior ministers are allowed (Verzichelli 2008; Strøm et al. 2010). In others, there is variation across cabinets; that is, some of the coalitions have appointed junior ministers, whereas others do not. This is for example the case in Germany and Sweden. In some of the countries, parties can also agree on non-cabinet appointments. As can be seen in Table 20.7, in all coalition cabinets formed in Denmark and Finland, the governing parties also agreed on non-cabinet appointments. In other countries, such as Iceland, Netherlands, and Portugal, the coalitions have not reported on any appointments outside of cabinet. We now turn to describing the discipline that coalition governments invoke in the parliamentary legislative process in the different Western European countries. Table 20.8 describes whether or not there is an informal rule that all parties and their members of parliament (MPs) should vote as a coalition bloc. In some countries, such as Austria, Denmark, and Ireland, commitment to coalition discipline in legislation is always or almost always the case. At the other extreme, there are some countries, more specifically, in France and Greece, where coalition discipline in legislation is the exception, or no such discipline is invoked. There are also a number of countries that fall somewhere in between these extremes, with some cabinets invoking discipline whereas others allow exceptions to discipline. Table 20.8 also provides information on how tightly coalitions have committed the partners in conducting other parliamentary business, such as asking parliamentary questions and committee work, which says something about how

23 38 14 24 48 33 27 8 32 14 50 30 14 10 11 1

Austria Belgium Croatia Denmark Finland France Germany Greece Iceland Ireland Italy Netherlands Norway Portugal Sweden United Kingdom

21 9 9 24 0 0 20 0 9 12 18 0 5 0 3 0

2 26 0 0 28 0 6 0 23 2 32 29 7 9 7 1

0 0 0 0 20 33 0 4 0 0 0 0 0 1 0 0

0 2 5 0 0 0 0 4 0 0 0 0 0 0 0 0

No discipline 21 0 9 24 0 0 25 0 10 14 17 0 2 0 3 1

2 0 0 0 28 0 1 0 22 0 0 0 10 4 7 0

Exceptions from discipline

Always

Discipline as exception

Always

Exceptions from discipline

Parliamentary behaviour

Legislation

Note: Data on coalition discipline is not available for all cabinets.

Coalitions

Country

Table 20.8 Coalition discipline in legislation and parliamentary behaviour

0 1 0 0 20 0 0 0 0 0 32 0 0 6 0 0

Discipline as exception

0 36 5 0 0 33 0 8 0 0 1 29 0 0 0 0

No discipline

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

705

coalition partners can check ministers through the parliamentary process (e.g. Martin and Vanberg 2011). A strong commitment to coalition discipline in various parliamentary matters means that governing parties are not able to use the parliamentary process to constrain their coalition partners’ ministers. Table 20.8 shows that in particular in Austria, Denmark, Germany, and Ireland, the coalition parties commit to joint parliamentary behaviour. Belgium, France, Greece, and the Netherlands are at the other side of that spectrum, allowing coalition partners to make use of the parliamentary process as a means to check on their partners.

The size and content of coalition agreements One of the most important ex ante control mechanisms that parties can use to control its partners when power is delegated to individual ministers, and to avoid that ministers ‘drift’ from the coalition policy, is to write a contract describing the policies that should be implemented during the government’s time in office. There is a growing literature focusing on this particular control mechanism (see e.g. Klüver and Bäck 2019). Most work on this topic has focused on the size of coalition agreements, that is on their length. For example, Indridason and Kristinsson (2013) argue that when the risk of ‘ministerial drift’ is severe, we should expect coalition agreements to be more extensive. Analysing coalition agreements in Western Europe, they show that the length of coalition agreements for example increases as the ideological diversity between partners increases. In Figure 20.5, we present updated information on the average length of coalition agreements (in number of words) in the Western European post-war cabinets. The figure shows that there is substantial variation across countries. Some countries have very high averages; for example, in Italy, Germany, Norway, and Belgium the average coalition agreement is over 18,000 words. In other countries, these contracts are significantly shorter; for example, the average size of these documents in Iceland, Greece, Finland, Sweden, and France is below 5,000 words. Here, it should be noted that in some cases, such as Sweden, the documents also serve as government declarations presented in parliament, rather than separate and full-fledged coalition agreements, which partly explains why they are shorter than the agreements in other countries. Previous research has shown that coalition agreements in Western Europe have become longer over time (see Strøm et al. 2008; Klüver et al. 2021). In Figure 20.6, we present the variation over time in the length of an average coalition agreement in Western Europe using our updated data provided by our country experts. Here, we can distinctly see that there has been a substantial increase in the number of words used by coalition partners when writing these contracts. In the 1950s, the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

706

  .

Average number of words

30,000

(6) 27760

(15) 19643 (14) (22) 18536 18073 (17) 16568 (1) (21) (12) (17) 14977 14749 14298 14083

20,000

(14) (10) 12114 11860

Total average: 11086

10,000 (11) (48) (12) 4056 3891 3836

(3) (28) 2409 2312

G Ita er ly m N any or Be way Un Net lgiu ite her m d lan Ki d ng s d Au om D str en ia m a Cr rk oa Ire tia l Po and rtu Fr gal a Fi nce nl a Sw nd ed Gr en ee Ic ce ela nd

0

Figure 20.5 Size of agreements

1

25000 .8 20000 15000

.6

10000 .4 5000 0 55 19 6 19 0 65 19 70 19 75 19 80 19 85 19 90 19 95 20 00 20 05 20 10 20 15 20 20

19

19

19

50

.2

Share of cabinets with coalition agreement

30000

45

Average length of coalition agreements

Notes: The figure shows the average length of coalition agreements in each country, and the number in parentheses above the bars shows the number of coalition agreements in each country.

Year Length (average number of words) Share (average proportion)

Figure 20.6 Length of coalition agreements in Western Europe Notes: All lines are moving three-year averages to account for an uneven number of agreements on a yearly basis. “Length” shows the trend for the average number of words in coalition agreements. Share refers to the proportion of cabinets with a written coalition agreement.

average length of a coalition agreement was about 800 words, whereas such documents are over 20,000 words on average since the end of the 2000s. This may suggest that coalition agreements have become a more important control mechanism for parties forming multiparty cabinets.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

707

100

Percent

75

50

25

Ita ly Au str ia

Ire lan d Fi nl an N d or wa y Ic ela nd Sw ed en Fr an Be ce lg iu m G er m N et any he rla nd Po s rtu ga l Gr ee ce Cr oa D tia en m ar k

0

Other (in %)

Distribution of offices and competences (in %)

Procedural rules (in %)

Policies (in %)

Content type

Figure 20.7 Content of coalition agreements Notes: The figure shows the average share of coalition agreement text focusing on either policy, procedural rules or portfolio allocation.

As described in Chapter 2 of this volume, Müller and Strøm (2000) introduced a comparative data set on coalition agreements in the Western European countries and showed that three different points are typically settled in coalition treaties: policy agreements, portfolio allocation, and procedural rules. Following this seminal work, we here present updated information on the content of coalition agreements, as classified by our country experts. In Figure 20.7, we describe the average share of agreement text focusing on policy, procedural rules, and portfolio allocation. Figure 20.7 shows that most agreements focus on policy—on average across all the Western European countries, 90 per cent of the texts deal with policy issues. This suggests, as the previous literature has shown, that parties in coalition governments use these documents to agree on what policies the government will pursue during its term of office and to constrain coalition partners from pursuing their own agenda when controlling a specific department (see e.g. Indridason and Kristinsson 2013; Klüver and Bäck 2019). In some countries, the parties in multiparty governments also agree on and present the distribution of ministerial portfolios in these agreements; for example, in Austria and Croatia, on average, over 25 per cent of the agreements focus on portfolio allocation. Very little content of the agreements in the Western Europe

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

708

  .

Table 20.9 Policy agreements Country

Comprehensive

Variety of issues

Only a few issues

No policy agreement

Coalitions

Austria Belgium Croatia Denmark Finland France Germany Greece Iceland Ireland Italy Netherlands Norway Portugal Sweden United Kingdom

12 22 8 24 25 0 9 2 19 3 4 14 9 10 7 1

2 6 2 0 6 9 8 2 7 8 16 3 1 0 3 0

9 7 3 0 14 5 9 4 6 1 30 7 2 0 0 0

0 1 1 0 0 19 0 0 0 2 0 1 0 0 0 0

23 38 14 24 48 33 27 8 32 14 50 30 14 10 11 1

Note: Data on policy agreements is not available for all cabinets.

focuses on procedural rules, but there are some exceptions; for example, in Austria, Croatia, and France, on average more than 10 per cent of the coalition contracts focused on such rules. For example, as described by Müller (Chapter 3), the agreement negotiated by ÖVP and FPÖ in 2000 (Schüssel I) contained specific descriptions of rules on inter-party coordination, for instance including commitments to closely monitor budgetary developments. As described by Müller and Strøm (2008: 174–6), coalition agreements also differ in their completeness, where some coalitions are based on a ‘comprehensive policy programme, in which the participants commit themselves to a broad range of policy initiatives’’, whereas other agreements include commitments only on a few selected issues. Whether an agreement is comprehensive or not is likely to matter for how effective it is in constraining individual ministers in multiparty cabinets. For example, as shown by Bäck et al. (2017) when analysing spending behaviour in 17 Western European countries, comprehensive agreements significantly reduce the effect of government fragmentation on spending. In Table 20.9, we present updated information on the comprehensiveness of coalition agreements in the Western European countries on the basis of information provided by our country experts. Here, we can see that there is clearly variation both across and within countries. In a small number of countries, the agreements are (almost) always

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

709

comprehensive. This is the case in Denmark, Norway, and Portugal. In most of the countries, there is variation across cabinets; for example, in Austria, Germany and the Netherlands, some coalitions have negotiated a comprehensive agreement, whereas others have relied on contracts covering only a few policy issues, perhaps relying more on other control mechanisms instead.

Cabinet termination and duration As showed by Taagepera and Sikk (2010), there is strong association between the electoral system and how long cabinets last. Although it is beyond this study to investigate this relationship in detail, here we show and discuss some important cross-country differences and how the general trend has changed over time. Figure 20.8 shows the share of early terminations in Western Europe. The figure shows so-called ‘discretionary terminations’, that is cabinet terminations that are both political and discretionary divided into early election or a non-electoral replacement (i.e. changes in the governing parties or of the PM). In the figure only so-called ‘technical terminations’ are excluded, that is cabinet terminations

Share of early terminations (%)

73 14

80

10

62 62

42

18

60 46

13

37 16

49 26

47

48 17 32 27

42

40 26

20

26

23

11

6

33

31

10

13

14 0

3

It Gr aly ee Ire ce l Be and lg D ium en m a N S rk et pa he in rla n Au ds str Fi i a nl Un a P ite o nd d rtu Ki ga ng l do Ic m ela G nd er m a Fr ny an N ce or w Sw ay ed en

0

Early elections

Replacements

Figure 20.8 Share of early terminations in Western Europe, 1945–2019 Notes: In the figure so called “technical terminations” are excluded. These are termination due to nonpolitical reasons such as regular elections, other constitutional reasons, and the death of the head of government. Rather the figure shows so called “discretionary terminations’, that is, cabinet terminations that are both political and discretionary. That is, early terminations because of the strategic choices of politicians which results in either an early election or a non-electoral replacement (i.e. changes in the governing parties or of the prime minister). This includes, for example, parliamentary defeats at the hands of the opposition, voluntary enlargements of coalitions or defections from the incumbent coalitions (due to, for instance, conflicts between coalition parties over policy).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

710

  .

due to non-political reasons such as regular elections, other constitutional reasons, or the death of the head of government. In 10 out of 16 countries it is more common that a cabinet terminates early than serve the full term. However, there are large difference between countries. In Italy less than 20 per cent of the governments have served the full term, whereas in Sweden the reverse is true. Only 16 per cent of the Swedish cabinets have terminated for reasons unrelated to the end of the parliamentary term or similar constitutional reasons. In some countries early elections are more common than non-electoral replacements. In several countries, not least the United Kingdom (up to 2010) and Denmark, the absence of fixed election dates encourages opportunistic election timing. For instance, governments may seize the opportunity to call for early elections during times of relatively strong economic performance (see e.g. Kayser 2005; Walther and Hellström 2019). In contrast, the Norwegian constitution does not permit early dissolutions, and the Swedish constitution states that an early election does not start a new term, which significantly reduces the propensity for early elections. Looking more closely at the different mechanisms of cabinet termination, Table 20.10 shows that about four out of ten cabinets have terminated due to technical reasons, while the majority have terminated for various discretionary reasons. However, as indicated earlier there are substantial cross-country differences. Where early discretionary elections account for a vast majority of all terminations in the United Kingdom, Ireland, Greece, Spain, Denmark, and Austria, this is almost unheard in Norway, Sweden, Finland, and Italy. The second most common mechanism of cabinet termination is conflict between coalition partners. In Belgium, Croatia, Finland, Iceland, Italy, and Sweden this accounts for most of all discretionary terminations (however, in Sweden, this has happened only three times). Figure 20.9 shows the trend in cabinet ‘stability’ over the entire post-war period up to 2019. The line shows the so-called relative duration, or the proportion of the term of office spent in office, that is the time elapsed between the cabinet’s appointment and termination as a percentage of its maximum feasible duration (Müller and Strøm 2000:16–17). Thus, a cabinet that serves the full term has a relative duration of one. In the figure we can see that an average cabinet between the 1940s to the 1960s serves a little more than a half term. The stability of cabinets then increased until the mid-1990s to about two-thirds of the term when the trend levelled out. From the mid-2000s we can see a clear trend towards more instability and that the average contemporary cabinet does not last much longer than a cabinet before the 1970s. However, the general trend does not say anything about the large cross-national differences, nor the origin of this trend. In this respect, several countries have exhibited a larger degree of early terminations in the last two decades (e.g. in Austria, Greece, and the United Kingdom). In Table 20.11 we show some of these changes but most importantly the differences between the

7

137

21

50

0 1 1 10 2 1 3 0 1 3 14 2 4 4 3 0 1 109

8 19 7 2 18 3 5 3 8 4 16 8 3 2 0 3 0 39

4 1 3 0 1 0 2 1 2 5 17 0 0 3 0 0 0 67

5 3 0 2 3 7 8 2 1 5 22 3 2 0 4 0 0

313

18 29 11 25 28 15 13 16 17 21 54 18 11 10 9 5 13

Continued

64.3 67.4 55 64.1 54.9 41.7 43.3 76.2 48.6 75 84.4 62.1 33.3 52.6 60 16.7 48.1

157

1 2 1 2 4 1 3 0 1 0 3 0 2 0 0 1 0

555

13 11 2 18 3 5 3 13 11 18 7 7 0 5 7 1 13

8 9 5 7 15 9 15 1 13 2 7 10 18 7 5 20 6

28 43 21 39 51 37 31 21 36 29 64 30 33 19 15 31 27

Austria Belgium Croatia Denmark Finland France Germany Greece Iceland Ireland Italy Netherlands Norway Portugal Spain Sweden United Kingdom Totals

0 0 0 3 0 0 0 0 1 0 0 0 0 1 0 2 0

Conflict between Intra- Number of Share Number Regular Death Early Voluntary Cabinet of parliamentary of PM parliamentary enlargement defeated by coalition parties party discretionary discretionary conflict terminations terminations cabinets election election of coalition opposition in parliament Policy Personnel conflict conflict

Country

Table 20.10 Mechanisms of cabinet termination

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

1.3

28.6

25

25.6 3.8

3.8 9.1

9.3 19.9

19.3 7.1

6.8 12.2

12.7

57.1

57.2

Note: The sum of the number of mechanisms can be higher than the number of cabinets because the mechanisms are not mutually exclusive. Non-partisan cabinets are excluded.

1.3

28.8

Conflict between Intra- Number of Share Cabinet Voluntary Death Early Number Regular parliamentary of PM parliamentary enlargement defeated by coalition parties party discretionary discretionary of conflict terminations terminations of coalition opposition in election cabinets election parliament Policy Personnel conflict conflict

Means/% of 33.4 all excl. Croatia Means/% of 32.6 all incl. Croatia

Country

Table 20.10 Continued

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

713

Relative duration (0–1)

.7

.65

.6

.55 1940

1950

1960

1970

1980

1990

2000

2010

2020

Year

Figure 20.9 Trend of cabinet duration Western Europe, 1945–2019 Notes: The figure shows the trend for the average proportion of the term of office spent in office for cabinets in Western Europe between 1945 and 2019. A relative duration of 1 means that a cabinet serves the full term of office and 0 that the cabinet terminates the first day in office. The line is a socalled loess curve and shows the underlying trend in the data.

countries. For descriptive purposes, the table shows the mean and median duration of all cabinets in the respective countries. We also show the standard deviation of the mean duration (i.e. the amount of variation around the mean value). As there are substantial differences in the maximum constitutional term of parliament (from three to five years), we also show the relative duration for the two time periods. In the table we can see that an average Italian cabinet only lasts about a year, but more recent Italian cabinets last longer. In Table 20.11 we show some of these changes but most importantly the differences between the countries. For descriptive purposes, the table shows the mean duration in days, the mean relative duration in percentages (as described earlier), and the median relative duration in percentages. The first of these measures provides an overview of the average number of days a government stays in power in the respective countries. However, as there are substantial differences in the maximum constitutional term of parliament (from three to five years), the remainder of the table shows the relative duration, which makes cross-national comparisons more straightforward. Since some countries have individual governments that do not follow the general pattern (e.g. resigned atypically quickly after they took office), we also report the median value as it is not as sensitive to outliers. In the table we can see that an average Italian cabinet only lasted about a year, but comparing the columns for two periods (1945–1989 and 1990–2019), we can

Number of cabinets 69.6 47.9 40.8 56 57.6 75.4 67 56.8 61.3 59.4 33.1 65.4 74.7 55.2 77 82.9 64 60.2 59.4

684.7 677.2

Mean relative duration (%)

906.1 561.6 474.8 681.9 443.9 588.8 806.2 628.9 738.8 936.3 368.7 806.1 806.7 673 980.7 883.7 1,000.7

Mean duration (days)

Note: Non-partisan cabinets are excluded.

