The Selection of Ministers in Europe: Hiring and Firing (Routledge Advances in European Politics) [1 ed.] 041543081X

Citation preview

The Selection of Ministers in Europe

This volume discusses the formation of government cabinets within twenty European democracies, providing the institutional background to the selection and de-selection of ministers. Exploring the historical and constitutional context to cabinet formation, this volume proceeds to provide vital data on the strategic issues that affect the selection of ministers. Covering states from all over Europe, the authors examine trends from the post-war period up to the present day, with specific focus on recent decades for the newer democracies in political transition. The volume includes: • • •

pioneering new research into the hiring and firing of government ministers vital information on appointments, dismissals and resignations within government cabinets succinct constitutional data relating to ministerial selections across a number of European states

The book is the first output of the Selection and De-selection of Political Elites international network of scholars (SEDEPE) and will provide a major source of information for all scholars interested in the formation, maintenance and termination of cabinets and the nature of ministerial government. The Selection of Ministers in Europe: Hiring and firing will also be of broader interest to students of European government and political institutions. Keith Dowding is Professor of Political Science and Head of the Political Science Programme at the Research School of Social Sciences, Australian National University in Canberra. He has published widely in comparative politics, rational and social choice theory, public administration and public policy, urban political and political philosophy. Patrick Dumont is Researcher in Political Science at the University of Luxembourg. His main research is in comparative politics, studying executive coalitions and their relations with legislative and party politics, and the impact of European integration on national political systems.

Routledge Advances in European Politics

1 Russian Messianism Third Rome, revolution, Communism and after Peter J.S. Duncan 2 European Integration and the Postmodern Condition Governance, democracy, identity Peter van Ham 3 Nationalism in Italian Politics The stories of the Northern League, 1980–2000 Damian Tambini 4 International Intervention in the Balkans since 1995 Edited by Peter Siani-Davies 5 Widening the European Union The politics of institutional change and reform Edited by Bernard Steunenberg 6 Institutional Challenges in the European Union Edited by Madeleine Hosli, Adrian van Deemen and Mika Widgrén 7 Europe Unbound Enlarging and reshaping the boundaries of the European Union Edited by Jan Zielonka 8 Ethnic Cleansing in the Balkans Nationalism and the destruction of tradition Cathie Carmichael 9 Democracy and Enlargement in PostCommunist Europe The democratisation of the general public in fifteen Central and Eastern European countries, 1991–1998 Christian W. Haerpfer

10 Private Sector Involvement in the Euro The power of ideas Stefan Collignon and Daniela Schwarzer 11 Europe A Nietzschean perspective Stefan Elbe 12 European Union and E-Voting Addressing the European Parliament’s internet voting challenge Edited by Alexander H. Trechsel and Fernando Mendez 13 European Union Council Presidencies A comparative perspective Edited by Ole Elgström 14 European Governance and Supranational Institutions Making states comply Jonas Tallberg 15 European Union, NATO and Russia Martin Smith and Graham Timmins 16 Business, The State and Economic Policy The case of Italy G. Grant Amyot 17 Europeanization and Transnational States Comparing Nordic central governments Bengt Jacobsson, Per Lægreid and Ove K. Pedersen 18 European Union Enlargement A comparative history Edited by Wolfram Kaiser and Jürgen Elvert

19 Gibraltar British or Spanish? Peter Gold

31 Europeanizing Social Democracy? The rise of the party of European socialists Simon Lightfoot

20 Gendering Spanish Democracy Monica Threlfall, Christine Cousins and Celia Valiente

32 Conflict and Change in EU Budgetary Politics Johannes Lindner

21 European Union Negotiations Processes, networks and negotiations Edited by Ole Elgström and Christer Jönsson

33 Gibraltar, Identity and Empire E.G. Archer

22 Evaluating Euro-Mediterranean Relations Stephen C. Calleya 23 The Changing Face of European Identity A seven-nation study of (supra)national attachments Edited by Richard Robyn 24 Governing Europe Discourse, governmentality and European integration William Walters and Jens Henrik Haahr 25 Territory and Terror Conflicting nationalisms in the Basque country Jan Mansvelt Beck 26 Multilateralism, German Foreign Policy and Central Europe Claus Hofhansel

34 Governance Stories Mark Bevir and R.A.W Rhodes 35 Britain and the Balkans 1991 until the present Carole Hodge 36 The Eastern Enlargement of the European Union John O’Brennan 37 Values and Principles in European Union Foreign Policy Edited by Sonia Lucarelli and Ian Manners 38 European Union and the Making of a Wider Northern Europe Pami Aalto 39 Democracy in the European Union Towards the emergence of a public sphere Edited by Liana Giorgi, Ingmar Von Homeyer and Wayne Parsons

27 Popular Protest in East Germany Gareth Dale

40 European Union Peacebuilding and Policing Michael Merlingen with Rasa Ostrauskaite

28 Germany’s Foreign Policy Towards Poland and the Czech Republic Ostpolitik revisited Karl Cordell and Stefan Wolff

41 The Conservative Party and European Integration since 1945 At the heart of Europe? N.J. Crowson

29 Kosovo The politics of identity and space Denisa Kostovicova

42 E-Government in Europe Re-booting the state Edited by Paul G. Nixon and Vassiliki N. Koutrakou

30 The Politics of European Union Enlargement Theoretical approaches Edited by Frank Schimmelfennig and Ulrich Sedelmeier

43 EU Foreign and Interior Policies Cross-pillar politics and the social construction of sovereignty Stephan Stetter

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

44 Policy Transfer in European Union Governance Regulating the utilities Simon Bulmer, David Dolowitz, Peter Humphreys and Stephen Padgett 45 The Europeanization of National Political Parties Power and organizational adaptation Edited by Thomas Poguntke, Nicholas Aylott, Elisabeth Carter, Robert Ladrech and Kurt Richard Luther 46 Citizenship in Nordic Welfare States Dynamics of choice, duties and participation in a changing Europe Edited by Bjørn Hvinden and Håkan Johansson 47 National Parliaments within the Enlarged European Union From victims of integration to competitive actors? Edited by John O’Brennan and Tapio Raunio

48 Britain, Ireland and Northern Ireland since 1980 The Totality of Relationships Eamonn O’Kane 49 The EU and the European Security Strategy Forging a global Europe Edited by Sven Biscop and Jan Joel Andersson 50 European Security and Defence Policy An implementation perspective Edited by Michael Merlingen and Rasa Ostrauskaitò 51 Women and British Party Politics Descriptive, substantive and symbolic representation Sarah Childs 52 The Selection of Ministers in Europe Hiring and firing Edited by Keith Dowding and Patrick Dumont

The Selection of Ministers in Europe Hiring and firing

Edited by Keith Dowding and Patrick Dumont

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

First published 2009 by Routledge 2 Park Square Milton Park Abingdon Oxon OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 Routledge is an imprint of the Taylor & Francis Group, an informa business.

This edition published in the Taylor & Francis e-Library, 2008. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” © 2009 Keith Dowding and Patrick Dumont selection and editorial matter; individual contributors, their contributions All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data The selection of ministers in Europe : hiring and firing / Keith Dowding and Patrick Dumont [editors]. p. cm. – (Routledge advances in European politics ; 52) Includes bibliographical references and index. 1. Cabinet officers–Selection and appointment–Europe. 2. Cabinet officers–Europe. I. Dowding, Keith M. II. Dumont, Patrick. JN94.A69A67 2008 352.24’3–dc22 2008023457

ISBN 0-203-88690-9 Master e-book ISBN

ISBN 10: 0–415–43081–X (hbk) ISBN 10: 0–203–88690–9 (ebk) ISBN 13: 978–0–415–43081–4 (hbk) ISBN 13: 978–0–203–88690–8 (ebk)

Contents

List of illustrations Preface Notes on contributors 1

Structural and strategic factors affecting the hiring and firing of ministers

ix xiii xv

1

K E I T H D O WD I NG A N D P A TRI CK D U M O N T

2

Hiring and firing ministers under informal constraints: Germany

21

JÖ R N F I S C H E R A N D A N D RÉ K A I S ER

3

Cabinet dynamics and ministerial careers in the French Fifth Republic

41

C H R I S T O P H E R K A M A N D I N D RI D I I N D RI D A S O N

4

Choosing, moving and resigning at Westminster, UK

58

S A MU E L B E R L I N S K I , TO RU N D EW A N , K EI TH D O W DING AND GITA SUBRAHMANYAM

5

Italy: the difficult road towards a more effective process of ministerial selection

79

L U C A V E R Z I C H ELLI

6

Cabinet dynamics in democratic Spain (1977–2008)

101

JO S É R E A L - D ATO A N D M I G U EL J EREZ- M I R

7

Belgium: ups and downs of ministerial careers in a partitocratic federal state P A T R I C K D U MON T, S TEF A A N F I ERS A N D RÉG I S DANDOY

125

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

viii Contents 8 Portugal: the primacy of ‘independents’

147

AN T Ó N I O C O S TA P I N TO A N D P ED RO TA V A RE S DE AL M E IDA

9 In tranquil waters: Swedish cabinet ministers in the postwar era

159

H A N N A B Ä CK , TH O M A S P ERS S O N , K Å RE V ER NB Y AND H E L E N A WOCK ELBERG

10 Constructing and maintaining Irish governments

179

E O I N O ’ MA LLEY

11 More safe than sound? Cabinet ministers in Iceland

194

GU N N A R H E LG I K RI S TI N S S O N

12 Learning to rule: ministerial careers in post-communist democracies

204

K A T JA F E T T ELS CH O S S A N D CS A BA N I K O LENYI

Name index Subject index

228 231

Illustrations

Tables 3.1 3.2 3.3 3.4 4.1 4.2 4.3 5.1 5.2

5.3

5.4 6.1 6.2 7.1 7.2 7.3 7.4 8.1 8.2

Educational background, gender and political experience of French ministers, 1959–2005 44 Average time in the French government, 1960s–2000s 48 Career advancement and mobility among French ministers, 1959–2005 51 Reasons for the end of French ministerial appointments, 1959–2005 52 Size of cabinets and governments in Britain, 1900–2005 (as at 1 January) 62 Resignations and non-resignations in British governments from Attlee to Blair, 1945–2007 72 Ministerial resignations by government in Britain, from Attlee to Blair, 1945–2007 73 Cabinet formation in Italy before 1992 and after 1996 85 Insiders, outsiders and other types of political backgrounds among Italian ministers. First and Second Republic (1948–1992 compared with 1996–2006) 89 Resignation, reallocation of ministerial tasks and other changes in government composition during the ministerial mandate, First and Second Republic in Italy (1948–1992 compared with 1996–2006) 90 Composition of Prodi II government in Italy, May 2007 94 Three decades of democratic cabinets in Spain, 1977–2008 102 Individual characteristics of cabinet members in Spain, 1977–2008 108–109 Distribution of ministerial positions by party in Belgium, 1946–2007 130 Average age and experience per ministerial position in Belgium, 1946–2007 132 Average age and political experience of ministers in Belgium according to ministerial departments, 1946–2007 136 Reasons for resignation in Belgium, 1946–2007 138 Type, party composition and size of governments in Portugal, 1976–2005 150 Social background of ministers in Portugal, 1976–2005 152

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

x Illustrations 8.3 Political background of ministers in Portugal, 1976–2005 152 8.4 Reasons for termination of governments in Portugal, 1976–2005 154 8.5 De-selected ministers without political experience in Portugal, 1976–2005 155 8.6 Causes of ministerial de-selection in Portugal, 1976–2005 156 9.1 Swedish cabinets, 1945–2007 162 10.1 Irish governments since 1973 180 10.2 Characteristics of TDs in all Dála, 1973–2002 186 11.1 Ministerial risk and resignations in Iceland, 1944–2006 200 11.2 Public satisfaction with ministers and their fates in Iceland, 1993–2006 201 12.1 Government characteristics in post-communist democracies, 1990–2007 212 12.2 Constitutional provisions on key aspects of cabinet life in post-communist democracies, 1990–2007 214 12.3 The length of government formation periods in post-communist democracies, 1990–2007 216 12.4 The average age of prime ministers in post-communist democracies, 1990–2007 219 12.5 Reasons for resignations in post-communist democracies, 1990–2007 221 12.6 Reasons for resignations in post-communist democracies, by country, 1990–2007 224–225

Figures 2.1 The selection of cabinet ministers in Germany from different recruitment pools, 1949–2007 (first-time appointments) 27 2.2 Ministerial discontinuations in Germany, 1949–2007 32 3.1 Number of offices held by ministers in the French Fifth Republic, 1959–2005 49 3.2 The length of ministerial careers in the French Fifth Republic, 1959–2005 50 4.1 Ministerial survivor functions by government in Britain, 1945–2007 65–67 4.2 Resignations versus calls for resignation in Britain, excluding reshuffles and adjusted for size of government payroll, from Attlee to Blair, 1945–2007 74 5.1 Government ministers with experience of junior minister in Italy, 1948–2006 87 5.2 Parliamentary origin of ministers in Italy, 1948–2006 88 5.3 Female ministers in Italy, 1980–2006 89 5.4 Average duration (in years) of ministerial career in Italy, 1948–1992 91 6.1 Cabinet dynamics in Spain: cabinet size and organization (by events of change) 106

Illustrations xi 6.2 Cabinet dynamics in Spain: entries, exits, portfolio changes and continuity, 1977–2008 111 6.3 Ministerial survivor functions by prime ministerial term in Spain, 1977–2008 112 7.1 Average age and political experience at the start of each cabinet in Belgium, 1946–2007 133 7.2 Parliamentary origin of ministers in Belgium, 1946–2007 135 7.3 Number of resignations in Belgium, 1946–2007 140 9.1 Social representativeness of Swedish ministers, 1945–2007 164 9.2 Parliamentary and party background of Swedish ministers, 1945–2007 166 9.3 Occupational background of Swedish ministers, 1945–2007 167 9.4 Ministerial survival functions by government in Sweden, 1945–2007 170–171 9.5 Ministerial survival functions by background characteristics in Sweden, 1945–2007 173 10.1 Ministerial resignations in Ireland, 1945–2007 190

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

Preface

Keith Dowding first started collecting data on ministerial resignations in the UK in 1991 when Brunel University gave him some money to employ a student for work placement duty. Many research assistants later it was in 1998 that the data collection exercise finally led to a publication (Dowding and Kang 1998). Attempts to find funding for further data collection proved fruitless until 2002 when his project took on a new lease of life. Patrick Dumont supervised similar data collection by Régis Dandoy for Belgium starting 2000, which also led to a publication (Dumont et al. 2001). Torun Dewan introduced the two editors, who met in a café in Bloomsbury, London, for an hour. There it was that they decided to work together to encourage similar data collection around the world. This led to an APSA panel in Philadelphia in 2003 and then an ECPR Workshop at the Joint Sessions in Granada in 2005. It was at this meeting that the contributors to this volume met, with others, giving papers on many states not only from Europe but also Canada, Israel and other countries. At the end of the workshop the contributors agreed to form the Selection and De-selection of Political Elites network (SEDEPE). The idea behind SEDEPE is for members to collect similar data on elite selection and de-selection, not only of ministers, but also of EU commissioners and mayors thereby also facilitating comparisons with studies on the profile of legislators at various levels of governments. Data collection regarding senior civil servants and judges is also envisaged by the network in the future. It was agreed that whilst there is a large literature on government formation seen in terms of party coalition, there was no systematic comparative literature on the selection of the elites that form the heart of the state. SEDEPE hopes to fulfil the aim of discovering the institutional and personal factors that lead to elite selection, and to elite failure. This book is the first output from SEDEPE. It provides institutional background for each country in terms of ministerial selection, important for understanding the context of comparative analysis, and some data on the selection and de-selection forces that exist in many European countries. It is not the collection of papers presented at that first workshop in 2005, but rather papers written to a specific plan and structure to provide the background for further comparative analysis. The draft introductory chapter was presented at the ECPR General Conference in Pisa in 2007, in the section devoted to the Quality of European Political Elites. As mentioned above, SEDEPE is not simply concerned with Europe, nor with

xiv Preface government ministers. It is of course open to scholars other than those contributing to the present book, who can sign up at its website http://www.mzes.unimannheim.de/sedepe. Here scholars can find details of the network and comparative dataset being compiled (and to be compiled) to be made available for all scholars in the network. The website is run by Hanna Bäck, and the other important organizer in SEDEPE is Luca Verzichelli. Sign up for your own country if it is not already being covered. The editors would like to thank their families for giving them time to work on this project and this book. Patrick dedicates this book to Véronique, his patient and supporting wife, and Théo and Sylvain for the colourful life they bring. Keith dedicates it to Lagavulin Blurg who provided inspiration at the very end of this project.

Notes on contributors

Hanna Bäck is a Junior Professor at the Mannheim Centre for European Social Research (MZES), University of Mannheim. Her current research focuses mainly on coalition formation and portfolio allocation in Western European parliamentary democracies. Samuel Berlinski is a Lecturer in Economics at University College London and Research Fellow at the Institute for Fiscal Studies (IFS). He is also an editor of the IFS journal, Fiscal Studies. His main research interests are in public policy and political economy. Régis Dandoy is a Teaching Assistant in Political Science and Researcher at the Cevipol of the University of Brussels (ULB). He is also secretary of the Belgian Association of Political Science – French-speaking Community (ABSP-CF). Torun Dewan is a Lecturer in Comparative Politics at the London School of Economics and Political Science. His main research is in political economy and in the formal analysis of parties and coalitions. His current research focuses on political leadership and legislative and executive behaviour. Keith Dowding is a Professor and Head of the Political Science Program at the Research School of Social Sciences, ANU. He has published extensively in comparative and urban politics, public choice and public administration and political theory. He has been editor of the Journal of Theoretical Politics since 1996. Patrick Dumont is a Researcher in Political Science at the University of Luxembourg. His main research is in comparative politics, studying executive coalitions and their relations with legislative and party politics, and the impact of European integration on national political systems. Katja Fettelschoss is an Assistant Professor at the Centre for the Study of Democracy, Leuphana University of Lüneburg, Germany. Her research focuses mainly on career patterns of Central Eastern European ministers. Stefaan Fiers is an Associate Professor of Comparative Politics at the Katholieke Universiteit Leuven, Belgium. His research interests are the study of political elites with a focus on Members of Parliament and civic education.

xvi Contributors Jörn Fischer is a Ph.D. student at the Chair of Comparative Politics at the University of Cologne and holds a Diploma (M.A. Equivalent) in Regional Studies Latin America. Indridi Indridason is a Career Development Fellow in Formal Analysis at University of Oxford and Associate Professor in Political Science at the University of Iceland. Miguel Jerez-Mir is a Professor of Political Science at the University of Granada and former Prince of Asturias Chair at Georgetown University. His current research interests focus on Spanish political and economic elites and political science as a discipline. André Kaiser is a Full Professor and Chair of Comparative Politics at the University of Cologne and a faculty member in the International Max Planck Research School ‘The Social and Political Constitution of the Economy’. Christopher Kam is an Assistant Professor of Political Science at the University of British Columbia. His current research examines the logic of cabinet reshuffles in parliamentary democracies. Gunnar Helgi Kristinsson is a Professor of Political Science at the University of Iceland. He has written extensively on Icelandic politics (political parties, constitution, the executive and public administration). Csaba Nikolenyi is an Associate Professor at the Department of Political Science, Concordia University. He is interested in the comparative study of electoral and party politics, especially the impact of electoral systems on party competition, and the formation and durability of electoral and executive coalitions. Eoin O’Malley is a Lecturer in Political Science in the School of Law and Government in Dublin City University. His main research interest is in executive power in parliamentary democracies. Thomas Persson is a Postdoctoral Fellow at the Department of Government, Uppsala University. His research interests include executive-legislative relations and the impact of European integration on national political systems. António Costa Pinto is a Professor of Modern European History and Politics at the Institute of Social Science, University of Lisbon. He co-edited recently Southern Europe and the Making of the European Union (2002) and Who Governs Southern Europe? (2003). José Real-Dato is a Lecturer at the University of Almería (Spain). His current research interests are the study of political careers and Spanish and European Union research training policies. Gita Subrahmanyam is a Research Associate with the London School of Economics and Political Science Public Policy Group. Her research examines the impact of constitutional arrangements and elite collusion on state operations in Europe, Africa and Asia.

Contributors xvii Pedro Tavares de Almeida is a Professor of Political Science at Universidade Nova de Lisboa. He has written extensively on political elites and elections. Recent books (co-edited) include Who Governs Southern Europe? (2003). Kåre Vernby is a Postdoctoral Fellow at the Department of Government, Uppsala University and visiting scholar at Harvard University during 2007–8. His research interests lie in the fields of political economy and comparative politics. Luca Verzichelli is a Professor of Political Science at the University of Siena. He has extensively published in the fields of comparative political elites and political institutions. Recent books include Political Institutions in Italy (2007), co-authored with Maurizio Cotta. Helena Wockelberg is a Postdoctoral Fellow at the Department of Government, Uppsala University. Her research interests include the relationship between cabinets and public administration, accountability procedures and the impact of the EU on national agencies.

1

Structural and strategic factors affecting the hiring and firing of ministers Keith Dowding and Patrick Dumont

The centrepiece of representative democracy is the accountability of the government through the electoral process. However, if accountability was only composed of a chance to throw the rascals out every three to five years or so, especially if one only has rascals with which to replace them, then representative democracy would not provide much accountability. In reality, there are many other ways in which governments can be, and are, held to account. Constitutions specify the limits of governmental power; courts ensure that legislation as written and as implemented conforms to constitutional and legal requirements; the press and mass media keep an eye on the actions of members of the government; and of course parliament, even when controlled by a majority party, performs an oversight role. When accountability affects the whole government we can talk about collective responsibility; sometimes individual ministers are held to account for the policies and actions of their department, as well as for their own actions. Part of accountability is that ministers inform the public about what their departments are doing and explain when things go wrong, and to make changes when these are necessary. When a minister proves incapable of making amends then the rest of the cabinet might withdraw protection and here the minister is held to personal account. Underlying collective responsibility is the threat that the government might fall – through a vote of no confidence; through constitutional coup forcing an election; or simply through losing a regular election. That threat is ever-present in the mind of government ministers, especially the prime minister who is herself accountable to her cabinet as well as parliament and the people.1 For that reason politicians are constantly interested in opinion polls to see how the government is faring today; and in focus group or polling research suggesting how future policies might fare in the arena of public opinion. Not that the fall of the government is always considered a disaster by all members of a government. Different parties within a coalition government may view its fall in a somewhat different light, and one coalition partner might choose to withdraw from government in order to hasten that fall. The eyes that politicians keep pinned to public opinion are directed not only at the popularity of the government but also at the relative popularity of the partner parties and factions that compose it; and the popularity of the prime minister and other leading ministers who might be challengers for that top position. Choosing when to be

2 The selection of ministers in Europe loyal to a given prime minister and when to withdraw will rely upon such strategic considerations as much as real policy concerns. Democratic politicians may be driven by ideology, or principles, or the desire to promote equality and prosperity, in short to be policy-oriented. But they cannot ignore how those policies are viewed or how they work out in practice. Some prime ministers might try to control the whole of government output; others prefer to allow their ministers to take the lead on strategy and simply to manage conflicting departmental policy initiatives and facilitate coherent government. Either way round prime ministers want policies that fit with their guiding principles and can be practically implemented, and, at least relatively, popular. In that sense prime ministers can be viewed as the principal with ministers as the agent. However, the complex strategic relationship between ministers within a cabinet, and between ministers and the prime minister, ensure that this is not a simple principal–agent relationship. The agents do not simply have interests at variance with the principal, but have competing interests and allegiances with other agents involved in the government.2 Governments are formed of cabinets of ambitious politicians. Some of these politicians will see heading a department as the pinnacle of their career; others might hope for further preferment, either to a more important post or to the premiership (or presidency) itself. When forming a cabinet the chief executive ideally wants team members who will complement each other. She might want some ministers who will be dynamic and produce new policy initiatives. These she might choose for posts in areas where it is widely agreed that there has been government failure or where she herself believes a new agenda is required. She might want a department to be reorganized and needs a person prepared to fight and defeat a recalcitrant bureaucracy; or a person adept at handling a sceptical media. She will also want some safe hands, ministers who will be able to handle their department quietly and efficiently and not desire to make waves. It is likely that the latter are those who are content at their stage of the careers, with the former being ambitious and perhaps future rivals to the chief executive herself. A chief executive might have an ideal cabinet in mind, but the availability of personnel and party political and constitutional constraints might not allow this. In some countries the constitution specifies that specific demographic groups must be represented or balanced in cabinet. Party constraints also take several forms. Powerful people with the backing of sections of the party need to be found jobs. Different factions within a party must be satisfied. In coalition governments the process of putting a cabinet together sometimes requires long discussion with coalition partners about the shape of the cabinet, which party will get which posts and who will fill those posts. Putting together a cabinet can be a tortuous process. Politicians invited to take on a government post must also consider their options carefully. Coalition negotiations often involve discussions over specific policy portfolios and the offer to a given politician for one of those portfolios might already have been bargained and refusal difficult. However, politicians do turn down posts in a number of circumstances: if they feel the post is too junior or unimportant, in policy areas that they do not feel is the right area of expertise for them, when they do not support the type of coalition formula or the government

Structural and strategic factors 3 agreement arrived at during negotiations, or because they feel the post is a poisoned chalice. A cabinet must work together for the collective good – that is the idea behind collective cabinet responsibility – and is required for the government to remain popular and be re-elected. But there are always tensions and rivalries amongst cabinet ministers. Even in single-party governments there are personal rivalries for the top jobs and ideological tension between wings and factions within the party. For coalition governments these tensions are exacerbated by cross-party conflict. The games that are played across cabinets are complex and intense. Ministers may well brief against each other and the prime minister, if not openly then through friends and colleagues. In single-party governments perhaps no member wants the government to fail as a whole, though many cabinet ministers might enjoy seeing others including the prime minister in trouble. They must calculate how much they might be personally damaged by publicly speaking out against colleagues, and so how much support they must give at any one time. And then calculate when to pounce, when to speak and suggest a colleague needs to go. The strategic issues are more complex in coalition governments. Here it cannot be assumed that all actors do not want the government to fail. Parties are more interested in the poll rating of the parties than the government as a whole, since new elections may strengthen their bargaining power. Coalition partners may get together for necessity rather than anything else, and each party in the coalition wants to get credit from the public for successful policies and may well desire their partners to fail in their own portfolios. Here the hiring and firing of ministers may well take a very different form from single-party governments. A chief executive might well enjoy seeing a rival party put an incompetent into an important portfolio and be reluctant to see that minister resign even after repeated failure. His failure might well be bad for the government as a whole, but such failure might well be accompanied by a rise in popularity of another party of the coalition. It can be seen therefore that there is a large number of strategic issues that enter into decisions both to hire and fire ministers – and for ministers to decide to accept appointment and leave the government of their own accord – and these interact with constitutional and party political constraints. Until relatively recently there has not been a great deal of dedicated research into the selection mechanisms for cabinet ministers, or over the causes of their removal. In comparison, for example, there is a broad body of both theoretical and empirical work examining the formation and dissolution of governments (for example, Riker 1962; Schofield 1987; Laver and Schofield 1990; Laver and Shepsle 1996; Warwick 2005). Theoretically and empirically much of this work examines parties as single unitary actors although it is widely recognized that coalitions can be broken up by splits caused within parties (Cox 1997; Giannetti and Laver 2001; Mair 1990, 1997; Cotta 1996; Damgaard 2008).3 Of course, governments are made up of individuals and there is a great deal of literature on the representative role of political elites, especially examining the personal characteristics and background of ministers (Blondel and Thiébault 1991; de Almeida et al. 2003) and even more literature examining the policy differences that leaders (Blondel 1987, Kavanagh 1990, Elcock 2001, Elgie 1995) and ministers make (Headey 1974, Blondel and

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

4 The selection of ministers in Europe Müller-Rommel 1993; Chabal 2003). However these examine the impact on policy that personality and structure can have. There is less on personnel hiring and firing. De Winter’s (2002) ‘state of the art’ piece on parties and government states that of all aspects of government formation – party composition, portfolio allocation, policy definition and personnel selection – the last has received the least attention from scholars. In that he reiterates the claim of Blondel (1985, p. 8) that the study of ministers and ministerial careers is still in its infancy. And overall, as argued by Strøm et al. (2008), what happens between government formation and collective termination is still poorly documented and understood. Recently however there has been a flowering of articles examining precisely the issues of ministerial selection and de-selection using dedicated datasets (see Dowding and Kang 1998; Dumont et al. 2001; Huber and Martinez-Gallardo 2004; Dewan and Dowding 2005, O’Malley 2006; Fischer et al. 2006; Berlinski et al. 2007a; 2007b; Indridason and Kam 2008). Most of these studies are single-country as complete cross-country comparative work on features common across the selection and de-selection of ministers has still to be done. Examination across these single-country studies for common features is possible, though researchers must be aware of institutional differences in the hiring and firing process that make comparison problematic. This book is a contribution towards the examination of the strategic issues involved in the hiring and firing of ministers in a comparative context. It emerged from an ECPR Workshop held in Granada, Spain, in 2005, though none of the papers included here was presented in this form at that conference. One of the conclusions of that workshop was the formation of a network of scholars dedicated to collecting data on the hiring and firing of ministers and other political elites. This book constitutes the first output of the Selection and De-selection of Political Elites (SEDEPE) network. The chapters in this book all examine institutional features of the respective European countries that affect the hiring and firing of ministers. Any comparative analysis of ministerial selection and de-selection must take into account specific structural factors, which we have broken down further into constitutional and party political issues. Both these structural factors impact upon the strategies that actors might adopt. Our third category is thus strategy as impacted by structure and the events which agents use for strategic advantage. Rather than introducing the chapters individually in the rest of this introduction we will examine these structural and strategic factors in a broad brushstroke under the two headings of ‘Hiring’ and ‘Firing’. Constitutions directly affect the hiring and firing of ministers in terms of what is allowed or required in the construction of a cabinet. The organizational rules and conventions within parties also enable and constrain the construction of cabinets in ways which vary across countries, and across parties within countries. The third element concerns the strategies that might be adopted by prime ministers or presidents in constructing cabinets. Different constitutional and party rules, and the nature of the specific bargaining process across factions or parties in coalition governments, affect how prime ministers might go about trying to construct their preferred cabinet. Comparative analysis across countries can reveal the structural factors (both formal in terms of constitutions and rules, and informal in terms of expectations and conventions) which cause variation in the strategies available.

Structural and strategic factors 5

Hiring Theoretically we can consider a government formed by a chief executive – a prime minister (in most cases in Europe) or a president – who can be considered the principal who hires agents – her ministers – to devise and oversee implementation of policies. How much discretion is given to ministers to devise policy and how much they are directed by the chief executive (who we will often assume is a prime minister) will depend in part upon her particular personality type. Some prime ministers will be more ideologically or policy-driven than others and want to get involved in the policy detail, becoming occupied and associated with specific policies. Others may also want to take the lead in some policy areas if they expect it to be rewarded in terms of personal popularity and votes. Yet others still are much happier to delegate, either because they are less policy-driven or because they simply believe that governing involves consensus and teamwork. A large part of the type of leadership chief executives bring might well depend on whether they occupy the median position in their cabinet or whether they are cabinet or party outliers needing to force others into their way of thinking. Of course, how successful chief executives can be in this process depends upon the constitutional, party and practical or strategic environment within which they operate. It is easier in some countries than others for a prime minister to become more ‘presidential’ (indeed easier than it is for some presidents) than it is in others (see Foley 1993, 2000; Poguntke and Webb 2005). Taking the amount of delegation a prime minister gives to ministers as a given, what would a chief executive want from his ministers? We might consider what the ideal minister might look like. Those on the inside proclaim that a good minister must be able to make decisions; he should be able to handle his civil servants and manage his department, read briefs quickly and take advice (Kaufman 1997). He must be able to deal with questions in parliament and politicians generally; and deal with the media, coming across well on television – this latter aspect especially in the most important posts. He ought to have some policy ideas of his own, and be able to develop them without pushing impractical policies against civil service advice. He would do well to have a stable and supportive family, and keep well away from financial, social and sexual scandals. A serious mind and ready wit would also do him well, and for some more technical portfolios a certain degree of expertise (either acquired outside the realm of politics or in parliamentary committees or even in similar executive positions at other government levels) may be regarded as an asset, though the German Chancellor Helmut Schmidt allegedly once pretended that ‘with a little more than average intelligence anyone can do that’ (in Bahr 1996: 463, cited by Fischer and Kaiser in Chapter 2 below). In any case, not all ministers can fit the ideal, and, indeed, a Prime Minister might not want them all to live up to such an ideal, unless she is ready to face so many potential rivals. In fact she will probably also want (or will be forced by constitutional or party constraints to look for) a mix of different types of ministers in her cabinet, not least in order to meet representative requirements, in terms of age, gender or geographic origin. A balance in personality types may also be sought. A prime minister might want some ministers who are dynamic in the sense

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

6 The selection of ministers in Europe they provide policy initiatives, lead the political agenda and respond to crises effectively. However, dynamic and charismatic ministers might also provoke greater policy controversy and lead more hectic personal lives. So she might also want some more statesmanlike, constant, steady and secure ministers who are seen as safe hands. One might think a chief executive would also want a cabinet that works together well as team players, supporting each other. However, as we mentioned above, strategic considerations enter in, and prime ministers often seem to thrive on playing off factions against each other, and once we have coalition governments all sorts of strategic games might come into play. It is not necessarily the case, especially in faction-ridden or coalition governments, that a prime minister will want a minister in every portfolio who carries out policy that the prime minister ideally wishes. She might have had to buy the coalition by allowing a party to run a particular portfolio in order to pursue a certain policy. The prime minister might be happy to see that policy fail. She might be happy to see someone who is less competent fulfilling a role in a particular position in order to see that policy fail and tar the faction or party of the minister. In other words it is not clear how far traditional principal–agent models hold for these complex strategic considerations. It also means that it is not at all clear what general statements can be made about the desirable form of cabinet agents that all prime ministers might desire. And whatever a chief executive may ideally desire, she is hemmed in by constitutional, party and strategic considerations with regard both to hiring and to firing. We will base our discussion on those three types of considerations illustrating with examples from different countries. Constitutional constraints Assuming that a chief executive wants to provide as talented a cabinet as possible, the size of the ministrable is important. In some countries ministers are drawn from the legislature, though it varies cross-nationally whether in bicameral legislatures they can be drawn from either house. In the UK for example, ministers can be drawn from either the House of Commons or the House of Lords. Most are however drawn from the Commons though spokesman for each ministerial department will be appointed for the second chamber. In Ireland the incoming prime minister is allowed to nominate eleven of the sixty senators, which allows her, at least in theory, to enlarge her recruitment pool which is constitutionally limited to the legislature.4 In many countries ministers will retain their position as members of the legislature, but in some countries they will resign representative posts when they serve as ministers. For example, in Bulgaria, Estonia, Slovakia or Belgium since 1995 ministers must take a ‘sleeping mandate’ and temporarily relinquish their legislative roles, which they can retrieve when they leave government. Similarly special conditions apply in some countries during caretaker governments in countries such as Italy and Portugal. In other countries ministers cannot be members of the legislature (for example, the United States) and therefore must resign from their elective mandate owing to incompatibility of functions, and in others still they can be drawn either from the legislature or from outside it (for

Structural and strategic factors 7 example, Germany and Spain).5 There might be advantages in drawing ministers from outside the legislature in terms of their experience and knowledge of affairs and in broadening the quality base, though there can be big disadvantages if those without political experience face a hostile parliament and media. Prime ministers may think twice before hiring such outsiders with technical-professional expertise, given the potential for informational asymmetry and difficulties in controling their action even within cabinet. It may therefore be safer to look for some form of expertise amongst professionals of politics, such as through experience in parliamentary committees, or in political positions at supra- or sub-national levels. In some countries there are rules specifying the composition of the cabinet. In Belgium for example a constitutional provision aimed at protecting the Frenchspeaking minority ensures that the Council of Ministers is composed of the same number of French-speaking as Dutch-speaking ministers. The prime minister is excluded from the calculation of this ‘linguistic parity’ rule, and so are junior ministers (secretaries of state) who are not formally part of the Council of Ministers. Since 2002 the Belgian constitution also imposes the representation of both genders in the Council of Ministers. The size of government may also be constitutionally specified (for example in Ireland as well as in Belgium since 1995, where the number of ministers cannot exceed 15), a feature that may constrain prime ministers trying to find a balance between sometimes numerous coalition parties (see Mershon 2002 who develops the idea of expanding the number of cabinet posts to cope with office costs of including additional coalition parties). Another constraint facing a chief executive is whether the cabinet will face a formal vote of confidence once it has been formed, as in Belgium, Ireland, Italy, Portugal, Spain and Sweden.6 Of course, no-confidence votes against the government as a whole can be taken in all legislatures at any subsequent time, but in countries such as Belgium, Iceland, Italy, Sweden or the Baltic States of Estonia, Latvia, Lithuania and also Poland votes of no confidence can be taken over individual ministers as well as the cabinet as a whole (see below). These features can make a big difference to the discretion of a prime minister who can be constrained in choosing ministers who might court controversy or be unpopular within the parliament. The role of the head of state in cabinet appointment may also matter. In most countries the chief executive (whether prime minister or president) can hire largely without constraints even though officially ministers are appointed by the head of state (for example in the UK; in Sweden since the 1970s the king is not even part of the formal process any more). Notice that even in parliamentary monarchies or republics with a non-elected head of state, some choices may be vetoed or influenced by the latter (this has sometimes been the case for instance in Belgium). In other countries, for example in France, the choice of ministers is more of a bargained affair between the president and the prime minister and who dominates depends upon other practical political constraints (for instance the president clearly has the upper hand in times of unified government). Indeed, broader constitutional provisions regarding executive–legislative relations influence the selection (and de-selection) of cabinet ministers. Whether

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

8 The selection of ministers in Europe government arises out of the legislative assembly which can at all times put an end to executive power, as in parliamentary democracies, or is named and directed by a chief executive popularly elected and does not rely on parliamentary confidence, as in presidential democracies, has implications on the composition of cabinets (Shugart 2006). In the latter, chief executives’ (elected presidents’) legitimacy emanates directly from voters, and their quest for the preferred policy positions of the median voter may well diverge from that of their party fighting parliamentary elections in multiple constituencies. As a consequence, choosing ministers exclusively within their party ranks may not serve their interests. The absence of mutual confidence mechanisms between executive and legislative powers also frees presidents from the constraint of appointing MPs in order to ensure good relationships and smooth working relationship with the legislature. Semi-presidential regimes, which have grown in number following the democratization of the former Communist bloc, combine popularly elected presidents with sometimes considerable constitutional authority (such as in France the right to dissolve the assembly) and a cabinet headed by a prime minister accountable to parliamentary majority. Such constitutional features may also lead to the selection of greater proportions of ministers with no party background (see Amorim Neto and Samuels 2003; Schleiter and Morgan-Jones 2007). The organization of the state may also have an impact on personnel selection, as in federal systems the existence of sub-national executive positions may dry out the pool of ‘ministrables’ at the national level because of the incompatibility between these executive offices, but may on the other hand represent a springboard for ambitious politicians (for example in Belgium, Germany and Spain). Hence, even without considering strong constitutional customs or practices which are widespread in government formation, we can see that formal constitutional provisions may constrain chief executives and party leaders in their task of cabinet composition. We now turn to less formal, but in some settings possibly more effective, constraints. Party political constraints Where single-party governments dominate, as in the UK and France, the choice of ministers is more straightforward than with coalition governments. That is not to say that party political factors do not play a major role in selection however (Alderman 1976). In some parties the initial choice of the cabinet is dominated by party considerations. In 1933 the National Executive Committee (NEC) of the British Labour Party decided it would influence ministerial choice (constitutionally it could not choose it as that formally belongs to the monarch) but Attlee ignored that in 1945, as have subsequent Labour prime ministers (McKenzie 1963). In most countries parties do not have much formal control over choice of ministers, but there are practical or strategic party matters. Chief executives need to keep the different wings and factions in the party happy. They cannot ignore powerful colleagues that command great party or parliamentary support and are well advised to heed Don Corleone’s advice to his son Michael to ‘keep your friends close, but keep your enemies closer’. And such choices are not simply subject to the choice

Structural and strategic factors 9 of whom to include into the cabinet, but its composition also. Often powerful politicians can refuse posts they do not fancy and demand specific portfolios. In Australia for example, the Deputy Leader (who is often a losing rival in leadership battles) is traditionally allowed to choose the post he wants (often the Treasury), as the leader of the party fills the position of chief executive. In coalition governments the party effect obviously is much greater. Here the very formation of the coalition itself is founded upon the distribution of seats amongst parliamentary parties, and therefore party leaders often have a greater say in ministerial selection than chief executives. Interestingly, despite the predictions of formal coalition theory, the proportion of government seats parties receive tends to reflect their contribution in parliamentary seats to the coalition, with some bias towards smaller parties (Gamson 1961, Browne and Franklin 1973, Budge and Keman 1990).7 The distribution of seats also reflects both power balances and policy interests; the trade-off for the premiership for one party might be reflected in other powerful posts such as Finance or Foreign Affairs (the traditional post of the German FDP for example) going to coalition partners. Although important efforts have recently been made to weight a wide range of portfolios (Druckman and Warwick, 2005; Druckman and Roberts 2008), we still know little about how parties in practice weigh portfolios and which criteria they use to do so: policy relevance according to their electoral manifestos and campaign slogans, interests of traditional or new clienteles, patronage opportunities, prestige, visibility, distributive versus redistributive policy departments, goodness of fit with the qualifications of their pool of ministrables available in the party, personal tastes of party heavyweights, and so on, even though we know that core ideological orientations and social roots still influence portfolio allocation, with parties of the left demanding welfare posts and agrarians or Christian Democrats often found at the head of agriculture ministries (see Budge and Keman 1990). An instance of such a colonization was found in Belgium from 1958 to 1999, with the CVP constantly in power and at the helm of the agriculture ministry, mainly owing to the presence in the party of a faction (Boerenbond) representing the interests of the farmers. Factions like these can also play a major strategic role. As Laver and Shepsle (1996) suggest, an apparent weakness such as an ideological split in a party can be a bargaining strength, as a leader might be able to demand more and specific posts in order to be able to deliver his support in parliament. They may also enable their leaders to balance the cabinet and make the median policy position of the coalition acceptable to all intra-party factions, by hiring for instance prominent personalities representing the left wing of a centre party in coalition with a party located on the right of the political spectrum, as was often the case in Germany. Aside from horizontal party organization, intra-party democracy also matters, as party congress may refuse governmental participation or constrain the choice of cabinet ministers (for instance in Belgium in 1999, the Green leader who was the main negotiator of his party delegation to coalition talks was denied a ministerial position by his own party congress). Finally, especially in coalition cabinets (as no party is supposed to be so large as to form a majority government on its own), the size of governing parties may also constrain chief executives and party leaders in their ministerial personnel choices (Laver and Shepsle 2000) – even though the

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

10 The selection of ministers in Europe positions to be filled will be roughly proportional to their parliamentary contribution to the coalition – as smaller parties will arguably have a more restricted pool of talent to recruit from. Strategic issues Some of the party political issues we have identified are strategic, but other strategic factors also come into play. We mentioned above the need for chief executives to balance high-flying, ambitious and dynamic policy-makers with more stable and statesmanlike ministers. A chief executive faces a strategic problem when putting together a cabinet. She will want ministers, particularly in some problematic issue areas, who can develop new and perhaps radical policy initiatives in order to manage problems that are in the public spotlight. However, policy initiatives will always generate some controversy. Entrenched interests are likely to oppose radically new policies unless they can be bought off somehow, and any such rents are likely to be controversial themselves. Thus new policy initiatives in some issue-areas need to be balanced with a quieter approach in other areas. If the cabinet is full of too many dynamic and ambitious ministers then they will fight over the legislative timetable and perhaps brief against and try to block the others. Thus a prime minister needs to counter-balance her dynamic ministers with a few quiet ones. This is especially so, if, as seems plausible, the more ambitious and dynamic the personality the more likely they are to have a complex social and financial life. We have not mentioned junior ministers so far. In the postwar period the number of ministers chosen across democracies has risen as the scope of state government has grown. The numbers of full cabinet ministers have not grown nearly so far, rather junior minister posts (and intermediate or ‘non-full-cabinet-ministers’) have been created, escalated in number and taken over many responsibilities. The increase in the number of junior ministers opens up new strategic possibilities to a prime minister.8 First, it enables her to promote up-and-coming politicians who then owe her some loyalty. She might even buy the loyalty (given collective responsibility – not voting against the government one is a member of – which operates in most countries) of otherwise potential hostile or disappointed backbenchers, as is arguably the case in Ireland. Having junior positions also allows a chief executive to gauge how well a candidate is likely to perform at more senior levels (Kaufman 1997; Manow 2005) helping to overcome adverse selection problems. Secondly, there is a tradition in many countries (such as Italy) with coalition governments having junior ministers from one party ‘shadow’ the ministers of another. In single-party majority cabinets the prime minister can also keep an eye on senior ministers she might mistrust by placing a trustee in the junior position. It is open to question how much office-holding (that is, satisfying parties and factions in coalition bargaining and fine-tuning balance in coalition using these positions as spare change to round off otherwise imperfect deals) and how much policy questions (that is, shadowing ministers to see what they are up to) dominate handing out junior positions (Manow and Zorn 2004). In a similar fashion, if chief executives are allowed to appoint ministers without portfolio, they have additional

Structural and strategic factors 11 opportunities either to promote political friends to cabinet rank or on the contrary to involve rivals in cabinet if they are forced to select heavyweights, but this way at no office cost, and leaving them with no credit for specific actions in a ministerial department.

Firing Of course the hiring and firing decisions are closely connected. For every firing between electoral changes of government there will almost certainly be a hiring, even though replacements close to elections (when the end of the parliamentary term is near or the cabinet functions as caretaker during the formation of a new one) are not systematic; but powers of chief executives to fire are not precisely symmetric with their powers to hire. The most obvious asymmetry is that ministers may resign at their own discretion but cannot be hired at their own discretion. Thus whilst the strategic considerations over hiring are almost entirely in the hands of the chief executive, within party leaders in coalition systems and sometimes head of states’ constraints, resigning can involve strategic interaction between various agents.9 Ministers can choose to go at times that will embarrass the prime minister. They may choose to resign over some policy that will gain them kudos within their own party,10 or to enhance their own party in coalition governments. Just like the calling of anticipated elections by (usually single-party) incumbent governments, the unilateral mass resignation of ministers to breakup a coalition can be timed for party advantage if an election can be forced, though such manoeuvres can backfire if the public believe they are too crudely based on party-power considerations rather than what is really good for the country. The prime minister can also choose the time when ministers go, often through a strategic reshuffle of the cabinet. Again we discuss these issues under the headings of constitutional, party political and strategic considerations. Constitutional constraints Similarly to hiring, some chief executives have almost complete de facto if not de jure (usually kept by the head of state) power over firing ministers. In other countries (this is quite clearly the case in semi-presidential regimes, such as the French case) the decision to fire, as to hire, needs to be negotiated between the prime minister and the head of state. In presidential and semi-presidential regimes, where the head of state has a separate political legitimacy from the legislature, we mentioned that the hiring of less politically experienced ministers was more likely than in parliamentary democracies. One can expect that ministers with such an ‘outsider’ profile may be easier to sack, as they do not have the same kind of electoral legitimacy as prominent MPs for instance, or the support of a political party or a faction that may be able to veto such a decision. Legislatures may sometimes have the right to veto sackings but generally speaking legislatures can only show displeasure at the prime minister or government itself following any reshuffle. However, legislatures may themselves trigger the resignation of ministers, by voicing criticism and putting pressure on certain ministers, and for

1 2 3 4 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

12 The selection of ministers in Europe countries in which no-confidence votes directed at individual ministers exist (see above) this pressure may be much more effective than where such an instrument is absent. In countries where ministers are drawn from parliament, this constitutional requirement may interact with strategic issues and reduce the chief executive’s ability to sack ministers or reshuffle. This is clearly the case in Ireland where the occurrence of minority or bare majority cabinets makes prime ministers well aware that dismissing individual ministers may cause cabinet severe trouble in parliament, as the sacked minister is likely to retaliate against the government through his voting behaviour. The small size of parliament in Ireland reinforces this constraint, which also materializes in Iceland which, despite the absence of formal constitutional provision of choosing ministers from parliament, has the same practice. There is a small pool of ministrables and ministers therefore enjoy a relatively high degree of job security. Where rules exist that specify group representation, such as in Belgium, firing a minister can involve a major reshuffle as a new minister from the same language group needs to be hired and the group belonging of the best possible substitute may not always fit into the post that has been vacated. It can even involve asking another minister to resign at the same time simply to enable a reshuffle that fits constitutional requirements of parity in the Council of Ministers, or sacking the same number of ministers from each linguistic group in case of decisions to cut down the size of cabinet (there were instances of each phenomenon in 1976 and in 1973). Of course, these considerations mean that a prime minister is greatly constrained in firing, and, at least in theory, ministers could well survive scandals that they would not survive in other countries simply because of the added constitutional problems involved in forcing resignations. Finally, the existence of multiple levels of government may also induce different patterns of ministerial resignations. Ministers sometime move to prestigious positions in international organizations: for example the Swedish Dag Hammarskjöld resigned in 1953 to become General Secretary of the United Nations; P.-H. Spaak in 1957 and Willy Claes in 1994 both from Belgium resigned to take up the post of General Secretary of NATO; whilst Franco Frattini, Italy’s Foreign Minister, became vice-president of the European Commission in 2004. Furthermore, the availability of politically relevant sub-national functions may well have an impact on the ability of chief executives to keep even their most talented ministers. In Spain some national ministers resign to fight electoral campaigns in their region with the ambition of becoming minister-president, whilst in Belgium they tend to resign after elections are held and their executive position in regional government is secured. Overall then, one must not forget that, even when we do not count age or health issues as reasons for termination, ministerial resignation from a national mandate does not always equal political punishment, as important positions at other levels of government may be considered as even more prestigious than that of line minister in a national executive.

Structural and strategic factors 13 Party political constraints In majority-rule governments, prime ministers can usually sack and reshuffle knowing that they can command the support of their party in doing so. In coalition governments such assumptions may well not hold. Certainly, generally speaking prime ministers would fire ministers from another party only after discussion with the relevant party leader. Where scandal is damaging the minister’s party his party leader may well agree or even instigate the resignation. Threats to a minister of another party in a coalition usually lead to a strong defence, so we expect less ministerial turnover in coalition governments (Budge 1985; Huber and MartinezGallardo 2008). Even deciding to reshuffle ministers from her within own party might prove difficult for a chief executive heading a coalition government, as sometimes equilibriums reached in terms of policy programme and personnel during cabinet formation may be dangerous to alter. For example, the tendency of large centre parties to include prominent left-wing figures from their own ranks when coalescing with junior partners on the right of the political spectrum – and vice versa in Germany for instance – can make it more difficult to remove these outliers if they get into trouble. Furthermore, unilateral changes may disrupt mutual trust amongst coalition partners. Also, in countries like Iceland and in the past Labour governments in Australia the prime minister’s decision to reshuffle cabinet posts amongst ministers needs the approval of her party appointing body (parliamentary group, caucus or central committee). Of course, such threats may well be tactical as a prime minister may appeal to a different wing or faction in her coalition partner, in order to destabilize the current leader of that party, but conversely junior coalition partners may also sometimes force the resignation of high profile ministers of the prime minister’s party (as has happened in Ireland). In coalition systems in particular, whether or not party leaders who decide on the party delegation in government enter cabinet may also make a difference in terms of cabinet stability and ministerial turnover. Especially for junior partners it is tempting for the party leader to send another heavyweight as deputy or vicePM and monitor government action from the party headquarters. As members of the party team in government owe their position to a party leader overseeing them, they have few incentives to depart from the positions defended by the leader during the formation and the life of government. Turnover may therefore be even less likely than a cabinet breakdown in case of conflict between coalition parties. When the party leader enters cabinet and leaves party chairmanship, cabinet stability may be strengthened but line ministers of the party become subject to two potentially conflicting loyalties – that towards the leader who chose them (and the cabinet as a whole to which that leader has tied his reputation) and that towards the new leader – thereby increasing the risk of ministerial turnover. This logic also applies for single-party governments where ministers chosen by a previous prime minister and ‘inherited’ by another taking over this responsibility during a legislative term (without new elections) face more risks of being de-selected than ministers who are selected by their current chief executive.

14 The selection of ministers in Europe Strategic issues There are at least three reasons why a prime minister will sack a minister. First because of a policy disagreement: a minister who does not toe the governmental line may be asked to leave. Of course, as we have mentioned above, ministers can also use policy disagreement as a means of damaging the prime minister and promoting their own cause. If the minister is relatively young he can return and further his career after displaying some ideological or policy purity. Older ministers may also choose to resign over policy, realizing that this is the end of their career, but do so in order to pay back old debts, or simply because they feel the party and government has moved too far away from the principles on which they joined (this was twice the case of Jean-Pierre Chevènement in France). Secondly, a minister might resign or be sacked following some personal or policy scandal. Dewan and Dowding (2005) have shown that in the UK scandals reduce the popularity of a government and getting rid of the minister involved in the scandal has such a ‘corrective effect’ in terms of government poll ratings that the prime minister may even be given credit for doing this. Of course, this means that a prime minister cannot always sack a minister who is criticized. If she did so the public could not trust the prime minister’s firing rule – she could be doing it just for popularity’s sake and be simply considered weak or ‘trigger happy’. And if other parties or the media learned they could force ministers’ out easily through criticism this would encourage such criticism. Thus prime ministers must learn to defend their ministers who are in trouble even when criticism damages the government and decide to sack only when they feel it is strictly necessary. For instance Belgium in the second part of the 1990s was characterized by a crisis of confidence in political institutions and several scandals led to the resignation of individual ministers. The context suggests the chief executive was prone to reactive behaviour when accusations of sexual affairs were levelled against the French-speaking socialist party vice-PM by the press with further reverberations in parliament. Prime Minister Dehaene, leader of the Flemish Christian Democrats, kept his confidence in his vice-PM and the affair died out when allegations proved to be false a few weeks later. In some cases, one way around the problem of not being able to sack a minister who the prime minister feels is incompetent at the time of a scandal is to defend him then, but then ease him out under the cover of a general reshuffle a few months later. This leads to the third reason why prime ministers sack ministers. Reshuffling the cabinet is a way of revitalizing the government. It can involve shifting some of the chairs, but also bringing in new faces, possibly through retirement of older ministers who may be offered other job sinecures outside the cabinet, Berlinski et al. (2007a) for example demonstrate that older ministers and those who have served in previous cabinets have a higher hazard than younger, less experienced ministers. The prime minister will also strategically analyse the amount of trouble a minister faces. A naive firing rule would be to rate ‘troubles’ (personal scandals, implementation fiascos, policy controversy or simple long-term criticism) on a scale and once a minister had passed a given ‘trouble threshold’ move him out. However, a prime minister might well provide greater protection for his

Structural and strategic factors 15 friends than his enemies. On the other hand, a close rival might suffer more longterm damage from being kept in cabinet (and under scrutiny) when doing badly than being sacked early (‘keep your enemies closer’). The prime minister could then get rid of his rivals when they are weakest and no longer seen as potential successors. A prime minister might choose to sacrifice political friends who have not been shown to be adept, or to clear out ministers from rival factions when she feels stronger and more secure in her position after some time in power. A prime minister can also be hemmed in by squabbling factions and find it very difficult to fire anyone lest it be seen that she is backing one group against another that could lead to coalition, government or party break-up. Again getting rid of ministers is easier when the ministrable is larger or of higher quality. Thus we might expect to find that the larger the pool of candidates relative to the size of the cabinet the faster the turnover, all other aspects being equal. Thus where ministers are drawn exclusively from parliaments the larger the parliament (and the larger the government majority) the faster the turnover, so small countries with smaller legislatures might have lower turnover than larger countries (see Dumont and Varone 2006 on the direct and indirect impact of size of political systems in the delegation and accountability chain of parliamentary democracy). Where data exists over relatively long periods of time this seems to be true. Iceland and Ireland seem to have lower rates of turnover than the UK or Germany for example. We might expect even higher rates of turnover when ministers are drawn from an even wider pool of candidates outside the legislature. Of course other factors might compound such comparisons – such as the political skill of non-politicians and the degree of protection that is afforded by prime ministers dependent upon the factors discussed above.

Conclusions In this introductory chapter we have started the process of examining the sorts of issues that will affect the hiring and firing of ministers. We have started to examine how the three sets of institutional features we have identified will affect those strategies. Constitutional factors constrain the ability of chief executives to put together the cabinet they might otherwise covet. Party-political constraints also affect chief executives of government. The specific organization requirements of parties may give chief executive more or less discretion in choosing ministers, as do the reality and power of party factions. A clear distinction needs to be made between single-party and coalition governments, as chief executives’ leeway in hiring and firing ministers is arguably quite different in these two types of contexts. We have examined a few of the strategic considerations that affect prime ministerial hiring and firing rules and how these are affected by the environmental and structural features of different governments. For example, accountability seems more straightforward in governments where one-party rule is the norm since the ‘corrective effect’ of firing incompetent ministers seems straightforward, but correcting for the errors of other parties in a coalition complicates the rational firing rule. For this reason, Berlinski et al. (2007a) suggest that the British Westminster system might be more accountable than other systems as evidenced in the shorter

16 The selection of ministers in Europe span that ministers seem to be in post relative to other countries (Herman 1975). Whether the greater ability to fire ministers in these systems is in practice used only to sack ‘bad’ ministers or to reflect presidential tendencies of strategic chief executives (suggested to be on the rise by Poguntke and Webb 2005) driven not only by general but also personal interests in mind is still an open question. In order to investigate how strategic features actually operate in different nations, and to see what features of ministerial selection and de-selection occur across all nations, we need dedicated datasets of all ministerial movements in and out of government, in addition to all relevant background variables. Only by analysing issues pertaining to the selection and de-selection of ministerial personnel using dedicated cross-national datasets and identifying the constraints chief executives face when hiring and firing ministers will we start understanding how these constraints really operate and understand the strategic nature of cabinet formation. Furthermore only with such data can we grapple with more normative questions such as the desirability of stability or turnover of ministerial personnel. Is there a trade-off between efficiency and accountability; or does stability merely come through the constraints and problematic nature of complex bargaining environment of cabinet formation? It is for this reason that the SEDEPE network has been created to encourage the collection and sharing of data on the turnover of ministerial and other elite personnel around the democratic world. However, in order to accomplish fully comparative analysis the institutional details of different nations need to be understood. In the following chapters the institutional details are given to provide some background for researchers and preliminary analysis of the nature of ministerial hiring and firing is described for a set of twenty European nations. The smallest of these is also the only one that is not a member of the European Union (Iceland), but four are founding members of it (Belgium, France, Germany, Italy) and ten come from the 2004 enlargement, being more specifically all the Central and Eastern Europe new entrants. The UK and Spain add to the large member states already cited, whereas Ireland, Portugal and Sweden add to the medium to small ones. Altogether our sample includes one of several cases representing Southern, Scandinavian, Benelux, non-continental and, as mentioned earlier, Central and Eastern European countries in our sample.

Notes 1 In common with most formal principal–agent models we refer to the principal as a ‘she’ and the agents as ‘he’. 2 Some aspects of these relationships are discussed by Andeweg (2000). 3 In a sample of more than four hundred western European governments, Damgaard (2008) finds that almost a quarter of non-technical cabinet resignations (i.e. those triggered by mandatory elections) are due to intra-party conflicts, and that this proportion is the same amongst single-party and coalition governments. 4 Eoin O’Malley however reports in his chapter that in practice only three senators have been appointed to cabinet from 1973 to 2007. 5 We also find a variety of partial constraints enshrined in constitutions, such as in the Czech Republic where MPs may remain in the legislature and occupy a government post, but then cannot be chairs or vice-chairs of both houses of the parliament, any parliamentary committee or commissions of inquiry. In Ireland, even though all

Structural and strategic factors 17

6

7

8 9 10

members of the cabinet must come from either of the two houses of parliament (with no more than two ministers drawn from the Senate), the Prime Minister, the Deputy Prime Minister and Minister of Finance must be drawn from the Chamber (Dáil). Confidence ‘investiture’ votes vary in terms of majority and quorum requirements (for instance there is a ‘negative confidence rule’ in Sweden, where a cabinet can be inaugurated by a vote in parliament as long as an absolute majority does not vote against the Prime Minister proposed by the speaker of parliament) and whether it is the PM (in Sweden, Ireland, Germany) or the full cabinet (in Belgium and Italy) that is voted on. Co-operative game theory would predict that parties get pay-offs proportional to their bargaining power rather than their contribution in parliamentary seats and noncooperative models suggest that formateur parties should receive a disproportional amount of portfolios owing to their advantageous position in the bargaining process. Note that in several countries, however, the choice of junior minister is more in the hands of the senior minister. There has not been a great deal of consideration of these strategic issues, apart from a short book about British ministers published forty years ago: Alderman and Cross (1967). Three British ministers who chose to resign over policy later came back to lead their parties as Prime Minister – Churchill, Eden and Wilson – and these resignations certainly did not harm their progress. Resigning in order to set up a leadership challenge is well known in many other countries including France and Australia.

References Alderman, R. K. (1976) ‘The Appointment of Ministers: An Exercise in Political Bargaining’. Parliamentary Affairs 29: 101–134. Alderman, R. K. and Cross, J. A. (1967) The Tactics of Resignation: A Study in British Cabinet Government London: Routledge and Kegan Paul. Almeida, P. T. de, Pinto, A. C. and Bermeo, N. (eds) (2003) Who Governs Southern Europe? Regime Change and Ministerial Recruitment, 1850–2000. London: Frank Cass. Amorim Neto, O. and Samuels, D. (2003). ‘Cabinet Partisanship and Regime Type in Contemporary Democracies’. Paper presented at the annual meeting of the American Political Science Association, Philadelphia, PA. Andeweg, R. B. (2000) ‘Ministers as Double Agents? The Delegation Process Between Cabinet and Ministers’. European Journal of Political Research 37: 377–395. Bahr, E. (1996) Zu meiner Zeit Munich: Blessing. Berlinski, S., Dewan, T. and Dowding, K. (2007a) ‘The Length of Ministerial Tenure in the United Kingdom, 1945–1997’. British Journal of Political Science 37: 245–262. Berlinski, S., Dewan, T. and Dowding, K. (2007b) ‘Performance Evaluation of British Ministers, 1945–1997’. Manuscript. Blondel, J. (1985) Government: Ministers in the Contemporary World London: Sage. Blondel, J. (1987) Political Leadership: Towards a General Analysis. London: Sage. Blondel, J. and Müller-Rommel, F. (eds) (1993) Governing Together: The Extent and Limits of Joint Decision-Making in Western European Cabinets. Basingstoke: Macmillan. Blondel, J. and Thiébault, J-L. (eds) (1991) The Profession of Government Minister in Western Europe. London: Macmillan. Browne, E. and Franklin, M. (1973) ‘Aspects of Coalition Payoffs in European Parliamentary Democracies’. American Political Science Review 67: 453–469. Budge, I. (1985) ‘Party Factions and Government Reshuffles’. European Journal of Political Research 13, 3: 327–333.

18 The selection of ministers in Europe Budge, I. and Keman, H. (1990) Parties and Democracy: Coalition Formation and Government Functioning in Twenty States. Oxford: Oxford University Press. Chabal, P. M. (2003) ‘Do Ministers Matter? The Individual Style of Ministers in Programmed Policy Change’. International Review of Administrative Sciences 69: 29–49. Cotta, M. (1996) ‘Structuring the New Party Systems after the Dictatorship: Coalitions, Alliances, Fusions and Splits During the Transition and Post-Transition Phase’. In Pridham, G. and Lewis, G. (eds) Stabilising Fraglie Democracies: Comparing New Party Systems in Southern and Eastern Europe. London: Routledge 69–99. Cox, G. W. (1997) Making Votes Count. Cambridge: Cambridge University Press. Damgaard, E. (2008) ‘Cabinet Termination in Western Europe’. In Strøm, K., Müller, W. C. & Bergman, T. (eds) (2008) Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press. Dewan, T. and Dowding, K. (2005) ‘The Corrective Effect of Ministerial Resignations on Government Popularity’. American Journal of Political Science 49: 46–56. De Winter, L. (2002) ‘Parties and Government Formation, Portfolio Allocation and Policy Formation’. In Luther, K. R. and Müller-Rommel, F. (eds) Political Parties in the New Europe. Oxford: Oxford University Press. Dowding, K. and Kang, W.-T. (1998) ‘Ministerial Resignations 1945–97’. Public Administration 76: 411–429. Druckman, J. N. and Warwick, P. V. (2005) ‘The Missing Piece: Measuring Portfolio Salience in Western European Parliamentary Democracies’. European Journal of Political Research 44: 17–42. Druckman, J. N. and Roberts, A. (2008) ‘Measuring Portfolio Salience in Eastern European Parliamentary Democracies’. European Journal of Political Research 47: 101–134. Dumont, P., De Winter, L. and Dandoy, R. (2001) ‘Démissions gouvernementales et performances électorales des majorités sortantes (1946–1999)’. Courrier hebdomadaire du CRISP 1722. Dumont, P. Varone, F. (2006) ‘Delegation and Accountability in Parliamentary Democracies: Smallness, Proximity and Shortcuts?’ In Braun, S. and Gilardi, F. (eds) Delegation in Contemporary Democracies. London: Routledge, 52–75. Elcock, H. (2001) Political Leadership. Cheltenham: Edward Elgar. Elgie, R. (1995) Political Leadership in Liberal Democracies. Basingstoke: Macmillan. Fischer, J., Kaiser, A. and Rohlfing, I. (2006) ‘The Push and Pull of Ministerial Resignations in Germany, 1969–2005’. West European Politics 29: 709–735. Foley, M. (1993) The Rise of the British Presidency Manchester: Manchester University Press. Foley, M. (2000) The British Presidency: Tony Blair and the Politics of British Leadership Manchester: Manchester University Press. Gamson, W. (1961) ‘A Theory of Coalition Formation’. American Sociological Review 26: 373–382. Giannetti, D. and Laver, M. (2001) ‘Party System Dynamic and the Making and Breaking of Italian Governments’. Electoral Studies 20: 529–553. Gibb, C. A. (1968) ‘Leadership: Psychological Aspects’. In Sills, D. (ed.) International Encyclopedia of the Social Sciences. New York: Free Press. Headey, B. (1974) British Cabinet Ministers London: Allen & Unwin. Herman, V. (1975) ‘Comparative Perspectives on Ministerial Stability in Britain’. In Herman, V. and Alt, J. E. (eds) Cabinet Studies: A Reader. London: Macmillan, 55–76.

Structural and strategic factors 19 Huber, J. D. and Lupia, A. (2001) ‘Cabinet Instability and Delegation in Parliamentary Democracies’. American Journal of Political Science 45: 18–33. Huber, J. D. and Martinez-Gallardo, C. (2004) ‘Cabinet Instability and the Accumulation of Experience: The French Fourth and Fifth Republics in Comparative Perspective’. British Journal of Political Science 34: 27–48. Huber, J. D. and Martinez-Gallardo, C. (2008) ‘Cabinet Turnover in Parliamentary Democracies’. American Political Science Review 102: 169–180. Indridason, I. H. and Kam, C. (2008) ‘Cabinet Shuffles and Ministerial Drift’. British Journal of Political Science 38: 621–656. Kam, C. (2000) ‘Not Just Parliamentary “Cowboys and Indians”: Ministerial Responsibility and Bureaucratic Drift’. Governance 13: 365–392. Kaufman, G. (1997) How to Be a Minister London: Faber and Faber. Kavanagh, D. (1990) Personalities and Politics London: Macmillan. Laver, M. and Schofield, N. (1990) Multiparty Government: The Politics of Coalition Formation in Europe. Oxford: Oxford University Press. Laver, M. and Sheplse, K. A. (1996) Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies Cambridge: Cambridge University Press. Laver, M. and Shepsle, K. A. (2000) ‘Ministrables and Government Formation: Munchkins, Players and Big Beasts of the Jungle’. Journal of Theoretical Politics 12: 113–124. McKenzie, R. T. (1963) British Political Parties, 2nd edition. London: Heinemann. Mair, P. (1990) ‘The Electoral Payoffs of Fission and Fusion’. British Journal of Political Science 20: 131–142. Mair, P. (1997) Party System Change: Approaches and Interpretations. Oxford: Oxford University Press. Manow, P. (2005) ‘“Will Junior Ever Grow Up?” or, Where and When Do Junior Ministers Become Senior Ministers? An Empirical Investigation for 23 OECD Countries, 1949–2003’. ECPR Joint Sessions of Workshops. Granada. Manow, P. and Zorn, H. (2004) ‘Office versus Policy Motives in Portfolio Allocation: The Case of Junior Ministers’. MPIfG Discussion Paper 04/9. Mershon, C. (2002) The Costs of Coalition. Stanford: Stanford University Press. Müller, W. and Strøm, K. (eds) (2003) Coalition Governments in Western Europe, Oxford: Oxford University Press O’Malley, E. (2006) ‘Ministerial Selection in Ireland: Limited Choice in a Political Village’. Irish Political Studies 21: 319–336. Poguntke, T. and Webb, P. (eds) (2005) The Presidentialization of Politics. Oxford: Oxford University Press. Riker, W. H. (1962) The Theory of Political Coalitions. New Haven: Yale University Press. Schleiter, P. and Morgan-Jones, E. (2007) ‘Party Control in European Cabinets?’ Paper presented at the annual meeting of the American Political Science Association, Chicago. Shugart, M. S. (2006) ‘Comparative Executive-Legislative Relations’. In Rhodes, R. A. W., Binder, S. A. and Rockman B. A. (eds) The Oxford Handbook of Political Institutions. Oxford: Oxford University Press, 344–365. Strøm, K., Müller, W. C. and Bergman, T. (eds) (2008) Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press.

20 The selection of ministers in Europe Schofield, N. (1987) ‘Stability of Coalition Governments in Western Europe’. European Journal of Political Economy 3: 555–591. Warwick, P. V. (2005) ‘When Far Apart Becomes Too Far Apart: Evidence for a Threshold Effect in Coalition Formation’. British Journal of Political Science 35: 383–402.

