Nonviolent Resistance and Democratic Consolidation 3030393704, 9783030393700

Citation preview

Nonviolent Resistance and Democratic Consolidation Daniel Lambach · Markus Bayer · Felix S. Bethke · Matteo Dressler · Véronique Dudouet

Nonviolent Resistance and Democratic Consolidation

Daniel Lambach · Markus Bayer · Felix S. Bethke · Matteo Dressler · Véronique Dudouet

Nonviolent Resistance and Democratic Consolidation

Daniel Lambach Goethe University Frankfurt Frankfurt, Germany Felix S. Bethke Peace Research Institute Frankfurt Frankfurt, Germany

Markus Bayer Institute of Political Science University of Duisburg-Essen Duisburg, Germany Matteo Dressler Flemish Peace Institute Brussels, Belgium

Véronique Dudouet Conflict Transformation Research Berghof Foundation Berlin Berlin, Germany

Additional materials to this book can be downloaded on https://www.springer. com/gp/book/9783030393700. ISBN 978-3-030-39370-0 ISBN 978-3-030-39371-7 (eBook) https://doi.org/10.1007/978-3-030-39371-7 © 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

Preface

This project, like so many others, is the product of a chance observation. Markus Bayer, while writing his doctoral dissertation on nonviolent movements, noticed that much of the literature he was reading was in agreement that nonviolence is not only a useful strategy for achieving a movement’s goals but also lays the groundwork for long-term societal peace. However, empirical research concentrated heavily on short-term outcomes while mostly neglecting long-term impact. And thus, the idea for this research project was born. We are indebted to a great many persons and institutions who have made the research that went into this monograph possible. The University of Duisburg-Essen (UDE), where Daniel Lambach and Markus Bayer were situated at the time, provided seed funding in 2013 through its Main Research Area Transformation of Contemporary Societies, which allowed us to commission Felix Bethke to conduct the first exploratory tests. Results from these tests were very encouraging and enabled the submission of a fully fledged grant proposal to the German Research Foundation (DFG), which was funded under grant number LA 1847/9-1. The project was conducted at UDE from April 2015 to March 2019, chiefly involving Markus Bayer, Felix Bethke, and Daniel Lambach. Mathieu Rousselin also briefly worked in the project as cover for parental leave, conducting a case study of Mali as an example of unsuccessful democratic consolidation after nonviolent resistance. At UDE, we received support from the team of the Chair of International Relations and Development v

vi

PREFACE

Policy, especially Professor Tobias Debiel, Inge Fischer, and Julia Nachtigall, and from the Institute for Development and Peace. The Institute of Political Science provided office space and further administrative support. We are also greatly indebted to our research assistants Cemal Öztürk, Lena Pohl, Katrin Grätz, Leah Ngaba, and Ibrahim Alhadjiui, who assisted in preparing the grant proposal, helped to organize field trips, wrote minutes of our team meetings, coded data, and also completed the arduous task of transcribing the many hours of interviews that Markus collected during fieldwork. The Berghof Foundation was involved in the project from its inception. Véronique Dudouet had supported the initial grant proposal and contributed to the research project throughout the entire process, including during her research fellowship at the United States Institute of Peace, where discussions on the project findings influenced her contributions to this book. Katrin Planta conducted the very first case study on El Salvador before going on maternity leave; her tasks were taken up by Matteo Dressler, who did all of the remaining work, contributing to the project even after he moved from the Berghof Foundation to the Flemish Peace Institute in Brussels. Markus Bayer conducted field research in Namibia between November 2015 and March 2016 in cooperation with the Namibian Institute for Democracy and the Institute for Public Policy Research. In Namibia, Henning Melber, Naita Hishoono, and Dennis Zaire were not only competent experts but also helpful advisors and opened some closed doors and established important contacts. Field research in Benin was conducted between June and November 2016 through contacts first established via the Gesellschaft für Internationale Zusammenarbeit (GIZ) and the Friedrich Ebert Foundation (FES). We thank Wim Deckers and Toni Kaatz-Dubberke from the GIZ Program for Decentralization and Communal Development and Simon Asoba and Dr. Kalus-Peter Treiydte from FES. The final field research phase in Cape Verde took place between February and May 2017. There, Markus worked with Emiliano Moreno and Omarú Djaló Abreu, who acted as translators and go-betweens and were instrumental in making this visit a success. Peter Meyns, Aristides Lima, and André Corsino Tolentino kindly shared their immense knowledge and their helpful contacts. The desk studies on Chile, El Salvador, and Paraguay benefited from expert reviews by Alberto Martín Alvarez, Claudio Fuentes, and Magdalena López, and interviews in Chile and

PREFACE

vii

Paraguay were conducted by Josefina Abarzúa Varela, Camilo Jose Caballero Ocariz, and Edith Arrúa. Aaron Griffiths provided proofreading services. We presented interim results of our research at various conferences and thank audiences, participants and discussants in these events for their feedback. We are also grateful to the International Center for Nonviolent Conflict for giving us the opportunity to present some of our results in an entry to the Minds of the Movement blog. Also, we’d like to thank Aries Arugay, Janet Cherry, Christopher Clapham, Anita Gohdes, Carrie Manning, Henning Melber, Peter Meyns, Jonathan Pinckney, Janjira Sombatpoonsiri, Johannes Vüllers, Nils Weidmann, and Stephen Zunes for comments on various pieces of writing and for advice regarding the case studies. We are also grateful to Palgrave Macmillan for giving us the opportunity to publish this monograph with them. Anca Pusca, the Senior Editor for International Relations and Security Studies, has shown interest in this project for years. Katelyn Zingg, her editorial assistant, has made the publication process as smooth as possible. Frankfurt, Germany Duisburg, Germany Frankfurt, Germany Brussels, Belgium Berlin, Germany

Daniel Lambach Markus Bayer Felix S. Bethke Matteo Dressler Véronique Dudouet

Contents

1

1

Introduction

2

Theory

15

3

Statistical Analysis

41

4

Mechanisms

63

5

Inching Towards Theory

133

6

Conclusion

175

Index

189

ix

List of Figures

Fig. Fig. Fig. Fig. Fig. Fig.

2.1 3.1 3.2 3.3 3.4 4.1

Fig. 4.2 Fig. 4.3 Fig. 4.4 Fig. 4.5

Horizontal and vertical relations of political regimes Kaplan-Meier survivor functions Peaceful turnovers of power for regimes without NVR Peaceful turnovers of power for regimes with NVR Average levels of democracy before and after transition Legislative constraints of the executive for regimes with and without NVR-induced transition Freedom of association and expression for regimes with and without NVR-induced transition Political environment for CSOs for regimes with and without NVR transition Probability of coups after transition Success rate of coups

21 49 51 51 54 71 73 75 108 109

xi

CHAPTER 1

Introduction

A Story of Three African Democracies In 1989, the small West African country of Benin entered its 18th year of autocratic rule by the socialist single-party regime of President Mathieu Kérékou. The country had experienced a number of military coups in the 1960s and 1970s, including the one that brought Kérékou to power in 1972. Benin was then, as now, very poor—it has been on the United Nations’ list of Least Developed Countries since the list was first published in 1971. According to World Bank data, Benin’s literacy rate in 1992 (the closest year for which data was available) was just 27.2% of people aged 15 and above, compared to an average of 53.4% in sub-Saharan Africa. In short, Benin was just about the most unlikely place for democracy to emerge and take root. But that is precisely what happened. Resistance to Kérékou’s socialist one-party regime germinated in the mid-1980s, mainly among student groups and university teachers. With the economic situation deteriorating and wages being paid irregularly, more and more of the urban population joined the ranks of the opposition and pushed Kérékou to liberalize the political system. Restrictions on the press and trade unions were lifted in 1988, but protests continued unabated. The year 1989 saw huge nonviolent demonstrations and strikes. By December, Kérékou was left with no option but to announce the end of Marxism–Leninism as the state ideology and to call for the appointment of a National Conference. The

© The Author(s) 2020 D. Lambach et al., Nonviolent Resistance and Democratic Consolidation, https://doi.org/10.1007/978-3-030-39371-7_1

1

2

D. LAMBACH ET AL.

National Conference worked out a new constitution, installed a provisional government, set out a timetable for democratic elections and paved the way for Benin’s ‘renouveau démocratique’—its democratic renewal. Kérékou was voted out of office in 1991, but the first extraordinary event of Benin’s nascent democracy took place in 1996. The Beninese voted out the first democratic government of Nicéphore Soglo in favour of Kérékou. Soglo’s economic policies had been met with widespread disapproval and he alienated many voters when he tried to pass the 1994 budget via executive decree—a move that was blocked by the constitutional court. From 1996, regular and peaceful turnovers of power were to become a feature of Benin’s democracy. The presidential elections in 2016 were the sixth of their kind and resulted in the fourth handover of power. Benin passed Samuel Huntington’s famous ‘two-turnover test’, whereby democracies are seen as consolidated after the second electoral turnover, in 2006 when Kérékou had to leave office at the end of his two-term limit. That constituted the second extraordinary event. Leading up to 2006, rumours had been spreading that Kérékou would seek a third term in office. But the constitutional consensus reached by the National Assembly had remained highly valued by the population, and people took to the streets under the slogan ‘touche pas à ma constitution’—‘don’t touch my constitution’. The same thing happened in 2016, when Kérékou’s successor, Thomas Boni Yayi, also sought to circumvent the two-term limit. Again, citizens rushed to the defence of the constitution using the same slogans and banners as in 2006, with an even broader coalition this time. Benin, this small, impoverished country with a chequered past of military rule and one-party autocracy now has an unbroken history of democracy that is about to celebrate its 30th anniversary. And it is not that this 30-year history has been easy. Far from it, democracy in Benin has been threatened time and again, but it has proven to be remarkably resilient in spite of weak state institutions and a fragmented party system, thanks mostly to the pro-democratic attitudes and activist stance of Beninese civil society. Compare this with Liberia and Guinea-Bissau, two other small West African countries. In 1989, Liberia too was poor and had a history of single-party dominance and military rule. But instead of a peaceful revolution, it experienced a violent insurrection by the National Patriotic Front of Liberia (NPFL) led by Charles Taylor. The NPFL quickly attracted support from a disaffected populace while the Liberian state, weakened

1

INTRODUCTION

3

through decades of autocratic misrule by President Samuel Doe and his predecessors, was unable to quash the rapidly growing insurgency. In July 1990, a decisive victory for the NPFL seemed at hand but a Nigerianled military intervention by the Economic Community of West African States (ECOWAS) prevented it from capturing the capital, Monrovia. Six years of stalemated conflict and a dozen failed peace agreements later, the Nigerian leadership and Taylor reached a compromise deal to introduce a caretaker government followed by presidential elections, which Taylor and the still fully armed NPFL would go on to win with a decisive majority. But Taylor’s accession to the presidency in 1997 did not bring peace. Taylor ran the country like a kleptocracy for the benefit of himself and his closest supporters. There was no reconciliation or reconstruction and society’s wounds were left to fester. In 2000, militias mobilized against Taylor, whose government proved to be no more capable of counterinsurgency than its predecessor. By 2003, with the militias advancing on Monrovia, Taylor agreed to go into exile in Nigeria. This paved the way for a UN-supervised peace process, backed by a large contingent of peacekeepers, that led to democratic elections in 2005 and 2006. Democracy has held since then, even as each new election tests the stability of the new regime. Estimates of casualty figures from the first Liberian war (1989–1996) vary between 150,000 and 250,000 people dying on and off the battlefield (Ellis 2006, pp. 312–316), with a smaller number in the second Liberian war (2000–2003). According to official statistics of the UN High Commissioner for Refugees, during the mid-1990s there were up to 800,000 Liberian refugees in the neighbouring countries of Sierra Leone, Guinea, and Côte d’Ivoire, plus an estimated one million internally displaced persons. Given that the pre-war population was only about three million, these are staggering numbers. The Liberian economy was devastated and the country has needed massive injections of aid to finance the recovering state apparatus. Guinea-Bissau, like Benin, had been run by a socialist regime that had come to power in a military coup. Under the leadership of João Bernardo Vieira, the Revolutionary Council governed through the only party, the Partido Africano para a Independência da Guiné e Cabo Verde (PAIGC), for 14 years. Faced with a worsening economic crisis and abandoning socialism, Vieira started to liberalize the regime in 1991, legalizing the activities of opposition parties. The first multiparty elections were held

4

D. LAMBACH ET AL.

in August 1994. Vieira narrowly won the presidential election and the PAIGC retained control of parliament. But democratization did little to assuage social and economic grievances or to mediate elite power struggles. The post-transition Vieira government was tainted by corruption, patronage, and economic stagnation. Finally, in 1998, the army chief of staff Ansumane Mané deposed Vieira after a bloody 11-month civil war that necessitated the deployment of ECOWAS and UN peacekeepers. And even though Mané’s military junta appointed a transitional government afterwards and Guinea-Bissau has had no fewer than five multiparty elections since then, it has never managed to attain even a semblance of political stability. There have been successful coup d’états in 2003, 2009, and 2012, plus an unsuccessful attempt in 2011. Vieira returned to power in the 2005 elections but was assassinated while in office during the 2009 coup. Politics in the country is characterized by political factions in government and the military engaging in all-out power struggles. Civil society, especially in rural areas, is weak and disenfranchised, preferring to keep its distance from the state (Forrest 2003). Benin, Liberia, and Guinea-Bissau represent three different ways of democratizing a country. In Benin, democratic transition was forced upon the regime through large-scale nonviolent resistance. In Liberia, the government was toppled by armed insurrection, with democracy only coming about via a mediation process under heavy international pressure. In Guinea-Bissau, democracy, such as it was, was installed top-down by the incumbent authoritarian government. Comparing these three modes of transition shows that events leading up to and during transition will affect the long-term viability of democracy for years or even decades to come. The first point is that transition without popular mobilization robs democracies of some of their lifeblood. In cases like Guinea-Bissau, citizens and civil society are mostly demobilized and have little leverage over the government. The political, military, and economic elites retain their perks and veto positions. Democratic reform is shallow and vulnerable to military or executive coups. In contrast, revolutions, whether violent or peaceful, can constitute a vibrant and capable civil society and generate a feeling of efficacy among participants and observers and can be a powerful symbol for the future generations. But not all revolutions have the same effects. Violent insurgency leads to thousands—sometimes hundreds of thousands—of battle-related and indirect deaths. It destroys people’s livelihoods and displaces them from

1

INTRODUCTION

5

their homes. Also, whereas successful nonviolent resistance inaugurates political transition in a divided but broadly still intact society, successful armed insurrection leads to a transition under the worst possible circumstances. The country is left in ruins, there are severe humanitarian challenges needing immediate attention, and society is deeply split. Democratization is a risky process even under favourable conditions; to have to do it in a post-war context makes it even less likely to succeed (Cederman et al. 2010).

Thinking About Nonviolence From an ethical perspective, nonviolent strategies are clearly the superior choice. In a disciplined movement, violence by protesters is rare. What few casualties there are during nonviolent resistance occur mostly at the hands of oppressive governments. There is much less disruption of the everyday lives of citizens and displacement is kept at a minimum. Societies undergoing peaceful revolutions are not as divided as those that emerge traumatized from violent conflict. Principled nonviolence is often associated with popular figures like Mahatma Gandhi or Martin Luther King and religions like Hinduism or Christianity, but the norm to refrain from violence is anchored in each of the major faiths and most major philosophies through some variant of the Golden Rule: ‘Do unto others as you would have them do unto you’ (Küng 1998). But there are also pragmatic reasons for choosing nonviolent forms of resistance (Sharp 1973). Study after study has found that nonviolent protest is much more effective at achieving its goals than violent forms of dissent. This has been demonstrated by, among others, a landmark study by Erica Chenoweth and Maria Stephan, Why Civil Resistance Works (2011). Using detailed data on 323 resistance campaigns collected in the Nonviolent and Violent Conflict Outcome database (NAVCO), Chenoweth and Stephan find that nonviolent campaigns are the most frequent form of resistance campaigns and have a higher probability of success than other forms of resistance. In contrast, violent transitions to a stable democracy are extremely rare. In 218 instances of violent resistance, democracy emerged in only 5% of cases (Chenoweth and Stephan 2011, p. 209). In contrast, nonviolent campaigns often trigger a transition towards democracy in autocratic regimes (Celestino and Gleditsch 2013) and have a significant and positive impact on the probability of a democratic regime persisting five years later (Chenoweth and Stephan

6

D. LAMBACH ET AL.

2011, p. 213). Clearly, nonviolence works even when it is used in a purely instrumental manner and without an ethical commitment to pacifism. Chenoweth and Stephan’s book was instrumental in pushing what had been a relatively obscure research field into the public limelight. Policymakers and the media started to pay attention to nonviolent resistance and academic researchers were galvanized into action. Since its publication eight years ago, the book has been cited widely and is rapidly becoming a modern classic of political science scholarship. In its wake, dozens of follow-up studies were conducted and the interdisciplinary field of Resistance Studies emerged, with its own research community, journals, and conferences. But praising the impact of Why Civil Resistance Works should not distract us from recognizing the long and deep intellectual traditions upon which the field rests. Prior empirical works, for example by Ackerman and Karatnycky (2005) and Johnstad (2010), had already produced similar results, albeit based on more limited datasets. There are also scores of studies on single cases or discussing particular aspects on nonviolence (for an overview, see Ackerman and Rodal 2008; Dudouet 2011). Much of this earlier research was inspired by the work of Gene Sharp, who took a pragmatist approach to the issue of nonviolent protest. With his seminal three-volume Politics of Nonviolent Action (1973), Sharp became one of the academic pioneers in the study of nonviolent action. Building upon the insight that the power of an authoritarian ruler is based on the consent of the population, Sharp not only developed a theory of power but also compiled 198 techniques of nonviolent resistance suitable for challenging authoritarian rulers. In contrast to Gandhi or King, Sharp did not make a moral or religious argument about the necessity of nonviolence. Instead, his aim was to develop a strategic approach towards nonviolence suitable for people around the world, earning him the nickname ‘the Clausewitz of Nonviolence’. His works had a tremendous impact: From Dictatorship to Democracy (Sharp 2008) has been translated into more than 30 languages and served as a source of inspiration for activists in places as diverse as Serbia, Burma, and Egypt. Aside from Sharp, issues of nonviolence are also discussed prominently in many works of political thought (Boersema and Brown 2006; Atack 2012). The earliest theory of nonviolent resistance was elaborated in the sixteenth century by Étienne de la Boétie (1975 [1576]). Having witnessed the bloody absolutist rule of Henry II of France, de la Boétie laid the groundwork for most modern theories on nonviolent resistance.

1

INTRODUCTION

7

In his Discourse on the Voluntary Servitude, posthumously published in 1576, he tried to answer the question of how so many people can suffer from a single tyrant who has no more power than each of them individually. He concluded that it is the voluntary servitude of the people that allows a single person to rule over so many, an idea that had a substantial impact on later scholars of nonviolent action and civil disobedience. It is clear that there is an active and expanding research agenda for the field of nonviolent resistance that is able to draw on a rich tradition of scholarship. But within the context of this unfolding agenda, there are still many unanswered questions. We are interested in two issues in particular: does nonviolent resistance have long-term benefits for democracies, as the examples at the beginning of this chapter suggest? And if so, what are the underlying causal mechanisms?

The Case for Democracy The focus of the nonviolent resistance literature is on a particular technique rather than the aims which actors are trying to achieve. Previous work has covered resistance against various forms of injustice and oppression and for all kinds of goals, like national independence, gender equality, ethnic autonomy, religious freedom, or the redress of socio-economic grievances. Why, then, do we focus on the impact of nonviolent resistance on democracy? The answer has scholarly, normative, and practical elements. From a scholarly perspective, we have chosen to ask these questions simply because they have not been answered before. Many contributions have asserted that peaceful resistance can help instal democratic regimes that are more resilient and more likely to consolidate than democracies which came about in other ways (Ackerman and Duvall 2001; Ackerman and Rodal 2008; Dudouet 2011; Sharp 2005, 2008), but these claims have not been systematically evaluated. In this sense, analysing the longterm impact of nonviolent resistance on democratic consolidation represents an intriguing puzzle, the answers to which might move the research programme on nonviolent resistance forward. In a normative sense, the authors of this book share an ethical commitment to democracy that we want to be transparent about. We view the spread of democracy over the past centuries as a political achievement that is worth defending, in spite of all its faults. As such, we agree with Winston Churchill that ‘democracy is the worst form of government, except

8

D. LAMBACH ET AL.

for all the others’, but we also believe that there are more positive reasons to be in favour of democracy. Citizens in democracies live more freely and have greater agency and capacity to shape their own lives. This becomes very obvious when citizens become dissidents and choose to resist certain laws or policies of their government. While freedom can be construed as a developmental goal in itself (Sen 1999), citizens in democracies are also better off in material terms. Democracies are wealthier and show better progress in human development than non-democracies (Gerring et al. 2012). Which way the causal arrow runs has been the subject of much controversy: Does democracy cause well-being, or does a well-off citizenry opt for democracy (Przeworski 2000)? Whichever answer is preferred, both are arguments in favour of democracy. Either democracy is the cause of socio-economic development or it is seen as a worthwhile goal by ordinary people around the world who find themselves in a position to influence their political system. Beyond whether democracy is desirable at all, the quality of democracy matters a great deal (Munck 2016). Formal democracy only requires competitive and open elections, but the boundary between such a minimalist conception of democracy and electoral autocracies, where elections are also held without ever challenging ruling interests, is thin. So we want democracies that are not just democracies in name only but full, consolidated, high-quality democratic systems that are characterized by a separation of powers, the rule of law, unconditional respect for political rights and civil liberties, functioning state institutions, an active and engaged citizenry, and a clear and unequivocal commitment to democratic ideals among societal elites. High-quality democracies are inclusive: they have institutions that are representative of the demos and that provide for the basic needs of all citizens. We expect that nonviolent resistance is helpful in moving democracy towards this ideal. In practical terms, our research may also be of interest to actors and activists within social and resistance movements, and to policymakers engaging with them. This is a particularly timely issue now that we are firmly in the backlash phase of the third wave of democratization (Huntington 1991; Diamond 2008). After large gains in the number of democracies in the 1990s, the tide has turned, although there is considerable debate how broad and deep this reversal truly is (see the controversy between Foa and Mounk 2016, 2017 and Inglehart 2016), with Waldner and Lust pointing out that while there is some backsliding in terms of the

1

INTRODUCTION

9

quality of democracy, there are few instances of true democratic breakdown in terms of reversion to autocracy (2018, p. 94). Since reversals typically occur in ‘young’, less consolidated democracies, we wish to learn more about the process of consolidation and what role mass movements can play therein. We therefore echo the sentiment of Kadivar, Usmani, and Bradlow who point out: ‘In an era of anxiety about the antidemocratic proclivities of the mass public, our results are a reminder that ordinary people have advanced the cause of democracy and not hindered it’ (Kadivar et al. 2019, p. 2).

Research Design and Results The question of the long-term effects of nonviolent resistance has received comparatively little attention. Chenoweth and Stephan analyse the impact of nonviolent resistance on democratic survival up to five years post transition but are silent on the longer-term prospects. The mechanisms that make nonviolent resistance work are also still shrouded in mystery. Chenoweth and Stephan argue that nonviolent campaigns are more attractive than violent resistance to large segments of the population owing to their less extreme means (see also Schock 2005). They offer ‘an opportunity to people to participate with varying levels of commitment and risk tolerance’ (Chenoweth and Stephan 2011, p. 37). Accordingly, the larger number of participants as compared to violent insurrection is crucial for a nonviolent movement’s success (Chenoweth and Stephan 2011, pp. 39–40). A broad movement also requires consensus-seeking among disparate opposition groups and an inclusive approach in their demands. Furthermore, Sharp (2005) expects that government repression of peaceful protest would increase mobilization and give the opposition access to external support. Ackerman and Rodal (2008, p. 118) argue that the moral superiority of nonviolent resistance may help fracture regime coalitions and entice security forces to defect or to remain neutral. However, many of these assumptions have not yet been fully substantiated by empirical research and even Chenoweth and Stephan’s monograph does not fully answer its own title, Why Civil Resistance Works. In this book, we take a slightly different approach, comparing different modes of transition. We distinguish cases based on whether transition was induced by civil resistance, violent resistance, or through a top-down process of elite-led liberalization. We conduct a range of statistical analyses to gauge the effect of nonviolent resistance on different

10

D. LAMBACH ET AL.

measures of democratic consolidation. Based on a dataset that combines information on democratic regimes with information on the presence of nonviolent resistance during the transition of these regimes, we estimate the effect of nonviolent-resistance-induced transition on three different outcome measures of democratic consolidation: (1) democratic survival, (2) accomplishing the two-turnover test, and (3) quality of democracy. Our findings demonstrate that nonviolent resistance during the transition to democracy has a long-term beneficial effect on the odds of democratic survival, on the likelihood of a democracy seeing two peaceful turnovers of power, and on the quality of democracy. For some indicators, these effects persist for over a decade. This is particularly intriguing—how can the form of resistance have a stabilizing effect on democracies years, even decades, after the resistance movement has demobilized? How are these stabilizing effects transmitted and the legacy of resistance kept alive in post-transition democracy? To answer these questions, we combine the statistical work with a series of case studies of consolidating democracies in sub-Saharan Africa and Latin America. Using a ‘Most Different, Similar Outcome’ logic (De Meur and Berg-Schlosser 1994), we selected two cases where democracy resulted from nonviolent resistance (Benin and Chile), two cases where democracy resulted from violent struggle (Namibia and El Salvador) and two cases of top-down transition (Cape Verde and Paraguay). Data for the African cases was collected by Markus Bayer via interviews and archival research during ten weeks of fieldwork in each country. The Latin American cases were conducted as desk studies by Matteo Dressler and Katrin Planta at the Berghof Foundation, supplemented by interviews with country experts and in-country key informants. Our findings from a comparison of these cases and from more detailed statistical tests indicate that nonviolent resistance stabilizes democracies and helps them consolidate via three key mechanisms. First, nonviolent resistance ‘levels the political playing field’. It is effective at dislodging incumbents without replacing them with another dominant political force. This uncertainty about power relations during the transition process leads to the creation of more inclusive institutions and procedures of government and to more effective checks and balances on the executive. Second, successful nonviolent protest promotes civic activism and creates a more democratic political culture. The resistance movement also leaves behind a symbolic reference point for future remobilization. The general

1

INTRODUCTION

11

population sees political activism in a more positive light and is supportive of crucial political rights like freedom of expression and freedom of association. Third, nonviolent resistance avoids the ‘praetorian problem’ of the military getting involved in politics by working towards healthier civil–military relations. In contrast, democracies installed by violent resistance and or through top-down resistance tend to have more military involvement in politics and run a higher risk of being toppled via coups d’état. Given the small number of case studies, we need to be cautious about overgeneralizing from our findings. However, evidence from other successful and unsuccessful cases of nonviolent resistance suggests that the results are also applicable to other circumstances. At the very least, our results suggest avenues for further research, opening the way towards a more nuanced discussion of the causal mechanisms underpinning the effectiveness of nonviolent resistance. We see particular potential in disentangling the drivers of political transition from its process. Our findings suggest that whereas resistance movements have great potential to pressure regimes into political reforms, elites still exercise greater agency in shaping the specific institutional process of transition.

Outline of the Book The book proceeds as follows. Chapter 2 sets out our argument on how nonviolent resistance fosters democratic consolidation. It briefly surveys the literature on nonviolent resistance and democratic consolidation, explains democratization in terms of critical junctures and path dependence, and introduces our model of transition as an interaction between incumbents, oppositional elites, the security forces and citizens. Chapter 3 presents results from quantitative tests of the first part of our argument. Comparing cases of nonviolent resistance with cases of violent resistance and top-down liberalization shows that the former produce democracies with much higher odds of survival and significantly better indicators of democratic quality. Chapter 4 presents our findings on the underlying mechanisms of nonviolent resistance’s long-term democratizing effect. The chapter focuses on the two case studies of nonviolent resistance, Benin and Chile, giving brief narrative histories for each. It then discusses individual causal mechanisms and their interaction which we contrast with cases of top-down and violent transition to highlight the specific effect of nonviolent resistance on democratic consolidation. Chapter 5 brings

12

D. LAMBACH ET AL.

together the findings of the previous chapter into a theoretical synthesis. It summarizes our findings, situates them in the literature, and offers new research perspectives. Chapter 6 concludes the book with some observations about the study of nonviolent resistance and some implications for political practice.

References Ackerman, P., & Duvall, J. (2001). A Force More Powerful: A Century of Nonviolent Conflict. New York: Palgrave Macmillan. Ackerman, P., & Karatnycky, A. (2005). How Freedom Is Won: From Civic Resistance to Durable Democracy. New York: Freedom House. Ackerman, P., & Rodal, B. (2008). The Strategic Dimensions of Civil Resistance. Survival, 50(3), 111–126. Atack, I. (2012). Nonviolence in Political Theory. Edinburgh: Edinburgh University Press. Boersema, D., & Brown, K. G. (Eds.). (2006). Spiritual and Political Dimensions of Nonviolence and Peace. Amsterdam: Rodopi. Cederman, L.-E., Hug, S., & Krebs, L. F. (2010). Democratization and Civil War: Empirical Evidence. Journal of Peace Research, 47 (4), 377–394. https:// doi.org/10.1177/0022343310368336. Celestino, M. R., & Gleditsch, K. S. (2013). Fresh Carnations or All Thorn, No Rose? Nonviolent Campaigns and Transitions in Autocracies. Journal of Peace Research, 50(3), 385–400. https://doi.org/10.1177/0022343312469979. Chenoweth, E., & Stephan, M. J. (2011). Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. New York: Columbia University Press. de la Boétie, E. (1975 [1576]). The Politics of Obedience: The Discourse of Voluntary Servitude. New York: Free Life Editions. De Meur, G., & Berg-Schlosser, D. (1994). Comparing Political Systems: Establishing Similarities and Dissimilarities. European Journal of Political Research, 26(2), 193–219. https://doi.org/10.1111/j.1475-6765.1994.tb00440.x. Diamond, L. (2008). The Spirit of Democracy: The Struggle to Build Free Societies Throughout the World. New York: Times Books. Dudouet, V. (2011). Nonviolent Resistance in Power Asymmetries. In B. Austin, M. Fischer, & H. J. Giessmann (Eds.), Advancing Conflict Transformation (pp. 237–264). Opladen: Barbara Budrich. Ellis, S. (2006). The Mask of Anarchy, Updated Edition: The Destruction of Liberia and the Religious Dimension of an African Civil War. London: Hurst & Company. Foa, R. S., & Mounk, Y. (2016). The Danger of Deconsolidation: The Democratic Disconnect. Journal of Democracy, 27 (3), 5–17.

1

INTRODUCTION

13

Foa, R. S., & Mounk, Y. (2017). The Signs of Deconsolidation. Journal of Democracy, 28(1), 5–16. Forrest, J. B. (2003). Lineages of State Fragility: Rural Civil Society in GuineaBissua. Athens: Ohio University Press. Gerring, J., Thacker, S. C., & Alfaro, R. (2012). Democracy and Human Development. The Journal of Politics, 74(1), 1–17. https://doi.org/10.1017/ s0022381611001113. Huntington, S. P. (1991). The Third Wave: Democratization in the Late Twentieth Century. Norman: University of Oklahoma Press. Inglehart, R. F. (2016). How Much Should We Worry? Journal of Democracy, 28(1), 18–23. Johnstad, P. G. (2010). Nonviolent Democratization: A Sensitivity Analysis of How Transition Mode and Violence Impact the Durability of Democracy. Peace & Change, 35(3), 464–482. Kadivar, M. A., Usmani, A., & Bradlow, B. H. (2019). The Long March: Deep Democracy in Cross-National Perspective. Social Forces. https://doi.org/10. 1093/sf/soz050. Küng, H. (1998). A Global Ethic for Global Politics and Economics. Oxford: Oxford University Press. Munck, G. L. (2016). What Is Democracy? A Reconceptualization of the Quality of Democracy. Democratization, 23(1), 1–26. https://doi.org/10.1080/ 13510347.2014.918104. Przeworski, A. (2000). Democracy and Development: Political Institutions and Well-Being in the World, 1950–1990. New York: Cambridge University Press. Schock, K. (2005). Unarmed Insurrections: People Power Movements in Nondemocracies. Minneapolis: University of Minnesota Press. Sen, A. (1999). Development as Freedom. Oxford: Oxford University Press. Sharp, G. (1973). The Politics of Nonviolent Action (3 vols.). Boston: Porter Sargent. Sharp, G. (2005). Waging Nonviolent Struggle. Boston: Albert Einstein Institution. Sharp, G. (2008). From Dictatorship to Democracy. A Conceptual Framework for Liberation. Boston: Albert Einstein Institution. Waldner, D., & Lust, E. (2018). Unwelcome Change: Coming to Terms with Democratic Backsliding. Annual Review of Political Science, 21(1), 93–113. https://doi.org/10.1146/annurev-polisci-050517-114628.

CHAPTER 2

Theory

This chapter sets out our argument on how nonviolent resistance (NVR) fosters democratic consolidation. First, building on existing theoretical accounts of democratic transitions and democratic consolidation, we develop a relational approach of democratic regimes. This means that we assume that democracy depends on the establishment and maintenance of stable and cooperative relations between four key actors: (1) the government, (2) the opposition, (3) security forces, and (4) citizens.1 During democratic transition, a specific form of relationship between these four actors is established and preserved in political institutions such as constitutions, party systems, and electoral franchises. These institutions structure civil–military relations, state–society relations and electoral competition between the incumbent government and the opposition. Together, these relations constitute a democratic equilibrium, where the interests of all major political forces converge on maintaining democracy. Second, we use this framework to conceptualize democratic transitions as critical junctures with lasting and path-dependent effects on subsequent political development. Specifically, we argue that the mode of transition determines how stable or fragile the equilibrium of power relations

1 This does not imply that other kinds of actors, such as business elites or societal authority figures, are unimportant. But we expect that their importance varies greatly from case to case which is why we do not include them in this general model for the sake of parsimony. For the same reason, our treatment of these groups as singular actors glosses over the possibility of internal factionalism within each group.

© The Author(s) 2020 D. Lambach et al., Nonviolent Resistance and Democratic Consolidation, https://doi.org/10.1007/978-3-030-39371-7_2

15

16

D. LAMBACH ET AL.

between key actors is, which in turn has a substantial impact on the odds of democratic consolidation or breakdown and the quality of the resulting democracy. We therefore distinguish elite-led transitions from those initiated by the masses through either violent or NVR. Third, we describe the mechanisms of how these different modes of transition affect political developments in a path-dependent way, arguing that NVR-induced transitions are more beneficial for democratic consolidation and avoiding democratic breakdown than elite-led transitions and democratization brought about by violent rebellion. Specifically, we focus on the effect of NVR on three observable implications of democratic consolidations: (1) democratic survival, (2) peaceful turnover of power, and (3) democratic quality.

Democratic Consolidation Since our main interest is the political development of democratic regimes, we use a categorical classification of political regime that only distinguishes between democratic and non-democratic regimes. We follow Skaaning’s definition of a political regime as an ‘institutionalized set of fundamental formal and informal rules structuring the interaction in the political power centre (horizontal relation) and its relation with the broader society (vertical relation)’ (Skaaning 2006, p. 13). Thus, the distinction between democratic and non-democratic regimes is mainly based on the question of how vertical and horizontal relations are structured. We follow the comparative politics literature on categorical regime classifications (Cheibub et al. 2010, p. 69; Geddes et al. 2014, pp. 317–318; Ulfelder 2010, p. 56; Boix et al. 2013, p. 9) to argue that a regime can be described as democratic if political relations are marked by contestation (in the horizontal dimension) and participation (vertical dimension). Contestation consists of two conditions: (1) the executive is directly or indirectly elected in popular elections and is responsible either directly to voters or to a legislature and (2) the legislature (or the executive if elected directly) is chosen in free and fair elections. Participation refers to a minimal level of suffrage (i.e. a majority of adult men have the right to vote). Political regimes that do not fulfil all of these criteria are considered non-democratic.2

2 The differences among the cited works are rather subtle, although the conceptual distinction between contestation and participation is emphasized most prominently by Boix et al. (2013).

2

THEORY

17

When switching between democracy and non-democracy, regimes undergo a political transition.3 Regarding the direction of transition events, we focus on the transition from an autocratic regime to a democratic regime. Siaroff (2008, pp. 274–277) distinguishes several different ways in which these democratic transitions come about. First, the ruling elite of an autocratic regime may itself decide to democratize (e.g. when a military government returns to the barracks). Second, the process of democratization may be initiated through negotiations between the autocratic regime and opposition groups. Third, autocratic regimes may collapse because of external or internal pressure, for instance if a dictator goes into exile because of mass protest, losing a civil war or when foreign powers conquer an autocratically ruled state and impose a democratic system. However, our research interest lies not with the transitions themselves but with what happens afterwards. Are ‘new’ democracies able to survive and thrive, do they break down through a military or executive coup (i.e. a self-coup, or autogolpe) or do they ‘muddle along’ somewhere in between? Democracies clearly differ in terms of their quality: where some regimes only fulfil the bare minimum of the participation and contestation criteria, others go beyond that by expanding suffrage, adhering to higher standards of electoral integrity or providing and safeguarding civil liberties (Merkel 2004). To account for these differences in democratic quality, theories of political transition distinguish different phases of transition, from the liberalization of the old regime via the institutionalization of the democratic regime and its subsequent consolidation as democracy becomes ‘the only game in town’ (Linz and Stepan 1996b, p. 15). Democratic consolidation is therefore mainly concerned ‘with making democratic institutions both enduring and functional and connecting them to civil society and citizenship’ (Encarnación 2000, p. 485). Consolidation refers to the intuitively compelling notion that as democracies get older, they become more rooted in society and therefore better able to withstand challenges. Unfortunately, there is no consensus in the literature as to how consolidation should be conceptualized (see Schneider 1995, for an overview 3 A dichotomous conceptualization of regime type is necessary so that we can identify transition events. For a more elaborate discussion of the advantages and disadvantages of dichotomous and continuous approaches to specify regime types, see Collier and Adcock (1999).

18

D. LAMBACH ET AL.

of key debates). Linz and Stepan (1996b, pp. 15–16) provide probably the most influential approach by distinguishing three dimensions of consolidation: the behavioural dimension, when no political significant group aims to overthrow the political system; the attitudinal dimension, when the overwhelming majority of the population supports democracy even during periods of crisis; and the constitutional dimension, when all political actors are accustomed to managing conflicts through democratic institutions and procedures (similarly Schedler 2001). ‘In short, with consolidation, democracy becomes routinized and deeply internalized in social, institutional, and even psychological life, as well as in political calculations for achieving success’ (Linz and Stepan 1996b, p. 16). Others equate consolidation with the likelihood of democratic survival, referring ‘to the process by which a newly established democratic regime becomes sufficiently durable that democratic breakdown – a return to nondemocratic rule – is no longer likely’ (Gasiorowski and Power 1998, p. 743). Schedler speaks of it as ‘expected regime stability’ (2001, p. 69; see also Beetham 1994; Svolik 2015). O’Donnell (1996) criticizes such approaches for tending to treat consolidation as little more than the opposite of democratic breakdown, robbing it of crucial depth. To compare these different standpoints, Schedler (1998, pp. 94–95) distinguishes ‘negative’ understandings of consolidation, which focus on survival and the avoidance of democratic breakdown, from ‘positive’ understandings, which highlight improvements in democratic practice. Moreover, there are different viewpoints on whether consolidation should be seen as an end state, i.e. whether democracies can ever be fully consolidated, or whether it should be seen as an open-ended process, although O’Donnell (1996) rightly points out that either notion carries an unspoken teleology that all democratic regimes should consolidate after transition. Given this lack of consensus, we take a pragmatic approach to the concept of consolidation. To make this complex, latent process visible, we focus on its observable implications at the level of the political regime. We ask, in other words, what effects we should be able to observe as a democratic regime consolidates. Three indicators are particularly important for our quantitative research (see Chapter 3): (1) democratic survival, (2) peaceful turnovers of power, and (3) democratic quality. Democratic survival is a necessary condition of consolidation—if a democracy breaks down, it cannot consolidate. Also, the sheer persistence of democracy implies that it has been able to weather various challenges along the way. Long-standing democracies are also better able to create civic political

2

THEORY

19

cultures that support their stability. Peaceful turnovers of power are often cited as a proxy for regime consolidation (Huntington 1968; Gasiorowski and Power 1998). Two peaceful turnovers of power imply that the party that won the founding election after democratic transition loses a subsequent election and hands over power to another party (first turnover); when this other party peacefully transfers power to another party after subsequent elections, two turnovers of power have taken place.4 Finally, there are various ways of measuring democratic quality that offer a more detailed if more complex picture of democratic practice than the first two indicators (Coppedge et al. 2015). The consolidation literature provides some guidance for our qualitative case studies (see Chapter 4). We find Linz and Stepan’s (1978, 1996a) approach useful in that they highlight two important dimensions of consolidation that are otherwise lost in a more results-oriented framework: the roles of actors and political culture (similarly Merkel 2010). If we wish to explain why and how particular democracies consolidate while others do not, we cannot just look at system-level determinants. Instead, Linz and Stepan (and other transitologists) remind us that political systems are continually reproduced through the interactions of political agents that are embedded in a complex web of institutions and state– society relations. However, similar to many other classic authors, Linz and Stepan (1978) highlight the role of elites in upholding the democratic consensus and preventing the breakdown of democracy. Such an elite-centred approach can also be found in works that understand democratic transition (and consolidation) and war-to-peace transition in terms of ‘settlements’, whether they be peace settlements (Barma 2016), political settlements (Khan 1995) or elite settlements (Burton et al. 1991; Higley and Burton 1998). In contrast to these approaches, we argue that the broader population is quintessential not only for bringing about democratic change (a point forcefully made in classic works on NVR like Chenoweth and Stephan [2011], Ackerman and Karatnycky [2005]) but also for upholding and defending it. As Rueschemeyer, Huber and 4 The ‘two-turnover test’ was first proposed by Huntington (1991) and is admittedly crude. Schneider points out some of the problematic classifications thrown up by this approach: ‘Japan barely met the two-turnover test in the 1990s, the United States did not meet it until 1840, and Chilean democracy was consolidated in 1970, only shortly before collapsing’ (Schneider 1995, p. 220). Power and Gasiorowski (1997, p. 132, Fn. 116) argue that a ‘first-turnover test’ should be sufficient, especially when analysing young democracies in developing countries.

20

D. LAMBACH ET AL.

Stephens argue, ‘(t)hose who have the most to gain from democracy will be its most reliably promoters and defenders’ (Rueschemeyer et al. 1992, p. 57). Similarly, Bermeo (2003, p. 19) points out that democratic breakdown does not usually occur due to the actions of ordinary people but rather due to their inaction in the face of elites deviating from the democratic path. We therefore supplement the literature on democratic transition with published research on NVR and social movements. As highlighted by scholars of NVR like Etienne de la Boétie (1975/1576) or Gene Sharp (1973), at its most basic every regime relies on the acceptance and obedience of its population. This actor-centred perspective gives citizens a substantial amount of agency which is rooted in their potential to refuse obedience. We understand NVR to be a ‘reversible’ (Galtung 1996) type of resistance that disrupts given relations, strives for constructive conflict settlement (Dudouet 2008), and makes it possible to negotiate a new social contract based on more equal terms (Cervellati et al. 2014). We therefore assume that NVR and the involvement of large parts of the population in ‘eventful transitions’ (della Porta 2014) not only transforms relations among elites (i.e. between incumbents and the opposition) but also between elites and the broader population (Vinthagen 2015). This generally leads to (a) more stable political arrangements than mere elite settlements (Bethke 2017; Bayer et al. 2016) and (b) more inclusive institutions (Pinckney 2018; Bayer 2018) and greater respect for civil rights (Bethke and Pinckney 2019) than top-down transitions or those effected by violent resistance (VR). Furthermore, NVR-induced transitions also help shape the democratic practices and the political culture of a country (della Porta 2016). We therefore take a relational approach to political regimes and their consolidation that includes the broader population and highlights its role during transition and the subsequent process of deepening and consolidating democracy.

A Relational Approach to Political Regimes To explain how democratic transitions occur and how democracy sustains, breaks down, or consolidates, we use the concept of democratic equilibrium, which refers to a situation where the interests of the major political actors converge on establishing and/or maintaining democratic institutions (Przeworski 2005). We model these situations as a set of strategic interactions between four key actors: incumbent elites (the government),

2

Opposition

4

1

Government

2

THEORY

21

Security forces

3

5

Citizens

Fig. 2.1 Horizontal and vertical relations of political regimes

elites outside the regime (opposition), the security forces, and citizens. These relations are depicted in Fig. 2.1. Between these actors, we identify three primary (1–3) and two secondary relations (4–5) that determine the emergence, stability, and quality of the democratic equilibrium. Two of these relations are horizontal (1–2), defining relations between relevant elites. The other three are vertical relations (3–5), meaning that they describe relations between elites and the broader population. To specify how these relationships determine whether a democratic equilibrium is established and maintained, we partly build on related work from formal political theory (Fearon 2011; Przeworski 1991, 2005; Weingast 1997). These theoretical models stress that democracy is not just brought about by changing norms and culture within societies, but that key actors must have strategic motivations to initiate a transition from autocratic to democratic rule. We view political actors as rational agents but apply a broad understanding of rationality (Levi 1990). In other words, actors’ preferences are shaped as much by socialization, appropriateness, ideology, and political culture as they are by economic benefits or a will to power. Since we are not chiefly interested in what drives actors but what comes out of their interaction, we make no strong assumptions about what motivates actors except that they will generally act in such a way as makes sense to them. Therefore, actors may decide to support democracy for both highminded ideals and narrow self-interest. Their decision is based on a strategic calculation in the sense that each actor opts to support or oppose the democratic equilibrium based on their own interest but also takes into account their expectations about the strategy of the other main actors. If

22

D. LAMBACH ET AL.

each actor views supporting democratic institutions as their best response to the expected decisions of the other actors, a democratic equilibrium is established. The equilibrium is maintained if none of the actors has a strong enough incentive to change their preference from democracy to a different system of government. Thus, democratic equilibrium is the result of self-interested strategic compliance (Fearon 2011; Przeworski 2005). Correspondingly, the equilibrium breaks down primarily when the interests and expectations of the four main actors change.5 The breakdown of democracy is the result of multiple decisions taken by relevant actors within the democratic regime. Either the ruling elite itself abandons democratic rule and decides to govern autocratically or factions from within the regime or from outside depose the democratic government through a coup d’état, rebellion, or popular mobilization. Actors do not operate in a void but within and through institutions. We follow Levi in treating institutions as ‘social bargains’ (Levi 1990, p. 409) between actors that lock in particular distributions of resources, rights, and obligations. But although we view actors as ontologically prior to institutions, institutions still have an effect on actors’ preferences, subjectivities, and scope of action (Scharpf 1997). For example, institutions have a socializing function on their members, inculcating them with shared values (Schimmelfennig 2000). They also provide transparency about roles, rights, and mutual expectations among actors. Furthermore, institutions have a high degree of inertia: once created, they are hard to change or get rid of. This makes political institutions such as the constitution, parliament, and the judiciary important pillars and foundations of democratic regimes. Similarly, political culture matters for actors’ preferences and behaviours and thus for the odds of democratic consolidation (Welzel 2007). 1. Relationship between government and opposition The relationship between government and opposition is one of the defining differences between democracies and autocracies. Democracy depends on having a meaningful and independent opposition. Autocracies repress, disallow, or otherwise stymie the activities of the opposition. 5 Of course, this requires the simplifying assumption that each group is homogenous in their assessment and action. In truth, each group will contain factions and subgroups, but accounting for such complexity would overload the model.