Austria 28 Belgium 43 Croatia 21 Denmark 39 Finland 51 France 37 Germany 31 Greece 21 Iceland 36 Ireland 29 Italy 64 Netherlands 30 Norway 33 Portugal 19 Spain 15 Sweden 31 United 27 Kingdom Totals 555 Means excl. Croatia Means incl. Croatia

Country

Table 20.11 Cabinet duration

63.5 61.2

76.9 35.3 33.8 54 63.7 92.7 86 56.6 72 59.7 22.8 72.1 100 43.4 87.3 100 78.2

Median relative duration(%)

50.6

73.1 40.4 – 48.9 49.1 72.8 59.1 60.1 58.5 57.8 31.1 63.9 71 37.8 70.1 79.6 61.6

Mean relative duration (1945–1989)

68 65.4

64.2 65.4 40.8 70.3 76.3 78.3 88.7 55.5 66.5 62.5 36.9 68.7 82.1 70.9 81.7 92.1 68.2

Mean relative duration (1990–2019)

54

78.2 24.6 – 47.9 46.4 81.7 58.2 75.9 73.4 59.4 20.9 67.8 92.2 36.7 79.3 100 73.9

Median relative duration (1945–1989)

78.2 74.7

57.7 86.2 33.8 69.2 90.4 94.3 99.4 53.3 68.9 67.1 24.1 80.1 100 94.5 97.6 100 80.1

Median relative duration (1990–2019)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

715

see that Italian cabinets last longer than before. Excluding Italy, the least durable cabinets on average can be found in Croatia, Belgium, and Portugal. In contrast, Sweden, Spain, France, and Norway have the most durable cabinets on average. Focusing on the differences between the two time periods, only Austria, Greece, and to a lesser extent Iceland have cabinets that were more short-lived in the latter time period. In fact, regardless whether we look at the mean or the median values, cabinets in Western Europe are significantly more durable over time. But as showed previously in Figure 20.10 this is not the case since mid-2000s, as several additional countries have experienced several rather short-lived cabinets since then (in addition to the countries already mentioned—in Belgium, Ireland, and Norway). Whether this is a temporary trend is too early to say.

Concluding discussion Modes of coalition governance After having followed the full life cycle, we now turn back to the three models (or types) of coalition governance that we described in the introductory chapters. Recall that one, the ministerial government model, is based on a division of power between coalition partners and their ministers, giving individual ministers a great deal of autonomy (e.g. Laver and Shepsle 1996). A second, the coalition compromise model, stresses inter-party compromise among the government parties (Müller and Strøm 2000; Martin and Vanberg 2014). A third, the dominant prime minister model, captures when a cabinet is dominated by the leading party and in particular the PM (Dunleavy and Rhodes 1990; Müller 1994). Available research has linked most countries in Western Europe to the coalition compromise type (Müller and Strøm 2000; Thies 2001; Strøm et al. 2008; Barbieri and Vercesi 2013; Martin and Vanberg 2014). In that model, credible commitment among coalition partners requires cabinet consent. Although ministers through their ministry are likely to have considerable advantages to shape the policy process, these are mitigated via mechanisms of coordination, such as coalition contracts and policy agreements, various means of mutual control to reduce the ministers’ informational advantage (Müller and Strøm 2000; Thies 2001; Strøm et al. 2008). However, there has actually been little systematic comparison between different types of governance processes in the Western European countries. An initial analysis by Andeweg and Timmermans (2008) outlined the various mechanisms and arenas for conflict management. From here we also have the distinction between internal arenas such as the inner cabinet, mixed arenas, and external arenas, such as party summit. They also found a correlation between on the one hand the degree of party system fragmentation, comprehensive coalition

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

716

  .

agreements, and a context of negative parliamentarism and on the other hand the use of internal arenas in conflict management. Another exception is a paper by Ecker et al. (2016) where the combination of mechanisms is at the core of the analysis. They argue that each mechanism in isolation is unlikely to be effective and that political parties thus devise complementary and synergetic coalition governance mechanisms. Their results support the idea that this matters for cabinet stability and duration. We also know that the empirical record from Central Eastern Europe points to a previously unobserved and undocumented diversity between different systems (Bergman et al. 2019: 555–8). To exemplify, in that region, Hungary and Poland have represented a rather pure form of the dominant prime minister model, and Lithuania and Latvia come closer to the ministerial government model, whereas Slovenian governance is characterized by the coalition compromise model. Our volume represents a move towards a systematic analysis also of Western European cabinets.¹ The extent to which the three governance models— ministerial, coalition compromise, and dominant prime minister—apply to the political systems in Western Europe, or to particular cabinets, is one of the novel aspects of our research. Of course, neither a pure coalition compromise model nor a strict ministerial government or dominant prime minister model is something that we actually expect to find. Instead, the actual governance architecture can be more or less similar to our models. In the foregoing discussion, and in our country chapters, we have covered several indicators on coalition governance. For example, we included the dispersion of leading ministerial positions. We also discussed the institutional powers of the PM. The use of written and publicly available coalition agreements is another important indicator, and we considered if these written coalition contracts have been comprehensive or at least included a wide variety of policy issues. The governance mechanisms also included other conflict management mechanisms, such as an inner cabinet and a coalition committee, or junior ministers. In this section, we bring these multiple aspects together to highlight the coalition governance patterns that are identifiable at the country level across Western Europe. Albeit tentative, the analysis strives to be systematic and transparent. A combined analysis of these indicators provides useful information on which model a case approximates. By such an analysis, we can compare the main characteristics of the 17 governance systems that we cover. In this, we build on the analysis of governance in Central Eastern Europe in Bergman et al. (2019: 553–8). Turning to Table 20.12 and the combined practice of different models of coalition governance, seven indicators help identify how well the three different models

¹ We have to leave some aspects for future research. For example, we do not have systematic data for aspects of the ‘shadowing’ in parliamentary committees, for example via a committee chair from a different party (Carroll and Cox 2012).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

717

Table 20.12 Three models and sixteen coalition governance structures Model and country Compromise AT GE IR NL NO PM DK UK Ministerial FI Mixed BE FR GR HR IS IT PT SE

Compromise

PM

Ministerial

5 5 5 5 6

3 3 3 1 2

3 1 1 2 2

3 2

5 6

3 2

3

3

5

5 2 2 3 4 4 3 4

1 4 4 3 4 2 2 4

5 2 4 3 4 4 3 2

Source: Table 20.13.

match our sample of countries. During the observation period, Spain did not have a single formal cabinet coalition, which leaves us with 16 coalition systems to compare. From Table 20.3 we can identify the countries with the highest PM Powers (with scores of 5 or more of the 7). These are Croatia, Denmark, France, Germany, Ireland, Italy, and Portugal. Likewise, in Table 20.7, we can identify the countries in which party summits are used to manage conflicts between coalition partners. These are the countries with the least hierarchical procedures in favour of the PM and the inner cabinet. In the same table (Table 20.7), we can identify the countries where at least some coalitions use junior ministers to help manage their coalition. When we look back at Table 20.8, we can identify where parliamentary discipline is less strict. Belgium and the Netherlands stand out for a particular reason. Here, party discipline can at times, and in some coalitions, be quite strict with regard to legislative proposals but often lax when it comes to other parliamentary discipline. Portugal too has had patterns of discipline with regard to legislative matters but less so towards other matters. Nevertheless, taken together, there is a qualitative difference between the eight countries that are classified as having less strict discipline and the other countries in our sample.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

718

  .

To our help we also have three figures. Figure 20.5 records which are the countries that unusually have lengthy and detailed coalition agreements. Figure 20.7 identifies those countries that on average include more policy guidelines in the coalition agreements, before or just at the beginning of their tenure. For a final indicator, we return to Figure 20.4, which records if there is a practice that one party, for example the PM party, also controls other important ministries, in this case finance and foreign affairs. In Table 20.13, we combine these indicators into a measure of the goodness of fit between each of the three models and our 16 countries. We use a binary coding for each variable. This masks some variation, for example with regard to parliamentary discipline and coalition discipline in other parliamentary matters (e.g. debates). However, in most countries, the patterns are quite stable over time. In order to compare the information (Table 20.13) against our three models, three sets of combinations can be identified. The coalition compromise model is one that can be understood as having the following combination of the seven indicators: a relatively low PM Powers index does not put primacy on the PM as the undisputed leader; party summits sometimes move decision-making outside of the cabinet (Andeweg and Timmermanns 2008); likewise, parliamentary discipline is not strict in the legislature, and junior ministers help manage conflict. In this practice, coalition agreements are both lengthy and detailed and provide guidance for both the cabinets as a whole and for individual ministers. Finally, there is very little power concentration in the sense that one party by itself controls several of the most important ministries. As for the PM dominant model, in countries that fit the model, a number of formal PM Powers exist, conflicts are largely managed among party leaders in the inner cabinet, and party discipline in parliament is tight (see also Vercesi 2020: 446–9). There is no heavy emphasis on junior ministers or they are not allowed, and to provide the leaders with flexibility, coalition agreements are relatively short and lack detail. In addition, the PM party is typically dominant also in that it controls several of the most important ministries. Finally, the ministerial government model suggests that a system that is a perfect match has low PM Powers; occasional party summits solve conflicts between coalition partners; there is low party discipline in parliament, which does not use watchdog junior ministers. These all gives the individual ministers considerable leeway in acting on their own (see also Andeweg 2020: 465). The freedom of manoeuvre for individual ministers is increased by short and not very detailed coalition agreements, which makes it difficult to monitor compliance. Nor is there a single party that controls several of the most important portfolios. In the three right-hand columns in Table 20.13, which are the numbers shown in table 20.12, we have, counted the number of times a country matches each model. This provides us with a measure of fit in terms of coordinates. These coordinates (e.g. for Germany: 5, 3 and 1) are used to place the countries in four

Out of 7 PM Powers: Score 5 or above? Table 20.3 (last row per country)

No No Yes Yes No Yes Yes Yes No Yes Yes No No Yes Yes

Country (last row)

Austria Belgium Croatia Denmark Finland France Germany Greece Iceland Ireland Italy The Netherlands Norway Portugal Spain

No Yes Yes No No No No Yes No Yes Yes No No Yes -

PS is a governance mechanism? Table 20.7

Yes Yes Yes No No Yes Yes Yes No Yes Yes Yes Yes Yes -

Junior ministers are used? Table 20.7

No Yes Yes No Yes Yes No Yes No No Yes Yes No Yes -

Parliamentary discipline often less strict Table 20.8

Yes Yes Yes Yes No No Yes No No Yes Yes Yes Yes Yes -

Above average length of coalition agreements? Figure 20.5

No No No No Yes Yes Yes No Yes Yes No Yes Yes No -

Above average policy content (%) in agreement? Figure 20.7

Table 20.13 Coalition governance models, empirical manifestations at the country level

No No Yes No No Yes No Yes No Yes No No No No -

Power concentration in cabinet (PM, finance and foreign affairs) above country mean?Figure 20.4

5 5 3 3 3 2 5 2 4 5 4 5 6 3 -

Summary points:No. of Coalition compromise model matches

3 1 3 5 3 4 3 4 4 3 2 1 2 2 -

No. of PM dominance model matches

3 5 3 3 5 2 1 4 4 1 4 2 2 3 Continued

No. of Ministerial governance model matches

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

No

Yes

Yes

No

No

No

Yes Yes Yes

Junior ministers are used? Table 20.7

Yes

No

No No No

Parliamentary discipline often less strict Table 20.8

No

No

No No Yes

Above average length of coalition agreements? Figure 20.5

No

No

Yes No Yes

Above average policy content (%) in agreement? Figure 20.7

No

Yes

Yes Yes No

Power concentration in cabinet (PM, finance and foreign affairs) above country mean?Figure 20.4

Note: Bold summary points indicate a close match or a particularly distant relationship with the pure model.

No No Yes

No Yes No

Sweden U.K. Coalition compromise model PM dominance model Ministerial governance model

PS is a governance mechanism? Table 20.7

Out of 7 PM Powers: Score 5 or above? Table 20.3 (last row per country)

Country (last row)

Table 20.13 Continued

4 2

Summary points:No. of Coalition compromise model matches

4 6

No. of PM dominance model matches

2 2

No. of Ministerial governance model matches

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

721

categories, three corresponding to our three basic models and the fourth is a category that we label as ‘mixed’. Table 20.12 groups our 16 countries after how well they match each one of the three models by the structure of cabinet governance. Again, this is not exactly the same as the governance practice. Depending on the party system, or perhaps a very strong long-term PM, or majority/minority status of the government, the actual governance patterns might seem different. For example, we have accounts from the country chapters that emphasize the strong position of the PM in some countries (e.g. Croatia, Denmark, France, and Portugal). Others stress the coalition compromise nature of coalition governance relations (Finland, Italy, and Sweden). Iceland, on the other hand, is one of the few countries in these accounts that seem to be dominated by ministerial government. In an attempt to capture the variation in governance structure, Table 20.12 suggests a somewhat different pattern. First, we have five countries that all load high on the compromise model, Austria, Norway, Germany, Ireland, and the Netherlands. As high, we count a score of 5 or 6. These countries also have lower values for indicators belonging to the PM and ministerial models. Of these, it is perhaps Germany, with the strong chancellor position, for which the compromise placement is a bit surprising. However, a tradition of grand coalitions might over time have manifested itself in more elaborate governance structures than are in place in other countries. The other countries in this category seem to fit well with the model, both here and in the country chapters. The two countries where the PM has a dominant role are Denmark and the United Kingdom. At least in Denmark, there are also important elements of compromise and ministerial independence, but the overall finding is in line with the strong position of the PM. Until the UK Fixed-term Parliaments Act 2011, in our sample of countries, there were also the only two PMs that in practice on their own held the power to dissolve parliament and call an extra election. The table also shows that Belgium and the Netherlands have governance structures that puts the PM in a weak position (a score of 1). As mentioned, the Bergman et al. (2019) analysis of Central Eastern Europe found that ministerial government seems to exist in a relatively pure form in Lithuania and Latvia. In Western Europe, it is ministers in Finland and to a lesser degree Greece, Iceland, and Italy that have the most leeway, that is are the least constrained by structures favouring the PM. We place Belgium in the ‘mixed’ category. Interestingly, Belgium represents a case of both high coordination and simultaneously high autonomy for the individual ministers. As the country chapter and the low score on the PM dominated model indicates, there is perhaps instead little centralized decision-making. Notably, Germany and Ireland, two coalition compromise countries, have very little of ministerial government. That is, ministers work in structures that provide little autonomy, because they are constrained either by the PM or by an elaborate set of coalition governance mechanisms.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

722

  .

Other country structures have a less pronounced affinity with the three models. Or rather, they display traits from all three models. In a similar fashion, the country chapter on Austria speaks of governance in terms of ‘constrained ministerial government’. Thus even in the relatively clear cases, there are features of other models. Still, France and Greece lean more towards the PM model than the compromise one, and ministers in Greece do have some autonomy. Italy combines some ministerial autonomy with compromise rather than PM dominance. Portugal does not seem to lean toward either of the models, whereas Sweden leans in favour of coalition compromise with a fairly strong position for the PM. The true ‘mixed’ cases are Croatia and Iceland. Here, a simultaneous combination of all three models of governance can be observed.