2

Hiring and firing ministers under informal constraints Germany Jörn Fischer and André Kaiser

Constitutional framework One short sentence in the German Basic Law is dedicated to the appointment and dismissal of ministers: ‘Federal Ministers shall be appointed and dismissed by the Federal President upon the proposal of the Federal Chancellor’ (Article 64 I of the Constitution).1 This norm leaves ample room for inter- and intra-party bargaining when it comes to the process of cabinet formation and the selection of ministers. It also grants the Federal Chancellor formally unlimited decision-making autonomy concerning the sacking of ministers.2 Article 66 of the Basic Law states that neither the Federal Chancellor nor ministers may hold any other salaried office or engage in any trade or profession. According to Article 4 of the Federal Minister Law,3 a member of the federal government must not be member of a state government at the same time. Furthermore, it is shared understanding, although nowhere clearly stated, that ministers are subject to the same conditions of eligibility as the Federal Chancellor (Schröder 2005: 1588). They have to be German citizens and at least 18 years of age. Except for these requirements the Chancellor is formally free in her choice of cabinet members. However, informal rules of cabinet selection and de-selection have developed since the founding of the Federal Republic in 1949 – we will discuss these in detail below. The Constitution also leaves room for interpretation on the role of the Head of State in the selection of the cabinet. Most constitutional lawyers nowadays agree that the Federal President is obliged to follow the Federal Chancellor’s proposal concerning her ministerial personnel – unless there are legal grounds which would justify the rejection of a nominee. Yet, if the Federal President comes to the conclusion that the appointment of a minister violates Federal Law or the Basic Law, the only legal norm on which he could base such a refusal is the very general Article 20 III of the Constitution (Herzog 1983: 7), which states that the executive shall be bound by law and justice. So far, no President has ever vetoed the Chancellor’s proposal for such legal reasons. Although a presidential right to veto a minister for political reasons cannot be derived from the constitutional text, the President can express his political concern or discontent with a proposal, and is free to ask the Chancellor to revise a decision regarding her ministerial personnel. This has not occurred in recent years, but back in the 1950s and 1960s,

22 The selection of ministers in Europe when the interpretation of the Basic Law was not so clear-cut, several presidential attempts to veto a minister were carried out: the Federal President Theodor Heuss was successful in preventing a second tenure for Thomas Dehler (Freie Demokratische Partei, FDP) as Minister of Justice after the 1953 general election. During his first term, the latter had publicly criticized the Federal Constitutional Court, thereby affecting his relationship with the president of the Court but also with the Head of State, leading Heuss to threaten not to sign the letter of appointment in informal talks with the Chancellor Konrad Adenauer (Christlich Demokratische Union Deutschlands, CDU). Although Federal Chancellor Adenauer did not acknowledge a legal foundation for this presidential intervention, he complied.4 This successful veto however did not serve as a precedent. His successor, Heinrich Lübke, known for his attempts to be a political president (an unusual interpretation of the role of the Head of State in Germany), tried in 1959 to influence the choice for the post he himself had left vacant when elected as Federal President, the Ministry of Agriculture, but failed to veto the appointment of a candidate minister who did not share his policy preferences. He also intervened in the formation of the cabinet in 1961 und 1965, when again out of differences in policy positions he publicly opposed the candidacy of Gerhard Schröder (CDU) for the Foreign Office, thereby provoking a crisis during the cabinet formation process. Nevertheless, in the end the Head of State had to accept the Chancellor’s choice of personnel. Against the background of the Constitution of the Weimar Republic (1919–33), which had granted substantially more powers to the Head of State than the 1949 Constitution does, one can view these episodes as attempts to test the limits of presidential powers under the Basic Law. Winkler concludes in a 1967 monograph that a successful presidential veto is possible only if there are grave reasons at hand, but also that the abdication of the minister is not tantamount to the termination of the governing coalition (Winkler 1967: 36). Nowadays, it seems highly unlikely that a Federal President would intervene and actually try to veto a ministerial appointment using his political weight. Cabinet formation is formally a domain of the Federal Chancellor, which in fact is constrained only by the coalition partner and by factions in her own party, as we will see below. With regard to the formation of government as a whole, the Federal President has ‘next to no discretionary powers’ (von Beyme 1983: 16). As the Basic Law does not provide any rules for the process of government formation, coalitions are the result of freestyle bargaining between parties. The President’s tasks are basically of a ceremonial kind. Only if the candidate for chancellorship does not achieve an absolute majority in the first round of the investiture vote and in as many rounds as the Bundestag wishes to hold within 14 days of the first unsuccessful attempt, does the President have to choose whether to accept a Chancellor elected into office on the basis of a plurality of votes or whether to dissolve the Bundestag and to have another general election. So far this has never been the case, even though it seemed possible for a few days after the extremely close 2005 Bundestag election, when no clear majority government could be formed. Except for these majority requirements and the deadline for the election of the Chancellor there are no constitutional rules concerning coalition formation. Although the Bundestag is

Hiring and firing ministers: Germany 23 not formally involved in the appointment of ministers, to avoid negative surprises in the Chancellor’s investiture vote the list of ministers composing the cabinet has at times not been published until after the vote. The duration of this process is comparatively short. Measured from the general election to the swearing-in of the cabinet, it took between 24 (1969 and 1983) and 65 (2005) days. The Basic Law, unlike the Constitution of the Weimar Republic and a few constitutions of the states, the Länder, does not require an individual or a collective investiture vote on ministers or the cabinet. Only the Federal Chancellor is formally elected by the legislature. Similarly, there is no such thing as a vote of no confidence against a minister or the cabinet as a whole. Yet, if the Chancellor loses a constructive vote of no confidence or forfeits the chancellorship in any other way, the whole cabinet is obliged to resign. The fact that parliament cannot vote a minister out of cabinet is related to the doctrine of ‘political responsibility’. The ‘departmental principle’ (Ressortprinzip) in Article 65 states that within the government’s general guidelines ‘each Federal Minister shall conduct the affairs of his department independently and on his own responsibility’. It serves as the constitutional link to the doctrine of individual ministerial responsibility. However, the doctrine specifies neither to whom the minister is actually responsible nor which consequences can be drawn from this principle. Unlike in the Weimar Constitution (which specified ‘on his own responsibility to the Reichstag’ (Article 54)), the responsibility’s addressee is left open in the Basic Law. Some scholars focus on a responsibility to the Chancellor (Conradt 2005: 203), basing their point of view on the fact that she is the only one equipped with the possibility of the final sanction: dismissal (Saalfeld 2003a: 36). The prevailing opinion, however, is that of a responsibility to the Bundestag, justified with the parliamentary control which the minister’s administration is subject to (Hesse and Ellwein 2004: 274; Mehde 2001; Schröder 2005: 1615; Wengst 1984: 539–40). However, unlike the Weimar Reichstag (Article 54 of the Weimar Constitution) and a few Länder parliaments today, the Bundestag does not have the power to dismiss an individual minister. The Bundestag can use several instruments, such as questions for the minister, committees of inquiry and the summoning of a minister to parliament to claim the minister’s accountability and hence scrutinize his doings. Reprobation votes and even motions which prompt the Chancellor to dismiss a minister are also possible, yet they have no legal impact whatsoever. In a nutshell, the doctrine of individual ministerial responsibility in Germany is rather vague and related to political pressure instead of constitutional powers. Moving on to the collective responsibility of the cabinet as a whole, things become more difficult. Undoubtedly, one constitutive feature of parliamentary systems is that the government is responsible to parliament, but the Basic Law lacks concrete rules in this respect. The ‘cabinet principle’ (Kabinettsprinzip), as stipulated across several articles in the Constitution is, broadly interpreted, an indication of collective responsibility for governmental decisions. Yet, as Schröder rightly argues, cabinet responsibility may also develop in practice beyond the written Constitution (2005: 1616). As the political practice in Germany is the formation of coalition governments, public disagreements between cabinet members are more likely, thereby undermining cabinet solidarity (Conradt 2005: 207).

24 The selection of ministers in Europe The ‘chancellor principle’ (Kanzlerprinzip) is considered to be the strongest of the three principles determining the organizational structure of the German federal executive. According to Article 65 of the Constitution, ‘the Federal Chancellor shall determine and be responsible for the general guidelines of policy’. Her dominant role in cabinet decision-making is deduced from this principle. In case of conflict among cabinet members she must try to arrive at a conclusion and ultimately formulate what she perceives to be the majority view, which will become the cabinet decision. Generally, before a minister takes a policy proposal officially to the cabinet, he usually tries first to reach an agreement beforehand with the ministries involved.5 Although the cabinet is an important arena of conflict management in a coalition, it is in practice not the most important one (Saalfeld 2003a: 60), as informal structures have developed and conflicts are often dealt with in ‘coalition talks’ between government members and leaders of the governing parties’ parliamentary groups. Also, representatives of the Länder governments are often included in these meetings, as the second chamber, the Bundesrat, must approve a good portion of federal law.6 Saalfeld considers these more informal coalition talks to be ‘the most important steering bodies’ (Saalfeld 2003b: 366). Cabinet committees also play a role in the interdepartmental co-ordination of policy, although they are often dominated by civil servants and thus further away from the political sphere (Saalfeld 2003a: 60). Because of informal decision-making mechanisms, conflicts in cabinet meetings are rare, and, even though the formal decision rule is majority voting (GeschOBReg,7 Article 24 (2)), quick unanimity decisions are common. Cabinet’s role is indeed ‘essentially limited to that of a final political check on the general line of governmental policy’, and in cabinet meetings ‘decisions are approved rather than made’ (Müller-Rommel, 1994: 167). Based on the British role model, the so-called ‘Parlamentarische Staatssekretäre’8 were created in 1967 as junior ministers whose mission was intended to strengthen the political control of departments and liaison between parliament and the top level of Federal Ministries (Saalfeld 2003b: 367–8). To achieve this closer connection between legislature and executive, junior ministers, with the exception of those in the Federal Chancellery, are required to hold a seat in the Bundestag.9 Just like cabinet ministers, junior ministers are appointed by the Federal President on the proposal of the Chancellor, but, according to the Law on Parliamentary State Secretaries, this is done in agreement with the senior minister in whose portfolio a junior minister is going to serve (ParlStG, Art 2).10 In constitutional terms, they are not members of the government (Article 62 of the Basic Law, Schröder 2005: 1549), but politically they are considered to be part of it (Ismayr 2000: 209–10; Patzelt 2005: 245) because of their function of standing in for the cabinet minister in the legislature. They also attend cabinet meetings when deputizing for the relevant minister in case of absence. Yet junior ministers do not carry a voting right in cabinet. The creation of this new position had a temporary effect on the size of German cabinets: starting with a small pool of 14 ministers in the postwar years their number increased to 22 in the 1960s but decreased with the introduction of junior ministers. The number of ministers, however, climbed again to 20 in the Kohl era, but dropped to a record low of 13 in Gerhard Schröder’s second term.

Hiring and firing ministers: Germany 25

Selecting ministers The choice of ministers formally rests exclusively with the Chancellor, but in fact her powers of appointment are constrained by her own party and, concerning the ministers of the coalition partner, even more so by the other governing party. Within the Chancellor’s own party, i.e. the CDU/CSU (Christlich Demokratische Union Deutschlands/Christlich-Soziale Union in Bayern) or the SPD (Sozialdemokratische Partei Deutschlands), a small group consisting of parliamentary and party leaders, including the most influential leaders of Länder party associations, decide on the cabinet composition. Of course, this selectorate must take into account the different interests and demands within the parliamentary parties which ultimately have to sustain the government and the Chancellor and keep her in office. These interests have sometimes been included rather peacefully in an anticipating manner, while at other times there has been massive pressure by different factions in the parliamentary parties (Schüttemeyer 1998: 242). Except for grand coalitions between Germany’s large parties, the SPD and the CDU/CSU, between 1969 and 1972 and from 2005, governments have consisted of a large party (the Chancellor’s party) and a small coalition partner (the FDP or the Green Party). As an unwritten rule, the coalition partners are quite free in the choice of ‘their’ ministers. The influence of the parliamentary party in the FDP’s internal selection of its cabinet candidates is rather direct and formal. Decisions are made in joint meetings of the parliamentary group with the party executive, and narrow votes between candidates are common (Schüttemeyer 1998: 246). When the Green Party participated in the federal government for the first time in 1998, it was assigned three portfolios. The selectors, notably the parliamentary and party leaders, had the difficult task of finding a balanced representation of the party’s two main wings11 in government. This was essential not least because the party convention had to approve the cabinet composition. The introduction of junior ministers is a recent development in Germany. Even though these are seen as apprentices from among whom the more successful ones would later be promoted to a senior post, this expectation has not been fulfilled. Whereas a third of those appointed before 1977 later became senior ministers (Schindler 1999: 1114), the overall share covering the years from 1967 to 2003 dropped to only 14 per cent (Manow 2005: 18), showing that from 1977 onwards this rate was very low. One reason is the increase in the number of junior ministers whereas the number of senior ministers remained more or less stable. In addition, most of the promoted junior ministers were appointed to a department in which they had not served before (Schindler 1999: 1114). The number of junior ministers has increased greatly since 1967, outnumbering the number of portfolios for the first time in 1982. From this point onwards every department possessed between one and three junior ministers,12 ‘depending on the size and importance of the department’ (Saalfeld 2003b: 368). Junior ministers are often regarded as an instrument of mutual policy control by placing them in a portfolio led by the coalition partner. Whether they should ‘keep tabs on partners’ (Thies 2001) had been an issue during their introduction (Eschenburg 1966: 19) but it was only the case during the chancellorship of Helmut

26 The selection of ministers in Europe Schmidt (1974–82) where departments were actually led by junior ministers alongside a cabinet minister (see the data in Schindler 1999: 1062–107). Under Helmut Kohl (1982–98), the number of junior ministers peaked at 33, partially functioning as watchdogs not only regarding the coalition partner (FDP) but seemingly also as a means of two-way control between the respective Christian Democratic sister parties, the CDU and the Bavarian CSU, which together constitute the parliamentary party of the Union on the federal level (Schindler 1999: 1062–107). In some cases in the 1970s, junior ministers were intentionally placed in a department led by a senior minister belonging to a different party wing for intra-party balance (Kröger 1974: 589). The three most recent coalition agreements in 1998, 2002, and 2005, however, more or less left the nomination power over junior ministers to senior ministers. A different approach interprets the introduction of junior ministers as a resource for solving or easing distributive conflicts between coalition partners over office or policy resources (Laver and Shepsle 1996; Mershon 2002). This highlights the relevance of portfolio allocation, which is the result of intense negotiations between parties during the process of government formation. Usually the distribution of portfolios takes place after the main negotiations regarding policy issues have come to an end. During the coalition governments under Konrad Adenauer, from 1949 to the early 1960s where more than two parties sometimes formed the government in a not yet consolidated party system, the CDU/CSU granted between five and eight cabinet posts to their coalition partners. Even when the CDU/CSU had obtained an absolute majority of votes and seats after the Bundestag elections in 1957, Adenauer preferred to govern in a coalition with the Deutsche Partei, which supplied three ministers who, after a while, switched to the CDU anyway. Since the election of Ludwig Erhard as Federal Chancellor in 1963, the smaller coalition partner always obtained three or four cabinet posts – with the exception of Erhard’s first and Helmut Kohl’s fifth cabinet, when the CDU generously granted five ministries to the FDP. In the two cases of grand coalitions, both under CDU chancellorships, the SPD received nine out of 19 (1969–72) and eight out of 15 (!) portfolios (from 2005 onwards). By and large Germany confirms the pattern proposed by Gamson (1961) and empirically observed by Browne and Franklin (1973) for European parliamentary democracies that parties in a coalition tend to receive portfolio pay-offs in almost perfect proportions to their seat share. Portfolio allocation German-style is also consistent with the ‘small-party bias’ found elesewhere (Browne and Franklin 1973; Schofield and Laver 1985; Warwick and Druckman 2001), as the FDP and other smaller parties have frequently been slightly over-represented in cabinet (Saalfeld 2003a: 65; for respective data see Schindler 1999: 1143–1145). Testing to what extent the inclusion of junior ministers has any corrective effect regarding the deviation from the perfect proportionality between the seat and the cabinet shares of coalition partners, Manow and Zorn indeed found some of this effect, but to a much smaller degree than in other European countries like France, Italy and Portugal (Manow and Zorn 2004: 16). The distributional perspective is concerned not only with how many ministerial posts each coalition partner gets but also with which portfolios they hold (Budge

Hiring and firing ministers: Germany 27 and Keman 1990: 89–133; Laver and Shepsle 1996). The main claim of the portfolio allocation approach is that it matters for parties which portfolio they hold, because cabinet ministers pursue the implementation of their party’s policy. Consequently, parties seek to spot their ministers in departments for which they dispose of a high party-specific salience value resulting from the party’s ownership of a specific issue. Comparing the salient issues of Germany’s parties, calculated by Franzmann (2006), with the allocation of portfolios between parties in government reveals very clear patterns. Whenever the CDU/CSU is in government, it holds the Ministry of Defence and the Home Office, with defence and law and order belonging to its most important issues. Except in the years between 1966 and 1969 the SPD, when governing, has always been in charge of the Department of Social Affairs, focusing on social justice as a core issue. The Ministry of Economic Affairs has in more than two-thirds of its time in government been under control of the FDP, with free enterprise and economic orthodoxy being among their core issues. And, needless to say, the Greens obtained the Department of Environmental Affairs during their seven years of government participation at the federal level. Germany has often been characterized as a ‘party state’ (Grewe 1951) or “party democracy” (Gabriel, Niedermayer and Stöss 1997). Indeed, parties ‘provide the most important screening mechanisms for parliamentary and ministerial candidates’ (Saalfeld 2003b: 349). Since 1949, only three ministers have never joined a party (Ludger Westrick, Hans Leussink, Werner Müller), and two more became party members shortly after their appointment (Siegfried Balke, Klaus Kinkel). For a long time the public believed that Minister of Economic Affairs Ludwig Erhard did not officially affiliate to the CDU until shortly before his 30

25

20 Outsider Land executive

15

Parliamentary party Standing committee

10

5

Merkel I

Schröder II

Kohl V

Schröder I

Kohl III

Kohl IV

Kohl I

Kohl II

Schmidt III

Schmidt I

Schmidt II

Brandt I

Brandt II

Kiesinger I

Erhard I

Erhard II

Adenauer IV

Adenauer II

Adenauer III

Adenauer I

0

Figure 2.1 The selection of cabinet ministers in Germany from different recruitment pools, 1949–2007 (first-time appointments) Note: The categories ‘Land Executive’ and ‘Parliamentary Party’ are not mutually exclusive. Therefore, some first-time appointments may be counted twice.

28 The selection of ministers in Europe chancellorship. Only recently was it revealed that, in fact, Erhard never signed his party membership card. Some ministers changed their party affiliation while being in office or in between two portfolios, a temporary phenomenon due to party dissolutions in the early postwar years. Before a politician in Germany reaches a cabinet post, he often has to work his way up the hard way through the party organization, a career pattern termed Ochsentour. Reliable data on ministers with a career stage on the local level is hard to obtain, but 55 per cent of all chancellors, senior, and junior ministers had a party function (party executive) on the state level before becoming a cabinet member (Syed Ali 2003: 143). Following from the CDU’s Länder party associations having a comparatively strong intra-party position (Schmid 1990), 65 per cent of the CDU cabinet members have filled a party post in a Landesverband, whereas only 42 per cent of the SPD ministers had a significant past as a party official in a Land association (Syed Ali 2003: 143). Looking at the national party level, only 38 per cent of all Chancellors and ministers held a party leadership position before joining the cabinet (Syed Ali 2003: 144). Yet we have good reason to assume that this share is significantly higher if junior ministers are excluded from the calculation. In a nutshell: a party career via a high position in the party hierarchy is one possible career path but not a ‘must’. A more common springboard to the cabinet is the Bundestag. Parliaments in Western Europe serve for would-be ministers as ‘major socialization agent’ (De Winter 1991: 44), providing the knowledge and skills which are potentially useful for future cabinet tasks. The recruitment of ministers from parliament is said to guarantee the responsive and representative characteristics of cabinet government by building a personal link between executive and legislature. In Germany, this link is strengthened by the fact that most of the cabinet members whilst in office also belong to the Bundestag, even though the Basic Law does not require ministers to be parliamentarians. Resigning ministers who have been MPs during their ministership usually keep their seat in parliament after losing the cabinet office. Out of 174 first time appointments (senior ministers only) in the period from 1953, i.e. since the second post-war cabinet,13 to 2007, 134 ministers (77 per cent) were drawn from the Bundestag (all calculations in this section are based on Schüttemeyer 1998: 233–6, or Kempf 2001: 19–23, and own data). For all parties the parliamentary party constitutes the most important pool from which to draw cabinet members (FDP: 92 per cent of a total N of 26; CDU/CSU: 85 per cent, N = 73; SPD: 73 per cent, N = 62; Greens: 50 per cent, N = 4). Parliamentary seniority is clearly a criterion: 70 per cent of those who were MPs before their ministerial office had served in parliament for eight years or more. In a comparative perspective, De Winter calculates a rather short ‘waiting time’ of 7.7 years on average for German ministers (De Winter 1991: 48, covering the years 1949 to 1984). A strong hierarchy exists within the parliamentary parties. Belonging to the leadership circle of the parliamentary groups enhances the chances of being selected as a minister: 66 per cent of those who were picked from the legislature had been member of the parliamentary group’s standing committee;14 this share drops to 41 per cent if we exclude the extended standing committees.

Hiring and firing ministers: Germany 29 Are local and regional politics an avenue to Germany’s cabinet? According to Thiébault, Germany is an outlier in the group of Western European countries, with no fewer than 81.5 per cent of all cabinet members having held a local or regional elective position before starting off their ministerial career (1991: 33).15 From 1949 to 2007, 40 out of 190 ministers (21 per cent) had been members of a Land government, either as minister-president (or mayor, in the cases of the city states Berlin, Hamburg and Bremen) or as a minister (or senator). Given the joint character of German federalism, this is no surprise: the minister-presidents of the Länder are constantly becoming involved in national politics. Many of them acquire a national profile and actually exercise a strong influence in federal politics while being the chief of a Land executive. A very recent phenomenon remains to be mentioned. Land executives seem to be gaining importance as a recruitment pool, whereas the parliamentary party is losing its relevance. This trend is observable for all parties participating in government since 1998 (SPD, Greens, CDU/CSU). Nearly half of the newly appointed federal ministers were once members of a Land executive (16 out of 35). Some of them were drawn virtually directly from there (Karl-Heinz Funke) and others with a time-lag of up to several years (for example Joschka Fischer, who left the Ministry for Environmental Affairs in Hesse in 1994 and became Foreign Minister in 1998). The explanation seems to be that in recent years Germany experienced one complete and one partial government change, which made the selection of new teams necessary. Gerhard Schröder, assuming the Chancellor’s office on the basis of a successful past as Minister-President of Lower Saxony, relied to some degree on former ministers of his regional cabinet (Jürgen Trittin, Karl-Heinz Funke). A wave of ministerial resignations in his first term forced him to continuously reshuffle his cabinet. At the same time, some SPD minister-presidents at the Länder level were voted out of office, so that Schröder had to find posts for these top politicians (Hans Eichel, Reinhard Klimmt, later Sigmar Gabriel). Angela Merkel, although being firmly rooted in her parliamentary party which she had headed from 2002 to 2005, also opted for some novices in national politics. Annette Schavan and Ursula von der Leyen, both with histories as Land ministers, gained a national profile for the portfolios they were selected for. Regional politics is for many politicians no more than a stepping stone towards national parliament, from where the career might continue towards a federal cabinet post (Stolz 2003). The direct leap to the federal executive, however, is the exception rather than the rule. In addition, the avenues, via the legislature on the national level and via the Land executives respectively, are not mutually exclusive. Helmut Schmidt and Gerhard Schröder are prominent cases of politicians who switched from the national parliament to a regional executive and finally back to the national level as Chancellor. Which paths then lead outsiders who neither have a parliamentary past on the national level nor have obtained a post in a government at the Land level to the cabinet? Selecting side-entrants for cabinet has been a rare phenomenon during the entire history of the Federal Republic. Outsiders like Ursula Lehr, Rita Süssmuth (both CDU, both Ministers of Family Affairs), Hans Leussink (non-party

30 The selection of ministers in Europe member, Minister of Science) and Horst Ehmke (SPD, Minister of Justice) had a marked professional and scientific background related to the portfolio they were later asked to lead. The nomination of Werner Müller, a non-party member with a professional background as a manager, to Gerhard Schröder’s cabinet can be interpreted as a symbolic act of openness towards economic interests.16 These are, however, exceptional cases. There is little incentive for the Chancellor to appoint outsiders, as their nomination can cause disappointment in and always has to be justified vis-à-vis the parliamentary party as the main recruitment pool.17 Composing a cabinet is a tricky task. When scanning the pool of ‘ministrables’, the wish for representation of very different groups and interests in the parties has to be taken into account alongside skills and experiences. Party wings, regional origin, and, although decreasingly, religious affiliation are the most important criteria which the nominators of ministerial personnel have to consider. The key word is Proporz. Including representatives from the various ideological party factions in the cabinet is necessary to avoid discontent within the party and a strategy of appealing to a wide spectrum of voters. All parties dispose of formal or informal platforms for intra-party groups on the basis of ideology. Not surprisingly we find prominent ‘leftist’ politicians in CDU-led governments and ‘rightist’ ministers in SPD-led cabinets.18 To what extent the representation of different party groups is actually practised is a matter of the prime minister’s personal leadership style. The same holds for the inclusion of the Chancellor’s intra-party rivals and potential trouble-makers. Regional representation is an issue because of the power of the parties’ Länder associations, especially with regard to the more densely populated states. The presence of regional leaders (or their confidants) in the cabinet assures them direct or indirect influence in national politics. Vice versa, the integration of regional leaders allows the Chancellor to exercise a certain degree of control over their actions.19 Since the unification, including at least one politician from East Germany in the government is a must. Between 1991 and 2007, eight senior ministers came from East Germany and 60 from West Germany, with one minister (Thomas de Maizière) having a mixed background. East German ministers are often in charge of the Ministry of Transport, which also carries the task of the new Länder’s infrastructural modernization. Until 2007 the three most populated Bundesländer provided the largest numbers of senior ministers: North Rhine Westphalia: 45, Bavaria: 35, and Baden-Württemberg: 24 (Kempf 2001: 11, and own data). Religious denomination was a very important factor in the arithmetic of cabinet composition in the first two decades. Catholic and Protestant ministers were almost equally represented in the first five cabinets (Lange 1973: 138). This shows how important religious affiliation was in CDU/CSU-led cabinets and how much stress was put on representing their Protestant minority wing. In those days religious denomination could turn out to be the decisive factor in starting (or preventing) a ministerial career. The other parties paid (and are paying) less attention to religious aspects in cabinet composition and alongside with societal changes nowadays it has become a minor factor, even though the CDU/CSU especially still takes it into account.

Hiring and firing ministers: Germany 31 ‘With a little more than average intelligence one can do that’ – this is allegedly Chancellor Helmut Schmidt’s reaction to Egon Bahr’s argument that ministers leading the portfolios of finance, justice or agriculture would require a certain professional background (Bahr 1996: 463). This quotation sheds light on the minor importance of expertise. Nowadays, the prevailing concept of a minister is that of a generalist, but this has not always been the case. We observe a slow shift towards generalism. For example, according to Blondel, German cabinets until 1981 exhibited a comparatively high share of 43 per cent specialists (1985: 277), and Armbruster (1973) in his early analysis of German governments complained about a lack of generalists in the cabinet. However, today, not even the Ministry of Agriculture, which was for more than fifty years an agricultural expert’s domain, is led by a farmer any more. Only Justice is left as a domain where a professional background in law is seemingly a necessary condition for the minister. Even though in Germany the percentage of ministers who served in two or more portfolios (28 per cent, from 1949 to 2007) is rather low, this figure is no proof for a lack of generalist profiles, but can rather be interpreted as an expression of general reservation towards rotating ministers. The first female federal minister was appointed in 1961 to the Department of Family Affairs (Elisabeth Schwarzhaupt). Up until 1998, there were 16 women at the cabinet level (Syed Ali 2003: 114). Usually they held portfolios considered to be ‘women-specific’, like Family Affairs or Health. The Ministry of Justice was the first classical department led by a woman (Sabine Leutheusser-Schnarrenberger from 1992 to 1995). Since 1998, the situation has completely changed. The red– green government of 1998 included five female cabinet members, i.e. a record share of one-third. Female representation even rose to 45 per cent in Schröder’s second cabinet, but has dropped to one-third in the grand coalition installed in 2005 and led by the first female Chancellor in German history, Angela Merkel. The appointment of young ministers is very rare, with only one senior minister being under 30 years of age (Claudia Nolte, 1994). The appointment age groups 40–9 (43.6 per cent) and 50–9 (38.6 per cent) were by far the strongest (Syed Ali: 2003: 108) between 1949 and 2002 (junior ministers included). The average age of all senior ministers at the moment of their first appointment is approximately 49 years (Kempf 2001: 14). Members of SPD-led governments are on average about three years younger than those under CDU leadership (Kempf 2001: 14). Leaders of the governing parties are often cabinet members, but there is variance between parties. The Chancellor is usually also head of his party, with Helmut Schmidt (SPD), Gerhard Schröder (SPD) and partially Ludwig Erhard (CDU) being the exceptions. Whereas the CSU chairman participated in cabinet during only 15 of its 38 governing years between 1949 and 2006, FDP leaders were almost always present (Schindler 1999: 1160–2, and own data). The two co-leaders of the Greens may not become members of the cabinet for reasons of their party’s statute.20 From a Chancellor’s perspective, governing is often easier if the coalition partner’s leader is involved in and disciplined through cabinet work. A similar logic stood behind the appointment to the Ministry of Finance of the CSU chief and Helmut Kohl’s Union intern rival, Theo Waigel, whose involvement has been

32 The selection of ministers in Europe considered to be a ‘masterstroke of Helmut Kohl’ (long-time Foreign Minister Hans-Dietrich Genscher, cited in Kempf 2001: 27).

De-selecting ministers Formally, the Chancellor is totally free in her decision not only to hire but also to fire ministers. Analogous to the interpretations of the Head of State’s right regarding an appointment (i.e. no intervention for political reasons), according to most constitutional jurists the Federal President must also follow the Chancellor’s proposal of a dismissal (Schröder 2005: 1590, Schneider and Zeh 1989: 1319) irrespective of her reasons to do so. In fact, as we will see below, a Chancellor faces numerous constraints regarding the dismissal of her cabinet members. Ministerial discontinuations can be classified into five types. First, a minister may fall victim to a general cabinet reshuffle. Secondly, the minister may lose office as a consequence of his party leaving government. Thirdly, the minister may step down because of some kind of pressure resulting from a resignation discussion in public (‘push resignation’). Fourthly, a minister can be pulled out of office if he 60

50

40

30

20

10

Figure 2.2 Ministerial discontinuations in Germany, 1949–2007

Party leaves government

Abolition of department

Reshuffle/change of portfolio – after election

Reshuffle/change of portfolio – during term

Shuffled out after election

Shuffled out during term

Resignation for miscellaneous reasons

Protest resignation

Pull resignation

Push resignation

0

Hiring and firing ministers: Germany 33 is offered a political post outside cabinet – for example a leading party position or an important job in an international organization (‘pull resignation’21). Finally, the minister might disagree with government policy or have other, politically neutral, reasons for resigning (‘protest’ or ‘neutral resignation’). As we define a resignation to be a premature, non-scheduled discontinuation of a minister’s term in cabinet (Fischer et al. 2006: 712; for a similar delineation see Sutherland 1991: 110), only types three to five are actually resignations. In contrast, we term the loss of an office due to a reshuffle or a new government ‘ordinary discontinuation’. Concerning the de-selection of ministers, Germany appears to be a boring country. Of a total of 236 office discontinuations22 from 1949 to 2006, 180 (76 per cent) are ordinary in the above sense. In 55 cases, ministers left cabinet because of a reshuffle23 after an election (with their party being still in power). In 49 cases, a ministerial term ended because the party left government; 29 ministers moved to another portfolio after an election. In nine cases, office discontinuation was due to the abolition of the department. This leaves 19 cases where ministers lost office and 19 cases where ministers moved to another portfolio as a consequence of reshuffling during a running term. Between 1949 and 2005, only six major reshuffles took place (1956, 1962, 1978, 1982, 1989 and 1993). There is therefore no tradition of systematic mid-term reshuffles. This instrument is, however, used in more specific situations, either in order to get rid of ministers weary of their office or to reallocate promising talents to more prominent portfolios. Individual resignations, making an adjustment of power and cabinet Proporz necessary, sometimes trigger major reshuffles. This might also be a welcome opportunity to silently get rid of badly performing ministers if a ‘health problem’ is not available or thought inappropriate to fire them individually. Hence, reshuffles are used from time to time as an elegant instrument for sacking unsuccessful or tainted ministers, i.e. those who survived one (or more) resignation debates and thus might face the increased hazard of an individual resignation (Fischer 2004, Dewan and Myatt 2007). Marie Schlei (1978, SPD, Minister of Economic Co-operation and Development) and Rupert Scholz (1989, CDU, Minister of Defence) are typical illustrations. Marie Schlei’s visit to some African countries in 1977 caused calls for resignation for a bundle of reasons: alleged lack of knowledge in development aid policy, participation in an inappropriate tourist programme and quarrels with her spokesperson – the media coverage of this trip was disastrous, although the harsh attacks especially from CDU politicians did not seem to be justified in each point. Being the first female minister in the Federal Republic in a portfolio other than health or family affairs in those times might have contributed to her difficult standing. In February 1978 she was shuffled out of the cabinet. Rupert Scholz was called to Helmut Kohl’s cabinet in 1988. Within less than a year he accumulated three resignation issues. The several military plane crashes during his tenure with numerous fatalities surely cannot be blamed on him, but in the follow-up activities of these accidents he acted unfortunately. He lost his office in a cabinet reshuffle just 11 months after his appointment. In a theoretical perspective, reshuffles may be carried out with the hope of an electoral reward as well as with the intention of better policy performance in key portfolios. Both factors play a role in Germany. So far, the argument that

34 The selection of ministers in Europe Prime Ministers use reshuffles to reduce agency loss resulting from the delegation of power to ministers and civil servants (Indridason and Kam 2008) has not been tested for the German case. However, the unsystematic pattern of reshuffling activities puts its validity for Germany in doubt. Reshuffling activities seem to depend to a large extent on the Chancellor’s personal leadership style, with four Chancellors not reshuffling their cabinet at all (Ludwig Erhard, Kurt-Georg Kiesinger, Willy Brandt, Gerhard Schröder).24 Coalition government, of course, is an important constraint, limiting the Chancellor’s freedom to select and de-select ministers in a reshuffle. Shuffling a coalition partner’s ministers without prior consultation with the coalition partner’s leadership is quite unthinkable. Politically more interesting are the cases of preliminary and non-scheduled discontinuations, thus the actual resignations. With 57 cases in 58 years (1949– 2007), these are a relatively rare phenomenon in Germany. Among the 57 cases there were 17 push and 18 pull resignations, 11 happened as a sign of protest or expression of discontent and 11 for miscellaneous reasons. The latter include three ministers who died whilst in office. Interestingly, ten push resignations, i.e. resignations preceded by some kind of scandal, occurred between 1990 and 2002, with Gerhard Schröder’s first term being the administration struck by the most push resignations of all cabinets (five cases). A complete alternation of government and, in effect, an inexperienced coalition consisting of the SPD (after 16 years on the opposition benches) and the Greens (for the first time in government on the federal level) may explain this unusual accumulation. In stark contrast, the second Schröder government finished its term with exactly the same team as it had started with. When does a minister have to resign? The doctrine of ministerial responsibility, as underlined in Article 65 of the German Constitution, does not help to explain in which cases cabinet ministers are actually obliged to step down. The simple reason for this is that there are no formal rules on what this rather vague doctrine implies. Therefore, it is a mixture of external pressure and internal interpretations of this doctrine by the minister concerned as well as by other actors in government and political considerations by the Federal Chancellor which ultimately decides whether a minister has to go. In an extensive statistical analysis of 111 resignation debates during the years 1969 to 2005, Fischer et al. (2006) find that the outcome of resignation discussions, that is whether or not ministers must resign, is determined mainly by two factors: the specific resignation issue and the interplay between the position taken by the Federal Chancellor and the minister’s party. Resignations follow systematic patterns. However, these patterns are not compatible with the ministerial responsibility doctrine for two reasons. First, errors in the minister’s actual area of responsibility account for only seven out of the 14 resignations in the period covered here. Secondly, and more interestingly, personal errors have only once led to an actual resignation, while departmental errors have done so in six cases. Hence, resignation events do not follow a managerial logic with the minister concerned losing office automatically after an error has been detected and publicized. We find that ‘extrapolitical’ resignation issues, having mainly to do with personal enrichment and favouring a third party, are much more dangerous.

Hiring and firing ministers: Germany 35 ‘Of course, resignations are not always what they seem’ (Marshall 1989: 127). The true reasons for a ministerial resignation are often difficult to identify, even if the occasion might appear obvious at first sight. It seems especially tempting to look for the ‘true’ reasons behind the 18 pull resignations in Germany. But in this respect, Bodo Hombach, Head of Gerhard Schröder’s Federal Chancellor’s office, is indeed the only one where the official reasons given for his resignation are doubtful and the connection between his leaving the cabinet and his intra-party standing (as well as personal enrichment allegations) is rather obvious. He was unpopular with the left wing of the SPD who blamed him for poor coordination and communication within the, at the time, crisis-prone government. Furthermore, allegations about financial irregularities regarding the construction of his private home came up. After a short decency period he was posted to Sarajevo as Special Co-ordinator of the Stability Pact for South-East Europe. The reasons for the eleven protest resignations are very diverse. Most of the ministers concerned stepped down because of differences with the government’s policy plans regarding their own portfolios (Heinrich von Brentano, CDU, Foreign Minister, 1961; Ewald Bucher, FDP, Minister of Justice, 1965; Paul Lücke, CDU, Minister of Home Affairs, 1968; Alex Möller, SPD, Minister of Finance, 1971; Karl Schiller, SPD, Minister of Finance and Economic Affairs; Erhard Eppler, SPD, 1974, Minister for Economic Co-operation and Development; Antje Huber, SPD, 1982, Minister for Youth, Familiy and Health). Only one of those occurred because of conflicts regarding the portfolio’s budget (Erhard Eppler). A few ministers resigned as a consequence of a policy clash in a field outside their responsibility (Gustav Heinemann, CDU, Minister of Home Affairs, 1950; Christian Schwarz-Schilling, CDU, Minister of Postal Affairs, 1992), and Sabine Leutheusser-Schnarrenberger (FDP) left her office in 1996 to circumvent a policy clash with her own party regarding the use of wiretapping in the fight against crime, an issue she was in charge of in her function as Minister of Justice. One of the most spectacular protest resignations was surely that in 1999 of Oskar Lafontaine (SPD, Finance Minister), who, after resigning, kept absolute silence for days before he complained about the intra-governmental lack of team spirit. Interestingly, the Ministry of Finance seems to be particularly susceptible to protest resignations. Comparing the parties’ share of push resignations over all ministerial discontinuations, the percentages range from 6 per cent (CDU/CSU) and 7 per cent (FDP) to 9 per cent (SPD). The Green Party’s share of 25 per cent has to be interpreted against the background that there have been only four ministerial discontinuations so far. Regarding ministers shuffled out during term, the percentages are similar among parties, again with the exception of the Green Party with no such case so far. These numbers suggest that big and small parties do not differ much when it comes to the potentially conflictive issue of sacking a minister.

Conclusion Although Germany’s Constitution provides next to no formal constraints with regard to selecting and de-selecting ministers by the Federal Chancellor, her room

36 The selection of ministers in Europe for manoeuvre is in fact restricted in a number of respects. Two factors are very important. First, the German government is a coalition government. The Federal Chancellor may want to consult with the leaders of the coalition partners regarding promising ministerial personnel but in the end has to accept their decisions. Secondly, Germany is a federal political system. This has a direct and an indirect effect on the selection of ministers. The direct effect is related to the specific type of joint federalism where the federal level to a large degree depends on the willingness of the Länder governments to implement policy decisions and therefore has to get their agreement in the Bundesrat, Germany’s quasi-second chamber. This implies that Federal Chancellors are well advised to include some personnel in their government with experience and expertise on the level of Länder governments. However, there is also an indirect effect of federalism. Germany’s major political parties are federalized organizations with resourceful Länder associations. Under these circumstances, Federal Chancellors can employ two different strategies. One is to ignore the wishes of Länder party associations and to build a power base on the parliamentary party in the Bundestag. The other option is to include regional party barons in the cabinet in the hope that this has a stabilizing effect. This seems to be the strategy that Angela Merkel has chosen (see Figure 2.1), although she interestingly ignored the wishes of the largest regional party association from North Rhine Westphalia. These informal constraints make composing a government acceptable for all actors capable of influencing the survival of the government a delicate task. However, these same constraints are mainly responsible for the fact that those actually chosen to serve as minister cannot easily be shuffled out.

Notes 1 The German constitution is called Grundgesetz (Basic Law). An English version is available online at http://www.bundestag.de/htdocs_e/parliament/function/legal/ germanbasiclaw.pdf.. 2 This section deals exclusively with senior ministers. Junior ministers as non-cabinet members are subject to different rules which will be introduced below. 3 The Federal Minister Law (Bundesministergesetz, BMinG) is available online at http:// bundesrecht.juris.de/bundesrecht/bming/gesamt.pdf (German version). 4 As a consequence, some MPs of the coalition partner FDP abstained from voting for Heuss’s 1954 re-election as Federal President. This episode however did not cause serious harm to inter-party relations within the coalition. 5 For empirical evidence see Müller-Rommel 1994 and Merz 2001: 65–73, both based upon surveys of ministers. 6 Before the reform of German federalism in 2006, more than 50 per cent of the Bundestag’s bills had to seek approval by the Bundesrat. One aim of the reform was to reduce this rate, but there is disagreement whether this can be achieved. Georgii and Borhanian (2006) predict a rate between 25 per cent and 30 per cent, Burkhart and Manow (2006) consider a rate of 35 to 40 per cent to be more realistic. 7 The standing orders of the federal government are called Geschäftsordnung der Bundesregierung and abbreviated as GeschOBReg. A German version is available online at http://www.bundesregierung.de/Content/DE/__Anlagen/geschaeftsordnungder-bundesregierung.html. 8 Parlamentarische Staatssekretäre may also carry the denomination Staatsminister, but

Hiring and firing ministers: Germany 37

9 10 11

12

13 14 15 16 17

18

19

20 21 22 23 24

this does not change their formal status. State Ministers can be found in the Foreign Office, but also in the Federal Chancellery. The exception was introduced in 1999, when the new Schröder government installed a state minister for cultural affairs in the Federal Chancellery who was not a MP. The Law on Parliamentary State Secretaries is called Gesetz über die Rechtsverhältnisse der Parlamentarischen Staatssekretäre and abbreviated as ParlStG. A German version is available online at http://bundesrecht.juris.de/bundesrecht/parlstg_1974/gesamt.pdf. The division of the Greens in ‘Realos’ and ‘Fundis’ has its origin in the 1980s, when the more fundamentalist members with a radical anti-system ideology formed the ‘Fundis’ and the group more orientated towards realpolitik was called ‘Realos’. This denomination still persists but has lost importance. The Green Party’s ideological wings are more pronounced than in other German parties. They still play a role in internal selection processes for any office or mandate. There were very few exceptions. The Ministry of Environmental Affairs operated in the first nine months of its existence in 1986–87 without a junior minister, and the ministries without portfolio which were created as a result of the German reunification in October 1990 and lasted only until the investiture of a new government three months later did not establish junior ministers either. For the first cabinet there was no parliament from which ministers could be recruited. This contains the acting standing committee (Geschäftsführender Fraktionsvorstand) as well as the extended standing committee (Erweiterter Fraktionsvorstand); the FDP works only with an acting standing committee. The overall share in 14 West European countries is 52.3 per cent. This was at the same time balanced by the selection of the trade unionist Walter Riester (SPD) as Minister of Labour. Symptomatically the main headline in the Frankfurter Allgemeine Zeitung the day after Ursula Lehr’s appointment reads: “Enttäuschung in der Unionsfraktion nach Kohls Entscheidung für Frau Lehr” (Disappointment in the CDU/CSU’s parliamentary party after Kohl’s decision for Mrs Lehr), 30 November 1988, p. 1. Typically, they lead a portfolio allowing them to sharpen their profile. The longtime CDU minister Norbert Blüm, once head of his party’s employees’ association, was considered to be the ‘social conscience’ of the Kohl government and held the Ministry of Labour and Social Affairs from 1982 to 1998. Likewise, Otto Schily (SPD) appealed to more right-wing voters with his law and order agenda as Home Secretary from 1998 to 2005. When composing the cabinet in 2005 Angela Merkel failed to select a CDU minister from North Rhine Westphalia. Since then the regional party leader and ministerpresident Jürgen Rüttgers is said to be on collision course with regard to both federal government and intra-party policy. A German version of the Satzung des Bundesverbandes Bündnis 90/Die Grünen is available online at http://www.gruene.de/cms/files/dokbin/32/32483.die_satzung _von _buendnis_90die_gruenen.pdf, paragraph 14 (4). We borrow these terms from migration research where push and pull motives of migration are distinguished: see Lee (1972). ‘Office discontinuation’ means leaving a portfolio, but not necessarily the cabinet. A minister can be shuffled to a different department. We consider a reshuffle a change of at least three ministers at the same time. Cynics might argue that Gerhard Schröder’s first term was one large reshuffle, with no fewer than eight individual resignations.

References Armbruster, F. (1973) Ressort-Rotationen in Grossbritannien und in der Bundesrepublik. Minister: Fachmann oder Politiker?’ Zeitschrift für Parlamentsfragen 4: 95–110.

38 The selection of ministers in Europe Bahr, E. (1996) Zu meiner Zeit. Munich: Blessing. Beyme, K. von (1983) ‘Coalition Government in Western Germany’. In Bogdanor, V. (ed.) Coalition Government in Western Europe. London: Heinemann, 16–37. Blondel, J. (1985) Government Ministers in the Contemporary World. London: Beverly Hills and New Delhi: Sage. Browne, E. C. and Franklin, M. (1973) ‘Aspects of Coalition Payoffs in European Parliamentary Democracies’. American Political Science Review 67: 453–69. Budge, I. and Keman, H. (1990) Parties and Democracy: Coalition Formation and Government Functioning in Twenty States. Oxford: Oxford University Press. Burkhart, S. and Manow, P. (2006) ‘Was bringt die Föderalismusreform? Wahrscheinliche Effekte der geänderten Zustimmungspflicht’. MPIfG Working Paper 06/6, Max Planck Institute for the Study of Societies Cologne (available online at http://www.mpifg.de/ pu/workpap/wp06-6/wp06-6.html), (downloaded 27 August 2007). Conradt, D. P. (2005) The German Polity, 8th edition. New York: Longman. Dewan, T. and Myatt, D. P. (2007) ‘Scandal, Protection, and Recovery in the Cabinet’. American Political Science Review 101: 63–77. De Winter, L. (1991) ‘Parliamentary and Party Pathways to the Cabinet’. In Blondel, J. and Thiébault, J.-L. (eds) The Profession of Government Minister in Western Europe. Basingstoke: Macmillan, 44–69. Dowding, K. and Kang, W.-T. (1998) ‘Ministerial Resignations 1945–97’. Public Administration 76: 411–429. Eschenburg, T. (1966) Zur Politischen Praxis in der Bundesrepublik, vol. 2. Munich: Piper. Fischer, J. (2004) Ministerrücktritte in der Bundesrepublik Deutschland von 1983–2002. Eine quantitative Analyse. Unpublished diploma thesis, University of Cologne. Fischer, J., Kaiser, A. and Rohlfing, I. (2006) ‘The Push and Pull of Ministerial Resignations in Germany, 1969–2005’. West European Politics 29: 709–735. Franzmann, S. (2006) ‘Parteistrategien auf oligopolistischen Issue-Märkten’. Politische Vierteljahresschrift 47: 571–594. Gabriel, O. W., Niedermayer, O. and Stöss, R. (eds) (1997) Parteiendemokratie in Deutschland. Bonn: Bundeszentrale für politische Bildung. Gamson, W. A. (1961) ‘A Theory of Coalition Formation’. American Sociological Review 26: 373–382. Georgii, H. and Borhanian, S. (2006) Zustimmungsgesetze nach der Föderalismusreform: Wie hätte sich der Anteil der Zustimmungsgesetze verändert, wenn die vorgeschlagene Reform bereits 1998 in Kraft gewesen wäre? Wissenschaftliche Dienste des Deutschen Bundestages, Vol. WD 3 – 37/06 und 123/06. Berlin: Deutscher Bundestag (available online at http://www.bundestag.de/bic/analysen/2006/zustimmungsgesetze_nach_der_ foederalismusreform.pdf) (downloaded 27 August 2007). Grewe, W. (1951) ‘Parteienstaat – oder was sonst?’. Der Monat 3, 36: 536–537. Herzog, R. (1983) Art. 64, in Maunz, Theodor and Dürig, Günter (eds) Grundgesetz. Kommentar, vol. 4. Munich: Beck, 1–24. Hesse, J. J. and Ellwein, T. (2004) Das Regierungssystem der Bundesrepublik Deutschland, vol. 1, 9th edition. Berlin: de Gruyter. Indridason, I. H. and Kam, C. (2008) ‘Cabinet Shuffles and Ministerial Drift’. British Journal of Political Science 38: 621–656. Ismayr, W. (2000) Der Deutsche Bundestag im Politischen System der Bundesrepublik Deutschland. Opladen: Leske + Budrich. Kempf, U. (2001) ‘Die Regierungsmitglieder als soziale Gruppe’. In: Kempf, U. and Merz, H.-G. (eds) Kanzler und Minister 1949–1998. Wiesbaden: Westdeutscher Verlag, 7–35.

Hiring and firing ministers: Germany 39 Kröger, K. (1974) ‘Der Parlamentarische Staatssekretär – Gehilfe oder Mimikry des Ministers?’ Die Öffentliche Verwaltung 27: 585–90. Lange, R.-P. (1973) ‘Auslesestrukturen bei der Besetzung von Bundesämtern’. In Dittberner, J. and Ebbighausen, R. (eds) Parteiensystem in der Legitimationskrise. Opladen: Westdeutscher Verlag, 132–171. Laver, M. and Shepsle, K. A. (1996) Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Lee, E. S. (1972) ‘Eine Theorie der Wanderung’. In Széll, G. (ed.) Regionale Mobilität. Munich: Nymphenburger Verlagshandlung, 115–129. Manow, P. (2005) ‘Will Junior Ever Grow Up?’ or, Where and When Do Junior Ministers Become Senior Ministers? An Empirical Investigation for 23 OECD Countries,1949–2003’. Unpublished paper presented at ECPR Joint Sessions of Workshops, Granada. Manow, P. and Zorn, H. (2004) ‘Office versus Policy Motives in Portfolio Allocation: The Case of Junior Ministers’. MPIfG Discussion Paper 04/9, Max Planck Institute for the Study of Societies Cologne (available online at http://www.mpifg.de/pu/mpifg_dp/dp 04-9.pdf) (downloaded 20 November 2006). Marshall, G. (1989) ‘Individual Responsibility: Some Post-war Examples’. In Marshall, G. (ed.) Ministerial Responsibility. Oxford: Oxford University Press, 127–133. Mehde, V. (2001) ‘Die Ministerverantwortlichkeit nach dem Grundgesetz. Dogmatischer Kernbestand und aktuelle Herausforderungen’. Deutsches Verwaltungsblatt 1: 13–19. Mershon, C. (2002) The Costs of Coalition. Stanford: Stanford University Press. Merz, H.-G. (2001) ‘Regierungshandeln im Lichte einer Befragung deutscher Bundesminister’. In Kempf, U. and Merz, H.-G. (eds) Kanzler und Minister 1949–1998. Wiesbaden: Westdeutscher Verlag, 36–81. Müller-Rommel, F. (1994) ‘The Role of German Ministers in Cabinet Decision Making’. In Laver, M. and Shepsle, K. A. (eds) Cabinet Ministers and Parliamentary Government. Cambridge: Cambridge University Press, 150–168. Patzelt, W. J. (2005) ‘Die Bundesregierung’. In Gabriel, O. W. and Holtmann, E. (eds) Handbuch Politisches System der Bundesrepublik Deutschland, Munich and Vienna, 3rd edition, Oldenbourg, 233–61. Saalfeld, T. (2003a) ‘Germany. Stable Parties, Chancellor Democracy, and the Art of Informal Settlement.’ In Strøm, K. and Müller, W. C. (eds) Coalition Governments in Western Europe. New York: Oxford University Press, 32–85. Saalfeld, T. (2003b) ‘Germany: Multiple Veto Points, Informal Coordination, and Problems of Hidden Action’. In Strøm, K., Müller, W. C. and Bergman, T. (eds) Delegation and Accountability in Parliamentary Democracies. New York: Oxford University Press, 347–375. Schindler, P. (1999) Datenhandbuch zur Geschichte des Deutschen Bundestages 1949–1999, complete edition in three volumes, vol. 1. Baden-Baden: Nomos Verlagsgesellschaft. Schmid, J. (1990) Die CDU. Organisationsstrukturen, Politiken und Funktionsweisen einer Partei im Föderalismus. Opladen: Leske + Budrich. Schneider, H.-P. and Zeh, W. (1989) ‘Koalitionen, Kanzlerwahl und Kabinettsbildung’. In Schneider, H.-P. and Zeh, W. (eds) Parlamentsrecht und Parlamentspraxis in der Bundesrepublik Deutschland. Berlin and New York: de Gruyter, 1297–1324. Schofield, N. and Laver, M. (1985) ‘Bargaining Theory and Portfolio Payoffs in European Coalition Governments 1945–1983’. British Journal of Political Science 15: 143–164.

40 The selection of ministers in Europe Schröder, M. (2005) ‘Art. 62–65a’. In Mangoldt, H. von, Klein, F. and Starck, C. (eds) Kommentar zum Grundgesetz, vol. 2, 5th edition. Munich: Vahlen, 1545–1628. Schüttemeyer, S. S. (1998) Fraktionen im Deutschen Bundestag 1949–1997. Empirische Befunde und theoretische Folgerungen. Wiesbaden: Westdeutscher Verlag. Stolz, K. (2003) ‘Moving Up, Moving Down. Political Careers across Territorial Levels’. European Journal of Political Research 42: 223–248 Sutherland, S. L. (1991) ‘Responsible Government and Ministerial Responsibility. Every Reform Is Its Own Problem’. Canadian Journal of Political Science 24: 91–120. Syed Ali, A. (2003) Karrierewege und Rekrutierungsmuster bei Regierungsmitgliedern auf Bundesebene von 1949–2002. PhD thesis, Martin-Luther-University HalleWittenberg. Thiébault, J.-L. (1991) ‘Local and Regional Politics and Cabinet Membership’. In Blondel, J. and Thiébault, J.-L. (eds) The Profession of Government Minister in Western Europe. Basingstoke: Macmillan, 31–43. Thies, M. F. (2001) ‘Keeping Tabs on Partners: The Logic of Delegation in Coalition Governments’. American Journal of Political Science 45: 580–598. Warwick, P. V. and Druckman, J. N. (2001) ‘Portfolio Salience and the Proportionality of Payoffs in Coalition Governments’. British Journal of Political Science 31: 627–649. Wengst, U. (1984) ‘Ministerverantwortlichkeit in der parlamentarischen Praxis der Bundesrepublik Deutschland: Eine historische Bestandsaufnahme’. In Zeitschrift für Parlamentsfragen 15: 539–551. Winkler, H.-J. (1967) Der Bundespräsident. Repräsentant oder Politiker? Opladen: C. W. Leske Verlag.

3

Cabinet dynamics and ministerial careers in the French Fifth Republic Christopher Kam and Indridi Indridason

Introduction France has operated under two constitutional regimes since the end of the Second World War, the purely parliamentary Fourth Republic (1946–58) and since then the semi-presidential Fifth Republic. The immediate cause of the Fourth Republic’s demise was the open refusal by the French military to countenance the decolonization of Algeria, but the Algerian Crisis was itself rooted in the structural weakness of the Fourth Republic: the country’s deep-seated social cleavages combined with the Fourth Republic’s permissive proportional representation electoral system to produce legislative assemblies that were too fragmented and polarized to maintain stable cabinets. The chronic cabinet instability of the Fourth Republic thus left French civil servants and military officers in the regions and colonies to their own devices, unconstrained by and ultimately resentful of civilian executive authority. The crisis was averted by de Gaulle agreeing to return to power, but he did so on the condition he be given a free hand to draft a new constitution. This draft constitution was put to referendum on 28 September 1958, and endorsed overwhelmingly by the French electorate. The Fifth Republic came into force on 4 October 1958. The Fifth Republic is an innovative fusion of presidential and parliamentary government. The Constitution outlines a dual executive composed of a President, popularly elected since 1962, and a Prime Minister who has the confidence of the majority of the legislature. Article 8 of the Constitution establishes the relationship between the two executive officers, empowering the President to appoint the Prime Minister and, on the advice of the Prime Minister, other cabinet ministers. Article 8 also allows the President to terminate the Prime Minister’s appointment when the latter tenders the government’s resignation.1 Similarly, the President terminates the appointments of other cabinet ministers on the Prime Minister’s advice. In addition to the powers of appointment and dismissal provided by Article 8, the President is granted the power to chair meetings of the Council of Ministers (i.e. the cabinet) (Article 9) and dissolve the National Assembly (Article 12). Finally, the Constitution designates the President as the ‘guarantor of national independence, territorial integrity and observance of treaties’ (Article 5) and commander-in-chief of the armed forces (Article 15). These last two articles allow the President to dominate the conduct of French foreign policy.

42 The selection of ministers in Europe The Prime Minister’s powers flow from Articles 20 and 21 (Chagnollaud and Quermonne 1996: 337–8). Article 20 empowers the government ‘to determine and conduct the policy of the Nation’, whilst Article 21 places the government under the direction of the Prime Minister. In constitutional theory, then, the Prime Minister recommends cabinet ministers to the President, sets governmental policy and determines the boundaries of departmental competencies (Elgie 1993: 10). The chief constraints on these Prime Ministerial powers are, first, that the ministerial appointments require the President’s approval, and secondly, that ministerial office is incompatible with parliamentary office (Article 23), i.e. ministers cannot sit in parliament and, should they hold a parliamentary seat, must resign it before taking up ministerial office.2 There are several classes of ministers within the government. The lowest rank is comprised of junior ministers, Secrétaires d’Etat. Above Secrétaires d’Etat one finds three classes of ministers, Ministres d’Etat, Ministres délégués and Ministres (Chagnollaud and Quermonne 1996, 295–8).3 The first of these titles is an honorific reserved for prominent or long-serving politicians, and its usage has declined over time. Ministres délégués are ministers who are responsible for a specific policy or function, often within the bailiwick of a larger portfolio (e.g. European Affairs within the Foreign Affairs portfolio). Ministre is the more typical title, identifying the holder as responsible for a specific portfolio. French cabinets averaged 36 ministers (all ranks) between 1968 and 2002 (16.5 Ministres, 4.5 Ministres délégués, 2.4 Ministres d’Etat, and 12.5 Secrétaires d’Etat), but have also grown slightly larger as time has passed. Indeed, while much was made of Sarkozy’s intention to reduce the size of the cabinet, Fillon’s second cabinet (June 2007) grew to 32 members, his first having included only 21 members.

The informal rules of the game in French politics Thanks to the bipolar nature of the French party systems and the important role of pre-electoral agreements, French governments are usually formed within a couple of days of an election (Thiébault 2000). A strict reading of the Constitution suggests that the Prime Minister controls the management and direction of the cabinet. The President, after all, cannot appoint or remove ministers without the Prime Minister’s initiative and, moreover, the President is required to do so upon the recommendation of the Prime Minister. Things work differently in practice, however. The crucial variable is whether the President’s party (or coalition) controls the National Assembly. A President who enjoys the support of a parliamentary majority is the cabinet’s dominant political figure, in charge of its composition, membership and political direction. In 1976, for example, Giscard d’Estaing created a post for a junior minister of industrial affairs without informing the Prime Minister. Ministers have also been appointed by the President against the advice of the Prime Minister (Safran 1998). For example, in 1974, Giscard d’Estaing appointed Jean-Jacques Servan-Schreiber against Prime Minister Chirac’s wishes (Knapp and Wright 2006). Of course, the membership of the cabinet is typically the result of consultation between the President and Prime Minister (Wright 1989: 86). However, these

Cabinet dynamics in the French Fifth Republic 43 examples show that during periods of unified government it is the President who has the upper hand. The premiership in these periods takes on the character of a presidential deputy, and certainly Presidents at the head of unified governments have not shied away from sacking their Prime Ministers. De Gaulle set the precedent early on by sacking his first two Prime Ministers, Debré and Pompidou. Pompidou’s case is especially illustrative in that he was sacked after just having led the Gaullists to a landslide victory at the legislative elections of 1968 (Time 1968). Pompidou’s transgression was to hint at his succession to the presidency (Wright 1989), and his subsequent sacking reveals an important maxim of French politics under unified government, namely, that Presidents dismiss Prime Ministers who are either too unpopular or too popular. As Elgie (1993: 1) notes: When things go well, the President often receives the credit. When things go badly, the Prime Minister usually takes the blame. If things go very badly, and the President starts to be criticised, then the Prime Minister is replaced. If things go very well and the Prime Minister starts to be praised, then the Prime Minister is also replaced. Things change markedly under cohabitation, the Prime Minister becoming the dominant player in the cabinet and in domestic politics more broadly, the President retiring into his domaine réservé of defence and foreign affairs. Prime Ministers do not have untrammelled power to appoint, dismiss or reshuffle ministers during cohabitation (divided government), however. Elgie (1993: 50) notes, for example, that Mitterrand vetoed Chirac’s appointment of Léotard as Minister of Defence in March 1986. Now, Léotard did get into cabinet as minister of Culture and Communication, and Chirac managed to replace Jacques Fournier, a Mitterrand ally, with Renaud Denoix de Saint Marc, a gaullist, as head of GSG (the French Secret Service). These examples indicate that even during cohabitation Prime Minister Chirac’s ability to appoint and dismiss ministers was constrained by President Mitterrand. Stevens (1992: 103) suggests that in these situations reshuffles will take the form of a bargain between the two executives. The dual executive nature of the Fifth Republic would seem to place French ministers in the position of having to serve two masters, especially during periods of cohabitation. The safety valve in the system is a convention of ministerial autonomy – which is itself less a constitutional ideal than a tacit recognition of a particular set of incentives and constraints. First, the Fifth Republic’s dual executive facilitates this sort of ministerial autonomy because it allows ministers to circumvent the Prime Minister by appealing to the President directly (Elgie 1993: 32–3). Secondly, many French ministers have their own power bases as heads of party factions (courants).4 In the Parti Socialiste (PS), for example, Chevènement, Rocard, Mitterrand, Strauss-Kahn, Jospin, Fabius, Hollande, and Dray were all ministers and factional chieftains (Knapp 2004: 167). Ministers have also been able to amass independent political capital via the French device of cumul des mandats, that is, the accumulation of other elected offices. Many deputies and ministers seek mayoral or other positions to secure a power base upon which to solidify their political careers.5 Thirdly, French ministers are provided with the

44 The selection of ministers in Europe institutional resources to operate independently. French ministers have cabinets of advisers loyal to them personally, head highly centralized departments and are less subject individually to parliamentary scrutiny than ministers in purely parliamentary systems (Thiébault 1994: 140–1). Finally, mutual self-interest induces ministers to keep to their own portfolios, a tradition labelled cloisonnement or ‘compartmentalization’ (Stevens 1992: 104; Elgie 1993: 30–1).

Ministerial selection French Prime Ministers are not restricted to choosing ministers from the membership of the National Assembly, and frequently look to the civil service, academia or the private sector for ministerial recruits. This flexibility means that the set of people from whom French Prime Ministers select their ministers is neither strictly defined nor fully observed. Indeed, one observes only the people to whom the Prime Minister has offered (and who have accepted) cabinet office. This selection bias complicates efforts to make accurate statements about ministerial selection in France. Even if we observe a high proportion of lawyers among French ministers, for example, it remains possible that the proportion of lawyers is just as high in the (unobserved) set of those eligible for cabinet office, and that practising law is statistically independent of cabinet membership. We require some other frame of reference with which to put our data on ministerial selection in context. One comparison to draw is between the Fifth Republic’s first cabinet, Debré I (1959–62), and its successors. Debré I was dominated by politicians who had built their careers under the parliamentary institutions of the Third and Fourth Republics, and so comparisons between Debré I and later cabinets provide a sense of how the complexion of the French cabinet has changed over time and of how it has responded to the transition to semipresidential government. We focus on three ministerial characteristics, academic training (at the Ecole Nationale d’Administration (ENA), in particular), gender and political experience prior to obtaining ministerial office. Other traits are almost certainly relevant to ministerial selection, but these three characteristics afford us some basis for inferring whether longitudinal changes in the make-up of the French cabinet reflect the changing importance of certain selection criteria or simply mirror changes in the composition of the recruitment pool. Table 3.1 shows the distribution of these traits across French cabinets. The figures for the Debré cabinet in the first row are followed by the figures for the remaining 30 cabinets clustered in two ways. First, we group the governments by periods of left or right government. Secondly, to provide a more balanced picture of long-term development, we group governments into 15-year periods. To maintain comparability across cabinets and to avoid double-counting individuals, we confine our sample to the initial membership of each cabinet and record data only for an individual’s first ministerial office.6

1.8 10.0 24.8 12.7

0.0 3.0 14.6 5.0 16.1 18.9 33.3 26.5

Women

75.0 63.3 63.4 65.8

61.2 73.1 61.0 75.0 51.6 83.8 55.6 50.0

Parliamentary (deputy or senator)

1.8 4.4 9.7 5.4

0.0 3.0 4.9 5.0 3.2 5.4 0.0 17.7

Local politics (regional Councillor or mayor)

Prior political experience

7.1 7.8 8.6 9.2

23.8 7.5 4.9 15.0 12.9 2.7 11.1 11.8

Ministerial cabinet

5.4 3.3 3.2 3.9

4.8 4.5 4.9 0.0 3.2 2.7 11.1 2.9

Party position

10.7 20.0 11.8 14.2

9.5 11.9 22.0 5.0 29.0 5.4 0.0 17.7

None

56 90 93 260

21 67 41 20 31 37 9 34

N

Note: To avoid double counting and ensure a fixed numerical base we have constructed these categories so that they are nested in one another from right to left in the table. The broadest category is therefore parliamentary experience; ministers counted in this column may also have experience in local politics, whereas those listed with experience in local politics are those without any parliamentary experience. Similarly, those who are counted as having local political experience may also have experience as cabinet advisers, but not the reverse. These percentages do not add up to 100 because a small percentage of ministers remaining have experience as Members of the European Parliament.

16.1 10.0 20.4 16.5

Pompidou I – Chirac I 14.04.62–25.08.76 12.5 Barre I – Rocard II 27.08.76–15.05.91 25.6 Cresson I – Raffarin III 15.05.91–31.05.05 20.4 Total 18.9

IEP

28.6 16.4 4.9 15.0 12.9 24.3 11.1 20.6

09.01.59–14.04.62 14.04.59–13.05.81 21.05.81–20.03.86 20.03.86–10.05.88 10.05.88–29.03.93 29.03.93–02.06.97 02.06.97–06.05.02 06.05.02 –31.05.05

Debré I Pompidou I – Barre III Mauroy I – Fabius I Chirac II Rocard I – Bérégovoy I Balladur I – Juppé II Jospin Raffarin I – III

ENA

Educational background

0.0 13.4 17.1 50.0 29.0 13.5 22.2 20.6

Dates

Cabinets

Table 3.1 Educational background, gender and political experience of French ministers (percentages of initial ministers), 1959–2005

46 The selection of ministers in Europe Ministerial selection and the ENA The ENA is one of several Grandes Ecoles which serve as training grounds for France’s administrative and political elite. What sets the ENA apart from its competitors and what we find useful, however, is the fact that the ENA was founded only in 1945; hence its first cohort of 100-odd graduates (known as énarques) was just entering the ministerial recruitment pool at the advent of the Fifth Republic.7 In contrast, many of the other Grandes Ecoles are over a century old, and so the number of their graduates in the ministerial recruitment pool can be assumed to have been stable throughout the Fifth Republic. Using one of these schools as a comparative base (we chose the Institut d’Etudes Politiques de Paris (IEP) founded in 1872) allows us to assess just how quickly the énarques penetrated the cabinet and how relatively advantageous an ENA education has been to ministerial aspirants. It took some time for a critical mass of énarques to build up in the recruitment pool and enter the cabinet. Indeed, no énarque held a portfolio in Debré’s initial cabinet, though one, Giscard d’Estaing, was the junior finance minister and was soon promoted.8 By the next wave of cabinets, however, 12.5 per cent of ministers were énarques, almost the same percentage as IEP graduates. Even so, it would take another 10–12 years for the first cohort of énarques to reach a retirement age of 65 and for their number in the recruitment pool to stop growing. This suggests that énarques enjoyed a success rate in securing ministerial office in the 1970s that was out of all proportion to their numbers in the recruitment pool. The fact that énarques and IEP graduates remained in parity in the ministry in the 1990s indicates that this advantage has subsided, though given the ENA’s small size (producing less than a fifth of the number of graduates as the IEP) it remains very large.9 Ministerial selection and gender Article 23 makes ministerial office incompatible with parliamentary office, but in doing so it frees French Prime Ministers to look outside parliament for their ministers. This removes two barriers to ministerial office that operated under the Fourth Republic: the need for ministerial aspirants to be selected as a candidate for parliamentary office and to win a seat in the National Assembly. These barriers have proved difficult for French women to overcome, and so one might expect women’s cabinet representation to have outstripped their parliamentary representation in the Fifth Republic.10 This did not occur. Debré’s initial cabinet contained no female ministers, with Nafissa Sid Cara joining the cabinet only as a junior minister via a reshuffle. The situation remained largely unchanged through the 1970s and 1980s. The breakthrough (relative to past levels of female cabinet and parliamentary representation) came in 1991 with the elevation of Edith Cresson to the premiership, the first woman to occupy the post. Even so, only 24.8 per cent of new ministers were women over the period 1991–2005 and the percentage of women in the French cabinet remains below parity, achieving a maximum of 34.4 per cent of the posts in Fillon’s present cabinet.

Cabinet dynamics in the French Fifth Republic 47 Ministerial selection and political experience One reason why women’s cabinet representation remains anchored to their lower levels of parliamentary representation is that the recruitment pool for ministers remains dominated by parliamentarians. Article 23 notwithstanding, 60–75 per cent of French ministers have prior experience as a deputy or senator. This far exceeds the proportion of ministers who arrive only with experience in local politics, as adviser in a ministerial cabinet or after having held an extraparliamentary party position. Of course, these positions are not mutually exclusive, and many ministers with parliamentary experience were also mayors, cabinet advisers or party executives. This fact hardly dilutes the essential message of these data: parliamentarians claim two of every three ministerial positions. As Gaxie noted in reaching a similar conclusion to ours, winning a national election (i.e. a seat in the National Assembly) is almost a necessity for advancement to the cabinet in the Fifth Republic (Gaxie 1986: 66). In addition, ministers who come to the cabinet from politics (rather than from the civil service) also tend to secure more politically important posts (Gaxie 1986: 68–9). The Fillon cabinets deserve a special mention with regard to political experience. Fillon’s first cabinet included among others a former socialist minister, Bernard Kouchner, as well as Hervé Morin of the Union pour la Démocratie Française (UDF) who had lent his support to Bayrou on the first ballot in the preceding presidential election. The composition of the first Fillon cabinet was perceived to be a strategic move to split the opposition and neutralize arguments that Sarkozy’s right-wing agenda needed to be balanced by a left-leaning legislature in the upcoming legislative election.