2

THEORY

23

But even though authoritarian regimes rarely allow a meaningful political opposition to operate, they still have to deal with political elites that are not part of the regime coalition, typically resorting to repression or co-optation. Liberal democracies rely on political pluralism, which is reflected in multiparty systems. However, we cannot infer from the fact that multiple parties exist anything specific about the relations between government and political opposition within a multiparty system. The relation is determined, first, by the relative strength of the opposition versus the ruling party and, second, by the degree of polarization between the two. To capture both, scholars of party systems distinguish dominant-authoritarian, dominant, non-dominant, and pulverized systems (Sartori 1976). All are multiparty systems but they vary greatly in the degree of party dominance and oppositional checks on the government and thus in the quality of democracy. Less polarized politics with a more equitable distribution of power gives the opposition more scope for democratic control of the executive and also makes institutional deadlock in parliament less likely. The founding elections—the first executive elections after transition— are a particular test for this relationship because this is when the roles of government and opposition are first assigned. Parties, even those with pro-democratic convictions, might disagree with unfavourable election results, and parties with anti-democratic ideals will see elections as merely a means to an end. Once elections are held, compliance with the outcome is a crucial factor that determines the establishment and stability of a democratic equilibrium. The losers will have to decide whether to accept the result or rebel against the nascent democracy (Höglund 2009). Thus, democracy depends on whether government and opposition comply with the rules of the electoral competition. Accordingly, a peaceful turnover of power, where an incumbent party loses elections and peacefully hands power over to another party, can be considered a strong indicator of democratic consolidation. It signals that a major political actor complies with the democratic rules of the game and values the prospects of winning the next elections higher than the odds of a rebellion. 2. Relationship between government and security forces With the security forces controlling the means of violence, they represent a considerable threat to the survival of democratic civilian

24

D. LAMBACH ET AL.

governments. Accordingly, all governments have a keen interest in stable civil–military relations. Tight civilian control of the security forces is crucial to prevent military coups or the emergence of so-called ‘praetorian regimes’ (Huntington 1991, pp. 231–253). Civilian control consists of two aspects: authority and oversight. Authority means that civilians make binding decisions over military matters. Oversight refers to the government’s ability to control and monitor whether the military fulfils its delegated functions as intended. However, as Finer (2009, p. 3) points out, civilian control over the military is not unique to democratic systems but can also be found in nonmilitary subtypes of authoritarian regimes. Thus, the second dimension of civilian control becomes important: the question how closely security forces are aligned with the government and the state. This is a question of professionalism (Huntington 1957) and the ethos of the military. Some security forces may seek political power, others may subscribe to political ideologies and be closely aligned with the governing regime, while yet others disavow any interference in civilian politics. In Perlmutter’s (1977, 1986) typology of soldiers, these would respectively be labelled the ‘praetorian’, the ‘revolutionary’, and the ‘professional’ model. And whatever their political stance, most militaries like to maintain a certain degree of autonomy and access to resources. Whether a democratic equilibrium can be established and maintained between the government and the security forces is thus also dependent on the military’s ethos and self-image. If the military corps is professional, a stable democratic equilibrium is easier to achieve. In contrast, the praetorian soldier seeks to retain his autonomy and represents a threat to democracy. While praetorian security forces may return to the barracks after democratic transition, their preference will be to overturn the democratic equilibrium through a military coup. Finally, the revolutionary soldier maintains close ties to their political faction and keeps it in power. In this setting it is the government’s dominance of the security forces that becomes a threat to democracy. With the coercive power of the security apparatus at its disposal, the incumbent government may expand its power and restrict institutionalized checks on the executive, repress the opposition, rig elections, and/or limit the provision of public services to its supporters instead of all citizens. Therefore, we assume that only civil– military (and civil–police) relations (Schmitter 1994) that are based on professionalism stabilize the democratic equilibrium. Praetorian relations create incentives for security forces to defect from democratic equilibrium

2

THEORY

25

by means of coup. Revolutionary civil–military relations create incentives for the incumbent elite to defect from democratic equilibrium by means of a self-coup, or autogolpe. 3. Relationship between government and citizens The vertical relation between the government and citizens is of central importance for a democracy since democratic governments are legitimized through elections. Thus, a democratic government has to show sufficient levels of responsiveness and accountability for a credible claim to represent the people, who are sovereign. Free, fair, and regular elections are one way to influence the government and to hold it accountable. Lobbying, petitioning, and protest are other options that are used by citizens to influence the course of government outside of election times. Non-democratic regimes are faced with similar challenges but work with different instruments. Well-organized authoritarian regimes use limited elections, petitions, a semi-free media, or clientelism to structure their interactions with society, thereby also achieving a modicum of responsiveness to public grievances. Less well-organized regimes will resort to heavy-handed repression or will try to ‘buy off’ disaffected parts of the population with occasional displays of magnanimity and largesse (Göbel 2011; Gerschewski 2013). Maintaining the democratic equilibrium between government and citizens is a challenge. A government is entrusted with enforcing democratic rules but is at the same time capable of transgressing these rules since it controls a powerful administrative and security apparatus. The democratic equilibrium therefore depends on there being limits to government which enforces rules but does not abuse power for its own advantage. Systems of accountability, institutional checks and balances, a vigilant civil society, and a free press are mechanisms through which this can be achieved, but an active citizenry with pro-democratic attitudes and a political culture that values participation and engagement are the foundation on which these mechanisms rest. From a rational perspective, a government is more likely to account for citizens’ demands if citizens can credibly threaten to punish it if it breaks the democratic compact (Weingast 1997). From a values perspective, a civic culture ‘produces’ political elites with prodemocratic attitudes through manifold processes of education and socialization (Almond and Verba 1963; Putnam 1993).

26

D. LAMBACH ET AL.

The other relationships—(4) between the opposition and citizens and (5) between the security forces and citizens —are of somewhat lesser importance for democratic consolidation. The relationship of citizens to the opposition is crucial for the embeddedness of elite regime outsiders in society. Having a working opposition is dependent on the opposition being able to pick up on public grievances that the government is not responsive to. During transition, the ability of opposition politicians to insert themselves into negotiations is strongly influenced by the degree of public backing they can claim to enjoy. The relationship between the security forces and citizens can also be considered part of civil–military relations since the population can actively contribute to civilian control of the military. Here, of great importance is how the security forces perceive themselves and their role. Are they ‘citizens in uniform’, or are they strictly highly professional forces obeying orders from the government? The relation between security forces and citizens determines how far the government can use the security forces as a repressive instrument against its own population (Croissant et al. 2018). In other words: if the government orders its soldiers to put down a demonstration by force, will the soldiers obey? Or might the security forces even, as Johnson and Thyne (2018) observe, use protests as cover for a coup d’état? To summarize, we would describe the characteristics of consolidated democracies in relational terms as follows: • The government enjoys civilian control over a military with a professional ethos. • Government and opposition parties adhere to democratic rules. There are frequent and peaceful changes of power after elections. • The government is responsive and accountable to citizens. Citizens are able to express their interests and grievances freely and use available avenues for political participation. • The security forces enjoy public trust and will defy government attempts to instrumentalize them against the citizenry and the opposition. • The opposition articulates public grievances that the government is not addressing and can credibly claim to represent part of the citizenry.

2

THEORY

27

Critical Junctures, Path Dependence, and Modes of Transition Democracies do not emerge fully formed. Instead, they need time to grow and consolidate. While emerging and consolidating democracies can—in theory—break down at any point of their evolution, these risks recede as democracies get older. This is due to a set of self-reinforcing and selfstabilizing factors that are described by the concept of path dependence. Broadly, path dependence means that decisions and events in the present are influenced by decisions and events in the past (historical dependence). More precisely, however, it describes how although several causal paths are available (causal possibility), some causal paths become less possible or impossible over time (closure) and that certain processes keep actors on the chosen track (constraints) (Bennett and Elman 2006). Examples of such constraints would be sunk costs, which discourage path shifting, or increasing returns that make the current path more attractive. Path dependence can also create lock-in situations (Page 2006) through, for example, network effects where a decision appears better in relation to other alternatives because a sufficient number of actors have already made the same choice. Through these mechanisms, deviation from the chosen path becomes less and less likely. Yet, paths can be changed at critical junctures. These are typically described as ‘formative or founding moments’ (Munck and Leff 1997, p. 343) or periods of ‘significant change’ (Collier and Collier 1991, p. 29). Critical junctures occur when stabilizing mechanisms fail or are overcome. They are characterized by a high level of contingency and uncertainty and the possibility of a new path being established. We consider political transitions to be critical junctures which set a political system on a new path. Accordingly, choices made during transition will have a lasting impact on subsequent political developments (Capoccia and Kelemen 2007, p. 341; Soifer 2012, pp. 1572–1573; Zielinski 2002). After transition, political systems, whether democratic or not, develop self-stabilizing features. Constitutional, political, and legal institutions provide barriers against irregular challenges or attempts to subvert or topple the regime. There is a popular demobilization after the heady days of change, thereby raising opportunity costs of defection or rebellion. As a regime consolidates, citizens become invested in the system even if they do not consider it just or legitimate.

28

D. LAMBACH ET AL.

We therefore argue that the conditions under which democracies emerge are crucial for their long-term health and viability. This is in line with previous research on modes of transition. For example, Huntington (1991) describes transitions during the third wave of democratization as either ‘transformations’, which are initiated by reform-oriented authoritarian leaders, ‘replacements’, where governments are overthrown by mass protests, or ‘transplacements’, where the government and the opposition negotiate the terms of transition. These different forms of transition matter for the long-term trajectory of democracy. In similar typologies, Linz (1978) distinguishes between reforma, ruptura, and ruptforma, while Share and Mainwaring (1984) speak of ‘transaction’, ‘breakdown’ and ‘extrication’. These typologies differentiate between various modes of transition according to the main drivers of the transition. In other words, transitions are initiated either by regime insiders, by the opposition or a mix thereof, such as through negotiations between moderates from both camps. These modes of transition have different long-term effects on the stability and quality of the subsequent democracy, for example through the form of democracy (presidential or parliamentary), the design of political institutions, and electoral systems (majoritarian or proportional). Thus, the question who is involved at which point of the transition process and who dominates these becomes crucial. Most transitologists have traditionally held that elite-led, negotiated transitions are favourable for the long-term stability of democracy because they involve the most powerful actors in the country and can enshrine elite bargains into political institutions (O’Donnell et al. 1986; Linz and Stepan 1978; Rustow 1970). But it is not just about who gets to participate but also about the means by which transition is achieved. Guo and Stradiotto (2014) focus on whether transition was initiated violently or peacefully and distinguish two subtypes of each variant. Violent modes of transition include (a) the collapse of the old regime through rebellion or coup d’état and (b) foreign intervention. Both of these are marked by the lack of a reform faction within the regime, hence transition occurs without negotiation. In peaceful transitions, there are (a) regime conversions where incumbents hold a power advantage over the opposition and (b) cooperative transitions where the opposition has the upper hand. Like Guo and Stradiotto, we assume that a peaceful transition is favourable for democratic transitions and the subsequent consolidation of democracy. However, we depart from their approach through three key assumptions. First, peaceful

2

THEORY

29

transitions do not only occur when the opposition is too weak to challenge the government in a military contest. Peaceful protest can be the strategic choice of a strong and diverse opposition. Second, the mode of transition is, accordingly, itself the outcome of a strategic decision made by the opposition, i.e. whether its strategies of resistance were appropriate and sufficiently powerful to bring down the regime. Third, the mode of transition influences not only the institutional structure of the new democracy but also the relationships between key actors.

Causal Assumptions and Mechanisms We expect that the mode of transition influences the prospects of democratic consolidation. Elite-led or pacted transitions (‘No resistance’, NoR) are deleterious for democratic consolidation. We agree with Karl’s critique of elite pacts in Latin America in which she argued that ‘the very modes of transition that appear to enhance initial survivability by limiting unpredictability may preclude the future democratic self-transformation’ (Karl 1990, p. 13). Such elite bargains demobilize social forces and exclude parts of the opposition while preserving the prerogatives of the military and protecting the veto positions of incumbent elites (Hagopian 1990). Under these circumstances, while there might be regular elections, they will typically not be allowed to meaningfully affect the elite bargain. There are few cases of VR leading to democracy. However, where such transitions succeed, VR has lasting, mostly negative consequences. Where civil war was concluded through peace negotiations, power-sharing arrangements preclude the evolution of meaningful party competition. Such situations are characterized by low trust between former warring sides, even as those same sides continue to dominate the political scene. Parties that emerge from rebel movements are typically marked by an authoritarian style of leadership and limited intra-party democracy (Lyons 2016). When the rebels actually manage to overthrow the authoritarian government, the emerging democracy will be hampered by a dominant party with an authoritarian internal culture (Dorman 2006). NVR gives us the best odds of getting a mode of transition that is favourable to democracy in the long term. NVR advances the stability and quality of all three crucial relationships necessary to maintain democratic equilibrium—civil–military relations marked by professionalism and trust, free and constructive democratic competition between government and opposition, and state–society relations that are marked by an active civil

30

D. LAMBACH ET AL.

society holding the government accountable. This is down to three main mechanisms: (1) NVR is better at ‘levelling the political playing field’ after the transition by creating a political environment more conducive to multiparty competition; (2) NVR shifts the political culture towards more democratic attitudes and civic activism, providing a powerful symbol for future activists to reference; and (3) NVR avoids the ‘praetorian problem’ of the military getting too involved in politics.6 Levelling the Political Playing Field We argue that NVR-induced transitions lead to a more constructive and equitable relationship between government and opposition by levelling the playing field for electoral competition between political parties. In other words, NVR balances the electoral chances among different political parties since the dominance of the former elites is broken without creating a new organization able to dominate the political sphere. As a result, parties, even if they lose an election, value the prospects of winning future elections higher than disputing election results and attempting a rebellion. Since NVR movements tend to be very diverse and are mostly united by their goal to unseat an authoritarian regime, these movements do not constitute a homogeneous bloc able to take over state power (Carter et al. 1970; della Porta 2014). As a result, post-transition power relations are balanced at the time when democracy is being institutionalized, and this balance of power will be ‘baked into’ the emerging democratic institutions. As outlined above, transitologists argue that pacted transitions enable elites to control the transition so that ‘institutions are custommade for a particular person, party or alliance’ (Przeworski 1991, p. 82). In contrast, if no single faction controls the transition process, all actors involved will be more likely to opt for a more inclusive and pluralist institutional design, similar to the Rawlsian ‘veil of ignorance’ (Rawls 1999, p. 118). The protest movement will disintegrate into various factions, proto-parties, and demobilized citizens (Grodsky 2012), but by discrediting and delegitimizing the previous ruling party, NVR will have ‘levelled the political playing field’ (Bayer 2018) for new parties which profit from 6 We also checked for a range of hypotheses drawn from the existing literature but could find little evidence that they had a systematic causal effect beyond isolated instances (see Chapter 5).

2

THEORY

31

the liberal institutional design. In contrast, elite-led transitions privilege parties connected to the old regime and disadvantage or co-opt the opposition. VR typically grants the armed movement a dominant position in the post-transition party system. In NVR cases, the greater dispersal of power among parties and the resulting need for coalition-building among them stabilizes and deepens democracy in the long run. Advancing a Democratic Political Culture NVR-induced transition also creates favourable conditions for civil society. According to Gene Sharp (2005, p. 424), nonviolent struggle has lasting effects on the activists themselves as well as on society at large. This is in line with findings from Mazumder (2018) that the impact of protest activity on political attitudes can be felt for generations. NVR builds on individual and collective action, working only when people become active and rethink their previously submissive and compliant behaviour. This process of mobilization is described by Yousef Abduljalil, a teacher from Sanaa (Yemen), who describes the effects of participating in protests against the regime of Ali Abdullah Saleh: ‘Now life is not what it used to be – there is no fear, no despair, no submission or surrender […] It would be impossible to return to the dark past and equally impossible to accept ever again a regime like that of Ali Abdullah Saleh’ (Abduljalil 2015, p. 180). Acting in concert with others enhances the political efficacy of the participating individuals (van Stekelenburg and Klandermans 2013, p. 892). We also assume that people who have previously fought for democracy are more committed to democratic values (Sharp 2008, p. 53) and willing to hold elites accountable. NVR thus advances a democratic political culture.7 It also provides a powerful symbol for future acts of resistance against any authoritarian backlash, especially if earlier demands remain unfulfilled. This democratic political culture is sustained by civil society organizations whose relations with the government will be different than they were before the transition (Adler and Webster 1995, p. 77). 7 It is possible that the causal relationship works both ways. Contrary to earlier assumptions (Almond and Verba 1963), more recent research finds that a democratic political culture is often already present when democratic transition occurs. For instance, Welzel and Inglehart argue that in countries of the former Eastern Bloc like Poland, Hungary, and Estonia, high intrinsic support for democracy had emerged prior to the transition to democracy. Thus, they argue, democratic values within the population made democratization possible, not the other way around (Welzel and Inglehart 2009, p. 138).

32

D. LAMBACH ET AL.

In sum, NVR-induced transition leaves behind democratic citizens who can credibly threaten protest, non-cooperation, and resistance against the incumbent government if it abuses power and does not comply with democratic rules. Citizens under an elite-led transition cannot send such a strong signal of credible commitment to protest because there is no history of collective action to refer to. After violent transitions, civil society is often marginalized and only veterans of the armed movement can send credible signals of collective action. These credible signals of former veterans, however, often take the form of serious threats to take up arms again and result in a de facto veto power which is not conducive to liberal democracy. Avoiding the Praetorian Problem Democracies emerging from an NVR-induced transition are less hampered than other democracies by what Huntington has called the ‘praetorian problem’—the question of how to limit political influence of the military (Huntington 1991, p. 231). This is achieved by changing relations between the security forces and the government and between security forces and citizens, respectively. According to Helvey ‘[any] regime will rely on some pillars of support’ (Helvey 2004, p. 9). The goal of NVR is to separate the regime from its pillars of support by undermining the security forces’ willingness to back the regime any further. The military—the ‘trump card’ (Helvey 2004, p. 11) of the regime—thus plays a pivotal role in making democratic transition possible. Soldiers typically live in barracks with little contact to the civilian population and are often seen as more reliable than local police forces when it comes to using violence against civilians. Nonviolent action aims to win over the security forces by (a) establishing direct contact between citizens and soldiers, (b) providing professional soldiers with a future in a democratic system, and (c) avoiding any violent threats that could lead to a ‘rally round the flag’ effect among soldiers. Once the military has sided with ‘the people’ by staying neutral or even defecting during the transition phase, it has fewer incentives to intervene in politics in the future or to serve as a repressive organ of the new elites. Thus, we expect the relation between the security forces and the broader population to be at least neutral following NVRinduced democratization. In contrast, violent resistance movements usually stack the security apparatus with movement veterans, creating a loyalist and politicized army, police, and other security forces. This imbalance

2

THEORY

33

in civil–military relations creates incentives for the ruling elite to abuse its power and deviate from the democratic equilibrium. In cases of eliteled transition, the security forces retain their close connection to the old regime, which also raises doubts about their commitment to democracy. To conclude, we adopt a relational approach to explain the stability and quality of democracy. The fate of democracy depends on the relations between its four key actors, namely the government, the opposition, the security forces, and citizens. We argue that contrary to elite-led and violent transitions, NVR has a positive effect on both the stability and the quality of democracy through three key mechanisms. First, if larger parts of the population are involved in the transition, the political environment and the political institutions are more conducive to multiparty competition. Second, NVR has a lasting effect on the key relation between the broader population and political and military elites. It empowers citizens and shifts the political culture towards more democratic attitudes. Further, it provides a powerful symbol for future activists to refer back to and remobilize the population to defend or advance democracy. Last but not least, NVR avoids the ‘praetorian problem’ of the military getting overly involved in politics. We test and further explain these propositions in Chapters 3 and 4.

References Abduljalil, Y. (2015). Killing the Rose but Not the Spring. In A. Al Saleh (Ed.), Voices of the Arab Spring: Personal Stories from the Arab Revolutions (pp. 175– 180). New York: Columbia University Press. Ackerman, P., & Karatnycky, A. (2005). How Freedom Is Won: From Civic Resistance to Durable Democracy. New York: Freedom House. Adler, G., & Webster, E. (1995). Challenging Transition Theory: The Labor Movement, Radical Reform, and Transition to Democracy in South Africa. Politics & Society, 23(1), 75–106. Almond, G. A., & Verba, S. (1963). The Civic Culture. Political Attitudes and Democracy in Five Nations. Princeton, NJ: Princeton University Press. Barma, N. H. (2016). The Peacebuilding Puzzle: Political Order in Post-Conflict States. Cambridge: Cambridge University Press. Bayer, M. (2018). The Democratizing Effect of Nonviolent Resistance: How Nonviolent Resistance Featured Democratic Consolidation in Benin (Working Paper Vol. 3). Basel: Swisspeace.

34

D. LAMBACH ET AL.

Bayer, M., Bethke, F. S., & Lambach, D. (2016). The Democratic Dividend of Nonviolent Resistance. Journal of Peace Research, 53(6), 758–771. https:// doi.org/10.1177/0022343316658090. Beetham, D. (1994). Conditions for Democratic Consolidation. Review of African Political Economy, 21(60), 157–172. Bennett, A., & Elman, C. (2006). Complex Causal Relations and Case Study Methods: The Example of Path Dependence. Political Analysis, 14(3), 250– 267. Bermeo, N. (2003). Ordinary People in Extraordinary Times: The Citizenry and the Breakdown of Democracy. Princeton: Princeton University Press. Bethke, F. S. (2017). Nonviolent Resistance and Peaceful Turnover of Power. Peace Economics, Peace Science and Public Policy, 23(4), 1–5. https://doi. org/10.1515/peps-2017-0022. Bethke, F. S., & Pinckney, J. (2019). Non-violent Resistance and the Quality of Democracy. Conflict Management and Peace Science. https://doi.org/10. 1177/0738894219855918. Boix, C., Miller, M., & Rosato, S. (2013). A Complete Data Set of Political Regimes, 1800–2007. Comparative Political Studies, 46(12), 1523–1554. https://doi.org/10.1177/0010414012463905. Burton, M., Gunther, R., & Higley, J. (1991). Introduction: Elite Transformations and Democratic Regimes. In J. Higley & R. Gunther (Eds.), Elites and Democratic Consolidation in Latin America and Southern Europe (pp. 1–38). Cambridge: Cambridge University Press. Capoccia, G., & Kelemen, R. D. (2007). The Study of Critical Junctures: Theory, Narrative, and Counterfactuals in Historical Institutionalism. World Politics, 59(3), 341–369. Carter, A., Hoggett, D., & Roberts, A. (1970). Non-violent Action: A Selected Bibliography. London: Housmans. Cervellati, M., Fortunato, P., & Sunde, U. (2014). Violence During Democratization and the Quality of Democratic Institutions. European Economic Review, 66, 226–247. https://doi.org/10.1016/j.euroecorev.2013.12.001. Cheibub, J. A., Gandhi, J., & Vreeland, J. R. (2010). Democracy and Dictatorship Revisited. Public Choice, 143(1–2), 67–101. Chenoweth, E., & Stephan, M. J. (2011). Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. New York: Columbia University Press. Collier, D., & Adcock, R. (1999). Democracy and Dichotomies: A Pragmatic Approach to Choices About Concepts. Annual Review of Political Science, 2, 537–565. Collier, R. B., & Collier, D. (1991). Shaping the Political Arena. Princeton: Princeton University Press. Coppedge, M., Gerring, J., Lindberg, S. I., Skaaning, S.-E., & Teorell, J. (2015). V-Dem Codebook v5. Gothenburg: Varieties of Democracy (V-Dem) Project.

2

THEORY

35

Croissant, A., Kuehn, D., & Eschenauer, T. (2018). The “Dictator’s Endgame”: Explaining Military Behavior in Nonviolent Anti-Incumbent Mass Protests. Democracy and Security, 14(2), 174–199. https://doi.org/ 10.1080/17419166.2017.1423471. de la Boétie, E. (1975/1576). The Politics of Obedience: The Discourse of Voluntary Servitude. New York: Free Life Editions. della Porta, D. (2014). Mobilizing for Democracy: Comparing 1989 and 2011. Oxford: Oxford University Press. della Porta, D. (2016). Where Did the Revolution Go? Contentious Politics and the Quality of Democracy. Cambridge: Cambridge University Press. Dorman, S. R. (2006). Post-Liberation Politics in Africa: Examining the Political Legacy of Struggle. Third World Quarterly, 27 (6), 1085–1101. Dudouet, V. (2008). Nonviolent Resistance and Conflict Transformation in Power Asymmetries. In Berghof Handbook for Conflict Transformation. Berlin: Berghof Foundation. Encarnación, O. G. (2000). Beyond Transitions: The Politics of Democratic Consolidation. Comparative Politics, 32(4), 479–498. Fearon, J. D. (2011). Self-Enforcing Democracy. The Quarterly Journal of Economics, 126(4), 1661–1708. https://doi.org/10.1093/qje/qjr038. Finer, S. E. (2009). Men on Horseback. New Brunswick: Transaction Publishers. Galtung, J. (1996). Peace by Peaceful Means: Peace and Conflict, Development and Civilization. London: Sage. Gasiorowski, M. J., & Power, T. J. (1998). The Structural Determinants of Democratic Consolidation: Evidence from the Third World. Comparative Political Studies, 31(6), 740–771. Geddes, B., Wright, J., & Frantz, E. (2014). Autocratic Breakdown and Regime Transitions: A New Data Set. Perspectives on Politics, 12(2), 313–331. Gerschewski, J. (2013). The Three Pillars of Stability: Legitimation, Repression, and Co-optation in Autocratic Regimes. Democratization, 20(1), 13–38. https://doi.org/10.1080/13510347.2013.738860. Göbel, C. (2011). Authoritarian Consolidation. European Political Science, 10(2), 176–190. Grodsky, B. K. (2012). Social Movements and the New State: The Fate of ProDemocracy Organizations When Democracy Is Won. Stanford: Stanford University Press. Guo, S., & Stradiotto, G. A. (2014). Democratic Transitions: Modes and Outcomes. New York: Routledge. Hagopian, F. (1990). “Democracy by Undemocratic Means”?: Elites, Political Pacts, and Regime Transition in Brazil. Comparative Political Studies, 23(2), 147–170. https://doi.org/10.1177/0010414090023002001. Helvey, R. L. (2004). On Strategic Nonviolent Conflict: Thinking About the Fundamentals. Boston: Albert Einstein Institution.

36

D. LAMBACH ET AL.

Higley, J., & Burton, M. (1998). Elite Settlements and the Taming of Politics. Government and Opposition, 33(1), 98–115. https://doi.org/10.1111/ j.1477-7053.1998.tb00785.x. Höglund, K. (2009). Electoral Violence in Conflict-Ridden Societies: Concepts, Causes, and Consequences. Terrorism and Political Violence, 21(3), 412–427. Huntington, S. P. (1957). The Soldier and the State: The Theory and Politics of Civil-Military Relations. Oxford: Oxford University Press. Huntington, S. P. (1968). Political Order in Changing Societies. New Haven: Yale University Press. Huntington, S. P. (1991). The Third Wave: Democratization in the Late Twentieth Century. Norman: University of Oklahoma Press. Johnson, J., & Thyne, C. L. (2018). Squeaky Wheels and Troop Loyalty: How Domestic Protests Influence Coups d’état, 1951–2005. Journal of Conflict Resolution, 62(3), 597–625. https://doi.org/10.1177/0022002716654742. Karl, T. L. (1990). Dilemmas of Democratization in Latin America. Comparative Politics, 23(1), 1–21. https://doi.org/10.2307/422302. Khan, M. H. (1995). State Failure in Weak States: A Critique of New Institutionalist Explanations. In J. Harriss, J. Hunter, & C. M. Lewis (Eds.), The New Institutional Economics and Third World Development (pp. 71–86). London, New York: Routledge. Levi, M. (1990). A Logic of Institutional Change. In M. Levi & K. S. Cook (Eds.), The Limits of Rationality (pp. 402–418). Chicago: Chicago University Press. Linz, J. J. (1978). Crisis, Breakdown and Reequilibration. In J. J. Linz & A. Stepan (Eds.), The Breakdown of Democratic Regimes: Crisis, Breakdown and Reequilibration (pp. 14–124). Baltimore: Johns Hopkins University Press. Linz, J. J., & Stepan, A. (Eds.). (1978). The Breakdown of Democratic Regimes— Crisis, Breakdown and Reequilibration. Baltimore: Johns Hopkins University Press. Linz, J. J., & Stepan, A. (1996a). Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe. Baltimore, London: Johns Hopkins University Press. Linz, J. J., & Stepan, A. (1996b). Toward Consolidated Democracies. Journal of Democracy, 7 (2), 14–33. Lyons, T. (2016). From Victorious Rebels to Strong Authoritarian Parties: Prospects for Post-war Democratization. Democratization, 23(6), 1026–1041. https://doi.org/10.1080/13510347.2016.1168404. Mazumder, S. (2018). The Persistent Effect of U.S. Civil Rights Protests on Political Attitudes. American Journal of Political Science, 62(4), 922–935. https://doi.org/10.1111/ajps.12384. Merkel, W. (2004). Embedded and Defective Democracies. Democratization, 11(5), 33–58. https://doi.org/10.1080/13510340412331304598.

2

THEORY

37

Merkel, W. (2010). Systemtransformation: Eine Einführung in die Theorie und Empirie der Transformationsforschung (2nd ed.). Wiesbaden: VS. Munck, G. L., & Leff, C. S. (1997). Modes of Transition and Democratization: South America and Eastern Europe in Comparative Perspective. Comparative Politics, 29(3), 343–362. O’Donnell, G. (1996). Illusions About Consolidation. Journal of Democracy, 7 (2), 34–51. O’Donnell, G., Schmitter, P. C., & Whitehead, L. (Eds.). (1986). Transition from Authoritarian Rule: Comparative Perspectives. Baltimore: Johns Hopkins University Press. Page, S. E. (2006). Path Dependence. Quarterly Journal of Political Science, 1(1), 87–115. Perlmutter, A. (1977). The Military and Politics in Modern Times: On Professionals, Praetorians and Revolutionary Soldiers. New Haven: Yale University Press. Perlmutter, A. (1986). The Military and Politics in Modern Times: A Decade Later. Journal of Strategic Studies, 9(1), 5–15. https://doi.org/10.1080/ 01402398608437245. Pinckney, J. (2018). When Civil Resistance Succeeds: Building Democracy After Popular Nonviolent Uprisings. Washington, DC: International Center on Nonviolent Conflict. Power, T. J., & Gasiorowski, M. J. (1997). Institutional Design and Democratic Consolidation in the Third World. Comparative Political Studies, 30(2), 123– 155. Przeworski, A. (1991). Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America. Cambridge: Cambridge University Press. Przeworski, A. (2005). Democracy as an Equilibrium. Public Choice, 123(3–4), 253–273. Putnam, R. D. (1993). Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press. Rawls, J. (1999). A Theory of Justice (Rev. ed.). Cambridge: Harvard University Press. Rueschemeyer, D., Huber, E., & Stephens, J. D. (1992). Capitalist Development and Democracy. Chicago: University of Chicago Press. Rustow, D. A. (1970). Transition to Democracy: Toward a Dynamic Model. Comparative Politics, 2(3), 337–363. Sartori, G. (1976). Parties and Party Systems: A Framework for Analysis. Cambridge: Cambridge University Press. Scharpf, F. W. (1997). Games Real Actors Play: Actor-Centered Institutionalism in Policy Research. Boulder: Westview.

38

D. LAMBACH ET AL.

Schedler, A. (1998). What Is Democratic Consolidation? Journal of Democracy, 9(2), 91–107. Schedler, A. (2001). Measuring Democratic Consolidation. Studies in Comparative International Development, 36(1), 66–92. https://doi.org/10.1007/ bf02687585. Schimmelfennig, F. (2000). International Socialization in the New Europe: Rational Action in an Institutional Environment. European Journal of International Relations, 6(1), 109–139. https://doi.org/10.1177/ 1354066100006001005. Schmitter, P. C. (1994). Dangers and Dilemmas of Democracy. Journal of Democracy, 5(2), 57–74. Schneider, B. R. (1995). Democratic Consolidations: Some Broad Comparisons and Sweeping Arguments. Latin American Research Review, 30(2), 215–234. Share, D., & Mainwaring, S. (1984). Transitions Through Transactions: Democratization in Brazil and Spain. In W. A. Selcher (Ed.), Political Liberalization in Brazil: Dynamics, Dilemmas, and Future Prospects (pp. 175–216). Boulder: Westview. Sharp, G. (1973). The Politics of Nonviolent Action (vols. 3). Boston: Porter Sargent. Sharp, G. (2005). Waging Nonviolent Struggle. Boston: Albert Einstein Institution. Sharp, G. (2008). From Dictatorship to Democracy: A Conceptual Framework for Liberation. Boston: Albert Einstein Institution. Siaroff, A. (2008). Comparing Political Regimes: A Thematic Introduction to Comparative Politics. Toronto: University of Toronto Press. Skaaning, S.-E. (2006). Political Regimes and Their Changes: A Conceptual Framework (CDDRL Working Papers) Stanford University: Freeman Spogli Institute for International Studies. Soifer, H. D. (2012). The Causal Logic of Critical Junctures. Comparative Political Studies, 45(12), 1572–1597. https://doi.org/10.1177/ 0010414012463902. Svolik, M. W. (2015). Which Democracies Will Last? Coups, Incumbent Takeovers, and the Dynamic of Democratic Consolidation. British Journal of Political Science, 45(4), 715–738. https://doi.org/10.1017/ S0007123413000550. Ulfelder, J. (2010). Dilemmas of Democratic Consolidations: A Game-theory Approach. Boulder, CO: First Forum Press. van Stekelenburg, J., & Klandermans, B. (2013). The Social Psychology of Protest. Current Sociology, 61(5–6), 886–905. https://doi.org/10.1177/ 0011392113479314. Vinthagen, S. (2015). A Theory of Nonviolent Action: How Civil Resistance Works. London: Zed Books.

2

THEORY

39

Weingast, B. R. (1997). The Political Foundations of Democracy and the Rule of the Law. American Political Science Review, 91(2), 245–263. Welzel, C. (2007). Are Levels of Democracy Affected by Mass Attitudes? Testing Attainment and Sustainment Effects on Democracy. International Political Science Review/Revue Internationale de Science Politique, 28(4), 397–424. Welzel, C., & Inglehart, R. F. (2009). Political Culture, Mass Beliefs and Value Change. In C. W. Haerpfer, P. Bernhagen, R. F. Inglehart, & C. Welzel (Eds.), Democratization (pp. 127–144). Oxford: Oxford University Press. Zielinski, J. (2002). Translating Social Cleavages into Party Systems: The Significance of New Democracies. World Politics, 54(2), 184–211.

CHAPTER 3

Statistical Analysis

In this chapter, we evaluate the effect of nonviolent resistance (NVR) on democratic consolidation for a comprehensive sample of historical cases using quantitative data and statistical analysis. We compare the consolidation of democratic regimes where transition was achieved by means of NVR to regimes without this feature (i.e. cases of violent transition and elite-led transitions). As discussed in Chapter 2, the concept of democratic consolidation is ambiguous and contested (Schneider 1995). It refers both to the condition that democracy is completed on various dimensions and a return to autocracy is virtually impossible and to the process of consolidating democracy after transition. To address this ambiguity, we evaluate three different aspects of democratic consolidation, which emphasize different properties of the concept: (1) democratic survival, (2) the two-turnover test, and (3) the quality of democracy. The first conceptualization refers to the resilience of democracy. Once a democratic transition occurs, democratic regimes may either endure, which we refer to as democratic survival , or they break down at some point in time. Using the statistical method of survival analysis, we evaluate whether democratic regimes for which transition was brought about by an NVR campaign survive longer than democratic regimes without this feature. The second conceptualization of democratic consolidation uses the two-turnover test proposed by Huntington (1968). Here, we specify an

© The Author(s) 2020 D. Lambach et al., Nonviolent Resistance and Democratic Consolidation, https://doi.org/10.1007/978-3-030-39371-7_3

41

42

D. LAMBACH ET AL.

event—a democracy experiencing a second peaceful turnover of power— as a necessary condition for democratic consolidation. The test of two turnovers of power implies that the party that won the founding election during democratic transition loses a subsequent election and peacefully hands over power to the party that won (first turnover). If this other party again peacefully transfers power to another party after another subsequent election, two turnovers of power have taken place. From this perspective, we evaluate whether NVR-induced transitions lead to democratic regimes that are more likely to pass the two-turnover test relative to democratic regimes that were not induced by NVR. As a third conceptualization of democratic consolidation, we look at the development of democratic quality after transition. Thus, we evaluate to what extent democratic regimes are able to ‘complete’ democracy, i.e. how close they get to an ideal type of democracy. Whereas the concept of democratic transition is based on meeting necessary conditions, such as a minimum level of suffrage and free and fair elections, the quality of the resulting democratic regime is a matter of degree. For instance, a free and fair founding election is a necessary condition for democratic transition but regimes show different degrees of freedom and fairness at the ballot. Furthermore, suffrage can be expanded from a limited franchise (for example with only adult men allowed to vote) to a universal suffrage. We explore whether NVR-induced democratic regimes achieve higher levels of democratic quality and improve more substantially in their democratic quality relative to regimes where the transition from autocracy to democratic rule occurred without NVR. The rest of this chapter is structured as follows. The next section covers our research design in more detail, including information on the data sources used in the empirical analysis as well as the statistical procedures and methods used to evaluate the hypotheses stated above. In section “The Effect of NVR on Democratic Survival” we describe the findings related to the first hypothesis on democratic survival. Sections “The Effect of NVR on Achieving Peaceful Turnovers of Power” and “The Effect of NVR on Democratic Quality” present the results related to two turnovers of power and democratic quality, respectively. In the final section, we summarize the results of the empirical analysis, discuss our findings, and highlight areas for further research.

3

STATISTICAL ANALYSIS

43

Research Design Measuring Resistance Campaigns and Democratic Transitions For the quantitative analysis, we created a dataset that combines information on democratic regimes with information on the presence of NVR during the transition from which they emerged. The data on political regimes is provided by Ulfelder (2012). This dataset contains information on episodes of democracy and autocracy in countries worldwide for the period 1955–2010. It specifies whether political regimes in countries across the world meet the minimum requirements to be considered democracies as discussed in the preceding chapter. If political regimes do not completely fulfil these criteria, they are considered autocracies. Furthermore, the data provided by Ulfelder (2012) also identifies transition events where political regimes switch between the categories of autocracy and democracy.1 To measure the influence of NVR on the democratic transition process, we use the Non-Violent and Violent Conflict Outcome database (NAVCO) compiled by Chenoweth and Lewis (2013), which provides information on the occurrence, goals, and success of resistance campaigns. To combine the data on political regimes and transitions with the data on resistance campaigns, we applied original coding using case-specific material and historical sources. Multiple coders evaluated the relevance of resistance campaigns for each transition event. We coded a campaign as relevant for the transition process if it was present in the year of the transition or the year before the transition, if it aimed at political change in the incumbent autocratic regime, and if it seemed to have a causal effect on the transition.2 Our final coding distinguishes between (1) regimes whose transition process was induced without a resistance campaign, (2) regimes whose transition process was induced by a violent resistance campaign, and (3) regimes whose transition process was induced by a NVR campaign. The first category essentially refers to elite-led democratic transitions, where resistance campaigns did not substantially affect political developments. A prominent example of this kind of transition is Spain, where King Juan 1 We only consider cases of endogenous democratic transition. Cases of foreign powers imposing democratization and cases where democratization occurred as part of decolonization are not included in the analysis. 2 Note that we excluded campaigns whose goals were secession or policy change.

44

D. LAMBACH ET AL.

Table 3.1 Frequency of resistance campaigns during transitions

No resistance campaign Violent resistance campaign Nonviolent resistance campaign Total

Freq.

Percent

66 9 37 112

58.93 8.03 33.04 100

Carlos and reform-oriented civilian elites were the main drivers of democratization after the death of the dictator Franco in 1975. The second category includes cases of democratization where a violent campaign or rebel group caused democratic transition. For instance, violent insurgencies were instrumental for democratization in El Salvador and Guatemala in the 1980s. While these transitions may also have involved civil forms of resistance, violent campaigns were the main drivers of democratization. Finally, the third category captures transitions that were induced by an NVR campaign. Most prominently, the Solidarity campaign in Poland in 1989 or the Carnation Revolution of 1976 in Portugal brought about democratization predominantly by means of civil resistance.3 Table 3.1 describes the frequency distribution of these categories in our global sample of cases. It shows that few regimes in our sample experienced violent resistance campaigns during their transition phase. More than half of the sample consists of regimes that emerged without any resistance campaign and about one-third were regimes established via a NVR campaign. Given the rarity of democratic transitions driven by violent resistance, we conducted most of our empirical analysis using a combined category for transitions that were induced without a resistance campaign or by a violent resistance campaign. Thus, in most analyses we only distinguish whether the transition was induced by an NVR campaign or not. To ensure the robustness of our findings, we employed two alternative datasets for the specification of transition events. Each analysis described in this book was conducted not only with the data provided by Ulfelder (2012) but also repeated with alternative datasets compiled by Boix et al.

3 These are, of course, ideal-types. NAVCO distinguishes campaigns as ‘primarily’ violent or nonviolent so the differences between NVR and VR may well be a matter of degree.

3

STATISTICAL ANALYSIS

45

(2013) and Geddes et al. (2014).4 This procedure ensures that our findings are not driven by specific measurement choices about what constitutes democracy, autocracy, and transition events. Alternative Explanations for Democratic Consolidation To account for alternative explanations of democratic consolidation, we employ multiple additional covariates in our statistical models. In general, we focus on potential confounding variables correlated with both the occurrence of NVR and democratic consolidation and so-called prognostic factors—pre-treatment measures of the outcome of interest. Specifically, we employ the following covariates identified as most important in previous studies on democratization and NVR (Bayer et al. 2016; Chenoweth and Stephan 2011; Guo and Stradiotto 2014; Teorell 2010): GDP per capita, military legacy, previous instability, proportion of neighbouring democracies, and urbanization. We consider GDP per capita as a proxy measure of economic development which research on modernization theory has found to be an important explanatory factor for democratization, democratic survival and democratic quality (Boix and Stokes 2003). Moreover, resistance campaigns may be more likely to occur when economic grievances are more prevalent among the population (Celestino and Gleditsch 2013; Chenoweth and Lewis 2013; Chenoweth and Stephan 2011; Haggard and Kaufman 2016). Our variable measuring the level of GDP per capita uses an updated version of the ‘Expanded Trade and GDP Data’ compiled by Kristian Skrede Gleditsch (2002), transformed using natural logarithms. 4 We opted for the Ulfelder (2012) data for three reasons. First, it explicitly specifies transition events while the other two datasets are mainly interested in specifying the presence of democratic and autocratic regimes. Moreover, the Ulfelder data contains detailed descriptions of transition events which makes the coding of NVR influence more accurate and valid. Given our focus on transition events, we prefer the Ulfelder dataset as our main measure for political regimes. Second, the Ulfelder data uses a less demanding definition of democratization and therefore includes many cases not covered by the other two datasets. Third, in the subsequent empirical analysis we utilize matching procedures which try to optimize balance of covariance to advance the causal interpretation of the estimated effects of NVR. Among the three datasets used in the analysis, we achieved the best covariate balance in the Ulfelder dataset. Complete lists of cases included in all three datasets along with the respective coding decisions on NVR influence on democratic transitions are provided in the appendix section A1.

