The Politics of Recall Elections 9783030376093, 9783030376109

Citation preview

The Politics of Recall Elections Edited by Yanina Welp · Laurence Whitehead

The Politics of Recall Elections “The recall of elected officials is often neglected in democratic theory. Yet if the ideal of democracy is grounded in our political equality, and if it also applies to the relationship between ordinary citizens and elected officials, then the politics of recall is crucially important. In this volume, leading scholars in the field bring a variety of perspectives to bear on recall elections and provide a global, state-of-the-­ art overview. Highly recommended!” —Steffen Ganghof, Chair of Comparative Politics, University of Potsdam, Germany “This is the first contribution that seriously takes recall referenda under the radar of academic research. The book addresses the comparative description of the features of different institutional designs, causes and consequences of the global diffusion of recalls, the gap between theoretical aspirations and the practical use of recalls through case-studies. This is an indispensable contribution at times of a generalised citizen disenchantment with traditional models of electoral representation.” —Mariana Llanos, GIGA Institute of Latin American Studies, Germany “In critical phases of representative democracy, a significant device to overcome it and reconnect democratic institutions to citizens is the recall, a procedure to remove elected politicians from office. Thus, for anyone interested in our democratic future, a full awareness of how this device works in USA, France, Germany, Japan, Eastern European countries, Latin America, seems necessary. The authors provide that awareness with lucid, in-depth accounts that eventually give new hopes for democracy.” —Leonardo Morlino, Emeritus Professor of Political Science, Luiss G. Carli, Italy

Yanina Welp  •  Laurence Whitehead Editors

The Politics of Recall Elections

Editors Yanina Welp Albert Hirschman Centre on Democracy Graduate Institute Geneva, Switzerland

Laurence Whitehead Nuffield College Oxford, UK

ISBN 978-3-030-37609-3    ISBN 978-3-030-37610-9 (eBook) https://doi.org/10.1007/978-3-030-37610-9 © The Editor(s) (if applicable) and The Author(s), under exclusive licence to Springer Nature Switzerland AG 2020 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover illustration: © Alex Linch / shutterstock.com This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Foreword

The Politics of Recall Elections interrogates a relatively underexplored topic that cuts across a range of concerns for anyone interested in democratic politics and, no less importantly, in the very processes and problematics of governance. Recall is itself an irregular micro-procedure of removing elected politicians from office, inevitably overshadowed by the macro-­ status and impact of the major performative rituals discharged through general or local elections. But—as this sophisticated and innovative collection of essays demonstrates—recall is packed with resounding political significance, and it rightfully demands our scholarly attention. In a series of skilled and challenging analyses combining theoretical reflections and tellingly varied case studies, the editors of the volume and its contributors address central political themes: ephemerality and precarity, duration and curtailment, immediacy and impatience, accountability and power—all of which intersect with, and are fine-tuned by, the practice of recall. Democracy may be about the exercise of popular sovereignty, and for many it may even hold out the promise of good government, but it is also about containing the wilfulness of political leaders, the taming of electoral impulsiveness and the management of political time. Governments, seemingly, are there to govern, yet democracy as a set of politico-ethical principles prefers slow cooking to fast food, diligent deliberation to instant response. Recall is one major device aimed at mitigating the time lags imposed by a democratic constitutional order and the pressures such an order tends to store up. It responds dynamically to localized forms of serious discontent by engineering shorter-term adjustments—vital at periods of accelerated socio-political change or in dealing with corruption. v

vi 

FOREWORD

Concurrently, it may channel small-scale attempted power grabs by oppositional groups and endeavour to dislocate the allegiances of representatives. The normalization of disruption introduces an edge that serves democratic practice well—after all, democracy is crucially about the institutionalization of tolerable mistrust as a spur to governmental propriety and answerability. Yet enhancing unpredictability can chip away at the modicum of continuity and efficiency that justifiable authority requires. The comparative lenses employed throughout this book thoroughly and astutely set out the intricate complexity of recall arrangements and intentions. Their rich diversity reveals an intriguing bundle of measures suspended between a commitment to democratic reform, an ethos of grassroots political participation, a brutalism of power-wielding against political opponents, the contested imperative of external control over agents’ conduct and the temporal resilience that serves as a prerequisite for legitimacy. The great merit of this study is not only to shed light on the manner in which recall acts as a catalyst and conveyor of those themes but through them to provide a nuanced example of state-of-the-art contemporary political science. Mansfield College Oxford University Oxford, UK

Michael Freeden

Acknowledgements

This book began with a project focused on Latin America developed at the Center for Democracy Studies, University of Zurich. The exchange with scholars from different places of the world made it evident that recall was not only a device in development in the Americas. We thank our colleagues there for bringing information and cases to build a much bigger project such as the one we are presenting here. Nuffield College supported Laurence Whitehead allowing him to participate in a series of meetings. Conferences such as the RedGob in Rome in 2013 and the IPSA World Congress 2016 in Poznan gave the opportunity to scale up our approach. We thank our colleagues in all these events and in workshops and seminars where they provided useful comments which helped us to further improve this project. Yanina Welp thanks her kids Mora and Mateo, who are honestly neither patient nor supportive but they are anyway lovely, and with more merits on his own she thanks her husband for backing her always. Laurence Whitehead thanks his wife Linette, who adapts to his irregular Skyping habits and who allows him to carry on with ‘recall’ tasks when she would prefer to recall him to the dining table.

vii

Contents

1 The Politics of Recall Elections  1 Yanina Welp and Laurence Whitehead 2 Recall: Democratic Advance, Safety Valve or Risky Adventure?  9 Yanina Welp and Laurence Whitehead 3 The Political Theory of the Recall. A Study in the History of the Ideas 29 Matt Qvortrup 4 The Recall in France: A long standing and unresolved debate 49 Clara Egger and Raul Magni-Berton 5 Recall Elections in the US: Its Long Past and Uncertain Future 73 Joshua Spivak 6 Recall in Japan as a Measure of Vertical Accountability 95 Mitsuhiko Okamoto and Uwe Serdült

ix

x 

Contents

7 Explaining Institutional Change Towards Recall in Germany117 Brigitte Geißel and Stefan Jung 8 Recall Practices in Central and Eastern Europe: From Citizen Accountability to Partisan Account Settling143 Sergiu Mișcoiu 9 The Recall Revival and Its Mixed Implications for Democracy: Evidence from Latin America159 Yanina Welp and Laurence Whitehead 10 Reselection and Deselection in the Political Party179 Jonathan White and Lea Ypi 11 Narratives of Executive Downfall: Recall, Impeachment, or Coup?201 Aníbal Pérez-Liñán 12 On Reconciling Recall with Representation229 Laurence Whitehead Index249

Notes on Contributors

Clara  Egger is Assistant Professor of Globalisation Studies and Humanitarian Action at the University of Groningen, The Netherlands. She has held various teaching and research positions in France (Sciences Po Grenoble), Canada (University of Québec in Montréal), Germany FSU (Friedrich-Schiller-University) Jena and Switzerland (Geneva). Her research interests concern the comparative analysis of democratic institutions and their impacts on international cooperation in global public goods provision. She has started a project on the impact of mandatory referendums on the entry and exit strategies in international organizations. Her research received several awards and appeared in Peacebuilding, the Cambridge Review of International Affairs and Studies in Conflict & Terrorism. Her last book co-authored with Raul Magni-Berton is entitled Initiative and Referendum Explained: Inside Direct Democracy (2019, in French). Brigitte Geißel  is Professor of Political Science and Political Sociology and Head of Research Unit ‘Democratic Innovations’ at Goethe University Frankfurt, Germany. She was Speaker of the ECPR (European Consortium for Political Research) Standing Group ‘Democratic Innovations’ from 2010 to 2015 and held fellowships, research and teaching positions at several universities and institutes such as Harvard University (USA); Social Science Research Center Berlin (WZB); Center of Excellence on Democracy at Åbo Akademi (Finland); universities of Muenster, Berlin and Illinois (USA); and TU (Technical University) Darmstadt as well as the Vietnamese German University, Saigon (Vietnam). Her research interests include democratic innovations, new forms of governance and politixi

xii 

NOTES ON CONTRIBUTORS

cal actors (new social movements, associations, civil society, parties, political elites, citizens). Her articles have appeared in, for example, International Political Science Review, Comparative Sociology, West European Politics and European Journal of Political Research. Stefan  Jung  is doctoral candidate at the Faculty of Social Sciences at Goethe University Frankfurt, Germany. He studied at Goethe University Frankfurt, University of Wisconsin-Milwaukee (USA) and Universidad de Granada (Spain) and holds a master’s degree in Political Science from Goethe University Frankfurt. His research focuses on democratic innovations, direct democracy, local politics and social inequality in Germany and the USA. Raul  Magni-Berton is Professor of Political Science at Sciences Po Grenoble, France, and researcher at PACTE and at the Laboratory for Interdisciplinary Evaluation of Public Policies (LIEPP). He holds a PhD in Sociology from the University of Paris-Sorbonne. He has previously taught at the universities of Paris, Montréal, Bordeaux and Geneva. His publications are mostly based on international comparisons, individual surveys and national institutions. His articles have appeared in journals such as Political Research Quarterly, Political Studies, Public Choice, The Social Science Journal, International Review of Law and Economics, Social Justice Research, Current Sociology and other journals of political science, sociology and economics. He has also published several books. The latest is entitled Initiative and Referendum Explained: Inside Direct Democracy (with Clara Egger, 2019, in French). Sergiu Mișcoiu  is Professor of Political Science at the Faculty of European Studies, Babeş-Bolyai University in Cluj-Napoca, Romania, where he serves as Director of the Centre for International Cooperation. He holds a PhD in Political Science (University of Paris-Est), a PhD in History (Babeş-Bolyai University) and a habilitation in Political Science (University of Paris-Est). He is a member of the Laboratoire Interdisciplinaire d’Étude du Politique-Hannah Arendt (LIPHA) at the University of Paris-Est (France) and associate professor of the University of Szeged (Hungary). He wrote four books, edited and co-edited 20 volumes and wrote 50 articles, mainly in English, French and Romanian. His main research interests are the constructivist and the alternative theories applied to the nationbuilding processes, to populism and to the political transitional dynamics of the Central-­Eastern European, French and African public spaces.

  NOTES ON CONTRIBUTORS 

xiii

Mitsuhiko Okamoto  is professor and head of the Department of Political Science at Tokai University, Japan. He was a visiting researcher at the Center for Democracy Studies Aarau (ZDA) in Switzerland for the academic year 2014–2015. He holds a PhD in Political Science from Waseda University in Tokyo. His dissertation focused on the urban politics and local autonomy in Switzerland. Previously, he was also a visiting professor at the University of Vienna, Austria. His specialized field is public administration and local autonomy. He is interested in comparative direct democracy at the local level in Japan and other countries. Aníbal Pérez-Liñán  is Professor of Political Science and Global Affairs at the University of Notre Dame, France. His research focuses on political institutions, executive-legislative relations and the rule of law in new democracies. He is the author of Presidential Impeachment and the New Political Instability in Latin America (2007) and Democracies and Dictatorships in Latin America (with Scott Mainwaring, 2013). He is editor-­in-chief of the Latin American Research Review, the scholarly journal of the Latin American Studies Association (LASA), and co-editor with Paolo Carozza of the Kellogg Series on Democracy and Development. His research has been funded by the National Science Foundation, the United States Agency for International Development, the Inter-­ American Development Bank, Uruguay’s Agency for Innovation and Development and Paraguay’s National Council for Science and Technology, among other institutions. Matt Qvortrup  is Professor and Chair of Political Science at Coventry University, UK. He studied politics at Oxford University and earned his doctorate from Brasenose College in 1999 on a study entitled ‘Constitutional Implications of the Use of the Referendum’. He has previously taught at the London School of Economics and University College, London. As a practitioner, he was a special envoy for the US State Department in Cyprus (2004–2005) and in Sudan (2009) and a special advisor to the House of Commons Constitutional Affairs and Public Administration Committee (2015). He is author of a dozen books including Referendums and Ethnic Conflict (2014) and Government by Referendum (2018). Since 2016, he has been co-editor of The European Political Science Review. Uwe  Serdült  is working as a full professor since 2017 at Ritsumeikan University, Japan, in the College of Information Science and Engineering

xiv 

NOTES ON CONTRIBUTORS

while keeping some projects as principal investigator at the Center for Democracy Studies Aarau (ZDA) at University of Zurich, Switzerland. In this dual position, he is teaching and doing interdisciplinary research in several domains of participatory and digital democracy. Within digital democracy he is especially interested in internet-based platforms and tools for citizens (e-participation) as well as public administrations (e-government) in order to enhance transparency and deliberation in an information society. In the realms of direct democracy, he is focusing on the study of recall institutions and practice in polities as diverse such as Japan, Peru, Poland and Switzerland. He studied history, political science and computer science at the universities of Zurich and Geneva and received his doctoral degree in Political Science from University of Zurich, Switzerland. As a postdoc and lecturer, he used to teach and work at various universities in Switzerland (ETH Zurich, Uni Zurich, Uni Geneva). Research stays and guest lecturer positions lead him to Hungary (Andrassy University, 2016), Austria (Karl-Franzens-Universität Graz, 2015), Poland (PU Cracow, 2013/2014), Japan (Waseda University, Tokyo, 2002/2003) and the USA (University of Pittsburgh, PA, 1999/2000). Joshua  Spivak  is a senior fellow at the Hugh L. Carey Institute for Government Reform at Wagner College. He has written the Recall Elections Blog, which has examined and analysed recalls since 2011. He has written academic papers and opinion pieces for newspapers around the world on recalls as well as been interviewed and cited by newspapers and TV and radio stations. In addition to studying the recall, he writes on a host of political process and historical topics. He works as a public relations executive. Spivak received a JD from Columbia University Law School, a MA in history and a BA from Brooklyn College. Yanina Welp  is associate researcher at the Albert Hirschman Centre on Democracy, Graduate Institute, Geneva, Switzerland. She finished her habilitation (highest degree in the Swiss German academic system) in 2015 at University of St. Gallen, with the venia legendi in Latin American Studies, and she holds two bachelor degrees in Social Communication and Political Science from the University of Buenos Aires (Argentina) and a PhD in Political and Social Sciences from the Pompeu Fabra University (Spain). Her main areas of study are the introduction and practices of mechanisms of direct and participatory democracy and

  NOTES ON CONTRIBUTORS 

xv

digital media and politics, that is, ‘democratic innovations’. She has published extensively on these topics in academic journals and books. Jonathan  White is Professor of Politics at the London School of Economics and Political Science, UK. He has held visiting positions at the Berlin Institute of Advanced Studies, Harvard, Stanford, Humboldt University, Sciences Po in Paris and Australian National University. His books include Politics of Last Resort (2019), The Meaning of Partisanship (with Lea Ypi, 2016) and Political Allegiance After European Integration (2011). He was awarded the 2017 British Academy Brian Barry Prize for Excellence in Political Science and the 2015 Harrison Prize for the best paper published in Political Studies. Laurence  Whitehead is senior research fellow in Politics at Nuffield College, Oxford, UK, and until 2015 was senior fellow of the college. During 2005–2006 he served as Acting Warden there. In 2011–2012 he served as Senior Proctor of the University. Among his books are Latin America: A New Interpretation (2006 second revised updated edition 2010) and Democratization: Theory and Experience (2002). His most recent edited publications include Let the People Rule? Direct Democracy in the 21st Century (with Saskia Ruth and Yanina Welp, 2017), Illiberal Practices: Territorial Variance Within Large Federal Democracies (with Jacqueline Behrend, 2016) and Democratization in America: A Comparative-Historical Analysis (Desmond King, Robert C.  Lieberman and Gretchen Ritter, 2009). His most recent articles include ‘Recall of Elected Officeholders: The Growing Incidence of a Venerable, but Overlooked, Democratic Practice’ (Democratization, October 2018). He is editor of the book series Studies in Democratization, Chair of Research Committee 51 (International Political Economy) of the International Political Science Association, Chair of the Board of the Oxford Institute for Energy Studies and Honorary President for life of the Conseil Scientifique of the Institut des Ameriques. Lea  Ypi is Professor of Political Theory at the London School of Economics and Political Science (LSE), UK, and Adjunct Professor of Philosophy at Australian National University. A native of Albania, she has degrees in Philosophy and Literature from the Sapienza University of Rome, and she completed her PhD at the European University Institute. Prior to joining the LSE, she was a research fellow at Oxford University, the Italian Institute for Historical Studies and the Institute for Advanced

xvi 

NOTES ON CONTRIBUTORS

Studies in Berlin. She is the author of Global Justice and Avant-Garde Political Agency (2011) and, with Jonathan White, The Meaning of Partisanship (2016). She has co-edited Migration in Political Theory (with Sarah Fine, 2016) and Kant and Colonialism (with Katrin Flikschuh, 2015).

List of Figures

Fig. 4.1

Fig. 6.1

Fig. 9.1

Disproportionality under the French Fifth Republic. Disproportionality represents the ratio between the percentage of seats received by the majority coalition and the percentage of registered voters who opted for this coalition in the first round of legislative elections. When this ratio equals 1, these percentages are identical, meaning that the representation is genuinely proportional 59 Three-year moving averages of recall attempts, recall votes and successful recalls for mayors, assembly members as well as assemblies (1947–2014). Sources: MHA (n.d.), “Chiho-jichi Geppo” nos. 6/7, no. 9, no. 12, no. 23, no. 29, no. 37, no. 38, nos. 41–43, nos. 45–52. MIC (n.d.), “Chiho-jichi Geppo” nos. 53–59112 Legal provisions to activate recall regulated from the national and subnational level in Latin American (1923–2016). Source: Own elaboration. (1) In Argentina recall is regulated at the provincial level in Chaco (1957), Chubut (1994), Córdoba (1923, 1987), Corrientes (1960), La Rioja (1986), Rio Negro (1988) and Santiago del Estero and Tierra del Fuego (1991); other provinces include it for their municipalities, namely, Entre Ríos (1933, not included in later constitutions), Neuquén (1957), Misiones (1958), San Juan (1986) and San Luis (1987). It is also included in Ciudad de Buenos Aires (1996) (Arques 2014). (2) In Mexico it is regulated in Oaxaca (1998), Morelos (2011), Guerrero (2013), Zacatecas (2014), Aguascalientes (2014), Nuevo León (2016) and Mexico City

xvii

xviii 

List of Figures

(2017). Interestingly the two attempts registered in Yucatán and Chihuahua led to interventions of the National Judicial Court and the declaration of illegality of the recall, which was removed from the two constitutions but introduced in many other states (Limón 2016) Fig. 9.2 Municipalities with recall attempts in Colombia, 2017. Source: Welp and Milanese (2018: 1388) Fig. 11.1 Typology of mechanisms to remove the chief executive

163 170 205

List of Tables

Table 2.1 Table 5.1 Table 5.2 Table 5.3 Table 6.1 Table 6.2 Table 6.3 Table 7.1 Table 7.2 Table 7.3 Table 9.1 Table 9.2

Regulation of recall in a selection of cases 16 Recall election requirements in all US states 75 Total number of recalls that made the ballot or led to official resignation in the US, 2011–2018 80 US recalls by election date 84 Direct recall vote options and signature requirements in Japan (since 2012) 99 Direct demands on the local level, 1947–2018, attempts (A) and votes (V) 107 Reasons for the 106 recall votes from January 1989 to March 2018 (Heisei era) 109 Institutional rules for recall in Germany 118 Analytic framework for institutional change towards direct recall123 Factors for institutional change towards direct recall in three German Federal states 131 Recall regulation and practice in Latin America 166 Bolivia 2018: Number of recall requests by departments and over the total of municipalities 169

xix

CHAPTER 1

The Politics of Recall Elections Yanina Welp and Laurence Whitehead

Traditionally positions of political authority were subject to no fixed time limits. They might be accessed through inheritance, conquest or by virtue of some exceptional power or merit (prophets and sages). In all these cases, leadership tenure would extend indefinitely at the will of the incumbent, subject only to defeat by a rival challenger, permanent incapacity or death. In principle such rulers might voluntarily abdicate or nominate a successor, but these were risky and infrequent alternatives, since the survival of a former ruler would inevitably cast a shadow of the authority of the newcomer. Representative government provides the standard modern and time limited alternative to this situation. Officeholders gain access to defined parcels of authority for fixed periods of time, subject to procedures and constraints that are not (at least directly) under their control. Challengers with a wider range of attributes can put themselves forward under more predictable and safer conditions, and may try again at a later time if initially unsuccessful. The advantage for incumbents is that they exchange Y. Welp (*) Albert Hirschman Centre on Democracy, Graduate Institute, Geneva, Switzerland L. Whitehead Nuffield College, Oxford, UK e-mail: [email protected] © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_1

1

2 

Y. WELP AND L. WHITEHEAD

the uncertain benefits of life tenure for the greater security of a defined term followed by an orderly handover. The people they “represent” also gain the prospect of a peaceful and regular rotation of incumbency, with more scope for the recruitment of “new blood”, the correction of arbitrary policy decisions, and some ethic of accountability (under which rulers heed the concerns of their subjects, and may even start viewing them as citizens). As these principles of representation spread through the society, it becomes easier to extend the parcellization of political authority, both within the executive offices of the state, and through a separate legislature and local forms of government. In its full modern form this involves a constitution (rules specifying the relationships between the various parcels of authority) and a rule of law (specialized judicial arrangements for the impersonal arbitration of disagreements between these elected authorities). The idea of “recall” elections is not a last minute “add on” to these principles of representative government, but rather a logical strand of thought interwoven into this foundational reasoning. Once political authority is no longer grounded on divine right, conquest, hereditary entitlement or prophetic visions, it must find some more performative justification. Officeholders can serve to the extent that they are duly elected, and that they “represent” those over whom they exercise temporary authority. This provides some protection against the arbitrary excesses and insecurities of the traditional formulae, but these defences can only be partial. The very notion of a “time limit” discloses the inherent cross-­ pressures. Ex ante it cannot be known whether a short or a long term of office will be more performative. Too little time renders the officeholder impotent; too much risks despotism. The right balance will partly depend upon the challenges that arise during the incumbency, but will also be affected by the parcellization of power between the various positions to be filled, the rules concerning re-election, the scope of the constitutional provisions concerning accountability and so on. In view of these complexities (which help explain why despite centuries of experience no single template for best practice representation has ever won general assent), all systems of office-holding must contend with the hypothesis that under certain circumstances an elected representative might need to be recalled before time. The easiest cases to envisage are the most extreme—treason, insanity, blatant criminality and so on. But once the possibility of in extremis recall is acknowledged then the possibility arises that a broader set of more conventional political failings might be

1  THE POLITICS OF RECALL ELECTIONS 

3

added. At one end of the spectrum lies a view of political office that resembles the “bourgeois” model of private property—a set of entitlements, social obligations and revenue streams that should not be prematurely confiscated, except in cases of clear and serious legal violations and under due process of law. A clear example of this variant is the UK’s 2014 “Recall of MPs Act”, which confines the procedure to cases where an offence has been committed, and at least 10% of the electorate can start a petition for a recall. Constitutional provisions concerning the “impeachment” of sitting presidents mostly operate more or less along these lines (despite the inescapably “political” elements of judgement involved). At the other pole of this debate lies the “imperative mandate” argument (for the most part, although not exclusively, a “left-wing” or even “socialist” alternative to the bourgeois ideal.). This views the officeholder as agent of the electors, charged with transmitting their instructions to the assembly or governing authority to which they are delegated. According to this position, any failure to discharge the mandate as instructed would be cause for instant recall. (Again some “political” elements of judgement are inescapable here— mandated instructions cannot cover all details of a decision, unexpected contingencies may intervene, “instant” recall is an unattainable ideal). Some chapters included in this volume trace these lines of thought back through the long record of Western debates over political representation (see also Whitehead 2018), but our central concerns are to identify the current conditions that are encouraging widespread new experiments in “recall”, and to draw from both historical and contemporary evidence a clear balance sheet mapping the pitfalls—as well as the potential benefits— of recourse to this long subterranean strand in the representative government tradition of thought. This is the first volume dealing with recall referendums with a global approach (Zimmerman (1997) and Cronin (1989) published on the US experience). The cases have been selected to offer a general overview and to cover still understudied cases such as Japan or Moldova, Mexico and France. Latin American cases are better covered by previous works, and accordingly we opted for a regional approach in place of presenting case  studies. In the case of Switzerland the work of Serdült (2015) is up to date.1 1  Previous works include Geißel and Jung (2018), Qvortrup (2011), Welp (2016), Welp and Milanese (2018), Welp and Castellanos (2019), Whitehead (2018) and Mișcoiu (2019).

4 

Y. WELP AND L. WHITEHEAD

1.1   Structure of the Volume In a stage setting chapter, “Recall: Democratic Advance, Safety Valve, or Risky Adventure?” (Chap. 2), the editors propose that in principle the recall referendum is expected to provide a “safety valve” that could help to re-legitimize democratic politics in an era of dissatisfaction. Then, they explore the available empirical evidence to highlight substantial differences in institutional designs both between and within countries as well as evolving episodes of reform over time which affect the performance of the institution. In “The Political Theory of the Recall. A Study in the History of the Ideas” (Chap. 3), Matt Qvortrup traces the development of the recall in the history of political philosophy from the Roman Republic to present-­day republican thought. The chapter outlines how the institution was discussed by Karl Marx before it became an integral part of the political doctrines of respectively Vladimir Lenin and Antonio Gramsci, how the recall was championed by American Centre-left progressives, and how the institution was paradoxically utilized by Republican Arnold Schwarzenegger. “The Debate on the Recall in France: Imperative Mandate or Political Irresponsibility Under the Fifth Republic?” (Chap. 4) by Clara Egger and Raul Magni-Berton focuses on a case understudied to describe the French experience of recall mechanisms considering the historical diffusion of the idea of recall, its introduction during the French Revolution, as imperative mandate and the most recent developments, especially focusing on the demand which has emerged out of the yellow vests movement, including a discussion of the possible design a recall process could take in France. “Recall Elections in the US: Its Long Past and Uncertain Future” (Chap. 5), by Joshua Spivak, looks back at US history to show that the recall has been popular before, only to crash back to disuse. But right now, the recall is currently in a boom phase, as jurisdictions in the US have adopted the recall over the last decade and the recall may be growing in use throughout the country. The author analyses the reasons behind this grow as well as the practical hurdles that will prevent the recall from becoming a major player in US politics. The following chapter moves to the east. “Recall in Japan as a Measure of Vertical Accountability” (Chap. 6) by Mitsuhiko Okamoto and Uwe

1  THE POLITICS OF RECALL ELECTIONS 

5

Serdült showed that between 1947 and 2014 there were at least 1527 signature collections aiming to trigger a recall and 485 binding recall votes. The study shed some light on the less known history, practice, functions and consequences of recall referendum procedures in Japan offering a general overview followed by three case studies examined in more detail (Maki 1996, Akune 2010 and Nagoya 2011). Then, in “Explaining Institutional Change Towards Recall in Germany” (Chap. 7), Brigitte Geißel and Stefan Jung develop a comprehensive framework based on different institutionalist views (historical, sociological, rational choice) and insights from theories of democratic innovations to explain the adoption and diversity of local recall in Germany. “Recall Practices in Central and Eastern Europe: From Citizen Accountability to Partisan Account Settling” (Chap. 8) by Sergiu Mișcoiu explores the experience of the post-communist Eastern European countries with recall referendums by bringing evidence from several such procedures that took place in the region and by exploring in a more detailed way the 2017 recall referendum in Chisinau (Moldova). Latin America is probably the most prolific region regulating and activating recall, with Peru as the world’s most intensive user of recall referendums. “The Recall Revival and Its Mixed Implications for Democracy: Evidence from Latin America” (Chap. 9) by Yanina Welp and Laurence Whitehead focuses on the region to identify different intraregional paths and analyse under what conditions recall has an effect (positive or negative) on democracy. The closing chapters move on to include different settings and consider processes of deselection within political parties and the potential connections between impeachments and recall referendums. “Reselection and Deselection in the Political Party” (Chap. 10) by Jonathan White and Lea Ypi suggests that recall mechanisms in political parties can be used to reinforce their programmatic basis and to consolidate the reasons people have to associate with parties in the first place. They then defend a principled use of reselection and deselection mechanisms in political parties and answer four main criticisms that the view typically attracts: the ­constituency objection, the responsibility objection, the incentives objection and the efficiency objection. “Narratives of Executive Downfall: Recall, Impeachment, or Coup?” (Chap. 11) by Aníbal Pérez-Liñán compares historical examples from the US and Latin America to show that partisan considerations, rather than

6 

Y. WELP AND L. WHITEHEAD

prosecutorial zeal, have been the driving force behind historical episodes of impeachment. This reality creates a recurrent legitimacy crisis. Underpinning such legitimacy crisis lies a crucial question: Whether legislative majorities (or public opinion) at the time of impeachment should have greater authority over the president’s tenure than electoral majorities at the time of the election. The chapter discusses whether ratification of impeachments via popular referendum would offer a solution to this institutional conundrum. Finally, Laurence Whitehead stands back to conclude the volume by reintroducing the concept of “caesaristic democracy” as a still available alternative to representative and popularly accountable forms of government. If inherited models of electoral democracy no longer attract sufficient citizen allegiance, and are increasingly perceived as self-serving and dysfunctional, then popular support may migrate to more unconstrained, imposed and even arbitrary varieties of rule. The promise of “recall” and related forms of direct democracy is that they may reconcile dissatisfied citizens to reformed versions of representative politics. The potential pitfall is that such procedures carry the potential to clash with established institutional practices and to aggravate the collective sense of misgovernment by an out-of-control “political class”. The concluding chapter explores the scope for avoiding these risks by reconciling recall with representation under three headings: healing divisions, addressing “buyer’s remorse” and promoting “losers’ consent”.

References Cronin, T. (1989). Direct Democracy: The Politics of Initiative, Referendum and Recall. Cambridge and London: Harvard University Press. Geißel, B., & Jung, S. (2018). Recall in Germany: Explaining the Use of a Local Democratic Innovation. Democratization, 25(8), 1358–1378. Mișcoiu, S. (2019). ‘Never Just a Local War’: Explaining the Failure of a Mayor’s Recall Referendum. Contemporary Politics 25(1), 47–61. Qvortrup, M. (2011). Hasta La Vista: A Comparative Institutional Analysis of the Recall. Representation, 47(2), 161–170. Serdült, U. (2015). A Dormant Institution – History, Legal Norms and Practice of the Recall in Switzerland. Representation, Journal of Representative Democracy, 51(2), 161–172.

1  THE POLITICS OF RECALL ELECTIONS 

7

Welp, Y. (2016). Recall Referendums in Peruvian Municipalities: A Political Weapon for Bad Losers or an Instrument of Accountability? Democratization, 23(7), 1162–1179. Welp, Y., & Castellanos, A.  S. (2019). Understanding the Use of Recall Referendums: Evidence from Ecuador. International Political Science Review. https://doi.org/10.1177/0192512119830373. Welp, Y., & Milanese, J. P. (2018). Playing by the Rules of the Game: Partisan Use of Recall Referendums in Colombia. Democratization, 25(8), 1379–1396. Whitehead, L. (2018). The Recall of Elected Office Holders. The Growing Incidence of a Venerable, but Overlooked, Democratic Institution. Democratization, 25(8), 1341–1357. Zimmerman, J. F. (1997). The Recall: Tribunal of the People. New York: Praeger Publishers.

CHAPTER 2

Recall: Democratic Advance, Safety Valve or Risky Adventure? Yanina Welp and Laurence Whitehead

2.1   Introduction A key argument for the adoption of a recall process is that it can provide a safety valve allowing discontented voters to feel that they could in extremis remove a representative who had let them down and could therefore exert influence over an elected body even between regular elections by altering its composition (see Lijphart 1984). Such a procedural rule might counter citizen disillusion with representative democracy and help a disaffected electorate to remain basically loyal to democratic principles. The recall of California Governor Gray Davis in 2003 can be viewed as a good illustration of how this might come about after a century of low visibility use (Qvortrup 2011). But if that is the raison d’etre of existing recall mechanisms, then the fact that such rules and procedures are mostly so little known and so infrequently applied would indicate that recall is not much of a safety valve after all. Perhaps the designers of the system were ­providing Y. Welp (*) Albert Hirschman Centre on Democracy, Graduate Institute, Geneva, Switzerland L. Whitehead Nuffield College, Oxford, UK e-mail: [email protected] © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_2

9

10 

Y. WELP AND L. WHITEHEAD

for a contingency that they hoped would not arise, so once it was on the books, they were happy to have it fall into desuetude? This is one plausible reading of the Swiss or the Cuban cases where recall is regulated but almost never used (see Serdült 2015 and Guzmán 2014). But despite a few examples of recall as a ‘dormant institution’, most instances display a fair range of activations—or at least of attempts at activating recall. During the early twentieth century, such legal provisions were confined to the subnational level (e.g. Swiss cantons, US states, Argentinian provinces) and only used with frequency in some US municipalities. But during the 1990s (following the fall of the Berlin Wall), this little known institution began to spread in different regions of the world, such as Poland and Colombia (both in 1991), Peru (1993), British Columbia (1995), Venezuela (1999) and many German länder (see Serdült and Welp 2017). The recall cases which have attracted most media attention are the votes against governor of California Gray Davis in 2003, president of Venezuela Hugo Chávez in 2004, president of Bolivia Evo Morales and eight governors in 2008 and president of Romania Traian Băsescu in 2012. Recently there have been high profile recall challenges in big cities such as Duisburg (Germany) in 2011, Lima (Peru) and Warsaw (Poland) both in 2013 and Bogota (Colombia) (the vote was scheduled but cancelled when he was removed by the Procuraduría), Nagoya (Japan) in 2014 and Chisinau (Moldova) in 2017. As the most recent case, in Kaohsiung (Taiwan) a petition with 30,000 signatures to recall the mayor was presented to the central election committee on Dec 26th 2019.1 Peru has the record with an estimated 20,000 recall attempts (nearly all very local), but Poland, Colombia and Bolivia have also registered hundreds of attempts in recent decades. These cases highlight substantial differences in institutional design both between and within countries (e.g. contrasting formally unitary states like Japan, Colombia and Peru and federal systems like Germany and Argentina) and evolving episodes of reform over time which affect the performance of the institution (e.g. in Colombia, in Ecuador). Our goal here is to survey the spread and operation of recall referendums in contemporary representative democracies. In doing so we consider the following four dimensions: diffusion (how and why recall is 1  The number of signatures represents more than 1% of the city electorate of 2.28 million. Defenders of the mayor say the petition is invalid because the signatures were collected before the 12 month grace period was up. The law is not clear on this point, so the commission will have to make a ruling, and that may be challenged in a court.

2  RECALL: DEMOCRATIC ADVANCE, SAFETY VALVE OR RISKY ADVENTURE? 

11

spreading), institutional design (impact of designs on activations), outcomes (both short and longer term) either as safety valves or as new sources of contention and functions within the representative democratic model. This chapter is structured around these dimensions.

2.2   Diffusion: Why and How Recall Has Been Introduced? An early strand of thinking about political representation concerned the ‘imperative mandate’ according to which an elected delegate was expected to transmit the instructions of his electors without modification. The implication was clearly that a non-compliant representative would immediately lose his mandate and become subject to recall. This was usually just a tacit assumption. For example, the English Chartists envisioned annual elections, and given the slow communications of the 1840s, it was hardly thought necessary to prescribe faster recall. In the Paris Commune, the imperative mandate was combined with a recall provision. Since they were defeated there was no follow-through in either of these cases, but the arguments given then are not so different from the arguments in favour of recall given by contemporary movements, for example La France Insoumise during the 2017 campaign or Manuel López Obrador during the Mexican electoral campaign of 2018 (see also Chapter 4 on the current debate in France in this volume). From the very earliest days of discussion about electoral representation, the possibility of recall was among the most contentious themes of debate. During the English Revolution, the Leveller Lt Colonel John Lilburne championed the linked causes of delegation, tight control and recall as correctives to the risk of oppression by an overbearing Parliament. Lilburne was initially a hero of the Puritans for his defiance of the Star Chamber in 1638, but by 1645 he was expelled from Parliament and sent to Newgate Prison. His ideas lived on in the colonies through William Penn and Roger Williams. They also resurfaced among English Radicals at the time of the ‘Wilkes and Liberty’ agitation, for example, in the 1782 election for the constituency of Westminster (de Grazia 1951: 47) This was also a decade of intense debate about the principles of political representation in North America. “It seems an evident dictate of reason, that when a person authorises another to do a piece of business for him, he should retain the power to displace him”. Thus wrote ‘Brutus’, one of the opponents of the American Constitution, when he advocated the inclusion of the ‘recall’.

12 

Y. WELP AND L. WHITEHEAD

There was provision for the recall of congressmen by their voters was in the first draft of the American Constitution written by James Madison, though it was later dropped, owing to objections in the ratification debates from New York. The leading anti-federalist Melancton Smith argued that “senators would be rendered too independent” without recourse to the recall (quoted in Ketcham 2016: 377). But two different ideas need to be distinguished here. The recall of senators would have been by the state assemblies that selected them, in the same way that recall had been envisioned under the previous confederacy. New  York anti-federalist Smith wanted recall for senators, but not for directly elected congressmen, since state legislators would not “possess the qualities of a mob”, whereas he did not want members of the House to be subject to “the impulses of the multitude” (quoted in Klarman 2003: 377). In the event the Philadelphia Convention ruled against both versions of recall. Hence, Switzerland became the first modern liberal democracy introducing recall—although only at the subnational level, addressed to the whole council and not to single individuals, and in few cantons (i.e. Bern in 1846, Schaffhausen in 1876, Solothurn in 1869, Ticino in 1892, Thurgau in 1869 and later on in Uri in 1915; see Serdült 2015). Parallel to these Swiss provisions, although without documented connection, in the United States in 1892 the Socialist Labor Party and the Populist Party re-introduced the topic, this time only at the state and city levels. The recall was subsequently included in the new charter of the city of Los Angeles in 1903, but the author of this initiative (popularly known as ‘Recall Haynes’) seems to have been unaware of any of these precedents (see Spivak 2004). In the United States, the process of institutional diffusion started at the local level, and within a decade the Los Angeles example had been taken up by maybe 200 cities as well as 3 states (Cronin 1989). After a century it had spread to the 26 states which provide some legal means to activate it. So was this the product of independent invention by California progressives, or did it result from a process of international imitation, inspired from Europe, and in particular from Switzerland? According to the main promoter of recall in the United States, Dr J. R. Haynes, he got the idea by reading The City for the People by Parsons (1901). However, that book offers a lot of references to the Swiss example in general—although it does not mention recall in particular (Rappard 1912). The earlier US debates on this subject cannot have been entirely absent from the background thinking of the progressives, even if they were not explicitly present either.

2  RECALL: DEMOCRATIC ADVANCE, SAFETY VALVE OR RISKY ADVENTURE? 

13

In Japan, discussions among intellectuals about direct citizen participation can be traced back to the time before the establishment of the Meiji Constitution (see Okamoto and Serdült 2016). Already in 1902 an article by Shūsui Kō toku dwelt on the importance of the mechanisms of referendums and initiatives, while other publications analysed the Swiss case. The Japanese Social Democratic Party was a main defender of direct democratic institutions as can be seen in its electoral programme of 1901. However, shortly after the party was banned, and so it was not until after the Second World War that the topic returned to the sphere of public debate. Under the supervision of the US military occupation, the Japanese were required to democratize the country, and at that time direct democracy was included as a credible component of a more decentralized state (contrary to Japanese elite preferences for a more centralized model like the one implemented by the Empire). The Social Democrats became the strongest party in 1947 and then promoted the introduction of a series of direct democratic rules, including the right not only to dissolve the assembly (as in Switzerland) but also to remove assembly members and mayors (the three types of recall regulated in Japan). The cases of Germany and Poland are both related to the fall of the Berlin Wall. However, while in Germany recall has been only introduced regionally by some länder (acting under their decentralized powers), in Poland it was introduced through the Local Referendum Act of 1991 (Piasecki 2011). In Germany most länder adopted recall in the 1990s (see Geissel and Jung in this volume). All German federal states introduced direct election of mayors, citizen initiatives and referendums at the local level. In addition, 11 states complemented these instruments with provisions for indirect recall (i.e. activated by the council), and crucially, 4 of these also included direct recall (Brandenburg, Schleswig-Holstein, North Rhine-Westphalia and Saxony). In Latin America recent developments in the Andean countries provide a clearer pattern of regional diffusion. From the 1990s on, several republics replaced (or even ‘refounded’) their constitutions, and some introduced recall referendums together with various other mechanisms of direct and participatory democracy. This happened in Colombia in 1991, in Ecuador in 1998 and 2008 (two constitutions were approved in this period), in Peru in 1993, in Venezuela in 1999 and Bolivia in 2009 (Welp 2016, Welp and Whitehead in this volume). Do these radical examples mean that the recall referendum must be associated with the left of the ideological spectrum? The empirical evidence

14 

Y. WELP AND L. WHITEHEAD

we have suggests a negative answer. Latin American countries share various common features prompting these innovations, for example erosion of political party systems, disillusion with parliaments and political parties; but there were also important contrasts, for example, in Colombia the change was promoted under a liberal government, in Peru by an authoritarian regime (Alberto Fujimori), whereas in Bolivia, Ecuador and Venezuela it was led by new parties broadly following the lines of Venezuela’s Bolivarian Revolution (see Welp 2016). But the recall referendum certainly received considerable attention from prominent writers on the left in the early part of the twentieth century. It was championed by Gramsci, Lenin, Rosa Luxemburg—and the institution was even used in the early years of the Soviet Union (see White and Ronald 1996). In the same vein recall was regulated in Cuba in the Constitution of 1976, although it had been introduced earlier in the Constitution of 1940, but never used (Guzmán 2014). But, as already noted, it also has a long genealogy in Switzerland and the United States and has even been adopted—in an extremely conservative version—by the United Kingdom. It seems that there are several progenitors of the recall procedure, with quite contrasting intentions and expectations. Overall the speed of diffusion has been very slow and the coverage extremely patchy, at least until recently, when—even if still fitful and mostly ‘below the radar’—dissemination has accelerated and moved onto a new path. What might account for this shift? In a large and probably growing set of ‘really existing’ electoral democracies, citizen dissatisfaction with their elected representatives is sufficiently acute and widespread to generate persistent pressure for the introduction of more direct forms of accountability. This includes experimentation with various versions of the recall process. Some of these activities have involved the reactivation of long latent mechanisms; some have been spread through networks of similarly minded opinion formers; and some may be impromptu local reactions to specific pressures or scandals, uninformed by any broader comparative perspective. For example, the UK law was prompted by the ‘MPs expenses scandal’ of 2010. In general, recall practices appear to be spreading more widely, but without any high visibility coordination or direction. Any full account of their distribution would need to distinguish between the different levels and formats involved, and to explain not just where they are legal but when they are active.

2  RECALL: DEMOCRATIC ADVANCE, SAFETY VALVE OR RISKY ADVENTURE? 

15

2.3   Institutional Design and the Incidence of Recall Activation Recall is more frequent at the subnational level of government. The cases considered here confirm that the introduction of recall quite often accompanies reforms aimed at the direct election of mayors and the promotion of processes of political decentralization (see the cases of Japan, Germany, Central and Eastern European countries and Latin America in Chaps. 6, 7, 8 and 9 in this volume). In the German länder the introduction of recall went hand in hand with the establishment of constitutionally comprehensive options for self-­management, meaning that local decisions, such as the election or the recall of a mayor, do have important consequences for the further development of political life at the level of the municipality. Devolution is also part of the setting for recall in Poland and in the Andean countries of Latin America. This suggests recall may sometimes operate as a new form of vertical control (by the national government seeking leverage over local governments) rather than as a transfer of more political powers downwards into hands of regional and local governments. Focusing on direct recall (i.e. that activated by signature collection), we should note the multiplicity of institutional designs adopted. They diverge over which authorities can be removed by recall, for what reasons, the period during which activation is allowed, the number of signatures required, the time given to collect signatures and the actions taken if the authority is removed (see Table 2.1 as example of this diversity). All these considerations (in combination with other variables) affect which attempted recalls end in a vote. One should also add partisan motivation into the mix. That is why design matters, and in this business ‘the devil is in the detail.’ All these rules were decided by someone, so the interests and intentions of the designers need to be understood, even when the results prove far from different from their initial expectations. Since most experiments of this kind fall below the radar, and are little studied beyond an immediate circle of practitioners, there is little scope for incremental learning. Instead half-thought through provisions, trial and error and local improvisation seem to guide a great deal of the design rules adopted in most recall episodes. That has typically left limited scope for rationalistic calculus or the orderly diffusion of lessons from elsewhere. In some cases it is the national constitution that governs recall rules at the subnational level. This applies, for instance, to Cuba, Bolivia, Colombia, Peru, Poland, Ecuador, Venezuela and Japan. In Colombia

16 

Y. WELP AND L. WHITEHEAD

Table 2.1  Regulation of recall in a selection of cases Country/ unit

Reasons

Ecuador 1998

Corruption; programmatic vote –

Period of activation

Any time for corruption Ecuador After the 2008 first year, before the last Ecuador Programmatic After the since 2011 vote first year, before the last

Signatures

Time to collect signatures

If the authority is removed…

30%

–

Substitutes assume

10%

60–180 days

Substitutes assume

From 25% (5000 registered voters) to 10% in + 300,000 voters 25% but a maximum of 400,000

60–180 days

Substitutes assume

Unlimited

Peru (until – 2013)

After the first year, before the last

Venezuela

–

Middle of the term, before the last year

20%

3 days

California

–

Any time

12% voters on the last election

160 days

Minnesota Malfeasance/ Any time nonfeasance a ‘serious’ crime North – Any time Dakota

25% voters on the last election 25% electors

90 days

New elections if more than 1/3 is removed. If less, substitutes assume Depending on when it takes place, new elections or substitutes take the position Simultaneous election to remove or elect a new authority New elections

–

New elections

only local executive authorities can be removed by recall. In Peru both executive and legislative representatives (mayors and councillors) can be removed by a recall vote. By contrast, in federal democracies, the rules are determined at the subnational level, so that they may vary among units within the same country, as happens in Germany, Argentina, the United

2  RECALL: DEMOCRATIC ADVANCE, SAFETY VALVE OR RISKY ADVENTURE? 

17

States, Switzerland and Mexico where länder, provinces, states or cantons offer differing procedures for the recall of mayors and/or legislative authorities. The reasons required to activate a recall referendum also vary and reflect diverse models of representation. One possible motive is ‘dissatisfaction’, which is too vague to provide a serious basis for political accountability. An alternative concerns a programmatic vote, which implies a delegate model of representation. This contrast carries strong implications for the ensuing democratic performance. Whatever the stated grounds for initiating a recall, the procedure can be perverted if the agency in charge of evaluating the recall does not perform as a technical and therefore neutral arbiter on such questions. The most recent experience of Venezuela, where on 20 October 2016 the recall of President Nicolás Maduro was blocked by a co-opted National Electoral Council, provides a cautionary example. Temporal restrictions are normal to give an elected representative enough time to act in accordance with citizens’ expectations before facing a challenge. In the countries analysed here at least a year must elapse before a recall can be activated. This is the most common regulation although exceptionally (in some US states like California) a recall can be activated at any point during the mandate. A more critical consideration may be the number of signatures required. Geissel and Jung (2018) suggest that since higher signature requirements and shorter time limits increase the costs of a successfully initiated recall, such rules could be expected to cause less frequent recall challenges, while lower signature thresholds and more campaigning time will incentivize citizens to initiate more recalls. In their study of the German länder, the authors found that recall activations are most frequent when low institutional requirements and the relevant political (longer terms) and attitudinal conditions (higher dissatisfaction) come together. Brandenburg exemplifies this. It has by far the highest recall use among all the länder with a total of 44 recalls. Comparative study of the Andean countries also confirms the extent to which institutional designs do have an influence in the incidence of activations, although no linear effect could emerge (see Welp 2016). In Colombia, a new law approved in 2015 reduced the number of signatures required to activate a recall referendum (from 40 per cent to 30 per cent of the total of votes obtained by the elected authority) and the threshold for success (dropping from 50 per cent to 40 per cent of valid

18 

Y. WELP AND L. WHITEHEAD

votes originally cast) and also speeded up the registration of promoters. This led to an abrupt increase in the number of recall attempts which in only 5 months doubled the total recorded over the 15 previous years (Welp and Milanese 2018). Precise regulations (the targets at risk, the quorum for signatures, the grounds for recall, the methods of implementation, the consequences of an adverse outcome, etc.) make a great deal of difference, and such provisions are not necessarily stable or well understood. In addition, it is misleading to assess any recall mechanism in isolation from the broader repertoire of direct democracy ideas and options that give it meaning. For example, a practice may be advocated as a corrective to perceived failures of democratic accountability, but it could also be promoted by others aiming for a direct democracy gesture that they can curb or control.

2.4   Varied Outcomes When examining the range of outcomes that may follow from the introduction of recall, we need to start from what all variants have in common. All variants of recall involve potential curbs on the previous timescale of office-holding and an altered structure of career incentives and penalties. All options also reshape the machinery of accountability and provide new avenues of influence over the conduct of political incumbents. In constitutional theory, all this can be justified as a method for enhancing citizen involvement in politics and improving the responsiveness of politicians to their electors. But such theoretical claims should always be tested against the empirical record. Does recall necessarily empower the citizenry as a whole, or might it rather provide new channels through which powerful minorities may exert leverage over those in office? Does it enhance ‘convivencia’ and citizen confidence in the fidelity of their representatives or, to the contrary, might it add to divisions and reinforce voter disenchantment boosting distrust in the democratic process? Are public policies rendered more accountable and user-friendly through the availability of recall (an ‘error correcting’ mechanism), or might they be distorted, foreshortened or even rendered impractical by the uncertainties and instabilities inherent in this process (‘error-compounding’)? Alfred de Grazia arrived at the following assessment of the first half century of recall experiences in the United States:

2  RECALL: DEMOCRATIC ADVANCE, SAFETY VALVE OR RISKY ADVENTURE? 

19

The recall, therefore, though in principle a pure manifestation of the psychology of direct democracy, can hardly be said to have restored the direct communication between elected and electors so dear to the heart of direct democrats. It is possible, if the complaints of many politicians are even partially believed, that the recall was another step toward preventing the representative from acquiring an integrated moral in the performance of his specialized tasks. Again we have a case where increasingly complex politics is faced with a simple device aimed at complete control (de Grazia 1951: 155)

The more extensive comparative evidence now available indicates that the recall mechanism can serve a range of different political functions: direct or indirect accountability, institutional struggle or party competition (see Serdült and Welp 2017). It can operate as a mechanism of direct accountability, as a catharsis in the case of scandals or mismanagement by political elites. This assumes that voters are able to make judgements about representatives’ performances and remove them from office if they choose to do so. As an example of this consider Duisburg (a German city of 500,000 inhabitants). There the activation of recall was connected to the Love Parade 2010 which took place in an area surrounded by rail tracks that was only accessible via tunnels. Overcrowding and the following panic among visitors trying to leave the area led to the deadly catastrophe in which 21 people died and 500 were injured in a stampede. The mayor Adolf Sauerland from the Christian Democrats (CDU) was held responsible for having allowed the festival to take place in an unsuitable area. A first attempt to remove the mayor came in the form of a parliamentary motion to trigger the recall vote in 2010, but it did not receive the required two-thirds majority. Then, in May 2011 by a simple majority the länder Parliament changed the Local Government Act in order to make this recall possible. Thus North Rhine-Westphalia became the fourth land in Germany to introduce a citizen-initiated recall procedure. Almost immediately after the approval of the new Act, a committee called ‘New Start for Duisburg’ started to collect signatures in order to demand a recall referendum and soon collected signatures from 15 per cent of the electorate. So, on 12 February 2012 the mayor was recalled by 85.8 per cent of the votes (Serdült and Welp 2017). Recall can function also as means of indirect accountability. This happens when the process turns into a so-called second-order recall. In that format, the recall vote is actually directed against a specific policy rather than a named officeholder. The case of Japan shows a peak of recall activations linked to the emergence of city planning movements (machizukuri)

20 

Y. WELP AND L. WHITEHEAD

and environmental movements (see Okamoto and Serdült 2016). In the 1980s tensions between residents and regional governments increased mainly due to the construction of nuclear power plants, US military bases and also landfills. What these authors suggest is that it was the high hurdles Japanese citizens have to overcome in order to implement a local referendum ordinance allowing for a (non-binding) vote on such policy issues that drove them to move from indirect to direct procedures and therefore to challenge the parliament or the mayor through a recall vote. Although such indirect operations were easier to mount and could produce some striking immediate effects, they may not adequately address the longer-term policy objectives of their promoters. The introduction of mechanisms of direct democracy in Switzerland and the United States could be understood as a process of democratization in which recall was seen as an institutional option both to produce accountability and also to prevent revolts and revolutions by offering a ‘safety valve’ to defuse crisis and extreme polarization. When no institutional channel is available, informal demonstrations or even violence could offer the only outlet for protest. But when recall is available, it could also provide options of interest to diverse actors with a stake in the existing political system. Not all of these need be that keen on defusing protest, enhancing accountability or advancing democratization. After all, the process of recall can be designed so that it can be activated not only by disaffected voters but also by organized political parties. New or smaller parties might use the recall process to gain visibility, whereas larger parties might see a recall as an opportunity to challenge the party in power, particularly in the context of uncertainty regarding the survival of a given party. The chance to force a new election is not the only incentive available to opposition parties; rather, recall as a political weapon could also become part of a strategy to keep a party’s electorate mobilized or to compel internal discipline in what can be considered an extension of the ‘permanent campaign’ (Conaghan and de la Torre 2008). Such political actors might view recall as a new means to direct their fire against rival parties. Or a given party might even use this procedure to discipline dissidents from within its own ranks. Recall could also be a tool used by political parties or political leaders to unblock an institutional gridlock (legislature vs. executive or mayors vs. parliaments or even politicians vs. judicial forces). False trails, unintended consequences and perversion of the reforms are quite as common as successful experiences of democratic innovation.

2  RECALL: DEMOCRATIC ADVANCE, SAFETY VALVE OR RISKY ADVENTURE? 

21

Recalls are in any case to be expected both in the context of scandals and where there is intense political distrust and in the context of extreme political polarization. Even if they might serve as a safety valve in such conditions, their longer-term legacy could be to aggravate the instabilities they are supposed to rectify. And although institutional reformers may intend to use such devices to save the system of representation from collapse, they can also serve short-term partisan objectives which can deepen the electorate’s cynicism and distrust of all politicians. In the longer run the erosion of traditional political party systems and the spread of social digital networks set the scene for further risky experiments. Such factors seem likely to generate a growing number of recall activations in the near future, where the legal provisions for this exist, and to boost popular demand for including it where it is not available. But unless carefully designed, such innovations are as likely to backfire as to provide relegitimization or a safety valve. Much depends both on the details of the rules and on broader contextual factors which are also highly variable. Turnout, Scale and Strategic Adaptation When assessing the outcome of recall mechanisms, it is essential to specify whether the analysis concerns a ‘large N’ review of all (including the small and peripheral) cases, or whether it is the politically salient major units that are of interest. In the second instance, as in the first, although the available evidence remains incomplete, we now have enough material to affirm that the resulting patterns of political response are complex and diverse. There is no single dominant story to report. At the same time, there are some recurrent features. At least three variables can be identified that help explain the outcomes to be observed in the more salient minority of recall experiences: voter turnout, unit scale and the learning processes/adaptive practices of key actors. Once a recall vote has been approved, most countries have established turnout requirements or approval quorums to validate the verdict of the electorate. Turnout quorums require a certain proportion of citizens to turn out to vote, and the level of this threshold will influence expectations of the likelihood of achieving a recall. Colombia stands out as a case in point here. There were 169 attempts registered between 1991 and 2013 of which 54 ended in votes, but none of them was validated because the 40 per cent of threshold was never attained, either in minor contests or in

22 

Y. WELP AND L. WHITEHEAD

major cities. Both formal rules and informal conventions must be considered to explain this case. The abstention rate in Colombia is the highest in Latin America (contrary to most of the countries in the region, voting is not compulsory in Colombia). Abstention levels are especially high when it comes to local elections and recall referendums. In this context, the strategy of mayors facing recall is to do nothing and rely on the protection given by a low level of turnout (Welp and Milanese 2018). Another topic refers to adaptive strategy by political actors who anticipate the disruptions of the recall process and therefore act to head it off before their security of tenure is put at risk. For example, frequent use of other direct democratic instruments in Germany could make recall obsolete as conflicts can be solved by less drastic measures (see Geissel and Jung 2018). A related theme requiring further research arises from the fact that if either councillors or citizens succeed in initiating recall, mayors can avoid the final vote if they resign after the successful initiation of recall. Thus, the actual effect of recall might well be stronger than it appears in the numbers of recall votes.

2.5   Representative Democracy and Recall Among the factors determining how recall initiatives play out in practice, there is how campaigns are to be financed, what role the judiciary or the electoral authorities may play in either supporting or thwarting them and whether they can be adequately protected against partisan capture. In general the gains from grafting direct democracy innovations onto previously established systems of representation will depend on how carefully these two potentially conflicting sources of democratic legitimacy can be harmonized (see Whitehead 2017). There is some serious empirical support for the proposition that recall mechanisms, like other variants on the theme of direct democracy, can indeed provide genuine improvements to the quality and credibility of democratic institutions, when introduced and integrated into the rest of the representative system in a careful and constructive manner. For example, the 2003 replacement of Governor Davis by Governor Schwarzenegger in California proved to be a legitimate and popular decision that enhanced rather than destabilized democratic governance there. To some extent it can be understood as a ‘safety valve’, although this was not a moment of acute polarization or severe political crisis (Qvortrup 2011). A significant set of recall episodes have empowered voters to redirect public affairs in a

2  RECALL: DEMOCRATIC ADVANCE, SAFETY VALVE OR RISKY ADVENTURE? 

23

democratically desired direction, between regular election dates and without inflicting significant institutional damage. Such successful experiences help to explain the current appeal of the recall idea, in particular at a time when citizen distrust and dissatisfaction with their elected officials is on the rise, and when new more horizontal and instantaneous forms of political communications are undermining traditional party hierarchies. Thus recall can be a political advance. But the scope conditions for this seem quite restrictive. Much of the evidence reviewed here arises from other less favourable settings where the risks of a damaging outcome are more prevalent. Extreme examples are provided by Venezuela and Bolivia. In Venezuela, in August 2003, in a context of high polarization, the opposition collected 3.2 million signatures. However, the National Electoral Council (CNE) rejected it on the grounds that they had been collected prematurely (before the midpoint of the presidential term). In November, once the midterm was reached, the opposition collected 3.6 million signatures in a few days, taking advantage of massive demonstrations that made the collection of signatures an unusually easy task. The process of validation was long, highly controversial and characterized by irregularities. International organizations did play a role in forcing a dialogue between the government and the opposition. But one of the most flagrant irregularities was the online publication of the petitioner names, to put pressure on public servants (see McCoy 2006). Finally the vote took place, and Chávez was confirmed in office. The case of Bolivia shares with Venezuela the context of polarization. Note that the mechanisms had not yet been introduced (the constitutional convention was working between 2006 and 2009 when the new constitution finally entered into force). The recall activated in 2008 in Bolivia has been qualified as ‘vote of confidence’ given that the referendum was called by President Evo Morales as a strategy to bypass the deadlock between the President (controlling the executive and the lower house) and the opposition (controlling the senate and some important territories at the subnational level, especially Santa Cruz). Morales doubtless considered the Venezuelan precedent when opting for such a risky move. The threshold for success was set at above the percentage of voters originally voting for the incumbents, meaning that to recall Morales and Vice-President Álvaro García Linera, more than 53 per cent of the votes was required (i.e. a majority of 51 would have not sufficed). The same rule applied for the governors of eight departments (excluding Chuquisaca, where the

24 

Y. WELP AND L. WHITEHEAD

governor had only just been elected). The referendum was held on 10 August 2008 and confirmed Morales and seven governors in office (one was removed). These Latin American experiences suggest that populist leader may embrace mechanisms of direct democracy but then use it in a limited way. In a modern world, complex societies are less loyal than in the past (as Chávez experienced in 2004, although he managed to survive the recall referendum against himself). In 2016 Morales lost a referendum on standing for a fourth term, but he proposes to ignore that direct democracy verdict. Even where political polarization has become acute and policy crisis may be looming, our evidence still indicates that—under the right circumstances—dissatisfied voters may indeed use recall as a safety valve, an opportunity to clear out a few bad representatives, thus potentially enhancing the quality of representative democracy overall. The very notion of a ‘safety valve’ presumes the existence of uncontrolled pressures and risks of mishap, so it would be asking too much to require that such a mechanism must always produce a positive outcome. But this comparative survey also contains much evidence of recall processes that are relatively unlikely either to contain short-term tensions or to defuse longer-run threats to the credibility of the representative system. Instead the provisional findings so far include fairly frequent instances where the apparent outcome of recall has been to distract political leaders from the responsibilities of government, shortening their time horizons and intensifying partisan in-­ fighting. It seems that quite often recall experiments deliver more of a ‘risky wrench’ to a political system under stress than the relief expected from an escape valve.