A changing coalition life cycle? As mentioned in this chapter, many of the Western European party systems have experienced significant changes in the past decades, for example with increasing party system fragmentation and a rise of populist radical right-wing parties. In Chapter 2 of this volume, we discussed some potential implications of these party system changes, suggesting that such party system changes are likely to influence various stages of the coalition life cycle. For example, we would expect these party system changes to result in longer bargaining duration, and in a higher cabinet instability, or shorter cabinet duration. In this concluding chapter we have presented some initial empirical results evaluating such hypotheses. The analyses presented in this chapter are descriptive, and only provide some hints as to whether we see an impact of party system change, and more sophisticated multivariate analyses are clearly needed before we can say anything conclusive about the impact of party system change on the coalition life cycle—the data provided here by our country experts should allow for such analyses in future research. When analysing government formation, we have here shown that there is a clear trend towards longer bargaining duration in the Western European countries, something which could be attributed to the party system changes described earlier, resulting in higher levels of bargaining complexity and uncertainty for the parties negotiating a coalition deal. At the government formation stage, we also see that there is a decreasing share of single-party minority cabinets and a more frequent occurrence of minimal winning and minority coalitions. Political parties that used to have a strategic advantage because of their median position in policy space are still common coalition partners, but slightly less often end up in cabinet. However, larger parties seem to hold on to their bargaining advantage. When they form coalitions, political parties tend to write increasingly long coalition agreements, but the share of coalitions that have such agreements is in recent years

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

723

somewhat declining. The increasing fragmentation of the party system and a loss of dominance of larger Social Democratic and Christian Democratic parties may account for such changes at the first stage in the coalition life cycle. If we instead look at the last stage of the coalition life cycle, that is on the survival of governments, we have here shown that that there is significant variation across countries in how long cabinets last but that there is little evidence of any clear trends over time. The party system changes described here, with increased fragmentation, and a rise of populist parties could potentially lead to more government crises, and a lower cabinet survival rate, but we do not find much evidence that this has yet occurred in the Western European countries. There is a trend of declining duration in the most recent data, but during the preceding decades the trend was rather that cabinets stayed longer in power. However, as mentioned earlier, it is too early to say if this is a temporary trend, and future studies should analyse whether the increasing bargaining complexity of the more fragmented and polarized political systems of Western Europe comes with a higher ‘mortality rate’ of the cabinets that form, resulting in shorter coalition life cycles.

References Andeweg, Rudy (2020). ‘Parties and Executives in Parliamentary Systems—From Party Government to Party Governance’. In Rudy Andeweg, Robert Elgie, Ludger Helms, Juliet Kaarbo, and Ferdinand Müller-Rommel (eds), The Oxford Handbook of Political Executives. Oxford: Oxford University Press, 460–80. Andeweg, Rudy, and Timmermans, Arco (2008). ‘Conflict Management in Coalition Government’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 269–300. André, Audrey, Sam Depauw, and Kris Deschouwer (2015). ‘Changing Investiture Rules in Belgium’. In Bjørn Erik Rasch, Shane Martin, and José Antonio Ceibub (eds), Parliaments and Government Formation: Unpacking Investiture Rules. Oxford: Oxford University Press, 49–66. Bäck, Hanna and Royce Carroll (2020). ‘The Distribution of Ministerial Posts in Parliamentary Systems’. In Rudy Andeweg, Robert Elgie, Ludger Helms, Juliet Kaarbo, and Ferdinand Müller-Rommel (eds), The Oxford Handbook of Political Executives. Oxford: Oxford University Press, 314–35. Bäck, Hanna, and Patrick Dumont (2008). ‘Making the First Move. A Two-Stage Analysis of the Role of Formateurs in Parliamentary Government Formation’. Public Choice, 135(3–4): 353–73.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

724

  .

Bäck, Hanna, Wolfgang C. Müller, and Benjamin Nyblade (2017). ‘Multiparty Government and Economic Policy-Making: Coalition Agreements, Prime Ministerial Power and Spending in Western European Cabinets’. Public Choice, 170(1–2): 33–62. Barbieri, Cristina, and Michelangelo Vercesi (2013). ‘The Cabinet: A Viable Definition and Its Composition in View of a Comparative Analysis’. Government and Opposition, 48(4): 526–47. Bergman, Torbjörn (1993). ‘Formation Rules and Minority Governments’. European Journal of Political Research, 23(1): 55–66. Bergman, Torbjörn, Gabriella Ilonszki, and Wolfgang C. Müller (eds) (2019). Coalition Governance in Central Eastern Europe. Oxford: Oxford University Press. Bergman, Torbjörn, Svante Ersson, and Johan Hellström (2015). ‘Government Formation and Breakdown in Western and Central Eastern Europe’. Comparative European Politics, 13(3): 345–75. Bowler, Shaun, Thomas Bräuninger, Marc Debus, and Indridi H. Indridason (2016). ‘Let’s Just Agree to Disagree: Dispute Resolution Mechanisms in Coalition Agreements’. The Journal of Politics 78(4): 1264–78. Browne, Eric C., and Mark N. Franklin (1973). ‘Aspects of Coalition Payoffs in European Parliamentary Democracies’. American Political Science Review, 67(2): 453–69. Carroll, Royce, and Gary W. Cox (2012). ‘Monitoring Partners in Coalition Governments’. Comparative Political Studies, 45(2): 220–36. De Swaan, Abram (1973). Coalition Theories and Cabinet Formations. Amsterdam: Elsevier. Döring, Holger, and Johan Hellström (2013). ‘Who Gets into Government? Coalition Formation in European Democracies’. West European Politics, 36(4): 683–703. Doyle, David (2020). ‘Measuring Presidential and Prime Ministerial Power’. In Rudy Andeweg, Robert Elgie, Ludger Helms, Juliet Kaarbo, and Ferdinand MüllerRommel (eds), The Oxford Handbook of Political Executives. Oxford: Oxford University Press, 382–401. Dunleavy, Patrick, and R. A. W. Rhodes (1990). ‘Core Executive Studies in Britain’. Public Administration, 68(1): 3–28. Ecker, Alejandro, Thomas Meyer, and Wolfgang C. Müller (2016). ‘The Architecture of Coalition Governance’. Paper presented at the ECPR General Conference, Charles University, Prague, 7 September to 10 September. Indridason, Indridi H., and Gunnar H. Kristinsson (2013). ‘Making Words Count: Coalition Agreements and Cabinet Management’. European Journal of Political Research, 52(6): 822–46. Kayser, Mark Andreas (2005). ‘Who Surfs, Who Manipulates? The Determinants of Opportunistic Election Timing and Electorally Motivated Economic Intervention’. American Political Science Review, 99(1): 17–27.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

     

725

Klüver, Heike, and Hanna Bäck (2019). ‘Coalition Agreements, Issue Coverage and Cabinet Governance’. Comparative Political Studies, 52: 1995–2031. Klüver, Heike, Hanna Bäck, and Svenja Krauss (2021). ‘Coalition Agreements as Control Devices: Coalition Governance in Western and Eastern Europe’. Unpublished book manuscript. Laver, Michael, and Kenneth A. Shepsle (1996). Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Laver, Michael, and Norman Schofield (1990). Multiparty Government. The Politics of Coalition in Europe. Oxford: Oxford University Press. Lijphart, Arend (2012). Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, 2nd edn. New Haven and London: Yale University Press. Martin, Lanny W., and Georg Vanberg (2011). Parliaments and Coalitions: The Role of Legislative Institutions in Multiparty Governance. Oxford: Oxford University Press. Martin, Lanny W., and Georg Vanberg (2014). ‘Parties and Policymaking in Multiparty Governments: The Legislative Median, Ministerial Autonomy, and the Coalition Compromise’. American Journal of Political Science, 58(4): 979–96. Mitchell, Paul, and Benjamin Nyblade (2008). ‘Government Formation and Cabinet Type in Parliamentary Democracies’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 201–36. Müller, Wolfgang C. (1994). ‘Models of Government and the Austrian Cabinet’. In Michael Laver and Kenneth A. Shepsle (eds), Cabinet Ministers and Parliamentary Government. Cambridge: Cambridge University Press, 15–34. Müller, Wolfgang C., and Thomas M. Meyer (2010). ‘Meeting the Challenges of Representation and Accountability in Multi-party Governments’. West European Politics, 33(5): 1065–92. Müller, Wolfgang C., and Kaare Strøm (eds) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Müller Wolfgang C., and Kaare Strøm (2008). ‘Coalition Agreements and Cabinet Governance’. In Kaare Strøm, Wolfgang C. Müller and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe, Oxford: Oxford University Press, 159–99. Sjölin, Mats (1993). Coalition Politics and Parliamentary Power. Lund: Lund University Press. Strøm, Kaare (1990). Minority Government and Majority Rule. Cambridge: Cambridge University Press. Strøm, Kaare, Ian Budge, and Michael J. Laver (1994). ‘Constraints on Cabinet Formation in Parliamentary Democracies’. American Journal of Political Science, 38(2): 303–35.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

726

  .

Strøm, Kaare, Wolfgang C. Müller, and Daniel Markham Smith (2010). ‘Parliamentary Control of Coalition Governments’. Annual Review of Political Science, 13: 517–35. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (2008). Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press. Taagepera, Rein, and Allan Sikk (2010). ‘Parsimonious Model for Predicting Mean Cabinet Duration on the Basis of Electoral System’. Party Politics, 16(2): 261–81. Thies, Michael F. (2001). ‘Keeping Tabs on Partners: The Logic of Delegation in Coalition Governments’. American Journal of Political Science, 45(3): 580–98. Thürk Maria, Johan Hellström, and Holger Döring (2020). ‘Intuitional Constraints on Cabinet Formation: Veto Points and Party System Dynamics’. European Journal of Political Research, 60(2): 295–316. Timmermans, Arco (2003). High Politics in the Low Countries. An Empirical Study of Coalition Agreements in Belgium and the Netherlands. Aldershot: Ashgate. Vercesi, Michelangelo (2020). ‘Cabinet Decision-Making in Parliamentary Systems’. In Rudy Andeweg, Robert Elgie, Ludger Helms, Juliet Kaarbo, and Ferdinand MüllerRommel (eds), The Oxford Handbook of Political Executives. Oxford: Oxford University Press, 438–59. Verzichelli, Luca (2008). ‘Portfolio Allocation’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 237–68. Walther, Daniel, and Johan Hellström (2019). ‘The Verdict in the Polls: How Government Stability Is Affected by Popular Support’. West European Politics, 42(3): 593–617. Warwick, Paul V., and James N. Druckman (2001). ‘Portfolio Salience and the Proportionality of Payoffs in Coalition Governments’. British Journal of Political Science, 31(4): 627–49. Warwick, Paul V., and James N. Druckman (2006). ‘The Portfolio Allocation Paradox: An Investigation into the Nature of a Very Strong but Puzzling Relationship’. European Journal of Political Research, 45(4): 635–65. Zubek, Radoslaw (2015). ‘Coalition Government and Committee Power’. West European Politics, 38(5): 1020–41.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

APPENDIX

On Definition and Measurement Torbjörn Bergman, Hanna Bäck, Johan Hellström, and Jonas Lindahl This appendix introduces our empirical study of the coalition life cycle.¹ The structure of each country chapter is based on six core tables, each one capturing a major stage in the coalition life cycle.² In this chapter, we use the tables from the case of Sweden to illustrate how the tables in the different country-specific chapters are structured in terms of variables, columns, and cells. In the book chapters, the tables help structure the analysis and provide a wealth of information and details of the coalition cycle. They also ensure that the chapters are comparable as each of the contributors for the respective chapter investigates a set of common indicators. Nevertheless, the book chapters and this appendix only present a sample of the data and information that we have collected for this research project. For reasons of space, we have had to limit the presentation in the book chapters in terms of time periods and variables.³ Nonetheless. this appendix presents details that are important for readers who are not familiar with the definitions used in coalition research or interested in how particular variables were measured.

Governments, institutions, and party systems In the country chapters, the contributors provide historical and cultural information that can help the reader understand coalition politics in the specific country, while using strict cross-national criteria. To begin with, note our definition of our main unit of analysis, a ‘cabinet’, which is the unit of analysis for our tables (Table A.1b being the sole exception). A new cabinet is counted every time there is (a) a change in party composition, (b) a change in prime minister (PM), or (c) a general election, and ends when the cabinet officially resigns or (if sooner) on the date of the general election that ends its tenure (Müller and Strøm 2000: 12). Our strict cross-national counting rule does not include cabinets that, for example, hand in resignations but never resign. This practice means that we do not always follow particular national counting rules in each and every respect. Instead we aim for cross-national validity and comparability. Also note our common starting point: we begin with the first cabinet that emerged from the first democratically elected parliament at the end of the Second World War or the first during the current democratic regime. (This means 1944 in the cases of Iceland and Ireland and, for example, from the beginning of the Fifth Republic, 1959, for France.) We include all cabinets that formed all the way up through January 2019. For the cabinets that were in ¹ This appendix builds on Müller and Strøm (2000), Strøm et al. (2008), and Bergman et al. (2019) volumes on coalitions. Our structured collaboration approach is outlined in Bergman et al. (2005). ² Each country expert has also worked with a table on the institutional make-up of each country. As there are not much within-country variations in these tables, these data are only presented in the introductory and concluding chapters. ³ An online appendix with complete tables with all of the time periods and variables is available at www.erdda.org. Here we also provide an open-access full data set and a detailed codebook for empirical analyses.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

728

  .

power at that date we do not of course have certain types of data, such as information on their termination. As the first day of the cabinet, we count the first day in cabinet life according to official criteria (sworn-in, passed parliamentary inauguration vote, etc.). More specifically, the cabinet ‘date-in’ variable is defined as the date that the PM or cabinet was appointed by the head of state, alternatively the date of the investiture vote in parliament or, if that is not applicable, the date of the general election. The end date is the date that the cabinet resigned or, if it is constitutionally required, when its resignation is formally accepted by the head of state or parliament or the date of a general election, whichever comes first. In two-round elections, we use the second date to determine the date-out. To determine the formal date of resignation, we have looked for the date of the acceptance of the resignation by the head of state. If no such date can be found, we record the date of a vote of no confidence. These strict criteria allow for cross-national comparisons. Note that our cross-national definition of the date of resignation might not be the one that is officially recognized in the national tradition as the event or date that ended the cabinet. The official records of individual countries are often not very good for exact comparative purposes as they are based on different criteria. Our country chapters build on the following template: all chapters start with a brief discussion of the institutional (constitutional) set-up of each country and the major changes that have occurred in the party system since the late 1980s and early 1990s. We provide a shorthand history on (a) the number and nature of relevant policy dimensions and (b) important party placements on these dimensions (this is further discussed later). While necessary for an understanding of the parties’ coalition behaviour, the historic section will be relatively short.

Institutional background Each chapter outlines the institutional context for coalition politics. The account summarizes the main developments and the main institutions of representative democracy prior to the 1990s. For the period after that, major changes in electoral systems, parliamentary design (e.g. unicameralism or bicameralism), and the powers of the head of state are presented in more detail. Given that we are concerned with national government, each chapter also introduces the core government institutions and the government cabinet structure. This section also provides information on the rules for government formation and dissolution as well as the rules for early dissolution of parliament. The following are the core institutions that we discuss in terms of constraints, or enablers, of coalition formation (cf. Laver and Schofield 1990; Strøm et al. 1994; Müller and Strøm 2000: 565–9). Semi-presidentialism: If applicable, here we present the role of the head of state in government formation and government termination. We make clear whether the respective country has a ‘pure’ parliamentary system or if it is a parliamentary democracy with semipresidential features. We only count as semi-presidential those constitutions that (a) provide for a directly (popularly) elected president and (b) allow the president a direct influence on coalition formation and/or termination (Elgie 1999; Strøm et al. 2003: 27). Note that we are interested in regime types and not in the behaviour of the head of state in extraordinary circumstances. Bicameralism: Here we present the unicameral or bicameral character of each country’s parliament. In particular, we discuss the powers of the second chamber to influence the fate

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

   

729

of the government by blocking or delaying the government’s policies. In our sample of 17 countries, we only include those bicameral systems where both chambers vote on the inauguration and the dismissal of the government. In the context of Table A.1a, we present the seat distribution for the upper chamber in Belgium, Italy, and Spain in 1977 and Sweden before 1971 only. We consider these to be the only cases of strong (or pure) bicameralism in our sample of countries. Investiture rules: Investiture rules are important for the coalition game, and the question is, are votes of inauguration required and which are the decision rules under which they are held (e.g. relative or absolute majority)? We detail the decision rules on investiture. However, sometimes these decision rules involve some judgement calls. For example, Ireland has a two-step procedure in the investiture vote, where first the PM and then the list of cabinet ministers are voted on in parliament. This makes it more difficult to say when the cabinet passed the investiture vote (the first or second date). In these difficult cases we have relied on the criterion that we consider the decisive vote to be the one after which the government, in effect, takes office (e.g. it can propose a huge tax reform or even start a military conflict). This in turn has consequences for the ‘length of the formation period’ and other indicators down the coalition life cycle. As regards the formation process, sometimes constitutional rules designate a procedure that uses informateurs and formateurs (Strøm 1990: 25–7), whereas other systems rely on ‘freestyle bargaining’ among the political parties (Bergman 2000). We are also interested in whether or not the formal rules state that a particular party (e.g. the largest party in parliament) will be asked to begin the process of formation, that is be designated as the first formateur party. Here, we also want to be particularly careful to distinguish what is a legal rule from what is a constitutional convention (perhaps upheld by the courts), and thus in turn from a norm for behaviour that is usually observed but can more easily be ignored. Vote of (no) confidence: As shown by Sieberer (2015) the general trend is that countries with demanding investiture requirements (i.e. where an absolute or relative majority support in parliament is required) also have higher requirements for no confidence motions to pass, and the other way around. Here we discuss ‘no confidence’ vote (when the parliament initiates a vote on the cabinet), confidence votes (when the cabinet/PM initiates a vote to prove that it has the support of the parliament), and whether or not the constitution calls for a (German type of) constructive vote of no confidence. Parliamentary dissolution powers: We also include a discussion of parliamentary dissolution (early election) rules. The power to dismiss the parliament and call an early (or extra) election has been shown to be an integral part of coalition politics in some countries (Strøm and Swindle 2002), but not all countries allow for such dissolutions, or they have constitutional rules that effectively prohibit the use of this strategic instrument, and we want to account for that variation. Cabinet rules/PM Powers: Since we want to highlight the coalition governance part of the life cycle, we include a discussion of what the constitution and other legal texts have to say about cabinet decision-making. Typically, constitutions are relatively ‘silent’ on such matters. As we introduce the basic features of the rules and procedures of cabinet decision-making, we also detail what they contain. While we do not present a table on this, we are interested in the powers of the PM over the cabinet agenda and his or her authority over the other ministers (Strøm et al. 2003). How these formal rules actually play out in different cabinets and countries is discussed in the governance section of each chapter, and we return to this in the concluding chapter. In Chapter 20, with regard to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

730

  .