46

D. LAMBACH ET AL.

Military legacy is a binary variable indicating whether the pre-transition autocratic regime was a military regime, as coded by Geddes et al. (2014). We consider military legacy as a prognostic factor predetermining the outcome variable (i.e. democratic consolidation). Previous studies have shown that in cases of democratic transition from a military regime there is an increased risk of a coup attempt in the future, which impedes democratic survival and democratic quality (Cheibub 2007). A history of political instability is potentially related to both the onset of resistance campaigns (Chenoweth and Lewis 2013) and difficulties in democratic political development after transition (Boix and Stokes 2003). We measure previous instability by counting the number of regime changes from 1900 until the transition in question. Empirical studies found that democratic development is also affected by international and regional factors. Democratic quality is generally higher in a geographic environment where democracy is widespread— that is, where most of the neighbouring countries are democratic states. Neighbouring countries serve as role models and thereby trigger the diffusion and spillover of democratic ideas and norms (Gleditsch and Ward 2006). Thus, we consider the regional democratic environment of a state as a prognostic factor. To measure how widespread democracy is in a regime’s geographic environment, we use the variable neighbouring democracies, which is the proportion of states in the region that are democratic.5 Modernization theory considers urbanization to be a social prerequisite for democracy (Lipset 1959), indicating that citizens develop liberal preferences and beliefs more quickly in urban settings than in rural society. Moreover, urbanization facilitates the mobilization of participants in resistance campaigns by mitigating communication and coordination problems (Chenoweth and Stephan 2011). Urbanization is defined as the share of the total population living in cities with more than 100,000 inhabitants. For measurement, we use the National Material Capabilities dataset version 4.0 (Singer 1987). For similar reasons, we include a measure of the total population of the country using the same data source. Depending on the statistical model, we measure covariates either during

5 We use a tenfold politico-geographic classification of world regions from the V-DEM dataset (Coppedge et al. 2015).

3

STATISTICAL ANALYSIS

47

the year before the transition occurred or consider them as time-varying, i.e. measuring them for the years before and after the transition.6 Estimation Procedures In the empirical analysis that follows, we estimate the effect of NVRinduced transition on three different outcome measures of democratic consolidation: (1) democratic survival, (2) accomplishing the twoturnover test, and (3) quality of democracy. While each of these outcome measures requires different estimation procedures, our general empirical strategy is the same for each one, proceeding in three steps. First, we use plots to describe the respective outcome variable for regimes with and without NVR-induced transition graphically. We can thereby visualize any obvious differences between regimes with different transition processes. However, this procedure, though appealing in its simplicity, has the disadvantage that we are unable to account for confounding factors that possibly drive the relationship between NVR and democratic consolidation. Second, we use a regression-based approach that allows us to account for the alternative explanations for democratic consolidation and/or factors that could be confounding the relationship between NVR and democratic consolidation. Furthermore, regression models are able to estimate the average size of the effect of NVR-induced transitions on the respective measure of democratic consolidation. However, regression models suffer from the disadvantage of imposing some very restrictive modelling assumptions (Ho et al. 2007; Aronow and Samii 2016).7 Third, to account for some of these issues related to regression models, we employ matching methods. Matching is a procedure of pre-processing data to create a balanced dataset in which the treatment and control group are as similar as possible (Ho et al. 2007). Cases of NVR-induced transition are matched with cases without NVR on a set of baseline characteristics. Based on matching procedures we discard those cases without NVR-induced transition for which no comparable match exists among 6 Descriptive statistics for all variables are reported in the appendix section A2. 7 Most frequently scholars point to issues of extrapolation and model dependence.

Extrapolation refers to the assumption that the results generalize beyond the range of the available data. Model dependence means that the functional form of the regression model parameters impacts the estimated effects.

48

D. LAMBACH ET AL.

the NVR cases. Thus, we adjust the data in a way that for each NVR case a most similar non-NVR case exists. The basic idea is that if two cases are sufficiently similar on observed characteristics but differ in terms of NVR influence then the selection process of NVR influence can be considered to be ‘as good as random’ (Sekhon 2009, p. 495). We use three different matching schemes to create samples, where the treated cases (NVRinduced transitions) and the control cases (transition without NVR) are as similar as possible with regard to the covariates described in the previous subsection.8 Accordingly, we implement each aspect of the analysis for three different regime datasets using the full sample of cases and the three matched samples.

The Effect of NVR on Democratic Survival As discussed before, democratic survival refers to the duration a democratic regime remains in place after transition until a democratic breakdown occurs. Democracies typically break down via either an executive coup, where a democratically elected government or a faction of the government extends its rule via unconstitutional means and begins to govern autocratically, or a (military) coup d’état, where elites and/or actors from within the armed forces depose the elected government. Occasionally, democratic breakdown also involves popular rebellion, where the elected government is toppled by the masses (Ulfelder 2010, p. 3). The Ulfelder (2012) data, which we use to measure democratic survival and the timing of democratic transitions, contains 112 democratic regimes. Of these, 69 experienced a democratic breakdown. Accordingly, 43 regimes are right-censored, meaning that they did not experience a failure event until the end of the year 2010. Because of this characteristic of the data, we use survival analysis estimators to analyse the effect of NVR-induced transitions on democratic survival. Figure 3.1 shows the survivor functions for democratic regimes induced by NVR and for those without this feature. For each post-transition year, Fig. 3.1 shows the proportion of democratic regimes that survived up to that point. It is evident that regimes which experienced NVR during the transition phase on average survive longer than regimes without this feature. Regimes with NVR-induced 8 Additional details of the matching procedures are described in the appendix section

A3.

49

STATISTICAL ANALYSIS

0.50 0.00

0.25

proportion surviving

0.75

1.00

3

0

10

20

30

40

50

1 1

0 1

survival time Number at risk NVR No NVR

37 75

27 31

15 10

3 3

Fig. 3.1 Kaplan-Meier survivor functions Notes: The dashed line represents regimes where the transition was induced by NVR and the solid line refers to regimes without NVR-induced transition

transitions have a median survival rate of 47 years. By contrast, democratic regimes that were not induced by NVR have a median survival rate of just nine years. Accordingly, many of these regimes fail within the firstdecade transition, which is illustrated by the steep slope of the survival line during this time period. These results indicate that NVR during the transition phase is beneficial for the survival of the subsequent democratic regime. At the same time, Fig. 3.1 shows that this finding is restricted to a time period of up to 30 years after transition. After about 30 years, the sample becomes too small to make any reasonable comparison and the slope of the survival curves of the remaining regimes does not differ across groups anymore. We obtain similar results when using alternative regime datasets.

50

D. LAMBACH ET AL.

As described above, we further evaluate this descriptive finding with regression models and propensity score matching.9 We employed Coxproportional hazard models to estimate the effect of NVR-induced transitions on the hazard of democratic breakdown. The results show that NVR significantly and substantially reduces the hazard of democratic breakdown. Depending on the regime dataset used and different aspects of the estimation procedure, the reduction is estimated to be at least 53% and at most 72%.10 These estimates describe relative hazards comparing the risk of breakdown in NVR-induced regimes to those without this feature. This means that at each point in time after democratic transition, regimes that were induced by NVR are at least 53% less likely to experience a democratic breakdown than regimes without this feature. In sum, the empirical analysis provides strong support for the hypothesis that NVR-induced transitions are beneficial for subsequent democratic survival. Using different datasets and estimation procedures we find a significant and substantial effect of NVR, which highlights the robustness of this result. Specifically, our findings are not affected by particular coding decisions of political regime datasets (i.e. timing of transition events and subsequent survival), functional form assumptions of estimators, and covariate imbalance.

The Effect of NVR on Achieving Peaceful Turnovers of Power In this section, we analyse how NVR-induced transition influences the probability of achieving peaceful turnovers of power. The data on peaceful turnovers of power is taken from the Varieties of Democracy project (Teorell et al. 2016). As discussed above, turnovers occur according to a sequential process. The process begins with a regime experiencing a

9 Detailed results are reported in the appendix section A4. Also note that the results on the effect of NVR on democratic survival draw upon an earlier study conducted by the authors (Bayer et al. 2016). 10 For each of the three datasets, we estimated Cox models with robust standard errors clustered by regime and Cox models with shared regime frailties. For the Geddes data, some of the models yielded a potential violation of the proportional hazard assumption and subsequent analysis indicated that the effect of NVR on democratic survival could be diminishing over time.

3

STATISTICAL ANALYSIS

51

democratic transition which entails successful founding elections. Afterwards, the regime may experience its first peaceful turnover of power or not. Once a regime has successfully accomplished the first turnover, it may achieve the second peaceful turnover of power or not. In other words, achieving the first turnover of power is a necessary condition to realize democratic consolidation through the second peaceful turnover of power. In Figs. 3.2 and 3.3 we describe this sequential process with empirical data and show if and how regimes induced by an NVR campaign and regimes without this feature differ regarding the achievement of the first and second turnover. Taking the full sample of cases, out of the 112 regimes, 44 experienced the first turnover of power. Of these 44 regimes, 29 achieved the second turnover. Of 75 regimes that democratized without NVR influence, only 24 (32%) achieved the first turnover of power. By contrast 20 (54%) of the 37 NVR-induced democracies achieved the first turnover. The difference between the two groups gets even more

Democratic transition (75)

First turnover (24)

No turnover (51)

Second turnover (12)

No Second turnover (12)

Fig. 3.2 Peaceful turnovers of power for regimes without NVR

Democratic transition (37)

Second turnover (17)

First turnover (20)

No turnover (17)

No Second turnover (3)

Fig. 3.3 Peaceful turnovers of power for regimes with NVR

52

D. LAMBACH ET AL.

pronounced when looking at the second turnover. Half of the 24 regimes without NVR influence that achieved the first turnover also saw the second peaceful turnover of power. Of the 20 regimes that were induced by NVR and achieved the first turnover, 17 (85%) also completed the second turnover. A similar pattern also appears when using different datasets and/or matched samples. We further evaluate this descriptive result with regression analysis, using a sequential logit model which accounts for the sequential nature of the two-turnover measure of democratic consolidation (Buis 2013).11 The results are inconclusive regarding the achievement of the first turnover of power. We find no significant effect of NVR on the first turnover of power in any of our specifications. However, in line with the pattern described above in Figs. 3.2 and 3.3, our results indicate that, given a successful first turnover, NVR improves the probability of accomplishing the second turnover of power. Substantially, NVR increases this probability by more than 30% relative to cases with elite-led or violent transitions. However, these findings are not entirely robust. We obtain a significant effect of NVR on the second turnover of power only in half of our specifications, which vary with regard to the sample of cases used in the analysis and the measurement of political regimes and transition events.12 In sum, our findings highlight that NVR-induced regimes more frequently achieve two peaceful turnovers of power relative to regimes without this feature. While we find no substantial effect of NVR on the first turnover, our estimates suggest a substantial effect on the second turnover. Since the estimated effect from the regression analysis is not robust across different datasets and matching procedures, we remain sceptical with regard to its causal interpretation. We cannot rule out that confounding factors and alternative explanations drive the relationship between NVR and achieving two peaceful turnovers of power. Considering the sequential nature of the phenomenon, these findings suggest

11 The model estimates a separate logit for the first and second turnover, accounting for the characteristic of the data that while all regimes in the sample can experience the first turnover of power, only those regimes that had a successful first turnover may achieve the second turnover of power. 12 Detailed results are reported in the appendix section A5. Note that preliminary results on this topic were also reported by Bethke (2017).

3

STATISTICAL ANALYSIS

53

that NVR-induced transitions alone do not explain the accomplishment of two peaceful turnovers of power.

The Effect of NVR on Democratic Quality We now look at how NVR influences the quality of democracy. Here, we focus on democratic consolidation not as a condition but as a continuous variable, i.e. a matter of degree. Accordingly, we analyse whether democratic regimes that came about through NVR campaigns attain a higher degree of democracy after transition than regimes without this feature. To measure democratic quality, we use a polyarchy index proposed by Teorell et al. (2016). This measure of democratic quality is based on expert coding by more than 2600 country experts and gathers hundreds of indicators for almost every country in the world. The polyarchy index is based on the following five components: (1) elected officials, (2) free and fair elections, (3) freedom of expression, (4) associational autonomy, and (5) inclusive citizenship. Expert ratings for these dimensions are combined in an index that ranges from zero to one, with higher values indicating a higher quality of democracy. Figure 3.4 describes the average pre- and post-transition trends in polyarchy levels for the regimes with NVR and without NVR. Figure 3.4 shows that before experiencing a democratic transition, regimes that later democratized with the help of NVR had slightly lower levels of democracy relative to regimes that democratized without NVR. However, the overlapping confidence intervals indicate that this pattern is not significant. Furthermore, the mostly parallel lines suggest that both groups of regimes follow a similar time trend until the transition event. Just one year before the transition the groups converge at almost the same level of democratic quality. After transition, a different pattern emerges. Regimes that experienced an NVR-induced transition improve more quickly and more substantially regarding their democratic quality than regimes without this feature. One year after the transition, NVRinduced regimes on average reach a level of democracy of 0.61, whereas regimes that democratized without the help of NVR on average only achieve a level of 0.46. This substantial difference remains for years, until after about five years the group of regimes without NVR slowly begins to catch up. But even at the end of the time series, 10 years after transition, we can still observe a higher level of democracy for regimes where the transition was achieved by means of NVR. For comparison, states in

54

.4 0

.2

polyarchy

.6

.8

D. LAMBACH ET AL.

−5

−4

−3

−2

−1

0

1

2

3

4

5

6

7

8

9

10

time before and after transition

Fig. 3.4 Average levels of democracy before and after transition Notes: The grey line represents the average level of the polyarchy index for regimes where the transition was induced by NVR and the black line refers to regimes without NVR-induced transition. Spikes describe 95% confidence intervals

Western Europe and North America, which are usually considered consolidated democracies, achieve scores between 0.8 and 0.9 on the polyarchy index. After ten years some of the NVR-induced regimes come close to this level, whereas democracies that came about without NVR mostly do not even reach a score of 0.6. Results from regression analysis using the full and the matched samples largely confirm these patterns.13 We conducted two kinds of regression analysis. First, we analysed the effect of NVR on post-transition levels of democracy up to 10 years after the transition. The results confirm that there is robust evidence for a positive effect of NVR on post-transition 13 Detailed results are reported in the appendix section A6. Also note that this analysis of the effect of NVR on democratic quality draws upon an earlier study conducted by Bethke and Pinckney (2019).

3

STATISTICAL ANALYSIS

55

levels of democratic quality. For the first five years after transition, the difference in democratic quality between NVR-induced regimes and regimes without this feature is significant and substantial across all specifications. Moreover, for most specifications the findings also suggest a substantial long-term effect of NVR on democratic quality up to 10 years after transition. Second, we evaluated how NVR affects the degree of improvement in democratic quality from pre- to post-transition. This is called a difference-in-difference (DiD) approach. Instead of estimating the difference in the average levels of democratic quality after transition between regimes with NVR-induced transitions and regimes without this feature, we compare the difference in their improvement in democratic quality from pre- to post-transition. One main advantage of this approach is that it accounts for all static unmeasured differences between countries/regimes.14 To implement the DiD estimation, we use measures of the difference between the level of polyarchy one year before the transition and up to 10 years after the transition as the outcome variable in a regression analysis. The results largely corroborate the pattern described in Fig. 3.4: that democratic regimes with NVR-induced transition improve more substantially in democratic quality from pre- to post-transition than regimes without this feature. On average, NVR-induced regimes improve 0.1 to 0.2 units more than regimes without this feature from the year before until three years after transition. This is a substantial difference, given that the scale for change in polyarchy ranges from −1 to 1. For comparison, a difference of 0.18 units is roughly equivalent to the difference in the level of democracy between the US and Romania in 2015. Again, estimates are substantially similar across a variety of regression and matching specifications. However, the DiD results also show that regimes without NVR that survive manage to catch up after some time. Especially after five years, the pre-to-post difference in improvement becomes substantially smaller and in some cases results are even not significant. This highlights that 14 For instance, one potential unmeasured factor which we can account for with this procedure is democratic culture or, more specifically, the differences in the predisposition of elites and the population towards democratic values. Countries with positive attitudes towards democracy among elites and the population should be more likely to experience the occurrence of NVR-induced transitions and attain a higher level democratic quality after transition. The DiD approach accounts for such confounding factors by comparing differences in improvement of democratic quality.

56

D. LAMBACH ET AL.

the positive effect of NVR on democratic quality is not based on small incremental changes. Instead, NVR-induced regimes improve more substantially than regimes without this characteristic during transition and an initial period of democracy lasting two to three years. However, the difference in democratic quality established in the early post-transition years gets smaller over time. In sum, the results from the DiD support a causal interpretation of the effect of NVR on democratic quality. The DiD effect of NVR is robust and substantial for the early time periods after transition, but our results are inconclusive for the later time periods. NVR-induced regimes establish and sustain high levels of democratic quality but some cases without NVR also attain a high level of democratic quality.

Discussion of the Results Our analysis reveals that NVR influences all three aspects of democratic consolidation. Democratic regimes that evolved from an NVR campaign survive longer than regimes without this feature. They are also more likely to pass the two-turnover test and achieve higher levels of democratic quality. What are the implications resulting from these findings? Results of statistical analysis of macro-comparative indicators may appear crude and superficial when applied to complex social processes such as democratic consolidation. Moreover, the estimated average effects offer little guidance when evaluating actual empirical cases and contemporary political developments in specific countries. Therefore, it seems necessary to translate the estimated results into more meaningful implications and comparisons and highlight how they are relevant for societies. The finding that NVR advances democratic survival implies that these regimes are more resilient and more capable of avoiding democratic breakdown. As discussed above, the effect of NVR often translates into decades without democratic breakdown. The post-communist transitions and subsequent political developments illustrate the relevance of this finding. Only one post-communist country that achieved democracy by means of NVR experienced democratic breakdown (Albania). In all other cases democracy survived. By contrast, democracies in post-communist countries that were established without an NVR campaign mostly did not last long (e.g. Azerbaijan, Georgia, or Russia). In general, when matching cases of NVR-induced transitions to similar cases without this feature,

3

STATISTICAL ANALYSIS

57

most of the time the NVR cases outperform their counterparts in terms of democratic survival. These implications appear particularly relevant to current debates about democratic backsliding and shrinking spaces for civil society organizations in many regions of the world. Countries where democracy emerged from an NVR campaign may still suffer from a decline in democratic quality but their core democratic institutions are more likely to survive than those of other democracies. In the context of recent development of democratic regress in Eastern European countries such as Hungary or Poland, these findings should give us some confidence that democracy will nonetheless survive. The finding that NVR increases the probability that democracies achieve consolidation, measured in terms of two peaceful turnovers of power, was not entirely robust and should therefore not be seen as conclusive. Moreover, the complicated structure of the two-turnover process also implies that the results should be interpreted with caution. However, even after considering these caveats, an important implication remains. Although there is no evidence that NVR advances the probability of a first peaceful turnover, the probability of subsequently achieving a second peaceful turnover is substantially higher for NVR-induced regimes relative to democracies without this feature. More specifically, for regimes that were not induced by NVR, the chance of a second turnover is more or less a coin flip, whereas NVR-induced regimes are highly likely to accomplish the milestone of the second turnover if they managed to complete the first.15 This highlights the long-term enduring effect of NVR-induced transitions on political development. Furthermore, the finding implies that especially for democracies that came about by means of NVR, the first turnover is a bigger hurdle than the second one. Activists and pro-democratic groups and individuals should be aware of the pitfalls of the first peaceful turnover and also of the opportunity it represents on the path to the second turnover of power. For example, Tunisia, one of the contemporary democracies that achieved a transition after a campaign of NVR, has just seen its first peaceful turnover. After

15 This also contradicts Gasiorowski and Power (1998, p. 133, Fn. 116), who argue that the two-turnover test is ‘excessively confining’ for Third World countries who have a mixed record of democracy. It turns out that the second peaceful turnover reveals a crucial difference between NVR-induced democracies and those that came about through other modes of transition.

58

D. LAMBACH ET AL.

the death of President Beji Caid Essebsi, presidential elections were held in September 2019. No candidate achieved a majority in the first round of voting, so a runoff was held between two political outsiders, Kais Saied and Nabil Karoui. In the second round, Saied convincingly defeated his opponent, the ‘Tunisian Berlusconi’ Nabil Karoui, a media mogul who spent most of his campaign in jail on charges of fraud and money laundering. Although it is hard to say at this point, we expect that the independent Saied with his grassroots democratic orientation and lack of an institutional power base will be more capable of restoring the dwindling faith in democratic accountability than Karoui would have. Domestic civil society organizations and political leaders as well as external actors involved in promoting democracy should concentrate their efforts on supporting democratic governance in Tunisia to help the country achieve the critical milestone of the second peaceful turnover. Our findings on the effect of NVR on democratic quality suggest that NVR-induced transitions manage to establish regimes of higher democratic quality during the early (post-)transition years relative to cases where transition occurred without an NVR campaign. Especially during the early post-transition years, the difference in democratic quality between regimes established via NVR campaigns and those that were not is quite substantial. Afterwards the gap begins to close, which means that regimes that achieved democracy without an NVR campaign but managed to avoid democratic breakdown are able to attain levels of democratic quality that are closer to NVR-induced democracies. Thus, the difference in levels of democratic quality between NVR-induced regimes and regimes without this feature is not due to a constantly higher improvement rate of the former. Instead, the rate of improvement is much higher in NVR-induced regimes during the early years but their advantage over other democracies diminishes as time goes by. (Although this effect might also be partly due to badly performing non-NVR democracies breaking down quickly and thereby being selected out of the sample.) But even ten years after transition, NVR-induced regimes on average still have slightly higher levels of democratic quality than regimes without this feature. This implies that NVR-induced regimes from the very beginning establish core features of polyarchy such as extensive freedom of expression and association and free and fair elections. Democratic regimes without NVR either break down early or need more time to attain norms and institutions of similar quality.

3

STATISTICAL ANALYSIS

59

In sum, our results show that NVR advances democratic consolidation across the three dimensions of democratic survival, peaceful turnovers of power, and democratic quality. This supports our argument about the path-dependent effect of NVR: that the mode of transition determines the trajectories of democratic transition and its subsequent consolidation. However, the effect of NVR varies in size and robustness depending on the specific indicator representing the concept of democratic consolidation. This suggests that NVR is not a catch-all solution for democratic consolidation. Instead, these results suggest that NVR has a strong positive effect on some aspects of democratic political development, while it has little or no effect on others. Finally, we want to highlight two important caveats related to the findings presented in this chapter. First, an important issue in the estimation of the causal effect of NVR on democratic consolidation is that we consider all NVR campaigns as equal with respect to their influence on the transition process and subsequent political development. Our data only measures whether a resistance campaign influenced the transition process or not. Thus, we ignore potential variation in the intensity of our treatment indicator (NVR-induced transitions). From a methodological perspective, this relates to the stable unit treatment value assumption (SUTVA), which is important for the estimation of causal effects. Among other things, the SUTVA requires that the treatment is the same for all subjects and there is no variation in treatment severity or quantity (Guo and Fraser 2010, pp. 35–36). In our case, SUTVA problems may arise due to various factors, including variation in the size of campaigns and their organizational capacity to influence the transition process and subsequent political developments. We opted for this approach because we wanted to compare NVR-induced transitions with those that occurred without a civil resistance campaign and were less interested in teasing out differences among NVR campaigns. As a result, we cannot measure the size and scope of a nonviolent campaign where no such campaign was present and argue that it would be inappropriate to assign a score of zero. Thus, we chose to prioritize proper counterfactual comparison over considering treatment intensity in our analysis. We go some way to exploring the variation between different cases of NVR-induced transition in Chapters 4 and 5 based on our analysis of the qualitative evidence. Second, while some of our results are very robust and substantial, which in some instances even justifies a causal interpretation of the estimated effects, they do not tell us how the positive effect of NVR on

60

D. LAMBACH ET AL.

democratic consolidation comes about. In other words, the quantitative analysis alone provides no answer as to through which mechanisms and processes NVR influences political development. Moreover, we did not analyse the various dimensions of democratic consolidation in great detail, and therefore cannot tell anything about which aspects of political development—such as elections, civil society, constitutions, or organs of the state—are affected by the legacy of NVR. These issues will be addressed by discussing case-specific evidence on the mechanisms of NVR in subsequent chapters.

References Aronow, P. M., & Samii, C. (2016). Does Regression Produce Representative Estimates of Causal Effects? American Journal of Political Science, 60(1), 250– 267. https://doi.org/10.1111/ajps.12185. Bayer, M., Bethke, F. S., & Lambach, D. (2016). The Democratic Dividend of Nonviolent Resistance. Journal of Peace Research, 53(6), 758–771. https:// doi.org/10.1177/0022343316658090. Bethke, F. S. (2017). Nonviolent Resistance and Peaceful Turnover of Power. Peace Economics, Peace Science and Public Policy, 23(4). Bethke, F. S., & Pinckney, J. (2019). Nonviolent resistance and the quality of democracy. Conflict Management and Peace Science. https://doi.org/10. 1177/0738894219855918. Boix, C., Miller, M., & Rosato, S. (2013). A Complete Data Set of Political Regimes, 1800–2007. Comparative Political Studies, 46(12), 1523–1554. Boix, C., & Stokes, S. C. (2003). Endogenous Democratization. World Politics, 55(4), 517–549. Buis, M. (2013). SEQLOGIT: Stata Module to Fit a Sequential Logit Model. EconPapers. https://econpapers.repec.org/software/bocbocode/s456843. htm. Celestino, M. R., & Gleditsch, K. S. (2013). Fresh Carnations or All Thorn, No Rose? Nonviolent Campaigns and Transitions in Autocracies. Journal of Peace Research, 50(3), 385–400. https://doi.org/10.1177/0022343312469979. Cheibub, J. A. (2007). Presidentialism, Parliamentarism, and Democracy. Cambridge University Press. Chenoweth, E., & Lewis, O. A. (2013). Unpacking Nonviolent Campaigns: Introducing the NAVCO 2.0 Dataset. Journal of Peace Research, 50(3). Chenoweth, E., & Stephan, M. J. (2011). Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. New York: Columbia University Press. Coppedge, M., Gerring, J., Lindberg, S. I., Skaaning, S.-E., & Teorell, J. (2015). V-Dem Codebook v5. Gothenburg: Varieties of Democracy (V-Dem) Project.

3

STATISTICAL ANALYSIS

61

Gasiorowski, M. J., & Power, T. J. (1998). The Structural Determinants of Democratic Consolidation: Evidence from the Third World. Comparative Political Studies, 31(6), 740–771. Geddes, B., Wright, J., & Frantz, E. (2014). Autocratic Breakdown and Regime Transitions: A New Data Set. Perspectives on Politics, 12(2), 313–331. Gleditsch, K. S. (2002). Expanded Trade and GDP Data. Journal of Conflict Resolution, 46, 712–724. Gleditsch, K. S., & Ward, M. D. (2006). Diffusion and the International Context of Democratization. International Organization, 60(4), 911–933. Guo, S., & Fraser, M. W. (2010). Propensity Score Analysis: Statistical Methods and Applications. Thousand Oaks: SAGE Publications. Guo, S., & Stradiotto, G. A. (2014). Democratic Transitions: Modes and Outcomes. New York: Routledge. Haggard, S., & Kaufman, R. R. (2016). Dictators and Democrats: Masses, Elites and Regime Change. Princeton: Princeton University Press. Ho, D. E., Imai, K., King, G., & Stuart, E. A. (2007). Matching as Nonparametric Preprocessing for Reducing Model Dependence in Parametric Causal Inference. Political Analysis, 15(3), 199–236. Huntington, S. P. (1968). Political Order in Changing Societies. New Haven: Yale University Press. Lipset, S. M. (1959). Some Social Requisites of Democracy: Economic Development and Political Legitimacy. American Political Science Review, 53(01), 69–105. Schneider, B. R. (1995). Democratic Consolidations: Some Broad Comparisons and Sweeping Arguments. Latin American Research Review, 30(2), 215–234. Sekhon, J. S. (2009). Opiates for the Matches: Matching Methods for Causal Inference. Annual Review of Political Science, 12, 487–508. Singer, J. D. (1987). Reconstructing the Correlates of War Dataset on Material Capabilities of States, 1816–1985. International Interactions, 14, 115–132. Teorell, J. (2010). Determinants of Democratization: Explaining Regime Change in the World, 1972–2006. Cambridge University Press. Teorell, J., Coppedge, M., Gerring, J., & Lindberg, S. (2016). Measuring Electoral Democracy with V-Dem Data: Introducing a New Polyarchy Index (Working Paper No. 25). Varieties of Democracy Institute. Ulfelder, J. (2010). Dilemmas of Democratic Consolidations. A Game-theory Approach. Boulder, CO: First Forum Press. Ulfelder, J. (2012). Democracy/Autocracy Data Set. Harvard Dataverse, V1(hdl:1902.1/18836).

CHAPTER 4

Mechanisms

Building upon the preceding chapter’s findings from the quantitative analysis on the general positive effect of nonviolent resistance (NVR)-induced transition on subsequent democratic consolidation, we now investigate the causal mechanisms that produce this effect. NVR research has generally not thus far taken an empirical approach to the question of exactly how NVR works. As discussed in Chapter 2, previous works make theoretically founded assumptions but these are often based on the idiosyncrasies of particular cases, with little systematic theory-building. Only recently has the study of the mechanisms underpinning NVR received some attention (Kim and Kroeger 2019). The notion of ‘mechanisms’ has been much contested in the philosophy of science (see e.g. Bennett 2013; Gerring 2010), a discussion we do not have space to enter into here. For us, mechanisms are processes that link cause (e.g. NVR) and effect (democratic consolidation). This linkage encompasses causal processes in a narrow sense (of some part of the process setting off the next stage) and in a wider sense (of some part of the process making the next stage possible). In the social world, mechanisms and causal pathways are typically complex with many variables and mechanisms interacting in complicated ways. This means that any theory about causal mechanisms will necessarily be reductionist. What is more, our aim is not just to theorize about what produced the outcome in a particular case, as many process-tracing studies aim to do, but to identify more general patterns of causality that are useful beyond single-case explanations. Hence, we take a comparative approach to the study of mechanisms to © The Author(s) 2020 D. Lambach et al., Nonviolent Resistance and Democratic Consolidation, https://doi.org/10.1007/978-3-030-39371-7_4

63

64

D. LAMBACH ET AL.

identify ‘systematic ideal-types’ of causal configurations (Jackson 2011, p. 150). That said, the sample size of our comparison cases is deliberately small to allow for an in-depth analysis. There are two related reasons for this: First, given the dearth of existing work on mechanisms, the nature of our work was more theory-building than theory-testing. Second, this approach required a deep immersion into cases to identify and sift through possible explanations. This is why most of the discussion in this chapter is based on case evidence of democratic transition and consolidation. We mostly draw on the cases of democratic transition in Benin (1991) and Chile (1990). These are both cases of democratic transition and some degree of consolidation achieved by means of NVR. However, they also differ in terms of the impact that NVR had on the transition process. While transition in Benin was driven by a peaceful pro-democratic movement, transition in Chile was dominated more by political elites. One result is that while Benin managed something of a clean break with autocracy, it took Chile more than a decade for the last authoritarian holdovers to be displaced and for civil society to mobilize again. From an ex post perspective, we can justifiably call these two cases most (Benin) and least likely (Chile) for our theoretical predictions about how NVR advances democratic consolidation. Where appropriate, we bring in additional and ‘contrastive’ evidence (Grynaviski 2013) from two case studies on Namibia and El Salvador which witnessed a democratic transition after an armed struggle, and from two case studies on Cape Verde and Paraguay which saw top-down transitions in 1990. These four cases served as controls to check whether potential causal mechanisms were specific to NVR-induced democracies or could also be found in non-NVRinduced cases.

Democratization and Democratic Consolidation in Benin Our first case is what the Beninese called the ‘renouveau démocratique’, the 1990/1991 democratic renewal of the country after widespread and strong civil resistance against the socialist regime of Mathieu Kérékou. The transition ended with the first peaceful electoral turnover on the African mainland, making Benin the first of the new democracies in Africa (Decalo 1997) and up to now one of Africa’s most advanced and stable ones.

4

MECHANISMS

65

Due to the strength of the NVR movement and the absence of many other factors typically associated with democratization and democratic consolidation, Benin can be considered as an ‘ideal case’ (Seawright and Gerring 2008) to study the effect of NVR on democratic consolidation. With its level of economic development ‘well below the levels commonly associated with democratic success’, Benin is frequently described as an instance of democratic consolidation ‘against the odds’ (Gisselquist 2008, pp. 789–794; Magnusson 2001) that contradicts the assumptions of modernization theory (e.g. Lipset 1959; Rostow 1971). Further, the transition took place shortly before the famous La Baule conference of French and African heads of state (19–21 June 1990), where the French President François Mitterrand announced that French development aid would be conditional on democratic improvements in the future. Thus, foreign influence was a less decisive factor in Benin than in other cases of democratization in sub-Saharan Africa (Banégas 1997). Last but not least, Benin’s political history was very chequered and riddled with periods of instability. The small country located between Togo and Nigeria gained independence from France in 1960 and after a short period as a democratic republic soon acquired a ‘sad reputation of being famous for successive military coups’ (Koko 2008, p. 4). From its independence in 1960– 1972, the country saw 11 presidents, six different constitutions, and 12 attempted and five successful coups d’état (Bierschenk 2009, p. 13). The last of this series of coups brought to power Kérékou, a major in the armed forces and commander of a paratrooper unit. He declared himself president and introduced Marxism as the official state ideology. From 1974 onwards, Benin became a single-party state with Kérékou acting as President of the Politbureau of the newly founded single party, the Parti de la Révolution Populaire du Bénin (PRPB). This brought some stability to the country since Kérékou’s relatively moderate socialism helped to align the left with the regime and ‘provided a way of explaining past failure while promising a viable and effective development path’ (Allen 1992, p. 44). From 1985 onwards, however, ever-growing parts of the population distanced themselves from the regime. First, resistance came from left-wing students and teachers and was directed against education reforms (Nwajiaku 1994, p. 435; Allen 1992). After independent unions had been forbidden and forced to join the national union—the Union Nationale des Syndicats des Travailleurs du Bénin—the resistance grew as more and more syndicalists joined. With the national economy deteriorating and terms of trade worsening, the financial situation of the regime

66

D. LAMBACH ET AL.

declined to the point where public servants were paid only irregularly. This drove the urban population and even parts of the bureaucracy itself to join the ranks of the opposition. Under increasing pressure, Kérékou liberalized the media in 1988 in an attempt to appease the opposition. This, however, backfired. Far from reducing public grievances, the emergence of a free media further contributed to anti-regime mobilization by publicizing the rampant corruption (Robinson 1994, p. 594). Over time, the goals of the opposition became more radical: while the initial protests aimed at only minor political reforms, the opposition later rallied under the call to ‘Rise up to get rid of Kérékou and his clique’ (Houngnikpo and Decalo 2013, p. 12). The year 1989 was first marked by work slowdowns (Seely 2009, p. 39) and later by a general strike with ‘overwhelming national proportions’ (Koko 2008, p. 44).1 In July 1989 employees from 13 out of 16 state ministries went on strike (Bierschenk 2009, p. 3). The mainly urban protests in Porto Novo and Cotonou were complemented by communistled acts of civil disobedience such as tax boycotts in the more rural areas of the north (Akindes 2015, p. 54). With many state employees joining the protests, the Kérékou regime could not withstand such an assault for long (Eboussi Boulaga 2009, p. 32; Heilbrunn 1993). On 7 December 1989, Kérékou officially announced the end of Marxism–Leninism in Benin and called for the appointment of an Assemblée Nationale des Forces Vives de la Nation—a National Assembly of the Active Forces of the Nation, or ‘National Conference’ for short. What was originally meant as a symbolic exercise to discuss some minor reforms and as ‘a forum of ideas’ (Kérékou 1994, p. 19) to ‘tackle the objective of structural adjustment’ (p. 14) developed its own dynamic. The opposition successfully coordinated and joined forces, which gave them the upper hand in the National Conference, whose delegates finally declared it to be a constituent assembly (Seely 2009, p. 42). The National Conference worked out a new constitution, installed a provisional government and set the terms for democratic elections. What Kérékou labelled as ‘civilian coup d’état’ (Adamon 1995, p. 59) became known to the public as the ‘renouveau démocratique’ (Adamon 1995), Benin’s democratic renewal. Since large segments of society participated in the protests and strikes, it was also considered a ‘People’s 1 Banégas (1995) reports of multi-sectorial protests including the hospitals, the taxi drivers, lawyers, workers from the influential post and telecommunication enterprises and even some police units.

4

MECHANISMS

67

Revolution’ (Koko 2008, p. 43). Owing to its constructive and cooperative nature, Johansen and Vinthagen (2019) speak of the transition as a case of ‘revolutionary constructive resistance’. The first free presidential and parliamentary elections were held in 1991. In the presidential election, Kérékou lost to a rival, Prime Minister Nicéphore Soglo, and became the first president on the African mainland to resign peacefully after a defeat at the polls. Since then, Benin has often been portrayed as a role model for democratization (Decalo 1997; Seely 2009; Magnusson 2001). So far, Benin has witnessed six presidential and eight parliamentary elections which have all been rated as mostly free and fair (Houngnikpo and Decalo 2013, p. 14). The polls resulted in four peaceful political turnovers. Benin passed the two-turnover test in 2006 when Kérékou, whose re-election in 1996 marked the first peaceful turnover, handed over power to Thomas Boni Yayi. After two terms in office, Boni Yayi stepped down in 2016 for another peaceful turnover to Patrice Talon, the winner of the presidential elections.

Democratization and Democratic Consolidation in Chile From 1973 to 1989 Chile was ruled by the autocratic regime of army general Augusto Pinochet. The overthrow of the democratically elected socialist government by a military Junta in 1973 broke with Chile’s sustained democratic development since the nineteenth century (Huneeus 2009, p. 199). The coup reflected the Chilean military’s praetorian selfunderstanding of being the custodian of the nation and its constitution by installing a ‘protected democracy’ (Constable and Valenzuela 1989, p. 172). More specifically, the Chilean military saw its role as ‘guarding’ Chile against what it perceived as the chaos of political polarization and economic struggles under the incumbent socialist government. The prize for this ‘guardianship’ was that the Junta only allowed minimal public participation and governed through severe repression, including the systematic murder and torture of dissidents (Huneeus 2014; Delamaza 2015; Angel 2007). Open resistance against the regime started in 1983 and was mainly carried out by two groups. The first were social movements consisting of a multitude of actors including slum dwellers, trade unions, and students, and the second were (re)emerging oppositional political elites rooted in democratic parties that the regime had pushed underground. These

68

D. LAMBACH ET AL.

groups developed diverging strategies and goals to topple the military dictatorship. The social movements mostly utilized NVR techniques of street demonstrations, boycotts, and strikes, demanding Pinochet’s resignation and a complete democratic transformation of society (Kurtz 2009; Nepstad 2011). The second group coalesced around several centre-left multiparty initiatives in the early to mid-1980s, demanding new elections and seeking negotiations with regime representatives (Huneeus 2014; Hecht Oppenheim 2007). However, both the protest movement and the party initiatives failed to seriously weaken the regime and from 1986 onwards the attention of pro-democratic forces turned to the plebiscite scheduled for 1988. The plebiscite was based on the 1980 constitution through which the military dictatorship had sought to legitimize its rule. The constitution foresaw an eight-year ‘transition period’ after which the plebiscite would be held to answer the question of whether to keep the military regime in power or to return to civilian rule. The electorate was presented with the question of whether Pinochet should remain head of government for another eight years (the ‘yes’ vote) or not (the ‘no’ vote). In the case of the ‘no’ vote winning, elections would be held (Huneeus 2014; Fuentes 2013). While the plebiscite itself was anything but free and fair, it presented an opening that the opposition could exploit. The outfit running the campaign for the vote against Pinochet was called the Concertación Partidos por el No and consisted of a broad alliance of centre-left opposition parties, which had an experience of working together in previous political initiatives. Over 90% of eligible citizens registered for the referendum, and various rallies, massive grassroots voter registration and influential TV spots accompanied the campaign (Nepstad 2011; Boas 2015). The vote ended in a success for the ‘no’ camp (54.7 vs. 43%). As Ibarra writes, ‘the triumph of the “No” was considered the result of a joint effort by a political elite [the oppositional parties] that was capable of organizing and creating alliances and a civil society that was active in organization and large-scale street protests’ (Ibarra 2016, p. 89). In the 1989 general election, the coalition of centre-left parties born out of the ‘no’ campaign changed its name into Concertación para la Democracia (Coalition for Democracy). The military itself mostly stayed out of the elections but the right-wing elites which had been its closest political allies set up an alliance of parties (Alianza). However, while the Concertación united most of the left, the right was undercut by an independent campaign of right-wing populist Francisco Talavera. As a result,

4

MECHANISMS

69

the candidate of the Concertación, Patricio Aylwin, won 55% of the votes. No major electoral irregularities, such as repression of political parties or civil society, were reported and the campaigns and vote were mostly free and fair (email conversation with Professor Claudio Fuentes, December 2016). In 1993, in the second national elections after the plebiscite, the Concertación was again successful and Eduardo Frei Ruiz-Tagle was elected president. The results of the first two elections were not challenged by the political right, as the constitutional setting still allowed for substantial veto powers for the military and right-wing political coalition, which will be described in detail later. This has led some authors to denote the transition as pacted (Fuentes 2015, p. 102; Rovira Kaltwasser 2007). By the end of the 1990s, the electoral dimension of Chile’s democracy could be described as consolidated, having gone through a set of free municipal and national elections without any grave irregularities. It is against that background that the post-Pinochet turnovers of power between the main political blocs have to be seen. After two decades of centre-left government, the 2009 elections brought to power Sebastián Piñera, a representative of the more liberal wing of the right-wing parties—through free and fair elections. It was a measure of the consolidation of Chilean democracy that the right-wing coalition’s victory did not signify a return to more autocratic politics (Navia and Godoy 2014). Instead, a second turnover of power in 2014 returned Michelle Bachelet of the Concertación (now called Nueva Mayoría) to the presidency, after a first stint from 2006 to 2010. In parallel to the electoral dimension, Chile adopted major changes to its constitution in 1989, 2006, and 2017, tackling many of the constitution’s remaining authoritarian tendencies over a drawn-out period (Huneeus 2014; Fuentes 2015). Now, Chile is often described as one of the wealthiest and most democratically stable states in the Latin American region (see for example Castiglioni and Rovira Kaltwasser 2016, p. 3; Bertelsmann Stiftung 2018), although its consolidation of democracy is by no means spotless, as the coming sections will show.

Levelling the Political Playing Field As described in Chapter 2, we assume that an NVR-induced transition has a positive effect on democratic consolidation because it levels the political playing field. This refers to horizontal power relations between government and opposition as well as the vertical relationship between

70

D. LAMBACH ET AL.

political elites and civil society. We begin our investigation of this mechanism by analysing the effect of NVR on levelling the playing field for the whole sample of cases, before providing detailed case-specific evidence from Benin and Chile. For comparative analysis, we focus on two aspects of this mechanism. Regarding the horizontal power relations between government and opposition, we investigate whether and how NVR advances constitutional changes that constrain the executive and empower the parliamentary opposition to hold ruling elites accountable. In terms of the vertical relationship between political elites and civil society, we evaluate whether and how NVR fosters an institutional environment where civil society is capable of holding the government accountable. Checks and balances and the separation of powers are crucial elements of democracy that constrain elites in the executive, legislative, and judicial branches in order to prevent the abuse of power. For democratic consolidation, constraints on the executive are particularly important because an authoritarian backlash typically originates from this source, much more than from parliament or the courts. This is why democracies give the legislative, in which the opposition is represented, substantial powers to hold the government accountable. These ‘legislative constraints’ on the executive are a core element of a democratic relationship between government and opposition: they provide a limit to the government’s power and encourage it to behave responsibly, and they empower the opposition by giving it a meaningful role in the maintenance of democracy. We expect that NVR-induced regimes employ stronger constraints on the executive than regimes that democratized without NVR. Figure 4.1 describes the results of comparing the average degree of legislative constraints for regimes with and without NVR-induced transition. Our measure of legislative constraints considers to what extent parliament is capable of questioning, investigating, and exercising oversight over the executive. The indicator is drawn from the V-DEM dataset. It is scaled from zero to one with higher values indicating a higher degree of legislative power to check and balance the executive. Figure 4.1 shows that NVR has a substantial impact on the development of legislative constraints on the executive. Whereas before transition, cases of subsequent NVR-induced transition score lower on this indicator relative to regimes that went on to transition without NVR, the picture is reversed after transition. One year after transition, NVR-induced regimes attain a level of 0.68. Regimes without this feature only reach a level of

MECHANISMS

71

.4 0

.2

legislative constraints

.6

.8

4

−5

−4

−3

−2

−1

0

1

2

3

4

5

6

7

8

9

10

time before and after transition

Fig. 4.1 Legislative constraints of the executive for regimes with and without NVR-induced transition Notes: Grey markers show the average value for regimes that transitioned via NVR and black markers refer to transitions without NVR. Spikes show 95% confidence intervals

0.51 on legislative constraints of the executive. Both types of cases continue to improve after transition, with non-NVR cases only beginning to close the gap slightly towards the end of the first decade after transition. In sum, NVR-induced regimes improve more substantially, more rapidly, and attain higher levels on this indicator. These results also hold when we employ more sophisticated quantitative methods that account for confounding and alternative explanations.2 This finding indicates that NVR campaigns succeed in establishing a balance in power relations which allows the opposition to constrain the executive through parliament and thereby impedes the evolution of dominant party regimes, which are more likely to fall back into authoritarian politics.

2 Detailed results are reported in the appendix section A7.

72

D. LAMBACH ET AL.

Next, we investigate if NVR achieved a levelling of the playing field regarding civil society’s ability to hold the government accountable. This concerns the development of two key political freedoms, the freedom of association and the freedom of expression, as well as the specific political environment in which civil society organizations (CSOs) operate. Again, we use different measures drawn from the V-DEM dataset to operationalize these variables. First, we use the ‘freedom of association thick index’, which measures the degree to which political parties and CSOs are able to form and operate freely. Specifically, this consists of indicators measuring party bans, barriers to parties, the autonomy of opposition parties, whether elections were contested by multiple parties, the independence of CSOs from government interference and the degree of CSO repression. Second, the measure of freedom of expression includes the following aspects: (1) media censorship efforts, (2) internet censorship efforts, (3) harassment of journalists, (4) media bias, (5) media selfcensorship, (6) the share of critical print/broadcast media, (7) the diversity of print/broadcast media perspectives, (8) freedom of discussion for men and women, and (9) freedom of academic and cultural expression. Both measures are scaled from zero to one with higher values corresponding to more freedom of association and expression respectively. Figure 4.2 displays the development of both measures over time for regimes that democratized via NVR and regime where transition occurred without NVR. Figure 4.2 shows that NVR has a substantial positive effect on the development of both freedom of association and freedom of expression. Regimes that democratized by means of NVR show a more substantial improvement and attain higher levels on these indicators than regimes without this feature.3 For example, one year after transition NVR-induced regimes attain a level of 0.82 on the indicator of freedom of expression whereas regimes that transitioned without NVR only reach a level of 0.65. Moreover, the relative positive effect of NVR only diminishes slightly over time. Even ten years after transition, NVR-induced regimes have higher levels of freedom of association and expression relative to regimes that transitioned by other means. Interestingly, there is a noticeable improvement in both scores in the years leading up to the transition, both for NVR-induced transitions, where the effect is stronger, and for 3 Note that Bethke and Pinckney (2019) report similar results for this effect using a different matching procedure.