Ministerial career patterns Taking ministerial tenure as a metric of regime stability, Huber and MartinezGallardo (2004) argue that the standard view of a chronically unstable Fourth Republic being replaced by a stable Fifth Republic is overdrawn. This is not because ministers in the Fourth Republic managed to accumulate more experience than one might expect given the regime’s high level of cabinet instability, but because the Fifth Republic’s stability has not extended to ministerial longevity. Our data echo this message, but there are two points on which we diverge. First, whereas Huber and Martinez-Gallardo suggest that the initial surge in ministerial experience and stability in the early years of the Fifth Republic has declined over time, our data show no such trend. Secondly, our data indicate that the Fifth Republic has succeeded in replacing the statesmen of the Fourth Republic with its own stable core of ministrables. Huber and Martinez-Gallardo argue that the relationship between ministers’ portfolio experience (time spent in a specific office) and ministers’ cabinet experience (time spent in any cabinet post) provides important information on the cabinet’s dynamics. Low levels of portfolio experience combined with high levels of cabinet experience, for example, suggest a cabinet of musical chairs with core ministrables regularly exchanging portfolios. The distinction between portfolio

48 The selection of ministers in Europe experience and cabinet experience is a useful one, but in focusing on just the ten most prominent portfolios Huber and Martinez-Gallardo ignore the experience that ministers accumulate as they climb up the ministerial pecking order. A full picture of ministerial careers requires that one tracks both the lateral movement of ministers across portfolios and their vertical movement in the ministerial hierarchy. We use a three-tiered ranking of ministerial positions to track the vertical movement of ministers. The lowest rank is comprised of junior ministers, Secrétaires d’Etat. For the ranks above Secrétaires d’Etat we use Warwick and Druckman’s (2006) portfolio salience measures to divide portfolios into two sets, one containing the twelve most salient ‘major’ portfolios, the other the remaining ‘minor’ portfolios.11 As Ministres d’Etat tend to hold major portfolios, we count them as a major portfolio in cases where the title appears without a portfolio. Similar reasoning leads us to count Ministres délégués as minor portfolios unless there are clear indications to the contrary (e.g. Laurent Fabius as Ministre délégué auprès du ministre de l’économie et des finances, chargé du budget in Mauroy’s first cabinet). This leaves us with three ranks of ministers: junior ministers, minor ministers and major ministers. With this ranking system in place we can measure first the time spent in any specific post, secondly the time spent at a given rank, and thirdly the overall length of ministers’ careers, from the time they first enter the government until they depart from their final posts. We present these data in Table 3.2, broken down by decade over the course of the Fifth Republic so as to be comparable with Huber and Martinez-Gallardo’s data. The average number of months ministers spend in any specific office is between 10 and 15, with no evidence of systematic increases or decreases during the Fifth Republic, not even with the alternance of 1981. Much the same is true of ministerial careers, which range only between four and five years on average. One pattern that we observe in the data is that ministers spend about twice as long at a given rank as they do in any specific post, and further that their careers span twice as long again as the average time they spend at a given rank. This pattern suggests that a typical ministerial career in the Fifth Republic involves a minister gaining 24–30 months’ experience in two different posts before advancing in rank to spend an additional 24–30 months at two more posts. This picture of a typical ministerial career is somewhat misleading, however. Notice that the overall means in Table 3.2 are high relative to the decade-by-decade means. This suggests the presence of a coterie of experienced ministers scattered Table 3.2 Average time in the French government, 1960s–2000s Decade

Months in post

Months in rank

Months in government

1960s 1970s 1980s 1990s 2000s

14.3 10.2 12.6 12.8 13.4

26.6 24.6 25.0 26.7 22.0

57.2 48.9 51.8 64.6 54.0

Overall

15.3

28.2

58.3

Cabinet dynamics in the French Fifth Republic 49 amongst a large group of ministerial transients; only when this long-lived coterie is collected together, not dispersed across the decade-by-decade subgroups, does it noticeably affect the overall means. This view is buttressed by evidence set out in Figure 3.1, which shows the number of posts held by ministers through their careers. Most ministers hold only a single post, and the median number of posts held is just two, but about 10 per cent of ministers hold five or more posts in their careers. A similar picture emerges from Figure 3.2, which shows the length of ministerial careers in the Fifth Republic. The career length distribution is highly skewed to right with most careers lasting less than 50 months. Nevertheless, some 20 per cent of ministers have careers that are at least twice as long. The ministers in this class are among the most prominent figures of modern French politics: Giscard d’Estaing, Chirac, Faure, Raffarin, Rocard, Bérégovoy, Couve de Murville, Cresson, Juppé, Sarkozy, and Fabius to name a few. It is worth considering the career path of one of these long-lived ministers in detail. Jacques Chirac’s career serves as a good example both because Chirac built his career entirely within the confines of the Fifth Republic and because he served in so many capacities.12 Chirac’s ministerial career began on 6 April 1967 with his appointment as Secrétaire d’Etat and it was only after his fourth stint as a junior minister that he was promoted to Ministre délégué. Eighteen months later, in 1972, Chirac was given his first major portfolio, becoming Ministre de l’agriculture et du développement rural. Chirac then served briefly as Ministre de l’intérieur in 1974 before securing the premiership itself. On resigning the premiership on 25 August 1976, Chirac’s career trajectory takes a turn that would be considered unusual in many democracies. Chirac 40

Per cent

30

20

10

0 1

2

3

4

5

6

7

8

9

10

11

N posts held over career

Figure 3.1 Number of offices held by ministers in the French Fifth Republic, 1959–2005

50 The selection of ministers in Europe 40

Per cent

30

20

10

0 0

50

100

150

200

250

300

350

400

450

Months in government

Figure 3.2 The length of ministerial careers in the French Fifth Republic, 1959–2005

returned to parliament, secured the chairmanship of the Rassemblement pour la République (RPR), won and gave up a seat in the European Parliament, and become mayor of Paris (an office he retained until 1995). When he returned to government on 20 March 1986, Chirac returned as Prime Minister. Chirac’s second term as premier (and France’s first experience with cohabitation) lasted two years until socialist electoral gains forced him out of office. Even this was not the end of Chirac’s political career: Seven years later, on 7 May 1995, he was elected to the presidency. What we find instructive here, and quite unique to French politics, is the capacity of a small subset of French ministers to build up and fall back on alternative bases of political power so that they can leave and return to ministerial office, repeatedly and over long periods of time. How typical is Chirac’s career? How many ministers fall by the wayside after holding just one or two government posts for every Jacques Chirac who manages to move from portfolio to portfolio while climbing steadily up the ministerial pecking order? Table 3.3 helps us answer this question. For each minister we know what position the minister held just previously (i.e. p–1) and what position he held next (i.e. p+1). Cross-tabulating every minister’s past and future positions as in Table 3.3 thus provides an overall sense of ministerial advancement and mobility in the Fifth Republic.13 The bottom row of Table 3.3 indicates that any given point in time we can expect 36.9 per cent of French ministers to be exiting the government, 17.4 per cent to be acting as Secrétaires d’Etat, 24.1 percent to be holding minor portfolios, and 21.6 per cent to be holding major portfolios. Three things stand out in Table 3.3. First, between 35 and 40 per cent of all ministers are out of government within two periods, and, except for ministers in major portfolios,

Cabinet dynamics in the French Fifth Republic 51 Table 3.3 Career advancement and mobility among French ministers, 1959–2005. Vertical axis: position at p – 1. Horizontal axis: position at p + 1. Cell entries are row percentages, with the raw numbers shown in parentheses Outside

Secrétaire d’Etat

Minor portfolio

Major portfolio

Outside

36.7 (181)

22.2 (108)

26.7 (131)

14.3 (70)

490

Secrétaire d’Etat

37.4 (104)

34.9 (97)

17.6 (49)

10.1 (28)

278

Minor portfolio

38.2 (110)

5.2 (15)

32.3 (93)

24.3 (70)

288

Major portfolio

35.3 (85)

2.1 (5)

16.2 (39)

46.5 (112)

241

Total

36.9 (479)

17.4 (226)

24.1 (312)

21.6 (280)

1297

N

exit from the government is the most likely destination. Indeed, 14 per cent of all French ministers (181 of 1,297) are out of government after serving in a just single post. Secondly, alongside this picture of transience is one of stability: Ministers who still hold a post (i.e. p+1 is not ‘outside’) are most likely to be in a position of the same rank. This is especially true of holders of major portfolios, 46.5 per cent of whom retain a high-level cabinet position (though perhaps not the same one). Thirdly, fewer than 25 per cent of junior ministers or minor portfolio holders at any given point in time move on to a higher office. The corollary of most ministers exiting or remaining in place is that only a minority of French ministers advances in office.

Departing from government The short spells that French ministers spend in any specific post are largely due to frequent government terminations. What tends to happen is that every 18 months (on average) the Prime Minister brings his or her government to a formal end. Sometimes the Prime Minister stays in place, and the government is remade within days with many of the same ministers reappointed. On other occasions, however, the Prime Minister is replaced and a wholly new cabinet is formed. This process is often triggered by election results of some sort, and between presidential, parliamentary, cantonal, municipal, and European elections there is almost always an election for a French Prime Minister to respond to. Some recent examples show the range of conditions under which this process of government termination and reformation unfolds: 1. The Socialist victory at the June 1997 parliamentary elections. In 1997, Chirac dissolved the National Assembly to trigger early parliamentary elections. Chirac’s gambit backfired, however, and the PS and their allies won the election. Faced with a hostile parliamentary majority, Chirac was forced to accept Alain Juppé’s resignation and install Lionel Jospin as the new Prime Minister.

52 The selection of ministers in Europe 2. The Presidential and parliamentary elections of May–June 2002. Following his re-election and Jospin’s resignation, Chirac appointed an interim government led by his own party’s Jean-Pierre Raffarin. Notwithstanding his party’s success in the legislative election the following month, Raffarin tendered his government’s resignation, but Chirac immediately reappointed Raffarin and the government was reformed. 3. The regional elections of March 2004. The Union pour un Mouvement Populaire (UMP) suffered a significant setback at the March 2004 regional elections. Raffarin responded by tendering his resignation, but again Chirac reappointed Raffarin to the premiership and the government was reformed, though with many personnel changes. 4. The referendum to ratify the EU constitution in May 2005. Chirac and the government championed the ratification of the EU constitution and, following its defeat, Raffarin tendered his resignation for a third time. This time Chirac accepted and appointed Dominique de Villepin as Prime Minister. The first and fourth cases involve the constitutional replacement of one government by another. The second and third cases, on the other hand, are effectively cabinet reshuffles in the French style. In all these cases, however, ministers’ appointments are formally terminated. Consequently, it is not surprising to learn that, of the 1,307 ministerial posts in our dataset, 1,131 (86 per cent) were ended by a government termination. A more accurate account of ministerial attrition in France requires tracking how many ministers are reappointed to the government after these terminations. We do this in Table 3.4 which lists the six ways in which the ministerial appointments in our data set end: A cabinet termination – followed by reappointment or not, a reshuffle from one post to another exclusive of those that take place as a result of termination, the end of an interim appointment, resignation and death. Most ministers (54.2 per cent) are reappointed following a government termination, with two-thirds of these reappointments seeing the minister take up a different office in the government. Another 8 per cent of cases are shuffled independently of a cabinet termination. A further 32.4 percent of ministers are not reappointed. In many of these cases, the ministers are effectively sacked, but in others the ministers are not reappointed because their government has suffered an electoral defeat (see example 1 above). The remaining appointments are ended because their holder dies, has completed an interim appointment or resigns. Table 3.4 Reasons for the end of French ministerial appointments, 1959–2005

Cabinet termination without reappointment Cabinet termination followed by reappointment Reshuffle Resignation End of interim appointment Death Total

N

%

423 708 105 60 6 5 1307

32.4 54.2 8.0 4.6 0.5 0.4

Cabinet dynamics in the French Fifth Republic 53 Reasons for resignation Resignations are only a small fraction (4.6 per cent) of all ministerial exits, but because they are often connected to important political events they deserve attention. Policy differences are one reason for resignation, and several cases can be cited, e.g. Pierre Sudreau’s resignation as Minister of Education in 1962 in protest against the direct election of the President and Jean-Pierre Chevènement’s resignations, first as Minister of Research and Industry in 1983 in protest against Mitterrand’s changed course on economic policy and, then, from his post as Minister of Defence in 1991 in protest over the Gulf War. A resignation on the grounds of policy disagreement is also a means by which ministers distance themselves from the government and establish themselves as independent electoral actors. Rocard’s resignation from Fabius’s government in April 1985 is one example of this sort of resignation, as he intended to be a Presidential candidate for the Left had Mitterrand retired. Ministers are also forced from office by policy failures, though again there is a significant difference in the way these events are handled in France as opposed to Britain. In France, Article 68-1 of the Constitution makes ministers criminally liable for their actions in office. Article 68-1 has only been invoked once to date, in 1999 when Laurent Fabius, Georgina Dufoix and Edmond Hervé were charged with manslaughter on account of having failed in their official capacities to protect France’s blood supply from HIV contamination in the mid-1980s.14 All three were out of office by the time the trial occurred, and so one can only conjecture as to how the process might have affected their ministerial tenures; Hervé and Dufoix have not held ministerial office since. Resignations have also been guided by informal rules adopted by government leaders. Bérégovoy initiated the principle that ministers placed under investigation were required to resign from their post, and this principle has remained in place with Bernard Tapie resigning in 1992, Alain Carignon, Gérard Longuet and Michel Roussin in 1994, Dominique Strauss-Kahn in 1999 and Renaud Donnedieu de Vabres in 2002.15 Other prime ministers, e.g. Fillon, have insisted that ministers who run for election resign if they fail to win a seat. Alain Juppé, minister of the environment, transportation and energy and number two in Fillon’s cabinet, was forced to resign after losing the Bordeaux constituency in 2007. Personal scandals provide a third ground for resignation, and on this front we do not lack for examples. Financial scandals are especially frequent in France, and a party-financing scandal in the 1990s that encompassed almost every political party suggests that some of the problems are systemic rather than personal in nature.16 It is not just the number of ministers allegedly involved in corrupt activities that is startling, but their stature: Dominique Strauss-Kahn, a former finance minister, Alain Juppé and Pierre Bérégovoy, former premiers, and Chirac, himself, have all been accused of financial wrong-doing. One should not come away thinking that French ministerial careers are especially vulnerable to personal scandals, however. They are, in fact, fairly resistant: Of the 39 scandals we documented involving sitting ministers, only 13 led to resignation.17

54 The selection of ministers in Europe

Conclusions The Fifth Republic is a hybrid of parliamentary and presidential government. The introduction of a dual executive has certainly changed French cabinet dynamics. This is especially noticeable during periods of unified government, when the Prime Minister and the government are driven by the President’s political agenda. Thus, we observe prime ministers handing in their resignations and their governments terminated to protect the President. In other ways, however, the Fifth Republic has left longstanding cabinet dynamics intact. French ministerial careers have remained stubbornly parliamentary in nature, a repeatedly vacated and reoccupied seat in the National Assembly (or failing that, in the Senate) remaining the primary power base of most French ministers. And French ministers, at least those who hope to become part of the regime’s core set of ministrables, must develop alternative power bases. French ministerial careers are not long, but they are flexible and the government permeable; provided a ministrable remains politically active, in parliamentary, municipal or European politics, a French minister can repeatedly enter and exit government.

Notes 1 An English translation of the French Constitution of 4 October 1958 can be obtained online from the National Assembly’s website, http://www.assemblee-nationale.fr/ english/8ab.asp (last visited 9 March 2007). 2 Zarka (2000) is a useful guide to the constitutional arrangements and norms of France’s semi-presidential regime. 3 Garde des sceaux (Keeper of the Seals) is another ministerial title, but one that is held alongside the justice portfolio. Debré’s initial cabinet also listed four ministres conseillers, all of whom were indigenous political leaders of major French colonies. Once these colonies became independent, the title disappeared. We exclude these ministres conseillers from our analysis. 4 This is a more recent development; prior to Mitterrand’s presidency ministers tended to emerge from the civil service rather than from party politics (Thiébault 1994). About 35 per cent of the ministers since 1959 have had background in the civil service (Lacam 2000: 115). 5 Prime ministers Mauroy (Lille in Nord), Mollet (Arras in Pas-de-Calais) and Fabius (Grand-Quevilly in Seine-Maritime) all held mayorships, and among ministers Deffere (Marseille in Bouches-du-Rhône), Soldani (Draguignan in Var), and Chevènement (Belfort) held mayorships. Article 23 forbids members of the government from simultaneously holding another public office, so ministers must resign from any other such office before assuming government office. 6 At this stage of the analysis we exclude junior ministers. 7 If the first cohort of énarques graduated in 1948 (coursework at the ENA takes three years) at about 25 years of age, they would have required some 10 years of professional and political seasoning before they were eligible for ministerial office. (Only two ministers out of the 314 in our full sample assumed office before they were 35; Jacques Soustelle and François Mitterrand were 33 and 31 years old, respectively, on first assuming ministerial office.) This would mean that the first cohort of énarques would have been eligible for ministerial office in 1958. Although our sample contains ministers during the Fifth Republic, some ministers may have first entered office during the Fourth Republic. 8 Giscard d’Estaing was also an alumnus of the Ecole Polytechnique, another of the established Grandes Ecoles, however.

Cabinet dynamics in the French Fifth Republic 55 9 In the absence of exact numbers on the number of graduates from each school we cannot say precisely how much more likely an énarque was to be selected as a minister than an IEP graduate, and we should keep in mind that over time the number of ministers who have attended both schools has also grown. 10 Matland’s (1998) data show that only 4.3 per cent of National Assembly deputies were women in 1980, and only 10.9 per cent in 1997. As of October 2007, 13.3 per cent of French deputies were women (http://www.assemblee-nationale.fr/12/tribun/gs2.asp#P -1_0 (visited 7 October 2007)). 11 The twelve major portfolios are: the premiership, finance, interior, justice, foreign affairs, education, defence, budget, agriculture, labour, industry and planning. 12 An English translation of Chirac’s biography can be found at http://www.elysee.fr/ elysee/anglais/the_President/biography/biography.39706.html (visited 20 March 2007). 13 It may seem counter-intuitive not to use ministers’ present positions (i.e. p) in the cross-tabulations – but by definition a minister cannot have been outside the government at both p–1 and p. Of course, cross-tabulating ministers’ immediate past and future positions blinds the analysis to cases in which an individual moves from a low-ranking position at p–1 to a higher position at p, and then down again at p+1. Empirically, however, this is a minor issue: in only 17 cases did a government member at p–1 assume a higher office at p followed by a lower office at p+1. Similarly, in only 16 cases did a minister move from a higher office to a lower office and then back to a higher office. Thus, these irregular career trajectories represent just 2.5 per cent of the data. 14 Fabius, Dufoix and Hervé were Prime Minister, Social Affairs Minister and Secretary of State for Health, respectively, in Fabius’s government. See, e.g., Bosia (2007), Chalaby (2004) and British Medical Journal (1998). 15 See http://www.rfi.fr/actufr/articles/049/article_26173.asp (visited 6 February 2007). 16 The scandal resulted in a number of convictions of ministers or former ministers, including Henri Emmanuelli of the PS, François Léotard of the Parti Républicain (PR) and the UMP, Pierre Méhaignerie of the Centre des Démocrates Sociaux (CDS) and Gérard Longuet of the PR. 17 Half the scandals involved financial misconduct. Many instances of financial misconduct come to light only after the ministers have left office. Of 43 ministers placed under legal investigation for financial misconduct between 1990 and 2000, only five were in office at the time (Adut 2004).

Acknowledgments and note on sources The authors wish to thank Philippe Poirier for his comments on an earlier version of the chapter. Our data on ministerial appointments are derived from Les gouvernements et les assemblées parlementaires sous la Vème République 1958–2004 (Paris: Assemblée nationale, July 2004). Biographical data come primarily from Who’s Who in France, though for ministers who held office early in the Fifth Republic, Benoit Yvert’s Premiers ministres et présidents du conseil: histoire et dictionnaire raisonné des chefs du gouvernement en France, 1815–2002 (Paris: Perrin, 2002) and websites at the Académie française (http://www. academie-francaise.fr/) and Ordre de la Libération (http://www.ordredela liberation.fr/) proved helpful. Information on political scandals is taken from Keesing’s Contemporary Archives.

56 The selection of ministers in Europe

References Adut, A. (2004) ‘Scandals as Norm Entrepreneurship Strategy: Corruption and the French Investigating Magistrates’. Theory and Society 33: 529–578. Amorim Neto, O. and Strom, K. (2006) ‘Breaking the Parliamentary Chain of Delegation: Presidents and Non-partisan Cabinet Members in European Democracies’. British Journal of Political Science 36: 619–643. Bosia, M. J. (2007) ‘From Contention to Engagement: French Social Science and the Politics of HIV/AIDS’. French Politics 5: 96–105. British Medical Journal (1998) ‘Ex-ministers to Face Trial in Tainted Blood Scandal’. 7154, 1 August 1, 317: 302. Butler, D. (1994) ‘French Ministers Face Public Trial in HIV Blood Affair’. Nature 370, 6487: 243. Chagnollaud, D. and Quermonne, J.-L. (1996) Le gouvernement de la France sous la Ve République, nouvelle édition. Paris: Librarie Arthème Fayard. Chalaby, J. K. (2004). ‘Scandal and the Rise of Investigative Reporting in France’. American Behavioral Scientist 47, 9: 1194–1207. Conley, R. S. (2006) ‘From Elysian Fields to the Guillotine? The Dynamics of Presidential and Prime Ministerial Approval in Fifth Republic France’. Comparative Political Studies 39, 5: 570–598. Dogan, M. (1979) ‘How to Become a Cabinet Minister in France: Career Pathways 1870–1978’. Comparative Politics 12, 1: 1–25. Elgie, R. (1993) The Role of the Prime Minister in France, 1981–91. New York: St Martin’s Press. Gaxie, D. (1986) ‘Immuables et changeants: les ministres de la Vème République’. Pouvoirs 36: 61–78. Huber, J. D. and Martinez-Gallardo, C. (2004) ‘Cabinet Instability and the Accumulation of Experience: The French Fourth and Fifth Republics in Comparative Perspective’. British Journal of Political Science 34: 27–48. Knapp, A. (2004) Parties & the Party System in France. A Disconnected Democracy? New York: Palgrave Macmillan. Knapp, A. and Wright, V. (2006) The Government and Politics of France, 5th edition. London: Routledge. Lacam, J.-P. (2000) La France, une République de mandarins? Les hauts fonctionnaires et la politique. Paris: Éditions Complexe. Lewis, E. G. (1970) ‘Social Background of French Ministers, 1944–1967’. The Western Political Quarterly 23, 3: 564–578. Matland, R. (1998) ‘Women’s Representation in National Legislatures: A Comparison of Democracies in Developed and Developing Countries’ Legislative Representation’. Legislative Studies Quarterly 23: 109–125. Safran, W. (1998) The French Polity. New York: Longman. Stevens, A. (1992) The Government and Politics of France. New York: St Martin’s Press. Thiébault, J.-L. (1994) ‘The Political Autonomy of Cabinet Ministers in the French Fifth Republic’. In Laver, M. and Shepsle, K. A. (eds) Cabinet Ministers and Parliamentary Government. New York: Cambridge University Press, 139–149. Thiébault, J.-L. (2003) ‘France: Forming and Maintaining Government Coalitions in the Fifth Republic’. In Müller, W. C. and Strøm, C. (eds) Coalition Governments in Western Europe. Oxford: Oxford University Press. Time (1968) ‘A Sudden Parting: How Pompidou Was Fired’ 19 July. Warwick, P. V. and Druckman, J. (2006) ‘The Portfolio Allocation Paradox: An

Cabinet dynamics in the French Fifth Republic 57 Investigation into the Nature of a Very Strong but Puzzling Relationship’. European Journal for Political Research 45: 635–665. Wolfreys, J. (2001) ‘Shoes, Lies and Videotapes: Corruption and the French State’. Modern and Contemporary France 9, 4: 437–451. Wright, V. (1989) The Government and Politics of France, 3rd edition. London: Routledge. Zarka, J.-C. (2000) Fonction présidentielle et problématique majorité présidentielle / majorité parlementaire sous la cinquième République (1986–1992). Paris : LGDJ.

4

Choosing, moving and resigning at Westminster, UK1 Samuel Berlinski, Torun Dewan, Keith Dowding and Gita Subrahmanyam

Constitutional situation The UK is a constitutional monarchy. The current queen, Queen Elizabeth II, ascended to the throne in 1953 following the death of her father. There are two Houses of Parliament at Westminster, the House of Commons and the House of Lords. Throughout the twentieth century (certainly since 1911 when powers were transferred) the Commons has been the dominant house, and not since Lord Salisbury (who resigned in July 1902) has there been any Prime Minister from the Lords. Though ministers are still drawn from the Lords, the major offices of state sit in the Commons (the appointment of Lord Carrington as Foreign Secretary in 1979 was considered unusual). The Commons is now elected from around 650 constituencies in England, Scotland, Wales and Northern Ireland.2 From the time of the 1911 Parliament Act the House of Lords has provided oversight of legislation, able to amend and reject legislation but without the ability to block money bills or to delay other public bills beyond a maximum of two parliamentary sessions (one session since the 1949 Parliament Act).3 Non-controversial public bills and private bills can also be initiated there. It was constituted of hereditary peers and, since the 1958 Life Peerages Act, peers ennobled by the Queen on the advice of the Prime Minister could sit, speak and vote in the Lords but could not pass their peerages on to their heirs. The hereditary nature of the Lords gave the Conservatives an inbuilt majority. The 1999 House of Lords Act removed the right of hereditary peers to sit and vote in the Lords with the exception of 92 of them (elected by their fellows from the 759 hereditary peers in 1999) as an interim measure to full reform.4 From 1999 the total membership of the Lords was reduced from 1,295 to 695. Following an election the outgoing Prime Minister will have an audience with the monarch. If the Prime Minister has lost her majority in parliament she will resign and the monarch will ask another member of parliament whether she can secure enough support in parliament to form a government.5 There is no automatic vote of confidence in an incoming government. All departmental responsibilities are covered by ministers (usually junior or ministers of cabinet rank – see below) in the Lords as well as a Leader in the Lords to speak on behalf of the government (prime minister). The opposition parties also appoint shadow spokesmen (a shadow cabinet) to cover all areas of responsibility. Only the leaders of the opposition in

Choosing, moving and resigning, UK 59 both houses and the Chief Whips receive state funding (beyond their parliamentary salaries). These days, with a strong party system, the monarch will ask the leader of the majority party to form a government. Failing that she will ask the leader of the party with the most MPs. As with much else in the British Constitution, the monarch invites the majority party leader by convention, rather than constitutional provision. Outside of some parliamentary rules governing payments for expenses, there is no mention of political parties in the British Constitution. However, throughout the twentieth century there was a strong party system and de facto the structures and rules of the political parties determined whom the monarch should ask to form a government on the resignation or death of the Prime Minister. Very rarely in the UK has there been an occasion where there is not a majority party. At the election of 28 February 1974 Edward Heath’s Conservatives gained a plurality of the votes but fewer seats than Labour. After the Unionists from Northern Ireland refused to support Heath and following failed negotiations with the Liberal Party, Heath resigned as Prime Minister on 4 March. Harold Wilson formed a minority government though everyone understood that this would be a temporary or interim government until new elections were called in October 1974, and no major pieces of legislation were attempted. There is no formal investiture of ministers or government and no confidence vote in new ministers or government. Governments can face a motion of no confidence tabled by the opposition, or a Prime Minister can let it be known that a vote on a given bill is a vote of confidence. However, it is known that to lose a vote on some aspects of legislation such as a budget would, in practical terms, be equivalent to a vote of no confidence.

Choosing ministers Ministers, by convention, are also appointed by the monarch, but of course these days they are chosen by the Prime Minister who will announce their appointment through the press office without consulting the monarch.6 There are no rules governing whom a prime minister can choose as a minister, though as they are accountable to parliament they should be drawn from it. In 1964 Wilson named Patrick Gordon Walker as his Foreign Secretary despite Gordon Walker having been defeated at Smethwick. Two Labour MPs stood down so that Gordon Walker and Frank Cousins could enter the House of Commons. However Gordon Walker lost his by-election in January 1965 and had to stand down as Foreign Secretary. A Prime Minister might ennoble someone in order to give them a government post. For example, following the 2005 general election Andrew Adonis, special adviser to Tony Blair, was ennobled allowing him to become Minister of State for Higher Education. Both Houses of Parliament are set up with the government benches facing the opposition benches, and in principle anyone sitting on the government side could be chosen as a minister. In practice, however, the prime minister will only normally draw ministers from her own party except in coalition or national governments.

60 The selection of ministers in Europe We can sensibly draw four categories of people on the government payroll: full cabinet ministers, ministers of cabinet rank, junior ministers and parliamentary private secretaries. In this chapter we also include in the data a fifth category – whips – whose resignations can embarrass the government. The whips’ role is to ensure that party members vote the correct way in any voting division and to negotiate ‘pairing’ – that is, to ensure that any minister who cannot be present for a vote has a pair drawn from the opposition who will promise not to vote either. This enables ministers to miss many votes (including key ones) whilst they go about their governmental duties.7 Only the Chief Whip is on the government payroll and these days tends to be a member of the cabinet, demonstrating how important that role has become. Officially a civil service department is defined in terms of the minister who is held accountable to parliament for that department’s activities (Dowding 1995: 17). Large departments will be headed by a cabinet minister, while smaller and less important departments will be headed by ministers of cabinet rank. Large departments will also have ministers of state taking on specific responsibilities (and these days ministers of state are usually named in terms of the responsibilities they hold). Ministers of state constitute (though not exclusively) what we are referring to as ministers of cabinet rank. In large departments junior ministers are appointed with specific responsibilities; they are sometimes drawn from the Lords in order to give that House a spokesperson for that departmental brief. They will report to their cabinet minister or minister of state. The Prime Minister is responsible for appointing all these posts. She might consult her cabinet colleagues about lower appointments in their department or ministry though that is not necessary. There has been a strong convention in the Labour Party (and a weaker one in the Conservative Party) for an incoming Prime Minister to appoint members who have been shadow ministers – that is, spokesmen for the opposition in the relevant ministerial roles. The whips from both sides of the house are also appointed by the leader of their party, and here the Prime Minister will be strongly advised by the Chief Whip as to whom to appoint as whips. Parliamentary Private Secretaries (PPSs) are also officially appointed by the Prime Minister, but, as these people are closely linked to their minister (the minister’s ‘ears’ in parliament), it is normal for ministers to choose their own PPSs. At a reshuffle ministers will be called to Number 10 (Downing Street (the Prime Minister’s office) and the Prime Minister will thank them and ask them to stand down or to serve in another position. Sometimes these offers, when considered a demotion, are turned down, and sometimes someone offered a post in government will turn it down for the same reason. Some posts are considered difficult. Traditionally the Minister of Defence was a difficult position for Labour ministers in a party that once had a large pacifist component. A number of factors come into play when the Prime Minister chooses her ministers. Obviously she wants a talented and efficient cabinet. Quite what a talented and efficient minister means can vary quite a bit however. Ministers need to develop policies, and so those with policy ideas may be welcome, though of course, the Prime Minister does not want to appoint ministers with ideas at variance with her own. Loyalty to the party is important, and voting against the party line

Choosing, moving and resigning, UK 61 is strongly correlated with MPs who have either left office, have had office pass them by or have no desire for office (see Benedetto and Hix 2007). Over the past fifty years, policy-making has become more centralized, with the Prime Minister’s office and the Cabinet Office taking a much larger role in policy-making. Under Gordon Brown as Chancellor of the Exchequer, the Treasury (following a course set by earlier Chancellors) has taken on an important policy role, notably in the areas of social welfare, but also through the line-by-line reading of department budgets, thus enabling it to affect policy formation and implementation throughout the government machine. A minister must also be able to handle civil servants and be on top of his brief. An important aspect of any minister’s job is his ability to handle parliament. All ministers are required to answer questions in parliament about anything within their departmental remit. This occurs both at regular question times but also if issues are raised within parliament, especially when things seem to be going wrong. In recent times, being able to handle the media is also important, though under Blair greater centralized control meant that ministers appeared on the television much less often than in the past. The Prime Minister would ideally like a cabinet of friends, both politically and personally, but some people will have a strong backing of factions within their party and be deemed too important not be appointed. John Major rewarded his friends: Norman Lamont was given the Treasury and Chris Patten the Duchy of Lancaster (an honorary title usually associated with a minister without portfolio or given together with other departmental responsibilities). But he also had to find posts for rivals. Hence Michael Heseltine was given a major policy role as a figure too important to ignore. Peter Lilley, another rival, was made Trade Secretary. Major’s government started out following almost eleven years of Conservative rule and lasted another seven years. His was a very troubled administration with his cabinet and party split over many issues, though most notably their attitude towards the European Union. He also had many ministers forced out through scandal. His problem was that with so many casualties, so many former ministers on his backbenches, he was short of high-quality people both for the full cabinet and the government machine itself.8 Blair’s government was also bedevilled with problems over appointments, as some ministers were seen as Blair’s people and others as owing loyalty to Gordon Brown. Their continuing feud, which arose from Blair standing for the leadership of the Labour Party while Brown stepped aside for his fellow modernizer, was the defining feature of Blair’s entire premiership and lasted until Blair’s final nine months in power, when he reluctantly accepted that Brown would take over. Blair had to perform a delicate balancing act – although the longer he was in power, the less he felt the need to protect Brownites. Out of office parties have tended to have ‘shadow cabinets’, and until recently it was the norm that a newly elected government was formed by ministers taking on the roles they had shadowed. There was no shadow cabinet in the Labour Party until 1955 when Attlee introduced the institution. Prior to that, spokesmen were elected by the Parliamentary Labour Committee. The Conservatives have also had fluid shadow cabinets with some acting as front-bench spokesman without constituting the full shadow cabinet. Heath reduced his shadow team down from

62 The selection of ministers in Europe 72 in 1966, which can be compared to David Cameron’s shadow cabinet of only 16 members in 2006. The size of the full cabinet has not varied over the fifty or so years from 1945 to 2007, and has ranged from 18 to 23 members. But the size of the government as a whole leapt from 81 in 1950 to 111 in 2005 (see Table 4.1).9 Governments have grown as state business has increased, but slight variations in cabinet size are partly due to the personality of the PM and in part due to political exigencies. Attlee’s cabinet was large following the overblown size of the wartime cabinets. Heath wanted a small cabinet and created larger departments, toying with the idea of super-ministries for efficiency. In fact he had a cabinet larger than he ideally desired, feeling he had to give posts to certain influential people. Prime Ministers have continually tinkered with departments in responding to circumstances. Churchill in 1951 introduced ‘overlords’ who oversaw departments, something which had been successful in his wartime cabinet but signally failed in peacetime. Some reorganizations are purely political: John Prescott was made Deputy Prime Minister in 1997 and given an enormous department in his role as Minister and First Secretary of State for Environment, Transport and the Regions. Blair bowed to Prescott’s demands given how important Prescott was in taking the Labour Party along with the modernizers (Seldon 2005). The size of the government payroll overall has increased by about 37 per cent since 1945 – see Table 4.1.

Moving ministers In recent years, certainly since the time of Margaret Thatcher, all government posts have been announced quickly both at the beginning of a government or during a cabinet ‘reshuffle’ (see below), rarely taking longer than two days to announce. Table 4.1 Size of cabinets and governments in Britain, 1900–2005 (as at 1 January)

Year

Cabinet ministers

Non-cabinet ministers

Junior ministers

No of MPs in paid govt posts

No of peers in paid govt posts

Total paid in govt posts

1900 1910 1917 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005

19 19 5 19 19 9 18 19 21 22 22 22 21

10 7 33 15 9 25 20 20 33 38 33 34 36

31 36 47 47 30 40 43 43 48 47 48 49 52

33 43 60 58 50 58 68 65 85 86 80 82 91

27 19 25 23 8 16 13 17 17 21 22 24 20

60 62 85 81 58 74 81 82 102 107 102 106 111

Source: Butler and Butler, 2000: 71; Butler and Butler, 2006: 26

Choosing, moving and resigning, UK 63 Throughout the 1940s and into the late 1950s such reshuffles were more leisurely affairs, sometimes taking as long as a month for all posts to be filled. The Prime Minister may move ministers around at any time, promote or demote, fire and bring in new faces as political and other circumstances allow. These ‘shuffles’ occur reasonably frequently in British government. There is no official definition of a shuffle. If a minister is forced to resign in circumstances of scandal, ill-health or whatever, then the Prime Minister may take the opportunity to make wider changes than simply replacing the minister. Political circumstances may mean that personnel may be moved up the informal hierarchy. Whilst there is no official recognition of a hierarchy of ministers, the four great offices of state – Prime Minister, Home Secretary, Foreign Secretary and Chancellor – have been the most important posts, together with Deputy Prime Minister, a recent innovation that perhaps keeps the politically important mollified whilst keeping them away from central government business. Following implementation disasters, several functions have been taken away from the Home Office. Notably a Ministry of Justice has been set up. The civil service produces a list of ministers; currently the full list of ministers is found on the Number 10 website under the ‘Guide to Government’ section. Previously ministerial lists were printed four or five times a year in the HMSO (Her Majesty’s Stationery Office, now renamed the Office of Public Sector Information) publication H.M. Ministers and Heads of Public Departments, as well fortnightly in Hansard (printed transcripts of parliamentary debates) during parliamentary sessions. The first Minister of Justice was Jack Straw, a heavy hitter having headed up both the Home Office and the Foreign Office in the past. Perhaps it is not insignificant that Jacqui Smith became the first female (excepting Margaret Thatcher as PM) to take one of the more important cabinet posts after the Home Office lost its pre-eminent position. Jack Straw as Secretary of State for Justice and Lord Chancellor was listed above Jacqui Smith as Home Secretary on the website at 10 Downing Street (at fourth and fifth positions respectively) upon appointment. The first woman to be appointed to a ministerial post was Margaret Bondfield, first serving as Private Secretary in the 1924 Labour government, returning as a full cabinet minister in the Labour government of 1929–31. The next full female cabinet minister was Ellen Wilkinson in the next Labour government serving from 1945 to 1947. The appointment of women as full cabinet minsters has been sporadic and, discounting Margaret Thatcher (Prime Minister through the 1980s), it is not until the 1990s that we have a continuous run of cabinets with at least one female minister (with two, Virginia Bottomly and Gillian Shepherd) from 1992 to 1997. The Blair government promoted more women, with five in his cabinet (about 25 per cent). Gordon Brown’s first cabinet had six women (out of 23) plus four out of five ‘ministers also attending cabinet’ (who can be considered important ‘ministers of cabinet rank’). Examination of ministerial careers does not show a standard pattern. Being in government is not like other jobs, where a normal career trajectory might see one start at the bottom and move up, with some high-flying executives flown in from similar positions in similar organizations. Some ministers start at the very top. They will have served in parliament and will have led their parties and the opposition,

64 The selection of ministers in Europe but their first job as minister is one of the top ones. Tony Blair had never served in any government prior to taking office on 2 May 1997, and only one cabinet member had served in a previous government (Jack Cunningham served as Undersecretary of State under Callaghan). Blair named three cabinet ministers who had served in the Lords. Of course, many ministers start in junior positions and move up depending on how long their party has been in power. ‘Enforced’ cabinet shuffles (or reshuffles as they are often called) usually involve relatively few personnel. Some prime ministers, perhaps because of personality or circumstance, shuffle regularly. Wilson’s first term of office (1964–70) was thought to be particularly fluid: Richard Crossman (Crossman 1975: 78) felt that the constant ‘fiddling’ had ‘undermined the central strategy of government’. Our data suggest that Wilson’s 1966–70 government was shuffled a lot. Margaret Thatcher’s second and third terms also demonstrated great fluidity, though not more than Attlee’s or Macmillan’s first administrations (see the graphs in Figure 4.1). Figure 4.1 shows the Kaplan-Meier survivor function by prime ministerial term. These pictures sum up ministerial duration over each prime ministerial term.10 They contain in boiled-down form prime ministers’ individual histories, problems and cabinet relations in each government, at least with regard to how these impacted on changes in cabinet composition.11 In many ways these twentytwo pictures look broadly similar, with differences largely explained by the length of the government term. Where the term is very short, for example in 1950–51, 1955, 1963–64, 1974, 1974–76 and 1990–92, then the survivor function has a very different shape to when the prime minister has had a full term of between four and five years. In the former cases the survivor function is relatively constant, indicating a bedding-in period before elections are called and a new government formed. For example Wilson’s 1974 government had only three resignations in the eight month period before the next election. The survivor function for this government is flat, indicating that nearly all ministers survived the term. The three small steps in this function represent the resignations of the three ministers. A similar pattern is evident for Wilson’s government of 1964–66, whereas the survivor function for the Wilson government which formed in 1966 shows a steeper decrease even after the reshuffle at around the ten-month mark. The length of the term varies considerably. Major’s second administration is the only one to go beyond the full five years (61 months), though Attlee’s first administration was also into its fifth year (at 56 months). Other governments went the full five years – Macmillan and Douglas-Home (60 months) – whilst Eden– Macmillan (53 months) and Wilson–Callaghan (55 months) were in their fifth year by the time the election was called. Prime ministers in control such as Thatcher and Blair now choose to go in their fourth year. Qualitative histories give the detail of cabinet relations while the survivor function graphs provide only a broad brush stroke. Nevertheless they can reveal aspects that might be missed in more detailed historical accounts. For example, Harold Wilson has the reputation as a shuffler of his cabinet pack, but Figure 4.1 reveals that his reputation is based on his 1966–70 government only. Margaret Thatcher is also a premier with the reputation as someone who lost or fired

5

0.00 0.25 0.50 0.75 1.00

0

0

10

0.00 0.25 0.50 0.75 1.00 40

10 Months

Churchill

Eden

1951–55

1955

20 Months

60

0

5

0.00 0.25 0.50 0.75 1.00

Months

30

40

0

.5

15

1 Months

Eden

Macmillan

1955–57

1957–59

10 Months

15

20

0

10

20 Months

2

30

40

Douglas-Home

1959–63

1963–64

20 30 Months

40

50

0

5

10

20

1.5

Macmillan 0.00 0.25 0.50 0.75 1.00

10

0.00 0.25 0.50 0.75 1.00

0

Attlee 1950–51

20

0.00 0.25 0.50 0.75 1.00

0

Attlee 1945–50

0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00

Choosing, moving and resigning, UK 65

15

Months

Figure 4.1 Ministerial survivor functions by government in Britain, 1945–2007 (continued on pp. 66–7)

10

5

0.00 0.25 0.50 0.75 1.00

0

0

10

10 Months

0.00 0.25 0.50 0.75 1.00

1966–70

15

20

0

10

20

30 Months

Wilson 1974

0.00 0.25 0.50 0.75 1.00

Heath 1970–74

20 30 Months

0.00 0.25 0.50 0.75 1.00

0

Wilson

1964–66

40

50

0

2

4 Months

Wilson

Callaghan

1974–76

1976–79

10 Months

0.00 0.25 0.50 0.75 1.00

5

0.00 0.25 0.50 0.75 1.00

0

Wilson

15

20

0

10

20 Months

Thatcher

Thatcher

1979–83

1983–87

20 30 Months

Figure 4.1 continued

0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00

66 The selection of ministers in Europe

40

50

0

10

20 30 Months

40

50

6

8

30

40

40

50

0

10

Thatcher

Major

1987–90

1990–92

0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00

Choosing, moving and resigning, UK 67

20 Months

30

40

0

5

0

10

Months Months

40 40

60 60

0.00 0.000.25 0.250.50 0.500.75 0.751.00 1.00

20 20

20

1997-2001 1997–2001

0 0

10 10

20 30 20 30 Months Months

Blair

Blair

2001–5

2005–7

20 30 Months

0.00 0.25 0.50 0.75 1.00

0 0

15

Blair Blair

1992-97 1992–97

0.00 0.25 0.50 0.75 1.00

0.000.25 0.250.50 0.500.75 0.751.00 1.00 0.00

Major Major

10 Months

40

50

0

5

10

15 Months

40 40

20

50 50

25

Figure 4.1 continued

ministers. Whilst this seems confirmed in her second and third terms, ministerial durability under her is not significantly different from the Attlee governments of 1945 and 1950 or Macmillan’s first administration (it is his second that saw drastic changes). Another feature picked up by these graphs is the timing of government reshuffles. Officially a person remains in a government post until the incoming prime minister has announced her cabinet. On a change of party this might occur up to two days after the day of the election. In practice ministers are clearing their offices as soon as the result is known. Even after winning an election most prime ministers change their cabinet from what it had been. Some ministers might have stood down as MPs at the election, others told the prime minister they wanted to leave the cabinet, and she might want to remove others to have a clean sheet after the election. We start each graph in Figure 4.1 with government as announced by the prime minister in the few days after the election, so these first ‘reshuffles’

68 The selection of ministers in Europe following an election are not included (the minister’s tenure is censored a few weeks before the election.) However, you will see that some ministers leave government very early on even so, in Attlee’s administration and in Thatcher’s third administration for example. As well as providing a picture of British political history which provides novel insights, the survivor functions illustrate the extent to which political and economic circumstances are reflected in terms of ministerial turnover. For example, the declining survivor functions mid-term for the Heath government (1970–74) and the second Major administration (1992–97) reflect the turmoil of these periods of British political life. The Heath government collapsed under the pressures of massive world oil price increases and industrial unrest. Major won a fourth electoral victory for the Conservatives but Black Wednesday (16 September 1992, where a run on sterling caused the withdrawal of the pound from the European Exchange Rate Mechanism and an increase in interest rates from 10 to 12 per cent) saw the end of the electorate’s belief that the economy was safe under the Conservatives. His administration then seemed to stagger for five years as he desperately tried to keep a disparate and despairing party together with claims of sleaze continually dogging his administration (Denver et al. 1998, King 1998). In comparing the graphs for these administrations we find that, during the Heath government, a quarter of the ministers had lost their jobs within the first three years; during the Major government this benchmark was reached within 28 months of the beginning of the term. More generally, one might suggest that these graphs illustrate the extent to which the Prime Minister is in control of events. A Prime Minister in control of events may initiate a lot of changes in her first few years and then stabilize the ship in the run-up to the election, hence the decrease in the rate at which the survivor function falls at around four years.12 Attlee savaged his first cabinet in the first two years, and Blair lost a quarter of his ministers in the first 30 months in his first administration, as did Thatcher in her second. But of course this has not always been the case. Major certainly hung on for as long as he could till 1997, since the polls were bad and he was hoping that time would heal the economy (or something) in his favour. Callaghan was also in severe difficulties and surprised everyone by not calling an election in the Autumn of 1978. However, there is no general pattern discernible across these 22 pictures. We rather see them as reflecting the particular history of each government, and the particular personality of the prime minister trying to orchestrate her cabinet in light of the problems and challenges she faced. Thus these 22 graphs illustrate the hazards of being a minister within 22 histories all much written about. British commentators have often bemoaned the short span of ministers in office. Our data show that the mean tenure of all ministers is 28 months for Conservatives and 24 for Labour. But British rates of ministerial turnover are rarely compared with those of other countries so it is difficult to get a sense of what an optimal tenure rate would be. Alderman (1995) defends the British practice of relatively quick ministerial turnover, whilst Berlinski et al. (2007) argue that, given that the route to the top is crowded, there are diminishing marginal returns in a ministerial career in the context of one-party rule. Attlee and Blair had the least experienced

Choosing, moving and resigning, UK 69 governments upon taking office, while John Major had the most experienced. Nevertheless commentators have characterized Attlee’s and Blair’s governments as efficient (though turnover was relatively high in both) while Major’s was seen as tired and inefficient.

Resigning When reshuffles occur at the pleasure of the Prime Minister, either after an election or around mid-term, the changes can be made largely as the Prime Minister wishes. She may believe that she needs to enliven the government, at mid-term perhaps to look towards the general election, and after an election for a bright start to a new term of office. Some ministers will be damaged goods, ones who have proved themselves incapable of running a large department, disappointing in policy initiatives or unable to handle parliament. Others may have been involved in controversy – perhaps their departments have bungled or they may have made some personal error, been ridiculed for some policy speech or become embroiled in a personal scandal or problem. Such events can lead to resignation almost immediately when the issue arises, as we discuss below; but ministers often survive such ‘resignation issues’. However, they will have weakened their positions, lowered their political capital and may be quietly shuffled out months or even years later. Other ministers might be moved on as the Prime Minister, who perhaps relied on that person early on in government, is now strong enough to manage without them. Rivals in their own party might require a cabinet post early in a government, but as the Prime Minister secures her position she can shuffle the minister out. Margaret Thatcher was at first viewed as a weak Prime Minister in a hostile cabinet (though she ensured that all of the economic positions were held by those who shared her views). The so-called ‘doves’ (those who followed the postwar interventionist and social conservatism rather than hard-line monetarism) were culled in the reshuffle of 1981, and then culled again when she was returned to power in 1983. Following Carrington’s resignation after the Falklands invasion in 1981, Thatcher put in dove Francis Pym as Foreign Secretary as a steady hand, but removed him from the cabinet when she named her cabinet following the 1983 general election with her much increased majority. The big problems for prime ministers come with forced shuffles and resignations. We can make a distinction between ministers who are shuffled out quietly and those who are forced to resign because of some current event. Forced resignations often entail a reshaping of the cabinet or a mini-reshuffle. There are two aspects of ministerial responsibility: individual ministerial responsibility and collective cabinet responsibility. The first relates to what the minister is responsible for in terms of his departmental responsibilities. It can be simply defined as those aspects of government business for which he may answer questions in parliament. (The Prime Minister might be asked anything of course, which demonstrates their overall responsibility for all government business.) A minister is also responsible for his own personal conduct. Collective cabinet responsibility means that every cabinet minister is responsible for the conduct of the cabinet as a whole; that is, if you sign up to be a minister then you are duty bound not to criticize any aspect

70 The selection of ministers in Europe of government policy. Collective responsibility has its origins in the eighteenth century through the practice of collusion between ministers in the advice they gave to the sovereign. Such collusive practices acted as a check on the sovereign, allowing the Prime Minister and the cabinet to develop and pursue specific policy agendas. Moreover, it provided protection for individual ministers, limiting the monarch’s ability to single them out for blame. By the nineteenth century Dicey had identified a convention of collective ministerial responsibility which involved unanimity in the advice given to the Crown. It developed further under Lord Salisbury, who set out the formulation of joint or collective responsibility in a speech to parliament: For all that passes in Cabinet every member of it who does not resign is absolutely and irretrievably responsible and has no right afterwards to say that he agreed in one case to a compromise, while in another he was persuaded by his colleagues . . . It is only on the principle that absolute responsibility is undertaken by every member of the Cabinet, who, after a decision is arrived at, remains a member of it, that the joint responsibility of Ministers to Parliament can be upheld and one of the most essential principles of Parliamentary responsibility established. (Hansard, vol. 239, cols 833–4) Salisbury instantiated this understanding of collective responsibility in order to stop what today would be described as ministers briefing against each other, though not through friends to newspapers as such, but rather to their dinner party guests in their grand houses. Such collective responsibility also implies solidarity with one’s fellow ministers. Because each minister takes responsibility for the policies of the government as a whole and so is expected to defend their colleagues, all receive at least some protection should a problem arise. As Turpin (1993: 58) argues: ‘Collective responsibility provides a shield for the individual minister, only rarely so emphatically withdrawn as to leave the minister no alternative but resignation.’ It also means that some ministers resign when they find they cannot bring themselves to defend a policy or vote for it in the parliament. There have been 68 resignations due to policy disagreement from 1945 to 2007, some of the notable ones including Short and Cook over Iraq in 2003 and Thorneycroft over economic policy in 1958. Junior ministers often find it easier to resign over policy disagreement than senior ones.13 A senior minister leaving over an issue of collective responsibility is essentially ending his ministerial career. One notable exception was Michael Heseltine, who resigned over the Westland Affair in 1986 when he believed that the cabinet committee minutes were inaccurate (Hennessy 1986).14 He remained on the backbenches until, after he lost the contest for the leadership of the Conservative Party to John Major in November 1990, Major brought him back into the cabinet as Environment Secretary and demonstrated his importance by making him Deputy Prime Minister in a reshuffle in July 1995. Younger ministers might calculate that it is optimal to quit a government with unpopular policies, as this may enhance their chances of achieving high office in a later government, particularly

Choosing, moving and resigning, UK 71 if the issue they choose is unpopular with their party. The most obvious example is when Harold Wilson resigned over policy as a junior minister under Attlee, only to return as leader of his party and gaining the premiership thirteen years later. Resignation over failure in one’s departments – that is, departmental error – is rare. In our dataset we have coded only four ministers who have resigned and these involve only two cases: the Crichel Down affair in 1953 (over the sale of land compulsorily purchased during the war) when Sir Thomas Dugdale resigned, and the invasion of the Falklands Islands (when the UK was unprepared for the Argentine action) when Lord Carrington resigned along with two of his foreign ministers, Minister of State Richard Luce and Lord Privy Seal Humphrey Atkins. Even these cases are contested. The Crichel Down affair is a complicated story, and quite why Dugdale resigned is questionable (Wheare 1975); whilst Carrington went in order to save the Prime Minister (see also Dowding and Kang 1998). Without either Defence or Foreign Minister going, Margaret Thatcher would undoubtedly have had to resign. Carrington fell on his sword, though how far the Foreign Office was to blame is again questionable.15 The cardinal sin for ministers is to be found to have misled parliament: a minister who makes a statement to the House that he knows to be wrong, or should have been briefed not to have made, will except in the most unusual of circumstances have to resign. We code these instances as ‘personal error’. Few ministers have resigned when accused of a personal error – only one in five since 1945 – and the range of offences vary greatly in scope and severity. In 2001 Peter Mandelson left Cabinet after allegedly lying about his role in helping an Indian businessman and Dome donor, Srichand Hinduja, to obtain a British passport (Mandelson was later cleared of all charges). In 2004 David Blunkett resigned as Home Secretary amid allegations that he misused his position to fast-track his ex-lover’s nanny’s visa application. Also coded as ‘personal errors’ are cases where errors of judgement are attributable directly to the minister. Hence we count as a personal error Edwina Currie’s resignation as junior health minister following her statement that most of Britain’s egg production was contaminated with salmonella.16 Another form of personal error is one we code as ‘performance’ – that is, where ministers are judged to have made mistakes or to have been ineffective in discharging their portfolio responsibilities. Our data show that since 1945 only four ministers have resigned because of performance, which is usually the result of a series of purported errors culminating in a resignation: for example Norman Lamont’s 1993 exit as Chancellor of the Exchequer two months after Black Wednesday. Personal errors also include sexual scandals and financial scandals, together commonly referred to as ‘sleaze’. The former were until recently exclusively linked to Conservative governments and always used to lead to resignations. However, since the mid-1990s some calls for resignation over personal affairs have not led to resignation. For example, Robin Cook did not resign when his affair with his secretary leaked out, nor did David Mellor after reports of his affair – though in Mellor’s case the press pursued him on other matters until he eventually resigned. Financial scandals can lead to resignation if they are deemed serious even if unproven, as can be seen with Reginald Maudling, but not if considered mild, such as with John Prescott.17

72 The selection of ministers in Europe Table 4.2 shows the number of individual resignations that have occurred in these categories since 1945. It also shows the number of ‘non-resignations’. A nonresignation is coded where the minister is asked to resign, or to ‘consider his position’ or some similar phrase used in the House or said by some reputable commentator following an event.18 Looking at the weight of non-resignations visà-vis resignations for reasons other than reshuffle, retirement or death, we can see the protection that ministers are given over these controversies. Table 4.3 shows how individual administrations have dealt with calls for resignation and gives an indication of how prime ministers in control behave. Three trends emerge. First, prime ministers in control tend to support and retain ministers accused of scandal. Blair, Thatcher, Churchill and Wilson sacrificed fewer than 30 per cent of ministers subject to public criticism, while Major, Heath and Callaghan sacked around 40 per cent. Second, strong prime ministers do not bend to ‘sleaze’ allegations as much as weak ones, though these cause resignations more than other public disquiet. Blair, Thatcher and Wilson forced resignations on less than half their ministers accused of sexual or financial scandals. By contrast, Major and Heath sacked over 60 per cent. Finally, powerful leaders show loyalty to lowerlevel ministers, dispensing only with those at the higher levels, where damage to the government is greatest. A large chunk of the resignations under Blair, Thatcher and Wilson were attributable to cabinet ministers. Meanwhile, only 12 per cent of those who resigned under Major were cabinet ministers. Figure 4.2 shows the annual count of resignations and non-resignations since 1945 smoothed over a five-year interval. We can see clearly that the number of calls for resignation has gone up over time. There was a sharp increase in the 1960s starting with scandals in the Conservative government increasing during Wilson’s second government in 1966; the graph then falls slightly before increasing again throughout the 1980s, 1990s and into the new century, tailing off towards the end of the Blair government. Whilst not included in our data set, Gordon Brown, after a honeymoon period, has also found his cabinet colleagues facing resignation calls over alleged incompetence and other scandals. We control the calls for the size of the government payroll, but that does not make a great deal of difference. Table 4.2 Resignations and non-resignations in British governments from Attlee to Blair, 1945–2007 Reason

Resignation

Non-resignation

Total

Personal error Departmental error Sexual scandal Financial scandal Policy disagreement Personality clash Performance Other controversy Retirement/other appointment/reshuffle/death

13 4 12 12 68 3 4 7 468

64 52 8 23 33 1 69 23 0

77 56 20 35 101 4 73 30 468

Total

591

273

864

Resignation reasons Personal error Departmental error Sexual scandal Financial scandal Policy disagreement Personality clash Performance Other controversy Total resignations Non-resignation reasons Personal error Departmental error Sexual scandal Financial scandal Policy disagreement Personality clash Performance Other controversy Total nonresignations Resignation as % of calls Sexual scandals Financial scandals ‘Sleaze’ resignations Cabinet minister resignations

0 1 0 0 0 0 0 0 1

1 2 0 0 1 0 1 1

6

14 0 0 0

1

2 3 0 0 1 0 0 5

11

42 0 1 1

3

0

40 0 0 0

3

1 0 0 0 0 0 1 1

0 0 0 0 2 0 0 0 2

2

29 2 0 2

20

8 4 0 0 0 0 4 4

2 0 2 0 4 0 0 0 8

2

50 0 0 0

2

1 1 0 0 0 0 0 0

0 0 0 0 0 2 0 0 2

Churchill III Eden Macmillan Douglas-Home

1 0 0 1 6 0 0 0 8

Attlee

6

29 0 0 0

25

4 2 0 0 9 0 9 1

1 0 0 0 6 1 0 2 10

Wilson I

Table 4.3 Ministerial resignations by government in Britain, from Attlee to Blair, 1945–2007

2

46 2 1 3

7

1 1 0 0 0 0 4 1

0 0 2 1 2 0 0 1 6

Heath

0

29 0 3 1

12

1 1 0 2 5 0 3 0

0 0 0 1 4 0 0 0 5

Wilson II

1

36 0 0 0

7

1 0 0 0 0 0 5 1

0 0 0 0 4 0 0 0 4

10

29 2 0 1

39

8 11 1 0 11 1 7 0

3 3 1 0 7 0 1 1 16

Callaghan Thatcher

3

37 9 9 11

43

13 12 3 4 2 0 9 0

3 0 6 5 8 0 1 2 25

10

29 5 21 5

88

23 15 4 17 4 0 16 9

3 0 1 4 25 0 2 1 36

Major Blair

–

5

10

15

20

25

30

35

40

45

50

1945–49 1950–54 1955–59 1960–64

Calls for resignation

Five years’ period

1970–74 1975–79 1980–84 1985–89 1990–94 1995–99

Resignations

1965–6

2000–4

2005–7

Figure 4.2 Resignations (lower line) versus calls for resignation (upper line) in Britain, excluding reshuffles and adjusted for size of government payroll, from Attlee to Blair, 1945–2007

Count

Choosing, moving and resigning, UK 75 Figure 4.2 should not be interpreted as suggesting that ministerial standards have fallen, rather that scrutiny through the mass media has increased. Many personal matters are now reported that would have proceeded without comment in the past; and scrutiny of government policy and potential failure has increased enormously.

Conclusions In comparison with many other nations, in the UK the Prime Minister is much less constrained in choosing, moving and sacking ministers. Whilst she does have to take account of intra-party conflicts, she does not – at least not during the sixty years considered here – have to take any account of coalitional politics. For that reason the accountability of ministers in terms of their efficiency is perhaps greater or at least more direct than in other countries. Of course we have to understand what efficiency means here. Whilst the Prime Minister has to take account of powerful figures within her party, any minister whose actions are thought to damage the government, as judged by the public (or at least the mass media) faces the danger of being moved or sacked. Ministers who enhance government popularity are likely to be supported. Dewan and Dowding (2005) have shown that forced resignations tend to enhance the popularity of the government, relative to what its reputation would have been had such resignations not taken place. Scandals damage government popularity, but resignations following them can more than correct for that effect. Resignations due to policy disagreement are exempt from the corrective effect, possibly because policy disagreement signals dissension in government which deleteriously affects its popularity. But resignations following other kinds of problems signal that the Prime Minister is holding ministers to account and is on top of her job. Prime ministers in control shuffle their cabinets regularly, partly tinkering and partly responding to events. A prime minister who has just won an election is able to remodel her machine. However, we can see that events can spiral out of control especially once a party has been in power for a long time. Newly elected governments seem to have a period of grace where calls for resignation are lower than later on in their term, although this effect is lost for re-elected governments. In recent years, scrutiny of government has been much higher than it used to be, and calls for ministers (and the prime minister) to resign occur much more frequently than in the past. Prime ministers adopt their own style for dealing with their cabinets, so that ministerial turnover differs from one administration to another. However, as we have shown in this chapter, many aspects of hiring, moving and firing ministers remain the same across the decades.

Notes 1 The data used in this chapter have been compiled from various sources including Butler and Butler (2000, 2006), The Times newspaper online and on microfiche, and the online database Nexis. These data have been compiled by Joseph Angolano, Françoise Boucek, Chloe Cangardel, Helen Cannon, Norman Cooke, Won-Taek Kang and Gita Subrahmanyam, overseen by Keith Dowding. We are grateful to David Butler

76 The selection of ministers in Europe

2 3 4

5 6

7 8

9

10

11 12

13 14

for sending us information on ministerial movements subsequent to his published data. The precise number has varied. The lowest number in the postwar period is 625 in the elections of 1950 and 1951, and the highest 659 in 1997 and 2001. There were 646 MPs elected in May 2005. A public bill is one which is in the general public interest and usually introduced into parliament by a minister, though backbenchers can introduce public bills. Private bills refer to some specific individual or organization. Some bills are hybrid. Further reform has never taken place; see Judge 2005: 64–77; Shell 2000; McLean et al. 2003 for an account of the 1999 reform and the failure to take this interim reform further. For a more complete account of the duties and abilities of the Lords consult Walters 2003. Following principal–agent convention we will use the female pronoun to refer to the prime minister (and monarch) and male pronoun for other ministers. Senior ministers and senior members of the opposition parties are sworn in as Privy Councillors at an archaic ceremony presided over by the Queen. Unless removed for dishonourable behaviour, once sworn in as a Privy Councillor one remains so. As a Privy Councillor one might be consulted or informed about issues of state that should not be used for party gain, such as information about Britain’s independent nuclear deterrent or about terrorist activity. Ordinary MPs also try to pair up so they can miss votes when they have duties elsewhere. Tristen Garel-Jones, a whip, scanned a list of 15 potential candidates for a junior ministerial post in the 1980s and thought that he would not employ any of them. ‘You have to find maybe ninety people to form a government. You have perhaps 350 people to choose from. Once you’ve eliminated the bad, mad, drunk and over-the-hill, you’ve got rid of a hundred. You then have to pick ninety people out of a pool of 250’ (cited Paxman 2003: 209–210). In Major’s time the majority was smaller, and the number of ex-ministers larger so that the pool to choose from was much smaller. The full cabinet has been as small as five in 1917 and nine in 1940, with size of the government at 85 (a wartime government) and 74 respectively (Butler and Butler 2000: 71; 2006: 26). At the beginning of the twentieth century the cabinet was about the same size as today, with 19 members but with 60 in the government as a whole. We analyse the length of time that elapses from when a minister enters government until he leaves or the government terminates. We treat the end of a government term as occurring either when there is an election or when there is a change of prime minister. We treat the starting day for each minister as occurring two weeks from the day the government is formed thus allowing for a period during which the prime minister might shuffle cabinet. Similarly, we censor everybody two weeks before the end of government just to avoid problems generated by coding errors at the end of governments. See Thomas 1998: chs 1–2 for Atlee to Major, or Hennessy 2000 and Leonard 2005 for Atlee to Blair (first government). Temple (2006) covers Blair until 18 months before he resigned. Whilst in many ways Blair appeared in control of his government, the number of scandals, individual resignations and reshuffles made his administration seem out of control. However, the poor state of the opposition and his personal popularity until the Iraq war explains more about his character as Prime Minister than does the perception of his government as flooded by resignations. Only 17 full cabinet members have resigned over policy, while 25 junior ministers and 19 PPSs have left because of a policy disagreement. The Westland affair concerned differences of opinion about the relative merits of a US or European takeover of a helicopter company. Heseltine resigned when he claimed that the minutes of a cabinet meeting were not accurate; later the Trade and Industry

Choosing, moving and resigning, UK 77

15 16 17

18

Secretary Leon Brittan also resigned when it emerged he had misled parliament over the leaking of a letter about the affair from the Attorney General. Interestingly Carrington, who was a junior minister, also offered to resign over Crichel Down. It might be that Carrington is the only politician who actually believes in the convention that ministers resign over departmental error. She actually resigned because she made a statement that was technically correct, but was taken to mean that most of Britain’s eggs (rather than egg production) were contaminated with salmonella. Maudling had been named as chairman of an associated company by John Poulson during his bankruptcy hearing. When the metropolitan police, for whom Maudling as Home Secretary had responsibility, started to investigate Poulson, Maudling resigned, though there is no evidence Maudling had done anything wrong. Prescott had contravened ministerial code, meeting with a US billionaire to discuss casinos and the London Dome and failed to report the trip in the register of MPs’ interests. For full coding details see http://polsc.anu.edu.au/staff/dowding/datasets.htm.

References Alderman, R. K. (1995) ‘A Defence of Frequent Ministerial Turnover’. Public Administration 73: 497–512. Alderman, R. K. and Carter, N. (1992) ‘The Logistics of Ministerial Reshuffles’ Public Administration 70: 519–534. Benedetto, G. and Hix, S. (2007) ‘The Dejected, the Rejected and the Ejected: Explaining Government Rebels in the 2001 to 2005 British House of Commons’. Comparative Political Studies 40: 755–781. Berlinski, S., Dewan, T. and Dowding, K. (2007) ‘The Length of Ministerial Tenure in the United Kingdon, 1945–97’. British Journal of Political Science 37, 2: 245–262. Butler, D. and Butler, G. (2000) Twentieth-century British Political Facts 1900–2000. Houndmills: Macmillan. Butler, D. and Butler, G. (2006) British Political Facts Since 1979. Houndmills: Macmillan. Crossman, R. (1975) Diaries of a Cabinet Minister. London: Cape. Denver, D., Fisher, J., Cowley, P. and Pattie, C. (eds) (1998) British Elections and Parties Review: The 1997 General Election. London: Frank Cass. Dewan, T. and Dowding, K. (2005) ‘The Corrective Effect of Ministerial Resignations on Government Popularity’. American Journal of Political Science 49, 1: 46–56. Dowding, K. (1995) The Civil Service. London: Routledge. Dowding, K. and Kang, W.-T. (1998) ‘Ministerial Resignations 1945–97’. Public Administration 76, 3: 411–429. Hennessy, P. (1986) ‘The Westland Affair’. Journal of Law and Society 13: 423–432 Hennessy, P. (2000) The Prime Minister: The Office and Its Holders Since 1945. London: Penguin. Judge, D. (2005) Political Institutions in the United Kingdom. Oxford: Oxford University Press. King, A. (ed.) (1998) New Labour Triumphs: Britain at the Polls, New Jersey: Chatham House. Leonard, D. (2005) A Century of Prime Ministers: Salisbury to Blair. London: Palgrave. McLean, I., Spirling, A. and Russell, M. (2003) ‘None of the Above: The UK House of Commons Votes on Reforming the House of Lords, 2003’. Political Quarterly 74: 298–310. Paxman, J. (2003) The Political Animal. Harmondsworth: Penguin.

78 The selection of ministers in Europe Seldon, A. (2005) Blair. London: Free Press. Shell, D. (2000) ‘Labour and the House of Lords: A Case Study in Constitutional Reform’. Parliamentary Affairs 55: 290–310. Temple, M. (2006) Tony Blair. London: Haus Publishing Ltd. Theakston, K. (1987) Junior Ministers in British Government. Oxford: Basil Blackwell. Thomas, G. P. (1998). Prime Minister and Cabinet Today. Manchester: Manchester University Press. Turpin, C. (1993) British Government and the Construction. Chicago: Northwestern University Press. Walters, R. (2003) ‘House of Lords’. In Vernon Bogdanor (ed.) The British Constitution in the Twentieth Century. Oxford: Oxford University Press, 189–236. Wheare, K. C. (1975) ‘Crichel Down Revisited’. Political Studies 26: 390–408.

5

Italy The difficult road towards a more effective process of ministerial selection Luca Verzichelli

Introdution For more than forty years, republican Italy was constantly cited as a singular case of parliamentary democracy, where the patent instability and weakness of a headless collegial cabinet is balanced by a remarkable continuity in the electoral behaviour and stable political elite. Then in 1992 came a sudden, profound crisis causing a revolution at the core of the political system. The explosion of the previous party system and the almost complete disappearance of the ruling class came together with demands for a more accountable and effective system of government. Among other things, the claims concerned a new constitutional setting; more transparent criteria for ministerial recruitment; more talented and skilled people to form the ruling class; and, above all, greater power in the hands of the chief executive in order to provide a constant evaluation of ministerial behaviours and performance. Some fifteen years after the crisis, the degree of political change one can measure in the process of government formation, keeping the period 1948–92 as point of reference, is rather disappointing: the basic constitutional rules have remained unaltered and the political fragmentation within the executive is still quite high, although it is a fact that several interesting new features have emerged in the Italian system of government. Indeed, recent analyses have pointed out that the executive institution in Italy has been at the core of a relevant process of transformation, stressing three major changes.1 First, the emergence of a preelectoral mode of government formation (De Winter 1995), not only in terms of formal electoral cartel and coalition agreements on policy objectives but also with the explicit indication of the candidature of a coalition leader as prospective chief executive. Secondly, a significant increase in the average duration of governments, which has moved from 324 days for the period 1948–92 to 600 days for the period 1996–2008. Thirdly, a stronger role for the President of the Council of Ministers in conducting coalition negotiations on governmental policy, in the distribution of portfolios and competences between parties and between the single ministerial aspirants, as well as during the effective management of cabinet activities. Even though since 1992 the figure of the President of the Council of Minister (from now on PM) has been continuously referred to as a premier, the holders of this office have had a considerably different political influence and surely a different degree of electoral legitimation.2

80 The selection of ministers in Europe The question at the core of this chapter is to what extent has the transition to another model of democracy (what the journalistic jargon has defined the passage to the Second Republic) impacted on the processes of government formation and ministerial selection? Conversely, did the recent uncertainties (among them, the return of the proportional rule in 2006, after three legislatures elected by means of a mixed-majoritarian system, and the difficulties of the Prodi II cabinet, formed by nine parties and defeated by a no-confidence vote in January 2008) completely frustrate the expectation of a more efficient and accountable method of ministerial selection? 3 This chapter will try to answer these questions, focusing mainly on the differences between the period of the First Republic and the last two decades. The choice of such a diachronic strategy of comparison, due to the relevance of the political changes, imposes a specific focus on the most recent events and, consequently, a less attentive analysis of the processes of government formation in historical prospective.

Constitutional framework Italy is a stereotypical example of ‘positive’ parliamentary democracy (Bergman et al. 2003), since, according to the constitution (Article 94), the government must obtain the explicit confidence of both the parliamentary branches, which share equivalent electoral rules and similar legislative and non-legislative powers (‘symmetrical’ parliamentary system), following a procedure strictly regulated by the primary principles of Section III of the 1948 Constitution, a section which has been never significantly amended.4 With regard to the appointment of the PM and ministerial selection the Constitution is however laconic: ‘The President of the Republic appoints the PM and the ministers who are proposed by him’ (Art. 92). In practice, the government formation nevertheless follows a much more complex procedure which can be summarized in three phases: (1) a consultative phase under the direction of the Head of State; (2) the appointment of a prospective PM until his formal acceptation; (3) ministerial selection and preparation of governmental agenda. The consultative phase is aimed mainly at clarifying the existence of a political majority and a significant convergence on the name (or even a short list of names) of a PM candidate. The President of the Republic is conducting the negotiations, having as counterparts the past presidents of the Republic, the speakers of both the chambers and the representatives of all the parliamentary parties. Once the round of consultations is over, the President convokes the candidate PM at the Quirinale, conferring upon him a specific mandate: if the indication from the consultations is clear and politically viable, the mandate conferred is that of formateur. The holder is then actually already a PM in pectore, with reasonable expectations of success that a first round of contacts with the representatives of the different coalition parties will materialize, allowing him to start working on ministerial selection. Sometimes however attempts to form an executive fail and successive formateurs are appointed, or an informateur (for instance, the Chairman of one of the parliamentary branches) is called to the Quirinale with the explicit aim to take time and look for a solution.