PM Powers, we produce an index based on seven different forms of power that the PM may or may not have.⁴

Dimensions of competition, party alignments, and median parties After having provided a brief account of the most important and basic historical and institutional contexts, we next turn to the basic actors, the political parties. Each chapter begins by outlining the national party system under the era when it was were primarily thought of as ‘frozen’ (Lipset and Rokkan 1967), which in some countries goes back to the early 1990s. From that starting point, we discuss newcomers and major parties that have left the scene. To understand coalition politics, one should consider the number of political parties in a political system, their relative size, and the main dimensions of competition among them. We use Table A.1a to structure the discussion on cabinets and party systems. This table is when applicable presented in two forms (a and b). Table A.1a contains the basic data on cabinets formed and Table A.1b covers the existence of electoral alliances and pre-electoral coalitions, described in more detail later. In those cases where the cabinet is also directly responsible to the upper chamber, we include data on the upper chamber in Table A.1a as well. The seat distribution is of course essential as it provides the parties with bargaining weight in coalition negotiations. We record the total number of seats of the parliamentary parties when the cabinet forms.⁵ We also identify the seat share based on the number of seats held by the cabinet parties after each general election, presented as a percentage of the total seats in parliament in Table A.1a. Changes during a sitting parliament will result from defections by individual members of parliament (MPs) or from party mergers. Important such changes during the constitutional inter-election period (CIEP) are noted in the text or in the table notes. In this study, the constituent unit when discussing political ‘parties’ and their seats is the parliamentary party group (PPG). In our counting of political parties and calculations for Table A.1a, a party is recorded for the first time when it gains three seats; otherwise ‘groups’ with one or two MP seats fall in the category of ‘Other’. Once a party has gained three seats and thus has been registered in the table, any seat in future elections will always be coded for that party even if it just gains one or two seats. Furthermore, once a party has gained the sufficient number of seats to be coded as a distinct party, it is also coded retroactively. However, we allow for a few exceptions from these general rules. Parties with cabinet posts are always recorded even if they have only one seat. We also allow for the experience, good judgement, and skills of the country expert(s) to decide on whether a party is an important actor to record—if a very small party or a single MP had an important impact on the party system, whether this is due to its impact on the left–right placement or relevant second dimension, or on coalition formation. The number of parties that have fulfilled the above criteria for a given cabinet is presented in the number of parties column in Table A.1a. Thus, this column does not include any of the parties recorded in the ‘Other’ category. ⁴ A single point, to a maximum of seven, is added based on the presence of the following attributes: the PM has the right to appoint ministers, the PM has the right to dismiss ministers, ministers can only be removed by parliament if parliament dismisses the PM/full cabinet, the PM has the right to decide ministry jurisdiction, the PM has steering or co-coordinating rights vis-à-vis ministers, and the PM has control over cabinet agenda. The seventh mechanism is when the PM has a bureaucratic structure in his or her own office which is designed to monitor other ministers and ministries. ⁵ A full account of seat shares for the entire period of study is only reproduced in the online appendix, as well as in the data set available at http://www.erdda.org.

44.7 48.7 46.1 46.1 37.5 41.3 51 49.6 39.5 33.2

349 349 349 349 349 349 349 349 349 349

6 7 7 7 7 7 7 8 8 8

3.67 4.19 3.50 3.50 4.29 4.23 4.15 4.54 4.99 5.63

S C S S MP S KD SD SD SD

C, L

V, MP V, MP

Median party Formal support in the first parties policy dimension

Cabinet types: maj = Single-party majority cabinet; min = Minority cabinet (both single-party and coalition cabinets); mwc = minimal winning coalition.

The number of parties only includes parties that have held more than two seats in parliament when a cabinet has formed.

The first policy dimension is economic left–right.

Median parties are retrieved from the Comparative Parliamentary Democracy Data Archive (http://www.erdda.se), gathered for Müller and Strøm (2000) and Strøm et al. (2003), for the 1945–1998 period. Median parties for the subsequent period are based on data from Polk et al. (2017) and Bakker et al. (2015).

min min min min min min mwc min min Min

Type Cabinet Number of Number of ENP, Party parties in parliament strength seats in composition of parliament parliament (first chamber cabinet (%) of cabinet in parentheses)

1988-09-18 S 1991-09-15 M, C, KD, L 1994-09-18 S S 1998-09-20 S 2002-09-15 S 2006-09-17 M, C, KD, L 2010-09-19 M, L, C, KD 2014-09-14 S, MP 2018-09-09 S, MP

Election date

For a list of parties, consult the chapter appendix.

Notes:

1988-09-18 1991-10-04 1994-10-07 1996-03-22 1998-09-20 2002-09-15 2006-10-06 2010-09-19 2014-10-03 2019-01-21

22 23 24 25 26 27 28 29 30 31

Carlsson II Bildt Carlsson III Persson I Persson II Persson III Reinfeldt I Reinfeldt II Löfven I Löfven II

Date in

Cabinet Cabinet number

Table A.1a Swedish cabinets since 1988

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

732

  .

In Table A.1a, in addition to listing the parties that are included in the cabinet (party composition), we include the election date for post-electoral cabinets. This distinction allows for an analysis of cabinets that form immediately after a general election and those who form further along the CIEP. Several scholars have shown that bargaining complexity, such as a high number of parties and ideological polarization in the parliament, can lead to longer government formation periods (see e.g. De Winter and Dumont 2008). This is particularly true in the period immediately following a new election. Coalitions that form during the CIEP have often more information (or fewer options), which can lead to shorter formation periods. However, as the number of parties is a crucial ‘structural attribute’ of the party system, the parliament, and the resulting cabinet (Strøm et al. 1988), it is important to be clear about how and when we count a political party and how we determine how many seats it has. We follow closely in the footsteps of the information provided in the Müller and Strøm (2000) volume on coalition governments. As in that book, we have to make a difficult choice and look at the seats distribution at the formation of the cabinet rather than after any party switches one or two years later into the tenure of the parliament. This ‘snap-shot’ principle is necessary or the tables would not be possible to present in a brief chapter format. We apply the same (snapshot) principle to all our tables, as is the standard in the field of study.

Conflict dimensions and the median party The important question of the main dimension(s) of policy competition is even less straightforward to answer. In doing so, we face a choice between various existing crossnational measures based on expert surveys (e.g. Benoit and Laver 2006 or the expert surveys conducted out of Chapel Hill, North Carolina, e.g. Bakker et al. 2015; Polk et al. 2017), party manifestos (Klingemann et al. 2006), elite interviews (Kitschelt et al. 1999), and mass surveys. All these measures have their distinct advantages and disadvantages and they relate to one or several measurement points (see e.g. Hooghe et al. 2010). Here, we have jointly decided to build on the Chapel Hill Expert Survey (CHES) data.⁶ Also, for parties that are not included in each CHES data set, other sources have been used, which are outlined in the respective country chapters. The Chapel Hill data do not cover all parties. Many PPGs, mostly smaller, are simply missing from the data. When there is no placement for a particular party available in the data, we have asked our country experts to place the remaining parties on the ordinal alignment, thus complementing the Chapel Hill survey data. We have chosen this procedure because we want the party alignment to be based on expert judgement rather than common wisdom. When placing the political parties that are missing from the Chapel Hill surveys but still have had parliamentary seats, we have asked our contributors to use data from the best possible sources—expert surveys, manifesto or programme analysis, elite interviews, and mass surveys—to provide a full picture.⁷ Substantively we are interested in

⁶ The CHES data sets for 1999, 2002, 2006, 2010, 2014, and, for a subset of our included countries, 2017 are available online at: https: /www.chesdata.eu/. ⁷ For all this data, and for the purpose of coalition studies, we should be aware of the circularity problem: actual coalition behaviour may impact on the measures. Experts might think that parties are close in policy space because they cooperate, which does not necessarily follow. Different parties might have different motives for their behaviour. This argument has been held against expert surveys in particular. But, again, all measures have their distinct problems and our collective judgement is that the Chapel Hill dimensions cover what we want to capture fairly well.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

   

733

policy proximity: to what extent do the parties differ in their preferred public policies, as this is important for the government formation process and cabinets durability.⁸ Our placement of the political parties on various conflict dimensions, including the socio-economic left–right dimension, is useful for the respective chapter and these placements can be very useful for further comparative studies. We will also use this to determine the median party on both dimensions (for reasons of space, only the median party in the first dimension is presented in Table A.1a). As we stress in this book, the ‘median party’ is a very important concept in coalition analyses. While we use the same left–right dimension for all 17 countries in the first dimension (with a single cabinet in the United Kingdom as the sole exception), they can and do differ on the second dimension. Our contributors noted which the most salient dimension is after the left–right dimension, There is of course a possibility of a third salient dimension as well. Thus, in cases where another dimension of conflict would provide a more accurate picture of party competition in any of the 17 countries, parties have been aligned on a third dimension.⁹ Small parties that have never gained more than two seats or have otherwise been noteworthy for inclusion as distinct parties in our data pose a potential issue for identifying the median party. Often these parties are difficult to position on an ideological space, a difficulty that becomes even more acute when there are large numbers of independent parliamentarians that do not belong to any party groups in parliament, with Ireland being a prime example. There are a number of possible approaches—such as positioning all such parliamentarians to the far ‘right’ or far ‘left’ or distributing them randomly on any given scale to rectify this problem (see Bergman et al. 2019), but here we have opted to exclude these parties and/or parliamentarians from the calculation. Table A.1a also contains the ‘effective number’ of legislative parties, which is a measure of party system fragmentation (Laakso and Taagepera 1979). This measure takes both the number of parties and their relative size into account (see e.g. Lijphart 2012: 63–7). If the index increases, this means that the party system has become more fragmented and complex. We base the index on the seat distribution available in the online appendix. The category of ‘Other’ parties is counted as a single party for these calculations. Following up on the discussion on institutional arrangements, in those countries where the government needs a majority in both houses of parliament for the purpose of coming to office and/or surviving as a cabinet we also include relevant figures for the upper chamber in Table A.1a. Among our 17 countries this is only the case for Belgium, Italy, Spain (in 1977), and Sweden (before 1971). Finally, following the criteria and typology established by McDonnell and Valbruzzi (2014), by indicating a limited policy remit, we also note in Table A.1a any technocratic and caretaker status of cabinets. Notably, due to our coding rules, in a few cases we include in our data what Conrad and Golder (2010) label continuation caretaker cabinets. A recent and notable case is the Rajoy II caretaker cabinet in Spain, in office between the 2015 and 2016 parliamentary elections. Table A.1b describes the existence of electoral alliances and pre-electoral coalitions, which is measured during the election campaign and thus not at the level of the resulting

⁸ We here confine ourselves to present the ordinal alignment and we do not claim to provide a conceptual map of the more exact distances between all political parties in policy terms on that ordinal alignment. We do the same for the second dimension, when applicable. ⁹ For the second (and possibly third) dimension, when applicable, a discussion of this is found in the country chapter. In that context, we ask our contributors to make a qualitative judgement about the presence and content of such dimensions. These can be, e.g., the ‘environment vs. growth’ scale from Laver and Hunt (1992: 307), ‘permissive vs. restrictive immigration’ from Benoit and Laver (2006), and GAL–TAN from the Chapel Hill Expert Surveys (Bakker et al. 2015; Polk et al. 2017).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

734

  .

Table A.1b Electoral alliances and pre-electoral coalitions in Sweden, 1988–2018 Election date

Constituent parties

Type

Types of pre-electoral commitment

1991-09-15 2006-09-17

M, L M, C, L, KD

PEC PEC

2010-09-19

M, C, L, KD

PEC

2014-09-14

S, MP, V M, C, L, KD

PEC PEC

2018-09-09

S, MP M, C, L, KD

PEC PEC

Written contract, joint press conference Written contract, joint press conference, other* Written contract, joint press conference, other* Joint press conference, other Written contract, joint press conference, other* Joint press conference Joint press conference

Notes: Type: Electoral alliance (EA) and/or Pre-electoral coalition (PEC) Types of pre-electoral commitment: Written contract, Joint press conference, Separate declarations, and/or Other * Joint electoral manifesto

cabinet. Political parties sometime contest election and gain seats in parliament under joint labels with other political parties. We label such arrangements electoral alliances. An electoral alliance is here defined as two or more parties running a joint list in the election or agreeing on mutual withdrawal of candidates with the purpose of maximizing their seat share. As mentioned earlier, if they then, after the election, split up and form different PPGs, we count the PPGs as individual parties. We also identify pre-electoral coalitions. A pre-electoral coalition is defined as an agreement between parties (before the election) to form a coalition government after the election. In both instances, this needs to be a two-sided commitment (i.e. both parties have to agree). It may be a contract or something less formal (e.g. a joint press conference, separate declarations of party bodies). Both these phenomena are coded ex ante, that is before the election takes place. Both electoral alliances and pre-electoral coalitions are recorded in Table A.1b on an election-by-election basis. We record both the constituent parties and how their support is expressed publicly before the election. The parties may or may not have a written pre-electoral coalition agreement, but here we do not count one-sided statements, that is that one party issues a preference for another party as future coalition partner but receives no commitment in return, as pre-electoral coalition pacts. Moreover, we only consider coalition statements that are official and made on behalf of the party (typically based on a party executive or party congress decision). Hence, we do not simply count individual politicians’ statements of coalition preferences or preferred outcomes but instead code pre-electoral alliances only when there is evidence of a mutual commitment from multiple parties.

Coalition formation Next, we turn our attention to the formation of coalitions and governments. Here we present the empirical record in terms of bargaining. How straightforward is government formation in each country? How many cabinets formed on first attempt? How many

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

   

735

bargaining efforts were required otherwise? Are there general patterns, for example do governments formed immediately after elections take more attempts to form than cabinets formed in between elections? In the context of discussing difficult or straightforward government formation processes, we account for the number of days required: on average, the minimum amount of time required and the lengthiest bargaining period. Table A.2 structures the presentation of the process of government formation in our respective countries. In the chapter texts, we distinguish a formateur-led process from freestyle bargaining. These two types can help us distinguish between countries but can also be used to characterize phases in one and the same process. Recall that an informateur has the function of identifying a viable coalition (but not to become PM him/herself), while a formateur is intended to form the government him/herself (and to become PM). In the tables, we focus more on the resulting process in terms of conclusive and inclusive bargaining rounds. In the formation process, before a successful attempt at formation, there can be inconclusive bargaining rounds. We identify which parties, including only those aiming at cabinet-level positions, were involved in government formation attempts and which parties were left out in these attempts. A ‘bargaining round’ is defined by any change in the party composition and any change in the formateur or informateur. When one party drops out of the negotiations and/or another party enters, we count a new bargaining round. We also only count formation attempts, or bargaining rounds, that are ‘publicly known’. Thus, we record when it is generally known (e.g. via media) that parties are sitting down and negotiating over government formation. Furthermore, simultaneous negotiations of different sets of parties for the purpose of forming alternative coalitions are counted as different bargaining rounds.¹⁰ When there were inconclusive bargaining rounds, we list the parties involved in these bargaining rounds. We begin with the party that has the formateur status (and hence would nominate the PM). In freestyle bargaining systems, we try to identify the party that is the (biggest) driving force—the informal formateur—behind the formation attempt. We also record the time it took to complete the government formation attempt(s), counting the days that pass between the first and the last day of negotiations.¹¹ To capture the length of the full bargaining process, we calculate the time between official ‘date-out’ of the previous cabinet and the official ‘date-in’ (see the earlier discussion on these measures). As an alternative measure we estimate the time between the first (potentially inconclusive) and the last day of the entire set of bargaining rounds. Thus, in contrast to the formal ‘number of days required for cabinet formation’, the column Total bargaining duration indicates the actual length of the bargaining process in terms of how long it actually took for the parties involved to agree on a new coalition. Another discrepancy that can arise relates more specifically to our coding rules. In a few countries, for example Denmark, Norway, and Sweden (before 2011), a government can, in theory, remain in office in perpetuity absent any successful votes of no confidence. If an

¹⁰ We count a new bargaining round when one in/formateur is replaced by another. Note, however, that when one informateur is replaced by a formateur who then succeeds in forming the government, we do not count this as a new bargaining round, provided that the formateur indeed forms the coalition that the preceding informateur suggested. In contrast, we count a new bargaining round if the informateur’s proposed coalition failed to materialize but the formateur then succeeded with a new party composition. ¹¹ We include in the count both the first and the last day. The minimum number of days of a bargaining round is one.