2.6   For or Against Recall? The balance between such positive or negative outcomes depends partly on the underlying legitimacy and robustness of the system in question, partly on the precise design features of the recall mechanism itself and partly on particular context in which it was first adopted and then adapted. As regards pre-existing legitimacy, the crucial issue is whether the result of the recall is accepted by the losers, as well as the winners, and that depends not solely on whether the process is procedurally correct and generates a public benefit but also on whether the electorate is inclined to demand institutional compliance from all its political operators. Context matters when determining the effectiveness and acceptability of any given

2  RECALL: DEMOCRATIC ADVANCE, SAFETY VALVE OR RISKY ADVENTURE? 

25

procedural rule, since each society develops its own traditions and collective understandings about the rightness of claims to political authority. So the question is never just does recall work well, but also does it mesh with societal expectations and assumptions, how well is it understood and approved by our people. In the United States it is understood and accepted. In Venezuela it has toxic associations. The best organized actors will invariably try to game the new system before the general public have learnt how to make good use of it. But more systematic comparative study and greater public education into the merits of well-structured recall might shift the balance back towards a more positive pattern. One suggestion implicit in several of the case studies is that recall may work better when it is embedded in a broader process of public deliberation and policy debate, rather than simply operating as a device available to be used for the purpose of partisan aggression. On the other hand, another theme requiring more attention is the possibility that some (more radical) variants of recall may be so entangled with more far-­ reaching programmes of social transformation that they can be overused to undermine the status quo, but then get entirely set aside once the radicals are in the ascendant and might be vulnerable to the same sanctions that they had favoured before their ascent. Mexico under López Obrador could prove a test case of these two alternatives. Comparative experience can be useful as a source of guidance to minimize unintended or damaging effects, and to increase the chances of successful institutional design. In view of the broader forces at work challenging classical models of political representation (which include the instantaneous citizen responses facilitated by advances in information and communications technology and the reduced authority of national parliaments and traditional hierarchical political parties), it would be unrealistic to expect that the direct democracy surge can simply be blocked or disregarded. From a democratic development point of view, finding recall procedures that work well seems a better option than hoping that they disappear. Experience does indicate that under some circumstances they may indeed subside into latency. If they are on the books and not needed, that renders them harmless. But professional politicians risk a backlash if they are seen to be neutering such provisions in order to shore up their own privileges. There may well be a ‘ratchet’ effect, such that they are easier to introduce than to repeal. If so, some long historical sequences show that

26 

Y. WELP AND L. WHITEHEAD

even when recall mechanisms fall into disuse for decades, they still remain available for potential reactivation if triggering conditions reawaken them.

References Cronin, T. (1989). Direct Democracy: The Politics of Initiative, Referendum and Recall. Cambridge and London: Harvard University Press. de Grazia, A. (1951). Public and Republic: Political Representation in America. New York: Alfred A. Knopf. Geissel, B., & Jung, S. (2018). Recall in Germany: Explaining the Use of a Local Democratic Innovation. Democratization, 25(8). Guzmán Hernández, Y. (2014). Cuba: deudas pasadas y retos presentes desde la norma. In Y. Welp & U. Serdült (Eds.), La dosis hace el veneno: la revocatoria del mandato en Suiza, Estados Unidos y América Latina (Serie Ciencia y Democracia) (pp. 187–205). Quito: Consejo Nacional Electoral. Ketcham, R. (Ed.). (2016). The Anti-Federalist. New York: Signet Classics. Klarman, M. (2003). The Framers’ Coup. Oxford: Oxford University Press. Lijphart, A. (1984). Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries. New Haven: Yale University Press. McCoy, J. (2006). The 2004 Venezuelan Recall Referendum. Taiwan Journal of Democracy, 2(1), 61–80. Okamoto, M., & Serdült, U. (2016). Recall in Japan. Paper prepared for presentation at the 24 IPSA World Congress, July 22–28, Poznan, Poland. Piasecki, A. (2011). Twenty Years of Polish Direct Democracy at the Local Level. In T.  Schiller (Ed.), Local Direct Democracy in Europe (pp.  126–137). Germany: Springer. Qvortrup, M. (2011). Hasta la Vista: A Comparative Institutional Analysis of the Recall. Representation, 47(2), 161–170. Rappard, W. (1912). The Initiative, Referendum and Recall in Switzerland. Annals of the American Academy of Political and Social Science, 43, 110–145. Serdült, U. (2015). A Dormant Institution – History, Legal Norms and Practice of the Recall in Switzerland. Representation  – Journal of Representative Democracy, 51(2), 161–172. Serdült, U., & Welp, Y. (2017). The Levelling Up of a Political Institution. Perspectives on the Recall Referendum. In S. P. Ruth, Y. Welp, & L. Whitehead (Eds.), Let the People Rule? Direct Democracy in the Twenty-First Century. Colchester: ECPR Press. Spivak, J. (2004). California’s Recall: Adoption of the ‘Grand Bounce’ for Elected Officials. California History, 81(2), 20–63. Welp, Y. (2016). Recall Referendums in Peruvian Municipalities: A Political Weapon for Bad Losers or an Instrument of Accountability? Democratization, 23(7), 1162–1179.

2  RECALL: DEMOCRATIC ADVANCE, SAFETY VALVE OR RISKY ADVENTURE? 

27

Welp, Y., & Milanese, J. P. (2018). Playing by the Rules of the Game. The Partisan Use of Recall Referendums in Colombia. Democratization, 25(8), 1379–1396. White, S. L., & Hill, R. J. (1996). Russia, the Former Soviet Union and Eastern Europe: The Referendum as a Flexible Political Instrument. In P. V. Uleri & M. Gallagher (Eds.), The Referendum Experience in Europe. London: Macmillan. Whitehead, L. (2017). Between the ‘Fiction’ of Representation and the ‘Faction’ of Direct Democracy. In S.  Ruth, Y.  Welp, & L.  Whitehead (Eds.), Let the People Rule? Direct Democracy in the Twenty-First Century. Colchester: ECPR Press.

CHAPTER 3

The Political Theory of the Recall. A Study in the History of the Ideas Matt Qvortrup

3.1   Introduction In 2015, shortly before the election of that year, a majority in the British Parliament enacted the Recall of MPs Act 2015 (c. 25). The new law stipulated that henceforth voters would be able to recall their constituency Member of Parliament. Though, unlike in other countries (see Welp and Whitehead, in Chap. 2 in this volume), the Act does not allow constituents to begin proceedings unless the MP is found guilty of a wrongdoing that fulfils certain criteria. This was seen as a major shortcoming. Zac Goldsmith, a Conservative MP supporting the recall, believed the Act to be ineffective and too vague to have any effect (House of Commons Debates September 27, 2014, Coll.74). And, so it would seem. According to Section 1 of the Act, a Recall would only be triggered if the MP had received a custodial prison sentence of a year or less (longer sentences automatically disqualify MPs without

I am grateful to Laurence Whitehead and Yanina Welp for their helpful suggestions. The usual disclaimer applies.

M. Qvortrup (*) Centre for Trust, Peace and Social Relations, Coventry University, Coventry, UK © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_3

29

30 

M. QVORTRUP

need for a petition) or was suspended from the House of Commons for at least 10 sitting days or 14 calendar days, or if the MP was convicted for providing false or misleading expense claims. If these criteria were met, a recall election could be triggered by 10 per cent of the registered voters in the constituency. To the surprise of many, even the British Act proved effective. Three recall processes had been initiated. Ian Paisley Jr, a member of the Democratic Unionist Party, survived after he had been suspended from the Commons. Labour MP Fiona Oluyinka Onasanya was recalled after she had been found guilty of perverting the course of justice in 2019 (27 per cent had signed the petition—but Labour retained the seat). And, in the same year, Conservative MP Christopher Davies incurred the wrath of the voters after he was accused of forging invoices. Nineteen per cent of the voters signed a petition for a recall election. Unlike Ms Onasanya, Mr Davies fought to keep his seat. To no avail, he was recalled and the Liberal Democrat candidate Jane Dodds won with a majority of 1425 on a swing of 11.96 per cent. So, despite being a weak mechanism, the Recall of MPs Act 2015 has had some effect. But what does the history show? How has it worked before? And, indeed, what is the intellectual history of this mechanism? These questions will be sought and answered below.

3.2   The Early History of the Recall The Recall—here defined as the voters’ revocation of one or more elected representatives’ mandates before an ordinarily scheduled election—has a long history, although its practice has been patchy. The Romans in the first century BC, who had a more representative system of government than the Greeks in the third century BC, provided for an early form of the Recall, though this was “confined mainly to the last century of the Republic” (Abbott 1915: 89). Thus, the historian Plutarch (AD 46–120) recounts how the Tribune Octavius was removed from office because he had ill-advisedly vetoed a Senate Bill. According to a contemporary account, Tiberius, the other Tribune (they served co-­ equally), proposed that his colleague be recalled. Consequently, the latter “summon[ed] the people to pass a verdict upon him [Octavius]…averting that he would willingly relinquish his authority if the citizens wanted it” (Plutarch quoted in Zimmerman 1997: 7). We do not know the finer administrative or operational details of the enabling legislation—most Roman law was based on unwritten

3  THE POLITICAL THEORY OF THE RECALL. A STUDY IN THE HISTORY… 

31

c­ onvention—but we do know that the Recall was successful and “the law for his deprivation being thus voted Tiberius ordered one of his servants to remove Octavius”, and “this being done, the law…was ratified and confirmed” (Plutarch in Zimmerman, ibid.). When in the career of an elected politician might the possibility of recall (or demotion) arise? Already in antiquity elected officeholders were subject to the possibility of sanction before, during and after the completion of their mandates. For an example of before, in 62 BC Roman consul-elect Lucius Licinius Murena was prosecuted for electoral malpractice (bribery) by a defeated rival, but the author of the bribery law, former consul Marcus Tullius Cicero, secured his acquittal despite the offence, on the grounds of military necessity. As for after, a year later the Tribune of the People, P.  Clodius Pulcher, secured passage of a law that would have outlawed Cicero retrospectively for his conduct when consul, had the ex-consul not gone into exile.1 Consuls and Tribunes were the highest elected officeholders in the Roman Republic (with 12-month terms), but positions lower in the hierarchy were also vulnerable to sanction. For example, the presiding magistrate of a popular assembly was normally entitled to immunity for decisions taken during his mandate but would lose this sacrosanctitas if he failed to suspend its meeting in the face of violence. So in broad terms the practice of recall stretches back into the earliest days of electoral politics. From its outset the practice of a representative form of government was enmeshed with ideas about the accountability of term-limited officeholders—both to monitoring authorities and also more directly to ‘the people’ who had chosen them to serve as their temporary representatives. In various forms, and at different stages in the mandate of an elected official, such answerability remains an inherent feature of the role to this day. This legacy from the Roman Republic resurfaced in later European history. For example, the Polish kings after the last of the Jagiellons died could also be recalled. From 1573, under the Acta Henriciana and the Pacta Conventa, successive kings swore “not to intervene in the Seym’s election of the successor not to wage war or levy taxes without the consent of the Seym which he had to convene every two years…if he defaulted on 1  The law targeted Cicero in person. His political career was based on his success as prosecutor of individual officeholders for misconduct. He then wrote The Laws clarifying that whereas prosecutors must act on individual cases, all laws should be impersonal. That was his lasting contribution to our concept of the ‘rule of law’.

32 

M. QVORTRUP

any of this… he could forfeit his throne if he did not abide by the terms of his employment” (Zamoyski 1987: 98). At this time the Poles were following the Renaissance and imitating the style and ideals of the Roman Republic. This was not an idle talk. In 1831, the Seym dethroned Nicholas. Though this did not lead to government by the people, but rather direct rule from the Moscow overlords (Zamoyski 1987: 272). This did not change much before the eighteenth century. Overall the Recall—and discussions about this institution—was rare until the American Revolution. Yet, it is worth mentioning a number of prominent exceptions. Thus Marsilius of Padua (1275–1342) acknowledged the citizens’ right to “remove rulers from office who betrayed their trust” (Marsilius 1988: 45), and there are indications the Levellers in England in the seventeenth century espoused the device—some even believed members of the House of Commons should be subject to revocation (Foxley 2013). For example, in the Agreement of the People (1647), the Levellers mentioned the power of “removing and calling to account magistrates” (Foxley 2013: 179), though in the end these ideas came to naught. The same was true for the espousal of the Recall by Jean-Jacques Rousseau. In The Social Contract, the Swiss philosopher stressed that “the holders of executive office are not the people’s masters but its officers [and] the people can appoint them and dismiss them as it pleases—les diśtituer quant il lui plait” (Rousseau 1964a: 434). Unlike Montesquieu who held that “the people ought to have no share in government but for the choosing of representatives” (Montesquieu 1989: 153), Rousseau actively advocated the recall and criticised the English Parliament for its lack of accountability in Considerations on the Government of Poland— though he was rather evasive about the practical details (Rousseau 1964b: 978).

3.3   The Recall in the United States It has become well established that Rousseau’s writings were available to American readers at the time of the Revolution in 1776 (Spurlin 1969). Yet it is difficult to argue that the right of Recall was directly inspired by le philosophe. Rather, the American experiences with the Recall of representatives had very practical origins and do not seem to have been inspired or influenced by philosophical considerations. The first instance of its use can be traced back to the laws of the General Court of the Massachusetts Bay Colony of 1631, which held “the right

3  THE POLITICAL THEORY OF THE RECALL. A STUDY IN THE HISTORY… 

33

course to beholden when they, said commoners, shall see any defect of misbehave [is] to remove any of them” (Ordinance, cited in Luce 1931: 488). A similar institution was mentioned in the Massachusetts Charter from 1691, where “Councillors… may at any time be removed” (ibid.). These ideas found their way into the Articles of Confederation (1781–1789), which, in Art. V, provided for the Recall and replacement of delegates even within their “one-year term” (Cronin 1999: 129). The recall was included in James Madison’s first draft of the US Constitution. According to the so-called Virginia Plan, “members of the National Legislature [should be] subject to recall” (Madison 2005: 179). That this proposal was rejected at the Constitutional Convention in Philadelphia in 1787 was one of the main objections raised by the Anti-Federalists who criticised the elitism of the proposed constitution. ‘Brutus’, one of the most celebrated Anti-Federalists, was an example: “It seems an evident dictate of reason, that when a person authorises another to do a piece of business for him, he should retain the power to replace him” (Brutus in Ketcham 2003: 356). And in the heated debates in the New  York Ratifying Convention, Melancton Smith—believed to be Brutus’ alter ego—again defended the recall, noting that it would be used sparingly, “the power of the recall would not be exercised as often as it ought. It is highly improbable that a man, in whom the state has confided, and who has an established influence, will be recalled, unless his conduct has been notoriously wicked” (Melancton Smith in Ketcham 2003: 377). This was to no avail; the proposal for a recall was rejected as too radical. The Anti-Federalist’s argument was by no means revolutionary or unusual. For example, in the Declaration of the Rights of the Inhabitants of the Commonwealth or State of Pennsylvania, a section in the Constitution of Pennsylvania, September 28, 1776, provided: That those who are employed in the legislative and executive business of the State, may be restrained from oppression, the people have a right, at such periods as they may think proper, to reduce their public officers to a private station, and supply the vacancies by certain and regular elections. (Constitution of Pennsylvania 28th September 1776, Art. VI)

In the light of this it was hardly surprising that the Recall was discussed at the Constitutional Convention in Pennsylvania in 1787, though it failed

34 

M. QVORTRUP

to win majority support and support for it died out (Cronin 1999: 129). Discussions about it were not revived until after the American Civil War. In the 1880s, as a result of what was perceived corruption of the political system and the undue influence of the party machine, so-called populists championed the use of referendums, initiatives, direct elections of US Senators, primary elections and the Recall from the 1880s (see Welp and Whitehead in Chap. 2 in this volume). The movement in favour of these reforms had a distinct left-wing tenor. To wit, “Populist and Socialist Labor parties [in the United States] urged the adoption of the Recall in several of their national and state platforms in the 1890s” (Cronin 1999: 130). It is relatively well established that these ideas were inspired by the Swiss experiences. For example, the American Labour leader J.W. Sullivan’s Direct Legislation by the Citizenship Through the Initiative and Referendum was explicitly based on the author’s impressions from a trip to Switzerland and wrote enthusiastically about how “the people may recall their servants at brief intervals” (Sullivan 1892: 39). Sullivan—like other Americans advocating the use of the Recall—was almost exclusively basing his advocacy on the Swiss practice; they were not revolutionaries in the Marxian sense. Theirs was an argument for establishing more efficient checks and balances. Or, as William B. Munro, another American populist, suggested, “the chief argument in favour of the recall as advanced by friends of this expedient, is its efficacy as an agent of unremitting popular control over men in public office” (Munro 1912: 46). In a creative analogy which was later cited in court cases supportive of the Recall (see, e.g. Hilzinger v. Gillman 1909), Munro found that the Recall was an: Application, in the wider sense of the principle of ministerial responsibility, which is a feature of the English government, and which enables the course of public policy to be altered at any moment by the recall of the Cabinet at the hands of the House of Commons. (Munro 1912: 46)

The Recall was “a means of keeping officials responsible and responsive to public opinion” (Munro 1912: 47). Using another metaphor, Thomas A. Davies wrote of the Recall in terms of contract law, “if a man employs an agent for a term of years by contract, and the agent betrays his principal, the principal may terminate the contract and get rid of the faithless one” (Davies in Munro 1912: 314).

3  THE POLITICAL THEORY OF THE RECALL. A STUDY IN THE HISTORY… 

35

Or, in the words of Delos F. Wilcox, another activist and advocate, the representative was to be compared to an ambassador, “a diplomat”, he wrote, “is a servant with power but he has specific instructions or is presumed to be acquainted with his master’s will. If he fails to recognize his responsibility or if he misinterprets his instructions he may be recalled at any time” (Wilcox 1912: 171). Not surprisingly, the established politicians were severely critical of the recall. President William Taft was one of the strongest critics of the device. Indeed, the 27th President of the United States made a point of vetoing the proposed constitution for Arizona in 1911 (the year before the territory became the 48th state) because the document contained a provision for the recall (Congressional Record, August 15, 1911, 3964). The provision was removed from the draft and immediately instated once statehood had been granted (Zimmerman 1997: 16). It is a measure of the importance attached to—and the dangers associated with—the Recall that Taft continued his crusade against the device after he left the White House. In a series of lectures delivered at Yale University, the former president criticised the Recall, which—in his view— would create a “nervous condition of resolution as to whether he [the representative] should do what he thinks he ought to do in the interest of the public” (Taft quoted in Munro 1912: 83). Of course, this potential for keeping politicians in check, for preventing legislative activism, was the chief reason behind the populist espousal of the device, “the recall”, noted William A. White, the editor of a Kansas City newspaper, “should make …statesmen nervous” (White 1910: 60). And, of course, that is its very raison d’être. In America, the supporters of the Recall won the day, and it was implemented in many states at the instigation of politicians and advocates on the left. The same was not uniformly the case in Europe at the same time. In Europe too, it was the left side of politics who promoted the recall. But there was—as we shall see—an ideological gulf between this pragmatic and bourgeois conception of the Recall as it developed in America and the socialist and Marxist advocacy of the institution in Europe at the same time. In the old continent, the Recall was broadly associated with revolutionaries rather than with those advocating piecemeal reform. To be sure, several German cities and nine of the Länder adopted the recall during the years of the Weimar Republic often at the instigation of moderate socialist politicians, but few representatives were recalled, and the provision was of little practical importance (Wells 1929: 33).

36 

M. QVORTRUP

3.4   Marx and the Recall in Europe The Recall—as well as the referendum—was adopted in several American states in the first decades of the twentieth century. The recall was introduced in California in 1911. It was introduced in Arizona, Colorado, Nevada and Washington in 1912 (Cronin 1999: 132), but it remained an unimplemented goal and a cherished ideal in Western Europe. In Western Europe the device was associated with Marxist thinkers and can only be understood in the historical context of developments pertaining to revolutionary and communist movements of the latter part of the nineteenth century and the first decades of the next. Its origins in Marxist theory can be found in the political writings of Karl Marx (1818–1883) and were later developed into a more systematic theory by Gramsci and above all by Lenin who made the Recall into a central theme in his theory of the communist state. Marx developed his considerations about the revocability of the workers’ representatives in his account of the 1870–1871 Paris Commune. Following Napoléon III’s defeat by the Prussians at Sedan, French citizens stormed the National Assembly and declared a republic. In elections shortly afterwards, moderates led by Louis-Adolphe Thiers won a general election. Radicals in Paris did not accept this result. Thiers responded by sending in the army. The Parisians responded in kind and in addition elected their own assembly, which was dominated by radicals and even communists. One of the distinguishing features of the short life of the Commune (it was established in February 1871 and ended after the violent suppression by the Thiers government during the so-called La semaine sanglante (‘The Bloody Week’) beginning on May 21) was the provision for the Recall. This appealed to Karl Marx. He welcomed that the elected representatives were “at all times revocable” [jederzeit absetzbar] and noted approvingly, “all public servants, magistrates and judges were to be elective, responsive and revocable” (Marx 1973: 339–340). In his essay, he linked this system of accountability to his ideal of a decentralised political system. Combining factual empirical description, and emotive outbursts, with recommendations, Marx outlined how representatives were to be part of a decentralised structure under which “the rural communes of each district were to administer their own affairs by an assembly of delegates in the central town”, which in turn “were to send their delegates to the National Delegation” (Marx 1973: 340).

3  THE POLITICAL THEORY OF THE RECALL. A STUDY IN THE HISTORY… 

37

But, Marx evidently feared that some representatives could develop interests different from those of their constituents and hence stressed that each of the delegates should be “revocable and bound by the mandat impératif (formal instructions) of his constituents” (Marx 1973: 341). In his writings about the Commune, it was not entirely clear if Marx was merely commenting or whether he was recommending a system to be emulated by other revolutionaries. The historical evidence does not suggest the Paris Commune worked anything like as efficiently as seemingly sketched out by Marx. Rather his reflections were prescriptive rather than descriptive, something which was most explicit in a passage from The Civil War in France later quoted and used by Lenin (see below). In his reflection Marx stressed that under a communist system: The merely repressive organs of the old government were to be amputated, its legitimate functions were to be wrestled from an authority usurping pre-­ eminence over society itself and restored by responsible agents of the self-­ same society. (Marx 1973: 342)

However, with the provision for the Recall, participation would not be limited to “deciding once every three or six years which member of the ruling class was to represent the people in parliament” (Marx, ibid.). Rather “universal suffrage was to serve the people constituted in communes” (Marx, ibid.). While Marx was certainly committed to these principles, his writings on the subject of representative institutions never displayed anywhere near the theoretical rigour so characteristic of works like The Capital or Grundrisse. It was left to other Marxist writers to develop the argument into a coherent political theory.

3.5   Lenin and the Recall Marx’s lack of theoretical elaboration prompted Vladimir Ilyich Lenin (1870–1924) to develop—or, as he saw it himself, to reconstruct—a Marxist theory of the state based on Marx’s writings on the Paris Commune, “In 1916 and 1917 Lenin rescued from oblivion into which it had fallen Marx’s account of the [Paris] Commune as the administrative structure proper to Socialism” (Harding 1996: 282). The Soviets were based on immediate democracy “with delegates subject to recall” (Harding, ibid.). In Lenin’s own words, Marx’s writings were not underdeveloped

38 

M. QVORTRUP

but had been deliberately “forgotten and opportunistically distorted” (Lenin 1977: 6). For him, therefore, Marx’s theory of revocable representatives was a central part of the communist theory of the state. As in the case of Marx, Lenin’s thinking about the recall must be reconstructed and understood in its own unique political context. In Lenin’s case, it was imperative to develop an institution that made government responsive to the will of the proletariat. After the revolution, he wrote, “the organ of suppression is now the majority of the population, and not the minority” (Lenin 1977: 37). Lenin’s target in the State and Revolution was left-leaning Social Democrats like Karl Kautsky (1854–1938) who believed that socialism could be brought about under “fully democratic and parliamentary institutions” (McLellan 1991: 267). Lenin did not deny the need for representative systems, but the system should be under the control of the “vanguard of the proletariat, capable of assuming power and of leading the whole people to socialism” (Lenin 1977: 23–24). Writing shortly before his return to Russia as the leader of the Bolshevik Revolution in April 1917, Lenin was not writing about an abstract problem but about a concrete situation. He was adamant that he and the Bolsheviks had the “task of directing and organising the new order, of being the teacher, guide and leader of all the toiling and exploited in building up their social life” (Lenin 1977: 24). Yet, whether for opportunistic reasons or out of conviction—it is difficult to tell which—he still regarded representative institutions as indispensable. The representatives of the proletariat still had to be accountable. As he bluntly stated, “without representative institutions we cannot imagine democracy not even proletarian democracy” (Lenin 1977: 41). But this was a different sort of democracy from the bourgeois variety— one that left no room for checks and balances, “representative institutions remain but parliamentarianism, as a division of labour between the legislative and executive functions…no longer exists” (Lenin 1977: 41). Under the system championed by Lenin, there would be no “privileged positions for deputies”, rather it would be a system under which “all officials, without exception [are] elected and subject to Recall at all times” (Lenin 1977: 38). Such a system would be—what Lenin called—a ‘fuller democracy’. In a deliberate echoing of Marx, this system would be an improvement upon the bourgeois model “because all officials [would] be fully elective and subject to recall”, which was not true under the system of bourgeois

3  THE POLITICAL THEORY OF THE RECALL. A STUDY IN THE HISTORY… 

39

democracy, under which the voter was given the right to decide once every few year “which member of the ruling class is to represent and oppress the people through parliament” (Lenin 1977: 40). Thus with the addition of the Recall, the representative system’s wayward representatives could be held to account and “a way out of parliamentarianism could be found…not in the abolition of representative institutions…but in the conversion of the representative institutions from mere ‘talking shops’ to working bodies” (Lenin, ibid.), where the representatives through the Recall were “directly responsible to their electorate” (Lenin 1977: 41). That the Recall was a central plank perhaps even the lynchpin of Lenin’s theory of representation is also evidenced by a short piece he wrote weeks after the revolution. Drawing on the tenor of the argument in the State and Revolution, Lenin went on: Democratic representation exists and is accepted under all parliamentary systems, but this right of representation is curtailed by the fact that the people have the right to cast their votes once in every two years, and while it often turns out that their votes have installed those who oppress them, they are deprived of the democratic right to put a stop to that by removing these men. (Lenin 1917a: 338)

Sounding almost like a comparative political scientist, Lenin cited “cantons in Switzerland and some States of America” as places where “this democratic right of recall has survived”. But, perhaps more interestingly, Lenin also suggested that the recall could have made the revolution less violent and that the right of recall held out the prospect of a peaceful means of resolving political problems. Rather than “a rather stormy revolution”—his description of the storming of the Winter Palace—there could have been a peaceful change of power, “if we had had the right of recall [of the Provisional government headed by Alexander Kerensky] a simple vote would have sufficed” (Lenin 1917a: 339). When penning these words in December 1917, Lenin was faced with a particular problem that could be opportunistically resolved by the Recall. At the time there were two types of elected bodies: the local Soviets of soldier and peasant representatives (which were established after the abdication of the tsar in February of the same year) and the All-Party Constituent Assembly (which was elected as a multi-party representative institution). While the Recall already existed for the Soviets, “the truly

40 

M. QVORTRUP

popular nature” of which, according to Lenin, was, “evident by the fact that every peasant sends his representatives…and is also entitled to recall them” (Lenin 1917a: 339), this right of revocation did not exist for the Constituent Assembly. Lenin described this as a problem, though the reasons for his concern were seemingly practical and partisan rather than based on theoretical or idealistic concerns. The All-Party Constituent Assembly was elected through proportional representation (Douglas 1923: 129). This created a problem for Lenin who readily admitted that “under this system [Proportional Representation] is may be somewhat difficult to introduce the right of Recall”. Though, being a man of action, he also added that the “difficulties entailed are purely technical and are fairly easy to overcome” (Lenin 1917a: 339). Lenin went on to say that there is no contradiction between proportional representation and the “right to Recall” (ibid.). In this Lenin was fundamentally right. While political scientists have tended to the view that the recall “is incompatible with proportional and semi-proportional representation” (Lijphart 2012: 200), concrete case studies show that the right of Recall extended to whole councils in proportionally elected councils in Switzerland at the time—something Lenin as a long-time resident of the Alpine Republic conceivably could have known (Serdült and Welp 2017). But, once again, Lenin’s problem was not theoretical. His was a solution to specific problem and to a particular situation. The Bolsheviks had failed to win a majority of the seats in the All-Party Constituent Assembly— the Socialist Revolutionary Party (SR) won 45 per cent of the seats against a paltry 25 per cent for Lenin’s party (Dando 1966: 315). But one of the reasons for this was that out-of-date ballot papers had failed to take account of defectors from the SR to the Bolsheviks. As a result of this, some voters had voted for Bolshevik candidates but the votes had gone to parties from which they had defected (Fitzpatrick 2017). In the light of this Lenin demanded that new elections be held in areas where SR had split, as this was “the only chance of securing a…solution to the crisis that has arisen owing to the divergence between the elections to the Constituent Assembly, on the one hand, and the will of the people” (Lenin 1917b: 383). There should, more specifically, be a right of recall where voters were “misled by the parties” (Lenin 1917a: 339), and in the present circumstance in these constituencies, it was “for the people to elect its constituent members anew” (Lenin 1917b: 383).

3  THE POLITICAL THEORY OF THE RECALL. A STUDY IN THE HISTORY… 

41

Lenin’s solution to Recall under proportional systems was to hold new elections in which all the candidates in the district were up for new elections. Lenin set out his thoughts at some length at a meeting of the Central Executive Committee of Soviets of Workers’, Soldiers’, Peasants’ and Cossacks’ Deputies, second convocation, in Moscow in November 1917. The passage deserves to be quoted verbatim: Because the system of proportional representation is more democratic than the majority system, it demands more complex measures for the exercise of the right of Recall, that is, the actual subordination of the elected to the people. But it would be betraying democracy and abdicating the basic principles and tasks of the socialist revolution, which has begun in Russia, to refuse, on that ground, to practise the right of Recall, or to hamper or restrict its exercise in any way. What proportional representation demands is not a curtailment of the right of recall but a mere change of form. Since the system of proportional representation is based on acceptance of the party system and the conduct of elections by organised parties, any major change in the balance of class forces and the relation of classes to parties, especially in the event of splits within major parties, necessarily produces the need for a re-election in any electoral district where there is a clear and obvious discrepancy between the will of the various classes and their strength, on the one hand, and the party composition of those elected, on the other. True democracy makes it imperative that the appointment of re-elections should not depend only on the institution to be re-elected, that is, that the desire on the part of those elected to retain their seats should not work against the exercise of the people’s will to Recall its representatives. (Lenin 1918: 336)

Some Recall elections were held to the All-Party Constituent Assembly. And relatively prominent bourgeois representatives were recalled, including the Liberal Pavel Miliukov (Lenin 1918: 337). However, the Bolsheviks did not secure a majority in by-elections held in December 1917. Having failed to overturn the result through the ballot box, the Bolsheviks closed down the All-Party Constituent Assembly after one day in January 1918, and henceforth all parties other than the Communist Party were banned (Figes 1996: 518). Power reverted back to the Soviets and especially their representatives in the ‘Vanguard of the Party’. Yet—perhaps surprisingly—the nominal right of Recall was kept throughout the Soviet Union’s 75 years of history, though it lay dormant until Gorbachev instituted his policy of Glasnost in the 1980s. Having been used in the early years of the Soviet Union, the

42 

M. QVORTRUP

Recall was only used “to recall an elected deputy in Sverdlovsk” in the dying years of the USSR (White and Hill 1996: 158).

3.6   Gramsci, Luxemburg and Other Socialist Advocates of the Recall Karl Marx and Friedrich Engels had asserted in The Communist Manifesto, “the first step in the revolution by the working class is to raise the proletariat to the position of ruling class, to win the battle of democracy” (Marx and Engels 2005: 470). But it was never entirely clear how this was to be achieved. While communists declared their commitment to democracy, most theoreticians were critical of direct democracy in the form of the referendum. Thus, the German Social Democrats’ proposal for “direct legislation by the people” was dismissed by Karl Marx in the Critique of the Gotha Programme as an institution that was “exaggerated in fantastic imagination” (Marx 1970: 29). Karl Kautsky was similarly critical of referendums, maintaining that it would drive a “wedge between the party and the people” (Kautsky 1911: 133). The same was true for Arturo Labriola (1843–1904), the foremost Italian communist theoretician who wrote a book entitled Contro il referendum (Labriola 1897). In the light of this, it was hardly surprising that Antonio Gramsci (1891–1937)—the founder of the Italian Communist Party—also dismissed the referendum. While he was willing to concede that “a referendum [could] be called in exceptional circumstances” (Gramsci 1978: 50), he was adamant that “Communists are on principle opposed to the referendum since they place the most advanced and active workers on the same plane as the lazy, ignorant and idle workers” (Gramsci, ibid.). But this did not imply that Gramsci—let alone Marx—was critical of popular involvement in politics. Like Marx and Lenin, Gramsci shared the view that legitimacy was needed even, or, perhaps, especially, under a system of democratic centralism. The workers’ representatives could err, and under these circumstances the Recall could be used as a remedial mechanism. Hence, in a socialist or communist system, “the delegate is, or should be elected …imperatively mandated and instantly recallable”, and, since “the mandate is imperative and revocable it can also be assumed that the Delegates’ Assembly represents the workers at all times” (Gramsci 1978: 50), wrote Gramsci.

3  THE POLITICAL THEORY OF THE RECALL. A STUDY IN THE HISTORY… 

43

However, his ideas came to naught after Mussolini’s takeover in 1922 and Gramsci’s subsequent imprisonment. Gramsci and Lenin both espoused democratic centralism in varying degrees. However, not all Marxist theoreticians advocated centralism along the lines advocated by the Bolsheviks. One notable exception was Rosa Luxemburg (1871–1919). However, despite her disagreements with the leaders of the Russian Revolution, she too championed the recall. While references to the recall are absent from Luxemburg’s more theoretical writings such as The Accumulation of Capital or even in The Crisis in the German Social Democracy, which is more commonly known as The Junius Pamphlet (Weitz 1994: 27), she made references to the device in her more activist writings such as, above all, the 1918 pamphlet What Does the Spartacus League Want?. In this short piece, she stressed the necessity of the “election of all officers by their units, with right of immediate recall at any time” (Luxemburg 1984a: 354). In the light of Luxemburg’s general writings, and in particular her opposition to the centralism of Lenin, there was a certain logic to her stance. Early in her career, Rosa Luxemburg had advocated a stronger role of the people and had criticised Lenin’s “line of thought [for being] concerned principally with the control of party activity and not with its fertilization, with narrowing and not with broadening, with tying the movement up and not with drawing it together” (Luxemburg 1984b: 256). In her view it was necessary to find mechanisms for rectifying the inevitable mistakes made by the workers’ representatives. At that stage, however, her views on socialist democracy had yet to develop. Her commitment to popular involvement grew stronger in the wake of the Russian Revolution. She once again stressed, “The whole mass of the people must take part of it [the development of a socialist society, MQ]. Otherwise, socialism will be decreed from behind a few official desks, by a dozen intellectuals. Public control is indispensable necessary. Otherwise the exchange of experiences remains only with the closed circles of the officials of the regime. Corruption becomes inevitable”, she wrote in her manuscript on The Russian Revolution (Luxemburg 1984c: 306). In order to provide this control of the officials, it was necessary that all positions were subject to revocation; hence there should be a “right of immediate recall by the local workers’ and soldiers’ councils and replacement of their representatives in the central council, should these not act in the interests of their constituents” (Luxemburg 1984a: 354). Due to her

44 

M. QVORTRUP

premature death in 1919, Rosa Luxemburg’s writings on the recall were tragically interrupted at the time when she began to develop them. This does not mean that the recall disappeared from theoretical discussions on among socialists and Marxists. While it is difficult to draw a connecting line between Lenin and Gramsci and the democratic socialists in Britain, it is noteworthy that Harold Laski (1893–1950)—the far-left political science professor who served as chairman of the Labour Party— wrote favourably about the recall in his chief work A Grammar of Politics. In this he argued that this institution “would be a valuable addition to our elected machinery”, though he was at pains to stress that it “ought not clearly to be a weapon of easy use”. Yet, this caution notwithstanding he felt that “proper safe-guards can be had, and they enable the electorate to have a closer check both upon the member and his party than the present system admits” (Laski 1938: 320). But, like in the other cases, very little became of the recall in practical politics, and other socialist theoreticians in England did not discuss it at the time or indeed later. After this Marxist theoreticians became more wary of the recall. Insofar as it was debated at all, Western Marxist theoreticians agreed with Nicos Poulantzas’ view that while there should be direct, universal and secret voting, “there should not be imperative mandates” [ne soient pas régies uniquement par le mandat imperative] “the recall at any time” [la révocabilité à tout instant] (Poulantzas 1985: 279).

3.7   Conclusion “The rich may be made poorer, and the poor richer. In other words… socialistic” (Taft 1913: 90). This is how President Taft denounced the recall shortly after stepping down as the 27th President of the United States. It is perhaps ironic that the political institution that brought Republican Governor Arnold Schwarzenegger to the Governor’s Mansion in Sacramento in 2003 was denounced as a threat to “the right of property”. But seen in the perspective of the history of the ideas, the Recall has consistently been championed by theoreticians of the left. While traces of the recall can be found in Ancient Rome and among the Levellers in seventeenth-­century England, the institution only gained practical prominence in the wake of the American Revolution. Inspired by earlier constitutions, the Articles of Confederation contained provisions for the Recall.

3  THE POLITICAL THEORY OF THE RECALL. A STUDY IN THE HISTORY… 

45

Although the Recall was discussed at the Constitutional Convention in Philadelphia in 1787, it did not become part of the US Constitution. A hundred years later, the recall re-emerged in two different contexts: in America by left-leaning farmers and organised labour and in Europe by prominent communist writers, including Karl Marx, Rosa Luxemburg and Antonio Gramsci, but especially by Vladimir I. Lenin. Inspired by experiments with the recall in Switzerland, so-called populists in the United States championed the recall as a check upon the party machine and in order “to break the power of a politically entrenched plutocracy” (Weyl 1919: 298). These groups succeeded in winning the right of recall in ten US states by 1914. In Europe, Karl Marx first championed the recall in his account of the Paris Commune in 1871. Based—and elaborating—on this, Vladimir Lenin developed a case for the recall in his State and Revolution and went, opportunistically, on to further elaborate on this in occasional writings following the October Revolution in 1917. Characteristically, Lenin elaborated on this theory to make a case for the recall under proportional representation after the Bolsheviks failed to win a majority of the votes in the elections to the Constituent Assembly. Following the Russian Revolution, the Recall remained a nominal part of the Soviet political system, though it remained unused until the 1980s. Elsewhere in Europe other prominent socialist writers such as Rosa Luxemburg in Germany, Harold Laski in England and Antonio Gramsci in Italy wrote approvingly about the Recall and linked this institution to their respective conceptions of socialist democracy. However, by the 1970s support for the recall had waned, and Marxist critics like Nicos Poulantzas rejected the need for the recall—notwithstanding Marx’s support for this institution. A concept in the history of political ideas can be compared to a thread. There is no fibre running through the entire string, only different fibres which together constitute it. While writers from Plutarch through the American populists, from the Levellers to Lenin and Rosa Luxemburg and from Marsilius of Padua to Marx have provided different fibres, the essence of the recall has remained the same—to provide a mechanism for holding elected representatives accountable. Benjamin Constant, the French liberal writer, concluded in his famous essay ‘The Liberty of Ancients Compared with that of Moderns’ that “the people who, in order to employ the liberty that suits them, resort to the representative system, must exercise an active and constant surveillance

46 

M. QVORTRUP

over their representatives and reserve for themselves…the right to discard them if they betray their trust” (Constant 1988: 616). The recall is nothing more than putting this into practice.

References Abbott, F. (1915, January). The Referendum and Recall Among the Ancient Romans. Sewanee Review, pp. 84–94. Constant, B. (1988). The Liberty of the Ancients Compared to that of the Moderns. In M.  Gauchet (Ed.), Benjamin Constant: Écrits politique (pp. 589–619). Paris: Gallimard. Cronin, T.  E. (1999). Direct Democracy: The Politics of Initiative, Referendum, and Recall. Cambridge, MA: Harvard University Press. Dando, W. A. (1966). Map of the Election to the Russian Constituent Assembly of 1917. Slavic Review, 25(2), 310–325. de Montesquieu, C. (1989). The Spirit of the Laws. Cambridge: Cambridge University Press. Douglas, P.  H. (1923). Occupational versus Proportional Representation. American Journal of Sociology, 29(2), 129–157. Figes, O. (1996). The People’s Tragedy: The Russian Revolution, 1891–1924. London: Jonathan Cape. Fitzpatrick, S. (2017). The Russian Revolution. Oxford: Oxford University Press. Foxley, R. (2013). The Levellers: Radical Political Thought in the English Revolution. Manchester: Manchester University Press. Gramsci, A. (1978). Selections from Political Writings. London: Lawrence and Wishart. Harding, N. (1996). Leninism. London: Macmillan. Kautsky, K. (1911). Parlamentarismus und Demokratie. Stuttgart: Dietz. Ketcham, R. (2003). The Anti-Federalist Papers. New York: Signet Books. Labriola, A. (1897). Contro il referendum. Roma: Uffici della critica sociale. Laski, H. (1938). A Grammar of Politics. London: George Allen and Unwin. Lenin, V.  I. (1917a). Report on the Right of Recall. In Lenin: Collected Works (pp. 338–339). Moscow: Progress Publishers. Lenin, V.  I. (1917b). Theses on the Constituent Assembly. In Lenin: Collected Works (pp. 379–383). Moscow: Progress Publishers. Lenin, V. I. (1918). Draft Decree on the Right of Recall. In Lenin: Collected Works (pp. 336–337). Moscow: Progress Publishers. Lenin, V. L. (1977). The State and the Revolution. Moscow: Progress Publishers. Lijphart, A. (2012). Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. New Haven, CT: Yale University Press. Luce, R. (1931). Legislative Principles. The History and Theory of Law Making by Representative Government. Boston and New York: Houghton, Mifflin Company.

3  THE POLITICAL THEORY OF THE RECALL. A STUDY IN THE HISTORY… 

47

Luxemburg, R. (1984a). Organizational Questions of Russian Social Democracy. In P. Hudis & K. B. Anderson (Eds.), The Rosa Luxemburg Reader (pp. 48–65). New York: NYU Press. Luxemburg, R. (1984b). What Does the Spartacus League Want? In P. Hudis & K. B. Anderson (Eds.), The Rosa Luxemburg Reader (pp. 312–374). New York: NYU Press. Luxemburg, R. (1984c). The Russian Revolution. In P. Hudis & K. B. Anderson (Eds.), The Rosa Luxemburg Reader (pp. 281–311). New York: NYU Press. Madison, J. (2005). The Constitutional Convention. New  York: Modern Library Classics. Marsilius of Padua. (1988). The Defender of Peace. Cambridge: Cambridge University Press. Marx, K. (1970). Critique of the Gotha Program. In K. Marx & F. Engels (Eds.), Selected Works (pp. 13–30). Moscow Progress Publishers. Marx, K. (1973). Der Bürgerkrieg in Frankreich. In K. Marx & F. Engels (Eds.), Werke 17 (pp. 313–365). Berlin: Dietz Verlag. Marx, K., & Engels, F. (2005). Manifest der Kommunistischen Partei. Frankfurt aM: Fischer Verlag. McLellan, D. (1991). Karl Kautsky. In D. Miller (Ed.), The Blackwell Encyclopedia of Political Thought (pp. 267–268). Oxford: Blackwell. Munro, W. B. (1912). The Initiative, Referendum and Recall. New York: Appleton. Poulantzas, N. (1985). L’État, le pouvoir, le socialism. Paris: PUF. Rousseau, J.-J. (1964a). Du Contrat Social. In B.  Gagnebin & M.  Raymond (Eds.), Rousseau, Jean-Jacques: Oeuvres complétes III (pp.  349–471). Paris Gallimard: Écrites Politique. Rousseau, J.-J. (1964b). Considérations sur le Gouvernement de Pologne. In B.  Gagnebin & M.  Raymond (Eds.), Rousseau: Oeuvres complétes III (pp. 953–1044). Paris Gallimard: Écrites Politique. Serdült, U., & Welp, Y. (2017). The Levelling Up of a Political Institution: Perspectives on the Recall Referendum. In S. Ruth, L. Whitehead, & Y. Welp (Eds.), Let the People Rule? Direct Democracy in the 20th Century (pp. 137–154). Colchester: ECPR Press. Spurlin, P. M. (1969). Rousseau in America, 1760–1809, Tuscaloosa Al. University of Alabama Press. Sullivan, J. W. (1892). Direct Legislation by the Citizenship. New York: Twentieth Century Publishing Company & Co.. Taft, W. H. (1913). Popular Government: Its Essence, Its Permanence and Its Perils. New Haven: Yale University Press. Weitz, E. D. (1994). “Rosa Luxemburg Belongs to Us!” German Communism and the Luxemburg Legacy. Central European History, 27(1), 27–64. Wells, R.  H. (1929). The Initiative, Referendum and Recall in German Cities. National Municipal Review, 18(1), 29–36.

48 

M. QVORTRUP

Weyl, W. E. (1919). The New Democracy. New York: Macmillan. White, W. A. (1910). The Old Order Changeth: A View of American Democracy. New York: Macmillan. White, S., & Hill, R.  J. (1996). Russia, The Former Soviet Union and Eastern Europe: The Referendum as a Flexible Political Instrument. In M. Gallagher & P. V. Uleri (Eds.), The Referendum Experience in Europe. London: Macmillan. Wilcox, D. (1912). Government by All the People: The Initiative, the Referendum and the Recall as Instruments of Democracy. New York: Macmillan Co. Zamoyski, A. (1987). The Polish Way. In A Thousand-Year History of the Poles and their Culture. London: Murray. Zimmerman, J. F. (1997). The Recall. New York: Praeger.

Legal Cases: Hilzinger v. Gillman 56 Washington 228 (Washington Supreme Court 1909), 233.

CHAPTER 4

The Recall in France: A long standing and unresolved debate Clara Egger and Raul Magni-Berton

4.1   Introduction France is one of the rare cases of uninominal electoral systems with an elected President and no recall procedures. The Constitution of the Fifth Republic offers very little room for citizens to demand accountability from people voted in power. Yet, possibilities of shortening the term of elected officials exist. In particular, the President holds a right to dissolve the National Assembly and to trigger new elections, a right already activated five times over the 60-year-long history of the Fifth Republic. The Assembly can in turn remove the President but only  under exceptional circumstances and after a demanding procedure. Moreover, with the change of the electoral calendar in 2000, which aims to limit the competition between the President and the Assembly, the activation of such procedure is even more unlikely.

C. Egger (*) IRIO Department, Faculty of Arts, University of Groningen, Groningen, The Netherlands e-mail: [email protected] R. Magni-Berton Univ. Grenoble-Alpes, Sciences Po Grenoble, PACTE, Grenoble, France e-mail: [email protected] © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_4

49

50 

C. EGGER AND R. MAGNI-BERTON

This institutional set-up is key to understand why a majority of French citizens declares feelings of distrust or aversion towards their politicians in national opinion polls (Cheurfa and Chanvril 2019). This explosive political context has given rise, in November 2018, to the Yellow Vests movement, which has put recall practices back on the agenda two centuries plus after its defence by French revolutionary movements. This chapter aims to explain the specificity of the French experience of recall procedures and practices. To do so, the analysis is structured along four axes. First, the chapter traces the origins and historical diffusion of the idea of recall, introduced during the French Revolution in the unique form of imperative mandate. Second, it analyses current constitutional provisions and discusses their compatibility with some forms of recall processes. Third, it examines the various social demands for recall, especially focusing on the one which has emerged out of the Yellow Vests movement. Lastly, the chapter ends with a discussion of the future and the possible design of recall procedure in France.

4.2   The History of the Demand for Recall Elections in France Although some of its close neighbours (Germany) or inspiration (United States) have introduced recall mechanisms over the course of their history, France has constantly resisted it. A tentative explanation could lie in the absence of a political demand for this tool. Yet, historical accounts largely invalidate this explanation. As in other countries, recall elections have been regularly in the agenda, especially in times of democratic advance (Serdült and Welp 2017: 142). Several outsider parties have also promised its introduction if elected, especially as the discontent towards the irresponsibility of elected officials grew. In the French case, this demand has however long focused on the introduction of the imperative mandate, a close but different form of mechanism to ensure political accountability and representatives’ responsiveness. Yet, this demand has largely stayed at the margins of French political theory and ideologies. Contrary to other forms of direct democracy such as citizens’ initiative and referendum (hereafter: I&R), the demand for imperative mandate or forms of recall elections has failed to gain support outside of radical leftist parties. This may explain why recall elections are largely absent from existing accounts of French political history. The following section attempts to fill this

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

51

gap by focusing on the key moments and specificity of French historical trajectory. The Framing of the Demand: Imperative Mandate during the French Revolution (1789–1793) The specificity of French historical debates on the control mechanisms of elected representatives lies in their focus on imperative mandates. Although imperative mandate is considered as a form of recall practice (Council of Europe 2009), it is different from standard recall procedures in that it explicitly specifies the criteria to launch a recall process. With an imperative mandate, elected officials commit to implement specific measures and can be removed if they depart from this original agreement. In contrast, recall election gives a greater room of interpretation for citizens who can remove their representatives when they feel that they are not meeting their obligations. Imperative mandates have a long history and their use can be traced back to the Middle Ages. During the French Ancien Régime, the role of deputies was limited by the cahier de doléances, a list of specific demands that members of different orders were required to transmit to the King. It is only during the French Revolution that the role of deputies evolves towards a broader and less constrained representative role despite the fierce opposition of French aristocracy advocating for keeping imperative representation (Zaidman 2011: 9–10). Debates on imperative mandates during the first Constituent Assembly played a structuring role in the debates on recall procedures and, more largely, on the limits to be put on the action of elected representatives. Although Rousseau is considered as the father of imperative mandate, his opposition to representation did not lead to the design of specific procedures as he was more in favour of I&R (Rousseau 1762). This opposition to imperative mandate coupled with the defence of direct democratic procedures also characterized some of the most influential thinkers of the Constituent Assembly. Condorcet who designed citizens’ I&R procedures on constitutional matters is, for example, famous for his opposition to imperative mandate, stated in these terms: “the people sent me not to support his opinions but to defend my own” (quoted in Picard 2014: 17). Supporters of the imperative mandate at that time mainly came from popular Parisian movements, supported  by the most radical stream of the revolutionaries. The Cercle Social especially aimed at developing and

52 

C. EGGER AND R. MAGNI-BERTON

­ isseminating a political manifesto on direct democracy and used their d journal, La Bouche de Fer, to do so. Some local sections of the movement adopted imperative representation at that time based on this manifesto. A few leading figures of the Enragés party, such as Jean-François Varlet or John Oswald, used their position in the newly elected French National Assembly to voice this demand (Zaidman 2011: 33–35). The isolation of the Enragés party and the opposition of both democrats and liberals to imperative mandates will bury their recommendations. Imperative mandates were paradoxically perceived as not ambitious enough—as they add checks and balances to representation but do not fundamentally alter it—and too risky. The French Constituent Assembly hence opted for national representation and adopted representational mandates perceived as more likely to create unity in the new Republic by avoiding the voicing of particular and at times clientelist interests. Yet, the short-term failure of the Enragés in the National Assembly should not hide their influence on popular movements. In particular, the 1848 revolutionary spring successfully reintroduced their demand back on the public agenda. A Timid Progression in Times of Revolutionary Ideals (1848–1871) The French revolutionary debates embedded the issue of imperative representation at the extreme left of the political spectrum. This political colour explains the success of the demand during revolutionary times and experiences. Recall procedures received a considerable support in the nineteenth century. Three events illustrate such a revival. First, as the Revolution of 1848 and the subsequent Second Republic instituted adult male suffrage, it also revived debates about the role and limits of representation. Advocates of a direct government became more vocal. Victor Considérant especially claimed that deputies should only act as delegates of the people and that the executive power should be under permanent removal by the Delegates Assembly (Zaidman 2011: 44). The Constitution of the 4 November 1848 invoked the American precedent of 1787 to introduce these concerns into French constitutional law. It is the only one in French history that introduced a penal and political responsibility of the French President. Interestingly, the Constitution states that

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

53

“every measure by which the President of the Republic shall dissolve or prorogue the Assembly, or interpose any obstacle to the exercise of its public trust, shall be deemed a crime of high treason” (art. 68). This definition of high treason is unique in that it is understood as any obstacle put to the democratic exercise of power by citizens. The proposed procedure involves a High Court of Justice which implements—after a request by the National Assembly—recall procedures as well as decides upon the exact charges and associated sentences. The competence of this High Court— composed of 5 judges, of 36 jurymen—gives a central role to citizens and is explicitly detailed in nine dedicated articles (Articles 91 to 100). Ironically, this procedure was never applied as Louis-Napoléon Bonaparte successfully seized power in the coup of the 2 December 1851. The democratic spirit of 1848 however persisted among popular movements. The failure of the Second Republic led Proudhon to advocate for an independent organization at the communal level. In his Principe fédératif, Proudhon (1863) supports the idea of a federation of free communes giving limited and clearly defined powers to a central authority. This idea surely influenced the insurrectionists of the Commune de Paris. Following the principles of the First International, the Central Committee instituted an elected Commune on 26 March 1871 in these terms: “The members of the municipal Assembly, constantly controlled, monitored and discussed by the opinion, are removable and responsible”. Confronted by a civil war, the revolutionaries were not always able to fully apply these principles. In particular, between 28 March and 13 April, the debates of the Municipal Assembly remained secret as their publicity was considered as too dangerous for the survival of the Commune. Despite its violent repression, the memory of the Commune experience and the legacy of Proudhon have had a strong influence on French socialist movements. Although recall procedures and imperative mandates are absent in all the subsequent French Constitutions, they are integrated in the practices of the French union movement. Since the nineteenth century, French labour law has acknowledged the fact that “each member of the personnel delegation to the social and economic committee can be removed during its mandate upon a request from the union organization that appointed her” (Article L 2314-36). Beyond this specific and limited case, recall practices disappeared from French political system after the short-lived revolutionary experiences of the nineteenth century.

54 

C. EGGER AND R. MAGNI-BERTON

The Implantation of Political Unaccountability (1875–1958) The support for recall procedures in popular movements and among leftist intellectuals did not suffice to ensure its inscription in French constitutional law following the Restoration of the Republic in 1875. The Third Republic’s Constitution limited the responsibility of the President and of his ministers to cases of high treason. Citizens had no role in the procedure. Its application was left to the Senate in the Third Republic and to the National Assembly in the Fourth. The 1875 and 1946 Constitutions did not draw on the precedent set by the Second Republic and left the criteria for the procedure and its implementation very vague. Moreover, they resolutely omitted any form of imperative mandate.1 This does not mean the Third, Fourth and Fifth Republics contain no provisions for elected officeholders to be removed. There are legal rules limiting the access to the representational mandate in certain cases. They concern issues such as the incompatibility of electoral mandates with other responsibilities, the interdiction of cumulating mandates or cases of ineligibility. Some debates regularly emerged concerning cessation of office for health reasons (Houillon 2006), but no formal rule was established to limit the ability of the French President to stay in office. However, few procedures enable elected officials to remove other officials or to bring forward the date of new elections. These procedures can be viewed as indirect forms of recall because citizens can put pressure on elected officials to obtain a change. The following section details what the current French Constitution states on this matter.

4.3   Removing Elected Officials Before Completion of Their Term in France The Constitution of the Fifth Republic adopted in 1958 offers many of the preconditions for classic recall procedures. Its electoral system is uninominal. The President is directly elected (from 1965 onwards). At the local level, mayors and their councillors are jointly elected, and the representatives of départements (counties) are elected in a binominal electoral system. In the French political system, there are many elections, and in almost all of them, each elected official is chosen by her own g ­ eographically 1  The current French Constitution even states “no Member shall be elected with any binding mandate” (art. 27).

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

55

located electorate.2 Despite these features, France is one of the few democracies sharing these institutional features that does not provide voters with any mean to remove an elected (national or local) official from office. Citizens can neither initiate a removal nor approve it through a vote. Yet, the 1958 Constitution provides a right to remove officials to two elected entities: the President can dissolve the National Assembly (the lower house), and under specific circumstances, the Parliament can remove the President. The President’s Right to Dissolve the National Assembly As it is the case in other European semi-presidential regimes—such as Austria, Iceland or Portugal—the French President has the power to dissolve the Assembly. The 1958 Constitution states: “The President of the Republic may, after consulting the Prime Minister and the Presidents of the Houses of Parliament, declare the National Assembly dissolved. A general election shall take place no fewer than twenty days and no more than forty days after the dissolution” (Article 12). The “consultation” of the Prime Minister and the Presidents of the Houses of Parliament does not involve any veto power or consensus requirements. “Consulting” merely means “informing” in this case. Two specific contexts however impede the President from making use of this right: during the year following the election caused by a dissolved legislature and during the exercise of emergency powers. In practice, these conditions rarely occur leading the presidential power to dissolve the Assembly to be almost of a discretionary nature (Goplerud and Schleiter 2016). Since 1958, the presidential right to dissolve the Assembly has been used five times. During the 1960s, dissolution was used to solve political crises. In 1962, the Parliament for the first time used its right to constrain the Prime Minister to resign.3 As the President supported the latter, dissolving the Assembly was a way of asking voters to solve the conflict between the President and the Assembly. The second dissolution took place in 1968, during the May civil unrest. Once again, the government 2   The only exception is the election of regional councillors, which is party-list proportional. 3  This right is usual in parliamentary and semi-presidential regimes. We do not analyse this procedure in this chapter, because we focus only on the ways to remove directly elected officials. Ministers are appointed by the Parliament or by the President, and they are therefore not directly elected.

56 

C. EGGER AND R. MAGNI-BERTON

faced street contestation, and triggering a new election was a way of letting voters decide of the resignation or confirmation of the government. While voters can neither initiate nor declare the dissolution, these two events bear similarities with recall practices. As opposition to the government grew, elections were triggered to allow voters to arbitrate the conflict. Following dissolution practices were different and mainly aimed at avoiding divided government. Given that, until 2001, the President had a seven-year term and the Assembly a five-year one, the partisan affiliation of the President and of his government did not necessarily match. As a result, in 1981 and 1988, the dissolution took place just after the presidential election, while in 1997 it occurred two years after, one year before the planned term of the legislature. This partisan use of removing MPs from office is also observed when the citizens have the right to recall. However, the main difference is that when the President has this right, it will be used only to avoid divided governments, while when this right is enlarged to voters, it can be also used to promote them (Welp and Milanese 2018). The Parliamentary Right to Remove the President Until the constitutional revision of 2007, the French President was almost irremovable. The only way of removing Presidents before the end of their term was the conviction for high treason, initiated by the majority of the MPs and arbitrated by the High Court, directed by 5 judges and composed of 12 MPs appointed by the Parliament. While the meaning of “high treason” is somewhat vague, many behaviours—including criminal acts—do not fall under this realm (de Cazals 2007). Moreover, the procedure included a combination of representatives from the legislative and judicial powers, a fact reflecting its exceptionality and its mixed character (between a criminal and a political judgement). After some revealed corruption cases that involved President Chirac at the end of the 1990s, a commission was set up to initiate a constitutional revision, in order to modernize the Article 68. The commission proposal was validated on February 2007, just before the new presidential election. The new constitutional Article 68 states: The President of the Republic shall not be removed from office during the term thereof on any grounds other than a breach of his duties patently incompatible with his continuing in office. Such removal from office shall be

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

57

proclaimed by Parliament sitting as the High Court. The proposal to convene the High Court adopted by one or other of the Houses of Parliament shall be immediately transmitted to the other House which shall make its decision known within fifteen days of receipt thereof.