73

.6 0

0

.2

.4

freedom of expression

.8

.8 .6 .4 .2

freedom of association

MECHANISMS

1

1

4

−5 −4 −3 −2 −1 0 1 2 3 4 5 6 7 8 9 10

−5 −4 −3 −2 −1 0 1 2 3 4 5 6 7 8 9 10

time before and after transition

time before and after transition

Fig. 4.2 Freedom of association and expression for regimes with and without NVR-induced transition Notes: Grey markers show the average value for regimes that transitioned via NVR and black markers refer to transitions without NVR. Spikes show 95% confidence intervals

those without NVR. We believe that this is due to the drawn-out nature of transitions, which rarely happen within a single year. If we take the example of Benin, where transition is coded for the year 1991 when the first national elections were held, there were noticeable improvements in civil liberties starting at least three years before, when the government lifted restrictions on independent media. Later, improvements continued under the caretaker government installed by the National Conference. Improvements in freedom of association and freedom of expression can therefore be partly explained by liberalizing moves by under-pressure autocrats trying to pacify public grievances and partly by institutional improvements in the early phase of transition. Taking a closer look at CSOs, we also analysed the effect of NVR on their post-transition political environment. We used the following four

74

D. LAMBACH ET AL.

indicators from the V-DEM dataset: (1) independence of CSOs, (2) freedom from repression for CSOs, (3) consultation of CSOs, and (4) participation in CSOs. The independence of CSOs refers to the ability of the government to control the involvement of CSOs in public life. This includes the licensing, founding, and banning of CSOs. The repression of CSOs considers whether CSO members are harassed, arrested, imprisoned, sanctioned, or in any way restricted in their scope of action. The consultation of CSOs refers to the degree of involvement of members of CSOs in policymaking. Finally, participation in CSOs considers the degree of popular involvement in CSOs. The first two indicators were already part of the freedom of association index discussed above. While that part of the analysis was more focused on the big picture of how freely citizens can associate in a variety of organizational forms, we also wanted to see how CSOs in particular are affected by NVR-induced transition. Again, all indicators are scaled from zero to one with higher values indicating better prospects for CSOs. Figure 4.3 describes the development of these different aspects of the CSO environment over time. As shown in Fig. 4.3, NVR leads to improvements in some scores but not in all of them. The clearest gains are in CSO independence. Both CSO consultation and CSO freedom from repression also show improvements, although these diminish over time. The effect on CSO participation is mixed, with NVR cases no longer trailing non-NVR cases but not improving significantly beyond them either. In sum, we interpret these to mean that the main effect of NVR is to provide a facilitating institutional environment for CSOs where they can work freely and autonomously, and where CSOs are more accepted as stakeholders in public governance. NVR does not lead to a measurably larger participation of citizens in CSOs. This runs counter to optimistic theories that participation in NVR movements will lead citizens to become more active in civic associations. However, these findings provide strong evidence that NVR has a substantial positive effect in fostering institutions favourable for civil society. Our results are also robust if we employ multivariate methods that account for confounding and alternative explanations.4 In sum, the comparative analysis highlights the different ways how NVR levels the political playing field with regard to both executive constraints and the institutional environment for civil society. Building on

4 Detailed results are reported in the appendix section A7.

.2

.4

.6

.8

1

75

−5 −4 −3 −2 −1 0 1 2 3 4 5 6 7 8 9 10

time before and after transition

time before and after transition

.8 .6 .4 .2 0

.2

.4

.6

.8

CSO participation

1

1

−5 −4 −3 −2 −1 0 1 2 3 4 5 6 7 8 9 10

0

CSO consultation

MECHANISMS

0

.8 .6 .4 .2 0

CSO independence

1

CSO freedom of repression

4

−5 −4 −3 −2 −1 0 1 2 3 4 5 6 7 8 9 10

−5 −4 −3 −2 −1 0 1 2 3 4 5 6 7 8 9 10

time before and after transition

time before and after transition

Fig. 4.3 Political environment for CSOs for regimes with and without NVR transition Notes: Grey markers show the average value for regimes that transitioned via NVR and black markers refer to transitions without NVR. Spikes show 95% confidence intervals

these findings, we now turn to the cases of Benin and Chile to analyse whether there is a causal mechanism connecting NVR movements to these constitutional and institutional improvements, and if so, how it works. How NVR Levelled the Playing Field in Benin The process of democratization in Benin vividly illustrates how NVR played a vital role in implementing executive constraints and improving the position of civil society vis-à-vis the government. Most importantly, civil society was directly involved in the National Conference, the main institutional body where the new rules for democracy in Benin were negotiated. The inclusive nature of the National Conference was achieved in the so-called ‘Quota War’ (Seely 2009, p. 41), where the opposition pressured the preparatory committee under Robert Dossou to

76

D. LAMBACH ET AL.

change the composition of the proposed National Conference to include more delegates from civil society. While the Kérékou regime was trying to fill delegate positions with loyal supporters, the opposition under the leadership of the unions demanded ‘[q]uotas that reflected the real situation’ (Léopold Dossou, secretary-general of the Syndicat National des Enseignants du Supérieur, cited by Banégas 1995, p. 15). These demands were supported by renewed street protests. Finally, an agreement was reached. The regime reduced the number of members of the army and the ruling party and increased the number of farmers and craftsmen who were designated by peasant unions and local development associations. Political parties were not represented in the National Conference since the unitary party had been dissolved. Opposition parties were either embryonic or nonexistent, with the exception of the Communist Party of Benin (PCB) which had worked clandestinely under the Kérékou regime but refused to take part in the conference. After the Quota War, the regime still held a formal majority since many of the new delegates came from parastatals. However, these delegates mostly represented softliners and the opposition took advantage of that. In talks preceding the National Conference, the opposition was able to win over many of these newly appointed delegates, altering the balance of power in their favour. Thus, civic forces, and specifically CSOs, gained substantial direct and indirect influence in the National Conference. In his commencement speech, the general rapporteur of the National Conference, Albert Tévoédjrè, highlighted the special role of the nongovernmental organisations [which] have become a very powerful force for development. They are faith-based or secular, and now they have drawing rights and credit lines in various international institutions. Briefly put, they have opportunities of direct support that can help us immediately, particularly with regard to the education sector, public health, urban environment or rural development.5

The strong position of civil society during transition is further evidenced by the many CSO representatives in the National Conference who later became representatives of the transitional government and the first government under President Soglo. Given their own background, these 5 Own translation. See http://democratiebenin.over-blog.com/article-rapport-generalde-la-conference-nationale-1990-73929678.html.

4

MECHANISMS

77

politicians did their best to realize the ideals of transparency and inclusivity. As Simon Asoba of the Friedrich Ebert Foundation puts it, this degree of public transparency and participation imbued the National Conference with a special legitimacy in the eyes of the watchful public: There was this enthusiasm of what they discuss[ed] […] in the national conference […]. So, everybody was watching them and they wanted to do well. They did wonderful work. (Interview with Simon Asoba, Friedrich Ebert Foundation, Cotonou, 19 October 2016)

Civil society leverage during the transition process had a substantial impact on Benin’s new constitution. As decided by the National Conference, the new constitution was to be drafted by the Commission des Lois et des Affaires Constitutionnelles (Commission for Laws and Constitutional Affairs) headed by Professor Maurice Ahanhanzo-Glèlè, a widely respected Professor for Law at the Université Paris-Sorbonne. However, the constitution was not simply to be drafted by constitutional experts and amended by the provisional government. Instead, Ahanhanzo-Glèlè, tasked with drafting the constitution by the National Conference, was eager to write what he called a ‘people’s constitution’. In an attempt to increase transparency and inclusivity, he instructed the members of the commission to head to their regions to discuss the Commission’s first draft of the new constitution with the population. The Commission members brought back additional recommendations from their trips which were discussed and included in the redrafting process (interview with Maurice Ahanhanzo-Glèlè, Cotonou, 10 November 2016). Further, the decision on three core principles of the constitution, namely the minimum age (40 years) and the maximum age (70 years) for presidential candidates and the two-term limit, were delegated to a public referendum. Public involvement in drafting the constitution also led to the implementation of several checks on executive power. The most important one is the constitutional court, which was established in 1993. It was designed as a political arbitrator both among political elites and between elites and citizens. This arbitrating role had hitherto been occupied by the armed forces, which ended political deadlock through military intervention in politics or outright coups d’état. Thus, the constitutional court taking on this role also represented a safeguard against renewed military rule

78

D. LAMBACH ET AL.

(Stroh 2013, p. 6).6 The court was indeed able to fulfil this task and played an important role in the consolidation of democracy in Benin (Seely 2009; Gisselquist 2008; Banégas 2003).7 In order to fulfil its first function to solve political deadlock and to counterbalance the presidency, the constitutional court was put in charge of supervising national elections and promulgating their results. Over the years, the court has settled several disputes over election results. In 1996 it rejected Soglo’s complaints that the presidential elections had been rigged. Through this verdict, which was rendered in spite of death threats against court judges, the constitutional court gained credibility and a reputation as an independent arbitrator (Kangnikoé 2014). This first constitutional court, presided by Elisabeth Pognon, also included AhanhanzoGlèlè, the ‘father of the constitution’. It was and continues to be widely perceived as having been the most legitimate and most impartial in the history of the new republic (Stroh 2016). Generally, the performance of the court has been quite impressive: since Benin’s transition to democracy all of the scheduled six presidential and eight legislative elections have taken place on time and have resulted in three peaceful turnovers, and the court has played an important part in that. The second function of the constitutional court is to prevent the abuse of power and to protect the individual citizen. Therefore, the fathers of the constitution, having a strong human rights background connected to their experiences of repression under the socialist regime, placed a great emphasis on individual rights. A total of 34 articles of the constitution are dedicated to human rights issues (Rotman 2003). Although most of the legal experts have been educated in France and were thus inspired by the French constitution and the French legal system, they decided to

6 To prevent military intervention, article 66 of the constitution further establishes

‘the right and the duty to make an appeal by any means in order to re-establish the constitutional legitimacy, including recourse to existing agreements of military or defense cooperation’. Further, article 66 stipulates that the right of any Beninese ‘to disobey and organize himself to put a check to the illegitimate authority shall constitute the most sacred of rights and the most imperative of duties’. 7 According to the constitution (Title V) the constitutional court shall ‘be the judge of

the constitutionality of the law [and] guarantee the fundamental human rights and the public liberties’. Furthermore, the constitutional court serves as ‘the regulatory body for the functioning of institutions and for the activity of public authorities’. This combination of constitutional jurisdiction (contrôle de constitutionnalité) and a duty to protect human rights is a unique feature to the Beninese constitutional court (Kangnikoé 2014).

4

MECHANISMS

79

deviate from it by granting every citizen the individual right to legal remedy before the constitutional court with respect to the constitutionality of laws. According to the legal expert Bado Kangnikoé, ‘one deliberately did not want to be guided by the French model. In France, the citizens have no direct appeal, and the National Assembly of the Active Forces wanted to give the citizens this opportunity to directly invoke the constitutional court. If the citizens have no appeal right, an unconstitutional court decision would be enforceable without any correction’ (correspondence with Bado Kangnikoé, Senior Researcher, Max Planck Institute for Social Law and Social Policy, 11 October 2016). This was anchored as the unconstitutionality exception in article 122 of the constitution. As another control function, the president and government ministers can be called before the legislature and questioned about their policies and actions. If the government does not respond to this call within 30 days, the constitutional court deals with the issue and may conclude that the president does not respect the legislature and even remove him from office (Constitution of Benin, article 71, articles 76–78). In cooperation with the constitutional court, the parliament is able to impeach the president. To ensure the independence of the constitutional court, four of the seven members of the constitutional court are appointed by an executive commission of the parliament and three by the president of the republic. The president of the constitutional court and their deputy are elected by their peers. Members of the constitutional court can serve for two terms, each lasting five years. Beyond the legal implementation of executive constraints8 and guarantees of freedom of association and expression,9 the nonviolent transition also fostered long-term gains for civil society via consultative practices. One particular mechanism that induced such long-term developments is the concept of ‘Estates-General’ which was established in Benin by the National Conference in 1989. Some authors argue that the idea 8 Article 71 of the constitution stipulates that every member of the National Assembly has the right to interpellate the president or any of his ministers. The respective president or minister has to respond to the interpellation. In case of conflict the National Assembly can serve as mediator and propose a solution. 9 Article 25 of the constitution guarantees the freedom of association, of assembly, of procession and of demonstration, article 23 guarantees religious freedom and freedom of expression, article 24 guarantees the freedom of press, and article 31 safeguards the right to strike.

80

D. LAMBACH ET AL.

of the National Conference was directly inspired by the French EstatesGeneral of 1789, which brought together clergy (First Estate), the nobility (Second Estate), and the commoners (Third Estate) (Fomunyoh 2001, p. 40).10 The historical reference to France provided the delegates with a ‘script’ and soon after the beginning, participants in the national conference began to refer to the forum as the États-Généraux (Robinson 1994, p. 579). This procedure of popular consultation and consensual decision-making in the National Conference had a huge symbolic influence. Early in the process, the preparatory committee under Robert Dossou had asked the public to send in ideas and proposals to set the agenda for the conference. Further, although the opinions on different topics were highly controversial within the National Conference, the delegates nonetheless developed a stance that a compromise needed to be sought ‘in order to maintain the momentum of transition’ (Seely 2009, p. 68). The whole conference was broadcast via radio and television and was eagerly followed by the population which is still proud to have managed the democratic transition peacefully (interviews with Alfred Deguenonvo and Comlan Théonas Moussou, consultants for civil participation and decentralization, Cotonou, 20 October 2016). The government news agency even produced a two-hour videotape of conference highlights which was pirated in ‘untold numbers’ (Robinson 1994, p. 576). In this sense, the National Conference was the first real democratic experience after 15 years of patronizing single-party rule and thus ‘set an institutional precedent by defining a method for the people to interact with the state and to express their sentiments to government’ (Heilbrunn 1999, p. 232). Although these public consultations in form of the Estates-General were not anchored in the constitution by the National Conference, they became an established democratic practice in Benin. According to Jennifer Seely the inclusive and consultative mode of the National Conference marked the beginning of a new era in Beninese politics, in which direct public appeals, whether to the population at large, or to specific elements of civil society, became a legitimate political tactic used by leadership to build

10 According to Robinson (1994, p. 591) the Benin-based Pan African Social Prospects Centre (PSPC) held a conference on ‘Africa and the French Revolution: Lessons for the 21st Century’ from 21 to 24 November 1989.

4

MECHANISMS

81

support for whatever project or policy question they were working on. Moreover, the technique was so popular that it became difficult for leaders to avoid this strategy when making important decisions, because it would appear that the will of the people was being ignored. (Seely 2009, p. 64)

Thus, public consultations in the form of Estates-General were held by most presidents on different topics. The first, the États Généraux de l’Education (Estates-General on education), already took place in 1990 under the provisional government. Another was organized by the Soglo government in January 1993 and covered the territorial administration of the country. Others were organized to discuss issues relating to the modernization of the public service, the reform of the military, judiciary, public health, the national economy, vaccination, or Catholic education. The most recent one, the Etats Généraux de la Société Civile was held in December 2018 to discuss the ‘strengthening and participation of civil society’. Overall, these Estates-General, each of them a mini National Conference, ‘enhanced political efficacy and improved popular awareness about specific problems’ (Heilbrunn 1999, p. 232). Each conference further ‘reinforced sentiments that the government was committed to engage in a dialogue with its constituents’ (Heilbrunn 1999, p. 232). Thus, the broad mobilization of civil society, eager to participate directly in politics, combined with the high level of procedural legitimacy which resulted from participatory decision-making, led to the establishment of practices of public-civil society consultation which no president has dared break. Over the years, this practice of public consultation became informally institutionalized without ever being formally adopted. How NVR Levelled the Playing Field in Chile In Chile’s democratic transition, the way in which NVR levelled the playing field through creating executive constraints and improving conditions for civil society was less straightforward than in Benin. Contrary to the assumption that after bringing down the autocratic ruler civil society would be a strong force during and after transition, capable of keeping political developments on a democratic track, Chile instead saw the demobilization and fragmentation of its civil society, especially among unions and student movements but also more widely (Delamaza 2015; OECD 2016). Camila Jara Ibarra describes this process in sobering terms:

82

D. LAMBACH ET AL.

Despite the effervescence of the end of the military dictatorship, a gradual eclipse of civil society movements and organizations occurred once democracy was restored. Thus, the return to democracy was characterized by the fragmentation, weakening, and crisis of social movements, the demobilization and deactivation of civil society, the decline of social movements, the citizen withdrawal, a development without citizen participation, and even electoral abstention. (Ibarra 2016, p. 89)

The decline of civil society activism can be traced back to 1986 when the Civic Assembly, a movement of trade unions and other civic groups, organized protests and called for a general strike. These acts of resistance were met with severe repression by the state, with several student leaders being killed (Nepstad 2011). An attempt on President Pinochet’s life by a communist guerrilla group and the discovery of a large shipment of Cuban arms to the same group provided the regime with an excuse for its brutal response to the protest. These events also weakened the capacity of the nonviolent movement to mobilize amid speculation among some of the protesters that the resistance could turn violent (Nepstad 2011; Huneeus 2009). After 1986, there was a sense among the more radical forces in the opposition (such as student unions or slum dwellers), which were influential in grassroots organizing, that NVR had failed as a method to achieve a full transformation to democracy. Some of the more progressive segments of the movement started to demobilize (Ibarra 2016). Whereas most of them eventually supported the ‘no’ vote and some joined demonstrations to support it and helped to restore destroyed electoral registers, they viewed the plebiscite as an elite-led process which aimed to continue a socio-economic path set out by the authoritarian government and which failed to achieve their more revolutionary goals (Ibarra 2016; Garretón 2016). In sum, there was a strong sentiment that popular mobilization had not been successful in pursuing a profound transformation of society and politics. There were other factors that contributed to a demobilization of civil society. One was the overall move from underground oppositional politics to a political system of representative democracy, which materialized after 1989, at least based on the minimal standard of fair and free elections. This move, together with a political culture dominated by elites and oriented towards stability and consensus, limited the entry points and

4

MECHANISMS

83

the potential for political participation for civil society during the transition (Ibarra 2016; Hecht Oppenheim 2007). The second cause of demobilization was the individualistic political culture created by the military regime’s extensive neoliberal reforms. This economic system had led to an ‘expansion of the market and the privatization of public services’ (Ibarra 2016, p. 90) and created a highly consumerist culture, which was associated with declining interest in politics and engagement in political movements (Garretón 2016; Ibarra 2016; Lechner and Guell 1998). Linked to this, and with the loss of a common agenda and the disappointment with the failure of a greater transformation, civil society could not sustain the unity of the 1986 Civic Assembly. Similarly, the heterogeneous coalition that was the 1988 ‘no’ campaign fragmented into different groupings, including a human rights movement, environmental movements and a movement for homosexual rights (Garretón 2016; Bertelsmann Stiftung 2016). As a result, the focus of the less radical part of the pro-democratic opposition, often represented by oppositional pro-democratic party elites, was to beat the regime at its own game by winning the plebiscite and the first election. And so they did. But this manner of ‘winning’ on the terms set out by the autocratic regime came at a cost. By dictating the rules of the game, the Junta managed to avoid abrupt change even though its candidates lost the plebiscite and the elections (Huneeus 2014). For the pro-democratic ‘winners’ of the transition, this created the onerous task of painstakingly negotiating incremental changes to level the playing field, a process that took decades, while disadvantaged by the electoral system, undemocratically elected senators, and powerful veto players such as the military and economic elites. Indeed, the developments closely resembled what comparative politics has to say about pacted transitions which enable elites to control the transition so that ‘institutions are custom-made for a particular person, party or alliance’ (Przeworski 1991, p. 82). Thus, due to its profound demobilization and fragmentation, civil society was in no position to pressure the key actors of the autocratic regime, which still held powerful positions in government, the economy, and the military and strangled democratic progress. There was a clear comparative advantage for actors close to the Junta in a transition which followed the timeline and procedure laid out in the 1980 constitution. Rather than leveraging the nonviolent pressure from the street, pro-democratic elites in government and the less radical segments of civil society favoured a

84

D. LAMBACH ET AL.

culture of compromise to accommodate old autocratic elites, while making incremental progress towards democratic consolidation in the first two decades after transition. In other words, the particular nature of the transition, built on elite negotiation with the political right, conflict aversion, and concern for overall political stability, led to favouring some dimensions of democratic consolidation (e.g. stability of political institutions) over others (e.g. participation). Moreover, the divide between those members of the pro-democratic movement who moved into institutional politics and those who either continued their work in NGOs or demobilized grew incrementally and led to a split between civil society and the political sphere. Important negotiations over constitutional changes occurred between the referendum on the extension of Pinochet’s rule in October 1988 and the presidential elections in December 1989. The main parties involved in these negotiations were the military government, moderate right-wing parties that supported the government, and the centre-left opposition. In contrast to Benin, there was no way for citizens and CSOs to participate and they had no influence on the outcome. They were not consulted during the negotiations but neither were they making their voice heard in the streets. In the constitutional referendum in June 1989, citizens could choose only between the status quo of the autocratic 1980 constitution and the proposed changes that came out of the elite negotiations. Perhaps civil society’s largest window of opportunity for real change during the early transition opened after the plebiscite in late 1988. With substantive input from civil society groups, a platform drafted the initial programme of the emerging centre-left party coalition (Concertación) for their electoral campaign and their first potential term in office. Their demands concerning the deep democratization of society and the party system were discussed with minimal programmatic interference by party elites. However, the restrictive type of consensus-based politics which elites preferred would undercut most of the more progressive proposals (Delamaza 2015, pp. 240–247). Thus, opposition elites were not particularly enthusiastic about bringing in the more far-reaching proposals from the consultative process into negotiations with right-wing parties and most radical demands were not included in the 1989 referendum on constitutional change. This limited involvement of civil society in transition politics had profound and lasting effects on the outcome of Chile’s transition to democracy and subsequent democratic consolidation. The constitutional

4

MECHANISMS

85

changes resulting from the referendum in 1989 did involve some executive constraints which were, however, designed to protect the prerogatives of the military and influence of right-wing parties and former authoritarian elites. As a result, the deal that was negotiated by political elites, which determined the post-transition political system, did not create an institutional environment where citizens and CSOs could operate freely. Instead, military repression continued, albeit with decreasing intensity, as were certain restrictions on media, which only later lifted. Especially in the first decade after the transition, Chilean politics was still gripped by a culture of fear that continued to weaken civic mobilization. The situation changed only slowly, with little citizen influence. The most significant changes regarding the overall balance of power between the government and the opposition came about with the 2006 constitutional reform which included abolishing the position of unelected senators, making the National Security Council more independent from the military, granting the president the right to appoint the leaders of the four main security branches, making appointments for members of the constitutional court more democratic, and strengthening the role of the court to exert jurisdiction concerning the constitutionality of laws (Hecht Oppenheim 2007, p. 242; Fuentes 2015; Bertelsmann Stiftung 2016). However, these reforms were not the result of path-dependent, long-lasting effects of the NVR campaign but of strategic interactions among political elites. They were undoubtedly significant for Chile’s democratic development but were made possible because the political right’s strategic interest had changed since the transition. In other words, the right ‘had stopped benefitting from the rules of the game it had designed’ before the transition (Siavelis 2016, p. 65). These strategic interests are well covered by Fuentes (2015), who outlines the manifold factors that allowed for these changes: a context which saw Pinochet’s and the military’s influence and reputation decrease after 1998, the anticipation that a right-wing victory in elections might be possible and thus that overly close links with the old military regime might be counterproductive, a loss of control over appointing senators by the right after successive electoral defeats and weakening ties to judicial actors and the military, constant negotiation efforts by pro-democratic political elites, and lastly the Chilean peculiarity of the so-called ‘binomial’ voting system which provided a safety net for the right-wing Alianza coalition. Thus, there was only a partial levelling of the political playing field at the beginning of the 1990s after a successful plebiscite and two elections.

86

D. LAMBACH ET AL.

Indeed, it took years of changing political context and constant negotiation by pro-democratic elites to rid the country of some of its autocratic vestiges and to instal further checks and balances. This of course does not diminish the importance of NVR efforts, which had helped to enable the transition. Pinochet and his confidants no longer ruled the country after 1988 and right-wing parties kept losing national elections for two decades after 1989. However, the direct influence of the NVR movements on the outcome of the transition in general and constitutional changes in particular was small. There were instances of civil society activists entering party politics, but these moves were often shaped more by their commitment to a particular party rather than by their civil society activism (Delamaza 2015). Concluding Summary of the Mechanism and Evidence from Other Cases Our analysis shows that NVR-induced transition usually sees a levelling of the playing field by means of checks on the executive and improvements in the institutional environment for civil society. However, casespecific evidence highlights that deposing a dictator by means of NVR is by itself not sufficient to generate a substantial levelling effect. Instead, it is important that NVR movements sustain mobilization and push for influence and inclusion in transitional negotiations about the new rules of the game. This appears to be the main difference between Benin and Chile, which we see as responsible for the different democratic trajectories of these countries after transition. However, as the example of Benin shows, even if NVR levels the political playing field during the transition period, other power disparities persist. There, economic elites poured into politics, converting their status and their financial resources into political capital, recreating established practices of political clientelism. Clearly, NVR is not a magic bullet solving all of society’s ills and inequities. Our two cases of top-down transition, Paraguay and Cape Verde, illustrate how the absence of popular mobilization leads to transitions where political power structures mostly persist. In Paraguay, the regime of General Alfredo Stroessner, which came to power through a military coup, ruled from 1954 until 1989 and was the longest serving dictatorship in Latin America during the twentieth century. During his rule, Stroessner established a tightly knit system connecting Paraguay’s Armed Forces, the Colorado Party, and the state administration. The regime provided

4

MECHANISMS

87

the façade of a representative democracy, yet was based on a permanent state of exception which limited civil and political rights (Powers 1992; Nickson 2010). The political culture of patronage and repression during this period is expressed by Stroessner himself: ‘rewards for friends, beatings for enemies, and for those who are indifferent, the law’ (quoted in Britez 2015). The decline of Stroessner’s regime started in the early 1980s due to a mix of factors. Alongside a deepening economic crisis, a lack of credible succession mechanisms for the ageing general is often described as the most important factor in this process (Abente Brun 2009). Disunity in the Colorado Party which had started to emerge at the beginning of the 1980s eventually led to a coup in February 1989 by a softline faction, the so-called ‘tradicionalistas ’ (López 2019). The short three-month timeframe allowed for holding the subsequent national elections and an unreformed political system meant that the 1989 elections were not fair, and only partly free. They were won by Andrés Rodríguez, the main protagonist of the coup, with a sweeping 74% of the vote. A new liberal constitution and a reformed electoral system provided an institutional base for more democracy in 1992. However, those reforms were only fully implemented and adhered to with the third presidential elections (again won by the Colorado Party) in 1998, which are seen as the first general elections which can be described as free, nonfraudulent, and competitive (Duarte Recalde 2017). It was only in 2008, 20 years after the coup, that Paraguay saw a peaceful turnover of power, when the Colorado Party lost the presidency to the Alianza Patriótica para el Cambio (APC), an alliance of left-wing parties backed by the centrist Partido Liberal Radical Auténtico (PRLA) (Abente Brun 2009; López 2019). In 2012 the power shifted back to the Colorado Party in free and fair elections. However, these elections were only called after a political crisis had led to an arguably unconstitutional impeachment procedure against President Fernando Lugo in 2012, which came at the cost of political sanctions by Paraguay’s neighbours (Ezquerro-Canete and Fogel 2017, p. 291). A multiparty system did emerge as democracy began to take root in Paraguay, although this was at least partly due to internal divisions within the Colorado Party. The Colorado Party nonetheless attempted to monopolize power and its entrenched position as the hegemonic party was eroded only slowly. It is still the most powerful political force today (Zubizarreta 2013; López 2019). The Colorado Party’s continuing dominance is complemented by a lack of accountability mechanisms for the

88

D. LAMBACH ET AL.

parliamentary opposition and civil society. Civil society, traditionally weak and fragmented, was unable to make its voice heard during the transition (Lambert 1997; Potthast and Sosna 2008) and struggles to this day with a lack of resources, a lack of connections between national and local activism, strong Colorado Party influence on many fields related to political life, and a media sector controlled by the political right (Sondrol 2007; interview with local civil society activist, Asunción December 2017). This can be partly explained by the elite-led transition because of which the military was still in a position to pressure people into submission until the early 2000s (Sondrol 2007; Frentes Carreras 2012). On the other hand, civil society has still played a vital part in several episodes of civic mobilization against democratic regression (see next section). Cape Verde provides an even clearer example of how top-down transitions leave the political playing field uneven. There, the ruling Partido Africano da Independência de Cabo Verde (PAICV) instituted the transition to democracy itself in 1990. PAICV had been unsettled by an internal generational conflict, the so-called ‘Trotskyite crisis’ between old cadres and younger, more educated party cadres coming back from their studies in Lisbon, as early as the late 1970s. Many of these younger cadres left the party and went abroad. The paternalistic leadership style of PAICV was also coming in for public criticism. For example, in 1981 the famous music group Bulimundo released their song ‘Dimõcracia’ in which they took a swipe at the government for prohibiting critical and independent thinking beyond the party lines (interview with Zeca Nha Reinalda, Bulimundo singer, Praia, 9 June 2017). PAICV’s ‘postindependence pragmatism’ (Meyns 2002, p. 158) made it sensitive to these signs of discontent and it started to liberalize the political system. In the 1985 parliamentary elections, independent candidates were allowed to run on the PAICV list. One of them, the future prime minister and leader of the Movimento para a Democracia (MPD) Carlos Veiga, was voted into the National Assembly. A programme of economic liberalization and electoral reform ‘to increase competitiveness’ was decided at the third PAICV congress in 1988 (Lima 1992, p. 296). The next two years mostly saw steps towards economic liberalization, followed by political liberalization from 1990 onwards (interview with Aristides Lima, former member of the National Assembly, Praia, 31 January 2017). This culminated in the decision of the fourth PAICV party congress to tolerate opposition parties, the adoption of a new party law, the foundation of the MPD, and finally National Assembly elections on 13 January 1991 (Lima 1992, p. 297).

4

MECHANISMS

89

The PAICV surprisingly lost the first free multiparty elections to the MPD. And although the electoral defeat represented a loss of power and economic resources for PAICV (interview with Pedro Pires, former Prime Minister and PAICV leader, Praia 26 May 2017) and resulted in some reshuffling within the party, it did not signify a major power shift in Cape Verdean politics as the MPD was mainly made up of ex-PAICV cadres who had left the party during the Trotskyite crisis, including the first MPD President and later Prime Minister Carlos Veiga and current President Jorge Carlos Fonseca. Since 1990, the political landscape has been divided between PAICV, the party which brought independence, and the MPD, which brands itself as the party which brought democracy. Other political parties play only marginal roles at the national level and none has won more than 10% in parliamentary elections. Civil society played no significant role in the transition or in post-transition politics. Our cases where violent revolutions brought about democratic transition, Namibia and El Salvador, add some ambiguous evidence to this mechanism. In Namibia, SWAPO (originally the South West Africa People’s Organisation) completely remade the political landscape in its own image whereas in El Salvador, the 1992 peace accord removed all veto powers or political enclaves, providing a relatively good basis for politics to not be dominated by a single actor. Clearly, the dynamics of transition, with the resistance achieving military victory in Namibia compared with a negotiated transition in El Salvador, had an impact on the post-transition political playing field. In both cases we see some effects on civil society that are in line with our theoretical expectations, especially in Namibia where the armed struggle had a negative effect on the development of an independent civil society. The national liberation movement tried to unite all political activity underneath the roof of the party structure since it feared a threat to its legitimacy as the sole and authentic representative of the people. Namibia and El Salvador provide some evidence that in violent resistance cases, resistance movements leave behind CSOs that are more supportive than critical of the new regime (e.g. veterans’ associations) or that are politically neutral (service delivery organizations). Having said that, there is a lively tradition of civic activism in El Salvador, where a generation of activists emerged in the 1980s and struggled for social justice despite the country’s history of military dictatorship.

90

D. LAMBACH ET AL.

Advancing a Democratic Civic Culture and Resistance as a Symbolic Reference Point Having looked at the establishment of institutions that enhance the capabilities of citizens to hold the government accountable, we now look at citizen’s attitudes towards democracy and their willingness to hold the government accountable. Beyond their effect on political institutions, we expect NVR movements to set the stage for a democratic civic culture to emerge after transition. Civic culture connects individual attitudes to the political institutions of a democratic regime. Ideally, the attitudes and values of citizens are consistent with and benevolent towards the political institutions and actors and thereby stabilize democracy. This relates to the argument that a democratic regime is considered consolidated if democracy is perceived as ‘the only game in town’, i.e. where citizens and elites have internalized democratic norms to such a degree that non-democratic forms of governance are not considered credible alternatives. We suggest that participating and experiencing the transformative power of NVR has a lasting effect on the political attitudes of citizens and elites. Large-scale formative events such as democratic transitions and revolutions foster a participatory identity and feelings of political efficacy among citizens, which in turn advance a positive attitude towards democratic norms and values. We expect that this emerging democratic civic culture translates into political behaviour: the nonviolent struggle to achieve democracy becomes a symbolic reference point which activates citizens for future resistance against authoritarian backlash. We did not engage in large-N comparative analysis for this mechanism. First, comprehensive survey data on political attitudes is not available with sufficient coverage across time and space. Second, survey response styles vary across cultures, which makes it difficult to compare data from different countries. Third, survey data suffers from desirability bias and only captures intentions and not actual political behaviour (e.g. Ostroot and Snyder 1985; Van Vaerenbergh and Thomas 2012). Therefore, we mostly rely on case-specific evidence to analyse the effect of NVR on civic culture. Numerous examples support our assumption that NVR advances a democratic civic culture. For instance, the nonviolent struggle of the Solidarno´sc´ movement in Poland, which eventually achieved democratic transition in 1990, became a part of the collective memory of Polish civil society that continues to resonate even today. Poland developed what

4

MECHANISMS

91

Ekiert and Kubik (2001) called a ‘rebellious civil society’ in which protest became a legitimate and regularly used method to articulate grievances and protect democracy (see Landé 2001 on a similar effect in the Philippines). In recent years Polish civil society, among other things, has mobilized against judicial reforms intended to weaken democratic checks and balances. At mass demonstrations, Lech Wał˛esa, the former Solidarno´sc´ leader, addressed the crowds and reminded them: ‘Our generation led Poland to freedom in an incredibly difficult situation and based it on the separation of powers. This is the most important thing that we managed to do. If anyone wants to disturb this most important victory, you, the young people, cannot let that happen. So that there is no doubt, I will always be with you, despite my condition, even if they arrest all of you here’.11 A similar development occurred in the Philippines, where the ‘people power revolution’ of 1986, which deposed President Ferdinand Marcos by means of NVR, served as a symbolic reference point for the later protests against the corruption of President Joseph Estrada in 2001, which forced him to step down (Landé 2001). We might also turn to Ukraine, where the 2004 Orange Revolution laid the groundwork for the 2014 Maidan protests. NVR need not even be successful to become an important cultural touchstone. The 2019 street protests in Hong Kong, still ongoing at the time of writing, drew inspiration from the ultimately unsuccessful 2014 ‘Umbrella Movement’. In our theoretical approach, we build on research from political psychology and political behaviour on how formative events can have a longterm effect on political attitudes and behaviour (Doherty et al. 2019; Opp 1998; Murdie and Purser 2017).12 Opp (1998) studied the impact of participation in the East German protests in Leipzig that led to the fall of the communist regime of the German Democratic Republic in 1989. His findings show that individual participation in anti-regime action in 1989 persistently advanced democratic political attitudes among citizens for years after the events occurred. Related to this, Murdie and Purser (2017)

11 CNN Transcript: http://edition.cnn.com/TRANSCRIPTS/1707/23/cnr.22.html. 12 Related research also shows that generally protest participation, especially if

nonviolent, can have long-term effects on political attitudes and behaviour (McAdam 1989; Mazumder 2018; Sherkat and Blocker 1997).

92

D. LAMBACH ET AL.

show that a country’s experience with nonviolent protests advances individual support for freedom of association. Successful democratic transitions induced by NVR become a collective memory that fosters individual willingness to protect democratic rights.13 To substantiate these findings, we will now investigate the effect of NVR on civic culture in more detail for the cases of Benin and Chile. The Development of Civic Culture in Benin After Transition In contemporary Benin, diffuse support for democracy is strong both in absolute terms and relative to other countries in the region. According to data collected by Afrobarometer, a steady 70–80% of the population preferred democracy over any other kind of government in successive surveys conducted between 2005 and 2018. In 2018, 79% of the population preferred an accountable government over one that ‘gets things done’ which is astonishing in view of Benin’s dire economic situation and given the fact that the general satisfaction with democracy is not very high (roughly 40%). Thus, it is definitely not the performance or the ‘output’ of the democratic governments that convinces the Beninese people and legitimizes the democratic rule, but rather the ‘input’ or the idea of democracy in principle (Scharpf 1999). This democratic orientation is driven by a widespread rejection of other forms of undemocratic government. In 2018, 71% of the population rejected military rule and 79% oneparty rule. A staggering 92% rejected one-man rule (Afrobarometer). This can be explained by Benin’s history of dictatorships and coups: ‘Because what the people experienced for 17 years has left a mark on the Beninese citizens who didn’t want to experience this period ever again’ (interview with Jean Baptiste Elias, Front des Organisations Nationales Contre La Corruption, Cotonou, 8 November 2016). This democratic conviction can be traced back to the history of NVR. One effect of the resistance to the regime of Kérékou was that people learned to ‘value institutional pluralism’ (Allen 1992, p. 48). For instance,

13 However, Doherty et al. (2019) show that political attitudes do not always translate

into changes in political behaviour. Using data on protest participation during the Tunisian revolution and voting in the subsequent Constituent Assembly elections, Doherty et al. (2019) find no evidence that such formative event activate citizens. Their results indicate that those who participated in the protests were not substantially more likely to vote than those who did not.

4

MECHANISMS

93

when wages went unpaid every other month at times in the 1980s, people turned towards the leaders of the state-led unified union. These leaders found themselves in the difficult situation of being accountable to their own members on the one hand, while being employees and members of the regime on the other. The union was thus unable to fulfil its function of addressing the grievances of its members, which served as an ‘objective lesson’ for many, illuminating the ‘necessity for the reform of the system’ (Allen 1992, p. 48). Since then, the Beninese have valued independent organizations highly. With rising frustration about the problem-solving capabilities of the socialist institutions and rising repression against what was believed to be justified demands, grievances became more and more politicized (Banégas 1995, p. 10). In the face of government repression, the rule of law and civil liberties such as the protection from arbitrary arrests gained particular significance for the population. In the course of the resistance campaign, the coordination and the effectiveness of the movement was significantly increased by the Bureaux de Liaison and the Comités d’Action (Banégas 1995, p. 13). Together, the successful resistance and the path-breaking National Conference led to an increased feeling of efficacy (Heilbrunn 1999, p. 232; Mbaku 2003, p. 39). Similar to most important transitions, many participants were proud of ‘what we achieved in terms of trying to overwhelm the political system that definitely failed and to put in place one that is completely viewed as a superior one’ (interview with Guillaume Moumouni, former student activist in 1989, now Lecturer at the Faculty of Law and Political Science, University of Abomey-Calavi, Cotonou, 15 October 2016). Pride in having been the first mainland African country to manage the transition to democracy peacefully prevails today. In his commencement speech, the general rapporteur of the National Conference, Albert Tévoédjrè, described the incredible nature of the process that had taken place: And yet, suddenly this whole country […] found itself in the GeneralEstates [National Conference]. Here, men and women who probably didn’t know each other, some of whom hated each other, who had diverging interests, and who were divided in their suffering, met for days and nights to cry out their hope together.14

14 Own translation. See http://democratiebenin.over-blog.com/article-rapport-generalde-la-conference-nationale-1990-73929678.html.

94

D. LAMBACH ET AL.

After the transition, the National Conference soon became the ‘anchor point’ and a ‘founding myth’ for the new democratic system (interview with Alexander Stroh, Chair of African Politics and Development Policy, University of Bayreuth, 5 October 2016). Until today, the constitution derives much legitimacy from the fact that it represents the consensus of the National Conference. Thus, both the National Conference and the constitution today serve as symbols for national unity, peace, and the empowerment of the people. As Célestine Zanou, a candidate in the 2006 presidential elections, later put it: (W)hat remains of the National Conference is the consciousness of a political class that knows how to reconcile with the people in times of great danger; it is the memory of a leadership that has been able to challenge itself to save the essential; it is faith in the creative genius of a community of destiny that refuses its fate. In short, it is a vast national commitment […] It must be taken action to safeguard this spirit. (Zanou 2016)

In recent years, there have been moves to preserve and formalize the remembrance of the National Conference so that: the Beninese [may] take inspiration thanks to the consensus and the spirit of dialogue inherited from the National Conference […]. May the rulers commit themselves to celebrate annually and around a general theme the National Conference with the aim to perpetuate this heritage which we must start to teach in our schools and universities so that its accounts and acknowledgement finally become part of our customs in Benin. (Zanou 2018)

To mark the occasion of the 30th anniversary of the National Conference in 2020, a civil society initiative is mobilizing support and lobbying the government to preserve the Hotel Alédjo, the venue of the National Conference, which has fallen into disrepair. In their statement they declare it ‘a sacred place that deserves to be preserved among our identity heritage. It is a place that must be visited by the younger generation and future generations […]. This place is indeed […] the symbol of national

4

MECHANISMS

95

unity’ and thus should be renovated and declared a national heritage and a UNESCO World Heritage Site.15 Benin has developed an engaged civil society willing to defend democratic institutions in general and the constitution specifically, and it frequently harks back to these symbolic reference points (Adjovi 2006, p. 190). According to Jean Baptiste Elias, president of the Front des Organisations Nationales Contre La Corruption (FONAC), civil society activists ‘know very well that the essentials should be preserved. The essential being absolutely the struggle for the perpetuation of democracy’ (interview with Jean Baptiste Elias, FONAC, Cotonou, 8 November 2016). This struggle to uphold democracy always refers to the original struggle for democracy. As Martin Assogba, president of the Association de Lutte Contre le Racisme, l’Ethnocentrisme et le Régionalisme (Association of the Fight against Racism, Ethnocentrism and Regionalism, ALCRER) put it when speaking of attempts by political elites to deviate from the democratic path: ‘we [organized civil society like ALCRER] remind them of what brought us to the conference of active forces and that we do not want to fall back into the habits which we had and which led us to rise up’ (interview with Martin Assogba, ALCRER, Cotonou, 21 December 2016). The first important civil society intervention in defence of democracy was the ‘touche pas à ma constitution’ (don’t touch my constitution) campaign by the NGO ELAN in 2003. The NGO was set up by Reckya Madougou who went on to occupy several ministerial posts in the Boni Yayi government between 2008 and 2013. ELAN appeared overnight and started a media campaign pasting ‘touche pas à ma constitution’ posters all over the city of Cotonou. Their goal was to prevent a constitutional amendment that would have allowed President Kérékou to run for a third presidential term. This sparked a noisy public debate and the mobilization of civic forces. Citizens from all social strata and diverse organizations including unions, women’s organizations, and academic groups joined the protests. Thirty-three of these organized themselves in the Front des Organisations de la Société Civile pour les Elections Transparentes et Démocratiques (Front of Civil Society Organisations for Transparent and Democratic Elections, Adjovi 2006, p. 191). In July 2005, Kérékou 15 Own translation. Original French petition online: https://www.change.org/p/ ressuscitons-l-h%C3%B4tel-plm-al%C3%A9djo?fbclid=IwAR3eW-mXfNAnxsJJqXcNnuN0m9CaUjIRvEiwv5cWaRTwXbrJDc9K68SWYE.

96

D. LAMBACH ET AL.

finally announced that he would not seek a third term in office. He substantiated his decision referring to the situation in 1990: ‘If we express the desire to stay in power, or that we persist to stay against the will of the people, we create the troubles that Benin avoided in 1990’ (cited in Gbadamassi 2015). Nevertheless, the situation escalated when the government announced that elections would be rescheduled due to financial difficulties. Again, CSOs mobilized in defence of democracy demanding that ‘the people want to vote in March 2006’ (Adjovi 2006, p. 193). These protests were further complemented by a 48-hour strike and an attempt to collect private money to organize the elections (Dossa 2005). The elections took place on 5 March 2006 and President Kérékou, who had finished his second term in office and had reached the age limit of 70 years, stepped down and peacefully handed over power to Thomas Boni Yayi. This experience of collectively defending the democratic consensus was burnt into the memory of many a pro-democratic citizen in Benin (Stroh 2013, p. 5) and serves as a narrative for remobilization. Between 2014 and 2016, civil society was back in action to pressure President Boni Yayi to hold elections on time and to refrain from revising the constitution. People rallied under the same slogan, ‘touche pas à ma constitution’, albeit this time under ALCRER’s leadership. Once again civil society ‘held marches […] to demand that the elections be held because we made decisions at the conference of the forces of the nation, this decision must be respected’ (interview with Martin Assogba, ALCRER, Cotonou 21 December 2016). Among the organizations mobilizing against a constitutional referendum and a possible third term of Boni Yayi was the newly founded Alternative Citoyenne, founded by Joseph Djogbenou.16 Inspired by the ‘colour revolutions’ in other parts of the world, the Alternative Citoyenne introduced a new disruptive technique to the existing repertoire of contention by organizing the so-called Mercredi Rouge (Red Wednesday), meaning that people regularly gathered on Wednesdays dressed in red. Like his predecessor, Boni Yayi finally stepped down in 2016, handing over power to his successor Patrice Talon.

16 Djogbenou, then a professor at the faculty of law and political science of the University of Abomey-Calavi and president of the NGO Droits de l’Homme, Paix et Développement (DHPD), later became president of the constitutional court. His movement transformed into a political party.