1988 1991 1994

1996 1998 2002 2006 2010 2014

2019

Carlsson II Bildt Carlsson III

Persson I Persson II Persson III Reinfeldt I Reinfeldt II Löfven I

Löfven II

4

0 0 0a 0 0 1

0 0 1

Number of inconclusive bargaining rounds S M, C, L, KD S (1) L, S S S S M, C, L, KD M, C, L, KD S, MP (1) S, MP, V S, MPb (1) M, C, L, KD (2) M, KD (3) S, MPb (4) M, KD

Parties involved in the previous bargaining rounds

16 0 1 0 0 0 17 0 17 1 24 35 9 12 16

Bargaining duration of individual formation attempt (in days)

0 0 0 17 0 17 80

134

16 1

Total bargaining duration

4 0 0 19 0 19

0 19 19

Number of days required in government formation

154

132

153

0

175

115

154

23 130

Abstention

178

163 180

Pro

77

49

169

0

147 26

Contra

Result of investiture vote

b

In addition to S and MP, C and L were also party to negotiations on future support and collaboration on the budget.

Before 2011 an incumbent PM could remain in office after an election with no requirement to resign or submit to an investiture vote. a MP held parallel negotiations with on the one hand S and L, and on the other C and KD. However, due to our coding rules, this does not constitute as inconclusive bargaining round due to S instead opting to continue in office (see discussion on investiture rules).

Notes:

Year in

Cabinet

Table A.2 Cabinet formation in Sweden, 1988–2019

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

   

737

incumbent government decides to remain in office after post-election negotiations, our coding rules stipulate that the new cabinet has a government formation duration of zero days—the preceding cabinet terminated with the election and the new cabinet is never formally reappointed or reinvested, leaving us with the election as the only satisfied criterion for coding a new cabinet. In these cases, we have opted to keep the bargaining duration as zero days, while including any potential negotiations as table notes as well as discussing them in the text.¹² The next section focuses on multiparty cabinet coalitions only. Table A.3 outlines the result of the coalition bargaining preceding the cabinet taking office: how the most important cabinet portfolios are distributed. One of the early coalition theorists, Gamson (1961) is famous for an early statement of the ‘parity rule’, that is that there is a proportionality between the number of ministerial portfolios and the number of parliamentary seats held by a coalition party. That is, if party A has 60 per cent of the coalition’s parliamentary seats, party B 30 per cent, and Party C 10 per cent, then minister portfolios tend to be distributed accordingly (i.e. 60 per cent, 30 per cent, and 10 per cent). Even if this is not a perfect predictor, because the largest party sometimes tends to get a somewhat less than proportional allocation (Mershon 2002: 65; see also Keman, 2006), the parity rule is empirically well established. It has been shown that parity distribution is more easily achieved under conditions of uncertainty and complexity when parties can have a harder time exploiting any differences in bargaining power (FalcóGimeno and Indridason 2013), but as a general rule, parity tends to be a cross-national outcome. In Table A.3, we include a tally of the number of cabinet posts, that is, ministers, held by each party of the government coalition as well as the number of independent ministers, if any. When political parties bargain for government office, they also have the option of increasing or decreasing the number of members (ministers) in the cabinet (Mershon 2002; Verzichelli 2008; Bergman et al. 2015). From the coalition literature on Western Europe it is also known that political parties do not consider all portfolios to be of equal worth (Blondel and Thiebault 1991; Mair 2007). Some positions, such as the PM position, the ministers for foreign affairs, finance, economy, and defence, often tend to be valued higher than other ministerial portfolios. It is also well known that political parties that belong to certain party families have their own favourite portfolios. The agrarian parties, for instance, will bargain hard to get the Ministry of Agriculture, while the Social Democrats are usually keen to get the Ministry of Labour Affairs (Budge and Keman 1990). These observations have recently been confirmed empirically for Western Europe (Bäck et al. 2011). In Table A.3, we have included the same five ministries (or their national equivalents) for all countries: the PM, Ministry of Finance, Ministry of Foreign Affairs, Ministry of Social Affairs, and Ministry of the Interior or Justice,¹³ in order of descending importance, based on our interpretation of Budge and Keman (1990: 95), Druckman and Warwick (2005), and Bäck et al. (2011). This is the section in which we discuss these patterns and critically evaluate how well these cross-national finding applies to the individual case. Thus, we discuss, among other

¹² If a cabinet already in power negotiates with an opposition party about forming a different cabinet with a different party composition, and this bargaining round in the end is incomplete, there can be an odd technicality. For example, in Norway, Stoltenberg III had individual bargaining attempt of 21 days, but since it never resigned and just continued in power, it also had zero (0) days of government formation. ¹³ Whichever ministry is in charge of the police.

1991

2006 2010 2014 2019

Bildt

Reinfeldt I Reinfeldt II Löfven I Löfven II

9 (M), 4 (L), 4 (C), 3 (KD), 1 (ind.) 11 (M), 5 (C), 4 (L), 3 (KD) 13 (M), 4 (C), 4 (L), 3 (KD) 18 (S), 6 (MP) 18 (S), 5 (MP)

Number of ministers per party (in descending order)

23 24 24 23

21

Total number of ministers

13 12 11 12

14

Number of ministries

M M S S

M

1 Prime minister

M M S S

L

2 Finance

M M S S

M

3 Foreign affairs

M M S S

C

4 Labour and equality affairs

M M S S

M

5 Justice

In a few of our countries, instead of the generic ministry ‘Social Affairs’, a ‘Labour and Equality’ ministry, or the equivalent, is listed as more relevant and important.

There have been no watchdog junior ministers in Sweden during the presented period.

Note:

Year in

Cabinet

Table A.3 Distribution of cabinet ministerships in Swedish coalitions, 1991–2019

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

   

739

country-specific observations, the quantitative allocation of ministries to coalition parties and if the distribution is in proportional (in quantitative terms, i.e. as proportion of parliamentary seats contributed to the coalition) or is there an ‘overcompensation’ of small parties. We also highlight the qualitative aspects of ministerial distribution. In each country, some ministries come with more power (e.g. finance) and prestige (e.g. foreign affairs). We discuss whether the quality of ministries increases or decreases the proportionality of allocation and, in particular, evidence in favour of (or opposed to) the conclusion that specific parties aim for (and regularly get) specific ministries (e.g. Christian parties might aim for education, Social Democrats might aim for social affairs, Liberals might aim for justice, etc.). We also discuss particular regularities: for example, some ministries may never be combined in the hands of the same party (e.g. interior and defence), some may always go together (e.g. PM and minister of finance), and some may never be given to some particular party even when it participates in the government (e.g. interior to Communist successor parties). In that context, we also stress the extent to which ministries that belong to different political parties ‘shadow’ or keep tabs on each other. Cross-sectional ministries (e.g. finance, justice, etc.) may be good for that purpose or ministries that have similar tasks (e.g. economics or budget may shadow finance). We also mention specialities such as two ministers (at the same time) in one ministry and discuss the distribution of ‘watchdog junior ministers’ and their actual role (based on qualitative information) in coalition governance. When applicable, we also identify junior ministers when they are from a different party than the head of the department (the minister), a watchdog junior minister. We define a junior minister as a political appointee in the chain of command of the minister with executive power, ideally just under the minister. Here the term ‘political appointee’ indicates that junior ministers come and go with a particular government/administration. This separates a case of a junior minister from a situation in which a politician is exercising political control over top-level bureaucratic appointments. As with cabinet ministers, we include a tally of watchdog junior ministers—if any—per party in Table A.3.

Coalition agreements Continuing on the ‘set-up’ of coalitions, Table A.4 details the written coalition agreements. When elaborating on this table we want to discuss the role of formal coalition agreements (clearly defined documents or other ex ante agreements). Coalition agreements are the most binding statements between the parties, that is the most authoritative document that constrains party behaviour. Some coalitions have documents specifically designed for that purpose (e.g. ‘coalition agreements’ in Germany, Austria, Belgium, the Netherlands, etc.). Other coalitions have functional equivalents such as ‘joint pre-election programmes’ of incoming coalitions or ‘joint budget proposals’ of the non-socialist parties (as in Norway). Some documents may involve only some of the government parties. Some coalitions split the functions of coalition agreements between different documents, for example between policy programmes, coalition agreements, and agreements on parliamentary cooperation of the government parties. We are interested in all such documents. With regard to the formal (written) agreements, each chapter discusses the relevance of the coalition agreements in coalition governance. Here we first address issues such as the enforceability of coalition agreements, for example by describing what happens if ministers or parties break the agreement. We also discuss the relevance and distribution of the specific mechanisms contained in Table A.4. Cabinets in Table A.4 are split over two or more rows, for the purpose of presentation, if the coalition agreement was substantially amended or coalition governance changed substantially during the tenure of the cabinet. Thus, a particular cabinet can have

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

740

  .

Table A.4 Size and content of coalition agreements in Sweden, 1991–2019 Coalition

Size

Distribution Distribution of Policies General Policy-specific (in %) competences procedural rules of offices rules (in %) (in %) (in %) (in %)

Bildt Reinfeldt I Reinfeldt II Löfven I

5,200 6,421 6,262 3,772a 4,961b 3,683

0 0 0 0 0 0

Löfven II

4 0 0 0 0 0

0 0 0 0 0 0

0 0 0 0 0 0

96 97.5 92.3 95.8 98 90.4

Notes: a

2014 Government declaration

b

2014 policy agreement between the Social Democrats and the Green Party

multiple rows if there were multiple and separate coalition agreements. Also note that the table columns do not necessarily sum to 100% for each and every cabinet. This is because we do not report the share of purely ceremonial passages in each agreement. Theoretically, the table is motivated by the existence of different attempts at structuring coalition governance. One is primarily based on procedural agreement. The coalition parties agree on how they will make decisions once they are together in cabinet but do not spell out in great detail the contents of the cabinet’s policies. The other form is primarily based on policy agreement. The parties negotiate and include in their agreement the master plan for government policy in terms of substantive agreement. Of course, mixed forms are possible. We discuss also how precise is the wording of the agreement and how broad is the coverage of policies/issues. The information in Table A.4 is based on the most authoritative document, that is preferably a document that is approved or signed by all party leaders. If a country expert determines that there are two or more documents that constitute the outcome of the interparty bargaining process, we list them in separate lines with distinct word counts and percentages. The same goes for instances in which there are substantial (major) amendments to the original agreement(s) during the tenure of a coalition cabinet. Table A.4 also indicates distribution of competences. This is our way to try to capture what the parties agree about ‘who gets what?’ in their formal coalition agreement. This could be a listing of already existing ministries, or some description of how existing ministries would be merged or split, and/or a description of how administrative units and partial jurisdictions would be shifted from one ministry to another as a result of the coalition negotiations. We detail this and discuss how coalition governance ‘by contract’ actually has worked by presenting both an overview and useful examples. The coalition agreements, though being incomplete and non-enforceable contracts, set the agenda of the incoming government and thus reduce both the uncertainty and the conflict potential between coalition partners (Timmermans 2006; Walgrave et al. 2006; Moury 2011). To the extent that political parties bear the costs of drafting and implementing coalition agreements, they are readily available to coordinate policies and reach collective decisions. Complementing their external function, coalition agreements are likewise used to contain and reduce conflicts within coalition parties (Müller and Strøm 2008). As coalition parties commit ex ante to a specific policy programme to be implemented over the course of the term, they seek to immunize themselves against defection and intra-party conflict.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

   

741

Coalition governance Some of the main features of coalition governance are recorded in Table A.5. Here we present in shorthand what the discussion in the chapter text then explains and elaborates. In this context, we also return to the formal rules as of cabinet decision-making and illustrate how they work in practice. Such formal rules and powers are partly helpful, partly disturbing for the practice of coalition governance. Junior partners (here, smaller parties) in a coalition, for instance, may dislike strong PM Powers and try to pre-empt other means and measurers. We also present information on when coalition agreements have been reached, if at all, in relation to their proximity to an election. It should also be noted that in Table A.5 we refer to the coalition governance mechanisms that exist in practice, that is the set of rules that were put in place at the start of the coalition. It is our coding of both formal and informal rules of coalition governance. While it may occasionally be difficult to distinguish informal rules from pure behavioural regularities, we do our utmost to do so. Rules have some power to constrain behaviour. That means they have to be enforceable in some way, however imperfectly. Even though coalition agreements are designed to contain conflicting preferences between and within coalition parties, they cannot address the universe of potential conflicts throughout a coalition’s lifetime. Rather, both internal and external shocks present governments with a myriad of unforeseen circumstances. Coalition parties therefore often turn to additional oversight mechanisms and conflict resolution arenas to further reduce agency loss and minimize the potential for lethal inter-party conflict. A first means of mutual control frequently employed are ‘watchdog’ junior ministers (i.e. a junior minister of one cabinet party shadowing the full minister of another). Especially in the absence of alternative institutional checks such as second chambers and strong PMs, coalition parties are likely to monitor each other’s ministers via ‘watchdogs’, who deliver information from the ministry (Müller and Strøm 2000; Thies 2001). Adding to these executive means of monitoring and mutual control, coalition parties may likewise resort to legislative oversight. In fact, parliamentary scrutiny has often proved an effective means of intra-coalition monitoring (Martin and Vanberg 2004, 2005). Similarly, parties will strive to regulate and scrutinize ministerial behaviour by appointing chairs of legislative committees from rival government parties (Kim and Loewenberg 2005; Carroll and Cox 2012). Finally, political parties may devise and agree upon additional mechanisms to resolve conflicts within coalition governments. While their organizational structure and their personnel composition vary from ‘internal’ over ‘mixed’ to ‘external arenas’ coalition governments devise and then resort to these arenas in order to settle inter-party conflicts that otherwise might potentially bring down the government (Andeweg and Timmermans 2008; Müller and Meyer 2010a, 2010b).¹⁴ As mentioned in Chapters 1 and 2, we use a set of indicators to capture and summarize the real type of coalition governance. To capture the extent to which the three governance

¹⁴ However, not specifically in Table A.5, but still very important, we record how coalition parties use parliamentary instruments and structures in their pursuit of coalition governance. They can also be constrained by such structures (Strøm 1990; Martin and Vanberg 2011; Zubek 2015). Parliamentary committees scrutinize ministers, but ministers also work with committee members to pursue policies and monitor other ministers. Other instruments, such as strategically placed parliamentary questions are in place in almost all parliamentary systems. Our contributors place the use of such parliamentary relationship mechanisms in the context of other instruments of coalition governance.

1991 2006 2010 2014

Post Pre, Post Pre, Post Post

Y Y Y Y

N N N N

IC, CaC IC, CaC IC, CaC IC, CaC

CaC CaC CaC CaC

IC IC IC IC

All used Most For common most serious conflicts

Year Coalition Agreement Election Conflict management in agreement public rule mechanisms

Y Y Y Y

Most/Most All/All All/All All/All

Y Y Y Y

No Comp. Comp. Comp.

Y Y Y Y

Policy agreement: Comp. = Comprehensive policy agreement

Y Y Y Y

Junior NonCoalition Freedom of Policy discipline appointment agreement ministers cabinet positions in legislation/ other parl. behaviour

Coalition discipline: All = Coalition discipline is always expected; Most = Coalition discipline is expected on all matters except those explicitly exempted

N N N

Personal Issues union excluded from agenda

Coalition agreement: PRE = pre-election; Pre, Post = pre- and post-election; POST = post-election Conflict management mechanisms: IC = Inner cabinet; CaC = Cabinet committee

Notes:

Bildt Reinfeldt I Reinfeldt II Löfven I

Coalition

Table A.5 Coalition governance mechanisms in Swedish coalition governments, 1988–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

   

743

models—ministerial, coalition compromise, and PM dominated—apply to the political systems in Western Europe, or to particular cabinets, we cover a series of indicators on coalition politics. In the country chapters, indicators such as the dispersion of leading ministerial positions, the existence of coalition agreements (contract), and if these written coalition contracts are comprehensive or not are the indicators that we build on. Other conflict management mechanisms include information on the existence of an inner cabinet and a coalition committee, or ‘watchdog’ junior ministers. Coalition behaviour in parliament, such as coalition discipline, and the role of individual ministers in cabinet decision-making and as parliamentary liaisons provide the basis for a conclusion in this respect. The summary analysis of these indicators provides useful information on which model a country (or a particular cabinet) approximates and possibly also on changes over time in a country. At the end of this section, and in the conclusion of each chapter, each contributor makes an assessment of how much (or little) each particular country corresponds to our three governance ideal types.