The main novelty of this article is that all jurisdictional aspects are deleted, making the decision exclusively political (de Cazals 2007). The Parliament (sitting as the High Court) now fully controls the process through which the President can be removed. The procedure needs to demonstrate the existence of a “breach” in “duties patently incompatible with his continuing in office”. This limits the discretionary power of the Parliament, but enlarges the possible reasons to initiate the procedure, including corruption or a mere inability to reduce social contestation putting the state at risk. Paradoxically, French Presidents—during their mandate—are now entirely unaccountable to judiciary powers (including civil justice) and can only be removed by the Parliament. The process requires that one-tenth of the Parliament signs a motived motion for a resolution, which must then be validated by the parliamentary standing committee for constitutional law. If both chambers—Assembly and Senate—adopt the motion, the High Court is set up and has to decide over a one-month period. A majority of two-thirds of the Parliament is then required to remove the President. This article has never been used despite an abortive attempt in 2016 signed by 79 MPs. This procedure is certainly much more demanding than dissolving the Assembly. It not only requires a specific justification but also rests on a large consensus among MPs. In addition, during the process, Presidents are not removed from office and can dissolve the Assembly. The timing of the revision also increases its unlikely nature: with the revision of the electoral calendar and the decision to align the duration of the President’s and of the Assembly’s mandates, both are no longer competing. Neither the presidential right of dissolving the National Assembly nor the parliamentary right to remove the President have been activated in recent years. An explanation for this political restraint lies in the 2001 reform of the electoral calendar (de Cazals 2007). Following the constitutional referendum held on 24 September 2000, the length of the presidential mandate was reduced from seven to five years and therefore aligned with the term of the Legislative Assembly. Moreover, legislative elections have been scheduled to immediately follow the election of the President.

58 

C. EGGER AND R. MAGNI-BERTON

This produces both a de facto impossibility of a divided government and a presidentialization of French political parties. Divided governments are unlikely because presidential and legislative elections are now so close in time that voter preferences are likely to stay globally stable. The output is similar so that the majority party is also the party of the President. Moreover, political parties presidentialize because the result of the presidential election, which precedes the legislative one, influences the way people vote at the legislative election. On average, winning the presidential election produces a 25% increase in the votes for the party of the President in legislative elections (Magni-Berton and Robert 2017). This allows Presidents to hugely influence the choice of their party’s candidates in legislative elections. When Presidents and National Assemblies do not conflict each other, the procedures to remove each other become inoperative. Therefore, while the Constitution gives more power to the Parliament to remove the President, it has also divested the Parliament of incentives to do so.

4.4   The Yellow Vests Movement and the Return of Recall to the Political Agenda This French institutional situation carries explosive potential consequences. Uninominal electoral systems are known to increase the likelihood of unrest (Reynal-Querol 2002). The main reason is that they produce stable governments at the cost of introducing a strong disproportionality in the electoral system allowing a party supported by a minority to control the legislature. The French institutional context is key to understanding why the newly elected Macron government faced an unprecedented social movement 18  months after its coming in power. The specificity of this social movement—referred to as the Gilets Jaunes (Yellow Vests) movement—is its focus on demands of institutional reform and, in particular, on the adoption of direct democracy procedures in France, including the right to initiate a recall election. This section presents the political context of the emergence of the movement, its proposals for institutional reform as well as the ensuing debates about the introduction of recall procedures in France.

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

59

The Political Context: Rising Demands for Political Accountability The growing disaffection of French citizens towards their political system has resulted in a decrease of turnout combined with an increased share of votes for losing parties since the beginning of the 2000s. Figure 4.1 shows the consequence of that: the disproportionality produced by the electoral system has hugely increased, and it reaches its maximum in 2017. This situation is partly responsible for the fact that governments are rapidly unpopular, with approval ratings lower than 30%. As a result, debates have started on the need to adapt French national institutions focusing on granting a stronger role to citizens in decision-making processes. In particular, the introduction of recall arrangements has gained support since the 2010s. The capacity of citizens or institutional counter-­ powers to dismiss elected officials or to anticipate elections was perceived as a way of moderating the negative effects of disproportionality produced by uninominal systems. The introduction of recall procedures initiated by 4 3.5 3 2.5 2 1.5 1 1962 1967 1968 1973 1978 1981 1986 1988 1993 1997 2002 2007 2012 2017

Fig. 4.1  Disproportionality under the French Fifth Republic. Disproportionality represents the ratio between the percentage of seats received by the majority coalition and the percentage of registered voters who opted for this coalition in the first round of legislative elections. When this ratio equals 1, these percentages are identical, meaning that the representation is genuinely proportional

60 

C. EGGER AND R. MAGNI-BERTON

citizens (but also by institutional counter-powers) could provide a serious threat to stable governments voted by a minority and might provide an incentive for the government to take into account the preferences of the median voter. Since the early 2010s, recall arrangements have especially been defended both by far-left and far-right political parties. The Front National (far right) was the first party to propose a constitutional reform to introduce a recall mechanism in the French Constitution. The proposition was that a recall election could be initiated by one fifth of registered voters. It aimed to remove a President in office without affecting her capacity to stand for another election (provided the limit on the number of terms is not reached) (Collard 2013). Among the leftist parties, the introduction of recall elections was initially promoted by the Parti de Gauche (now La France Insoumise, far left). Between 8 November and 11 November 2014, militants from this party sponsored an unofficial vote on this issue in Metropolitan France and its overseas territories. French citizens were asked to answer to the following question: Do you support a citizen’s right to recall elected officials? 500 ballot boxes were available on the French territory, enabling 180,000 citizens to participate in the poll. The results were largely in support of the introduction of recall (Garrido 2014). This proposition has also been timidly supported by Benoît Hamon, candidate of the Parti Socialiste (social democrat) during the 2017 presidential election. Yet mainstream parties rather supported the introduction of participatory democracy mechanisms resting on the non-binding consultation of citizens to inform legislative processes. These parties considered initiative and referendum (including recall) as too dangerous, too associated with the agenda of so-called populist parties and not adapted to the workings of French representative democracy. Experts and intellectuals echoed these concerns. In 2015, a parliamentary working group “Recreate democracy” rejected the measure after an internal vote and without providing substantiated arguments (Bartolone and Winock 2015). In contrast, the High Authority for transparency in public life suggested strengthening the capacity of MPs to sanction elected officials guilty of “grave breaches of ethics” such as tax evasion. The report stressed the fact that removal should not become a political weapon against dissident or minority political opinions but should be used in a cross-partisan manner under the control of the constitutional court (Nadal 2015).

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

61

Although the current President Emmanuel Macron stressed the need to make elected officials more responsible during his electoral campaign, his election in April 2017 did not deliver on these promises. His proposed measures mainly focused on enabling citizens to comment on the work of their MPs, but did not grant them a right to oppose and sanction elected officials. This neglect combined with the fact that parties proposing a right to recall obtained more than 40% of support in 2017 presidential election put the newly elected government at risk of popular contestation. The Yellow Vests Movement and Its Proposals for Institutional Reform in France In October 2018, 18 months after the election of Macron’s Republique en Marche, an unprecedented social movement started in France. After an online petition posted in May had attracted nearly a million signatures, mass demonstrations began on 17 November 2018. The movement initially started due to rising fuel prices and a high cost of living. Messages on social media especially denounced that a disproportionate burden of the government’s tax reforms was falling on the working and middle classes especially in rural areas. Yellow high-visibility vests, which French law requires all drivers to have in their vehicles and to wear during emergencies, were chosen as “a unifying thread and call to arms” because of their convenience, visibility and association with working-class industries (Friedman 2018). The protests started in the context of the low popularity of the recently elected President, nicknamed “President of the very rich” whose start of mandate was already tarnished by several scandals and the resignation of several Ministers. The movement emerged as a deliberately unorganized and leaderless grouping, proclaiming their repudiation of political representation, even towards would-be political leaders emerging from within their own ranks. Rapidly, the movement focused not only on economic demands but above all on institutional and political reforms. In the first week of the movement, these demands appeared as very diverse. Between November and December 2018, French roads and roundabouts started to be covered by messages demanding the resignation of Emmanuel Macron. Some messages called for the introduction of a seven-year mandate for the French President. Yet late November, political demands started to focus on a unified and very clear proposal: the introduction of the so-called

62 

C. EGGER AND R. MAGNI-BERTON

CARL citizens’ I&R which becomes the priority measure for the movement from mid-December. The CARL abbreviation refers to four direct democratic measures, conceived as indivisible package to increase the political rights of French citizens. All the proposed measures have in common to enable citizens to take decisions without relying, at any time, on elected representatives. “C” means constituant (constituent) and refers to the introduction of I&R in constitutional matters. This tool emerged from the 2005 French experience of the project of the European Constitution. French citizens were called to the ballot to validate the project—which required a modification of the French Constitution—and largely rejected the project. The Lisbon Treaty—a revised version of the rejected project—was then adopted by MPs. In the eyes of the Yellow Vests movement, this event reflected the lack of political power of French citizens. “A” means abrogatif (repeal), that is, the capacity of citizens to veto a law. This form of referendum was one of the most popular at the beginning of the movement as supporters of the Yellow Vests opposed the economic and redistributive policies of the French government which largely benefit the richest. “R” stands for révocatoire (recall). The proposition aims to allow citizen to recall any political officials, whether they hold a local or national mandate or belong to the executive or legislative branch. The last element of the RIC CARL refers to legislative I&R on ordinary law matters. While the linkages of these very diverse forms of referendum are unique to the Yellow Vests movement, the Swiss example appeared as a primary source of inspiration. Besides this focus on four interrelated direct democracy tools and on the definition of an overall procedure, the movement did not go as far as to suggest a specific way of introducing the measure into French law or to spell out a specific procedure for each of the forms of citizens’ referendum. In many respects, the proposition appeared as the lowest common denominator among a political diverse movement. Members of the Yellow Vests only have in common to be poor workers with low revenues.4 The movement was especially successful in rural areas that have faced a loss in public services over the past decades. Most of the members of the movement 4  Sciences Po Bordeaux (2018), “Gilets jaunes”: une enquête pionnière sur la “révolte des revenus modestes”, Le Monde.fr, 11 décembre 2018, https://www.lemonde.fr/idees/article/2018/12/11/gilets-jaunes-une-enquete-pionniere-sur-la-revolte-des-revenusmodestes_5395562_3232.html.

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

63

were politically active for the very first time. At the political level, the movement draws from very diverse backgrounds. A survey conducted in December 2018 on a limited number of Yellow Vests activists showed that most of them refused the right/left cleavage, while some claimed to feel closer to the far left or the far right.5 Given the political context of the emergence of the movement, it is not surprising that direct democracy demands became the priority demand as these tools reduce the power of elected officials while preserving the unity of the movement. A drawback of the multifaceted nature of the demand is the fact that it can be easily manipulated by political parties and elected officials who can pick the form of referendums which affects the lesser of their powers. Depending on the political affiliation of members, different options were proposed. Some integrated citizen I&R within a broader agenda of constitutional reform. Constituent workshops started to be organized in different places to write a new Constitution including but not limited to direct democracy tools. Others stressed the need to adapt some articles of the current French Constitution with two proposals particularly standing out. One concerns the modification of Article 3. The proposition suggests adding after the existing article—“National sovereignty shall vest in the people, who shall exercise it through their representatives and by means of referendum”—“initiated by citizens in all matters, included constitutional and related to treaty ratification. This article can only be modified by referendum” (Article 3 2019). The second suggests a specific focus on the introduction of citizen initiative in constitutional matters through the modification of Article 89 (Magni-Berton and Egger 2019). Initially this difference of strategy did not affect the unity of the movement. The situation changed due to the adverse reception of the proposition by political parties. Everything but Citizen I&R: The Reception of the Demand The proposition to introduce direct democracy procedures—including but not limited to recall elections—was received in a very negative manner among political and media circles. On 10 December 2018, President Macron tried to appease the movement in the most viewed political speech in French history. His answer mainly focused on economic measures but left aside the Yellow Vests’  Idem.

5

64 

C. EGGER AND R. MAGNI-BERTON

demand for democracy. Instead of granting citizen I&R, a Great National Debate was organized aiming to restore dialogue between French citizens and the President. This debate was supposed to inform an agenda of reform, but the procedures through which data and input were to be aggregated was not transparent. Besides this measure, the Interior Ministry ordered French police forces to engage in a brutal repression of the movement, causing thousands of people to be injured. The reaction of opposition parties differed across the political spectrum. Two parties—the Rassemblement National (ex Front National, far right) and the French Insoumise (far left)—tried to align the demands of the movement with their own political agendas. This strategy—especially marked for the France Insoumise—succeeded in creating divisions among the Yellow Vests movement. Parts of the movement focused on a sovereigntist agenda—making the priority a referendum on international and European treaties—whereas other groups advocated for the introduction of referendums on non-constitutional matters. Overall, within the movement, recall elections progressively lost appeal. A tentative explanation for this is that it was perceived as too associated with the agenda of the France Insoumise and not empowering enough for citizens. In that regard, the Yellow Vests movement was not just seeking a veto power, but more ambitiously advocated for a right of initiative. Mainstream political parties stressed the risky character of citizen initiatives, associating the demand for recall with far-right populist and anti-elite demands. They reshaped the Yellow Vests proposition by focusing on the need to develop more participatory tools at the local level. The media coverage of the demand for direct democracy further revealed the divorce between popular and intellectual classes in France. The Yellow Vests movement was portrayed as a violent, far-right and ill-­ informed movement. The media gave intense coverage of minority racist or violent incidents taking place on the very margins of the movement (ACRIMED 2018). Mainstream media emphasized the risky and uncertain impact of the introduction of all forms of recall and other direct democracy mechanisms. Very few experts on direct democratic procedures were invited to comment on the proposition of the Yellow Vests. The use of comparative evidence—focusing on the use of citizens’ I&R in other countries—was rather minimal. As a result, the latest opinion poll on the confidence in media revealed a record loss of confidence from French citizens. Findings especially emphasize that media are not perceived as independent from political (69%) and financial (62%) pressure. Some (51%) of

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

65

the respondents stated that the Yellow Vests movement was badly covered by the conventional media, which over-dramatized the event (67%) and did not fully allow to understand the demands of the movement (54%).6

4.5   Is There a Future for Recall Procedures in France? Following the Yellow Vests movement, the future of recall elections in France remains open. While in 2018 most of French citizens had not ever heard about the existence of such a procedure, in 2019, almost all of them know what it is. Therefore, for the very first time, its introduction in the French Constitution is now a conceivable scenario. Two factors will determine the future of the tool: first, will these reforms be supported enough to become a strong electoral argument? Second, what procedure of recall could be adapted to the French system of representation? A Large but Fragile Support for Recall Procedures in French Political Opinion The Yellow Vests movement has increased the salience of direct democracy in general and of recall procedure in particular. However, it is difficult to assess whether this popularity will lead to an institutional reform. The most detailed survey on citizens’ I&R revealed that 60% of respondents declared “knowing exactly what the citizen initiative and referendum is”. This percentage was much higher among people who declared supporting the Yellow Vests (72% vs. 54%).7 Three instruments were detailed: I&R to introduce a law (supported by 77% of the respondent), I&R to abrogate a law (72%) and recall of officeholders (67%). Interestingly, references to constituent I&R were absent from the poll. Recall elections hence are the least supported instrument but remain popular in public opinion. Except among the electorate of the President’s party and among those who are strongly opposed to the Yellow Vests, the support is majority in all social and political groups. 6  Carasco A. (2019), “Baromètre médias, les journalistes sommés de se remettre en question”, 24 janvier 2019, La Croix.fr, https://www.la-croix.com/Economie/Medias/ Barometre-medias-journalistes-sommes-remettre-question-2019-01-24-1200997667. 7  Ifop pour Valeurs Actuelles (January 2019). Les Français et le référendum d’initiative citoyenne.

66 

C. EGGER AND R. MAGNI-BERTON

While the Yellow Vests movement has strongly contributed to the return of this issue to the public agenda, it is probably not responsible for a rise in support for direct democracy. Some months before the beginning of the movement, the Pew Research Center measured this support in several countries. In France, it had already reached 74%.8 In 2011, 72% of French people supported I&R.9 Past surveys demonstrate that the level of support for direct democracy in France has not significantly increased after the Yellow Vests. Despite its popularity, direct democracy has not yet become a solid electoral argument. Beyond the extent of support, the intensity of such support is determinant for influencing the agenda of political parties. Many clues suggest the intensity of public support for direct democracy is still low. First, only 29% of respondents “strongly support” recall, lower than the share of respondents who (strongly or moderately) oppose it (33%). Hence, the majority of people who supports recall does so in a lukewarm fashion. Second, institutional issues are systematically not considered as priority. For example, a survey shows that economic issues are much more salient than institutional ones.10 Purchasing power (cited by 48%), taxes (44%), unemployment (29%), inequalities (28%), public spending (25%) and pensions (23%) are all more chosen than “the citizens’ participation in political decisions” (16%). Even among respondents who declare themselves “Yellow Vests”, only 27% consider direct democracy as priority (far behind purchasing power and taxes). These results may explain why political parties have so far been little responsive to the institutional demand of the movement, in particular with regards to the introduction of recall mechanisms. This weak responsiveness of political parties has opened a space for the common mobilization of associations defending direct democracy. In August 2019, ten associations allied to develop a political strategy to obtain the introduction of

8  Pew Research Center (October 2017), “Globally, Broad Support for Representative and Direct Democracy”. The question was: “Would a democratic system where citizens, not elected officials, vote directly on major national issues to decide what becomes law be a good or bad way of governing this country?” 9  Ifop pour l’Observatoire de la Fiscalité et des Finances Publiques (March 2011). Les Français et le referendum d’initiative populaire. 10  Institute Elabe pour BFMTV (January 2019). Les Français et les gilets jaunes (1003 respondents).

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

67

I&R and recall mechanisms in France.11 This strategy included a project to develop a concrete reform, direct and grassroots lobbying activity and, potentially, direct involvement in elections. What Might a French Recall Mechanism Look Like? Following the Yellow Vests, only one proposal to introduce recall elections in the French Constitution was submitted to the Parliament in January 2019 which rejected it 44  days later. This proposal aimed to introduce many I&R reforms and included an article on the possibility of removing elected officials. The draft article provided that the President, the MPs and local elected officials could be removed after a vote triggered by 5% of registered voters. The election would have to be held within a maximum timespan of six months after the receipt of the petition reaching the required threshold of support signatures. The procedure could only be activated after the first third of the term. Within the Parliamentary Commission, debates were quite poor. Few discussions focused on the concrete design, with only the general principle of direct democracy receiving consideration. The mainstream parties rejected everything. They argued that I&R mechanisms lead to political instability, incentivize opportunistic policies and strengthen the influence of lobbying activities. Three proposed amendments were however interesting. The first one suggested giving voters the possibility to dissolve the Assembly as a whole, rather than removing single MPs in their electoral district. The second one aimed to increase the threshold of signature for the dismissing of the President, based on the argument that removing the President is a more critical decision than changing the law. For stability motives, the third amendment suggests to delete the possibility to remove the President. All these amendments were rejected (Lachaud 2019). Some elements of this debate reflect two general issues pertaining to recall procedures that are of particular relevance in the French case. The first issue deals with the alternative between removing single MPs and removing the Assembly as a whole. The first option is possible when there is a single-member district electoral system as in some American presidential democracies. The second option is used in some European 11  These associations are Article 3, CLIC-RIC, Culture RIC, Dauphiné démocratique, Démocratie d’abord, Faites des RIC, Mouvement pour L’Initiative Citoyenne, Objectif RIC, Opération article 3, Scrutin National RIC.

68 

C. EGGER AND R. MAGNI-BERTON

parliamentary systems with either party-list or mixed electoral systems. In some ways, France is between these two models. On the one hand, the presidential right to dissolve the Assembly is a common trait of European Republics and means that the Assembly is viewed as a whole. On the other hand, France is, in the European Union, the only country with both a single-member district electoral system and a directly elected President. This makes the recall of individual representative possible. However, the proposed law would force France to keep its electoral system, a proposal which runs counter to political parties’ propositions to establish a proportional system (demanded by La France Insoumise, the same party that put forward the recall proposal) or a mixed electoral system (largely supported across the political spectrum). Moreover as the French electoral system is not specified in the Constitution, it can easily be changed. For example, during the 1980s, a proportional system was introduced and then abandoned. If the recall of single MPs was adopted, any change of the electoral system requiring multiple-district membership would either violate the Constitution or allow some MPs to be protected from recall procedures, or require the introduction of a different electoral rule in the case of recalled officials, which produces a clear electoral incentive to abuse this procedure. Therefore, the amendment consisting in enlarging the right to dissolve the whole Assembly is more adapted to the French context. The second issue lies in the required threshold of signatures, especially for a presidential recall. The argument that this threshold should be higher than for a simple legislative initiative is reasonable. This principle is already implemented in many places where it incentivizes its use in exceptional cases. In the specific case of the President, countries with such a provision (Bolivia, Ecuador, Venezuela) have particularly high threshold requirements. One can however wonder whether the quality of the procedure is to be assessed against its exceptionality. Contrary to other popularly initiated referendums, recall votes are not expected to improve the quality of debates. In policy referendums the focus is on specific public policy issues, and this promotes an in-depth public debate (see, e.g. Smith and Tolbert 2009). In contrast, recall votes frame the debate on the global performance of an official and produce a state of permanent campaigning, which is deleterious for legislative work (Welp 2016). Therefore, high thresholds of signatures for recall procedures are justified. We have to consider, however, that the higher the threshold, the less useful the device. For example, at the extreme, if 51% of voters were required to sign the petition to bring forward new elections, the vote itself

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

69

would become redundant. Considering that in the last French legislative election the turnout was of 48.7%, even 25% of signatures could be enough to directly trigger a new election, without needing to proceed on to a vote. This exactly corresponds to the governing recall procedure in Arizona, in Michigan or in Wisconsin. There are convincing arguments to avoid the elections even when the threshold is lower—say 15% or 20% of voters—especially in France, which is much more populated than US states. The main argument is based on the problem of strategic or partisan use of triggering new elections. Some opposition parties may initiate a recall procedure (to dissolve the Assembly or remove the President) when the polls are in their favour. This incentive is not entirely negative, because it encourages ruling parties to take account of the preferences of other parties. However, if minority parties have nothing to lose when initiating a recall procedure, this could produce a state of permanent campaigning that deteriorates the quality of the government. We argue that recall provisions without votes minimize the use of recalls in a strategic way, especially in the case of dissolution of the Assembly. Opposition parties, which would strategically use the dissolution procedure through a petition, are likely to be punished by voters, exactly as President Chirac was sanctioned in 1997 for having dissolved the Assembly for strategic reasons. Such parties could then lose seats and reinforce the party in power in the election they triggered. In contrast, if the petition triggers a recall vote instead of an election, the opposition party that would initiate the petition cannot lose. If voters decide to punish the initiative, the referendum is lost, no election is triggered and each party keeps its seats. On the contrary, if voters support the initiative, the vote is won, the Assembly dissolved and the new elections will see the former ruling party penalized. Therefore, the worst scenario for the opposition party that initiates the dissolution procedure is the status quo. This certainly does not deter strategical and partisan use of recall procedures. Considering that elections also have an economic cost, relatively high thresholds of signatures without vote could be a relevant formula for the French system.

4.6   Conclusion In France, recall elections have never been introduced. This does not mean that their introduction is unlikely. For more than two centuries, regular demands for removing officials have been voiced triggering intense

70 

C. EGGER AND R. MAGNI-BERTON

debates which uniquely focused on the role of imperative mandates. The latest resurgence of this debate occurred in December 2018. Instead of citizens’ recall, the French Constitution provides national officials with the right to remove other elected officials: the President of the Republic can dissolve the Assembly, and since 2007, the Assembly can dismiss the President. However, the 2001 reform of the electoral calendar has removed incentives to use these rights by substantially reducing the likelihood of conflict between Presidents and Assemblies. During this period, the French government has managed to obtain the majority of seats with less and less votes. In 2017, only 15% of registered voters were sufficient for the winning coalition to obtain 61% of seats in the National Assembly. This has produced high levels of discontent, which resulted in the birth of the Yellow Vests movement on 17 November 2018. Rapidly, the demand for direct democracy (including but not limited to recall) has risen up the movement’s agenda. It is difficult to forecast whether recall will be institutionalized in the French case. For sure, issues raised by electoral system-induced disproportionality, combined with the heterogeneity of the French electorate, have to be addressed. Recall elections are currently the most popular way of moderating this concentration of power in a few hands, and many groups in civil society are mobilized to promote it. However, most French politicians strongly oppose it and prefer modifying the electoral system in order to introduce more proportionality. The specific provisions needed to make a lasting success of recall reform are little studied by even specialists. The future of recall in France is therefore still uncertain.

References ACRIMED. (2018). Mobilisation des Gilets Jaunes. Les médias et les mobilisations sociales. Retrieved from https://www.acrimed.org/-Mobilisations-des-giletsjaunes-2018-. Article 3. (2019). Pétition pour l’instauration en France du référendum d’initiative citoyenne. article3.fr. Retrieved from https://www.article3.fr/actions/ petition-pour-l-instauration-du-referendum-d-initiative-citoyenne-en-france. Bartolone, C., & Winock, M. (chairs). (2015). Refaire la Démocratie, XIV e legislature, Groupe de travail sur l’avenir des institutions de l’Assemblée Nationale. Paris: La Documentation Francaise. Retrieved from https://www.ladocumentationfrancaise.fr/var/storage/rapports-publics/154000692.pdf.

4  THE RECALL IN FRANCE: A LONG STANDING AND UNRESOLVED DEBATE 

71

de Cazals, M. (2007). La Ve République face à l’instauration d’une destitution politique inédite du Président de la République. Revue française de droit constitutionnel, (3), 451–470. Cheurfa, M., & Chanvril, F. (2019, January). 2009–2019: la crise de la confiance politique. Baromètre de la confiance politique. Paris: CEVIPOF. Retrieved from https://www.sciencespo.fr/cevipof/sites/sciencespo.fr.cevipof/files/ CEVIPOF_confiance_10ans_CHEURFA_CHANVRIL_2019.pdf. Collard, G. (2013, April 24). Proposition de loi constitutionnelle organisant le referendum révocatoire du Président de la République. Rassemblement National. Proposition de loi constitutionnelle organisant le referendum révocatoire du Président de la République. Council of Europe. (2009, June 16). Report on the Imperative Mandate and Similar Practices. Strasbourg. Study n. 488/2008. Friedman, V. (2018, December 4). The Power of the Yellow Vest. The New York Times. Garrido, R. (2014, November 20). La révocabilité des élus est-elle une ambition réalisable? L’Humanite.fr. Retrieved from https://www.humanite.fr/ la-revocabilite-des-elus-est-elle-une-ambition-realisable-558120. Goplerud, M., & Schleiter, P. (2016). An Index of Assembly Dissolution Powers. Comparative Political Studies, 49(4), 427–456. Houillon, G. (2006). A propos de la santé du Président de la République. La Revue administrative, 59(354), 620–628. Lachaud, B. (2019). Rapport fait au nom de la commission des lois constitutionnelles, de la législation et de l’administration générale de la république sur la proposition de loi visant à instaurer la possibilité de référendums d’initiative citoyenne (n° 1558). Magni-Berton, R., & Egger, C. (2019). RIC: le Référendum d’initiative citoyenne expliqué à tous. Au cœur de la démocratie directe. Limoges: FYP éditions. Magni-Berton, R., & Robert, M. V. (2017). Maximizing Presidential Coattails: The Impact of the Electoral Calendar on the Composition of the National Assembly. French Politics, 15(4), 488–504. Nadal, J. (2015). Renouer la confiance publique  – Rapport au Président de la République sur l’exemplarité des responsables publics. Haute Autorité pour la transparence de la vie publique. Paris: La Documentation Francaise. Retrieved from https://www.ladocumentationfrancaise.fr/rapports-publics/154000023/ index.shtml. Picard, D. (2014). L’urne imprévisible ou Le malheur de la République. Paris: Société des Ecrivains. Proudhon, P. J. (1863). Principe Fédératif. Paris: E.Dentu Editeurs. Reynal-Querol, M. (2002). Political Systems, Stability and Civil Wars. Defence and Peace Economics, 13(6), 465–483.

72 

C. EGGER AND R. MAGNI-BERTON

Rousseau, J. J. (1762). Du Contrat Social. Genève, version du 7 octobre 2012. Retrieved from https://www.rousseauonline.ch/pdf/rousseauonline-0004. pdf. Serdült, U., & Welp, Y. (2017). The Levelling Up of a Political Institution. Perspectives on the Recall Referendum. In S. P. Ruth, Y. Welp, & L. Whitehead (Eds.), Let the People Rule? Direct Democracy in the Twenty-First Century (pp. 137–154). Colchester: ECPR Press. Smith, D. A., & Tolbert, C. (2009). Educated by Initiative: The Effects of Direct Democracy on Citizens and Political Organizations in the American States. Ann Arbor: University of Michigan Press. Welp, Y. (2016). Recall Référendums in Peruvian Municipalities: A Political Weapon for Bad Losers or an Instrument of Accountability? Democratization, 23(7), 1162–1179. Welp, Y., & Milanese, J. P. (2018). Playing by the Rules of the Game: Partisan Use of Recall Referendums in Colombia. Democratization, 25(8), 1379–1396. Zaidman, P.  H. (2011). Le mandat impératif. De la Révolution Française à la Commune de Paris. Paris: Les Editions Libertaires.

CHAPTER 5

Recall Elections in the US: Its Long Past and Uncertain Future Joshua Spivak

5.1   Introduction The start of the twenty-first century has seen the recall election become an international phenomenon, with presidents, governors and mayors and other lower-level officials facing recalls (see Whitehead 2018; Welp 2018). While the United States of America has limited the scope of the recall— the President is not subject to a recall, and most likely neither is any member of the federal government—the country is arguably one of the most notable users of this particular electoral weapon. This is not just because of its frequency but because of one notable use. Thanks to the recall of California Governor Gray Davis in 2003, and his replacement by international movie star Arnold Schwarzenegger, the recall gained a spotlight that it has not since yielded. The recall in the US has waxed and waned over the years. A look at US history shows that the recall occasionally has had bursts of popularity, only to crash back into disuse. For example, California had recalls against three state legislators from 1913 to 1915. It then went 79 years without, until another California state legislator faced a recall. Even the study of the

J. Spivak (*) Hugh L. Carey Institute for Government Reform, Wagner College, New York, NY, USA © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_5

73

74 

J. SPIVAK

recall goes through peaks and valleys. Decades have frequently gone by without serious examination of the device (Sych 1996). The result is that incorrect facts and statistics on the recall have become lore and are repeatedly cited by various writers. The recall appears to be currently in a boom phase. The US has maintained a role as a leader in the use of the recall, one that has been disseminated throughout the globe. In the US the practice of recall only exists at the state and local levels, where there is a wide variety in these laws. In some countries a recall requires a voter turnout threshold in order to go into effect. In the Romanian Presidential recall in 2012, the recall failed because only 46.24% voted. This provision is rare in the US (only Idaho requires a form of the “absentee veto” provision on the state level), perhaps not a surprise in a country that has been enamoured with the First Past the Post system. What the US instead has are  variations on similar provisions. For example: 1. To get a recall on the ballot, petitioners need to collect signatures from anywhere between 10 and 40% of a specific target group—in many cases, the number needed is the turnout from the last election (i.e. signatures equal to 25% of the total turnout for the seat being contested). Sometimes, the signature total is tied to total registered voters, which is a much higher amount. 2. The resulting ballot can be either (1) a brand new election, where the elected official faces off against one or more candidates who meet the eligibility requirement; (2) an up or down (yes or no) vote to remove the official; or (3) an up or down vote and then a replacement race, in which the official is frequently, but not always, barred from running if they lost the up or down vote. 3. The replacement can be chosen by voters on the same day, at a different day or through appointment by an elected body. 4. In one state, Virginia, when a recall qualifies for the ballot, the judge holds a “recall trial” and decides if the official should be ousted. The following table provides a look at some of the diversity recall requirements in all US states (Table 5.1). The way that the US states are using the recall may be a sign for how other countries will be using it in the future. The recall is also changing in ways that present an exciting and at times unsettling new method of using direct democracy to enact changes.

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

75

Table 5.1  Recall election requirements in all US states State

Officials facing recalla

Political (P) or malfeasance standard (M)

Type of recall

How a replacement is chosen

Alabama

Local only

P

Yes or no

Alaska Arizona Arkansas California

State/no judges State Local only State

M P P P

Yes or no New election New election Yes or no

Colorado

State

P

Yes or no

Different-day election Appointment New election New election Same-day replacement race Same-day replacement race

Connecticut District of Columbia Florida

Local Local

P P

Yes or no

Local

Georgia

State

Mix— Yes or no malfeasance requirement on state level M Yes or no

Hawaii Idahob

Local State

P P

Yes or no Yes or no

Illinois

Local + Governor

Yes or no

Kansas Louisiana

State State

P (M for Governor) M P

Maine Marylandc

Local Local

P P

Yes or no Yes or no

Massachusettsc Local

P

Yes or no

Michigan Minnesota

State State

P M

New election Yes or no

Missouri

Local

Montana

State

Split between P Yes or no and M M Yes or no

Nebraska

Local

P

Yes or no Yes or no

Yes or no

Different date election Different date election

Different date election Unclear Appointment— must be in the same party Same-day election Appointment Different-day election Appointment Different-day election Different-day election New election Different-day election Appointment Different-day election Both (continued)

76 

J. SPIVAK

Table 5.1 (continued) State

Officials facing recalla

Political (P) or malfeasance standard (M)

Type of recall

How a replacement is chosen

Nevada

P

New election

New election

New Jersey

State (no judge based on State Supreme Court decision) State

P

Yes or no

New Mexico

State

M

Yes or no

North Carolina North Dakota Ohio Oklahoma Oregon Rhode Island

Local

P

Yes or no

Election (can be same or different day) Both election and appointment Appointment

State Local Local State State

New election Yes or no New election Yes or no Yes or no

South Dakotad Local Tennesseee Local

P P P P P/malfeasance mix M P

Texas Virginia

P M

New election Yes or no, but Chattanooga is new election Yes or no Uses both Trial New election

M P P P

Yes or no Yes or no New election New election

Washington West Virginia Wisconsinf Wyomingf

Local Local + recall trial for state officials and some local State Local State Local (limited to commission form of government)

New election Appointment New election Appointment Replacement race—different day New election Both

Appointment Appointment New election New election

Source: Author’s own research No provisions in Delaware, Indiana, Iowa, Kentucky, Mississippi, New Hampshire, New  York, South Carolina, Pennsylvania, Utah and Vermont b Vote total in favour of removal must top the amount won by the candidate in the last election c Candidate can run to replace self d The malfeasance standard includes a “gross partiality” requirement e City Council and School Board recalls require a 66% supermajority for removal f Includes a primary a

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

77

Right now, the recall is currently in a boom phase, with numerous countries that do not have recalls starting to consider adopting it, including Australia (Nicholls 2011), Canada (Henton 2016), Germany (Geissel and Jung 2018), several Latin American countries (Welp and Milanese 2018), Russia (Stulov 2012) and the UK (Electoral Commission), with the UK and the British Columbia province of Canada adopting a form of the recall. Numerous jurisdictions in the US have also added a recall law over the last decade, and the recall may be growing in use throughout the country. In this chapter, we shall look at (1) how the recall spread in the US, (2) how the recall has been used in over the years, (3) the controversy surrounding the recall of judges, (4) the technological changes that may ensure that the recall does not effectively disappear in the future and may continue to grow and (5) the practical hurdles that will prevent the recall from becoming a major player in US politics, even in many of the states that possess what this part of what the Los Angeles Times once called “freak legislation” (Los Angeles Times 1911) or the “Grand Bounce” (Sitton 1992).

5.2   US’s Long Flirtation with the Recall As also noted by Matt Qvortrup in Chap. 3 in this volume, the recall in the US goes back to the very beginning of the colonial era, going back to the Massachusetts Bay Colony in 1631 and 1691 (Luce 1930: 488; Thorpe 1906: 1879). The earliest forms of the recall were different in that they gave local government bodies the right to recall the officials they sent as a representative to another, more senior body. During the Revolutionary War Era, both Pennsylvania and Vermont allowed for the recall of members of the Continental Congress (Thorpe 1906: 3085, 6578). The Articles of Confederation, the precursor to the Constitution, also had a recall provision (Article V), though John Lansing noted that the recall was never exercised throughout the brief Confederation era (Elliot 1891: 2:299–300). The recall was also included in the original draft of the Constitution. The Virginia or Randolph Plan contained a very modern-style recall proposal. It would have allowed the recall for the members of the first house of the legislature, who were directly elected by the people (Elliot 1891: 5:127). The recall provision did not last long at the convention, as it was struck out by a vote on June 12. From Madison’s record, the recall was

78 

J. SPIVAK

only brought up again at the end of the convention, as delegates Elbridge Gerry and Luther Martin cited the lack of a recall provision as one of his proofs that the convention exceeded its mandate (Elliot 1891: 5:422). However, the recall became a powerful component of the argument of the anti-Federalists, who were the opponents to the adoption of the Constitution. The Constitution required that either legislatures or special conventions in 9 of the 13 states vote in favour of the document before it would go into force. Luther Martin stressed the lack of the recall, and the freedom from popular control that this absence represented, as a reason to reject the document (Elliot 1891: 1:361). Martin’s idea was taken up by other opponents, though their plan for a recall eventually focused on senators, who at the time were elected by the state legislature. The topic was heatedly debated at the New York Ratifying Convention and came up at the Massachusetts and Rhode Island Conventions. Alexander Hamilton, strongly opposing a recall law, noted that it “will render the senator a slave to all the capricious humors among the people” (NY State Convention 1788: 63, Hamilton 1962: 5:69). After the adoption of the Constitution, the recall faded into the background. The legislature in Virginia attempted to bring the topic up as a constitutional amendment in 1803 and 1808. The 1808 amendment was met by resolutions of disapproval from six states (Luce 1930: 490). There was also an idea, known as Instruction, which held that Senators should resign from office if they disobeyed direct orders from their state legislature (Farmer 2001: 8; Whitaker 1992: 5). This idea had significant support, but it never got on the books. The recall effectively lay dormant for a century. It began to reappear on the scene in 1892 and 1896, when the platforms of two political parties, the Socialist Labor Party and the Populist Party, touted support for the idea of what they called the “imperative mandate” (Zimmerman 1997: 9). But the recall failed to make headway. The National Direct Legislation League refused to support the recall as part of their plans because, ­according to President Eltweed Pomeroy, the recall “involves the personal element in a manner allowing for reprisals and political revenge” (Sitton 1992: 35–36). When states started adopting other direct democracy provisions, starting with South Dakota in 1898, the recall was nowhere to be seen. Eight states adopted direct democracy legislation before the first one adopted the recall in 1908. The recall took off thanks to the efforts of Dr John Randolph Haynes in Los Angeles. Though several small localities are said to have adopted

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

79

the recall prior to 1900, they are unknown (Cronin 1989: 130; Farmer 2001: 11). Haynes first set his objectives on the city of Los Angeles, which had spent several years attempting to draft a new charter. He managed to be appointed in 1900 to the 15-member committee for the revision of the Los Angeles charter, where he proposed a recall. Los Angeles adopted it as part of its new charter in 1903. In 1904, Councilman James Davenport was voted out of office. In 1909, Mayor A.C. Harper was tossed out. The recall started taking off in different states. Oregon became the first state to adopt it in 1908, followed by California in 1911; Arizona, Colorado, Nevada and Washington in 1912; Michigan in either 1908 or 1913; Kansas and Louisiana in 1914; North Dakota in 1920; Wisconsin in 1926; and Idaho in 1933. Following Idaho, there was a 26-year delay for the next state to take up the recall, Alaska in 1959. Since then only seven states have adopted any form of the recall. However, when given a chance, voters usually adopt the law overwhelmingly. In 1993, New Jersey voters passed a recall law with almost 75% in favour; in 1996 Minnesota voters pass the recall law with almost 90% in favour. While there was initially significant attention paid to the recall, the device soon lost its lustre. In 1921, the Governor of North Dakota, the Attorney General and the Agriculture Commissioner were ousted in recall votes. No other state-wide official faced enough signatures to be placed on the ballot for another 67 years, and no recall against a Governor took place for another 82 years. While there were three recalls in the first four years of its existence in California, it quickly lost popularity. Between 1915 and 1971, there were only two state legislative recalls in the whole country—in 1933 in Oregon and in 1935 in Wisconsin. Recalls simply faded from the political agenda.

5.3   The Comeback of the Recall and Its Current Use The California gubernatorial recall of 2003 and a subsequent one against Wisconsin Governor Scott Walker in 2012 were just the most notable episodes in an upsurge in the use of the recall over recent decades. On the state level, the recall has been used in total against 39 state legislators in eight states. Six other state-level officials have faced recalls—three Governors, one Lieutenant Governor, one Attorney General and one Commissioner of Agriculture. Additionally, a recall was set to take place against Arizona Governor Evan Mecham in 1988, before he was

80 

J. SPIVAK

Table 5.2  Total number of recalls that made the ballot or led to official resignation in the US, 2011–2018a Year

Recalls or resignation in total

Removal

Survived vote

Resigned

2018 2017 2016 2015 2014 2013 2012 2011

150 102b 119 109 126 107 166 151

85 39 60 66 61 50 82 78

37 29 42 28 45 35 58 66

28 32 17 15 20 22 26 7

Source: Author’s own research The list does not include Native American tribes Two officials were qualified for recalls, but no one ran against them so the election was cancelled. One official did not contest the recall and instead ran for mayor and won a

b

impeached. Two of the Governors, the Lieutenant Governor and 32 of those legislative recalls took place since 1981. Perhaps more importantly, between 2011, when I started compiling yearly data on the recall, and 2018, at least 861 recalls have taken place, with 521 officials losing their jobs and 340 surviving the vote. Another 167 officials have resigned in the face of a recall.1 Previous data, based possibly on nothing but conjecture, claimed that between 4000 and 5000 recalls had taken place in the US, between 1904 and 1989 (Cronin 1989: 142). That would total far less than the more than 100 per year that we are now averaging. A different study by Professor Charles M. Price claimed over 1180 recalls between 1970 and 1979 (Zimmerman 1997: 106), which seems much closer to today’s numbers (Table 5.2). The recall is not just being used regularly. It is also constantly being threatened. Between two-thirds and three-quarters of all recall attempts never get to the ballot. Most of these failed recall attempts simply wither on the vine, with signatures never handed in for review. In California, there were 31 attempted recalls of the Governor before one got on the ballot. But the threat of the recall works. As the above table shows, many officials resign under threat of the recall. Oregon’s Governor John Kitzhaber was one such official. Oregon does not have an impeachment 1  Additionally, at least eight officials who were facing recall attempts in the last two years were thrown out of office by a legislative or executive body.

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

81

procedure. Kitzhaber could have stayed on indefinitely once a scandal regarding his girlfriend came to light, but the very real threat of a recall forced him out of office (Johnson 2015). Is the Recall Designed for Political Use or to Punish Malfeasance? The recall is generally thought of by members of the public and editorial boards as being designed for use against crimes, egregious situations and other malfeasance in office. For example, one editorial states: “Recalls should be reserved for situations where elected officials are guilty of serious offenses—financial misconduct, ethical improprieties, continued offensive behavior, and unwillingness to carry out one’s duties as an elected officials and the like” (Norfolk 2015). Another columnist in California states: “As a rule, California’s institution of recall should be reserved to remove elected officials who are criminally corrupt or mentally ill. Occasionally, a recall is mounted when the motivation is about policy or the officeholder’s behavior” (Spotswood 2019). As we will see below in the discussion of political recalls, this is for the most part a misreading of the law and its effect. Recalls were not designed just for egregious conduct. California Governor Hiram Johnson wanted the recall (and other direct democracy weapons) to “kick the Southern Pacific out of politics” (Bird and Ryan 1930: 42)—a very clear policy goal. And recalls are usually about political, though not necessarily partisan, or policy fights. The idea that recalls are just supposed to be about corruption-­ type issues continues to permeate any public discussion of the recall and is the basis of most defence campaigns by targeted officials. But they are not used that way. Among the state legislative recalls, I would say only four can truly be considered to have been launched over clear criminal or ­malfeasance action, in California in 1913, Oregon in 1985 and 1988 and Wisconsin in 2003. The rest are clearly about political or policy reasons. A look below at the use of the recall in each state shows that the overwhelming majority of recalls take place in political recall states. Partisan Breakdown of Recalls Many recalls take place for officially nonpartisan local-level position, aimed at school board members, mayors, council members, sheriffs, county clerks and other positions of mainly local interest. Local towns are less likely to be evenly divided on partisan lines—often one political party will

82 

J. SPIVAK

dominate the local jurisdiction. But recalls still occur with marked frequency. These recalls are usually launched over localized subjects, be it a policy disagreement, personal vendetta or a desire to capture an office. They often have nothing to do with national or even state politics, and there is not much information to be gleaned by examining the sparse partisan record. The partisanship in recalls does appear more strongly on the state level. Nearly every state legislator in the country belongs to either the Democratic or Republican Party, and which party controls the leadership of the legislature is of critical importance. The result is that recalls are likely to have a strong partisan lean on the state level. This is also a reason why there are few state legislative recalls. Professor Lawrence Sych, in his paper examining the results of these recalls, notes “Signing a petition for recall is often more than repudiating a single state official—it is also a rejection of his or her party” (Sych 1996). The result is these recalls are not only rare, but they appear to take place when they may have the maximum impact—namely, in a closely divided state legislature with an either explicit or implicit goal of flipping control of the body to the other party—or they target high-ranking leaders of one party. The 3 in Colorado in 2013, the 13 recalls in Wisconsin in 2011 and 2012, the 1 in Wisconsin in 1996, the 3 in California in 1995, the 2 in Michigan in 1983 and the 1 in Washington in 1981 all either switched the chamber or had the potential to do so. Three threatened recalls in Nevada in 2018 and Colorado in 2019 all had a very blatant similar aim. Two other recalls, in California in 2008 and 2018, were launched to either gain or strip one party of a veto-proof majority. Three other recalls, in Arizona in 2011, in Michigan in 2008 and California in 1994, specifically targeted the leaders of either the Senate or the House.2 So 28 out of 39 recalls had a clear state-wide partisan gain from the recall. Since the recalls take place over the course of more than a century, as well as in different states, it is of limited use to consider which party benefits more. However, with this caveat, we shall still look at this data. According to my research, more Republicans have faced recalls in state legislative elections. This data is very skewed by the Wisconsin recalls of 2011 and 2012, when nine Republicans faced recall votes, as did four Democrats. Including those, 22 of the recalls were against Republicans, 2  House leaders also faced recalls in California in 1995 and in Colorado in 2013, though those all had other policy as well.

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

83

12 of whom lost their seats. One fact that limits the partisan discussion is that for the most part, the Republican legislators were not replaced by Democrats. In only four cases, Wisconsin in 1996, and three Wisconsin Senators in 2011 and 2012, were the recall replacements Democrats. The other eight were replaced by fellow Republicans—though Oregon law mandates an appointment replacement procedure and requires that the replacements be from the same party. Thirteen Democrats have faced recalls, with seven being removed. Five of them were replaced by Republicans. Additionally, two party flippers who switched from Republican to Independent faced recalls. Both were removed and replaced by Republicans. The fact that so many officials are replaced by members of their own party suggests that recalls should not be considered an effective method for partisan warfare. Special Elections Versus General Elections Recalls can take place on either the same date as a previously scheduled election (either a general or primary) or on a special election date. A number of states require recalls to take place on the same date as a general election.3 Almost all jurisdictions require recalls that qualify for the ballot near the date of regularly scheduled elections to take place on the general election date. Special elections have come under scrutiny in the US as the losing parties question whether special elections are an accurate look at the wider electorate or are instead a kind of race that gives a special advantage to challengers in a what is typically very low-turnout balloting. Despite some rare examples of high-turnout special elections, such as the gubernatorial recalls in California in 2003 and Wisconsin in 2012 and the 2017 race to fill a vacant Alabama US Senate seat, special elections generally take place in the shadows of the news cycle. Much as with the arguments that special elections are a unique, atypical portrait of the electorate, there is a question of whether stand-alone recalls would be more likely to result in an incumbent losing office than those that take place on a regularly scheduled election date. According to this theory, the fact that it is a special election means that the more motivated, “angrier” party will have an easier time turning out its base—which would

3  In 2012, Michigan changed its laws to remove special election recalls and require all recalls to take place on either a primary or general election date.

84 

J. SPIVAK

Table 5.3  US recalls by election date Recall type, 2011–2018

Recalls

Removal

Survived vote

Special election date General or primary

529 332

310 211

219 121

Percentage 58% 64%

explain why the out-of-office party has done better in many special elections. However, the statistics suggest otherwise. Sitting officials, it turns out, are not disproportionately hurt—and sometimes may actually be helped— when a recall election is held. Moreover, the more motivated, angrier voters do not seem to skew special elections in an obviously predictable way. Of the 861 recalls that took place in the US from 2011 to 2018, 529 were stand-alone special elections, where 58% of the incumbent officials lost their vote. But the remaining 332 recalls, which took place on the same day as a general or a primary election, saw a 64% removal rate. Every recall is different—though despite perception, only a small few are about corruption issues—but apparently having a special election where the presumably angrier group of voters is motivated to turn out to vote for a single issue may not help. On the whole, it seems that special election recall may not always benefit the more motivated party (Table 5.3).

5.4   Recalls of Judges The most controversial and least used recall is those targeting elected judges. Unlike many countries and unlike the federal government, most US states elect judges to office. The wisdom of electing judges has been a source of significant debate in the US. Similarly, the question of whether judges should be subject to the recall has been bitterly fought over. California’s adoption debate hinged on the question of whether judges should be legally included or exempt from the recall law. The Progressive proponents of the recall argued that it was critically important to include the judiciary. The State Supreme Court had found 57 of cases in favour of the Southern Pacific railroad, and Hayne believed that the “Southern Pacific’s domination of no other arm of government was as complete or pernicious” (Mowry 1951: 14). A late-breaking State Supreme Court scandal during the debate appeared to tip the scales in favour of the recall. On the federal front, the recall of judges became a national issue in the 1912 election. Former President Theodore Roosevelt was a leading pro-

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

85

ponent of the recall of judges, as well as a separate provision that has fallen out of the discussion after 1930, the recall of judicial decisions.4 Roosevelt’s support of the recall of judges ran smack into Republican Party orthodoxy, which was one of the issues that helped precipitate a split in the party (New York Times 1912). President William Howard Taft, who later served as Chief Justice of the US Supreme Court, was the leading opponent of the recall. Taft vetoed the Arizona Constitution in 1912 specifically because it included a recall of judges’ provision (Dodd 1911: 83). Following his defeat in 1912, Taft wrote a book attacking direct democracy, and especially the recall, noting that “Could there be a system more ingeniously devised to subject judges to momentary gusts of popular passion than this?” (Taft 1913: 171). After this debate, the recall of judges never took off on the state level. While 2018 saw a successful recall of Santa Clara Judge Aaron Persky in California, this was the first recall of a judge to get on the ballot anywhere in the US since a recall in 1982 in Wisconsin. The last time there was a successful recall and removal of a judge in the US before Persky was in 1977 in Wisconsin. California had many attempts for a recall of a judge before Persky, including five alone against liberal State Supreme Court Chief Judge Rose Bird in the 1980s,5 but they all failed. Prior to the Persky recall, no California judge had faced a recall vote since 1932. The lack of interest in the recall of judges is also apparently how often states leave it out. Of the 19 states to adopt the recall law, 9 of them have a carved-out exception for judges. And in Nevada, one of the ten states that allowed such a recall, the State Supreme Court threw out the portion of its recall law that could be used against judges (Ferrara 2017). As the US has recently seen an upsurge in interest in judicial elections (Barnes 2007) coupled with the success of the Persky recall, there is some reason to believe that the recall of judges may be a potential growth area. However, the strict limits put on by states, and the limited partisan benefit of recalling lower-level judges (who rarely rule on political issues), will put a limit on how significant an impact the recall can have against judges. 4  The recall of judicial decisions was an attempt to allow the populace to prevent judges from declaring acts unconstitutional. According to Bird and Ryan, after the Supreme Court of Colorado declared it void in 1921, the issue disappeared. 5  Bird and fellow Supreme Court Justices Cruz Reynoso and Joseph Grodin were all eventually ousted in a mandatory retention election. Many observers have confused judicial retention elections with recalls, but the retention elections are simply regular elections.

86 

J. SPIVAK

5.5   Potential of Technological Changes Responsible for the Growth of the Recall While a sharpened and more divisive political environment is frequently credited or blamed anytime there is an increased use in the recall, there are real questions as to whether this matters in the overall use of the recall. A politically charged atmosphere is nothing new in American politics. During the 1930s or the 1960s, the political divide was very stark. And yet the recall does not appear to have been used in great force during those decades. In the 1930s, two state legislators faced recall votes and none in the 1960s. If recalls are expanding, there are other potential factors that would explain its growth. Simple awareness may be an important factor. The Gray Davis recall was a major political story as was the Scott Walker recall. The recall has penetrated the consciousness of the nation enough that two popular comedy shows, The Simpsons and Parks and Recreations had episodes devoted to recalls. The fact that the recall is more well known may make voters more willing to turn to its device. Another important change that may be the critical piece in the recall puzzle is the revolutionary changes that have taken place in technology. One of the basic challenges facing recalls is the financial cost. Historically, the recall has been more of a threat than a weapon because of this barrier. Alerting fellow voters to a campaign, gathering signatures, defending them in court and running a campaign against an entrenched incumbent are all expensive and time-consuming propositions. Raising funds for these actions was extremely difficult. In the past, fundraising for these battles was strictly a local endeavour. Unless a deep-pocketed backer pushed the recall, money would be limited. But that is no longer the case. Fundraising on a national level is now often a matter of good marketing. Since the Internet has become ubiquitous, rumour campaigns and directed e-mails have become potent tools. Local news stories now reach a far wider audience. Thirty years ago, few in California or Florida might have closely followed the goings-on in the Wisconsin Legislature. The Internet and 24-hour cable news channels have changed all that. Consider the case of Congressman Joe Wilson effect. The little-known South Carolina congressman screamed “You lie!” during President Barack Obama’s address on health-care reform to a joint session of Congress in 2009. In the past, Wilson would have been an embarrassment to the

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

87

Republicans and would have been pushed into the background. Instead, he was flooded with unsolicited donations. Despite being in a relatively safe seat, he raised $1.8  million in one week. His opponent raked in $1.6 million (Bailey 2009). The Wisconsin gubernatorial and legislative recalls show how financing of recalls can change almost instantly. The Wisconsin recalls saw spending by both sides totalling at least $137.5  million—that we know of—in a state of less than 6 million people (Wisconsin Democracy Campaign 2012). On a similar front, emerging communications technology has made organizing and running signature-gathering efforts significantly easier. Basic computing power, cell phones and everyday technological changes have on the electoral process. For instance, the power of the original killer app, the spreadsheet, has transformed campaigning. It allows campaigns to compile lists of likely supporters and opponents, with an ability to parse great sets of data easily. It is a programme available for free to any person with a computer with Internet access. It is also rarely mentioned as a campaign tool. So too, cell phones let you keep in touch with the troops in the field. Items as simple as colour printers and word processing programmes have helped lower the cost of a political campaign and make it more accessible to grassroots voters. The recent growth of the recall has also taken place at the same time as the birth of social networks such as Twitter and Facebook and sites like YouTube. The social networks and YouTube allow an easy way to spread the word of a recall without any filtering and without the need to convince local reporters that the campaign is serious. One of the Colorado State Senate recalls in 2013, against State Senator Angela Giron, highlighted the value of technological changes. The petitioners used a new technology available from the website and smartphones of Colorado’s Secretary of State to check the eligibility of petition signers. The result was that the petition against Giron had an extremely low 6% rejection rate. As a comparison, the recall against Giron’s fellow State Senator, Majority Leader John Morse, witnessed a 37.5% reject rate (Lee 2013). There is another non-technology reason for the recall’s expansion: in some jurisdictions, apathy, not anger, has helped push recalls to the fore. With few exceptions, voter turnout has dropped precipitously in the past several decades. In these states and localities, the number of signatures needed to qualify a recall for the ballot is directly tied to the number of voters in the previous election. So a decrease in turnout means it takes a

88 

J. SPIVAK

smaller portion of signatures from the overall electorate to qualify a recall. The fact that California experienced the lowest voter turnout in its history in the 2002 gubernatorial election helped make the recall of Gov. Gray Davis in 2003 that much easier to get on the ballot (Spivak 2004).

5.6   Barriers for the Expansion of the Recall Even though these numbers suggest a growth spurt, the US has strict limits on the use of the recall. As pointed out in the chart above, not every 1 of the 50 states allows its use, and the majority of the states restrict recalls to lower-level officials. While 39 states allow the recall on the lower level, only 19 allow it for some or all state-level officials. Nine states allow it for federal officials such as US Senators and Congressional Representatives. However, it is likely that a recall against a federal official will be thrown out by the courts. A 1994 footnote in the 1994 Supreme Court term limits decision (U.S. Term Limits v. Thornton n.d.: 890) and a New Jersey State Supreme Court ruling (the Committee to recall Robert Menendez 2010) both reject federal-level recalls. Based on the term limits decision, and on all subsequent actions by federal- and state-level judges, I believe that it is likely that if a recall of a federal-level official got to the ballot, the Supreme Court as currently composed would overwhelmingly if not unanimously find it unconstitutional. Even within the states that have the recall, there is a significant divide that ensures that many jurisdictions would not be able to use the device. There are two broad categories of states with recall laws. Eleven states have what is called a political recall law. This means that an official can face a recall for almost any reason. There is no need to prove a cause of action, such as criminal behaviour, for the recall to move forward. Essentially, all famous recalls in the US have taken place under these laws. The 11 states are California, Wisconsin, Michigan, Colorado, Arizona, Oregon, New Jersey, Idaho, North Dakota, Nevada and Louisiana. The other eight states, Kansas, Minnesota, Georgia, Alaska, Washington, Montana, Rhode Island and Illinois, have a form of a recall called either a malfeasance standard recall or a judicial recall (not to be confused with a recall of a judge). Additionally, Virginia has something called a recall trial that will be discussed below. For recalls to take place in these states, the petition must show a violation of either law or in some cases of incompetence. It is usually up to an election commission or a court to decide whether this standard is met. If the commission finds that the standard is

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

89

met, the targeted official can appeal to a court to overturn the commission’s ruling. The result is that the process not only must meet a procedural hurdle, but it also must take a significant outlay of time and money on the part of the petitioner to get the recall onto the ballot. The recall laws in malfeasance standard states vary widely. In Illinois, it is only the Governor who is covered by recall, and even then the legislature plays a significant role in getting it on the ballot. In Virginia, there is no election but rather a recall petition triggers a judicial hearing. The judge decides whether to kick out the official. But all require an agency or the courts to hold that a specific, statutorily delineated bad act was performed by the elected official. The practical impact is that few recalls take place in the malfeasance standard states. Of the 39 state legislative recalls, only 1, against Washington State Senator Peter Von Reichbauer in 1981 over his decision to switch parties, made the ballot. The other 38, as well as the ones against state-­ wide officials, were all in political recall states.6 Even in states with political recall laws, there are numerous requirements that make a recall almost impossible to get off the ground. States have widely varying signature requirements to get recall on the ballot. Most of the states that see frequent use of the recall, such as Michigan, Wisconsin and Oregon, require petitioners to get signatures from voters totalling 25% of the voter turnout for that seat in the previous election. The difference between the voter turnout and the registered voter requirement can be thousands of signatures for state legislative and mayor and city council positions. By comparison, New Jersey requires 25% of registered voters to sign, and the state rarely has a recall get on the ballot. Louisiana requires signatures from one-third of registered voters in the affected jurisdiction to get a recall on the ballot. The result is that over the last five years, six recalls have gone to the voters in all of Louisiana. All of these have been in small jurisdictions, where the amount of signatures is a low number. The state has had a plethora of corruption scandals, but the laws are too onerous to get recalls on the ballot. By comparison, in California, a state whose population was eight and a half times the size of 6  State legislative recalls have taken places against 1 Arizona State Senator; a California Governor and 9 state legislators; 2 Colorado Senators; 2 Idaho legislators; 4 Michigan legislators; 3 Oregon legislators; the Governor, Attorney General and Commissioner of Agriculture in North Dakota; 16 Wisconsin state legislators as well as the Governor and Lieutenant Governor; and 1 Washington State Senator.