4

MECHANISMS

97

The Development of Civic Culture in Chile After Transition On citizens’ commitment to democratic values, the picture emerging from survey results for Chile is mixed. Approval for democracy as the preferred political system over any other system fluctuated between 50 and 65% between 1995 and 2018 and was slightly below the regional average until 2003. After civic remobilization and substantial constitutional reforms in 2005, it has been at or above the Latin American average for 9 out of 13 years since then. The most recent data from 2018 shows a relative consolidation of democracy. For example, only 6% of Chileans believe that their political system is not a democracy (second lowest in the region), and 43% believe that their democracy has only small problems (best value in the region). This does not translate into any great satisfaction with the actual functioning of democracy in Chile (only 43% of respondents in the 2018 survey). But even this has been at, or above the regional average for all but three years since 2004—the 2018 average for Latin America was only at 23%—while it had been mostly below the regional average before 2004 (Latinobarómetro 2017, 2018, 2019). In short, in recent years, Chileans seem generally more content with the functioning of their democracy than their neighbours, although their level is satisfaction is not particularly high in global terms. Other trends show a generally declining approval for autocracy as a viable form of government and a continuously declining opinion of Pinochet’s former regime and its human rights record. They further show declining acceptance of military coups, which seems to underline strong and improving democratic attitudes among Chileans (Navia and Osorio 2019). Clearly, pro-democratic attitudes slowly gained ground in Chile since 1989, although their development is less impressive in absolute than in relative terms. However, as Navia and Osorio (2019) suggest in a detailed analysis of attitudes towards democracy, approval for democracy has remained more or less stable across a longer time span (pre-Pinochet, during Pinochet’s reign, and after the re-transition to democracy). The authors argue that this development can be explained by a range of factors such as Chile’s long democratic experience in the pre-Pinochet era, and stable support for democracy across time among Chile’s middle-class and left-wing voters. The experience of autocracy from 1973 to 1989 and the transition to democracy did not seem to substantially affect these attitudes. Figures on pro-democratic attitudes after the transition also have to be seen against the background of 43% of voters who still voted for Pinochet in the 1988

98

D. LAMBACH ET AL.

referendum. These findings indicate, overall, that there appears to be no direct link between NVR and democratic attitudes in Chile. Instead, one negative trend—the crisis of political representation— clearly stands out and can be qualitatively linked to the pacted nature of the transition, which shows the limited impact of Chile’s nonviolent protest on political attitudes. Citizens’ trust in political parties has deteriorated constantly, dropping from 32% in 1995 via 22% in 2001, 16% in 2008, and 15% in 2014 to 11% in 2017 (Latinobarómetro 2018, 2019). These figures are below the regional average for a majority of years since 2004. Such a disconnect between the social and political party spheres in Chile is also well demonstrated in a quote by a former Concertación politician: [The] impressive level of detachment of the political parties from their role as intermediary with the public […] is a product of the rules of this constitutional system [which made and partly still does make it hard to achieve structural change] and the long duration that the political left remained in power …. The Concertación parties became close to the state because they were in government, and lost proximity to the public [and] their social expressions and the capacity for dialogue [with the public]. (Interview with former Concertación politician, Santiago de Chile, December 2016)

Furthermore, as the previous sections have shown, civil society was unable to mitigate this deficiency of political representation because it was to a large extent demobilized and fragmented, but also split between those who joined the political ranks and those who stayed outside of formal politics. Civil society could thus not act as an intermediary between the social and political spheres. This estrangement is well expressed by a top Socialist Party leader talking about his relation to former comrades in the pro-democracy NGO he left behind when he joined the government after the 1989 election: ‘Informally we met once a month to talk, […] certain sorts of relationship remained, but definitely, when you change jobs, you create other groups and other ties’ (quoted in Delamaza 2015, p. 238). A prominent human rights activist presents another side to the story: ‘the parties have captured some people who came from the resistance movement and the rest of us became marginalized even though we remained motivated, but left out…’ (interview with human rights activist, Santiago de Chile, February 2017). The latter depicts a trend

4

MECHANISMS

99

examined in a large empirical study by Gonzalo Delamaza that demonstrated that civil society activism and representativeness of particular segments of society were not decisive criteria in recruitment for positions in the Concertación after the transition. Rather, prior political commitment, party colour and technocratic knowledge were regarded as more important (Delamaza 2015, p. 233). This trend of detachment between citizens and institutionalized politics seemed to take a turn with the 2006 elections won by Michelle Bachelet Martínez, a socialist party leader with a background in the antiPinochet resistance who had campaigned for more participatory democracy (Navia 2008; Weeks and Borzutzky 2012). However, she largely failed to implement her participation agenda due to a lack of support from within the Concertación coalition and in the face of a number of political crises. This dashed renewed hopes for meaningful democratic reform and led to increasing political discontent and magnified emerging student protest at the time (Cummings 2015). This notwithstanding, since the mid-2000s various welfare and education reforms have been implemented or discussed which aim at moderating the strongly neoliberal economic model (Bertelsmann Stiftung 2016, 2018). In sum, Chilean society displays a general belief in democracy as the best system, more so than its neighbours and particularly since the mid2000s, but Chileans’ overall attitudes towards democracy have remained remarkably stable over decades and across regime types. Therefore, it is difficult to establish a link between NVR and political attitudes. Furthermore, we find no evidence that the capacity of post-transition civil society to threaten the regime with NVR leads to higher government responsiveness, or that close links between political parties and civil society allows for participatory politics. A crisis of representation has been haunting Chile for many years, which can be explained by the outcome of a partially pacted transition which led to institutional restraints, elite detachment from society, and the politics of stability rather than participation. Initially launched by middle school students in 2006 and later spreading to universities, post-transition protests reached their peak between 2011 and 2013 (Ibarra 2014; Garretón 2016). Citizens’ support for such protests was significant, hovering between 70 and 80% (Delamaza 2015, p. 286), suggesting a general acceptance for methods of NVR 20 years after the original transition. An example is the unprecedented solidarity student movements experienced from broad parts of society after their demonstrations were met with a particularly harsh police response on 4

100

D. LAMBACH ET AL.

August 2011, reminding many Chileans of the military dictatorship. In a spontaneous action, organized through various social networks, thousands of people in Santiago de Chile and throughout the country started a Cacerolazo (noise riot), in which wooden spoons were beaten against iron pots and pans to express solidarity with the protesters. This form of protest had been frequently used under the military dictatorship (Labarca 2016, p. 623). The protests were led by a new generation which had come of age after the fall of the Pinochet regime. This generation was different from the ‘civil society generation’ of their parents which had mostly succumbed to the 1990s consensus of stability politics and economic growth to avoid a backlash by the military. This new ‘fearless’ generation (‘la generación sin miedo’), had no direct experience with dictatorial repression and harsh authoritarian restraints on everyday life, and did not think that largescale civic mobilization would destabilize the democracy their parents had fought for. This fearlessness empowered protesters and also gave them a common identity in the face of institutional politics (Cummings 2015). It may be possible to trace discourses of empowerment among the current generation back to the legacy of the NVR movement, especially among more left-wing groups, although this would require further and more detailed research. The impetus behind the protests resonates closely with the transformative goals of the more radical segments of the pretransition resistance, in that the demonstrations were directed against Chile’s entrenched neoliberal economic model, an ideological project of the military Junta’s rule which had endured for 20 years after the transition to democracy. As one of the leaders of the student protests states: In what I advocate, I feel like I continue the fight that they [the movements in the 1980s] were fighting. We feel that we are a part of this legacy of socio-political struggle that, although it was dispersed [after 1990, …] is now revitalized. (Interview with former student activist, Santiago de Chile, November 2016)

One of the most prominent issues identified by the protesters was inequality, particularly in relation to social services (e.g. education). In a system introduced during the dictatorship, the lower-income population gets free but low-quality public services, the middle class have access to better public services through co-payment, and the wealthy can afford

4

MECHANISMS

101

high-quality private services (Ibarra 2014). In 2013, on the 40th anniversary of the 1973 coup, a spokesperson of the Coordinating Assembly of High School Students stated: The slogan for commemorating this date is ‘40 years of resistance’. We students have been affected by the educational system we inherited from the dictatorship and that the Concertación and the Alliance perfected in recent years. (El Mostrador 2013)

So while it is possible to find strands that connect contemporary civil society activism to the resistance movement of the 1980s, there are two major caveats. First, it is important to note that like with memories of repression, memories of the resistance differ widely among socio-political camps. Generally speaking, right-wing perspectives tend to stress incidences of violence within the resistance movement whereas left-leaning opinions emphasize the movement’s heroics. Moderate viewpoints tend to focus on the pacific nature of the transition to democracy (email conversation with Professor Claudio Fuentes, December 2016). Second, as much as the new generation of protesters might have drawn inspiration from their predecessors, their mobilization was driven by current developments. Thus, the fact that poverty was decreasing and that expectations for better education were rising ties in with classical theories of social mobilization which point to gaps between expectations and capabilities as drivers of mobilization (e.g. Gurr 1970, see Cummings 2015 for an elaboration on this argument for Chile). The failure of the first Bachelet government to meet its commitments on education reform further mobilized students (Cummings 2015). In broader terms, it also meant a clash between the institutionalized elite politics of stability and governability and the lived reality of the younger student generation. The elite’s political style of not rocking the boat was widely accepted when the military still had substantial veto powers in the 1990s. However, after significant constitutional reform and declining military influence, the new movements questioned whether the old compromise politics were still relevant and called for popular participation to set the political agenda (Castiglioni and Rovira Kaltwasser 2016, p. 18). Lastly, internal dynamics within the student protest also contributed to their success, with protesters able to connect themselves across different political and economic backgrounds (Donoso 2013).

102

D. LAMBACH ET AL.

In sum, a new generation of civil society rediscovered the tactics of mass mobilization from 2006 onwards and directed their discontent against aspects of inequality produced and sustained through the neoliberal economic system inherited from the former autocratic regime. This system had been sustained by both right- and left-wing party coalitions in parliament after 1989. Indeed, as suggested by the theory mapped out in Chapter 2, the newly emerging student movements drew on the ‘spirit’ of resistance which had made democracy possible in the first place. This link was mostly of a symbolic nature, surfacing in discourses legitimizing the new resistance. Moreover, solidarity for the young resisters among older generations was arguably also influenced by their experience of past dictatorial repression. On the other hand, young protesters’ ability to mobilize was in part influenced by their lack of direct experience of repression: this made them less fearful of a backlash from the security forces. When protests reappeared, particularly after 2010, conflict lines from the 1990s transition had shifted and political polarization and division can be found between cycles of social movement mobilization and institutional party politics more broadly, rather than between pro-democratic and autocratic forces (interview with former student activist, Santiago de Chile, November 2016). Therefore, we find that civil society mobilization after 2006 draws on the social justice and pro-democracy discourse of the anti-Pinochet resistance but that this represents only one of the narratives of the protest movement. Concluding Summary of the Mechanism and Evidence from Other Cases In sum, the evidence from Benin and Chile underscores the variation in the strength of impact that NVR may have on a transition. Political developments in Benin indicate the NVR movement’s lasting impact on improvements in civic culture through its involvement in transition politics. The movement created a symbolic reference point of empowerment, which advanced political efficacy and the willingness of citizens to defend democracy and hold elites accountable. In Chile, by contrast, the NVR movement had less impact on the pacted transition and similar direct and path-dependent developments are more difficult to identify. While the NVR-induced transition in Benin produced a positive narrative of empowering ordinary citizens, NVR did not have the same effect in Chile. Here, people power was not sufficient to achieve full democratic

4

MECHANISMS

103

change. It took 15 years after transition for civil society to remobilize on a sizable scale. In terms of the timeframe, this might seem comparable to remobilization in Benin, where such instances also occurred more frequently from the 2000s onwards. But the difference between the two cases lies in the nature of the mobilization. Where Beninese civil society mobilized in defence of democracy against an executive that was challenging the institutional checks and balances, mobilization in Chile coalesced around demands for more democracy. This fits with our findings of the development of democratic quality post transition (see Fig. 3.4). While Benin, with its more impactful resistance movement, saw rapid democratic gains, these same gains occurred more slowly in Chile. By comparison, democratic transitions that came about through violent resistance campaigns have also created symbolic reference points but not the kind of inclusive and participatory political culture that is conducive to civil society activism. For instance, in Namibia, which achieved democratic transition along with its independence in 1990, the armed struggle for independence led by SWAPO still represents a powerful symbolic reference point. However, contrary to Benin, where the discursive reference to the NVR against Kérékou and the peaceful transition mostly serves as a reminder to stick to the constitutional rules, discourse on the armed struggle for independence and democracy in Namibia is instrumentalized by SWAPO to uphold its dominant position among political parties. Leading SWAPO politicians frequently use the rhetorical question ‘where were you when we were fighting?’ to silence criticism from the opposition or even younger generations of party activists. The heroic deeds of the past serve to claim absolute moral authority and to deny political opponents the right to criticize certain policies or personal opinions. A leading figure of the opposition All People’s Party (APP) says: Particularly the DTA [Democratic Turnhalle Alliance]17 has problems with engaging SWAPO because, as soon as the DTA starts with the work they

17 The DTA, renamed in 2017 to Popular Democratic Movement (PDM), is a political party in Namibia. It was originally founded after the Turnhalle Constitutional Conference in 1975 which was assembled to devise a constitution for an independent country South West Africa. At this time, the country was still occupied by South Africa. The conference failed and the DTA was formed as a political party that rallied the moderate forces of the country. The DTA participated in the 1978 elections and consequently governed the country in cooperation with the occupying forces of South Africa. Today, it is therefore often regarded as party that collaborated with the South African apartheid regime.

104

D. LAMBACH ET AL.

get reminded of us where were you? You were fighting against us and now you want to tell us and teach us things. You don’t have the moral authority to stand up and tell us, the freedom fighters. So that thing is still there. It is still used whether it is in the villages or wherever, that those ones that fought actually. Their only moral authority to dominate it is just the statement that I was in the struggle you cannot question my judgement, you cannot question me…. (Interview with a leading politician from the APP, Windhoek, 2 March 2016)

Related research by Isaak-Finhold (2018) shows that references to ‘struggle’, ‘fight’, or ‘liberation struggle’ are widely used in parliamentary debates in Namibia, usually by SWAPO members to undergird their presumed right to rule. SWAPO’s 1994 election manifesto, for example, reminds the electorate that: ‘it was SWAPO that was exiled, it was SWAPO-members who were imprisoned, and it was SWAPO-members who sacrificed their lives for the liberation of the country’ (Swapo 1994, p. 5). These feelings of entitlement based on heroic deeds in armed struggle undermine the idea of equal citizenship and therefore the basic idea of democracy itself (Bayer and Pabst 2018).18 In El Salvador, political culture is similarly exclusionary although this is less visible at the level of parliamentary politics. The democratic transition was a long, drawn-out process after a violent insurrection against the military dictatorship in the 1970s. From a constitutional assembly in 1982 and the first election of a civilian president to 1984 to the 1992 Chapultepec Peace Accords, there were many steps along the way (Córdova Macías et al. 2009; Zinecker 2004). While this has created a stable political system, El Salvadorean democracy was long dominated by two major parties, the Frente Farabundo Martí para la Liberación Nacional (FMLN) and Alianza Republicana Nacionalista (ARENA), representing the major forces of the civil war, with much of the senior personnel of these parties drawn from the former rebel movement and the military, respectively. In contrast, civil society has been fragmented and weakened since the end of the civil war. Different sectors of society often align with one of 18 This part of the Namibian experience is comparable to Zimbabwe where the ruling

Zimbabwe African National Union (ZANU) under Robert Mugabe used a similar strategy of political entitlement to fortify its rule (Scarnecchia 2006, pp. 233–234). In this case, the rhetoric of entitlement determined ZANU’s exclusive claim to power in Zimbabwe, which, among other things, eventually led to democratic breakdown in the form of an autogolpe in 1987.

4

MECHANISMS

105

the major political parties, e.g. labour with the FMLN and business with ARENA, providing little space for civic-minded mobilization and strategic alliances across social boundaries. This is not to say that civil society is entirely marginalized—the human rights movement has been particularly influential in pushing for transitional justice for official crimes and in memorializing the victims of the regime. Also, there was a surge of new organizations founded by former insurgents, many of them focusing on the old issue of land access but also covering other issues such as the environment and gender equality.19 This shows how not just NVR but also armed insurgency can bequeath ‘a culture of citizen mobilisation and participation among sectors of society that were previously excluded from political life’ (Alvarez 2010, p. 36). But these examples of civic activism aside, there are few opportunities for citizens to get involved in politics. Analysts mourn the closing down of spaces for political dialogue and consensus finding and highlight the continuing polarization and the lack of readiness to compromise among the dominant parties. As outlined by Córdova Macías, Loya Marín, and Ramos, there is a strong imbalance between political power holders and unrepresented voices in the country, whereby the former do not seem to see the necessity of including the latter, or at least of negotiating public policy with them: ‘one basic aspect is that the principal political actors have to recognise or accept the “other” actors as valid interlocutors, who have the right to have a different perspective and vision’ (own translation, Córdova Macías et al. 2009, p. 167). Apparently, the publicly lauded learning process triggered by the peace accord—the search for compromise and political negotiation—has not had long-term effects. Evidence from our cases of elite-led democratization also offers partial support of this mechanism, especially in Cape Verde where the broader population was not involved in the transition to democracy. Instead, the liberalization of the single-party regime was initiated by the ruling PAICV and later negotiated with the newly formed MPD. Although the democratic transition is celebrated every 13th of January as Freedom and Democracy Day, the transition had little influence on the political culture and did nothing to empower civil society as an autonomous force. According to a study based on Afrobarometer data, civil society is still

19 We are grateful to Alvaro Martín Alvarez for pointing this out to us during his review of our case study on El Salvador.

106

D. LAMBACH ET AL.

in its infancy in Cape Verde since ‘very few Cape Verdeans are members, active or inactive, of trade unions, professional associations, or community development associations. Citizens do not tend to get involved in longer lasting and more organized collectivities’ (Ames et al. 2003, p. 12). This is not necessarily a consequence of the top-down mode of transition and might also be explained by Cape Verde’s socialist past of obligatory collective organization in the state-led unified union. However, in Cape Verde reluctance about collective organization is paired with feelings of low political efficacy through which citizens ‘feel quite incapable of influencing the functioning of the political system’ (Ames et al. 2003, p. 13). As Leao de Pina from the Instituto de Ciências Jurídicas e Sociais says: ‘People didn’t have this culture to face the government and to do demonstration. So, after five hundred years of Portuguese [rule] without the participation of society and almost 15 years after independence without this kind of culture, it is a kind of cultural thing’ (interview with Leao de Pina, Instituto de Ciências Jurídicas e Sociais, Praia, 31 May 2017). CSOs in Cape Verde do not see themselves as an external check on government but rather as its ‘helping hand’. They are mostly active in the areas of development, education, or environmental protection. According to Mario Moniz, secretary of Plataforma das ONG’s de Cabo Verde, the national NGO platform, ‘all these actions complement the actions of public institutions’ and target regions where ‘public institutions have some difficulties in getting there’ (interview with Mario Moniz, secretary of the Plataforma das ONG’s, Praia, 5 May 2017). In Paraguay, there are multiple structural constraints on civic activism. First, civil society is mostly excluded from formal policy processes. Second, approval rates for democracy are lower than the regional average while approval rates for authoritarianism as a political regime are relatively high (Latinobarómetro 2016, p. 13). Third, the educational system does little to inculcate democratic values and glosses over Paraguay’s autocratic past. In short, Paraguay’s post-transition civil society is held back by several factors which can in part be attributed to the elite-led transition and persisting features of its former autocratic regime. However, although the odds seemed to be stacked against Paraguayan civil society, it has managed to mobilize multiple times at critical junctures of Paraguay’s democratic trajectory (Larrouqué 2019). Three instances stand out. First, during the so-called Marzo Paraguayo (Paraguayan March) in 1999, a wide array of cross-party societal groups assembled on the main square after the assassination of the

4

MECHANISMS

107

Vice-President, Luis Maria Argaña (Colorado Party). The killing was attributed to his intra-party rival President Cubas and his powerful backer, former army general Lino Oviedo. Sustained street protests led to the resignation of the government and Oviedo fled the country (López 2019). Second, the mobilization of left-leaning civil society and peasants also helped the first non-Colorado president, Fernando Lugo, to power in 2008. He was elected after a political crisis in which the outgoing Colorado President Nicanor Duarte Frutos had attempted to change the constitution to prolong his time in power (Uharte Pozas 2012). Third, in 2017 massive protests under the slogan ‘dictadura nunca más ’ (dictatorship never again) helped to prevent the Colorado President Horacio Cartes from amending the constitution to prolong his mandate (Larrouqué 2019). Thus, although civil society in Paraguay seems restrained in influencing democratic consolidation, it is strong enough to unite in critical moments to defend Paraguay’s democratic system, providing a ‘backstop’ for democratic survival. This suggests that although the struggle for a better democracy is far from over, a critical mass of Paraguayans is not interested in returning to an authoritarian past, challenging our assumption that NVR-induced transitions have a more positive impact on the development of a democratic political culture.

Avoiding the Praetorian Problem As our third mechanism of democratic consolidation, we evaluate whether and how NVR influences civil–military relations in a way that mitigates the praetorian problem. As described in Chapter 2, NVR movements win the support of the armed forces by establishing direct contact with soldiers, offering them a future in a democratic system while avoiding any threat of violence towards the military. Therefore, we expect that in a subsequent democratic regime, military influence in politics will be reduced. In contrast, violent resistance movements that achieve democratic transition usually create a loyalist and politicized army (i.e. revolutionary civil–military relations) stacked with their own cadres and constituencies. Such settings of revolutionary civil–military relations foster incentives to expand executive power and reduce democratic checks and balances, which may eventually lead to democratic breakdown by means of autogolpe. Finally, eliteled transitions often produce civil–military relations in which the armed forces remain aligned to the old authoritarian regime and its successor parties and are likely to defect from the democratic equilibrium by means of coup.

108

D. LAMBACH ET AL.

.1 0

probability of coup

.2

We begin our analysis of civil–military relations by analysing the frequency of coups and their success rate across different modes of transition. Our total sample of democratic regimes featured 77 coups d’état, of which 40 were successful. Of the 77, only 19 occurred in regimes that came about by means of NVR while 50 coups occurred after elite-led transitions and another eight after violent resistance transitions. However, as discussed before, NVR-induced regimes only make up about one-third of our sample and therefore these absolute numbers do not tell us much. To address this issue, we estimated the average probability of a coup attempt for each year after transition for regimes induced by NVR and regimes without this feature.20 Figure 4.4 describes the results for the first ten years after transition.

1

2

3

4

5

6

7

8

9

10

time after transition

Fig. 4.4 Probability of coups after transition Notes: Grey markers show the estimates for regimes that transitioned via NVR and black markers refer to transitions without NVR. Spikes show 95% confidence intervals 20 Detailed results are reported in the appendix section A7.

4

MECHANISMS

109

For the first ten years after democratic transition, the estimated probability of a coup attempt is lower in regimes that came about by means of NVR than in regimes that democratized without NVR. However, as the overlapping confidence intervals indicate, the data provide no conclusive evidence about this trend. Moreover, in absolute terms, the difference in probabilities is rather small. For instance, at five years after transition, the probability of a coup attempt is estimated at 4% for NVR-induced regimes and 7% for regimes without this feature. Thus, NVR seems to have little impact on the probability of coup attempts. However, when looking at the success rate of coup attempts, a different picture emerges. Figure 4.5 reports the frequency of successful and failed coups across regimes induced by NVR and regimes without this feature. The success rate of coups is substantially different for regimes with and without NVR-induced transition. In regimes that democratized without NVR, coups succeeded about 60% of the time. If democratization was NVR

40

No NVR

23 20

frequency

30

35

10

14

0

5

fail

success

Fig. 4.5 Success rate of coups

fail

success

110

D. LAMBACH ET AL.

achieved by means of NVR, the coup success rate is only 26%. This finding indicates that NVR-induced regimes are substantially more robust in terms of resisting democratic breakdowns caused by coups.21 This lower success rate of coups in NVR-induced democratic regimes cannot be attributed to more democratic attitudes among military personnel in these regimes, because in this case we would also expect a lower frequency of coups. Instead, we see two possible explanations. The first is that while changes in civil–military relations may not fully convert the armed forces to the cause of democracy, they create fissures in the military. If we model a coup as a ‘coordination game’ (Singh 2014), where the coup plotters have to act on incomplete information about who will support or oppose them, a divided military makes it more likely that wouldbe coup leaders overestimate the strength of their support. The second explanation is that there is broader resistance to coups from civil society and political elites. Coups are usually legitimized by referring to the ‘will of the people’. Such justifications are less plausible after successful democratization via civil resistance. Additionally, democracies that come about by means of NVR can rely on a stronger civil society which is both capable and willing to defend democracy against coups. Zunes (2017) describes civil resistance against coups in Argentina and Burkina Faso in support of this argument. After years of brutal military rule, Argentina achieved a democratic transition driven by NVR in 1983. The newly elected government of President Alfonsín launched a series of investigations and prosecutions against military officers suspected of involvement in the so-called ‘Dirty War’ (1976–1983), in which death squads went after leftists and other dissidents. These investigations led to the sentencing of several military officers that spurred parts of the army to revolt in April 1987, demanding an end to the trials. In response, civil society mobilized in large numbers in the streets to support the Alfonsín government and even marched into army bases, forcing officers to surrender. Protesters explicitly referenced their intention to fight for democracy and against military intervention into politics. Under societal pressure and with limited support from the rest of the armed forces, the mutineers caved in.

21 We obtain similar results if we consider violent transitions as separate category.

4

MECHANISMS

111

In Burkina Faso, a military coup was launched on 16 September 2015, just one year after a civil resistance campaign known as Le Balai Citoyen (The Citizen’s Broom) had ousted the long-term dictator Blaise Compaoré in October 2014, initiating steps towards democratic transition. The coup conspirators were members of an autonomous military unit named Regiment of Presidential Security, originally created by Compaoré to counterbalance the regular army. Prior to the coup, a civilian commission had recommended dissolving the unit, describing it as ‘an army within an army’ (Bonkoungou and Bavier 2015). However, facing intense pressure from civil society and eventually the regular armed forces, the Junta was not able to consolidate power. Civilian rule was finally restored on 23 September 2015. While coups d’etat represent extreme cases of the breakdown of civil– military relations, the problem of praetorianism can also manifest itself in other, more subtle ways that undermine democracy. These include military prerogatives such as the power to name representatives in the legislature, de facto or de jure veto positions on certain policy issues, the ability to control its own funding, or a lack of legal accountability. Looking at the transition and subsequent political development in Benin and Chile, we also find some effects of NVR on these aspects of civil–military relations. Civil–Military Relations in Benin Infamous for their multiple interventions into politics, the armed forces of Benin were an important veto player during the transition to democracy, especially because Kérékou himself had gained power through a coup d’état and his regime and the military were closely entangled. From 1976 onwards, military leaders occupied key positions in the bureaucracy and the economy (Akindes 2015, p. 51). However, through the ‘Quota Wars’, civil society was able to reduce the number of delegates from the armed forces to the National Conference, limiting the military’s ability to participate in, let alone dominate, the negotiations about the transition to democracy. But the danger of a military intervention was still present during the National Conference, as illustrated by a statement of retired Colonel Maurice Kouandété in front of the delegates. Kouandété, who himself had led the military coups against General Christophe Soglo in

112

D. LAMBACH ET AL.

1967 and against President Émile Zinsou in 1969, declared that: ‘The [National] conference is zero. Everyone wants to destroy Kérékou, even the ex-Presidents. I was astonished to learn that you blame the army. If you blame the army, I will respond tomorrow’ (cited by Seely 2009, p. 69). The following night, the conference room of the PLM Hotel Aledjo was surrounded by tanks, probably members of the Presidential Guard, who threatened to kill the participants if they did not refrain from declaring the National Conference a sovereign assembly or if they continued to push for regime change (Banégas 1995, p. 19). However, by this point, the balance of power had already tipped in favour of the opposition and the military units withdrew. Over the year 1989, the opposition, especially the unions, had shown that they were unified in their goals and capable of refusing cooperation with the regime quite effectively. With the ability to call a general strike at their disposal, the unions were able to plunge the regime even deeper into economic misery. Additionally, far more radical demands for a social revolution and holding regime members accountable for their actions hung in the air, leading to a ‘radical flank effect’.22 As the following open declaration of a National Conference delegate illustrates, everybody was well aware of this fact: We all know here that we are supported by the pressure of strikes which constitute supreme forces. Do not procrastinate (about sovereignty). We can do it because the Strike pressure is there and will not go away after the Conference. (Unknown delegate of the National Conference cited by Banégas 1995, p. 19)

By contrast, the army’s unity had dissolved. Previous attempts to use the military against protesters had proved that the military was far from unified and not a reliable pillar of support of the autocratic regime. There were several reasons for this. First, in an attempt to control and counterbalance the military, Kérékou had transformed the ceremonial Batallion de la Garde Présidentielle ‘into a highly trained guard, and equipped it

22 The radical flank effect refers to the positive (or negative) effect of a radical flank on the outcome of a conflict (Haines 1984; Tarrow 1998). In this case, the PCB, which did not participate in the National Conference, represented the radical flank of the movement, which had a positive effect on the National Conference and the more moderate parts of the movement.

4

MECHANISMS

113

with some of the best material the armed forces had to offer’ and created a new militia, the People’s Militia, which ‘was recruited from among civilian partisans of the regime for added protection’ (Morency-Laflamme 2018, p. 472). By contrast, the regular army was ‘proletarianized’, ‘illequipped’, and ‘rather inclined to protest’ (Banégas 1995, p. 5). The unity of the armed forces was further reduced because under Kérékou leading officers occupied influential positions in the bureaucracy and the economy while the younger officers and ordinary soldiers did not benefit to the same degree (Akindes 2015, p. 51). As one anonymous retired military officer put it: ‘Even within the army’s ranks, there were soldiers with grievances and individuals who sympathized with the protesters were easy to find’ (cited by Morency-Laflamme 2016, pp. 134–135). By the time of the National Conference, the army was already split between hardliners and moderates. The leadership and certain former officers like Colonel Kouandété were closely connected to the socialist regime, while younger officers and soldiers were not (Banégas 1995). The nonviolence of the resistance movement was therefore instrumental in deepening divisions in the armed forces and preventing a ‘rally around the flag’ effect in the military (Sharp 1973; Chenoweth and Stephan 2011, pp. 48–50). Also, in past military interventions, the army had typically intervened on behalf of civilian politicians or had stepped in after public protests by students and trade unions. However, this time, students and unions were challenging the regime but large parts of the armed forces were not willing to align with the ruling elite (Akindes 2015, p. 46). As Morency-Laflamme (2018) argues, the unified opposition movement made credible assurances to the armed forces that its personnel would not be prosecuted for their ties to the regime and that they would have a key role to play in the future. Assurances consisted of a blanket amnesty for all military and civilian members of the former regime and an honourable retreat for the army. Appeasement set the tone in the final declaration of the National Conference which ascribed the responsibility for the misdeeds of the socialist regime mainly to the already dissolved single party. In contrast, the defence forces’ involvement in politics from 1972 onwards was portrayed as having been ‘justified by a concern and imposed by duty’ (Declaration of the Armed Forces at the National Conference of the Active Forces, in: Friedrich Naumann Foundation 1994, p. 123). These credible assurances paved the army’s way back to the barracks. Acknowledging the historical role of the National Conference and in order to safeguard national unity, the army further

114

D. LAMBACH ET AL.

committed itself to the new ethics of political neutrality in the future (Declaration of the Armed Forces at the National Conference of the Active Forces, in: Friedrich Naumann Foundation 1994, p. 119). In a similar vein, Kérékou concluded the conference by saying that ‘this is no capitulation, it is a matter of national responsibility’ (Concluding speech to the National Conference by Kérékou, in: Friedrich Naumann Foundation 1994, p. 111), indicating that a return to the barracks should not be seen as a defeat. Several articles of the revised constitution explicitly address civil–military relations. Article 65 defines any attempt to overthrow the constitutional order by military personnel as a crime against the nation. Article 66 defines a duty to resist for any Beninese citizen and even allows them to call in external military support, a drastic step for Benin, which has suffered external interventions in the past. The constitution further prohibits candidates for an official (political) post, including the presidency, from holding a position in the military. Today, the norm that the military must submit itself to the elected civilian officials ‘is turning into reality in Benin’ (Akindes 2015, p. 39) and due to ‘an active opposition and civil society, this [military intervention] has not occurred’ after 1990 (Akindes 2015, p. 56). These developments have been safeguarded by professionalization: as a kind of democratic dividend, many democratic countries like France, the US, and Germany entered into military cooperation agreements with the Forces Armées Béninoises (FAB). In the course of such cooperation, many Beninese soldiers have been trained and educated abroad and have come into contact with soldiers from other democratic nations, their professional ethos and their doctrines of leadership. Furthermore, the FAB has participated in international peacekeeping and peace enforcement operations (in the Democratic Republic of the Congo, Côte d’Ivoire and Mali), which represented welcome opportunities for additional training and to bolster the wages of the participating soldiers, opportunities which would be at risk if the FAB were to be perceived as an unprofessional and unreliable force (correspondence with Colonel Brillsauer, German military attaché for Nigeria, Benin, Burkina Faso, Côte d’Ivoire, Ghana, Mali and Togo, 5 September 2016). In sum, the evidence from Benin highlights that NVR had a direct and lasting positive effect on civil–military relations.

4

MECHANISMS

115

Civil–Military Relations in Chile The evidence from Chile shows that NVR movements are not always able to exercise a lasting positive impact on post-transition civil–military relations. Like Benin, Chile had previous experience of military coups in its early history, such as in 1924 and 1925, and the military was instrumental both in bringing Pinochet to power and making his downfall possible. During the 1988 plebiscite, Pinochet’s most high-ranking confidants in the military rejected the General’s attempt to manipulate media coverage and public perception in the hours after polls closed (Nepstad 2011, p. 86). Among the many factors which led to the end of Pinochet’s reign, the timely disobedience of high-ranking commanders was critical to ensuring that Chile was able to restore its democracy. However, this disobedience was not necessarily caused by the influence of civil resistance on the armed forces. Internal cohesion within the security forces was very strong, and contact between army personnel and resisters minimal. Thus, the NVR movement was not able to divide the armed forces during transition. Two factors explain the military’s behaviour. First, the results of the plebiscite had been widely leaked before communication could be controlled by the regime and so military elites abstained from rigging the result in Pinochet’s favour in fear of provoking a revolutionary backlash from civil society. This can be interpreted as an (indirect) influence of the NVR campaign. A radio station sponsored by the opposition was crucial in spreading an offhand statement from a regime official that the ‘no’ vote had won. Moreover, Nepstad reports that the opposition campaign sent a credible signal to regime insiders not to defect from the agreed rules and procedures of the plebiscite (Nepstad 2011, p. 90). Second, army privileges were constitutionally protected even in the case of democratic transition (Nepstad 2011, pp. 90–91; Huneeus 2014). This included legal stipulations to prevent the prosecution of past human rights abuses as well as clauses safeguarding the military’s political influence (Huneeus 2014; Fuentes 2013). This second factor highlights the failure of the NVR movement to substantially influence negotiations between the main veto players during transition, which later turned out to be a burden for the

116

D. LAMBACH ET AL.

new democratic regime. Chile’s first two democratic Concertación governments (1990–1998) faced resistance from a persistently powerful military and a political right still aligned closely to Pinochet (Hecht Oppenheim 2007). Hence, little headway was made on reforming the undemocratic institutions set out by the 1980 constitution. The 1991 Commission on Truth and Reconciliation (TRC), however, did represent some progress and can be seen as the first serious challenge to the military. The TRC achieved little in terms of judicial proceedings against human rights abusers in the military, only a few of whom were prosecuted, but it did manage to establish an official number of victims. It also produced evidence of the system of torture present under the military regime (Delanó 2011; Huneeus 2014). Prior to the establishment of the TRC, the issue of repression had led to the mobilization of citizens’ groups, such as the Agrupación de Familares de Detenidos-Desaparecidos (Association of the Families of the Detained and Disappeared) (Huneeus 2014). In the aftermath of the commission’s work, then President Aylwin reinforced his presidential powers, ultimately vetoing the Supreme Court’s decision to disallow the legal review of human right abuses by the military on the grounds of an amnesty law enacted in 1978 (Hecht Oppenheim 2007). Subsequently, some of the more publicized murders and abuses under the Junta came to be investigated. The military tried to defend itself, for instance by locking down the area around its headquarters in downtown Santiago de Chile in 1993 in order to discuss TRCrelated matters. This action raised fears in government that the armed forces were preparing for a coup d’état, but it also forced the governing coalition into dialogue with the military leadership over investigations of human rights abuses (Huneeus 2014). The continued role of the military in politics in the 1990s had a direct impact on the political culture and perpetuated a legacy of fear long after the first democratic elections. This fear had been internalized throughout years of severe repression and often inhibited people from speaking openly or participating in associations without anxiety. Ibarra even speaks of a ‘social trauma generated up to 1990’ (Ibarra 2016, p. 96). Hence, a generalized anxiety over confrontation with what remained of the military regime and fears of a renewed Junta loomed large. Therefore, political elites, the public and most CSOs shared a preference for consensus politics aimed at avoiding overt conflict (Ibarra 2016; Hecht Oppenheim 2007). Moreover, right-wing parties could still use the military court system to exert

4

MECHANISMS

117

influence in the aftermath of the plebiscite. Although the Aylwin government had managed to shift the jurisdiction regarding alleged insults against the military from military to civilian courts, other crimes such as offences related to defamation of (conservative) state authorities were still handled by the military judicial system and were still used throughout the 1990s to levy fines or short prison sentences against overly vocal leftists. Even though the number of offences prosecuted was significantly lower than during the military regime, it sustained a culture of fear among critical journalists and was counterproductive for the resurgence of a free and independent press (Human Rights Watch 1998). It was only in 2001 that most of these relics of the autocratic regime were removed (Freedom House 2008). Another essential step towards democratic consolidation was made with the Valech Commission. This commission was designed to investigate torture under the Junta and its creation owed in part to the unrelenting advocacy by human rights groups and families of victims (Hecht Oppenheim 2007, p. 240). Adding to the findings of the 1991 TRC— which had conducted its work under less favourable circumstances—the final report of the Valech Commission, published in 2004, laid bare the extent of systematic torture during military rule, which made it all but impossible for the military to shift the blame to ‘bad apples’ and away from the institution itself. High-ranking military figures finally started recognizing the abuses committed. A simultaneous process of professionalization and a generational leadership change within the military helped to normalize civil–military relations (Hecht Oppenheim 2007, pp. 240– 241). Lastly, important figures in right-wing parties realized at this point that blind loyalty to Pinochet’s legacy would not help them win support or any future elections, as public opinion (even on the right) with respect to the former autocratic regime became ever more unfavourable. This led to gradual distancing of the ex-dictator’s human rights legacy (Fuentes 2015) and facilitated (among other factors) a partial democratization of the constitution. In sum, the impact NVR had on civil–military relations in Chile indeed appears to be limited. Because of the militaristic nature of Pinochet’s regime, it was hard to speak of the armed forces as a separate entity. Instead, the fortunes of the armed forces were closely tied into the ruling regime. Resisters had little opportunity to establish constructive contact with soldiers. Moreover, the military government took good care of its own cadres by institutionalizing military influence on politics, protecting

118

D. LAMBACH ET AL.

itself against human rights prosecutions and keeping the former dictator as its commander-in-chief. In contrast to our theoretical assumption about NVR transitions, the military was still strongly involved in politics, based on constitutional stipulations. Similarly, right-wing parties and economic veto players close to the former regime also utilized the threat of the military to maintain a political status quo favourable to them. In fact, the danger of a coup or some other kind of democratic breakdown was quite palpable during the first years of the transition as Edgardo Boeninger, a leading Concertación politician, expressed in 1989: ‘the fear of military regression, and the understanding of the risk of such events occurring, will be directly determined by the level of conflict that exists between political parties’ (cited in Siavelis 2016, p. 69). In fact, as Laurence Whitehead states: ‘Full civilian control was therefore delayed by almost a generation (1988-2005) beyond the short period of the transition’ (Whitehead 2019, pp. 177–178). On the positive side, pro-democratic forces were able to challenge the military on some fronts, particularly pushing for transparency regarding human rights violations. Thus, to some degree, civil society was able to overcome the military veto pressure. The shift to a professional military, in which civilians make binding decisions over military matters and in which the government is able to exercise control over it, only came about in 2005, due to an array of factors. Some of these changes, especially the focus on human rights abuses, could be attributed to legacies of the NVR movement, with civil society groups who had campaigned on this issue for decades maintaining the pressure, whereas others, such as civilian oversight over the military, were more influenced by post-transition factors largely unrelated to civilian pressure. Concluding Summary of the Mechanism and Evidence from Other Cases Our analysis highlights that NVR-induced transitions generally have a positive impact on post-transition civil–military relations. However, again, a case-specific comparison of Benin and Chile shows differences with respect how substantial the impact of NVR is. In Benin, the NVR movement was able to co-opt the armed forces in order to achieve the transition to democracy, and afterwards the movement successfully established firm civilian control over the military. Due to the limited leverage that the NVR movement had during the transition in Chile, the armed forces

4

MECHANISMS

119

kept many undemocratic prerogatives and threatened military intervention to pursue their interest. Therefore, important reforms of civil–military relations were delayed for years. To contextualize these findings, it is important to also consider evidence from comparable counterfactual cases and to analyse how violent and elite-led transitions respectively influence civil–military relations. The way in which pacted transitions can lead to a praetorian problem is illustrated by the transition in Paraguay, where major political and constitutional reforms were strongly driven by elite negotiations among factions of the Colorado Party. The armed forces remained one of the most important political actors, capable of influencing the developments within the party (such as the choice of leadership) and the future direction of the political system. Most notably, the military helped rig votes, for example the general elections in 1993 and the constitutional assembly elections of 1991 (Frentes Carreras 2012). Although the armed forces technically began to be depoliticized with the introduction of the 1992 constitution (article 173 of the constitution clearly prohibits active service members from any political activity), in practice the military remained deeply involved in politics (Abente Brun 2012). It would take an attempted coup d’état (1996) and a high-profile political assassination sparking the socalled Marzo Paraguayo (1999) to finally strip the military of significant political influence. Civilian rule over the armed forces was enforced more strongly, the connections between the political elite and the military were curtailed, and budgets for the country’s armed forces were cut, effectively removing the military from political life by the beginning of the 2000s (Abente Brun 2012). Thus, Paraguay clearly exemplifies the challenges of a praetorian military after an elite-led transition, but the ‘praetorian’ crises that followed were also the precursor for further institutional reform to limit the military’s role on politics. The difficulties of establishing civilian control over the military after democratization by means of violent revolution are well illustrated by the case of Namibia.23 Namibia’s transition to independence and multiparty 23 Although we wish to stress that the Namibian model is not applicable to all other cases. It shares several commonalities with other cases where liberation armies were transformed into official military outfits, such as South Sudan or Timor-Leste, but our control case of El Salvador followed a different path. There, the rebels of the FMLN argued that the military should be downsized or abolished altogether and that human rights violators should be punished. However, we see an important difference in the modes of transition in Namibia and El Salvador. While violent resistance was the crucial driver of transition

120

D. LAMBACH ET AL.

democracy took place under the auspices of the UN. After nearly three decades of the struggle for independence, the United Nations Transition Assistance Group (UNTAG) was tasked to watch over the ceasefire agreement and the elections called for in UN Security Council Resolution 435. UNTAG was also tasked with monitoring the retreat of the South African Defence Force (SADF) and the demobilization of the former combatants from the People’s Liberation Army of Namibia (PLAN), SWAPO’s armed wing, as well as the former South West African Territorial Force (SWATF) and Koevoet units fighting alongside the SADF. Due to its violent past, civil–military relations in post-transition Namibia were conditioned by the need to demobilize and reintegrate roughly 65,000 former combatants (Metsola 2006, p. 1120) and to build a neutral and functional democratic army out of the different units. This was a daunting task. After the 1989 election, the new Namibia Defence Force (NDF) was inaugurated in the Defence Amendment Act (No. 20 of 1990). Thereby, the two biggest political parties—SWAPO and the Democratic Turnhalle Alliance (DTA)—initially agreed that former PLAN and SWATF forces were to be equally integrated into the new force to balance each other (Lamb 1998). The overall strength of the new army was set to be 10,000 men. However, this policy was soon unofficially abandoned by the Ministry of Defence. SWATF forces never made up more than ten percent of NDF forces and were mostly deployed in unimportant positions (Preston 1997, p. 459). However, even if nearly 9000 former PLAN fighters were integrated into the new NDF, some 20,000 were left without employment. These were first integrated into so-called development brigades, in which they were to acquire necessary skills for their reintegration into civilian life (Griffiths 1996, p. 476), a plan that was mostly unsuccessful. After serious protests in which former PLAN fighters marched through the capital of Windhoek demanding jobs and pensions, the Special Field Force (SFF) was founded in the late 1990s as a new branch of the Namibian Police (Nampol), incorporating ‘thousands

in both cases, Namibia saw a clear victory (albeit under UN auspices) of the insurgents whereas democracy in El Salvador emerged out of peace negotiations between the government and the FMLN, more closely resembling the dynamics of a pacted transition. Hence, it was never a realistic option for the FMLN to ‘stack’ the armed forces with loyalists. Instead, FMLN veterans were mostly integrated into the police force.