Cabinet termination After governance comes, inevitably, termination. Government stability is a well-researched subject in the West European context. In the next section, we make an important distinction between technical and behavioural reasons for termination. Technical (non-political) terminations are caused by events exogenous to (beyond the control of) the cabinet or the parliament. These include the occasions when regular parliamentary elections are held at the end of the CIEP or when the PM gets seriously ill or dies in office, as well as other purely constitutional reasons for why a cabinet is required to terminate. If many, we code only one of the technical reasons. Note that technical and discretionary terminations are mutually exclusive categories. However, there can be more than one reason behind discretionary terminations. These include strategically timed early elections, a loss in a parliamentary vote of confidence, or when a cabinet ends because one of its members (a party) leaves or a new one joins the coalition. When it is possible to identify an event or a particular occasion that caused the downfall of a cabinet, we record this as the ‘terminal event’. All those instances that do not fall into any of the existing discretionary termination categories, for example voluntary resignation of the PM, may be coded as ‘other voluntary reason’. Moreover, when possible, our country experts present information on the particular policy area that was the most affected or the most significant contributing reason for the downfall of the cabinet. In our discussions of Table A.6, we summarize patterns and distinguish periods. Technical terminations are normally not a sign of government instability. With regard to political terminations the major analytical question is whether it is strategic from the point of view of the entire coalition (a good time to go to the polls or to take on board another party) or conflictual (intra-coalition conflict of some kind). We summarize the mechanisms employed in these cases and discuss whether these were triggered by a particular external event. In this context, our experts also summarize and reflect on cabinet duration in their country as well as the related topic of government stability. We present factual information on how long the cabinets remained in power, represented by each cabinet’s relative duration. To get a comparative assessment, we compare the number of days that the cabinet existed with the number of days that it could potentially exist. We count from the date of the formation until the next constitutionally mandated date of parliamentary election, or to the date of presidential elections if it is required or customary for governments to resign or be replaced at the time of presidential elections. In the case of two-round elections, we take the second-round election date. Likewise, for elections spanning several

1988-09-18 1991-10-04 1994-10-07

1996-03-22 1998-09-20 2002-09-15 2006-10-06 2010-09-19 2014-10-03

Carlsson II Bildt Carlsson III

Persson I Persson II Persson III Reinfeldt I Reinfeldt II Löfven I

1998-09-20 2002-09-15 2006-09-17 2010-09-19 2014-09-14 2018-09-09

1991-09-15 1994-09-18 1996-03-18

Date out

100 100 100 100 100 100

100 100 36.6

Relative duration (%)

1 1 1 1 1 1

1 1 9

Mechanisms of cabinet termination

14

Terminal events

Parties

Policy area(s)

PM Carlsson voluntarily resigned as leader of the Social Democrats and as PM.

Comments

14: Personal event

10: Elections, non-parliamentary; 11: Popular opinion shocks; 12: International or national security event; 13: Economic event;

Terminal events

9: Other voluntary reason

7a/b: Conflict between coalition parties: policy (a) and/or personnel (b); 8: Intra-party conflict in coalition party or parties;

Discretionary terminations 4: Early parliamentary election; 5: Voluntary enlargement of coalition; 6: Cabinet defeated by opposition in parliament;

1: Regular parliamentary election; 2: Other constitutional reason; 3: Death of PM

Technical terminations

Date in

Cabinet

Table A.6 Cabinet termination in Sweden, 1988–2018

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

   

745

days, we take the last date that ballots can be cast. Thus, we base the comparison on the last day the cabinet could stay in office provided the parliamentary term would be exhausted. In most cases (and according to our counting rules) this will be the latest possible election date. In the context of Table A.6, each chapter, where relevant, also contains an explicit and authoritative discussion on the practice of early (or extra) elections and the practice of parliamentary dissolution. While early accounts on cabinet duration emphasized the destabilizing effect of multiparty governments relative to single-party ones (Taylor and Herman 1971; Sanders and Herman 1977), more recent empirical studies find little difference in government stability. This is likely due to the prevalent incentives and discretionary power of single-party governments to strategically time legislative elections (Strøm and Swindle 2002; Saalfeld 2008; Schleiter and Morgan-Jones 2009). Thus, to the extent that strategic considerations outweigh the inherent stability of single-party governments, the stalemate between coalition partners to agree upon early election calling may prima facie countervail any inherent instability of coalition governments.

Appendix. List of political parties Abbreviation

Name

V

Left Party (Vänsterpartiet), 1990– Left Party Communists (Vänstperartiet kommunisterna), 1967–1990 Communist Party of Sweden (Sveriges kommunistiska parti), 1921–1967 Green Party (Miljöpartiet de gröna) Social Democratic Party (Socialdemokratiska arbetarepartiet) Liberals (Liberalerna), 2015– People’s Party (Folkpartiet liberalerna), 1934–2015 Centre Party (Centerpartiet), 1959– Agrarian League (Bondeförbundet), 1913–1959 Moderate Unity Party (Moderata samlingspartiet), 1969– Right Party (Högerpartiet), 1952–1969 Right Party (Högerns riksorganisation), 1938–1952 Christian Democrats (Kristdemokraterna), 1995– Christian Democrats (Kristdemokratiska samhällspartiet), 1987–1995 New Democracy (Ny demokrati) Sweden Democrats (Sverigedemokraterna)

MP S L C M

KD NyD SD Note:

Party names are given in English, followed by the party name in Swedish in parentheses. If several parties have been coded under the same abbreviation (successor parties), or if the party has changed its name, these are listed in reverse chronological order followed by the period during which a specific party or name was in use. In some of our countries, political parties that have no seats in parliament are mentioned, typically these are regional parties with an impact on electoral alliances, but they are not included in the party appendix. It includes only national level parties.

References Andeweg, Rudy B., and Arco Timmermans (2008). ‘Conflict Management in Coalition Government’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 269–300.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

746

  .

Bäck, Hanna, Marc Debus, and Patrick Dumont (2011). ‘Who Gets What in Coalition Governments? Predictors of Portfolio Allocation in Parliamentary Democracies’. European Journal of Political Research, 50(4): 441–78. Bakker Ryan, Catherine de Vries, Erica Edwards, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Vachudova (2015). ‘Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010’. Party Politics, 21(1): 143–52. Benoit, Kenneth, and Michael Laver (2006). Party Policy in Modern Democracies. London: Routledge. Bergman, Torbjörn (2000). ‘Sweden: When Minority Cabinets Are the Rule and Majority Coalitions the Exception’. In Wolfgang C. Müller and Kaare Strøm (eds), Coalition Governments in Western Europe. Oxford: Oxford University Press, 192–230. Bergman, Torbjörn, Svante Ersson, and Johan Hellström (2015). ‘Government Formation and Breakdown in Western and Central Eastern Europe’. Comparative European Politics, 13(3): 345–75. Bergman, Torbjörn, Gabriella Ilonszkiand, Wolfgang C. Müller (eds) (2019). Coalition Governance in Central Eastern Europe. Oxford: Oxford University Press. Bergman, Torbjörn, Wolfgang C. Müller, and Kaare Strøm (2005). ‘Comparative Parliamentary Democracy: A Project Report’. European Political Science, 4(2): 42–54. Blondel, Jean, and Jean-Louis Thiébault (eds) (1991). The Profession of Government Minister in Western Europe. Basingstoke: Macmillan. Budge, Ian, and Hans Keman (1990). Parties and Democracy: Coalition Formation and Government Functioning in Twenty States. Oxford: Oxford University Press. Carroll, Royce, and Gary W. Cox (2012). ‘Shadowing Ministers: Monitoring Partners in Coalition Governments’. Comparative Political Studies, 45(2): 220–36. Conrad, Courtenay Ryals, and Sona N. Golder (2010). ‘Measuring Government Duration and Stability in Central Eastern European Democracies’. European Journal of Political Research, 49(1): 119–50. De Winter, Lieven, and Patrick Dumont (2008). ‘Uncertainty and Complexity in Cabinet Formation’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 123–57. Druckman, James N., and Paul V. Warwick (2005). ‘The Missing Piece: Measuring Portfolio Salience in Western European Parliamentary Democracies’. European Journal of Political Research, 44(1): 17–42. Elgie, Robert (1999). ‘The Politics of Semi-presidentialism’. In Robert Elgie (ed.), Semipresidentialism in Europe. Oxford: Oxford University Press, 1–21. Falcó-Gimeno, Albert, and Indridi H. Indridason (2013). ‘Uncertainty, Complexity, and Gamson’s Law: Comparing Coalition Formation in Western Europe’. West European Politics, 36(1): 221–47. Gamson, William A. (1961). ‘A Theory of Coalition Formation’. American Sociological Review, 26(3): 373–82. Hooghe, Liesbet, Ryan Bakker, Anna Brigevich, Catherine de Vries, Erica Edwards, Gary Marks, Jan Rovny, and Marco Steenbergen (2010). ‘Reliability and Validity of Measuring Party Positions: The Chapel Hill Expert Surveys of 2002 and 2006’. European Journal of Political Research, 49: 684–703. Keman, Hans (2006). ‘Parties and Government: Features of Governing in Representative Democracies’. In Richard S. Katz and William Crotty (eds), Handbook of Party Politics. London: Sage, 160–74.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

   

747

Kim, Dong-Hun, and Gerhard Loewenberg (2005). ‘The Role of Parliamentary Committees in Coalition Governments: Keeping Tabs on Coalition Partners in the German Bundestag’. Comparative Political Studies, 38(9): 1104–29. Kitschelt, Herbert, Zdenka Mansfeldova, Radek Markowski, and Goba Toka (1999). PostCommunist Party Systems. Competition, Representation, and Inter-party Cooperation. Cambridge: Cambridge University Press. Klingemann, Hans-Dieter, Andrea Volkens, Judith Bara, Ian Budge, and Michael D. McDonald (2006). Mapping Policy Preferences II. Estimates for Parties, Electors and Governments in Central and Eastern Europe, European Union and OECD 1990–2003. Oxford: Oxford University Press. Laakso, Markku, and Rein Taagepera (1979). ‘ “Effective” Number of Parties: A Measure with Application to West Europe’. Comparative Political Studies, 12(1): 3–27. Laver, Michael, and Ben W. Hunt (1992). Policy and Party Competition. New York, NY: Routledge. Laver, Michael, and Norman Schofield (1990). Multiparty Government. The Politics of Coalition in Europe. Oxford: Oxford University Press. Lijphart, Arend (2012). Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, 2nd edn. New Haven & London: Yale University Press. Lipset, Seymour Martin, and Stein Rokkan (eds) (1967). Party Systems and Voter Alignments: Cross-national Perspectives. New York: Free Press. Mair, Peter (2007). ‘Party Systems and Alternation in Government, 1950–2000: Innovation and Institutionalization’. In Siri Gloppen and Lise Rakner (eds), Globalisation and Democratisation: Challenges for Political Parties. Bergen: Fagbokforlaget, 135–53. Martin, Lanny W., and Georg Vanberg (2004). ‘Policing the Bargain: Coalition Government and Parliamentary Scrutiny’. American Journal of Political Science, 48(1): 13–27. Martin, Lanny W., and Georg Vanberg (2005). ‘Coalition Policymaking and Legislative Review’. American Political Science Review, 99(1): 93–106. Martin, Lanny W., and Georg Vanberg (2011). Parliaments and Coalitions. The Role of Legislative Institutions in Multiparty Governance. Oxford: Oxford University Press. McDonnell, Duncan, and Marco Valbruzzi (2014). ‘Defining and Classifying Technocrat-Led and Technocratic Governments’. European Journal of Political Research, 53(4): 654–71. Mershon, Carol (2002). The Costs of Coalition. Stanford: Stanford University Press. Moury, Catherine (2011). ‘Coalition Agreement and Party Mandate: How Coalition Agreements Constrain the Ministers’. Party Politics, 17(3): 385–404. Müller, Wolfgang C., and Thomas M. Meyer (2010a). ‘Meeting the Challenges of Representation and Accountability in Multi-party Government’. West European Politics, 33(5): 1065–92. Müller, Wolfgang C., and Thomas M. Meyer (2010b). ‘Mutual Veto? How Coalitions Work’. In Thomas König, George Tsebelis, and Marc Debus (eds), Reform Processes and Policy Change: How Do Veto Players Determine Decision-Making in Modern Democracies. New York: Springer, 99–124. Müller, Wolfgang C., and Kaare Strøm (eds) (2000). Coalition Governments in Western Europe. Oxford: Oxford University Press. Müller Wolfgang C., and Kaare Strøm (2008). ‘Coalition Agreements and Cabinet Governance’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 159–99. Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher, Marco Steenbergen, Milada Vachudova, and Marko Zilovic (2017). ‘Explaining the Salience of Anti-elitism and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

748

  .

Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data’. Research & Politics, 4(1): 1–9. Saalfeld, Thomas (2008). ‘Institutions, Chance and Choices: The Dynamics of Cabinet Survival in the Parliamentary Democracies of Western Europe (1945–1999)’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 327–68. Sanders, David, and Valentine Herman (1977). ‘The Stability and Survival of Governments in Western Europe’. Acta Politica, 12(3): 346–77. Schleiter, Petra, and Edward Morgan-Jones (2009). ‘Constitutional Power and Competing Risks: Monarchs, Presidents, Prime Ministers, and the Termination of East and West European Cabinets’. American Political Science Review, 103(3): 496–512. Sieberer, Ulrich (2015). ‘Hire or Fire? The Link between Cabinet Investiture and Removal in Parliamentary Democracies’. In Björn Erik Rasch, Shane Martin, and José Antonio Cheibub (eds), Parliaments and Government Formation. Unpacking Investiture Rules. Oxford: Oxford University Press, 331–56. Strøm, Kaare (1990). Minority Government and Majority Rule. Cambridge: Cambridge University Press. Strom, Kaare, Ian Budge, and Michael Laver (1994). ‘Constraints on Cabinet Formation in Parliamentary Democracies’. American Journal of Political Science, 38: 303–35. Strøm, Kaare, Eric C. Browne, John P. Frendreis, and Dennis W. Glieber (1988). ‘Contending Models of Cabinet Stability’. The American Political Science Review, 82(3): 923–41. Strøm, Kaare, and Stephen M. Swindle (2002). ‘Strategic Parliamentary Dissolution’. American Political Science Review, 96(3): 575–91. Strøm, Kaare, Wolfgang C. Müllerand, Torbjörn Bergman (eds) (2003). Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Strøm, Kaare, Wolfgang C. Müllerand, Torbjörn Bergman (eds) (2008). Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press. Taylor, Michael, and Valentine Herman (1971). ‘Party Systems and Government Stability’. American Political Science Review, 65(1): 28–37. Thies, Michael F. (2001). ‘Keeping Tabs on Partners: The Logic of Delegation in Coalition Governments’. American Journal of Political Science, 45(3): 580–98. Timmermans, Arco (2006). ‘Standing Apart and Sitting Together: Enforcing Coalition Agreements in Multiparty Systems’. European Journal of Political Research, 45(2): 263–83. Verzichelli, Luca (2008). ‘Portfolio Allocation’. In Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman (eds), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 237–68. Walgrave, Stefaan, Frédéric Varone, and Patrick Dumont (2006). ‘Policy with or without Parties? A Comparative Analysis of Policy Priorities and Policy Change in Belgium, 1991 to 2000’. Journal of European Public Policy, 13(7): 1021–38. Zubek, Radoslaw (2015). ‘Coalition Government and Committee Power’. West European Politics, 38(5): 1020–41.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

Index accountability electoral 366 Adenauer, Konrad 265 Ahern, Bertie 365, 371, 386 Aho, Esko 172, 178, 181, 183, 198 Alliance 90/The Greens (GR) 247, 249–51, 255–6, 258–61, 266, 269, 274–6 Alliance for the Future of Austria (BZÖ) 5, 41, 45, 48, 55–6, 58, 64–5 alliances, pre-electoral: definition 734 Belgium 92 Croatia 651 France 218 Greece 289 Iceland 333 Ireland 368 Italy 405 Portugal 522 Spain 556 UK 617 Alternative for Germany (AfD) 5, 247–8, 254–5, 275–6 Amato, Giuliano 402–3, 413–4, 417–8, 434 Anti-Austerity Alliance–People Before Profit 362, 365, 367, 378 Austrian People’s Party (ÖVP) 41, 44–5, 48, 50–6, 58–9, 64–5, 69, 72–4, 708 Ayrault, Jean-Marc 226, 231, 240 Aznar, José Maria 52, 552, 559–60, 562–3, 568 Balkenende, Jan Peter 457, 460, 473–4 Balladur, Edouard 214, 232 bargaining advantage 22, 722 bargaining uncertainty 21 duration 2, 20, 31 situation/environment 15, 22, 32 bargaining rounds definition 735 Austria 49–50 Belgium 94–5, 101 Croatia 652–4 Denmark 137 Finland 178–9 France 221

Germany 257 Greece 292–5, 296 Iceland 334–7 Ireland 369 Italy 412–7 the Netherlands 457–9 Norway 494–5 Portugal 524 Spain 558 Sweden 586 the United Kingdom 618–9 Basque Nationalist Party (PNV) 550, 552, 557, 559–60, 563, 565 Beel, Louis 473 Beregovoy, Pierre 240 Berlusconi, Silvio 406–11, 416–8, 423–4, 426, 429, 431, 433–7, 439–40 bicameralism 8, 19, 30, 208, 406, 439, 484, 545, 686–7, 728–9 Bildt, Carl 593, 595, 600–2 Blair, Tony 386, 633 Blue Reform (SIN) 5, 176, 198 Bondevik, Kjell Magne 493, 496–8, 502–3, 505, 507, 509–10, 512 Bright Future (BF) 329, 333, 353 Brothers of Italy (FdI) 410–1 Brown, Gordon 622–3, 633 Bruton, John 370, 382 cabinet cabinet committee 62, 84, 107, 109, 146, 148–9, 193, 230, 232, 234, 268, 303, 307, 342, 345, 383, 390, 428–9, 431, 470–1, 503–4, 535, 599–600, 629–30, 662, 664, 701–2, 742 cabinet durability 32, 669 cabinet duration 2, 4, 8, 28–30, 115, 125, 154, 187, 198, 237, 272, 318–9, 349, 433, 473, 510, 537, 566, 592, 603, 632, 668–9, 713–4, 722, 743, 745 cabinet formation 1, 8, 11–2, 14, 166–7, 176, 178, 201, 207, 220–1, 250–1, 257, 260, 294, 335, 369, 391, 396, 448–9, 458, 493, 508, 524, 558, 574, 586, 588, 618, 653, 661, 684, 692, 694, 735–6 cabinet investiture 83