90 

J. SPIVAK

Louisiana, petitioners needed less total signatures (approximately 893,000) to get a recall of Davis on the ballot than petitioners would have needed to get a recall against a Louisiana Governor to the voters. The signature requirements, and the cost inherent in gathering signatures, make most recalls very difficult to get off the ground. This isn’t just for the most famous offices, such as Governor or state legislature. For example, in 2012 a California Moreno Valley school board member was indicted on attempted murder, pimping and insurance fraud charges for running a sex-for-money operation (he was convicted and is serving 14 years). Despite the lurid charges, the petitioners could not get the over 9000 signatures needed to get a recall on the ballot (Hines 2012). Elected officials will also attempt to stop recalls from occurring or throw up new barriers to their success. Following the Walker recall next door in Wisconsin in 2012 and a state House recall in 2011, Michigan changed its recall law to make it significantly harder to recall regular officials and made it virtually worthless to attempt a partisan recall of the Governor (who would now be replaced simply by the Lieutenant Governor) (Lawler 2016; Associated Press 2013). During a recall in 2018, California’s legislature changed the recall law to allow voters to remove their signatures from the petitions, a move designed to limit the power of the recall. In numerous jurisdictions, recalls have been delayed or killed by a government body’s refusal to schedule a recall on the ballot (Spivak 2014). A town called Keyser, West Virginia, took the extreme step of declaring that if a recall gets to the ballot and the official survives, the signers of the recall petition are liable for the $10,000 cost of holding the election, including a paid vacation day for workers (Blaisdell 2012). State officials claimed that the law was invalid, but even proving that would have required a lawsuit, presumably paid by those same voters who would have to pay the cost of the recall. The Keyser recall never got to the ballot (because signers did not list both their age and length of residency). Once again, these laws seem very successful in limiting the recall.

5.7   Conclusion From its early origins in the US colonial period, the recall has been a presence in the American political system. It was used to attempt to answer the question of whether an elected official is a delegate, serving as a representative of the people who should be subject to removal when they moved beyond the opinions of their constituents, or a trustee who is sought out

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

91

for the expertise and opinion. Throughout much of US history, the recall has been in eclipse, either not even in the game or as a weapon that is used as a spoken threat. In recent years, we have witnessed a change in the use of recalls in the US. Instead of a little noticed weapon, it has become a prime arsenal in the political world. The recalls in California, Wisconsin, Arizona and Colorado showed that the recall can become a central part of the political discussion. Thanks to technology and a more divisive political environment, the recall has been an easy-to-use part of the political landscape. However, even with this boost, the recall will run into natural barriers that can limit its growth. As in Michigan, and as threatened in Wisconsin, politicians have shown a willingness to push back against the use of recalls. More importantly, the existing laws prevent a challenge to the expansion of the recall. Since it cannot be used against federal officials, and since most states don’t allow the recall to be used against state-level officials, the recall can only affect change in a small grouping of states. Whether citizens in other states start clamouring for the recall and whether its use may expand are still to be determined.

References Associated Press. (2013, July 28). Michigan Law Could Lead to Fewer Recall Elections. Associated Press. Bailey, H. (2009, September 19). How Joe Wilson’s Heckle Became a Campaign Cash Cow. Newsweek. Barnes, R. (2007, October 28). Judicial Races Now Rife with Politics. Washington Post. Bird, F.  L., & Ryan, F.  M. (1930). The Recall of Public Officers. New  York: Macmillan. Blaisdell, E. (2012, May 17). Petition for Keyser Recall Election Given to City Official. Cumberland Times-News. Cronin, T. (1989). Direct Democracy: The Politics of Initiative, Referendum and Recall. Cambridge: Harvard University Press. Dodd, W.  F. (1911, December). The Recall and the Political Responsibility of Judges. Michigan Law Review. Elliot, J. (Ed.). (1891). The Debates in the Several State Conventions on the Adoption of the Federal Constitution by the General Convention at Philadelphia in 1787. Philadelphia: J.J. Lippincott Company. Farmer, R. (2001). Power to the People: The Progressive Movement for the Recall, 1890s to 1920. New England Journal of History, 57(Winter), 59–83.

92 

J. SPIVAK

Ferrara, D. (2017, April 13). Elected Judges Cannot Be Recalled, Nevada Supreme Court Rules. Las Vegas Review Journal. Geissel, B., & Jung, S. (2018). Recall in Germany: Explaining the Use of a Local Democratic Innovation. Democratization, 25(8), 1358–1378. Hamilton, A. (1962). The Papers of Alexander Hamilton (H.  Syrett, Ed.). New York: Columbia University Press, 5:69. Henton, D. (2016, March 14). Wildrose Takes Another Run at MLA Recall with Private Member’s Bill. Calgary Herald. Hines, L. (2012, October 30). Moreno Valley: School Board Member Rios Won’t Face Recall. Press Enterprise. Johnson, K. (2015, February 11). Love and Politics Collide as Scandals Plague Oregon’s Fourth-Term Governor. New York Times. Lawler, E. (2016, February 9). Recalling Gov. Rick Snyder Is Harder than Ever Under 2012 Law He Signed. MLive. Lee, K. (2013, July 27). Colorado Online Voter System Can Play Role in Future Recall Elections. Denver Post. Los Angeles Times Editorial. (1911, October 10). Luce, R. (1930). Legislative Principles. Boston: Riverside Press. Mowry, G. (1951). The California Progressives. Berkeley: University of California Press. New York State Convention. (1788). The Debates and Proceedings of the Convention of the State of New York (p. 63). New York: Francis Childs. New York Times. (1912, April 5). Roosevelt Assailed in House by Gardner. Nicholls, S. (2011, December 14). Let Voters Dump Bad Government, Experts Decide. Sydney Morning Herald. Norfolk Daily News Editorial. (2015, November 7). Recall Isn’t the Proper Response. Sitton, T. (1992). John Randolph Haynes: California Progressive (p. 85). Stanford: Stanford University Press. Spivak, J. (2004). California’s Recall: Adoption of the “Grand Bounce” for Elected Officials. California History. Spivak, J. (2014, June 10). Elections Laws That Prevent Elections. Reuters. Spotswood, D. (2019, June 15). Glickman Could Avoid Recall Effort by Running Again. Marin Independent Journal. Stulov, M. (2012, June 28). City Approves Reinstating Mayoral Elections. Moscow Times. Sych, L. (1996, October). State Recall Elections: What Explains Their Outcomes? Comparative State Politics. Taft, W. H. (1913). Popular Government: Its Essence, Its Permanence and Its Perils. New Haven: Yale University Press. The Committee to Recall Robert Menendez from the Office of Senator v. United States Senator Robert Menendez, 7 A3d 720. (2010)

5  RECALL ELECTIONS IN THE US: ITS LONG PAST AND UNCERTAIN FUTURE 

93

The Electoral Commission, Introduction to the Recall of MPs Act 2015. Thorpe, F. (Ed.). (1906). The Federal and State Constitutions Colonial Charters, and Other Organic Laws of the States Territories and Colonies Now or Heretofore Forming The United States of America. Washington, DC: Government Printing Press. U.S. Term Limits v. Thornton, 514 US 779. Welp, Y. (2018). Recall Referendum Around the World: Origins, Institutional Designs and Current Debates. In L. Morel & M. Qvortrup (Eds.), Compendium on Direct Democracy. Routledge. Welp, Y., & Milanese, J. P. (2018). Playing by the Rules of the Game. The Partisan Use of Recall Referendums in Colombia. Democratization, 25(8), 1379–1396. Whitaker, L. P. (coord.). (2012, January 5). Recall of Legislators and the Removal of Members of Congress from Office. Congressional Research Service. Whitehead, L. (2018). The Recall of Elected Office Holders. The Growing Incidence of a Venerable, but Overlooked, Democratic Institution. Democratization, 25(8), 1341–1357. Wisconsin Democracy Campaign. (2012, July 31). Recall Race for Governor Cost $81 Million: Fifteen Recall Races in 2011 and 2012 Cost $137 Million. Zimmerman, J. (1997). The Recall: Tribunal of the People. Westport: Praeger Publishers.

CHAPTER 6

Recall in Japan as a Measure of Vertical Accountability Mitsuhiko Okamoto and Uwe Serdült

6.1   Introduction Today in Japan the people can participate in politics between elections via petitions and referendums. Contrary to widespread belief, on the local level, there is direct democratic practice on a considerable scale (see Steiner 1965). When the government reformed the local political institutions in 1946, the so-called direct demand system was introduced. A direct demand is a mechanism for citizen participation, in some cases leading to a referendum vote, in some cases not. It allows citizens to collect signatures so that they can trigger procedures to enact, revise or abolish local ordinances (CLAIR 2019: 18), perform audits of government activities, dissolve the assembly and remove assembly members or the chief executive of M. Okamoto School of Political Science and Economics, Tokai University, Kanagawa, Japan e-mail: [email protected] U. Serdült (*) Center for Democracy Studies Aarau (ZDA), University of Zurich, Aarau, Switzerland College of Information Science and Engineering, Ritsumeikan University, Kusatsu, Japan e-mail: [email protected] © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_6

95

96 

M. OKAMOTO AND U. SERDÜLT

a constituency, which in most cases is the directly elected mayor (see Kobori 2009: 19–24, Numata 2006; Serdült and Welp 2017: 145; Shimizutani 2010; Steiner 1965: 464; Tsujiyama 2000: 28; Vosse 2000: 43–46). The bifurcation in Japan’s democratisation path regarding political participation took place after World War II.  Before 1947, under the Constitution of the Empire of Japan, the people had only very limited possibilities to take part in politics, either at the national or the local level. Subnational government was almost completely under the control of the central bureaucracy in the capital. The people could not directly vote either for a governor in the prefecture or for a mayor in a municipality. The governor was a bureaucrat appointed by the Ministry of Home Affairs, and the mayor was selected by the municipal assembly. Male residents could only elect the members of the assemblies, either at the prefectural or on the local level. Moreover, the Minister of Home Affairs had the power to dismiss a local assembly and remove a member of an assembly or a mayor. In our context, it is worth noting this latter point because it is this very competence of dismissal on the local level that turned into the recall procedures defined in the Local Autonomy Law (LAL) in 1947 (Government of Japan 2010). Following the defeat of Japan in World War II, the government was required to democratise. However, the Japanese bureaucracy intended to stick to the old political regime of the Meiji Constitution, in short, a centralised system with a tight control over local government. In particular, the Ministry of Home Affairs was hoping to keep the power to dismiss political bodies on the local level. The impetus for more direct citizen participation, putting the recall into the hands of the people, originated in the Diet when the LAL was debated. Members of parliament—especially those from the Social Democrats and the National Cooperative Party (Kokumin Kyodoto)—pointed out that the proposal for the new LAL made by the bureaucrats was equivalent to preservation of the regime under the Meiji Constitution before World War II. At a later reading, the Ministry then adapted the proposal. Eventually, this change was reinforced behind the scene by the administrative staff of the General Headquarters (GHQ) of General MacArthur. A direct transfer of US democratic institutions was, however, not on the agenda. If the GHQ had enforced such a system in Japan, it would have introduced initiatives and referendums at the local level as well, not only the recall. Furthermore,

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

97

elements such as the citizen-initiated audit of government activities as well as the ­dissolution of the assembly are not typical for the USA (Matsuno 1981; Steiner 1965: 451). Basically then, an already existing procedure and competence in the hands of the Ministry of Home Affairs—omnipotent on local affairs—was delegated to the citizenry. The transfer of three recall procedures into the hands of the citizens can thus be attributed to favourable domestic conditions for change at a major inflection point in Japanese political history. Additional direct democratic rights at the local level (such as an initiative for the enactment, revision and abolishment of local ordinances leading to policy referendums) were hardly discussed at the time.

6.2   Theoretical Framework and Working Hypotheses The recall of elected office holders is a measure of vertical accountability. Whereas vertical accountability implies the involvement of citizens or sometimes representatives acting on their behalf, horizontal accountability can be understood in the sense of checks and balances between ruling agents beyond the traditional branches of government (e.g. Central Banks, independent state agencies, etc.). With recall as a counterpower in the hands of the citizens, rulers can be sanctioned and directly voted out of their position  (Whitehead 2018). As Schmitter has suggested, the presence of the recall as a credible threat should make its application almost obsolete because office holders internalise expectations of those they are ruling (Schmitter 2005: 19–23). From a principal-agent perspective, vertical measures of accountability are desirable in general. Moreno et al. go so far as to claim that the fact that authority may be withdrawn from the principle by the agent is the very essence of accountability (Moreno et al. 2003: 83). The primary function of the recall is one of a last resort in the case of obvious mismanagement or scandals produced by political elites. Properly designed and not too easy to trigger, it can operate as a safety valve to prevent revolutions. In Switzerland, as well as in Peru, recall was also introduced as a measure to head off violence (Serdült 2015; Welp 2016). Furthermore, recall procedures can help to overcome institutional deadlock, particularly in the case of irreconcilable conflicts between the legislative and the executive, thus providing an option for a fresh start. However,

98 

M. OKAMOTO AND U. SERDÜLT

there are also unwanted effects of the recall. Smaller parties, runners-up, single-issue parties or parties that are almost equally strong as the ­incumbent can all be tempted to use the recall for less noble purposes, for example, to get public attention via media and ultimately to maximise votes in a future election. A further option of a perverted use of the recall is its use against unwanted policies rather than to sanction misconduct by the political authority per se. In the absence of alternative outlets such as policy referendums, and given the sheer availability of recall provisions, this option has been activated in Japan. As a starting point, we posit that the incentive structure of any institution will inevitably lead to primary and secondary unwanted outcomes, especially in the absence of an overall institutional balance (Hug 2004; Steiner 1965). If the institutional design of a constituency mainly draws on the recall as a mechanism for citizen-activated, vertical accountability, the danger for the political elite is positional. The risk for rulers is simply losing their job. When only policy referendums are available, the political elite initially faces reputational dangers. Properly implemented, policy referendum outcomes tend to satisfy the median voter and therefore make it unnecessary for political leaders to step down from office. The worst incumbents face in this case is being forced to shift their policy preferences. However, repeated reputational damage due to lost referendums will increase the risk of subsequent positional loss—usually the next election. But when both mechanisms (the recall and policy referendums) are available at a balanced cost, there is a strong tendency for them to be used in their primary way: recalls, only as a last resort against clear mismanagement and scandals, and policy referendums as punctuated corrections of policy incongruences between the political elite and the demos. In this chapter, we argue that Japan represents a case of imbalanced accountability because of its lack of consequential policy referendums (Tsujiyama 2000: 26). Citizens as well as opposition parties seeking to resist single policy measures, but aware of the non-binding nature of local referendums in Japan, have no other choice than to activate a recall procedure. Switzerland can be considered to represent the counter-factual case. With plenty of mechanisms of direct vertical accountability on the local level at hand to fight against policy measures, financial decisions and even whole budgets, the recall practically became obsolete, or more precisely, it was reduced to its primary function (Serdült 2015, see also Geissel and Jung in this volume).

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

99

6.3   Recall Regulations in Japan Recall of the Municipal Assembly, Individual Assembly Members and the Mayor In Japan, there are three types of recall that can lead to a binding vote by citizens in the respective constituency if a large enough part of the electorate supports the initiative with a signature: (i) the dissolution of the local assembly (Article 76.3 LAL), (ii) the recall of an individual member of the assembly (Article 80.3 LAL) and (iii) the recall of the mayor or governor (Article 81.2 LAL). Japan’s option for citizens to recall parliament is remarkable and exceptional from an international perspective. It is only known to exist for Liechtenstein on the national level and for some Swiss cantons (Serdült 2015; Serdült and Welp 2017). Recalled officials are eligible to run again for the office from which they have just been dismissed. However, there is a one-year grace period after general elections or a previous recall vote during which the same recall procedures cannot be triggered. It appears to be difficult to collect signatures for a recall, especially in the bigger cities. For example, in Kobe in 1999, signatures of more than one third of eligible voters were collected for a request demanding the introduction of a local referendum ordinance to allow for a policy referendum on the construction of a new airport. After the mayor disregarded the demand for a local referendum ordinance, the electorate began the collection of signatures to recall him; however, they did not manage to collect the required number. Only about one third of such recall attempts make it to the stage of an actual vote at the ballots. As we can see in Table  6.1, all three recall options need to fulfil the same signature requirements. The number of signatures depends on the size of the electorate. In the bigger cities, recall initiators need to come up Table 6.1  Direct recall vote options and signature requirements in Japan (since 2012) To recall

Signature collection threshold by the size of the electorate

Assembly Assembly member Mayor

Electorate e  400,000} Electorate e > 800,000: 1/3{400,000} + 1/6{400,000} + 1/8{e > 800,000}

100 

M. OKAMOTO AND U. SERDÜLT

with a higher absolute number of signatures but percentage wise the requirement decreases. The timeframe for signature collection is one month for smaller towns and two months for larger cities (more than 500,000 inhabitants) and prefectures. Before 2002 the LAL set the signature threshold at a third of the electorate for all constituencies irrespective of the size of the electorate. In 2002 this was changed in order to accommodate the fact that the absolute number of signatures required in bigger cities rendered a recall practically impossible. So, only a sixth of the electorate was required for units with above 400,000 citizens. In 2012, the threshold for larger cities was once more attenuated. For constituencies with up to 400,000 inhabitants, the signature threshold is again linear and still at a third of the electorate. Above 400,000 inhabitants, and even more so for more than 800,000, the percentage for the required signatures digressively decreases to reach a value of 15% at the level of 4 million inhabitants. For a city such as Nagoya with an electorate of 1.8  million, the threshold moved from a third to roughly 20% and currently 18%. A reduction from 20 to 18% signifies 40,000 less signatures to collect. Formally, the demand for a recall is addressed to the respective local election board. Once the necessary number of signatures is approved, a recall vote is organised and a simple majority of votes is enough for the recall to go through (CLAIR 2019: 18–19). Recall of Public Servants There is a fourth type of recall which does not lead to a popular vote. It allows the removal of important public servants, such as the vice-­governor, vice-mayor or members of school committees. Although the electorate can propose the removal, in the end only relevant assembly can decide. Such demands are addressed to the chief executive of the respective local government, that is, the mayor (or governor in case of a prefecture). The chief executive then asks the assembly to deliberate. If two-thirds or more of the assembly members are present and if the consent of at least three-­ fourths of them is obtained, the public servant in question is removed from office (CLAIR 2019: 19). Audit Request The fifth option for direct citizen participation on the local level addresses public management. With the signatures of one-fiftieth of the electorate

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

101

(Article 75 LAL), addressed to the audit committee members of the local government, a demand for an audit can be submitted. The audit should determine whether the affairs of the local government in question are being carried out properly or not (CLAIR 2019: 18).

6.4   Japan as a Case of Imbalanced Accountability Japan has no nationwide legislation regulating these local referendums. If a municipality intends to hold a local policy referendum (either proposed by the mayor and members of the assembly or initiated by citizens), it first needs to pass a local referendum ordinance providing for a legal basis. The foundations for such an ordinance are laid down in Article 74 of the Local Autonomy Law. According to Article 74, the request for a referendum ordinance, just like all mass petitions, requires signatures from at least one-­ fiftieth of the electorate. However, whether such a bill for a local referendum ordinance is passed or not is decided by the local assembly (although the mayor can issue a recommendation) (Okamoto et al. 2014). If a local legislature rejects such a demand, policy referendums cannot be held. The people can only ask the mayor to enact new local ordinances, or to revise or abolish existing ordinances in the form of a petition, but they cannot adopt or reject the local ordinances in a referendum vote. In contrast to the recall (which is available in all municipalities), only 56 municipalities out of 1700 have so far introduced legislation to allow for policy referendums (Okamoto et al. 2014). In general, such referendums, even if based on a local referendum ordinance, are not legally binding. Local referendum ordinances usually state that the mayor must respect (soncho suru) the referendum result (Okamoto et  al. 2014). However, the Ministry of Home Affairs (since 2001 Ministry of Internal Affairs and Communications) has reportedly instructed mayors not to use the term ‘to comply’ after a case in which a citizen attempted to include the explicit statement that the mayor must comply with the referendum result. Some mayors do respect and comply with the outcome, while others make a symbolic gesture but then carry on as usual. Residents of Nago in the Okinawa Prefecture, for example, sued the city in January 1998 for disregarding the result of a referendum against a plan to build a US military heliport. The lawsuit was dismissed by the Naha District Court in 2000, with the justification that the result of a referendum is not legally binding.

102 

M. OKAMOTO AND U. SERDÜLT

The stark difference between policy and recall elections leads us to label the case of the Japanese version of the recall as one of an imbalanced vertical accountability. From an institutional design perspective, one could make the argument that as soon as you introduce counterpowers such as the recall, they should be balanced and equally available to the electorate. Otherwise they become inoperative or start to generate perverse effects, such as going against a mayor in office while only being against a landfill, for example.

6.5   Prominent Recall Cases Out of the several hundreds of recall elections, we would like to examine three cases in more detail. These three municipalities experienced either the dissolution of the assembly, the removal of a mayor or both, following a recall vote. Town of Maki (1996) Maki-machi, which is today a ward of the city of Niigata, became known in Japan as the first town holding a policy referendum based on a local referendum ordinance. However, that required a parliament, and as a first step a mayor, willing to accept such a request from the citizens. Lacking other options, the political struggle leading up to the eventual introduction of such legislation in Maki-machi was fought out via a recall vote. The topic of dispute was the construction of a nuclear power plant by the Tohoku Electric Power Company. The plan to build a nuclear power plant in Maki-machi was debated intensely. In 1994, however, a mayor promoting the construction of a nuclear power plant was elected. As expected, when the residents of Maki-machi asked the mayor to hold a referendum on the construction of a nuclear plant, he turned them down. The only option left for the citizens was to collect signatures to remove the mayor in a recall election. On December 1995, however, the mayor resigned, before the recall vote could take place. The newly elected mayor was the leader of the citizens’ group pushing for the vote. On 4 August 1996, the vote on the construction of a nuclear plant in Maki-machi was held with a majority voting against the plant. Finally, Tohoku Electric Power Company officially withdrew the plan to build a nuclear power plant in Maki-machi in December 2003.

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

103

As the example of Maki-machi shows, the recall can be a tool in the political struggle for the implementation of a local referendum ordinance allowing for policy referendums. The City of Akune (2010) Akune in the Kagoshima Prefecture of the Kyushu area in the southern part of Japan is a small city with roughly 23,000 inhabitants under fiscal stress. In the 2008 mayoral election race, Shinichi Takehara became the new executive of the city with a platform suggesting financial cuts that would affect public servants as well as the members of the city assembly. The newly elected mayor proposed a reduction of the number of assembly members from 16 to 6. In addition, he publicised civil servants’ pay slips on his own blog, to demonstrate that their salaries absorbed a very large part of the tax income. He has furthermore been criticised for taking advantage of a provision in the LAL (Local Autonomy Law) allowing the mayor to act on important matters without convening an assembly session on several occasions. This emergency clause should only be applied if there is not enough time to let the assembly discuss and approve a proposal by the executive. For example, he unilaterally halved the bonus of city office workers, introduced a per diem system for assembly members and made a former police officer of the Ehime Prefecture his vice-mayor without the prior consent of the assembly.1 In reaction to such behaviour, the assembly passed a motion of no-­ confidence on 6 February 2009.2 Mayor Takehara then had to choose between his resignation and the dissolution of the assembly. The mayor decided to go against the city assembly and dissolved it on 10 February 2009. The new assembly again decided to pass a motion of no-confidence against the mayor on 17 April 2009. According to the LAL, a repeated no-confidence vote against the mayor automatically terminates his office and new elections must be held. However, Takehara was immediately re-­ elected as mayor on 31 May 2009.

1  Akune City Scandal, 2 October 2010, http://japanvisitor.blogspot.jp/2010/10/akunecity-scandal.html (last accessed 4 June 2017). 2  Attack on bureaucrat pay ousts city mayor, The Japan Times, 21 April 2009, http://www. japantimes.co.jp/news/2009/04/21/national/attack-on-bureaucrat-pay-ousts-citymayor/#.WTOOLcaB3BI (last accessed 4 June 2017).

104 

M. OKAMOTO AND U. SERDÜLT

In this situation of institutional deadlock, a group of anti-Takehara citizens collected the signatures to initiate a recall vote in October 2010. The recall vote took place on 5 December 2010, and Takehara was ousted from office by a tight margin of only 400 votes.3 The close recall vote results then lead Takehara to run again as a candidate for mayor in the follow-up elections in January 2011. But this time, he did not get re-­ elected. Instead, a member of the anti-Takehara citizen group was elected as the new mayor, which lead to a group of Takehara supporters collecting signatures and triggering a recall election against the assembly. The recall vote for the dissolution of the assembly was held on 20 February 2011, and a majority of eligible voters approved it.4 The assembly was dissolved, and on 24 April 2011, new assembly elections took place. The elections resulted in a majority for anti-Takehara members, and thus the saga of Akune came to an end. The example of Akune shows that where the recall option is not at the root of the problem, the mechanism might lead to a prolongation of the whole process but could also help to overcome a deadlock between two antagonistic political forces. The City of Nagoya (2011) Many Japanese local governments are experiencing financial problems due to a shrinking population and a lack of economic growth. In such an environment, conflicts between reformist, somewhat populist mayors and assemblies occur not only in smaller municipalities such as Akune but also in cities such as Nagoya with more than two million inhabitants (for further examples see Serdült and Welp 2017). Contrary to Akune, in Nagoya, the mayor together with his supporters, organised the recall vote and used it as a weapon against a council unwilling to support his reforms (Shimizutani 2010: 110). In April 2009 Takashi Kawamura won the elections in a landslide victory with a record of 510,000 votes and thus became the new mayor of 3  Recall of a mayor, The Japan Times, 10 December 2010, http://www.japantimes.co.jp/ opinion/2010/12/10/editorials/recall-of-a-mayor/#.WTOKtsaB3BI (last accessed 4 June 2017). 4  Wishy-washy Akune residents recall assembly in referendum, The Japan Times, 22 February 2011, http://www.japantimes.co.jp/news/2011/02/22/national/wishy-washyakune-residents-recall-assembly-in-referendum/#.WTOSJMaB3BI (last accessed 4 June 2017).

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

105

Nagoya. In his election campaign, he promised to cut the municipal tax by 10% and to reduce the assembly by half. In October 2009, when Kawamura proposed to cut the pay of councillors, the conflict with the assembly escalated. His promised tax cut bill was finally approved in an extraordinary session in December 2009 but only after Kawamura threatened the assembly with a recall if they voted the bill down. However, the city assembly in June 2010 limited the tax to one fiscal year only. In reaction to this trick by the assembly, he opted to trigger a recall election against the legislature and to step down in case of a successful vote voluntarily, only to seek re-­ election immediately thereafter. At a press conference on 2 August 2010, he stated: “I won’t be able to keep the promises I made to the citizens in the current municipal assembly. I would like to dissolve the assembly and leave the decision up to the people”.5 The biggest challenge in a large city such as Nagoya is to collect the required 366,124 (roughly 20% of the electorate) signatures in the time span of one calendar month. A signature consists of a voter’s full name, address, date of birth and a personal seal, or signature stamp. Since signature lists would be disclosed to the public, some voters might have been hesitant to sign. Until then, no recall had ever been held in a city with more than 200,000 voters in Japan. The signature collection started on 27 August 2010. The mayor’s support group collected about 465,000 signatures and submitted them to the election committee of the city of Nagoya. After some further quarrels6 concerning the validity of signatures, the electoral committee declared that the signature threshold was reached. As a result, both the elections of a new mayor and the recall vote for the dissolution of the assembly were held on 6 February 2011. Kawamura got re-elected and the assembly dissolved. In the subsequent assembly election on 13 March 2011, Kawamura’s local tax cut party, ‘Tax cut Japan’, became the leading political force in the Nagoya assembly. After the election, the assembly also accepted the proposal on the 50% reduction of the reward for assembly members.

5  Nagoya mayor petitions to recall city assembly, The Japan Times, 21 August 2010, http://www.japantimes.co.jp/news/2010/08/21/national/nagoya-mayor-petitions-torecall-city-assembly/#.VuKd7FQzrlZ (last accessed 4 June 2017). 6  Nagoya recall petition fails, The Japan Times, 25 November 2010, http://www.japantimes.co.jp/news/2010/11/25/national/nagoya-recall-petition-fails/#.WTTfIsaB3BJ (last accessed 4 June 2017).

106 

M. OKAMOTO AND U. SERDÜLT

However, the conflict between Kawamura, presently in his fourth term,7 and the assembly is still ongoing. After the Liberal Democratic Party (LDP) and Komeito party won the local assembly elections in 2015 with the help of the Democratic Party, they increased the financial compensation for members of the assembly from 8 million yen to 14.55 million yen per year, on 8 March 2016. Mayor Kawamura of course opposed such a steep increase. He and his supporters might try another recall vote against the assembly. However, his group is currently the minority in the assembly, and the signature collection might fall short. His enemies came to criticise him as a plebiscite dictator.

6.6   The Japanese Recall Experience in Numbers In general, the phenomenon of recall seems to be overlooked in the political science literature. This also holds true for the Japanese case. Several authors writing on local government in Japan are not paying much attention to the direct demand system or have stated that it is hardly being used (Haddad 2012: 40; MacDougall  2001;  Neary 2002: 153–154). But, with newly assembled data from official sources, it is now possible to come up with a fresh assessment and a more nuanced interpretation. This is the first time that recall statistics from Japan can be presented for such a long timeframe, although there are still gaps to fill in before more sophisticated analysis can be undertaken. For the period directly after World War II, the data quality is lower. Also, some of the official sources are not consistent over time. For the present chapter, only recall data from reliable sources with an exact location and date have been taken into account. Unclear cases, or mere hints of recall votes in summary statistics, were ignored so that the total numbers represent a rather conservative estimate of the recall phenomenon in Japan. However, compared to other countries, it is quite fortunate to have information—even including activation attempts—over a long timeframe. Commonly, such data is very difficult to obtain.

7  Tax cut advocate Kawamura cruises to fourth term as Nagoya mayor, The Japan Times, 24 April 2017, http://www.japantimes.co.jp/news/2017/04/24/national/politics-diplomacy/kawamura-looks-set-secure-fourth-term-nagoya-mayor/#.WTTYTMaB3BI (last accessed 4 June 2017).

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

107

Overview of Recall Attempts, Votes and Successful Outcomes In this first empirical overview, we highlight the magnitude of local political participation in Japan within the direct demand system. Descriptive Table  6.2 covers all activation attempts across the six types of popular demands (see Table 6.2). There are of course political grass-root activities and social movement activities in Japan taking place outside of the direct demand system as well (see Haddad 2012). Around the 1970s, with the appearance of city planning and environmental movements, participatory democracy started to gain importance (Numata 2006). One of the most influential forerunners of direct democratic processes in Japan was the movement against environmental pollution during the late 1960s and early 1970s. It was politically independent and involved all levels of society, not unlike in many Western countries. In the 1980s and 1990s, the demand for referendums in Japan increased amid growing tensions and conflicts between citizens and local authorities. Tensions between residents and regional governments increased mainly because of the construction of nuclear power plants, US military bases and landfills. This general strengthening of civil society is also partly reflected in the increased use of the channels for citizen participation offered by the direct demand system. Apart from the dismissal of public servants, all the main types of direct demand were activated quite frequently, particularly at the municipal level (Table 6.2). A total of 2133 attempts of policy requests top the list, followed by attempts to recall a mayor or governor in 720 cases. The comparatively high number of policy requests documents the high level Table 6.2  Direct demands on the local level, 1947–2018, attempts (A) and votes (V) Popular vote possible Recall of assembly A Prefecture 5 Municipality 493 Total 498

No popular vote possible

Recall of assembly member

Recall of the mayor

Government Public Policy audit servant request dismissal

V

A

V

A

V

A

A

A

0 193 193

3 321 324

1 101 102

12 708 720

0 191 191

14 656 670

2 14 16

137 1996 2133

Source: JPC-SED, 2001. MIC (n.d.), ‘Chiho-jichi Geppo’ nos. 53–59

108 

M. OKAMOTO AND U. SERDÜLT

of political pressure promoted by citizens through formal channels. But as we also see further below, they are turned down in most cases. The lowest incidence is of activations of the procedure to dismiss individual public servants. Most of the time, unruly or openly corrupt public servants resign by themselves or are put out of office by superiors before the need for a citizen-activated recall can occur. In general, the number of attempts potentially leading to a recall over almost seven decades and across the roughly 1700 municipalities in Japan does not seem to be disproportionate. Turning to the volume of actual votes taking place, the recall does not seem to be a method of vertical accountability that puts the stability of the political system in danger, as some voices feared when it was first introduced. Recall votes aimed at removing the entire assembly occurred 193 times, but this was less than 3 per  annum across almost 70  years. For members of assemblies and mayors, the totals amounted to 102 and 191, respectively (see Fig. 6.1). There may have been a few more in the early period right after the war, but they are not easy to identify anymore. One would have to search for them in archives. Although it is difficult to assess what a ‘healthy’ amount of recall activity should be, in all roughly 500 recall votes seem to suggest that the institution of the recall mainly fulfils its primary function as a measure of last resort for political conflicts. However, an important requirement for vertical accountability institutions is that they represent a credible potential threat to office holders. Since one can hardly observe such a feature directly, we resort to an indirect measure. To find out whether the threat of a recall is real enough, we calculated the average success rates for the three recall types for the years 1947 to 2018. Based on recall vote attempts, about a fifth of assemblies or individual assembly members have faced recall, but only 5% of the mayors. Based on recall vote outcomes, the success rate for assemblies and assembly members was 72% and 88%, respectively. The success rate for mayoral recalls was lower but still meaningful, at 26% (see Fig. 6.1). These recall success rates are stable over time. They do not fluctuate much. In total, we can identify 224 cases in which the assembly or office holders were ousted from their positions (115 assemblies, 70 assembly members and 39 mayors). Chief executives such as mayors therefore do not need to worry so much when a recall attempt is initiated. Even when the recall vote actually takes place, they have a relatively high chance of survival. So much for the recall as an at least partially credible but not overly abused institution of vertical accountability. By contrast, the success rates

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

109

for policy requests (except in municipalities with a referendum ordinance) are staggeringly low. According to governmental sources, 99% of the roughly 100 policy requests submitted between 1947 and 2012 at the prefecture level were rejected. At the municipal level, it does not look much better. From the nearly 2000 submitted policy requests during the same time span, 93% were rejected. Denied cases include demands supported by vast numbers of signatures. For example, more than 600,000 residents of Nagasaki Prefecture supported the demand for retaining a hospital, and more than 1 million Hokkaido residents asked for the creation of a local referendum ordinance allowing to hold a referendum on a nuclear power plant (Tsujiyama 2000: 28). For the votes dating from January 1989 to March 2018, we also know the reasons given by the initiators of a recall (see Table 6.3). With the help of this shorter but more detailed data set of 106 recall votes occurring in the Heisei era, we can investigate the degree to which the recall procedure is also used for purposes it was not designed for. In the theory part, we said those cases would represent the unintended, secondary function of the recall. The cases are first divided up into nominal categories of reasons given for the recall of the assembly, assembly members and mayors. The categories of reasons given are then assigned to two broad categories, depending on whether they fit the functional logic of a recall or a policy referendum or not. Unclear cases or the ones without a specific reason Table 6.3  Reasons for the 106 recall votes from January 1989 to March 2018 (Heisei era) Reasons given

Assembly

MP

Mayor

Corruption, scandal Conflict with parliament Personality Financial problem Merger Public facility/works Citizen demands Neglect of residents Other No reason Total

1 5 – 0 20 3 2 3 2 15 51

10 0 2 – 2 0 0 3 0 3 20

4 – 1 2 14 4 0 4 1 5 35

Subtotal

Fitting

Total

15 5 3 2 36 7 2 10 3 23 106

Recall

25

Referendum

55

None

26

Source: MHA  (n.d.), ‘Chiho-jichi Geppo’ nos. 49–52. MIC  (n.d.), ‘Chiho-jichi Geppo’ nos. 53–59. Based on the date of the demand

110 

M. OKAMOTO AND U. SERDÜLT

were put into residual categories (others, no reason). Arguably, the assignment of reasons to one of the two functions can be done slightly different. However, the criterion to define a case as fitting the referendum logic was whether the problem at hand could at least potentially have been decided with the help of a policy referendum. For scandals, the personality of politicians or conflicts with councils, this seems very unlikely. For financial problems, a referendum at an early stage might have been an option, but we assumed the damage was already done. The largest category is recalls over a merger proposal as the reason for the vote (36 cases). Further recall votes that could have been prevented can be traced back to the non-responsiveness of local authorities (citizen demands, neglect of residents). Public facilities or works are clearly also political issues that can be accepted or voted down with the help of policy referendums. It therefore seems safe to say that a relatively large proportion of recall votes would turn into policy referendums if this option was available (authorised by a local referendum order) and if it was allowed to produce a binding decision. Development Over Time To map recall practices over time, the annual raw frequency data were transformed to three-year moving averages (see Fig. 6.1). Choosing the time window of three years is admittedly arbitrary. However, we wanted to account for the fact that from its initiation the vote of a recall can potentially carry over into the next year. At least a two-year moving average would therefore seem in order. In addition, we aimed to present a reasonably smooth, easy-to-interpret timeline, without losing some of the smaller fluctuations in the data. Hence, a three-year moving average was applied. Frequencies for the three types of recall initiatives show two clearly identifiable peaks: the first one in the period right after World War II from 1948 to 1950 and a further substantial peak in the years from 2004 to 2006. The two peaks capture two instances of mobilisation against elected authorities representing the primary and secondary use of recall procedures as outlined in the theory chapter. The first one is related to the transition to democracy right after World War II, the second one to a large degree to municipal mergers.

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

111

 ecalls as a Contribution to Democratisation R The first wave of recall activities in Japan during the transition to democracy attained levels which thereafter have not been reached again. Initiatives to recall assemblies and mayors rose to more than twice the level that can be observed during the second highest peak in the years 2004–2006 (see Fig. 6.1). How can this be explained? To put it bluntly, because the new LAL recall procedures were at the disposal of citizens and they evidently made use of them right away. Whereas a certain ‘novelty effect’ can often be observed—be it in the case of technical or of democratic innovation— and might have been at play, we suggest there is more to it than just ‘testing’ the new device. The years after World War II were extraordinary times. As a hypothesis which would have to be investigated in more detail, we posit that recall procedures provided a necessary—but not sufficient— condition for Japan’s successful democratisation process. In a first phase of democratisation, the so-called purge by the GHQ managed to eliminate undesirable personnel from public office. Even though it took two years to administer the purge, 200,000 public servants were at least temporarily retired. For politicians, the focus of the purge was on the national level (Baerwald 1959: 83). In this context, the recall procedures at the local level served as a backup option. Some local politicians were still bound by feudalistic thinking. With the recall in their hands, citizens and modernising forces in general had the option to weed out anti-­ democratic politicians. On the one hand, the role and importance of the recall in the democratisation process of Japan should not be exaggerated. On the other hand, it might be an element which has been overlooked so far. As frequently stated in the literature, factors such as economic and social conditions favourable to sustainable democratisation, the intermediary sector and the crucial role of elites are certainly also part of the equation (Schmitter 2010; Buček and Smith 2000). However, all those conditions come into play at a slightly later stage. They are usually not operative right after a major crisis such as a war or a revolution. At an early stage of the transition, despite the purge of authoritarian elites, anti-­ democratic rulers do not only try to make it back into public office, at least some of them usually also succeed. Open and fair elections, eventually bringing parts of the old elites back, are one of the conundrums for new democratic orders in a transition phase (Rothstein 1996: 156; Croissant 2004: 169). It is certainly too early to draw far-reaching conclusions based on aggregate statistics alone. However, the Japanese evidence suggests that the

112 

M. OKAMOTO AND U. SERDÜLT Mayor

100 90 80 70 60 50 40 30 20 10 0 1949 1952 1955 1958 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 2012 2015 2018

Assembly Member 100 90 80 70 60 50 40 30 20 10 0 1949 1952 1955 1958 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 2012 2015 2018

Assemblies 100 90 80 70 60 50 40 30 20 10 0

1949 1952 1955 1958 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 2012 2015 2018

Fig. 6.1  Three-year moving averages of recall attempts, recall votes and successful recalls for mayors, assembly members as well as assemblies (1947–2014). Sources: MHA (n.d.), “Chiho-jichi Geppo” nos. 6/7, no. 9, no. 12, no. 23, no. 29, no. 37, no. 38, nos. 41–43, nos. 45–52. MIC (n.d.), “Chiho-jichi Geppo” nos. 53–59

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

113

recall and direct vertical accountability in general merit more detailed exploration in future studies. The question is how much and under which conditions direct vertical accountability contributes to more sustainable processes of democratisation.  ecalls Against Mergers and Other Policy Measures R Several peaks of recall activity in our longitudinal data are strongly related to phases of municipal mergers. The number of municipalities in Japan has decreased quite dramatically from 71,314 in 1888 to 1718 in 2016. The decrease occurred over three consolidation periods. The first one was the “Great Meiji Consolidation” 1888–1889, when the number of municipalities was reduced from 71,314 to 15,859. During the “Great Showa Consolidation” from 1953 to 1961, the number of municipalities decreased further from 9868 to 3472. We can observe increased activity for all three recall types in our data (see Fig.  6.1). Finally, during the “Great Heisei Consolidation” lasting from 1999 to the end of March 2010, the number of municipalities was again reduced from 3229 to 1756 (Yokomichi 2007). Those two most recent devolution laws dating from 1999 to 2000, respectively, are the cause for another series of recalls in the years 2004–2006, the second most active phase of the recall in Japanese history. Many municipalities went through a fierce confrontation between those in favour and those against the merger. There were activations of the recall from both sides.

6.7   Conclusions The transfer of recall procedures into the hands of the citizens can be attributed to favourable conditions at a major inflection point in Japanese political history. The Social Democrats were the strongest political party in the Diet at that time (and never again thereafter). From its foundation this party had adopted the idea of recall. Without their support the delegation of competence for local-level recalls from the Ministry of Home Affairs to the citizens would probably not have happened. After an initial upsurge of recall initiatives in the period right after World War II, the use of the recall subsided for a generation. It was around the 1970s, with the appearance of city planning movements (machizukuri) and environmental movements, when Japanese participatory democracy started to gain importance (Numata 2006). One of the most influential forerunners of direct democratic processes in Japan was the movement

114 

M. OKAMOTO AND U. SERDÜLT

against environmental pollution during the late 1960s and early 1970s. In the 1980s and 1990s, the demand for popular referendums in Japan increased dramatically amid growing tensions and conflicts between citizens and local authorities. In the 1980s tensions between residents and regional governments increased, mainly because of the construction of nuclear power plants, US military bases and landfills. Japan’s local autonomy system is still mainly based on the principles of representative democracy. However, the people are permitted not only to elect and to be elected as representative assembly members and chief executives but also have more direct means of political action. The three types of recall are very important measures for the citizens in prefectures and municipalities. A problematic mayor or a troublesome member of the assembly can be dismissed from office by means of a direct vote. For example, long-lasting conflicts between the assembly and a mayor can bring a lot of disorder to a municipality with negative effects for the inhabitants. If things come to such a pass, the voters themselves are able to start the recall process and finally to vote to resolve the deadlock. Residents have their treasured ‘family sword’ in the form of the recall option, and they can draw it. However, we should also issue a word of caution here. If political camps start using the recall procedures as a political weapon in the first place, the situation can easily deteriorate. However, up to now this does not seem to be the case in Japan. But there is also a dysfunctional element to the recall in Japan. It is due to the lack of effective policy referendums. Because of the high hurdles Japanese citizens have to overcome in order to trigger a local referendum ordinance allowing for a (non-binding) vote on policy issues, they often do not have any other choice than to confront parliament or the mayor with a recall. In his assessment of Japanese efforts to decentralise politics, and as part of this process to increase local direct democratic practice, Tsujiyama (2000: 28) also observes this shift from the primary function of the recall to a secondary one rather directed at policies (Jain 1991: 560). Epic battles such as the ones in Nagoya with two recalls and two by-­ elections within a short time span could probably be averted if there was the option of effective policy referendums. A general lesson we can draw from the Japanese case is that once mechanisms of direct citizen participation are introduced, all the other political institutions should accommodate that. Counterpowers such as policy referendums and recalls should be well balanced, just as the division of powers in a representative system should also be well balanced. Otherwise

6  RECALL IN JAPAN AS A MEASURE OF VERTICAL ACCOUNTABILITY 

115

there is a risk of more second-order resort to recall procedures, not for the vertical accountability of councillors, assemblies or mayors, but as an indirect way to oppose specific unacceptable policy moves.

References Baerwald, H.  H. (1959). The Purge of Japanese Leaders Under the Occupation. Berkeley and Los Angeles: University of California Press. Buček, J., & Smith, B. (2000). New Approaches to Local Democracy: Direct Democracy, Participation and the ‘Third Sector’. Environment and Planning C: Politics and Space, 18(1), 3–16. Croissant, A. (2004). From Transition to Defective Democracy: Mapping Asian Democratization. Democratization, 11(5), 156–178. Haddad, M.  A. (2012). Building Democracy in Japan. Cambridge: Cambridge University Press. Hug, S. (2004). Occurrence and Policy Consequences of Referendums: A Theoretical Model and Empirical Evidence. Journal of Theoretical Politics, 16(3), 321–356. Jain, P.  C. (1991). Green Politics and Citizen Power in Japan: The Zushi Movement. Asian Survey, 31(6), 559–575. Kobori, M. (2009). Referendums in Britain and Japan: Turnouts, Campaigns, and Systems. Ritsumeikan Law Review, 26, 1–25. MacDougall, T. (2001). Towards Political Inclusiveness: The Changing Role of Local Government. In M. Muramatsu (Ed.), Local Government Development in Post-War Japan (pp. 29–62). Oxford: Oxford University Press. Matsuno, M. (1981). Genko Chokusetsu Seikyu Seido no Seiteikatei [Decision Making Process of the Current Direct Demand System]. Shimadai-Hogaku, 24(2/3), 89–106. Moreno, E., Crisp, B. F., & Shugart, M. S. (2003). The Accountability Deficit in Latin America. In S. Mainwaring & C. Welna (Eds.), Democratic Accountability in Latin America (pp. 79–131). Oxford: Oxford University Press. Neary, I. (2002). The State and Politics in Japan. Cambridge: Polity Press. Numata, C. (2006). Checking the Center: Popular Referendums in Japan. Social Science Japan Journal, 9(1), 19–31. Okamoto, M., Ganz, N., & Serdült, U. (2014). Direct Democracy in Japan. C2D Working Paper Series, 47, 1–24. https://doi.org/10.5167/uzh-103334. Rothstein, B. (1996). Political Institutions: An Overview. In R.  E. Goodin & H.-D. Klingemann (Eds.), A New Handbook of Political Science (pp. 133–166). Oxford: Oxford University Press. Schmitter, P. C. (2005). The Ambiguous Virtues of Accountability. In L. Diamond & L. Morlino (Eds.), Assessing the Quality of Democracy (pp. 18–31). Baltimore: The Johns Hopkins University Press.

116 

M. OKAMOTO AND U. SERDÜLT

Schmitter, P. C. (2010). Twenty-Five Years, Fifteen Findings. Journal of Democracy, 21(1), 17–28. Serdült, U. (2015). A Dormant Institution: History, Legal Norms and Practice of the Recall in Switzerland. Representations, 51(2), 161–172. Serdült, U., & Welp, Y. (2017). The Levelling Up of a Political Institution: Perspectives on the Recall Referendum. In S. P. Ruth, Y. Welp, & L. Whitehead (Eds.), Let the People Rule? Direct Democracy in the Twenty-First Century (pp. 137–156). Colchester: ECPR Press. Shimizutani, S. (2010). Local Government in Japan: New Directions in Governance Toward Citizens’ Autonomy. Asia-Pacific Review, 17(2), 99–117. Steiner, K. (1965). Local Government in Japan. Stanford: Stanford University Press. Tsujiyama, T. (2000, Spring). Local Self Governance in Japan: The Realities of the Direct Demand System. Nira Review, 26–30. Vosse, W. (2000). The Emergence of a Civil Society in Japan. Japanstudien, 11(1), 31–53. Welp, Y. (2016). Recall Referendums in Peruvian Municipalities: A Political Weapon for Bad Losers or an Instrument of Accountability? Democratization, 23(7), 1162–1179. Whitehead, L. (2018). The Recall of Elected Officeholders: The Growing Incidence of a Venerable, but Overlooked Democratic Institution. Democratization, 25(8), 1341–1157. Yokomichi, K. (2007). The Development of Municipal Mergers in Japan. Tokyo: CLAIR.  Retrieved from http://www.clair.or.jp/j/forum/honyaku/hikaku/ pdf/up-to-date_en1.pdf.

Sources and Government Reports CLAIR (Council of Local Authorities for International Relations). (2019). Local Government in Japan 2016 (2018 Revised Edition), Tokyo. Retrieved from http://www.clair.or.jp/j/forum/pub/docs/jichi2018-en.pdf. Government of Japan. (2010). Chiho Jichi Ho [Local Autonomy Law, LAL], E-Gov. Retrieved from http://law.e-gov.go.jp/htmldata/S22/S22HO067. html. JPC-SED (The Japan Productivity Center for Socio-Economic Development). (2001). Chiho-bunken to Jumin-sanka wo kangaeru [Thinking on Decentralization and Citizens’ Participation]. The Japan Productivity Center for Socio-Economic Development, Tokyo. MHA (Ministry of Home Affairs). Chihojichi Geppo. No.6–9, 12, 23, 29, 37, 38, 41–43, 45–52. MIC (Ministry of Internal Affairs and Communications). Chihojichi Geppo. No.53–59.

CHAPTER 7

Explaining Institutional Change Towards Recall in Germany Brigitte Geißel and Stefan Jung

7.1   Introduction More than any other democratic innovation, recall has an immediate impact on elected political elites by deciding over the fate of their incumbency. Therefore, it is not surprising that elites in many countries hesitated to implement this institution for a long time. However, recall has regained popularity in both new and established democracies in the course of the third wave of democratization.1 As a case in point, eleven of the sixteen German federal states (Bundesländer) have introduced the local recall of mayors since the fall of the Berlin Wall. This convergence notwithstanding, these eleven federal states adopted a variety of institutional designs which in turn led to diverse practices of local recall (see Table 7.1; cf. Böhme 2008; Fuchs 2007; Geissel and Jung 2018). They all introduced indirect recall, which is initiated by the legislative (as opposed to direct recall, Qvortrup 2013: 77), with however 1  For an international overview of provisions for recall, see Qvortrup (2011), Serdült and Welp (2017: 143), and Welp (2018: 453).

B. Geißel (*) • S. Jung Institute of Political Science, Goethe University Frankfurt, Frankfurt, Germany e-mail: [email protected]; [email protected] © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_7

117

118 

B. GEIßEL AND S. JUNG

Table 7.1  Institutional rules for recall in Germany Federal state

Year of Direct Indirect recall: introduction recall Council majority requirement

Direct recall: Signature quorum (% eligible voters)

Approval quorum (% eligible voters)

Brandenburg

1993

Yes

23

10

25

Hesse

1998 1992

No

23

30

Lower Saxony

1996

No

34

25

Mecklenburg-­ Vorpommern

1997

No

23

50d,e

North-Rhine Westphalia

1994

No

2011

Yes

25a, 20b, 15c

2004

33.3e 25

23 20f, 17.5g, 15h

Rhineland-­ Palatinate Saarland

1993

No

23

30

1994

No

23

30

Saxony

1993

Yes

34

Saxony-Anhalt

1993

No

34

Schleswig-­Holstein

1995

Yes

23

2002 Thuringia

1993

No

2002 2016

33.3, 20i

50 30

25

33.3

20

20 30

12 23

Yes

35

Sources: Böhme 2012: 57; Fuchs 2007: 34F; Landtag Mecklenburg-Vorpommern 1997: 696; Landtag Mecklenburg-Vorpommern 2004: 62; Schleswig-Holsteinischer Landtag 1996: 43; Schleswig-­ Holsteinischer Landtag 2002: 133; Schmidt 2018: 167; Thüringer Landtag 1993: 510; Thüringer Landtag 2002: 470; Wollmann 2008: 89 a Municipalities with up to 20,000 inhabitants b Municipalities with over 20,000 up to 60,000 inhabitants c Municipalities with over 60,000 inhabitants d Only turnout required e 2 3 majority of voters required for approval f Municipalities with up to 50,000 g Optional for municipalities with over 100,000 inhabitants h Municipalities with over 100,000 inhabitants i Municipalities with over 50,000 up to 100,000 inhabitants

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

119

­ ifferent council majority requirements. Three federal states, Brandenburg, d Saxony, and Schleswig-Holstein, granted citizens the right to initiate recall in the 1990s. After more than fifteen years without significant institutional changes, two more federal states, North Rhine-Westphalia and Thuringia, also decided to supplement indirect recall with direct recall. Each of these five federal states adopted different quorums to initiate and approve recall (see Table 7.1). In this chapter, we are especially interested in explaining the introduction of (local) direct recall, because direct recall transfers even more power to citizens than indirect recall. Several attempts have been made to explain the implementation of direct democratic instruments and the direct election of mayors, but a comprehensive analysis of institutional change towards recall, the reasons for the varying institutional designs, and subsequent institutional changes is missing until now. In this chapter, we present a framework based on theories of institutional change and democratic innovations to explain the introduction of direct recall and the choice of quorums for its initiation and approval. Then we use the framework to answer the following two questions: (1) Why do states implement (local) direct recall? (2) What leads states to choose higher or lower quorums for recall? To answer these questions, we analysed protocols of plenary and committee sessions of the five state legislatures (Landtage) with direct recall and also considered the historical context. Four reasons seemed to be crucial to introduce direct recall in Germany. Direct recall was almost exclusively implemented in the course of broader democratic reforms, which corresponds to its strong association with the direct election of mayors and other direct democratic instruments. Our findings also suggest that direct recall is predominantly introduced by left-wing parties as an instrument to increase the accountability and limit the power of the executive.

7.2   An Analytic Framework for Institutional Change Towards Recall The implementation of recall and its diffusion around the globe since the 1990s is considered a result of democratization and a reaction to citizens’ dissatisfaction with their representatives (Welp and Whitehead 2020). In our contribution, we expand upon this assessment and aim to provide a refined framework for explaining institutional change towards

120 

B. GEIßEL AND S. JUNG

recall. In assembling the framework, we take on perspectives from institutionalist theories of institutional change (cf. Hall 2010), looking at key actors’ interests and ideologies, the institutional context, and processes of diffusion between states. In the following sections, we break down these three dimensions of the framework, starting with attributes of key actors that motivate institutional change. Actors’ Interests and Ideologies Outlining key actors’ strategic interests and attitudes towards institutions is crucial to understand why institutions change. Due to party competition, parties may induce institutional change for strategic reasons (Scarrow 1997: 456; Mahoney and Thelen 2010: 9). Generally, parties in government should have little incentive to implement recall on their level of authority because it endangers their power base and reduces their control over offices between elections.2 Nevertheless, we assume two strategic interests parties could pursue by introducing local direct recall: First, parties might be more willing to sacrifice control by introducing direct recall on the local level if they expect electoral benefits on higher political levels (cf. Scarrow 1997: 465). They might frame recall as an opportunity for citizen participation as a strategy to appeal to voters (cf. Scarrow 1997: 461). In this case, we expect parties to choose low quorums in order to attract voters. Second, governing parties may opt for direct recall especially when the opposition dominates the local level (cf. Klein and Lüdecke 2018: 131; Serdült and Welp 2017: 142; Welp 2016: 1174; Welp and Milanese 2018: 1392). This way, their local branches can bypass the council to recall mayors from rival parties (cf. Welp 2016: 1170). In this case, we should observe strong resistance by the opposition accusing the governing parties to use direct recall as a political weapon. Conversely, governing parties might criticize the opposition for rejecting direct recall only to protect their incumbents. Additionally, we expect the government to choose low quorums for direct recall to facilitate its use as a means to oust incumbent mayors. Governing parties might also be forced to implement direct recall as a means to regain citizens’ trust in the course of political scandals (Scarrow 2  A similar argument can be made for direct democratic instruments because they reduce party control over political decision-making (Scarrow 1997; Vetter 2009).

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

121

1997: 465; Vetter 2009: 133). In this case, we expect direct recall to be framed by legislators as a means to appease dissatisfied citizens, provide them with new participatory opportunities, and regain their trust in ­politics. This reasoning should entail low quorums for direct recall to enable frequent use by citizens. Institutional change can be a product of (altered) attitudes of governing parties towards an institution, which sets the focus on party ideologies (Hall 2010: 208, 212). There is a tendency of left-wing parties to support reforms for direct democratic instruments (Christmann 2009). Correspondingly, left-wing governments have often introduced recall in Latin America (Welp and Serdült 2014). Thus, we expect direct recall to be implemented predominantly by left-wing parties, promoting it as a new measure of citizen participation. Correspondingly, low quorums should be proposed to make it appealing for citizens. Institutional Context We propose that the institutional context where direct recall is situated could influence the decision to implement direct recall and the shape of its institutional design in three different ways. First, recall can be introduced to increase the accountability of an independent executive towards the citizens (Qvortrup 2013: 74–75; Serdült and Welp 2017: 141). Accordingly, Wollmann (2004: 159) argues that legislators implemented recall to balance the high legitimacy mayors receive from direct election with high accountability—and they might also want to limit the executive power of the mayors (cf. Egner 2013: 35–36; Serdült and Welp 2017: 139–142; Wollmann 2008: 92–93). Second, legislators might choose to introduce direct recall when they deem it suitable in the local institutional setting (cf. Holtkamp 2005: 22). We can differentiate German federal states local institutional settings being more majoritarian or consensual (Holtkamp 2008; Holtkamp and Bogumil 2016). If we consider direct recall establishing citizens as an additional veto player analogous to direct democratic initiatives, it would be more suitable to consensual settings (Lijphart 2012: 221–222; Vatter and Bernauer 2009: 338–339; but see Geissel and Michels 2018). Thus, we assume that direct recall was implemented in states with consensus democratic institutional settings.

122 

B. GEIßEL AND S. JUNG

Third, direct recall shares similar features with other direct democratic instruments regarding institutional design and application (cf. Welp 2018: 451). Thus, politicians might refer to direct recall as part of a broader programme of direct democratic instruments.3 Accordingly, we expect preexisting or parallelly introduced direct democratic instruments to influence the institutional design of direct recall through the choice of similar quorums. Policy-Learning Policy-learning is an essential factor for institutional change as governments often look for experiences with similar institutions implemented in other states when carrying out institutional reforms (Dolowitz and Marsh 2000; Elkins and Simmons 2005; Shipan and Volden 2008). After reunification, the five new federal states formed cooperative partnerships with West German federal states adopting institutional settings of their West German partner states (cf. Haus 2005). Accordingly, we expect legislators willing to implement direct recall to refer to positive experiences made with this institution by other states or to hint to cooperation with states functioning as a role model. Consequently, we assume legislators to choose similar quorums like these states. Institutional change towards recall may be part of a broader process of democratization (Welp and Whitehead 2020). This refers to the transformation of non-democracies to democracies but also to democratic improvements. Particularly in a federal system, different federal states might opt for similar democratic reforms (cf. Haus 2005). Therefore, the democratization processes in the new German states might have motivated the adoption of direct recall and to choose low quorums. Table 7.2 summarizes the different factors for institutional change towards recall. Before analysing institutional change towards direct recall, we provide an overview of the implementation, institutional design, and procedural similarities of recall in German federal states in the next section.

3  Direct recall might even be used as a ‘second order recall’ where recall ‘is actually targeted at a policy and not the political authority per se’ (Serdült 2015: 10).

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

123

Table 7.2  Analytic framework for institutional change towards direct recall Dimension

Factor

Actors’ interests and ideologies

Party competition Parties implement direct recall to gain votes on the state level or to break the opposition’s dominance on the local level.

Institutional context

Policy-­ learning

Implementation of recall

Choice of quorums

Parties choose lower quorums for direct recall to gain votes on the state level or to break the opposition’s dominance on the local level. Political scandal Parties implement direct Parties choose lower recall to regain trust after quorums to regain trust scandals. after scandals. Party ideology Left-wing parties Left-wing parties choose implement direct recall to lower quorums to promote citizen promote citizen participation. participation. Accountability of State legislators State legislators choose the executive implement direct recall to lower quorums to increase the accountability increase the and limit the power of the accountability and limit executive. the power of the executive. Consensus State legislators State legislators choose democracy implement direct recall in lower quorums in consensual institutional consensual institutional settings to establish settings to establish citizens as veto players. citizens as veto players. Direct democracy State legislators State legislators choose implement direct recall similar quorums for together with direct direct recall and direct democratic instruments or democratic instruments. where these already exist. Policy-learning State legislators State legislators adopt implement direct recall quorums from other due to cooperation with states they cooperate with or positive experiences of or that made positive other states providing experiences with direct direct recall. recall. Democratization States that recently States that recently transitioned to democracy transitioned to democracy implement direct recall. choose lower quorums.