Italy: the difficult road 81 When a solution over coalition formula and policy (the latter aspect does not, in general, take much time as government formation in Italy is largely about office distribution, see Verzichelli and Cotta 2000: 456) is finally found, the phase of ministerial selection begins, since the candidate PM has to go back to the President with a list of ministers, to take the formal oath before going to the parliament and to present the policy platform of the new executive. There are no time constraints for the choice of a PM and for the formal acceptance of the appointment of the ministers; a great deal of flexibility therefore characterizes the practice of government formation. In the past, the selection of ministers took an average of one month, while after 1994 this phase has been reduced to one or two weeks. In any case, the pre-parliamentary phases of government formation have to be concluded with the enactment of three decrees from the President of the Republic: (1) the formal decree of resignation of the former PM and ministers; (2) the decree stating the appointment of the new PM, who also signs the decree to testify his acceptation; (3) the decree stating the appointment of each single minister and also, if necessary, the appointment of each deputy-minister (vice-ministro), a figure introduced by a law of 1997. The formalization of all these documents occurs in a solemn ceremony ending with the oath of all the members of the new cabinet. This moment marks the official start of the life of a government and the acceptance of a full responsibility by each minister, although the political legitimation of the executive will be completed only after the inauguration confidence votes of both chambers. Usually, a meeting of the new cabinet takes place before the first of the two parliamentary debates. During this meeting, the list of junior ministers (sottosegretari) is defined and sent to the President of the Republic who has to produce another decree in order to enable the delegated administrative powers of each junior minister. Neither deputy-ministers nor junior ministers are considered in the constitutional provisions but they have been always considered as quasiconstitutional figures: junior ministers were already present in the pre-republican practice and they have been confirmed after the establishment of the new regime. They can stand in for their respective minister in parliament, in official meetings and negotiations and also within the ministry itself. However, they do not participate in cabinet meetings, and their authority depends upon an explicit act of delegation from their minister sanctioned by the cabinet. Concerning the deputy-ministers, according to the 2001 law regulating the matter, they have a full delegation on specific issues and they can participate to the cabinet meetings in order to report on their work, without having formal rights of vote. In fact, all of the spoils of government are extremely important in the context of Italian coalition governance. First of all, junior ministers are often allocated so as to ‘keep tabs on partners’ (Thies 2001), thus creating a sort of mutual control among coalitional actors in the most delicate policy domains. Secondly, the availability of a large number of positions to be allocated immediately after the initial ministerial selection provides the opportunity of an extension of the bargaining set, and therefore to solve problems and dissatisfactions among parties, factions and single political personalities. It therefore takes a collective vote of the cabinet to send the finalized list of junior ministers to the Head of State. In order to put the quasi-constitutional nature of such office in force, the recent amendments to the

82 The selection of ministers in Europe ordinary legislation regulating the functioning of the government introduced a formal oath for the junior ministers, in front of the President of the Republic. As already argued, the basic rules of the Italian positive parliamentarianism developed with the 1948 constitution have been unaltered up to now. This applies not only to the formal procedures of government formation but also to the rules concerning the no-confidence motions that can be presented in parliament against the cabinet or even against a single component of the cabinet. However, the changes in the electoral regime and the emergence of a bipolar party system with large coalitions have introduced fundamental elements of pre-electoral constraints, like the indication of the prospective PM, the pre-electoral formation of the coalitions and also the pre-determination of a team of ministeriabili (potential candidates). Moreover, the law introducing the regulation of the PM office (1988) and the rules reorganizing the whole ministerial structure and the role of PM office, during the 1990s, have reinforced the core executive in Italy by giving a stronger role to the PM and increased the salience of some central portfolios vis-à-vis others in the organization of government structure and work.

Selecting ministers In the traditional picture of the First Republic, ministerial recruitment was strongly impacted by a peculiar application of the party government model of democracy. Scholars are still debating the specificities of such a phenomenon, but there is no doubt that a strong control over the ministerial appointments was always exerted by top-national party elites (Pasquino 1986; Cotta 1988; Cotta and Verzichelli 2003). In their empirical analysis of ministerial recruitment and career during the age of partitocrazia, Calise and Mannheimer (1982) laid down four fundamental characteristics of ministerial selection of this period characterized as the ‘Christian Democratic thirty years’:5 (1) a predominant parliamentary origin of the ministerial elite; (2) the relevance of regional or territorial representation in the selection of ministers, which used to be the second criterion to be taken in account by the nominators, after the typical proportional norm among parties which characterized all coalition governments; (3) the tendency to create, at least among the Christian Democrats, an incompatibility between party leader positions and ministerial offices; (4) the system of incentives produced by the patronage potential of given portfolios. The eminent elite scholar Mattei Dogan, studying the secret of how to become a minister in Italy at the end of the 1980s, reached similar conclusions to those of Calise and Mannheimer. A number of unwritten rules were particularly evident in the recruitment of government ministers among the ranks of Christian Democrats: (1) the members of government were recruited among parliamentarians; (2) one did not make it to government during one’s first parliamentary mandate; (3) one did not become a minister without having previously been junior minister; (4) the position of junior minister was limited to a few years (typically, a legislature); (5) all the regions were represented in government; (6) expertise in a given policy area was not an essential requirement to be appointed in a given position, and competence is supposed to be acquired through political experience; (7) the

Italy: the difficult road 83 composition of government reflected the proportional strength of factions within the parties; (8) the ministerial positions were distributed to parties and factions, which in turn proposed their representatives to government; (9) factions had no right of veto against representatives of other factions; (10) the PM was an arbiter and not a selector; (11) the President of the Republic, despite his formal power, did not choose ministers. As a result, the chances of the aspirants to a ministerial career could be easily determined by a number of variables, including personal electoral success and parliamentary tenure. The first step of such a career was usually to achieve an appointment as a junior minister, which could be followed, usually after one legislative term, by the appointment in a second-rank position of the cabinet. The nature of the Italian model of party government automatically determined a predominance of generalist politicians within the cabinet, and the strategic incentives for the nominators to choose experts or ministers with some alternative profile were always reduced, although a minority of technocratic ministers – wellknown high public officials, economists or independent university professors – had been appointed from time to time, especially in particular offices like Finance, Treasury, Foreign Trade and also in the position of minister without portfolio for the reform of public administration. The poor sensibility of the Italian elite for equal opportunities and the limited presence of women in parliament – the main pool of aspirants for the ministerial career – have hampered the emergence of female representation during the First Republic: the first female minister was appointed only in 1976. Overall, between 1948 and 1992, only five women were admitted to the council of ministers, that is to say, 1.9 per cent of the overall ministerial population. On the whole, during the First Republic, the variability both in the behaviours of the ministerial selectorate and in the strategies of the aspirants was very reduced: once the phase of negotiation between the prospective PM and the parties was over, and therefore the number of coalition players who could be interested to share ministerial payoffs was definite, one could foresee the list of ministrables with a certain degree of certainty, guessing also the types of portfolios to be allocated to the different people. Usually the leaders of coalition parties preferred to monitor the government from outside, leaving ministerial offices to other party representatives – mostly generalist politicians from the parliamentary ranks who had collected a significant personal electoral strength – in proportion to the strength of coalition parties and party factions (but also taking into account the territorial representation). Finally, despite her formal role of nominator and variations in effective power depending on the moment of the political cycle and personal capabilities, the Head of State only had a weak role in the selection (or veto) of ministers during the period of classic partitocrazia. This pattern of ministerial selection following unwritten rules was suddenly challenged in the early 1990s. After the rapid dissolution of the classic partitocrazia, some of the basic ground rules of ministerial recruitment fade away: all the components of the old party system which had served as governmental parties disappeared, and new demands for a more efficient and responsible ruling class

84 The selection of ministers in Europe were, more or less explicitly, associated to claims for more majoritarian-type rules. Moreover, the parliamentary status of ministerial personnel was debated and the continuous dance from one post to another was deprecated as a negative habit of an incompetent and unproductive group of generalist politicians. 6 The data in Table 5.1 clearly show the truth of some of these claims, thereby triggering important changes in government formation and ministerial personnel (Cotta and Verzichelli 2003): comparing average values of all governments which formed in the first ten republican legislatures (1948–92) with those formed after 1992 (and more particularly after 1996, since the transition period was still characterized by a great deal of uncertainty) we see a substantial reduction of the duration of the process of government formation, due to the absolute certainty about the name of the future PM,7 especially when the government follows a legislative election, and about the political composition of the coalition. Also, the number of portfolios and, to a certain extent, the number of cabinet ministers have been reduced, but this has not determined a general reduction of the overall governmental apparatus, given the increase of junior ministers, the emergence of a handful of deputy-ministers and, recently, the divisions of some ministries which have increased again the number of portfolios available. However, the whole process of ministerial selection has not been fully revolutionized. To begin with, the formal procedures of government formation are unchanged. The main innovation is the emergence of a form of electoral selection of the chief executive, thanks to the electoral systems employed after 1992. Consequently, the role of the PM in choosing the ministers has of course increased, although the political conditions are decisive in determining which real opportunity of prime-ministerial government could be achieved from one case to another (Fabbrini 2001). In particular, the two cycles of centre-left governments (1996– 2001 and 2006–2008) have been characterized by a high degree of fragmentation in the coalition, which resulted in strong pressures by leaders of parties of the pre-electoral coalition on the PM. In the case of the cycle of the centre right (2001–6), a much more manifest PM leadership emerged, owing to the more pronounced cohesiveness within this coalition and even to the personal legitimacy of the premier, as the uncontested leader of the largest party of the coalition. Even the traditional prerogative of coalitional party leaders to set the ordering of ministeriabili seems to be somehow weakened. An interesting picture of this game has been provided by Calise in his essays about government transformation during the Second Republic (2005; 2006). According to Calise, a sort of partial presidentialization has characterized the Italian system in these years. In such a system, parties and presidents (the chief executives but, in some respects, even the presidents of the republic, when the legitimation of PM was not so evident, especially during the transition period 1992–96) are bound to fight each other and limit the other’s powers in turn. Executive leaders and party elites appear to play an adversarial role in government formation, but in practice some spoils seem to be clearly in the hands of the PM, while top party leaders try and usually manage to be decisive for other important governmental roles, mainly by themselves occupying top-rank cabinet seats like Deputy-PM or Minister of Foreign Affairs (see below).

32 47 7 48 26 19 12 4 8 9 3 15

1.7 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0

Number of attempts to form the new government 20 21 20 20 20 20 20 20 20 15 15 18

Number of Ministries 25 32 25 27 22 21 26 25 24 26 26 26

Number of ministers 69 59 58 69 59 64 82 91 78 86 98 100

Overall number of government members

8 17 5 21 9 10 6 8 9 5 7

(6.1) (25.0) (50.7) (19.2) (100) (39.1) (34.5) (23.1) (33.3) (30.0) (19.2) (26.9)

Non-parliamentary members (%)

24 24 22 21 22 27 22 20 26 22 23

(53.1) (75.0) (85.7) (84.6) (100) (95.7) (93.1) (84.6) (83.3) (89.7) (84.6) (88.5)

Ministers without experience of junior minister (%)

324 298 259 225 355 874 423 118 382 1409 375 616

Duration (days)

Note: The figures of the number of ministries, ministers, overall number of government members and the percentage of ministers not coming from parliamentary and junior ministerial backgrounds refer to the first day of the life of each government.

Governments 1948–1992 (average values) Amato I (1992) Ciampi (1993) Berlusconi I (1994) Dini (1995) Prodi I (1996) D’Alema I (1998) D’Alema II (1999) Amato II (2000) Berlusconi II (2001) Berlusconi III (2005) Prodi II (2006)

Time of government formation (days)

Table 5.1 Cabinet formation in Italy before 1992 and after 1996: some indicators

86 The selection of ministers in Europe The comparison between the central period of the First Republic and the last fifteen years over strategic incentives for the nominators reveals that the classic unwritten rule of the need for a junior ministerial apprenticeship is no longer at work (Figure 5.1). Moreover, the data clearly show some consequences of the changes, already mentioned, in the political backgrounds of the nominees: generally speaking, a stronger link between policy competencies of ministers and fields of ministerial delegation arises. In particular, the number of expert ministers – technicians or ‘semi-technicians’ (politicians with a strong policy profile) – has recently grown, mainly in portfolios such as Health, Communication, Public Transport, Research and University, or Agriculture. In some cases, a quota of two or three cabinet members has been left to the discretion of the PM to hire a task force of technocratic personalities. Finally, an important portfolio like the super-ministry of the Economy, created in the 1990s by merging three portfolios in the economic and financial area, has been often allocated to an independent personality. All these features have determined a remarkable, though not dramatic, reduction of the typical parliamentary origin of government ministers in Italy (Figure 5.2). Another feature is the greater political weight of the cabinet ministers hired: the PMs who were appointed thanks to their electoral victories, being the uncontested leaders of their own coalition (and, in the case of Berlusconi even the leader of the largest coalitional party), have favoured the appointment in particular governmental positions of other relevant party leaders, in order to maximize the prestige of the cabinet and the expectations of government durability. In fact, many party leaders have recently started their ministerial career immediately at top cabinet positions: among them, Berlusconi (PM in 1994); Prodi (PM in 1996 8), Veltroni (deputy-PM in 1996), D’Alema (PM in 1998), Fini (deputy-PM in 2001) and, in other important ministerial offices, Bossi, Follini, Fassino, Buttiglione and Rutelli. In other words, the whole leadership of the ‘new’ Italian party system has been recruited directly at the top of the government and without leaving their party responsibilities. This is a pretty evident sign of change in comparison to the First Republic. Concerning the rate of female representation within the ministerial elite, Italy has not followed the remarkable growth of some other European countries. However, in the last fifteen years the presence of women has passed from a symbolic quota (one or two out of 25–30 cabinet members) to a higher number: female ministers after 1996 have been indeed always between three and six, bringing the percentage rate to between 10 per cent and 20 per cent. The same trend emerges if we look at the larger group of government members, which includes vice-ministers and junior ministers: the number of female members in the Prodi II government is 20 (20 per cent), which sets a record high in the republican experience. In some sense, we can say that gender has finally become an informal constraint for nominators although the crucial role of such a criterion seems to emerge only when centre-left parties – usually the ones with higher rates of female parliamentarians – are in power. Another change regards the combination of political and parliamentary backgrounds of ministers. Using a simplified version of De Winter’s (1991)

Figure 5.1 Government ministers with experience of junior minister in Italy (%), 1948–2006

0

25

50

75

100

De Gasperi IV De Gasperi V De Gasperi VI De Gasperi VII De Gasperi VIII Pella Fanfani Scelba Segni Zoli Fanfani II Segni II Tambroni Fanfani III Fanfani IV Leone Moro Moro II Moro III Leone II Rumor Rumor II Rumor III Colombo Andreotti Andreotti II Rumor IV Rumor V Moro IV Moro V Andreotti III Andreotti IV Andreotti V Cossiga Cossiga II Forlani Spadolini Spadolini II Fanfani V Craxi Craxi II Fanfani VI Goria De Mita Andreotti VI Andreotti VII Amato Ciampi Berlusconi Dini Prodi D© Alema D© Alema II Amato II Berlusconi II Berlusconi III Prodi II

Figure 5.2 Parliamentary origin of ministers in Italy (%), 1948–2006

0

25

50

75

100

De Gasperi V De Gasperi VI De Gasperi VII De Gasperi VIII Pella Fanfani Scelba Segni Zoli Fanfani II Segni II Tambroni Fanfani III Fanfani IV Leone Moro Moro II Moro III Leone II Rumor Rumor II Rumor III Colombo Andreotti Andreotti II Rumor IV Rumor V Moro IV Moro V Andreotti III Andreotti IV Andreotti V Cossiga Cossiga II Forlani Spadolini Spadolini II Fanfani V Craxi Craxi II Fanfani VI Goria De Mita Andreotti VI Andreotti VII Amato Ciampi Berlusconi Dini Prodi D'Alema D'Alema II Amato II Berlusconi II Berlusconi III Prodi II

Italy: the difficult road 89 25

20

15

10

5

Prodi II

Berlusconi III

Amato II

Berlusconi II

D'Alema II

D'Alema

Dini

Prodi

Ciampi

Berlusconi

Amato

Andreotti VI

Andreotti VII

Goria

De Mita

Craxi II

Fanfani

Craxi

Fanfani V

Spadolini II

Forlani

Spadolini

Cossiga

Cossiga II

0

Figure 5.3 Female ministers in Italy (%), 1980–2006 Note: The percentage of female ministers is calculated at the beginning of each government mandate.

Table 5.2 Insiders, outsiders and other types of political backgrounds among Italian ministers: First and Second Republic compared (%) 1948–92 Outsiders Other types Insiders Total (N)

16.5 30.8 52.7 100.0 (263)

1996–2006 39.0 22.8 38.2 100.0 (138)

Note: Insiders are ministers with at least five years of parliamentary background and experience of national party officer. Outsiders are ministers with no parliamentary background and no party office experience at all.

classification between insiders (ministers with long parliamentary tenure and a noticeable party background) and outsiders (ministers with no party or parliamentary backgrounds), we see in Table 5.2 that the passage from the First to the Second Republic brought a noticeable increase of outsiders, owing to the number of technocratic personnel hired since the mid-1990s, and a substantial decrease of insiders. Overall, the choice of ministers has changed significantly, both in respect of the nature of the process and also in some fundamental characteristics of the aspirants to office. However rather than claiming a new pattern it would be better to say that some traditional specificities have changed. These were due to the particular nature of the old party system – notably the presence of peculiar,

90 The selection of ministers in Europe polycentric and factionalized parties – which have disappeared, creating the conditions for a much more complex and less constrained structure of opportunities for ministerial aspirants.

Reshuffling and de-selecting ministers During the First Republic, the dynamics of shuffling ministers from one portfolio to another and the patterns of ministerial de-selection were quite clear and relatively predictable. Ministerial circulation did not often take the form of midterm reshuffles and the partial modifications of the ministerial team (changes or withdrawal of some particular delegations to one or more ministers) were rather limited (Table 5.3). As a matter of fact, the frequent governmental crises were taken as an opportunity to rediscuss the distribution of roles within the ministerial elite. Broad and often chaotic circulation of a rather stable governmental elite with a rather unstable job profile within the cabinet generally ensued. Between 1948 and 1992, although 12.4 per cent of ministers were not able to remain in charge for at least one year and another 21.3 per cent had a ministerial experience of one year only, the average duration of ministerial experience was three years and eight month, that is no less than four times the average duration of the governments. Moreover, a good 25 per cent of ministers serving during the First Republic remained in office for at least five years, reaching in some cases more than ten years and, in two extreme cases, 26 and 28 years of experience.9 These represent what Calise and Mannheimer defined as the Italian ministerial inner circle: a selected group of eminent politicians who could be employed in a wide range of ministerial offices. Among these offices were of course the premiership, highly visible political departments (Interior, Foreign affairs, Defence, Justice) and other more substantial political payoffs (Education, Public Works, Post and Telecommunication). Even though policy disagreements, or more generally conflicts among coalition partners (including some cases of intra-party conflicts within the Christian Table 5.3 Resignation, reallocation of ministerial tasks and other changes in government composition during the ministerial mandate, First and Second Republic in Italy (1948–1992 compared with 1996–2006) Total number of ministerial Moved appointments ministers (A) (% of A)

Resigned ministers (% of A)

Ministers dead during mandate (% of A)

Latecomer ministers (% of A)

1948–1992

1,081 (mean = 24)

19.0 0.4

40.0 0.9

6.0 0.1

26.0 0.6

1996–2006

174 (mean = 25)

5.0 2.8

9.0 5.2

0.0 –

5.0 2.8

Note: Moved ministers are those ministers who remain in government for all the duration of the cabinet but with different jobs. Latecomer ministers are ministers appointed at a later stage to replace a resigned minister or to cover a newly created office.

Italy: the difficult road 91

60 Mean = 3.69 Std Dev. = 3.701 N = 258 50

40

30

20

10

0 0 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 27 28 29

Figure 5.4 Average duration (in years) of ministerial career in Italy, 1948–1992

Democrats), were very frequent in the First Republic, the limited autonomy of the PM did not allow him to force a single minister to resign. Instead, low-profile compromises were reached within the cabinet: for instance, the promotion or appointment of a politician from the party which had criticized the minister in trouble, or even a partial change in the delegation system, with the circulation of the minister to another portfolio, but without sacking anybody from the cabinet. On the other hand, the discussion on personal conflicts (or policy conflicts generated by the action of some ministers) could also lead to a wider discussion about the duration of the whole executive. In that case, the formal governmental crises were typically used as an instrument of conflict resolution allowing the requests of some of the coalition partners or party factions to be satisfied (on these points see Verzichelli and Cotta 2000). All in all, up to the end of the First Republic a clear lack of delegation control was evident in the ministerial de-selection process, as Italian PMs could not control

92 The selection of ministers in Europe this process from within, owing to their political weakness. On the other hand, the same party leaders who were, indeed, the real nominators, were most of the times unable to sack their own ministers (both because of the vetoes of internal factions and, in some parties, because of the electoral prestige of the minister involved). Ministerial conflicts were solved through policy compromises and through increasing the number of watchdog positions appointed at the governmental level. The literature on the Italian party government has stressed the diachronic transformation that occurred in this respect: while, at the beginning of the republican age, the predominant position of the main party actor and the prestige of a premier-style leader like Alcide de Gasperi were important factors of stabilization of a sufficient delegation between the PM and ‘his’ ministers, with the passing of time the control over such a system of relationships has moved out of the cabinet. This explains the emergence of external arenas for the solution of policy conflicts, namely the use of majority summits (Criscitiello 1993). The final period of the First Republic, from the age of pentapartito to the last governments led by Giulio Andreotti, can be seen, despite the presence of important politicians directly involved in governmental office and despite the attempts to give more formal power to the PM, as a phase of residual party government, with a paralysis of the programmatic actions in the decision making and a less and less legitimated action of promotion of ministerial personnel (Vassallo 1994). What happened after the disappearance of these parties? The increased autonomy of the PM and the more evident pattern of delegation within the cabinet team have some important consequences in terms of stability of the executive and persistency of the ministers in specific ministerial positions. Although we have to consider that we are comparing a long period of about 45 years with a much shorter one, the data on the number of different jobs accepted by ministers are clear: the average number of different jobs taken by those who had at least two ministerial offices in the First Republic was 2.65, and only 25 per cent of ministers kept just one ministership during their career. Since 1996, this rate goes up to 34 per cent, with an average number of different jobs which is now 1.94. We saw earlier that the job profile of most ministers of the Second Republic was much more linked to their policy competences, a characteristic which also leads to credible claims over the same portfolio once a minister is asked again to take part in cabinet. To give just a few examples, Giulio Tremonti served as Minister of Finance in 1994 and he was back to the same office (or better, to a larger ‘Economy’ portfolio which includes also the old finance ministership) in 2001 and in 2005, when he was recalled to government. The same occurred to Pier Luigi Bersani, who served as Minister of Industry, Trade and Craftsmanship in two centre-left governments between 1996 and 1999 and came back as Minister of Economic Development (a new aggregate portfolio which is covering the same policy area) in 2006. However, this does not mean that the process of selection de-selection is now following a totally new shape. Many political conditions are at work and the complexity of the coalition governance in Italy is still decisive in limiting the potential for change. The experiences of two governments like Berlusconi II (2001–5) and Prodi II (2006–8) are good examples of such a situation: Berlusconi

Italy: the difficult road 93 II was unusually able to persist for more than 1,400 days (thus setting a new record of stability for an Italian executive), but it was also characterized by an unusual number of reshuffles and internal changes. However, the PM did not prove to be always strong enough to decide when to fire and to hire. Among the cases of resignations, the Minister of Foreign Affairs, Renato Ruggero, probably left because of policy disagreements with the PM, but the dismissal of the Minister of Economy, Tremonti, was somehow forced by the coalition partners. At the same time, Berlusconi was strong enough to re-impose Tremonti when the technician appointed in his place (Domenico Siniscalco) also decided to leave because of policy disagreements. But the chief executive could not remove troublesome ministers from other parties (like the Northern League minister Calderoli, who caused a serious diplomatic case with the Islamic communities) and he preferred to shift, instead of sacking, the Minister of Interior (Claudio Scajola), whose affirmation concerning a victim of terrorism had provoked turmoil in the press. For Prodi II (Table 5.4), the PM had even less political autonomy than Berlusconi: the list of the ministers was defined after very long and delicate negotiations, and the presence of an extraordinary number of coalition parties (the highest of the whole republican experience) forced the PM to accept a number of compromises. The degree of proportionality in the allocation of the spoils, which is still quite high if one considers the saliency of the different offices, is however less optimal than the easier distribution which had characterized the centre-right governments of Berlusconi in the course of the previous legislature. At the same time, Prodi managed to appoint a little core group of ministers not responding to any of the several coalition partners, thereby reinforcing his personal role of nominator and allowing a fully fledged policy delegation. However, once in office, because of the fragility of his compound coalition, Prodi could not reshape the structure of his cabinet in any way during the 617 days of life of the government, notwithstanding the emergence of continuous policy disagreements among the partners. In other words, the political autonomy and the policy delegation in different fields are not so strongly correlated. Berlusconi seems to have missed the second while the Prodi II cabinet is particularly weak in the first respect. Resignations are not a frequent event of ministerial life in Italy. Only 40 ministers left their seats during the life of a cabinet between 1948 and 1992, that is to say, on average, less than one minister resigned from each cabinet. The situation has slightly changed during the Second Republic, and the phenomenon of shuffling is also not very relevant (Table 5.3). In most cases, resignations were clearly due to inter-party conflicts, marking a change in the strategy of one of the smaller coalitional parties, passing from a full participation to a form of external support. In some other cases, individual resignations were necessary after the appointment of a given minister to other important offices such as the Head of State, European Commissioner, parliamentary speaker or even leader of a parliamentary group. If we exclude all these cases of necessary interruption of service, the examples of voluntary individual resignation due to personal conflict with the chief executive are really unusual. Even when a minister was involved in a scandal or was suddenly at the centre of a journalistic case, the extrema ratio of a resignation was seldom employed.10 This is not surprising, given the weak position of the PM in the Italian

Table 5.4 Composition of Prodi II government in Italy, May 2007 Office

Salience

Holder

Party affiliation

Prime Minister Minister of Interior Minister of Foreign Affairs Minister of Economy Minister of Justice Minister of Health Minister of Defence Minister of Telecommunications Minister of Education Deputy PM Deputy PM Minister of Labour and social security Minister of Social solidarity Minister of Transport Minister of Infrastructures Minister of Economic Development Minister of Agricolture Minister of International trade Minister of Environment Minister of University and Research Minister of Culture Minister of Public administration reform (without portfolio) Minister of Relations with parliament (without portfolio) Minister of Regional affairs (without portfolio) Minister of Implementation governmental programme (without portfolio) Minister of Equal opportunities (without portfolio) Minister of Family (without portfolio) Minister of Youth and sports (without portfolio) Deputy Ministers

2.48 1.78 1.69 1.64 1.23 1.19 1.19 1.18 1.10 1.10 1.10 1.06 1.06 0.87 0.87 0.87 0.83 0.83 0.69 0.69 0.65

Prodi Amato D’Alema Padoa Schioppa Mastella Turco Parisi Gentiloni Fioroni D’Alema Rutelli Damiano Ferrero Bianchi Di Pietro Bersani De Castro Bonino Pecoraro Scanio Mussi Rutelli

Independent Independent DS Technician UDEUR DS DL DL DL DS DL DS RC PdCI Idv DS DL RnP Greens DS DL

0.56

Nicolais

Technician

0.56

Chiti

DS

0.52

Lanzillotta

DL

0.48

Santagata

Independent

0.48 0.48

Pollastrini Bindi

DS DL

0.48 0.47

Meandri

Ds 5 DS, 3 DL, 1 RnP, 1 RC 23 DS, 17 DL, 5 Independent, 5 DS, 3 DL, 2 Idv, 2 PdCI, 2 UDEUR, 2 Greens, 3 from minor parties

Junior ministers

0.46

Note: The measures of salience are those calculated by Druckman and Warwick (2005). In the cases of a new office I have taken the measure of salience of the most similar ministership. The measure of the salience of the deputy ministers has been calculated as the average between the low-rated minister (0.48) and the junior minister (0.46). DL (Democracy and Freedom – Daisy, Democrazia e Libertà – La Margherita), DS (Democratic Left, Democratici di Sinistra), IdV (Italy of Values, Italia dei Valori), PdCI (Party of Italian Communists, Partito dei Comunisti Italiani), RC (Communist Refoundation, Rifondazione Comunista), RnP (Radicals and Socialists, Rosa nel Pugno), UDEUR (Democratic Union for Europe, Unione Democratica per l’Europa)

Italy: the difficult road 95 institutional setting and the limited autonomy of each coalition actor to accept the abandonment of an important political spoil by a single minister. Even party leaders could it find extremely difficult, during the First Republic, to force the resignation of a minister who had not proved loyal in his role of partisan ‘watchdog’. Indeed, the factional nature of the coalition parties (particularly, but not exclusively, in the case of Christian Democrats) and the territorial representative role of many ministers made such an eventuality very infrequent. On the contrary, party leaders used to leave the choice of substitutes to the leaders of the factions which had to be represented, and very often they tried to convince the dissident ministers to remain within the cabinet (perhaps in another position, in order to preempt new conflicts), thus preserving the cabinet’s internal equilibrium. An extreme example of this kind of agency loss due to the polycentric system of representation within the cabinet can be seen in the case of the resignation of five ministers of the left wing of the DC (July 1990). The PM (Andreotti) could not personally select the alternatives, but even the party was in difficulty since a withdrawal of a whole faction was so unusual.11 The solution was a strange kind of reshuffle, with the appointment of a couple of personalities from the same political area of the dissidents, and three “technicians” also appointed with the consensus of the Christian Democratic leadership. Given the persistence of rather fragmented coalitions, we can argue that the situation has not changed much in the last fifteen years. Whatever the source of their legitimacy and the extent of delegation awarded, each minister tended to believe that he was there to stay, without considering the possible collapse in their agency relationship with the chief executive. The principle of collective responsibility of the government seems therefore to be much more critical and solid in comparison to the delegation power attributed to the PM by other constitutional provisions. In reality, the situation looks quite complex from a purely legal point of view. If sufficiently supported by his majority, the PM has adequate legal instruments to sack ministers or, alternatively, to adopt a censure against the members of his cabinet who are not respecting their role. In republican history, there has been a recurrent claim by PMs who aimed at creating a more transparent system of delegation and threatened to use their powers of firing.12 These powers, however, are unreliable given the political conditions, and there has been no example, up to now, of a personal and clear action of sacking exercised by the chief executive. The peculiar case of Minister of Justice Mancuso, who was replaced in 1995 after a conflict with the Head of State and the PM, is just an isolated example of ministerial removal, which is remembered by the constitutional lawyers as evidence of the effective possibility of judging the individual accountability of ministers. The minister was a pure technician (like all the members of that peculiar caretaker government), and the hard conflict with the Head of the State had induced some parliamentary groups to present a motion of personal no confidence, thus helping the PM to create the conditions to fire the dissident minister. However, this example of sacking a minister was a viable solution only because of a number of political conditions which seem to be very unlikely in the ‘normal’ conditions of coalition governments.

96 The selection of ministers in Europe On the other hand, during Second Republic the enhanced popular legitimacy and more pronounced personalities have made for stronger PMs able to replace some ministers, forcing some to resign or to move to other positions. After his large electoral victory in 2001, Berlusconi could easily form the new government, choosing personally some of his ministers. Also, he could have fired some of them (technicians or ministers coming from his own party) and improved the system of delegation inside the cabinet, creating a minister without portfolio for the implementation of the governmental programme and monitoring the actions of each single minister by means of an ad hoc department. Moreover, he explicitly referred to a British-style kind of reshuffle, announcing that the government would be ‘rearranged’ after two and a half years of service, following an accurate examination of each minister’s work. In reality, the life of the cabinet was progressively marked by the traditional problems of coalition governance, and the weakness of the PM in controlling ministerial selection soon re-emerged. For instance, the withdrawal of the Minister of Economy, Tremonti (July 2004), was somehow a price paid by Berlusconi to his allies. Much more evidently, after the electoral defeat of the regional elections, in 2005, a formal crisis was open, notwithstanding the resistance of Berlusconi who would have preferred to perform a light reshuffle. The consequences of this formal crisis were a little dance of ministers in the First Republic style, and the renouncement of some technical ministers in order to compensate the appointment of new representatives of some parties (including two micro-parties which had been represented only by junior minister in the previous executive). Even more evidently, Prodi could not respond to the high degree of conflict that immediately emerged after the beginning of his second cabinet (2006) by menacing sanctions against dissenting ministers. The only tool he could employ was the moral persuasion against the several parties supporting his cabinet (all of them necessary to provide a majority in the upper chamber), stressing his role of mediator among all the voices of the coalition. Overall, as Donati (1997) argues, in Italy unlike other parliamentary regimes, the PM does not have adequate instruments to limit or block the actions of individual ministers acting in a way inconsistent with their delegation, which makes them unaccountable for their deeds. In particular, still in this phase of the so called Second Republic, the sacking of dissenting ministers, although fully compatible with the current constitutional setting, is difficult owing to the political situation of weak party coalition.

Conclusion This chapter firstly aimed at reaching a cognitive goal – to answer the question about the extent of change produced by the processes of ministerial selection and de-selection by the recent shift to the Second Republic – but was also aimed at starting a discussion about the persistence or innovation in patterns of ministerial selection and de-selection in Italy, at the interpretative level. The first question can be answered with some degree of confidence: although the overall profile of government formation in Italy has changed significantly, most of the Italian specificities in the processes of selection and de-selection of ministers are still at

Italy: the difficult road 97 work, mainly because of the persistence of a difficult coalition game which makes the role of the PM very weak and increases the range of variability in terms of ministerial background, ministerial continuity and extent of ministerial delegation. The formation of a new Berlusconi government in May 2008, as a consequence of a new evident electoral victory of the centre-right camp, seems to have brought to a decisive change. The new government is indeed formed by only three parties: two of them (National Alliance and Forza Italia) are undertaking a process of fusion, while the third, the Northern League, remains mainly a significant political actor only in a part of the national territory. Moreover, Berlusconi could limit this time the ministerial team to just 21 ministers, and the overall government to only 60 persons. In fact, the consolidation of a new form of party government in a context of a bipolar party system has determined a multifaceted pattern of government formation, which is, in a sense, simplified by the stronger popular legitimacy of the chief executive and by the introduction of pre-electoral modes of government formation. This is somehow consistent with the political expectations of the early 1990s. Moreover such a simplification recalls the hypothesis of a more “proportional” distribution of portfolios due to the existence of a pre-electoral process of coalition formation (Carroll and Cox 2007). On the other hand, the processes of ministerial selection and de-selection seem to be even more complex that what we have observed in the past. Looking at ministerial selection, the changes concern the possible role of PM, who surely is more central in the conduct of the game than in the past, but can still be vetoed in different ways and therefore sometimes end up in the same frustrating weak position that characterized the chief executive during the First Republic. If to choose is difficult, to move and to sack seem to be almost impossible for the PM. However, also at the level of ministerial circulation and ministerial career the changes we have observed are surely not insignificant. A logic of more direct delegation from PM to the line ministers has been affirmed, in a way or another. More importantly, the ministerial pathways within the inner circle look simplified and more coherent with the skills of the personalities who have been appointed. Nevertheless, the period of the Second Republic covered by our data is still rather short, and the continuous noises within the party system, as well the neverending discussion about the bi-partisan transformation of an extremely fragmented bipolarism,13 prevent us from making forecasts about the form these processes will take in the future. What we can argue at the moment is that the party government paradigm has to be re-considered in the light of a number of evident factors of change: the crucial role of the PM as nominator and the increase of his expectation about each minister’s accountability; the reduction of the veto-power of minor parties, balanced by an increase of authority of all the party leaders of the winning coalition, who are normally more inclined to take direct government responsibilities. Finally, we should note a much more visible role of a number of ‘inner circle’ ministerial actors, in comparison to the First Republic style, due to a number of reasons, including the evolution of governmental organization and communication style, the complex nature of some policy issues and probably also

98 The selection of ministers in Europe the effect, favorable for government ministers, of the Europeanization of some domestic policies. Our impressionistic account cannot be refined at the moment. To complete this analysis, an accurate qualitative look at the division of work within the cabinet and at the system of incentives and rewards for governmental ministers acting in different roles is needed. This would improve our measurement of the salience of different ministerial positions, together with our understanding of the relevance of ministerial circulation in the different government roles and the real extent of party gatekeeping vis-à-vis the institutional selectorate. But this is another story and another problem.

Notes 1 A broader analysis of recent institutional change in Italy can be found in Cotta and Verzichelli (2007). 2 Berlusconi in 1994, 2001 and 2008, and Prodi in 1996 and 2006, were clearly chosen by the voters since they had guided their respective coalitions in the legislative elections. D’Alema (1998 and 1999) and Amato (2000) were designated by the parliamentary majority following the traditional scheme of the First Republic. 3 The electoral system used between 1994 and 2001 was a mixed-plurality system with 75 per cent of the seats allocated by a first-past-the-post-system and 25 per cent by a PR system. The system adopted before the 2006 election is a mixed PR system allocating a majority bonus to the winning coalition nationwide (in the lower chamber) or a number of small bonuses to the coalition which wins in each region (Senate). Despite the different outcomes in degrees of proportionality and fragmentation, both force the parties to join one of the two bipolar cartels in order to maximize their parliamentary representation. 4 The procedure is the following: once the cabinet has been nominated and sworn in by the Head of State, it will have to appear before each chamber within ten days to get a vote of confidence. In both chambers there is a minimum quorum requirement (50 per cent + 1 MP) and a simple majority is sufficient in order to get the confidence vote. 5 The research of Calise and Mannheimer was indeed particularly devoted to the peculiar power system within the Christian Democratic party elite. Generalizing from it seems to be reasonable however, since Christian Democrat ministers represented a large majority of the Italian ministers (about 75 per cent between 1948 and the beginning of the 1980s, when the book by Calise and Mannheimer was published; and still about 60 per cent if we consider the whole First Republic period, that is 1948–92). Furthermore, the other governmental parties tended to follow a very similar recruitment process, even changing somehow their original system of selection to adopt that of the DC, as argued by the same authors and even by Cotta (1979). 6 During the crisis of 1992, the Christian Democrats, who were still the largest governmental party, compelled their ministers to resign from their parliamentary mandate, once appointed in cabinet. 7 Not surprisingly, after 1992 there were no cases of exploratory informateurs or formateurs who failed in their attempt to form a new cabinet. 8 Actually Romano Prodi had a short experience as substitute (technical) Minister of Industry at the end of the 1970s. 9 Emilio Colombo remained 26 years in government, achieving 30 ministerships between 1955 and 1993. He got eight different jobs, becoming PM in 1970 and then returning within the inner circle a few times more. Giulio Andreotti had a 29-year experience, getting 28 ministerships (among them, seven prime ministerships) between 1954 and 1992. Considering that both of them had entered the sphere of

Italy: the difficult road 99

10

11

12

13

government some years before as junior ministers (Andreotti in 1947, Colombo in 1948), we can use these two examples as ‘perfect’ illustrations of the stereotypical ministerial career during the First Republic. A rather isolated example was that of Attilio Piccioni, one of the most eminent leaders of the first generation of the Christian Democratic Party. He resigned from the office of Minister of Foreign Affairs in 1954, after his son had been involved in a police investigation following the murder of a young woman. In other circumstances the left wing of the DC had behaved as an autonomous party, giving its support to a cabinet without accepting any office within it. But this was the first time in which a whole group of ministers from a party faction decided to resign. It is worth remembering that this event was somehow anticipating the fragmentation of the party, which would take place three years later, after the scandals of Tangentopoli, in many political formations. Significantly, this first strong division occurred on a specific policy issue, the decision of the government to ask a confidence vote on a new decree concerning the national private television network, which would have favoured the quasi-monopoly of Silvio Berlusconi, the future leader of the centre-right coalition, on this business sector. These claims were particularly recurrent after the age of consociativism in the 1970s: Spadolini (PM between 1981 and 1982) drafted a Decalogue including a number of codes of conduct for individual ministers. Craxi (1983–86) and De Mita (1988–1989) – two leaders who were able to cumulate the office of party leader and that of PM – implemented a strong ‘vertical style’: Craxi created a ‘council of Cabinet’ with a few ministers representing the coalition partners thus avoiding the implementation of external coalition summits. De Mita passed the ordinary law regulating the PM office and reinforced the power of agenda of the PM vis-à-vis the collective cabinet. In particular, the effective reduction of fragmentation in the centre-left coalition would be a decisive thrust to assure a better governability when this camp is majoritarian. The Prodi II cabinet (2006–8) was formed by at least ten different political parties (without considering some micro-formations represented in the residual group of the chambers).

References Bergman, T., Müller, W. C. and Strøm, K. (2003) ‘Democratic Delegation and Accountability: Cross-national Patterns’. In Bergman, Müller and Strøm (eds) Delegation and Accountability in Western Democracies. Oxford: Oxford University Press. Calise, M. (2005) ‘Presidentialization, Italian Style’. In Poguntke, T. and Webb, P. (eds) The Presidentialisation of Politics: A Comparative Study of Modern Democracies. Oxford: Oxford University Press, 88–106. Calise, M. (2006) La terza repubblica. Partiti contro presidenti. Rome and Bari: Laterza. Calise, M. and Mannheimer, R. (1982) Governanti in Italia. Un trentennio repubblicano. Bologna: Il Mulino. Carroll, R. and Cox, G. W. (2007) ‘The Logic of Gamson’s Law: Pre-election Coalitions and Portfolio Allocation’. In American Journal of Political Science 51, 2: 300–313. Cotta, M. (1979) Parlamento e classe politica in Italia. Bologna: Il Mulino. Cotta, M. (1988) ‘Italy: A Fragmented Government’. In Blondel, J. and Müller-Rommel, F. (eds) Cabinet Government in Western Europe. London: Macmillan. Cotta, M. and Verzichelli, L. (2003) ‘Ministers in Italy: Notables, Partymen, Technocrats and Mediamen’. Southern European Politics and Society, 4: 117–152. Cotta, M. and Verzichelli, L. (2007) Political Institutions in Italy. Oxford: Oxford University Press.

100 The selection of ministers in Europe Criscitiello, A. (1993) ‘Majority Summits: Decision-Making Inside the Cabinet and Out: Italy, 1970–1990’, West European Politics 4: 581–594. De Winter, L. (1991) Parliamentary and Party Pathways to the Cabinet’. In Blondel, J. and Thiebault, J. L. (eds), The Profession of Government Ministers in Western Europe. Basingstoke: Macmillan. De Winter, L. (1995) ‘The Role of Parliament in Government Formation and Resignation’. In Doering, H. (ed.) Parliaments and Majority Rule in Western Europe. New York: St Martin’s Press, 115–151. Dogan, M. (1989) ‘How to Become a Cabinet Member in Italy. Unwritten Rules of the Political Game’. In Dogan (ed.) Pathways to Power. Selecting Rulers in Pluralistic Democracies. Boulder: Westview. Donati, F. (1997) La responsabilità politica dei ministri nella forma di governo italiana. Turin: Giappichelli. Druckman J. N. and Warwick, P. V. (2005) ‘The Missing Piece: Measuring Portfolio Salience in Western, European Parliamentary Democracies’. European Journal of Political Research 44: 17–42. Fabbrini, S. (2001) Tra pressioni e veti. Il cambiamento politico in Italia. Rome and Bari: Laterza. Pasquino, G. (1986) ‘Party Government in Italy: Achievements and Prospects’. In Katz, R. S. (ed.) Party Governments: European and American Experiences. Berlin and New York: Walter de Gruyter, 202–242. Thies, M. (2001) ‘Keeping Tabs on Partners: The Logic of Delegation in Coalition Governments’. American Journal of Political Science 45, 3: 580–598. Vassallo, S. (1994) Il governo di partito in Italia. Bologna: Il Mulino. Verzichelli, L. and Cotta, M. (2000) ‘Italy: From Constrained Coalitions to Alternating Governments?’ In Müller, W. C. and Strøm, K. (eds) Coalition Governments in Western Europe. Oxford: Oxford University Press, 433–497.

6

Cabinet dynamics in democratic Spain (1977–2008) José Real-Dato and Miguel Jerez-Mir

Introduction The origins of the present Spanish political system must be found after the death of Franco – who had managed to assert his rule over the country after the Civil War (1936–39) that put to an end the liberal democratic regime of the Second Republic (1931–36). Following Franco’s demise in November 1975, his successor as Head of State, King Juan Carlos I, opened the transition to democracy. To lead the process the King appointed in July 1976 as Prime Minister a young member of the reformist sector within the regime, Adolfo Suárez. In his first government, Suárez had to cope with strong pressures coming from both the more recalcitrant sectors of the Franco regime and the democratic opposition – the former opposed to any kind of move towards democracy, and the latter demanding a complete rupture with the previous regime. The Transition became a process of reforma pactada/ruptura pactada through elite settlement allowing the institutions of the Franco regime to be substituted for democratic ones with the new status quo accepted by the opposition as it was comparable to many contemporary western democracies. Democracy was finally re-established with the general election of June 1977, inaugurating a constituent period ending up with the institutionalization of parliamentary monarchy in the 1978 Constitution (Linz et al. 2003; Linz and Stepan 1996: ch. 6). Since the restoration of democracy until March 2008, there have been nine legislative terms (one constituent plus eight ordinary terms), ten governments, and five presidents of the government: Adolfo Suárez, Leopoldo Calvo-Sotelo, Felipe González, José M. Aznar, and José L. Rodríguez Zapatero (Table 6.1). In four of the legislative terms the governing party enjoyed an absolute majority: the Socialist Party (PSOE) in 1982, 1986 and 1989; and the Popular Party (PP) in 2000. In the other cases, the winning parties obtained a large plurality of seats in the lower chamber, allowing the Prime Minister (PM) of the largest party to form one-party cabinets whilst leaning on regular or ad hoc support from other parliamentary parties (see Reniu 2002). PMs, ministers and governments have been extensively studied during this time (Aragón and Gómez 2005; Bar 1983, 1997; Botella 1997; Heywood 1991, 1995; Linz et al. 2003; López 1996; Montabes 2001; Olías 1994; Pérez Francesch 1998; Rodríguez 2004; Román 2002; Van Biezen and Hopkin 2005). However, none of

102 The selection of ministers in Europe Table 6.1 Three decades of democratic cabinets in Spain, 1977–2008 Legislative term

Party in government

Prime Minister

Parliamentary support

Government duration

Constituent

UCD

Adolfo Suárez (I)

Minority

First

UCD

Adolfo Suárez (II)

Minority

1977-07-05 > 1979-04-05 1979-04-06 > 1981-02-25

First

UCD

Minority

Second

PSOE

Leopoldo Calvo-Sotelo Felipe González (I)

Third

PSOE

Felipe González (II)

Majority

Fourth

PSOE

Felipe González (III)

Majority

Fifth

PSOE

Felipe González (IV)

Sixth

PP

José M. Aznar (I)

Seventh

PP

José M. Aznar (II)

Minority + parliamentary support (nationalist parties) Minority + parliamentary support (nationalist parties) Absolute Majority

Eighth

PSOE

José L. Rodríguez Zapatero

Majority

Minority + parliamentary support (nationalists parties + IU)

1981-02-26 > 1982-12-02 1982-12-03 > 1986-07-25 1986-07-26 > 1989-12-06 1989-12-07 > 1993-07-13 1993-07-14 > 1996-05-04

1996-05-05 > 2000-04-27

2000-04-28 > 2004-04-17 2004-04-18 > 2004-04-13

UCD: Unión de Centro Democrático; PSOE: Partido Socialista Obrero Español; PP: Partido Popular; IU: Izquierda Unida

these works has dealt in depth with the dynamics of the selection and de-selection of ministers, a gap that we aim to fill with this chapter.

Constitutional framework The tradition of the Spanish constitutional government, inaugurated by the Constitution of Cádiz (1812), is to divide executive power between the head of state and the government. The latter was always conceived restrictively, being identified with the cabinet, and consisting in the President of the Council of Ministers (equivalent to the Prime Minister)1 and the ministers. In the twentieth century, under King Alfonso XIII (1902–23), the head of state’s interventionism relegated the PM to a weak institutional position within the political structure, mainly manifested in the short duration of cabinets. During the Second Republic,

Cabinet dynamics in democratic Spain 103 and until the outbreak of the Civil War, competing competencies of the offices of the President of the Republic and the PM contributed to even more ephemeral cabinets (Linz et al. 2003: 43, 62–3; Heywood 1995: 88–9).2 Under the Franco regime (1939–75), the formal division between head of state and PM was blurred as both offices were held by Generalisimo Franco until June 1973, when Admiral Carrero Blanco was appointed President of the Government. The traditional secondary role of the government was reversed in the parliamentary monarchy system instituted by the 1978 Constitution. It establishes that the government is in charge of ‘conducting the domestic and foreign policy, civil and military administration and the defence of the State’ (sect. 97). Regarding the government’s organization, the Constitution only confirms that ‘[it] shall consist of the President, Vice-presidents, when appropriate, Ministers and other members as may be created by law’ (sect. 98), leaving to the legislature the task of establishing its content. In practice, governments since 1977 have tended to reproduce the traditional restricted conception, which was finally confirmed by the Government Act passed in 1997. This act limited the composition of the government to the President, the Vice-president(s) when appropriate and the ministers. This configuration excludes the alternatives existing in other countries where the government expands to lower decisional levels, as in the British case. Secretaries of State – the highest echelon in the state’s administration, immediately below ministers – are in charge of governmental action in a specific policy area and may attend the Council of Ministers only if they are summoned to do so (Government Act, sect. 5.2). Therefore, in practice, the Spanish government is equated with the cabinet (the Council of Ministers). Nevertheless, despite their overlapping membership, both concepts are formally differentiated. The Council of Ministers is the formal collegiate gathering of the government’s members and is in charge of the most important decisions of the executive power (such as approving bills, passing legislative decrees, law decrees and regulations or deciding to negotiate international treaties (Government Act, sec. 5)). In contrast, the government as a body also includes the actions of its individual members. Finally, the Government Act also allows government to create delegated committees –chaired by the PM and formed by a subset of ministers in order to co-ordinate specific policy areas, an inheritance from the government organization during Franco’s regime. As in every parliamentary system, the Spanish government depends on the support of the parliament (Cortes Generales), consisting of two chambers, the Congreso de los Diputados and the Senado. In this bicameral system, the Congreso or lower chamber has a superior standing in the legislative process, playing also an exclusive role in the formation and termination of the government. The 1978 Constitution regulates the vote for investiture (sect. 99), which falls only on the President. Following the renewal of the Congreso after a general election, or where the PM loses parliamentary confidence (see below), resigns or dies, the King shall nominate through the Speaker of the Congreso, and after personal consultation with the representatives appointed by the political groups with parliamentary representation, a candidate for the Presidency of the Government. Considering the politically neutral and mostly symbolic role the Constitution attributes to the monarch, the practice is that the nominated candidate (not

104 The selection of ministers in Europe necessarily a member of the parliament) is designated by the party with a plurality of seats at the Congreso. After the submission of the political programme by the candidate and the corresponding debate,3 the Congreso proceeds to vote. If candidate does not obtain an absolute majority, the vote is repeated forty-eight hours later, this time the confidence being granted by a plurality of MPs. 4 If he fails in the second vote, another candidate is nominated by the King. If within two months no candidate obtains the confidence of the Congreso, the monarch dissolves both chambers and calls for new elections. There are two constitutional mechanisms for withdrawing parliamentary confidence to the government. The first is the ‘motion of censure’ (sect. 113), proposed by at least a tenth of the members of the Congreso. It is defined as a ‘constructive’ motion of censure, since it has to include an alternative candidate for the PM. If the motion succeeds (supported by an absolute majority of the members of the Congreso), the alternative candidate is appointed by the King as President of the Government. This constructive character has largely restricted its utilization, since it is difficult for candidates to obtain the required parliamentary support. When it has been presented – twice in thirty years, both unsuccessfully (one in 1980 by the Socialist Party, PSOE, and the other in 1987 by the Popular Alliance, AP) – it was aimed basically at increasing the public visibility of the challengers (Felipe González and Antonio Hernández-Mancha, respectively) and not to provoke the fall of the cabinet. The second mechanism for withdrawing confidence is the ‘confidence vote’ (sect. 112), which is presented by the President of the Government after deliberation by the Council of Ministers. This vote may be referred either to the entire programme of government or to a general policy statement. If the PM does not get a plurality of favourable votes, he is deemed to have lost the confidence of the chamber, and forced to resign. Confidence votes have also been infrequent. Only two have been presented since 1978, the first one in 1980 by Adolfo Suárez, who aimed to reinforce his weak political position after the PSOE’s motion of censure; and the other in 1990 by Felipe González, who aimed to confirm his legitimacy as the previous vote of investiture after the 1989 election took place while electoral results were still in dispute in three constituencies. Therefore, the PM has a higher political-institutional position compared to other cabinet members (Heywood 1991: 98; Pérez Francesch 1998: 61; Zapatero 2003: 932). This is confirmed by the extensive powers and functions the Constitution grants to the PM – he can dissolve (formally through the King) the parliament and call for new elections; he also summons the cabinet, sets the agenda, gives instructions and solves conflicts between ministers. This institutional centrality transmits into parliamentary elections which are mainly perceived as elections focused on selecting the PM (Van Biezen and Hopkin 2005: 120ff.; see also Heywood and Molina 2000). One of the powers of the PM is deciding cabinet composition. Whilst it is the King who appoints or dismisses cabinet members, in practice this is merely ceremonial. Since 1985, the Budgetary Act (later confirmed by the 1997 Government Act) empowers the PM to create, modify and suppress ministries by decree on his sole responsibility.

Cabinet dynamics in democratic Spain 105 Ministers are usually at the head of ministerial departments, though there have been ‘ministers without portfolio’ too. Vice-presidents, if appointed, stand in for the PM in case of absence, or illness (Government Act, sect. 13). Their other powers are established by the PM when appointed, including assigning them to specific portfolios. There are no formal rules regulating the process of cabinet formation. As the vote of investiture involves only the PM, and the final list of ministers depends on his sole will, the timing of the final constitution of the cabinet after an election is linked to the duration of the period between the election and the investiture. The average duration of this period is 38 days, the quickest cabinet formation in 1979 (29 days after the election) and the slowest in 1996 (63 days). In contrast, the average time from investiture to cabinet formation is three days, with a maximum of seven (in 1979) and a minimum of one (1996). Formal requirements to be appointed a minister are limited to having Spanish nationality, being entitled to full electoral rights, and not being barred by judicial sentence from occupying public offices. Once appointed, the only position a minister may simultaneously hold is that of member of the national parliament (MP),5 being excluded from any other public or private posts (Government Act, sect. 14). Finally, there is no institutional mechanism through which the chamber – either the Congreso or the Senado – can claim formal individual ministerial responsibility ending in forced resignation. The only alternative provided by the Constitution is that of the motion of censure, which is based on the principle of collective responsibility of the cabinet, as it affects the cabinet as a whole. Nevertheless, even though since 1981 a ‘motion of reprobation’ referring to one or more ministers can be presented in parliament, such a motion, even if successful, does not involve any direct consequence for the reprobated minister. Thus, forcing a resignation depends only on the will of the PM (Montero and García Morillo 1984: 192–9).

Structure and composition of cabinets6 Constitutional and legal arrangements grant the PM complete control over cabinet structure and individual composition. Excluding the PM, the average size of cabinet is 16.9, ranging from 23 in Suárez’s first cabinets and 14 in the first Aznar government.7 Variations in cabinet size are related mostly to the presence of Vice-Presidents and ministers without portfolio, since the number of departments has remained more stable. In this sense, the average number of departments is 15.6 (with a standard deviation of 1.05). Ten departments appear in every government (sometimes with changing subheadings): Agriculture, Defence, Economy, Education, Foreign Affairs, Health, Interior, Justice, Labour and Public Works. Most of the governments also included other departments, such as Industry (present in nine out of ten governments), Culture (eight governments) and the Ministry of Presidency (eight governments), the latter being usually in charge of horizontal political co-ordination among departments, and between the government and the supporting parliamentary group (during the third and fourth

Jul 95(**+) Dec 95(*)

May 94(**+)

Nov 93(*+)

Jun 92(*+) Jul 93

Jan 92(*+)

Mar 91(**)

Jan 91(+)

May 90(*)

Dec 89

Jul 88(**)

Jul 85(**+) Jul 86

Dec 82

Sept 82(*)

Jul 82(**)

Dec 81(**)

Sept 81(**+)

Feb 81

Sept 80(**+)

Jan 80(**+)

May 80(**+)

Apr 79

Feb 78(**+)

Feb 78(*)

Sept 77(+)

July 77 Total departments

Jan 99(**)

Ministers without portfolio

Apr 99(*)

Vice-presidents with portfolio

L7 Aznar II

Vice-presidents (total)

Apr 00

Cabinet size

Feb 00(*+) L6 Aznar I

Feb 01(*)

L5 González IV

Jul 02(**)

L4 González III

Feb 03(*)

L1 Calvo-Sotelo I L2 Glez L3 Glez I II

Apr 04

Sep 03(**)

May 96

Note: Data include only ministers and, when appropriate, vice-president(s). Key: (*) Cabinet adjustment; (**) Cabinet reshuffle; (+) One or more resignations; L = legislative term

L8 Zapatero I

Apr 06(**+)

L1 Suárez II

Feb 07(*)

Constituent Suárez I

Sep 06(*)

Figure 6.1 Cabinet dynamics in Spain: cabinet size and organization (by events of change)

0

5

10

15

20

25

Jul 07 (**)

Cabinet dynamics in democratic Spain 107 legislative terms this latter function was accomplished by the Ministry for Relationships with the Parliament). Table 6.2 displays aggregate data about some relevant variables over the composition of cabinets. Governments’ mean age (computed as the average age of ministers at the moment of their appointment) is 46. Zapatero, González III and Aznar II are the governments with the highest average age. The lowest corresponds to González I, which introduced a brand new generation of politicians to the executive. Between 1977 and 1993 there were only three female ministers: the first in 1981 in the Calvo-Sotelo government; and two in González II, who continued in González III. In the fifth legislative term, the proportion of women increased in the González cabinets, although it never surpassed 20 per cent. Aznar reached his top of female representation in his second government (26.7 per cent), and, for the first time in Spain, a first-order portfolio like Foreign Affairs was assigned to a woman. In contrast, women in González governments were assigned politically less important departments, such as Social Affairs, Culture, or Health In this respect, the Zapatero government represents a further step, as it has been the first one formed on a parity basis and where a woman has occupied a Vice-Presidency. A large majority of ministers are usually members of the party in government. Only in González IV and the last Zapatero cabinet is the proportion substantially low compared with the general trend. Concerning party leadership positions, the much lower percentages in the PSOE governments (particularly during González’s premiership) are due to the fact that, from the 1984 Thirtieth PSOE Conference on, party statutes establish the incompatibility of member of the party’s Executive Committee (excepting the president, the general secretary, and the executive secretaries) with any executive public office (including cabinet positions), a constraint not appearing in other parties. Before Aznar’s second cabinet, with a few exceptions (Suárez’s last cabinet, González I, and the second cabinet in González II) more than 65 percent of the ministers had some kind of expertise regarding their portfolio (including those who had held the same portfolio in a previous term). This trend reverses in Aznar I. From his second cabinet on, the proportion of experts decreases considerably, particularly during the Zapatero government. Regarding those cabinet members who simultaneously occupy a seat in the parliament, they do so usually in the Congreso – more than a half in all cabinets, with exceptions in the Suárez I, CalvoSotelo, González IV, and Aznar I governments. In contrast, senators have been infrequent among ministers (just one in Aznar I since the Constitution was passed, and four during Suárez I). More than 85 per cent of ministers in every cabinet had some kind of previous political experience (Table 6.2). This trend was broken only in González IV and the two last Zapatero cabinets. National politics appear to be the main source of recruitment. Except for González IV and Zapatero, more than half of the ministers had some prior parliamentary experience before their current appointment (for ministers during the constituent legislative term, such experience refers to membership in Franco’s corporatist Cortes). Executive political positions in state administration are also an important source of ministers, particularly during the UCD governments – since a significant number held political office either

46.0

44.5

42.8

Apr 79 Jan 80 (**+) May 80 (**+) Sept 80 (**+)

Feb 81 Sept 81 (**+) Dec 81 (**) Jul 82 (**) Sept 82 (*)

Dec 82 Jul 85 (**+)

Jul 86 Jul 88 (**)

44.1

44.8

July 77 Sept 77 (+) Feb 78 (*) Feb 78 (**+)

Mean age

L1 Calvo-Sotelo

0.0 11.1

0.0 0.0

0.0 0.0 5.6 6.3 6.3

0.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0

Women

87.5 77.8

87.5 87.5

100.0 100.0 100.0 93.8 87.5

91.3 91.3 95.7 95.5

73.7 72.2 73.7 83.3

Affilt

18.8 33.3

25.0 31.3

60.0 53.3 61.1 50.0 50.0

69.6 69.6 56.5 54.5

63.2 61.1 63.2 77.8

Lead

87.5 61.1

50.0 43.8

86.7 80.0 72.2 81.3 81.3

65.2 69.6 73.9 59.1

68.4 72.2 68.4 72.2

Expert

87.5 61.1

50.0 62.5

53.3 46.7 61.1 50.0 43.8

65.2 65.2 65.2 63.6

47.4 44.4 47.4 50.0

Congr

Table 6.2 Individual characteristics of cabinet members in Spain (%), 1977–2008

Constituent Suárez I

L1 Suárez II

L2 Glez I

L3 Glez II

0.0 5.6

12.5 12.5

0.0 0.0 0.0 6.3 6.3

0.0 0.0 0.0 0.0

10.5 11.1 10.5 5.6

No

12.5 11.1

12.5 18.8

6.7 6.7 5.6 6.3 0.0

4.3 4.3 8.7 4.5

5.3 5.6 5.3 5.6

Loc

6.3 5.6

6.3 6.3

0.0 0.0 5.6 6.3 6.3

0.0 0.0 0.0 4.5

0.0 0.0 0.0 0.0

Reg

Prior political experience

37.5 33.3

37.5 31.3

73.3 73.3 72.2 62.5 62.5

65.2 69.6 78.3 72.7

68.4 72.2 73.7 72.2

Nat Adm

62.5 50.0

56.3 56.3

53.3 46.7 61.1 56.3 56.3

69.6 69.6 65.2 59.1

47.4 50.0 52.6 50.0

Parl

75.0 50.0

0.0 6.3

100.0 93.3 72.2 62.5 56.3

56.5 60.9 47.8 54.5

36.8 38.9 42.1 38.9

Gov

47.6

47.5

46.3

47.6

49.4

Dec 89 May 90 (*) Jan 91 (+) Mar 91 (**) Jan 92 (*+) Jun 92 (*+)

Jul 93 Nov 93 (*+) May 94 (**+) Jul 95 (**+) Dec 95 (*)

May 96 Jan 99 (**) Apr 99 (*) Feb 00 (*+)

Apr 00 Feb 01 (*) Jul 02 (**) Feb 03 (*) Sep 03 (**)

Apr 04 Apr 06 (**+) Sep 06 (*) Feb 07 (*) Jul 07 (**)

L4 González III

L5 González IV

L7 Aznar II

L8 Zapatero

50.0 50.0 50.0 50.0 43.8

12.5 12.5 13.3 20.0 26.7

21.4 14.3 7.1 7.1

17.6 17.6 18.8 20.0 20.0

11.1 11.1 11.8 11.8 11.8 11.8

75.0 75.0 75.0 68.8 56.3

87.5 87.5 93.3 93.3 93.3

78.6 78.6 78.6 78.6

58.8 58.8 62.5 60.0 60.0

68.8 83.3 82.4 88.2 88.2 88.2

37.5 37.5 31.3 25.0 25.0

56.3 50.0 53.3 60.0 46.7

64.3 57.1 50.0 57.1

11.8 11.8 12.5 20.0 20.0

27.8 22.2 17.6 17.6 17.6 17.6

43.8 37.5 37.5 37.5 43.8

56.3 50.0 53.3 53.3 46.7

71.4 64.3 57.1 57.1

82.4 82.4 81.3 80.0 80.0

100.0 94.4 94.1 82.4 88.2 88.2

62.5 75.0 68.8 62.5 62.5

62.5 56.3 60.0 60.0 60.0

35.7 42.9 50.0 57.1

41.2 41.2 43.8 40.0 33.3

66.7 66.7 64.7 52.9 52.9 52.9

12.5 12.5 12.5 18.8 31.3

0.0 0.0 0.0 0.0 0.0

7.1 7.1 7.1 7.1

17.6 17.6 18.8 20.0 20.0

0.0 0.0 0.0 0.0 0.0 0.0

12.5 12.5 12.5 12.5 18.8

25.0 25.0 33.3 33.3 33.3

28.6 21.4 21.4 21.4

5.9 11.8 6.3 0.0 6.7

11.1 11.1 11.8 5.9 5.9 5.9

31.3 25.0 25.0 25.0 12.5

50.0 50.0 33.3 26.7 13.3

42.9 50.0 57.1 57.1

17.6 17.6 25.0 26.7 33.3

5.6 5.6 5.9 5.9 11.8 11.8

50.0 56.3 56.3 56.3 56.3

37.5 31.3 26.7 33.3 46.7

28.6 35.7 42.9 42.9

52.9 58.8 56.3 53.3 60.0

38.9 38.9 41.2 52.9 47.1 47.1

50.0 43.8 43.8 37.5 37.5

62.5 62.5 60.0 60.0 66.7

64.3 64.3 64.3 71.4

35.3 29.4 25.0 26.7 20.0

66.7 66.7 70.6 58.8 58.8 52.9

6.3 12.5 12.5 12.5 18.8

50.0 43.8 40.0 40.0 26.7

7.1 7.1 7.1 7.1

52.9 47.1 50.0 40.0 33.3

100.0 94.4 94.1 64.7 64.7 58.8

Note: See Figure 1. Key: Affilt = Affiliated to party in government; Lead = Party leadership; Expert = Expertise; Congr = MP for the Congress at the moment of the appointment; Loc = Experience at the local level; Reg = Experience in regional governments or parliament; Nat Adm = Political appointment at the national administration; Parl = Parliamentary experience; Gov = Governmental experience (ministers)

L6 Aznar I

110 The selection of ministers in Europe during the Franco regime or during the Transition – and, once more, in González IV. In contrast, local and regional experiences are comparatively less frequent, although they play an important role (particularly regional experience) in the Aznar cabinets. In the case of the UCD governments and González I, the absence of ministers with regional background is due to the fact that autonomous communities introduced by the 1978 Constitution were still under construction. Before the Aznar premiership, governmental experience mainly reflects patterns of continuity along governments of the same party. In this sense, only two individuals (Arias Salgado and Fernández Ordóñez) have been ministers with two different parties (UCD and PP, and UCD and PSOE respectively). Regarding the Zapatero government, only two cabinet members were also ministers with González.

The dynamics of the selection and de-selection of Spanish ministers To understand patterns in the size, structure and composition of cabinets we need to pay attention to the context in which cabinet changes take place and the general dynamics of the selection and de-selection of Spanish ministers. Figure 6.2 gives the basic data on these dynamics (entries, exits, portfolio changes and continuity of ministers related to events of change). Figure 6.3 completes the picture by displaying the Kaplan-Meier survival function graphs by prime ministerial term. These graphs summarize the dynamics of the different governments in terms of global duration and stability in ministerial positions. They also show the average ministerial duration at each legislative term.8 In order to identify patterns in the selection and de-selection of ministers, we need to pay attention to the inner factors affecting cabinet dynamics. The institutional centrality of the PM in Spain entails that cabinet composition is, first of all, derived from the PM’s calculations regarding a series of goals (promoting certain policies, obtaining public acceptation, increasing electoral expectations, maintaining coalition-party stability) and beliefs about the suitability of potential appointees. Limitng the PM’s formal autonomy are a number of constraints deriving from his party, other supportive parliamentary parties and public opinion. In short, cabinet dynamics depend on the particular configuration of these constraints influencing the PM. We can classify the different prime ministerial periods in two broad categories. First, stable governments (Suárez I, González I and II, Aznar I and II and Zapatero). Here the PM has a wide margin of manoeuvre, based on a solid parliamentary support and a widely accepted or strong leadership within his party. What in Figure 6.3 could be interpreted as cabinet instability (Aznar II and Zapatero) largely results from the PM’s strategic calculations. Secondly, unstable governments (Suárez II, Calvo-Sotelo and González III and IV). Here, instability results from a PM’s weak or questioned party leadership, internal party struggles or the influence of major external shocks (mainly scandals) undermining the PMs’ autonomy.

Feb 00(*+)

Jan 99(**)

May 96

Dec 95(*)

Jul 95(**+)

May 94(**+)

Nov 93(*+)

Jul 93

Jun 92(*+)

Jan 92(*+)

Mar 91(**)

Jan 91(+)

May 90(*)

Jul 88(**)

Jul 86

Jul 85(**+)

Dec 82

Sept 82(*)

Jul 82(**)

Dec 81(**)

Sept 81(**+)

Feb 81

Sept 80(**+)

May 80(**+)

Jan 80(**+)

Apr 79

Feb 78(**+)

Sept 77(+)

Feb 78(*)

July 77

Portfolio changes

L8 Zapatero I

Continue from last legislative term

Apr 00

Exits

Sep 03(**)

Entries

Apr 99(*)

L7 Aznar II

Feb 01(*)

L6 Aznar I

Feb 03(*)

L5 González IV

Jul 02(**)

L4 González III

Apr 04

L2 L3 Glez I Glez II

Apr 06(**+)

L1 Suárez II L1 Calvo-Sotelo I

Feb 07(*)

Constituent Suárez I

Sep 06(*)

Dec 89

Note: See Figure 6.1. (*) Cabinet adjustment; (**) Cabinet reshuffle; (+) One or more resignations. Key: L = legislative term.

Figure 6.2 Cabinet dynamics in Spain: entries, exits, portfolio changes and continuity, 1977-2008

0

5

10

15

20

25

Jul 07 (**)

0

500

1000

GD = 1415 days AMD = 710.0 days

L VII Aznar II

GD = 1190 days AMD = 832.5 days

L III González II

GD = 603 days AMD = 437.8 days

L Constituent Suárez I

1500

0

500

1000

GD = 1420 days AMD = 884.8 days

L VIII Zapatero

GD = 1276 days AMD = 736.0 days

L IV González III

AMD = 345.0 days

GD = 691 days

L I Suárez II

1500

0

500

1500

0

500

1000

1500

GD = 1406 days AMD = 1035.7 days

L VI Aznar I

GD = 1296 days AMD = 942.3 days

L II González I

Note: See Figure 6.1. Key: Vertical axis = survival function; Horizontal axis = analysis time (days). GD = Global duration; AMD = Average ministerial duration

1000

GD = 962 days AMD = 643.2 days

L V González IV

GD = 608 days AMD = 341.9 days

L I Calvo-Sotelo

Figure 6.3 Ministerial survivor functions by prime ministerial term in Spain, 1977–2008

0.00

0.25

0.50

0.75

1.00

0.00

0.25

0.50

0.75

1.00

0.00

0.25

0.50

0.75

1.00

Cabinet dynamics in democratic Spain 113

Stable governments In this category governments have long global and the highest average ministerial durations (Figure 6.3). Suárez first democratic government ‘anomaly’ is basically the result of the particular nature of the term, as the PM called for elections as soon as the Constitution was passed at the end of 1978. Cabinet stability in these governments is associated with the PM’s ability to control cabinet inner dynamics, facilitated by a solid parliamentary support and a widely accepted or strong party leadership. González I and II, and Aznar II, indisputably had large parliamentary support backed by an absolute majority of seats at the Congreso. Besides, although Suárez I, Aznar I and Zapatero did not enjoy such a majority, they were highly stable governments. The Aznar government in 1996 relied on formal post-electoral parliamentary pacts signed with the moderate Catalan (CiU), Basque (PNV) and Canary Islands (CC) nationalists, in exchange for advancements in devolution for these regions (Reniu 2002). The Zapatero government managed to manufacture ad hoc agreements with other parliamentary forces (apart from the PP) on the basis of an open, flexible negotiating style, and thanks to the programmatic coincidence of many policy proposals with left-wing parties, mainly IU (a coalition including the Communist Party) and the left Catalan (ERC) and Galician (BNG) nationalist parties (partners of the PSOE in their respective regional governments). The minority UCD government formed by Suárez in June 1977 did not count on other parties’ parliamentary support, but given the constituent character of this term and the consensual style politics it induced – the main manifestation of which was the participation of almost all parliamentary forces in the drafting of the Constitution – this did not affect cabinet stability. Another common feature in these stable governments is that PMs exerted strong or widely accepted leadership over their party. Such leadership is indisputable in the cases of González, Aznar and Zapatero. With González elected in 1974 as general secretary, the PSOE overcame its internal divisions – originally between the exile and the internal organizations, and then between Marxists and social democrats – managing to integrate all the social-democrat family under the same banner (Méndez 2000; Sánchez Cervelló 2004). His charismatic presence was also important for the PSOE’s mass public support – after the 1977 election it was the second parliamentary political force – and to consolidate an image of moderation to appeal as a solid alternative government. Such leadership was complemented by the tight control exerted by the vice-general secretary Alfonso Guerra – Vicepresident of the government between 1982 and 1991 – over the party organization, making it a highly disciplined electoral machine (López Garrido 1985; Heywood 1991: 102). During their time in office, González and Guerra’s control over the party was reinforced by the fact that from 1984 on (see above) the presence of members of the PSOE executive committee in the cabinet was significantly restricted (Maravall 2003). After his election as president of the Popular Party, in 1990, Aznar transformed the legacy of the Popular Alliance (AP) – the conservative party from which the PP was re-founded – into a modern, moderate party, which managed to unify the entire Spanish centre right, incorporating a great deal of the inheritance from

114 The selection of ministers in Europe the UCD (García-Guereta 2001). Over time Aznar consolidated exclusive leadership over the PP, giving him absolute control over party decisions, including the selection of his successor – Mariano Rajoy – in 2003. Similarly, Zapatero’s election as general secretary in 2000 ended the internal crisis the PSOE had experienced since the early 1990s, exacerbated by the 1996 electoral defeat and González’s resignation as general secretary in 1997. The favourable electoral results obtained by the PSOE in the 2003 local and regional elections, and the unexpected victory at the 2004 general election,9 consolidated Zapatero’s leadership over the party. Suárez’s leadership over the UCD during his first democratic government never was complete. Appointed PM by the King in July 1976, Suárez stood aside from party politics until the 1977 election. At that time, as he wanted to run for office in order to complete the transition, he needed an electoral platform (Martín Villa 1984: 76). It was provided by the Union of the Democratic Centre (UCD), a coalition of 15 small liberal, Christian democratic, social democratic and regionalist parties, formed around a handful of relevant political personalities (barones), together with independents coming from the reformist sector of the Franco regime (Huneeus 1985; Hopkin 1999). The alliance was a win-win solution for both partners, providing Suárez with an electoral label to run for office (even though he was not a formal member of any party) and the UCD benefited from the wide popularity of the President and the support of the state bureaucratic structure. After the election, Suárez’s popularity and his commanding role during transition, along with the favourable electoral results, helped consolidate his leadership of the coalition (Linz 2001) which was transformed into a single party in December 1977 with Suárez elected as president (Huneeus 1985; Hopkin 1999). The consensual character of parliamentary politics during the constitutional term and the continuity of general public support clearly favoured Suárez’s position before the barones, who had conserved their control over the ideological party families that formed the UCD. The PM’s autonomy granted by stable parliamentary support and leadership over the party organization manifests itself in the process of cabinet formation. Although Suárez adjusted to the UCD’s coalitional nature – eight of the barones and other UCD significant members were appointed ministers – he also kept a significant degree of autonomy in designing the cabinet, maintaining in the cabinet six out of the 20 ministers of his 1976 government. In addition, the vicepresidencies with portfolio (Defence and Economy) were given to independent personalities (a military officer and a prestigious economist, respectively) appointed in order to deal with two of the major problems during the constituent period – namely, military distrust of the newborn democratic regime and economic crisis. As third vice-president, in charge of political co-ordination with the party and the parliamentary group, Suárez appointed his right-hand man, Fernando Abril Martorell. The need to combine internal representation of the UCD families with an inner circle of collaborators is the main reason for the comparatively large size of Suárez’s cabinets and the higher number of ministers without portfolio (usually charged with a specific task, such as regional devolution, EEC negotiations or administrative reform).