4

MECHANISMS

121

of ex-combatants [who] now [outnumber] other Nampol personnel by nearly three to one’ (Metsola 2006, p. 1127). This policy to integrate more and more ex-PLAN fighters into the security forces raised concerns about the neutrality of the army and newly created security forces, especially in the wake of SWAPO’s ‘spy crisis’ during its Angolan exile in the 1980s (Lamb 1998), which saw PLAN forces torture and kill suspected South African spies in its Lubango detainee camp (Saul and Leys 1995, 2003). These crimes were later covered by a blanket amnesty (Akawa and Silvester 2015) and one of the main culprits, Samuel Hawala, became commander of the NDF (Griffiths 1996, pp. 478–479). Since the SFF is almost exclusively made up of former PLAN fighters it is widely seen as an extension of SWAPO and sometimes even as a presidential armed force (Metsola 2006, p. 1127). It reinforced this reputation during the suppression of the Caprivi Rebellion in 1998 and 1999 (Metsola 2006, p. 1128). The security forces, foremost the SFF, used ruthless and brutal tactics against the Caprivi Liberation Movement (CLM), leading some 2500 local people to flee over the border into neighbouring Botswana (Melber 2009, p. 466). As a result of the armed struggle and the subsequent integration of PLAN fighters into the armed forces, civil–military relations in Namibia are marked by strong ties between the ruling party and the security forces. The loyalty of the defence and security forces to the Namibian state, as opposed to SWAPO, has never been put to the test, so it is difficult to determine whether this loyalty would be transferred to a different government if SWAPO were ever voted out of office. Relations between the armed forces and the local population are ambivalent. On the one hand, the army long enjoyed the cachet of being the unofficial successor to the liberation movement; on the other, it is seen as too close to SWAPO. Public perceptions deteriorated markedly when soldiers harassed and shot at innocent civilians during Operation Hornkranz, a law enforcement operation in 2018–2019. This sparked a public outcry and protests in the capital (Namibian Sun 2019). Also, the last round of the Afrobarometer in 2017/2018 shows a marked decline in public trust in the army with the number of respondents trusting the NDF ‘not at all’ more than doubling from the previous survey in 2014/2015.

122

D. LAMBACH ET AL.

General Remarks on Mechanisms and Case Studies Our assumptions about causal mechanisms are mostly supported by empirical evidence, both in terms of quantitative data and case-specific evidence. Our findings suggest that a levelling of the political playing field, the advancement of a democratic political culture and an avoidance of the praetorian problem are the mechanisms that connect NVR to democratic consolidation, although this does not rule out the possibility that additional mechanisms exist which we were unable to identify due to our selection of cases. Clearly, additional research would be useful to contextualize our results and identify possible scope conditions. Furthermore, our case studies reveal that there is substantial variation between cases of the same type. The differences between Benin and Chile (as NVR cases) were spelled out in this chapter. El Salvador and Namibia (as violent resistance cases) also diverged substantially in their democratic development, with a dominant party system emerging in Namibia and a polarized two-party system in El Salvador. Paraguay and Cape Verde (as cases of top-down liberalization) also followed somewhat different paths, with Paraguayan politics being much less stable, at least in the first decade after transition, than the relatively placid politics of Cape Verde. This might lead us to question the validity of our basic typology of cases (i.e. by form of resistance). It might also indicate that there are unobserved variables which we failed to include in our research design that explain some of this diversity. Based on our case studies, we believe that the variation in the ‘intensity’ (campaign size, duration, etc.) of NVR could be a likely candidate for such a variable, although conceptualized somewhat differently than in the existing studies. While we do not observe that aspects such as campaign size, duration, or organizational structure are relevant for subsequent democratic consolidation, the campaign’s capacity to influence transition politics appears of crucial importance for further political developments. Whereas in Benin, civil society determined the transition process and the design of new institutions such as the constitution, the resistance campaign in Chile had only a marginal influence on the transition process. This difference largely explains the variation in democratic consolidation with regard to the three mechanisms. Moreover, in Benin we observed an almost perfectly linear process of path dependency where NVR fostered events and decisions during the transition phase that shaped the process of democratic consolidation thereafter. By contrast, in Chile the traces of the NVR movement are less

4

MECHANISMS

123

clear-cut and do not necessarily imply a path-dependent effect. Our interpretation of this finding is that NVR movements need to sustain transitional mobilization to have a lasting impact on democratic political development. Moreover, this finding also relates to the results of the quantitative analysis. As described in Chapter 3, our quantitative analysis relies on the stable unit treatment value assumption, which requires that the ‘intensity’ of NVR-induced transitions is largely the same across cases. The case studies highlight that for the important dimension of how much NVR shaped the transition process this assumption appears questionable. While this means that our quantitative findings will have to be approached with some scepticism, it warrants further research to evaluate whether there is a general pattern suggesting that those NVR campaigns more involved in the transition process are better able to foster subsequent democratic consolidation relative to campaigns with less involvement. We explore this and other possible extensions of our research in the following chapter.

References Abente Brun, D. (2009). The Unraveling of One-Party Rule. Journal of Democracy, 20(1), 143–156. Abente Brun, D. (2012). Estalidad y calidad de la democracia. América Latina Hoy, 60, 43–66. Adamon, A. D. (1995). Le Renouveau démocratique au Benin. Paris: Édition L’Harmattan. Adjovi, E. V. (2006). Mobilisations citoyennes et démonopolisation du travail politique au Bénin. Perspective Afrique, 1(3), 187–223. Akawa, M., & Silvester, J. (2015). Waking the Dead: Civilian Casualties in the Namibian liberation Struggle. In J. Silvester (Ed.), Re-viewing Resistance in Namibian History (pp. 192–206). Windhoek: University of Namibia Press. Akindes, S. (2015). Civil-Military Relations in Benin: Out of the Barracks and Back—Now What. In M. Rupiya, G. Noyo, & H. Laugsen (Eds.), The New African Civil-Military Relations (pp. 38–63). Pretoria: APPRI. Allen, C. (1992). Restructuring an Authoritarian State. Democratic Renewal in Bénin. Review of African Political Economy, 54, 47–48. Alvarez, A. M. (2010). From Revolutionary War to Democratic Revolution: The Farabundo Martí National Liberation Front (FMLN) in El Salvador (Berghof Foundation Transitions Series No. 9). Berlin. Ames, B., Renno, L., & Rodrigues, F. (2003). Democracy, Market Reform, and Social Peace in Cape Verde (Afrobarometer Paper No. 25). Cape

124

D. LAMBACH ET AL.

Verde. https://afrobarometer.org/sites/default/files/publications/Working% 20paper/AfropaperNo25.pdf. Accessed 1 October 2019. Angel, A. (2007). Democracy After Pinochet: Politics, Parties and Elections in Chile. London: Institute of the Americas. Banégas, R. (1995). Action collective et transition politique en Afrique. La conférence nationale du Bénin. Cultures & Conflits, 17 (1), 1–26. Banégas, R. (1997). Retour sur une transition modèle: Les dynamiques du dedans et du dehors de la democratization béninoise. In J. Daloz & P. Quantin (Eds.), Transitions démocratiques africaines (pp. 23–94). Paris: Èdition Karthala. Banégas, R. (2003). La démocratie à pas de caméléon. Transition et imaginaires politiques au Bénin. Paris: Karthala. Bayer, M., & Pabst, A. (2018). Heroes and Victims: Economies of Entitlement After Violent Pasts. Peacebuilding, 6(1), 49–64. Bennett, A. (2013). The Mother of All Isms: Causal Mechanisms and Structured Pluralism in International Relations theory. European Journal of International Relations, 19(3), 459–481. Bertelsmann Stiftung. (2016). Bertelsmann Stiftung’s Transformation Index (BTI) (Chile Country Report). Gütersloh: Bertelsmann Stiftung. Bertelsmann Stiftung. (2018). Bertelsmann Stiftung’s Transformation Index (BTI) (Chile Country Report). Gütersloh: Bertelsmann Stiftung. Bethke, F. S., & Pinckney, J. (2019). Non-violent resistance and the quality of democracy. Conflict Management and Peace Science, forthcoming. https:// doi.org/10.1177/0738894219855918. Bierschenk, T. (2009). Democratization Without Development: Benin 1989– 2009. International Journal of Politics, Culture, and Society, 22(3), 337–357. Boas, T. C. (2015). Voting for Democracy: Campaign Effects in Chile’s Democratic Transition. Latin American Politics and Society, 57, 67–90. Bonkoungou, M., & Bavier, J. (2015, September 14). Burkina Faso Presidential Guard Should Be Disbanded, Panel Says. Reuters. https://uk.reuters. com/article/uk-burkina-army/burkina-faso-presidential-guard-should-bedisbanded-panel-says-idUKKCN0RE2BN20150914. Accessed 1 October 2019. Britez, E. (2015, August 16). Deberes. Abc color. http://www.abc.com.py/ edicion-impresa/opinion/deberes-1398350.html. Accessed 1 October 2019. Castiglioni, R., & Rovira Kaltwasser, C. (2016). Introduction: Challenges to Political Representation in Contemporary Chile. Journal of Politics in Latin America, 8(3), 3–24. Chenoweth, E., & Stephan, M. (2011). Why Civil Resistance Works: The Strategic Logic of Nonviolent Resistance. New York: Columbia University Press. Chilean National Commission Truth and Reconciliation. (1991). Report of the Chilean National Commission on Truth and Reconciliation. http://www.usip. org/sites/default/files/resources/collections/truth_commissions/Chile90Report/Chile90-Report.pdf. Accessed 1 October 2019.

4

MECHANISMS

125

Constable, P., & Valenzuela, A. (1989). Chile’s Return to Democracy. Foreign Affairs, 68(5), 169–186. Córdova Macías, R., Loya Marín, N., & Ramos, C. G. (2009). La contribución del proceso de paz a la construcción de la democracia en El Salvador (1992– 2004). San Salvador: FundaUngo. Cummings, P. M. M. (2015). Democracy and Student Discontent: Chilean Student Protest in the Post-Pinochet Era. Journal of Politics in Latin America, 7 (3), 49–84. Decalo, S. (1997). Benin: First of the New Democracies. In J. F. Clark & D. E. Gardinier (Eds.), Political Reform in Francophone Africa (pp. 43–61). New York: Routledge. Delamaza, G. (2015). Enhancing Democracy: Public Policies and Citizen Participation in Chile. New York: Berghahn Books. Delanó, M. (2011, August 20). Chile reconoce a más de 40.000 víctimas de la dictadura de Pinochet. El país. http://elpais.com/diario/2011/08/20/ internacional/1313791208_850215.html. Accessed 1 October 2019. Doherty, D., Schraeder, P. J., & Dobbs, K. L. (2019). Do Democratic Revolutions ‘Activate’ Participants? The Case of Tunisia. Politics, forthcoming. https://doi.org/10.1177/0263395719840240. Donoso, S. (2013). Dynamics of Change in Chile: Explaining the Emergence of the 2006 Pingüino Movement. Journal of Latin American Studies, 45(1), 1–29. Dossa, A. (2005, December 30). Financement de la Cena: Alcrer propose un guichet d’emprunt obligataire. Fraternité (Cotonou). https://fr.allafrica.com/ stories/200512300332.html. Accessed 1 October 2019. Duarte Recalde, L. R. (2017). Democracy and Representation in Paraguay. Partecipazione e Conflitto, 10(1), 65–88. Eboussi Boulaga, F. (2009). Les conférances nationales en afrique noire. Une affaire à suivrer. Paris: Édition Karthala. Ekiert, G., & Kubik, J. (2001). Rebellious Civil Society: Popular Protest and Democratic Consolidation in Poland, 1989–1993. Ann Arbor: University of Michigan Press. El Mostrador. (2013, September 11). Ocho colegios fueron tomados en Santiago para conmemorar los 40 años del golpe de Estado. El Mostrador. https://m. elmostrador.cl/noticias/pais/2013/09/11/ocho-colegios-fueron-tomadosen-santiago-para-conmemorar-los-40-anos-del-golpe-de-estado/. Accessed 1 October 2019. Ezquerro-Canete, A., & Fogel, R. (2017). A Coup Foretold: Fernando Lugo and the Lost Promise of Agrarian Reform in Paraguay. Journal of Agrarian Change, 17 (2), 279–295. Fomunyoh, C. (2001). Democratization in Fits and Starts. Democratization, 12(3), 37–50.

126

D. LAMBACH ET AL.

Freedom House. (2008). Chile: Freedom of the Press. https://freedomhouse.org/ report/freedom-press/2008/chile. Accessed 1 October 2019. Frentes Carreras, A. L. (2012). La Consolidacion Democratica en Paraguay. América Latina Hoy, 60, 67–82. Friedrich Naumann Foundation. (1994). Les actes de la conference nationale. Cotonou: Edition ONEPI. Fuentes, C. (2013). El Fraude. Santiago de Chile: Hueders. Fuentes, C. (2015). Shifting the Status Quo: Constitutional Reforms in Chile. Latin American Politics and Society, 57 (1), 99–122. Garretón, M. A. (2016). Crises de representación, movilizaciones sociales y eleccions presidentiales 2013 en Chile. In F. Mayorga (Ed.), Elecciones y legitimidad democrática en América Latina (pp. 15–37). La Paz: Plural Editores. Gbadamassi, F. (2015, October 16). Mathieu Kérékou, l’homme de la Conférence nationale. Franceinfo afrique. https://www.francetvinfo.fr/monde/ afrique/benin/benin-mathieu-Kérékou-lhomme-de-la-conference-nationale_ 3067493.html. Accessed 1 October 2019. Gerring, J. (2010). Causal Mechanisms: Yes, But…. Comparative Political Studies, 43(11), 1499–1526. Gisselquist, R. (2008). Democratic Transition and Democratic Survival in Benin. Democratization, 15(4), 789–814. Griffiths, R. J. (1996). Democratisation and Civil-Military Relations in Namibia, South Africa and Mozambique. Third World Quarterly, 17 (3), 473–486. Grynaviski, E. (2013). Contrasts, Counterfactuals, and Causes. European Journal of International, Relations, 19(4), 823–846. Gurr, T. R. (1970). Why Men Rebel. Princeton: Princeton University Press. Haines, H. H. (1984). Black Radicalization and the Funding of Civil Rights: 1957–1970. Social Problems, 32(1), 31–43. Hecht Oppenheim, L. (2007). Politics in Chile: Socialism, Authoritarianism, and Market Democracy (3rd ed.). Boulder: Westview Press. Heilbrunn, J. R. (1993). Social Origins of National Conferences in Benin and Togo. Journal of Modern African Studies, 31(2), 277–299. Heilbrunn, J. R. (1999). Corruption, Democracy, and Reform in Benin. In A. Schedler, L. Diamond, & M. Plattner (Eds.), The Self-Restraining State: Power and Accountability in New Democracies (pp. 227–244). Boulder: Lynne Rienner. Houngnikpo, M. C., & Decalo, S. (2013). Historical Dictionary of Benin. Lanham: Scarecrow Press. Human Rights Watch. (1998). Limits of Tolerance: Freedom of Expression and the Public Debate in Chile. Washington, DC: Human Rights Watch. https:// www.hrw.org/legacy/reports98/chile/index.htm. Accessed 1 October 2019. Huneeus, C. (2009). Political Mass Mobilization Against Authoritarian Rule: Pinochet’s Chile, 1983–88. In A. Roberts & T. Garton Ash (Eds.), Civil

4

MECHANISMS

127

Resistance and Power Politics: The Experience of Non-violent Action from Gandhi to the Present (pp. 197–212). Oxford: Oxford University Press. Huneeus, C. (2014). La democracia semisoberana. Santiago de Chile: Taurus. Ibarra, C. J. (2014). Democratic Legitimacy Under Strain? Declining Political Support and Mass Demonstrations in Chile. Revista Europea de Estudios Latinoamericanos y del Caribe (97), 25–50. Ibarra, C. J. (2016). The Demobilization of Civil Society: Posttraumatic Memory in the Reconstruction of Chilean Democracy. Latin American Perspectives, 43(6), 88–102. Isaak-Finhold, S. (2018). Namibia und Südafrika—Befreiungsbewegungen an der Macht. Maastricht: Shaker Verlag. Jackson, P. T. (2011). The Conduct of Inquiry in International Relations: Philosophy of Science and Its Implications for the Study of World Politics. London: Routledge. Johansen, J., & Vinthagen, S. (2019). Revolutionary Constructive Resistance— Benin 1989 in Context and Perspective. Nössemark: Irene Publishing. Kangnikoé, B. (2014). Verfassungsgerichtsbarkeit und Demokratisierung im frankophonen Westafrika. Länderstudie Benin (Franz von Liszt Institute Working Paper 2014/05). Giessen. http://intlaw-giessen.de/fileadmin/ user_upload/bilder_und_dokumente/forschung/westafrikaprojekt/ workingpapers/Draft_WP_2014_benin.pdf. Accessed 1 October 2019. Kérékou, M. (1994). Discours du président Kérékou à l’ouverture de la Conférence des Forces Vives de la Nation. In Friedrich Naumann Foundation (Ed.), Les Actes De La Conference Nationale. Cotonou, du 19 au 28 Février 1990 (pp. 14–21). Cotonou: Edition ONEPI. Kim, N. K., & Kroeger, A. M. (2019). Conquering and Coercing: Nonviolent Anti-regime Protests and the Pathways to Democracy. Journal of Peace Research, 56(5), 650–666. Koko, J. L. (2008). National Conference as a Strategy for Conflict Transformation and Peacebuilding. The Legacy of the Republic of Benin Model. London: Adonis & Abbey. Kurtz, L. (2009). Chile: Struggle Against a Military Dictator (1985– 1988). Washington, DC: International Center on Nonviolent Conflict. https://www.nonviolent-conflict.org/wp-content/uploads/2016/02/ Chile-2.pdf. Accessed 2 October 2019. Labarca, J. T. (2016). El ‘ciclo corto’ del movimiento estudiantil chileno: ¿conflicto sectorial o cuestionamiento sistémico? Revista Mexicana de Sociología, 78(4), 605–632. Lamb, G. (1998). Civil Supremacy of the Military in Namibia: An Evolutionary Perspective (Dissertation). Master of Social Sciences, University of Cape Town. https://open.uct.ac.za/bitstream/handle/11427/9552/thesis_hum_ 1998_lamb_g.pdf. Accessed 1 October 2019.

128

D. LAMBACH ET AL.

Lambert, P. (1997). The Regime of Alfredo Stroessner. In P. Lambert & A. Nickson (Eds.), The Transition to Democracy in Paraguay (pp. 3–23). Basingstoke: Macmillan. Landé, C. H. (2001). The Return of ‘People Power’ in the Philippines. Journal of Democracy, 12(2), 88–102. Larrouqué, D. (2019). Introduction: Le Paraguay peut-il encore être qualifié de ‘semi-autoritaire’? Cahiers des Amériques latines (90), 21–38. Latinobarómetro. (2016). Informe 2016. http://www.latinobarometro.org/ LATDocs/F00005843-Informe_LB_2016.pdf. Accessed 1 October 2019. Latinobarómetro. (2017). Informe 2017. http://www.latinobarometro.org/ LATDocs/F00006433-InfLatinobarometro2017.pdf. Accessed 1 October 2019. Latinobarómetro. (2018). Informe 2018. http://www.latinobarometro.org/ latdocs/INFORME_2018_LATINOBAROMETRO.pdf. Accessed 1 October 2019. Latinobarómetro. (2019). Online Database. http://www.latinobarometro.org/ latOnline.jsp. Accessed 9 August 2019. Lechner, N., & Guell, P. (1998). Construcción social de las memorias en la transición chilena. http://www.cholonautas.edu.pe/modulos/biblioteca2. php?IdDocumento=0271. Accessed 1 October 2019. Lima, A. R. (1992). Grundzüge des politischen Reformprozesses in Kap Verde. In P. Meyns (Ed.), Demokratie und Strukturreformen im portugiesischsprachigen Afrika (pp. 286–303). Freiburg: Arnold-Bergstrasser-Institut. Lipset, S. M. (1959). Some Social Requisites of Democracy: Economic Development and Political Legitimacy. American Political Science Review, 53(1), 69–105. López, M. (2019). Transición y democracia en Paraguay (1989–2017). ‘El cambio no es una cuestión electoral’. Buenos Aires: Sb editorial. Magnusson, B. A. (2001). Democratization and Domestic Insecurity: Navigating the Transition in Benin. Comparative Politics, 33(2), 211–230. Mazumder, S. (2018). The Persistent Effect of US Civil Rights Protests on Political Attitudes. American Journal of Political Science, 62(4), 922–935. Mbaku, J. M. (2003). Transition to Democratic Governance in Africa: Learning from Past Failures. In J. Ihonvbere & J. M. Mbaku (Eds.), Political Liberalization and Democratization in Africa: Lessons from Country Experiences (pp. 17–42). Westport: Praeger. McAdam, D. (1989). The Biographical Consequences of Activism. American Sociological Review, 54(5), 744–760. Melber, H. (2009). One Namibia, One Nation? The Caprivi as Contested Territory. Journal of Contemporary African Studies, 27 (4), 463–481.

4

MECHANISMS

129

Metsola, L. (2006). ‘Reintegration’ of Ex-Combatants and Former Fighters: A Lens into State Formation and Citizenship in Namibia. Third World Quarterly, 27 (6), 1119–1135. Meyns, P. (2002). Cape Verde: An African Exception. Journal of Democracy, 13(3), 153–165. Morency-Laflamme, J. (2016). Regime Crises in Africa: A Study of Armed Forces’ Behaviour (Doctoral thesis). Department of Political Science, University of Montreal. https://core.ac.uk/download/pdf/84095448.pdf. Accessed 1 October 2019. Morency-Laflamme, J. (2018). A Question of Trust: Military Defection During Regime Crises in Benin and Togo. Democratization, 25(3), 464–480. Murdie, A., & Purser, C. (2017). How Protest Affects Opinions of Peaceful Demonstration and Expression Rights. Journal of Human Rights, 16(3), 351– 369. Namibian Sun. (2019, May 3). March Against NDF Brutality. Namibian Sun. https://www.namibiansun.com/news/march-against-ndf-brutality2019-05-02. Accessed 1 October 2019. Navia, P. (2008). Top-Down and Bottom-Up Democracy in Latin America: The Case of Bachelet in Chile. Stockholm Review of Latin American Studies (3), 109–130. Navia, P., & Godoy, R. (2014). The Alianza’s Quest to Win Power Democratically. In K. Sehnbruch & P. M. Siavelis (Eds.), Democratic Chile: The Politics and Policies of a Historic Coalition, 1990–2010 (pp. 43–68). Boulder: Lynne Rienner. Navia, P., & Osorio, R. (2019). Attitudes Toward Democracy and Authoritarianism Before, During and After Military Rule: The Case of Chile, 1972–2013. Contemporary Politics, 25(2), 190–212. Nepstad, S. E. (2011). Civil Resistance in the Late 20th Century. Oxford: Oxford University Press. Nickson, A. (2010). Political Economy of Policymaking in Paraguay. In A. Berry (Ed.), Losing Ground in the Employment Challenge: The Case of Paraguay (pp. 285–314). New Brunswick: Transaction Publishers. Nwajiaku, K. (1994). The National Conference in Benin and Togo Revisited. Journal of Modern African Studies, 32(3), 429–447. OECD (2016). Trade Union Density. https://stats.oecd.org/Index.aspx? DataSetCode=UN_DEN. Accessed 14 November 2016. Opp, K. D. (1998). Does Antiregime Action Under Communist Rule Affect Political Protest After the Fall? Results of a Panel Study in East Germany. The Sociological Quarterly, 39(2), 189–213. Ostroot, N. M., & Snyder, W. W. (1985). Measuring Cultural Bias in a CrossNational Study. Social Indicators Research, 17 (3), 243–251.

130

D. LAMBACH ET AL.

Potthast, B., & Sosna, A. (2008). Das politische System Paraguays. In K. Stüwe & S. Rinke (Eds.), Die politischen Systeme in Nord- und Lateinamerika: Eine Einführung (pp. 468–488). Wiesbaden: Springer VS. Powers, N. (1992). The Transition to Democracy in Paraguay: Problems and Prospects (Kellogg Institute Working Paper No. 171). Notre Dame. https:// pdfs.semanticscholar.org/8aae/f34bdbdff7a667908a00a4a487f778d1d798. pdf. Accessed 1 October 2019. Preston, R. (1997). Integrating Fighters After War: Reflections on the Namibian Experience, 1989–1993. Journal of Southern African Studies, 23(3), 453–472. Przeworski, A. (1991). Democracy and the Market. Cambridge: Cambridge University Press. Robinson, P. T. (1994). The National Conference Phenomenon in Francophone Africa. Comparative Studies in Society and History, 36(3), 575–610. Rostow, W. W. (1971). Politics and the Stages of Growth. Cambridge: Cambridge University Press. Rotman, A. (2003). Benin’s Constitutional Court: An Institutional Model for Enforcing Human Rights (Bepress Legal Series Working Paper No. 104). https://law.bepress.com/expresso/eps/104/. Accessed 1 October 2019. Rovira Kaltwasser, C. (2007). Chile: transición pactada y débil autodeterminación colectiva de la sociedad. Revista Mexicana de Sociología, 69(2), 343–372. Saul, J. S., & Leys, C. (1995). Namibia’s Liberation Struggle: The Two-Edged Sword. London: James Currey. Saul, J. S., & Leys, C. (2003). Lubango and After: ‘Forgotten History’ as Politics in Contemporary Namibia. Journal of Southern African Studies, 29(2), 333– 353. Scarnecchia, T. (2006). The ‘Fascist Cycle’ in Zimbabwe, 2000–2005. Journal of Southern African Studies, 32(2), 221–237. Scharpf, F. W. (1999). Governing in Europe: Effective and Democratic? Oxford: Oxford University Press. Seawright, J., & Gerring, J. (2008). Case Selection Techniques in Case Study Research: A Menu of Qualitative and Quantitative Options. Political Research Quarterly, 61(2), 294–308. Seely, J. C. (2009). The Legacies of Transition Governments in Africa. The Cases of Benin and Togo. New York: Palgrave Macmillan. Sharp, G. (1973). The Politics of Nonviolent Action (3 Vols.). Boston: Porter Sargent Publishers. Sherkat, D. E., & Blocker, T. J. (1997). Explaining the Political and Personal Consequences of Protest. Social Forces, 75(3), 1049–1070. Siavelis, P. M. (2016). Crises of Representation in Chile? The Institutional Connection. Journal of Politics in Latin America, 8(3), 61–93. Singh, N. (2014). Seizing Power: The Strategic Logic of Military Coups. Baltimore: Johns Hopkins University Press.

4

MECHANISMS

131

Sondrol, P. C. (2007). Paraguay: A Semi-authoritarian Regime? Armed Forces & Society, 34(1), 46–66. Stroh, A. (2013). Stabilitätsanker Verfassung: Kleine Reformen und große Ängste in Benin (GIGA German Institute for Global and Area Studies Focus Africa No. 04). Hamburg. https://www.ssoar.info/ssoar/handle/document/ 35077. Accessed 1 October 2019. Stroh, A. (2016). The Consequences of Appointment Policies for Court Legitimacy in Benin: A Network Analysis Approach (German Institute for Global and Area Studies Working Paper No. 281). Hamburg. https://www.giga-hamburg.de/ sites/default/files/publications/wp281_stroh.pdf. Accessed 1 October 2019. Swapo. (1994). Better Opportunities for All Namibians (Swapo Election Manifesto). Windhoek. Tarrow, S. (1998). Power in Movement: Social Movements and Contentious Politics. Cambridge: Cambridge University Press. Uharte Pozas, L. M. (2012). El Proceceo de democratización Paraguayo: Avance y resistencia. América Latina Hoy, 60, 17–42. Van Vaerenbergh, Y., & Thomas, T. D. (2012). Response Styles in Survey Research: A Literature Review of Antecedents, Consequences, and Remedies. International Journal of Public Opinion Research, 25(2), 195–217. Weeks, G., & Borzutzky, S. (2012). Michelle Bachelet’s Government: The Paradoxes of a Chilean President. Journal of Politics in Latin America, 4(3), 97– 121. Whitehead, L. (2019). Democratization in Chile: A Long-Run and Comparative Perspective. In U. van Beek (Ed.), Democracy Under Threat. Challenges to Democracy in the 21st Century (pp. 173–197). Cham: Palgrave Macmillan. Zanou, C. (2016, January 19). La démocratie est une culture et l’alternance est son expression. Notre Afrik. http://www.notreafrik.com/archives-celestinezanou-la-democratie-est-une-culture-et-lalternance-est-son-expression/. Accessed 1 October 2019. Zanou, C. (2018, February 28). 28ème anniversaire de la conférence nationale au Bénin. La Nouvelle Tribune. https://lanouvelletribune.info/2018/ 02/28eme-anniversaire-conference-nationale-temoignage-celestine-zanou/. Accessed 1 October 2019. Zinecker, H. (2004). El Salvador nach dem Bürgerkrieg. Ambivalenzen eines schwierigen Friedens. Frankfurt am Main: Campus Verlag. Zubizarreta, Z. D. (2013). Disentangling the Fall of a ‘Dominant-Hegemonic Party Rule’: The Case of Paraguay and Its Transition to a Competitive Electoral Democracy (International Development Working Paper Series No. 13-144). London School of Economics. Zunes, S. (2017). Civil Resistance Against Coups: A Historical and Comparative Perspective. Washington, DC: International Center on Nonviolent Conflict.

CHAPTER 5

Inching Towards Theory

This chapter lays out what we believe our main contributions to current debates are, and how these contributions move the research programme forward. We compare, contrast and integrate our findings with other literatures. In short, we engage in theory-building without indulging in the illusion that we can actually achieve substantive and consistent theory. The core motivation of our project was to elucidate the long-term effects of nonviolent resistance (NVR) on democratic consolidation. We have made strides to clarify both the causal impact as well as the underlying mechanisms. Nonetheless, further work will be necessary here. We have attempted to bridge the more movement-centred NVR literature with social movement research and comparative politics scholarship of democracy and democratic transition, but these fields are still too often treated as separate from one another (della Porta 2016). The long-term effects of NVR remain an understudied subject, although we find that resistance has path-dependent effects that persist for decades or even longer. In this chapter, we start by bringing together our findings from Chapters 3 and 4 and then revisiting our initial assumptions. We then discuss the limitations of our results before comparing how our findings resonate with other literature from the fields of comparative politics, social movement research, resistance studies and the scholarship of democracy and democratic transition. The chapter concludes with suggestions on how to refine our approach further and move NVR scholarship forward and proposing questions for future research.

© The Author(s) 2020 D. Lambach et al., Nonviolent Resistance and Democratic Consolidation, https://doi.org/10.1007/978-3-030-39371-7_5

133

134

D. LAMBACH ET AL.

Summary of Findings Our statistical analysis supports the assumption that NVR has a positive impact on democratic consolidation, whether measured as democratic survival, peaceful turnovers of power, or democratic quality. First, NVR during transition makes democracies more resilient and much less likely to break down. Regimes with NVR-induced transitions have a much longer median survival rate (47 years) than regimes that came about by other means. Second, there is evidence that NVR-induced democracies have better odds of meeting the classic two-turnover test of democratic consolidation. While there seems to be little difference between NVR and violent resistance (VR) or no resistance (NoR) cases in terms of the likelihood of a first peaceful turnover, differences emerge when it comes to the second turnover. Whereas NVR cases are very likely to survive the second turnover, for other cases that democratized without the help of NVR it is essentially a flip of the coin. This again underscores the long-term stabilizing effects of NVR. Third, NVR-induced democracies improve more substantially and attain a higher level of democratic quality after transition compared to democracies brought about by violent revolution or elite-led liberalization. Notably, most of these relative gains are frontloaded into the first few years after transition, whereas non-NVR democracies either accrue gains more slowly, somewhat closing the quality gap over time, or break down entirely. In sum, our results show that NVR advances democratic consolidation across the three dimensions of democratic survival, peaceful turnovers of power and democratic quality. This supports our argument about the path-dependent effect of NVR: that is, that the circumstances of transition determine the trajectories of democratization and subsequent consolidation of democracy. The findings also provide a glimpse into when NVR-induced democracies fail. Specifically, we find some cases of breakdown during the first decade after transition and shortly thereafter but very few at later stages, e.g. only a single one beyond the 20-year mark. This is consistent with our finding regarding the turnover measure. When NVR-induced democracies fail, they tend to do so before the first electoral turnover. If they survive this milestone, their odds of survival are much better than those of non-NVR democracies. Evidently, the risk of democratic breakdown in NVR-induced democracies is concentrated in their early years, roughly the first decade (usually the first two legislative periods and terms of office

5

INCHING TOWARDS THEORY

135

for the head of the state). Our findings regarding the gains in democratic quality suggest a possible reason for this. A democratic transition can be a revolutionary and empowering event in which citizens claw back power from the elites, or, at the other extreme, it can be mostly cosmetic, as political and economic elites maintain their status and privileges. We believe that NVR pushes transition more towards the revolutionary end of this spectrum. This leads to faster and larger gains in key indicators of democratic quality but is also destabilizing, as new institutions and norms have to be developed and power redistributed among political actors. In other words, NVR may provide short-term improvements and long-term consolidation at the prize for some short-term instability. A comparison of our case studies suggests three likely mechanisms which distinguish NVR-induced transitions from those cases where democracies was installed either by violent revolution or through topdown liberalization: NVR levels the political playing field, promotes civil society activism and a democratic political culture, and avoids the pitfalls of praetorianism. First, NVR-induced democracies have a more even political playing field, both in horizontal and vertical terms. Horizontally, there are more legislative constraints on the executive in NVR cases than in non-NVR cases. Vertically, NVR democracies score substantially higher on measures of freedom of association and freedom of expression, and offer a much more conducive environment for civil society activity. However, it should be noted that while NVR pushes transitions slightly more to the revolutionary end of the spectrum, they rarely lead to wholesale transformations in society. NVR’s effect lies specifically in the creation of a more equitable political playing field, not in levelling social and economic inequalities (although della Porta 2016 is more optimistic about the impact of social movement protests on social rights). Second, NVR promotes a democratic political culture and civil society activism. Although a systematic check of these claims is not possible due to lack of comprehensive cross-national data, they are supported by wellestablished theories and by evidence from the cases. Ordinary citizens are more politically engaged in Benin and Chile than they are in Namibia, El Salvador, Cape Verde or Paraguay. Beninese citizens in particular express a high level of support for pro-democratic ideas. Both Benin and Chile have seen multiple episodes of mass mobilization and civil society activism

136

D. LAMBACH ET AL.

after transition. Beninese protesters have consciously referred to the history of NVR in their pro-democratic ‘touche pas à ma constitution’ demonstrations. Third, the NVR cases mostly avoided the praetorian problem of military veto players. Mass-led transition also did not create any new undemocratic veto players, although it should be noted that the revolution did not fundamentally change traditions of clientelist politics in Benin. In Chile, NVR was successful in forcing democratic elections, but it was only partly successful in stripping existing elites of their power. More broadly, we find that democratization via NVR on average does not lower the likelihood of coup attempts after transition relative to cases of democratization without NVR. However, NVR has a stabilizing effect by lowering the risk that these coup attempts are successful. While these findings indicate that there are distinct causal mechanisms underlying the long-term stabilizing effects of NVR, it is also worthwhile to explore the variation between our two NVR cases, Benin and Chile. While both can be considered consolidating democracies and both score similarly on common indicators of democracy, NVR played out differently in each case, leading to somewhat different processes and outcomes of transition. A key difference between Benin and Chile lies in the early demobilization of the Chilean resistance movement. During transition, the mass movement was progressively replaced by a more partydominated coalition. In the end, this gave the democratization process an elite-centred nature whereby democratic progress emerged as the result of drawn-out negotiations between right-wing and left-wing political, economic and military elites. In contrast, the Beninese movement had a much greater impact on how transition was negotiated, continuing to exert pressure even after President Kérékou had agreed to a National Conference. It was this pressure that was instrumental in empowering the National Conference to overhaul the constitution and the political system. These diverging trajectories are also visible in other ways. First, the peaceful movement in Benin created a powerful symbol of resistance and empowerment which continues to inspire citizens to hold their politicians accountable. While there are similar narratives in Chile, these are arguably less influential for civic activism. In addition, the Beninese movement was able to co-opt the armed forces while in Chile citizens lacked the leverage to install civilian control over the military. As a result, the prerogatives of the armed forces had to be chipped away over a number of years.

5

INCHING TOWARDS THEORY

137

In terms of our relational concept of democracy, NVR exerts most of its beneficial effects on the vertical accountability of ruling elites to their citizens. Citizens enjoy more political rights and empowerment and civil society is less constrained in NVR-induced democracies. The effects on horizontal relations between government, the opposition and the military are less pronounced, with legislative constraints on the executive being the most notable benefit. Beyond that, there is little that differentiates NVRinduced democracies from others in terms of horizontal accountability. NVR does not reduce the probability of military coup attempts after transition, indicating that militaries are not per se less likely to agree to civilian oversight. While NVR does reduce the likelihood that coups are successful, we interpret this to mean that such coups are met with disapproval and resistance by citizens, which again underscores the importance of the vertical dimension, or that fissures in the armed forces lead would-be coup plotters to overestimate support for their cause. This interpretation seems to be consistent with results obtained by Svensson and Lindgren (2010), who argue that NVR against autocratic governments is more likely to succeed when challenging the vertical rather than the horizontal legitimacy of the state.

Revisiting Our Assumptions Our results generally indicate support for our starting assumptions. However, they also provide grounds to revisit and update some of them, especially in light of the material obtained from the case studies which offers a rich if somewhat complex empirical picture. This is arguably more important for the causal mechanism side of our explanation, as the statistical analysis paints a relatively clear picture that is consistent with other recent work (Pinckney 2018; Kadivar 2018; Kadivar et al. 2019; Butcher et al. 2018). Moreover, there has been much less work on the causal mechanisms (but see Kim and Kroeger 2019), so we had less settled work to draw on when we started our research. Indeed, many of the mechanisms that are prominent in the theoretical literature found little or mixed support in the case evidence.1 In other words, there is a greater degree of 1 We checked several other possible mechanisms drawn from the more theoretical literature against evidence from our case studies and found no or only inconsistent support. However, this should be understood with the important caveat that case selection likely had a strong impact on these results. (Statistical analysis was not possible due to lack of

138

D. LAMBACH ET AL.

uncertainty about our results regarding the causal mechanisms, hence a greater need to critically revisit our assumptions. Regarding the first mechanism, NVR levelling the political playing field, we initially assumed that this would be most visible in two ways. Our first expectation was that NVR cases had a more pluralistic party system. The rationale was that NVR campaigns represent coalitions of diverse antiregime interest groups, all of which would be empowered to form political parties after the transition to democracy. In contrast, we expected VR cases to be dominated by parties emerging from the armed movements and NoR cases to see a strong involvement of former single parties. However, the evidence on these is mixed. For our six case studies, these assumptions mostly held up. While Benin as our African NVR case currently has approximately 120 parties, Namibia (VR) is dominated by SWAPO (the former liberation movement) and Cape Verde by the former single party (PAICV) and its challenger (MPD). Similarly, El Salvador (VR) is dominated by two former conflict parties, ARENA and the FMLN, while Chile (NVR) has a multitude of political parties organized in two coalitions. Paraguay, despite the transition in 1990, was ruled by the former single party, the Colorado Party, until 2008 when the Patriotic Alliance for Change won the elections. However, when we look at a larger sample of post-transition elections, we do not find a significant difference in the effective number of parties across modes of transition.2 In addition, a higher number of parties does not necessarily translate into more democratic politics. If anything, NVR may have a levelling effect

suitable data.) These purported mechanisms include, first, that NVR inculcates a ‘culture of compromise’ among movements having to balance between different interests within the coalition. This culture was then supposed to carry over into democratic politics as former activists enter into politics (Chenoweth and Stephan 2011; Kinsman and Bassuener 2013). We found no evidence that such aspects of ‘political culture’ were in any way shaped by NVR. Second, it is often claimed that ‘veterans’ of the movement that are committed to democratic ideals go into politics and safeguard democracy from inside its institutions (Kadivar 2018). While there are certainly instances of that, they were mostly anecdotal for the cases we analysed. Third, Sharp (1973b) argued that people trained in techniques of NVR are more capable of preventing democratic backlash. It is true that our NVR cases showed instances of remobilization but these were typically not undertaken by former participants of NVR movements themselves but by a new generation of activists, casting doubt on the ‘training’ aspect of this explanation. 2 We used data on the effective number of parties in the first election after transition provided by V-DEM. On average, the effective number of parties is 3.1 with top-down transitions and 3.6 for both violent and nonviolent transitions.

5

INCHING TOWARDS THEORY

139

on the party system by not leaving behind a single political actor with an authentic narrative of having liberated the country and thus the opportunity to monopolize the historical achievement. The downside, however, is that NVR does not establish a stable party system but may even create a more volatile one. In Benin, parties are fractionalized and weak to the point of dysfunction. The party system reflects, however, the heterogeneity of local identities and interests in society. Most parties were created around old economic elites and not from the resistance. The same applies to Chile, where the transition opened up space for the older parties— some of whom had supported the NVR campaign—to re-emerge. Our second expectation was that NVR would lead to a greater level of ‘churn’ among elites, as entrenched politicians and economic elites were unseated and replaced with new elites from outside the traditional circles. We find some anecdotal evidence of this but were not able to analyse this systematically (see below). This seems to be worthy of further research, especially since it fits well with recent work by Michael Albertus. In his book with Victor Menaldo, he argues that democracies are often undercut by authoritarian elites continuing to enjoy their perks, privileges, and veto positions well into the new regime: ‘When the time comes and they bargain their way out of dictatorship to democracy, they then leverage these advantages under elected rule to shape the rules of the game and public policy’ (Albertus and Menaldo 2018, p. 8). Although their model of politics is more elite-centred than ours, Albertus and Menaldo acknowledge that popular protest and revolution are a key mechanism in disrupting these elite networks (Albertus and Menaldo 2018, pp. 49–50). Our second mechanism of NVR advancing a democratic civic culture needs more elaboration. The finding itself is consistent with other works, for example García-Ponce and Wantchekon’s analysis of the longterm effects of anti-colonial resistance. They find that countries that saw nonviolent urban protest rather than rural armed insurgency under colonialism tend to have more democratic institutions today, and that this is due to peaceful protest leading to shared ‘norms of peaceful political expression and compromise, which provided cultural bases for liberal democracy’ (García-Ponce and Wantchekon 2017, p. 2). Their findings are echoed by Kodila-Tedika and Asongu (2018), who demonstrate that African independence movements’ strategic choices about nonviolence continue to affect present-day quality of governance. But the underlying

140

D. LAMBACH ET AL.

mechanisms are still somewhat fuzzy, especially since this requires tracing a difficult-to-measure causal effect over many decades. Our interpretation of this mechanism focuses on two connected factors. The first is that NVR empowers civil society to perform its democratic functions to ‘1) provide a space in which citizens can meaningfully practice democracy on a day-to-day basis, 2) anchor the legitimacy of political practices and state institutions in vigorous public debate and 3) serve as a counter balance to the aggregative logic of political society’ (Heller 2009, p. 359). This empowerment comes not necessarily from civil society organizations emerging from the movement (although this is also a possibility) but from the resistance providing a powerful symbol of ‘people’s power’, a narrative that civil society can draw on to mobilize citizens against anti-democratic backsliding.3 However, civil society matters not just in organizational and institutional terms but also influences informal, social, and behavioural legacies, which leads us to our second factor. In a comparison of democratic transition in Spain and Portugal, Fishman makes the point that civil society has a ‘capacity to shape prevailing cultural assumptions about who deserves to be heard in democracy’s central “conversation”’ (Fishman 2017, p. 392) and that the cultural processes set in motion during transition create social legacies that persistently shape democratic practice. That political culture is shaped by the past is well-established. Voting maps of Poland, the Ukraine, Germany and other countries show how countries whose borders have shifted over the past centuries are still broken into distinct political regions. Putnam’s (1993) classic study of democracy in Italy displays a similarly persistent break in the civic traditions between the north and the south of the country. These differences in political culture are sustained not just through institutions (e.g. discriminatory laws) or structures (e.g. unequal distribution of capital) but also through ‘behavioral path dependence’, i.e. informal institutions like ‘family socialization and community norms’ (Acharya et al. 2018, p. 5). More specifically, resistance influences political culture in many ways. Protests are formative socializing experiences and may impact the mobilization of affected groups for a long time. Enos et al. (2019) find that exposure to the 1992 Los Angeles Riots led to a political mobilization of both African–American and white voters in favour of the Democrats 3 Recent work by Thurber (2019) also suggests a possibility of reverse causality. He finds that the choice of strategy by resistance movements is influenced by their social ties to various communities in the state.

5

INCHING TOWARDS THEORY

141

that was still visible in 2005, more than a decade later. Fishman argues that even low-level participation in revolutionary protest remade participants’ ‘taken-for-granted sense of how politics works, or at least how it can work’ (Fishman 2017, p. 403). Social movements create and anchor collective identities by engaging in memory politics such as telling stories, performing songs or erecting monuments. NVR leaves behind a symbol of resistance that can inspire and sustain future movements, as della Porta argues: ‘The degree to which a positive memory of the role of contention in the democratic transition is shared affects the capacity of social movements to refer to it, especially in critical moments, in order to claim legitimacy’ (della Porta 2018, p. 13). These narratives can then be transmitted through formal institutions, inasmuch as resistance movements have been able to reshape these on their own terms (Fishman 2017). In contrast, violent movements, including various liberation movements, tend to glorify the violence that paved their way to power, thereby establishing a basis for a persistent culture of violence. Our third mechanism posits that NVR-induced democracy avoids the praetorian problem that continues to haunt VR and NoR cases. Much like Sherlock Holmes’s dog that didn’t bark in the night, we construct our argument on the basis of something not occurring. Obviously, this is difficult to establish empirically, relying entirely on comparison and counterfactual reasoning. That notwithstanding, the mechanism finds support in the breadth of evidence about the risks associated with praetorianism. Haggard and Kaufman (2016, pp. 244–245) find a significant linkage of praetorianism, measured as a history of coups, with democratic reversion. It is, however, more difficult to establish whether NVR-induced democracies indeed have healthier civil–military relations. On the one hand, our analysis of the likelihood of military coups (see Chapter 4) seems to contradict this claim. On the other, the fact that most of these coups fail in NVR-induced democracies suggests that the military is less able to unseat democratic governments in these cases. And while in Benin a historically praetorian military has never threatened democratic stability, it took a long while to whittle away the armed forces’ influence in Chilean politics. However, while we are reasonably confident in the validity of this mechanism, more research on civil–military relations after NVR-induced transition is necessary.