Source: Own depiction

124 

B. GEIßEL AND S. JUNG

7.3   Institutional Designs of Recall in German Federal States Direct recall was institutionalized first in 1993 in Brandenburg and Saxony and later in Schleswig-Holstein in 1995. It took another fifteen years until North Rhine-Westphalia and most recently Thuringia adopted direct recall as well. What is striking are the substantial differences in quorums adopted by the first two federal states that introduced direct recall. Signature quorums ranged from 10% in Brandenburg to one-third of eligible voters in Saxony and approval quorums from 25% in Brandenburg to 50% in Saxony. Schleswig-Holstein following a couple of years later chose a middle road between these two extremes. North Rhine-Westphalia decided to take a more nuanced approach, differentiating signature quorums based on municipality size, whereas Thuringia opted for a stricter signature quorum in 2016. Direct recall proceeds similarly in all five federal states that adopted recall until today.4 First, citizens need to collect a certain number of signatures within a limited period to initiate direct recall (Witte 2001: 60). If the signature quorum is reached, citizens decide on recall in a popular vote (Witte 2001: 60–61). Each state installed approval quorums in addition to requiring a majority of voters to dismiss the mayor, meaning that a certain proportion of eligible voters needs to vote in favour of recall. Mayors can avoid the final vote if they resign after the successful initiation of direct recall.

7.4   The Introduction of Direct Recall in the 1990s The first introduction of direct recall in German federal states took place in a period of profound institutional change in the wake of German reunification in 1990. The reconstituted East German federal states had to enact new municipal codes replacing the ones of the German Democratic Republic (GDR), which included direct democratic instruments but not recall (Volkskammer 1990; Holtmann et al. 2017: 90f).5 4  In the case of indirect recall, the council needs to pass a resolution with a supermajority to initiate the popular vote (Witte 2001: 59). 5  For different traditional types of West German municipal codes, see Holtmann et  al. (2017: 90) and Woyke (2013: 213ff).

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

125

Around the same time, most West German federal states implemented major reforms of municipal codes and municipal election laws including cumulative voting and panachage, direct democratic instruments, and the direct election and the recall of mayors (cf. Geissel 2017; Haus 2005; Scarrow 1997; Vetter 2009). While most federal states only introduced indirect recall, Brandenburg and Saxony went further by implementing direct recall almost simultaneously in April 1993.6 In the next sections, we will take a closer look at the early implementation of direct recall by Saxony, Brandenburg, and Schleswig-Holstein followed by a short analysis of later changes towards direct recall. Saxony The reform of the municipal code leading to the implementation of direct recall in Saxony began in 1992. In the reform of the municipal code, the CDU, which governed with an absolute majority since 1990 largely, maintained the regulations for direct democratic instruments from the GDR municipal code but introduced the direct election of mayors as well as direct and indirect recall. Examining the parliamentary debates, we find little evidence that party competition was an essential factor as no party advertised for direct recall as a new opportunity for citizen participation. However, party competition may have had some weight for the CDU dominating both the state and local level at that time. The first draft bill proposed by the CDU dispensed entirely with the recall of mayors, and the party introduced direct recall only reluctantly choosing high quorums (Sächsischer Landtag 1992a). This reluctance of the CDU and the high hurdles they set for direct recall might have resulted from the CDU’s intention to protect their mayors. However, in the parliamentary debates, the opposition did not accuse the government of inhibiting direct recall to protect their mayors. The need for a new municipal code signalled by mayors and municipal councils notwithstanding, there was no political scandal that forced the government to implement direct recall, and there were no references to political scandals during the debate. The fact that it took the CDU almost

6  When implementing indirect recall, most states chose the same majority requirements that previously existed for the councils’ vote of no confidence against the mayor (Lenhof 2013: 57ff).

126 

B. GEIßEL AND S. JUNG

three years after their election into government to pass the legislation supports the lack of pressure. In line with the argument that parties from the left are more likely to support direct recall, the first bill for a reform of the municipal code of Saxony that included direct recall was proposed by the left-wing Linke Liste/PDS in 1991.7 Thus, party ideology was a relevant factor as the CDU did not include direct recall in its bill, but later ‘gave in.’ While SPD and Bündnis 90/Grüne favoured council-elected mayors in the first place (Sächsischer Landtag 1992b: 2855), they criticized the high quorums together with Linke Liste/PDS for making recall an unusable, merely theoretic institution and proposed amendments to lower the quorums (Sächsischer Landtag 1993b: 4519f). Increasing the accountability of the executive was a primary reason for the governing CDU to implement direct recall. While the CDU did not consider recall at the beginning, they later regarded direct recall as necessary to limit the power of the mayor and claimed that citizens’ right to directly elect mayors should also entail the right for direct recall (Sächsischer Landtag 1993a: 40). However, the CDU conceived direct recall only as a preventive instrument that should not represent a permanent threat influencing the mayors’ decision-making (Sächsischer Landtag 1993b: 4520). The choice of high quorums suggests that the CDU saw direct recall more as an ultima ratio setting limits to the mayors’ power rather than an instrument that could be frequently used by citizens to hold mayors accountable (cf. Sächsischer Landtag 1993b: 4505). The new municipal code enacted by the CDU was much stronger oriented towards consensus democracy than the municipal code of the GDR (Holtkamp 2008: 121).8 While direct recall fitting into this institutional setting might have been a reason why the CDU finally accepted to implement direct recall, the CDU effectively prevented citizens from becoming another veto player by establishing high quorums. Regarding direct democracy, there are general references to other direct democratic instruments in the debates on direct recall, but this did not affect the choice of quorums by the government. 7  Linke Liste/PDS emerged from the dominant party of the GDR, SED (Sozialistische Einheitspartei Deutschlands). 8  Most East German stats implemented municipal codes with features of consensus democracy which some scholars see as a result of the political culture of the GDR and the experience of reunification (Holtkamp and Bogumil 2016: 26).

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

127

During the reform of the municipal code, policy-learning had significant effects on the Saxon CDU government but also on the opposition through the cooperation with its partner state Baden-Württemberg (Maaß 2014: 216; Sächsischer Landtag 1991). The CDU’s reluctance to implement direct recall and its decision for high quorums suggests a strong orientation towards the Baden-Württemberg model which did not feature direct recall. Generally, the democratization of Saxony provided the background for the reform of the municipal code and the implementation of recall. The experience with autocratic leadership in the GDR was present in the parliamentary debates on the direct election of mayors, and direct recall might have been a means to alleviate the fear of creating too powerful executives. Together with a general openness regarding citizens’ involvement, these considerations may have affected the CDU’s position towards direct recall even if it did not result in low quorums. Brandenburg Brandenburg found itself in a similar position as Saxony with the provisional municipal code of the GDR requiring the state parliament to author a new municipal code. Unlike Saxony, the state election in 1990 made it necessary for the winning SPD to coalesce with the much smaller FDP and the Greens (Lorenz 2016: 29f). The coalition government in Brandenburg also chose another path to introduce recall, enacting a new municipal election law in April 1993 before passing the new municipal code half a year later (Landtag Brandenburg 1993c, 1993d). Party competition was probably not an essential factor for the implementation of direct recall in Brandenburg. The new municipal election law in Brandenburg was subject to fierce debates between the government coalition and the opposition, but the main lines of conflict did not refer to recall but instead to the proceeding of the reform which included early elections. There was also no reference to political scandals during the debates. Party ideology was an important factor for the implementation of direct recall in Brandenburg. It was again the PDS/Linke Liste to first draft a bill for a new municipal code in January 1993 planning to introduce the direct election of honorary mayors who should be subject to direct recall (Landtag Brandenburg 1993a). The government coalition found it more difficult to agree on a common position regarding the election of mayors

128 

B. GEIßEL AND S. JUNG

with the SPD favouring election by council and FDP and the Greens supporting direct election. The compromise in the first draft bill planned to let all municipalities hold a referendum to choose between direct and council election of mayors concurrently with the early municipal elections (Landtag Brandenburg 1993b). Nevertheless, the coalition also agreed to introduce direct recall with rather low quorums in the first draft bill. When the resistance of the SPD against the direct election of mayors later broke down under the pressure of the coalition partners (Landtag Brandenburg 1993f: 5056), direct recall now affected all mayors (Landtag Brandenburg 1993c). In sum, left-wing parties in government and opposition played a crucial role in implementing direct recall and setting low quorums. Increasing the accountability of the executive was a motivation for government and opposition. Both government and opposition saw direct recall as a logical consequence of the direct election of mayors and as the only legitimate measure to dismiss mayors (Landtag Brandenburg 1993b: 126). The government opted for a relatively low signature quorum providing citizens with extensive opportunities to hold mayors accountable (Landtag Brandenburg 1993f: 5094). While tending towards consensus democracy, Brandenburg also implemented features of majoritarian democracy, and therefore the introduction of direct recall is less straightforward regarding the broader institutional setting (cf. Bathge et al. 2016: 187). Direct democracy turned out to have a much stronger effect in Brandenburg compared to Saxony. The coalition government copied the signature and approval quorums of 10% respectively 25% of eligible voters set for direct democratic instruments in the municipal code of the GDR for direct recall in the new municipal election law. The unaltered adoption of the institutional design of direct democratic instruments from the municipal code of the GDR can also be interpreted as a consequence of policy-learning due to the positive experiences with these institutions (cf. Landtag Brandenburg 1993e: 4835). Furthermore, cooperation between Brandenburg and its partner state North Rhine-Westphalia influenced the reform of the municipal code: the SPD was cautious to give mayors too much power (Holtkamp 2005: 28), which led to low quorums for direct recall. Similar to Saxony, the implementation of recall in Brandenburg can be seen as a continuation of the democratization of Brandenburg that involved new opportunities of participation for its citizens, but in the debates,

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

129

direct recall was not explicitly situated in this context (Künzel 2010: 86; Reutter 2016: 172). Schleswig-Holstein Until 1995 direct recall was confined to East German states but with the reform of its municipal code, Schleswig-Holstein became the first West German state to implement direct recall. The SPD government first met with resistance of its state party conference when proposing the direct election of mayors. Reform then only got on the agenda when the opposition parties CDU and FDP launched an agenda initiative (Volksinitiative) for implementing the direct election of mayors in 1994 (Vetter 2009). This strategy proved to be successful as the SPD started to prepare a reform after the agenda initiative reached the required signature quorum. The initiative can be interpreted as motivated by party competition (Scarrow 1997: 461), but neither the opposition nor the government mentioned direct recall in the initiative or the first draft bills. Thus, the agenda initiative succeeded in building pressure for reform, but direct recall itself was not subject to strategic party interests. Accordingly, there were only slight complaints by the opposition regarding too high quorums in the final parliamentary debate (Schleswig-Holsteinischer Landtag 1995c: 7394, 7396). Political scandals were also not an important factor for the government and the main opposition parties. For instance, no other party reacted to the proposal of a representative of the extreme-right DVU to introduce direct recall to solve a political conflict in the capital city (Schleswig-­ Holsteinischer Landtag 1995b: 6571). Party ideology played a more ambiguous role, although it was again a left-wing party that implemented direct recall. Neither a first motion by the CDU that asked to implement the direct election of mayors nor a similar bill drafted later by the government included direct recall (Schleswig-­ Holsteinischer Landtag 1993; Schleswig-Holsteinischer Landtag 1995a). However, the SPD changed its initial position and implemented direct recall (Schleswig-Holsteinischer Landtag 1996). Regarding the accountability of the executive, we find again arguments presented by the state government that establish a logical link between the legitimacy of the direct election of mayors and the requirement for direct recall. However, the SPD first justified indirect recall in the draft bill as a necessary filter for the sometimes-irrational public opinion in contrast to

130 

B. GEIßEL AND S. JUNG

its later statement declaring it indispensable to provide citizens with direct recall (cf. Schleswig-Holsteinischer Landtag 1995a: 111, 1995c: 7394). Due to different local institutional settings depending on municipality size, Schleswig-Holstein traditionally had no strong tendency to majoritarian or consensus democracy (Bathge et al. 2016: 179). Therefore, it is difficult to assess which role the institutional setting of Schleswig-Holstein played in the reform. Schleswig-Holstein adopted direct democratic instruments on the local level already in 1990 as the first German federal state after Baden-­ Württemberg (Kellmann 2010: 333f). While this prior reform might have made it easier for the government to justify direct recall, it did not have a direct effect on quorums, which were set much higher for recall.9 Legislators in Schleswig-Holstein may have considered the early experiences with direct recall in Brandenburg and Saxony, but in the parliamentary debates we found no evidence for policy-learning and cooperation was probably not a significant factor—Schleswig-Holstein had no partnerships with these East German states. Finally, Schleswig-Holstein might have started the wave of reforms by implementing direct democratic instruments in 1989 (Scarrow 1997: 463; Vetter 2009), but democratization was not a reason for the introduction of direct recall—the state was part of the Federal Republic of Germany since its constitution (Table 7.3). Later Institutional Changes Towards Direct Recall in Germany From 1995 to 2010 there were no significant institutional changes concerning direct recall except for Brandenburg raising the signature quorums in 1998 and Schleswig-Holstein lowering signature and approval quorums in 2002. In Brandenburg, the government decided to raise signature quorums after relatively frequent recall use in the past years which the press described as ‘mayor bowling’ (Bürgermeisterkegeln) (Witte 2001: 69f). Here, public pressure both from the press and the Association of Towns and Municipalities was probably the main reason for change, while the opposition criticized the reform for making direct recall unusable (ibid.: 70).

9  In a later reform in 2002 that decreased quorums for direct democratic instruments, the quorums for direct recall were adjusted to an equivalent level.

Saxony

First proposed by left-wing party. Conservative government implements high quorums.

Source: Own depiction

Democratization Experience with democratization favours direct recall.

Institutional Accountability of Intention to increase accountability context the executive and limit executive power but only preventively. Consensus Part of reform to consensus democracy democracy but choice of high quorums. Direct democracy Reference to direct democratic instruments but no correspondence with quorums. Policy-­ Policy-learning First state to implement direct recall. learning Cooperation encourages the choice of high quorums.

Party ideology

No competition concerning direct recall but high quorums preserve dominance of governing party on municipal level. Political scandal No references to political scandals.

Factor

Actors’ Party interests and competition ideologies

Dimension

Schleswig-Holstein

Reference to direct democratic instruments but no correspondence with quorums. No reference to direct recall in other states. No evidence for an influence of cooperation.

Reference to direct democratic instruments and equivalent quorums. No reference to direct recall in other states. Cooperation encourages the choice of low quorums. Experience with democratization favours direct recall.

No influence of experience with democratization.

No references to political scandals. First proposed by left-wing party. Left-wing government implements intermediate quorums. Intention to increase accountability and limit executive power. No consensus democracy.

No references to political scandals. First proposed by left-wing party. Left-wing government implements low quorums. Intention to increase accountability and limit executive power. No consensus democracy.

No competition concerning No competition concerning direct recall. direct recall.

Brandenburg

Table 7.3  Factors for institutional change towards direct recall in three German Federal states 7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

131

132 

B. GEIßEL AND S. JUNG

In contrast, the lowering of signature and approval quorums in Schleswig-Holstein was closely related to concurrent changes of quorums for direct democratic instruments. Both signature and approval quorums for direct recall were set to correspond to the reduced approval quorum of 20% for popular votes (Schleswig-Holsteinischer Landtag 2002). North Rhine-Westphalia became the fourth federal state to implement direct recall in 2011 after a disaster in Duisburg in July 2010 and the subsequent failure of indirect recall in the municipal council (Böhme 2012; Serdült and Welp 2017). The political scandal led to high public pressure on the state government mainly built by a citizen initiative that collected thousands of signatures for the withdrawal of the CDU mayor of Duisburg, Adolf Sauerland. Furthermore, party ideology played a significant role. The minority government of SPD and the Greens that had replaced the coalition of CDU and FDP already pledged to introduce direct recall in their coalition agreement just a few weeks before the disaster (SPD NRW and Bündnis 90/Die Grünen NRW 2010). Direct recall was enacted based on a revised bill drafted by the Left (Die Linke) who supported the minority government in this matter (Landtag Nordrhein-Westfalen 2011: 3316–3322). The CDU strictly opposed direct recall, accusing the government of creating a ‘Lex Sauerland’ only to remove the CDU mayor. While the coalition rejected this accusation by referring to their coalition agreement, party competition may have contributed to the implementation of direct recall, because CDU mayors led most municipalities in North Rhine-Westphalia at that time. Thus, there were more opportunities than risks for the SPD when implementing direct recall. Finally, policy-learning and cooperation had some influence on the choice of the signature ­quorums, which were graded according to municipality size following the example of partner state Brandenburg (ibid.: 3318). The most recent institutional change towards direct recall took place in Thuringia in 2016 after the first coalition between the Left, SPD, and Greens came into power in 2014. Party ideology was a primary factor for implementing recall. The Alliance for More Democracy in Thuringia (Bündnis für mehr Demokratie in Thüringen) formed by the coalition partners together with a broad range of interest groups induced and starkly influenced the implementation of direct recall. This advocacy coalition had already been successful in building public pressure on the previous CDU governments to liberalize the initially strict rules for direct democratic instruments by way of two popular initiatives in 2000 and 2008 (Franke-Polz 2010: 345ff). After getting into power, the left-wing parties

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

133

heavily leaned on the Alliance to continue the reform of the municipal code. Following recommendations of the alliance, the coalition government enacted a new law that summarized and further liberalized the rules for direct democratic instruments, now also featuring direct recall (Thüringer Landtag 2016). However, the governing coalition set much stricter rules for direct recall with the signature quorum of 35% of eligible voters being the highest one in Germany to date.

7.5   Discussion Comparing the findings in Table 7.3, we see that there is no single path to implement direct recall followed by all federal states analysed. Nevertheless, some general tendencies and mechanisms can be identified, partly contradicting those we described in our analytic framework. In line with the literature on recall in Latin America and prior analyses of institutional change towards direct democracy and the direct election of mayors, we assumed that party competition is a strong factor for implementing direct recall. However, there is little evidence to suggest that state governments introduced recall to seek electoral advantage or break the dominance of the opposition on the municipal level. Generally, neither the government nor opposition seems to expect political gains from demanding or introducing direct recall. However, the case of Saxony suggests that governing parties dominating both state and municipal level have incentives to implement high quorums to avoid competition. Moreover, the more recent case of North Rhine-Westphalia shows some signs of party competition being a factor when direct recall is not embedded in a broader institutional reform, and the office of a mayor in a bigger city is at stake. Regarding political scandals, the first three instances of institutional change reveal no concrete public pressure to implement direct recall. However, the more recent case of North Rhine-Westphalia shows that political scandals can trigger institutional change towards direct recall even independent from broader reforms. We found party ideologies to play an important role even if direct recall was not an issue of fierce ideological antagonism in parliamentary debates. This more pragmatic treatment of direct recall notwithstanding, we find a clear preference of left-wing parties to demand and introduce direct recall in each case investigated. There is also a tendency for left-wing parties to favour lower quorums even if the recent example of Thuringia contradicts this finding.

134 

B. GEIßEL AND S. JUNG

We expected existing institutional norms and practices to influence institutional change towards recall. In each state, increasing the accountability of executives was an important justification for promoting direct recall. Legislators regarded direct recall, providing citizens with the sovereignty to withdraw electoral decisions, as a logical consequence of mayors holding the only directly elected executive office. Lower quorums were adopted depending on the extent the state government wanted to put limits on the mayors’ power. Thus, legislators perceived direct recall as necessary due to the new legitimacy conferred with direct election and the change in the municipal power structure relating to it, whereas its potential function as a safety valve or opportunity for citizen participation was less relevant. In our analysis, we did not find consensus democracy to encourage the implementation of direct recall. Direct recall was implemented both in consensual (Saxony) and majoritarian (North Rhine-Westphalia) institutional settings (cf. Holtkamp and Bogumil 2016: 30), and the role of direct recall of establishing citizens as an additional veto player was not an essential motivation for legislators. Direct democratic instruments were either already existing or introduced together with direct recall in each case. We observed references to direct democratic instruments in each legislative process and final legislation. These references notwithstanding, direct recall is attributed with a different quality and function which becomes apparent in the much higher quorums chosen by legislators. The wave of reforms towards recall in Germany and other countries around the world during the 1990s suggests that policy-learning could explain institutional changes, especially between federal states. We could find no explicit reference that policy-learning influenced the implementation of recall, but in some specific cases, experiences of other federal states influenced the choice of quorums. Furthermore, cooperation in the form of partnerships between federal states affected the choice of quorums depending on the institutional setting of the partner state that served as a role model. Whereas direct recall was first a unique East German phenomenon, West German states followed not much later. Thus, recent democratization might have set the stage for direct recall in Germany but did not factor in the following institutional changes. The fact that only two of five East German federal states implemented direct recall shortly after the transition

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

135

to democracy supports the argument that democratization created a favourable environment for institutional change but was not instrumental.

7.6   Conclusion Our analysis of institutional changes towards the direct recall of mayors in five German federal states highlights three significant factors for the introduction of direct recall and the shape of its institutional design. First, party ideology played a significant role as direct recall was advocated and implemented predominantly by left-wing parties. However, most of these left-wing parties did not stress direct recall as a new participatory opportunity but rather as a necessary addition to the direct election of mayors. Second, increasing the accountability and limiting the power of executive leaders  (mayors) provided significant motivations for implementing direct recall (cf. Wollmann 2004: 159, 161). Considerations of providing a safety valve for dissatisfied citizens or promoting citizen participation were less relevant. Third, direct recall was almost always part of broader institutional reforms of direct elections of the mayor, direct democratic instruments, or both. Accordingly, the institutional design of direct recall was influenced by direct democratic instruments implemented concurrently or by the anticipated increase of institutional power to the directly elected mayor. Only once, a federal state implemented direct recall because of a political scandal independently from broader reforms. In our framework to explain institutional change towards direct recall, we also considered party competition, political scandals, consensus democracy, as well as policy-learning and recent democratization as possible factors for the introduction of recall. When analysing parliamentary debates and the historical context of the reforms, we could only find limited evidence that these factors influenced the implementation of recall. Our findings provide some guidance for future research on institutional change towards recall and democratic innovations. While our analysis confirms that left-wing parties often promote democratic innovations such as direct recall, this does not necessarily mean that increasing opportunities for citizen participation is their main motivation. Paying more attention to different motivations might help to explain why left-wing parties in other context fail to implement democratic innovations. One important motivation for legislators was to increase the accountability of the executive and the (perceived) function of direct recall to limit executive power.

136 

B. GEIßEL AND S. JUNG

Consequently, direct recall always should be studied with its institutional context, which is underlined by the fact that direct recall in Germany was almost always part of broader democratic reforms. Future research should pay special attention to the (perceived) functions ascribed to these institutions in a specific political system and its anticipated effects on the distribution of power to explain institutional change towards direct recall and other direct democratic instruments.

References Bathge, T., Bogumil, J., Ebinger, F., Fuhrmann, T., Holtkamp, L., & Seuberlich, M. (2016). Zur anhaltenden Varianz kommunaler Entscheidungsstrukturen in Deutschland. In J.  Bogumil & L.  Holtkamp (Eds.), Kommunale Entscheidungsstrukturen in Ost- und Westdeutschland (pp.  179–189). Wiesbaden: Springer Fachmedien Wiesbaden. https://doi.org/10.1007/ 978-3-658-11847-1_7. Böhme, D. (2008). Die Abwahl von Bürgermeistern. Verwaltungswissenschaftliche Beiträge 38. Bamberg: Universität Bamberg. Böhme, D. (2012). Die direkte Abwahl von Bürgermeistern. Die Öffentliche Verwaltung, 65(2), 55. Christmann, A. (2009). In welche politische Richtung wirkt die direkte Demokratie? Rechte Ängste und linke Hoffnungen in Deutschland im Vergleich zur direktdemokratischen Praxis in der Schweiz. Baden-Baden: Nomos. Dolowitz, D. P., & Marsh, D. (2000). Learning from Abroad: The Role of Policy Transfer in Contemporary Policy-making. Governance, 13(1), 5–23. https:// doi.org/10.1111/0952-1895.00121. Egner, B. (2013). Stellung und Funktion der Räte im Bundesländervergleich. In B. Egner, M.-C. Krapp, & H. Heinelt (Eds.), Das deutsche Gemeinderatsmitglied (pp. 17–56). Wiesbaden: Springer Fachmedien Wiesbaden. Elkins, Z., & Simmons, B. (2005). On Waves, Clusters, and Diffusion: A Conceptual Framework. The ANNALS of the American Academy of Political and Social Science, 598(1), 33–51. https://doi.org/10.1177/0002716204272516. Franke-Polz, T. (2010). Kommunalpolitik im Freistaat Thüringen. In A. Kost & H.-G.  Wehling (Eds.), Kommunalpolitik in den deutschen Ländern: Eine Einführung (pp.  337–349). Wiesbaden: VS Verlag für Sozialwissenschaften. https://doi.org/10.1007/978-3-531-92034-4_17. Fuchs, D. (2007). Die Abwahl von Bürgermeistern  – ein bundesweiter Vergleich. KWI Arbeitshefte Nr. 14. Potsdam: Univ.-Verl. Geissel, B. (2017). Direct Democracy and Its (Perceived) Consequences: The German Case. In S. P. Ruth, Y. Welp, & L. Whitehead (Eds.), Let the People

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

137

Rule? Direct Democracy in the Twenty-first-century (pp. 157–170). Colchester: ECPR Press. Geissel, B., & Jung, S. (2018). Recall in Germany: Explaining the Use of a Local Democratic Innovation. Democratization, 25(8), 1358–1378. https://doi.org /10.1080/13510347.2017.1398735. Geissel, B., & Michels, A. (2018). Participatory Developments in Majoritarian and Consensus Democracies. Representation, 54(2), 129–146. https://doi.org/10 .1080/00344893.2018.1495663. Hall, P.  A. (2010). Historical Institutionalism in Rationalist and Sociological Perspective. In J.  Mahoney & K.  A. Thelen (Eds.), Explaining Institutional Change: Ambiguity, Agency, and Power (pp.  204–224). Cambridge and New York: Cambridge University Press. Haus, M. (2005). Lernen im Föderalismus? Die Reform der Kommunalverfassungen in Deutschland in institutionenpolitischer Perspektive. In M.  Haus (Ed.), Institutionenwandel lokaler Politik in Deutschland: Zwischen Innovation und Beharrung (pp.  56–84). Wiesbaden: VS Verlag für Sozialwissenschaften. https://doi.org/10.1007/978-3-322-80740-3_2. Holtkamp, L. (2005). Reform der Kommunalverfassungen in den alten Bundesländern—eine Ursachenanalyse. In J.  Bogumil & H.  Heinelt (Eds.), Bürgermeister in Deutschland: Politikwissenschaftliche Studien zu direkt gewählten Bürgermeistern (pp.  13–32). Wiesbaden: VS Verlag für Sozialwissenschaften. https://doi.org/10.1007/978-3-322-85142-0_2. Holtkamp, L. (2008). Kommunale Konkordanz- und Konkurrenzdemokratie: Parteien und Bürgermeister in der repräsentativen Demokratie. Wiesbaden: Verlag für Sozialwissenschaft. Holtkamp, L., & Bogumil, J. (2016). Ost- und westdeutsche Kommunen zwischen Konkordanz- und Konkurrenzdemokratie. In J.  Bogumil & L.  Holtkamp (Eds.), Kommunale Entscheidungsstrukturen in Ost- und Westdeutschland (pp.  7–47). Wiesbaden: Springer Fachmedien Wiesbaden. https://doi. org/10.1007/978-3-658-11847-1_1. Holtmann, E., Rademacher, C., & Reiser, M. (2017). Kommunalpolitik: eine Einführung. Wiesbaden, Germany: Springer VS. Kellmann, K. (2010). Kommunalpolitik in Schleswig-Holstein. In A.  Kost & H.-G.  Wehling (Eds.), Kommunalpolitik in den deutschen Ländern: Eine Einführung (pp.  325–336). Wiesbaden: VS Verlag für Sozialwissenschaften. https://doi.org/10.1007/978-3-531-92034-4_16. Klein, M., & Lüdecke, Y. (2018). Ent-Parteipolitisierung und faktischer Konkurrenzausschluss bei Bürgermeister- und Landratswahlen. Zeitschrift für Politikwissenschaft, 28(2), 125–146. https://doi.org/10.1007/s41358-0180134-3. Künzel, W. (2010). Kommunalpolitik in Brandenburg. In A. Kost & H.-G. Wehling (Eds.), Kommunalpolitik in den deutschen Ländern: Eine Einführung

138 

B. GEIßEL AND S. JUNG

(pp.  79–119). Wiesbaden: VS Verlag für Sozialwissenschaften. https://doi. org/10.1007/978-3-531-92034-4_5. Landtag Brandenburg. (1993a). Drucksache 1/1607. Retrieved September 19, 2019, from https://www.parlamentsdokumentation.brandenburg.de/starweb/LBB/ELVIS/parladoku/w1/drs/ab_1600/1607.pdf. Landtag Brandenburg. (1993b). Drucksache 1/1652. Retrieved September 19, 2019, from https://www.parlamentsdokumentation.brandenburg.de/starweb/LBB/ELVIS/parladoku/w1/drs/ab_1600/1652.pdf. Landtag Brandenburg. (1993c). GVBl. Nr. 7, S. 110. Retrieved September 19, 2019, from https://www.parlamentsdokumentation.brandenburg.de/starweb/LBB/ELVIS/parladoku/gvbl/1993/7.pdf. Landtag Brandenburg. (1993d). GVBl. Nr. 22, S. 398. Retrieved September 19, 2019, from https://www.parlamentsdokumentation.brandenburg.de/starweb/LBB/ELVIS/parladoku/gvbl/1993/22.pdf. Landtag Brandenburg. (1993e). Plenarprotokoll 1/63. Retrieved September 19, 2019, from https://www.parlamentsdokumentation.brandenburg.de/starweb/LBB/ELVIS/parladoku/w1/plpr/63.pdf#page=29. Landtag Brandenburg. (1993f). Plenarprotokoll 1/66. Retrieved September 19, 2019, from https://www.parlamentsdokumentation.brandenburg.de/starweb/LBB/ELVIS/parladoku/w1/plpr/66.pdf#page=41. Landtag Mecklenburg-Vorpommern. (1997). GVOBl. M-V.  Nr. 18, S. 694. Retrieved September 19, 2019, from https://www.landtag.nrw.de/portal/ WWW/dokumentenarchiv/Dokument/XKVGVB9718.pdf. Landtag Mecklenburg-Vorpommern. (2004). GVOBl. M-V.  Nr. 3, S. 61. Retrieved September 19, 2019, from https://www.landtag.nrw.de/portal/ WWW/dokumentenarchiv/Dokument/XKVGVB043.pdf. Landtag Nordrhein-Westfalen. (2011). Plenarprotokoll 15/34. Retrieved September 19, 2019, from https://www.landtag.nrw.de/portal/WWW/ dokumentenarchiv/Dokument/MMP15-34.pdf. Lenhof, A. (2013). Die Abwahl des Bürgermeisters. Frankfurt/M.: PL Acad. Research. Lijphart, A. (2012). Patterns of Democracy: Government Forms and Performance in Thirty-six Countries. New Haven: Yale University Press. Lorenz, A. (2016). Demokratisierung und ihre Herausforderungen an die Landespolitik. In A.  Lorenz, A.  Anter, & W.  Reutter (Eds.), Politik und Regieren in Brandenburg (pp.  25–41). Wiesbaden: Springer Fachmedien Wiesbaden. https://doi.org/10.1007/978-3-658-07226-1_3. Maaß, A. (2014). Kommunalpolitik und Kommunale Selbstverwaltung. In E. Jesse, T. Schubert, & T. Thieme (Eds.), Politik in Sachsen (pp. 209–228). Wiesbaden: Springer Fachmedien Wiesbaden. https://doi.org/10.1007/ 978-3-531-19110-2_10.

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

139

Mahoney, J., & Thelen, K. A. (Eds.). (2010). A Theory of Gradual Institutional Change. In Explaining Institutional Change: Ambiguity, Agency, and Power (pp. 1–37). Cambridge and New York: Cambridge University Press. Qvortrup, M. (2011). Hasta la Vista: A Comparative Institutionalist Analysis of the Recall. Representation, 47(2), 161–170. https://doi.org/10.1080/00344 893.2011.581067. Qvortrup, M. (2013). Direct Democracy. Manchester: Manchester Univ. Press. Reutter, W. (2016). “Ländersachen”  – Landesverwaltung und kommunale Selbstverwaltung. In A.  Lorenz, A.  Anter, & W.  Reutter (Eds.), Politik und Regieren in Brandenburg (pp.  167–186). Wiesbaden: Springer Fachmedien Wiesbaden. https://doi.org/10.1007/978-3-658-07226-1_10. Sächsischer Landtag. (1991). Drucksache 1/789. Retrieved September 19, 2019, from http://edas.landtag.sachsen.de/viewer.aspx?dok_nr=789&dok_art= Drs&leg_per=1&pos_dok=1&dok_id=undefined. Sächsischer Landtag. (1992a). Drucksache 1/1634. Retrieved September 19, 2019, from http://edas.landtag.sachsen.de/viewer.aspx?dok_nr=1634&dok_ art=Drs&leg_per=1&pos_dok=1&dok_id=undefined. Sächsischer Landtag. (1992b). Plenarprotokoll 1/43. Retrieved September 19, 2019, from http://edas.landtag.sachsen.de/viewer.aspx?dok_nr=43&dok_ art=PlPr&leg_per=1&pos_dok=1&dok_id=undefined. Sächsischer Landtag. (1993a). Drucksache 1/2939. Retrieved September 19, 2019, from http://edas.landtag.sachsen.de/viewer.aspx?dok_nr=2939&dok_ art=Drs&leg_per=1&pos_dok=1&dok_id=undefined. Sächsischer Landtag. (1993b). Plenarprotokoll 1/64. Retrieved September 19, 2019, from http://edas.landtag.sachsen.de/viewer.aspx?dok_nr=64&dok_ art=PlPr&leg_per=1&pos_dok=1&dok_id=undefined. Scarrow, S. E. (1997). Party Competition and Institutional Change: The Expansion of Direct Democracy in Germany. Party Politics, 3(4), 451–472. https://doi. org/10.1177/1354068897003004001. Schleswig-Holsteinischer Landtag. (1993). Drucksache 13/1479. Retrieved September 19, 2019, from http://lissh.lvn.parlanet.de/shlt/lissh-dok/ infothek/wahl13/drucks/1400/drucksache-13-1479.pdf. Schleswig-Holsteinischer Landtag. (1995a). Drucksache 13/2806. Retrieved September 19, 2019, from http://lissh.lvn.parlanet.de/shlt/lissh-dok/ infothek/wahl13/drucks/2800/drucksache-13-2806.pdf. Schleswig-Holsteinischer Landtag. (1995b). Plenarprotokoll 13/95. Retrieved September 19, 2019, from http://lissh.lvn.parlanet.de/shlt/lissh-dok/ infothek/wahl13/plenum/plenprot/XQQP13-95.pdf. Schleswig-Holsteinischer Landtag. (1995c). Plenarprotokoll 13/106. Retrieved September 19, 2019, from http://lissh.lvn.parlanet.de/shlt/lissh-dok/ infothek/wahl13/plenum/plenprot/XQQP13-106.pdf.

140 

B. GEIßEL AND S. JUNG

Schleswig-Holsteinischer Landtag. (1996). GVOBl. Nr. 2, S. 33. Retrieved September 19, 2019, from http://lissh.lvn.parlanet.de/shlt/lissh-dok/ infothek/gvb/1996/XQQGVB962.pdf. Schleswig-Holsteinischer Landtag. (2002). GVOBl. Nr. 8, S. 126. Retrieved September 19, 2019, from http://lissh.lvn.parlanet.de/shlt/lissh-dok/ infothek/gvb/2002/XQQGVB028.pdf. Serdült, U. (2015). The History of a Dormant Institution. Representation, 51(2), 161–172. https://doi.org/10.1080/00344893.2015.1056219. Serdült, U., & Welp, Y. (2017). The Levelling Up of a Political Institution: Perspectives on the Recall Referendum. In S. P. Ruth, Y. Welp, & L. Whitehead (Eds.), Let the People Rule? Direct Democracy in the Twenty-first-century (pp. 139–156). Colchester: ECPR Press. Schmidt, C. (2018). Die Entwicklung von Bürgerbegehren und Bürgerentscheid seit 2016. KommJur, 5, 165–169. Shipan, C.  R., & Volden, C. (2008). The Mechanisms of Policy Diffusion. American Journal of Political Science, 52(4), 840–857. https://doi. org/10.1111/j.1540-5907.2008.00346.x. SPD NRW, & Bündnis 90/Die Grünen NRW. (2010). Gemeinsam neue Wege gehen. Koalitionsvertrag zwischen der NRWSPD und Bündnis 90/Die Grünen NRW. Retrieved September 19, 2019, from https://www.nrwspd.de/wp-content/uploads/sites/2/2010/07/doc_30009_2010713112745.pdf. Thüringer Landtag. (1993). GVBl. Nr. 23, S. 501. Retrieved September 19, 2019, from www.parldok.thueringen.de/ParlDok/dokument/8681/gesetz_und_ verordnungsblatt_nr_23_1993.pdf. Thüringer Landtag. (2002). GVBl. Nr. 14, S. 467. Retrieved September 19, 2019, from http://www.parldok.thueringen.de/ParlDok/dokument/20291/gesetz_ und_verordnungsblatt_nr_14_2002.pdf. Thüringer Landtag. (2016). GVBl. Nr. 9, S. 506. Retrieved September 19, 2019, from www.parldok.thueringen.de/ParlDok/dokument/60185/gesetz_und_ verordnungsblatt_nr_9_2016.pdf. Vatter, A., & Bernauer, J. (2009). The Missing Dimension of Democracy: Institutional Patterns in 25 EU Member States between 1997 and 2006. European Union Politics, 10(3), 335–359. https://doi.org/10.1177/ 1465116509337828. Vetter, A. (2009). Citizens Versus Parties: Explaining Institutional Change in German Local Government, 1989–2008. Local Government Studies, 35(1), 125–142. https://doi.org/10.1080/03003930802574524. Volkskammer. (1990). Gesetzblatt der DDR I., Nr. 28, S. 255. Retrieved September 19, 2019, from http://www.verfassungen.de/ddr/kommunalverfassung90.htm. Welp, Y. (2016). Recall Referendums in Peruvian Municipalities: A Political Weapon for Bad Losers or an Instrument of Accountability? Democratization, 23(7), 1162–1179. https://doi.org/10.1080/13510347.2015.1060222.

7  EXPLAINING INSTITUTIONAL CHANGE TOWARDS RECALL IN GERMANY 

141

Welp, Y. (2018). Recall Referendum Around the world. In L. Morel & M. Qvortrup (Eds.), The Routledge Handbook to Referendums and Direct Democracy (1st ed., pp. 451–463). Routledge. https://doi.org/10.4324/9780203713181-27. Welp, Y., & Milanese, J. P. (2018). Playing by the Rules of the Game: Partisan Use of Recall Referendums in Colombia. Democratization, 25(8), 1379–1396. https://doi.org/10.1080/13510347.2017.1421176. Welp, Y., & Serdült, U. (Eds.). (2014). La dosis hace el veneno: análisis de la revocatoria del mandato en América Latina, Estados Unidos y Suiza. Quito: Consejo Nacional Electoral. Welp, Y., & Whitehead, L. (2020). Recall: Democratic Advance, Safety Valve, or Risky Adventure? In Y.  Welp & L.  Whitehead (Eds.), Representation and Recall – Promise and Pitfalls.Wiesbaden: Springer Witte, J. (2001). Der kommunale “Recall” in Deutschland: erste Anwendungserfahrungen. Zeitschrift für Parlamentsfragen, 32(1), 57–71. Wollmann, H. (2004). Urban Leadership in German Local Politics: The Rise, Role and Performance of the Directly Elected (Chief Executive) Mayor. International Journal of Urban and Regional Research, 28(1), 150–165. Wollmann, H. (2008). Reformen in Kommunalpolitik und-verwaltung. Wiesbaden: VS, Verl. für Sozialwiss. Woyke, W. (2013). Stichwort: Wahlen. Wiesbaden: VS Verlag für Sozialwissenschaften. https://doi.org/10.1007/978-3-531-19928-3.

CHAPTER 8

Recall Practices in Central and Eastern Europe: From Citizen Accountability to Partisan Account Settling Sergiu Mișcoiu

8.1   Introduction In Central and Eastern Europe (CEE), revised electoral codes and the other related procedures have been gradually established after 1989. Recall principles and procedures were generally among the last to be adopted and this for two reasons. Firstly, there was an inevitable delay in establishing new institutional and electoral structures after the fall of the communist regimes, and the local political frameworks were most frequently among the last to be reconfigured (Holmes 1997). Naturally, after the introduction of political pluralism, the main preoccupation was the organization of the first free general elections, followed by the making and the adoption of a new (or radically revised) Constitution. In most cases, it was only after these first steps that local, county and/or regional elections were organized, based on incipient (and naturally incomplete) electoral codes. Consequently, for instance, the first local elections took place as late

S. Mișcoiu (*) Department of International Relations and German Studies, Babeş-Bolyai University, Cluj-Napoca, Romania © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_8

143

144 

S. MIȘCOIU

as in October 1991 in Bulgaria, in February 1992 in Romania, in 1994 in Ukraine and in 1995 in the Republic of Moldova. Secondly, delays arose because recalling a popularly elected official was perceived as being at least problematic, as this came after half a century of authoritarian appointment of all the decision-makers, from local councillors to the head of the state. The very idea of meddling with the will of the people as expressed through their votes and of initiating a recall procedure encountered strong opposition as it reminded citizens of the top-down practices of the past. This opposition was stronger in the South-Eastern European countries (Romania, Bulgaria, Serbia, Albania) and in the former Soviet countries (Latvia and Ukraine, for instance), where the democratic enthusiasm of the early 1990s prevented a solid reflection about the virtual situations where recall could have been needed precisely as tool for democracy (McFaul 2002). In most cases, the procedures allowing for mayoral recall came later than the electoral codes themselves and were subjected to parliamentary (and to a lesser extent to public) debates. There was no consensus with regard to the right, the need, the opportunity, the morality and the degree of democracy involved in the principle of recall itself, nor with regard to other collateral aspects. In what follows, I will briefly discuss and contextualize these disputed aspects, before showing the progressive shift from a legal-administrative approach to a strategic-operational practice that I will wider illustrate by exploring the case of the 2017 former Chișinău’s mayor recall.

8.2   Why Could a Mayor Be Recalled? In the CEE countries, the institution of recall was copied and adapted from different other models, mainly from the German and the US legislative and procedural frameworks. However, the nature and the meaning of the mayoral recall were under debate. Among the key issues, the reasons allowing for the defenestration of a sitting mayor became a particular source of initial concern for the leaders of these new democracies (Mișcoiu and Păun 2016). In Hungary, the early 1990 debates under the Jozsef Antal reformist government first dealt with the differentiation between the American-like practice of impeachment and the quasi-unaccounted-for institution of recall (which had existed in the ‘Socialist’ Constitution of 1949, a copy of a provision in the 1936 Soviet Constitution, though it was never activated

8  RECALL PRACTICES IN CENTRAL AND EASTERN EUROPE: FROM CITIZEN… 

145

under communist party rule) (Visegrády 1992). It was during that period that the Central and Eastern Europeans internalized the logic according to which while impeachment was exceptional and determined by some acts of constitutional violation, recall was a tool in the hands of the local community in order to withdraw confidence from an inefficient and careless mayor. As Hungary was a parliamentary republic and impeachment was not applicable, the Hungarian MPs concentrated on recall, which was nevertheless also considered in the light of the signification of impeachment: the violations of the law in the capacity of mayor prevailed over the lack of competence in explaining why mayors could be recalled by their citizens (Trabucco 2017). Concerning the causes that might justify a recall process, there were debates between the defenders of the sacredness of the full term mandate, who intuitively feared the transformation of recall into political vendetta, and promoters of the citizens’ right to a minimal good governance and effective administration. In Romania, parallel debates displayed a similar cleavage between these two positions (Mișcoiu 2012: 22–24). The parliamentary debates sought to distinguish between dismissal (undertaken by the courts in the event of a mayor’s illegal actions in his official capacity) and recall (completed via a popular vote and motivated by the severe managerial failures of a mayor). In practice, there was nevertheless a constant correlation between the two: one of the main reasons for instigating mayoral recalls has been the actual or potential indictment of mayors, even if legally they were not necessarily at risk of dismissal or in some cases not even temporarily suspended. All in all, these first debates of the early and mid-1990s influenced in a decisive way the politicians’ and the citizens’ attitudes with regard to recalls. As a consequence, there were countries, such as Poland, where recall initiatives became a constant preoccupation, being at the same time rather frequent and specifically framed. In Poland recall is one of the most intensively used instruments of direct local democracy. According to Serdült (2016), it was practised since the beginning of the democratic transformation and had the most profound political consequences. Out of all legislative and executive bodies that could be put under recall, the most frequent scenario concerned the recall of municipal councils (rada gminy). In the first cycle of local government (1990–1994), there were 76 recall initiatives although 28 out of them did not progress beyond the signature collecting phase (Węglarz 2013: 55). In some cases the consensus between initiators and local

146 

S. MIȘCOIU

administration was reached at early stage but, in most of the cases, the conflict ended at the polls. But there were also countries such as Bulgaria, Romania, Slovakia or Serbia (in Serbia, the recall procedure was initiated in 2004 and then abolished in 2008), where this instrument was underused and poorly institutionalized. For example, between 1998 and 2008, there were almost 200 recall activations in Poland and 150 in Hungary and less than 40 in Serbia and Bulgaria altogether (Smith 2013). Moreover, almost one fifth of the total localities in Poland have been involved in successful or failed recall attempts, compared with less than 10% in Slovakia (Report of The Venice Commission 2019).

8.3   Who Is Entitled to Recall a Mayor? In Estonia, one of the earliest post-Soviet democratizing countries, there was an initial rejection of the principle of recall. But later legislation allowed for the recall of a mayor during his term in office by the city council, followed by a local popular vote. This was a logical consequence of his/her initial election by the same council (councillors are directly elected by their constituents, while mayors are designated by the majority of the local councillors). In all the other countries of the region, mayors are directly elected. So if only the constituents of the locality where the respective mayor was elected had the right to recall him/her, the participation of other institutions, such as the local councils, the national governments or the judicial courts remained an issue of disagreement. In Croatia and in Lithuania, a rather restrictive legislation was initially adopted. This conditioned the initiation of the recall processes to approval or mandatory advice of various institutions. Such provisions made them unlikely to take place (Podolnjak and Gardašević 2013). Elsewhere, such as in Romania and Lithuania, recall was understood as the expression of the citizens’ will which should therefore be entirely at their disposal. On this view the role of the state was limited to providing the infrastructures needed, without limiting or conditioning the free expression of the popular will (Janušauskienė 2011: 44–65). The particular conditions legally required for a valid recall process also vary from one CEE country to another. More precisely, discussions concerned the minimum number or proportion of citizens required to sign the recall demand: the threshold for the validation of the results and the relationship between such a threshold with either the total vote cast or the

8  RECALL PRACTICES IN CENTRAL AND EASTERN EUROPE: FROM CITIZEN… 

147

number of votes the incumbent mayor obtained during the preceding elections. Poland reflects the complex and oscillatory treatment of these issues. Initially the 1991 Law of Local Referendums specified that 20% of the total population should sign for the recall proposition and required 50% of voter participation for the referendum to be validated. In the first version, the act neither specified rules for the conduct of the campaign nor was there any restriction regarding the number of recalls. The initiators had 60 days in order to collect the signatures of eligible voters after formally having announced the plan for holding a recall to the provincial election commissioner (Piasecki 2011: 128). In subsequent months, because of heavy criticism, the law was amended and these restrictive conditions were softened (to 10% and 30%, respectively). Given the huge volume of ensuing recall initiatives, the Parliament amended the law in 1995: it was made impossible to recall a mayor during his/her first six months and during the last year in office (Piasecki 2011). Then, a new law in 2000 extended the recall to the county and regional levels. Finally, a rather criticized 2005 law abolished the 30% turnout condition and replaced it with a minimum threshold of 3/5 of the total turnout in the prior regular elections (Smith 2013: 200). In spite of these controversies, there was a high amount of consensus in the CEE countries with regard to recall activations: from the outset, there was a clear majority in favour of the idea that the results of such votes should be mandatory and not consultative, as it was sometimes the case in Western Europe or elsewhere. However, in an original way, the Bulgarian legislation distinguished between two cases: when participation exceeded 50%, the results of the vote were mandatory, while when it was less than 50%, the results were optional. This should have prevented abuses of using recall procedures; but in practice, it severely limited the appetite for local civic initiatives (Dawson 2016).

8.4   Citizens’ Empowerment, or a Tool for Settling Political Accounts? Behind this consensus over the legal consequences of the valid public consultations for dismissing the incumbent mayors lay an issue that was from the outset the most difficult practical aspect of recall: is the recall process really about cutting short a corrupt or inefficient mayor unwanted by his constituents? Or is it a manoeuvre prepared by some adversaries to

148 

S. MIȘCOIU

­ olitically neutralize a normal incumbent? The ongoing practice of launchp ing recall initiatives in the CEE countries dramatizes this tension, especially in the case of major municipalities. And it also reveals a switch from the quasi-legal-administrative reasons for recalling mayors to the alternative strategic-political ones, which in terms of democratic consolidation could be interpreted as a setback. This is mainly due to the fact that citizens tend to overrate the importance of their city and, as a consequence, they are likely to believe that gaining control over it is an objective of the whole political establishment (Welp and Milanese 2018). Thus, citizens tend to act in order to block attempts at removing the mayor and to maintain control over the city. The people’s perception about the high importance of the political disputes over their city’s control influences their behaviour with respect to the recall. Previous experience shows that citizens’ tendency to inflate the role of their city and, implicitly, that of their mayors’ recall elections has been constant (LeDuc 2003). The importance people attach to their cities, especially when mayoral recall appears on the agenda, is likely to induce solidarity the incumbent officeholder and the consequent rejection of the recall process as an attempt to deprive them of their elected official. In Poland as well as in many other countries providing for recall as a direct democratic instrument, we can observe the practice moving up from the local level to medium-sized and bigger cities and in 2013, finally up to the capital. The Warsaw recall vote of 13 October 2013 is the latest and so far most spectacular case (Serdült and Welp 2017). On that occasion the Mayor Hanna Gronkiewicz-Waltz faced dismissal via a recall vote provoked by the conservative opposition. It was initiated by the mayor of one of the capital’s boroughs and rather small, local citizen groups opposing such policies as the rising price of local transport tickets, the cost of rubbish collection as well as the delays with the construction of the second metro line—all projects under the responsibility of the mayor of the political party in power also on the national level, the Civic Platform (PO). The opposition party, Law and Justice (PiS), later on jumped on board and contributed to the collection of 134,000 signatures as well (Serdült 2016). Donald Tusk’s Government and the Civic Platform (of which the incumbent mayor was one of the prominent leaders) succeeded in demobilizing the electorate by exposing the opposition’s charges (failures and delays in public works, utility price increases, etc.) as mere covers for the real political objective: the elimination of a possible successor of Mr Tusk

8  RECALL PRACTICES IN CENTRAL AND EASTERN EUROPE: FROM CITIZEN… 

149

as Prime Minister or even of the future presidential candidate of the civic liberals (Moskalewicz and Przybylski 2017: 202–208). The recall was timed for one year before national parliamentary elections were due. The ballot stated only one question: ‘Should Mrs. Hanna Gronkiewicz-Waltz be recalled from the position of President of Warsaw before the end of the term of office?’. However, due to the required turnout threshold, the recall vote was not valid (25.66% instead of the needed 29.01%) despite the fact that a large majority of 94% of the participants voted against her (Serdült 2016). More generally the switch from legal and incompetence considerations to a political instrumentalization of the recall process is often driven in part by the mayor’s partisan alignment with the national/regional governmental majority (Uleri 2002: 863–865). Indeed, several recall elections in Central and Eastern Europe have been prompted by the incumbent’s different political views, his/her opposition with regard to the central government or to his/her potential to become a national leader (Mișcoiu 2019a). This was the case recently with the 2018 recall process initiated in Slănic Moldova, a small but wealthy resort town in Romania, where the liberal mayor denounced the ‘harassment and torture’ inflicted by the social-democratic county and national governments in order to deprive him of his elected office. One mayor’s campaign against national or regional majorities is generally seen as a marker of his city’s particular orientation and distinct individuality. In 2011, in order to mobilize the citizens in his favour, the mayor of the Czech city of Liberec, Jan Korytář, countered a recall process by revealing in a public campaign a local ‘putsch’ prepared by the ‘famiglia’ of the opposition leaders hand in hand with various ‘real estate tycoons’, because he tried to cancel ‘illegal municipal contracts’ signed by the previous majority. Korytář claimed as the incumbent mayor he had ‘stood alone’ and defended the local interests against the dominant parties and lobby groups. His membership of the independent party Change for Liberec and his recall at the initiative of the mighty Social Democratic Party comforted his victimization strategy (Radio Praha 2011). Overall, at a regional level, there is an observable correlation between the mayors’ status as a member of the opposition and the probability of transforming recall into a political tool in the hands of the ruling parties. This is not a specific feature for the CEE countries but a constantly observable phenomenon in the case of recall processes (Mendelsohn and Parkin

150 

S. MIȘCOIU

2001), although it seems that it became particularly important in Eastern Europe. Finally, recall practices in Central and Eastern Europe have been affected by the citizens’ belief that there might be a wider regional, national or even international stake involved. In Slovakia, most recalls take place in the south-eastern regions where the Hungarian and the Roma minority proportions are important (Smith 2013). A series of successful recall votes resulted in banning the ethnic minority mayors and administratively replacing them temporarily with majority or alternative minority public officials. This could be interpreted as a setback in terms of minority rights and could have wider national and even international implications, as was the case in the mid-1990s under the nationalist government of Vladimir Mečiar, when relations with Hungary became regularly aggravated by such measures. This factor is more salient in major cities, such as country capitals, whose mayors are generally important national political leaders, or in divided societies, where issues such as the national political strategy or the country’s geopolitical position are regarded as contentious (Welp 2013, 2016). Generally, citizens are likely to reject governmental attempts to recall mayors who support alternative geopolitical orientations that run counter to the government’s direction in foreign policy. In the following section, I will offer a detailed analysis of the 2017 (failed) recall election of the mayor of Chișinău (Republic of Moldova), which is symptomatic of wider CEE and post-Soviet political tendencies, in particular in terms of societal context and long-term consequences.

8.5   When a Mayoral Recall Bid Becomes a Geopolitical Contest: The Case of Chisi̦ nău The Republic of Moldova is one of the ex-Soviet countries that has been facing numerous challenges related to the very definition of its collective identity, as it is divided between a Romanian (Moldovan)-speaking majority, a Russian-speaking large minority and several other ethno-linguistic minorities (mainly Ukrainian and Gagauz), not to mention more general transition-related difficulties (Henry and Mișcoiu 2016: 209–214). After the break-up of the USSR, the Republic of Moldova obtained its independence but had to immediately pay the price of a short yet traumatic war with the secessionist region of Transnistria, later on transformed into a

8  RECALL PRACTICES IN CENTRAL AND EASTERN EUROPE: FROM CITIZEN… 

151

frozen conflict and a permanent casus belli with Russia (Secrieru 2011: 241–243). Following a first period of relative dominance of the largely anti-Soviet pro-Romanian majority (1991–1996), Moldova passed successively from a more consensual but still pro-Western government (1996–1999) to a decade in which the Communist Party (PCRM) built its hegemony in a semi-consolidated authoritarian regime that reoriented Moldova geopolitically towards the Russian Federation (Way 2003; Quinlan 2004; March 2007: 599–601). In this context, the victory of the Liberal Party’s Vice-President, 29-year-old Dorin Chirtoacă, in the second round of 2007 municipal elections held in Chișinău (more developed and more westernized than the other Moldovan cities and regions) and his subsequent re-election in 2011 and 2015 with the support of all the pro-European parties was perceived as a major setback for the PCRM and the other pro-Kremlin political forces. Chirtoacă’s initial win also contributed to the successful contestation of the April 2009 parliamentary election results and the organization of new elections (in September), which finally ended with the victory of a pro-Western coalition. Thereafter, Moldova officially re-embarked on its pro-Western course. In spite of its heterogeneity, this governmental coalition resisted and implemented a programme of liberal reforms inspired and supported by the European Union, then Moldova became the first post-Soviet republic (except for the Baltic states) to be invited to sign the EU pre-accession agreement, in November 2013 (Mișcoiu 2014: 431–446). In 2014–2015, two key players emerged: (i) within the opposition, the Socialist Party (PSRM), a dissident group of the PCRM, led by the populist leader Igor Dodon, who progressively swallowed the outdated Communist group, and (ii) within the majority, the Democratic Party (PDM), led by the businessman and media tycoon Vladimir Plahotniuc, who unofficially started to gain control over the parties of the pro-­ European coalition, and over an increasing number of state institutions. The November 2015 parliamentary elections resulted in an almost hung assembly; the new governmental coalition, consisting of the same three parties as before, faced numerous difficulties in reaching consensus: no less than three different Prime Ministers and two other caretaker Prime Ministers held office from December 2015 to December 2017. In November 2016, the first presidential elections were organized after a constitutional reform that introduced the direct popular election of the President of the Republic. This involved a clash between the candidate of the new pro-European Party for Action and Solidarity(PAS), Maia Sandu

152 

S. MIȘCOIU

(endorsed by the remaining members of the incumbent pro-European coalition), and the Socialist candidate, Igor Dodon (endorsed by the remaining members of the PCRM). Although officially endorsing Maia Sandu, Vlad Plahotniuc and his wide networks of influence left numerous channels of communication open with Igor Dodon, while Sandu was much more bitterly criticized than Dodon in the Plahotniuc-controlled media during the campaign. As expected, Dodon finally won with a 6% margin, but lost by a narrow difference in the city of Chișinău, confirming the well-entrenched political dissonance between the capital and the rest of the country. This threw Moldova in a situation of cohabitation between the Socialist President and the Democrat-led legislative majority, which oscillated between phases of apparent inter-institutional conflict and phases of appeasement and consensus. In this context, in early 2017, an important corruption scandal erupted in relation to a contract for the management of Chișinău’s pay parking system. The granting procedure was contested as unlawful and opaque, while the beneficiary of the contract was denounced as a ‘ghost company’. In spite of denying any knowledge about this contract, Mayor Chirtoacă was also indicted in May 2017 (he was arrested and then placed under judicial control, meaning that he was obliged not to leave the Capital and to be available at any time if summoned by the investigators). In July, the Court suspended the mayor from office pending the investigation. As Chirtoacă refused to resign, the PSRM gathered signatures for a local recall vote. According to the Electoral Code, the mayor was to be recalled if this was favoured by more voters than the ones who had voted in his favour at the previous elections and, simultaneously, if they represented at least half of the total number of voters who had participated in the original election. This meant his recall required about 33% of the total number of registered voters. Because of that, the strategies of the two camps were different: the pro-recall camp wanted to mobilize the voters by showing them the need to replace the mayor, while the anti-recall coalition campaigned in favour of boycotting the vote. On 19 November 2017, only 17.5% participated in the recall vote (of whom 87% favoured the recall). Consequently, the revocation was invalidated. The failure of the recall bid or, in other words, the success of the boycott campaign involved a wide range of factors that I have discussed in detail elsewhere (Mișcoiu 2019b). For the purpose of this chapter, they can be condensed into two main factors: the perceived importance of the

8  RECALL PRACTICES IN CENTRAL AND EASTERN EUROPE: FROM CITIZEN… 

153

political control over the Capital city and the perception of the geopolitical issues at stake. Moldova is one of the countries with the lowest levels of confidence in public institutions and political parties. This is not only because the dominant public perception is that they are unable to properly fulfil their tasks but also because of their high degree of instability and inconsistency (IPP 2015). In Moldova, the dominant opinion is that both institutions and parties are extensively manipulated by interest groups which unofficially control them and that those interest groups are connected with foreign lobbying networks, especially those belonging to pro-Kremlin Russian oligarchs (IRI 2017). Nevertheless, most mayors benefit from their local entrenchment and are more trusted than any other officials or politicians, especially when their position or orientation opposes those of the central political establishment (LKP 2017). Under these circumstances, winning a third term in office in particularly difficult political conditions offered to Mayor Chirtoacă the chance of becoming the hope for change and resistance against the ‘anti-national’ forces. The major importance that citizens attach to the political orientation of their national capitals is often considered to be a key factor influencing their voting behaviour (Qvortrup 2011). One of Moldova’s particularities is the overwhelming importance of its capital, Chișinău, where almost one quarter of this country’s total population lives. The empirical research I conducted in the aftermath of the November 2017 vote found that citizens view their mayors as much more legitimate than the other elected or non-elected officials and that they were expected to directly represent and defend the interests of the local community. They also consider the real purpose of the recall procedure is not to counter frauds or corruption, but just to take control of the Capital. This widely explains why the 2017 boycott strategy was largely successful, and the mayor was reinstated by his citizens. Secondly, Moldovan citizens attributed a huge geopolitical importance to the possible replacement of the pro-Western mayor with a pro-Russian one. The outbreak of this corruption affair, which led to the indictment and suspension of Dorin Chirtoacă and ultimately the practice of recall, was initially presented by the media (and perceived by the public opinion) as yet another indication of the endemic corruption that permeates Moldovan institutions. However, perceptions rapidly switched to seeing the affair as a settling of scores between different political factions and between rival institutions. As the recall campaign advanced, it was even

154 

S. MIȘCOIU

rebranded as a critical choice between the West (boycotting the vote) and the East (voting for Chirtoacă’s recall). It is a widely held public belief that Dorin Chirtoacă resisted throughout time not mainly because of his administrative skills but rather because of his undisputable pro-Western and pro-Romanian commitments (LKP 2017). The field research showed that the ‘geopolitical factor’ proved critical, as the boycotters widely internalized the argument of the ‘decisive outcome’ of the recall vote for the ‘future of democracy in Moldova’. They detected a ‘putsch attempt against the elected mayor’ as ‘an extension of Russia’s new war against Eastern Europe’. While the anti-corruption argument was considered either fake or discretionarily applied, most citizens perceived an operation of political take-over and power projection disguised under the veil of a pro-integrity campaign. This geopolitical factor proved decisive: as fantastic as it may have appeared to some of the participants in the recall contest, the argument that the vote would decide not just the removal from office of an allegedly corrupt politician but the historic fate of the Republic of Moldova dissuaded the boycotters from going to the polling stations, and so ensured the failure of the recall campaign.