Cabinet dynamics in democratic Spain 115 The first González cabinet introduced a new ministerial elite into government, politically trained in opposition during the late years of the Franco regime and the Transition. This personnel proved efficient enough in the eyes of the PM, as in 1986, 12 out of 16 ministers of the González I last cabinet continued – including those heading the main portfolios, such as Vice-presidency, Interior, Economy, Defence and Foreign Affairs. The composition of Aznar’s first government was meant to send – both to the public and to his parliamentary allies – the message of the PP as a centrist and moderate party. In this sense, Aznar’s first cabinet carefully represented the different sensibilities integrating the party since its re-foundation (Tussell 2005: 90), and included members from the AP and the extinct UCD, as well as independents (one of them, Josep Piqué, Minister of Industry, an economist who formerly had related offices in the Catalan regional government, a clear signal to Aznar’s Catalan nationalists’ parliamentary partners). The structure of the government was reduced to 14 departments, following the PP electoral manifesto’s commitment to reduce public expending. In 2000, with an absolute parliamentary majority, and secured total control over the party, Aznar’s personal and political preferences prevailed over internal partisan equilibrium criteria. A general characteristic of PP governments, in contrast to the PSOE, is the overlap between the core of the cabinet and that of the party. In example, in his first government, Aznar appointed the PP general secretary (Álvarez-Cascos) as first Vice-president/Minister of Presidency; and the three vice-general secretaries (Rato, Mayor Oreja and Rajoy) as second Vice-president/Minister of Economy, Minister of Interior and Minister for Public Administration, respectively. The first Zapatero cabinet also had a strong symbolic component. As in Aznar’s governments, ministers were selected less on expertise (Table 6.2) than on political criteria. However, Zapatero appointed two experienced individuals to Foreign Affairs and Economy, with the latter assigned to Pedro Solbes, who had also held such position during the final González government. The composition of the first Zapatero cabinet was a manifestation of both electoral pledges (i.e. its gender parity nature) and the aim of reflecting the PSOE’s internal diversity. Within the cabinet there were members of the executive committee and the parliamentary group, along with representatives of regions where the PSOE was in office – Andalusia, Castilla-La Mancha, Catalonia and Extremadura – as a concession to the party’s more important territorial branches. Apart from some changes derived from electoral promises, the Zapatero cabinet maintained basically the same structure as the last Aznar cabinet, in particular the dual structure of vice-presidencies with portfolio (Presidency and Economy, the former being assigned for the first time to a woman). Cabinet changes also reflect the PM’s autonomy in these stable governments, for the cabinet reshuffles were mainly strategic, decided and controlled by the PM. In this sense, all cabinet reshuffles (except the one in Suárez I and some minor cabinet adjustments) took place in the second half of the constitutional interelection period (Figure 6.2), that is, more than two years after cabinet-formation. The strategic changes mainly aimed at conferring political momentum to the governments for electoral purposes. Notably this may be considered the main

116 The selection of ministers in Europe reason behind the González II 1988, Aznar II 2002, and Zapatero 2007 reshuffles which generally consisted in sacking or shifting the most discredited ministers because of policy failure and poor performance. During the Aznar and Zapatero governments cabinet changes were also used as an instrument linked to the party’s electoral strategy. This was clearly the case of the Aznar’s January 1999 and 2003 reshuffles. The latter prepared for both the November 2003 Catalan and the 2004 national elections. In September 2003 Aznar – who had decided not to run for a third term – designated Mariano Rajoy as his successor as candidate and general secretary of the PP. The second Vice-president and Minister of Economy, Rodrigo Rato, who was also candidate to succeed Aznar, was appointed in compensation first Vice-president, replacing Rajoy, who left the cabinet. Piqué, minister of Science and Technology, also left the cabinet to run for the presidency of the Catalan government. Javier Arenas – who had to quit his position as general secretary of the PP for Rajoy – was rewarded with the second Vice-presidency and Presidency portfolio. Minor cabinet adjustments during the Aznar (except the one in February 2000) and Zapatero governments were also dictated by partisan electoral strategy, where exiting ministers left the cabinet to run either for the presidency of a regional government (three during Aznar II and two in the Zapatero government);10 or, in the case of Loyola de Palacio in the Aznar April 1999 adjustment, to head the PP’s electoral list for the European Parliament elections. Gaining political momentum is not the only reason explaining cabinet changes in these stable governments. For instance, the Zapatero 2006 reshuffle took place just after the ETA terrorist group announced a permanent ceasefire. Under these extraordinary circumstances, the PM replaced the minister of Interior with a person of a higher political profile (Alfredo Pérez Rubalcaba, until then head of the socialist parliamentary group in the Congreso, and former minister under González governments) in order to manage the hypothetical dialogue with the ETA. Finally, the Suárez 1978 and González 1985 reshuffles were mainly related to internal cabinet disputes. Suárez took the resignation of the Minister of Economy (who did not find much support for his initiatives among his cabinet counterparts) as an opportunity to reinforce his dominant position within the cabinet, the reshuffle ending up with five Suárez’s collaborators entering the cabinet and Abril Martorell, appointed Vice-president/Minister of Economy (Lamelas 2004: 171–2; Abella 2006: 314). The González 1985 reshuffle was motivated by discrepancies between the ministers in charge of economic reforms – Boyer (Economy) and Solchaga (Industry) – and the ministers of Transport and Public Works. González tried to solve the crisis by appointing Boyer Vice-president to co-ordinate the entire economic area (González 2003: 816–18). This provoked Vice-president Guerra to threaten to resign, as González’s plans represented for him a loss of power within the cabinet (Guerra 2006: 186–9; García Abad 2006: 395–9). As Boyer learnt that he would not be appointed Vice-president, he resigned – despite the fact that González had dismissed Boyer’s opponents, including the minister of Foreign Affairs. Policy failures, some discontent regarding departmental policy, criticism in parliament or other forms of external shock did not represent a great threat to

Cabinet dynamics in democratic Spain 117 cabinet stability to these governments, mostly because of the large parliamentary and, in some cases, wide public support. Under these conditions, PMs usually protect their discredited ministers. For instance, during the first Suárez democratic government, the impact of policy failures was reduced by the style of consensual politics and the unique conditions of the Transition. This allowed the PM to protect threatened cabinet members, such as the Minister of Interior, Martín Villa, frequently under the opposition’s fire because the terrorist attacks of the ETA as well as the police’s overwhelming repression when facing popular unrest. Similarly, González and Aznar also protected ministers facing major crises, as in the case of the Minister of Education in 1987 and 1988 after massive teachers’ and students’ protests against the department’s policies, the Minister of Labour following the 1988 general strike, or the Ministers of Presidency, Defence and Public Works after the Prestige oil spill crisis in 2002.11 In these circumstances, the ‘sanction’ could be postponed to a later reshuffle, in order to reduce the reinforcing impact of the crisis that a quick response would involve. This ability to absorb external shocks and to protect ministers explains the absence of forced resignations during these governments. The only exception occurred in the first Aznar government, where the Minister of Labour resigned as a result of a scandal related to the allocation of public aids which involved some high officers of his own department.

Cabinet instability In governments characterized by cabinet instability (Suárez II, Calvo-Sotelo and González III and IV) three major factors undermined the PM’s autonomy in the selection and de-selection of ministers: a weak or questioned PM’s leadership of his party, internal party struggles, and the influence of major external shocks (mainly political scandals). The first two factors principally affected the Suárez II and Calvo-Sotelo governments. In the March 1979 election the UCD obtained in the Congreso only two more seats than in 1977 (168, eight MPs below the absolute majority). Once the Constitution had been passed, the new political conditions inevitably led to a period of ‘normal’ politics, putting to an end the consensual manners that had characterized the previous term. Since parliamentary parties’ strategies left no margin either for a coalition government or for post-electoral agreements (Reniu 2002: 204–14), Suárez finally formed a minority government. This new context was unfavourable for Suárez’s leadership style. Previously he had managed to impose his personal leadership over the UCD families – mainly as a result of his wide popularity in steering the transition process – transforming the original coalition into a unitary party. Nevertheless, Suárez – who always privileged his role as PM over that of President of the UCD – left organizational issues to his collaborators (particularly Abril Martorell and the UCD general secretary, Arias Salgado) (Lamelas 2004). After the 1979 election, in order to reinforce the unitary character of the UCD and his personal leadership against the party’s ideological families, Suárez excluded the most important barones from the new government. This was a strange decision, since the composition of the UCD

118 The selection of ministers in Europe parliamentary group still reflected the equilibrium among the foundational parties (Powell 2004: 164). Furthermore, Suárez’s leadership style was at odds with the requirements of parliamentary politics, where he always felt uncomfortable (Martín Villa 1984: 90). This was shown in two of the most important moments of that term, the investiture session in March 1979, and the motion of censure presented by the PSOE in May 1980. On both occasions, Suárez refused to speak before the chamber, being replaced in the debate of the motion of censure by his right-hand man, Vice-president Abril.12 This attitude and Suárez’s growing tendency to delegate to Abril, created an image of political passivity, skilfully exploited by the opposition, particularly by the PSOE, by then the second parliamentary party. During this first ordinary legislative term, Suárez’s popularity also experienced a continuous decay (Gunther 1986: 471) exacerbated by the difficulties of the economic and the political situation (Hopkin 1999: 127ff.), the latter mainly associated with the problems derived from the process of regional devolution and ETA terrorist attacks, which fed military unrest and the threat of a coup d’état. However, the main problems were inside the UCD, with a growing internal opposition to Suárez’s leadership. The PM’s inability to act as a factor of internal cohesion (Powell 2004: 160) set in motion struggles among party factions pervading both the cabinet and the parliamentary group. These conflicts about ideological and organizational issues later evolved into centrifugal forces that caused the disintegration of the UCD (Hopkin 1999; Gunther 1986). The three reshuffles during Suárez II reflected the PM’s progressive lack of control over the government. The one in January 1980 was a consequence of the resignation of the Minister of Culture who disagreed with his party’s position against granting the Andalusian region similar standing to that obtained by the Basque Country, Catalonia and Galicia. The devolution issue also brought the first electoral defeats to the UCD (in the Andalusian referendum,13 and the regional elections in the Basque Country and Catalonia) during the first months of 1980 (Abella 2006: 364–6), provoking a crisis within the cabinet and party that led to the following reshuffle. In May 1980, Suárez’s intention to create a new vice-presidency in charge of regional policy triggered dissensions within the PM’s inner circle, adding to the prior disputes between the Vice-president Abril and two ministers in the social-democrat family. To solve the crisis, Suárez proposed some of the barones to join the cabinet, but they refused as long as Abril still occupied a central position. In the end, the cabinet resulting from the reshuffle left almost everyone unsatisfied. Internal opposition and the motion of censure presented by the PSOE eroded Suárez’s popularity. In addition, his relationship with Abril worsened, ending up with the resignation of the Vice-president in August. The September 1980 reshuffle was Suárez’s last attempt to reverse the UCD crisis, bringing back the main barones into the cabinet. The reshuffle however did not stop internal criticisms but, on the contrary, fuelled the opposition within the UCD parliamentary group, where some relevant members felt excluded by the PM. Unable to cope with both growing internal and external pressures (mainly, economic crisis, terrorism, and military unrest), Suárez resigned in January 1981 (Huneeus 1985: 323–7; Oneto 1981).

Cabinet dynamics in democratic Spain 119 Suárez’s successor, Calvo-Sotelo, was not able to resolve problems. CalvoSotelo had seemed the only candidate with an acceptable profile for the different families in UCD (Christian democrats, liberals, social democrats, and suaristas – Suárez’s supporters) (Calvo-Sotelo 1990). However, he lacked the public charisma and support to construct an effective leadership that could restore both party and governmental stability. At the moment of his investiture, the separation between cabinet and party leadership (the latter still controlled by Suárez’s supporters) made even more difficult the co-ordination between the two and the control of centrifugal forces within the UCD (Huneeus 1985: 352ff.). An early manifestation of these forces was the resignation of the social-democrat baron Fernández Ordóñez to form his own party, which provoked a first reshuffle in September 1981. Three months later after the UCD’s fiasco in the October Galician regional election a new reshuffle was designed to save the critical situation. However, neither Calvo-Sotelo’s election as President of the UCD (Calvo-Sotelo 1990: 71–5) nor the appointment of the baron Martín Villa as Vice-president in order to increase political co-ordination between cabinet and party was enough to reverse the situation (Martín Villa 1984: 117). The last reshuffle, in July 1982 (also after a major electoral defeat in the Andalusian regional election), was a desperate attempt to rescue the UCD from the crisis before the unavoidable call for elections. The party’s electoral strategy prevailed over the cabinet. Some ministers left it in order to occupy relevant positions in the organizational structure to prepare the upcoming national election, and Calvo-Sotelo ceded the UCD presidency to the party candidate, the Christian democrat baron and Speaker of the Congreso, Landelino Lavilla. This could not stop the final disaster, confirmed in the October 1982 election, when the UCD obtained only 11 seats in the Congreso. González III and IV had the basic ingredients characterizing previous PSOE governments. González’s leadership over the party remained practically intact during this period. Besides, majority parliamentary support remained quite stable – the PSOE had an absolute majority in the fourth legislative term14 and a permanent support by the Catalan moderate nationalists of CiU in exchange for regional devolution during the fifth legislative term. Despite these favourable conditions, the González III and IV governments were highly unstable. This was largely due to a number of political scandals affecting ministers and the PSOE organization. In January 1991 Vice-president Guerra resigned after a political scandal involving his brother. Although he remained Vice-general Secretary, Guerra’s exit was seen by some cabinet and PSOE members as an opportunity to shift the power equilibrium within the party organization (Prego 2002: 137). The March 1991 reshuffle was interpreted in terms of the struggle between ‘guerristas’ (Guerra’s supporters) and ‘renovadores’ (the label identifying party members aiming at internal party renewal). González largely maintained the equilibrium between both factions balancing economic control given to Solchaga with promotion of some of Guerra’s supporters. The ‘Guerra affair’ was the first of a number of scandals affecting cabinet members and the PSOE organization between 1991 and 1996. A year after Guerra’s exit the Minister of Health also resigned because of irregularities while he was president of the national company of railways (RENFE). Other scandals

120 The selection of ministers in Europe affecting the PSOE (the ‘Filesa affair’ in May 1991, related to the illegal funding of the party) undermined the party’s public image. In the 1993 election the PSOE lost its absolute majority. The first cabinet tried to reflect the promises of ethical and political renewal made during the electoral campaign. González formed a cabinet with an eminently technical profile, with a lower proportion of members of the socialist parliamentary group, and seven portfolios being occupied by individuals not affiliated to the PSOE. In addition, the ministers from the guerrista faction were all dismissed, while Solchaga was appointed spokesman of the socialist parliamentary group at the Congreso. Despite these changes, cabinet instability continued in the following months. In November 1993, the minister of Interior (in office since 1988) resigned after the Constitutional Court declared unconstitutional part of the Citizen Security Act he had promoted during the previous legislative term. His successor also resigned a few months later as a consequence of the ‘Roldán affair’.15 At the same time, another scandal (the ‘Ibercorp affair’)16 forced the resignation of the recently appointed Minister of Agriculture, along with Solchaga. Both resignations ended up in a cabinet reshuffle only ten months after the formation of the government, where the main novelty was the fusion of the Departments of Justice and Interior, in order to control the corruption cases affecting top officers in the latter, which provoked the resignation of the former Ministers of Interior, Barrionuevo and Corcuera, as MPs. The troubles did not end there, as in July 1995 the Minister of Defence and the Vice-president resigned as a consequence of the illegal wiretappings by the Spanish Centre of Military Intelligence (CESID). 17 The coup de grâce was the re-opening of the judicial investigations on the ‘GAL affair’.18 In September 1995 CiU withdrew its parliamentary support, which hindered the passing of the 1996 budget. Finally, in December 1995 González decided to call for anticipated elections to be held in March 1996, which led to the triumph of the PP.

Conclusions The PM’s institutional centrality is usually argued as one of the manifestations of the ‘executive face’ (Poguntke and Webb 2005: 5) of the presidentialization in the Spanish political system. The analysis of the selection and de-selection of ministers can be, thus, interpreted as a test of this presidentialization hypothesis, given the formal autonomy granted to the PM in this area. Our conclusion agrees with Van Biezen and Hopkin (2005) and Heywood (1991) that the formal autonomy of the Spanish PM as head of the government largely depends on contingent political factors, such as his ability to impose his leadership over his party and gathering parliamentary support. When these elements are present, the PM has a large margin of manoeuvre in deciding cabinet composition, controlling at any moment the tempo of changes and protecting damaged ministers. Thus, the PM’s autonomy usually corresponds to cabinet stability. Nevertheless, as the cases of the González III and IV governments show, even when those conditions are present, exogenous shocks such as political scandals or major policy failures may greatly condition the PM’s decisions by their impact on public opinion and political allies, paving the way for cabinet instability.

Cabinet dynamics in democratic Spain 121

Notes 1 Here we will use both terms President (of the Government) and Prime Minister without distraction. 2 From 1902 to 1923 the average duration of a cabinet (with an unchanged composition) was 7.3 months. During the Second Republic cabinet instability was even higher, with an average duration of 3.4 months between 1931 and 1936 (Linz et al. 2003: 54). 3 There is no mention in the Constitution regarding the procedure of the investiture debate. The first one took place in February 1981 (Calvo-Sotelo), and it was formally established in the Regulations of the Congreso in February 1982. 4 This was the case of Calvo-Sotelo, the only candidate not elected in the first round. 5 Ministers are then not allowed to occupy a specific position (committe chair, etc.) in parliament other than ‘normal MP’. 6 The data used in this empirical section have been compiled by José Real using a variety of sources, such as press (mainly from the daily newspaper El País, the Lexis-Nexis database and the Linz’s Archive of the Spanish Transition, available online at www.march.es/ceacs/linz/); bibliographical sources (published biographies, political essays, academic monographs); and the database of the ELIPARSIS (Elites, partidos y systemas de partidos en España, 1868–2004) Project, funded by the Spanish Interministerial Committee for Science and Technology (BSO2003-07384). 7 Through the rest of the chapter, the term ‘government’ is used in the specific sense of prime ministerial term (that is, the period between two successive investitures, i.e. González I, Aznar II, etc.). Conversely, ‘cabinet’ refers to the specific ministerial composition of the government between events of change (i.e. during the González II government there were two cabinets, one after government formation, and the second after the July 1988 reshuffle). 8 Duration in ministerial position refers to the time a cabinet member keeps the same portfolio during a prime ministerial term. The term is considered to end the day before the next national election or, in case of government crisis, the election of the new PM. 9 The 2004 election took place only three days after the terrorist massacre of 11 March 2004. After the attack, the Aznar government blamed the Basque terrorist group ETA, thereby reinforcing PP’s anti-nationalist discourse, even though police investigations pointed to the Islamist track owing to Al-Qaeda’s repeated threats related to Aznar’s outstanding support to the Iraq invasion. This fed public suspicion that the government was manipulating the information in order to avoid an electoral defeat, which contributed decisively to increase turnout (particularly within the centre-left electorate) and strategic voting against the PP, allowing against all odds (polls were not favourable during the campaign) the PSOE to win the election (see Torcal and Rico 2004). 10 In contrast, during the González governments, only one minister – Manuel Chaves – left the cabinet to run for the presidency of the Andalusian government. 11 In November 2002, the sinking of the mono-hull tanker Prestige off the Galician coast provoked a massive oil spill that caused the greatest ecologic catastrophe in recent Spanish history. The questionable management of the crisis by the PP government triggered a wave of protests and social mobilization. 12 During the investiture, Suárez also refused to debate his programme with the rest of the parliamentary forces, behaviour that raised criticisms not only from the opposition parties but also from his own parliamentary group, including the Speaker (CalvoSotelo 1990: 39–40; Martín Villa 1984: 89–90; Oneto 1981: 50). 13 Held in February 1980, this referendum was aimed at deciding whether Andalusia would constitute its autonomy institutions in a similar manner to the so-called ‘historical’ regions. The UCD asked its voters to abstain, while the PSOE, the Communist Party and the Andalusian Party asked for an affirmative vote. 14 In fact, in the 1989 election the PSOE obtained 175 seats at the Congreso, one below the absolute majority. However, the absence during the term (for political reasons, in

122 The selection of ministers in Europe

15 16

17 18

principle related to the oath national MPs have to take) of the four MPs of the proindependence Basque party Herri Batasuna (HB) allowed the PSOE to keep its absolute majority in the lower chamber during the term. This affair refers to the illegal enrichment of the General Director of the Guardia Civil (one of the two major Spanish security corps), Luis Roldán. Being spotted by the press, Roldán escaped abroad, which motivated the resignation of the Minister of Interior. Use of privileged information in the stock market by the managers of the Ibercorp Group while Solchaga was minister of Economy. Amongst beneficiaries were the governor of the Spanish Central Bank and other relevant personalities close to the PSOE, such as the former Minister of Economy, Miguel Boyer. Amongst the wiretapped personalities were the King himself and party officials. This affair concerned the participation of top officials from the Ministry of Interior in the creation and funding of the Antiterrorist Groups for Liberation (GAL), which between 1983 and 1987 committed a series of terrorist attacks and assassinations against the environment of the Basque pro-independence left and the ETA terrorist group (see Magone 2004).

References Abella, C. (2006) Adolfo Suárez. El hombre clave de la Transición. Madrid: Espasa. Aragón, M. and Gómez, A. J. (coords) (2005) El Gobierno. Problemas constitucionales. Madrid: Centro de Estudios Políticos y Constitucionales. Bar, A. (1997) ‘Spain: A Prime Ministerial Government’. In Blondel, J. and MüllerRommel, F. (eds) Cabinets in Western Europe, 2nd edition. London: Macmillan, 116–135. Botella, J. (1997) ‘La élite gubernamentale espagnole’. In Mendras, H. and Suleiman, E. (eds), Le recrutement des élites en Europe. Paris: La Découverte. Calvo-Sotelo, L. (1990) Memoria viva de la transición. Barcelona: Plaza & Janés. García Abad, J. (2006) Las mil caras de Felipe González. Madrid: La Esfera de los Libros. García-Guereta, E. (2001) Factores externos e internos en la transformación de los partidos políticos: el caso de AP-PP. Madrid: Centro de Estudios Avanzados en Ciencias Sociales. González, F. (2003) ‘La rebelión contra la historia’. In Iglesias, M. A. (ed.) La memoria recuperada. Madrid: Aguilar, 789–875. Guerra, A. (2006) Dejando atrás los vientos. Memorias 1982–1991. Madrid: Espasa Calpe. Gunther, R. (1986) ‘El hundimiento de UCD’. In Linz, J. J. and Montero, J. R. (eds) Crisis y cambio: electores y partidos en la España de los años ochenta. Madrid: Centro de Estudios Constitucionales, 433–492. Heywood, P. (1991) ‘Governing a New Democracy: The Power of the Prime Minister in Spain’. West European Politics 14, 2: 97–115. Heywood, P. (1995) The Government and Politics of Spain. Basingstoke: Macmillan. Heywood, P. and Molina, I. (2000) ‘A Quasi-presidential Premiership: Administering the Executive Summit in Spain’. In Peters, G. B., Rhodes, R. and Wright, V. (eds) Administering the Summit: Administration of the Core Executive in Developed Countries. London: Sage. Hopkin, J. (1999) Party Formation and Democratic Transition in Spain. The Creation and Collapse of the Union of the Democratic Centre. Basingstoke: Macmillan. Huneeus, C. (1985) La Unión de Centro Democrático y la transición a la democracia en España. Madrid: Centro de Investigaciones Sociológicas. Lamelas, A. (2004) La Transición en Abril. Barcelona: Ariel.

Cabinet dynamics in democratic Spain 123 Linz, J. J. (2001) ‘El liderazgo innovador en la transición a la democracia y en una nueva democracia’. In Alcántara, M. and Martínez, A. (eds) Política y gobierno en España, 2nd edition. Valencia: Tirant Lo Blanch, 57–97. Linz, J. J. and Stepan, A. (1996) Problems of Democratic Transition and Consolidation. Baltimore: The Johns Hopkins University Press. Linz, J. J. and Jerez-Mir, M. (with the collaboration of S. Corzo) (2003) ‘Ministers and Regimes in Spain: from the First to the Second Restoration, 1784–2002’. In Tavares de Almeida, P. et al. (eds) Who Governs Southern Europe? Regime Change and Ministerial Recruitment, 1850–2000. London: Frank Cass, 41–116. López Calvo, J. (1996) La organización y funcionamiento del gobierno. Madrid: Tecnos. López Garrido, D. (1985) ‘Gobierno y Parlamento: dos modelos de relaciones internas (UCD y PSOE)’. In Alonso García, E. et al., El Gobierno en la Constitución Española y en los Estatutos de Autonomía. Barcelona: Diputació de Barcelona. Magone, J. M. (2004) Contemporary Spanish Politics. London: Routledge. Maravall, J. M. (2003) ‘Una peligrosa manera de pensar’. In Iglesias, M. A. (ed.) La memoria recuperada. Madrid: Aguilar, 33–92. Martín Villa, R. (1984) Al servicio del Estado. Barcelona: Planeta. Méndez Lago, M. (2000) La estrategia organizativa del Partido Socialista Obrero Español (1975–1996). Madrid: CIS. Montabes, J. (2001) ‘El Gobierno’. In Alcántara, M. and Martínez, A. (eds) Política y Gobierno en España, 2nd edition. Valencia: Tirant Lo Blanch. Montero Gibert, J. R. and García Morillo, J. (1984) El control parlamentario. Madrid: Tecnos. Olias, B. (1994) ‘Los gabinetes de los Presidentes del Gobierno en España’. Política y Sociedad 16: 257–272. Oneto, J. (1981) Los últimos días de un presidente. De la dimisión al golpe de Estado. Barcelona: Planeta. Pérez Francesch, J. L. (1998) El Gobierno. Madrid: Tecnos. Poguntke, T. and Webb, P. (2005) ‘The Presidentialization of Politics in Democratic Societies: A Framework for Analysis’. In Poguntke, T. and Webb, P. (eds) The Presidentialization of Politics. Oxford: Oxford University Press, 1–25. Powell, C. (2004) ‘Adolfo Suárez. El presidente que se hizo a sí mismo’. In Powell, C. and Bonin, P. Adolfo Suárez González. Barcelona: Ediciones B, 41–198. Prego, V. (2002) Felipe González. El presidente del cambio (1982–1996). Madrid: El Mundo/Unidad Editorial. Reniu, J. M. (2002) La formación de gobiernos minoritarios en España 1977–1996. Madrid: Centro de Investigaciones Sociológicas. Rodríguez Aizpeolea, L. (2004) Ciudadano Zapatero. Madrid: Espejo de Tinta. Rodríguez Teruel, J. (2004) Los ministros de la España democrática. Perfil, trayectorias y carrera ministerial de los miembros de Gobierno de Suárez a Zapatero (1976–2005). Unpublished doctoral dissertation. Universidad Autónoma de Barcelona. Román, P. (2002) ‘El Gobierno’. In Román, P. (ed.) Sistema político español. Madrid: MacGraw-Hill. Sánchez Cervelló, J. (2004) Felipe González. In Sánchez Cervelló, J. and Tubau, I. Felipe González Márquez. Barcelona: Ediciones B, 35–135.. Torcal, M. and Rico, G. (2004) ‘The 2004 Spanish General Election: In the Shadow of Al-Qaeda?’ South European Society & Politics 9, 3 (Winter): 107–121. Tusell, J. (2005) El aznarato. El gobierno del Partido Popular, 1996–2003. Madrid: Punto de Lectura. Van Biezen, I. and Hopkin, J. (2005) ‘The Presidentialization of Spanish Democracy:

124 The selection of ministers in Europe Sources of Prime Ministerial Power in Post-Franco Spain’. In Poguntke, T. and Webb, P. (eds) The Presidentialization of Politics. Oxford: Oxford University Press, 107–127. White, G. (2000) ‘Ministerial Musical Chairs: Patterns of Provincial Cabinet Shuffles’. Paper presented at the Annual General Meeting of the CPSA, Quebec City, 31 May. Zapatero, V. (2003) ‘El liderazgo del Gobierno’. In Iglesias, M. A. (ed.) La memoria recuperada. Madrid: Aguilar, 927–934.

7

Belgium Ups and downs of ministerial careers in a partitocratic federal state Patrick Dumont, Stefaan Fiers and Régis Dandoy

Constitutional framework Belgium is a constitutional monarchy. King Albert II succeeded his elder brother Baudouin (1951–1993) whose long reign followed difficult times for the royalty after the Second World War. Although the King is formally part of both the legislative and executive powers (but his acts are always since 1831 countersigned by a minister), his role in law-making and the day-to-day running of executives has at least since the Second World War been greatly reduced. 1 The King’s political role in contemporary Belgium is confined to cabinet formation and termination, during which he is given more leeway to take action aimed at stimulating political representatives to find a solution acceptable to a parliamentary majority. Even in this limited – but eminently important for a society characterized by multiple divisions and policy dimensions with at times extreme cabinet instability – remit, the King has increasingly been constrained by political parties’ strategies and agendas, as recently shown by the six-month-long cabinet formation of 2007. The current Constitution still stipulates that ‘the King appoints and revokes his ministers’, but since 1945 only the personality and longevity of King Baudouin made at times this formal power effective. Belgian parliamentary democracy as defined in the 1831 Constitution rested on a symmetrical bicameral system which gave the House of Representatives and the Senate equal powers regarding the making and control of governments as well as in law-making. The 1993 constitutional reform, which transformed the unitary state into a complex federal system (see below) ended this symmetry. Since its implementation in 1995, federal governments are accountable only to the House of Representatives. First, after the debate over the governmental declaration read by the new PM, a confidence vote is taken through which the government must receive the support of a majority of MPs voting (blanks excluded);2 whereas before 1995 this took place in both assemblies, since 1995 only the House of Representatives is asked to grant or not its confidence.3 Secondly, only members of the House can conduct interpellations that may lead to a vote on a no confidence motion targeted at an individual minister. Thirdly, only the House votes on motions of confidence and no confidence targeted at the government as a whole. Two innovations aimed at assuring greater stability to Belgian executives were introduced to prevent a few rebels from majority parties’ ranks from ‘accidentally’ bringing

126 The selection of ministers in Europe down a government on significant bills by voting against it. First, the adoption of a motion of no confidence by an absolute majority of Representatives, proposing at the same time the appointment of a new Prime Minister to the King, obliges the incumbent government to resign (constructive motion of no confidence). Secondly, when an absolute majority of the House rejects a motion of confidence introduced by the government and within three days passes a motion proposing a new PM, the government has to resign. Failing to find an alternative Prime Minister, Representatives may allow the incumbent government to continue or ask the monarch to dissolve Parliament and call for new elections. Through this constitutional reform the Senate has lost other important competencies as it is not involved in the approval or control of the budget and the House has the final say on most legislation. The major 1993 constitutional reform marked the passage from a unitary state to a fully fledged federal system based on three linguistic Communities (Flemish, French-speaking, and German-speaking) and three Regions (Flanders, Brusselscapital and Wallonia). From 1970 onwards institutional changes had gradually led to the creation of an autonomous political class at the sub-national level (Fiers 2001). In 1971 the so-called Cultural Councils, composed of members of the House of Representatives and directly elected senators, categorised by their membership of one of the two main language Communities, were created. Contrary to the latter, for which MPs have constantly been indirectly elected, (coming either from the national parliament or since 1995 from the regional ones), the German-speaking Community directly elected its own councillors – with only an advisory consultative role and no legislative power until 1980 – from 1974 onwards. These Councils enjoyed only limited competencies on cultural affairs, language use and certain aspects of education. The installation of the Regions took ten more years, with a transitional phase starting in 1974 with consultative Regional Councils composed of senators of the three respective regions but also regional ministerial committees within the central (national) government. Starting in December 1981 autonomous regional and community governments were created following the second important constitutional reform. Since 1995 all regional parliaments are made up of directly elected MPs (for a five-year fixed term as opposed to federal MPs with four-year terms); the ‘dual mandate’ therefore no longer exists. By now there is therefore an autonomous sub-national political class passing legislation – and there is no hierarchy of norms between federal laws and regional or community decrees – in policy fields delimited in the Constitution (or in specific bills voted on by a qualified majority, see below)4 which altogether represent about one-third of overall public spending. Executives headed by Minister-Presidents are elected by their respective parliament, and have their own civil service. This process has also led to more opportunities for professional politicians, as in two decades (1980–99) the aggregate number of executive positions at the federal and regional level increased by half (from 36 to 54). Being awarded a ministerial post at the federal level may not be the highest reward for Belgian politicians anymore, as for many of them the federal level is no longer the main locus of power. When cultural autonomy was granted to Communities to satisfy Flemish demands,5 protections for the French-speaking minority were constitutionally

Belgium: ups and downs of ministerial careers 127 introduced in 1970. Three measures were taken, necessitating defining linguistic groups in the House and the Senate. First, bills regarding the borders, competencies and statutes of the Communities and the Regions have to be voted on by a twothirds majority of the valid votes cast, and a majority of valid votes of each linguistic group in each House (with a majority of members of each linguistic group present). Secondly, an ‘alarm bell’ procedure is set up to delay or block legislation threatening the interests of a linguistic community. In both Chambers, if threequarters of the members of a linguistic group introduce such a motion after the committee report on a bill and before the final plenary vote, the legislative procedure is suspended and the issue dealt with by the Council of Ministers within thirty days. Thirdly, linguistic parity is imposed in the Council of Ministers with the Prime Minister possibly excepted from the counting.6 This was the first time, 140 years after Belgium’s independence, that the position of Prime Minister was enshrined in the Constitution, together with the first reference to the main collective decision-making body, the Council of Ministers, which does not include junior ministers. From 1995 ministers could not hold parliamentary seats though they may hold a ‘sleeping mandate’ (an MP promoted to the government may take her parliamentary seat back when her ministerial time is over). A clearer distinction between the executive and its controller, here the House of Representatives, is therefore introduced, following the principle of the separation of powers and intended at changing the image of an unaccountable political clique. Also the number of ministers is set at a maximum of 15,7 reflecting in part the loss of policy competencies to Regions and Communities but also ending the inflation in size of the cabinet according to the number of coalition parties (Frognier 1993; Mershon 2001).8 Since a 2002 constitutional amendment the Council of Ministers must include members of both sexes. There is no quota specified for Secretaries of State,9 nor for the members of the so-called Kerncabinet, an inner cabinet committee that meets regularly and has become the main locus of power at the federal level. It resolves major conflicts between the coalition parties but also takes all important decisions formally ratified by the full Council. The Kerncabinet is made up of the PM and the Vice-PMs who, as leaders of their party delegations in government, act as guardians of the coalition agreement and monitor the actions of ministers of the other parties.10 The process of cabinet formation is not spelled out in the Constitution, but strong customary practice has developed a ritual coming in successive phases. It starts with the King’s consultations, usually on the day after elections, and the presentation of the government’s resignation by the PM. The King usually accepts it and charges the incumbent government with the implementation of ‘current’ affairs. He then consults the Chairs of the two Chambers, the presidents of all democratic parliamentary parties and socio-economic corporatist organizations. Since 1831 the King then appoints a formateur, generally the leader of the largest party, with the mission of forming a cabinet. Since 1935 the King has sometimes appointed an informateur first, when the political situation is exceptionally difficult and a first round of talks between parties is needed to inform the King of the possible coalition

128 The selection of ministers in Europe formulas.11 When a formateur likely to be successful is finally found, party delegations of a specific coalition formula convene to negotiate and draft a usually quite comprehensive coalition policy programme, which since the 1960s is rendered public when signed by all partners and is for the life of the cabinet considered as the governmental bible (De Winter et al. 2003). If the negotiations collapse, the process starts again with the King’s consultations and the appointment of an informateur or a new formateur. When eventually signed by the leaders of the partner parties, the governmental agreement is then presented to each of the coalition parties’ congresses to be voted on by these sovereign party bodies, so that not only do party leaders commit to the negotiated deal but all intraparty components of each partner are bound by the contract. Portfolio distribution is the final stage in the formation process. Party leaders with the help of the formateur – the Prime Minister to be12 – first distribute the portfolios amongst themselves, taking into account the constitutional constraints described earlier and within these constraints adopting a pure proportional key to allocate executive seats between parties, then the names of ministers are proposed to the King. Once the PM and ministers have been sworn in by the King, there is the customary parliamentary investiture won by all governments except one since 1945. On average it took 33 days to form the 37 governments of the the postwar period (from 1946 to 2003), but formation duration rises if we consider only post-electoral formations (52 days) or the post-1968 period where the number of parties required for a majority rose (the average number of coalition parties rose from 1.8 before 1968 to 4.4 after) and institutional concerns became more relevant (43 days and no fewer than 66 days for post-electoral governments). The 2007 government formation crisis set a new record of 193 days to form a tentative cabinet designed for only three months.13

Choosing ministers The selection process The influence of the Head of State in government formation has clearly decreased since 1831. Whereas King Leopold I drafted a governmental pact with the person he appointed formateur and sometimes suggested individuals as ministers and picked his Minister of War, his successors never had the same impact. Since the end of the nineteenth century ministers are not considered as mere servants of the King but as leaders of the parliamentary majority. During his long postwar reign Baudouin discreetly and sporadically influenced personnel selection.14 The atypical 2007 government formation process put royal preferences – supposed to be kept secret – in the limelight, causing press debate about the role of the monarch. Ironically, Albert II had previously been perceived as less interventionist than his predecessor.15 Since the Second World War Belgium has become a ‘partitocracy’ (Dewachter and De Winter 1981; De Winter et al. 1996; De Winter and Dumont 2006): a form of party government characterized by the overwhelming role and omnipresence of disciplined political parties. From the late 1950s onwards party executives have

Belgium: ups and downs of ministerial careers 129 increasingly dominated parliamentary groups and influenced cabinet decisionmaking, sometimes coming to decisions (political pacts) drafted outside of cabinet and parliament and amendable by neither of them. The detailed coalition agreements have not been voted down by any party congress in the postwar era (for potential reasons see De Winter and Dumont 2006: 960). Rather than being policy dictators (Laver and Shepsle 1996), cabinet ministers are reduced to pure party agents (De Winter and Dumont 2003; 2006) as they are supposed to follow scrupulously the coalition agreement and keep the party line for issues falling outside this document. They also report almost constantly to their party executive and more specifically to their party president, who often does not enter government, preferring to monitor (and, since the arrival of autonomous regional governments, increasingly co-ordinate) the activity of his ministers. In practice, after the party congresses have voted on participation, the distribution of portfolios amongst coalition partners (the formateur and the party presidents) is decided first. From 1980 a weighting rule appears to have developed, aimed at compensating for the constitutionally required parity of ministers in the Council by awarding points to specific positions and enlarging the number of positions bargained over. The PM counts for three points, ministers and chairs of the two federal Chambers (as well as the position of European Commissioner) for two, and junior ministers only one. Applying the D’Hondt method to allocate points in proportion to the parliamentary seat contribution of each coalition partner, each party knows rather precisely the amount of office pay-offs it can claim. Each party leader is then invited to state her first portfolio preference, starting with the largest party (expected to claim the PM usually going to the formateur) and then every other partner in rank order of parliamentary strength. After this first round a second starts with the largest party, until all points have been allocated. After a first evaluation of the result, multilateral or bilateral talks may ensue until all partners are satisfied. More positions may enter the game (increasing the number of junior ministers or taking sub-national positions into account) or portfolio competencies alter. The formateur is the co-ordinator of this complex allocation process between parties. Although also responsible for presenting a list of ministers to the King, he does not have much leeway regarding names proposed by other coalition partners.16 Even with regard to his own party, the choices of the party president (if he is not the formateur) may prevail. Size of government and political completion Since the war most Belgian governments have been coalitions with as many as 36 members (the Leburton five-party coalition of 1973 and the Martens III sixparty coalition of 1980). Since the Second World War 326 individuals have been ministers or junior ministers, filling a total of 965 ministerial appointments of which 882 were made at the start of a new government (thus 83 substitutions or new positions made during the course of governments).17 Almost half have been filled by politicians from the Christian Democratic parties, one third by Socialists, less than one-fifth from the Liberals. The dominance by the Christian Democratics is due to their electoral strength and position at the centre of the socio-economic

130 The selection of ministers in Europe Table 7.1 Distribution of ministerial positions by party in Belgium (days) (percentages given in parenthesis), 1946–2007

Christian Democrats Socialists Liberals Flemish nationalists French nationalists Greens Communists Technicians Total

All positions (%)

At start of the government (%)

Individuals (%)

448 296 165 11 14 5 8 18 965

425 (48.2) 263 (29.8) 144 (16.3) 10 (1.1) 12 (1.4) 4 (0.5) 8 (0.9) 16 (1.8) 882

116 104 75 8 8 5 4 9 329

(46.4) (30.7) (17.1) (1.1) (1.5) (0.5) (0.8) (1.9)

(35.3) (31.6) (22.8) (2.4) (2.4) (1.5) (1.2) (2.7)

Average duration per individual 5.1 4.0 3.5 2.3 2.2 3.1 1.4 1.4

Note: 326 individuals have become minister or junior but three have served in cabinet for two different parties. They have been counted twice in this table, and therefore the number of individuals on party basis is 329 instead of 326.

divide permitting greater presence in power (in 31 governments, compared to 25 for Socialists and 17 for Liberals) and also dominance in cabinet (reflected usually in the PM but also in the number of ministerial positions). Christian Democratic ministers also stay longer in office whilst Liberals resign more frequently than ministers from the other two main party groups. The difference materializes also in terms of cabinet rank: 70 per cent of the Socialist and Christian Democratic cabinet positions are ministers, 19 per cent junior ministers and 11 per cent PM and Vice-PM. Positions for the Liberals are respectively 64 per cent, 27 per cent and 8 per cent. Earlier governments did not have as many Vice-PMs, and Liberals remained out of office from 1987 to 1999. With the exception of technicians and Communists (all at minister rank) who were recruited early in the period, smaller parties have a smaller share of ministers (25–63 per cent), and a larger proportion of juniors (33–50 per cent). Only twice (Flemish Nationalists and Greens) has a Vice-PM not come from one of the three traditional parties. The profile of Belgian ministers Though party presidents have discretion over choice of ministers, 18 they face intra-party constraints (Fiers 1998). Most visibly these come from internal factions, such as the standen (estates) in the Flemish Christian Democratic party, which has been in power from 1947 to 1954 and from 1958 to 1999. There, the leaders of the three main factions (trades unions, farmers’ league and middle-class associations) hand in their list of ministrables in preference order so that the party president is charged with balancing the internal forces amongst factions and ‘non-aligned’ heavyweights. The balance has to reflect the overall strength of factions but sometimes leaders try to over-represent some factions to compensate choices made

Belgium: ups and downs of ministerial careers 131 in coalition partners. This helps appease potential internal turmoil and prevent the cabinet from being too leftist or rightist.19 Party presidents also have to balance experience (outgoing ministers, heavyweights) and new faces. The geographic representation of provinces, of powerful constituencies and the proportion of female ministers are also considered by party presidents, who also have face tradeoffs in terms of expertise and more general profiles. Some form of balance between Representatives and Senators has generally been sought, although nowadays with new pools of ministrables at the sub-national level less attention is paid to this. Four broad aspects of the profile of ministers will be empirically assessed. First, 55 per cent of the ministers in the 16 governments between 1946 and 1970 were French-speaking. The first postwar cabinet where Flemish ministers outnumbered French-speaking ones, thereby reflecting the demographic reality of the country, was led by Gaston Eyskens in 1958; cabinets formed during the 1960s were roughly based on linguistic parity. From 1970 to 2007 however, because most PMs were Flemish (and in the event of a French-speaking minister the latter was excluded from the computation of the parity in the Council), 53 per cent of the senior ministers appointed were Flemish.20 The nomination of junior ministers has also been used to translate the dominance of Flemish population into a larger number of governmental positions: from the setting up of the parity rule in 1970, the proportion of Flemish junior ministers is 59 per cent, thereby clearly indicating the effectiveness of the compensation mechanism provided by the unconstrained nomination of junior ministers.21 Note that on several occasions the parity rule has led to ‘forced’ appointments or movements. For example in 1974, when the Walloon regionalists enlarged the Tindemans government, the nomination of a French-speaking newcomer in the Council of Ministers implied the appointment of a new Flemish minister; in 1977, when the French-speaking nationalists were revoked, two politicians of the remaining French-speaking partners had to be promoted minister. Secondly, Belgian politics has always suffered from poor female representation. Women first stood for election in 1921, and have been heavily under-represented in parliament (Fiers and Gubin 2003). This striking under-representation particularly holds true for the executive positions. In the postwar period 30 women (9 per cent of all) have served as minister or junior minister. The first female minister was selected in 1965, when Marguerite De Riemacker-Legot (CVP) became minister of Family and Housing. In the first 40 years after the Second World War, only three women were appointed as full minister (Das 1987), the others being awarded only junior positions. Only three women have been made Vice-PM, and Laurette Onkelinx (PS) was the first in 2003 to keep this position for more than one term. Out of 85 appointments amongst these 30 individuals, 59 portfolios belonged to the ‘socio-cultural’ category (BEIS-categorization, see Berkmann, n.d.)22 whilst only 11 were in Infrastructure and 12 in Economy23 and merely three in basic functions of the State (Interior, Foreign Affairs, Justice and Defence). Yet we find a clear progress in the presence of female politicians in the cabinet. Following the 2002 obligation of appointing at least one woman in government, no fewer than one-third of the ministers appointed during Verhofstadt II (2003–7) were women.

132 The selection of ministers in Europe The average age of cabinet members at their first appointment is 47.9 years, with an average parliamentary experience of exactly seven years; socialist parties traditionally select younger politicians (46.4 years on average) than Liberals (48.4) and Christian Democrats (48.5).24 With the exception of the Communists (whose limited participation in government dates from shortly after the Second World War), smaller parties typically choose experienced heavyweights to enter the cabinet (for instance the average Flemish nationalist minister was first appointed at over 49, after 8.7 years spent in parliament). When we take into account every cabinet member at the start of a government, as in Table 7.2, the average age rises to 50.5 years. However, there is considerable variation in average age according to ministerial position, PMs typically being older and more experienced. Over the years, the average age of members of cabinet remained stable until the 1990s (Figure 7.1) when the average age reduced, leading to a remarkably young Verhofstadt II cabinet in 2003, with an average age of 42.5 years. The two peaks of ‘old’ cabinets with numerous heavyweights in the 1950s can be related to the particular circumstances in which they were installed, with the King’s crisis in 1949–50 and the first anti-Christian Democrats coalition in 1954. Other ‘experienced’ cabinets are mere temporary continuations of previous governments after one party left the coalition (in the 1980s). Parliamentary experience remains as one of the necessary prerequisites to be selected for a governmental position (Blondel 1988; De Winter et al. 1996). After 1946, 81 per cent of ministers were federal MP at the time of their first appointment and overall 92 per cent of ministerial positions at the start of the cabinets were attributed to individuals with a federal parliamentary seat. Interestingly, throughout the period, ministers appointed during the lifetime of a government (in the context of a large reshuffle or to replace a resigning minister) on the other hand display a distinct profile, as no fewer than one-third of them did not come from the federal parliament. Party presidents often choose to launch the career of less experienced individuals as these new ministers are slightly younger and have less executive experience (one year as opposed to three for ministers chosen at the start of a government). We also notice a clear evolution in time:25 overall, no fewer than 35 per cent of ministers and junior ministers selected since the beginning of the 1990s were not federal MPs at time of first appointment whilst at start of government the proportion of extra-federal parliamentarians has tripled with regard to the 1970–85 period. These changes include appointing both from regional assemblies Table 7.2 Average age and experience per ministerial position in Belgium, 1946–2007

Overall totals Prime Minister Vice-Prime Minister Minister Junior minister

Number

Average age

Average parliamentary experience

Average ministerial experience

882 37 58 611 176

50.5 52.1 48.2 50.7 50.3

10.6 15.1 13.0 10.5 9.0

2.7 5.5 4.1 2.7 1.6

Huysmans

Eyskens

Pholien

Eyskens II

Hamel

Eyskens IV

Lebutton

Age Parliamentary exp Ministerial exp

Figure 7.1 Average age and political experience at the start of each cabinet in Belgium, 1946–2007

0.0

10.0

20.0

30.0

40.0

50.0

60.0

Spaak Van Acker III

Spaak II Spaak III

Duvieusart

Van Houtte Van Acker IV

Eyskens III Lefevre

Van den Boeynants I

Eyskens V

Tindemans I Tindemans II Tindemans III Tindemans IV Van den Boeynants II Martens I Martens II Martens III Martens IV M Eyskens Martens V Martens VI Martens VII Martens VIII Martens IX Dehaene I Dehaene II Verhofstadt I Verhofstadt II

134 The selection of ministers in Europe (17 per cent of first appointments since 1999) and from outside (28 per cent in the same period, a rate more typical of the 1940s and 1950s). In the eyes of their party president, MPs’ qualities may lie in their knowledge of parliamentary routines, their capacities at handling multiple policy fields but also, in the context of coalition governments, their socialization amongst MPs of other parties. In the aftermath of the Second World War the appointment of extra-parliamentary specialists has been valued, for instance for the Ministries of Defence and Economic Recovery. With complex issues constrained by the European Union, party presidents are nowadays also tempted to recruit outside the parliamentary pool. Such ministers might be easier to control as they owe more to the party president in particular (Poguntke and Webb 2005) though their informational advantage might work in the opposite direction. Aside from specialists, party presidents may want to recruit outside MP ranks to impose the figure of a strong, independent-minded leadership to their party troops. This may also be a risky move, especially after an electoral defeat, triggering rebellion amongst party MPs, as experienced by the French-speaking Socialist party president in 2008. Notwithstanding the informal rule that ministers are recruited from both the Senate and the House, the latter dominates (see Figure 7.2). On six occasions more than seven out of ten ministers stemmed from the House of Representatives whereas the highest share of senators in a cabinet (Martens III) was a moderate 44 per cent. This situation did not change dramatically after the institutional changes of 1995, which also enlarged the Senate electoral districts, making elections for this assembly a semi-direct election for the position of PM as each party list is pulled by its electoral heavyweight. The level of experience required depends on the job. There is a clear hierarchy between the various ministerial departments, with Foreign Affairs, Justice, Defence and Social Affairs/Public Health being typically filled by professional politicians with long years of parliamentary and ministerial experience. Other departments (Culture, Commerce, Colonies, Regional Affairs, Co-operation and Development) serve as training grounds for relatively young and less experienced politicians or – as in the case of Colonies – are awarded to technicians without any parliamentary experience. Table 7.3 also includes ‘regional affairs’, which consisted of a large number of positions within the federal government in the 1970s but with only sub-national competencies. These portfolios typically went to less experienced ministers, some of whom remained in office once the regional governments became fully autonomous (from 1981 onwards). Rather than creating linear careers from regions to the national level, federal Belgium has triggered quite unpredictable courses, as an increasing number of individuals have moved back and forth between levels of government making ministerial lives full of ups and downs. Finally, for an archetypical partitocracy, it is not surprising that, since the late 1970s, a majority of first appointees have been members of their party executive. A change of coalition partner typically provokes a rush of members of the party executive to the government, as the heavyweights of the junior party, who sometimes were ministers before their party went to opposition, enter cabinet (Martens V and Martens VIII are cases in point). In 1981 party presidents jointly

Huysmans

Eyskens

Pholien

Eyskens II

Lefevre

Extra parliament

Senate

House

Leburton

Regional

Figure 7.2 Parliamentary origin of ministers in Belgium, 1946–2007 1946–2007

0

10

20

30

40

50

60

70

80

90

100

Spaak Van Acker III

Spaak II Spaak III

Duvieusart

Van Houtte Van Acker IV

Eyskens III

Harmel Van den Boeynants I Eyskens IV Eyskens V

Tindemans I Tindemans II Tindemans III Tindemans IV Van den Boeynants II Martens I Martens II Martens III Martens IV M Eyskens Martens V Martens VI Martens VII Martens VIII Martens IX Dehaene I Dehaene II Verhofstadt I Verhofstadt II

136 The selection of ministers in Europe Table 7.3 Average age and political experience of ministers in Belgium according to ministerial departments, 1946–2007

Department

Number

Average age

Foreign affairs Agriculture and fishing Post telecom Justice Defence Social/health/family Public works Interior Education Labour/employment Economy Civil service Budget Transport/marine Finance/treasury Foreign trade Construction/housing International aid Regional affairs Colonies & overseas Commerce Culture

37 31 14 37 38 57 33 37 46 38 40 15 20 33 38 25 12 11 30 11 15 15

53.9 57.1 49.3 52.2 51.9 51.4 49.3 50.5 50.3 49.9 46.2 49.2 46.5 49.5 50.0 51.0 44.7 48.0 48.4 49.3 49.8 49.2

Average parliamentary experience

Average ministerial experience

17.2 14.8 13.9 13.2 12.3 11.9 11.7 11.5 11.1 10.9 10.0 9.9 9.5 8.8 8.4 8.2 8.1 7.8 7.4 6.9 6.8 6.6

7.3 3.5 2.6 3.5 3.6 3.4 2.5 2.1 1.7 3.1 2.6 1.7 2.5 3.3 3.7 2.3 1.0 1.7 1.8 1.1 0.4 1.8

Note: This table does not apply to junior ministers. We have listed only a selection of ministerial departments, and taken into account only the ministers at the start of the government.

decided to become cabinet ministers in order to maximize cabinet durability of Martens V after a period of instability (eight governments between 1977 and 1981). The unprecedented Liberal-Socialist-Green coalition of Verhofstadt I (1999–2003) was also almost exclusively composed of party executive members. Dejaeghere (2007) points out that the Socialists (especially the Flemish-speaking Socialists) heavily relied on their party executive as a pool of ministrables, though our own data show that over the years this phenomenon has become more widespread. By splitting traditional unitary parties, the number of party executive members has increased, opening their doors to lower-level party politicians, whilst the number of cabinet positions available for each party has been reduced by the greater number of coalition partners and the restriction of ministerial positions. Together with the need for selecting heavyweights able to defend party positions and assure cabinet stability, and notwithstanding the recent temptation to recruit outside the traditional parliamentary pools, these elements account for the fact that being member of the party executive has increasingly become an important prerequisite for selection.

Belgium: ups and downs of ministerial careers 137

Ministerial turnover According to Höjer (1969: 315), interwar politics in Belgium has been characterized by governmental instability compensated by the fact that a high proportion of individual ministers remained in office over long periods, in part because there was no complete alternation in party composition of coalitions. Between 1946 and 2007, despite an average cabinet duration of 567 days,26 complete ‘make-overs’, when no single minister of the preceding government remains in office, happened twice,27 whilst three governments were composed of exactly the same people as at the start of the previous government, though all of these were interim governments. In general, owing to the permanence of the Christian Democrats over 45 consecutive years and for the last 20 years the presence of the Socialists in power, more than half of the members of any new government (on average 56 per cent) had already been part of the ministerial team starting the preceding government. In only 16 out of 37 governments was this not the case. For the same reason, the average ministerial duration varies between parties, the overall cabinet experience of Christian Democratic politicians being 5.1 years, 4 for Socialists and 3.5 for Liberals. Altogether, the average duration of an individual in cabinet is 4.1 years. Compared to the data compiled in previous studies (Philippart 1962; CRISP 1972; Dewachter 1982; Blondel 1988; Frognier 1988; 1997), this average and sometimes discontinuous duration in government is slightly on the rise. Given the weak alternation of parties in power and the often long-standing colonization of specific portfolios by some of them (for instance Agriculture and Health for the Christian Democrats who count on powerful Catholic organisations in these fields (De Winter and Dumont 2003)), the heavy rotation of ministerial posts amongst a core group of non-specialized ministers may seem puzzling, but this does not appear to have changed much either, though some ministers (for instance Finance) have kept their portfolio for more than two consecutive governments. Also, regardless of the specific portfolio, only 36 per cent of Belgian ministers are appointed only once in the federal government. Notwithstanding the progress made regarding gender balance, it is striking that female ministers have had a harder time building long careers in the government, as more than half of them do not return after first appointment. Finally, we observe a negative correlation between the number of government positions and cabinet duration (r = –0.29 when computed with all positions and r = –0.37 with only senior ministers), suggesting that the larger the cabinet, the shorter its survival. These figures also reflect the fact that from 1995 to 2007 all cabinets have completed their four-year term in office (regardless of the number of junior ministers appointed), whilst the period of high instability of the 1970s corresponds to the fragmentation of the party system and the corresponding inflation of the number of offices.

Resigning In Table 7.4 we distinguish seven different motives explaining the 128 individual resignations out of 882 appointments at start of government. 28 Death whilst in office only imposes minimal changes inside the cabinet (usually, only the deceased

138 The selection of ministers in Europe Table 7.4 Reasons for resignation in Belgium, 1946–2007 Reason for resignation

N

%

Death or health reasons Incompatibility with other mandates

9

7

56

43.8

Individual responsibility Cabinet disagreement

19 16

14.8 12.5

Cabinet rebalancing Adaptation of the government

8 20

6.3 15.6

Specific reason

N

%

Local mandate Regional mandate International mandate Party executive mandate Others

9 24 11 8 4

7.0 18.8 8.6 6.3 3.1

Protest resignation Revocation

10 6

7.8 4.7

minister is replaced). Throughout the period five members died in office, while illhealth forced another four to resign. Invoking health may conceal the real motives of a resignation as they might be an escape route for the minister and colleagues over some scandal.29 Death and ill-health account for 7 per cent of all ministerial resignations. The category ‘mandate incompatibility’ is by far the largest for several reasons (see also Dumont et al. 2001). First, Belgium is now a federal country, with fully fledged regional levels of governments. Secondly, the local level is considered as highly valuable by both politicians and citizens. Thirdly, in a small country that was amongst the founding fathers of several international organizations, national politicians have had a disproportional chance of being recruited for important international posts. Finally, in a partitocracy, giving up a ministerial seat in order to become party president is widely considered as a promotion in one’s career, especially in parties that rely on a strong party structure (eight cases, all Socialist and Christian Democratic ministers). Whilst there is no legal incompatibility, the key party role is not normally held by ministers. Ministerial resignations taking place just after local elections or to become province governor (often seen as a reward at the end of a political career) are slightly less frequent than those in order to occupy a position at the international or supra-national level. But the different state reforms have paved the way for a new type of transfer: the departure of a federal minister towards federated entities. In July 2004 five ministers left their federal position to occupy executive mandates at the regional and community level. The overall amount of resignations due to incompatibility of the function between a federal ministerial mandate and another political mandate reaches more than two-fifths of the total. Within this category, transfers to a regional government account to almost 20 per cent of all resignations. Since late 1981, when regional governments became fully autonomous, 24 resignations occurred in order to occupy a position at the regional level, representing almost 38 per cent of the total between 1981 and 2007. Not all these can be seen as demotions, even though some may take it as an acceptable exit after difficult times at the federal level or for the party president to hide internal conflicts or sanctions.

Belgium: ups and downs of ministerial careers 139 In recent times, the formation of a government at any level had important repercussions on the composition of the cabinet of other levels. As parties are usually represented at both levels, the game of ministerial allocation, with its gender, geographical, experience and expertise constraints, is played with the whole number of ministerial positions that one party is allowed to fill. Hence, a resignation in one cabinet may imply several resignations in others in order to balance the representation of the party. With the third reason for individual resignation we come to ministerial accountability. Although parliamentary control over individual ministers (interpellations and parliamentary investigation committees) have increased over the past two decades (De Winter and Dumont 2006; Nandrin 2003), formal votes or recommendations taking place in parliament have not led to ministerial resignations since the late 1940s. Parties, media and public opinion seem to have considerable influence in the 19 (14.8 per cent) resignations in the postwar period. Debate around ministers’ political responsibility reappeared with the traumatic affairs – Dutroux, Adamu or dioxine – that emerged in the 1990s, leaving the country in a deep crisis (confidence in political institutions reached the lowest rates ever recorded in Eurobarometers). These affairs led to major policy changes like the police and justice reforms and impacted on the individual careers of some ministers. In these times of low trust in politicians, mistakes committed by a minister’s services or subordinates, once public, make it difficult for the implicated minister to remain. It is worth underlining that the choice to resign or not still ultimately constitutes a personal choice of the minister, but for the 1990s cases the role of the PM appeared predominant. Although Prime Minister Dehaene was at the same time the undisputable leader of his party and therefore enjoyed considerable leeway with regard to his party’s ministerial delegation, he was also the driving force behind three resignations of ministers who did not belong to his party by convincing his partners that the stability of the government and of the country was at stake. Collective decisions are binding, ministers must ‘submit or resign’. If unhappy with a decision of a fellow colleague or the Council of Ministers, a minister can either give in, protest through resignation or be forced to resign (revocations presented by the PM to the Head of State). Altogether 12.5 per cent of resignations were due to breaches of cabinet solidarity, with ten protest resignations and six revocations, the latter corresponding to the ousting of a small coalition party by the PM, in accordance with remaining partners, through the forced exit of all its ministers. Owing to the coalitional nature and the constraints over the composition of Belgian cabinets, individual ministerial resignations often impose a rebalancing of the cabinet. The changes, mainly due to the application of the parity rule, that might trigger promotions of junior ministers to full minister account for 6.4 per cent of all individual resignations, whilst adaptations of the structure of the government to a changing environment (international event, institutional reform) represent another 15.6 per cent. Between 1946 and 1972, we count only 29 resignations, that is more or less one resignation per year, before a short period of high instability with a five times

Figure 7.3 Number of resignations in Belgium, 1946–2007

0

2

4

6

8

10

12

1946 1948 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Belgium: ups and downs of ministerial careers 141 greater annual rate (1973 and 1977), followed again by a more stable decade. Between 1988 and 2007, we however notice again an important increase in the number of ministerial resignations with an average rate of three resignations a year, mainly triggered by moves to the regional level and the revived question of the individual responsibility of ministers. Eighteen governments did not lead to any ministerial resignation, the 128 cases being therefore spread on the remaining 19 governments for an average of 3.5 resignations per government. Those where resignations were most frequent were coalition governments composed of five different parties,30 whilst the few single-party cabinets in Belgium had the lowest number of resignations. Contrary to what could be expected from the potential role of the Prime Minister, by far the smallest proportion of resignations during the course of a government is recorded amongst the Christian Democrats (both the Socialists and the Liberals had more than twice their rate). The PM’s party may have protected its ministers in order to avoid threatening governmental stability not the least because of the factional character of the long-time party of the chief executive (the CVP): factions pressurize on both the PM and the party executive to keep their men in place, and this internal structure adds to the complexity of reshuffling the party team. Conversely, coalition partners may also refrain from calling for the resignation of ministers of the PM party for this same reason, as an internal crisis in the CVP might bring the cabinet down. In other traditional but junior parties, such as the Flemish Socialist Party, even heavyweights have had to resign at some point (and come back again later). As expected, smaller, single-issue and firmly policyoriented parties reach the highest average of resignations per governmental participation and per number of positions held. In terms of portfolio, law and order positions appear to be less sensitive than those pertaining to the economy, international relations or the welfare state.

Conclusions Over the postwar period, Belgium has arguably evolved into the most complex cabinet formation system, owing to the combined effects of the ‘community/ linguistic’ divide on party system fragmentation and on the structure of the state. Strong disciplined parties have created an elaborate set of delegation control mechanisms over governmental action (De Winter et al. 2003; 2006; De Winter and Dumont 2006). Government formation is therefore characterized by long negotiations and detailed coalition agreements. The potential for governmental downfall in case of breaches of subtle coalition equilibriums reached at cabinet formation amongst up to six parties has for long accounted for the limited number of individual resignations. An attack on an individual party minister was often interpreted as an attack on her party as a whole, and led therefore to the resignation of all party ministers, triggering the downfall of the government. The factional character of the long-time dominant Christian Democratic party, and the fact that often PMs were outranked by their party presidents in the battle for true leadership over the party, has led to the CVP having a lower turnover rate.

142 The selection of ministers in Europe Institutional constraints have had both a direct and an indirect impact on the selection and de-selection of ministers. Aside from specific rules such as the linguistic parity of ministerial personnel in the Council, the federalization of the country has also created a larger opportunity structure so that becoming minister at the federal policy level is not always seen as the ultimate prize of a political career actor. The room for strategic action of party presidents has been further enlarged as a result of their role in co-ordinating government action at the different levels. Depending on the internal state of the party and its electoral prospects, ministerial personnel may be moved from one level to another irrespective of their competencies or quality. However, the last two decades of governmental stability, probably sustained by the introduction of a constructive vote of confidence since 1995, have corresponded with a tremendous increase in individual resignations. The erratic character of current ministerial careers is not entirely due to the federalization of the country and/or party presidents’ strategies but can be attributed also in some cases to the influence of the PM when questions over the individual responsibility of ministers threatened the stability of his government. The action of recent Prime Ministers who were undisputed leaders of their own party both at the time of their accession to the government and during their mandate however does not compensate for the overall image of the almost non-existent power of chief executives regarding ministerial appointment and de-selection in a partitocratic federal state.

Notes 1 An exception occurred in 1990, when moral preferences led King Baudouin to refuse to sign the bill liberalizing abortion. The government found a constitutional ‘trick’ by declaring (with his consent) Baudouin temporarily unable to exercise his powers for a few days to allow the Council of Ministers to sanction and promulgate the law. 2 This vote takes place after the formal appointment of a new government by the King and therefore is more a test of viability than a genuine parliamentary investiture prescribed by the Constitution, conditioning the coming to power of the new executive team. However, this vote on a confidence motion presented by the new PM, which is enshrined in the standing orders of the House, is customarily considered as an investiture vote as the newly appointed government is not supposed to govern without this explicit parliamentary assent. It has constrained the formation of all Belgian governments since 1919 when the combined effects of proportional representation (1899) and universal suffrage (for males aged over 20) made coalition cabinets the rule. 3 In March 1946 the cabinet of Prime Minister Spaak, made up of Socialist and nonpartisan (technicians) ministers, which theoretically relied on only about one-third of all Belgian MPs, failed to gain parliamentary support only for one vote (there was a draw between supporters and opponents); a vote in the Senate was therefore unnecessary and Spaak presented the resignation of his government to the King. This need for majority support in both assemblies also constrained the formation of other cabinets: for instance in 1949 and 1958 the Christian Democrats had to form a coalition (and therefore enlarge their single-party transitory cabinet which was in minority in the House) despite relying on a majority in the Senate. 4 Or in bills implementing specific constitutional rules regarding the borders, competencies, statutes of the Communities and the Regions, which have to be voted on

Belgium: ups and downs of ministerial careers 143

5

6

7

8

9

10

11

12 13

by a two-thirds majority of the valid votes cast, and also by a majority of valid votes in each linguistic group in each House (with a majority of members of each linguistic group present). From the creation of the state, French had been the language of the elite. Linguistic laws in the 1870s and the 1930s gradually recognized the usage of Flemish (Dutch), the language spoken by a majority of the population. In 1963 a bill instituting a ‘linguistic border’ was voted on, delimiting three unilingual regions (French-speaking, Dutch-speaking and German-speaking) and the bilingual (French and Dutch) region of Brussels (made up of 19 communes). This delimitation and the setting up of electoral constituencies still causes governmental crises. Regional parties were created and gained momentum throughout the 1960s and in 1968 the crisis over the transfer of the French-speaking section of the University of Louvain (situated in Flanders) caused a government breakdown and the split of the first traditional party group (Christian Democrats) into two (one French-speaking, one Dutch-speaking) autonomous parties. The Liberals (1972) and the Socialists (1978) split in turn so that the effective number of parties represented in the House rose from 2.7 in 1961 to 7.7 twenty years later. The linguistic category of ministers without parliamentary background is inferred from official documents such as ID card or membership of unilingual organizations. Although the German-speaking area (71,500 out of more than 10,500,000) is recognized as an autonomous Community, with its own parliament and government, it is not recognized as a linguistic group at the national level. The position of adjunct to a minister was created in 1960 but came under different names until the 1970 constitutional revision completed the existing article stipulating that the King, as for his ministers, also appoints and revokes the ‘Secretaries of State’. These junior ministers are not part of the Council of Ministers (and therefore do not count for the ‘parity rule’), though they are invited when issues falling into their sphere of competence are discussed and their decision-making autonomy can be relatively large (their remit and the limits of their power to countersign legislation are specified by a royal decree at the time of government formation). For instance the Martens III 1980 coalition was made of six parties (both the Frenchspeaking and Flemish sides of the three traditional parties) and counted 27 ministers plus 11 junior ministers. Only one year later but with a hard economic austerity mission ahead, the same PM composed a cabinet of four parties with only 15 ministers (and 10 junior ministers). Anticipating the constitutional reform in 1992, Dehaene formed his first cabinet also with only 15 ministers (and one junior minister). It is necessary to go back to the 1950s to find such a reduced form of government. Nor for the holders of the honorary, permanent title of Minister of State who are generally former Prime Ministers or party leaders. These are not part of the government but may gather in a Council of the Crown to give advice to the King in times of international crises (this has happened only five times in Belgium’s history). The need for the appointment of the leader of the junior coalition party as deputy Prime Minister materialized in 1958 but the title of Vice-Prime Minister was first given to Spaak in 1961. From the 1980s onwards each coalition partner is awarded a Vice-PM and since 1988 usually the party of the PM also receives such a position allowing it to have two representatives in the Kerncabinet. Other specific missions or titles for their holder have been given since the Second World War, when especially community conflicts made the formation of a cabinet quite complicated: ‘mediators’, ‘negotiators’ and in 2007 even an ‘explorer’ were appointed by the King. There were exceptions in 1950 and in 1968, and in both 1979 and 1988 a successful formateur was replaced by the man who became PM a couple of days before the swearing-in ceremony of the new ministerial team. Even though Verhofstadt III formed in late December 2007 was invested with full powers (unlike a dismissed cabinet) it had a limited coalition programme and an ‘expiry date’ specified, the Flemish Liberal PM Verhofstadt being committed to

144 The selection of ministers in Europe

14 15

16

17 18

19

20 21

22 23 24

25

26 27

28 29 30

handing his position over by the end of March 2008 to Flemish Christian Democrat Leterme. Stengers (1996: 45–60) refers to at least four cases during Baudouin’s reign. He is also said to have been influential in the nomination of certain Prime Ministers, the last case being the appointment of Martens in 1988. Albert II is said to have suggested Louis Michel as Foreign Affairs Minister (despite the latter’s preference for Interior) to improve Belgium’s image abroad after the scandals of the 1990s that made the small kingdom (in)famous on the international scene. In 1999, the formateur Verhofstadt apparently pleaded for a feminized government and managed to have the French-speaking Socialist Party choose a female Vice-PM, whereas a male designate would normally have taken the position (Van Peteghem 2001). Three people have served under two different parties, and 20 ministers served in cabinet before the first general elections of 1946. Note however that in 1999 the French-speaking Greens (Ecolo) decided during the participation congress on the names of ministerial candidates to be presented to the formateur. This highly democratic exercise led to the exclusion of the main negotiator (and ‘leader’) of the party by the members from this Green ministerial delegation. Interestingly, amongst the 14 ministers of the Verhofstadt III coalition based on a centre-right coalition to which was added the French-speaking Socialists, there were no fewer than five left or centre-left ministers, as in addition to the three PS ministers each of the Christian Democratic parties sent a former (Catholic) trade unionist to cabinet. All these figures refer to raw numbers, no weighting of positions applied. The agreement in 2008 to have five French-speaking and only two Flemish junior ministers in the cabinet Leterme I is in this respect unusual. So is the asymmetric party composition of the coalition (unlike a traditional tripartite coalition which would have counted six parties, the Flemish Socialists did not enter government, leaving it with five parties), inherited from the Verhofstadt interim government. This includes Social Affairs, Health, Children’s Policy, Family, Housing, Education, Science, Culture, Labour and Sports. ‘Infrastructure’ includes Transport, Communication, Environment, while ‘Economy’ includes Economy, Industry, Commerce and Agriculture. These averages are calculated on the basis of 308 individuals who were appointed for the first time as minister or junior minister in the period after 13 March 1946. They do not include the additional 22 individuals who had their first ministerial appointment before this date. When looking at aggregate parliamentary experience in the graph, one has to bear in mind that since 1995 ministers drawn from parliament have to give up their seat in parliament for as long as they remain in cabinet, thereby no longer accumulating experience. High instability however mainly concerned the second half of the 1940s and the 1972–81 period. The Van Acker IV government (1954–58) was an anti-clerical government formed by Socialists and Liberals after the Christian Democratic single party majority of 1950–54. The succeeding government, Eyskens II (1958), which was a short term minority cabinet aimed at being enlarged, consisted of only Christian Democrats. These 128 individual resignations concern 98 ministers and 30 junior ministers. For instance Edouard Anseele (Socialist) officially resigned for health reasons, but was also suspected in a financial scandal. No fewer than 9.5 resignations per government. Instability of cabinets with an odd number of parties shows the importance of having party groups (re-)united in government, and to avoid as much as possible including an additional one without any counterpart in the other Community.