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

750



cabinet (cont.) cabinet reshuffle 600, 644 cabinet stability 28, 30, 274, 510, 710, 716 caretaker 101, 103, 114–5, 118, 126–8, 136, 167, 293, 296, 312, 315–6, 319, 326, 330, 332, 350, 398, 407, 412, 455, 457, 473–6, 548–9, 557–8, 562, 567, 569, 593, 603, 613, 733 Callaghan, James 617, 620, 633 Calvo-Sotelo, Leopoldo 557, 568 Cameron, David 622–4, 626, 629–31, 633, 635 Canary Islands Coalition (CC) 552, 559 Canary Islands Independent Groups (AIC) 559 Carlsson, Ingvar 603–4 Catalan Convergence And Union (CiU) 552–4, 557, 559–60, 563, 568 Catalan Republican Left (ERC) 552, 554, 560 Catholic People’s Party (KVP) 474 Cavaco Silva, Aníbal 518 CDS-People’s Party (CDS) 518–20, 523, 525–6, 528–30, 533–4, 536–7, 539–41, 543 Centre Party (Finland) (KESK) 166–7, 169, 174, 183, 197 Centre Party (Sweden) (C) 577, 580–1, 583, 587–93, 601, 606 Chirac, Jacques 52, 208, 212, 214–6, 219, 225, 227, 240 Christian Democracy (DC) 398–9, 408–9, 417, 431 Christian Democracy for the Autonomies (DCA) 410 Christian Democratic Appeal (CDA) 451, 455, 457, 459–60, 462, 477 Christian Democratic Centre (CDC) 409 Christian Democratic and Flemish (CDV) 85–6, 92–3, 97–100, 107 Christian Democratic Union of Germany & Christian Social Union of Bavaria (CDU/ CSU) 247, 249–51, 254–6, 258–61, 265–6, 268–9, 271–2, 275–6 Christian Democrats (Denmark) (KrF) 129, 131 Christian Democrats (Finland) (KD) 176, 187, 198 Christian Democrats (Sweden) (KD) 575, 577, 580, 583–4, 587, 589–90, 593, 606 Christian People’s Party (KrF) 483, 486, 488, 491–2, 495–7, 503, 505, 508–9 Christian Union (CU) 451, 455, 457, 560 Civic Choice (SC) 410–1 Citizens (Ciudadanos, C’s) 552–4, 560–1, 566 Clann na Poblachta (CnP) 362, 376 Clann na Talmhan (CnT) 362, 376 Clegg, Nick 621–2, 626, 630–3 clientelism 44, 289, 297

coalition alternatives 54, 184, 260, 359, 361, 399, 735 coordination 69, 662, 667 duration 28, 111, 320, 385 formation 1, 8–9, 16, 19, 42, 45, 48, 50, 53–4, 56, 134, 207, 212, 247, 249, 255, 269, 275, 293, 327–8, 330, 334, 367, 391, 449, 455–7, 466, 474, 557, 640, 650, 676, 681, 686, 696, 728, 730, 734 life-cycle 1–4, 6, 8–10, 15, 24, 28, 30–2, 250, 285, 319–20, 439, 606, 668, 680, 687, 691, 699–700, 722, 727, 729 membership 115, 206, 220, 222, 226, 229, 232, 235, 406, 412, 418, 432–3, 536, 672 renegotiation 359 coalition agreements content of 60, 105, 144, 185, 228, 264, 301, 340, 375, 424, 463, 465, 501, 531, 596, 627, 659, 705, 707, 740 formal 226, 319, 423, 739–40 length 27, 263, 265, 276, 465, 705–6 coalition bargaining 12–4, 29, 31, 81, 136, 206–8, 212, 225, 251, 266, 337, 391, 440, 485–6, 491, 589, 641, 652, 661, 691, 737, 745 coalition committee definition 703 Austria 61–2, 67 Belgium 108–9 Croatia 663–8 Germany 267–71 Greece 303 Iceland 344–5 Ireland 380 Italy 427–8 the Netherlands 467–8 Portugal 535 the United Kingdom 627, 629–32 coalition discipline definition 703 Austria 61–2, 64–5 Belgium 108–9, 119 Croatia 663–5, 675 Denmark 146 Finland 192–3 France 233–4, 236 Germany 267–8, 270 Greece 303, 308 Iceland 344–5 Ireland 380 Italy 427–8, 432 the Netherlands 467–8 Norway 504, 509 Portugal 535

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 Sweden 599, 602 the United Kingdom 629 coalition governance definition 24–6 Austria 58, 61, 63–4, 66 Belgium 107–10, 114 Croatia 662–8 Denmark 145–7, 149, 153 Finland 187–93 France 229, 231, 233–5, 237 Germany 265–8, 271–2 Greece 302–3, 306 Iceland 341–4 Ireland 379–80, 382 Italy 425–8 the Netherlands 465, 467–70, 472 Norway 500, 502, 504 Portugal 532, 535 Sweden 598–601 the United Kingdom 628–30, 632 Coalition of the Left and Progress (SYN) 286, 289–90, 296–7, 299–301, 306 Coalition of the Radical Left (SYRIZA) 5, 290–2, 297, 299–300, 307, 310, 319 coalition theory 3, 6, 16, 276, 680–1 Communist Party (France) (C) 213–5, 218, 222, 227, 229, 230, 240 Communist Party of Greece (KKE) 289, 291–2, 297 complexity 5, 15, 21, 31–2, 81, 142, 260, 263, 423, 440, 448, 680, 722, 732, 737 confidence motions see: no-confidence motion conflict management mechanisms 25, 62, 109, 146, 182, 189, 193, 202, 232, 234, 237, 268, 303, 345, 380, 428, 468, 477, 504, 534–5, 599, 629, 662, 664, 666–8, 675, 701–2, 716, 742–3 Conservative Party (Norway) (H) 483, 488, 495, 498, 503, 505–9 Conservative Party (UK) (Con) 611–3, 616–7, 620–3, 636 Conservative People’s Party (KF) 129–31, 142 constitutional inter-election period 275, 730 Conte, Giuseppe 411, 416, 418–9, 423, 426, 429, 431, 438–9 cooperation inter-party 507 Costa, António 518, 525, 536 Costello, John A. 376, 381, 383 Craxi, Bettino 429–30 credibility 50, 226, 329, 367, 382, 470, 497 credible commitment 25, 379, 715 Croatian Democratic Union (HDZ) 640, 642–5, 648–50, 652, 655, 658, 661, 666–70

751

Croatian Party of Pensioners (HSU) 650, 655 Croatian Peasant Sarty (HSS) 650 Croatian People’s Party (HNS) 650, 658, 669 Croatian Social Liberal Party (HSLS) 648, 650, 655, 658, 666, 669, 672 Danish People’s Party (DFP) 5–6, 124, 131, 139, 152–3, 158 decision rules 19, 43, 58, 167, 195, 548, 568, 685, 689–90, 729 Dehaene, Jean-Luc 81, 85, 97, 104, 111, 113, 115 delegation 111, 189, 269, 397, 429, 432–3, 483, 497, 512 De Mita, Ciriaco 430 democracy parliamentary 1–2, 9–14, 32, 47, 75, 90, 124–5, 133, 165, 173, 206, 211, 261, 263, 288, 332, 354, 357, 391, 397, 404, 454, 483, 485, 490, 512–3, 521, 550, 579, 585, 615, 640, 681, 687–8, 728, 731 Democratic Centre (CD) 410 Democratic Party (PD) 407–8, 410–1, 416, 418, 426 Democrats 66 (D66) 451, 455, 459–60, 462, 477 Di Rupo, Elio 96, 98–9, 107, 111, 113 discretionary termination definition 743 Austria 70–1 Belgium 116–7 Croatia 671–4 Denmark 155–7 Finland 199–200 France 238–9 Germany 273 Greece 311–7 Iceland 350–1 Ireland 387–8 Italy 435–8 the Netherlands 475–6 Norway 511 Portugal 539 Spain 567 Sweden 605 the United Kingdom 634–5 Durão Barroso, José Manuel 523, 525, 528–30, 532–4, 537–8, 540 D’Alema, Massimo 412, 418, 434 Ecolo (E) 91–3, 96–7, 101, 103, 106, 118, 122 elections campaign 188, 198, 247, 373, 381, 575, 582, 589–90, 622, 733

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

752



elections (cont.) early 4, 8, 29–32, 41, 54, 59, 65, 69–74, 96, 117, 126, 154–7, 159, 200, 239, 251, 273–4, 312, 315–9, 329, 342, 350–1, 360, 378, 387–9, 412, 416, 434–6, 438, 473–4, 476, 511, 537–9, 552, 564, 566–9, 603–5, 633–6, 643, 645, 672, 674, 709–10, 729, 743–5 strategic timing 4, 13, 386 electoral alliances 92, 134–5, 176, 213–5, 217–20, 222, 241, 249, 254–5, 286, 289, 333, 367–8, 396, 405, 408, 411, 455–6, 491, 493, 520, 522–3, 555–6, 584, 617, 650–1, 730, 733–4 electoral performance 45, 91, 222, 254, 648 electoral systems 125, 139, 206–7, 212–3, 237, 240, 286, 289, 292, 333, 358, 360, 367, 390, 398–9, 406, 409–10, 450, 510, 518, 546, 551, 575–6, 613, 616–7, 620, 640, 645, 648, 693, 709, 728 majoritarian 202, 206, 213, 216, 237, 240, 286, 399, 406, 418, 546 proportional 82, 124–5, 130, 139, 215, 286, 289, 292, 358, 360, 398–9, 406, 419, 449–50, 485–6, 518, 551, 575–6, 616, 640, 642, 644–5, 648, 694 Erlander, Tage 593, 603 European Union European Parliament 131, 175, 254, 519, 553, 616 European Commission 100, 526, 539–40 Fälldin, Torbjörn 578, 593, 600–1, 603 Fanfani, Amintore 430, 434 Faymann, Werner 41, 69, 72, 74 federalism 91, 119, 122, 249, 436, 484 bicameralism 8, 19, 30, 484, 687, 729 second chamber 19, 43, 248–51, 275, 643, 728, 741 Fianna Fáil (FF) 357–62, 365–72, 377–9, 386, 389–91 Fillon, François 217, 229, 231 Fine Gael (FG) 357, 360–2, 365–8, 370, 372–3, 376–9, 381, 383, 389–91 Finns Party 5, 174–7, 180, 183, 198, 201 Five-Star Movement (M5S) 5, 407–8, 410–1, 416, 418, 424–5, 429, 439 Flemish Interest (VB) 5, 85–6, 91–2, 98 Flemish Liberals and Democrats (VLD) 81, 85, 96–101, 103, 115 formal support parties 5, 46, 88, 132, 170, 172, 209–10, 287–8, 400, 402, 404, 454, 490, 521, 550, 564, 578, 614, 643, 646 Forum for Democracy (FvD) 451–2 Francophone Democratic Federalists (FDF) 85, 98, 122

Free Democratic Party (FDP) 44, 247, 249–51, 255–6, 258–61, 272, 275–6 Freedom Party of Austria (FPÖ) 5, 41, 43–5, 48, 50–6, 58–9, 63–5, 67, 69, 72–4 Galician Nationalist Bloc (BNG) 552, 560 Gamson’s Law 21–2, 140, 419, 423, 498, 512, 697 Gentiloni Silveri, Paolo 418, 434 Go Italy (FI) 399, 408–11, 418, 433 grand coalition 41, 43–5, 50–6, 58–9, 63–4, 66–9, 72–5, 251, 254, 259, 275, 417, 652, 655, 721 Green! (G) 81, 85, 91, 96–7, 103, 106, 122 Green League (VIHR) 166, 174, 182, 184, 192, 198 GreenLeft (GL) 451, 455 Green Party (Ireland) (GR) 362, 365–6, 386, 390 Green Party (Sweden) (MP) 575, 577, 580–1, 583–5, 587–8, 590–3, 597, 604, 606 Greens (V) 214–5, 217, 219, 222–3, 226–7, 229–30, 236 Gregurić, Franjo 642, 645, 655, 658, 665 Grillo, Beppe 407 Grímsson, Ólafur Ragnar 324–7, 352–3 Grivas, Ioannis 296, 319 Gunnlaugsson, Sigmundur Davíð 329, 342, 352–3 Gusenbauer, Alfred 55–6, 64–5, 69, 74 Guterres, António 523, 538 Haarde, Geir 342 Halla-aho, Jussi 175–6 Haughey, Charles 386 head of state 8, 42, 84, 127, 166, 195, 208, 248–9, 285, 411, 416, 449, 456–7, 459, 484, 545, 547, 549, 575, 687, 690, 696, 728 Heath, Edward 620 Human Shield (ŽZ) 641, 644, 649, 676 illiberal democracy 640 incumbency incumbency advantage 73 Independence Party (IP) 324–5, 328, 330, 333–4, 337–8, 341, 347, 352 Independent Democratic Serb Party (SDSS) 648, 655 Independent Greeks (ANEL) 5, 291, 297, 299–301, 307, 310, 319 informateur 98, 128, 292, 325, 370, 416, 456–7, 459, 493, 621, 729, 735 inner cabinets Austria 62, 67 Belgium 107, 109–13, 118 Croatia 662, 664, 666, 675

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 definition 743 Denmark 146 Finland 193 Germany 268–9 Greece 306 Iceland 343, 345 Ireland 383 Italy 428–9 the Netherlands 468 Norway 482, 503–4, 507 Portugal 534 Sweden 599–601 the United Kingdom 629, 631 interest groups 66, 180, 263, 338 International Monetary Fund 290, 366, 373, 517, 526 investiture requirement 7, 370, 684, 729 rule 84, 449, 576, 591, 687, 729 vote 19, 30, 43, 84, 126, 136, 138, 167, 235, 249, 259, 285, 370–1, 378, 397, 406, 411, 417, 424–5, 434, 449, 474, 546–7, 559–61, 576, 590, 643–4, 655, 658, 686–7, 728–9 Istrian Democratic Assembly (IDS) 650 Italian Communist Party (PCI) 398 Italian Liberal Party (PLI) 408, 424, 446 Italian Popular Party (PPI) 409 Italian Republican Party (PRI) 398, 408, 424, 446 Italian Social Democratic Party (PSDI) 398, 408, 424 Italian Social Democrats (SDI) 410 Italian Socialist Party (PSI) 398, 408–9, 424 Italian Social Movement (MSI) 398 Italy of Values (IdV) 407, 409–10 Jakobsdóttir, Katrín 333 Jóhannsson, Sigurður Ingi 352 Jospin, Lionel 212, 214, 216, 227, 229, 231–2 junior ministers definition 703, 739, 741 Austria 57, 61–2, 68 Belgium 99–102, 108, 113 Croatia 657, 662–5, 667 Finland 194 France 224–7 Germany 267–9, 271 Greece 297–300, 303–5, 307 Ireland 379–81, 384 Italy 398, 420, 422, 426–8, 457 the Netherlands 457, 461, 467–8, 471 Norway 483, 499, 503–4, 506–7, 509, 512 Portugal 527, 529–30, 533, 535 Sweden 585, 594, 599–601 the United Kingdom 625–6, 628–30

753

Juppé, Alain 232 Jørgensen, Anker 128, 130 Karamanlis, Kostas 310, 318–9 Katainen, Jyrki 176–7, 183–4, 187, 194–5, 198 Kekkonen, Urho 166, 182, 184, 198 Kenny, Enda 368, 370–2, 377–9, 385–6 Kern, Christian 47–9, 57, 59–60, 62, 69, 71–2 Kiesinger, Kurt Georg 259 Klima, Viktor 48, 52, 66, 69, 74 Kohl, Helmut 272, 274 Kok, Wim 459, 473 Kosor, Jadranka 655, 661, 665–6, 670 Kreisky, Bruno 55 Kurz, Sebastian 45, 48, 58, 64, 67, 69, 72–3 Labour Party (Ireland) (Lab) 360–1, 367, 373, 376, 383 Labour Party (Norway) (A) 483, 485–8, 491–3, 495–6, 498, 505–6, 509 Labour Party (The Netherlands) (PvdA) 451–2, 455, 457, 459–60, 462, 466, 473, 477 Labour Party (The United Kingdom) (Lab) 611–3, 616–7, 620–4, 633, 636 League (LN) 406–11, 416, 418, 424–5, 429, 439 League of Communists of Croatia (SKH) 645 Left Green Movement (LG) 327–8, 330, 333–4, 337–8, 347–9 Left Party (V) 577, 580–1, 583–5, 587–93 Le Pen, Jean-Marie 216 Le Pen, Marine 217 left–right dimension economic 5, 44, 97, 175–6, 201, 254, 328, 333, 365, 409, 451–2, 455, 488, 491, 551, 581, 584, 587, 696, 733 sociocultural 44–5, 176, 184, 201 Left Socialists (VS) 131 legislator, median 16, 45, 86, 254, 411, 491, 587, 696–7 Leterme, Yves 85, 93, 96, 101, 103, 106–7, 111, 115 Letta, Enrico 408, 411, 416–8, 426, 431, 439 Liberal Alliance (LA) 124, 129, 131, 136, 139–40, 151 Liberal Democrats (Lib) 611, 613, 617, 620–4, 626, 628, 630–3, 636 Liberal Forum (LIF) 45 Liberal Party (Croatia) (LS) 650 Liberal Party (Denmark) (V) 129–31, 139–40, 142 Liberal Party (Norway) (V) 483, 485–6, 488, 491–2, 495–8, 502–3, 507–8 Liberals (L) 575, 577, 580–3, 588–90, 592–3, 595, 597, 606