8.6   Conclusions The 2017 Chișinău example highlights the fact that recall processes in the CEE countries are rarely all about the mayor’s administrative activity and moral integrity. More often, they are perceived as being hyper-politicized and disrespectful of the factual evidence revealed during the recall campaign. The situation varies from one country to another, with Poland being the champion in terms of the number of recall attempts and of the complexity of the legislative procedures and the degree of ‘localness’ of these processes. Nevertheless, the trend towards the transformation of these municipal recalls from instruments for local direct democracy into arenas or even battlegrounds for mainstream political competition can be observed transversally across the entire CEE and post-Soviet region. Further research is needed in order to better explain and more accurately contextualize the reasons of this at least relative degradation of a formerly promising democratic instrument—especially in the contemporary conditions of contestation of representative democracy and of reconfiguration of the various forms of direct democracy. And, finally, the tension between local and national legitimacy may help explain the general return to mechanisms of grass-roots accountability, more favourable to the

8  RECALL PRACTICES IN CENTRAL AND EASTERN EUROPE: FROM CITIZEN… 

155

locally entrenched elected officials. Hence, the value of studying the evolution of direct recall practices in Central and Eastern Europe.

References Dawson, J. (2016). Cultures of Democracy in Serbia and Bulgaria: How Ideas Shape Publics. Oxford: Routledge. Henry, V., & Mișcoiu, S. (2016). Le discours politique et la quête identitaire en République de Moldavie. In S. Mișcoiu & N. Păun (Eds.), Intégration et désintégration en Europe Centrale et Orientale. Cahiers FARE no. 9 (pp. 221–254). Paris: l’Harmattan. Holmes, L. (1997). Post-Communism. An Introduction. Durham, NC: Duke University Press. IPP. (2015). Barometrul de Opinie [Barometer of Public Opinion]. Institutul de Politici Publice. Retrieved from www.ipp.md. IRI. (2017, February–March). Opinion Survey on the Trust in Institutions. International Republican Institute  – Magenta. Retrieved from www.iri.org/ country/moldova. Janušauskienė, D. (2011). Post-Communist Democratisation in Lithuania: Elites, Parties, and Youth Political Organisations: 1988–2001. Amsterdam: Rodopi. LeDuc, L. (2003). The Politics of Direct Democracy. Toronto: Broadview Press. LKP. (2017, May). Extensive Research on the Public Support for Institutions. Lake Research Partners. Retrieved from www.lkpsec.com. March, L. (2007). From Moldovanism to Europeanization? Moldova’s Communists and Nation Building. Nationalities Papers, 35(4), 601–626. McFaul, M. (2002). The Fourth Wave of Democracy and Dictatorship: Noncooperative Transitions in the Postcommunist World. World Politics, 54(2), 212–244. Mendelsohn, M., & Parkin, A. (Eds.). (2001). Referendum Democracy. Citizens, Elites and Deliberation in Referendum Campaigns. Palgrave Macmillan. Mișcoiu, S. (Ed.). (2012). Transitions et démocratisation en Roumanie. Illusions, mythes et défis/Transitions and democratisation in Romania. Illusions, Myths and Challenges. Cluj-Napoca: Printing House of the Foundation for European Studies. Mișcoiu, S. (2019a). Le spectre du peuple politique: l’éternelle transition roumaine. In C.  Delsol & G. de Ligio (Eds.), La démocratie dans l’adversité. Enquête international (pp. 925–944). Paris: Editions du Cerf. Mișcoiu, S. (2019b). ‘Never Just a Local War’: Explaining the Failure of a Mayor’s Recall Referendum. Contemporary Politics, 25(1), 47–61. Mișcoiu, S., & Păun, N. (Eds.). (2016). Intégration et désintégration en Europe Centrale et Orientale. Cahiers FARE no. 9. Paris: l’Harmattan.

156 

S. MIȘCOIU

Moskalewicz, M., & Przybylski, W. (2017). Understanding Central Europe. Oxford: Routledge. Piasecki, A. (2011). Twenty Years of Polish Direct Democracy at the Local Level. In T.  Schiller (Ed.), Local Direct Democracy in Europe (pp.  126–137). Wiesbaden: VS Verlag fiir Sozialwissenschaften. Podolnjak, R., & Gardašević, Đ. (2013). Directly Elected Mayors and the Problem of Cohabitation. Revus, 20(1), 79–96. Quinlan, P. (2004). Back to the Future: An Overview of Moldova under Voronin. Demokratizatsiya, 12(4), 485–504. Qvortrup, M. (2011). Hasta la Vista: A Comparative Institutionalist Analysis of the Recall. Representation, 47(2), 61–70. Radio Praha. (2011). Radio Praha Report. Retrieved from https://www.radio. cz/en/section/news/liberec-mayor-recalled-from-post. Secrieru, S. (2011). The Transnistrian Conflict  – New Opportunities and Old Obstacles for Trust Building (2009–2010). Southeast European and Black Sea Region Studies, 11(3), 241–263. Serdült, U. (2016). Recall in Poland. Paper prepared for presentation at the 24 IPSA World Congress, Poznan, Poland, July 22–28. Serdült, U., & Welp, Y. (2017). The Levelling Up of a Political Institution: Perspectives on the Recall Referendum. In S. P. Ruth, Y. Welp, & L. Whitehead (Eds.), Let the People Rule? Direct Democracy in the Twenty-first-century (pp. 139–156). Colchester: ECPR Press. Smith, M. (2013). The Local Referendum Revisited: An Old Tool For Inclusive and Effective Participation in New Democracies. In B.  Geissel & M.  Joas (Eds.), Participatory Democratic Innovations in Europe. Improving the Quality of Democracy? (pp. 9–32). Budrich: Opladen. The Venice Commission. (2019, July). Report on the Recall of Mayors and Local Elected Representatives. The Venice Commission. Retrieved from https:// www.venice.coe.int/webfor ms/documents/default.aspx?pdf file= CDL-AD(2019)011rev-e. Trabucco, F. (2017). The Evolution of Referendum Experience in Hungary. The Journal of Comparative Law, 12(1), 173–197. Uleri, P. V. (2002). On Referendum Voting in Italy: Yes, No, or Non-Vote? How Italian Parties Learned to Control Referendums. European Journal of Political Research, 41(6), 863–883. Visegrády, A. (1992). Transition to Democracy in Central and Eastern Europe: Experiences of a Model Country – Hungary. William & Mary Bill of Rights Journal, 1(2), 245–266. Way, L. (2003). Weak States and Pluralism: The Case of Moldova. East European Politics and Societies, 17(3), 454–482. Węglarz, B. (2013). Ewolucja lokalnej demokracji bezpośredniej w Polsce po 1989 roku. Kraków: Księgarnia Akademicka.

8  RECALL PRACTICES IN CENTRAL AND EASTERN EUROPE: FROM CITIZEN… 

157

Welp, Y. (2013). ¿Por qué Perú? Análisis de la revocatoria del mandato en perspectiva comparada [Why Peru? An Analysis of Recall in Comparative Perspective]. Elecciones, 12(13), 13–51. Welp, Y. (2016). Recall Referendums in Peruvian Municipalities: A Political Weapon for Bad Losers or an Instrument of Accountability? Democratization, 23(7), 1162–1179. Welp, Y., & Milanese, J. P. (2018). Playing by the Rules of the Game. The Partisan Use of Recall Referendums in Colombia. Democratization, 25(8), 1379–1396.

CHAPTER 9

The Recall Revival and Its Mixed Implications for Democracy: Evidence from Latin America Yanina Welp and Laurence Whitehead

9.1   Introduction Latin America is the world region that has been most prolific in regulating and activating recall, and Peru is the world’s most intensive user of recall elections, with almost half of the municipalities experiencing such votes between 1997 and 2014 (after which, a legal reform in 2015 had the effect of slowing down the number and success of activations; see Welp 2016). The Andean countries introduced the recall in a wave initiated by the Colombian new Constitution in 1991 and closed by the Bolivian one in 2009. However, the history of recall in the Latin American region began much earlier and is not restricted to the Andean countries. Why and how do political struggles over ideas and power shape institutional changes, notably the implantation of so-called democratic

Y. Welp (*) Albert Hirschman Centre on Democracy, Graduate Institute, Geneva, Switzerland L. Whitehead Nuffield College, Oxford, UK e-mail: [email protected] © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_9

159

160 

Y. WELP AND L. WHITEHEAD

i­nnovations? The topic is a key one, since so many democracies (both old and new) and even some non-democratic regimes—sometimes in response to growing citizen dissatisfaction, sometimes just in order to project an image of democratization—are undergoing shifts towards the inclusion of institutions of participation. In recent decades, first, the post-communist wave of democratization and then successive episodes of financial international crisis paved the way for alternative forms of political participation, not only in Latin America but also in many other areas of the world. This cast doubt on linear approaches to democracy as well as classical theoretical approaches to citizen participation. It indicated that the legal introduction of new mechanisms of participation might prove an additional route towards more democracy (see Almond and Verba 1963 and Norris 1999; for a critical view, see Whitehead 2011). In practice it is not that clear that “democratic innovations” (including recall within the definition) are necessarily and invariably instruments either of automatic or linear democracy advances, nor do political advocates with the power to introduce these new institutions always intend to deepen democracy as their main goal. In reality Institutions of Citizen Participation include a broad and diverse array of different mechanisms leading to different, sometimes contradictory results (Zaremberg and Welp 2019). In general, democratic innovations are the result of institutional change processes in which political actors struggle over two main things: different ideas concerning the value of citizen participation (and democratic legitimacy) and also the balance of power between actors in a political system. Thus, any institutional change is shaped in a process of negotiations, bargaining and competition in which advocates and opponents of increased political participation struggle over rival concepts (e.g. about what constitutes democratic legitimacy) as well as for power; and opportunity structures play an important role in explaining outcomes in the short and middle term (Scarrow 2001; Negretto 2013; Whitehead 2017). In particular, recall proposals bring traditional ideas about representative government into question—and of course they can only be legislated by the very officeholders whose tenure they weaken. The recall elections arrived in Latin America soon after the turn of the century progressive wave that disseminated the practice in the United States at the subnational level. By this time, it was first regulated in 1923 and 1933  in Córdoba and Entre Ríos provinces in Argentina and in Yucatán, Mexico, in 1938. It was also introduced in the Constitution of 1940 in Cuba (and subsequently reintroduced after the 1959 Revolution,

9  THE RECALL REVIVAL AND ITS MIXED IMPLICATIONS FOR DEMOCRACY… 

161

in the Socialist Constitution of 1976) (see Fig. 9.1). The dominant model in the Western Hemisphere is the direct recall. This variant is activated by a collection of signatures, whereas indirect recall is initiated by a legislature or a council after a process similar to an impeachment (but with the extra provision of a popular vote). Our focus here is on the direct model, which displays two contrasting outcomes or path: dormancy and activism. For long periods in the relevant provinces of Argentina and the comparable states of Mexico, dormancy has been the norm in striking contrast to the intensive use of recall promoted by partisan actors in Peru and Colombia. This chapter builds on previous findings with the aims of both identify different intraregional paths and to analyse under what conditions recall has an effect (positive or negative) on democracy. In doing so, three aspects arise. First, there are origins and terms of the legal provisions for recall; second, there are the actual arrangements that emerge; and third, the discursive level requires separate consideration since recall is not only an institution and a practice but also “a challenge” and a citizen’s demand. This threefold approach provides the basis for a concluding discussion about the relationship between recall and broader processes of democratization.

9.2   The Legal Provisions: Explaining Institutional Change The introduction of legal provisions of recall (or any other institutions of participation) does not necessarily mean that significant direct democracy outcomes are intended or expected. After all, if the legal requirements for activation are set sufficiently high, the actual effect can be to block rather than to advance such practices. On one hand the institutional rules can be designed to avoid rather than to promote activations, although on the other hand any durable provision for direct democracy needs to be embedded within the rule of law (see Guzmán Hernández 2014; Welp and Castellanos 2019). Various scholars have tried to reconstruct the pathways to the promotion of Mechanisms of Direct Democracy (MDDs) and recall initiatives. Early case studies focussed on the United States and suggested that the introduction of MDDs at the subnational level was to be explained by social pressure organized around the “progressive movement”, with the

162 

Y. WELP AND L. WHITEHEAD

goal of preventing corruption (Cronin 1989). In both the Swiss and the US cases, power struggles have been identified as central to at the introduction of MDDs and in both this has also been related to a process of democratization (Auer 2008). While the Argentinian provincial constitutions can also be explained in these ways, an additional factor seems to have been the desire to import the most modern ideas and practices. In this case, diffusion seems to be part of the explanation. Indeed in Latin America as a whole (“contagion”) is an additional factor that requires attention. Figure 9.1 shows the spread of recall provisions along time, displaying an increasing speed in recent times. The first case is registered in 1923 (Córdoba province in Argentina), the last one is Mexico 2019, from the national level. In nine cases the regulation comes from the national laws or Constitution (e.g. Bolivia, Colombia, Costa Rica, Cuba, Mexico, Panama, Peru and Venezuela), while in two of the three federal countries, the subnational units—provinces or states—decide on the introduction of the device. Most of the provinces in Argentina and the states in Mexico have done so. It has to be noticed that despite Brazil—the third federal country—being a well-known “laboratory” of participatory innovations, recall is not included at any level. And Uruguay, the most prodigious user of MDDs at the national level, has not regulated recall at any level. National regulations are quite diverse. First of all, in Latin America generally the direct model prevails. The exceptions are Costa Rica (where indirect recall is available at the local level to ratify the removal of executive authorities by a municipal Council), Panama (where recall is designed to resolve internal conflicts within a party, with the aim of forcing party discipline) and a few local cases in Argentina (such as the city of Bariloche, where recall is indirect and there was one activation on April 7, 2013, which ended with the removal of the mayor). Second, the authorities able to be removed through a recall device diverge. Interestingly, in only six countries in the world, all the elected authorities could be removed by direct recall elections and five are in Latin America (these cases are Bolivia, Cuba, Ecuador, Mexico, Venezuela and Taiwan). Then, there are cases in which the national constitution regulates recall only for the subnational level. These are the cases of Colombia and Peru. In Colombia only executive authorities can be removed by recall, while in Peru executive and legislative authorities could be removed by a vote. In federal countries, the regulation is coming from the subnational level as happens in Argentina and Mexico.

Fig. 9.1  Legal provisions to activate recall regulated from the national and subnational level in Latin American (1923–2016). Source: Own elaboration. (1) In Argentina recall is regulated at the provincial level in Chaco (1957), Chubut (1994), Córdoba (1923, 1987), Corrientes (1960), La Rioja (1986), Rio Negro (1988) and Santiago del Estero and Tierra del Fuego (1991); other provinces include it for their municipalities, namely, Entre Ríos (1933, not included in later constitutions), Neuquén (1957), Misiones (1958), San Juan (1986) and San Luis (1987). It is also included in Ciudad de Buenos Aires (1996) (Arques 2014). (2) In Mexico it is regulated in Oaxaca (1998), Morelos (2011), Guerrero (2013), Zacatecas (2014), Aguascalientes (2014), Nuevo León (2016) and Mexico City (2017). Interestingly the two attempts registered in Yucatán and Chihuahua led to interventions of the National Judicial Court and the declaration of illegality of the recall, which was removed from the two constitutions but introduced in many other states (Limón 2016)

9  THE RECALL REVIVAL AND ITS MIXED IMPLICATIONS FOR DEMOCRACY… 

163

164 

Y. WELP AND L. WHITEHEAD

The Spread of Recall in the Andes Looking at the national level, a pattern of regional diffusion emerged from the 1990s onwards, however even sharing similarities it seems more a result of the zeitgeist (the spirit of the time) than the result of specific variables shaping the institutional change. The similarities refer to the process: the Andean republics replaced (or even “refounded”) their constitutions, and some introduced recall elections together with various other mechanisms of direct and participatory democracy. This happened in Colombia in 1991, in Ecuador in 1998 and in 2008 (two constitutions were approved in this period), in Peru in 1993, in Venezuela in 1999 and in Bolivia in 2009. But while in Ecuador or Bolivia the demand for more participation was on the cards and new parties or coalitions were heading it, in Peru the social pressure is much less evident. Leftist and rightist coalitions introduced it. Even more, the Peruvian context was even not democratic, under Fujimori (Welp 2016; see Welp and Whitehead, Chap. 2, in this volume). Ecuador is an especially interesting case given that the recall election was identified as a direct answer to an institutional breakdown, and there was social pressure towards the introduction of the mechanisms. In February 1997, the country experienced the unconstitutional removal of President Abdalá Bucaram Ortiz. The Congress declared him mentally unfit to govern, in a context of strong social protests motivated by perceived high levels of corruption and increases in the cost of public services. Bucaram’s successor, interim President Fabián Alarcón, called for a referendum to amend the Constitution and introduced the recall device at the local level. One year later, a new constitution was approved introducing recall among other mechanisms of citizen control and participation (Castellanos 2014). However, the 1998 Constitution failed to provide the political system with legitimacy and stability, and a new constitution in 2008, promoted by President Rafael Correa (Alianza País), introduced recall at all levels and changed the regulation to make the activation easier (Welp and Castellanos 2019). Recall in Federal Countries: Argentina and Mexico The Argentinian provinces—as well as the Mexican states—later on displayed a wave of introduction of recall. In both cases it happened despite the national level did not introduce any mechanism of direct democracy. In Argentina it changed in 1994, although recall is not regulated at the

9  THE RECALL REVIVAL AND ITS MIXED IMPLICATIONS FOR DEMOCRACY… 

165

national level and other mechanisms do not allow the citizens to be initiators. In Mexico, the reform of 2012 included MDDs which can be activated by all actors, despite the concrete regulation that does not stimulate the use and the recall was introduced in 2019. In Mexico in a wave starting in 1994 and closing in 2016, all the states introduced mechanisms of direct democracy (the case of Yucatán (1938) appears as an isolated case deserving further research). The Mexican experience introducing MDDs is unique to explore political variables such as the influence of the party in government, social pressure, coalitions and majorities, among others, on explaining such institutional change. The study of Martha Sandoval systematizes it to explore to what extent the party in the government, the citizens demand for more direct participation, the democratic or non-democratic setting (many states introduced it before 2000) influences the change in regulation. Her findings show that 17 states introduced MDDs under the government of the Institutional Revolutionary Party (PRI), 9 states did it under the government of the National Action Party (PAN), 5 states under the government of the Democratic Revolutionary Party (PRD), while 1 state introduced MDD being governed by an independent candidate. It means that all parties were supportive to these mechanisms (however, the picture is different in the recent parliamentary discussion of recall, where the PAN was against; see Sect. 9.3). At the subnational level, the institutional change has been carried out both in unified government contexts and in divided government contexts. The influence of citizens in the introduction of MDDs at the subnational level was not very significant, for example, there is no evidence of citizens’ claims for recall or more direct democracy. Besides that, no single activation has succeeded in arriving at a vote until now and not all states regulated recall. It exists in the City of Mexico, Aguascalientes, Zacatecas, Sinaloa, Guerrero, Oaxaca and Nuevo León. Also, notice that three states in which it was regulated have derogated it as soon as an attempt of activation was registered. These are the cases of Yucatán, Chihuahua and Morelos. Summarizing, there is not a common path towards the introduction of recall. There is more diversity than factors in common in the cases at hand. Centralized and federal countries introduced it, with and without citizens’ demand, from both—both leftist governments (Venezuela, Ecuador and Bolivia) and centre-right governments (Colombia) and even in authoritarian contexts (Peru). Different combinations of factors can be identified, as the subnational experience in Mexico shows.

166 

Y. WELP AND L. WHITEHEAD

9.3   Evaluating Practices The data on recall use offers two contrasting outcomes or path: dormancy and activism (see Table  9.1). In the first group, it is included Cuba, Argentina, Mexico, Costa Rica and Panama. In the second group are placed the Andean countries. Fourth sets of—sometimes—interrelated variables have been offered to explain the frequency of recall attempts and votes: attitudinal variables, that is, citizen discontent; institutional variables, that is, the ease of activation; actor-centred variables, particularly partisan motivations; and the Table 9.1  Recall regulation and practice in Latin America Country

Type of recall

Argentina

Bolivia

Direct and 1923–1996 Executive and legislative indirect (several cases) authorities at the subnational level Direct 2009 All elected authorities

Colombia

Direct

1991

Executive authorities at the subnational level

Costa Rica Cuba

Indirect

1998

Executive authorities

369 1 (2012–2018) +300 ~41 (1994–May 2018) N/d 0

Direct

All elected authoritiesa

N/d

Ecuador

Direct

All elected authorities

Mexico

Direct

1940 and 1976 1998 and 2008 1938 to 2019 (several cases)

+900 79 (1998–2018) 2 0

Panama Peru

Indirect Direct

Venezuela Direct

Year of introduction

2004 1993

1999

Authorities

Executive and legislative authorities at the national and subnational level Legislative authorities Executive and legislative authorities at the subnational level All elected authorities

Attemptsa

Votes

12 ~3 (1923–2014)

N/d

N/d N/d 22,000 ~5304 (1997–2016) 167 10 (1999–2013)

Source: Updated from Welp and Milanese (2018: 1382) a The attempts refer to the commissions registered or kits sold (in the case of Peru) oriented to start a process. Numbers are approximate giving the difficulties of accessing to complete and accurate information

9  THE RECALL REVIVAL AND ITS MIXED IMPLICATIONS FOR DEMOCRACY… 

167

role of Electoral Management Bodies (EMB)  in allowing or preventing recall attempts, a role conditioned by the law and the level of autonomy of the institution.1 As for citizen discontent, the recall election is seen as offering a safety valve to resolve the problems of political representation (e.g. distrust in elected authorities and lack of accountability). Its activation is expected to be exceptional, occurring only in contexts of deep dissatisfaction with a government’s performance and/or high polarization in which a group of citizens would have sufficient incentives to engage in a process of signature collection (Whitehead 2018). However, far from proving exceptional, in some circumstances empirical evidence suggests that recall can become a relatively common practice (as in Peru between 1997 and 2013; see Welp 2016; Tuesta Soldevilla 2014). And still, the question remains open of why it is not activated in some contexts of high-level discontent. Then, the number of signatures and the time given to collect them, the reasons legally required to initiate a recall process or the actions to be taken if the authority is removed are all key components of the recall institutional design. For instance, the relatively low number of signatures and no reason required to activate it (recall as a political right) in Peru until 2013 explain why the mechanism has been so frequently activated there (and probably why it has slow down after an amendment despite the mechanism is highly popular among citizens). In contrast, a high signature requirement discouraged the use of this tool in Colombia before 2015. The third set of explanations is based on the incentives for political actors, what we define as partisan motivations. According to this argument, new or smaller parties might use the recall process to gain visibility, whereas larger parties might see a recall as an opportunity to challenge the party in power, particularly in the context of uncertainty regarding their own survival (Tuesta Soldevilla 2014; Welp 2016). The latter is expected in contexts of low political party institutionalization, accordingly we expect to find in such contexts a higher frequency of activation, driven by parties. But even in these cases, a level of citizen support is required to collect the necessary signatures as well as to succeed in removing an authority. This is the case for Colombia, where partisan motivations are behind most of recall attempts. Interestingly, there is no clear evidence suggesting that some parties (e.g. new parties vs. traditional) or movements (e.g. indigenous) are more likely to face recall elections. Even if some parties face 1

 This section is based on the arguments developed in Welp and Castellanos (2019).

168 

Y. WELP AND L. WHITEHEAD

more recalls than others, it is linked to the number of local governments in the hands of these parties (see Welp and Milanese 2018). Finally is the role of the electoral management body. Recall is commonly introduced in highly regulated settings. Its supervision and implementation is based in laws applied by specific institutions. The competences given to the EMB by law, together with its level of autonomy (understood as the extent to which laws and regulations are enforced without political intervention (Brinks 2012)), could have a strong influence in preventing or allowing activations. In other words, an EMB could be responsible for administrative and technical aspects of the oversight and validation of a proposed recall but could also be responsible for political judgment and interpretation. For instance, when the reason given for pursuing a recall must be demonstrated and evaluated by an electoral body, the independence of this body becomes crucial in enabling or preventing recall efforts. In extreme situations the electoral body would become an instrument of the executive in the application of recall against political “enemies” and the prevention of this application against allies. Such an example is offered by the attempt to activate recall against President Nicolás Maduro in Venezuela in 2016, systematically blocked by the Venezuelan National Electoral Council. Thus, dormancy can be explained by different specific, but in cases of high citizens’ dissatisfaction, it can be said that institutional designs on one hand and the intervention of the courts on the other play an important role in blocking usage. For instance, in Cuba only the delegates at the lowest state level can be removed by a recall election because the rest are elected by delegates and not by the people, what among other reasons, could explain while in Cuba recall is a “dormant” institution (see Guzmán Hernández 2014). In Argentina (see Arques 2014) and Mexico (see Limón 2016), the very few attempts of activating recall ended with a judicial process. Maybe not a surprise, the indirect recall in the city of Bariloche proceeded and the mayor was removed. The revocation is not activated more frequently by the existence of formal or informal obstacles that hinder or prevent it. Our recent research on Ecuador shows that recall has been transformed into an indirect instrument of political struggle and/or negotiation controlled by the EMB. The Ecuadorian National Electoral Council (CNE) is responsible both for implementing the formal legal requirements governing recall and for judging whether the reasons presented for the recall are sufficient. The new rules have granted power to non-elected authorities to evaluate and decide on the validity of a politically based request. Even

169

9  THE RECALL REVIVAL AND ITS MIXED IMPLICATIONS FOR DEMOCRACY… 

though some studies may find that the CNE has been controlled by actors close to government, we did not find evidence of its use against opponents. We did identify a strategy oriented to avoiding recall votes (Welp and Castellanos 2019). Diversity Within a Country: Bolivia and Colombia One step forward, if the national regulation can create a set of incentives to allow or avoid recall activations, looking at how it plays at the subnational level we find more differences. This can be observed in the cases of Bolivia (see Table 9.2) and Colombia (see Fig. 9.2) where recall attempts are not equally distributed within the country. In Bolivia, while in 57.9% of the total number of Beni municipalities, there were attempts of activating recall votes, only in 13.8 % of municipalities in Chuquisaca such attempts were registered. We need further research to understand these patterns which clearly stress that even if institutional designs play an important role this is not determinant (e.g. all Bolivian municipalities can activate recall under the same regulation). The map of Colombian attempts in 2017 (see Fig.  9.2) shows that attempts are especially—but not only—concentrated in the centre and the Caribbean coast, regions characterized by the presence of small ­municipalities and the strong presence of patronage structures. According to the literature, this region is particularly characterized by the presence of electoral machinery and patronage (Langbein and Sanabria 2013; Archer 1990). There electoral and political support occurs due to a delicate balance based on the sum of various political leaders (e.g. congressmen) or brokers (e.g. councillors and neighbourhood leaders) that mobilize the machinery “to the highest bidder” according to the potentialities of mobiTable 9.2  Bolivia 2018: Number of recall requests by departments and over the total of municipalities Department

Total number of municipalities

Municipalities with recall requests

%

La Paz Oruro Potosi Cochabamba Chuquisaca Tarija Santa Cruz

87 35 40 47 29 11 56

24 7 7 11 4 4 16

27.6 20 17.5 23.4 13.8 36.4 28.6

Source: Own elaboration based on Ascarrunz and Welp (2019)

170 

Y. WELP AND L. WHITEHEAD

Fig. 9.2  Municipalities with recall attempts in Colombia, 2017. Source: Welp and Milanese (2018: 1388)

lization and delivering of benefits (naturally also of the volume of this type of resources). Thus, these findings invite further reflection on the potential risks that well-organized special interests may capture and distort recall procedures (see Welp and Milanese 2018).

9  THE RECALL REVIVAL AND ITS MIXED IMPLICATIONS FOR DEMOCRACY… 

171

The Mexican Recall and Popular Consultation Law (2019) The Mexican recall and popular consultation law of October 2019 is the product of some very distinctive features of Mexican constitutional development. Ever since the 1917 Constitution, the crucial slogans have been “effective suffrage and no re-election”. But up to the end of the last century, this was accompanied by one-party dominance at all levels of government, both federal and local. To manage the “no re-election” part of the mantra, every executive officeholder and senator was elected for a single six-year term. But once multi-party competition and electoral alternation became the norm since the turn of the century, this inherited structure needed a major overhaul. It was, of course, the role of those in office to legislate the reforms that would alter their career prospects. In recent years the key incentive for embracing such reform was the possibility of including a re-election reward for incumbent legislators. With effect from 2018, that became possible (although the authors of the reform did not benefit, since the electorate turned out all the old-established parties). This exceptional background provides the context for the ambitious 2019 recall law (which to some extent involved transferring to the national-­ level ideas first adopted at the level of the Federal District in 2017). It is generally argued by expert observers that if re-election to office is to become the norm, then the term of office needs to be shortened (probably from six years to four). But since that was not included in the 2015 reform, it is no longer a politically saleable option. In this setting the danger is that elected legislators with six-year mandates and the possibility of re-election might easily become even more of a self-perpetuating elite removed from accountability to their voters. This goes far to explaining why a provision for recall half-way through these six yearly mandates has gained such widespread acceptance and will now be tried out—even for top executive officeholders such as the President and the state Governors, who are still barred from re-election.

9.4   Discursive Context

and Political Institutionalization

The diffusion and embedding of recall practices in Latin America has been a response to a variety of political and ideational challenges. Models and outcomes are correspondingly heterogeneous, but it is possible to identify a few main contexts and pathways. The implications for the quality and

172 

Y. WELP AND L. WHITEHEAD

durability of the encompassing democratic regimes vary in accordance with these political settings and their accompanying discourses and presuppositions. Thus, in Argentina the initial recall experiments were a response to ending of oligarchical dominance and the rise of the Radical Party (plus the vanguardism of the intellectual climate of the times) in the case of Córdoba, in 1923. However, ten years later, when was introduced in Entre Ríos the scenario was controlled by the militaries at the national level. A similar contradictory scenario can be observed in Mexico, as described above. In any case, both of these participatory bursts of enthusiasm were shortlived, and the accompanying institutional reforms were quickly deactivated (although they remained on the books). Similar considerations help explain the initial adoption and subsequent non-activation of recall processes in both the 1940 and the 1976 Constitutions in Cuba. A somewhat different context arose in Colombia after the constitutional reform of 1991. In that (early post-Cold War) setting, the priority was to broaden the popular base of the electoral regime, after the “National Front” pact between the Conservatives and the Liberals had given rise to a deep divorce between elected officeholders and Colombian voters, with anti-system crime and violence a looming alternative option. In such a setting, it was important not only to legislate for inclusionary political participation but also to develop that formula in practice in order to co-opt dissenters. The decay of the one party PRI regime in Mexico in the 1990s created some parallel incentives, although the initial effect was to motivate a reinvigoration of Congress and related representative institutions. It was only when they too become all visibly divorced from the people and mired in corruption and mounting social violence that MDD experiments came to the fore (as discussed more fully in the conclusion). The so-called pink tide elections in South America after 2000 represented a very different political and ideological agenda, and that was what gave rise to much of the recall activism recorded in this chapter. The essential difference arose from the impact of “neoliberal” reform policies in the 1990s and in particular how they were understood, and countered, by the parties of the left that returned to office with the help of a surge in primary product export earnings. From their perspective it had seemed all too possible that their pro-market opponents would become the natural electoral victors once neoliberal adjustments were made secure and that the left’s main sources of support—public employees, unionized workers, and rural sindicatos—would all be permanently hobbled. From that standpoint, the

9  THE RECALL REVIVAL AND ITS MIXED IMPLICATIONS FOR DEMOCRACY… 

173

unexpected good fortune of a favourable election was an opportunity to turn the tables. But to do so could not involve simply a return to the statist and corporatist status quo ante. If social power was to be re-appropriated from the private sector elites enthroned by marketization, some serious institutional re-engineering would be required. MDDs were a natural option in this setting. However, if at first the embrace of recall processes promised to create new avenues for bottom-up mobilization against privileged incumbents, after a while the left parties themselves became vulnerable to challenge from below. That both generalized the appeal of such practices and gave rise to backsliding, most notably in Venezuela, where in the absence of the charismatic Hugo Chávez recall procedures became a further threat to rule by his followers. This institutional and ideational panorama is still in flux, and diffusion effects continue to complicate the task of tracing the pathways to recall. Thus, today’s Mexico (where the initial recall legislation based in the capital city has only just spilt over to the national level) is out of sync with the retreat of the pink tide in the southern half of the subcontinent. In the ongoing turbulence, it is too soon to tell how much of the recent MDD innovation will prove durable. However, what this section does serve to demonstrate, at least, is that a full comparative analysis of recall experiments must extend beyond the legal and pragmatic aspects of the phenomenon and take into account the larger ideational (and indeed even ideological) dimensions involved. In many recent institutional crises, proposals for direct recall have been brought forward as a promise or as a demand. This pattern of behaviour is not confined only to Latin America. The gilets jaunes (yellow vests) in France provide a vivid illustration (see Chap. 3 in this volume). It also surfaces in unexpected locations within the Western Hemisphere. For example, it entered public discussion during the mid-2019 institutional crisis in Puerto Rico. Here, the governor of the island was forced into an unprecedented mid-term resignation due to a corruption scandal. Amid mass protests this gave rise to complex constitutional manoeuvres to identify a legitimate successor, during which a variety of proposals for a democratic revival, including recall, received attention for the first time. Neither in France nor in Puerto Rico is such a reform likely any time soon, but in both settings, it is now an option for consideration if easier solutions prove wanting. The credibility of these options is much strengthened by the speed and extent of the recent spread of actual recall experiments, especially in South

174 

Y. WELP AND L. WHITEHEAD

America. The necessary legal provisions have been introduced with increasing frequency since the 1980s (see Chap. 2 in this volume) and whereas in the past activations were mostly restricted to small municipalities in recent years they have spread to bigger units such as Lima and Bogota.

9.5   Discussion Scholarly research has associated the increasing use of MDDs in general with a crisis of representative democracy (Dalton et al. 2001). Within this overall context new Latin American democracies are often particularly susceptible, and the recall version of MDD initiatives has attracted particular interest because it can be seen as reinforcing, rather than competing with the accountability and responsiveness of the representative system. Indeed, to the extent that a recall campaign draws attention to some policy failure or governance dysfunction, it can be regarded not just as a democratic refinement but as a “safety valve” that may serve to protect the system from further damage. Certainly it has been a common experience of many Latin American democracies that if such problems are allowed to fester, they may eventually be addressed by undemocratic means. On this basis recall could help to head off a democratic breakdown. On the other hand, the idea of “common people” or individuals activating direct democracy has been challenged there is a strand in the academic literature (Kriesi 2006; Garret 2004; Serdült and Welp 2012) that rejects the assumption that recall necessarily reflects the unfiltered will of the “people”. This school points out that the activation of a recall process requires organization and resources. In this regard organized political parties (and other well-structured vested interests) will usually have the edge inactivating direct democratic procedures and may also have more incentives to do so. Recalls are particularly likely in a climate of scandal and/or in the context of extreme political polarization, both of which can be expected to motivate strong elite mobilization, as well as popular engagement. Latin America presents an abundance of such situations. Currently the erosion of traditional political party systems and the spread of social digital networks add to the factors triggering direct recall campaigns where the law allows them and also prompts interest in establishing this option where it is not yet available. But occasional and mostly localized recall initiatives are

9  THE RECALL REVIVAL AND ITS MIXED IMPLICATIONS FOR DEMOCRACY… 

175

insufficient to provide any overall remedy for the conditions undermining support for representative democracy in much of contemporary Latin America. It is also questionable whether more ambitious and comprehensive variants of recall (notably those associated with “pink tide” governments of the past decade) will prove constructive for regional democracy, on balance, and over the long run. If recall is only available at certain lower levels of government (affecting some officeholders in municipal or provincial positions), design errors are not too difficult to identify and correct through intervention from a higher level of authority and the political interests at stake are relatively modest and scattered. Such recall arrangements may provide a limited and partial “safety valve” for specific sources of discontent, but they cannot supply an overall corrective to generalized citizen dissatisfaction with an entire “political class” of professional officeholders. Moreover, such patchy recall provision is likely to disturb the balance of power between levels of government that are placed at risk of a foreshortened mandate and those (usually higher placed) officeholders whose positions are not in play. For a stark illustration of what this might mean, consider a system in which the mayor of the capital city could be targeted for recall but not the President of the republic. (This is the case in Colombia and Peru and will be true for Mexico between 2018 and 2022.) In many Latin American republics, the former office is a prominent stepping stone to the top position, and it is not uncommon for the effective “leader of the opposition” to run the main metropolis. The unintended consequences of such imbalances require more systematic study, both at the highest level and throughout the hierarchy of elected offices. On the other hand, where this procedure is on offer at all levels throughout a national electoral system, bad design choices can be both far more consequential and also much harder to diagnose and then reverse. At least in principle the head of state (and commander of the Armed Forces) is as open to challenge and replacement as a municipal mayor or a provincial assemblyman (that at least is what the Mexican recall law of 2019 appears to provide). But once such a system has been established the process of reviewing its results and correcting any failings is likely to prove intractable. Too many powerful actors have too much at stake to allow consensual fine-tuning. Authors of the legislation who develop second thoughts later on may have to go to extra-constitutional extremes in order to neutralize their own handiwork. As noted in Sect. 9.3, there are already indications that some “pink tide” reformers have begun to view MDDs in general and

176 

Y. WELP AND L. WHITEHEAD

recall laws in particular in a different light once they became more securely entrenched. Nationwide direct recall arrangements are therefore both more resistant to revision and more prone to powerful backlash than earlier scattered “safety valve” experiments. The potential legitimacy gains of a well-designed and widely supported recall system are much greater when the provision is evenly distributed and affects all officeholders in the same way. But the consequences of defenestrating a serving head of state are of a different order of magnitude from other recalls, and the potential for governance dysfunctionality is considerable. Since the political interests at stake are so high, the incentives to game the system are magnified accordingly. Both the promise and the pitfalls of direct recall are redoubled when it is made universal. In summary, Latin America has become the leading laboratory for experimentation with MDDs in general and with recall procedures in particular. The range of experiences is diverse, and the results so far are mixed (although more time is needed for the full story to emerge). This diversity reflects not just inexperience and lack of comparative learning but also the very divergent political and ideational contexts involved. Some of the boldest innovations are also the most recent as shown by the Mexican case.

References Almond, G., & Verba, S. (1963). The Civic Culture: Political Attitudes and Democracy in Five Nations. Newbury Park, CA: Sage Publications. Archer, R. (1990). The Transition from Traditional to Broker Clientelism in Colombia: Political Stability a Social Unrest. Kellogg Institute Working Paper #140. Arques, F. (2014). Argentina: una herramienta de los gobernados en manos de los gobernantes. In Y. Welp & U. Serdült (Eds.), La dosis hace el veneno: la revocatoria del mandato en Suiza, Estados Unidos y América Latina (pp. 159–186). Consejo Nacional Electoral: Serie Ciencia y Democracia. Quito. Ascarrunz, J., & Welp, Y. (2019). La revocatoria de mandato en Bolivia: análisis de limitaciones y propuesta de reformas [Recall election in Bolivia: limitations analysis and reform proposal]. Red de Participación Ciudadana y Control Social, Documento de Análisis Nº 01/2019 (p. 24). Bolivia. Auer, A. (2008). Una mirada suiza sobre la democracia directa en América Latina. In A.  Lissidini, Y.  Welp, & D.  Zovatto (Coords.), Democracia Directa en Latinoamerica. Buenos Aires: Prometeo. Brinks, D. (2012). The Transformation of the Latin America State-as-law: State Capacity and the Rule of Law. Revista de Ciencia Política, 32(3), 561–583.

9  THE RECALL REVIVAL AND ITS MIXED IMPLICATIONS FOR DEMOCRACY… 

177

Castellanos, A. S. (2014). Ecuador: la transformacion de las reglas del juego y sus consequencias (1998–2013). In Y. Welp & U. Serdült (Eds.), La dosis hace el veneno: la revocatoria del mandato en Suiza, Estados Unidos y América Latina (pp. 83–110). Serie Ciencia y Democracia. Quito: Consejo Nacional Electoral. Cronin, T. E. (1989). Direct Democracy: The Politics of Initiative, Referendum and Recall. Cambridge and London: Harvard University Press. Dalton, R., Bürklin, W., & Drummond, A. (2001). Public Opinion and Direct Democracy. Journal of Democracy, 12, 141–153. Garret, E. (2004). Democracy in the Wake of the California Recall. University of Pennsylvania Law Review, 153, 239–284. Guzmán Hernández, Y. (2014). Cuba: deudas pasadas y retos presentes desde la norma. In Y. Welp & U. Serdült (Eds.), La dosis hace el veneno: la revocatoria del mandato en Suiza, Estados Unidos y América Latina (pp. 187–205). Serie Ciencia y Democracia. Quito: Consejo Nacional Electoral. Kriesi, H. (2006). Role of The Political Elite in Swiss Direct-Democratic Votes. Party Politics, 12(5), 599–622. Langbein, L., & Sanabria, P. (2013). The Shape of Corruption: Colombia as a Case Study. The Journal of Development Studies, 49(11), 2–14. https://doi.org /10.1080/00220388.2013.800858. Limón, W. (2016). Revocación del mandato en México. C2D Working Paper Series 51/2016. ISSN 1662-8152. Retrieved from http://www.zora.uzh.ch/view/ subjectsnew/C2D.html. Negretto, G. (2013). Making, Constitutions. Presidents, Parties and Institutional Design in Latin America. New York: Cambridge University Press. Norris, P. (1999). Critical Citizens Global Support for Democratic Government. Oxford: Oxford University Press. Scarrow, S. (2001). Direct Democracy and Institutional Change: A Comparative Investigation. Comparative Political Studies, 34(6), 651–665. Serdült, U., & Welp, Y. (2012). Direct Democracy Upside Down. Taiwan Journal of Democracy, 8(1), 69–92. Tuesta Soldevilla, F. (Ed.). (2014). Una onda expansiva. Lima: Jurado Nacional de Elecciones. Welp, Y. (2016). Recall Referendums in Peruvian Municipalities: A Political Weapon for Bad Losers or an Instrument of Accountability? Democratization, 23(7), 1162–1179. Welp, Y., & Castellanos, A.  S. (2019). Understanding the Use of Recall Referendums: Evidence from Ecuador. International Political Science Review. https://doi.org/10.1177/0192512119830373. Welp, Y., & Milanese, J. P. (2018). Playing by the Rules of the Game. The Partisan Use of Recall Referendums in Colombia. Democratization, 25(8), 1379–1396.

178 

Y. WELP AND L. WHITEHEAD

Whitehead, L. (2011). Enlivening the Concept of Democratization: The Biological Metaphor. Perspectives on Politics, 9(2), 291–299. https://doi.org/10.1017/ S1537592711000909. Whitehead, L. (2017). Between the ‘Fiction’ of Representation and the ‘Faction’ of Direct Democracy. In S.  Ruth, Y.  Welp, & L.  Whitehead (Eds.), Let the People Rule? Direct Democracy in the Twenty-first Century. Colchester: ECPR Press. Whitehead, L. (2018). The Recall of Elected Officeholders the Growing Incidence of a Venerable, But Overlooked, Democratic Institution. Democratization, 1341–1357. Zaremberg, G., & Welp, Y. (2019). Beyond Utopian and Dystopian Approaches to Democratic Innovation. Recerca. Online first.

CHAPTER 10

Reselection and Deselection in the Political Party Jonathan White and Lea Ypi

10.1   Introduction Questions of re- and deselection tend to be approached with a focus on practical issues, notably how these mechanisms are legally defined, how they are conducted from one country to another, and who takes advantage of them and with what results. While these aspects are clearly central to any analysis of the phenomenon, it is important also to consider the more general questions of principle that arise. The recall of representatives impinges directly on the ethics of partisanship and citizenship, and to understand the behaviour, attitudes and debates of party members, still more to reach a rounded normative assessment of their activities, requires one to engage these matters more closely.

J. White European Institute, London School of Economics and Political Science, London, UK e-mail: [email protected] L. Ypi (*) Department of Government, London School of Economics and Political Science, London, UK e-mail: [email protected] © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_10

179

180 

J. WHITE AND L. YPI

Especially important is to consider how recall mechanisms can be squared with the norms of representative democracy. Re- and deselection are often treated as mechanisms of direct democracy, at odds with any meaningful transfer of decision-making authority to others. Sometimes they are presented as threats to the principle of representation, at other times as remedies to correct for its failings, in particular to compensate for its inherent elitism. Our own view is that the two may be more complementary than this suggests, and indeed that their complementarity needs to be fostered. Correctly configured, recall mechanisms can be used to reinforce the programmatic basis of political parties on which representative democracy depends. Our distinctive focus is on how these mechanisms can be developed for use within parties, and how this is consistent with, and may even buttress, the normative reasons for which people may be drawn to become members and supporters of parties. We write against the backdrop of real-world debates amongst party activists about the merits of these arrangements and efforts to make them central to the life of a party. Of particular interest to us is the subset of cases where these are tied to parties of strong partisan traditions with long-standing organisational structures. Notable here is the case of the UK Labour Party, one of the major European parties where these questions have been salient of late. In September 2018, after a summer-long campaign promoted by grassroots activists on the left of the Labour Party, the party’s annual general conference discussed a series of measures that would give ordinary members greater say on the mechanisms of selection and recall of party representatives in Parliament. The proposals ranged from lowering the threshold of members needed to trigger votes of deselection for sitting MPs, to the suggestion of open contests for selecting candidates at each General Election. The measures were the subject of heated debate among members, union representatives, campaign groups, party officials and policy advisors, and in the end only a handful were approved. Some party officials emphasised the potentially “destabilising” nature of the proposals, mindful of the risk of distracting elected MPs from the seemingly more important task of challenging the Tory government.1 Others warned of “catastrophic consequences” for securing the commitment of floating voters if the more ambitious measures of mandatory

1

 https://www.bbc.co.uk/news/uk-politics-45621354.

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

181

r­eselection were to pass.2 For those on the opposing side, the measures were intended to enable greater participation in politics, thus raising democratic accountability. As one Labour MP argued, “only by empowering grassroots members can Labour remain the party that does politics with the community rather than to it”.3 The campaign to bolster such mechanisms within the party would later continue to be maintained by “Momentum”, a social movement with close ties to Labour aimed at improving democracy, accountability and participation within the party.4 One way to read this debate, as much of the British press did at the time, is as an instance of petty fighting between rival groups in the party, revealing the divide between a more institutionally oriented, centrist parliamentary faction and a more radical body of party members committed to the purity of principles. Following the election of Jeremy Corbyn as party leader in summer 2015 and re-election in 2016, and the corresponding expansion of Labour’s membership, intra-party tensions became apparent on several questions. Disputes about the appropriate organisational form of the party might be seen in this sense as just extensions of a wider schism—a proxy war for more substantive disputes about how to interpret the meaning of democratic socialism and how to evaluate the pre-Corbyn “New Labour” era. But this would be reductive. The debate on mandatory reselection raises questions of self-standing importance. Examining them takes us beyond matters of parliamentary tactics and strategy into the heart of debates about the nature of democracy, the relationship between representation and self-government, and the contribution of partisanship to it. They are worth looking at in more detail also because the Labour Party has by no means been the only contemporary party to engage with them. Those that have emerged in Europe around a critique of mediation, seeking technological ways to make office-holders more responsive to a wider movement and developing mechanisms of deselection to underpin them, are testament like Labour’s experiences to the relevance of questions of representation, participation and recall (Urbinati 2015). Parties such as 2  https://www.theguardian.com/politics/2018/aug/27/roy-hattersley-urges-corbyn-tointervene-in-labour-deselection-row. 3  https://labourlist.org/2018/07/chris-williamson-mandatory-reselection-is-comingand-it-will-be-key-to-labours-success/. 4  See, for example, the Momentum statement of 12 July 2019: https://www.facebook. com/PeoplesMomentum/posts/a-statement-fr om-momentums-ncg-on-thebeginning-of-labours-selection-process-mom/919583388387049/.

182 

J. WHITE AND L. YPI

Podemos, La France Insoumise and Movimento 5 Stelle, sometimes referred to as a new breed of “digital party”, have experimented with demanding new mechanisms by which to keep office-holders in line, expelling MPs and Senators in the process (Gerbaudo 2019). However mixed their success as democratic innovations so far, and however much they depart from their principles once in government (as may be the case with Movimento 5 Stelle), such experiments carry wider resonance and are apt to be copied elsewhere. The chapter starts by presenting recall mechanisms as an instance of democratic partisanship that may usefully bridge the divide between representative and direct mechanisms of popular participation. It examines the specificities of mandatory reselection and deselection as they apply to intra-party democracy, and the prima facie merits they display. It goes on to discuss some problems related to the recall of representatives in liberal democracy, before suggesting possible answers and responding to possible criticisms. Our main contention is that issues around the ethics of recalling representatives provide an important basis for exploring how a particular conception of partisanship (which we have elsewhere called the democratic conception (White and Ypi 2016a)) can help mediate some of the tensions between representative institutions and direct democracy. Before proceeding with the main argument, one point needs to be underlined. The practices of representative recall, like those of initiative and referendum, are often analysed under the rubric of direct democratic measures designed to empower ordinary citizens by enabling them to influence government. Through these measures, the wider public can exercise control over particular decisions (e.g. by making new proposals for legislation or by voting in plebiscites) or over particular individuals (by filing petitions to remove particular officials from their position in public office). Experiments with such initiatives are increasingly widespread in contemporary democracies around the world (Whitehead 2018). Our focus in this chapter is related but narrower. Since there is by now a small but relevant literature on the issues of initiative, referendum and recall in their connection to direct democracy more generally (Cronin 1989; Altman 2011), our topic is the more limited but as yet underexplored one of recall mechanisms within parties.5 Analysing the role of recall ­mechanisms as they relate to the distinctive ethics of partisanship 5  For one exception, emphasising also the wider use of these mechanisms at local level, see Bowler (2004).

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

183

goes to the heart of the relationship between representative institutions and democratic participation. It raises questions that are different in focus from those that arise in the ethics of direct democracy more generally. Although, as we shall shortly see, some of the tensions we explore also apply to other mechanisms of direct democracy, focusing on the particular practice of recall from within partisanship goes to the heart of the relation between representative institutions and its realisation of the ideal of self-government.6

10.2   Partisanship and Recall Elsewhere we have argued that a conception of partisanship, understood as an associative political practice intended to promote and support political commitment, is essential to reviving the ideal of self-government at the heart of the radical democratic tradition (White and Ypi 2016). What distinguishes this democratic conception of partisanship from its liberal counterparts is the extent to which it rests on a normatively significant distinction between factions and parties, whereby the former are seen as promoting self-interested principles and aims, whereas the latter seek to realise principles and policies that could be generally and reciprocally justified. This implies an analysis of partisan activity irreducible to what parties are most commonly associated with, that is, the development of public policy proposals and the selection of candidates for election. It requires looking beyond the party as a formal organisation and considering the benefits and obligations entailed by a type of political association that promotes and sustains political commitment over the long term. But the relationship between partisanship—understood as a mode of political activity that involves people committed to certain generalisable principles and aims—and parties, as the institutional reflection of their will, requires further examination. One difficult question concerns the extent to which institutional constraints enable or reinforce that relation. Whether democratic partisanship actually realises the value of self-­ government depends very much on the modalities of political ­participation and the ongoing involvement that parties, and the political systems by which they are constrained, grant to activists. 6  Here we use the terms self-government, self-mastery and non-domination as synonymous. What is at stake in the democratic tradition is not being subject to the will of another and the extent to which collective political institutions enable this.

184 

J. WHITE AND L. YPI

While the value of self-government is generally celebrated in every liberal democracy, the relationship between representatives and the represented tends to be one of independence rather than ongoing scrutiny. The two most obvious, and historically debated, ways of preserving the sovereignty of the people while delegating decision-making power to representatives are the imperative mandate and the discretionary revocability of delegated powers (recall mechanisms). Yet neither of these mechanisms is deeply entrenched in the institutions of contemporary liberal democracy. Despite the consistent effort of radical political movements to incorporate such measures in their projects of institutional reform, their mobilisation has largely been unsuccessful. The result is that contemporary systems inherit much the same hostility to recall mechanisms as their nineteenth-­ century counterparts (Manin 1997: 163). For the champions of radical democracy, the emphasis on recall mechanisms was part of a more general struggle to realise the democratic ideal of self-government which the liberal analysis of representation seemed to have left to one side. In commenting on the Paris Commune’s measures to establish a system of permanent recall for members of the Council, Marx echoed Rousseau’s criticism of representation. “Instead of deciding once in three or six years which member of the ruling class was to misrepresent the people in Parliament, universal suffrage was to serve the people, constituted in Communes, as individual suffrage serves every other employer in the search for the workmen and managers in his business”. As Marx puts it, “it is well known that companies, like individuals, in matters of real business generally know how to put the right man in the right place, and, if they for once make a mistake, to redress it promptly. On the other hand, nothing could be more foreign to the spirit of the Commune than to supersede universal suffrage by hierarchic investiture” (Marx 2000 [1871]: 588). Marx’s allusion to hierarchic investiture emphasises the pre-modern nature of a conception of political representation that seeks to exclude the masses from important moments of public decision-making. Ultimately, it is a critique of the elite/mass divide in liberal politics that the current system of political representation entrenches without being able solve. Political scientists often refer to the different ways of thinking about the relationship between representatives and represented in terms of a “trustee” versus “delegate” model. The former seeks to preserve the ­independence of representatives from the people; the latter suggests that the role of elected representatives just is that of articulating the voice of

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

185

people in relevant institutional sites to which they would otherwise have no access (Marx 2000 [1871]).7 Recall mechanisms are an important part of the process through which elected representatives remain accountable to partisans. They provide institutional guarantees that articulate the political commitments in executive channels of political will formation. At a general level, political parties can be credited with making crucial contributions to the realisation of democracy under modern conditions. Parties mediate between the plurality of interests and principles characterising civil society and the unity of purpose required when institutions make laws that are coercively binding for all. They give definition and political focus to normative principles, connecting them on the one hand to the intuitions of a wider constituency, and on the other to the exercise of legislative and executive power. They aim to develop explicitly political identities, irreducible to the concerns of a sectional grouping, and towards which citizens may orientate themselves reflexively based on an evaluation of the associated objectives (Urbinati 2015: 36). Rather than just aggregations of societal interests, parties are agents fundamental to the exercise of popular sovereignty (Wolkenstein 2019; White and Ypi 2016, 2010). But how effectively parties and partisanship contribute to the active exercise of popular sovereignty depends in crucial part on the availability of scrutiny mechanisms that link the views of partisans to those expressed by their elected officials. Recall mechanisms empower partisans beyond the limited timing and mobilisation typically concentrated in elections. While a vibrant public sphere also gives partisans a vehicle through which the activities of public officials can be assessed and criticised, recall mechanisms are more effective in linking the process of selection to the purposes and ends that activists hope their elected representatives will promote. They seem essential to renewing what a collective association stands for and to reaffirming the political project on whose behalf partisans remain continuously engaged in politics. Recall mechanisms are especially important in moments of crisis in liberal representation, when citizen apathy and a perceived divide with elites are often cited as reasons for a decline in mass parties and the traditional channels of popular will formation (Mair 2013). The contemporary world offers plenty of examples of political fragmentation along the lines of masses and elites. The former increasingly drift towards a politics of personal charisma based on facile (mostly exclusive) keywords: “Make 7

 For the distinction between the two, see Pitkin (1967), esp. Chaps. 6 and 7.

186 

J. WHITE AND L. YPI

America Great Again”, “Take Back Control”, “Prima gli Italiani”. The latter take refuge in a politics of expertise that weakens the representative link and consolidates the role of increasingly unaccountable bodies: courts, central banks, international organisations and the like. In some ways, the current vision of representation is similar to that wished for by early champions of liberalism like Siéyès and Madison, who both emphasised the importance of a politics conducted by professional politicians with a sharp divide between representatives and the represented. In our time, oligarchic and technocratic decision-making is on the rise, while large swathes of the public remain alienated from parliamentary politics and drift towards a politics of demagoguery. But while at the origin of liberal representation there may have been affinities between the class interests of representatives and those of the voting subset (due to the selective criteria for voting), the spectrum of diversity between representatives and represented based on class background, cultural identification, social orientation and so on is now much wider. In this context, the gap between representatives and the represented must be urgently closed. Otherwise there is every risk that the mass of effectively disenfranchised citizens permanently drifts away from traditional channels of representation to the more populist forms of activism whose recent rise has been widely documented (White 2019, Chap. 6).

10.3   Mandatory Reselection, and Deselection A distinction can be made between those forms of partisan recall that map onto the rhythms of electoral politics and those that can be activated by partisans at any point in time. Mandatory reselection is of the first kind. The practice (sometimes called “open selection”) entails that, should an incumbent parliamentarian hope to be re-elected at the next election, they must be willing to face a competitive nomination process within their party.8 Should the party’s selectors vote in favour of another candidate, the incumbent representative is effectively recalled. The practice synchronises with the electoral cycle, and can be seen as part of the more general effort to renew the electoral campaign. Here it is only in 8  In the UK Labour Party, the current rules are more qualified: where an incumbent continues to enjoy significant majority support amongst the selectorate (>67% amongst relevant party bodies including the local party branch and affiliated trade unions), the process may be waived and the incumbent is automatically retained.