Belgium: ups and downs of ministerial careers 145

References Berkmann, M. (n.d.) Validation of Concepts, Database on Women and Men in DecisonMaking, n.p.: European Commission DG on Employment and Social Affairs. Blondel, J. (1988) ‘Ministerial Careers and the Nature of Parliamentary Government: The Cases of Austria and Belgium’. European Journal of Political Research 16, 2: 51–71. Centre de Recherche et d’Information Socio-Politiques (CRISP) (1972) ‘La stabilité gouvernementale en Belgique 1946–1971’. Courrier hebdomadaire du CRISP 578. Das, E. (1987). ‘Het Belgisch ministerieel carrièrepatroon. Proeve tot internationale vergelijking’. Res Publica 29, 2: 207–229. Dejaeghere, Y. (2007) ‘Etre membre d’un bureau de parti: la voie royale pour devenir ministre?’ In Dewachter, W. and Depauw, S. (eds) Bureaux de partis, bureaux du pouvoir. Cinquante ans d’histoire. Brussels: Editions de l’Université de Bruxelles, 183–194. Dewachter, W. (1982). ‘Crises macro-sociales et stabilité de l’élite’, Res Publica 24, 2: 305–325. Dewachter W. and De Winter, L. (1981) ‘Over particratie’, Special issue Res Publica 23, 1: 1–151. De Winter, L. (1991) ‘Parliamentary and Party Pathways to the Cabinet’. In Blondel, J. and Thiébault, J.-L. (eds) The Profession of Government Minister in Western Europe. London: Macmillan, 44–69. De Winter, L., Della Porta D. and Deschouwer, K. (1996) ‘Comparing Similar Countries: Italy and Belgium’. Res Publica 38, 2: 215–236. De Winter, L., Timmermans, A. and Dumont, P. (2003) ‘Belgium: On Government Agreements, Evangelists, Followers and Heretics’. In Strøm, K. and Müller, W. (eds) Coalition Governments in Western Europe. Oxford: Oxford University Press, 300–355. De Winter, L. and Dumont, P. (2003) ‘Belgium: Delegation and Accountability under Partitocratic Rule’. In Strøm, K., Müller, W. C. and Bergman, T. (eds) Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press, 253–280. De Winter, L. and Dumont, P. (2006). ‘Do Belgian Parties Undermine the Democratic Chain of Delegation?’ West European Politics 29, 5: 957–976. De Winter, L., Swyngedouw, M. and Dumont, P. (2006). ‘Party System(s) and Electoral Behaviour in Belgium: From Stability to Balkanisation’. West European Politics 29, 5: 933–956. Dumont, P., De Winter, L. and Dandoy, R. (2001) ‘Démissions gouvernementales et performances électorales des majorités sortantes (1946–1999)’. Courrier hebdomadaire du CRISP 1722. Fiers, S. (1998). Partijvoorzitters in België, of ‘Le parti c’est moi?’ Leuven: Afdeling Politologie. Fiers, S. (2001). ‘Carrièrepatronen van Belgische parlementsleden in een multilevel omgeving (1979–99)’. Res Publica 43, 1: 171–192. Fiers, S. and Gubin, E. (2003). ‘La physionomie de la Chambre des Représentants’. In Gubin, E., Nandrin, J.-P., Gerard, E. Witte, E. (eds) L’Histoire de la Chambre des Représentants, Brussels: Chambre des Représentants, 89–128. Fiers, S. and Krouwel, A. (2005) ‘The Low Countries: From Prime Minister to PresidentMinister’. In Poguntke, T. and Webb, P. (eds) The Presidentialization of Politics: A Comparative Study of Modern Democracies. Oxford: Oxford University Press, 126–156.

146 The selection of ministers in Europe Frognier, A.-P. (1988) ‘The Mixed Nature of Belgian Cabinets between Majority Rule and Consociationalism’. European Journal of Political Research 16, 2: 207–228. Frognier, A.-P. (1993) ‘The Single Party/Coalition Distinction and Cabinet DecisionMaking’. In Blondel, J. and Müller-Rommel, F. (eds) Governing Together: The Extent and Limits of Joint Decision-making in Western European Cabinets. Basingstoke: Macmillan, 43–73. Frognier, A.-P. (1997) ‘Belgium: A Complex Cabinet in a Fragmented Polity’. In Blondel, J and Müller-Rommel, F. (eds) Cabinets in Western Europe, 2nd edition. London: Macmillan, 75–97. Höjer, C. K. (1969) Le régime parlementaire en Belgique de 1918 à 1940. Brussels: CRISP. Laver, M. and Shepsle, K. (1996) Making and Breaking Governments. Cambridge: Cambridge University Press. Mershon, C. (2001) ‘Contending Models of Portfolio Allocation and Office Payoffs to Party Factions: Italy, 1963–1979’. American Journal of Political Science 45, 2: 277–293. Nandrin, J.-P. (2003) ‘Le droit d’enquête parlementaire. Perspective historiennes’. In Gubin, E., Nandrin, J.-P., Gerard, E. and Witte, E. (eds) L’Histoire de la Chambre des Représentants. Brussels: Chambre des Représentants, 291–308. Philippart, A. (1962) ‘Analyse statistique de la stabilité ministérielle en Belgique de 1830 à 1961’. Res Publica 4, 3: 239–276. Poguntke, T. and Webb, P. (2005) The Presidentialization of Politics: A Comparative Study of Modern Democracies, Oxford: Oxford University Press. Stengers, J. (1996) L’action du Roi en Belgique depuis 1831. Pouvoirs et influence. Brussels: Editions Racine. Van Peteghem, B. (2001) Paars met Groen. Tielt: Lannoo. Waleffe, B. (1971) ‘Le Roi nomme et révoque ses ministres’. La formation et la démission des gouvernements en Belgique depuis 1944. Brussels: Bruylant.

8

Portugal The primacy of ‘independents’ António Costa Pinto and Pedro Tavares de Almeida

Democratization and the semi-presidential government Portugal inaugurated the third wave of democratization with a bloodless military coup on 25 April 1974, putting an end to four decades of dictatorship (1926–74). Unshackled by international pro-democracy forces and occurring in the midst of the Cold War, the coup led to a severe crisis aggravated by the concomitance of two processes: the transition to democracy and the end of what was the last European colonial empire (Pinto 2003). As Juan Linz and Alfred Stepan have noted, ‘we all too often tend to see [Portugal] in the framework set by later transitions processes’, forgetting the greater degree of uncertainty and the ‘extreme conflict path’ of the regime change (Linz and Stepan 1996). The decolonization process was the main reason for the conflict that broke out between some conservative generals and the Armed Forces’ Movement (MFA – Movimento das Forças Armadas) – which had planned and executed the coup – in the immediate wake of the regime’s collapse: a conflict that was also at the root of the military’s active intervention in political life following the dictatorship’s overthrow. The revolutionary period of 1974–75, when powerful tensions emerged within Portuguese society, was the most complex phase of the transition to democracy. Following the military coup of 25 November 1975, which was led by moderate officers, tensions began to subside, and in 1976 a new Constitution was approved and the first legislative and presidential elections were held, opening the political arena for democratic consolidation. Alone out of the four principal founding parties of Portuguese democracy, the Communist Party (PCP – Partido Comunista Português) had a long history of clandestine organization within the country. The Socialist Party (PS – Partido Socialista), founded by Mario Soares in Germany in 1973, was heir to the republican and socialist elements of the electoral opposition to Salazarism. The parties that were to represent the right and centre-right, the Social Democratic Centre (CDS – Centro Democrático Social) and the Popular Democratic Party – later renamed Social Democratic Party (PSD) – were formed after 25 April 1974. In spite of periodic challenges,1 these parties have structured, in a stable way, Portuguese democracy (Bruneau 1997; Jalali 2007).

148 The selection of ministers in Europe The nature of the transition had several legacies for the political system, namely the initially prominent constitutional role of the military (Pinto 2006). After difficult negotiations, and as a result of a pact between the military and the parties, the 1976 Constitution established the Council of the Revolution (CR – Conselho da Revolução). The creation of the CR, which was to be led by the head of state, was closely connected with the implicit agreement that the first president would be a military officer – as actually happened in July 1976, when General Eanes was elected with the support of PS, PSD and CDS. The CR was granted relatively extensive powers. It had exclusive legislative powers over the organization, operation and discipline of the armed forces, and could approve international agreements on military matters via decree. Moreover, the CR was the guarantor of the Constitution, and had to be consulted by the president before the appointment of a new prime minister. The temporary ascendancy of the military was also of particular relevance for the constitutional design of a mixed form of government: the semi-presidential system. The dual responsibility of government towards the president and the parliament, which is common to all semi-presidential regimes, meant that the president could withdraw his political confidence, forcing the government to resign even if it enjoyed parliamentary support (Freire and Pinto 2005). Moreover, the president could dissolve parliament at will. During President Eanes’s first mandate, he oversaw the formation of three successive governments (1978–79) led by independents (see Table 8.1). The president was granted veto powers over both parliamentary and government bills, and could also request the verification of the constitutionality of the bills, either ex ante or ex post.2 The parliament and the CR also enjoyed powers that constrained the government’s overall steering function. Yet the 1976 Constitution placed the government at the helm of policy-making and of public administration. The president’s powers vis-à-vis the government were, in turn, limited in several respects. First, the president needs the prime minister’s countersignature in such matters as the appointment of government ministers, the dissolution of the governments of the autonomous regions (Azores and Madeira) and in the declaration of a state of emergency. Secondly, the president might preside over the Council of Ministers only if asked to do so by the prime minister. Lastly, the president had no formal capacity to influence the selection of ministers in the formation of party governments, and informal vetoes on individuals were rare. In 1982, when the balance of forces had evolved considerably with the decline of the military’s power, the Constitution was revised – largely to circumscribe the powers of the president and subordinate the military to political control. This was achieved with the support of the parties of the right in parliament – the PSD, CDS and Popular Monarchist Party (PPM – Partido Popular Monárquico) – and the PS. The CR was dissolved and its powers redistributed amongst two new bodies, the Constitutional Court and the Council of State. With respect to presidential power, the most important change was that the government became politically responsible only to parliament. This meant that, in contrast to the original text, the president could no longer dismiss the government by claiming a lack of political trust. Nevertheless, the president retained limited authority to dismiss the government

Portugal: the primacy of ‘independents’ 149 in ‘exceptional’ political circumstances. As to parliamentary accountability, the original rules were kept unchanged, but for one procedure.3 Hence, the government is dismissed under three circumstances: (1) the government’s programme is rejected by the absolute majority of deputies; (2) a confidence vote (moção de confiança) is not approved; and (3) a no-confidence vote4 (moção de censura) is approved by an absolute majority. The 1982 constitutional revision reduced the president’s powers to a level that is below average in semi-presidential regimes (Siaroff 2003). Following General Eanes’s two mandates (1976–80 and 1980–86), Mário Soares – historic leader of the PS – became the first civilian president of Portuguese democracy. Soon after he took office the PSD won the first of two absolute majorities (1987–95). Thus, Mário Soares’s presidency of 1986–96 was almost entirely one of cohabitation with a single-party majority government. Indeed, it seems that the consensus surrounding the reduction in presidential powers may have been cemented not on the decrease in constitutional powers per se but rather on the changes in the party system that produced stable government majorities. From 1987 until 1995, stable single-party absolute majorities, coupled with a president who regarded himself as a referee rather than as a policy-maker, combined to frame the Portuguese president as an active observer, but not as the locus of executive power, which rested firmly with the prime minister and his government (Lobo, Pinto and Magalhães 2008). President Jorge Sampaio also served two mandates (1996–2001 and 2001– 2006). The first one coincided with a single-party minority government led by António Guterres (see Table 8.1). Thus, between 1996 and 2001 both the government and the presidency were held by the Socialists. Following Guterres’s resignation at the end of 2001, new elections were held and a right-wing coalition government was formed by the PSD and the CDS. In July 2004, the abrupt resignation of Prime Minister Barroso, then appointed president of the European Commission, opened up a political crisis that was aggravated by the erratic performance of his successor (Santana Lopes). A few months later, the president decided to dismiss the government, to dissolve parliament and call for elections. This event shows clearly that constitutional powers still allow great presidential authority during times of political instability.5 In 2005, one year before the end of his second mandate, Sampaio saw the PS return to power, with its first absolute majority (Tavares de Almeida and Freire 2005). As regards the government structure, the 1976 Constitution established the post of vice-prime minister, but it has rarely been implemented. 6 Besides this position there is no formal hierarchy between ministers; yet the title of ‘ministro de Estado’ has been used to describe those who occupy pivotal positions in the cabinet (e.g. the Minister of Finances). According to the constitutional rules, junior ministers (secretários de Estado) depend on the respective minister and are forced to resign with him. They might also occasionally participate in cabinet meetings as substitutes for their ministers, but the recent institutionalization of the Meeting of Junior Ministers (Reunião de Secretários de Estado) – which meets regularly under the leadership of the Minister of the Presidency of the Council of Ministers – made this procedure quite rare. There is a formal incompatibility between a parliamentary seat and a ministerial office, hence deputies who are appointed

150 The selection of ministers in Europe ministers should be replaced, although maintaining a ‘sleeping mandate’. Afterwards, if they leave the government, they can resume their parliamentary seats. The overall size of the government has had some significant variations, namely as a result of the expansion or reduction of the number of distributable positions to junior ministers (see Table 8.1). The portfolio allocation of senior positions is limited, and most governments comprised between 16 and 18 ministers (including Table 8.1 Type, party composition and size of governments in Portugal, 1976–2005 Government

Type

Ministers

Junior ministers

Total

23 July 1976 M. Soares 23 January 1978 M. Soares 27 August 1978 Nobre da Costa 21 November 1979 Mota Pinto 31 July 1979 Pintasilgo 3 January 1980 Sá Carneiro 9 January 1981 F. Balsemão 4 September 1981 F. Balsemão 9 June 1983 M. Soares 6 November 1985 Cavaco Silva 17 August 1987 Cavaco Silva 31 October 1991 Cavaco Silva 28 October 1995 A. Guterres 25 October 1999 A. Guterres 6 April 2002 J.M. Barroso 17 July 2004 Santana Lopes 12 March 2005 J. Sócrates

Single party PS Coalition PS+CDS Independents

18

45

63

16

36

52

15

31

46

Independents

16

35

51

Independents

17

33

50

Coalition PSD+CDS+PPM Coalition PSD+CDS+PPM Coalition PSD+CDS+PPM Coalition PS+PSD Single party PSD Single party PSD Single party PSD Single party PS Single party PS Coalition PSD+CDS/PP Coalition PSD+CDS/PP Single party PS

15

39

54

18

41

59

15

46

61

17

40

57

14

32

46

16

36

52

17

51

68

18

39

57

18

43

61

18

34

52

20

38

58

17

36

53

Notes: Number of ministers and junior ministers at the time of government’s appointment. ‘Ministers’ includes the prime minister. ‘Junior ministers’ includes both Secretários de Estado and Subsecretários de Estado.

Portugal: the primacy of ‘independents’ 151 the prime minister). There is no correlation between changes in government size and changes in the number or the ideological orientation of government parties (both the smallest and the largest ones were single-party cabinets of the PSD).

The selection of Portuguese ministers As argued above, the transformations in executive power that have occurred over the last three decades have served first of all to extricate the military from the political system, and to subordinate them clearly to the civilian power. Presidential powers have also been circumscribed, at least at times of political stability. Hence, in Portugal’s democracy party leaders who become prime ministers have increasingly enjoyed a great deal of autonomy in the selection of their ministers. At the same time, parliamentary groups have tended to occupy a subordinate position within the parties’ internal power structures and it is normal for parties in power to suffer a ‘governmentalization’ of their leadership (Lobo 2005). Beginning by analysing the socio-demographic background of the ministers who have served during the present democratic period, we are able to identify four essential traits. First, we can see there have been very few women ministers. Despite a woman having been prime minister in 1978, the presence of women in the executives from 1976 to 1999 corresponds to approximately 4 per cent of the total.7 While remaining at a very low level, the proportion of female ministers has increased in the last four executives. As regards gender, it is worth noting that, unlike the parliament, where the parties of the left have been the major suppliers of women deputies, at ministerial level there is not a relevant difference between right and left. A second important aspect, alongside the virtual non-existence of ministers without university education, is the relative hegemony of ministers with a law degree, a trend that is more accentuated in governments of the right. Yet ministers with technical backgrounds (engineering, economics and management) are in the whole predominant, which is an indicator of a preference for selecting ‘specialists’ rather than ‘generalists’. Thirdly, the geographic centralization of ministerial recruitment is clearly visible by the overwhelming majority of ministers who completed their studies at the University of Lisbon. Finally, with respect to occupational profile, ministers tend to emerge from three basic categories: university professors; public and private managers; and lawyers. The importance of universities as a hotbed of ministerial recruitment is strictly correlated with the increased protagonism of ‘independents’ and/or technocrats in the executives. With respect to the political cursus honorum, there are two essential aspects that deserve to be highlighted (see Table 8.3). On the one hand, and contrasting with many other European democracies, the parliament does not have a strategic role in the ministerial selection process. Between 1976 and 2005 only 49 per cent of all first-time ministers had been elected to parliament, with a significant proportion of the ex-deputies having little effective experience as parliamentarians. With respect to the length of parliamentary service, it should be stressed that a large proportion of those who were deputies (38 per cent between 1976 and 1999) were elected only once, some of them having never actually served in parliament owing

152 The selection of ministers in Europe Table 8.2 Social background of ministers in Portugal, 1976–2005 (%) Degree

Main occupation Lisbon Female Governments Law Engineering Economics Univ. Professor Manager Lawyer 1976–99* Average

35.6 29.3

20.6

70.2

32.1

24.5

19.5

4.3

Guterres II 1999–2002

26.9 19.2

26.9

64.0

31.0

17.2

10.3

11.5

Barroso 2002–2004

43.5 13.0

30.4

81.0

30.4

17.4

13.0

17.4

Lopes 2004–2005

65.0 10.0

20.0

84.2

20.0

20.0

35.0

15.0

Sócrates 2005–

29.4 23.5

29.4

82.3

52.9

11.8

5.9

11.1

* Source: Tavares de Almeida and Pinto 2003.

Table 8.3 Political background of ministers in Portugal, 1976–2005 (%) Political experience Governments 1976–99** Average Guterres II 1999–2002 Barroso 2002–04 Lopes 2004–05 Sócrates *** 2005–

Junior ministers

Deputies

Party leadership

Yes

No

0.0

46.0

51.5

49.1

76.7

23.3

13.8

55.2

79.3

65.5

86.2

13.8

17.4

43.5

65.2

65.2

82.6

17.4

15.0

20.0

55.0

55.0

75.0

25.0

7.6

41.2

52.9

52.9

64.7

35.3

Mayors

Notes: ‘Deputies’ includes those who were elected MPs, regardless of having a seat or not in parliament. Figures for Socrates include only those appointed on 12 March 2005. Source: Tavares de Almeida and Pinto 2003. Since some individuals were ministers more than once, and had a discontinuous ministerial career, we considered their political experience at first-time appointment.

to their promotion to the cabinet within a matter of weeks after their election (Tavares de Almeida and Pinto 2003). Simultaneously, the position of junior minister – where, in general, the technical talents and skills are more valued8 – has been an important step in the rise to the rank of cabinet minister. On the other hand, during the same period (1976–2005) – and despite the prominence of party governments – around 44 per cent of all first-time ministers had never been members of national party bodies.

Portugal: the primacy of ‘independents’ 153 The importance of ministers without a political background (either as deputies or as party leaders), which represents almost one-quarter of all ministers between 1976 and 1999, reached its peak in the three governments overseen by President Eanes in 1978–79. In fact, approximately two-fifths (or 41.7 per cent) of the ministers who served in these governments had no political experience. What is more significant, however, is that the non-political ministers have continued to have a significant weight in party governments, regardless of its type (coalition versus single-party) or ideological orientation (centre right versus centre left). Thus, for instance, the single-party majority Socialist government appointed in 2005 included 35 per cent of ministers without party or representative experience, while in the previous centre-right coalition government the corresponding figure was 25 per cent. Hence, when we look at the socio-demographic and political backgrounds of the ministers who have been in office during the current democratic regime there is a major element of divergence with the majority of European cabinets: the weaker parliamentary socialization along with the presence of a substantial number of ministers without any party-political experience (Blondel and Thiébault 1991; Tavares de Almeida and Pinto 2003). The salient role of these ministers without ‘political experience’ in Portuguese party governments, particularly after the brief experience of presidential cabinets, can be explained by three interrelated factors. First, the trend to reinforce the ‘technical’ legitimacy of the executive results from the growing complexities of governance (e.g. the impact of Europeanization) and is intended to countervail the strong bias of ordinary citizens against the ‘political class’. In fact, public opinion polls both on the prestige of professional groups and on the role played by the major political institutions show the persistence of a negative evaluation of full-time politicians and of the performance of parliament. The second factor is the weak institutionalization and penetration in civil society of the political parties, a phenomenon that is illustrated by the increasing decline in electoral turnout and in party loyalties of voters, since the mid-1980s, as well as by the fact that many people with high professional and academic credentials went away from parties. Hence, the attempts made by parties to promote their ‘openness towards civil society’ through a variety of initiatives aiming to bring together independents who are experts in various policy areas. The PS was the first to initiate this process by launching in 1993 the so-called Estados Gerais (Etats Generaux), an all-encompassing platform gathering Socialists and independents, which played an active role in the establishment of the party’s electoral programme in 1995 and, after the election, became an important springboard of opportunity for cabinet recruitment (Ferreira-Pereira 2005). This ‘recipe’ would be replicated, and the two major parties (PS and PSD) generated pools of ‘fellow travellers’, mainly formed by university professors and managers, who become ‘ministeriáveis’ without a previous political career. The third factor is, regardless of the personality of officeholders, the increasing leadership power and autonomy enjoyed by prime ministers, namely in the formation of their cabinets, a development that is closely connected with the growing personalization of electoral politics (McAllister 2007) and, more generally, with

154 The selection of ministers in Europe the ‘presidencialization’ of governance in modern democracies (Poguntke and Webb 2005). The political memories published by former prime minister Cavaco Silva (Silva 2002), a recent case-study on Guterres’s first cabinet (Ferreira-Pereira 2005) and interviews with former ministers provide evidence for this argument.

Termination of governments and de-selection of ministers The formation of the first single-party majority government in 1987 marked a decisive turning point in the pattern of government stability. Between July 1976 and August 1987 there were ten governments – five coalitions, three non-partisans and two single-party minority – and none of them ended the four-year term between elections. As Table 8.4 shows, the lack of consistent parliamentary majorities, and intra- and inter-party dissensions, as well as the conflicts between party governments and the president, accounted for the low rate of cabinet survival and duration. After 1987, there was a reversal in terms of the predominant government type, from coalition to single-party majority, and furthermore three out of six governments ended their constitutional mandate – a predictable scenario also for the cabinet currently in office. This new pattern of stable governments from 1987 onwards did not result from any major change in the institutional rules (e.g. an electoral reform), being closely connected both with vote concentration in the two major parties (PS and PSD) and with ideological deradicalization of adversarial party politics. From 1995 to 2002 the Socialist Prime Minister Guterres led two successive single-party minority governments, and he interrupted his prime ministership not because of any threat of a non-confidence vote in parliament but by voluntary resignation following the PS defeat in local elections. As in government formation, the will of the prime minister has been a decisive factor as regards cabinet reshuffles and the de-selection of individual ministers. Regardless of the pressures he might exert, the president cannot directly intervene in the hiring and firing of the ministers, and the parties that support the incumbent governments are often caught by surprise by the timing and contents of reshuffles. While the leadership style of prime ministers may entail a solitary approach to the decision of reshuffle or de-selection, they often get the advice of an ‘inner circle’ of cabinet members. Seemingly, the composition of this ‘inner circle’ has varied according to the government type. In the last centre-right coalition Table 8.4 Reasons for termination of governments in Portugal, 1976–2005 Elections (end of mandate) Elections (before schedule) Voluntary resignation of PM Dissension within government parties Lack of parliamentary support Intervention of the President

4 1 3 3 3 2

(Cavaco II; Cavaco III; Guterres I) (Pintasilgo) (Mota Pinto; Guterres II; Barroso) (Balsemão I and II; Soares III) (Nobre da Costa; Soares I; Cavaco I) (Soares II; Santana Lopes)

Death of PM

1 (Sá Carneiro)

Portugal: the primacy of ‘independents’ 155 governments it was formed by the leading figures of both parties who were simultaneously prominent ministers; in the single-party cabinets of PSD and of PS the ‘inner circle’ seems to be less exclusive, comprising party leaders or activists who are mere junior ministers. Taking into account that in the cabinets of Portuguese democracy there is a significant proportion of first-time ministers without previous political experience, namely as representatives and party leaders, it is worthwhile to see if they were ‘preserved’ or ‘doomed’ whenever a reshuffle or de-selection took place. Looking at Table 8.5, data from 1976 to 2005 show that only six out of a total of 63 de-selected ministers were independent and/or specialists without a former political career – i.e. less than one-tenth. This suggests that the leadership power and autonomy of prime ministers, even in critical junctures, is not necessarily constrained by party pressures or demands. It should be added that in single-party governments the proportion of de-selected ministers who had no political experience was 7.0 per cent (3 out of 43), while in coalition governments the corresponding figure is 15.0 per cent (3 out of 20). As regards the causes of ministerial de-selection, in some cases it is not difficult to establish the determinant one, but in many others there are multiple factors intervening. Also, the real motives are often dissimulated in public (e.g. negative performances or personality clashes are usually justified as policy disagreements by the fired ministers). Hence, data presented in Table 8.6 should be considered cautiously. Table 8.5 De-selected ministers without political experience in Portugal, 1976–2005 Ministers de-selected

Governments Period

Prime minister

Party

N

%

Total

23/07/1976 – 23/01/1978 23/01/1978 – 29/08/1978 29/08/1978 – 22/11/1978 22/11/1978 – 07/07/1979 07/07/1979 – 03/01/1980 03/01/1980 – 09/01/1981 09/01/1981 – 04/09/1981 04/09/1981 – 09/06/1983 09/06/1983 – 06/11/1985 06/11/1985 – 17/08/1987 17/08/1987 – 31/10/1991 31/10/1991 – 28/10/1995 28/10/1995 – 25/10/1999 25/10/1999 – 06/04/2002 06/04/2002 – 17/07/2004 17/07/2004 – 12/03/2005

Soares Soares Mota Pinto Nobre da Costa Pintasilgo Sá Carneiro Balsemão Balsemão Soares Cavaco Silva Cavaco Silva Cavaco Silva Guterres Guterres Barroso Santana Lopes

PS PS/CDS Nonpartisan Nonpartisan Nonpartisan PSD/CDS/PPM PSD/CDS/PPM PSD/CDS/PPM PS/PSD PSD PSD PSD PS PS PSD/CDS PSD/CDS

0 0 – – – – 0 0 1 – 0 1 1 1 2 0

0.0 0.0 – – – – 0.0 0.0 16.7 – 0.0 10.0 11.1 9.1 40.0 0.0

4 1 0 0 0 0 3 4 6 0 9 10 9 11 5 1

Note: Includes those who were de-selected during the life of governments, and had no experience as party leaders, mayors, deputies or junior ministers. Figures for ministers de-selected are for those without political experience.

156 The selection of ministers in Europe Table 8.6 Causes of ministerial de-selection in Portugal, 1976–2005 Causes

N Ministers

(%)

Performance Policy disagreements Intra-party conflicts Personal error Financial scandals Death and health problems Departmental error Personality clashes To party leadership Other No information Total

14 6 5 5 4 4 3 3 2 6 11 63

22.2 9.5 7.9 7.9 6.4 6.4 4.8 4.8 3.2 9.5 17.4 100.0

The predominant cause of de-selection is a negative performance, which includes ministers who had to carry on unpopular and complex policy reforms, and/or were unable to handle their relationship with the media. Intra-party conflicts occurred mainly until the mid-1980s, when the subordination of the party in government towards the prime minister was not so accentuated. ‘Financial scandals’ – mostly connected with tax evasion and sometimes notoriously inflated by the media – forced the resignation of a few influential ministers.

Conclusion The semi-presidential system of government adopted by the 1976 Constitution has been maintained, some important changes in the balance of power having nonetheless occurred. The 1982 constitutional revision limited the intervening role of the president, and subsequently the leadership of the prime minister was enhanced, actually leading to a prime ministerial government. As in other European democracies, the prime minister is the leader of the wining party in the elections, but once nominated he tends to enjoy a greater autonomy from his party, specifically as regards the selection and de-selection of ministers. This tendency explains to some extent an important feature of Portuguese democratic cabinets: the unusual number of ministers who are ‘independents’ or do not have a political background as party leaders or as representatives, and are chosen because of their alleged expertise. Indeed, ‘party governments’ became the rule, but a party-cum-parliamentary route is not a standardized cursus for ministerial recruitment. The salient role played by non-political ministers is also shown by their rarely being stepped down on the occasion of cabinet reshuffles. A significant proportion of those who ascend to executive offices are drawn from the universities or managerial positions, as specialists with high academic credentials and/or technical competences. This strong presence of non-political ministers is also related with the attempts made by parties to promote their ‘openness towards civil society’ in a political culture with strong feelings against the ‘political class’, as well as with the increasing complexity and technocratic

Portugal: the primacy of ‘independents’ 157 nature of policy-making. Although with less autonomous political power than party leaders, ‘independents’ became so important to ‘quality’ of cabinets that prime ministers think twice before sending them back to ‘civil society’.

Notes 1

2 3

4

5 6 7

8

The major challenge was the founding of the Democratic Renewal Party (PRD), under the auspices of General Eanes, in 1985. This party was a short-lived one: having won almost 20 per cent of the votes cast in the legislative elections of 1985, it dramatically dropped to 5 per cent in 1987 and disappeared in the early 1990s. Until 1982 it was the Council of the Revolution that verified the constitutional validity of the law, whereas thereafter it was the Constitutional Court. In order to dismiss a government, the original text of the 1976 Constitution required the approval of two no-confidence votes, the second one to be held at least one month after the first one; the 1982 revision simplified this procedure, demanding just one no-confidence vote. The no-confidence vote is exclusively aimed at the cabinet collectively. When it is rejected, its proponent (a group of deputies or a parliamentary party) cannot submit another one in the remaining sessions of the legislative period. In principle, the full term of a legislature is four years, comprising four legislative periods. Notice also that all presidents have actively used their power to refer legislation to the Constitutional Court and have used their power of veto to influence policy-making. There was a vice-prime minister in three coalition governments. The position was allocated to the leader of the second largest party in the coalition, who usually combined it with the portfolio of Defence or of Foreign Affairs. Maria de Lourdes Pintassilgo was the first Portuguese woman to become both minister (17 July 1974, as Minister of Social Affairs) and prime minister (31 July 1979). The first woman to hold a ministerial position, as Undersecretary of State for Health, was appointed (21 August 1970) in the last cabinet of the authoritarian regime. Symptomatically, only around 20 per cent of the junior ministers appointed between 1976 and 2005 had been previously elected deputies or members of a party (executive, deliberative or jurisdictional) committee at national level (Ruivo 2007).

References Blondel, J. and Thiébault J.-L. (eds) (1991) The Profession of Government Minister in Western Europe. New York: St Martin’s Press. Bruneau, T. C. (ed.) (1997) Political Parties in Portugal: Organisations, Elections and Public Opinion. Boulder, CO: Westview Press. Ferreira-Pereira, L. (2005) ‘Mapping the Formative Dynamics of the XIII Constitutional Government: From “Estados-Gerais” to the Governnmental Mandate and Beyond’. ECPR Joint Session of Workshops, Granada. Freire, A. and Pinto, A. C. (2005) O Poder dos Presidentes. A República Portuguesa em debate. Lisbon: Campo da Comunicação. Jalali, C. (2007) Partidos e Democracia em Portugal, 1974–2005. Lisbon: Imprensa de Ciências Sociais. Linz, J. J. and Stepan, A. C. (1996), Problems of Democratic Transition and Consolidation: Southern Europe, South America and Post-Communist Europe. Baltimore: The Johns Hopkins University Press. Lobo, M. C. (2005) Governar em Democracia. Lisbon: Imprensa de Ciências Sociais. Lobo, M. C., Pinto, A. C. and Magalhães, P. (2008) ‘Le Portugal’. In De Waele, J-M. and Magnette, P. (eds) Les démocraties européennes. Paris: Armand Colin, 330–346.

158 The selection of ministers in Europe McAllister, I. (2007) ‘The Personalization of Politics’. In Dalton, R. J. and Klingemann, H.-D. (eds) The Oxford Handbook of Political Behavior. Oxford: Oxford Univeristy Press, 571–588. Pinto, A. C. (ed.) (2003) Contemporary Portugal. New York: SSM-Columbia University Press. Pinto, A. C. (2006) ‘Authoritarian Legacies, Transitional Justice and State Crisis in Portugal’s Democratization’. Democratization 13, 2: 173–204. Poguntke, T. and Webb, P. (eds) (2005), The Presidentialization of Politics. Oxford: Oxford University Press. Ruivo, J. P. (2007) Os Secretários de Estado em Portugal, 1976–2005. Unpublished paper. Siaroff, A. (2003) ‘Comparative Presidencies: The Inadequacy of the Presidential, Semipresidential and Parliamentary Distinction’. European Journal of Political Research 42: 287–312. Silva, A. C. (2002) Autobiografia política. Lisbon: Temas e Debates. Tavares de Almeida, P. and Pinto, A. C. (2003) ‘Portuguese Ministers, 1851–1999: Social Background and Paths to Power’. In Tavares de Almeida, P., Pinto, A. C. and Bermeo, N. (eds) Who Governs Southern Europe? Regime Change and Ministerial Recruitment, 1850–2000. London: Frank Cass, 25–40. Tavares de Almeida, P. and Freire, A. (2005) ‘Two Overwhelming Victories of the Portuguese Left: The 2004 European Election and the 2005 Legislative Election’. South European Society and Politics 10, 3 (November): 451–464.

9

In tranquil waters Swedish cabinet ministers in the postwar era Hanna Bäck, Thomas Persson, Kåre Vernby and Helena Wockelberg

One of the most striking features of Swedish society during the postwar era is the Social Democrats’ strong hold over the country’s political life. This dominance – as often bemoaned as extolled – is perhaps nowhere more evident than in the Swedish cabinet: since the Second World War, the Social Democrats have ruled 52 out of 64 years. For many comparative politics scholars, this is, and has always been, the most salient stylized fact about Swedish politics. In this chapter we ask: what hides beneath the surface in this remarkably stable setting of nearly continuous Social Democratic control of the cabinet?

Constitutional framework Sweden is a constitutional monarchy. However, the King lost his political powers when parliamentarism, in its modern form, was introduced in 1917, followed by democracy shortly thereafter (male suffrage in 1909 and then universal suffrage in 1921).1 Yet, it was not until 1969 that the first revisions were made to harmonize the Constitution with existing practices of parliamentarism. The Instrument of Government of 1974 transferred the monarch’s role in cabinet formation to the Speaker of the parliament, the unicameral (since 1971) Riksdag. 2 According to the new Constitution, bargaining over cabinet formation begins with the Speaker holding bilateral consultations with the leaders of all parties that are represented in the parliament. Based on these consultations, he or she then proposes a Prime Ministerial candidate. Four days after the candidate is nominated, at the latest, the parliament votes on the Speaker’s proposal. If more than half of the parliament’s members vote against the proposal, it is turned down. Thus, it suffices that the candidate is tolerated by the parliament; a constitutional feature that has been termed negative, as opposed to positive, parliamentarism (see Bergman 1995). The Prime Ministerial candidate – if tolerated by parliament – is then free to staff the cabinet pretty much as he or she sees fit, after which the government is formally installed. There are, indeed, very few restrictions on who is ‘selectable’ as a minister. A cabinet minister must have been a Swedish citizen for at least ten years, and must give up any employment before entering government office. Except for this, ministers can be of almost any walk of life, and membership neither in parliament nor in the governing party is required. When appointed, ministers who are MPs are required to leave parliament temporarily.3 Their seat is filled by

160 The selection of ministers in Europe a replacement, which means that, on resigning from a ministerial post, a minister can return to parliament if he or she so wishes. Historically, the number, the professional background and the tasks of members of government were specified in the Constitution.4 Over time, however, the trend has been to award the PM and his government more and more power over their own organization, in order to improve flexibility. This was the explicit motivation behind the elimination of most details concerning ministers in the 1974 Instrument of Government. In the latter, for instance, the previous requirement that the cabinet must include at least two legal experts disappeared. Constitutionally, all ministers (including the PM) have the same weight in cabinet decision-making. Still, we can distinguish between ministers who are heads of departments and those not in charge of a department, that is, ministers ‘without portfolio’, or, as they were previously called in Sweden, ‘consultative ministers’ (konsultativa statsråd). There is also a third category, the so called junior ministers, or state secretaries (statssekreterare), which is an administrative post typically second in command in the ministry. According to Bergman (2000: 211), the tradition has been to appoint an accomplished administrator with party political ties to the department head. What is important to note is that these ministers are not part of the cabinet. As has been made clear, bargaining over cabinet formation is rather loosely regulated. The purpose of the consultations with the Speaker is simply to inform the latter of which party leader is the most viable Prime Ministerial candidate. The rest is up to the parties in parliament. Typically, the PM candidate is recruited from the largest party in the legislature (in about 70 per cent of the cases), just as in most other Western European democracies (see Bäck and Dumont 2008). The proportional representation system, with its 4 per cent threshold, has promoted the emergence of a rather stable multi-party system. And according to surveys of voters’ and MPs’ perceptions, the ordering of the parliamentary parties is consistent across time, and runs as follows (from left to right): the Left Party, the Social Democrats, the Green Party, the Centre Party, the Liberal Party, the Christian Democrats and the Conservatives (Holmberg 2000; Brothén 2003; Bäck 2003). In cabinet formation, these parties have typically banded together in two blocs, with the Social Democrats, the Left Party and, since the late 1990s, the Greens comprising the ‘Socialist bloc’, and the Centre Party, the Liberal Party, the Christian Democrats and the Conservatives making up the ‘Bourgeois bloc’. The Social Democrats’ co-operation with the Centre Party in the 1950s is an important exception to this rule.5 Usually, negotiations over cabinet formation take only a few weeks (Bergman 2000). The typical outcome is a Social Democratic single-party minority government supported by the Left Party, and lately also the Greens. Since the Second World War, the centre-right parties have managed to form governments only on six occasions. The most immediate reason for the Social Democrats having governed for more than 83 per cent of the time since the Second World War is the fact that they collected an average of 44 per cent of the ballot in this period. The regular occurrence of minority governments is harder to explain, but Bergman (1995) attributes it to the aforementioned principle of negative parliamentarism.

In tranquil waters: Swedish cabinet ministers 161 The size of Swedish cabinets during the postwar era ranges from 16 ministers (Hansson 1945–46) to 22 (Reinfeldt 2006–), with an average size of 19 cabinet members (see Table 9.1). The number of consultative ministers varies across governments. Most PMs appoint about half as many consultative ministers as department heads. The last two cabinets (Persson III and Reinfeldt) however deviate from this pattern, appointing more consultative ministers than department heads. In addition, since each department head is typically accompanied by one or more junior ministers, the increase in the size of the cabinet also means that the number of junior ministers has increased over time. Constitutionally, there are two main routes to cabinet dissolution. Either the PM voluntary resigns or a majority of the members of parliament support a vote of no confidence directed against the PM. In both cases, the Speaker dismisses the remaining cabinet ministers. Since the Second World War, there have been four attempts to remove the PM and his cabinet through a vote of no confidence. None, however, has been successful.6 In addition, the Constitution prescribes that the cabinet shall be dissolved in the event of the death of the PM. The individual cabinet minister is subject to the same rules of resignation as the PM, with two exceptions. First, the PM may dismiss an individual minister, irrespective of whether the latter wishes to be separated from his or her position or not. Secondly, when a minister other than the PM chooses to step down, or is subjected to a successful vote of no confidence, he or she does not bring down any of the other cabinet ministers.7 In Table 9.1 we present an overview of the governments that are covered in this chapter, as well as some of their characteristics.

Selecting ministers As mentioned above, the typical Swedish government has historically been of the Social Democratic one-party kind. Hence, negotiations are, if existing, usually limited to the intra-party arena. Yet few Social Democratic PMs have experienced the need to get their staffing plan approved by the party or fractions of it. The most common way has instead been to discuss appointments in an informal setting, and with a limited number of members of the party elite (for example other party representatives with long experience of both party and government activities). The former PM Göran Persson, perhaps being the most secretive of them all, claims that he consulted only a handful of people among his closest friends (Fichtelius 2007: 56). Erlander is the exception, in the sense that he was very careful to make sure that he discussed appointments with the rest of the outgoing government. PMs heading coalition governments have much less leeway. For instance, the Centre Party leader Thorbjörn Fälldin, who headed the bourgeois coalitions in the 1970s, had little influence over the appointments to portfolios that had been allocated to parties other than his own (Larsson 1993, 166; Premfors et al. 2007: 97ff.). Usually, the number of portfolios allocated to each coalition member is made dependent on the relative size or strength of each party. In fact, portfolio allocation in Sweden is one of the most ‘proportional’ in Western Europe (see Bäck et al. 2009). In general, the largest party will get both the PM post and the largest

Per-Albin Hansson Tage Erlander Tage Erlander Tage Erlander Tage Erlander Tage Erlander Tage Erlander Tage Erlander Tage Erlander Tage Erlander Tage Erlander Olof Palme Olof Palme Olof Palme Thorbjörn Fälldin Ola Ullsten Thorbjörn Fälldin Thorbjörn Fälldin Olof Palme Olof Palme Ingvar Carlsson Ingvar Carlsson Carl Bildt

Ingvar Carlsson Göran Persson Göran Persson Göran Persson Fredrik Reinfeldt

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 Bildt

Carlsson III Persson I Persson II Persson III Reinfeldt

Social Democrats Social Democrats Social Democrats Social Democrats Conservatives

Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Centre Party Liberal Party Centre Party Centre Party Social Democrats Social Democrats Social Democrats Social Democrats Conservatives

PM’s party

Source: Own data, see also Bergman (2000).

Name of PM

Government

Table 9.1 Swedish cabinets, 1945–2007

06.10.94–18.03.96 21.03.96–20.09.98 20.09.98–15.09.02 15.09.02–17.09.06 05.10.06–

31.07.45–06.10.46 11.10.46–19.09.48 19.09.48–01.10.51 01.10.51–21.09.52 21.09.52–26.09.56 26.09.56–26.10.57 31.10.57–01.06.58 01.06.58–18.09.60 18.09.60–20.09.64 20.09.64–15.09.68 15.09.68–14.10.69 14.10.69–20.09.70 20.09.70–16.09.73 16.09.73–19.09.76 07.10.76–05.10.78 13.10.78–16.09.79 11.10.79–08.05.81 19.05.81–19.09.82 07.10.82–15.09.85 15.09.85–01.03.86 12.03.86–18.09.88 18.09.88–15.09.91 03.10.91–18.09.94

Date in–Date out Social Democrats Social Democrats Social Democrats Social Democrats, Centre Party Social Democrats, Centre Party Social Democrats, Centre Party Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Social Democrats Centre Party, Liberal Party, Conservatives Liberal Party Centre Party, Liberal Party, Conservatives Centre Party, Liberal Party Social Democrats Social Democrats Social Democrats Social Democrats Centre Party, Liberal Party, Conservatives, Christian Democrats Social Democrats Social Democrats Social Democrats Social Democrats Centre Party, Liberal Party, Conservatives, Christian Democrats

Parties in cabinet Yes Yes Yes No No No Yes Yes Yes Yes No No Yes Yes No Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No

21 22 22 22 22 22

Minority cabinet

16 16 16 18 16 16 18 15 15 16 18 19 19 20 20 18 21 20 19 19 18 19

Size

In tranquil waters: Swedish cabinet ministers 163 number of portfolios. For instance, the four parties forming the Reinfeldt coalition government in 2006 reportedly applied a distributional system that rendered Reinfeldt’s Conservative Party half the ministerial posts, as well as the first choice of two portfolios. This allowed the Conservative party to become the first coalition member ever to occupy the posts both of PM and of Minister of Finance. When Reinfeldt had made his first pick(s), the rest of the parties were allowed to choose one position each in an order based on the parties’ vote shares. Reinfeldt also asked the respective party leaders if they wished to match any of their political profiles with a certain portfolio. Furthermore, Reinfeldt’s government repeated the ideal introduced by Bildt (1991–94) of trying to keep departments of government ‘party homogeneous’: that is, to staff departments with ministers from the same party (Premfors et al. 2007: 97ff.). Despite there being few formal restrictions on who can become a minister, informal rules or norms limit the list of candidates and number of feasible compositions of the cabinet (Bergström 1987; Larsson 1993). These norms vary across different types of cabinets. For instance, the task facing a Social Democratic PM has been described as solving ‘a puzzle in a tight frame, a compromise between competence and different sorts of representativeness: male or female, union or party, Stockholm or the countryside’ (Elmbrant 1989: 264). In the following, we document the types of people who staff Swedish cabinets. In particular, we focus on three types of ministerial features: social representativeness, parliamentary and party background, and occupational background. In Figure 9.1 we present our results regarding the social representativeness of Swedish ministers. Although Swedish cabinets are fairly equal in composition between men and women nowadays, it was not until the 1970s that the number of female ministers started to increase more rapidly. The first woman to become a cabinet minister was the Social Democrat Karin Koch-Lindberg. She was appointed by Erlander in 1947 to a post without a portfolio. In total, 29 per cent of all ministers during the post-war era have been women, which can be compared to an average of 6.2 percent in postwar cabinets (1945–85) in Western Europe reported by Thiébault (1991: 20). For the same period, the Swedish average was 19 per cent female ministers, according to our own data. The former PM Ingvar Carlsson claims that his Social Democratic predecessor, Olof Palme, was the first PM to acknowledge the gender dimension in the recruitment of ministers. Carlsson made an official promise in the 1994 election campaign that he would appoint as many women as men to his new cabinet (Carlsson 2003: 550f.), and he kept his promise. At the time of first appointment, the average age of Swedish ministers is 48 years, which is the average age reported by Thiébault (1991: 21) for the Western European cabinets. During the period we study, only 20 per cent of ministers have been 40 years or younger. Although there has been an increase of younger ministers in recent years, this is not part of a clear trend. There seems instead to be a pattern of ministers getting older and being replaced in a cyclical way (see Figure 9.1). Social representativeness is, of course, one reason given for recruiting younger ministers. But several former PMs also stress that allowing young and promising individuals to take part in the cabinet serves to secure the succession

164 The selection of ministers in Europe 100 90 80 70 60 50 40 30 20 10

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 Bildt Carlsson III Persson I Persson II Persson III Reinfeldt

0

Women

Age 40 or younger

University education

Figure 9.1 Social representativeness of Swedish ministers, 1945–2007 Note: Each minister is only counted once per cabinet, which gives us a total of 593 observations.

of party leaders and, hence, potential PMs (Bohman 1984: 238; Carlsson 2003: 608ff.; Fichtelius 2007: 59ff., 109ff.). Examples of PMs to be who were recruited to cabinet at an early age are Olof Palme, who was appointed by Tage Erlander at the age of 36, and Ingvar Carlsson, who was appointed by Olof Palme at the age of 34. Most Swedish ministers (66 per cent) have graduated from higher education (university level), while 11 percent have undergone higher education without graduating. Taken together, then, 77 per cent of Swedish ministers have some form of university education, which is close to the Western European average (Thiébault 1991: 21). For the period studied by Thiébault (1945–85), about 73 per cent of the Swedish ministers have a university education. Finally, in recent years ethnicity has become a more important consideration. With the Instrument of Government of 1974 it became possible for people born abroad to become ministers. Before this, individuals who were born in another country could not become cabinet ministers, even if they had been Swedish citizens for a long time. So far, only four individuals with such a background have made their way into the cabinet.8 Social Democratic and bourgeois governments differ to some extent when it comes to social representativeness.9 First, the average level of university education is higher among bourgeois ministers (81 per cent) than among Social Democratic ministers (73 percent). Secondly, bourgeois governments are more gender

In tranquil waters: Swedish cabinet ministers 165 balanced (33 per cent women compared to 26 per cent). To some extent, this is an artefact of the long-lasting rule of the Social Democrats during the 1940s, 1950s and 1960s: a period when gender issues did not attract as much attention as they do today. If one instead compares figures from the 1970s and onwards – that is, starting with Olof Palme’s second government (since he inherited most ministers from Erlander in his first cabinet) – bourgeois and Social Democratic governments contain the same share of women, 36 per cent. Thirdly, the average age of a Social Democratic cabinet minister is slightly lower (47 years compared to 49 years). When compared to the population at large, Swedish cabinets can hardly be regarded as socially representative. Swedish cabinets have simply contained too many highly educated, Swedish-born, elderly men to be seen as representative. However, comparative studies have shown that Swedish cabinets appear to be more socially representative than most other cabinets in Western democracies (Blondel and Thiébault 1991). Turning to another important aspect of ministerial backgrounds, recruitment of members of parliament to the cabinet is often considered as a defining feature of parliamentary systems (see, e.g. Verney 1992) as previous research finds that the parliament is the main career path for becoming a minister. As has been shown by De Winter (1991), 75 per cent of Western European ministers (1945–85), on average, had been members of parliament before joining the cabinet. For the years 1945–2007, we find that 58 per cent of appointed ministers, counting first-time appointments only, had been a member of parliament (counting both chambers before the move to a unicameral legislature, see Figure 9.2).10 Thus, Sweden has had a comparatively low proportion of ministers with parliamentary background. Despite these comparatively low figures, Swedish PMs frequently underline the importance of parliamentary experience among their cabinet members. For instance, former PM Ingvar Carlsson argues that a ‘large share’ of ministers should have a parliamentary background, simply because the government is dependent upon the Riksdag (Carlsson 2003: 550). However, the stress placed on parliamentary experience varies across PMs. For instance, while Sweden’s most recent Social Democratic PM, Göran Persson, acknowledges the need for a parliamentary connection, he also stresses the need to broaden the recruitment base. One reason is that the Riksdag, according to Persson, lacks enough ‘ministerial material’, referring to the shortage of media experience among Swedish parliamentarians. Comparing himself to Tage Erlander, who was careful to maintain a tight parliamentary connection, Persson states that ‘Erlander lived in another Sweden’ (Fichtelius 2007: 64ff., quote on p. 64). That a cabinet minister has held a leading position in his or her party is also considered an important feature of parliamentary government. In Figure 9.2 we present the share of ministers (first-time appointments) held by individuals with a leading position in the national, regional or local organizational levels of their party, where background within the national party organization is most important. During the postwar era, 67 per cent of the ministers have had a party background. It is interesting to note that the most recent cabinet is the most party-politicized of the ones formed during the entire period studied; as we have shown elsewhere, it is even the most party-politicized since the introduction of

166 The selection of ministers in Europe 100 90 80 70 60 50 40 30 20 10

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 Bildt Carlsson III Persson I Persson II Persson III Reinfeldt

0

Parliamentary background

Party background

Figure 9.2 Parliamentary and party background of Swedish ministers, 1945–2007 Note: Each minister is only counted once per cabinet, which gives us a total of 593 observations. By parliamentary background we refer to ministers who have been members of parliament, considering both chambers before the abolishment of the bicameral legislature, before joining the cabinet. By party background we refer to ministers who have held leading positions in the national, regional or local organs of their party before entering the cabinet.

parliamentary democracy in Sweden in 1917 (Bäck et al. 2007a). This should give pause to scholars who claim that we have witnessed a decrease in the party professionalization of Swedish cabinets (e.g. Beckman 2007). Once again comparing Social Democratic and bourgeois governments, our data suggest that the latter place a stronger emphasis on parliamentary and party background. In bourgeois cabinets, 72 per cent of ministers have parliamentary experience and 81 per cent have a party background, while the corresponding figures for Social Democratic governments are 49 per cent and 60 per cent. We should however note that the Swedish bourgeois governments are typically coalition cabinets (all but one short-lived Liberal one-party cabinet), as opposed to the single-party Social Democrat cabinets, and it can be expected that coalition governments will include a higher proportion of ministers with parliamentary experience than single-party cabinets. The main reason for this is that each coalition party has fewer executive positions to fill and will tend to select its parliamentary heavyweights (see Bäck et al. 2007b). While PMs and ministers tend to agree that the party and parliament are important recruitment spots for the cabinet, it is also common to cite knowledge and

In tranquil waters: Swedish cabinet ministers 167 policy sector experience (expertise) as reasons for handing a particular individual a certain ministership (Carlsson 2003, 550; Feldt 1991, 48f.; Bohman 1984: 157f.). Studying ministers in the fields of economic and social policy, Beckman (2006) concludes that Swedish governments seldom contain true amateurs. On the contrary, expertise is and has always been present, at least if measured as professional and political experience (through participation in parliamentary committees or previous cabinet appointments) of a certain policy area. According to Beckman’s calculations, 88 per cent of ministers had at least some prior experience of the economic and social policy field before entering the cabinet. We have collected information only on general levels of occupational background, defined as experience from leading positions within trades unions, the private sector and the public sector. At a minimum, these measures indicate the extent to which experience from different societal sectors is sought after when cabinets are staffed, and document the degree of ‘corporatist’ representation in Swedish cabinets. Our data on occupational background, which are summarized in Figure 9.3, show that a large share of Swedish ministers have some form of background within these sectors. Many have held a high position in the public sector (64 per cent), but a sizeable portion also has experience from the private

100 90 80 70 60 50 40 30 20 10

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 Bildt Carlsson III Persson I Persson II Persson III Reinfeldt

0

Women

Age 40 or younger

University education

Figure 9.3 Occupational background of Swedish ministers, 1945–2007 Note: Each minister is only counted once per cabinet, which gives us a total of 593 observations. By union background, and private and public sector background, we refer experience from leading positions within trades unions, the private sector and the public sector.

168 The selection of ministers in Europe sector (33 per cent). Even though this is somewhat less common, Swedish cabinets have also included many former trades union officials (19 per cent). Looking at occupational background within the private and the public sector, we see a slight increase over time – a change that seems primarily to have occurred in the 1970s and 1980s. Note, however, that there was a remarkable drop in the share of ministers with public sector background in the government formed in 2006 (down to 52 per cent), and that the cabinet did not include any ministers with a trades union background. With regards to trades union background, this partly reflects the fact that Sweden’s most recent cabinet is a bourgeois one. But it is also part of a general trend towards recruiting fewer former union officials; especially if one compares with the 1960s. There is, however, a striking difference between Social Democratic governments, with a relatively high level of ministers with trade union background (29 per cent), and bourgeois governments with a considerably lower level of trades unionists in cabinet (7 per cent). And even in the most recent Social Democratic government (Persson III), a fair share of former union officials were included (22 per cent). It therefore seems premature to conclude, as some previous researchers have done, that the strong ties between the Social Democratic party and the trade union movement are vanishing (see e.g. Aylott 2003). Finally, and while many Social Democratic cabinet ministers have been recruited from the trades union movement, private sector experience appears to be more important when bourgeois governments are staffed – 44 per cent of the bourgeois ministers have held a high position in the private sector, compared to 25 per cent of the Social Democratic ministers. This can of course be related to the fact that, historically, the Social Democrats have defended a large public sector, whereas the bourgeois parties have closer ties to the private sector. To conclude our analysis of ministerial background, it is apparent that new ideals have been added to the list of desirable qualities of Swedish cabinet ministers. According to the former Prime Minister Göran Persson, he always had too many ministers in his government, since the only way to meet all new demands was to enlarge the government (Fichtelius 2007: 59). In 1945, Per-Albin Hansson formed a postwar cabinet consisting of 16 ministers contrasting with Göran Persson’s last cabinet, of 22 ministers when he resigned after the general election of 2006. Hence, whereas there are few formal requirements, there are a number of informal restrictions (norms and pressures) to which Swedish PMs must submit when staffing their cabinet. The nature of these restrictions appears to mirror developments in society at large, making the recruitment of ministers a dynamic and multifaceted exercise.

Cabinet reshuffles and dissolutions Once selected, the appointee has to face the reality of life as a cabinet minister. First, there is the risk of being reshuffled; some ministers are moved to different cabinet posts and many ministerial careers end with a, more or less, forced resignation. Secondly, and even if the minister is never shuffled, he or she will, sooner or later, experience events that lead to the resignation of the PM, and the dissolution of the entire cabinet. The most short-lived ministers during the period studied herein were Fredrik Reinfeldt’s Minister of Culture, Cecilia Stegö Chilò,

In tranquil waters: Swedish cabinet ministers 169 and Minister of Trade, Maria Borelius; both were shuffled out of office owing to personal economic scandals after only ten and eight days, respectively, in 2006. By contrast, the most long-lived ministers have lasted the whole 48 months of the legislative mandate (before 1998, general elections were held every third year).11 Figure 9.4 shows the Kaplan-Meier survivor functions for all postwar governments. The horizontal axis gives the number of months, and the vertical axis indicates how large a share of the ministers ‘survive’ this long, or longer. Dents in the graphs indicate that one, or several, ministers have been shuffled to another post, or out of the government. When looking at the graphs, the reader should be aware that, as we write this, the Reinfeldt government is still in office. When it comes to the question of which cabinets have the more turbulent inner lives, the panels of Figure 9.4 are highly suggestive. Cabinets led by Erlander, Persson and, to some extent, Carlsson stand out. The most recent Social Democratic PM, Göran Persson, lives up to his reputation as a comparatively ‘shuffle-happy’ PM. While this reputation is certainly well-deserved if one considers only his third term in office, the overall picture shows that he is not as exceptional as some Swedish journalists and political scientists would have us believe (e.g. Möller 2006; Ruin 2004; Svenning 2005; Ögren 2005). The fact of the matter is that, when reading Tage Erlander’s memoirs, it becomes apparent that the hiring and firing of ministers was almost constantly on his mind (see e.g. Erlander 2001a, 2001b, 2003). And, indeed, the survival graphs show that he reshuffled liberally and consistently in several of his periods in office. By and large, the same conclusion holds for Ingvar Carlsson. That Carlsson reshuffled his first and second governments on several occasions may come as a surprise to some; among Swedish scholars and pundits he has a reputation for having managed his cabinets with unusual delicacy. However, one has to keep in mind that he inherited most of his cabinet ministers from Olof Palme, whose murder in 1986 made Carlsson the PM. As should be clear by now, the graphs in Figure 9.4 do not exhibit any clear trend over time. Rather, it appears that the extent to which the PM reshuffles has a great deal to do with the type of government he or she is leading. The survival graphs of the bourgeois coalition cabinets look strikingly similar;12 there is not much turnover in cabinet personnel. By comparison, Social Democratic cabinets – all except Erlander III, IV, V, X and Palme I were single-party minority governments – were reshuffled more frequently. Our findings thus dovetail nicely with those in previous, and more comparatively oriented, studies, where reshuffles have been shown to be less common in coalition governments (Budge 1985; Huber and Martinez-Gallardo 2003). What, then, are the principal reasons for turnover in cabinet personnel? The larger dents in the survival graphs in Figure 9.4 represent extensive reshuffles; some members are transferred to new posts while others are shuffled out. Of course, many ministers do not leave the government in an extensive reshuffle, but one at a time. According to a recent study (Werner 2007), focusing on all cabinet exits during the years 1945–2007, the reasons vary a great deal. Close to 40 per cent step down because of personal reasons, illness or advanced age. Around 6 per cent leave because they have received more prestigious offers; as when Björn von Sydow left his post as Minister of Defence in 2002 to become Speaker of parliament, or when

0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00

0

0

0

0

24 Months

36

24 Months

36

24 Months

36

12

24 Months

36

Palme II 1970–73

12

Erlander VIII 1960–64

12

Erlander IV 1952–56

12

Hansson 1945–46

48

48

48

48

0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0

0

0

0

24 Months

36

24 Months 36

24 Months

36

12

24 Months

36

Palme III 1973–76

12

Erlander IX 1964–68

12

Erlander V 1956–57

12

Erlander I 1946–48

48

48

48

48

0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0

0

0

0

24 Months

36

24 Months 36

24 Months

36

12

24 Months

36

Fälldin I 1976–78

12

Erlander X 1968–69

12

Erlander VI 1957–58

12

Erlander II 1948–51

48

48

48

48

0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0

0

0

0

24 Months

36

24 Months

36

24 Months

36

12

24 Months

36

Ullsten 1978–79

12

Palme I 1969–70

12

Erlander VII 1958–60

12

Erlander III 1951–52

48

48

48

48

0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00

0

0

0

24 Months

36

24 Months

36

12

24 Months

36

Persson I 1996–98

12

Carlsson I 1986–88

12

Fälldin II 1979–81

48

48

48

0

0

0

24 Months

36

24 Months 36

12

24 Months

36

Persson II 1998–2002

12

Carlsson II 1988–91

12

Fälldin III 1981–82

48

48

48

0

0

0

24 Months

36

24 Months 36

12

24 Months

36

Persson III 2002–6

12

Bildt 1991–94

12

Palme IV 1982–85

Figure 9.4 Ministerial survival functions by government in Sweden, 1945–2007

0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00

48

48

48

0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0

0

0

24 Months

36

24 Months

36

12

24 Months

36

Reinfeldt 2006–

12

Carlsson III 1994–96

12

Palme V 1985–86

48

48

48

172 The selection of ministers in Europe Dag Hammarskjöld was elected General Secretary of the United Nations in 1953. Several ministers also leave because of what Dowding and Kang (1998) call ‘resignation issues’. In comparison, personal errors and economic scandals account for a larger proportion of ministerial exits in Sweden than in the UK. Examples of publicized resignations due to personal errors include Minister of Justice Sten Wickbom’s in 1987, after the escape of the Soviet Union spy Stig Bergling, and Minister of Justice Anna-Greta Leijon in 1988, who was accused in the media of violating the Constitution in her efforts to contribute to the criminal investigation on the Palme murder, and then threatened by a declaration of no confidence from the parliamentary opposition. In a more recent affair, Foreign minister Laila Freivalds resigned following critique in the media, by journalists and victims’ relatives, and later also by the parliamentary Committee on the Constitution, for her handling of the 2004 tsunami catastrophe. When she shortly thereafter was accused of being involved in the shutting down of a Swedish website reprinting the infamous cartoons of the prophet Muhammad, she was forced to resign. As for personal economic scandals, affairs related to taxes and finances are politically sensitive in Sweden (see e.g. Kratz 1996: 232ff.). Even formally legitimate economic transactions have led to ministerial resignations. In 1983, Olof Palme’s Minister of Justice Ove Rainer was accused in the media of largescale (but legal) tax planning. And in 2000, the aforementioned Freivalds had to resign, mainly as a result of internal critique following the media attention after she used the potentially lucrative option of buying her municipally owned apartment in Stockholm, which was problematic since the Social Democratic Party in Stockholm was against the selling-out of municipally owned apartments. Among the resignations due to financial scandal, we also find ministers who have been involved in outright illegal transactions. Present PM Reinfeld quickly fired two ministers (respectively for Culture and Trade, Stegö Chilò and Borelius) from his party because of journalists pointing to the fact that neither had paid employer taxes when hiring nannies, and for the former also because she had not paid her television licence fee. The same Reinfeld lost another member of his party one year after the formation of his cabinet, owing to disagreement between on the one hand the PM and the Minister of Finance and on the other the Minister of Defence over the budget of the latter’s department. The Minister of Defence had to go, thereby confirming previous research arguing that ministers in coalition governments may be as short-lived as in single-party cabinets if they belong to the PM’s party. Although this last resignation was forced because of policy disagreement between a minister and the PM, it is notable that this reason for ministerial termination is seldom notified (Werner 2007). In the literature on the Social Democratic Party and its leadership, one often refers to a norm of loyalty and the ambition to keep up the appearance of unity to external observers (see e.g. Barrling Hermansson 2004). For instance, Erlander was reportedly very upset when, in 1966, Ulla Lindström suddenly left the cabinet because of her dissatisfaction with the government’s foreign aid-policy (Elmbrant 1989: 66f.). Therefore, it is probably sensible to assume that reasons other than disagreements are sometimes used as window-dressing when, in reality, individual ministers leave because their political views are at odds with that of the government.

In tranquil waters: Swedish cabinet ministers 173

0.00 0.25 0.50 0.75 1.00

Age

0.00 0.25 0.50 0.75 1.00

Sex

Months Men

Months Women

40

Months

Months

No

Yes

No

Party background

Yes

0.00 0.25 0.50 0.75 1.00

0.00 0.25 0.50 0.75 1.00

Union background

Months No

Months Yes

No

Public sector background 0.00 0.25 0.50 0.75 1.00

Privte sector background 0.00 0.25 0.50 0.75 1.00

Yes

Months No

Months Yes

No

Yes

Figure 9.5 Ministerial survival functions by background characteristics in Sweden, 1945–2007

174 The selection of ministers in Europe If we take a step back – away from the more proximate causes of ministerial turnover – we can take a look at the more general patterns in who is reshuffled. In Figure 9.5 we show the Kaplan-Meier survivor functions of ministers grouped by the background characteristics discussed in the previous section. Overall, there does not seem to be any strong relationships between background characteristics and ministerial turnover. But, and on a purely descriptive level, the variables related to ministerial survival seem to be party and parliamentary background, sex and, as one might expect, age. These results do suggest that PMs are more reluctant to reshuffle ministers who have a certain type of political experience and, perhaps, a strong base in the party. We can think of (at least) two explanations for the fact that female ministers are less likely to be reshuffled if there does, in fact, exist a causal relationship. One is that recent pressures to include as many women as men in the cabinet have made PMs wary of reshuffling female ministers. The other is that female ministers are, on average more competent than their male counterparts. That is, if one makes the reasonable conjecture that several reasons have converged to make it harder for female politicians to ‘rise to the top’, we should expect only the most ambitious and competent to make it there. Looking at our data, and some variables that could be argued as gauging ‘competency’ (for instance, education, parliamentary and party experience), we find some support for this explanation. A higher share of female ministers (82 per cent) have undergone higher education than male ministers (71 per cent). Moreover, 63 per cent of female ministers have parliamentary experience while the corresponding figure for male ministers is about 57 per cent, and 71 per cent of the female ministers have held a leading position in the party they belong to, while only 60 per cent of the male ministers have a similar background.13 Finally, and as mentioned above, the majority of ministerial careers come to a halt in a cabinet dissolution. Usually, this will occur in the aftermath of an electoral defeat significant enough to remove the parliamentary support of the cabinet, but there exist exceptions to this rule. On a couple of occasions, PMs have handed in their resignation following policy conflicts among the parties in the cabinet. Disagreements over pensions policy ended the Social Democrats’ coalition with the Centre Party in 1957, and the conflict over nuclear energy brought an end to the Fälldin I government. In 1990, Carlsson I was dissolved after the PM made clear that a vote on a bill to freeze wages and prices was a vote of confidence; a threat he would subsequently have to go through with. In addition, during the postwar years, Sweden has seen two PMs pass away while in office, leading to cabinet dissolutions in 1946, when Per-Albin Hansson died from a heart attack, and in 1986, when Olof Palme was murdered.

Conclusions At the beginning of this chapter we asked what hides beneath the surface of the durable and, to a large part, continuous Social Democratic rule that characterizes postwar Sweden. It goes almost without saying that a large majority of Swedish cabinet ministers have been Social Democratic partisans. Our data – which cover

In tranquil waters: Swedish cabinet ministers 175 all ministerial appointees in Sweden during the years 1945–2007 – show that far from all ministers have had previous experience from the parliament, or held important positions in their party prior to joining the cabinet. What is more – although PMs and ministers tend to agree that the party and the parliament are important recruitment spots for the cabinet – PMs appear to differ widely when it comes to the stress they place on this type of experience. Another important selection criterion for PMs – for some more than others – is experience from working within a certain policy sector. All in all, we do not find a clear trend with respect to changes in party, parliamentary and expert background among Swedish ministers. For instance, we report historically high levels of parliamentary and party experience, and unusually low levels of policy sector expertise, in the most recently formed government. Using the same data, we also chart the durability of ministers; the fluidity with which personnel, once selected, are moved around, as well as out of, the cabinet. As we look at the durability of ministers, rather than cabinets, we find that the stylized view of Swedish cabinets has to be nuanced. There is a great deal of personnel turnover in Swedish cabinets. Again, PMs differ widely. While some are extremely ‘shuffle-happy’, others make only marginal adjustments to their cabinet personnel. However, there appears to be no long-term trend. Rather, the extent to which the PM reshuffles seems to have a great deal to do with the type of government he leads. In particular, our findings mesh well with those in comparative studies, where reshuffles are shown to be less common in coalition governments (Budge 1985; Huber and Martinez-Gallardo 2003). If nothing else, our findings should caution scholars not to make too much of the fact that the Social Democrats have ruled most of the years since the Second World War. Even in this stable setting, ministerial selection criteria vary a great deal; some over time, others between PMs. And, even if most ministers end their careers in a cabinet dissolution, many careers are cut short when appointees are moved about, or shuffled out of, the cabinet.

Notes 1 As early as the eighteenth century the King and his council were in fact dependent upon the parliament for their political power. This changed when Gustav III ‘convinced’ the members of parliament to accept a new constitution by demonstrating his military powers outside parliament. Even though this Constitution prescribed the separation of powers, in practice the King centralized political power to himself. Parliamentarism ‘as we know it’ is, however, usually regarded as having been firmly established in 1917 (see Larsson 1994: 53f., 171). 2 The Swedish Constitution consists of four fundamental laws: the Instrument of Government (Regeringsformen), the Act of Succession (Successionsordningen), the Freedom of the Press Act (Tryckfrihetsförordningen), and the Fundamental Law on Freedom of Expression (Yttrandefrihetsgrundlagen). The most important of the fundamental laws is the Instrument of Government, where chapters 6 and 7 describe the rules for government formation and the work of the government. 3 Before 1974, a minister did not have to resign temporarily from parliament. However, a decision from 1972 to use a system of replacements for MPs who were appointed Speakers or ministers who had to be temporarily freed of their duties for other reasons was incorporated into the new Constitution of 1974.