142

D. LAMBACH ET AL.

Limitations of Our Results We also need to acknowledge certain limitations of our results. Given that our exploration of the causal mechanisms rests on detailed case studies of a limited number of NVR-induced democracies, we have to be careful about overgeneralization. Cases were chosen with care so as to maximize analytical leverage, but it would be unreasonable to expect that case selection did not affect our results. For instance, had we chosen an example of NVR-induced democracy like Poland, where the leader of the NVR movement became president after democratic transition, we would probably highlight the importance of movement leaders staying involved in democratic politics. In addition, the temporal context probably had an impact. The NAVCO 2 dataset covers cases from the mid-twentieth century onwards but our case studies are mostly post-1989, placing them squarely in the third wave of democratization. This wave was characterized by more prodemocratic international environment following the end of the Cold War. In addition, donors like World Bank bound economic and financial assistance to conditionalities like democratic and economic liberalization. And while the impact on democracy of Structural Adjustment Programs in particular was ambiguous, we do think that on balance the conditions for democracy were more favourable in this period than in earlier ones. This is consistent with the finding of Kendall-Taylor and Frantz (2014) who show that NVR has become a more commonly successful means of forcing autocrats out of office since the 1990s. In short, had we chosen cases from other periods, it is likely we would have got different results. Hence, the mechanisms we put forward should be understood as propositions derived from a limited empirical base. While evidence suggests that these mechanisms apply more broadly (see below), this still has to be substantiated through further research. We also were unable to include variation among NVR campaigns in our research design, even though this is often flagged as relevant in other research. In the quantitative analysis, we determined our population of cases of NVR-induced transition solely by the criterion of whether the nonviolent campaign brought about democratization. As such, we assumed that differences among NVR campaigns do not systematically influence the outcome of democratic consolidation. Our case studies revealed some problems with this approach.

5

INCHING TOWARDS THEORY

143

As the literature shows, NVR campaigns come along in many different forms. First, campaigns vary in duration, size, and scope. For instance, millions of citizens participated in civil resistance campaign in South Africa and Poland that brought about the end of autocratic regimes. Other resistance campaigns were considerably smaller. Related to this, Chenoweth and Stephan (2011) advanced the so-called ‘3.5% rule’, referring to the empirical observation that once 3.5% of a country’s population participates in a resistance campaign against a government, the campaign is guaranteed to succeed. Recent research by Chenoweth and Belgioioso (2019) qualifies this rule by highlighting that successful resistance movements have ‘the momentum of dissent’, which is defined as combination of the number of participants and the frequency of protest events. However, notwithstanding the relevance of these aspects for the success of campaigns, we find little evidence that these aspects of campaigns affect subsequent political developments after transition. Second, regardless of their size, scope or momentum, resistance campaigns may differ in their impact on the transition process because they vary in access to and influence in political negotiations, constitutional revision, or other aspects of institutional engineering. This aspect was very relevant in our case studies. In Benin, civil society was crucially involved in almost all aspects of the transition and the movement was able to constrain and determine elite politics on various occasions, such as the process of choosing participants for the national conferences. Moreover, these interferences by civil society had a lasting impact and influenced post-transition politics in a path-dependent way. In Chile, by contrast, the NVR movement mostly demobilized early in the transition process and left politics to established pro-democratic elites, who in turn negotiated the new rules of the game with elites close to the former autocratic regime, and chose a non-confrontational, incremental approach to democratization. Nevertheless, in our quantitative analysis we treat these campaigns as equal and thus ignore the difference in involvement of the campaign in transition politics. We chose this approach because our main comparison cases are elite-led transitions that occurred without any campaign influence. Since these cases are characterized by having no political mobilization, we logically cannot measure the size and scope of the campaign there. Some researchers deal with this problem by assigning a score of zero to the relevant variable, which we consider inappropriate for our

144

D. LAMBACH ET AL.

research design (see Chapter 3). Thus, for our research interest, considering treatment intensity was less important than applying proper counterfactual comparison. However, we accept the need to pursue further research on this topic with a view to bringing together our findings with research on campaign intensity and campaign involvement in the transition process (see below). Finally, we still know little about why some cases of NVR fail to consolidate. Take the example of Mali, where popular protests in 1991–1992 managed to unseat President Moussa Traoré after the defection of key parts of the Malian military. For two decades, Mali was seen as a relative success story of democratization in the Sahel region before the government of President Amadou Toumani Touré was overthrown by mutinous soldiers who criticized the government’s handling of the Touareg rebellion in the country’s north-east. In spite of the NVR-induced transition, democracy in Mali was a mixed bag. Support for democracy was shallow among citizens, voter turnout typically low, and civil society remained mostly passive after 1992. On the other hand, the 1992 constitution had been instituted in a transparent and participatory way and commanded a lot of respect. The army had a praetorian streak but saw itself as acting in the interests of democracy, both in supporting the 1991 protests and in deposing Touré in 2012, and enjoyed a certain degree of public support. So while democracy in Mali was vulnerable in some respects, it is hard to construct a comprehensive and theoretically compelling argument as to why it broke down while Benin’s democracy has endured. The failure of NVR-induced democracy is an issue that has received little attention in previous research or in our own, so we can only sketch out a possible answer to this question. We follow Pinckney (2018) in suggesting that a too-rapid demobilization of the resistance movement after the autocratic incumbents are dislodged presents a risk for completing the transition to democracy. This is often due to the ‘negative coalition’ of groups united by common grievances against the regime falling apart once its singular joint goal has been achieved. Similar to the argument by Albertus and Menaldo (2018), Pinckney argues that a too-early exit of the movement from the political stage allows political elites to dominate the transition process, as happened in Chile. Maintaining public mobilization throughout the transition ‘is thus crucial to maintaining a degree of public scrutiny and accountability over those who may not have underlying preferences for the goals of the revolution to continue to support it’

5

INCHING TOWARDS THEORY

145

(Pinckney 2018, p. 45). Demobilization also means that the skills and traditions of NVR will fall from use and leave fewer traces in social groups and political culture. We return to this theme below when we outline directions for further theoretical development and research.

Comparing Our Results A comparison with other successful and unsuccessful cases of NVR suggests that our results are applicable beyond the cases we studied. In addition, our findings align well with those of other literatures from comparative politics, social movements research and contentious politics research as well as recent works on NVR. Finally, we see the emergence of many works championing relational approaches to democratic politics. At best, we can interpret these developments as support for our findings, increasing our confidence in their reliability. At the very least, our results suggest avenues for further research, opening the way towards a more nuanced discussion of the causal mechanisms underpinning the effectiveness of NVR. There are many examples of nonviolent movements not only being instrumental in the transition to democracy but also leaving legacies that continued to stabilize those same democratic regimes. In Nepal, the 1990 Jan Andolan (People’s Movement) forced the monarchy to agree to democratic reforms. In the early to mid-2000s, in response to King Gyanendra’s electoral interference, suspension of the constitution and attempt to reinstate direct royal authority, he was resisted by Nepalese political parties and civil society, again referring to themselves as Jan Andolan, which ultimately led to the complete downfall of the monarchy (Pinckney 2018, p. 50). Fishman (2017) recounts how the 1974 Carnation Revolution led to Portugal having substantially more inclusionary and equitable democratic practices than its neighbour Spain, where transition to democracy was more elite-dominated. And Diamond (2008, pp. 153–168) argues that democracy in India survived President Indira Gandhi’s attempts to weaken the judiciary in the mid-1970s thanks to a democratic political culture and vibrant civil society which were in no small part inspired by the history of peaceful resistance during the struggle for independence. The case of the Philippine transition underscores the long-term legacy of grassroots social mobilization well beyond NVR-induced democratic

146

D. LAMBACH ET AL.

openings, but also the ambivalent role of the military in influencing political processes. During the authoritarian rule of Ferdinand Marcos (1972– 1986), a nonviolent pro-democratic opposition progressively organized and coalesced around the Catholic Church and grassroots organizations of women, peasants, workers, students, and the urban poor (Schock 2005, pp. 68–79). Following a mass campaign of nonviolent action in 1985, two million citizens rallied to protest a fraudulent election in February 1986 and organized a general strike. The combined forces of civil society mobilization and significant military defections toppled the Marcos regime and instituted a democratic government followed by a long process of democratic consolidation. The Philippines passed the threshold of two peaceful presidential turnovers after the elections of 1992 and 1998, and the country did not encounter any major democratic backslide, despite being identified as only ‘partly free’ by Freedom House and as a ‘flawed democracy’ by the Economic Intelligence Unit in 2019, largely due to persistent armed conflict with communist, separatist and Islamist insurgencies. The three mechanisms discussed in Chapter 4 can all be observed in the Philippines, albeit in differing degrees. In terms of levelling the political playing field, the transition process saw the replacement of the dictatorship by a broadly unified anti-Marcos elite, although the leverage of left-leaning progressive elements of the opposition was undercut by their boycott of the (pre-revolution) 1986 elections and the formation of a US- and military-backed reformist government (Schock 2005, pp. 68– 81). Nevertheless, the new political opening led to a redistribution of power both horizontally and vertically. The post-revolution constitution ratified in 1987 re-instituted a strict separation of powers and a bicameral legislative. Moreover the Local Government Code passed in 1991 decentralized power to local governments and institutionalized the participation of civil society in local governance (Rivera 2002). With regard to the promotion of a democratic political culture, analysts have also noted the sustained role of civil society activism in articulating societal interests and monitoring government performance in the postMarcos era (Igarashi 2008): ‘An outstanding feature of the consolidation process […] has been the unusual ability of militant social movements to be part of the broader process of democratic incorporation through the electoral party-list system while maintaining their contentious politics of claim-making on behalf of the marginalized sectors’ (Rivera 2002, p. 466). Indeed, electoral reforms in the 1990s encouraged social movements to enter the political arena, leading to an expansion of democratic

5

INCHING TOWARDS THEORY

147

spaces. At the same time, the mass-based nonviolent mobilization which led to the ousting of elected incumbent President Joseph Estrada in 2001 demonstrated the resilience of a vibrant and contentious civil society long after the restoration of democracy. Finally, the ambivalent role of the military since 1986 indicates that the NVR-driven transition did not fully remove the pitfalls of praetorianism. Several coup attempts against democratically elected executives occurred during the Corazon Aquino administration (1986–1992) and following the 2001 ‘people power II’ revolution. The continued politicized mindset of the Filipino army poses a problem to democratic consolidation—all the more as the 1987 constitution upholds the military as the ‘protector’ of the people and the state. But the fact that none of the coup attempts were successful also confirms the finding presented in Chapter 4 that the legacy of NVR lowers the risk of successful military takeovers by building citizens’ capabilities to oppose and resist anti-democratic moves. This brief scan of other cases suggests that our results have some wider relevance beyond our two case studies. They are in agreement with findings from comparative politics, social movement research and more recent works on NVR. We briefly survey these fields below to highlight areas of agreement and disagreement so as to identify directions for productive dialogue and future research. We then compare our relational approach with similar approaches that have recently emerged in those fields. In comparative politics, few writings on democratic consolidation account for mass mobilization. Pro-democracy movements are often seen as epiphenomena of fracturing elite networks where the really important politics happen. Early influential works of transitology (e.g. O’Donnell et al. 1986; Linz and Stepan 1996) focus on these elite interactions, often arguing for pacted transitions to ensure democratic survival (see Carothers 2002, for a critique of the transition paradigm). And while the concepts of democratic transition and consolidation have somewhat fallen out of favour in comparative politics (Diamond et al. 2014, pp. 93–94), recent work is underpinned by the same kind of formal institutionalist/rationalist approaches in which democracy is modelled as a game between voters and competing political elites (Hollyer et al. 2018; Little et al. 2015; Fearon 2011). The work of Milan Svolik is a case in point. He offers a detailed empirical analysis of the odds of democratic consolidation but omits mass mobilization and the mode of transition as potential explanatory variables

148

D. LAMBACH ET AL.

(Svolik 2015). The related literatures on democratic breakdown, backsliding and regression also do not mention civil resistance, modes of transition or the resulting path dependencies (see the reviews in Waldner and Lust 2018; Tomini and Wagemann 2018). However, recent work by Stephan Haggard and Robert Kaufman has shown the way towards a more relational approach that takes the role of citizens in democratic transition seriously. Their acclaimed 2016 monograph Dictators and Democrats: Masses, Elites and Regime Change has received a lot of attention in the field. It starts from the premise that prior research on democratic transition and survival has been overly fond of structural explanations, focusing on social inequality, economic development and the like (e.g. Boix 2003; Acemoglu and Robinson 2006). They argue for a ‘more political’ approach ‘rooted in factors such as the nature of authoritarian and democratic institutions, regime performance, and capacities for collective action on the part of civil society’ (Haggard and Kaufman 2016, p. 3). Haggard and Kaufman distinguish two types of transition: distributive conflict transitions and elite-led transition (Haggard and Kaufman 2016, pp. 37–38). Distributive conflict transitions have two necessary characteristics: the mobilization of redistributive grievance by disadvantaged groups posing a threat to incumbent elites, and incumbents being ousted through mass mobilization. Elite-led transitions are all cases where such socio-economically motivated threats from below do not play an important role in overturning autocracy. Haggard and Kaufman find that theories of distributive conflict transitions, which argue that transitions are driven by grievances about socioeconomic injustice, do not hold up empirically. In their analysis of transition to democracy and democratic reversion (i.e. breakdown of democracy), they find that structural factors like the level of inequality have only a marginal impact. But that does not mean that mass mobilization does not matter—quite the contrary: ‘the mobilized de facto power’ (Haggard and Kaufman 2016, p. 14) of the masses accounts for half of the transitions in their sample. So the hypothesized mechanism (mass mobilization) works but the presumed motivation (grievances about economic inequality) does not. Their analysis of distributive conflict transitions offers additional detail that is relevant to our work. Haggard and Kaufman identify two ways in which mass mobilization ousts incumbents. In 30% of cases, incumbents were forced into resignation or exile by mass protest, leaving caretaker governments or national conferences in charge. The other 70% were cases

5

INCHING TOWARDS THEORY

149

where incumbents ‘retained at least some leverage in negotiations over the terms of their withdrawal’ (Haggard and Kaufman 2016, p. 105). But while Haggard and Kaufman (2016, pp. 105–106) dismiss the differences between these modes of transition, we have a hunch that these can be consequential. The cases of Benin and Chile fit these two profiles very well and although both can be considered reasonably consolidated, the transition and consolidation of Chile was noticeably more difficult and drawn-out than it could have been with a cleaner break at the end of the Pinochet era. Either way, this is an important reminder that transitions driven by mass mobilization can still allow for great influence by incumbent elites, an aspect we explore in more depth below. Our results are generally in line with those of Haggard and Kaufman. However, we differ in a few conceptual and theoretical aspects. First, we consider their approach to concept-building to be unsatisfactory. Haggard and Kaufman define distributive conflict transitions through two necessary conditions, mass mobilization and socio-economic grievances, which jointly lead to the removal of autocratic incumbents. All cases without both of these conditions are categorized as elite-led. But this results in cases where mass mobilization occurred but was not motivated by social injustice being erroneously classed among elite-led transitions. It is also questionable whether distributive conflict and elite-led transitions can be legitimately conceived as exclusive ideal-types. One example where this taxonomy runs into trouble is Chile, which Haggard and Kaufman code as elite-led, arguing that elite negotiations were the primary cause of democratic transition (Haggard and Kaufman 2016, pp. 205–207). Arguably, these negotiations would never have occurred without prior mass mobilization, even though popular protest was receding in the final years of the Pinochet regime. If anything, Chile displays aspects of both ideal-types of Haggard and Kaufman’s taxonomy just as it proved to be a hard case for our theoretical assumptions about the effects of NVR. This underscores that we should start disentangling the drivers or causes of transition, i.e. those factors that lead to transition or make it possible, from the modes of transition, i.e. how it plays out specifically and who calls the shots. Second, Haggard and Kaufman use a rationalist approach. Their explanation of why mobilization works strongly focuses on its effects on the balance of power in society. This is intelligible to a relational perspective, indicating a power shift away from elites towards civil society, but a rationalist approach underestimates ‘soft’ factors like political culture, collective memory and other crucial mechanisms we identified in our case

150

D. LAMBACH ET AL.

studies. Despite these limitations, Haggard and Kaufman’s work has been very helpful in moving comparative politics towards a greater appreciation of the long-term force of mass mobilization for democracy. We hope that further work in comparative politics expands upon this approach. The social movement and contentious politics literatures also offer helpful comparisons for our findings. We find ourselves in broad agreement with classic works, e.g. by Tarrow (1998), who shows how experiences of protest predispose citizens to mobilize again in future. Also, Grodsky (2012) echoes our finding that a too-rapid demobilization of mass movements hands the reins of the transition process to political elites and can have deleterious effects on the development of democracy. This point is explored in greater depth across several works by Donatella della Porta (2014, 2016, 2018) which resonate with our findings. She picks up on Grodsky’s point that revolutionary movements often disappear rapidly after transition to democracy. But, she argues, this does not mean ‘that the revolution is gone, but rather that it has complex paths of resilience’ (della Porta 2016, p. 346). In other words, resistance movements leave manifold legacies even after the movements themselves are long gone. Like us, della Porta sees protest as a potential critical juncture for society. Taking a path dependence approach, she disaggregates the impact of protest into ‘a sequence of processes of cracking, as the production of sudden ruptures; vibrating, as contingently reproducing those ruptures; and sedimenting, as the stabilization of the legacy of the rupture’ (della Porta 2018, p. 4). Della Porta expects that social movement participation in democratic transition will have long-running consequences for politics ‘as the call for a break with the past and increased rights for the citizens will be louder than in regime transitions that happen mainly through elite pacts’ (della Porta 2016, p. 2). This means that the effects of these movements continue to ripple throughout society long after the movements themselves have disappeared, in areas such as ‘the recognition of citizens’ rights to protest, the presence of channels of institutional access, and sensitivity to social justice’ (della Porta 2016, p. 2). The main effects are not to be found at the policy level but at the level of civic culture and civil rights. Della Porta even sees a positive impact of social movement on social rights and equity down the line. The main mechanism is that a political system that comes about with the involvement of social movements becomes more inclusive, both in terms of institutional design and political culture, than one that emerges out of a pacted transition (della Porta 2016, pp. 6–7). Finally, della Porta also agrees with our finding

5

INCHING TOWARDS THEORY

151

that mass movement participation in transition levels the political playing field, arguing that such ‘eventful democratizations […] leave more space for in-depth, unconstrained transformation’ (della Porta 2016, p. 19). Recent works on NVR highlight interesting extensions to our work. While our statistical findings are in line with those of similar research projects (Pinckney 2018; Kadivar 2018; Kadivar et al. 2019; Butcher et al. 2018), two issues have recently received particular attention that are of special interest to us: the role of organizations in resistance movements and the effects of movement duration. Both issues have seen recent advances that go beyond what we included in our research design and highlight ways to theorize the impact of NVR further. The role of organizations has been the focus of recent work by Charles Butcher (Butcher et al. 2018; Butcher and Svensson 2016) in particular. He finds that organizations add leverage to resistance movements and make government concessions more likely. The participation of national trade unions has been shown to have a positive effect on the odds of movement success and of post-conflict democratization (Butcher et al. 2018). This is in line with Haggard and Kaufman, who argue that organizations play vital leadership and coordinating roles even in otherwise spontaneous protests: organizations increase turnout, facilitate coordination across groups, and link protest issues to collective identities (Haggard and Kaufman 2016, pp. 66–69). From these results we might surmise that resistance campaigns behave and work differently depending on the role of the organizations—unions, nongovernmental organizations, traditional associations, collectives, women’s groups, etc.—involved in them. We might even try to distinguish between ‘mass-driven’ and ‘organization-driven’ movements. This would resonate somewhat with empirical evidence from Benin and Chile. In Benin, the movement was mass-driven, in spite of the role of the communist party in the initial mobilization and, later, the influence of trade unionists. In Chile, the movement was driven mostly by established organizations like political parties and unions, especially in its latter stages. This had repercussions for post-transition democracy in these countries, e.g. the structural weakness and fragmentation of Benin’s party system or the fragmentation of Chilean civil society. We may therefore infer that mass-based movements are more volatile because they lack the necessary political, administrative and ideological capital to stay engaged in post-revolutionary politics (Muriaas et al. 2016; but see Vinthagen 2015, for a more optimistic take).

152

D. LAMBACH ET AL.

Organizations, especially established political parties, know how ‘normal politics’ works. However, it is unwise to construct another dichotomy here. First of all, it probably makes more sense to think of ‘mass-driven’ and ‘organizationdriven’ movements not as opposites but as ideal-types on a spectrum, with most of the empirical cases clustering somewhere in the middle. Second, a dichotomy would just invite further conceptual problems— does it make a difference whether a movement is dominated by a single organization or a collection of organizations? Third, there are also examples of movements, for example the indignados, or the Occupy movement, that explicitly reject hierarchical forms of organizing, enacting a constructive/obstructive programme (Gandhi 1948) and upending traditional models of collective action. At the very least, we believe that the internal dynamics of movements, the role of organizations and the implications for democratic transition are worthy of further scrutiny. Recent work has also highlighted the positive impact of movement duration on the effects of NVR. This represents a break from the landmark study by Chenoweth and Stephan (2011, p. 213) who found no effect of the duration of resistance on the level of democracy in various models. It is also not something that we considered during our research, and our case studies are not suitable to test this proposition, but it represents a useful extension of the research programme, which is why we wish to highlight it here. If we bracket out the circular element of the argument—longer-running movements are more effective, more effective movements can sustain themselves longer—the core findings of the movement duration literature are in line with our own results. The work by Mohamed Kadivar is especially instructive here (Kadivar 2018; Kadivar et al. 2019). Much like us, Kadivar tests the impact of transitions induced by mass mobilization on both democratic survival and democratic quality (what he calls ‘substantive democracy’). However, both his studies focus on the duration of resistance campaigns. Interestingly, he finds that ‘(t)he length of unarmed mobilization is associated with a decrease in the chance of democratic breakdown, but the mere occurrence of a mobilization does not predict democratic survival. To observe the effects of mobilization, we must take into account its duration’ (Kadivar 2018, pp. 403–404). This challenges our results, which clearly identify the mode of transition as a substantial predictor of democratic survival and breakdown. However, there are conceptual and methodological reasons to question the approach taken by Kadivar

5

INCHING TOWARDS THEORY

153

and colleagues. Specifically, we did not consider the duration of resistance campaigns that led to democratic transition because it is unobservable for elite-led transitions, which are our main counterfactual. Kadivar and colleagues treat elite-led transitions as if the campaign duration is zero which, however, contradicts the linear additive logic of coefficients estimated by their statistical model. The one-unit change from zero to one year of campaign duration is of a different quality than the change from, say, seven to eight years of campaign duration. The first change refers to a change of transition processes while the second merely refers to a change of duration. Therefore, we suggest analysing the effect of campaign duration only for cases that actually feature a resistance campaign. Besides this crucial conceptual difference, Kadivar’s explanation of causal mechanisms often overlaps with our findings but also diverges in some significant aspects. He agrees that mass mobilization ‘creates a momentum that may marginalize authoritarian leaders in post-transition politics’ (Kadivar 2018, p. 394), echoing our ‘levelling the playing field’ argument. In addition, he supports Tarrow’s contentious politics argument that protest has a lasting impact on participants but also on political culture more widely: ‘Activism begets activism. A well-organized civil society also empowers excluded groups to organize themselves, articulate their interests, and to win inclusion and redistribution. Furthermore, the past success of contention legitimizes contentious methods as a major means of claim-making’ (Kadivar et al. 2019, p. 6). However, in contrast to our findings, Kadivar places great importance on movements training a cadre of democratic activists that replace authoritarian elites, first through ‘the day-to-day work and experience involved in building and maintaining complex movement organizations’ (Kadivar 2018, p. 384), and second by popularizing and endorsing leadership figures. Clearly, Kadivar is more bullish than we are on movement actors as the decisive vectors through which movements exert a democratizing influence. This may be due to the choice of case studies—there is little evidence of such careers having a strong influence on democracy in Benin and Chile, whereas Kadivar’s cases of South Africa and Brazil provide more prominent examples. There is some ambiguity in Kadivar’s argument as to whether it is movement activists in a narrow sense who enter the halls of power, or whether this also includes opposition politicians and other ‘regime outsider’ elites who have allied themselves with the protesters—an interpretation that would be more compatible with evidence from our cases. Interestingly, the case of Namibia would fit his

154

D. LAMBACH ET AL.

explanation very well, as SWAPO insiders basically took over the country after a prolonged struggle (Bayer 2017), although Kadivar maintains that his model does not apply to armed mobilization (Kadivar et al. 2019, p. 11). Kadivar also highlights the importance of organizations to movement success. Long-term campaigns cannot be waged without the participation of civil society organizations and without a degree of institutionalization within the movement itself. The long-term effect of movements rests on these networks of organizations and their accumulated capabilities surviving into post-transition democracy (Kadivar 2018, pp. 409– 410; Kadivar et al. 2019, p. 5). Our relational approach has proven very useful in conceptualizing transition as a series of strategic interactions among key actors. Importantly, we do not treat relations as separate from, or ontologically prior to, structures and institutions. This commitment allows for a theorizing how strategic interaction locks events and outcomes of political transition into institutions and sediments them into political culture. For instance, in Benin the constitutionalization of specific relationships among key actors during the National Conference made a huge difference for democratic survival. This makes it possible to bring together political agency and path dependence into the same model. Such an approach can also be used to bring in findings from other fields, such as work on the role of military defections during transition (Lutscher 2015; Barany 2016). Relational accounts of politics have become more prominent in comparative politics in recent years (see della Porta 2016, for a similar approach in social movement research). Transition research has historically tended towards elite-focused or structuralist theories in the past (Kadivar 2018; Haggard and Kaufman 2016), only according an ambivalent role to ordinary citizens and civil society (Bermeo 2003, pp. 3–20). In contrast, works on repression and revolution had traditionally taken a more interactionist approach, modelling protest as a strategic game between a repressive regime and a dissenting citizenry (Lichbach 1987). For instance, Boudreau (2004) offers an interactionist approach to protest and repression in South-East Asia. His model is that of state repression and public resistance as adaptive towards each other: ‘Over time, a relational logic emerges in the state and social sides of political struggle that informs authorities’ views about the difference between harmless and subversive mobilization, governs what challenges provoke state attacks, and structures consequent political contention’ (Boudreau 2004, p. 4). Reminiscent of later works by Gerschewski (2013), Boudreau emphasizes the

5

INCHING TOWARDS THEORY

155

wide variety of repressive tactics and instruments that authoritarian governments have at their disposal. Such interactionist and relational approaches are becoming more common in an increasing number of research fields, albeit with differing degrees of complexity. For the democratic backsliding literature, Waldner and Lust propose a conceptual framework centred around a relatively simple balance-of-power approach: ‘We can approach the study of vertical and horizontal accountability, for example, by considering the relative strength of actors who prefer greater or lesser degrees of constraint on executive power’ (Waldner and Lust 2018, p. 108). In his work on democratic quality, Usmani offers an updated version of earlier theories of class struggle, viewing democracy as ‘a contest featuring elites (landlords and industrialists) and nonelites (peasants, workers, and perhaps the middle class)’ (Usmani 2018, p. 669). He focuses on the distribution of power between these social forces and how ordinary citizens can force concessions from elites. Many works on civil–military relations take a similar approach, with most of them focusing on the institutional position of the armed forces and their relations to the civilian executive, and what this implies for the odds of democratic development (Powell et al. 2018; Kuehn 2017; Tusalem 2014). As an example of a more sophisticated approach, Heller proposes a very complex relational model to study democratic deepening. He argues that categories like the state, civil society or social movements should be disaggregated into relational clusters: ‘A relational approach as such calls for carefully unpacking the sometimes contradictory relationships between the state and the civil society and the way in which these shifting relationships both reflect societal power and shape the functioning of the state and civil society’ (Heller 2019, p. 358). While these examples underline the rising prominence of relational approaches to democratic survival and consolidation, they also offer suggestions as to how our own relational approach could be improved. For example, we have treated the four key actors in our relational model as relatively coherent and rational collective actors. This was mostly for the sake of brevity and so as not to make our theoretical approach overly complex, but it is clear that every single of these groups, in line with Heller’s argument, consists of factions, networks and flanks with their own interests and different degrees of cohesion. In addition, while the relational approach as detailed in Chapter 2 makes relatively few assumptions about what relations look like, we still have a theoretical predisposition towards citizens as the prime mover of social change. Instead,

156

D. LAMBACH ET AL.

we should emphasize the circularity and contingency of relations more. Popular protest does not emerge out of societal concerns alone but can also be conditioned or even caused by elite politics. To improve our relational approach, we need a better understanding of the mutual relationship between mass politics and elite dynamics, where both can be the cause of change in the other. Recent works by Hale, Kim and Kroeger, and della Porta offer suggestions on how this can be conceptualized. Hale (2019, pp. 6–8) distinguishes between three kinds of relationships between mass mobilization and elite splits. First, protest might be a catalyst for elite splits, encouraging regime insiders to defect from the governing coalition. Second, elite splits may open political opportunities for mass protest, although outcomes will still largely be determined by regime dynamics. Third, competing elite networks may actively foment protest as an instrument of leverage in their struggles over political and economic power. This ties into our argument that we should more carefully disentangle the causal role of resistance movements in the process of transition (see next section). Della Porta distinguishes ‘eventful democratization’, where regimes topple under public protest, from ‘participated pacts’, where elite softliners and movement representatives negotiate transition, and ‘troubled democratization’, where repressive regimes prevent peaceful mobilization, often leading to an escalation of violence (della Porta 2016, p. 4). Kim and Kroeger (2019, p. 651) offer several other suggestions. In their approach, anti-regime protests can (1) directly overthrow autocratic regimes, (2) pressure incumbent elites into liberalizing reforms, (3) increase the risk of elite splits, and (4) encourage leadership change within the regime. This demonstrates that within the same general relationship of nonviolent revolution leading to regime change, there is tremendous scope for empirical variation at the level of mechanisms.

Towards a Theory of NVR and Democratic Consolidation What can be learned from the preceding discussions that can help us move towards a theory of NVR and democratic consolidation? Where are the promising directions for new research? First and foremost, we see a need to further explore the linkages of NVR research with related

5

INCHING TOWARDS THEORY

157

fields. Donatella della Porta often makes the point that the social movements and contentious politics literature and works from comparative politics only rarely engage with each other. We have the same impression even though all these fields share common concerns, have similar theoretical approaches and arrive at comparable results, as our brief survey above indicates. Fabrice Lehoucq has argued that NVR research cannot assess the effectiveness of NVR without also studying authoritarianism (Lehoucq 2016, p. 271). We, in contrast, focusing on the long-term outcomes of NVR, need a better understanding of politics in consolidating democracies. A theory of NVR and democratic consolidation would have to integrate these various fields. This goes beyond what we can do at this point, which is why we focus on identifying those issues that are particularly relevant for the outcome and impact part of the process. We see the biggest scope for advancement in theory-building in a more detailed specification of the transition process. (Several other aspects in need of further clarification are discussed below.) In our approach, we assign relatively little agency to incumbents, at least when compared to other models of democratic transition. Instead, we place the onus for democratic change on the masses mobilizing for democracy and treat governments as if they are mainly acting under pressure from below. This assumption is not unreasonable per se—there are cases where this is precisely what happened—but it is unreasonable to posit this as some kind of default scenario. This becomes evident if we recall the findings by Haggard and Kaufman discussed above, that only 30% of distributive conflict transitions were cases of ‘direct replacement’ whereas 70% were pacted transitions occurring under pressure (Haggard and Kaufman 2016, pp. 105–106). We do not need to go as far as elite-centred theorists who argue that it is the incumbent elites who make the meaningful decisions during transition (e.g. Lehoucq 2016). Instead, we need to acknowledge that incumbents can exert influence at several key points: whether to use force against protesters and thereby push for an escalation, whether (and with whom) to enter into negotiations, whether (and how) to implement agreements. Taking the interactions between masses and elites more seriously also alerts us to a possible paradox of NVR and democratic consolidation. A successful nonviolent campaign requires pro-democracy activists to forge a broad-based alliance powerful enough to challenge the regime, a point that Chenoweth and Stephan (2011) repeatedly stress. Hence,

158

D. LAMBACH ET AL.

NVR movements need to get segments of the political, military and economic elite as well as the factions with which they are aligned to defect. To make defection more attractive, the resistance movement may have to give assurances that elites’ interests, status and resources will not be negatively affected by a democratically elected government to prevent these elites from becoming spoilers after transition. This forces a resistance movement into political concessions and opens up the risk of their co-optation by these same elites, running counter to the prospects of democratic consolidation. Bringing these interactions to the fore requires a revision of our basic approach. If we wish to preserve the trichotomy of NVR, VR and NoR— and there are arguments for abandoning such a categorization scheme entirely—we need to expand them to account for the variety of subtypes that distinguish different transition processes. For example, NVR in Benin led to a replacement of incumbent elites through the National Conference process whereas in Chile it led to an elite-dominated process of negotiated transition which preserved the veto powers and privileges of incumbent elites and the military. In short, we need a model of transition that takes the interaction between the government, the opposition, the military and citizens seriously, much as we argued in the previous section. A core element of this model should be a differentiation between the drivers (or causes) and the process (or mode) of transition. The drivers of transition are the factors that initiate or cause: first, a change in relations between the actors involved and, second, the transition towards democracy. NVR and popular protest can be such a driver of transition, just as several other factors which are commonly referenced in statistical analyses, such as economic downturns, external shocks, civil war, and electoral irregularities. The process of transition refers to the actual conduct of democratization, e.g. whether there were negotiations, elections, constitutional changes or a national conference, and the thousands of decisions and micro-dynamics that led to this particular shape of things. In other words: the driver aspect refers to the why of transition, the process aspect to the how. NVR movements may be more important as a driver of transition than they are as active participants in the process, but there are many well-publicized examples of the latter (such as Tunisia or Poland), so this would be a matter of more comparative research. As recent NVR movement research has shown, it takes a very disciplined and well-organized movement to inject itself into the discussions and negotiations about the future shape of the political system. In most cases, the

5

INCHING TOWARDS THEORY

159

resistance movement works by putting pressure on elites, but these same elites will still have substantial agency in structuring the process of transition. The taxonomy of interactions that Hale (2019), della Porta (2016), and Kim and Kroeger (2019) develop (see previous section) provides a useful tableau of the many ways this relationship between a movement (as the driver of transition) and incumbent elites (as the actor who has great influence over the process) can play out. This way of distinguishing between the destabilization of the old political order and the process leading to the inauguration of a new one is reminiscent of classic approaches from comparative politics which are rarely cited in the NVR literature. Early transition theories (with Linz and Stepan 1978 probably the best-known example) frequently used a threestage model of transition that proceeded from authoritarian breakdown (or liberalization) via democratic institutionalization to democratic consolidation (Encarnación 2000). In these terms, mass nonviolent protest would be very effective in the breakdown phase but possibly less so in the institutionalization phase. But in contrast to Linz and Stepan and other transitologists, we argue that what occurs in these phases has long-term consequences for consolidation (see also Celestino and Gleditsch 2013). In sum, we arrive at the following interim theoretical assumptions: There is a systematic effect of NVR on the transition process. By weakening incumbents, NVR movements exert pressure on them to reach some kind of negotiated agreement with the opposition and/or movement representatives, or face the risk of a forceful removal from power. In most cases, movements are at a disadvantage when it comes to the choice of transition instruments—e.g. national dialogue, national assembly, pacted transition, elections, power-sharing, etc.—and the composition of participants. This is because movements may lack the skills, capacity and internal legitimacy to enter into negotiations, or because they may be unprepared or hostile towards elites (Dudouet 2017; Wanis-St. John and Rosen 2017). In contrast, elites arguably have more opportunities to shape the process as they have the political experience, the social capital, and the technical and legal expertise to design a transitional process. Incumbent elites still enjoy the advantage of having the formal authority to establish political processes and institutions. This is why we expect that the process of transition will, more often than not, be more closely attuned to the terms of political elites. Beyond that, it is difficult to make specific theoretical predictions about how NVR is likely to impact the process of transition. For the

160

D. LAMBACH ET AL.

moment, we expect that the involvement of NVR movements will lead to a more inclusive process, increasing the public legitimacy of the outcome. This link is well-established in research on inclusive peace processes (e.g. Papagianni 2009). This in turn will lead to a more inclusive democracy, similar to what Fishman (2017) found for Portugal. Benin is another good example of that, where the resistance movement first pressured President Kérékou into setting up a National Conference and then continued to protest for better representation in the conference. This can also be contrasted with VR transitions, where negotiations usually happen between the major armed actors (government and rebels) with little representation of interests outside the armed movement. In NoR cases, negotiations typically take place between elites via organized political interests, with little to no involvement of citizens or civil society groups. Our findings show that NVR has a stabilizing effect on democracy induced by popular protest. But if we separate out the protests (as the driver) from the actual process of transition, we need to revisit and reformulate our causal mechanisms. First, the levelling effect of NVR is very useful in weakening incumbent elites going into the transition, thereby making a more equitable process and outcome more likely. Second, NVR’s support for a more democratic political culture mostly sidesteps the transition process, largely working within civil society itself. Third, avoiding the praetorian problem would manifest itself in the military being more likely to stay out of the transition process, although there may be cases where high-profile military defections make some inclusion of the military in negotiations necessary. This would have to be further substantiated through more detailed empirical analyses, which we expect would unearth certain empirical regularities. If there was a dataset on modes of transition, we might be able to see whether NVR leads more frequently to particular modes of transition, or whether particular kinds of resistance (nonviolent, violent, no resistance) combined with specific modes of transition (negotiation, early elections etc.) produces specific outcomes, although any analysis is likely to be limited due to the small number of cases among such subtypes. It would also be interesting to test whether specific drivers of transition produce different degrees of inclusiveness of the transition mechanism, e.g. in terms of how many parties are involved and how decisions are taken. We agree with Pinckney that too-rapid demobilization of the NVR movement will lead to it being sidelined by political elites during the transition process: ‘This popular engagement is crucial for keeping decision-makers

5

INCHING TOWARDS THEORY

161

accountable as a country’s new political institutions are being created. If rapid demobilization takes place early in the transition, the balance of force shifts back to the elites, particularly to the remnants of the old regime. These transitions are then likely to lead to limited democratization with a higher possibility of democratic backsliding down the road’ (Pinckney 2018, p. 24). Our finding that even NVR cases face a challenge to survive the first electoral turnover after transition may indicate that those democracies where movements have disappeared too rapidly were insufficiently strengthened to survive electoral turmoil. Our tentative conclusion from the observations made above would be that resistance campaigns create political junctures where actions of elites, within the boundaries established by mass activism, shape the political system. Under pressure from the grassroots, elites (both incumbents and opposition) create institutions (such as constitutions) and organize the transition process (national assemblies, elections, etc.) which lock in democratic gains. In this way, NVR campaigns can lead to the creation of institutions, processes and elite bargains that make a country more resilient to democratic breakdown over the long term. This is consistent with the findings of Bethke and Pinckney (2019) that NVR-induced transitions were significantly more likely to lead to an improvement in freedom of expression scores than other forms of democratic transition. A free media would be an example of such an institution that makes democracy more resilient. There are several other aspects where we see a need for further clarification. These are resistance as a symbolic resource, elite turnover during and after transition, political efficacy beliefs, the frailty of democracy in the early post-transition phase, and the space(s) of resistance. All of these aspects emerged as possibly significant during our research but we were unable to fully investigate them for one reason or another. Some of these reasons were methodological, e.g. for the symbolic resource dimension which is difficult to conceptualize and study in a systematic way; some of them were related to data availability and a lack of time and resources to conduct further tests. This applies, for example, to the political efficacy aspect which could be tested through survey experiments, or the issue of elite turnover which first requires more detailed data collection. As such, we wish to highlight those areas as avenues for further research to move the research programme forward.

162

D. LAMBACH ET AL.

Acts of resistance sometimes become an important symbolic resource for civic mobilization and protest down the road. This seems to be particularly important for ‘mass-driven’ protests, where civil society organizations are less able to act as ‘stewards’ of collective identity and memory. In such cases, remobilization cannot depend on established formal networks and institutions but has to work through deploying well-known symbols of resistance. For example, protests in Benin consciously recycle a prominent slogan (‘touche pas à ma constitution’) from earlier rounds of protest against democratic erosion. In other cases, movement names, practices, colours and images are used to situate a particular protest in a historical continuity of resistance. The legacy of resistance lives on through narratives of heroism in the face of government oppression, through socialization and education of coming generations, and through commemoration and memorialization of events. But as yet, we do not have a good grasp of what makes resistance become part of a society’s collective memory and how that matters for later remobilization. How does the public ‘select’ which acts of resistance become part of collective memory? The self-immolation of Mohamed Bouazizi is widely known in Tunisia and abroad but what about the Algerian protesters Maamir Lotfi, Said H. or Afif Hadri who followed Bouazizi’s example? Are there any inherent or contextual characteristics that make an event more memorable? If so, this can be exploited by resistance movements. Of course, movements already use highly visible acts of resistance as a core tactic of protest, but this issue goes beyond short-term acts towards a long view of inspiring the next generation of protesters. It would seem useful to engage with the literatures on memory politics (Assmann and Shortt 2012) and transitional justice, where the establishment of ‘historical truths’ represents an important step towards overcoming social trauma (Mihr et al. 2018). Our cases demonstrate that NVR is superior to other forms of transition in levelling the playing field. As part of this mechanism, our cases also suggest that NVR leads to a higher degree of elite turnover. Intuitively, this makes sense. Revolutionary change, whether violent or nonviolent, should be more disruptive for incumbents than top-down liberalization. And while violent revolutions, as in Namibia, may install a new elite drawn from the revolutionary vanguard, nonviolent revolutions create a more inclusive political landscape. However, while theoretically compelling, we only have anecdotal evidence of this. For instance, an analysis of the anatomy of SWAPO undertaken by Melber, Kromrey and Welz (2017)

5

INCHING TOWARDS THEORY

163

shows that between 1990 and 2012 the respective ministerial cabinets in Namibia were almost exclusively (90–100%) made up people who had already been active in SWAPO before independence. This number has only started to drop since 2015 onwards due to generational changes. So far, each of the three post-transition Namibian presidents had been an activist in exile. If we view this in context, it becomes starkly visible just how entrenched the SWAPO political elite is. Our own analysis of mean resident times in office shows that the average minister in Namibia between 1990 and 2005 served 100 months (roughly 9 years) before being replaced or, more often, before being shifted to another (ministerial) position. This is a long time for a democratic system—even longer than in Cape Verde under single-party rule (1975–1990), when the mean resident time in cabinet was nearly eight years. Following the liberalization of Cape Verdean politics, this figure dropped to around three years. Nevertheless, some key figures of the PAICV, the former single party, even survived electoral defeat in 1990 and regained cabinet posts later on. In Benin, the average time in office prior to the transition was 50 months, dropping to 36 months after democratic transition. Furthermore, only three members of pre-transition cabinets (in addition to Mathieu Kérékou himself) were able to stay in or return to a cabinet post after the transition in Benin, meaning that a nearly complete replacement of the highest echelon of political elites had taken place. We also see potential in investigating the effect of NVR on beliefs of participants and bystanders. This represents a causal mechanism through which NVR could conceivably exert a stabilizing effect on democracy. Chenoweth and Stephan (2011, p. 32) expect that participation encourages citizens to protest again in future. Sharp (2008, p. 53) argues that resistance movements help spread techniques and practices of resistance throughout the population, preparing it for future mobilization against democratic backsliding if necessary. There is some research that supports these assumptions. Della Porta (2016) provides evidence how participation in protest leads to long-term cognitive and emotional change in citizens. Wang (2019) explores how personality traits affect the willingness to engage in protest. Vinthagen (2015) introduces the notion of ‘utopian enactment’ to describe how relationships within the resistance movement performatively provide social roles for protesters but also for society at

164

D. LAMBACH ET AL.

large. However, there is little systematic evidence because conducting survey research and recurring tests in societies undergoing contentious politics is often prohibitively difficult. In a rare example of such research, Beissinger (2013) finds little evidence of pro-democratic attitudes among participants in the Ukrainian ‘Orange Revolution’, both during protests and after transition. As a result, we are sceptical about the widespread but facile assumption that protesters against an autocratic regime have particularly democratic attitudes and that this continues into the post-transition phase (e.g. Welzel 2007). Instead, we expect the impact of protest participation or observation to be located not at the level of normative attitudes but at a psychological level. Based on the assumptions of Gene Sharp (1973a), we hypothesize that NVR encapsulates a more inclusive idea of citizenship that affirms people’s political efficacy, i.e. their innate belief that they are able to influence the direction of the political system. Political efficacy, as originally defined by Campbell, Gurin, and Miller refers to ‘the feeling that political and social change is possible, and that the individual citizen can play a part in bringing about this change’ (Campbell et al. 1954, p. 187) and is generally seen as a determinant of democratic participation. Political efficacy can be further disaggregated into internal and external efficacy: The first describes the ‘individuals’ self-perception that they are capable of understanding politics and competent enough to participate in political actions such as voting’ (Miller et al. 1980, pp. 273–274). The second measures the belief that political institutions and representatives are responsive and open to ‘input’ and criticism from below. Political efficacy, together with political grievances and resources, is seen as a predictor to explain the emergence of social movements and the inclination of individual to engage in civic activism (Klandermans 2004). However, some research also suggest that political efficacy is not only a precondition for participation, activism and protest but also its result. Social psychologists Drury and Reicher (1999) show that the mere expectation of mutual support from a larger group can lead to an enhanced feeling of internal political efficacy. Similarly, Cocking and Drury postulate that ‘the greater the participation, the greater the sense of efficacy’ (Cocking and Drury 2004, p. 419). It thus seems plausible that engaging in or observing NVR can have a substantial effect on individual’s political efficacy and consequently on the political culture of a respective country. However, it would have to

5

INCHING TOWARDS THEORY

165

be seen how much the effect of NVR differs from that of violent resistance, which is exercised and organized differently but which also elicits positive emotions from participants. Further, effects might be temporary since the external efficacy especially is extremely volatile and dependent on the assessment of respective political regime and its political opportunity structure (Tarrow 1998). At the time of writing, Tunisia—less than ten years after the peaceful revolution that sparked the Arab Spring—struggles with widespread voter apathy and political disillusionment. While this might not be an unusual process after a revolution, it might as well be the corresponding individual process to the broader political demobilization of civil society. Currently, we still lack basic research tracing the effect of different modes of resistance on political efficacy, especially on the question of how lasting these effects are. Our statistical tests demonstrate that democratic breakdown almost exclusively occurs within the first ten years after transition. This is true for all modes of transition but especially pronounced for NVR-induced transitions. This should not be surprising for students of comparative politics where the elevated risk of breakdown in ‘young democracies’ has been demonstrated again and again (Kapstein and Converse 2008). For example, Keefer (2007) shows that young democracies perform significantly worse than established democracies on a range of indicators, such as higher levels of corruption, lower quality of the rule of law, and less administrative capacity. Furthermore, Merkel (2004) argues that such defects in young democracies cannot be attributed to single causes but to combinations of factors, such as the path and level of modernization, economic trends, and social capital. What these and other works highlight is that political institutions matter for democratic survival and consolidation. In young democracies, formal institutions are still weak and stability has to be maintained through extra-institutional factors like public sentiment, political culture and elite behaviour. With NVR (a) helping to lock in democratic gains early on, (b) limiting the number of veto players and preventing the formation of dominant parties by levelling the political playing field, (c) having a positive effect on civic culture, and (d) leading to a more inclusive transition process, we might help explain this divergence. For example, we think that NVR is useful to indirectly limit the impact of spoiling behaviour after transition. Spoilers exert a particular influence during the early phase of democratic transition when institutions are being (re)built and the ‘rules of the game’ are still being negotiated. Over time, incentives to buy into

166

D. LAMBACH ET AL.

the system increase, turning spoilers into (reluctant) supporters of democracy. Consistent with our survival analysis, where fewer NVR cases broke down during the first decade, we expect that a more inclusive transition produces fewer potential spoilers, thus lessening the risk of breakdown. For this, as well as more generally, a process-tracing approach to consolidation would be helpful, especially from a comparative perspective. This would help us identify typical ‘bottlenecks’ of democratic development, i.e. events where significant numbers of democracies break down and the failure mode in this particular stage (e.g. autogolpe, electoral violence, military coup). Markus Bayer takes such an approach in his analysis of critical junctures in the democratic development in Benin. He identifies the period of NVR, the National Conference, the inaugural elections, the first peaceful turnover of power and the unsuccessful attempts of Presidents Kérékou and Boni to abolish or side-step the presidential two-term limit as critical junctures where democracy in Benin could have failed (Bayer 2018, pp. 18–21). It would also make sense to bring in additional contextual variables that measure crucial dimensions of political institutions. For instance, research on state capacity indicates that higher infrastructural capacity of the state makes democracies more resilient (Andersen et al. 2014; Fortin 2012). Finally, there is more work to be done on the space(s) of resistance. This refers, first, to the urban or rural situatedness of the movement. Nonviolent movements ‘tend to mobilize large numbers of people in urban areas, particularly in capital cities’ (Kim and Kroeger 2019, p. 654). Central squares in the capital like the Maidan in Kiev and Tahrir Square in Cairo have become symbols of citizens assembling in protest against authoritarian regimes. In contrast, rural resistance movements tend to take the form of violent insurgencies (García-Ponce and Wantchekon 2017), with obvious exceptions like Gandhi’s Civil Disobedience Movement, at least when it comes to resistance movements with maximalist goals. NVR as a technique is often used by rural populations struggling for more minimalist, local causes, such as protesting against state overreach (Scott 1985) and for their land rights (Schock 2015). The location of resistance matters for political transition and consolidation. First, being in the capital makes it easier for nonviolent movements to exert pressure on elites, most visibly by engaging in protest right outside the windows of the halls of power where the transition process is being negotiated. Second, if we presume that NVR affects the attitudes and feelings of political efficacy of participants and observers, then it matters which segments of

5

INCHING TOWARDS THEORY

167

the population are exposed to it. Protests in the capital reach different audiences than protests in the countryside. More theoretically, there is also a need to better understand how movements use space as part of their strategies of protest, e.g. through occupation, use or defacement. Routledge calls protesters ‘space invaders’ and offers an agent-structure logic how movements relate to space: ‘Particular places provide protesters with opportunities and constraints as they wage their struggles. Places can influence the character of protests as well as being transformed by them. Protesters make space, and in so doing they can imbue placed with different meanings and feelings’ (Routledge 2017, p. 1). There is a wealth of literature in critical, postcolonial and feminist geography covering the ‘production’ of space that could be used to inform further research (for an introduction, see Massey 2005). This also matters for the phase of democratic consolidation where practices of space-making become visible, for example through the erection of monuments or through holding memorial marches in public spaces on significant dates. These are important practices of memory politics that reify the status of NVR as a symbolic resource in political culture.