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

187

the run-up to elections that sitting representatives are challenged to make a new case for their re-election. Deselection is activated on an ad hoc basis, outside the time structures of electoral institutions. Representatives are continuously scrutinised and can be recalled at any point, for example, by initiating a petition process and deselection campaign. The party membership to whom these representatives are accountable can seek to recall them whenever their conduct is felt to be at odds with the aims or standards associated with the party. Recall happens not on a periodic basis, in other words, but whenever a certain threshold is crossed, for example, when a certain number of signatures has been collected. Both are potential ways of strengthening participatory democracy through the channel of political parties. Mandatory reselection can be seen as part of the process of preparing for elections and is generally less controversial (though it may be opposed for strategic reasons). It conforms to the general principle that representation is time bound and needs periodic renewal. Amongst the positive outcomes it promises is to counter tendencies towards professionalisation and the demographic homogenisation of representatives, and in so doing also to sharpen the programmatic profile of parties, ensuring they are represented by those genuinely committed to the principles espoused. The arguments made by Momentum for developing such mechanisms in the Labour Party are illustrative: For too long politics has been dominated by posh men from expensive private schools. … People feel deeply alienated by our broken political system, and campaigning for open selections across the country will help surface a new generation of young, BAME, working class leaders who will take on the political establishment and provide a genuine alternative. We must remember that our strength lies in [the] knowledge and talent of Labour’s half a million members who live and work in every community. Collectively, they understand the damage wrought by decades of neoliberalism. They know we need a bold, radical agenda in government to rebuild communities, take on the establishment and avert climate disaster. They are the kind of people we need as Labour MPs in a Corbyn led government, and Momentum will support its members to push for open selections across the country and give this new generation a shot at parliament.9 9  https://www.facebook.com/PeoplesMomentum/posts/a-statement-from-momentumsncg-on-the-beginning-of-labours-selection-process-mom/919583388387049/; see also: h t t p s : / / w w w. t h e g u a r d i a n . c o m / p o l i t i c s / 2 0 1 9 / j u l / 1 2 / m o m e n t u m - d r i v e labour-members-deselect-mps.

188 

J. WHITE AND L. YPI

As this suggests, mandatory reselection can be pursued not just for the benefits of social diversity itself (though this may be one part of the rationale) but so as to strengthen the party’s vitality as a principled association. By ensuring the views of party elites are not fundamentally out of step with those of the larger membership, and by ensuring that the party is represented by those with a proven capacity to generate support amongst peers (as shown in the selection process), such devices can augment the party’s clarity of vision. Clearly, there may be moments when this form of party-based representation may apparently clash with the political representation of citizens more generally. Some constituents content with an incumbent MP and keen to support their re-election may feel disenfranchised should that MP’s party decline to nominate the individual again. But such objections do not seem especially forceful. To the extent that a candidate for election wishes to be associated with a party and the programmatic agenda it embodies, it is logical that the party should be able to assess their actions in this light and recall them where appropriate. Likewise, should they avail themselves of the resources of the party, accepting the campaigning assistance of party activists to help them get elected, it is logical that the party be in a position to evaluate their record and decide whether to renew its support. Moreover, those who are not reselected by their party may often be able to stand as independent candidates for election instead. The possibility of citizen representation unmediated by a political party is thus retained. But while mandatory reselection is fairly accommodating of general criticisms to the recall mechanism, deselection is important precisely because of the radical degree to which it potentially empowers ordinary citizens with relation to their elected representatives. It is here that the advantages of party-based recall mechanisms are likely to reveal themselves most sharply. Whereas candidate (re)selection campaigns at election time can be (and sometimes are) conducted outside parties, albeit with all the attendant risks of personalised contests, recall mechanisms outside the electoral cycle face significant collective-action problems. Their initiation by individual citizens is challenging, given the thresholds of support needed. (Granting opposition parties a significant role in initiating such processes meanwhile creates a capacity liable to be abused.) Channelling deselection processes through the party structures of incumbent representatives promises to make far more meaningful the checks on their exercise of power. Even if rarely enacted, the prospect of such challenges may in

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

189

itself have valuable disciplining effects, encouraging party representatives to act consistently with the principles they claim to share with their fellow partisans, and for the sake of which the latter go campaigning on their behalf. While mandatory reselection alone offers notable benefits for parties and party democracy, deselection has its own distinctive qualities. Being possible to initiate at any moment, it creates more immediate opportunities for participation, debate and programmatic renewal. It allows divergences of view between party members and representatives to be tackled more promptly than is possible when recall mechanisms are anchored in the wider electoral process. When there are major rifts within a party, waiting until the next election to address them may be fatal to its programmatic unity and electoral success. As the recent experiences of Labour suggest, disaffected representatives may undermine the party from within, obstructing policies or threatening to break away from it, while new competitor parties may emerge to capitalise on its difficulties. Deselection enables such problems to be addressed early. To be sure, such mechanisms can sometimes misfire. Some of the contemporary parties to debate and embrace such methods—notably “digital” or “platform” parties on the model of Movimento 5 Stelle—have not always done so with great success. With rather weakly defined ideological commitments, other than their enthusiasm for participatory methods, and reluctant to adopt the organisational structures that might underpin a more defined programmatic identity, they have arguably used recall mechanisms in an unpredictable and reckless fashion (Gerbaudo 2019). But however problematic such recent experiences, the mechanisms in question are in principle adoptable also by parties with a more clearly defined normative project and more willing to retain structures of hierarchy and representation. This is what makes the case of the Labour Party an interesting and significant one.

10.4   Constituency and Responsibility Any proposal to make the current system of party representation more hospitable to mandatory reselection and deselection processes is likely to meet with a number of objections. In this section we respond to four of the most recurrent and weighty ones, what we call the constituency objection, the plural responsibilities objection, the incentives objection and the efficiency objection. The next sections consider them in turn.

190 

J. WHITE AND L. YPI

The constituency objection has to do with how to conceive the obligations a party has towards its partisans and to fellow citizens at large. The objection can take a more principled or more pragmatic form. On a more principled level, one might say that elected representatives have obligations both to fellow party members, whose views they represent and who campaign on their behalf, and also to constituents at large, to the nation or “the people” whose sovereignty they articulate and whom they are supposed to serve once they have been elected. On a more pragmatic level, the worry may be that a too active use of recall mechanisms might render a programmatic party attractive only to a few core activists but unattractive to the constituency at large. The question of how to navigate that tension, from the point of view of principles, is a difficult one. But in answering it, it is important first to point out that the problem of how to represent both fellow partisans and their commitment to certain principles and aims and to represent constituents who may not agree with the principles of the winning party is not specific to the debate around recall. It is a tension intrinsic to how liberal institutions realise the idea of representation. Nothing about this tension raises further difficulties for recall mechanisms over and above the more general problem that “no man can serve two masters”, that parties cannot be both representative of members and of constituents without some tension, as some political scientists have put it (Katz 2014). The more specific answer has to be seen in light of that more general problem. Here opinions will differ, depending on how concerned one is to see the radical democratic ideal of self-government realised by institutions. All those subjected to the exercise of political power are both in the position of makers and takers of laws and policies. But from a radical democratic perspective what makes the latter bearable is the extent of one’s contribution in the former. Hence, it is difficult to see how, in the case of a conflict, prioritising the party-constituent relation would not end up undermining the party-partisan relation. And the latter is critical for ensuring the preservation, strength and maintenance of a principled identity for the party in the long run. The key point here is that in liberal democratic societies the election of representatives is intrinsically connected to the party system. It is on the basis of particular party manifestos and programmes that elected ­representative are voted into office. Given this basic model, it is crucial to ensure that incumbents are responsive to members with regard to how they act on the basis of the party programme, and that the latter retain a

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

191

say when compromises are required or difficult choices need to be made. Recall mechanisms that enable members to deselect sitting MPs who fail to properly connect the party on the ground with the party in office are a crucial step towards ensuring the principles and aims expressed in the party programme combine democratic support with executive efficiency. Such a model need not discourage or undermine the participation of ordinary citizens: on the contrary, it encourages that participation in the form of partisan engagement. When citizens profoundly disagree with the values and principles of the public officials that represent them, it is because they profoundly disagree with the values and principles of the party that wins the election. The most democratic way to challenge that outcome and channel that disagreement is through partisan affiliation. Recall mechanisms do nothing to aggravate the basic tension; they simply ensure that a greater number of people can take responsibility for the outcomes that follow. To be sure, the sociology of party membership suggests that those who currently join parties are in some ways unrepresentative of the wider electorate (Bale et al. 2019). But how inclined citizens in general are to become partisans will depend, amongst other things, on the extent to which partisanship empowers them. Recall mechanisms are one way to make party membership more enticing to a wider array of citizens. This answer takes us to the more pragmatic aspects of the constituency objection: what if responding to the will of party members who may hold more radical views than non-partisan constituents were to render a party increasingly unelectable? Taken at face value, the objection stems from an understanding of the party as an institution whose main purpose is to win elections rather than to articulate the commitments of those who subscribe to it. Elsewhere we have expressed scepticism towards that view (White and Ypi 2016). But even if one accepts it for argument’s sake, one may still want to ask: what does it take for an electoral victory to be institutionally significant? For a party to make a lasting impact on institutions and public affairs, it needs to prevail not just in one electoral cycle but to maintain its power over a substantial stretch of time, as well as to maintain its identity in opposition (White forthcoming). But how does it keep alive the loyalty of its traditional supporters? What guarantees are there that the decisions it takes do not undermine the project in the long run? The current identity crisis of many recent mainstream European social democratic parties is a good case in point. Their increasing support of Third Way policies at the turn of the millennium, apparently neither left nor right, brought electoral advantages for a time. But if one looks at the current

192 

J. WHITE AND L. YPI

crisis of large, long-standing socialist parties such as the PS in France or the SPD in Germany, it seems warranted to question whether such ideological turns were desirable in the longer run. Let us now turn to a second objection to the empowering of recall mechanisms within parties, what we may call the multiple responsibilities objection.10 A critic might observe that elected representatives have multiple representational responsibilities: to their constituents, to party members, to party voters, to the citizens at large and, possibly also, if certain MPs come from marginalised and disadvantaged groups, to these too. Since these responsibilities are distinct, they may not align with each other. If there is a significant conflict, individual MPs may lose the discretion needed to decide how to balance them if recall mechanisms empower party members to a significant degree. In reality, MPs never have complete discretion concerning how to balance these responsibilities. Their judgements are oriented by the principles their party subscribes to, the legal and policy documents that reflect how such principles are embraced in practice, and the institutional ethos that shapes the structures within which their actions are inscribed. In the absence of strong recall mechanisms, that relationship has few intermediaries and unfolds in a centralised manner. The risk is then that in cases of conflict, either individual MPs will shoulder the burden of interpretation on their own or central decision-makers will have the upper hand. The promise of recall mechanisms is to decentralise this process, enabling MPs to share responsibility for how they discharge their different obligations with supporters and fellow partisans, and in light of a collective understanding of their shared project. This in turn implies that the epistemic advantages MPs gain in the course of navigating their responsibilities can be shared more widely and become a collective institutional resource.

10.5   Incentives and Efficiency A third objection to the expansion of recall mechanisms concerns the perverse incentives that the process may set in motion. One aspect of this is the risk of encouraging those who might otherwise not have been motivated to join the party to do so, thus changing the internal balance of forces: what is often labelled “entryism”. An example from the United Kingdom is the process leading up to the election of Jeremy Corbyn as  We are grateful to Stuart White for the objection and the label.

10

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

193

leader of the Labour Party in 2015, a process which was preceded by thousands of new members and supporters joining the party so as to exercise the right to vote in the election. Many, including prominent Labour members and MPs, wrote officially to the then interim leader Harriet Harman to voice concerns over “infiltration” and to complain that the party was in “danger of getting trumped by people who have opposed the Labour party and want to break it up”.11 Issues of entryism and infiltration have since been raised several times in British public life, with one of the most recent cases involving the prospect of the deselection of proRemain Tory MP Dominic Grieve after a vote of no confidence in his Beaconsfield constituency led by his former UKIP opponent, who responded to defeat at the last election by joining the local Tory party along with a large contingent of his followers.12 Rather than intrinsically undermining the case for recall, however, these examples illustrate the need for reflecting on the larger institutional implications of the process. What they show is the need to devise constraints that make that process more effective while limiting its destabilising effects. It is not our aim here to take a stance on the empirical cases just mentioned, including whether the concerns raised by the UK Labour and Tory cases were generally warranted or rather exaggerated by a media and establishment fearful of radical shifts in political activism. Rather, what is worth emphasising is that, in the Labour case at least, one implication of the 2015 leadership election was to encourage the party to come up with stronger rules requiring minimal periods of involvement prior to granting the right to vote, and to enforce more extensive barriers on access towards those with recent histories of active involvement in other parties. The more power is granted to partisan activists, the more important become rules concerning length of members and wider evidence of partisan loyalty as preconditions for exercising these capacities. It is not clear why such rules could not be extended to recall processes as well, and why the extension and consolidation of recall mechanisms could not go hand in hand with a more robust enforcement of rules intended to prevent the process generating perverse incentives. There is another common criticism to the practice of deselecting representatives, again related to the incentives that the process puts in motion, 11  https://www.theguardian.com/politics/2015/jul/26/jeremy-corbyn-genuine-laboursupporters-leadership-election. 12  https://www.bbc.com/news/uk-england-beds-bucks-herts-47759208.

194 

J. WHITE AND L. YPI

which concerns the pressure that the threat of deselection puts on public officials, who may become dependent on public opinion or who may attempt to second-guess the will of citizens to the point of failing to exhibit leadership on matters of principle. While on the surface plausible, this concern is premised on a certain understanding of elected officials as people who have inherently more skills, leadership or principled commitment than the people they ought to represent (Bowler 2004). In other words, it is premised on accepting an asymmetry between professional politicians with greater skills and knowledge and ordinary members of the public who may lack the same abilities or understanding of the constraints of political institutions. The implicit historical reference is here not the democratic forum of equals but the Roman division between the higher magistrates and the people: while the people could elect magistrates, they could not themselves be magistrates (Manin 1997). On this alternative model, the worry is that a constant fear of sanctions by partisans on the ground might create perverse incentives that force representatives to self-­ censor or fail to make decisions that would displease their supporters, even when the latter may be warranted. There are several replies one might give here. The first, more abstract one, is to question whether this degree of professionalisation of politics is itself intrinsically desirable. If there are certain substantive as well as formal conditions that ought to be met in order to enable all citizens to be self-­ governing, for example, conditions to do with the distribution of skills and knowledge required to be involved in politics effectively, or with the degree of familiarity with how institutions work, surely the right response is to distribute these skills and assets more widely rather than to isolate those who have them from those who do not. Just as elections and the choice of government should not be considered in isolation from the more general partisan practices and political engagement that precedes them, the practice of recall should not be considered in isolation from an ongoing process of deliberation between representatives and represented, geared to the political education of each.13 This is in fact precisely where politics differs from other realms of life in which a division of labour or an asymmetric distribution of decision-­ making capacity may be warranted. It is why politics is precisely not a 13  On the role of political education for ends that are neither depoliticised nor reducible to the immediacy of partisan conflict, see Einaudi “Conoscere per deliberare” (1954) in Einaudi (1964) and Bobbio, “Intellettuali e classe politica” (1995) in Bobbio, Norberto, Il dubbio e la scelta. Intellettuali e potere nella societa’ contemporanea (Rome: Carocci, 1993).

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

195

profession like any other. Politicians are not like doctors, in whom we believe “whether they cure us with our consent or without it, by cutting or burning or applying some other painful treatment” (Plato 1997: 293b). When one has a heart problem, a cardiologist may be needed to examine the heart and fix the problem; one can trust no other claim to expertise, and certainly not one’s own. But elected representatives have no particular fields of expertise. They are ordinary citizens with a history of mobilisation, campaigning or policy-making, and with particular profiles singled out as suitable to represent their fellow partisans. Even if they have a privileged epistemic viewpoint with regard to some aspect of decision-making (they may come from a particular career path with relevant exposure to certain areas of institutional politics, or they may belong to particular minorities that give them a valuable epistemic vantage point when it comes to particular issues), these will typically be insufficient to cover all the aspects of decision-making and spheres of law with regard to which they will be called to make decisions. Therefore, while the doctor’s claim to be able to fix one’s heart better than oneself can in general be taken at face value, there are reasons to doubt professional politicians’ claim to expertise on all particular applications of political decision-making. And even when someone does have an unusual knowledge and ability to navigate and respond to a broad and complex array of political questions, this may be only as a result of the experience accumulated through years of public office. In the latter case, from a democratic perspective, surely the right attitude is to see this as an unfortunate implication of insufficient rotation in office rather than as something intrinsically desirable and to be replicated. A second important point to make here is that the extent to which professional politicians make independent decisions and exercise leadership given the institutional constraints is vastly exaggerated. It is true that in the absence of recall mechanisms, professional politicians representing particular parties can safely ignore the views of their constituents, barring election time. But it is usually very difficult for them to ignore the views of bureaucrats, policy-makers, lobbyists, business people, the media, civil society associations and polling agencies. Such influence and pressure usually goes much beyond the timing of elections. Given these constraints, the idea that elected representatives would exercise their judgement and leadership if it were not for the influence of constituents is largely a myth. In liberal market societies, political influence is usually proportionate to the power of money. For radical democrats, the question is how to make

196 

J. WHITE AND L. YPI

sure that influence is at least balanced so that the decisions of politicians (including the compromises they are asked to make in this environment) approximate as much as possible the views and judgement of the ordinary people who elect them. Given incentives are never pure, recall mechanisms should be judged as a balancing rather than disruptive force. It is also important to emphasise that they should be seen as balancing measures taken in combination with other, also non-partisan, mechanisms necessary to temper these oligarchic tendencies, not as a substitute.14 So, for example, a stronger associative democratic life, citizens’ initiatives, and a more extensive use of sortition in the election of public officials, or a combination of these, are all important transitional measures to minimise the corrupting effects of inequality in the political system, a goal that should arguably animate the party system itself.15 Let us, finally, consider a fourth kind of objection to recall mechanisms, concerned not so much with the system of incentives within which elected representatives operate as with the efficiency of recall campaigns. Call this therefore the efficiency objection. The concern here is that recall mechanisms risk distracting sitting MPs from their daily task of representing constituents by forcing them to confront fellow partisans when the latter object to their actions. Recall mechanisms may be thought to harm practices of deliberation, depriving representatives of the buffer needed to reflect and take decisions without fear of immediate reprimand. This concern is probably also over-stated. Firstly, there are many kinds of obligation that might distract representatives from their daily activities, including their obligations to family, friends or institutions other than their own parties. The point of taking up representative responsibilities (and the desirability of limiting such responsibilities in time) is that one is prepared to navigate these increased burdens. Where the occupation of office is limited in time, elected representatives know that such increased burdens are only provisional. The second, more important, part of the answer to the efficiency objection is that not all recall campaigns risk reducing the efficiency of representation: they might also increase it. Much depends on how recall measures 14  For a discussion of some of these alternative mechanisms in connection to the worry about oligarchy, see White (2017). For a discussion of sortition but as a substitute to elections, see Guerrero (2014). 15  Indeed, the commitment to reducing these inequalities may well be one of the distinguishing features of factions from parties, as we argue in White and Ypi (2016), Chap. 3.

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

197

are organised, what constraints are in place, how frequently they can be invoked, where they fit in the schedule of representatives, and what tone should govern the debate surrounding them. Amongst the variations on the theme of recall available to parties are temporary sanctions (e.g. the suspension of the whip for a period of weeks), deferred sanctions and those dependent on the support of a super-majority. Rules concerning the conduct of party debates and the kinds of argument permissible as reasons for deselection are another area where there is scope for innovation. Where recall debates have proven distracting and destructive, the fault arguably has lain with the highly personalised or negative tone of the campaign, with the absence of sustained principled debate on particular issues, and with the lack of working channels of communication between representatives and citizens. Such negative recall campaigns have generated further mutual distrust between representatives and represented, with charges of corruption coupled with personal insults. The problem here, however, relates not to recall mechanisms as such as to a lack of discipline and a culture of respect amongst activists. There is no reason why more constraints on the process, as well as different arrangements for the timing and nature of recall campaigns, could not result in a more constructive contribution. One way or another, those who are defeated in a particular debate must be willing to accept the legitimacy of the decision-making process and the (at least provisional) conclusiveness of its outcome. Partisan institutions must be supported by an appropriate ethos.

10.6   Conclusion Democracy is in crisis, it is often said. Those who lament this state of affairs link it to the crisis of representation, the fact that few people trust political elites, and that even fewer are inclined to turn up and vote, or that when they do, they end up voting for politicians who undermine rather than support democratic values. If all this is true, it is imperative to seek to restore trust in the ability of institutional politics to respond to the ­concerns of ordinary people, and to narrow the gap between citizens and political elites. Strengthening recall mechanisms is an important step in that direction. Consolidating mechanisms of direct democracy is crucial to ensuring ordinary citizens can take political responsibility, seeing themselves as agents and not only recipients of political decision-making. That such mechanisms should be part of the organisational structure of political parties,

198 

J. WHITE AND L. YPI

and not just of citizen-representative relations more generally, seems important in two respects. First, if long-standing oligarchic tendencies within parties are one aspect of the democratic weakness and alienation in question, it makes sense to see the empowerment of partisans vis-à-vis their leaders as one necessary component in any wider effort to revitalise democracy. Strengthening partisan recall mechanisms is one way to counter the failings of existing parties. Second, and more positively, it is when channelled through the party form that recall mechanisms have particular potential to augment representative democracy. Such mechanisms need those who can initiate them and connect them to shared normative commitments. Embedding them in an associational context defined by shared ideas counters the risk that they are used in highly personalised ways to punish the character failings of individuals, or that—on account of the same risk—their application is restricted to the most egregious cases of corruption, and thus rendered of little relevance to day-to-day politics. It is as a wider and principled check on political power, channelled through clearly defined parties, that mechanisms for recalling representatives have greatest potential.

References Altman, D. (2011). Direct Democracy Worldwide. Cambridge: Cambridge University Press. Bale, T., Webb, P., & Poletti, M. (2019). Footsoldiers: Political Party Membership in the Twenty-First Century. Abingdon: Routledge. Bobbio, N. (1993). Intellettuali e classe politica [1995]. In N. Bobbio (Ed.), Il dubbio e la scelta. Intellettuali e potere nella società’ contemporanea. Rome: Carocci. Bowler, S. (2004). Recall and Representation: Arnold Schwarzenegger Meets Edmund Burke. Representation, 40(3), 200–212. Cronin, T.  E. (1989). Direct Democracy: The Politics of Initiative, Referendum, and Recall. Cambridge, MA: Harvard University Press. Einaudi, L. (1964). Conoscere per deliberare [1954]. In L. Einaudi (Ed.), Prediche inutili. Torino: Einaudi. Gerbaudo, P. (2019). The Digital Party: Political Organisation and Online Democracy. London: Pluto. Guerrero, A. (2014). Against Elections: The Lottocratic Alternative. Philosophy and Public Affairs, 42(2), 135–178.

10  RESELECTION AND DESELECTION IN THE POLITICAL PARTY 

199

Katz, R. (2014). No Man Can Serve Two Masters: Party Politicians, Party Members, Citizens and Principal–Agent Models of Democracy. Party Politics, 20(2), 183–193. Mair, P. (2013). Ruling the Void: The Hollowing of Western Democracy. London: Verso. Manin, B. (1997). The Principles of Representative Government. Cambridge: Cambridge University Press. Marx, K. (2000). The Civil War in France [1871]. In Selected Writings. Oxford: Oxford University Press. Pitkin, H. (1967). The Concept of Representation. Berkeley: University of California Press. Urbinati, N. (2015). Revolt Against Intermediary Bodies. Constellations, 22(4), 477–486. White, S. (2017). Parliament, Constitutional Conventions, and Popular Sovereignty. British Journal of Politics and International Relations, 19(2), 320–335. White, J. (2019). Politics of Last Resort: Governing by Emergency in the European Union. Oxford: Oxford University Press. White, J. (forthcoming). What Kind of Electoral System Sustains a Politics of Firm Commitments? Representation. White, J., & Ypi, L. (2010). Rethinking the Modern Prince: Partisanship and the Democratic Ethos. Political Studies, 58(4), 809–828. White, J., & Ypi, L. (2016a). The Meaning of Partisanship. Oxford: Oxford University Press. White, J., & Ypi, L. (2016b). The New Partisanship. Renewal, 24(6), 27–35. Whitehead, L. (2018). Recall of Elected Officeholders: The Growing Incidence of a Venerable, but Overlooked, Democratic Practice. Democratization, 25(8), 1341–1357. Wolkenstein, F. (2019). Agents of Popular Sovereignty. Political Theory, 47(3), 338–362.

CHAPTER 11

Narratives of Executive Downfall: Recall, Impeachment, or Coup? Aníbal Pérez-Liñán

11.1   Introduction Are some procedures to depose the chief executive easier to justify than others? If so, what are the criteria to validate their use? This chapter outlines a typology of mechanisms—both constitutional and non-­ constitutional—to oust national executives and maps the foundations of their legitimacy in contemporary democratic societies. The first part of the chapter develops this typology, placing special emphasis on constitutional mechanisms of removal. Recall elections benefit from the fact that voters, who select the chief executive, also withdraw the executive’s mandate. Parliamentary votes of no confidence benefit from the fact that parliamentary majorities both appoint and dismiss the cabinet, but they lack the legitimacy granted by direct popular participation. Finally, presidential impeachments involve a situation in which congress dismisses a president directly elected by the people. They enjoy neither symmetry with respect to the selection procedure, nor popular legitimacy. This makes impeachments easy targets of critical narratives

A. Pérez-Liñán (*) Keough School of Global Affairs, University of Notre Dame, Notre Dame, IN, USA e-mail: [email protected] © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_11

201

202 

A. PÉREZ-LIÑÁN

recasting them as “parliamentary coups.” Underpinning such deficit of legitimacy lies a crucial question: whether legislative majorities (and street protests) at the time of impeachment should prevail over electoral majorities at the time of the election. The second part of the chapter articulates three normative criteria to distinguish legitimate and illegitimate uses of removal procedures: legality, fairness, and due process. Although these criteria occasionally matter to assess recalls and no-confidence votes, they are central to debates about the use of asymmetric procedures, such as impeachments, pre-trials, and declarations of incapacity. This section builds on the fact that impeachments combine the institutional logics of the judiciary and the legislature. On the one hand, the process seeks to emulate a trial in which the “prosecution” is expected to present evidence against the president. On the other, all variants of presidential impeachment—no matter their constitutional differences—are subject to approval by legislative majorities. Because these two logics are in tension, the use of presidential impeachment has always been subject to politicization and controversy. Evidence from the United States and Latin America indicates that, in the midst of a political crisis, partisan conflict always prevails over strictly legal considerations. However, those legal considerations impose normative standards to the process, represented by the three criteria discussed. The conclusions highlight an emerging puzzle: despite their greater legitimacy, no recall elections had removed a national executive from office by the time of this writing. Recalls against presidents have been rare, and the ones that took place in Bolivia (2008), Romania (2007 and 2012), and Venezuela (2004) failed to dislodge the incumbents. This fact suggests that controversial procedures such as presidential impeachments will remain part of the global political landscape in the near future.

11.2   Getting Rid of the Chief Executive Elected executives often fail to complete their terms in office. This is not surprising in parliamentary systems, where prime ministers are responsible to parliament, but presidential systems are also subject to considerable turnover. In the United States, 2 out of 20 presidents in office between 1900 and 2018 were killed (McKinley) or resigned in the midst of an impeachment (Nixon), while 2 others died natural deaths in office. In Latin America during the same period, almost 30% of all presidents (some

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

203

835 leaders in total, elected and unelected) were removed from power against their will, and over 20% left office in an unconstitutional manner. Three questions underpin any narrative intended to justify (or challenge) a particular procedure employed to oust a chief executive: Was the procedure constitutional? Was the process driven by the people or by elites? And in the latter case, was the chief executive in any way responsible to those elites? Advocates combine and conflate those issues in political debates, but they deserve separate analytical treatment. The first question involves a distinction between constitutional and non-constitutional procedures. By this, I simply mean that the constitution institutes the procedure in question as an explicit mechanism to remove the head of the government. I discuss procedures failing to meet this condition as non-constitutional, rather than unconstitutional, to avoid two potential misunderstandings. On the one hand, critics often argue that the ousting of a chief executive was unconstitutional because, even though the procedure was listed in the constitution, it was employed against the spirit of the law. This criticism is raised in many cases of impeachment, for reasons that will become clear. On the other hand, the opposition may overthrow the chief executive without invoking a constitutional procedure for removal, but using a strategy that is not against the constitution. This is usually the case when popular revolts force the resignation of the government. Citizens have a constitutional right to protest against the government, even though no constitutional clause authorizes them to undo the outcome of an election in this way. With rare exceptions, constitutional procedures are easier to defend than non-constitutional ones. Actors invoking constitutional procedures enjoy what we may call Kelsenian standing—they overthrow the president claiming an authorization granted by the law (Kelsen 1967 [1934]). Opponents can only challenge this legitimacy by articulating an alternative constitutional interpretation after the facts. In contrast, leaders concocting non-constitutional actions must rely on a Schmittian claim, invoking a state of exception that justifies the overthrow of the executive even if it requires suspending deference to constitutional norms (Schmitt 2014 [1921]). The second question involves a distinction between vertical and horizontal forms of accountability (O’Donnell 1998). Discussions of accountability normally focus on institutional mechanisms, but I extend the question to address non-constitutional mechanisms as well (Smulovitz

204 

A. PÉREZ-LIÑÁN

and Peruzzotti 2000).1 Is the government terminated by actions emanating from the people (voters, social movements, demonstrators) or from elites (legislators, judges, military officers)? In practice, those forces can intertwine to undermine the executive. Social movements, media outlets, and judges contribute to the creation of “societal” accountability (Smulovitz and Peruzzotti 2000); mass protests reinforce the position of legislators pushing for impeachment (Pérez-Liñán 2007); and military officers form alliances with social forces to overthrow disliked rulers (Chilcote 2014). However, it is usually possible to identify a proximate cause for the downfall of the government that is—at a late stage in the process—sufficient to decide the outcome. For example, if mass protests call for the resignation of the prime minister and parliament approves a no-confidence motion, the action of parliamentary elites effectively terminates the government. In contrast, if mass riots force the resignation of a terrified president, the outcome follows from popular mobilization even if congress formally accepts the president’s resignation. Other things being equal, vertical actions have an intrinsic advantage over horizontal ones. Narratives invoking popular will are easier to support than narratives based on elite pacts. The will of the people is—in modern democratic societies—virtually uncontestable, but elites must claim a popular mandate or rely exclusively on their Kelsenian standing to justify their actions against the government. The role of elites is tainted by narratives emphasizing their vested interests, as elected executives claim that “my enemies are out to get me” (Hinojosa and Pérez-Liñán 2007). The ability of elites to claim a popular mandate depends largely on the third factor, namely, whether the procedure to oust governments is symmetric with the procedure to select them. In most parliamentary systems, a legislative majority selects the prime minister and can dismiss the prime minister at will. The chain of delegation runs uncontested from voters to legislators to the executive. The democratic legitimacy of the appointment is therefore symmetric with the legitimacy of the dismissal procedure. In presidential systems, in contrast, the constitution creates two parallel chains of delegation—as voters elect presidents and legislators separately and for fixed terms in office—and any congressional attempt to remove the executive, even if it is constitutional, must grapple with this fact. 1  Rulers can be (and often are) accountable to social and political actors that are not constitutionally empowered to oversee their actions, because those actors can impose major costs—including, in extreme cases, removal from office (Bueno de Mesquita et al. 2003).

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

205

Symmetric procedures are thus easier to defend than asymmetric ones; those who grant a mandate can easily claim authority to recall it. This distinction is central to debates about the legitimacy of constitutional mechanisms of horizontal accountability. The distinction is implicit for non-constitutional mechanisms of horizontal control (e.g., military conspiracies are evidently misaligned with democratic sources of legitimacy), and for constitutional mechanisms of vertical accountability (e.g., recall elections are evidently aligned with democratic sources of legitimacy). Attention to these three questions allows us to map seven alternative procedures to dismiss the chief executive, presented in Fig 11.1. Although the list is comprehensive, the seven procedures should be seen as ideal types. Constitutional as well as non-constitutional procedures often combine different models of removal. For example, indirect recall procedures require a vote in parliament before a popular recall election takes place. Revolutions blend an elite (“vanguard”) and a popular revolt component. Moreover, as I show below, narratives depicting those ideal types often blur the categories to encourage a particular interpretation. The bottom row of Fig 11.1 outlines a potential scale of democratic legitimacy for removal procedures. From left to right: executive removal through a direct recall election is, other things being equal, easier to defend than a no-confidence vote, which in turn is easier to defend than an impeachment process, which is (usually) easier to defend than a popular revolt, and certainly easier to defend than a coup. This implicit range helps

Fig. 11.1  Typology of mechanisms to remove the chief executive

206 

A. PÉREZ-LIÑÁN

explain the nature of discourses seeking to legitimize (or de-legitimize) particular instances of executive downfall.

11.3   Vertical and Symmetric Procedures The strongest—that is, least contested—mechanisms involve constitutional instruments of removal that mirror the form of executive selection. A recall election “allows the voters of a given unit to remove an elected official from office before his or her term is complete” (Whitehead 2018: 1341). The vertical nature of this procedure is unquestionable in direct forms of recall, which voters activate through the collection of signatures. Instances of indirect recall, where parliament initiates a procedure eventually submitted to a popular vote, introduce by contrast a hybrid element of horizontal accountability into the process (Qvortrup 2011: 163). Welp and Milanese (2018: 1379) suggest that “when recall is activated by the authorities (…) its logic is similar to an impeachment, but with the extra step of requesting popular legitimation of the decision taken by the parliament or council.” The principle of symmetry underpins the recall procedure, even when an element of horizontal accountability dominates the process. For instance, Spivak notes that “unlike the current recall, many of the seventeenth and eighteenth-century versions involved the removal of an official by another elected body, such as a state legislature recalling its United States Representative in the Articles of Confederation government” (Spivak 2004: 22). In such cases recall operated in a purely horizontal manner, but symmetry remained unaffected. Only those with the authority to grant a mandate have authority to recall it. A vote of no confidence allows legislators to dismiss the prime minister when the government has lost its parliamentary majority. The principle of parliamentary confidence was established in the United Kingdom in the late eighteenth century. After 1841, it became conventional practice for prime ministers to resign or—more commonly—to ask monarchs for a general election when the House of Commons expressed lack of confidence in the government. Successful no-confidence votes are rare in the United Kingdom, given the majoritarian logic of the Westminster system and two-party dominance (Lijphart 2012), but no-confidence votes ended the Baldwin (1924) and the Callaghan (1979) cabinets in the twentieth century. Other no-confidence motions (e.g., against the May government in 2019) failed to gather majority support. In 2011, the Fixed-term

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

207

Parliaments Act codified this practice as one of the two procedures to terminate a government before the end of a five-year term—the other being an anticipated election (Norton 2016). Although no-confidence votes benefit from constitutional symmetry (the parliamentary majority forms and dismisses the government), their elitist nature can haunt their legitimacy. In Germany, the institution of “constructive” no confidence guarantees that the parliamentary majority must agree on a new government before dismissing the previous one, but the incoming cabinet must still call an early election. Piersig surmises that, even “though constructive non-confidence creates a constitutionally legitimate government, the demos views the mid-term transition as an usurpation of its ability to select the government” (Piersig 2016: 9–10). The German Constitutional Court ruled in 1983 that an early election is the conventional way to legitimize the government resulting from a constructive no-confidence vote. For the sake of clarity, we must distinguish votes of no confidence from two related institutions. A confidence motion initiated by the government is quite different from a no-confidence motion initiated by the opposition. A confidence motion threatens legislators with anticipated elections unless parliament supports the government or its policies (Huber 1996). Similarly, although censure is often synonymous with no confidence, censure motions sometimes refer to milder parliamentary procedures that do not trigger the resignation of the government (Norton 2016: 6). Although some presidential systems allow for legislative censure of cabinet ministers, votes of no confidence against the head of the government are distinctive of parliamentary or semi-presidential regimes.2 Congressional votes of “censure” and “no-confidence” against the chief executive occasionally take place in presidential systems, but they represent symbolic declarations without constitutional standing to unseat the president. In the United States, censure resolutions often express contempt for the president’s behaviour. For example, in 1848, the House of Representatives censured President James Polk for starting the Mexican-American war “unnecessarily and unconstitutionally.” In 1860, the House adopted a 2  The South African constitution empowers the National Assembly to remove the president through a no-confidence vote supported by a majority of its members (art. 74.92-2), but it also empowers the Assembly to impeach the president with a two-thirds majority (art. 74.102). An impeachment requires charges of a serious violation of the Constitution or the law, serious misconduct, or inability to perform the functions of office.

208 

A. PÉREZ-LIÑÁN

resolution expressing “reproof” of President James Buchanan for corruption in government contracts (Fisher 2014: 45–51). Between 1973 and 2016, legislators introduced seven resolutions of “censure” against President Richard Nixon (in the context of the Watergate scandal), five against Bill Clinton (in the context of the Lewinski scandal), seven against George W. Bush, and seven against Barack Obama. Most proposed resolutions challenged executive policies without calling for the president’s resignation. Only four (addressing Nixon and Clinton) asked the president to resign. Even in those cases, resolutions usually include technical language indicating they reflect “the sense of Congress,” to signal that votes of censure have “symbolic, rather than determinative, effects” (Maskell and Beth 2016: 5).

11.4   Asymmetric Procedures While no-confidence votes, despite circumstantial controversy, are normal episodes in the life of parliamentary regimes, legislative ousters of the chief executive represent traumatic events in the life of presidential regimes. Because elected presidents derive their legitimacy from the popular vote, congress must offer special justifications to overrule the popular will and dismiss the chief executive via impeachment, a pre-trial, or a declaration of incapacity. The impeachment procedure originated in the British parliamentary context, but it is now dormant in the United Kingdom. In the late eighteenth century, the Philadelphia convention borrowed—and tailored— this instrument to serve as a check on the elected executive (Gerhardt 1999). Presidential impeachment, thus redefined in Article II, Section 4 of the US Constitution, disseminated to later presidential systems. The impeachment procedure usually follows two models: in bicameral legislatures, the lower house initiates the accusation and the upper house rules in the trial against the president. In unicameral systems, a legislative committee or some proportion of assembly initiates the accusation, and the full assembly rules on the case.3 Supermajorities are normally required 3  Other variants are possible. In Nigeria, both chambers vote to order an independent investigation with a two-thirds majority and approve the report and the removal of the president with the same threshold (Lawan 2010). In Romania, a semi-presidential regime, presidential impeachment is subject to a popular vote, making this an instance of indirect recall (Gherghina and Mișcoiu 2013; Qvortrup 2011: 163).

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

209

in one or both legislative votes. Impeachment charges are not criminal charges, and conviction leads to the president’s removal from office, and often disqualification from future office (Gerhardt 1999; Pérez-­ Liñán 2007). Only three American presidents have confronted an impeachment trial, Andrew Johnson in 1868, Bill Clinton in 1998,  and Donald Trump in 2020, and none was removed from office. A congressional committee approved articles of impeachment against Richard Nixon in 1974, but the president resigned before the House of Representatives voted on the charges. Like recall elections, impeachments saw a global revival in the 1990s, albeit for different reasons. Recall elections were rarely allowed until the end of the Cold War, but Colombia, Ecuador, Peru, Poland, Venezuela, British Columbia, and 11 German Länder adopted them in the 1990s (Geißel and Jung 2018; Serdült and Welp 2017). Impeachment, in contrast, already existed in most presidential constitutions, but remained in a state of institutional atrophy. Presidents José P. Guggiari of Paraguay (who survived in office) and José R. Guizado of Panama (ousted) were impeached in 1931 and 1955, respectively, but episodes of this nature were rare. Yet, the end of the Cold War made military coups unjustifiable and encouraged civilian elites to leverage constitutional procedures against unpopular presidents. Several leaders were impeached and removed from office, or they resigned anticipating the outcome: Fernando Collor (1992) and Dilma Rousseff (2016) in Brazil, Raúl Cubas (1999) and Fernando Lugo (2012) in Paraguay, and Pedro Pablo Kuczynski (2018) in Peru.4 Closely connected to presidential impeachment are two related institutions. In a pre-trial—often described as a “judicial model” of impeachment—congress removes executive immunity, suspends the president from office, and authorizes a trial conducted by the Supreme Court or by the lower courts (Pérez-Liñán 2007). Presidents Carlos Andrés Pérez (1993) in Venezuela, Otto Pérez Molina (2015) in Guatemala, and Park Geun-hye (2017) in South Korea were removed under this procedure. With a declaration of incapacity, congress determines that the president is physically, mentally, or (under some constitutions) morally unable to dispatch the duties of the office.5 Congress may also claim that the 4  President Rolandas Paksas of Lithuania—a semi-presidential regime—was also impeached by parliament and removed from office in 2004 (Norkus 2008). 5  In the United States, the Twenty-Fifth Amendment (1967) established that the vicepresident and a majority of the cabinet must declare that the president is “unable to discharge

210 

A. PÉREZ-LIÑÁN

­ resident has abandoned his post. For instance, Carlos Luz (1955) in p Brazil and Abdalá Bucaram (1997) and Lucio Gutiérrez (2005) in Ecuador were ousted under such procedures. Beyond any specific differences, two elements are common to all three procedures. First, the constitution authorizes the legislative branch to activate or block the process, incorporating an explicitly political component to the procedure. Second, the constitution lists a series of criteria that justify an impeachment or declaration of incapacity, imposing restrictions on the process. The tension between these two principles—subjective political will vs. objective legal requirements—constitutes the distinctive problem for all mechanisms listed in the central box of Fig. 11.1 These conflicting precepts establish the range of possible constitutional interpretations for the removal of the president. If we accept that criteria for impeachment are deliberately vague and that the political element dominates this institution, impeachment becomes a functional equivalent to the parliamentary vote of no confidence, constrained only by the majority (or supermajority) thresholds established by the constitution. If we accept, on the contrary, that hard criteria are the constitutional core of the procedure and that parliamentary approval is just a mechanism to endow the process with legitimacy, the impeachment process becomes a quasi-­ judicial procedure and the declaration of incapacity, a quasi-­forensic decision. The first presidential impeachment, conducted in the United States in 1868, already demonstrated that these tensions are intrinsic to the institution. President Andrew Johnson fired his Secretary of War Edwin M. Stanton without legislative approval, prompting an accusation of high crimes and misdemeanours. Could firing a secretary be an impeachable offence? Congress charged the president with violating the Tenure of Office Act of 1867, which—oddly enough—required Senate authorization to dismiss cabinet members.6 However, the politics of this impeachment were not about the cabinet, they were about the Civil War. Elected as vice-president for Abraham Lincoln’s second term, Johnson took office after Lincoln’s assassination in 1865. In aftermath of the Civil War, he sought to pacify the South through concessions that infuriated the majority of his party committed to social reform. Johnson the powers and duties of his office.” If the president challenges the decision, only congress can declare the president impaired by a two-thirds majority in both houses. 6  This rule, eliminated two decades later, played on the principle of symmetry, since US presidents appoint cabinet members with the Senate’s consent.

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

211

pardoned ­rebels, derailed land reform in the South, and appointed former Confederates as provisional governors. The so-called “Radical” Republicans pushed for impeachment, but moderates in the party insisted that there were no grounds for an accusation. To regain control of the Reconstruction process, congress placed provisional governors under the jurisdiction of five military regions in the South and legislated that the president could only issue orders to the army through the Secretary of War, who secretly aligned with the Radical cause (Lewis 1954: 17). Then, to protect the Secretary of War, congress passed the Tenure of Office Act, which required cabinet members to serve for a full presidential term unless the Senate agreed to their removal. In August 1867, while congress had adjourned, Johnson suspended the Secretary of War and proceeded to replace military commanders in the South. Irate, the Senate invoked the Tenure of Office Act, and Secretary Stanton returned to his post in early 1868. Undeterred, Johnson sought to create a parallel department of war to bypass Stanton, but General William T. Sherman refused to take command. The president then argued that the Tenure of Office Act could not protect his Secretary of War, because Lincoln had appointed Stanton during his first term in office. In February 1868, Johnson removed Stanton, prompting moderate Republicans to side with Radicals in their ongoing calls for impeachment. The House impeached Johnson, but the Senate was unable to remove the president, falling one vote short of the required two-thirds majority (Benedict 1998). The South had lost the war, but eventually won the peace.

11.5   Narratives of Downfall Non-constitutional procedures are not the focus of this chapter, but they are relevant for our discussion insofar as they inform many narratives intended to legitimize or (more often) de-legitimize other procedures for removal. A popular revolt takes place when important sectors of the population mobilize to demand a premature end to an administration. Revolts are often enough to overthrow the government. Protests forced the resignation of Presidents Alberto Fujimori in Peru (2000), Fernando de la Rúa in Argentina (2001), and Gonzalo Sánchez de Lozada in Bolivia (2003) and of Governor Ricardo Rosselló in Puerto Rico (2019). Legislators merely formalized the ousting of Fujimori (who submitted his resignation from

212 

A. PÉREZ-LIÑÁN

Japan), appointed a successor to replace de la Rúa, stayed on the sidelines in Bolivia, and reluctantly speculated about impeachment in Puerto Rico. Popular revolts are not unconstitutional, because citizens have the right to oppose the government, but they are non-constitutional because no formal rule authorizes mass demonstrations to oust the chief executive. In the absence of such rule, demonstrators must justify why the will of the people seeking to overthrow the government today should prevail over the will of the people who casted votes in favour of the government at the last election. This question has no easy answer in split societies where a counter-movement defends the incumbent, such as Brazil in 2016. Hochstetler (2006) acknowledged the non-constitutional effectiveness of protests by reference to Stepan’s (1971) classic work on military intervention, noting that—in an age when military coups are rare—social movements have become the new “moderating power” in South America. In an insightful attempt to address popular revolts’ lack of Kelsenian standing, Leon Zamosc characterized the popular overthrow of neo-­ liberal presidents as “popular impeachments.” The term describes episodes in which “presidents have been removed or forced to resign as a result of political interactions in which street protests demanding their dismissal played a central role” (Zamosc 2013: 237). For Zamosc, popular protests exercise a vertical form of accountability in a context in which horizontal mechanisms of control have failed (Zamosc 2013). However, in most narratives of legitimation popular revolts validate impeachments, not the other way around. The vertical nature of revolts inspires an epic in which the people rise against corrupt rulers. Such framing provides scaffold to constitutional procedures otherwise tainted by their horizontal—that is, elite-driven—nature. In fact, popular protests have been crucial to back almost every successful impeachment or declaration of incapacity in the past two decades (Pérez-Liñán 2014).7 In a more controversial pattern of alliances, social movements have also legitimized military uprisings, as when the National Indigenous Federation of Ecuador (Conaie) backed Col. Lucio Gutiérrez’s coup against President Jamil Mahuad in January 2000 (Herrera Aráuz 2001; Pallares 2006).

7  Notable exceptions have been the impeachments of Presidents Fernando Lugo of Paraguay (2012) and Pedro Pablo Kuczynski (2018) of Peru. Lugo and Kuczynski were not popular at the time of their ousting, but they did not confront the massive protests against the government that characterized other cases.

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

213

The last point leads to the contested issue of coups. Coups are blatantly non-constitutional and elite driven, producing the least legitimate form of removal. For that reason, narratives that oppose executive ousters often recast any form of removal in the shape of a coup. “Those who organized this failed coup should be held responsible before state institutions,” clamoured of Romanian President Traian Băsescu a day after surviving a recall election in 2012. In Canada, the Conservative government decried an ongoing “coup d’état” when a coalition formed by the Liberal Party, the New Democratic Party, and the Bloc Québécois proposed a no-­ confidence motion in late 2008 (Russell and Sossin 2009). Lavinia Stan notes that “no aspect of Romanian politics has been vilified more than the no-confidence censure motion, after two such motions unexpectedly unseated the center-right cabinets [in 2009 and 2012]. The center-left opposition was faulted for grabbing power through a ‘legislative coup,’ without winning the popular vote, by virtue of a defect of democracy” (Stan 2015: 291). Proponents of a coup narrative reversed ideological camps a few years later, during the impeachment of Brazilian President Dilma Rousseff in 2016. Rousseff, elected for the centre-left Worker’s Party, insisted that “an impeachment without a proven crime of responsibility is a coup.” Echoing this view, Santos and Guarnieri (2016) argued that the political crisis that undermined her administration began with mass protests against public services in 2013, accelerated with an economic meltdown in 2014–2015, and concluded in the midst of a massive corruption scandal involving politicians of all parties. The corruption investigation did not affect President Rousseff, but congressional leaders engaged in a “farce” to divert attention, charging her with manipulating the budget to cover up fiscal deficits. Santos and Guarnieri concluded that the process was a “parliamentary coup.” Avelar (2017) challenged this characterization, pointing to “a gap between the rhetoric of the coup and (…) the country’s coalitional presidentialism.” Nunes and Melo (2017) declined to call the impeachment a coup, but noted that treating budgetary manoeuvres as a crime of responsibility was legally dubious. The Brazilian case shows that the coup narrative can be particularly damning against impeachments, which are tainted by the asymmetric nature of the procedure. But whether the coup rhetoric is effective to de-­ legitimize an impeachment is far from clear. Critical narratives intended to convey that an impeachment is as bad as a military coup may inadvertently convey that a military coup is as good as an impeachment. In March 2018,

214 

A. PÉREZ-LIÑÁN

the Instituto da Democracia e da Democratização da Comunicação conducted a survey in Brazil asking respondents if they believed that the impeachment of Dilma Rousseff had been a coup.8 The survey also asked whether citizens would justify a military coup in a situation of “great corruption.” About 52% of respondents identified the impeachment as a coup. Of them, however, some 47% admitted that they would justify a military coup. This means that almost a quarter of the sample accepted the impeachment of Dilma Rousseff as a “good” coup against corruption. The election of President Jair Bolsonaro—who unabashedly defended the military dictatorship of 1964–1985—later that year gains new meaning from this perspective (Pérez-Liñán 2018: 8). Marsteintredet and Malamud (2019) note that the increasing denounciation of “parliamentary coups” is part of a broader trend involving a proliferation of coups with adjectives. Political and academic narratives increasingly refer to parliamentary, soft, constitutional, neo-liberal, market, electoral, slow-motion, civil-society, and judicial coups. Those adjectives hedge the noun (coup) to create diminished subtypes (e.g., soft coup). The authors argue that this trend reflects a pattern of prevalenceinduced concept change: because instances of traditional military coups have become less prevalent in recent decades, users have expanded the meaning of the concept to retain its utility. The authors properly note that the use of diminished subtypes risks causal and conceptual confusion. For instance, any analysis showing that military coups and impeachments have common causes requires a conceptual separation between the two outcomes (Alvarez and Marsteintredet 2010; Pérez-Liñán and Polga-­Hecimovich 2017). For that reason, I propose that we reserve the term coup to describe an episode in which elites forcefully remove the chief executive. In practice, this means that a military (or paramilitary) operation terminates the government. This criterion is easy to operationalize and sufficient to distinguish a coup from an impeachment, even an illegitimate one. A coup d’état effectively suspends the constitution; an illegitimate impeachment effectively expands its interpretation. In an illegitimate impeachment, congress affirms its constitutional authority to oust the president, even if grounds for impeachment are absent. It entails, in short, a reinterpretation of the constitution to define impeachment as a vote of no confidence against the government. Thus, illegitimate impeachments are an instance  https://www.institutodademocracia.org/a-cara-da-democracia.

8

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

215

of congressional abuse of power; they exploit legislators’ Kelsenian standing without invoking a state of exception. Sometimes congress grants a veneer of legitimacy to a coup by declaring that the president is unable to discharge the duties of the office, but the security forces—not congress—are in such cases the efficient cause of the ouster. For example, Brazilian military and legislators joined forces against interim President Carlos Luz when he backed a conspiracy against President-elect Juscelino Kubitschek in 1955. Aware of the plot against Kubitschek, the minister of the Army sent tanks to surround the presidential palace. Luz escaped and boarded a Navy vessel to leave Rio, but coastal artillery stopped the boat in the Guanabara Bay. In the meantime, supporters of Kubitschek in congress voted to “acknowledge” that Luz was unable to govern and appointed his replacement under article 79 of the 1946 constitution. The motion passed with supermajorities (185–72  in the Chamber of Deputies and 43–9 in the Senate), and the speaker of the Senate replaced Luz (Dulles 1970). Such coalition of soldiers and legislators produces a “legislative coup,” but this is in no way a diminished subtype. Under this definition, the congressional removal of Brazilian President Carlos Luz was a coup, but the impeachment of Panamanian President José Ramón Guizado on the same year was not; the removal of Honduran President Manuel Zelaya in 2009 was a coup, but the impeachment of Paraguayan President Fernando Lugo in 2012 was not. The adoption of this operational rule, focusing on the forceful removal of the executive, does not require the acceptance of congressional abuses. Guizado was unjustly accused of conspiring to kill his predecessor and incarcerated under this false pretence. Lugo was impeached within 48 hours, without any possibility of defence. Both episodes are blatant examples of congressional abuses in the use of impeachment, even if we do not label them as coups.

11.6   Judging Asymmetric Procedures Can we distinguish good impeachments from bad ones? And if so, how? Figure 11.1 illustrates why presidential impeachments (as well as pre-trials and declarations of incapacity) are particularly controversial. Impeachment occupies an ambiguous space in the legitimacy range: it is constitutional, but it is also an asymmetric tool of horizontal accountability. The legitimacy of a popularly elected president enters in conflict with the constitutional authority of the legislature to suspend or remove the chief executive.

216 

A. PÉREZ-LIÑÁN

This tension imposes distinctive challenges to those seeking to pass judgement on asymmetric processes. In many instances, normative doubts are of little relevance in the face of political reality. When the president is very popular, there is little that a zealous legislature can do to exercise oversight. When the president has lost all popular support, electoral legitimacy is no protection for a disgraced executive (Pérez-Liñán 2014). However, when society is divided, the use of asymmetric mechanisms to remove the government becomes highly contested. Citizens, scholars, and judges need common criteria to assess the validity of such procedures. This section advances three standards to assess the legitimacy of an asymmetric process of removal. I use the term legitimacy in a normative, not in a sociological sense. The question is not whether the process is subjectively legitimate for public opinion or for the legislators ousting the president, but whether we can establish external criteria to distinguish legitimate impeachment processes from episodes of congressional abuse of power. This is no easy task, even for seasoned jurists. In 2017, the Inter-­ American Commission of Human Rights asked the Inter-American Court to issue a consultative ruling addressing “the guarantees of due process and the principle of legality in the context of the impeachment of constitutional and democratically elected presidents.” At the time, the Commission was considering petitions related to the ousting of Manuel Zelaya in Honduras (2009), Fernando Lugo in Paraguay (2012), and Dilma Rousseff in Brazil (2016). Handed a hot potato, the Court politely declined to answer, to avoid “a premature opinion in cases that it could be called on to hear in the exercise of its contentious jurisdiction.” It also surmised that “the Constitutions of the OAS Member States contain numerous different provisions regulating impeachment proceedings (…) that make it very difficult to reduce the inter-American standards to a minimum common denominator.” Judge Patricio Pazmiño dissented, offering a lonely opinion in which he lamented that “the Court is foregoing an important opportunity to develop international human rights law.”9 Figure 11.1 helps us identify three additional questions to assess the controversial procedures located in the central box. Since those procedures are constitutional, can legislators prove that the activation of an impeachment or declaration of incapacity is consistent with the motives 9  Inter-American Court of Human Rights, “Request for an Advisory Opinion Presented by the Inter-American Commission on Human Rights,” May 29, 2018 (paragraphs 8 and 14). Available at http://www.corteidh.or.cr/solicitudoc/sor_01_18_ing.pdf.

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

217

stated by the constitution? Since those procedures are horizontal, can legislators prove that they act on behalf of their constituencies or their parties, without ulterior motives to oust the government? Finally, since legislators claim constitutional authority to exercise an asymmetric action against the executive, do they follow a reasonable process, balancing their constitutional authority against the rights of the chief executive and his or her voters? These questions translate into three normative standards: legality, fairness, and due process. Legality Although impeachment is not a criminal procedure, it remains subject to the principle of legality (nulla poena sine lege), that is, the idea that no person shall be punished “except in pursuance of a statute which fixes a penalty” (Hall 1937). The constitution—and in some cases, complementary laws—establishes boundaries to the meaning of impeachable offences. Bowing to this principle, the Romanian constitution requires a (nonbinding) opinion of the Constitutional Court before parliament can initiate an impeachment procedure against the president. The legality principle constrains congressional action in a broad sense. In cases of impeachment, legislators must prove that the president has committed an impeachable offence. Under a pre-trial, legislators must show sufficient evidence of a typified crime that justifies a judicial process. When congress is empowered to judge the executive’s physical or mental incapacity, the constitution implies criteria to define impairment. The principle of legality acquires in this context broader meaning, but congress is still bound to justify it reasons to declare the president incapacitated according to the law. In practice, however, the principle of legality imposes weak restrictions to the removal of the president, for at least two reasons. First, the grounds for impeachment articulated in constitutions are deliberately vague and in some cases contemplate criteria that explicitly resemble a vote of no confidence. The constitutions of Argentina and Paraguay, for example, recognize “misperformance in office” as justification for impeachment. In Nigeria, impeachment follows from an allegation of “gross misconduct.” At the US Constitutional Convention in Philadelphia, George Mason proposed “maladministration” as a reason for impeachment. James Madison objected that the term was too vague, and Mason offered “high crimes and misdemeanors” as an alternative. Thus, when Hamilton analysed the

218 

A. PÉREZ-LIÑÁN

meaning of this clause in Federalist No. 65, he concluded that such offences are political, not criminal, in nature (Fisher 2014: 41–42). Complementary legislation can address constitutional vagueness through clarification of impeachment criteria. For example, Brazil’s Law No. 1079 (1950) defined the range of crimes of responsibility in 12 extensive articles. Auxiliary laws are convenient because they restrict the legislature’s margin of appreciation in advance, but such codification is unusual. Moreover, a detailed legal framework does not prevent constitutional controversy in the context of a political crisis. The experience of Brazil in 2016 suggests that the debate on the legality of the impeachment will become in such cases a debate on the interpretation of the auxiliary law. Second, although the legality principle constrains the range of charges that congress can leverage against the president, it does not prevent false accusations. As mentioned before, Panamanian President José Ramón Guizado was falsely accused of conspiracy to murder his predecessor and impeached in 1955. Because some constitutions suspend the president for the duration of the impeachment process, legislators can achieve the goal of replacing the government even if accusations ultimately prove to be false. If the principle of legality imposes such weak restrictions to the acts of congress, how can it help us assess the legitimacy of an impeachment process? Grounds for impeachment are deliberately vague, and legislators can interpret them broadly. Yet, there is an intertemporal aspect of this principle that is worth emphasizing. Every time congress articulates an interpretation of impeachment charges, it establishes a precedent. Observers can judge the reasonableness of this interpretation in retrospective and in prospective terms. Looking back: Is the interpretation of impeachable offences consistent with interpretations invoked by congress on previous occasions? Looking forward: What would the consequences be if legislators applied this precedent consistently in the future? Could presidents repeatedly be subject to removal under arbitrary criteria? Would this precedent systematically undercut the will of the electorate? Answers to these questions will not define the outcome of a presidential crisis, but they can provide external observers with guidelines to assess the principle of legality. Concerns about the principle of legality matter for some symmetric forms of removal, but they are less prominent. For instance, some constitutions require specific “crimes” such as corruption or malfeasance, to activate recall elections. The Ecuadorian constitution of 1998 authorized recall elections only when officials deviated from the government programme and in cases of corruption, after a judicial ruling (Welp and

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

219

Milanese 2018: 1381). In the United States, 11 states have unconstrained recall laws, while only 8 states limit the procedure to cases of malfeasance (Whitehead 2018: 1350). Qvortrup asserts that “the recall mainly is used in states where a recall process is a political tool. In states where the representative must be guilty of some form of malfeasance or misconduct, the recall is less frequently used” (Qvortrup 2011: 165). Fairness In a trial, judges are expected to be impartial, but this criterion is particularly problematic for asymmetric forms of removal. Legislators have an electoral connection to voters; they perceive impeachment as a legitimate response to popular mobilization demanding the ouster of unpopular presidents. Indeed, multiple studies have shown that social protests against the government are one of the main factors predicting the activation of an impeachment (Hochstetler 2006; Pérez-Liñán and Polga-Hecimovich 2017; Zamosc 2013). Moreover, lawmakers belong to parties, and some are subject to formal rules of party discipline. Given that the constitution places congress—and therefore, political parties—at the centre of this process, it would be illogical to demand a non-partisan process. Acknowledging that legislators form their preferences in line with the preferences of their parties and their constituents precludes a naïve understanding of impartiality, but it does not preclude a call for fairness. The Nigerian constitution implicitly recognizes this issue, ordering an independent panel to investigate accusations against the president (Lawan 2010). Fairness implies that legislators must decide in accordance with their responsibilities as representatives and party members, but not instrumentalize the impeachment process for other purposes. Standards of fairness are openly questioned when parliamentary leaders raise the threat of impeachment as a negotiating strategy in their interactions with the executive branch. For example, the president of the Brazilian Chamber of Deputies brandished the threat of impeachment against President Dilma Rousseff, hoping (unsuccessfully) that the ruling party would protect him from corruption investigations. Rousseff declined the offer, and the impeachment process moved to the committee stage. In late 2017, an opposition faction in Peru deactivated the first impeachment against President Pedro Pablo Kuczynski in exchange for a presidential pardon benefiting Alberto Fujimori. The transactional nature of these processes exposes the partiality of legislative leaders, independently of the

220 

A. PÉREZ-LIÑÁN

validity of accusations against the president. If accusations were valid, these actions show that parliamentary leaders were willing to exchange executive impunity for their own impunity. If accusations were false, these actions show that legislators were using their constitutional power as an instrument of extortion. The question of fairness occasionally arises for recall elections, which can be “a weapon in the hands of partisan actors” (Welp and Milanese 2018: 1392). In Peru, local politicians who win by narrow margins and women are more likely to face recall attempts (Holland and Incio 2019; see also Welp 2016). The issue is less salient in parliamentary systems, where the vote of no confidence is, by definition, a partisan affair. No-confidence motions represent visible opportunities for opposition parties to highlight the government’s incompetence. The purpose of most motions is not to bring the cabinet down—they succeed only 5% of the times—but to improve the proponent’s position at the next election. Williams (2011) shows that incumbents suffer an electoral penalty when they confront no-confidence motions, even if the motions fail. However, concerns about ulterior motivations may arise in subtler forms. For example, Trinidad and Tobago’s prime minister announced in 2012 a cabinet reshuffle involving the appointment of 30 cabinet members. Because the House of Representatives only has 41 members, every legislator from the ruling coalition was either a cabinet minister or a deputy speaker. In this case, personal interests prevented the fair exercise of symmetric accountability. Monteil argued that the intentional absence of backbenchers made a vote of no confidence all but impossible in a Westminster system like Trinidad and Tobago. He concluded that, with a gigantic cabinet, “there is no responsible government as contemplated by the 1976 Constitution, because members of the Executive (…) account to themselves” (Monteil 2015: 254). Due Process The principle of due process is perhaps less ambiguous than the principles of legality and fairness. Although the impeachment procedure is rarely codified in detail, flagrant violations of due process are easy to pinpoint. These violations manifest themselves not only as deviations from reasonable practice in a (quasi) criminal procedure but also as deviations from standard parliamentary rules. Among the most visible manifestations of

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

221

this problem are the absence of sensible deadlines, the disruption of parliamentary procedures, and the selective use of constitutional norms. The absence of sensible deadlines is well illustrated by the ouster of President Fernando Lugo of Paraguay in 2012. The events triggering the process occurred on June 15, when a violent confrontation between landless peasants and the police produced several deaths. Lugo’s legislative coalition broke down, and the Chamber of Deputies approved impeachment charges within a week, on June 21. The following day, the Senate voted to remove the president from office. Congress completed the process within 48  hours, raising an international thunderstorm of criticism (Marsteintredet et al. 2013). The fast-track decision prevented the president from mounting a legal defence but more importantly, it prevented him from mounting a political defence. Lugo was not a popular president at the time, but a longer impeachment process would have granted him additional time to mobilize social forces, activate international alliances, or (most likely) to negotiate an alternative coalition in congress. Another example from Paraguay exemplifies the disruption of parliamentary protocol. Raúl Cubas Grau was charged with contempt for the Supreme Court, but the reason behind his removal was a widespread suspicion that he had conspired to kill his vice-president, shot in March 1999. In a rush to charge Cubas, the Chamber of Deputies modified its internal procedures to redefine a two-thirds majority as 66.5% of the votes rather than 67%.10 Upset with this lack of civility, the president’s supporters cut the cords of the chamber’s sound system to delay the decision. The session to vote on impeachment charges took place on short notice, and some deputies aligned with the president had to force their way into the chamber. According to some interviews, one of the deputies who opposed impeachment was locked in a restroom to prevent him from voting. The constitutional accusation ultimately passed in the lower house with a difference of just one vote (Pérez-Liñán 2007). Few episodes expose the selective use of constitutional criteria as neatly as the Ecuadorian crisis of 1997. Abdalá Bucaram was a flamboyant president who, seven months into his administration, already confronted multiple accusations of corruption and nepotism. He also adopted an unpopular economic policy that involved fiscal austerity and price increases 10  Along similar lines, Lawan (2010) notes that local legislatures in Nigeria declared some seats vacant in order to reduce the number of members present to achieve a two-thirds majority.