176 The selection of ministers in Europe 4 The Instrument of Government of 1809 prescribed that six members of government were to be ‘consultative’. An amendment in 1840 specified that three of the ministers should be consultative, of whom two were to have a professional background as civil servants. These were given the specific task of upholding the rule of law, i.e. to scrutinize the government’s proposals from a formal and judicial perspective. The specific demands concerning the consultants’ professional background in practice meant that they would have training in the field of law. In 1941 it was specified that the lowest number of ministers who were not heads of a governmental department should be three (Larsson 1990: 138ff., 1994: 169ff., 208; Holmberg et al. 2006: 285). 5 The Social Democrats and the Centre Party were also engaged in close co-operation in the 1990s. Although not forming a coalition government, Anders Ljunggren, a member of the Centre Party elite, was given an office inside the Ministry of Finance, alongside the other politically appointed civil servants (Lagercrantz 2005: 183ff.). The Ljunggren model has since been extended into what Bergman and Aylott call ‘contract parliamentarism’. Over time as increasingly intense and stable co-operation emerged between Social Democrats, the Green Party and the Left Party, whilst denied the role of coalition members, the Green and Left Parties were given the chance to staff the departments of governments with a number of politically appointed civil servants or advisers (Bergman and Aylott 2003; Ullström 2005; Olsen 2007; see also Bale and Bergman 2006). 6 The only PMs who have been the subject of votes of no confidence are Thorbjörn Fälldin (1980) and Göran Persson (1996, 1998 and 2002). 7 In addition, there exists the possibility that the Committee on the Constitution (a parliamentary committee with the task of scrutinizing the work of the Government and its ministers) may charge a minister with misconduct, and have the case tried by the Supreme Court. This option has never been used. 8 Anita Bondestam (born in Poland, appointed Minister of Communications in 1978); Laila Freivalds (born in Latvia, appointed Minister of Justice in 1988); Ibrahim Baylan (born in Turkey, appointed Deputy Minister of Education in 2004); and Nyamko Sabuni (born in exile in Burundi, her parents were refugees from Congo-Kinshasa, appointed Minister of Integration in 2006). 9 There have been 126 Social Democratic, 93 bourgeois and five ministers without party affiliation in Swedish governments between 1945 and 2007. 10 When studying the same period as De Winter, the results are similar, that is, 58 per cent had been a member of parliament. 11 The unit of analysis in the subsequent survival analyses is the ministerial spell. A ministerial spell is the time spent by an individual on a particular cabinet post in a particular cabinet (see Box-Steffensmeier and Jones 2004). A ministerial spell can end for two reasons: the minister (1) leaves the cabinet or (2) is shuffled to another cabinet post. A ministerial spell is regarded as being censored in all other cases, meaning those cases where the minister survives until the end of the cabinet. In case the ministerial spell is ended or censored within three weeks before or after a new cabinet is formed, the final date is set to three weeks prior to the formation of the new cabinet. Finally, a ministerial spell starts when an individual is appointed to a particular cabinet position, unless he or she is appointed to that position during the three weeks following the formation of a new cabinet, in which case the starting date is set to three weeks after the new cabinet is formed. Spells are thus defined the same way as in Berlinski et al. (Chapter 4 in this volume), except that their data does not analyse for moving ministerial position and the fact that they use a four-week window centred on the date of formation of a new cabinet, whereas we use a six-week window. The major reason for this divergence is that some major post-election reshuffles – most notably the largest post-election reshuffle in Swedish history, carried out by Göran Persson after his narrow electoral win in 1998 – occurred in the third week after the election. Note that we exclude PMs from our analyses of ministerial survival. Since the death of a minister necessitates a reshuffle, the spells that end in death are included in the data. 12 Except Reinfeldt’s bourgeois coalition cabinet, which is still in its infancy.

In tranquil waters: Swedish cabinet ministers 177 13 Regarding occupational background from trades unions, the private sector and the public sector, the results are somewhat more ambiguous. We find hardly any differences between male and female ministers with previous experience from the trades unions (22 per cent among male ministers and 21 per cent among female ministers) and the public sector (61 per cent among male ministers and 63 per cent among female ministers). The differences become larger when comparing ministers with private sector background (39 per cent among male ministers but only 8 per cent among female ministers).

References Aylott, N. (2003) ‘After the Divorce: Social Democrats and Trade Unions in Sweden’. Party Politics 9: 369–390. Bäck, H. (2003) Explaining Coalitions. Evidence and Lessons From Studying Coalition Formation in Swedish Local Government. Uppsala: Acta Universitatis Upsaliensis. Bäck, H., Persson, T., Vernby, K. and Westin, L. (2007a) Från statsminister till president? Sveriges regeringschef i ett jämförande perspektiv. Stockholm: Fritzes. Bäck, H., Dumont, P., Meier, H. E. Persson, T. and Vernby, K. (2007b) ‘Does European Integration Lead to a Presidentialization of Executive Politics? Ministerial Selection in Swedish Post-War Cabinets’. Paper presented at the ECPR General Meeting in Pisa, September 6–8. Bäck, H. and Dumont, P. (2008) ‘Making the First Move. A Two-Stage Analysis of the Role of Formateurs in Parliamentary Government Formation’. Public Choice 135: 353–373. Bäck, H., Persson, T. and Meier, H. E. (2009) ‘Party Size and Portfolio Payoffs. The Proportional Allocation of Ministerial Posts in Coalition Governments’. Journal of Legislative Studies, forthcoming. Bale, T. and Bergman, T. (2006) ‘Captives No Longer, but Servants Still? Contract Parliamentarism and the New Minority Governance in Sweden and New Zealand’. Government and Opposition 41: 422–449. Barrling Hermansson, K. (2004) Partikulturer. Uppsala: Acta Universitatis Upsaliensis. Beckman, L. (2006) ‘The Competent Cabinet? Ministers in Sweden and the Problem of Competence and Democracy’. Scandinavian Political Studies 29: 111–129. Beckman, L. (2007) ‘Professionalisation of Politics Reconsidered: A Study of the Swedish Cabinet 1917–2004’. Parliamentary Affairs 60: 66–83. Bergman, T. (1995) Constitutional Rules and Party Goals in Coalition Formation. Umeå University: Department of Political Science. Bergman, T. (2000) ‘When Minority Cabinets Are the Rule and Majority Coalitions the Exception’. In Müller, C. W. and Strøm, K. 2000 (eds) Coalition Governments in Western Europe. Oxford: Oxford University Press. Bergman, T. and Aylott, N. (2003) ‘Parlamentarism per kontrakt? Blir den svenska innovationen långlivad?’ In Riksdagens årsbok 2002/03. Stockholm: Riksdagen. Bergström, H. (1987) Rivstart? Om övergången från opposition till regering. Stockholm: Tidens Förlag. Blondel, J. and Thiébault, J.-L. (1991) The Profession of Government Minister in Western Europe. Basingstoke: Macmillan. Box-Steffensmeier, J. M. and Jones, B. S. (2004). Event History Modeling: A Guide for Social Scientists. New York: Cambridge University Press. Bohman, G. (1984) Maktskifte. Stockholm: Bonniers. Brothén, M. (2003) Riksdagsenkät 2002. Dokumentation, stencil. Göteborg: Department of Political Science.

178 The selection of ministers in Europe Budge, I. (1985) ‘Party Factions and Government Reshuffles’. European Journal of Political Research 13: 327–333. Carlsson, I. 2003. Så tänkte jag. Politik och dramatik. Stockholm: Hjalmarson & Högberg Bokförlag. De Winter, L. (1991) ‘Parliamentary and Party Pathways to the Cabinet’. In Blondel, J. and Thiébault, J.-L. (eds) The Profession of Government Minister in Western Europe. Basingstoke: Macmillan. Dowding, K. and Kang, W.-T. (1998) ‘Ministerial Resignations 1945–97’. Public Administration 76: 411–429. Elmbrant, B. (1989). Palme. Stockholm: Fischer & Co. Erlander, T. (2001a) Tage Erlander Dagböcker 1945–1949. Hedemora: Gidlunds förlag. Erlander, T. (2001b) Tage Erlander Dagböcker 1950–1951. Hedemora: Gidlunds Förlag. Erlander, T. (2003). Tage Erlander Dagböcker 1953. Hedemora: Gidlunds Förlag. Feldt, K.-O. (1991) Alla dessa dagar. I regeringen 1982–1990. Stockholm: Norstedts. Fichtelius, E. (2007) Aldrig ensam, alltid ensam. Samtalen med Göran Persson 1996– 2006. Stockholm: Norstedts. Holmberg, E., Stjernquist, N., Isberg, M., Eliason, M. and Regner, G. (2006). Grundlagarna. Regeringsformen, Successionordningen, Riksdagsordningen. Stockholm: Norstedts Juridik AB. Holmberg, S. (2000) Välja parti. Stockholm: Norstedts Juridik. Huber, J. D. and Martinez-Gallardo, C. (2003) ‘Cabinet Turnover in Parliamentary Democracies.’ Paper presented at the annual meeting of the American Political Science Association in Philadelphia, August. Kratz, A. (1996). Ingvar Carlsson. Erlanders sista pojke. Stockholm: Bonnier Alba. Lagercrantz, A. (2005) Över blockgränsen: samarbetet mellan Centerpartiet och Socialdemokraterna 1995–1998. Hedemora: Gidlunds Förlag. Larsson, T. (1990). ‘Regeringen och regeringskansliets organisationsstruktur, beredningsoch beslutsformer under 150 år’. In Larsson, U. (ed.) Att styra riket. Regeringskansliet 1840–1990. Stockhom: Allmänna Forlaget. Larsson, T. (1993). Det svenska statsskicket. Lund: Studentlitteratur. Möller, T. (2006). ‘Nödvändigt se över reglerna för Perssons maktställning’. Dagens Nyheter, 14 August. Ögren, M. (2005) Makten framför allt: En antologi om statsminister Göran Persson. Stockholm: Wahlström & Widstrand. Olsen, L. (2007) Rödgrön reda. Regeringssamverkan 1998–2006, Stockholm: Hjalmarsson & Högberg Förlag. Premfors, R. and Sundström, G. (2007) Regeringskansliet. Malmö: Liber. Ruin, O. (2004) ‘Statsministerämbetets förändring’. In Ruin, O. (ed.) Politikens ramar och aktörer. Femton uppsatser om vårt politiska system tillägnade Ingvar Carlsson. Stockholm: Hjalmarson och Högberg. Svenning, O. (2005) Göran Persson och hans värld. Stockholm: Norstedts Förlag. Thiébault, J.-L. (1991) ‘The Social Background of Western European Cabinet Ministers’. In Blondel, J. and Thiébault, J.-L. (eds) The Profession of Government Minister in Western Europe. Basingstoke: Macmillan. Ullström, A. (2005) Samarbetskanslierna – vårtor på, länkar till eller delar av Regeringskansliet? Score report no. 2005:3. Stockholm: Score. Verney, D. V. (1992 [1959]) ‘Parliamentary Government and Presidential Government’. In Lijphart, A. (ed.) Parliamentary Versus Presidential Government. Oxford: Oxford University Press. Werner, J. (2007). Ansvarsbefriade ministrar? En studie av ministeravgångar i svenska politik 1945–2006. Master’s Thesis, Department of Government, Uppsala University.

10 Constructing and maintaining Irish governments Eoin O’Malley

Constitutional framework Ireland is a parliamentary democracy which, until 1921, was a full, if not fully integrated part of the United Kingdom. In 1921, 26 of the 32 counties achieved effective independence and since 1949 Ireland is a republic. Ireland’s legislative institutions or Oireachtas consist of two Houses (Dáil Éireann and Seanad Éireann) and the Presidency.1 For legislation to become law it must be passed by both Houses and then enacted by the President (see Mitchell 2003 for a good overview of the Irish political system) The President is directly elected for a seven-year term with a two-term limit. It is a largely ceremonial post and elections are not always held as at times the political parties effectively agree on a candidate. The President does have some potentially important powers: s/he can refer a Bill to the Supreme Court; and refuse to dissolve the Dáil on the advice of a Taoiseach (prime minister) who has lost the confidence of the Dáil. The Dáil is the dominant of the two houses and can override Senate objections easily. Forty-three of the 60 Senators are indirectly elected by TDs (MPs), outgoing senators and local councillors in elections that directly follow the general election and as such the Seanad tends to mirror the party structure in the Dáil except that small parties find it difficult to achieve representation through election. Eleven senators are nominated by the incoming Taoiseach and six are elected by graduates of the two main universities (Trinity College, University of Dublin and National University of Ireland). This has given the government an inbuilt majority in the Senate, though this may not be that important as the Senate can only postpone legislation by 90 days. One hundred and sixty-six TDs are directly elected to Dáil Éireann from 43 three-, four- and five-seat constituencies using Proportional Representation by Single Transferable Vote. This electoral system has the effect of pitting candidates from the same party against one another, often for a more-or-less fixed party vote. This arguably has certain effects on the type of candidate running. Where no seat is ‘safe’, and candidates will run against candidates from their own party as much as against candidates of others parties, electoral competition is frequently based on local issues and rivalries. It is now very rare for candidates who are not from the constituency to run, and it is increasingly uncommon for TDs to live outside their constituency. This may deter suitable ‘ministrables’ from running and makes

180 The selection of ministers in Europe it more difficult for the party leadership to impose ‘notables’ against the wishes of the local party with any guarantee of success. The importance of local activism means that it is common for independent or non-party candidates to be elected. As this is the case – there were five independents returned in 2007 – parties forming a government can sometimes depend on support from independents, which usually comes at a cost of funding projects in the independents’ constituencies (Weeks 2004). Following an election, the new Dáil will meet and immediately appoint a Ceann Comhairle (Speaker). The allocation of this post is important because it frequently affects the arithmetic of forming a government. While the likely incoming government may want to appoint a member from its own ranks, it can neutralize one of its opponents’ votes by supporting the appointment of someone from outside the governing parties.2 The Ceann Comhairle votes only when there is a tied Dáil vote and by tradition will support the government in such cases. If it is not used in this way it can be a useful post to offer a senior TD who may have expectations of a cabinet position or to salve the disappointment of a minister not being reappointed. The Dáil will then vote to nominate a Taoiseach. The Taoiseach must receive a simple majority of those voting – 83 votes assuming that all but the Ceann Comhairle vote. The Taoiseach is then appointed by the President, who has no discretion. The Taoiseach is nearly always the leader of the largest party in government (see below for a discussion of the exception). In the postwar period Ireland has had one dominant party, Fianna Fáil, which ruled as a single party for long periods. Between 1944 and 1989 there was a pattern of Fianna Fáil single-party government alternating with coalition governments Table 10.1 Irish governments since 1973 Year 1973 1977 1979 1981 1982 (March) 1982 (December) 1987 1989 1992 (February) 1993 1994 1997 2002 2007 2008

Party or parties in government Fine Gael, Labour Fianna Fáil Fianna Fáil Fine Gael, Labour Fianna Fáil Fine Gael, Labour Fianna Fáil Fianna Fáil, Progressive Democrats (PDs) Fianna Fáil, PDs Fianna Fáil, Labour Fine Gael, Labour, Democratic Left Fianna Fáil, PDs Fianna Fáil, PDs Fianna Fáil, Greens, PDs Fianna Fáil, Greens, PDs

Taoiseach

Government type

Liam Cosgrave Jack Lynch Charles Haughey Garret FitzGerald Charles Haughey Garret FitzGerald Charles Haughey

Majority Majority Majority Minority Minority Majority Minority

Charles Haughey Albert Reynolds Albert Reynolds

Majority Majority Majority

John Bruton Bertie Ahern Bertie Ahern Bertie Ahern Brian Cowen

Majority Minority Majority Majority Majority

Note: In 1994 the Reynolds government was replaced by the Bruton government without recourse to an election.

Constructing Irish governments 181 of Fine Gael, Labour and others. Since 1989 coalition government has become the norm; Fianna Fáil abandoned its principle of not entering a coalition, forming a government with the Progressive Democrats, and later with the Labour Party. Even in 2002 when Fianna Fáil could have conceivably formed a stable singleparty minority government with the support of a few independents, it chose to maintain its coalition arrangement with the Progressive Democrats. Government formation usually takes place in the week or two between the election and the first meeting of the new Dáil. Only once has a Dáil failed to nominate a Taoiseach on its first sitting. The programme for government is negotiated between sets of representatives from each party who use the manifestos as a basis for the coalition agreement. Issues that cannot be agreed are then dealt with by the party leaders. The party leaders also meet to discuss the allocation of ministerial appointments between the parties, which includes the number of posts and which portfolios each party will be given. Having been elected, Taoisigh (the plural for Taoiseach) then put together their cabinets and must return to the Dáil for formal approval. It has never occurred that the Dáil has rejected a newly elected Taoiseach’s cabinet but this is a possibility, and a Taoiseach would have to resubmit a cabinet to which the Dáil did not object. However, on one occasion the debate raised information about the nominated minister which caused the Taoiseach to withdraw his nomination (see below). In effect the Taoiseach appoints ministers and his only constraint is his coalition partners or in the case of a faction, the leaders of factions in his party. Again the President formally appoints the ministers. There is another group of ministers – ministers of state or junior ministers who are appointed by the government on the nomination of the Taoiseach. So the posts that are in a government’s pay are numerous. There is a constitutional maximum of 15 cabinet ministers (including the Taoiseach) who head up departments of state and are accountable to parliament for the actions of their department. The minimum number of ministers allowed is seven, but from 1973 Taoisigh have filled all 15 posts allowed to them. Since 1939, under an amendment to the Ministers and Secretaries Act (1924), the government is free to change the titles and allocations of responsibilities of these departments. The ministers are also subject to the rules of individual and collective responsibility, so ministers, as ‘corporation sole’ are legally responsible for the actions of their departments, and are politically collectively responsible for the decisions of the government. There is legislative provision to appoint up to 20 ministers of state. While in the past cabinet ministers would grant junior ministers duties and responsibilities within the department, these now usually have some special pre-assigned responsibility within a single department or could be given a post that straddles departments with a co-ordinating function, such as Minister for European Affairs. Cabinet ministers do not have a choice as to who gets assigned to their departments, though the Taoiseach may consult them. The coalition party leaders however are asked as to who from their own party is chosen and there are cases which indicate that each party leader may have a veto on certain appointments though, as these are rarely important appointments, party leaders are seldom involved. The Taoiseach also appoints a government Chief Whip who is a junior minister, but sits at cabinet

182 The selection of ministers in Europe to advise cabinet on the backbench’s likely reaction to certain policy proposals. In addition the Taoiseach appoints an Attorney General or senior legal adviser to the government. The AG, usually a senior lawyer close to one of the government parties, sits at cabinet but can speak only when requested to do so. AGs can be a member of one of the houses, though this has not occurred for over thirty years. However in 1999 a former TD of the junior party was appointed Attorney General and he was later appointed as a cabinet minister. In all this gives the Taoiseach about 35 posts to distribute. Given the size of a government’s parliamentary support (about 85), this number amounts to a post for every two people – when you remove those ineligible through age (too young or old), infirmity or lack of interest. There are very strict and restrictive rules governing the set from which ministers can be chosen. Ministers must be members of one of the Houses of the Oireachtas and according to the relatively rigid Constitution all but two cabinet ministers must be members of the Dáil. Two ministers can be chosen from the Seanad or senate but these cannot be appointed to the three most senior posts of Taoiseach, Tánaiste (deputy prime minister) and Minister for Finance, who must be TDs. As the Taoiseach has the right to appoint eleven senators discretionally, in theory any Irish citizen over the age of 21 could be appointed as a minister. In practice ministers are rarely drawn from the Seanad.3 There are good reasons for this. It is impractical, as a senator cannot take up the post until the new Seanad meets, which can be a number of months after the government is formed. Given the tight arithmetic of Dáil majorities, backbench TDs will expect to be rewarded for their support and loyalty. This reward usually comes in the form of political positions. Giving a post to someone who is not a member of the Dáil could have the effect of creating a sense of independence between the parliamentary parties of the government and the government itself. Such independence may manifest itself in a willingness to vote against the government. Much of the job of a government minister is to guide legislation through the Dáil; this is much more difficult if the minister is not aware of the rules of the Dáil and less familiar with the people in the Dáil. One former Taoiseach, Charles J. Haughey, thought that ministers should be respected and known in the Dáil (O’Malley interview).4 A senator might be resented by government backbenchers who feel that the minister has taken one of ‘their’ jobs. Another, Garret FitzGerald, was told by the chairman of his parliamentary party that it was insulting that he thought none of them capable of doing the job when he appointed James Dooge to the government (O’Malley interview). It is possible to put down a motion of no confidence in a minister, though this has no constitutional standing. No such motion has ever been passed, though if it were this would represent dissent from within the government party or parties and the Taoiseach would presumably be under some political pressure to sack that minister.

Constructing Irish governments 183

Selecting ministers The politics of ministerial selection Though constitutionally the Taoiseach chooses the cabinet and junior ministers, certain other political actors influence and constrain this formal discretion. As a government needs to retain the confidence of the Dáil it is important that a government has a reliable majority in that house. It is necessary then to construct durable coalitions and to maintain a loyal party. Taoisigh and party leaders need to use all the resources at their disposal to control their colleagues in party and government. In a country where politics is increasingly seen as a well-paid career and ministers are very well paid, the ability to control the careers of TDs occurs through the selection and de-selection of ministers and the allocation of portfolios. This is especially pertinent in Ireland where the ‘supine’ parliament offers few opportunities for career advancement independent of government (see Gallagher 2005). Retired ministers and Taoisigh interviewed agree that the cabinet is the centre of power in Irish politics (O’Malley interviews). While Jones (1985: 208–10) argues that, to some extent, UK cabinets pick themselves, this overstates the case for Ireland. The importance of one’s position in the party is crucial. Leadership and control of the party, perhaps through electoral success, gives Taoisigh some discretion about who enters cabinet and where they go, and are able to leave out some ministers who would naturally expect to be in government (and there are many Irish examples of this). One former Taoiseach, Jack Lynch, noted that taking over as a result of an internal party election deprived him of ‘one major strength every Prime Minister enjoys: the power of appointment . . . I had merely to endorse the changes [his predecessor, Lemass] made . . . Of course I knew I had that power formally but in effect most ministers knew they owed their position to Seán Lemass, not to me’ (Lynch 1979: 43): excluding the ‘Big Beasts’ of the political jungle – leaders or prominent members of the party – is more difficult than getting rid of more junior people. A faction that is not represented in government and unconstrained by the strictures of collective responsibility would be able to organize opposition to the prime minister within the party. But on at least two occasions incoming Taoisigh used ministerial appointments to reward supporters who assisted their seizing control of the party. When Charles Haughey became Taoiseach in 1979 much of the impetus for his successful takeover came from impatience among backbenchers who felt the ministerial elite was aloof and unsympathetic to their views. It is said that ‘the five new ministers nominated in the formation of this government and the two subsequently appointed were all identified as loyal supporters in Haughey’s leadership bid’ (Farrell 1987: 147). When Albert Reynolds succeeded Haughey in 1992 following an internal party ‘heave’ brought about in part because there was a sense that Haughey was not promoting younger talent, he initiated sweeping changes to his immediate predecessor’s government. One analyst remarked that ‘he was ruthless in rewarding his supporters and getting rid of Haughey’s’ (Chubb 1992: 173). Not only will the party leaders in a coalition and senior party figures in the Taoiseach’s own party expect to be included in cabinet, they may also expect to

184 The selection of ministers in Europe be consulted. Indeed when one minister was dismissed in an early coalition government, it was questioned whether the Taoiseach actually dismissed him or even whether he wanted to dismiss him. In a speech to his party’s conference, the party leader, Seán MacBride, argued that the minister ‘was a member of the government as my nominee on behalf of the [party]’ (quoted in Farrell 1971: 51). MacBride wrote the letter requesting the minister’s resignation. That one’s freedom to appoint ministers emanates from and is constrained by one’s party leadership is further illustrated by this exceptional government. In the two immediately postwar coalition governments, Fine Gael’s leader, General Richard Mulcahy, an army general who had been prominent in the Civil War and its aftermath, was regarded as unacceptable for the post of Taoiseach to members of a coalition party, although he served as a minister. John A. Costello was selected as Taoiseach only after the ministerial selection process had been agreed between the parties (McCullagh 1998: 32). One can clearly see that the demands of party supersede the constitutional prerogatives given to the Taoiseach. As well as allocating posts to people within their own parties, there is evidence that party leaders may be able to veto appointments from other parties. Following a mini-reshuffle caused by the dismissal of two ministers in Haughey’s last government in 1991, it was revealed that the nominee for Minister for Defence, Jim McDaid, had been photographed, smiling, with a member of the Provisional IRA, who was also a constituent, outside a court following a successful appeal against extradition. Des O’Malley, the leader of the Progressive Democrats, the minor party in government, effectively vetoed McDaid’s appointment to that post (Collins 2001: 224–5). We also know that, within parties, one Taoiseach has granted a senior rival who competed for the leadership a veto over certain sensitive portfolios (O’Malley 2006: 327). The importance of ministerial selection and portfolio allocation to satisfying the demands of coalition government can be seen in that one coalition (FF–PD 1989) nearly did not take place because the Taoiseach was reluctant to cede to the PD demand for two cabinet places, disproportionate to the party’s size (Collins 2001: ch. 17). Apart from one exception, smaller parties are over-represented in terms of ministerial posts.5 In fact between 1973 and 2007 the ratio of cabinet posts to legislative seats for junior coalition parties is 1.5, so a party with 10 per cent of the government’s seats in the Dáil can expect to receive 15 per cent of cabinet posts (author’s own calculations – the start date of 1973 is relevant because it is only since then that Irish cabinets have reached their legal maximum size and coalition has become the norm). The price of coalition in ministerial posts is sometimes seen as high by backbench TDs in the larger party, which can cause tensions. Taoisigh use the posts of minister of state and a number of other posts to relieve these tensions (see Mershon 2002 on this topic). The number of junior ministers – 20 in 2007 – has increased steadily, from up to seven originally, to ten in 1977, 15 in 1980 and 17 in 1995. These increases represent the need to satisfy the demands of ambitious backbench TDs. In 1977 the large overall majority of Fianna Fáil had meant there was also a larger number of backbench TDs. The incoming Taoiseach in 1979 rapidly increased this in 1980 ostensibly as a means to reward the TDs who had supported his challenge to

Constructing Irish governments 185 the party’s leadership. In 1995 the number was increased to accommodate the construction of a new cabinet with three parties (see Garry 1995). Again in 2007 it was further increased to apparently to ‘quell unrest’ on the backbenches of the main government party (Irish Independent, 18 June 2007). Some backbenchers were unhappy at what they saw as their loyalty and patience being left unrewarded by ministerial appointment (Irish Times, 21 June 2007). Because of this need to satisfy the members of the parties maintaining the government, we rarely see cases where independent members of the Oireachtas are appointed as ministers.6 Other posts in the Taoiseach’s gift are the chairmanships of the thirteen parliamentary committees. Although these are not government positions, and have limited power (see Gallagher 2005) they are paid posts and can reward disappointed backbenchers. They can also provide a means for a backbench TD to develop some policy expertise and a forum to be noticed by party leaders. Characteristics of ministers After the demands of party and coalition are satisfied, Taoisigh still have some discretion. The small pool from which ministers are effectively chosen means ministers can only be generalist in nature. While there are examples of teachers being appointed to Education, medical doctors to Health and lawyers to Justice, it is not necessary for ministers to have any specialization in the area, though one often sees that ministers who are appointed have acted as opposition frontbench spokespersons in the area. A number of possible criteria might be used by Taoisigh when selecting ministers: competence, experience, electoral performance, position in the party and geography. Though the need to maintain a durable government is central and informs the choices Taoisigh make when constructing cabinets, when asked most claim that when choosing a cabinet they did so on the basis of competence or to choose ‘the best people’. The small pool makes this difficult. Though Bertie Ahern publicly argued that he could have easily filled his cabinet twice over (RTÉ Radio One News at One, 15 June 2007) most ex-Taoisigh interviewed mentioned the small pool of candidates as a problem. One, for instance, ‘envied the capacity the French and American president had to pick his cabinet, to get the best brains in the country’ (Farrell interview). Another wanted the number of TDs to be doubled in order to give Taoisigh a greater real choice. He complained that there were only a handful of ministers of real quality (O’Malley interview). Verifying whether competence is used as a criterion is impossible, but there are cases where the choices were obvious attempts to control a ministry by putting a malleable figure in it – the appointment by Charles Haughey in 1980 of Gene Fitzgerald to Finance is one. Choosing incompetent loyalists, though, might create more difficulties that they solve. Table 10.2 gives some information on the characteristics of backbench TDs, junior ministers and cabinet ministers. We can see that cabinet ministers selected between 1973 and 2002 are more likely to have a third-level qualification than either junior ministers or backbench TDs – though this, it must be said, is a very crude measure of competence. In the Ahern governments one can see that those

186 The selection of ministers in Europe Table 10.2 Characteristics of TDs in all Dála, 1973–2002

Age (average) Years since first elected Third level educated (%) Personal vote as proportion of electoral quota N

Backbench TDs

Junior ministers

Cabinet ministers

46.80 7.90 51.90 0.82 684.00

46.90 10.90 64.10 0.90 198.00

48.70* 13.50*** 80.50*** 1.06*** 195.00

The unit of analysis is TDs in each of the newly elected Dála. The stars indicate the significance of difference of the descriptive statistics. * indicates significance at the 0.05 level; **, at the 0.01 level and ***, at the 0.001 level. Source: O’Malley 2006.

who make ministerial rank were those who had previously been successfully used by the party for media performances during difficult events. In Irish politics, where the party in the country is not independent of the party in parliament, would-be ministers may find it difficult to shine outside the Dáil. Liam Cosgrave spoke of spotting talent in the Dáil chamber. Dáil experience seems to be very important. Only six people (including two senators) have been made minister without any Dáil experience and of these five were very well known to the Taoiseach who appointed them. A sixth was deemed to have been a mistake by the party leader who appointed him (McBride 2005; McCullagh 1998).7 Table 10.2 shows that cabinet ministers have on average thirteen and a half years’ Dáil experience, and are slightly older that backbench TDs. Garret FitzGerald, one of those who appointed a Dáil novice to cabinet, suggested that individuals need at least eight years’ experience in the Dáil before they are ‘truly and instinctively professional in their reaction to events’ (O’Malley interview). Liam Cosgrave gave as an excuse for not appointing a senior party figure to cabinet the fact that he was out of the Dáil for a number of years. There also seems to be a move to a position where one must serve a political apprenticeship, on a committee in one’s first term, then, if one has performed well there, as a junior minister, before finally being elevated to cabinet. There is an interesting difference between the parties on this factor. Fianna Fáil’s political apprenticeship is longer than that of other parties. In Fianna Fáil one cannot easily skip these stages, whereas in Fine Gael having served as a junior minister is no necessary prerequisite. Another important factor that Taoisigh may consider is electoral performance. It has often been stated that electoral considerations are important, and one former minister has argued that for Dáil seats, ‘who delivered them and who is expected to’ are central to many people’s promotion to cabinet (Geoghegan-Quinn 2000). This contention is easier to confirm. Table 10.2 shows that cabinet ministers’ personal vote is higher than the electoral quota for their constituency, compared to .82 for backbenchers. However between the parties there are differences. Fine Gael ministers are much more likely to have higher personal votes than their backbenchers than in Fianna Fáil. Fianna Fáil ministers on the other hand have a much higher party vote in their constituency compared to other TDs – Fine Gael ministers will not garner a large level of support for their party (O’Malley 2006).

Constructing Irish governments 187 This may reflect the importance that is attached to maximizing the number of seats senior politicians get for their party in a constituency. It was suggested that a minister who was not reappointed after the 2007 election had his ‘hopes of staying at cabinet . . . weakened considerably when [his party] Fianna Fáil failed to take two of the three seats in his home Kerry South constituency’ (Irish Times, 15 June 2007). The problem with this interpretation is that same TD was appointed to cabinet twice before having failed to take two seats and many other examples could be cited of TDs being appointed to cabinet having achieved less-thanadequate electoral success for their party. A criterion frequently mentioned in the media in determining ministerial appointments is geography. This may seem odd in a small homogeneous country, though less so where the political system promotes localism and gives the impression that ministers are there to bring back ‘booty’ to their constituencies. In a recent case of two ministers not being reappointed it was said that ‘there was a lack of a geographic imperative to retain them as ministers’ (Irish Independent, 15 June 2007). This means that ministers should be spread around the country and a minister in a part of the country with other ministers may be at risk. Linked to this ‘geographic imperative’ is that Cork, the second city and largest county in the Republic, is represented in cabinet and that ministers do not come from the same constituency. However, these established political ‘truths’ that certain areas in Ireland must have a minister are often wrong. When asked about the importance of geography in constructing cabinets, former Taoisigh all agreed the need to have a Cork TD in the cabinet. However, a number of modern Irish cabinets (Haughey 1987, 1989 and Bruton from 1995) had no cabinet minister from Cork. It is equally unlikely that the ministers from Cork in the other cabinets are there for geographic reasons, given that they have held posts such as Finance, Foreign Affairs and Agriculture. Equally false is that one does not appoint ministers from the same constituency: in 1987 three constituencies had two cabinet ministers. An analysis of the distribution of cabinet places by province does not point to obvious trends. The two main parties have images of representing different types of constituencies. Fianna Fail has a tradition of spreading its ministerial posts around the country so as to cover the country electorally, to favour certain local factions as against others and to ensure that the party retains its status as a general ‘country’ party . . . by contrast Fine Gael remained very much a ‘centre’ party at elite level, as did Labour. (Garvin 1981: 185) Indeed Fine Gael ministers are more likely to come from Dublin and the eastern counties of Ireland. Jack Lynch’s later cabinets had more TDs from Munster – his own area – than other cabinets. Haughey’s cabinets had a large number of ministers from Connaught and Ulster. This might be because he depended on TDs from these areas to take the leadership and rewarded them appropriately. Bertie Ahern has tended to make fewer appointments among Fianna Fáil TDs from Dublin – his own

188 The selection of ministers in Europe city – although in 2007 seven of the cabinet including himself represented Dublin constituencies whereas just two represented rural west of Ireland constituencies. In general Dublin is over-represented, which may be because smaller coalition parties are strongest there and because Dublin constituencies might be considered less demanding which may enable the TD to spend more time on national issues and therefore become prominent. The number of junior ministers means that it is likely that they can be spread around the country, and indeed they are. The geographic explanation, however, does not hold up to scrutiny: in 1987 the southeast was unrepresented and in 2007 the south-west is unrepresented. In interviews no Taoiseach could point to an appointment made on a geographic basis. Since 1977 it has become the norm to appoint at least one woman as a minister. While then it was as a junior minister, in 1979 a woman was appointed to cabinet for the first time since the Civil War. There has been an increasing presence of women in the Dáil, and increasingly these are not women who took their husband’s or father’s seat. It is now normal that at least two ministers will be women. In 2007 three women were appointed to full cabinet rank The ability to speak Irish was also seen as important for the posts of Education and is still important for the Minister for the Gaeltacht (Irish-speaking area). In terms of portfolio the most important position after the Taoiseach is Finance. This is confirmed both by expert surveys (Druckman and Warwick 2005; Laver and Hunt 1992) and by interviews with ministerial elites. You then tend to see that senior and experienced politicians are appointed to Finance (see above for one exception).

Cabinet reshuffles: dismissing or resigning? Huber and Martínez-Gallardo (2004: 39) note that the ratio of Irish ministers’ ‘political experience’ to their ‘portfolio experience’ is much greater than in other countries they study. This suggests that there is a high portfolio turnover in Ireland, that is, ministers change jobs in Ireland more than elsewhere. This is surprising as few commentators on Ireland would regard reshuffles as common. However reshuffles in Ireland are not as unusual as is commonly thought. If we define a reshuffle as the coterminous movement of two or more ministers within the lifetime of a government then we can identify relatively large number of reshuffles in Ireland. Of the governments that lasted more than a year all but one postwar government had a reshuffle of some type. But general mid-term reshuffles were rarely used to dismiss Irish ministers. To give an indication of the rarity of dismissals and the stability of ministers in Irish governments, a writer on ministerial turnover treated cases of ministers not being reappointed by one long-serving Taoiseach following an election as dismissals (Farrell 1971). Taoisigh are not ‘good butchers’. In the village atmosphere of Irish politics one tries to avoid making enemies unnecessarily. This may have induced the culture of non-resignation, where Taoisigh will tend to protect rather than dismiss a minister under pressure for policy or administrative errors. Taoisigh are also restricted in that they have few inducements to offer dismissed ministers. Where in the UK the honours list or the House of Lords soften the blow for many, no such facility is available in Ireland. However the mid-term reshuffle has become more

Constructing Irish governments 189 common and is used to ease older cabinet ministers from their posts. These are flagged so far ahead that the minister is let down very gently. There are extra difficulties, in that the small pool from which one can choose a replacement means that, unlike elsewhere, in Ireland one cannot easily find a competent replacement with similar policy views. Nor does dismissing ministers reflect well on prime ministers. It exposes either poor judgement for making the earlier appointment, or division within the cabinet so deep that it can be solved only by resignation or dismissal. One also has to consider the reaction of the dismissed minister. Many Irish governments are minority governments, and with two exceptions (1977 and 1992) majorities, if they exist, tend to be small. A dismissed minister could seriously damage the government by voting against it or resigning his or her seat. Ray Burke, who was forced to resign in 1997, also resigned his seat, causing further problems for that minority administration. In a recent reshuffle, one minister allegedly threatened to resign his Dáil seat if he was moved out of the cabinet, and was not dismissed (The Irish Times, 5 October 2004). A description by Garret FitzGerald of a reshuffle may also indicate why reshuffles are avoided. FitzGerald outlines how his great plans unravelled as one minister threatened to resign from cabinet rather than be moved. This would have caused a crisis within the coalition, so the Taoiseach relented (see FitzGerald’s autobiography 1991: 621–5). Using general mid-term reshuffles to dismiss ministers is fraught with difficulties, which could explain why Taoisigh have to be so conservative in their selections in the first place. Reasons for resignations or dismissals There have been 39 ministerial resignations (a minister or ministers leaving government within the lifetime of that government, for whatever reason except death) in the postwar period. We might code the different resignations into different categories – some have resigned to take up new, often prestigious posts. Some have resigned over scandals, often financial. Other reasons might be policy disagreements and a fourth category is party disagreements or coalition breakdown. Of these 39 cases, in ten the minister was moving or being moved to another position, such as being made EU Commissioner. Eight could be said to have resigned over a ‘scandal’, usually financial – never sexual.8 A Fine Gael Minister for Defence in the 1970s, Paddy Donegan, was reported to have called the President a ‘thundering disgrace’ (he actually used more colourful language) for having referred a Bill to the Supreme Court. Donegan offered his resignation but it was not accepted, though he was later moved from Defence. This caused the President to resign, he claimed, in order to protect the integrity of the office. In Fianna Fáil a sense of loyalty or perhaps a lack of willingness to admit error means that Fianna Fáil Taoisigh are less likely to demand resignations. It is significant that one highprofile resignation in 1990 of a Fianna Fáil minister, Brian Lenihan, following a minor scandal was forced upon Fianna Fáil by coalition partners.9 As the Lenihan case illustrates, if political pressure does force a resignation, these are usually reluctantly acceded to. The case of Ray Burke, a Minister for Foreign Affairs,

190 The selection of ministers in Europe about whom many allegations were made of accepting large amounts of money for political favours, is also revealing. Even when he was dismissed, months after the allegations surfaced in public, the Taoiseach Bertie Ahern claimed that the resignation was as a result of a ‘sustained campaign of incremental intensity’ against Burke rather than for what Burke might have done (Dáil Debates Vol. 481 cols 40–1, 7 October 1997). Ahern later acknowledged that he should not have appointed Burke (Dáil Debates Vol. 481 cols 994–5, 12 May 1998). Five ministerial resignations can be coded as policy-related. The Arms Crisis in 197010 was a scandal which caused four ministers to leave the cabinet, but perhaps two of these could be coded as policy disagreement as both left the party – Neil Blaney and Kevin Boland. The dismissal of Noël Browne (see note 7) is an example of a policy disagreement. Sixteen dismissals can be defined as partyrelated, six where ministers within the party seen as disloyal resigned or were removed and ten where a coalition partner pulled out of government In the early 1980s two ministers resigned over the issue of loyalty to the party leader. The dismissal of two senior ministers who challenged party leadership in 1991 also caused a reshuffle. In all of these cases the Taoiseach may have been seen to have had little option but to dismiss the ministers in the face of open dissent of him or of his government. If we look at the other reasons for the dismissal of ministers, few mid-term dismissals can be put down to performance. Personal loyalty and the risks associated with dismissal mean that Taoisigh often wait for a new government to be set up if they wish to reorganize their government. Taoisigh will prefer to keep a close watch on the relevant departments. If anything demotion is more common,

10

Number of resignations

8

6

4

2

0 Policy dispute

Other job

Scandal

Party division

Coalition break-up

Figure 10.1 Ministerial resignations in Ireland, 1945–2007

Constructing Irish governments 191 although the link between performance and promotion or demotion is sometimes difficult to discern. Whereas in the 1960s under Lemass there was a clear drive to promote young, energetic ministers, and a tolerance of some mistakes in the interests of action, an analysis of the changes in the last two governments might reveal that some ministers are rewarded for avoiding action.

Conclusion The small numbers in the Irish political class and comparatively low turnover (Matland and Studlar 2004) in parliament mean that some government TDs may represent a strong threat for their party leaders. Taoisigh are reluctant to dismiss a minister when that dismissal could upset the delicate parliamentary balance. Personal loyalties to colleagues with whom one has worked for many years also make wielding the butcher’s knife an unpleasant task. Resignations tend to be forced upon Taoisigh by events rather than as a result of planned management of the government. The increasing number of posts in international organizations such as the EU and national agencies whose appointment is at the government’s discretion means that it has become easier to remove some ministers. However, reshuffles must still be handled delicately. The former Taoiseach, Bertie Ahern, was more cautious than his predecessors. He perhaps was acutely aware of the need to develop long-term relationships in which TDs have a clear, if comparatively slow, career progression. This may have satisfied most senior party members and kept the ambitious junior ones loyal in anticipation. That his first two governments have lasted their full terms may lead him to feel his caution was worthwhile.

Notes 1 I will use the Irish terms when these are the terms commonly used in Ireland. 2 This happened a number of times in the state’s history: Frank Fahy in 1948; Patrick Hogan in 1951, 1961 and 1965; John O’Connell in 1981; Seán Treacy in 1987 and 1989; and Séamus Pattison in 1997. Fahy was a Fianna Fáil TD retained by the interparty government. Hogan was a Labour party TD. Both O’Connell and Treacy were independents who had previously been Labour Party TDs. The post is very attractive not just because the Ceann Comhairle is paid the equivalent of a cabinet minister, over €200,000 but also because she is automatically returned at the following election. 3 Three senators have been appointed to cabinet: Joseph Connolly, twice, in 1932 and 1933; Seán Moylan in 1957 and James Dooge in 1981. Moylan and Dooge were Taoiseach’s nominees, a facility which did not exist when Connolly was elected to the Seanad. 4 The author interviewed all but one living ex-Taoiseach on cabinet construction in 2004 and 2005. References to these interviews are given as O’Malley interview. The author was also given access to the recordings of interviews by Brian Farrell with ministers. These interviews are referred to as Farrell interview. 5 This case is in the 1994 ‘Rainbow’ coalition where Democratic Left received one cabinet minister, which was closely proportional to the size of their parliamentary group in the Dáil. Even still they also had a ‘super-junior’, a junior minister who was entitled to sit at cabinet. 6 There was one case in 1948 where James Dillon, who had resigned from Fine Gael in opposition to its position of neutrality during the Second World War, was made a minister as an independent and represented the independents that supported that

192 The selection of ministers in Europe

7

8

9

10

government. As he was later to become a leader of Fine Gael he could not be considered a typical independent. The former Taoiseach Garret FitzGerald claimed that in 1985 he offered an independent TD who had been expelled from his own party a cabinet post if he would join Fine Gael (O’Malley interview). Noël Browne was appointed Minister for Health in 1948 for the new Clann na Poblachta (republican) party aged just 32. An energetic and reforming minister, his resignation came when the cabinet (including his party leader) refused to back his Mother and Child Scheme in the face of opposition from the Catholic Church. This controversy is one of the most celebrated in Irish political history. In the one revealed sexual scandal in Ireland a junior minister who was alleged to have propositioned a rent-boy was not asked to resign. In fact all parties united to support him. This was despite his coming from a different faction in his party to its leader and having been seen as ‘difficult’ by the party leadership. Brian Lenihan, Tánaiste and candidate for the Presidency, claimed that he had never approached a previous President in an attempt to persuade him not to allow the dissolution of the Dáil. His categorical denials were refuted when a Ph.D. student produced a tape in which Lenihan admitted to having done so. His attempts to marry the two versions and his ultimate excuse that he was suffering from the side effects of drugs he was taking to help his recent liver transplant made his position as a government ministers difficult to defend, and Haughey, under pressure from the PDs, asked for his resignation, which Lenihan refused, only to be sacked. Lenihan failed to get elected to the presidency. The basic facts of this scandal are still disputed, but what is established is that some government ministers were involved in importing arms for use in Northern Ireland, then at the outbreak of the conflict there. Some of the ministers claimed they were acting on foot of a government (sub-committee) decision, but the Taoiseach and other ministers claimed this was a lie. Another minister, Haughey, claimed no knowledge whatsoever of the events. The ministers were brought to trial but acquitted.

References Chubb, B. (1992) The Government and Politics of Ireland, 3rd. edition. Harlow: Longman. Collins, S. (2001) The Power Game: Ireland under Fianna Fáil. Dublin: O’Brien Press. Druckman, J. N., and Warwick, P. V. (2005) ‘The Missing Piece: Measuring Portfolio Salience in Western European Parliamentary Democracies’. European Journal of Political Research 44, 1: 17–42. Farrell, B. (1971) Chairman or Chief? The Role of Taoiseach in Irish Government. Dublin: Gill and Macmillan. Farrell, B. (1987) ‘Government Formation and Ministerial Selection’. In Penniman, H. R. and Farrell, B. (eds) Ireland at the Polls 1981, 1982 and 1987: A Study of Four General Elections. Washington DC: American Enterprise Institute and Duke University Press, 131–155. FitzGerald, G. (1991) All in a Life: An Autobiography. Dublin: Gill and Macmillan. Gallagher, M. (2005) ‘Dáil and Seanad’. In Coakley, J. and Gallagher, M. Politics in the Republic of Ireland, London: Routledge, 211–241. Garry, J. (1995) ‘The Demise of the Fianna Fáil–Labour “Partnership” Government and the Rise of the “Rainbow” Coalition’. Irish Political Studies 10: 192–199. Garvin, T. (1981) The Evolution of Irish Nationalist Politics. Dublin: Gill and Macmillan. Geoghegan-Quinn, M. (2000) ‘Cabinet Reshuffle Made with an Eye to the Next Election’. The Irish Times, 29 January. Huber, J. D. and Martinez-Gallardo, C. (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.

Constructing Irish governments 193 Jones, G. W. (1985) ‘The Prime Minister’s Power’. In King, A. (ed.) The British Prime Minister. Durham, NC: Duke University Press, 195–220. Laver, M. and Hunt, W. B. (1992) Policy and Party Competition. New York: Routledge. Lynch, J. (1979) ‘My Life and Times’. Magill, November: 33–51. McBride, S. (2005) That Day’s Struggle: A Memoir 1904–1951. Edited by C. Lawlor. Blackrock, Co. Dublin: Currach Press. McCullagh, D. (1998) A Makeshift Majority: The First Inter-party Government, 1948–51. Dublin: IPA. Matland, R. E. and Studlar, D. T. (2004) ‘Determinants of Legislative Turnover: A Crossnational Analysis’. British Journal Of Political Science 34: 87–108. Mershon, C. (2002) The Costs of Coalition. Stanford: Stanford University Press. Mitchell, P. (2003) ‘Ireland: “O What a Tangled Web.” – Delegation, Accountability and Executive Power. In Strøm, K., Müller, W. C. and Bergman, T. (eds) Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press, 418–444. O’Malley, E. (2006) ‘Ministerial Selection in Ireland: Limited Choice in a Political Village’. Irish Political Studies 21, 3: 319–336. Weeks, L. (2004) ‘Independents in Government: A Case-study of Ireland’. Paper read at ECPR Joint Sessions of Workshops – New Parties in Government, at Uppsala Universitet.

11 More safe than sound? Cabinet ministers in Iceland Gunnar Helgi Kristinsson

Constitutional situation Parliamentary government was established in Iceland in 1904 but remains a constitutional convention which is nowhere explicitly mentioned in the constitution. The present constitution, which was adopted at the time of the dissolution of the union with Denmark in 1944, has some semi-presidential features. Thus, responsibility for government formation according to the constitution lies with the head of state (since 1944 the President). Although in fact largely a figurehead, the President of Iceland is popularly elected and has some potential power, as he showed in 2004 when he refused to countersign a legislative act which had been passed by parliament (with the effect that the act was withdrawn to avoid the constitutionally prescribed referendum).1 The president is responsible for designating the government formateur during talks on government formation and the president’s judgement could under conditions of uncertainty affect the outcome. Parliament itself has no say in selecting the formateur, nor does a new government have to go through a vote of investiture when taking over. Of course, the president is expected, in accordance with parliamentary practice, to choose the formateur most likely to succeed in forming a majority coalition, but in cases where such a government is not forthcoming the decisions of the president could be decisive. No single-party majorities have emerged since the interwar period. The threat of presidential activism in government formation (e.g. through the formation of an extra-parliamentary cabinet composed of government officials) has been present in Iceland since 1942–44 when the Governor of Iceland formed such a government. The threat of presidential activism (the Governor became President in 1944) has been among the factors contributing to the successful formation of majority governments in Icelandic politics (Jóhannesson 2006). Unlike in Scandinavia, minority governments are in practice not tolerated, except in a provisional capacity. Government coalitions usually contain two or three parties, but two-way coalitions are more stable and tend to be preferred. The president starts government talks by meeting with the political leaders individually to sound out their preferences. Subsequently, the role of formateur is given to the party leader considered most likely to form a majority government. This is often the leader of the largest party, but it need not be so (as indeed it was not after the elections in 1971, 1978, 1979 and 1987). If the first round of

More safe than sound? Iceland 195 negotiations fails to produce a new government the president decides on additional rounds, while in some cases also exploring or even threatening the possibility of an extra-parliamentary cabinet. Coalition formation in the period since 1944 has been known to take up to seven rounds of negotiations and as much as 117 days (in 1947), but on average the formation of a new coalition took between three and four rounds and 40 days (Indridason 2005). The number of cabinet posts is not fixed and can vary from one coalition to another – it is among the things decided in the negotiations between the parties. Cabinet posts are in great demand among politicians and there is a tendency over time for their number to increase. Since the end of the Second World War their number has gradually increased from six to 12. Apart from the special status of the Prime Minister there is no formal grading of ministerial posts (there are no vicePMs, junior ministers or ministers without portfolio), and obtaining appointment as minister is generally considered the crowning achievement of a political career. According to the letter of the constitution the PM (on behalf of the president) selects ministers and discharges them. In reality, however, the division of portfolios is decided by the parties involved in the coalition negotiation process. The ministries which each party gets are a very significant part of its pay-offs. The division of ministries between the parties is all the more significant since ministers in Iceland enjoy a large degree of independence from cabinet control (Laver and Hunt 1992). Although they could, in principle, be bound by formal coalition agreements, joint decisions of the cabinet or cabinet committees, they are in practice relatively free of such interference. Individual ministers are the effective agenda setters within their departments. Coalition agreements tend to be short and restricted to generalities on most issues while joint cabinet decisions or the use of cabinet committees are very rare. All government bills, nonetheless, must be accepted by the cabinet and the parliamentary groups of the coalition parties before formally being presented in parliament (Kristinsson 2005). The cabinet can often be the decisive instance in the preparation of a government bill and the parliamentary groups can create problems for the ministers, even if they are not routinely the most important instance in the process. In administrative affairs individual ministers enjoy a large degree of autonomy in the running of their departments and appear – moreover – to be largely free to decide if administrative matters are to appear before the cabinet or not (Helgason 2004, 70). The division of ministries between the parties has an element of horse trading. Each party gets primary control of the issue areas under its ministers but relinquishes others, except with regard to government bills (where it has effective veto power). Horse trading can reduce tensions within the cabinet, since the parties are only to a limited extent expected to be responsible for each other’s policies. But it can also, of course, lead to paradoxical results, e.g. when minor parties hold important ministerial positions under the protection of larger parties which may, however, feel inclined to wash their hands of them before the voters (Kristinsson 2006). The distribution of portfolios in government coalitions seems to conform to Gamson’s law, which states that each party in a coalition will get a share of portfolios proportional to the seats it contributes to the coalition (Gamson 1961).

196 The selection of ministers in Europe Thus, there is a strong correlation between the contribution of the largest partner within a coalition to its joint parliamentary backing and its (unweighted) share of portfolios (r = 0.91) in the period 1959–2007. In accordance with a well established pattern, however (e.g. Browne and Franklin 1973), smaller parties in Iceland tend to be generously overpaid. In 1959–2007 the largest party contributed on average 59 per cent of coalition seats while receiving only 47 per cent of the portfolios. Carroll and Cox (2007) point out that in the absence of pre-election pacts between parties the bargaining position of smaller parties may be somewhat stronger than suggested by Gamson, which might explain their disproportionate share of government portfolios. This may account for the Icelandic case, where the coalition game, as a rule, is strictly a post-election affair, played without any prior commitments. Coalition partners usually divide the three most important posts (PM, Finance and Foreign Affairs) between them so that each major partner gets at least one of them. Given its agrarian roots, the Progressive Party often gets Agriculture; the Social Democrats often get Social Affairs and so on. The party furthest to the left, however, never gets the PM, Foreign Affairs or the Ministry of Justice, and the largely urban-based Social Democrats have never been allocated the Ministry of Agriculture in a majority government. Once the parties have divided ministerial posts between them it is entirely up to them to choose their own ministers. The PM takes no part in selecting ministers not in his party. While the constitution does not restrict the choice of ministers the competition is effectively confined to serving parliamentarians (see below). Ministers also remain MPs during their ministerial mandate.

The selection of ministers Given the importance attached to cabinet posts in Icelandic politics, the power of selection is by no means an insignificant political resource. As serving MPs make up the reservoir of ‘ministrables’ – non-parliamentarians enter the race only under exceptional circumstances such as when party leaders are without a seat in parliament (this covers three out of the six instances in the postwar era when ministers have been appointed without a parliamentary seat) – the parliamentary groups have a key influence in the selection of ministers in all of the parties. The established parties in the centre right, the Independence Party and the Progressive Party, have always given their parliamentary group the final say in the selection of ministers. Here the ministers are selected in a secret ballot. Prior to the 1980s the vote sometimes even took place without a formal proposal from the party leader, but this has changed (see below). Thus, while British ministerial candidates are said to wait eagerly for the PM to phone during the selection process, their Icelandic counterparts have been known to pick up the phone themselves and start ringing round in the parliamentary group in search of support. In the parties of the centre left, the parliamentary group has probably been as decisive in selecting the ministers as in the other parties but a larger body (national executive committee) has formally been part of the selection process in some of them. Once the ministers have been selected, the party leader normally decides which

More safe than sound? Iceland 197 positions they will fill. In the early postwar period, when each partner in a coalition government could expect to get only two or three portfolios, a large proportion of ministerial posts were filled by leaders, deputy leaders or others in formal leadership positions in the parties. With the growing size of the cabinet, however, a greater number of selection criteria came into play. Thus, over the whole postwar period party leaders represent 24 per cent of all ministers and others in formal leadership positions in the parties another 21 per cent; a combined share of 45 per cent.2 Since the 1990s, however, people in such positions have been fewer than one-third of those taking ministerial office. On the whole, political background is far more important in the competition for ministerial position than expertise or prior experience of the issue area of the government department in question. Only a minority of those who become ministers have any experience of their issue areas in their pre-political careers, i.e. before entering parliament for the first time, while two-thirds have little or no such experience. Those who do are primarily concentrated in the ministries responsible for specific sectors of the economy. Around 83 per cent, however, have a strong party background, in the sense of having served in positions of responsibility3 for their party prior to entering parliament for the first time, and interestingly enough this proportion is considerably higher in the conservative Independence Party (96 per cent) than in the other parties. This reflects the organizational strength of the Independence Party, which has a much larger membership than the other parties and a higher membership density. On average 62 per cent of those who become ministers in any new government have been ministers before, while average ministerial duration in 1944–2008 (defined as the continuous inhabitation of a ministerial post irrespective of cabinet duration) was 4.3 years.4 A minister’s constituency can be a relevant factor in the selection process. Those leading the party lists in the Reykjavik area (where a steadily growing proportion of the electorate lives – nowadays close to two-thirds) generally make a strong case for ministership. MPs from provincial constituencies also attach great significance to ministerial positions, and claims to ministership are sometimes backed by reference to the need for provincial representation in the cabinet. In the postwar period 44 per cent of ministers have been from provincial constituencies but in the rural Progressive Party the share is close to 74 per cent. The provinces have always been strongly over-represented in parliament. Socially the cabinet ministers are an elite group, where various categories of employers, executive officers and professionals dominate (75 per cent were in such occupations prior to entering parliament for the first time) and two-thirds have a university degree. The average age of ministers is 50 years when coming into office and most of them are males (90 per cent in the postwar period). Gender, however, is an increasingly significant factor in the selection of ministers, and the number of females has gradually grown since 1970 (when the first female minister came into office). Thus 27 per cent of the ministers selected since the turn of the century are females. Another significant trend is towards a clearer distinction between a national and a local political career. Whereas overall close to half of those who become ministers have sat on local councils at some point (48 per cent), they were 55 per cent before 1980 down to 38 per cent since the 1990s.

198 The selection of ministers in Europe As the number of ministers has grown, the selection of ministers and other positions of influence in the parliamentary group has become an increasingly important tool for maintaining party cohesion. To a certain extent it has compensated for a tendency towards weaker party discipline which followed the opening up of the nomination process in the parties during the 1970s when more or less open primaries for parliamentary elections became the norm. A large proportion of the electorate votes in the party primaries; election research indicates that the proportion is 15–30 per cent in the elections 1983–2007. Candidates selected through primaries are not as dependent on the party leadership or the core party organizations as they used to be so long as they maintain a loyal personal following in their constituencies. An extreme example was the contest in the Independence Party between Gudmundsson (an independent-minded patronage politician within the party) and party leader Hallgrimsson for the first place on the list in Reykjavik in 1983 which resulted in Hallgrimsson falling to seventh place on the party list and losing his seat in the Althingi, the Icelandic parliament. Party leaders have a smaller chance of eliminating difficult parliamentarians through the nomination process than they had when it was confined to the party organization. Hence, the most important instrument of the party leadership for keeping the parliamentarians in line has become the hope of a ministerial position or other positions of power and prestige The role of the party leader in choosing ministers therefore varies according to his strength in the party. When there were only two or three positions to be filled, the party leaders’ scope for using the hope of a ministership as an incentive for compliance was relatively small since it affected a very small part of the parliamentary group. In recent years, it seems, with an increasing number of ministerial posts the party leaders have learned to use the positions at their disposal as an instrument of party cohesion. Given that 32 seats in the unicameral Althingi are the minimum for a majority and the present number of ministers is 12 the task of the party leaders is to find means of securing the loyalties of the remaining majority MPs, and to avoid having to deal with a large number of loose guns in the parliamentary groups. One way to tackle this problem is to hand out other positions of responsibility so as to appease the ambitions of non-ministerial parliamentarians. These include the position of the speaker, chairmanship of the parliamentary group, chairmanship of the 12 standing committees and a few committee posts in the public sector. After these responsibilities have been handed out there usually remain only a few junior members of the parliamentary parties without positions of influence. These have an incentive to behave well, if they wish to secure their future position within the party (e.g. a future ministership). In a few cases, however, more senior members of the parliamentary parties feel they have been passed over. These can prove difficult to control but sometimes the issue is solved by finding positions for them in the civil service so that they can leave parliament and embark on a second career. Diplomatic posts abroad are especially popular for this purpose. Since the 1980s or 1990s the party leaders always present a definitive proposal with a list of ministers to the parliamentary groups. This is done after an agreement has been reached on a new coalition and the number of ministers which the party

More safe than sound? Iceland 199 is going to receive is clear. The party leaders prepare this list by meeting with each member of the parliamentary group individually and in some cases they have to suggest ministers who may not be to their liking. In most cases, nonetheless, a moderately strong party leader is likely to be the decisive influence on how positions of power and prestige are distributed among the parliamentary group.

The risks of a minister Ministers are selected solely by their respective parties and de facto they represent their parties in the cabinet. The risks they face of losing their cabinet post are the risk on the one hand that their party has to leave the government (collective risk) and on the other that their party loses confidence in them (individual risk). The theoretical possibility that the PM might discharge a minister from a coalition partner is never realized. An attempt to do so would be considered interference in the affairs of a coalition partner and would in all probability lead to the breakdown of the coalition. For similar reasons, individual votes of no confidence never succeed in parliament and in fact no such vote has taken place since the early 1950s. Even within his own party a party leader would need the approval of the appointing body (parliamentary group or central committee) if he wanted to attempt a reshuffle of the cabinet posts held by his own party. Ministerial risk may be regarded as the likelihood for every year started that a minister will lose her cabinet position. A total of 515 ministerial years were entered by Icelandic ministers in the period 1944–2006, of which 113 resulted in a loss of cabinet office before the end of the year.5 Thus, ministerial risk per year was 22 per cent. In the majority of cases, 65 out of 113, the end of ministerships came about through collective resignations, as the ministers’ party left government. An interesting 9 per cent of minister years (i.e. 48 of 515), however, ended with individual resignations. The reasons for individual resignations were the following. First, there are those who lose their jobs as ministers against their will, e.g. when the party renews its group of ministers while continuing in government or in response to criticism. Second, there are those leaving politics, either at the end of their term in parliament or, more often, in mid-term to start a new career outside politics (with a likelihood of around 2 per cent per year). Finally, one minister resigned in protest, in 1994, leaving the cabinet and her own party to form a new party.6 As can be seen in Table 11.1, most of the individual resignations fall into the category of ‘lost ministerships’ and occur most often when a party which continues in government after an election decides to terminate the cabinet membership of one or more of its ministers. The risk of losing a ministership is around 6 per cent per year. Out of 30 ministers who were removed from office against their will only four were forced to leave during their mandate. On the other hand, there are only two instances of ministers resigning in response to criticism.7 The relatively weak tradition of meeting criticism with resignations in Iceland may be contrasted with political systems where such resignations are relatively common. According to Dowding and Kang (1998) there were 71 resignations in the United Kingdom in the period 1945–97, apart from changes following general elections, cabinet reshuffles and retirements. The corresponding figure for Iceland,

200 The selection of ministers in Europe Table 11.1 Ministerial risk and resignations in Iceland, 1944–2006 Minister years

Probability

Continue Collective resignations Individual resignations Removed from office Lost ministership Resign in resp. to criticism Leaving politics End of term Second (non political) career Resign in protest

402 65

0.78 0.13

30 2

0.06 0.00

3 12 1

0.01 0.02 0.00

Total

515

1.0

Note: Collective resignations includes cases where ministers from one or more party leave the cabinet while others may remain in government. Of the 30 lost ministerships four were forced to leave during their mandates while 26 were not hired again after an election even though their party remained in government.

three, is not really comparable because of a much greater number of ministers in the UK. In the British case 35 per cent of resignation issues actually led to the resignation of a minister (ibid: 419). Non-resignations are identified as instances where a resignation fails to occur even if the press, a non-political organization or MPs have suggested resignation. Identifying non-resignations in Iceland comparable to the British ones is not entirely straightforward since ministerial resignation is not among the regular tools of ministerial control in Icelandic politics and hence not as big a part of its vocabulary and rhetoric. Over a period of twenty years, in 1985–2005, the parliamentary records were searched for urgency debates on the administration of individual ministers. An issue was considered a potential resignation issue if the guilt of a minister was clearly established, if there were allegations of the suppression of information and if there were demands for a formal investigation. An issue was also considered a resignation issue if the resignation of a minister was raised in the parliamentary debate or in the media.8 The total number of resignation issues by this account is 38 – of which the issue of ministerial resignation was explicitly raised in 20 – and in addition to that, one resigned in protest. The number of resignation issues which occurred in Iceland in 1985–2005 was therefore on average 1.9 per year compared to 3.9 in the United Kingdom in 1945–97. Various factors may account for this difference, but the more obvious candidates include a greater number of ministers in the UK, the relative absence of policy disagreements as the cause of resignations in Iceland and possibly a different style of mass media. Of interest in the present context, however, is not only the occurrence of resignation issues but the ratio of actual resignations to resignation issue. With three resignations since the end of the Second World War – all of which occurred in a span of seven years (1987–1994) – the ratio which we obtain seems inevitably affected by the period which we decide to study. If we study only the period 1985–2005 the result is that 7.9 per cent of resignation issues actually led to

More safe than sound? Iceland 201 resignations, but extrapolation to longer periods (through different methods) would result in a ratio of approximately 3 per cent. Resignations are therefore much less common in Iceland than in the UK, not only in absolute terms but also in the likelihood of a resignation issue leading to actual resignation. The major difference between the two countries lies in the role of the PM or party leader in selecting and getting rid of ministers. Whereas in the UK – with single-party majorities – the PM has a clear responsibility and authority in this respect, the responsibility in Iceland is less clear, given that the ministers are in fact considered party representatives. None of the resignations in Iceland was in the PM’s party and the PM played no role in any of them. The party leaders did play a role, but they did not have the sole authority to discharge the ministers given the role of the parliamentary groups as selectors. The ministers’ own parliamentary groups are less efficient in enforcing the resignation of a minister who may be causing damage to the party than a party leader with full powers. This may be thought of as a collective action problem. While it may be in the collective interest of the party to get rid of an incompetent minister, it is only the party leader who realistically can be expected to act on the collective interest. The party leader, however, has to go through the cumbersome process of bringing the matter before the parliamentary party which gives the minister in question an opportunity to create confusion and disorder or even a split in the party (like the one in the Independence Party when Gudmundsson resigned in 1987). Suffering the incompetent minister until the end of the term may thus be a less costly strategy for the leader, even if it reflects badly on the party, than the alternative of internal factionalism. This does not mean, however, that ministers are immune from criticism. While resignations in response to criticism are not common, a minister who performs badly reduces her chances of being re-selected as minister considerably. Gallup (þjóð arpúls Gallup various issues) has gathered data on public satisfaction with the performance of individual ministers at regular intervals since 1993. On average 43 per cent of voters are satisfied with their performance. The relationship of public satisfaction with ministerial risk may be seen in Table 11.2. The table shows average satisfaction with ministers who (1) continued in politics, (2) left politics Table 11.2 Public satisfaction with ministers and their fates in Iceland, 1993–2006

Continue Leave politics Removed from office

Average share of respondents satisfied with minister (%)

N

Standard deviation

46 43 35

21 8 9

12 7 10

Note: Each minister is usually rated several times before leaving office and the table is based on averages of these. Logistic regression with satisfaction as independent variable and removal from office as dependent yields an estimated odds ratio for satisfaction of 0.902 (Sig. 0.036). Excluded from the analysis are the Social Democratic ministers who left collectively in 1995 and the minister who resigned in protest in 1994.

202 The selection of ministers in Europe (e.g. to embark on a second career) and (3) were forced to leave during the period 1993–2006. Low public satisfaction with the performance of ministers seems to increase their risk of being removed from office. Most often this occurs at the end of term in relation to the renewal of the ministerial group and revision of coalition co-operation. Thus, the ministers, who are well protected in parliament and in many respects fairly autonomous within the cabinet as well, in the end, have to answer to their own parties, which in turn are concerned with the public eye on their ministerial personnel actions.

Conclusion Icelandic cabinet ministers enjoy a relatively high degree of job security. They are never fired by the prime minister and never removed from office through a parliamentary vote of no confidence. They rarely resign in response to public criticism and individual ministers almost never resign in protest or in response to policy disagreements. As a result, Icelandic ministers tend to last longer than their colleagues in most other European countries (Herman 1975; Frognier 1991). Ministers in Iceland should be regarded primarily as the agents of their parties. They are selected by their parties and the risks they face are on the one hand that their party has to leave government and on the other that their party loses faith in them, although on the whole they enjoy considerable job security. In recent years, the selection of ministers and other important positions in the parliamentary group has become an increasingly important tool for maintaining party cohesion, which was weakened with the introduction of primaries in the 1970s, and the party leaders have assumed an increasingly assertive role in the selection process. The parliamentary groups, however, have kept the last word.

Notes 1

2 3 4

The act in question was the controversial Media Act in which Prime Minister Oddsson was suspected of instigating as a personal vendetta against a supposedly hostile group of businessmen. According to article 26 of the constitution a refusal by the president to confirm a legislative act should bring about an abrogative referendum. This is the only general provision in the constitution for holding referenda with binding effect (two other articles concern the abolition of state religion and removal from office of the president). The president never used the authority given to him via article 26 until 2004 and there have been no referenda in Iceland since the establishment of the republic in 1944. The government in 2004 avoided the referendum (which it seemed certain to lose) by having its legislative majority withdraw the act in question. Formalists insisted that the referendum should go ahead anyway but an abrogative referendum on non-existent legislation would clearly have been a farce. Based on an inventory of cabinet members since 1944 using information especially from Alþ ingismannatal (1996), government offices and the Althingi. Positions of responsibility here include chairmanships or membership of the executive of local party organizations, women’s organizations and the youth organizations or similar functions. The figure of 62 per cent includes any ministers who carry on directly from the previous cabinet and those who have been ministers in other cabinets at an earlier point in time.

More safe than sound? Iceland 203 5 6 7

8

Excluded from the data are five cases (minister years) of ministers who died in office, had to leave for health reasons or because a temporary arrangement was ended. Sigurdardottir had been engaged in a long-lasting struggle with party leader Hannibalsson and resigned from the government after the Social Democratic Party conference favoured him over her in a leadership contest. The two cases include Gudmundsson, who was forced to resign in 1987 (after an investigation uniquely initiated by parliament revealed that a business company of his had been guilty of tax fraud while he was Minister of Finance), and Stefansson, who resigned in 1994 after a prolonged period of public criticism of his conduct as minister. Based primarily on the website of the Icelandic parliament (Alþingi) (www.alþingi.is) and the website of the daily newspaper Morgunblað ið (www.mbl.is).

References Alþ ingismannatal (1996) Reykjavik: Skrifstofa Alþingis. Browne, E. and Franklin, M. (1973) ‘Aspects of Coalition Payoffs in European Parliamentary Democracies’. American Political Science Review 67, 2: 453–469. Carroll, R. and Cox, G. W. (2007) ‘The Logic of Gamson’s Law: Pre-election Coalitions and Portfolio Allocations’. American Journal of Political Science 51, 2: 300–313. Dowding, K. and Kang, W-T. (1998) ‘Ministerial Resignations 1945–97’. Public Administration Vol. 76, 411–429. Frognier A. (1991) ‘Elite Circulation in Cabinet Government’. Blondel, J. and Thiébault, J.-L. (eds) The Profession of Government Minister in Western Europe. Basingstoke: Macmillan, 119–135. Gamson, W. A. (1961) ‘A Theory of Coalition Formation’. American Sociological Review 26, 3: 373–382. Helgason, Á. (2004) ‘Stjórnskipuleg sta ð a stjórnarráð sins’. In S. Ísleifsson (ed.) Stjórnarráð ∂ Íslands 1964–2004. Reykjavik: Sögufélag. Herman V. (1975) ‘Comparative Perspectives on Ministerial Stability in Britain’. In Herman, V. and Alt, J. (eds) Cabinet Studies: A Reader. London: Macmillan. Indridason, I. H. (2005) ‘A Theory of Coalitions and Clientelism: Coalition Politics in Iceland, 1945–2000’. European Journal of Political Research 44: 439–464. Jóhannesson, G. Th. (2006) ‘Leikstjóri, Leikari eð a Áhorfandi? Forsetinn og Stjórnarmyndanir’. Stjórnmál og Stjórns?sla 2, 1: 75–98. Kristinsson, G. H. (2005) ‘Pólitiskt Umboð ∂ og Ábyrgð ∂ Rá∂herra á Íslandi’. R. Spano (ed.) Rannsóknir í Félagsvísindum VI, Lagadeild. Reykjavik: Félagsvísindastofnun, 213–228. Kristinsson, G. H. (2006) ‘Sjálfstæð i Ráð herra og þingræð isreglan’. Stjórnmál og Stjórnsy´sla 2, 2: 152–164. Kristjansson S. (2003) ‘Iceland’. In Strøm, K., Müller, W. C. and Bergman, T. (eds) Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press. Laver M. Hunt, W. (1992) Policy and Party Competition. New York: Routledge. þjóð arpúls Gallup (various issues 1993–2006), published in Reykjavik.