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

754



‘life cycle’ analogy 2–4, 727, 729 Liinamaa, Keijo 167 Lipponen, Paavo 167, 182–4, 188, 194–5, 198 List Dedecker (LDD) 86 Löfven, Stefan 576, 588–93, 595, 597, 600–2, 604, 606 Lubbers, Ruud 473 Macmillan, Harold 633 Macron, Emmanuel 206, 208, 217, 219, 227, 231 majority cabinets definition 3 Austria 46–7 Belgium 87–90 Croatia 646–7 Denmark 132–3 Finland 170–3 France 209–11 Germany 252–3 Greece 287–8 Iceland 331–2 Ireland 363–4 Italy 400–4 the Netherlands 453–4 Norway 489–90 Portugal 521 Spain 550 Sweden 578–9 the United Kingdom 614–5 Manolić, Josip 670 Martin, Micheál 366, 378 Mateša, Zlatko 669–70 Mauroy, Pierre 222, 231, 240 Merkel, Angela 254, 256, 259–60, 266, 269, 271–2, 274 Mesić, Stjepan 642, 670 Messmer, Pierre 222 Michel, Charles 81, 84, 96, 98–100, 104, 107, 113, 118 Milanović, Zoran 650, 652, 655–6, 660–1, 666, 668–70 minimal winning coalition definition 3 Austria 46–7 Belgium 87–90 Croatia 646–7 Denmark 132–3 Finland 170–3 France 209–11 Germany 252–3 Greece 287–8 Iceland 331–2 Ireland 363–4 Italy 400–4

the Netherlands 453–4 Norway 489–90 Portugal 521 Spain 550 Sweden 578–9 the United Kingdom 614–5 ministerial government 3, 25–6, 68, 75, 158, 201, 265, 307, 324, 338, 341, 353, 425, 482, 534, 541, 661, 715–6, 718, 721–2 minority cabinets definition 3 Austria 46–7 Belgium 87–90 Croatia 646–7 Denmark 132–3 Finland 170–3 France 209–11 Germany 252–3 Greece 287–8 Iceland 331–2 Ireland 363–4 Italy 400–4 the Netherlands 453–4 Norway 489–90 Portugal 521 Spain 550 Sweden 578–9 the United Kingdom 614–5 minority coalition 124, 139, 248, 353, 371, 555, 569, 592, 655 minority coalitions 8, 124, 139, 143, 182, 418, 570, 694, 722 Mitsotakis, Konstantinos 306, 310, 318 Mitterrand, François 208, 212–4 Moderates (M) 577, 580–4, 587, 589–90, 593, 597, 606 Monti, Mario 407–8, 410, 417, 419, 426 Moro, Aldo 429 Movement for the Autonomies (MpA) 410, 424 Movement of Left Radicals (MRG) 218 National Alliance (AN) 409–10 National Coalition (KOK) 166–7, 169, 174–6, 182–4, 198, 215 National Front (France) (FN) 215–7 National Labour Party (NL) 376 National Rally (RN), see National Front (France) (FN) negative parliamentarism 7, 19, 42, 84, 125, 158, 325, 575–6, 604, 681, 686–7, 716 New Democracy (Greece) (ND) 284, 286, 289–92, 296–7, 299–301, 305–6, 309–10 New Democracy (Sweden) (NyD) 5, 577 New Democratic Centre (NCD) 411

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 New Flemish Alliance (N-VA) 85–6, 91–3, 96, 98–101, 103, 107, 118, 122 no-confidence 84, 118, 198, 285, 359, 438, 576 no-confidence motion 84, 118, 125–6, 359, 398, 438, 460, 474, 510, 547, 576, 587–8, 729 non-cabinet positions: definition 3 Austria 61–2 Belgium 108–9 Croatia 663–4 Denmark 146 Finland 190–3 France 233–4 Germany 267–8 Greece 303 Iceland 344–5 Ireland 380 Italy 427–8 Netherlands 467–8 Norway 504 Portugal 535 Sweden 599 UK 629 Norwegian Communist Party (NKP) 486 office-seeking behaviour 16, 21, 118, 176, 644, 656, 658 Open Flemish Liberals and Democrats (VLD) 85, 92, 99–101 Orešković, Tihomir 640, 643, 652, 658, 660–1, 665–8, 670, 676 Panhellenic Socialist Movement (PASOK) 284, 286, 289–91, 296–7, 299–301, 305, 309 Papademos, Lucas 291, 296–7, 299–300, 306–9, 319 Papandreou, Andreas 290, 310, 318–9 Papandreou, Georgios 299, 318 parliamentarism see also negative parliamentarism see also positive parliamentarism parliamentary government 125–6, 128, 138, 158, 325 leaders 431, 471–2, 536, 666–7, 702 systems 2, 22, 26–7, 207, 223, 227, 237, 269, 324, 330, 686, 741 parliamentary party group definition 730, 732, 734 party: anti-system 398, 407 effective number of 4–5, 45, 86, 131, 291, 367, 451, 477, 520, 537, 552, 613, 681, 733 left-wing populist 329

755

median 16, 18, 23, 86, 138, 296, 365, 409, 491, 520, 583–5, 587, 645, 730, 732–3 pivotal 54, 93, 158, 215, 272, 328, 349, 398, 675 radical left 251, 289–91, 296, 299, 407, 553 radical right 215, 406, 452, 455, 460, 544, 551, 555, 574, 577, 580, 582, 587, 606, 616, 680, 722 right-wing populist 86, 276, 289, 296, 487 summit 111–2, 308, 425, 430–1, 534, 662, 667–8, 675, 715, 717–8 Party for Freedom (PVV) 5–6, 451–2, 455, 457, 460, 474, 477 Party for the Animals (PvdD) 451–2 Party of Liberal Democrats (Libra) 646, 648, 650–1, 653, 657, 676 party system fragmentation 81, 118, 169, 174, 290, 292, 320, 353, 367, 377, 390–1, 680, 715, 722, 733 party system polarization 20–1, 31–2, 91, 118, 131, 541, 649, 681, 732 Passos Coelho, Pedro 518, 523, 525–6, 528–30, 532–4, 537–8, 540 patronage 297, 300, 384, 658 People’s Party for Freedom and Democracy (VVD) 451–2, 455, 457, 459–60, 462, 466, 477 People’s Union (VU) 85, 92, 122 People of Freedom (PdL) 407–8, 410–1, 418, 424, 433 Pikrammenos, Panagiotis 296, 319 Pirate Party (P) 327, 329 Plenković, Andrej 643, 652, 658, 661–2, 666–70, 676 policy agreements: definition 739–40 Austria 60 Belgium 105 Croatia 659 Denmark 144 Finland 185–6 France 228 Germany 264 Greece 301 Iceland 340 Ireland 375 Italy 424 Netherlands 463 Norway 501 Portugal 531 Sweden 596 UK 627 policy dimensions 5, 23–4, 31, 45, 365, 487–8, 491, 520, 563, 645, 658, 680–1, 696, 728

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

756



Popular Orthodox Rally (LAOS) 5, 289, 291, 296, 299–300, 319 Popular Party (PP) 544, 546, 549, 552–5, 557, 559–66, 569 portfolio allocation 9, 11, 21–4, 27, 48, 55–6, 58, 75, 97, 99, 101, 136, 139, 147, 177, 184, 207, 226, 261, 266, 276, 297, 299–300, 337–8, 354, 370, 373, 419, 462, 482, 495, 497–8, 512, 525, 530, 658, 696, 707 salience 22 Portuguese Communist Party (PCP) 519–20, 536 positive parliamentarism 249, 285, 397, 546, 643, 686 pre-electoral coalitions definition 730, 733–4 Austria 44 Belgium 92 Croatia 650–2 Denmark 134–5 France 212, 215, 217–8 Germany 249, 251, 254–5 Greece 286, 289 Iceland 333–4 Ireland 367–8 Italy 396, 399, 405, 409–12, 439 the Netherlands 455–6 Norway 491–3 Portugal 520, 522 Spain 555–6 Sweden 580, 584 the United Kingdom 617 prime ministerial government 248, 640, 661, 675 prime ministerial powers 8, 28, 660–1, 687–8 Prodi, Romano 406, 410, 412, 416–9, 434 Progress Party (Denmark) (FrP) 131 Progress Party (Norway) (FrP) 5, 164, 483, 487–8, 490, 492, 494–9, 503, 507–9, 512 Progressive Party (PP) 324–5, 328–30, 333–4, 337–8, 341, 347, 352 proportional representation (PR) 124, 130, 286, 289, 292, 358, 449, 485–6, 551, 640, 642, 644–5, 694 public opinion 30, 354 Račan, Ivica 642–3, 652, 655–6, 658, 660–2, 665, 667–70 radical left parties 251, 289–91, 296, 299, 407, 553 radical right parties 215, 406, 452, 455, 460, 544, 551, 555, 574, 577, 580, 582, 587, 606, 616, 680, 722

Raffarin, Jean-Pierre 231, 240 Rajoy, Mariano 546, 549, 553–4, 557, 561–6, 569, 733 Rasmussen, Anders Fogh 150, 154 Rasmussen, Lars Løkke 129, 136–7, 139, 145, 151 Rasmussen, Poul Nyrup 143, 149 Red-Green Alliance see: Unity List (EL) Reformed Political Party (SGP) 451–2, 455 Reformist Movement (MR) 81, 85, 91, 97–8, 100–1, 103 Reinfeldt, Fredrik 581, 588, 592–3, 595, 600–2, 606 Renzi, Matteo 406, 411, 417–8, 430, 439 the Republicans (LR) 217, 219 Reynolds, Albert 359, 370–1, 385–6 Rocard, Michel 222 Rose in Fist (RnP) 410 Rutte, Mark 448–51, 455, 457, 459–60, 462, 466, 470–4 Salvini, Matteo 406 Samaras, Antonis 291, 296–7, 299–302, 306–9, 318–20 Sanader, Ivo 643, 648–9, 652, 655–6, 660–1, 665–6, 668–70, 673 Sánchez, Pedro 546, 557, 559, 561, 564–6, 569 Santana Lopes, Pedro 523, 526, 528–9, 534, 538, 540 Sarkozy, Nicolas 215–6, 219, 229, 231, 236 Schlüter, Poul 130, 143, 147 Schröder, Gerhard 247, 251, 256, 258–9, 261, 266, 269, 271–2, 274–5 Schüssel, Wolfgang 41, 45, 50, 52–5, 59, 63–4, 73–4, 708 Scottish National Party (SNP) 616 Segni Pact (PS) 409 semi-presidentialism 8, 19, 30, 42, 166, 206, 208, 240, 285, 325, 327, 518, 640–2, 648, 656, 686–7, 728 Sigurðardóttir, Jóhanna 330, 353 single-party governments majority 284, 290, 544, 591, 623, 693, 696 minority 84, 182, 214, 357, 361, 370, 544–5, 559, 694, 696, 722 Sinn Féin (SF) 362, 365–7, 372, 378, 390–1 Sinowatz, Fred 64 Sipilä, Juha 177, 183–4, 189, 194, 197 SMP Social Democratic Alliance (SDA) 327–30, 332–3, 335–6, 339, 350, 352 Social Democratic Party (Iceland) (SDP) 328 Social Democratic Party (PSD) 518–20, 523, 525–6, 528–30, 533–4, 536–7, 540–1

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

 Social Democratic Party (Sweden) (S) 580–1, 583–5, 587–93, 597, 601, 604, 606 Social Democratic Party (The United Kingdom) (SDP) 613, 616–7 Social Democratic Party of Austria (SPÖ) 41, 44–5, 48, 50–2, 54–6, 59, 65, 69, 72–4 Social Democratic Party of Croatia (SDP) 640, 644, 648–50, 652, 655, 666 Social Democratic Party of Finland (SDP) 166–7, 169, 174, 177 Social Democratic Party of Finland (SDP) 167, 169, 174–5, 182–3, 198 Social Democratic Party of Germany (SPD) 247, 250–1, 254–6, 258–61, 266, 269, 272, 274–6 Social Democrats (Denmark) (S) 124, 130–1, 134, 139–40, 142–3 Social Democrats (Ireland) (SD) 362, 367 Socialist Left Party (SF-SV) 486, 498, 506, 509 Socialist Party (Belgium) (PS) 85, 91, 93, 98–9, 111 Socialist Party (France) (PS) 206–8, 213–9, 222–3, 227, 229–31, 236, 240 Socialist Party (PSOE) 544, 546, 549, 552–5, 557, 559–62, 565, 568–9 Socialist Party (the Netherlands) (SP) 451, 455 Socialist Party Differently (sp.a) 92, 96, 118 Socialist People’s Party (Denmark) (SF) 124, 129, 131, 134, 139 Solberg, Erna 495–6, 498, 500, 502–3, 505–9, 512 Sorsa, Kalevi 195, 198 South Tyrolean People’s Party (SVP) 410 Spirit (Spirit) 92, 624 Stoltenberg, Jens 482, 491, 498, 502–3, 505–9, 512, 737 Stubb, Alexander 198 Suárez, Adolfo 547, 559, 562, 568–9 surplus coalitions definition 3 Austria 46–7 Belgium 87–90 Finland 170–3, 182 France 209–211 Germany 252–3 Greece 287–8 Iceland 331–2 Ireland 363–4, 370, 386 Italy 400–4 the Netherlands 453–4, 459 Portugal 521 Sweden Democrats (SD) 574, 577, 580–4, 587–90, 592–3, 606 technical termination definition 743

757

Austria 70–1 Belgium 116–7 Croatia 671–4 Denmark 155–7 Finland 199–200 France 238–9 Germany 273 Greece 311–7 Iceland 350–1 Ireland 387–8 Italy 435–8 the Netherlands 475–6 Norway 511 Portugal 539 Spain 567 Sweden 605 the United Kingdom 634–5 technocrat 291, 296–7, 319, 398, 416–7, 419, 423, 426, 562, 624, 640, 652, 733 termination definition 743 Thanou-Christophilou, Vassiliki 296, 319 Thatcher, Margaret 620, 633 The Greens—The Green Alternative (GA) 43–5, 51–2, 54–6, 65 The Left (Linke) 254–5, 260 The Republic in Motion (LREM) 206–7, 217, 219 Thorning-Schmidt, Helle 136, 140, 143, 145, 147, 150 transaction costs 465 Tsipras, Alexis 292, 296–7, 299–301, 307–8, 310, 318–20 Tuđman, Franjo 640–2, 648, 655 Tzannetakis, Tzannis 290, 296–7, 299–301, 305–6, 308, 310, 319 Ullsten, Ola 584 unanimity 53, 195, 662 Unbowed France (FI) 217 uncertainty 5, 15, 20–1, 31, 81, 130, 134, 222, 259, 293, 324, 341, 352, 386, 408, 440, 471, 722, 737, 740 Union for a Popular Movement (UMP) 212, 214–6, 219, 222, 229, 236, 246, 529 Union for French Democracy (UDF) 213, 215, 219, 227 Union for the New Republic (UNR) 213, 219 Union of Democratic Centre (UCD) 549, 559, 562 Union of Democrats for Europe (UDEUR) 410, 434 Union of Democrats for the Republic (UDR) 213

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 29/6/2021, SPi

758



Union of the Centre (UDC) 407, 409–11 United Kingdom Independence Party (UKIP) 5–6, 616 United Socialist Party (USP) 328 Unity List (EL) 131, 134, 139 Us with Italy (NcI) 411 Valls, Manuel 222, 225 Van Agt, Dries 457, 473 Vanhanen, Matti 184, 189, 197–8 Van Rompuy, Herman 93, 107, 115–6 Varadkar, Leo 377, 386 Verhofstadt, Guy 52, 81, 85, 91, 93, 96–7, 100, 104, 106, 113, 115 veto power 153, 158, 232, 449, 498, 530, 598

Vox (Vox) 544, 555 Vranitzky, Franz 41, 63–4, 66, 69 We Can (Podemos) 544, 552–5, 559–62, 564, 566, 569 Wilders, Geert 452 Willoch, Kåre 482, 493, 496, 503 Wilson, Harold 620, 632–3 World War II 6, 74, 103, 169, 174, 184, 187, 189, 201, 265, 328, 397, 448, 457, 474, 485–6, 496, 577, 591, 597, 604, 611, 632, 686, 727 Zapatero, José Luis Rodríguez 546, 552, 560, 562–3, 568 Zolotas, Xenofon 290, 293, 296–7, 299–302, 306–7, 309, 319