222 

A. PÉREZ-LIÑÁN

for public services. His actions invited an unlikely alliance of left-wing movements who opposed his programme, right-wing parties excluded from power, poor demonstrators hurt by economic adjustment, and middle-­class protestors upset with his outrageous style (Mejía Acosta and Polga-Hecimovich 2011). In February 1997, mass demonstrations called for the ouster of the president. Opposition leaders realized that this was a unique opportunity to remove Bucaram from office, but they could not muster the two-thirds majority required to activate an impeachment process. However, they selectively invoked article 75 of the 1979 constitution, which allowed congress to declare Bucaram mentally incapacitated by simple majority. Despite the president’s erratic behaviour—which included the release of his debut album, A Loving Madman—the evidence to argue that he was mentally ill was medically insubstantial. Concerns about due process are less sensitive for no-confidence motions or recall elections, as “due process” reduces to meeting the legal requirements for activation. For example, parliamentary constitutions usually demand that no-confidence motions be initiated by a fifth to a quarter of legislators, debated within a certain time frame, and adopted by a majority. If the motion fails, the constitution may limit the ability of parliament to raise a new motion for some time. Constructive votes of no confidence additionally require that the coalition dismissing the cabinet be ready to elect a new government (Piersig 2016). For executive recalls, constitutions often impose rules about the period of activation, the time to collect signatures, and the number of signatures required. Tighter requirements raise the costs for the opposition (Geißel and Jung 2018; Welp 2016), and incumbents may question whether the opposition has properly adhered to those rules. Issues of due process in recall elections often relate to the forgery of signatures and the ability of electoral bodies to derail the process (Kornblith 2005). However, the meaning of the rules is relatively uncontroversial in comparison to the question of due process in an impeachment or declaration of incapacity. Taken in isolation, the principles of legality, fairness, and due process are rather blunt criteria to assess the legitimacy of horizontal mechanisms of asymmetric accountability. Constitutional grounds for accusation are ambiguous, legislators’ true goals are unknown, and impeachment procedures are rarely well defined. Taken together, however, these three principles provide a set of benchmarks to assess whether the process to remove a chief executive is conducted reasonably. Because violations along one dimension often correlate with violations in other dimensions, these

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

223

c­ riteria help us to determine if, and to what extent, congress deploys its constitutional powers in an arbitrary manner. Unfortunately, very few impeachments or declarations of incapacity pass the test unharmed.

11.7   From Narratives to Standards This chapter has argued that recall elections represent the institutional design most likely to produce a legitimate process to oust an elected chief executive. Recall elections are validated by a symmetric design in which the same people who elected the head of the government also rescind his or her mandate. No-confidence votes, the closest alternative, enjoy a similar advantage—the parliamentary majority that appoints the prime minister also dismisses the cabinet—but they suffer from the lack of popular participation. In presidential regimes, absent a vote of no confidence, impeachment becomes the main constitutional instrument to remove the president. However, because congress decides on the charges against a popularly elected president, this procedure lacks the legitimacy granted by either popular participation or electoral symmetry. As a result, narratives of impeachment often recast the process as a “parliamentary coup.” In the light of their intrinsic advantage, it is surprising that recall elections have not emerged as an effective tool of vertical accountability for national executives after the third wave of democratization. Although recall elections can be frequent at the local level (Welp 2016), no recall election has ousted a national executive to this day. Recall elections against Presidents Hugo Chávez of Venezuela (2004), Evo Morales of Bolivia (2008), and Traian Băsescu of Romania in 2007 and 2012 consistently failed to dislodge the incumbents. Several reasons may explain the weak role of recall elections in national politics, including the fact that recall procedures are relatively new in many countries and remain absent from the institutional menu in many others. But an intrinsic feature of this procedure deserves attention: recalls are extraordinary elections, a fact that creates three institutional challenges. First, in the context of a national political crisis, recall elections lack the expediency of a no-confidence motion or an impeachment process. Second, political controversy may encourage voter turnout in some contexts but discourage turnout in others, making the procedure ineffective.11 11  In Romania, only 25% of voters supported the dismissal of President Traian Băsescu in the 2007 recall. In contrast, over 88% supported his recall in 2012, but turnout failed to

224 

A. PÉREZ-LIÑÁN

Third, the national government may have undue influence in the organization of the electoral process, derailing the recall or delaying it to gain advantage. The last point is particularly relevant. Referring to Ecuador, Welp noticed that “when electoral bodies evaluate reasons for activation, they tend to prevent the implementation of recall referendums” (Welp 2016: 1169). No case illustrates this problem better than the recall against President Hugo Chávez of Venezuela. In mid-2003, opposition parties presented more than three million signatures requesting a recall. Fearing defeat, the president adopted new social policies to mobilize popular support, but he needed time for his approval rates to go up. A delay of the recall process was essential. The ruling party thus stalled the nomination of new members for the National Electoral Council (CNE), which required consensus (two-third majorities) in the National Assembly. Invoking this inaction, the government-controlled Supreme Court unilaterally appointed the new board members and gave the CNE broad powers to oversee the referendum. The CNE rejected the initial signatures and requested a new signature drive. It then took a long time to verify the new signatures, declaring many of them invalid on technicalities and delaying recognition of sufficient signatures until mid-2004, when presidential approval ratings had recovered (Kornblith 2005). Once the date for the recall was set, the CNE tampered with the design of the ballot, altered prior voting procedures, and reshuffled polling stations, creating great uncertainty. In the end, only 41% of voters supported the recall. International observers verified that the vote count was fair (McCoy 2005), but critics doubt the results to this day (Delfino and Salas 2011). In later years, the administration blacklisted citizens who had signed in favour of the recall for public jobs and government contracts. In October 2016, the CNE blocked another recall election against President Nicolás Maduro. The Venezuelan regime was only willing to hold elections it could win. Thus, although recall elections offer “a sensible alternative to impeachment” (Welp 2016: 1164), it appears that asymmetric procedures of horizontal removal will remain part of the menu to depose unpopular executives for years to come. This makes a debate about the proper use of such procedures urgent and necessary. Any theoretical debate about proper use of impeachment must acknowledge that two problems may undercut is legitimacy (Kada 2003). The first achieve the 50% required for the election to be valid (Gherghina and Mișcoiu 2013).

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

225

one is an error of omission: legislators may shield the executive from an investigation when there is sufficient evidence to pursue it, creating impunity. The second one is an error of action: legislators may remove the executive under false accusations or illegitimate proceedings, undermining the rights of the president and his or her voters. Because this chapter compared alternative forms of removal, the discussion implicitly focused on the second problem, but debates about the valid use of asymmetric procedures should anticipate of both types of error. Legal scholars confront the challenge of articulating standards that are both realistic and able to expose abuse of power by legislators. Such standards must be developed and debated outside the pressing context of presidential crises. This chapter has outlined three principles—legality, fairness, and due process—to assess asymmetric procedures. Additional research will be needed to establish whether such criteria can be operationalized into legal tests and probe their application to historical cases.

References Alvarez, M.  E., & Marsteintredet, L. (2010). Presidential and Democratic Breakdowns in Latin America: Similar Causes, Different Outcomes. In M.  Llanos & L.  Marsteintredet (Eds.), Presidential Breakdowns in Latin America (pp. 33–52). New York: Palgrave Macmillan. Avelar, I. (2017). A Response to Fabiano Santos and Fernando Guarnieri. Journal of Latin American Cultural Studies, 26(2), 341–350. Benedict, M. L. (1998). From Our Archives: A New Look at the Impeachment of Andrew Johnson. Political Science Quarterly, 113(3), 493–511. Bueno de Mesquita, B., Smith, A., Siverson, R. M., & Morrow, J. D. (2003). The Logic of Political Survival. Cambridge, MA: MIT Press. Chilcote, R.  H. (2014). A Revolução Portuguesa: Estado e Classes Sociais na Transição para a Democracia. Porto: Edições Afrontamento. Delfino, G., & Salas, G. (2011). Analysis of the 2004 Venezuela Referendum: The Official Results Versus the Petition Signatures. Statistical Science, 26(4), 479–501. Dulles, J. (1970). Unrest in Brazil. InPolitical-Military Crises, 1955–1964. Austin: University of Texas Press. Fisher, L. (2014). The Law of the Executive Branch: Presidential Power. Oxford: Oxford University Press. Geißel, B., & Jung, S. (2018). Recall in Germany: Explaining the Use of a Local Democratic Innovation. Democratization, 25(8), 1358–1378. Gerhardt, M. J. (1999). Lessons of Impeachment History. George Washington Law Review, 67(3), 603–625.

226 

A. PÉREZ-LIÑÁN

Gherghina, S., & Mișcoiu, S. (2013). The Failure of Cohabitation: Explaining the 2007 and 2012 Institutional Crises in Romania. East European Politics & Societies and Cultures, 27(4), 668–684. Hall, J. (1937). Nulla Poena Sine Lege. The Yale Law Journal, 47(2), 165–193. https://doi.org/10.2307/791967. Herrera Aráuz, F. (2001). Los Golpes del Poder al Aire: El 21 de Enero a Través de la Radio. Quito: Abya Yala. Hinojosa, V.  J., & Pérez-Liñán, A.  S. (2007). Presidential Survival and the Impeachment Process: The United States and Colombia. Political Science Quarterly, 121(4), 653–675. Hochstetler, K. (2006). Rethinking Presidentialism: Challenges and Presidential Falls in South America. Comparative Politics, 38(4), 401. Holland, A., & Incio, J. (2019). Imperfect Recall: The Politics of Subnational Office Removals. Comparative Political Studies, 52(5), 777–805. Huber, J. D. (1996). The Vote of Confidence in Parliamentary Democracies. The American Political Science Review, 90(2), 269–282. Kada, N. (2003). Impeachment as a Punishment for Corruption? The Cases of Brazil and Venezuela. In J.  C. Baumgartner & N.  Kada (Eds.), Checking Executive Power: Presidential Impeachment in Comparative Perspective (pp. 113–135). Westport, CN: Praeger. Kelsen, H. (1967 [1934]). Pure Theory of Law. Berkeley: University of California Press. Kornblith, M. (2005). Elections versus Democracy. Journal of Democracy, 16(1), 124–137. Lawan, M. (2010). Abuse of Powers of Impeachment in Nigeria. The Journal of Modern African Studies, 48(2), 311–338. Lewis, H.  H. W. (1954). The Impeachment of Andrew Johnson: A Political Tragedy. American Bar Association Journal, 40(1), 15–87. Lijphart, A. (2012). Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries (2nd ed.). New Haven: Yale University Press. Marsteintredet, L., Llanos, M., & Nolte, D. (2013). Paraguay and the Politics of Impeachment. Journal of Democracy, 24(4), 110–123. Marsteintredet, L., & Malamud, A. (2019). Coup with Adjectives: Conceptual Stretching or Innovation in Comparative Research? Political Studies, forthcoming. https://doi.org/10.1177/0032321719888857. Maskell, J., & Beth, R.  S. (2016). Congressional Censure and “No Confidence” Votes Regarding Public Officials. Washington, DC: Congressional Research Service – RL34037. McCoy, J. (2005). One Act in an Unfinished Drama. Journal of Democracy, 16(1), 109–123. Mejía Acosta, A., & Polga-Hecimovich, J. (2011). Coalition Erosion and Presidential Instability in Ecuador. Latin American Politics and Society, 53(2), 87–111.

11  NARRATIVES OF EXECUTIVE DOWNFALL: RECALL, IMPEACHMENT, OR… 

227

Monteil, R.  L. (2015). Constitution Reform for Responsible Government in Trinidad and Tobago. Oxford University Commonwealth Law Journal, 15(2), 245–268. Norkus, Z. (2008). Carl Schmitt as a Resource for Democratic Consolidation Studies: The Case of the President’s Impeachment in Lithuania. East European Politics & Societies, 22(4), 784–801. Norton, P. (2016). The Fixed-Term Parliaments Act and Votes of Confidence. Parliamentary Affairs, 69(1), 3–18. Nunes, F., & Melo, C. R. (2017). Impeachment, Political Crisis and Democracy in Brazil. Revista de ciencia política (Santiago), 37(2), 281–304. O’Donnell, G. (1998). Horizontal Accountability in New Democracies. Journal of Democracy, 9(3), 112–126. Pallares, A. (2006). Mass Mobilization and Presidential Removal in Ecuador: Entre la Ira y la Esperanza. LASA Forum, 37(1), 22–25. Pérez-Liñán, A. (2007). Presidential Impeachment and the New Political Instability in Latin America. Cambridge: Cambridge University Press. Pérez-Liñán, A. (2014). A Two-Level Theory of Presidential Instability. Latin American Politics and Society, 56(1), 34–54. https://doi.org/10.1111/j. 1548-2456.2014.00220.x. Pérez-Liñán, A. (2018). Impeachment or Backsliding? Threats to Democracy in the Twenty-First Century. Revista Brasileira de Ciências Sociais, 33(98), 1–14. Pérez-Liñán, A., & Polga-Hecimovich, J. (2017). Explaining Military Coups and Impeachments in Latin America. Democratization, 24(5), 839–858. Piersig, E. (2016). Reconsidering Constructive Non-Confidence for Canada: Experiences from Six European Countries. Canadian Parliamentary Review, 39(3), 5–15. Qvortrup, M. (2011). Hasta la Vista: A Comparative Institutional Analysis of the Recall. Representation, 47(2), 161–170. Russell, P. H., & Sossin, L. (Eds.). (2009). Parliamentary Democracy in Crisis. University of Toronto Press. Santos, F., & Guarnieri, F. (2016). From Protest to Parliamentary Coup: An Overview of Brazil’s Recent History. Journal of Latin American Cultural Studies, 25(4), 485–494. Schmitt, C. (2014 [1921]). Dictatorship. Cambridge: Polity Press. Serdült, U., & Welp, Y. (2017). The Leveling Up of a Political Institution: Perspectives on the Recall Referendum. In S. P. Ruth, Y. Welp, & L. Whitehead (Eds.), Let the People Rule? Direct Democracy in the Twenty-First Century (pp. 137–154). Colchester, UK: Routledge. Smulovitz, C., & Peruzzotti, E. (2000). Societal Accountability in Latin America. Journal of Democracy, 11(4), 147–158. Spivak, J. (2004). California’s Recall. Adoption of the ‘Grand Bounce’ for Elected Officials. California History, 81(2), 20–37.

228 

A. PÉREZ-LIÑÁN

Stan, L. (2015). Between Democracy and Putsch? Censure Motions in Romania (1989–2012). Communist and Post-Communist Studies, 48(4), 291–300. Stepan, A.  C. (1971). The Military in Politics. Changing Patterns in Brazil. Princeton: Princeton University Press. Welp, Y. (2016). Recall Referendums in Peruvian Municipalities: A Political Weapon for Bad Losers or an Instrument of Accountability? Democratization, 23(7), 1162–1179. Welp, Y., & Milanese, J. P. (2018). Playing by the Rules of the Game: Partisan Use of Recall Referendums in Colombia. Democratization, 25(8), 1379–1396. Whitehead, L. (2018). The Recall of Elected Office Holders. The Growing Incidence of a Venerable, but Overlooked, Democratic Institution. Democratization, 25(8), 1341–1357. Williams, L.  K. (2011). Unsuccessful Success? Failed No-Confidence Motions, Competence Signals, and Electoral Support. Comparative Political Studies, 44(11), 1474–1499. https://doi.org/10.1177/0010414011407470. Zamosc, L. (2013). Popular Impeachments: Ecuador in Comparative Perspective. In M.  Sznajder, L.  Roniger, & C.  A. Forment (Eds.), Shifting Frontiers of Citizenship: The Latin American Experience (pp. 237–266). Leiden: Brill.

CHAPTER 12

On Reconciling Recall with Representation Laurence Whitehead

12.1   Introduction In this volume we have assembled a broad collection of up-to-date studies of presently ongoing experiments and discussions concerning the recall of elected officeholders in contemporary electoral democracies. The underlying issue linking these chapters is whether the “recall” strand in the long tradition of thinking about direct democracy can serve to reinvigorate the dominant model of representative government through party alternation via periodic elections. That model looks past its prime and is coming under mounting stress from multiple sources. These include new citizen expectations of voice and accountability (partly stimulated by new forms of political communication); the increased professionalization of electoral careers that can create distrust and an “us/them” divide between rulers and ruled; and rising dissatisfaction in many countries with what is thought to be officeholder self-serving and even “corruption”. In principle “recall” might provide a remedy to these perceived ills of democracy. This procedure (together with a variety of other “direct democracy” innovations that are also spreading) can tap into popular energies and enthusiasms and might help to relegitimize faltering traditional structures of representation. However, the direct “will of the people” should not be lightly L. Whitehead (*) Nuffield College, Oxford, UK e-mail: [email protected] © The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9_12

229

230 

L. WHITEHEAD

invoked for it not only activates creative new sources of political engagement, it also disturbs the delicate flow of policy negotiation and production that lubricates complex systems of mutually restraining representation. An ecology of reflection and deliberation is required to produce sound policy decisions and to adjust them in the light of criticism and feedback. Powerful officeholders need enough security to work through such matters and to dialogue about complex alternative options. “Recall” is a blunt instrument that could reinvigorate representative government, but that also risks disrupting the ecology of deliberation and the informed production of orderly policy outputs. A century ago Max Weber theorized about this balance in terms of the tension between a (plebiscitary) ethic of conviction and a (parliamentary) ethic of responsibility. In an era of mass politics, both rationales for action have their place in the architecture of a democratic regime. But as Weimar subsequently demonstrated, and as we can see again now, there is also a third possibility—frustration with the clash between these two democratic procedures can lead to demands for an extra-institutional shortcut—an ethic of unrestrained “decisionism”.1 As discussed below, Weber framed the issue in terms of “caesaristic democracy”, a mode of leadership that was becoming ubiquitous as rhetoric and emotion was needed to mobilize electoral support, even in parliamentary systems. This chapter revisits that theme in the light of the contemporary upsurge in “recall” experiments surveyed in this volume. Their promise is that they might reconcile direct and representative democracy practices. The pitfall to avoid is that the conflict between them could short-circuit democracy altogether. This chapter reviews the promise and pitfalls of recall under three headings. These are “healing” divisions, providing for “buyer’s remorse”, and promoting “losers’ consent”.

1  “Dezisionismus” appeared in Carl Schmitt’s Die Dictatur (1921) referring to rightful deviations from law by personal decision-making in concrete circumstances, where it was presented as an essential component of his anti-parliamentary legal philosophy. But since Schmitt went on to endorse the Third Reich and to praise its 1935 law stripping Germany’s Jews of their legal rights, his ethic can fairly be classified as opening the way to the extrainstitutional and unrestrained use of executive power.

12  ON RECONCILING RECALL WITH REPRESENTATION 

231

12.2   Background Although our focus in this volume is on twenty-first-century experiences, the foundational issues at stake go back to the very earliest days of representative government, and many of the key issues have been debated throughout history (see Chap. 3). Current institutional experiments are mostly very recent, and little studied from either a comparative or a theoretical perspective. Available evidence is very tentative and incomplete. For these reasons it is salutary to include an historical perspective when engaging with the contemporary literature and issues, before concluding with some more prescriptive reflections. Certainly initial experiments with ideas of recall were exceptional, fragile, and only loosely comparable to contemporary practices. Electorates were highly stratified and inegalitarian. Ideals of peaceful representation were abstract and often unrealistic and in any case had to compete with strongly supported rival theories of power, based on heredity, class, divine authority, and the straightforward law of the strongest. Short-term office-­ holding was seldom an altruistic activity, and legal or administrative restraints on abuse of office were at best patchy and incipient. Overall, then, the pre-conditions for orderly and legitimate recall were not yet present, and such experiments as did occur were either short-lived or in other ways highly circumscribed.2 Conditions gradually became more favourable in the nineteenth and twentieth centuries, although universal suffrage did not necessarily eliminate the appeal of authoritarian solutions.3 But the global spread of “recall” practices is mainly a feature of recent decades. Even then it has mostly taken place “below the radar” (Whitehead 2018; see also Welp 2018) and its dynamism is only now 2  Consider, for example, Sanz Yague’s careful reconstruction of the complex relation between political representation (biennial election of officeholders), direct participation (a holdover from medieval asambleismo), and public authority in the Castilian provincial city of Soria in the eighteenth century. While under Bourbon absolutism and local tradition the constraints were severe, it was occasionally possible for the represented to help cut short the tenure of their representatives (Sanz Yague 2018). 3  “The stronger the power of democratic feeling, the more certain is the awareness that democracy is something other than a registration system for secret ballots. Compared to a democracy that is direct…parliament appears an artificial machinery, produced by liberal reasoning, while dictatorial and Caesaristic methods not only can produce the acclamation of the people but can also be a direct expression of democratic substance and power”. Shades of Carl Schmitt (1988 translation: 16–17) live on today, for example, in Turkey, and such ideas still continue to offer a significant global challenge to liberal democracy.

232 

L. WHITEHEAD

becoming impossible to ignore. Moreover, although the factors that once restricted its diffusion have lost some of their edge, they still remain operative in many parts of the contemporary world. And if the current upsurge in recall experiments is badly conducted, the practice could easily be discredited or subjected to new lines of critical resistance. All modern variants of recall presuppose the prior existence of robust structures of constitutional order, including neutral accountability and rule of law institutions, electoral integrity practices, and independent news media. These background conditions for the establishment of a conventional “electoral democracy” are all the more necessary if those elected to office are to be effectively monitored (and potentially defenestrated) by their voters between regularly scheduled elections. Such representative institutions reflect a complex worldview that cannot be taken as given. Its assumptions require scrutiny from a distance. But most modern political science has tended to downplay the foundational issues in representation, turning its study mainly into issues of principal-­agent control and functional institutional design, and so neglecting the contingent, identitarian, emotional, and foundationally non-­ consensual aspects.4 Modern liberal theory is also abstract and decontextualized and so cannot cope with sociological change and diversity or technologically shifting means of political communication and their unsettling/disruptive dynamics. In reality there can be no certainty either that rulers can be held to public account (a mechanistic illusion), let alone that voter choices will be stable and consistent (convergent under the incentives of democratic design). A broad comparative account of representation also needs to give full weight all the diverse allegiances that can still intervene (class, patriotic, cosmopolitan, religious, regional, linguistic, deferential, generational, ideological, as well as atomized, individual, etc.).

4  Much mainstream political science works within a more precise theory of representation than that presented here. For example, the Downs theorem specifies that a majority voting system will select the outcome most preferred by the median voter. While it might be possible to shoehorn some recall experiences into that framework, the theorem rests on very restrictive assumptions that fail to capture the dynamics of our case material. In essence it pictures politics as a duopoly market. Voters choose between Coke and Pepsi, with both brands striving to attract the undecided consumer. A few party systems may once have mimicked this model for a certain part of the late twentieth century, but it was always a very exceptional case, and hardly applies anywhere today. It is no longer a good fit even for the US fizzy drinks market.

12  ON RECONCILING RECALL WITH REPRESENTATION 

233

So in order to ground the presuppositions framing the recall debate, a brief summary of the basic theory of representative government underlying governance in all the world’s diverse electoral democracies (and so embracing both parliamentarism and presidentialism) is required here.

12.3   The Theory of Representation At the most abstract level, representative government is a system in which public authority is divided into certain fixed and interdependent parcels of legal competence to be exercised by elected officeholders whose powers (as “representatives” of their constituents) are specified in advance and limited in duration. Such officeholders are required (usually through on oath of office) only to operate within the limits of their delegated authority, and only so long as provided under their fixed electoral mandates, after which they must either stand down or (when allowed) compete with challengers to their office for a further fixed term. They are subject not only to a temporal limitation in office but also to the prevailing “rule of law” and increasingly to other norms and expectations concerning accountability for their decisions and propriety in the exercise of their public duties (codes of conduct, oversight processes, etc.) (Keane 2009: xxvii). Recall is a possible additional provision within this overall system of representation. It follows the logic specified above, referring only to specific powers and predetermined mandates. It provides for the addition (under carefully controlled and legally specified conditions) of a further possible electoral process, to be triggered at some interval before the termination of the officeholder’s regular mandate. A soundly conducted process of recall will require a collectively recognized timetable, an agreed triggering procedure and campaign format, a legally prescribed voting and counting process with clear criteria for success or failure of the recall, and a predetermined formula for replacing the officeholder where the electorate so decides.5 Although recall is sometimes allowed without the initiators being required to give cause, more commonly some public reason or 5  In June 2019 the European Commission for Democracy through Law (The Venice Commission) adopted a Report on the Recall of Mayors and Local Elected Representatives that provides a thorough review of the currently available evidence and a detailed menu of recommendations for good practice. Their key conclusion is that, where allowed, recall “should be regulated very carefully and only used as an exceptional tool, as a complement to other democratic methods which are available in a representative system—including other direct and participatory democracy instruments, if available” (para 74).

234 

L. WHITEHEAD

justification (either formal or informal) for interrupting the normal term of the incumbent is required. Understood in this way, modern “recall” procedures are optional second-­order additions to a system of representative government, which take its basic provisions as given. Indeed the current spread of recall experiments is an outgrowth of the spread of representative democracy. Yet this representative system of government is a historical late-comer as compared to the major alternatives (such as monarchism, oligarchy, theocracy, imperialism, and military rule) that have competed for ascendancy across the millennia. And even though over recent decades the complex structures of divided power and mutual restraint that characterize the theory of representation have gained ground in the contest for public allegiance, this system is often more dominant on paper than as an actively hegemonic practice in the contemporary world. To graft on “recall” (or other direct democracy innovations such as referendum or participatory budgeting) is therefore somewhat of a risky gamble. At a minimum it should only be done for good cause and with due precautions. Otherwise it could overload the representative system to the point of tilting the balance of opinion back towards more primitive forms of governmentality.

12.4   The Still Limited Reach of Representative Government After all, until the late twentieth century such “electoral democracies” were exceptions rather than the norm. Although restricted forms of representative government had a longer history, it was not until after the Second World War that universal suffrage became the standard template for electoral representation in most nations. In fact the number of countries classified by Freedom House as minimal electoral democracies (using rather undemanding criteria that fall well short of the theoretical ideal) peaked in 2015 at 125 out of the 195 sovereign states under consideration.6 So even on this generously inclusive basis over one third of all independent political regimes remain outside the scope of representative government. Of these 125 only 89 were classified as “free”, meaning that on a definition that more closely approximates to a fully functioning representative system less than half the nations of the world would qualify. And even that list 6  Only 69 out of 165 countries were counted as “electoral democracies” in 1990, when Freedom House began this series.

12  ON RECONCILING RECALL WITH REPRESENTATION 

235

includes many countries where the basic presuppositions of representative theory are only partially operative and not fully trusted by the electorate (e.g. Brazil, Bulgaria, India, Indonesia, Mexico, South Africa). Indeed, the 2019 Freedom in the World report from Freedom House even contains a separate analysis of the USA tracing the deterioration of its democracy across a range of indicators over a number of years. In short, the growing incidence of “recall” experiments in the democratic world is occurring not in a setting of overflowing confidence in the ascendancy of representative principles of government, but rather in a global context of democratic underperformance and citizen dissatisfaction with the competence and trustworthiness of elected officeholders. “Self-serving” and “corrupt” are the bywords used to disqualify politicians in all too many contemporary democracies. If that is the diagnosis, “recall” is an apparently obvious remedy. Such sweeping generalizations about voter attitudes must always be cross-checked against specific contexts and situations. The scholarly literature on this is too vast for exploration here.7 Suffice it to say that there are mounting recent indicators not just of party dealignments but of much more far-reaching shifts in public attitudes towards fundamental pillars of the representative system. This applies across a remarkably wide and diverse array of contexts. The explosive spread of digital communications may be a big factor here, but disenchantment with the “political class” as a whole needs to be understood at least partly as a response to their own dysfunctions as well, together with the still suppurating aftermath of the 2018 world financial crisis. Even if this political turbulence and attitudinal volatility does give rise to any serious and permanent impairment of representational democracy, but it is enough of a shock to precipitate

7  One influential starting point is Mair (2013), who highlights the “passing of the age of party democracy” in Western Europe and in particular the evidence that many well-established parties have “lost their representational integrity” and are no longer identified with any particular social bases or policy orientations. Beyond party politics, digital technology may be contributing to more general shift in popular understandings of representation, promoting more individualism and single issue networking, shorter time horizons, less reliance on established authorities (“liquid modernity”). For an overview of this theory-oriented approach, see Bluhdörn and Butzlaff (2019). For a recent survey-based global review, see Inglehart (2018)—notably Chap. 9 on the rise of Trump. The Varieties of Democracy Project provides a further wealth of detailed expert and public opinion survey evidence from across the world.

236 

L. WHITEHEAD

s­ oul-­searching and the consideration of hitherto marginal remedial actions in a large number of countries.8 Notably, in the diverse democracies covered by this volume, there has been a growing political interest in the potential of recall mechanisms to head off voter dissatisfaction (perhaps through acting as a “safety valve”). That makes it timely to review and compare these recent experiences and to reflect on both how the current fashion for “recall” may unfold and what lessons can be drawn concerning the restoration of confidence in broadly “democratic” forms of representational politics. This allusion to “broadly” democratic, rather than “narrowly” electoral, politics also requires a brief comment. As we have seen, the historical of “recall” is closely enmeshed with the emergence of electoralism as a way to parcel out political authority. But twenty-first-century experiments in recall should not be assessed solely as add-ons to routine voting procedures. What citizens increasingly expect is not just that their officeholders serve out their time with honesty and dedication. Many also require direct public engagement with the laws and regulations that shape their lives. They want a “voice” in decision-making, they expect their representatives to listen and respond to their concerns between elections. John Keane (2019) has coined the term “monitory democracy” in an attempt to capture this development: “the unending quest, in the name of equality, to humble the powerful, to prevent their abuse of power by means of various platforms to render publicly accountable not only elected governments but also state institutions, corporations, and other organizations, even in cross-border settings. The age of monitory democracy features the birth of many scores of watchdog organizations that never existed before in the history of democracy, including integrity bodies, citizens’ juries, truth and reconciliation commissions, participatory budgeting, intergenerational panels, citizen science networks, and bio-regional assemblies.”9 This is the 8  Recall is just one out of a large variety of direct democracy innovations that have attracted growing interest among political entrepreneurs over the past generation. Standard political science used to promote a very sharp demarcation between direct and representative forms of democracy, but that was always too binary, as comparative historical institutionalism confirms. Petitions, referendums, participatory budgets, “people’s defenders”, and “right to information” laws are among the other citizen empowering devices that have flourished in response to voter dissatisfaction with the conduct of their professional politicians. Cf Welp and Whitehead (eds) (2011) and Ruth et al. (2017). 9  Needless to add, such monitory arrangements are equally available to nationalist, antiimmigrant, and security-minded citizens.

12  ON RECONCILING RECALL WITH REPRESENTATION 

237

“broadly” democratic context, reflecting currently flourishing ideas about rights and active citizenship, transparency and accountability, within which the appetite for recall experiments has grown. On the one hand, it is a climate of opinion that exerts new pressures and demands on elected officeholders, thereby increasing the appeal of new methods to challenge incumbents during the course of their mandates. At the same time, and on the other hand, this is also a context in which entrepreneurial politicians can hope to reinforce their popular support and thus evade the risks of recall (and party deselection) by taking on other aspects of the monitory democracy agenda.

12.5   Current Recall Experiences In the course of compiling this volume, an ever-increasing range of recall proposals and initiatives has come to our attention. This momentum is likely to continue for some time yet. Indeed, the fashion for “recall” could be on the threshold of breaking through into the awareness of even the most professionally secluded political science scholars. So even at this early stage there is a need to evaluate this ongoing surge of untested innovations. And if such experiments are on the rise (like it or not), there is a need to start considering what approaches best serve the objective of reconciling recall with established practices of democratic political representation. Chapter 2 in this volume reviewed the current incidence of “recall” provisions and outlined the complex and diverse procedural rules that structure this field. It also sketched some early inferences about which options were more likely to work smoothly within the existing system of representation and which ran greater risks of incompatibility and unintended consequences. Institutional design is clearly important in this area, and that chapter adds value because to date there is no cross-national inventory of best practices or errors to avoid. But a full evaluation of these experiences needs to look beyond such technical questions of rules and procedure. Contexts differ greatly, and timing can also affect the prospects for successful innovation. For example, the contrast between the Californian gubernatorial recall of 2003 (carried with 55% in favour) and the Venezuelan presidential recall of 2004 (58% against recall) was much more about the political climate and context than the procedural rules. That is what explains why in the aftermath Californian democracy emerged unscathed, whereas the Venezuelan system went into

238 

L. WHITEHEAD

a downward spiral. Timing effects are also important. Thus, as Chap. 8 argues, the main significance of the 2013 vote to recall the Mayor of Warsaw (which failed through insufficient turnout rather than support for the incumbent) was not her survival in office, but what it foreshadowed and assisted in the national election defeat of her party in 2015.10 Beyond procedures and contexts, wider normative and theoretical issues also come into play. Three general headings merit particular attention here: healing divisions, buyer’s remorse, and losers’ consent. None of these can be satisfactorily addressed from a purely procedural or institutional design standpoint of the kind peddled by most mainstream political science. Deeper questions about the nature and prospects of representative government require an approach that goes beyond narrow horizons of institutional design, principal-agent theory, political career incentives, and one shot vote maximization (although all these topics are worthy of inclusion within the larger framework). Healing Effects? Can recall procedures help to “heal political divisions” in society, or are they more likely to intensify zero sum conflict and polarization? A more precise formulation of this issue might be, under what circumstances is the former result more likely than the latter? The skeleton of an answer would need to specify not just what precise variant of “recall” is involved but also how the prevailing system of representation relates to the patterns of conflict under consideration in each case. For example, in contemporary Spain the sharpest source of division concerns the impasse between a unified national system of democratic representation and a Catalan regional legislature with a separatist majority of representatives, also elected in accordance with democratic practices. In this context the introduction of recall or other MDD procedures either at the Spanish or at the Catalan level of representation would assuredly fall victim to the underlying political polarization, no matter how carefully it was designed. In contrast, a wide variety of MDD procedures are clearly viable in Switzerland, given its decentralized and citizen-friendly institutional ecology, with more scope to r­ econcile

10  Spivak (2016) reviews the timing and broader political significance of several controversial recall procedures, including Warsaw.

12  ON RECONCILING RECALL WITH REPRESENTATION 

239

opposing tendencies, and far less risk of escalating conflict.11 A more challenging and finely balanced test of these effects would be if some form of recall was introduced in contemporary France, as discussed in Chap. 4 by Clara Egger and Raul Magni-Berton. This volume provides some early and provisional comparative evidence bearing on the potential of existing recall procedures to contribute to political “healing”.12 Contemporary Japan is not afflicted by irreconcilable political divisions to the same extent as other countries we examined. Accordingly, Chap. 6 by Mitsuhiko Okamoto and Uwe Serdült focuses on vertical accountability rather than the “healing” potential of recall. In this case neither the party system nor the (non-binding) referendum mechanisms available suffice to manage a range of local policy issues of importance to sections of the municipal electorate. Through the recall of local authorities they may seek redress for this omission. The device is used quite sparingly and does not provoke escalating polarization. In this case it seems fair to regard “recall” as a minor but useful “safety valve” helping to balance an essentially still legitimate system of party-based representation. Likewise Chap. 7 by Brigitte Geißel and Stefan Jung concerns a still fairly consensual system of party representation and also deals with recall at the municipal level. Although there has been increasing political conflict in recent years, with the rise of “outsider” challengers, especially in some eastern lander, this is neither caused by nor much reflected in the recall processes considered in the chapter. There have been many of these, concentrated in a few lander, with few or none in most of the Federal Republic. The frequent recalls in Brandenburg, for example, can be seen as a local phenomenon that requires some institutional fine-tuning, rather than as a source of general division. In this case too, “recall” can be viewed so far as 11  The Swiss economist Bruno Frey (1992) may have been misled by his country’s atypically positive experience of direct democracy when he argued (on the basis of a statistical analysis of over 600 referendums) that they may increase collective happiness and improve the efficiency of government. Certainly from a British perspective, at least one striking exception to this rule comes to mind. 12  Since “healing” is a powerful normative metaphor, its application to political matters should be treated with caution. In 54 BC Cicero defended the popular vote as a pacifier of discontent. But to call that “healing” would disguise his conservative intent: “thanks to my law, the appearance of liberty is given to the people, the authority of the aristocracy is retained, and the cause of quarrelling is removed” (Cicero 2008 edn: 167 (translating Laws, Book Three, 1979, p. 40)).

240 

L. WHITEHEAD

a rather marginal and localized phenomenon, not impacts severely on the overall system of representation. It probably provides more of a “safety valve” than a source of polarization. Turning to the USA, as analysed in Chap. 5 by Joshua Spivak, we again encounter a national system of party representation that may have become much more polarized and conflictual over past decades, but not because of recall procedures that operate only at state and local level, and that have functioned intermittently over a long period of time, mostly damping down confrontations rather than aggravating them. Finally, it is worth considering the early evidence beginning to accumulate from the UK.  Here the severe intensification of party conflict and societal polarization (including real risks of defection from the union in Scotland and Northern Ireland) can be traced to the 2016 Brexit referendum. The very restrictive recall law that was adopted in 2015 in response to the MPs’ expenses scandal certainly represented an attempt to “heal” a breach of trust that had arisen between the electorate and its parliamentary representatives. So far only three petitions for recall have been launched and only two MPs have lost their mandates. This is a modest but cautionary outcome. It cannot serve to counteract the intense divisions and distrust arising from the larger political crisis, but in a small way it may have a “soothing” and “error correcting” effect. There is also a second form of recall now underway in the UK candidate deselection (or reselection) by party militants, as discussed in the chapter by Lea Ypi and Jonathan White. In this case the Brexit effect does penetrate the recall process, with the result that initial outcomes may prove conflict-intensifying. Whether or not the positive features of deselection highlighted in Chap. 11 turn out to prevail in the longer run is a matter for future research. All in all, then, our evidence suggests that, at least in these four established democracies, the “healing” effects of recall are positive, but minor and local. However, in the more fragile and conflictual electoral democracies of post-communist Europe (Chap. 8) and Latin America (Chap. 9), a somewhat different picture is emerging. In these settings recall is not always confined to municipal and provincial levels of authority. When the mayor of a capital city (Chisinau, Lima, Mexico City, Warsaw) can be recalled, then major national political interests become engaged—indeed such officeholders are often seen as the effective “leaders of the opposition” to incumbent presidents. The polarizing effect is still more likely when the president himself is subject to recall (as provided for in Bolivia and Venezuela and now also in Mexico). On the one hand, recall processes

12  ON RECONCILING RECALL WITH REPRESENTATION 

241

of this kind will be more consequential for the balance and legitimacy of the representative system as a whole. On the other, the risks of destabilization through poorly designed or badly implemented recall procedures could be considerably more serious, while the “healing” and “safety valve” potential remains quite limited. The evidence on this is still very tentative, and the counter-factual (how much conflict would be avoided by doing without recall?) is difficult to assess. But in settings where representative institutions are not very securely implanted, the destabilizing potential of an over-reach of recall experiments is all too apparent. Buyer’s Remorse Are there recall procedures that limit the scope for destabilizing over-­ reach? This is partly a question of design, but also touches on questions of legitimacy, and the importance of deliberation rather than mere “choice”. This broader perspective suggests precautions that have not received much consideration until now. The idea would be to mitigate the danger of voters acting on impulse in a manner that did not reflect their long-term considered preferences. One model for consideration is the “cooling off” period that is customarily provided to protect consumers from making financial commitments on the spur of the moment and under the pressure of a salesman, but that they later regret. It is now standard in some countries for commercial law to require that the vendor of some products to offer the purchaser the right to a change of mind without penalty within a certain period after the sale. A similar provision would be simple to add to most recall processes. For example, when a certain threshold of voter signatures is required to trigger a new vote, the signatory could under current conditions be instructed to provide an email address and/or a mobile phone number. Within, say, two weeks of receiving the petition the electoral authority could inform all signatories that the threshold had been passed, and a fresh election will be triggered, unless within a further two weeks a specified number of signatories withdrew their support. The provision could even be strengthened by allowing the officeholder under threat to transmit an appeal for relief and a brief message of self-defence. If, as in Japan, the real motive of the signatory is to secure some policy reform, the officeholder could have the opportunity to address that issue and so save head off a recall. Similar provisions could apply to deselection procedures within political parties (perhaps afforced by rules against entryism—e.g. requiring at least one year of

242 

L. WHITEHEAD

prior party membership). As with all such procedures, unintended consequences would have to be evaluated. Would more signatories come forward if they thought they were just scaring an incumbent, rather than forcing a full recall? The practicalities need scrutiny in each system, but there is merit in upholding the principle of delaying a recall decision until the full situation is clear. Losers’ Consent Metaphors of “healing”, “safety valves”, and “cooling off” all draw attention to the underlying point that any democratic decision involves not just a choice (with winners and losers) but also a coming together to make the best of whatever the choice may be (a “legitimation” of the outcome making it tolerable to the opposition).13 Institutional design can incentivize or deter such responses, but loser consent also derives from the operation of a broader “ecology” of supportive norms and practices. In addition to a “calculus” there is also such a thing as a “culture” of consent. There are instances where direct democracy innovations have been added on to standard representative systems without disturbing this culture or disrupting this ecology. In the Swiss case the overall effect seems to have been positive for social cohesion, and in California and Japan it has long been at least compatible with loser acquiescence. But direct democracy in general, and recall experiments in particular, can also generate escalating conflict between decisions that are said to be “the will of the people” and the standard operations of a representative system. One striking illustration of this has been the UK’s 2016 Brexit referendum, which has not only pitted a one-time irreversible and narrow popular majority on an imprecise question against the considered preferences of two successive parliamentary majorities, but has also revived the drive for 13  The konstructives misstrauensvotum (article 67 of the federal German constitution of 1949) is probably the most celebrated example of a device that induces the opposition to consent to the continuation in office of a Bundeskanzler despite having the votes to remove him, because they lack a majority for his replacement. This “constructive vote of no confidence” was lacking in Weimar, with the result that its parliamentary government was continually destabilized (although it did exist and functioned well in the Free State of Prussia between 1919 and 1932). Postwar German oppositions have accepted the procedure because it provides the benefit of governmental continuity, even though it has also meant that all but one of their challenges have lost. Albania, Belgium, Hungary, Israel, Lesotho, Poland, and Slovenia have all followed this German example.

12  ON RECONCILING RECALL WITH REPRESENTATION 

243

independence in Scotland. All 32 councils and 62% of the Scottish electorate voted to remain in the EU, and the ruling SNP has no intention of consenting to that nation’s forced departure from Europe just because the larger English portion of the UK electorate chose to do so. Indeed, Scottish nationalists can reasonably claim when they argue that their campaign for independence in the 2014 referendum was defeated (by a 55/45 divide) in large part because of the assertion that a sovereign Scotland could not be a member of the EU. Losers’ consent is also a key issue for recall experiments. For example, it is questionably democratic to recall an officeholder on the basis of fewer votes than elected him/her in the first place. The targeted official and supporters may also reasonably refuse consent if no justification is provided for the recall demand, or if they are allowed no adequate opportunity to reply to such charges. Careful design of the rules for recall can reduce (though not eliminate) such objections. A further possible precaution against unjustified campaigns can be found in Colombia and Peru: after the failure of one recall attempt, the officeholder cannot be challenged a second time before completion of the mandate. The Venice Commission also advocates making a clear distinction between recall as a political act and destitution or loss of office through a judicial ruling. Taking care on such matters will improve the chances of securing losers’ consent. In general, losers’ consent is partly a function of the scale and character of the loss in question. The recall of a president is far more difficult for the ruling party to accept than the recall of an individual mayor or a single legislator. Who loses is also relevant—an individual career loss is not necessarily of great concern to a large party or a broad coalition. A losing party can perhaps be induced to accept an adverse outcome if the compensations and side payments are made sufficiently attractive (e.g. accept the recall of this mayor and gain some budgetary or nomination benefit in return). A “one shot” loss may be offset by the prospect of later electoral rewards. But all such “calculus” considerations also need to be situated in their broader political culture context. For example, in a basically consensual political system there will be real costs attached to appearing as a “sore loser”, whereas in polarized settings the bigger risk may be to seem conciliatory. Beyond merely cost-benefit calculations, a recall verdict may be judged an affront to dignity. The loss will have to be explained to supporters who may find it hard to accept given their ideational commitments. In short, durable losers’ consent requires a climate of trust and mutual tolerance, not just a direct pay-off.

244 

L. WHITEHEAD

12.6   Recall and Representative Government—or “Caesaristic Democracy”? Recall procedures presuppose the ascendancy of a stable structure of representative government. They are spreading in the twenty-first century, although electoral democracies are no longer on the rise, and there is growing evidence of widespread “backsliding”. In any case, most really existing representative systems fall well short of meeting the full requirements of the theoretical model, and many are experiencing widespread citizen dissatisfaction. One driver for the spread of recall and other direct democracy innovations is precisely this popular discontent with traditional practices of representation. If it is to hold its own in the world (let alone to gain new ground, or to prevail against the alternatives), representative government will need to become more socially responsive, better adapted to modern technology, and in general more innovative. Recall might flourish as a secondary component of that revival. But none of that is a given. There are strong and well-supported alternatives. Some dispense with representation altogether, but probably the most serious challengers are those that combine some superficial appearance of authority delegation from the people with a fundamental reliance on power concentration justified by the imperatives of “decisionism”. During the nineteenth century this was discussed under the heading of “bonapartism”, but in 1918 Max Weber recast it as “caesaristic democracy”, and Vallenilla Lanz then introduced this concept into Venezuela (with reference to then incumbent President Gomez). For Weber the crucial feature was the use of the plebiscitary power—“no ordinary vote or election, but a declaration of “faith” in the vocation for leadership of the person who claims this acclamation for himself”. Karl Loewenstein’s commentary on this continues as follows: “the core of the whole question is whether there is a legal way for removing a political leader who has come to power by plebiscitary means, once he has lost the confidence of the masses, and thus what Weber calls his “test” of legitimacy. For the essence of Caesarism consists in the impossibility of putting a legal end to absolute rule…. The Caesar cannot be deposed by the constitution because he himself is the constitution…. the people no longer have the means of showing their loss of faith in him. They cannot get rid of him without violence…the reign of a Caesaristic dictator can be brought to an end only by his death, by a revolution which is successful because the armed forces desert him, or by defeat in war” (Loewenstein 1966, cfr.

12  ON RECONCILING RECALL WITH REPRESENTATION 

245

Casper 2007, and also Valenilla Lanz 1919). Despite its democratic origins, caesarism tends to degenerate into lifelong personalist rule. Contemporary studies of such tendencies depersonalize the phenomenon. It can be discussed under the rubric of “delegative democracy”. Toward the end of the Cold War many new democracies came into existence, with more or less regular and competitive elections, but otherwise quite weak political institutions (partly overshadowed by previous authoritarian legacies). The parallels with Weimar are evident. In Guillermo O’Donnell’s formulation, “the deep social and economic crisis that most of these countries inherited from their authoritarian predecessors reinforces certain practices and conceptions about the proper exercise of political authority that lead in the direction of delegative, not representative democracy” (my italics). He proposed this as a new species of democracy (although today it is proving more contagious than his species metaphor indicated). “Delegative democracies rest on the premise that whoever wins election to the presidency is thereby entitled to govern as he or she sees fit, constrained only by the hard facts of existing power relations and by a constitutionally limited term of office… This kind of rule has been analysed as a chapter in the study of authoritarianism, under such names as caesarism, bonapartism, caudillismo, populism, and the like. But it should also be seen as a peculiar type of democracy that overlaps with and differs from those authoritarian forms” (O’Donnell 1994: 56–59). Since this was written, the literature on regime types has introduced many further refinements (partly structured around the notion of “hybrid” regimes), while in the real world the delegative alternative to representative democracy has erupted in some unexpected places (possibly even including Washington, DC). Even elected presidents can, in principle, be subject to recall procedures. Recall could be a direct mechanism for correcting the unaccountable personalism of a delegative ruler. This term refers to a chief executive elected on a genuine democratic mandate, but who the uses the resulting political capital to build a majoritarian electoral base. The delegative leader then proceeds to flout the “checks and balances” embedded in the separation of powers system and protected by “rule of law” court supervision. Citizens seeking to restrain such excesses could be given access to a recall process to remove an abusive president before the completion of a full term. This kind of recall has parallels with the “impeachment” provisions discussed (and viewed as inferior to recall) in the preceding chapter by Pérez Liñán. Whereas impeachment procedures involve a semi-judicial

246 

L. WHITEHEAD

element, and rest on the idea of “horizontal” controls on the executive branch exerted by the legislative and judicial branches, recall is a “vertical” form of political accountability operated directly by the electorate from below. It is also a two-edged instrument. Although it might sanction excesses of personalist rule, it could also serve to actually legitimize them. This second possibility arises if a recall process is triggered but then the challenged incumbent survives in office, and proceeds to treat the outcome as a mandate to further deinstitutionalize the democratic regime. We have seen that recall is a peaceful and institutionally respectful method that presupposes the existence of an electoral democracy.14 But it can also be embraced by presidents (such as Hugo Chavez) who then proceed to capitalize on the reinforced mandate arising from the defeat of a recall bid in order to concentrate power and “eternizarse” in office for life. Now that Mexico has followed the precedent set by Mexico City and introduced a presidential recall (grafted on to a constitutional system with a sacralized electoral calendar and a strong tradition of “no re-election” six yearly executive mandates), we can follow that experiment to observe whether this method can be used to improve representative government rather than to subvert it.  A parallel experiment is underway in Taiwan. There is plenty of scope for electoral democracy to innovate and display resilience around the world in the twenty-first century. But the darker alternative of caesaristic democracy and its dysfunctions is also available if that fails.

References Blühdorn, I., & Butzlaff, F. (2019). Rethinking Populism: Peak Democracy, Liquid Identity, and the Performance of Sovereignty. European Journal of Social Theory, 22(2), 191–211. Casper, G. (2007). Caesarism in Democratic Politics: Reflections on Max Weber. SSRN Paper, Stanford. 14  As noted above a bonapartist or caesaristic ruler can only be removed by extra-constitutional, probably violent, means. By contrast, in the delegative democracy model the personalist ruler is still ultimately required either to step down at the end of an electoral mandate or (if allowed) to win re-election in a genuine contest. But in practice there is a grey area between the two categories, where recall might also be found. Thus, in a “hybrid” regime context there could be recall without full electoral integrity; while at the other end of the authoritarian spectrum, even a lifetime dictator might resort to recall as an escape mechanism when all other routes to personal survival were exhausted.

12  ON RECONCILING RECALL WITH REPRESENTATION 

247

Cicero. (2008). The Republic and the Laws Oxford World Classics. Oxford University Press. Frey, B. (1992). Efficiency and Democratic Political Organization; The Case for the Referendum. Journal of Public Policy, 12(3), 209–222. Inglehart, R. (2018). Cultural Evolution; People’s Motivations Are Changing, and Reshaping the World. Cambridge University Press. Keane, J. (2009). The Life and Death of Democracy. London: Simon and Shuster. Keane, J. (2019). The Age of Monitory Democracy and the Greening of Politics. Taiwan Journal of Democracy, 15(1), 9. Loewenstein, K. (1966). Max Weber’s Political Ideas in the Perspective of Our Time. Amherst: Massachusetts University Press. Mair, P. (2013). Ruling the Void: The Hollowing of Western Democracy. London: Verso. O’Donnell, G. (1994). Delegative Democracy. Journal of Democracy, 5(1), 55–69. Ruth, S. P., Welp, Y., & Whitehead, L. (Eds.). (2017). Let the People Rule? Direct Democracy in the Twenty-First Century. Colchester: ECPR Press. Sanz Yague, A. I. (2018). Representacion Politica y Paticipacion Directa: El “policentrismo” politico de Soria y la supervivencia del Comun en el Siglo XVIII (pp. 178/90). Madrid: Ediciones Polifemo. Schmitt, C. (1921). Die Diktatur. Munich: Dunker & Humblot. Schmitt, C. (1988 Translation of 1926 Original). The Crisis of Parliamentary Democracy. Cambridge, MA: MIT Press. Spivak, J. (2016, June 14). Venezuela’s Risky Recall? Why a New Vote in Venezuela Could Backfire. Foreign Affairs. Valenilla Lanz, L. (1919). Caesarismo Democratico (1990 ed.). Caracas: Monte Avila. Welp, Y. (2018). Recall Referendum Around the World: Origins, Institutional Designs and Current Debates. In L. Morel & M. Qvortrup (Eds.), Compendium on Direct Democracy. Routledge. Welp, Y., & Whitehead, L. (Eds.). (2011). Caleidoscopio de la Innovacion Democratica en America Latina. Mexico City: FLACSO. Whitehead, L. (2018, October). The Recall of Elected Officeholders: The Growing Incidence of a Venerable, but Overlooked, Democratic Institution. Democratization, 25(8), 1341–1357.

Index1

A Abstention, 22 B Boycott strategy, 153 C Campaigns, 11, 22, 61, 81, 86, 87, 105, 147, 149, 152–154, 174, 180, 181, 186–188, 190, 196, 197, 233, 243 Commune de Paris, 53 Condorcet, 51 Confidence vote, 23 Consensus democracy/consensual democracy, 126, 126n8, 128, 130, 134, 135

Corruption, v, 34, 43, 56, 57, 81, 84, 89, 152, 153, 162, 164, 172, 173, 197, 198, 208, 213, 214, 218, 219, 221, 229 Courts, judiciary, 146 D Decentralization, 15 Deselection, 5, 179–198, 237, 240, 241 Direct recall, 13, 15, 99, 117, 119–136, 122n3, 155, 161, 162, 173, 174, 176, 205 Divided government, 56, 58, 165 E Electoral majorities, 6, 202 Electoral system, 49, 54, 58, 59, 67, 68, 70, 175

 Note: Page numbers followed by ‘n’ refer to notes.

1

© The Author(s) 2020 Y. Welp, L. Whitehead (eds.), The Politics of Recall Elections, https://doi.org/10.1007/978-3-030-37610-9

249

250 

INDEX

F Federalists, 218 anti-federalists, 12, 33, 78 Financing recall, 87 First Past the Post system, 74 France Insoumise, 11, 60, 64, 68, 182 French Revolution, 4, 50–52 G Gramsci, Antonio, 4, 14, 36, 42–45 H Horizontal/vertical accountability, 95–115, 203, 205, 206, 212, 215, 223, 239 I Impeachment, 3, 5, 6, 80, 144, 145, 161, 201–225, 245 Imperative mandate, 3, 4, 11, 44, 49–70, 78, 184 Indirect recall, 13, 117, 119, 124n4, 125, 125n6, 129, 132, 161, 162, 168, 205, 206, 208n3 Informal recall, 234 L Legislative majorities, 6, 152, 202, 204 Legitimation, 206, 212, 242 Lenin, Vladimir, 4, 14, 36–45 Local recall, 5, 117, 152 Luxemburg, Rosa, 14, 42–45 M Madison, James, 12, 33, 77, 186, 217 Majoritarian system, 206

Mandatory reselection, 180–182, 186–189 Marx, Karl, 4, 36–38, 42, 45, 184, 185 Mechanisms of direct democracy (MDDs), 20, 24, 161, 162, 165, 172–176, 180, 183, 197, 238 Motives for recall, recall reasons, 17 P Parliamentarism, 233 Parliamentary coup, 202, 213, 214, 223 Popular mandate, 204 Popular revolt, revolt, 203, 205, 211, 212 Populists/populism, 24, 34, 35, 45, 60, 64, 104, 151, 186, 245 Presidentialism, 213, 233 Progressive movement, 161 Proudhon, P. J., 53 R Representation, 2, 3, 6, 11, 17, 21, 22, 25, 39–41, 45, 51, 52, 59, 61, 65, 167, 180, 181, 184–190, 196, 197, 229–246 Revolution of 1848, 52 Rousseau, J. J., 32, 51, 184 S Safety valve, 4, 9, 97, 134, 135, 167, 174–176, 236, 239–242 Semi-presidential regimes, 55, 55n3, 207, 208n3, 209n4 Signature collections, 5, 15, 100, 105, 106, 167, 206

 INDEX 

T Theory of representation, 39, 232n4, 233–234 “trustee” vs. “delegate” models of representation, 184 Triggers for recall, 5, 19, 69, 89, 105

W Wide recall system, 176 Y Yellow Vest, 4, 50, 58–67, 70, 173

251