Internet sources Alþingi (Parliament), www.althingi.is Government Offices of Iceland, http://raduneyti.is/Rikisstjornartal/ Morgunblað ið (daily newspaper), gagnasafn, www.mbl.is

12 Learning to rule Ministerial careers in post-communist democracies Katja Fettelschoss and Csaba Nikolenyi

The Third Wave of global democratization swept through Eastern and Central Europe in the early 1990s. By 2007, ten of the former communist states of the region (Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia and Slovenia) had completed sufficiently thorough and credible reforms in their political and economic systems to warrant their inclusion in the European Union (Vachudova 2005). Yet, while the adoption of competitive democratic systems of government was common to all of these post-communist states, the types of democratic institutions they have adopted vary significantly. With regard to their electoral systems, all but two of the ten states adopted various forms of proportional representation to elect members of the national parliament; the two exceptions are Lithuania and Hungary, which have used different mixed-member systems.1 The national legislatures of all but four of these new democracies are unicameral; we find bicameral parliaments in Poland, Romania, the Czech Republic and Slovenia, although in the latter case it is not a popularly elected chamber. The ten new EU member states also vary considerably with regard to the distribution of powers between the legislative and executive branches of government. In more than half of them, the constitutions provide for a popularly elected President (Bulgaria, Lithuania, Poland, Romania, Slovakia since 1999, and Slovenia); however, the overall levels of presidential powers even in these instances are fairly moderate in comparison with the more powerful head of state in Russia or Ukraine (Krouwel 2003). Nonetheless, even though the constitutional prerogatives of the head of state are carefully circumscribed, presidents in the new democracies of East Central Europe have often played an important role in shaping the games of cabinet formation and composition (Protsyk 2005). In fact, some claim that, even notwithstanding the limitations of the official powers of their office, presidents continue to be politically no less powerful than prime ministers (Baylis 2007)! This chapter starts with a country-by-country description of the constitutional foundations of ministerial selection and de-selection in the ten post-communist member states of the European Union followed by a crossnational comparative assessment and analysis of discernible patterns of variations. The third section deals with empirical indicators on ministerial selection and de-selection in the region. All data are based on Keesing’s Contemporary Archives Record of World events, but cross-checked with the RFE/RL database, the annually published Political Data Yearbook of the European Journal of Political

Learning to rule: post-communist democracies 205 Research and various government web pages, as for example the East European Constitutional Review or the websites of the different governments under study. Using the definition of government in Woldendorp et al. (1998, 2000), according to which government is terminated when the prime minister changes, a party leaves the coalition or an election is held, we have counted 103 governments in the ten states (Müller-Rommel, Fettelschoss and Harfst 2004, updated). 2 The dataset includes all ministers who assumed office up to 31 December 2006, but ministers of governments still in office at that time were excluded.

Country-by-country overview Hungary The constitution of post-communist Hungary establishes the rules on government formation in its Article 33/3 which states that the Prime Minister is elected by a majority of the members of parliament upon the recommendation of the President. It is further stated that both the person of the Prime Minister and the programme of the new government are voted upon simultaneously by the legislature. Upon successful passage of this vote, the Prime Minister recommends ministers to the new government who are in turn appointed by the President (Article 33/4) without any further parliamentary consultation. It is important to stress that the investiture vote calls for a majority of all the 386 members of the newly elected parliament to express a positive consent to the person of the head of government and programme that he proposes. The constitution is also very clear about the consequences of a failed investiture vote. According to Article 28/3b, if Parliament does not elect the Prime Minister who was nominated by the President within 40 days after the President made his first such nomination, the head of state can dissolve the legislature and call for new elections. This constitutional provision has not had to be invoked on any single occasion so far, not even in those instances when coalition talks between the formateur and proposed coalition partners appeared to be more protracted than usual. The 40-day period officially starts after the President has made the first formal nomination for Prime Minister after the election, which means that, in cases of apparent deadlock of difficult negotiations, the head of state can wait to offer parties more time to hammer out the details of their coalition bargain. At the same time, the constitution also allows the President to put pressure on parties, if necessary, for, once he has made his recommendation about who the new head of government should be, parties have up to 40 days only to arrive at forming a coalition capable of winning the investiture vote or else risk new elections. The Hungarian constitution is often noted for having laid down the foundations of a German-style chancellor democracy early on after the transition to democracy. Indeed, the constitution does provide for a very wide scope and range of Prime Ministerial authority and powers. As mentioned above, ministers are appointed and dismissed by the President on the recommendation of the head of government who is also free to appoint ministers without portfolio. Parliament can pass only a constructive vote of no confidence (39/A), which further buttresses the position of the Prime Minister. Moreover, since the creation and reorganization

206 The selection of ministers in Europe of ministries is outside the list of powers reserved for parliament, the Prime Minister enjoys considerable discretion in reorganizing and reshuffling his cabinet. The constitution does not render the role of the parliamentary deputy incompatible with that of a minister. Estonia The Estonian constitution clearly establishes in its Article 4 that the principle of separation and balance of powers shall constitute the foundation of the relationship among the main branches of government. Accordingly, it is not surprising to find that the constitution explicitly forbids members of parliament to be members of the government simultaneously (63, 64/1), although, similarly to other countries such as Bulgaria and Slovakia, a parliamentary deputy can have his or her mandate restored upon leaving the government (64/1). The selection of the Prime Minister and the formation of the government are given detailed description in Article 89 of the constitution. The President of the Republic has two successive opportunities to nominate a person for the position of Prime Minister. Within 14 days of such a nomination, the candidate has to present to Parliament (Reiskag) the bases on which he or she would proceed to form a government, which is followed by an immediate parliamentary vote to authorize the candidate to proceed. If the vote fails, or if the vote passes but the government formation process fails within seven days, then the entire process is repeated one more time. If the candidate succeeds in forming a government, its members are immediately appointed by the President without further parliamentary input in the process. If, however, no government is formed after the first two attempts, then the right to nominate the Prime Minister is transferred to parliament. Failure to produce a government on this third trial results in new elections (89/6). The Estonian Prime Minister does not command complete authority over his council of ministers. Although changes to the composition in the ministerial team can be effected only upon his proposal to the President (90), parliament can also censure individual ministers by passing a vote of no confidence (97/1). The head of government is also constrained in his ability to reorganize ministries, which can be created and altered only by legislation. However, the Prime Minister can appoint ministers without portfolio to his cabinet (94/4). Latvia The Latvian constitution is the most laconic of all ten constitutions considered. With regard to government formation, it simply states that the person invited by the President should form a government (56). Although parliament (Saeima) does not need to pass an investiture vote in the newly appointed government, it can pass a vote of no confidence either in the entire council of ministers or in its individual members (59). Barring this vote of no confidence in ministers, the power of the Prime Minister over his ministers’ careers is virtually unconstrained since the constitution is completely silent about other possible ways in which changes can be made to the personnel composition of the government. Parliamentary authority

Learning to rule: post-communist democracies 207 over the government is buttressed by the provision that requires the passage of legislation in order to change either the number of ministries or the scope of their activity and authority (57). Although the constitution does not mention explicitly the incompatibilities that members of the government face, it does make it clear in Article 63, regarding ministerial initiatives, that ministers can but do not have to be members of the legislature. Lithuania The selection and appointment of the Lithuanian Prime Minister is at the discretion of the President. However, the new head of government, whose authority to nominate and choose his ministers is limited only by the provision that members of the government must be members of parliament (60/2), must present to parliament (Seismas) the programme of the government within 15 days of his appointment (92). If parliament fails to vote on this programme within 30 days or if it reject the programme twice within a 60-day period, then the President of the Republic may call for fresh parliament elections (58/2/1). The constitution imposes several checks on Prime Ministerial authority over the government. First, it stipulates that ministers must head individual ministries, which effectively prevents the Prime Minister from appointing loyal political allies to the government as ministers without portfolio (98). Secondly, the constitution also requires legislation for changing and establishing ministries, although this can only be done upon the recommendation of the government (67/8). Thirdly, although the Prime Minister is free to dismiss members of his government, via recommendations to the President, his government must receive a new vote of confidence from the legislature if more than half of the original members of the government have changed (101). Finally, parliament can pass a vote of no confidence not only in the government as a collective but also in the individual ministers (67/9). Poland Of the ten post-communist EU members considered, Poland has gone through the most extensive constitutional changes with direct effect on the institutional sources of the selection and de-selection of ministers. Articles 57–62 of Small Constitution of 1992 provided for multiple rounds of voting for Prime Minister to take place between the President and the Sejm, the lower house of the bicameral parliament. The first move of the process was given to the head of state who could appoint, as opposed to merely nominating, the Prime Minister and, on the latter’s motion, the members of the new government within 14 days after the first sitting of the Sejm. The newly appointed government, however, had only another 14 days to submit its programme to the Sejm and receive from it a positive investiture vote. Failure to do so would result in the Sejm receiving a second opportunity to nominate a Prime Minister and vote him or her as well as members of the new government into office within 21 days. Successful passage of this vote obliged the President to appoint the new government into office (Article 58), while a failed investiture vote resulted in the repetition of the entire process. In other words, the

208 The selection of ministers in Europe President would have one more opportunity to select a Prime Minister and, in case of a failed parliamentary vote, the Sejm would also have one more chance to select a new head of government within the same time frames as before. In case of yet another failed vote, the President could either immediately dissolve the Sejm and call for new parliamentary elections or he could appoint a new government which would either have to receive a vote of confidence from the Sejm or survive a no-confidence vote within six months or the chamber would be automatically dissolved (Article 62). The so-called Small Constitution, which was largely limited to regulating the relationship between the different branches, and levels, of government, was replaced by the new constitution of 1997 which was narrowly passed by a popular referendum in May of that year. The new constitution changed the investiture process in three significant ways. First, it reduced the time frame available to the two major players, i.e. the President and the Sejm, to arrive at an agreement on the new government. While the President still had 14 days to appoint a new Prime Minister after the first sitting of the Sejm, the latter would also have the same 14 day period, as opposed to 21 days under the Small Constitution, to make a counter-nomination in case the President’s choice cannot win the investiture vote. Secondly, the new constitution also limited the Sejm to making only one such counter-nomination, in contrast to two attempts under the Small Constitution. Thirdly, Presidential powers were also curtailed by removing the President’s discretionary authority to appoint a government, in case of a deadlocked investiture game, which would have to prove its parliamentary support within six months. The new constitution also increased the political costs of deadlock by making it obligatory to dissolve the Sejm immediately and hold new parliamentary elections. The 1997 constitution significantly increased the authority of the Prime Minister over his cabinet vis-à-vis the politically powerful Presidency in a number of ways. First, while the President was able to appoint three important members of the council of ministers under the Small Constitution (the three presidential portfolios were Defence, Foreign and Home Affairs), the new constitution gave the Prime Minister the unequivocal right to make changes in his ministerial team, although it still needs the consent of the President. The new constitution also established a constructive no-confidence mechanism; however, the Sejm may still censure individual ministers by a regular no-confidence vote (159). The constitution provides a detailed list of offices that are incompatible with that of a parliamentary deputy; however, it explicitly mentions that deputies can be appointed to government (103). Ministers can either be at the head of individual ministries or may perform particular functions assigned by the Prime Minister (149). The right to appoint such minister with special tasks, without heading an executive portfolio, further increases the power of the Prime Minister within the cabinet. Finally, the 1997 constitution also gave the government the right to control its own internal organization, as opposed to making it dependent on legislative authorization (146).

Learning to rule: post-communist democracies 209 Czech Republic Article 68/2 of the Czech constitution allows the President to appoint the Prime Minister, and other members of the government upon the latter’s motion, without specifying any time limit within which the appointment has to take place. However, once the appointment has been made, the new government has to get an ordinary vote of confidence from the Chamber of Deputies, the lower house of the bicameral Czech parliament. If the government fails to win a confidence vote, the President will appoint yet another government, which would have another 30 days to prove that it enjoyed the confidence of the Chamber. In contrast to Poland, the Chamber, therefore, does not have the power to counter-nominate a Prime Minister. However, it is important to note that a second failed appointment allows the Chairman of the Chamber (the speaker) to recommend the next candidate for Prime Minister who in turn is appointed by the President. Should this government also fail to win a confidence vote, the President may dissolve the Chamber of Deputies (Article 35). The Czech constitution allows the lower house of parliament to pass a vote of no confidence only in the government as a whole but not in its individual ministers (71). While the Prime Minister must submit his resignation directly to the President, other members of the cabinet have to do so through the Prime Minister (73). The Chamber of Deputies, however, is entirely removed from the process of effecting or confirming changes in the ministerial composition of the government, which, ceteris paribus, strengthens Prime Ministerial authority. However, the constitution limits the powers of the head of government by stressing that the government resolutions must be passed with the support of an absolute majority of its members (76/2), and that ministries and state agencies can be created only by law duly passed by parliament (79/1). The constitution allows cabinet ministers to be drawn from the legislature; however it also provides a carefully constructed list of the positions that parliamentary deputy who is also a member of the government cannot occupy. These positions are the chair- and vice-chairmanship of both houses of the Czech parliament, any parliamentary committee or commissions of inquiry (32) Slovenia The constitution of Slovenia provides for a multi-round process of choosing and selecting the Prime Minister that involves both the President, who is elected directly by the voting population, and the lower house of parliament, the National Assembly (111). The President has the right to nominate to the National Assembly, after he has consulted with political parties represented in parliament, a candidate for Prime Minister whose appointment is subject to a parliamentary vote of investiture. If this vote fails, then the President can nominate another candidate, who can be the same person, within 14 days. Should this nomination also prove unsuccessful, parliamentary party groups with at least ten deputies can also put forward their nominees who are voted upon in the order of their nomination. However, if the President wishes to nominate yet another candidate of his own choosing then this candidate will be voted upon first by the Assembly. The

210 The selection of ministers in Europe President can dissolve the Assembly and call for fresh elections if this third attempt also fails. However, the Assembly can counter the order of dissolution by passing a vote in favour of yet another round of balloting. Failure to choose a head of government on this fourth ballot, however, automatically triggers fresh elections. Although the Assembly has to vote only upon the person of the Prime Minister, individual ministers are formally appointed, and dismissed, by the Assembly upon the recommendation of the head of government (112). The Assembly, however, cannot directly dismiss a minister by passing a vote of no confidence; in fact, the Assembly can remove the Prime Minister only by passing a constructive vote of confidence against him (116). The National Assembly has the power to alter the number and the competency of the various ministries (113). The constitution is silent about the incompatibility of ministerial office with other positions, which leaves the ministerial ‘selectorate’ potentially very wide in its scope. Slovakia The Slovak constitution is similar to its Czech counterpart in two important ways: first, the President does not have a specific time limit during which he must appoint the new government, and, secondly, once appointed the new government must win a confidence vote from the National Council within 30 days (Article 113). In contrast to the Czech case, however, an unsuccessful formation attempt in Slovakia is not linked to the possibility of dissolving the legislature nor can it be followed by an alternative recommendation coming from the Council. However, it might be worth noting that the inability of the National Council to act on a confidence motion in general does allow the President to dissolve it (Article 102/1b). Duly elected members of the Slovak National Council are eligible to be appointed as government ministers; however, they cannot exercise their mandate while they hold executive office (77/2). The National Council has significant powers over both the organization and the ministerial composition of the government. First, the establishment of ministries and other state institutions is tied to the passage of relevant legislation (86/f) and, secondly, the Council can pass votes of no confidence both in the government as a whole and in its individual members (86/g). The authority of the Prime Minister is checked by the provision that allows individual ministers to submit their resignation directly to the President of the Republic rather than to the head of government (116/2). Bulgaria Although the Bulgarian constitution provides the directly elected President with the power of appointing the Prime Minister (Article 99/1), it also imposes significant constrains on presidential authority. The constitution stipulates that the President must appoint as Prime Minister that person who is nominated by the party that won the plurality of seats in the National Assembly. Should this candidate fail to form a government within a short period of seven days, the President must invite the nominee of the second largest party in parliament who also has seven days to form a council of ministers (99/2). In case of yet another

Learning to rule: post-communist democracies 211 failed attempt at government formation, the President is free to invite a candidate put forward by one of the smaller parties (99/3). Only when a candidate for Prime Minister succeeds in forming a government does the President of Bulgaria ask the National Assembly to formally elect the person as head of government together with the other ministers (99/4, 84/7). If all three rounds of government formation prove unsuccessful, the President can appoint a caretaker government and call for fresh parliamentary elections. Once elected, the Prime Minister becomes the head of government; however, the Assembly also retains a considerable degree of authority over the cabinet. The hiring and firing of deputy ministers remains the sole prerogative of the head of government (108/2) and the National Assembly can elect and dismiss ministers only upon his or her recommendation (84/6). Moreover, while the Assembly has the legislative power to create new ministries or terminate existing ones, it can do so only upon the proposal of the head of government (84/7). At the same time, the constitution does not allow the inclusion of ministers without portfolio in the council of ministers unless it has passed a resolution to this effect (108/3). A further important limitation on Prime Ministerial authority is the absence of a constructive no-confidence clause in the constitution. The National Assembly can dismiss the government by passing an ordinary no-confidence vote; however, it cannot censure individual ministers. Furthermore, it is only the council of ministers as a collective group, as opposed to the Prime Minister alone, that can rescind an act issued by an individual minister (107). The constitution makes explicit reference to the scope of the ministerial ‘selectorate’ by providing for the so-called ‘sleeping mandate’, according to which an elected member of the Assembly can become a minister however, during his or her service in government a substitute deputy would have to be provided for (68). Romania The constitution of Romania gives parliament the power to invest the Prime Minster in office (102). The directly elected President can nominate a candidate for head of government after he has consulted with the majority party in parliament or, in the absence of such a party, with all other parliamentary party groups. The Prime Ministerial candidate has ten days to present his new government and its programme for a vote of investiture to both houses of the Romanian parliament that have to vote in a joint session. If two such attempts at investing a Prime Minister fail or if Parliament has not passed a vote of confidence within 60 days after the first nomination was made by the president, the head of state can dissolve the legislature and call for new elections (89/1). The Prime Minister enjoys a considerable degree of authority over his government. Although ministries and state institutions can be created and changed only by the legislature (72/3d, h), the head of government can make personal changes in his government without seeking parliamentary approval (85.2). The two houses of parliament can pass a vote of no confidence only against the government as a whole but not against its individual members (112). Government ministers can be members of either house of parliament; however, they cannot occupy any other public office (104/1). Article 106 reaffirms the authority of the

212 The selection of ministers in Europe Prime Minister by stating that he ‘shall direct Government actions and co-ordinate activities of its members’ (106/1).

Comparative assessment Before delving into a more detailed comparative assessment of the constitutional constraints on cabinet formation, change and termination, it is worth noting a few important general characteristics of the actual governments that have been formed in the ten post-communist democracies. Table 12.1 provides information on the number of governments formed in each country, their average duration, the average duration of ministers in one position and the average number of ministerial positions and actual ministers in each case. According to the table, governments in the ten new democracies have been fairly unstable, as suggested both by their high average number and by their short duration. The average values, of course, conceal important variation across the cases. For instance, during the same period of time, Hungary has seen the formation of fewer than half of the number of governments in Poland, and, as a result, the average duration of Hungarian cabinets has been almost three times the length of their Polish counterparts. With regard to the number of cabinet posts, there is much less variation across the states, while the number of ministers does vary considerably. The average duration of governments and the average duration of ministers in one position are very closely connected (r2 = 0.964). The overwhelming majority of governments in these countries were formed by multi-party coalitions. Of the 46 elections that have been held in the region since the transition to democracy, only five resulted in an outright majority victory by a single party or electoral coalition running on the same list (Bulgaria 1994, 1997, Hungary 1994, and Lithuania 1992 and 1996), and only one of these elections (Hungary 1994) was followed by the formation of a coalition government involving the majority party, the Hungarian Socialists, and a junior partner, the Free Democrats, in an Table 12.1 Government characteristics in post-communist democracies, 1990–2007 Average number of ministerial positions

Average number of ministers

Country

Starting point

Number of governments

Average duration of governments

Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovakia Slovenia

11/08/91 01/01/93 10/21/92 05/23/90 07/04/93 12/02/92 12/23/91 11/13/92 01/12/93 01/12/93

8.0 8.0 10.0 7.0 15.0 10.0 16.0 11.0 9.0 9.0

1.97 2.02 1.39 2.67 0.95 1.51 0.98 1.41 1.68 1.49

18.0 17.1 15.0 17.9 14.9 16.2 20.6 20.3 17.2 18.1

18.5 13.8 10.1 16.9 7.5 11.8 11.5 12.6 11.9 10.1

10.3

1.61

17.5

12.5

Average

Learning to rule: post-communist democracies 213 oversized surplus majority government. In addition to these five cases, there were also five other elections followed by the formation of a single-party minority government (Bulgaria 1991, Czech Republic 1998, Poland 2005, Romania 1992 and 2000). Table 12.2 provides a comparative summary of the constitutional provisions in the ten post-communist EU member states that have direct bearing on the selection, removal and position of government ministers. The table identifies six main dimensions along which the ten constitutions are compared: • • • • • •

selection of the Prime Minister selection of cabinet ministers (in)compatibility of ministerial and parliamentary deputy positions constructive vote of no-confidence procedure against the Prime Minister no-confidence procedure against individual ministers control over the (re)organization of government, i.e. where the power to create and change government ministries is located.

We now review how these ten states compare on each of these dimensions. Selection of the prime minister With regard to the first dimension, four different types of Prime Ministerial selection processes can be identified in the constitutions of the ten post-communist democracies. In three cases (Hungary, Romania, and Slovenia) the role of the President in selecting the Prime Minister is limited to a formal nomination of a candidate for the office while the actual investiture in office is done via a positive parliamentary vote. In six other states (Bulgaria, Czech Republic, Estonia, Lithuania, Poland, Slovakia) the role of the legislature is still important; however, it is more of a reactive kind. In these cases, the President actually appoints the head of government in office without prior parliamentary authorization. However, within a clearly and expressly specified period of time, the newly appointed government must receive a vote of confidence from the legislature. It is further worth stressing that the constitutions of Estonia, Poland and Slovenia provide a positive role for parliament to make nominations to the office of Prime Minister in case the President’s choice, or choices, fail to be confirmed. In stark contrast, the constitution of Latvia offers an idiosyncratic way of selecting the Prime Minister by placing it completely at the discretion of the President. Although political parties clearly play a crucial role in the actual process of government formation, including the selection of both the Prime Minister and the rest of the government, their position and role are formally acknowledged and specified in the constitutions of only three of the ten cases (Bulgaria, Romania and Slovenia). In each of these cases, parties are mentioned only with respect to the selection of the Prime Minister. As mentioned, the Bulgarian constitution makes a particular point about stressing the role of parties by giving the party winning the most seats in the election the first mandate to form a government. While it may not be surprising that the plurality party in Bulgaria always cashed in on its first-mover

Appointed by President

Appointed by President, vote required Nominated by President, authorized by Assembly Nominated by President, elected by Assembly Appointed by President Appointed by President, vote required Appointed by President, vote required Nominated by President, elected by Assembly Appointed by President, vote required Nominated by President, elected by Assembly

State

Bulgaria

Czech Republic

Slovenia

Slovakia

Romania

Poland

Latvia Lithuania

Hungary

Estonia

Selection of Prime Minister

Prime Minister recommends to President Prime Minister recommends to President Prime Minister recommends to President Prime Minister Prime Minister recommends to President Prime Minister recommends to President Prime Minister recommends to President Prime Minister recommends to President Prime Minister recommends to Assembly

Prime Minister

Selection of ministers

Parliament

Parliament

Parliament

Government

Parliament Parliament

Government

Parliament

Parliament

Parliament

Control over (re)organization of government

Yes

No

No

Yes

No No

Yes

No

No

No

Constructive vote of no confidence against Prime Minister

Table 12.2 Constitutional provisions on key aspects of cabinet life in post-communist democracies, 1990–2007

No

Yes

No

Yes

Yes Yes

No

Yes

No

No

Vote of no confidence against individual ministers

Sleeping mandate Compatible

Compatible

Compatible

Compatible Compatible

Sleeping mandate Compatible

Sleeping mandate Compatible

Compatibility of ministerial appointment with being an MP

Learning to rule: post-communist democracies 215 advantage, and succeeded in forming the post-election government, the case of the formation of the Stanishev government after the 2005 elections may be worth noting in order to illustrate the difficulties that might emerge even in the presence of an explicit constitutional definition of the process (Spirova 2006: 620). The 2005 Bulgarian election resulted in a significant jump in parliamentary fragmentation to 4.8 (from 2.9 in the previous polls!) while also bringing a number of new parties to the legislature. Although the Socialist Party won a plurality of the seats and thus received the mandate to form a government, it failed to forge a majority solution, thus letting the next largest party, the National Movement Simeon II (NDSV), lead the next attempt. Curiously, the NDSV also failed at its own attempt, thus bringing the government-formation process to a halt. Eventually, 51 days after the election, a new Bulgarian government was formed under the leadership of Prime Minister Sergey Stanishev of the Bulgarian Socialist Party. The coalition included the three largest parties in parliament, the BSP, the NDSV and the Movement for Right and Freedoms (DPS), a party devoted to the representation of the interests of Bulgaria’s Turkish minority. The allocation of portfolios among the three coalition partners reflected very closely their proportional contribution to the total parliamentary size of the coalition.3 In spite of the absence of formal constitutional provisions on the role of political parties in most of these new democracies, there is a fairly well established convention according to which the plurality party gets the first chance to try to form a government after a parliamentary election. The main exceptions to this pattern are the following post-election governments: Olszewski (Poland, 1991), Tariceanu (Romania, 2004), Paksas (Lithuania, 2000), Kristopans (Latvia, 1998), Dzurinda-I and Dzurinda-II (Slovakia, 1998 and 2002 respectively) and Laar (Estonia, 1999). The specific reasons why the largest party failed to form a government after these elections vary considerably from case to case. However, in broad strokes we can sort these seven cases in two groups: the first one is made up by the Olszewski, Tariceanu and Paksas government in the formation of each of which, and thus in the prevention of the formation of an alternative government by the plurality party, the directly elected, and politically quite powerful President played a critical role (Downs and Miller 2006; Fitzmaurice 2003). The second group consists of the remaining four cases (Kristopans, Dzurinda I and II, and Laar) where the central reason for the inability of the largest party to form a government had to do with the relative marginalization of the plurality party (the Centre Party in Estonia, the Movement for Democracy in Slovakia, and the People’s Party in Latvia) in the bargaining space and their genuine inability to attract coalition partners. Selection of cabinet ministers The ten new democracies show somewhat less nuanced variation on the second dimension, i.e. the composition of the government and the selection of ministers. The definition of government is of two main types. In half of the cases (Bulgaria, Czech Republic, Hungary, Slovakia and Poland), the national constitutions make explicit mention of Deputy Prime Ministers as members of the governments, while

216 The selection of ministers in Europe in the other five cases Deputy Prime Ministers are not mentioned. Parties in postcommunist democracies, as elsewhere, often resort to cementing their coalition deals by allocating the more prestigious and influential Deputy Prime Ministership among leaders of the coalition partners. The cases of Hungary, where no actual Deputy Prime Ministers have been appointed, and that of the Slovak National Party, which did not receive such a post in Vladimir Meciar’s three-party coalition government after the 1994 election, are exceptions to this norm. Interestingly, we find Deputy Prime Ministers even in single-party governments where the post is used to accommodate rival factions and nurture future aspirants for leadership within the governing party. A good case in point is the single-party Social Democratic government of Milos Zeman, formed after the 1998 Czech elections, where the Deputy Prime Ministership was given to a senior party member, Vladimir Spidla, who indeed became the Prime Minister of the next government. The most frequently used process of ministerial appointment is that the Prime Minister makes a formal recommendation to the President regarding each member of his cabinet. The only exceptions in this respect are Bulgaria and Latvia, where the Prime Minister does not have to obtain the formal consent of the head of state to select and appoint members of his council of ministers, and Slovenia, where the Prime Minister has to make his or her formal recommendation to the National Assembly as opposed to the chief executive of the state. The role that political parties play in the selection of cabinet ministers can be appreciated by considering the importance of the formal coalition agreements that parties conclude upon agreeing to form a government. While these formal coalition agreements range significantly in terms of their depth and scope, they almost invariably provide detailed information about the allocation of particular ministerial portfolios to the particular coalition partners. In spite of the often complex nature of coalition bargaining, governments in the ten post-communist democracies have not taken a very long time to form. Table 12.3 reports on the average length of time that it has taken political parties to form a government after a general election in each country. The overall average for all governments across the ten countries is 44 days, and only two states, the Czech Republic and Slovenia, have significantly Table 12.3 The length of government formation periods in post-communist democracies (average days after election), 1990–2007 Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovakia Slovenia

37 78 33 44 46 32 40 37 37 62

Average

44

Learning to rule: post-communist democracies 217 higher averages than this. However, these deviations are single-handedly due to the anomalous cases of the Topolanek government in the Czech Republic after the 2006 election and the Drnovsek government in Slovenia after the 1996 elections which unusually took a very long time to form, 219 and 109 days respectively. In both cases, government formation was much more protracted than usual owing to the near-even split of parliamentary seats between the two contending blocs of parties in each country. Another important variable regarding the selection of ministers is the ability of the Prime Minister to appoint politically important figures to the government without assigning them to particular ministerial portfolios. In three cases (Estonia, Hungary and Poland), the constitution formally allows the Prime Minister to appoint to the government ministers with special tasks but without a particular portfolio assignment. In 2007, the Estonian government has one, and the Hungarian and Polish government include three such ministers. It is only the constitution of Lithuania that formally forbids the presence of ministers without a portfolio in the government, while the constitution of Bulgaria ties such a ministerial appointment to parliamentary authorization. Not surprisingly, there are no ministers without portfolio in the government of either country today. The constitutions of the other states are silent on this issue, which technically does not prevent the Prime Minister from making such appointments if and when he or she sees fit. Of the remaining countries only the government of Slovakia does not have at least one minister without a portfolio in 2007. Compatibility of executive and parliamentary posts While none of the constitutions of the ten states renders the office of a government minister incompatible with that of a parliamentary deputy, in three cases (Bulgaria, Estonia and Slovakia), the parliamentary mandate of the deputy who is appointed to government is suspended as long as the person is in executive office. In all other countries, however, the two offices are compatible with each other, which makes the legislature a natural and often-used pool to select ministers from. Indeed, 41.8 per cent of all ministers in Central and Eastern Europe have been members of parliament before reaching the office. As indicated earlier, the Czech Republic stands out from the rest of group by virtue of having a constitutional stipulation that forbids ministers from occupying key legislative positions such as the chairmanship of either house of parliament or membership on committees. While the other countries do not have such an explicit rule on the incompatibility of ministerial and legislative offices, the norm has been to keep ministers away from the demands of committee business.

Votes of no confidence While constitutions obviously cannot address the informal political ways in which ministers are de-selected and removed from office, they are very explicit about the role that legislatures can play in the process. In this regard, Table 12.2 helps us identify four groups of countries. The first group includes Hungary and Slovenia,

218 The selection of ministers in Europe where parliament can censure the government only by passing a constructive vote of no confidence against it and it is not allowed to censure individual ministers. The joint presence of these provisions renders the position of the head of government very powerful in the political systems of these two states vis-à-vis the legislature as well as in his or her own cabinet. Poland constitutes the second category on its own: here the power of parliament to dismiss the entire cabinet is limited by the constructive no-confidence mechanism; however, it can still dismiss individual ministers by passing a vote of censure against him or her. By comparison with Hungary and Slovenia, this latter provision renders the institutional foundations of the Polish head of government somewhat weaker. The third group consists of the two most recent members of the EU (Bulgaria and Romania) as well as the Czech Republic. In these cases, the legislature can dismiss the government more easily than in the previous three countries since the no-confidence vote does not have to be constructive. However, parliament cannot interfere with the composition of the cabinet by censuring its individual members. Finally, the constitution provides for a very powerful legislature vis-à-vis ministers in the fourth group of states which includes the three Baltic democracies (Estonia, Latvia and Lithuania) and Slovakia. Here the legislature can dismiss both the government as a whole and its individual ministers by passing a simple vote of no confidence. The rate at which legislatures have used their powers to replace governments varies considerably across the ten cases. On the basis of the information that Table 12.1 provides about average cabinet duration we can identify two groups among the ten post-communist democracies. The first group includes five cases where average cabinet duration is longer than the group average: these are Bulgaria, the Czech Republic, Hungary, Romania and Slovakia. The second group contains the three Baltic countries (Estonia, Latvia and Lithuania) as well as Romania and Slovenia. The relatively unstable executives of Slovenia and Poland are somewhat of a puzzle given that the constitutions of these countries, like that of Hungary, provide for a constructive vote of no confidence in the head of government, which, all else equal, should strengthen executive stability. Control of government reorganization The last dimension of analysis provides information about the extent to which Prime Ministers are constrained in their ability to reorganize the institutions of state administration in order to suit their particular political and programmatic needs. Only in two of the ten cases (Hungary and Poland) does the government actually have control over its own administrative organization; in all other cases parliament must pass legislation to authorize such changes. This might, at least in part, explain why the number of ministries has changed so much in Hungary (moving from 18 after 1990 election to 12 in 2007). Since it was not until 1997 that the Polish government acquired the power to control its own organizational structures, there has been less fluctuation in the number of Polish ministries (Zubek 2001).

Learning to rule: post-communist democracies 219

Empirical analysis Ministerial selection We examine the patterns of ministerial selection in the ten states along three dimensions: age, gender and educational background. In the discussion of each of these dimensions, we start our empirical assessment on ministerial selection and de-selection with the apex of the cabinet, the Prime Minister. Table 12.4 reports the average age of post-communist Prime Ministers per country. The data allow us to make a couple of important observations. First, Prime Ministers in general have been fairly young: the average age on becoming head of government in these countries has been under 50. While Latvia and Estonia have the youngest Prime Ministers, on average, Poland and Hungary have the most senior ones. Secondly, Bulgaria, Lithuania and Hungary stand out with having high double-digit standard deviation scores, suggesting an important generational diversity amongst Prime Ministers in these states. Indeed, the age gap between the youngest and oldest heads of government in these states is quite striking: the youngest Prime Ministers coming into office were Filip Dimitrov (UDF) in Bulgaria at the age of 36, Viktor Orban (Fidesz) at the age of 35 in Hungary and Gediminisu Vagnorius (Sajudis) in Lithuania at the age of only 34. By contrast, the oldest Prime Ministers in these states were 67 (Lyuben Berov of Bulgaria), 63 (Imre Boross of Hungary) and 69 (Algirdis Brazauskas of Lithuania). The latter three were also the oldest Prime Ministers not only in their respective states but also in the entire group of ten post-communist democracies. The average age of cabinet ministers is 47 years and therefore very similar to that of Prime Ministers. While Estonia and Latvia tend to have younger ministers, Hungary and Poland are at older end of the spectrum. With regard to gender, the new democracies present an appalling picture: of the 103 governments considered, only one head a female head: Hanna Suchocka of the Polish government formed in 1992. Although women are slightly better represented as cabinet ministers (9.6 per cent of all ministers were female) than Prime Table 12.4 The average age of Prime ministers in post-communist democracies, 1990–2007

Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovakia Slovenia ECE-10

Average age

Standard deviation

49.4 48.5 44.9 53.8 43.2 47.7 50.2 48.9 47.2 47.4 47.8

11.5 7.0 7.6 12.0 5.2 10.3 8.0 4.4 4.4 6.3 8.3

220 The selection of ministers in Europe Ministers, the legislature remains the most representative institution of women in these post-communist democracies (on average 14.3 per cent of parliamentary seats have been filled by female MPs). Nonetheless, we still find important differences across our cases with regard to the appointment of women to ministerial portfolios: the percentage of female MPs in Latvia (16.1 per cent), at the high end of the range, is more than twice as high as that in Hungary (6.8 per cent) and Lithuania (5.9 per cent), which are at the low end of the range. Most of the ten states already had female ministers either in their first (Bulgaria, Estonia and Slovakia) or second (Latvia, Poland, Slovenia) post-transition cabinets while the others (Czech Republic, Hungary, Lithuania, Romania) started to do so only with their fifth, or later, governments. All Prime Ministers who have assumed office up to now are in possession of a university degree. Likewise, nearly all ministers in Central and Eastern Europe (97.4 per cent) hold a university degree. Other degrees, such as primary or secondary education (0.9 per cent) military education (1.0 per cent) and technical education (0.6 per cent) play only a minor role. There is only slight variation between the countries, with Lithuania (100.0 per cent university degree) and Hungary (99.1 per cent university degree) being on the upper end of the scale and the Czech Republic (93.6 per cent) on the lower end. Ministerial de-selection There are six different groups of reasons why ministers leave office. This can be (1) because of the end of term, (2) for personal reasons, (3) because the minister is dissatisfied, (4) caused by the dissatisfaction of third parties, (5) scandals or (6) cabinet reshuffles. In addition, these groups are subdivided into smaller groups specifying the reason why a minister quits office. Table 12.5 represents the reasons for leaving office for all countries. The first figure to attract attention in Table 12.5 is the high rate of ministers who leave office simply as a result of government ending. More than 80 per cent of all ministers fall into this category. Additionally, only a small number of ministers definitely leave their office after termination of government. For about four out of ten ministers, the end of cabinet coincides with the (preliminary) end of their ministerial career. An even more interesting finding in category 1 is the high rate of immediate returnees to cabinet. More than one-third of all ministers who formally leave the cabinet because of its termination return to their position immediately after a new cabinet is formed. In other words: on average, in every new cabinet which comes into office, one minister out of three has already served in the preceding cabinet. In contrast, the number of persons who leave a cabinet and return to office at a later point is very small: only 6.41 per cent of all ministers belong to this group. Since the number of ministers who leave office regularly is high, only 18.9 per cent of all ministers are attributed to the remaining five groups which refer to reasons for leaving office during the term of government. These five groups split up into two large and three small groups. Most ministers who do not leave office at the end of their cabinet do so because others are dissatisfied with their work (6.6 per cent) or because of a cabinet reshuffle (5.3 per cent). The remaining three groups (personal reasons, dissatisfaction and scandals) score clearly below 5 per

Learning to rule: post-communist democracies 221 Table 12.5 Reasons for resignations in post-communist democracies, 1990–2007 Goup

Reason for leaving office

Number

End of term

conclusive returning immediately returning at later point subtotal

756 701 125 1582

38.47 35.67 6.36 80.51

Personal reasons

new job health private life subtotal

16 6 5 27

0.81 0.31 0.25 1.37

Minister dissatisfied

with government work with overall situation with position/success subtotal

20 3 20 43

1.02 0.15 1.02 2.19

Others dissatisfied or struggles with

prime minister other ministers president parliament party/coalition public/press others several groups subtotal

44 8 4 7 38 15 3 10 129

2.24 0.41 0.20 0.36 1.93 0.76 0.15 0.51 6.56

Scandal

political personal subtotal

48 18 66

2.44 0.92 3.36

Cabinet reshuffle

political reasons organizational reasons temporary position subtotal

43 26 34 103

2.19 1.32 1.73 5.24

1950

99.24

15

0.76

1965

100.00

Total not known Total

%

cent. With 3.4 per cent of all cases, scandals appear to be a more important reason for leaving office than the dissatisfaction of a minister (2.2 per cent). Personal reasons (1.4 per cent) are the least important motivation for giving up an executive career. What particularly stands out is the fact that most ministers who abandon office while the government is still running do so under pressure from others. Only 3.65 per cent of ministers leave their post of their own volition compared to 15.3 per cent who do so because of external pressure. Overall, the figures in Table 12.5 show that most ministers in Central and Eastern Europe reach the end of their cabinet. That means that either the number of conflicts is not very high or that most conflicts do not lead to the sacking of the minister involved. Furthermore, when a minister has to leave the cabinet during term, he usually does not do so out of his own decision. Very often, the institutions

222 The selection of ministers in Europe directly linked to the government (Prime Minister and parties) play an important role in the dismissal of a minister. But the importance of other bodies such as the media and the public should not be underestimated. Table 12.6 is dedicated to the differences between the single countries. The main question is whether the differences according to the constitutions are also to be seen in the empirical data. Referring to the observations made in the first part of this chapter, one can see that in countries where the Prime Minister is rather free in deciding the composition of his cabinet, the number of ministers who regularly finish a term is comparatively high. This is especially true for Latvia and Slovenia, but also applies to Bulgaria. Therefore, at the upper end of the turnover ranking, we find Latvia, Slovakia and Slovenia. In these cases, the number of ministers who leave office at the end of a term is especially high. In the cases of Latvia and Slovenia, this goes hand in hand with a high number of immediate returnees and a very low percentage of ministers who do not come back. Romania exhibits a similar pattern, even though it has much lower internal government stability. At the bottom end of the turnover ranking, there are three countries, the Czech Republic, Hungary and Romania, which show particularly high minister turnover figures. At the same time, these countries also exhibit a very low number of ministers who return to office at a later point. Slovakia stands out for the large number of ministers who return at a later point to reassume an executive position. Another country that stands out is Bulgaria. Here the number of ministers who do not come back after having once left is extremely high. Concerning the reasons why ministers leave their office before a cabinet term is finished, we observe a very large variation. From the figures exposed in Table 12.5 it is already clear that personal reasons are least important for abandoning a ministerial position. Nevertheless, the variation between the single countries allows a classification into two groups. Bulgaria, Estonia, Latvia and Slovakia, belong to the first group, where personal reasons virtually play no role at all for leaving office. In contrast, in the Czech Republic, Hungary, Lithuania, Poland, Romania and Slovenia, there are some cases where a minister decides to leave for personal reasons. In most countries, the cases are equally distributed between the subgroups, meaning that new jobs are as important as health problems or private reasons for abandoning a ministerial position. Hungary and Slovenia, where all ministers leave because of a new job offer, are exceptions. The next largest group, scandal, is one of the three groups that refer to situations when ministers leave involuntarily. Scandals are in all observed countries relevant for the sacking of ministers. However, we can note some differences. Scandals are more often a motive for firing ministers in Estonia and Lithuania but less frequently lead to a dismissal in Bulgaria and Poland. Cabinet reshuffles have an important role in the premature termination of ministerial careers. From Table 12.6 we can observe that reshuffles are common in some countries such as Bulgaria, Hungary and Romania and unusual in other countries, especially in Latvia, Slovakia and Slovenia. The numbers in the subcategories (political reasons, organizational reasons and temporary position) are quite equally distributed. The Czech Republic is an exception, given that no ‘real’

Learning to rule: post-communist democracies 223 reshuffles have taken place. All the frequencies in this case must be attributed to the exchange of ministers who have been temporarily in a position. In the largest group, which refers to the dissatisfaction of third parties, the numbers do not spread very much. This means that the rather high occurrences are due to the central relevance of third-party dissatisfaction in sacking ministers in all observed countries. Only one country, Slovenia, has considerably lower numbers than the others. Furthermore, there are three countries which show slightly higher figures: Bulgaria, the Czech Republic and Poland. Generally speaking, there are two sub-groups which are especially important in the process of ministerial dismissal: the Prime Minister and the party or coalition. The president has in three cases been the driving force in sacking ministers in Poland. Poland is, together with Estonia, the only country where the president has ever affected the fate of a minister. This goes in line with the important position the constitution attributes to the head of state in these two countries. Speaking for all countries, parliament does not play an important role in the firing of government members. Slovenia forms together with Poland and Romania the only cases where the parliament has actively exercised its influence. Given the relatively weak position of the parliament attributed by the Slovene constitution (the parliament cannot dismiss single ministers), this is a rather surprising finding. However, one can observe that, in the two countries (Hungary and Slovenia) where the parliament is relatively weak, the number of ministers who leave office because of struggles with groups involved in the democratic process is comparatively low. The constitutional strength of the Prime Minister, on the contrary, can be observed in the number of dismissals she has been responsible for. In Hungary, Latvia and, with some restrictions, Poland, where the Prime Minister has a very prominent position concerning the organization of the government, a number of changes have been implemented on her initiative. In addition, we will take a view at the relative numbers connected to the dismissal of ministers. The reference parameter is the number of ministers who leave office before the end of term. In subdividing the countries along the line of ministers having left office out of their own choice or not, three groups can be formed. In four countries, leaving office because of personal reasons or dissatisfaction is relatively common. Hungary comes first in this group: 34.6 per cent of the ministers who did not stay until the end of the government left out of their own decision. Slovenia follows with 31 per cent, then Poland (27 per cent) and finally the Czech Republic (24 per cent). The middle group is formed of Lithuania (20 per cent) and Estonia (16 per cent). In the other four countries, the dismissal of ministers because of struggles, scandals or reshuffles is far more common than resigning for personal reasons. In Bulgaria, only 10 per cent of the ministers leave without other groups being directly involved, in Slovakia it is 9 per cent, in Romania 8 per cent and the lowest numbers in this concern are to be found in Latvia (4 per cent). To conclude, one can say that by far the most frequent reason for ministers to leave office has been the end of the government they served in. Only few ministers make the active decision to leave office. In most cases, struggles with other groups, scandals or reshuffles lead to the dismissal of governmental personnel. In

Group

prime minister other ministers president parliament party/coalition public/press others several groups subtotal

0 3 0 0 0 7 1 0 11

0 0 2 2

government work overall situation position/success subtotal

6.5 80.5

10 124

0.0 1.9 0.0 0.0 0.0 4.5 0.6 0.0 7.1

0.0 0.0 1.3 1.3

0.0 0.6 0.0 0.6

7.8

12

0 1 0 1

66.2

102

conclusive returning immediately returning at later point subtotal

new job health private life subtotal

Bulgaria N %

Reason for leaving office

6 2 0 0 4 2 2 0 16

1 1 3 5

1 2 0 3

6 119

48

65

3.9 1.3 0.0 0.0 2.6 1.3 1.3 0.0 10.5

0.7 0.7 2.0 3.3

0.7 1.3 0.0 2.0

3.9 78.3

31.6

42.8

Czech Republic N %

3 0 1 0 7 0 0 0 11

3 1 0 4

1 0 0 1

12 131

54

65

1.9 0.0 0.6 0.0 4.3 0.0 0.0 0.0 6.8

1.9 0.6 0.0 2.5

0.6 0.0 0.0 0.6

7.4 80.9

33.3

40.1

Estonia N %

6 0 0 0 2 0 0 1 9

9 0 3 12

5 0 0 5

6 106

35

65

3.9 0.0 0.0 0.0 1.3 0.0 0.0 0.6 5.8

5.8 0.0 1.9 7.7

3.2 0.0 0.0 3.2

3.9 68.4

22.6

41.9

Hungary N %

7 0 0 0 6 0 0 0 13

0 0 1 1

0 0 0 0

17 195

96

82

3.2 0.0 0.0 0.0 2.7 0.0 0.0 0.0 5.9

0.0 0.0 0.5 0.5

0.0 0.0 0.0 0.0

7.7 88.6

43.6

37.3

Latvia N %

Table 12.6 Reasons for resignations in post-communist democracies, by country, 1990–2007

End of term

Personal reasons

Minister dissatisfied with

Others dissatisfied/ struggles with

4 0 0 0 3 0 0 2 9

1 0 1 2

1 0 3 4

9 144

62

73

2.2 0.0 0.0 0.0 1.7 0.0 0.0 1.1 5.0

0.6 0.0 0.6 1.1

0.6 0.0 1.7 2.2

5.0 80.4

34.6

40.8

Lithuania N %

11 2 3 2 7 0 0 2 27

3 0 8 11

3 2 2 7

20 291

161

110

3.0 0.6 0.8 0.6 1.9 0.0 0.0 0.6 7.5

0.8 0.0 2.2 3.0

0.8 0.6 0.6 1.9

5.5 80.6

44.6

30.5

Poland N %

3 0 0 1 4 6 0 2 16

1 1 0 2

2 1 0 3

11 192

117

64

47

69

1.2 0.0 0.0 0.4 1.6 2.3 0.0 0.8 6.2

0.4 0.4 0.0 0.8

0.8 0.4 0.0 1.2

2 1 0 0 5 0 0 2 10

1 0 1 2

0 0 0 0

1.2 0.6 0.0 0.0 3.1 0.0 0.0 1.2 6.2

0.6 0.0 0.6 1.2

0.0 0.0 0.0 0.0

13.0 84.6

29.0

42.6

Slovakia N %

4.3 21 74.4 137

45.3

24.8

Romania N %

2 0 0 4 0 0 0 1 7

1 0 1 2

3 0 0 3

13 143

69

61

1.2 0.0 0.0 2.5 0.0 0.0 0.0 0.6 4.3

0.6 0.0 0.6 1.2

1.9 0.0 0.0 1.9

8.0 88.3

42.6

37.7

Slovenia N %

Scandal

Cabinet reshuffle

Total

not known

0.0

154 100.0

0

1.3 8.4

2 13

3.9

3.2

6

political reasons organizational reasons temporary position subtotal

1.3 0.6 1.9

5

2 1 3

political personal subtotal

0.0

2.6 2.6

0.0

0.0

2.0 1.3 3.3

152 100.0

0

4 4

0

0

3 2 5

0.0

1.9 3.7

0.0

1.9

4.9 0.6 5.6

162 100.0

0

3 6

0

3

8 1 9

0.0

5.2 10.3

1.9

3.2

2.6 1.9 4.5

155 100.0

0

8 16

3

5

4 3 7

0.0

1.4 1.8

0.0

0.5

1.8 1.4 3.2

220 100.0

0

3 4

0

1

4 3 7

2.8

2.2 3.4

1.1

0.0

2.8 2.2 5.0

179 100.0

5

4 6

2

0

5 4 9

1.1

1.9 3.9

0.3

1.7

1.4 0.6 1.9

361 100.0

4

7 14

1

6

5 2 7

0.4

0.8 14.0

5.8

7.4

2.7 0.4 3.1

2

1 4

0

3

7 0 7

1.2

0.6 2.5

0.0

1.9

4.3 0.0 4.3

258 100.0 162 100.0

1

2 36

15

19

7 1 8

1.85

0.0 0.0

0.0

0.0

1.9 0.6 2.5

162 100.0

3

0 0

0

0

3 1 4

226 The selection of ministers in Europe the country-by-country comparison, we can observe a number of differences between the single countries. Some of these differences seem to be connected to constitutional rules. For example, in countries where the Prime Minister has a central position, she is relatively often the cause for the sacking of ministers. In countries where the parliament is, according to the constitution, comparatively weak we observe fewer struggles that lead to the dismissal of a minister. In some countries it seems to be much more common than in others to abandon a ministerial position for personal reasons. But in order to explain all the differences between the single countries, other variables such as the overall governmental stability, cultural factors, legacies or the impact of single individuals would have to be considered in a further analysis.

Notes 1

2

3

Note that Bulgaria also used a mixed system in the election of 1990, which, however, is considered the first genuinely post-transition parliament. Also, Estonia used the single transferable vote system to elect members of the Riigikogu in the election of 1992. We start counting in each state with the first founding government, the founding government for each country being defined as the first government formed after the first election after independence and/or the adoption of a new constitution. The Czech Republic and Slovakia are exceptions: in these cases the formation of a new state is taken as the starting point (for theoretical background see Kasapovic 1997: 50f.). The three parties together controlled 169 seats (70 per cent) in the 240-seat Bulgarian parliament. The distribution of seats among the three coalition partners was as follows: BSP – 82 seats (48 per cent of 169), NDSV – 53 seats (31 per cent of 167) and DPS – 34 seats (20 per cent of 169). Cabinet portfolios were allocated according to the following scheme: BSP – eight portfolios (50 per cent), NDSV – five portfolios (31 per cent) and DPS – three portfolios (19%) (Spirova 2006: 620).

References Anasolahebere, S., Snyder, J., Strauss, S. and Ting, M. (2005) ‘Voting Weights and Formateur Advantages in the Formation of Coalition Governments’. American Journal of Political Science 49, 3: 550–563. Baylis, T. (2007) ‘Embattled Executives: Prime Ministerial Weakness in East Central Europe’. Communist and Post-Communist Studies 40: 81–106. Downs, W. and Miller, R. (2006) ‘The 2004 Presidential and Parliamentary Elections in Romania’. Electoral Studies 25: 409–415. Fitzmaurice, J. (2003) ‘Parliamentary Elections in Lithuania, October 2000’. Electoral Studies 22, 161–165. Kasapovic, M. (1997) ‘Parliamentary Elections in Croatia: Electoral Models and Their Effects’. In: Siber, I. (ed.) The 1990 and 1992/93 Sabor Elections in Croatia. Berlin: Sigma Verlag. Krouwel, A. (2003) ‘Measuring Presidentialism and Parliamentarism: An Application to Central and Eastern European Countries’. Acta Politica 38: 333–362. McCarty, N. (2000) ‘Proposal Rights, Veto Rights, and Political Bargaining’. American Journal of Political Science 44, 3: 506–522. Morelli, M. (1999) ‘Demand Competition and Policy Compromise in Legislative Bargaining’. American Political Science Review 93, 4: 809–820.

Learning to rule: post-communist democracies 227 Müller-Rommel, F., Fettelschoss, K. and Harfst, P. (2004) ‘Party Government in Central Eastern European Democracies: A Data Collection (1990–2003)’. European Journal of Political Research 43: 869–894. Protsyk, O. (2005) ‘Prime Ministers’ Identity in Semi-Presidential Regimes: Constitutional Norms and Cabinet Formation Outcomes’. European Journal of Political Research 44: 721–748. Spirova, M. (2006) ‘The Parliamentary Elections in Bulgaria, June 2005’. Electoral Studies 25: 611–634. Vachudova, M. A. (2005) Europe Undivided: Democracy, Leverage and Integration after Communism. Oxford: Oxford University Press. Warwick, P. and Druckman, J. (2001) ‘Portfolio Salience and the Proportionality of Payoff in Coalition Government’. British Journal of Political Science 38: 627–649. Woldendorp, J., Keman, H. and Budge, I. (1998) ‘Party Government in 20 Democracies. An Update (1990–1995)’. European Journal of Political Research 33: 125–164. Woldendorp, J., Keman, H. and Budge, I. (2000) Party Governments in 48 Democracies (1945–1998). Dordrecht: Kluwer Verlag. Zubek, R. (2001) ‘A Core in Check: The Transformation of the Polish Core Executive’. Journal of European Public Policy 8, 6: 911–932.

Name index

The index does not include references to names appearing in tables and figures. Abril Martorell, Fernando 114, 116–18 Adenauer, Konrad 22, 26 Adonis, Andrew 59 Ahern, Bertie 185, 187, 190, 191 Andreotti , Giulio 92 Albert II, King of Belgium 125 D’Alema Massimo 86 Alfonso XIII, King of Spain 102 Ali, Syed 31 Alvarez Cascos, Francisco 115 Attlee, Clement 64, 67, 68, 71 Arenas, Javier 116 Arias Salgado, Rafael 110, 117 Atkins, Humphrey 71 Aznar, José María 101–2, 113–17 Baudouin, King of Belgium 125 Balke, Sigfried 27 Bérégovoy, Pierre 49, 53 Berlusconi, Silvio 86, 92, 96 Berov, Lyuben 219 Blair, Tony 59, 61, 62, 63, 64, 69, 72 Blaney, Neil 190 Blunkett, David 71 Boland, Kevin 190 Bondfield, Margaret 63 Boross, Imre 219 Bottomly, Virginia 63 Boyer, Miguel 116, 122 Brazauskas, Algirdis 219 Brown, Gordon 59, 61, 63, 72 Browne, Nöel 190 Bucher, Ewald 35 Burke, Ray 189, 190 Callaghan, Jim 64, 72 Calvo-Sotelo, Leopoldo 101–2, 114, 117, 119, 121

Cameron, David 62 Carrero Blanco, Luis 103 Carignon, Alain 53 Carlson, Ingvar 169, 174 Carrington, Lord 55, 69, 71 Chaves, Manuel 121 Chevènement, Jean-Pierre 12, 43, 53 Chirac, Jacques 43, 49–50, 51, 53 Churchill, Winston 17n10, 72 Claes, Willy 12 Cook, Robin 70, 71 Corleone, Don 8 Cosgrave, Liam 186 Costello, John A. 184 Cousins, Frank 59 Couve de Murville, Maurice 49 Cresson, Edith 46, 49 Crossman, Richard 64 Cunningham, Jack 64 Currie, Edwina 71 De Gaulle, Charles 43 Debré, Michel 43, 44, 46 Dehaence, Jean-Luc 14 Dehler, Thomas 22 De Riemacker-Legot, Marguerite 131 Dimitrov, Filip 219 Donegan, Paddy 189 Dooge, James 182 Douglas-Home, Alec 64 Dufoix, Georgina 53 Dugdale, Thomas 71 Eanes, Ramalho 148 Eichel, Hans 29 Eden, Anthony 17n10, 64 Ehmke, Horst 29 Erlander, Tage 169

Name index 229 Eyskens, Gaston 131 Elizabeth II, Queen of the United Kingdom 58 Erhard, Ludwig 27, 31, 35 Eppler, Erhard 35 Fabius, Laurent 53 Fälldin, Thorbjörn 161, 174 Fernández Ordóñez, Francisco 110, 119 Fischer, Joschka 29 FitzGerald, Garret 182, 186, 189 Fitzgerald, Gene 185 Frattini, Franco 12 Faure, Edgard 49 Fournier, Jacques 43 Funke, Karl-Heinze 27 Gabriel, Sigmar 29 Giscard d’Estaing 42, 49 González, Felipe 101–2, 104, 110, 113–14, 116–17, 119–20 Gordon Walker, Patrick 59 Gudmundsson, Magnus 198 Guterres, António 154 Hallgrimsson 198 Hammarskjöld, Dag 12 Hansson, Per-Albin 168, 174 Heath, Edward 59, 64, 72 Haughey, Charles 182, 183, 184, 185 Heinemann, Gustav 35 Hernández Mancha, Antonio 104 Hervé, Edmond 53 Heseltine, Michael 61, 70 Heuss, Theodor 22 Hinduja, Srichand 71 Huber, Antje 35 Jospin, Lionel 43, 52 Juan Carlos I, King of Spain 101, 114, 122 Juppe, Alain 49, 51, 53 Kinkel, Klaus 27 Klimmt, Renihard 29 Kohl, Helmut 24, 26, 31 Lafontaine, Oskar 35 Lamont, Norman 61, 71 Lavilla, Landelino 119 Lehr, Ursula 29 Lenihan, Brian 189 Léotard, François 43 Léopold I, King of Belgium 128 Leussnik, Hans 27, 29

Leutheusser-Schnarrenberger, Sabine 31, 35 von der Leyen, Ursula 29 Lilley, Peter 61 Longuet, Gérard 53 Lübke, Heinrich 22 Luce, Richard 71 Lücke, Paul 35 Lynch, Jack 183, 187 MacBride, Seán 184 McDaid, Jim 184 Macmillan, Harold 64, 67 Maizière, Thomas de 30 Mandelson, Peter, 71 Martín Villa, Rodolfo 117, 119 Martens, Wilfried 130, 131, 132, 134 Major, John 61, 64, 69, 70, 72 Maudling, Reginald 72 Mayor Oreja, Jaime 115 Mellor, David 71 Merkel, Angela 31 Mitterrand, François 43, 53 Möller, Alex 35 Morin, Hervé 47 Mulcachy, Richard 184 Müller, Werner 27, 29 Nolte, Claudia 31 Onkelinx, Laurette 131 Orban, Viktor 219 O’Malley, Des 184, 186 de Palacio, Loyola 116 Palme, Olof 169 Patten, Chris 61 Perez Rubalcaba, Alfredo 116 Persson, Göran 161, 168, 169 Piqué, Josep 115–16 Pompidou, Georges 43 Prescott, John 62, 72 Prodi, Romano 80, 86, 92, 93 Pym, Francis 69 Raffarin, Jean-Pierre 52 Reinfeldt, Fredrik 168 Rocard, Michel 43, 53 Rajoy, Mariano 114–16 Rato, Rodrigo 115–16 Roussin, Michel 53 Salisbury, Lord 70 Sampaio, Jorge 148 Sarkozy, Nicholas 42

230 Name index Schavan, Annette 29 Schlei, Marie 33 Schmidt, Helmut 5, 25–6, 29, 30, 31 Servan-Shreiber, Jean-Jacques 42 Solbes, Pedro 115 Solchaga, Carlos 116, 119–20, 122 Scaljoa, Claudio 93 Schiller, Karl 35 Scholz, Rupert 31 Schröder, Gerhard 22, 24, 29, 30, 31, 35 Schwarz-Schilling, Christian 35 Schwarzhaupt, Elizabeth 31 Shepher, Gillian 63 Short, Claire 70 Sid Cara, Nafissa 46 Siniscalco, Domenico 93 Smith, Jacqui 63 Spaak, Paul-Henri 12 Stegö Chilò, Cecilia 168 Strauss-Kahn, Dominique 53 Straw, Jack 63 Suchocka, Hanna 219

Sudreau, Pierre 53 Suares, Mário 148, 149 Suarez, Adolfo 101–2, 104, 113–14, 116–19, 121 Süssmuth, Rita 29 Tindemans, Leo 131 Thatcher, Margaret 62, 63, 64, 69, 71, 72 Tapie, Bernard 53 Tremonti, Giulic 92, 93 Thorneycroft, Lord 70 Trittin, Jürgen, 29 Vabres, Donnedieu de 53 Vagnorius, Gediminisu 219 Veltroni, Walter 86 Verhofstadt, Guy 137 Weigel, Theo 31 Westrick, Ludger 27 Wilkinson, Ellen 63 Wilson, Harold 17, 59, 64, 71, 72

Subject index

accountability 1–2, 139 Australia 9, 13 Belgium 6, 7, 8, 9, 12, 14, 17, 125–46 cabinet formation 127–9 constitutional framework 125–8 1831 constitution 125 1993 constitutional reform 125–8 dual mandate 126 factions 130–31 formateur 127, 128, 129 government size 129–30 informateur 127, 128 junior ministers 139 linguistic communities 126 linguistic parity 127 ministerial characteristics 130–36 ministerial de-selection 137–41 accountability 139, 41 death 137–8 mandate incompatibility 138–9 ministerial power 129 ministerial responsibility 129, 139–40 ministerial selection 130–6 gender 131 ministerial turnover 137–41 partitocracy 128 sleeping mandate 127 Bulgaria 6, 210–11, 212–27 constitution 210–11, 214 electoral system 204 government characteristics 212 ministerial de-selection 221–6 ministerial selection 215–17, 218–19, 219–20 gender 219–20 ministrables 217 parliamentary fragmentation 215 prime ministerial selection 213 sleeping mandate 217 vote of confidence 214, 217–18

choosing to resign 11 coalition theory 8, 26–7 constitutional constraints 6–8; 11–13 Czech Republic 16, 204, 209, 212–27 constitution 209, 214 electoral system 204 government characteristics 212 ministerial de-selection 221–6 ministerial selection 215–17, 218–19, 219–20 gender 219–20 ministrables 217 prime ministerial selection 213 vote of confidence 214, 217–18 Estonia 6, 7, 204, 206, 212–27 constitution 206, 214 electoral system 204 government characteristics 212 ministerial de-selection 221–6 ministerial selection 215–17, 218–19, 219–20 gender 219–20 ministrables 217 prime ministerial selection 213 sleeping mandate 217 vote of confidence 214, 217–18 France 7, 8, 14, 41–57 civil service 44 cohabitation 41–4 Constitution -Article 5 41 -Article 8 41 -Article 9 41 -Article 12 41 -Article 15 41 -Article 20 42 -Article 21 42 -Article 23 42, 46, 47 elections

232 Subject index parliamentary 1997 51 parliamentary 2002 52 presidential 2002 52 regional 2004 52 domaine réservé 43 Ecole Nationale d’Administration (ENA) 44, 46 énarques 46, 54 n7 Grandes Ecoles 46 informal rules 42–4 Institut d’Etudes Politiques de Paris (IEP) 46, 54 n8 ministerial career paths 47–51 ministerial de-selection 51–3 and elections 51–2 and reshuffle 51–2 scandal 53 ministerial selection 44–7 civil service 44 education 46 experience 47 gender 46–7 ministrables 44–7 Ministres d’Etat 48 Ministres délégués 48, 49 policy failure 53 portfolio experience 48–9 Secrétaires d’Etat 48, 49 Germany 7, 8, 9, 15, 17, 21–40 age at appointment 31 collective responsibility 23–4 Constitutional Framework 21–3 Article 4 Federal Minister Law 21 Article 20 III Constitution 21 Article 64 I Constitution 21 Article 65 Constitution 34 Article 66 Basic Law 21 rules governing investiture 22–3 rules governing ministrables 21 gender 31 Head of State 21, 22, 32 junior ministers 24, 25–6 ministerial de-selection 32–6 firing 32–5 individual resignations 34–5 protest resignations 35 ministerial selection 25–32 reshuffles 32–4 pre-ministerial career 28–30 Presidential veto 22 Weimar Republic 22 Head of State 7, 11,

hiring 5–11 Hungary 204, 205–6, 212–27 constitution 206, 214 electoral system 204 government characteristics 212 ministerial de-selection 215–17, 218–19, 219–20 gender 219–20 ministrables 217 prime ministerial selection 213 vote of confidence 214, 217–18 international organizations 12 Iceland 7, 12, 193–203 Althingi (parliament) 198 cabinet size 195 coalition theory 195–6 constitutional situation 194–6 formateur 194–5 ministerial de-selection 199–202 collective 199, 200 protest 199, 200 reshuffle 199, 200 ministerial selection 196–9 experience 197–8 gender 197 leader 196, 198 party 197 ministrables 196 non-resignations 199–201 resignation issues 200–1 Ireland 7, 13, 16–17, 179–93 cabinet reshuffles 188–9 cabinet size 181 constitutional framework 179–82 investiture 180–1 ministerial characteristics 185–8 age 186 education 186 electoral performance 196–7 experience 185–6 gender 188 geography 187 ministerial de-selection 188–91 disloyalty 190 party related 190 performance 190–1 policy 190 scandal 189–90 ministerial selection 183–8 party effects 183–5 politics of 183–5 ministrables 181–2, 185 junior ministers 188 Italy 6, 7, 17, 79–100

Subject index 233 Constitution Article 92 80 Article 94 80 formateur 80 government duration 79 informateur 80 junior ministers 81–2, 87 ministerial de-selection policy disagreement 90–2 reshuffles 90 role of parties 90–2 ministerial selection 82–90 experience 82, 86–9, 90 factions 82–3 gender 83, 86, 89 parliamentary background 82, 86, 88 territorial 82 ministrables 82–8 vice ministro 81 Latvia 7, 204, 206–7, 212–27 constitution 206–7, 214 electoral system 204 government characteristics 212 ministerial de-selection 221–6 ministerial selection 215–17, 218–19, 219–20 gender 219–20 ministrables 217 prime ministerial selection 213 vote of confidence 214, 217–18 linguistic parity rule 7 Lithuania 7, 204, 207, 212–27 constitution 207, 214 electoral system 204 government characteristics 212 ministerial de-selection 221–6 ministerial selection 215–17, 218–19, 219–20 gender 219–20 ministrables 217 prime ministerial selection 213 vote of confidence 214, 217–18 ministerial de-selection 11–15 party political constraints 8–10, 13 Poland 7, 204, 207–8, 212–27 constitution 207–8, 214 electoral system 204 government characteristics 212 ministerial de-selection 221–6 ministerial selection 215–17, 218–19, 219–20 gender 219–20

ministrables 217 prime ministerial selection 213 vote of confidence 214, 217–18 Portugal 6, 7, 147–58 cohabitation 149 Constitution 148–51 1976 provisions 148 1982 revisions 148–9 dual responsibility 148 junior ministers 152 ministerial de-selection 154–6 prime ministerial power 154–5 table of causes 156 ministerial selection 151–4 education 151 geographic 151 gender 151 occupational background 151, 153 sleeping mandate 150 Romania 204, 211–27 constitution 211–12, 214 electoral system 204 government characteristics 212 ministerial de-selection 221–6 ministerial selection 215–17, 218–19, 219–20 gender 219–20 ministrables 217 prime ministerial selection 213 vote of confidence 214, 217–18 SEDEPE xiii–xiv, 4, 16 sleeping mandate 6, 127, 217 Slovakia 6, 204, 210, 212–27 constitution 210, 214 electoral system 204 government characteristics 212 ministerial de-selection 221–6 ministerial selection 215–17, 218–19, 219–20 gender 219–20 ministrables 217 prime ministerial selection 213 sleeping mandate 217 vote of confidence 214, 217–18 Slovenia 204, 209–10, 212–27 constitution 209–10, 214 electoral system 204 government characteristics 212 prime ministerial selection 213 ministerial de-selection 221–6 ministerial selection 215–17, 218–19, 219–20 gender 219–20

234 Subject index ministrables 217 vote of confidence 214, 217–18 Spain 7, 8, 12, 101–24 Al-Qaeda 121 Bloque Nacionalista Galego (BNG) 113 cabinet composition 104–5 cabinet dynamics 110–12 cabinet formation 105 cabinet instability 110, 117–20 cabinet size 105–6 cabinet stability 110, 113–17 Coalición Canaria (CC) 113 Congreso de los Diputados 103 Convergència i Unió (CiU) 113, 119–20 constitutional framework 102–5 1978 Constitution Sect 97 103 1978 Constitution Sect 98 103 1978 Constitution Sect 99 103 1978 Constitution Sect 112 104 1978 Constitution Sect 113 104 1978 Constitution Sect 133 104 Citizen Security Act 120 Constitutional Court 120 constitutional mechanisms for withdrawing parliamentary confidence (confidence vote and motion of censure) 104 Council of Ministers 102–3 Government Act 103–5 Cortes Generales 103 Esquerra Republicana de Catalunya (ERC) 113 ETA (Euzkadi ta Askatasuna) 116–18, 121, 122 Filesa affair 120 Francoism 101, 103, 107, 114–15 GAL affair 120, 122 Guerra affair 119 Ibercorp affair 120, 122 ministerial characteristics 107–10 Partido Nacionalista Vasco (PNV) 113 Popular Alliance (Alianza Popular, AP) 104, 113, 115 Popular Party (Partido Popular, PP) 101–2, 110, 113–16, 120, 121 Prime Minister: constitutional powers 104 reforma pactada-ruptura pactada 101 requirements to be appointed minister 105 Roldán affair 120, 122 Second Republic (1931–1936) 101–2, 121

Senado 103, 105 Socialist Workers Party (Partido Socialista Obrero Espanol, PSOE) 101–2, 104, 107, 110, 113–15, 118–20, 121, 122 Union of the Democratic Centre (Unión de Centro Democratico, UCD) 102, 110, 114–15, 117–19, 121 vote for investiture 103–4 strategy 1–4; 10–11; 14–15, 61 Sweden 7, 159–78 cabinet dissolution 161; 168–74 cabinet reshuffle 168–74 constitutional framework 159–61, 175–6 ministerial de-selection 168–74 financial scandal 172 personal error 172 policy disagreement 172 scandal 168, 172 ministerial durability 170–1, 174–5 ministerial selection 159, 161–8 age 163–4 characteristics 159, 163–8 education 164 ethnicity 164 gender 163–4, 174 party effects 164 prime ministers’ role in 161–3 ministerial types 160–1 junior ministers 161 sleeping mandate 159–60 UK 6, 7, 8, 14, 15, 17, 58–78 cabinet size 62 choosing ministers 59–62 constitutional situation 58–9 Head of State 58–9 hereditary peer 58 House of Lords 58–9 investiture 49 length of ministerial term 68–9 ministerial de-selection 63, 69–75 departmental error 71, 72–4 financial scandal 71–4 length of government term 64–7 personal error 71–4 policy disagreement 70, 72–4 sexual scandal 71–4 ministerial non-resignation 72–5 ministerial responsibility 69–70 ministerial selection experience 62–3

Subject index 235 gender 63 ministrables 59 NEC (Labour Party) 8 reshuffles 62–9

size of parliament 58 types of ministers 60 vote of confidence 58 USA 6