References Acemoglu, D., & Robinson, J. A. (2006). Economic Origins of Dictatorship and Democracy. Cambridge: Cambridge University Press. Acharya, A., Blackwell, M., & Sen, M. (2018). Deep Roots: How Slavery Still Shapes Southern Politics. Princeton: Princeton University Press. Albertus, M., & Menaldo, V. (2018). Authoritarianism and the Elite Origins of Democracy. Cambridge: Cambridge University Press. Andersen, D., Møller, J., Rørbæk, L. L., & Skaaning, S.-E. (2014). State Capacity and Political Regime Stability. Democratization, 21(7), 1305–1325. https://doi.org/10.1080/13510347.2014.960204. Assmann, A., & Shortt, L. (Eds.). (2012). Memory and Political Change. London: Palgrave Macmillan. Barany, Z. (2016). How Armies Respond to Revolutions and Why. Princeton: Princeton University Press. Bayer, M. (2017). Swapo Forever? Prospect for Liberal Democracy or Prolonged One-Party Dominance in Namibia. Journal of Namibian Studies, 21, 27–54. Bayer, M. (2018). The Democratizing Effect of Nonviolent Resistance: How Nonviolent Resistance Featured Democratic Consolidation in Benin (Working Paper) (Vol. 3/2018). Basel: Swisspeace.

168

D. LAMBACH ET AL.

Beissinger, M. R. (2013). The Semblance of Democratic Revolution: Coalitions in Ukraine’s Orange Revolution. American Political Science Review, 107 (3), 574–592. https://doi.org/10.1017/S0003055413000294. Bermeo, N. (2003). Ordinary People in Extraordinary Times: The Citizenry and the Breakdown of Democracy. Princeton: Princeton University Press. Bethke, F. S., & Pinckney, J. (2019). Non-Violent Resistance and the Quality of Democracy. Conflict Management and Peace Science. https://doi.org/10. 1177/0738894219855918. Boix, C. (2003). Democracy and Redistribution. Cambridge: Cambridge University Press. Boudreau, V. (2004). Resisting Dictatorship: Repression and Protest in Southeast Asia. Cambridge: Cambridge University Press. Butcher, C., Gray, J. L., & Mitchell, L. (2018). Striking It Free? Organized Labor and the Outcomes of Civil Resistance. Journal of Global Security Studies, 3(3), 302–321. https://doi.org/10.1093/jogss/ogy010. Butcher, C., & Svensson, I. (2016). Manufacturing Dissent. Journal of Conflict Resolution, 60(2), 311–339. https://doi.org/10.1177/0022002714541843. Campbell, A., Gerald, G., & Miller, W. E. (1954). The Voter Decides. Evanston: Row, Peterson and Company. Carothers, T. (2002). The End of the Transition Paradigm. Journal of Democracy, 13(1), 5–21. Celestino, M. R., & Gleditsch, K. S. (2013). Fresh Carnations or All Thorn, No Rose? Nonviolent Campaigns and Transitions in Autocracies. Journal of Peace Research, 50(3), 385–400. https://doi.org/10.1177/0022343312469979. Chenoweth, E., & Belgioioso, M. (2019). The Physics of Dissent and the Effects of Movement Momentum. Nature Human Behaviour. https://doi.org/10. 1038/s41562-019-0665-8. Chenoweth, E., & Stephan, M. J. (2011). Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. New York: Columbia University Press. Cocking, C., & Drury, J. (2004). Generalization of Efficacy as a Function of Collective Action and Intergroup Relations: Involvement in an Anti-roads Struggle1. Journal of Applied Social Psychology, 34(2), 417–444. https://doi. org/10.1111/j.1559-1816.2004.tb02555.x. della Porta, D. (2014). Mobilizing for Democracy: Comparing 1989 and 2011. Oxford: Oxford University Press. della Porta, D. (2016). Where Did the Revolution Go? Contentious Politics and the Quality of Democracy. Cambridge: Cambridge University Press. della Porta, D. (2018). Protests as Critical Junctures: Some Reflections Towards a Momentous Approach to Social Movements. Social Movement Studies, 1–20. https://doi.org/10.1080/14742837.2018.1555458. Diamond, L. (2008). The Spirit of Democracy: The Struggle to Build Free Societies Throughout the World. New York: Times Books.

5

INCHING TOWARDS THEORY

169

Diamond, L., Horowitz, D. L., Fukuyama, F., & Plattner, M. F. (2014). Reconsidering the Transition Paradigm. Journal of Democracy, 25(1), 86–100. Drury, J., & Reicher, S. (1999). The Intergroup Dynamics of Collective Empowerment: Substantiating the Social Identity Model of Crowd Behavior. Group Processes & Intergroup Relations, 2(4), 381–402. https://doi.org/10.1177/ 1368430299024005. Dudouet, V. (2017). Powering to Peace: Integrated Civil Resistance and Peacebuilding Strategies (Special Report) (Vol. 1). Washington, DC: International Center on Nonviolent Conflict. Encarnación, O. G. (2000). Beyond Transitions: The Politics of Democratic Consolidation. Comparative Politics, 32(4), 479–498. Enos, R. D., Kaufman, A. R., & Sands, M. L. (2019). Can Violent Protest Change Local Policy Support? Evidence from the Aftermath of the 1992 Los Angeles Riot. American Political Science Review, 1–17. https://doi.org/10. 1017/S0003055419000340. Fearon, J. D. (2011). Self-Enforcing Democracy. The Quarterly Journal of Economics, 126(4), 1661–1708. https://doi.org/10.1093/qje/qjr038. Fishman, R. M. (2017). How Civil Society Matters in Democratization: Setting the Boundaries of Post-transition Political Inclusion. Comparative Politics, 49(3), 391–409. Fortin, J. (2012). Is There a Necessary Condition for Democracy? The Role of State Capacity in Postcommunist Countries. Comparative Political Studies, 45(7), 903–930. https://doi.org/10.1177/0010414011428587. Gandhi, M. K. (1948). Constructive Programme: Its Meaning and Place (2nd ed.). Ahmedabad: Navajivan Publishing House. García-Ponce, O., & Wantchekon, L. (2017). Critical Junctures: Independence Movements and Democracy in Africa. Gerschewski, J. (2013). The Three Pillars of Stability: Legitimation, Repression, and Co-Optation in Autocratic Regimes. Democratization, 20(1), 13– 38. https://doi.org/10.1080/13510347.2013.738860. Grodsky, B. K. (2012). Social Movements and the New State: The Fate of ProDemocracy Organizations When Democracy Is Won. Stanford: Stanford University Press. Haggard, S., & Kaufman, R. R. (2016). Dictators and Democrats: Masses, Elites and Regime Change. Princeton: Princeton University Press. Hale, H. E. (2019). How Should We Now Conceptualize Protest, Diffusion, and Regime Change? Journal of Conflict Resolution, 0022002719862427. https://doi.org/10.1177/0022002719862427. Heller, P. (2009). Democratic Deepening in India and South Africa. Journal of Asian and African Studies, 44(1), 123–149. https://doi.org/10.1177/ 0021909608098679.

170

D. LAMBACH ET AL.

Heller, P. (2019). Divergent Trajectories of Democratic Deepening: Comparing Brazil, India, and South Africa. Theory and Society, 48(3), 351–382. https:// doi.org/10.1007/s11186-019-09351-7. Hollyer, J. R., Rosendorff, B. P., & Vreeland, J. R. (2018). Transparency, Protest and Democratic Stability. British Journal of Political Science, 1–27. https:// doi.org/10.1017/S0007123417000308. Igarashi, S. (2008). The Dilemma of Democratic Consolidation in the Philippines: The Contested Role of Civic Organizations in Electoral Governance. Philippine Political Science Journal, 29(52), 79–116. https://doi.org/10. 1080/01154451.2008.9723509. Kadivar, M. A. (2018). Mass Mobilization and the Durability of New Democracies. American Sociological Review, 83(2), 390–417. https://doi.org/10. 1177/0003122418759546. Kadivar, M. A., Usmani, A., & Bradlow, B. H. (2019). The Long March: Deep Democracy in Cross-National Perspective. Social Forces. https://doi.org/10. 1093/sf/soz050. Kapstein, E. B., & Converse, N. (2008). The Fate of Young Democracies. Cambridge: Cambridge University Press. Keefer, P. (2007). Clientelism, Credibility, and the Policy Choices of Young Democracies. American Journal of Political Science, 51(4), 804–821. https:// doi.org/10.1111/j.1540-5907.2007.00282.x. Kendall-Taylor, A., & Frantz, E. (2014). How Autocracies Fall. The Washington Quarterly, 37 (1), 35–47. https://doi.org/10.1080/0163660X.2014. 893172. Kim, N. K., & Kroeger, A. M. (2019). Conquering and Coercing: Nonviolent Anti-regime Protests and the Pathways to Democracy. Journal of Peace Research, 56(5), 650–666. https://doi.org/10.1177/0022343319830267. Kinsman, J., & Bassuener, K. (2013). A Diplomat’s Handbook for Democracy Development Support (3rd ed.). Waterloo: Centre for International Governance Innovation. Klandermans, B. (2004). The Demand and Supply of Participation: SocialPsychological Correlates of Participation in Social Movements. In D. A. Snow, S. A. Soule, & H. Kriesi (Eds.), The Blackwell Companion to Social Movements (pp. 360–379). Malden: Blackwell. Kodila-Tedika, O., & Asongu, S. (2018). Resistance to European Domination: Institutional Mechanism (MPRA Paper) (Vol. 85237). Munich: Munich Personal RePEc Archive. Kuehn, D. (2017). Midwives or Gravediggers of Democracy? The Military’s Impact on Democratic Development. Democratization, 24(5), 783–800. https://doi.org/10.1080/13510347.2017.1324421. Lehoucq, F. (2016). Does Nonviolence Work? Comparative Politics, 48(2), 269– 287.

5

INCHING TOWARDS THEORY

171

Lichbach, M. I. (1987). Deterrence or Escalation? The Puzzle of Aggregate Studies of Repression and Dissent. Journal of Conflict Resolution, 31(2), 266– 297. https://doi.org/10.2307/174013. Linz, J. J., & Stepan, A. (Eds.). (1978). The Breakdown of Democratic Regimes. Crisis, Breakdown and Reequilibration. Baltimore: Johns Hopkins University Press. Linz, J., & Stepan, A. (1996). Problems of Democratic Transition and Consolidation: South America, Southern Europe, and Post-communist Europe. Baltimore: Johns Hopkins University Press. Little, A. T., Tucker, J. A., & LaGatta, T. (2015). Elections, Protest, and Alternation of Power. The Journal of Politics, 77 (4), 1142–1156. https://doi.org/ 10.1086/682569. Lutscher, P. M. (2015). The More Fragmented the Better?—The Impact of Armed Forces Structure on Defection during Nonviolent Popular Uprisings. International Interactions, 42(2), 350–375. https://doi.org/10.1080/ 03050629.2016.1093476. Massey, D. (2005). For Space. London: Sage. Melber, H., Kromrey, D., & Welz, M. (2017). Changing of the Guard? An Anatomy of Power Within SWAPO of Namibia. African Affairs, 116(463), 284–310. https://doi.org/10.1093/afraf/adw073. Merkel, W. (2004). Embedded and Defective Democracies. Democratization, 11(5), 33–58. https://doi.org/10.1080/13510340412331304598. Mihr, A., Pickel, G., & Pickel, S. (Eds.). (2018). Handbuch Transitional Justice: Aufarbeitung von Unrecht - hin zur Rechtsstaatlichkeit und Demokratie [Springer Reference Sozialwissenschaften]. Wiesbaden: Springer VS. Miller, W. E., Miller, A. H., & Schneider, E. J. (1980). American National Election Studies Data Sourcebook, 1952–1978. Cambridge: Harvard University Press. Muriaas, R. L., Rakner, L., & Skage, I. A. (2016). Political Capital of Ruling Parties After Regime Change: Contrasting Successful Insurgencies to Peaceful Pro-Democracy Movements. Civil Wars, 18(2), 175–191. https://doi.org/ 10.1080/13698249.2016.1205563. O‘Donnell, G., Schmitter, P. C., & Whitehead, L. (Eds.). (1986). Transitions from Authoritarian Rule: Comparative Perspectives. Baltimore: Johns Hopkins University Press. Papagianni, K. (2009). Political Transitions After Peace Agreements: The Importance of Consultative and Inclusive Political Processes. Journal of Intervention and Statebuilding, 3(1), 47–63. https://doi.org/10.1080/ 17502970802608175. Pinckney, J. (2018). When Civil Resistance Succeeds: Building Democracy After Popular Nonviolent Uprisings. Washington, DC: International Center on Nonviolent Conflict.

172

D. LAMBACH ET AL.

Powell, J., Faulkner, C., Dean, W., & Romano, K. (2018). Give Them Toys? Military Allocations and Regime Stability in Transitional Democracies. Democratization, 25(7), 1153–1172. https://doi.org/10.1080/13510347.2018. 1450389. Putnam, R. D. (1993). Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press. Rivera, T. C. (2002). Transition Pathways and Democratic Consolidation in PostMarcos Philippines. Contemporary Southeast Asia, 24(3), 466–483. Routledge, P. (2017). Space Invaders: Radical Geographies of Protest. London: Pluto Press. Schock, K. (2005). Unarmed Insurrections: People Power Movements in Nondemocracies. Minneapolis: University of Minnesota Press. Schock, K. (2015). Rightful Radical Resistance: Mass Mobilization and Land Struggles in India and Brazil. Mobilization: An International Quarterly, 20(4), 493–515. https://doi.org/10.17813/1086-671x-20-4-493. Scott, J. C. (1985). Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven: Yale University Press. Sharp, G. (1973a). The Politics of Nonviolent Action. Part Three: The Dynamics of Nonviolent Action. Boston: Porter Sargent. Sharp, G. (1973b). The Politics of Nonviolent Action. Part One: Power and Struggle. Boston: Porter Sargent. Sharp, G. (2008). Von der Diktatur zur Demokratie. Ein Leitfaden für die Befreiung. München. Svensson, I., & Lindgren, M. (2010). Community and Consent: Unarmed Insurrections in Non-Democracies. European Journal of International Relations, 17 (1), 97–120. https://doi.org/10.1177/1354066109350049. Svolik, M. W. (2015). Which Democracies Will Last? Coups, Incumbent Takeovers, and the Dynamic of Democratic Consolidation. British Journal of Political Science, 45(4), 715–738. https://doi.org/10.1017/ S0007123413000550. Tarrow, S. (1998). Power in Movement: Social Movements and Contentious Politics. New York: Cambridge University Press. Thurber, C. (2019). Social Ties and the Strategy of Civil Resistance. International Studies Quarterly. https://doi.org/10.1093/isq/sqz049. Tomini, L., & Wagemann, C. (2018). Varieties of Contemporary Democratic Breakdown and Regression: A Comparative Analysis. European Journal of Political Research, 57 (3), 687–716. https://doi.org/10.1111/1475-6765. 12244. Tusalem, R. F. (2014). Bringing the Military Back In: The Politicisation of the Military and Its Effect on Democratic Consolidation. International Political Science Review, 35(4), 482–501. https://doi.org/10.1177/ 0192512113496683.

5

INCHING TOWARDS THEORY

173

Usmani, A. (2018). Democracy and the Class Struggle. American Journal of Sociology, 124(3), 664–704. https://doi.org/10.1086/700235. Vinthagen, S. (2015). A Theory of Nonviolent Action: How Civil Resistance Works. London: Zed Books. Waldner, D., & Lust, E. (2018). Unwelcome Change: Coming to Terms with Democratic Backsliding. Annual Review of Political Science, 21(1), 93–113. https://doi.org/10.1146/annurev-polisci-050517-114628. Wang, A. H.-E. (2019). Patience, Dynamic of Protest, and Democratic Consolidation. European Political Science, 18(3), 473–490. https://doi.org/10. 1057/s41304-018-0177-4. Wanis-St. John, A., & Rosen, N. (2017). Negotiating Civil Resistance. Peaceworks (Vol. 12). Washington, DC: United States Institute for Peace. Welzel, C. (2007). Are Levels of Democracy Affected by Mass Attitudes? Testing Attainment and Sustainment Effects on Democracy. International Political Science Review/Revue internationale de science politique, 28(4), 397–424.

CHAPTER 6

Conclusion

Securing democracy is an ongoing process. This has always been evident from high-profile cases of democratic breakdown, with examples ranging from many European democracies of the interwar period to more recent cases like Thailand, Fiji, or Niger, and contemporary processes of democratic erosion in Hungary, Russia, Turkey, and Venezuela (Lührmann and Lindberg 2019). For most of the post-1945 period, Western countries saw themselves as shielded from such events and processes. Expectations of the inevitable march of democracy, liberalism and progress were widely shared, as exemplified by Fukuyama’s much-discussed and much-criticized paean to liberalism, The End of History and the Last Man (Fukuyama 1992). But this triumphalism, as misplaced as it always has been, is being shaken up since at least 2016 when the Brexit referendum and the election of Donald Trump came to represent the rising backlash against liberalism even within Western societies. This has led to widespread fears that we are seeing an erosion of democracy on a scale and quality that seemed inconceivable only a scant few years ago. Current events, as of the time of writing, only play into these anxieties. In the United Kingdom, Prime Minister Boris Johnson’s decision to prorogue parliament for a record-breaking five weeks during a key period of the Brexit process in September and October 2019 is being widely condemned as an assault on British democratic traditions. In response, Nigerian playwright Wole

© The Author(s) 2020 D. Lambach et al., Nonviolent Resistance and Democratic Consolidation, https://doi.org/10.1007/978-3-030-39371-7_6

175

176

D. LAMBACH ET AL.

Soyinka asked for a fact-finding mission to see whether Britain merits suspension from the Commonwealth ‘for setting a bad example in democratic practice’. Across the Atlantic, US President Trump, in addition to his many other acts undermining democracy from the top, regularly hints that he might not accept results in the 2020 presidential elections if he is not re-elected, warning of (non-existent) vote-rigging to legitimize his ongoing assault on democracy. In Germany, the rise of the Alternative für Deutschland, a right-wing populist party, has led to fears about the resilience of democracy. The debate focuses mostly on the eastern Länder where the AfD regularly polls at 20–25%, a success that is often connected to the image of these regions as the ‘loser’ of reunification, buffeted by deindustrialization, high levels of unemployment and emigration of the young (Heitmeyer 2018). To some degree, this is nothing new. For too long, Western societies have been all too happy to ignore the populist threat to democracy which is only now entering public consciousness. Some cases in point are electoral successes of parties with clear exclusionary visions of society in various European countries over the last 20 years: Vlaams Belang getting 24% in the 2004 Flemish parliamentary election, the Freiheitliche Partei Österreichs getting 26.9% in the Austrian national elections in 1999, or the Front National ’s candidate coming second in the 2002 French presidential elections. These anxieties about the resilience of democracy, slowly forgotten among Western societies as the memory of the interwar period has receded, might seem familiar to democrats in many non-Western countries where democracy has never really been ‘the only game in town’, in Linz and Stepan’s memorable phrase (Linz and Stepan 1996, p. 15). Even Benin, whose democratic credentials we have repeatedly touted throughout this book, cannot consider itself secure from democratic erosion and breakdown. Given the country’s still-low level of education and economic development, it should not be forgotten that it continues to be a democracy against the odds. So it should come as no great surprise that Benin continues to undergo periods of political crisis. That democracy has weathered all previous challenges does not mean that it will continue to do so in the future, just that it is slowly getting better at facing them. The most recent challenge to democracy in Benin emerges from President Patrice Talon’s attempts to centralize power. To limit the large number of political parties the electoral reform of 2018 drastically tightened the rules for party registration. Based on the new electoral law, which

6

CONCLUSION

177

requires political parties to pay more than US$400,000 to field a list for parliamentary elections, the National Election Commission (CENA) only registered two parties—the Union Progressiste and the Bloc Républicain—for the elections on 28 April 2019. Since both of them support the current President Talon, this decision sparked widespread protests spearheaded by former President Boni Yayi. The opposition finally called for a boycott of the elections. Consequentially, the turnout dropped to 23% (previously 65%), by far the lowest since Benin adopted democracy. In the wake of the elections, violence erupted between supporters of Boni Yayi, who was placed under house arrest, and the police in Tchaourou, Boni Yayi’s hometown. During the clashes, the national army was called in and fired on the protesters with live ammunition, leaving several of them injured and several dead according to witnesses (ZEIT Online 2019). In June 2019, President Talon called for a political dialogue and promised that the CENA will soon register all oppositional parties. This move towards national reconciliation however did little to remove the tensions since it made it obvious that the exclusion of the oppositional parties from the election had been a strategic move. Since the elections, the parliament is de facto without opposition, which means a serious limitation of checks and balances on the president. This is especially alarming since the constitutional court has been headed by a close ally of President Talon, Joseph Djogbénou, since June 2018. The current situation can be seen as a serious test of democratic consolidation in Benin. Until the presidential elections in 2021, President Talon can govern de facto without any parliamentary control and legislative oversight. This, compounded with his control over the constitutional court, puts him in the position to push through constitutional changes, although it is currently unclear whether he will seek to do so and if so, in which form. Overall, the separation of powers is currently severely limited. Should Talon, as announced in 2017, seek a second presidential term, the decision on the future of democracy in Benin will in all likelihood fall in the presidential elections in 2021. In the past, such situations have repeatedly served as rallying points for the mobilization of the opposition and civil society. It will have to be seen whether the people of Benin can again turn out in force to repel the authoritarian tendencies of the present government. Similar tales can be told about Chile, El Salvador, Namibia, Paraguay and Cape Verde. As these examples and the cautionary tales of the UK,

178

D. LAMBACH ET AL.

the US and Germany show, democratic consolidation is always an unfinished business and democracy periodically needs to renew itself from within. The failure of Weimar Germany is often attributed to it being a ‘democracy without democrats’. Clearly, democracy needs ‘democrats’, citizens and elites who are choosing democracy day after day. This need not be a high-minded affair. Of course, symbols and rituals of democracy are important for anchoring political culture but a constant, ostentatious display of what Jürgen Habermas called ‘constitutional patriotism’ (Habermas 1990; Müller 2007) would probably be rather off-putting to citizens of liberal democracies. It is much more important for democracy to manifest itself in everyday behaviour. Elites that stick to institutional rules even if they work to their disadvantage, that accept electoral defeat and that call out undemocratic behaviour by others sustain democracies. It requires civil servants like supreme court judge Lady Hale, who declared the prorogation of the British parliament to be illegal, or of high court judge and former Solidarno´sc´ member Małgorzata Gersdorf, who resisted the Polish President Andrzej Duda’s attempt to send troublesome judges into early retirement. Finally, it also needs citizens who stay informed about politics, are civically engaged and use the various formal and informal avenues of participation that the political system offers. In our understanding, this includes the various methods of nonviolent resistance, starting from nonviolent protest and nonviolent non-cooperation and ending with more radical forms like nonviolent interventions or what Paul Watson calls ‘aggressive nonviolence’ (Buckmaster 2019). In short: public-minded citizens are the lifeblood of any democracy. So, on the one hand, democracies need active engagement from everyone to work satisfactorily. This might seem like democracy is a fragile flower, easily crushed when the weather turns against it, as Acemoglu and Robinson (2019) argue in their most recent work. But democracies are more resilient than they sometimes get credit for. We are often too eager to view dissent and conflict as a threat to democracy. This is certainly true in some cases, especially when such dissent is voiced by extremists whose commitment to democracy is doubtful. But we should not be too quick to see all ‘contentious politics’ (McAdam et al. 2001) as crises of democracy. Far from it—contestation and conflict are at the very core of democratic practice. Democracy is a form of institutionalized conflict management between parties competing for political power. As long as this competition adheres to certain procedural norms on how such conflicts should

6

CONCLUSION

179

be contested, vigorous debate arguably strengthens democracies rather than weakening them. Furthermore, democracies also distinguish themselves and might even grow from showing resilience in the face of those who try to circumvent it. This should make us a bit more sanguine about the ability of democracy to withstand challenges. If the fundamentals are sound, democracies can survive even difficult periods. We realize that this goes against the prevailing zeitgeist. Books like How Democracies Die (Levitsky and Ziblatt 2018) are popular bestsellers and receive attention far beyond academic circles. The Trump/Brexit era may turn out to be the twilight of democracy in the West but we would caution against overly alarmist claims. This is not to say that there are no problems facing Western democracy. Globalization and deindustrialization are changing social milieus and have eroded trust in democracy (Foa and Mounk 2016; Inglehart 2016). As Stefanie Walter (2017) shows in a comparison of 16 European countries, exposure to globalization increases perceptions of risk among low-skilled individuals. There is a danger that such feelings of insecurity among the ‘losers of modernization’ translates into disaffection and withdrawal (Bisbee et al. 2019). The austerity discourse that conveys the message that states are unable to offer social protection to workers dislocated by globalization who are consequently forced into the ‘gig economy’ and ‘zero-hours contracts’ has also done great damage to the image of democracy. These issues are stated most forcefully in Thomas Picketty’s seminal book ‘Capital in the Twenty-First Century’, where he argues that the concentration and unequal distribution of wealth in Europe and the US will cause social and political instability (Piketty 2014). At the same time, Western states continue to abet the erosion of democracy in other countries through their continued support for pliant dictators and the dismantling of international liberal institutions (Klaas 2018). Democracy, in developed and developing countries alike, needs a new vision that speaks to the masses of people who feel commoditized, alienated and forgotten in a globalized economy. We believe that our research might provide some inspiration for such a vision. There is an undeniable grace to citizens protesting peacefully for their freedom and their rights. Our findings only underscore the potency of these acts of protest which continue to reverberate in society long after the protesters have gone back home. Resistance can not only topple dictators, it can also help construct institutional guardrails that continue to

180

D. LAMBACH ET AL.

stabilize democratic politics long into the future. It can become an important touchstone in political culture, a shared narrative of ordinary heroism in the face of government oppression. Clearly, nonviolent resistance for a noble cause resonates with observers, and not just in the countries where resistance occurred themselves. The Arab Spring received worldwide attention a few years ago, just as earlier episodes of civic resistance in Europe, Africa, Asia, and Latin America. Today it is the Hong Kong protests that get noticed by a global audience. From an academic perspective, this should prompt scholars to reevaluate their assessment of the role of ordinary citizens (Bermeo 2003). Comparative politics often shows great scepticism about ‘the masses’, treating them as easily (mis)led by political elites who are the main players in politics and political transition. Class-based approaches may well ascribe the middle class an important role in sustaining democracy but it is cautious about their revolutionary potential. Only civil society is lionized as this crucial check on politics, but the concept itself is loaded with assumptions, usually referring to formal organizations dedicated to goals that are consistent with liberal democratic ideals (Jones 1998). Recent works on populism are also deeply elite-centred. They describe populist references to ‘the common people’ as little more than a political strategy deployed by elite regime outsiders against incumbents, while the masses are portrayed as this amorphous, angry set of people who can be mobilized against supposedly venal politicians and ultimately against democracy itself (Mudde and Rovira Kaltwasser 2017; Müller 2017).1 All of these theories betray an elitist mindset of how democracy, or politics more generally, works, or should work. In contrast, we think that citizens are more important for politics than they are often perceived to be. The example of Benin shows how sustained protest with little central organization can bring down a dictatorship. Examples from other countries also support this point. Civic mobilization is crucial in what della Porta (2014) calls ‘eventful transitions’, i.e. those revolutionary moments when the traditional order is challenged and everything seems possible. These moments exert a great mystique.

1 As part of this re-evaluation we should reconsider the usage of the term ‘masses’ which implies an objective difference between elite and masses. In a democracy, however, every individual, at least in theory, can become part of the political ‘elite’. Thus, if we wish to acknowledge the agency of the citizen we should be cautious about reifying the duality of elite-versus-masses.

6

CONCLUSION

181

They are full of potential and are sometimes described as exhilarating by participants. But revolutionary moments are just that—extraordinary moments of popular mobilization. These intense times are inevitably followed by a normalization of politics and day-to-day life. Typically, the movements demobilize, activists return to their families and workplaces. Elites, whether old or new, take over and re-establish the principle of representation. Some revolutionary regimes have tried to resist this transition into normality, from the Chinese Communists and the Khmer Rouge to the Gaddafi regime, often at terrible human cost (Scott 1998). The revolution cannot be institutionalized, even though the Partido Revolucionario Institucional, the party governing Mexico from 1929 to 2000, claimed the opposite. But civic mobilization matters for these periods of normal politics, too. First and most obviously, a vigorous and activist public is important for the maintenance of democracy. People getting involved in local politics, joining political parties, setting up community associations and non-governmental organizations, sustaining a free press, and organizing petitions and protest are the foundation upon which democracy is built. Second, and perhaps less obviously, civic mobilization during transition also matters for the long-term viability of democracy. Our findings, and those of others working on similar issues, are clear on that: when ordinary people peacefully mobilize for change, they leave legacies that continue to impact democratic politics years, even decades, down the line. Of course, mass mobilization can also have destructive and deleterious consequences. Our aim is not to imbue citizens with some special place in democracy, accord their actions a special kind of legitimacy or to portray them as some higher fount of wisdom—this would be precisely the populist argument. Our point is more modest: we wish to give citizens their due place in models of political change and stability. Where they were often overlooked by the classics, we believe that a relational model of political regimes accords citizens an important function and focuses our attention on the interaction among elite factions and ordinary citizens.2 This entails a recognition of the collective agency of citizens without making strong assumptions about ontological priority or the balance of forces between elites and the citizenry. 2 This harks back to the classical literature on state–society relations (e.g. Migdal 1988; Azarya 1988) which, at least in our view, been somewhat sidelined as a research program in comparative politics.

182

D. LAMBACH ET AL.

In addition, even though this book is primarily aimed at engaging scholars across the fields of nonviolent action, social movements, comparative politics and democratic transition, its findings are also relevant for democrats worldwide—from pro-democracy activists claiming their basic rights to diplomats or NGOs seeking to support peaceful transitions of power in authoritarian regimes, ‘electoral autocracies’ (Seeberg 2014) or ‘defective democracies’ (Merkel 2004). On the one hand, our comparison between bottom-up political transitions driven by civil resistance and those initiated by elite pacts or transitions involving armed opposition groups has evidenced the long-term effects of nonviolent collective action on the quality of democracy. On the other hand, our paired analysis of democratization in Chile and Benin also highlighted the critical importance of short-term transitional dynamics in influencing the future legacy of peaceful struggles. Hence both perspectives—careful planning for the transition itself and a long-term vision for sustained democracy—warrant attention by nonviolent movements in ongoing struggles or recent people-power revolutions such as Sudan and Algeria. The mechanisms of change uncovered in preceding chapters point to the role of collective nonviolent action in opening up political space by levelling the playing field, promoting a culture of participation and restricting anti-democratic backlash by veto players such as the military. These findings carry a few implications for nonviolent movements, including the importance of sustained mobilization beyond the toppling of a dictator. As highlighted by Pinckney, activists must carefully ‘think through the political visions, organizational structures, and practices of resistance that will continue to push their country’s political system towards a democratic, representative future’ (Pinckney 2018, p. 74). Sudan offers a positive example in this regard. There, the civil society coalition Forces of Freedom and Change kept up the momentum of street pressure after President Omar al-Bashir was ousted until the negotiation of a transitional power-sharing agreement in the summer of 2019. Indeed, Sudanese activists also understood the importance of creating inclusive dialogue spaces with a wide spectrum of incumbent elites conducive to consensual solutions, to prevent either their being sidelined by their political allies (as in Chile) or the emergence of a divisive and polarized environment benefiting the old establishment (as in post-Mubarak Egypt). However, Sudanese nonviolent activists have only passed the first threshold of a successful democratic transition. Our findings point to the

6

CONCLUSION

183

need to anticipate and prepare for the challenges ahead while seeking to institutionalize and consolidate democratic gains. Such long-term thinking will require building structures that can sustain rapid remobilization in times of crisis, both to defend democracy against ‘spoiler’ backlash by the military or remnants of the old regime, and to push for the full implementation of agreed reforms. As we have seen throughout our case studies, the consolidation of a healthy and accountable democracy requires the cultivation and transmission of nonviolent skills, methods and mindset across generations. Activists should also work on building durable institutions. However, our research did not find any solid evidence that nonviolent movements create effective and durable political parties. Hence, their efforts might be better focused on creating independent watchdog organizations (e.g. ombudsmen, commissions and other oversight bodies) in order to hold the new elites accountable, and to safeguard the democratic aspirations that drove the transition. Finally, our findings also offer relevant implications for outside actors such as foreign governments and international organizations, even though most NVR-led transitions are internally driven. As highlighted by the authors of the Diplomat’s Handbook for Democracy Development Support, ‘the role of outsiders is never primary, but their catalytic support can be pivotal’ (Kinsman and Bassuener 2013, p. 24). In Benin, the massbased mobilization which drove the transition to multiparty democracy received some external support from France and the Catholic Church. In Chile both the Pinochet regime and the opposition coalition benefited from extensive support by international backers and sponsors (Kinsman and Bassuener 2013, pp. 417–432). As our book confirms the evidence that NVR-induced transitions are positively correlated with stable and vibrant democracy, we can confidently reinforce the widely held view that citizens, democracy activists and human rights defenders engaged in peaceful efforts towards democratic empowerment are worthy of responsive support by international bodies. We have documented elsewhere the full range of options available to governments, aid agencies or solidarity networks to encourage, recognize, amplify and protect grassroots prodemocracy movements (Dudouet and Clark 2009; Dudouet 2015). However, we wish to add a note of caution for two reasons. First, overt foreign support for civil resistance groups can backfire. Repressive governments can use evidence of such support as a delegitimizing narrative portraying dissidents as ‘foreign agents’, a charge increasingly levelled against civil society organizations and institutionalized in law, thereby providing

184

D. LAMBACH ET AL.

a handy rationale for the repression of activists. Second, foreign support for nonviolent resistance movements comes at a price. Sponsoring governments, even those who can credibly claim to care about democracy promotion, will use their support to pursue foreign policy goals, such as toppling an inconvenient government. Hence, movements have to be cautious not to become overly reliant on external support to avoid being instrumentalized by their political allies abroad. Essentially, these concerns mirror debates about the Responsibility to Protect (R2P), whose critiques can also be levelled against the Responsibility to Assist and similar approaches. But this should not be taken to mean that external actors have no place in nonviolent resistance and democratization. The experiences of transition in Sudan underscore this point. There, the African Union was swift in condemning the military coup that sought to curtail the April 2019 people-power revolution, and suspended Sudan’s AU membership following the brutal crackdown on unarmed protesters in early June. In doing so, it not only emboldened the nonviolent movement to pursue its mobilization for a full transition of power, but it also safeguarded its own role as ‘the primary custodian of continental norms and frameworks’ (PSC Report 2019)—in line with Article 4 of the AU’s Constitutive Act which criminalizes unconstitutional changes of government. There are growing voices in academia and NGO circles demanding that the ‘Responsibility to Protect’ imperative, with its focus on early action to prevent genocides and mass atrocities, should be accompanied by a ‘Responsibility to Assist’ grassroots nonviolent movements rising in opposition to systematic abuses of power (e.g. Lagon and McCormick 2015; Ackerman and Merriman 2019). Here again, based on our findings, such assistance should encompass measures to encourage and empower nonviolent movements to: (1) forge tactical alliances with all prodemocracy forces while retaining their autonomy and mobilizing capacity; (2) learn and practice the skills of dialogue and negotiation both within opposition coalitions and across the conflict spectrum; (3) connect with, and learn from, other movements that have successfully navigated bottom-up political transitions in the past. The emerging Tunisian democracy has just passed its first peaceful turnover of power after a relatively peaceful, inclusive and consensual regime change, and its civil society leaders and Nobel Prize recipients have become consultants and trainers for pro-democracy activists worldwide.

6

CONCLUSION

185

Finally, external actors also ought to stay engaged well beyond the critical juncture of a nonviolent transition of power in order to accompany the consolidation of democracy, and early signs of long-term commitment can help to incentivize elites or security services that there are many benefits to serving under democratic governments (Ackerman and Merriman 2019, p. 16). These include pledges of future economic assistance or institutional support, as well as engagement to deploy monitoring teams to prevent violent retribution on all sides and to oversee the materialization of the ‘new rules of the game’. Such support can reinforce the three mechanisms of democratic consolidation highlighted in this volume, by promoting the survival of a levelled political playing field, a culture of civic awareness and sustained societal participation, and a containment of spoilers and veto players along the way.

References Acemoglu, D., & Robinson, J. A. (2019). The Narrow Corridor: States, Societies, and the Fate of Liberty. New York: Penguin. Ackerman, P., & Merriman, H. (2019). Preventing Mass Atrocities—From a Responsibility to Protect (RtoP) to a Right to Assist (RtoA) Campaigns of Civil Resistance. Special Report (Vol. 3). Washington, DC: International Center on Nonviolent Conflict. Azarya, V. (1988). Reordering State-Society Relations: Incorporation and Disengagement. In D. Rothchild & N. Chazan (Eds.), The Precarious Balance: State and Society in Africa (pp. 3–21). Boulder and London: Westview. Bermeo, N. (2003). Ordinary People in Extraordinary Times: The Citizenry and the Breakdown of Democracy. Princeton: Princeton University Press. Bisbee, J., Mosley, L., Pepinsky, T. B., & Rosendorff, B. P. (2019). Decompensating Domestically: The Political Economy of Anti-Globalism. Journal of European Public Policy. https://doi.org/10.1080/13501763.2019.1678662. Buckmaster, L. (2019, July 26). Sea Shepherd Captain Paul Watson: ‘I Call What We Do Aggressive Non-Violence’. The Guardian. della Porta, D. (2014). Mobilizing for Democracy: Comparing 1989 and 2011. Oxford: Oxford University Press. Dudouet, V. (2015). Sources, Functions and Dilemmas of External Assistance to Civil Resistance Movements. In K. Schock (Ed.), Comparative Perspectives on Civil Resistance (pp. 168–200). Minneapolis: University of Minnesota Press. Dudouet, V., & Clark, H. (2009). Nonviolent Civic Action in Support of Human Rights and Democracy. Directorate-General for External Policies of the Union, EXPO/B/DROI/2008/69. Brussels: European Parliament.

186

D. LAMBACH ET AL.

Foa, R. S., & Mounk, Y. (2016). The Danger of Deconsolidation: The Democratic Disconnect. Journal of Democracy, 27 (3), 5–17. Fukuyama, F. (1992). The End of History and the Last Man. New York: Free Press. Habermas, J. (1990). Verfassungspatriotismus – im Allgemeinen und im Besonderen. In J. Habermas (Ed.), Die nachholende Revolution. Kleine politische Schriften VII (pp. 147–174). Frankfurt: Suhrkamp. Heitmeyer, W. (2018). Autoritäre Versuchungen. Frankfurt: Suhrkamp. Inglehart, R. F. (2016). How Much Should We Worry? Journal of Democracy, 28(1), 18–23. Jones, D. M. (1998). Democratization, Civil Society, and Illiberal Middle Class Culture in Pacific Asia. Comparative Politics, 30(2), 147–169. https://doi. org/10.2307/422285. Kinsman, J., & Bassuener, K. (2013). A Diplomat’s Handbook for Democracy Development Support (3rd ed.). Waterloo: Centre for International Governance Innovation. Klaas, B. (2018). The Despot’s Accomplice: How the West Is Aiding and Abetting the Decline of Democracy. Oxford: Oxford University Press. Lagon, M., & McCormick, P. (2015). The Responsibility to Accompany: A Framework for Multilateral Support of Grassroots Nonviolent Resistance. Ethics & International Affairs. https://www.ethicsandinternationalaffairs. org/2015/the-responsibility-to-accompany-a-framework-for-multilateralsupport-of-grassroots-nonviolent-resistance/. Levitsky, S., & Ziblatt, D. (2018). How Democracies Die. New York: Broadway Books. Linz, J. J., & Stepan, A. (1996). Toward Consolidated Democracies. Journal of Democracy, 7 (2), 14–33. Lührmann, A., & Lindberg, S. I. (2019). A Third Wave of Autocratization Is Here: What Is New About It? Democratization, 26(7), 1095–1113. https:// doi.org/10.1080/13510347.2019.1582029. McAdam, D., Tarrow, S., & Tilly, C. (2001). Dynamics of Contention. New York: Cambridge University Press. Merkel, W. (2004). Embedded and Defective Democracies. Democratization, 11(5), 33–58. https://doi.org/10.1080/13510340412331304598. Migdal, J. S. (1988). Strong Societies and Weak States: State-Society Relations and State Capabilities in the Third World. Princeton: Princeton University Press. Mudde, C., & Rovira Kaltwasser, C. (2017). Populism: A Very Short Introduction. Oxford: Oxford University Press. Müller, J.-W. (2007). Constitutional Patriotism. Princeton: Princeton University Press. Müller, J.-W. (2017). What Is Populism? London: Penguin.

6

CONCLUSION

187

Piketty, T. (2014). Capital in the Twenty-First Century. Cambridge: Harvard University Press. Pinckney, J. (2018). When Civil Resistance Succeeds: Building Democracy After Popular Nonviolent Uprisings. Washington, DC: International Center on Nonviolent Conflict. PSC Report. (2019). Dangers Ahead for Sudan After Its Suspension from the AU. ISS PSC Insights. Pretoria: Institute for Security Studies. Scott, J. C. (1998). Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven: Yale University Press. Seeberg, M. B. (2014). State Capacity and the Paradox of Authoritarian Elections. Democratization, 21(7), 1265–1285. https://doi.org/10.1080/ 13510347.2014.960210. Walter, S. (2017). Globalization and the Demand-Side of Politics: How Globalization Shapes Labor Market Risk Perceptions and Policy Preferences. Political Science Research and Methods, 5(1), 55–80. https://doi.org/10.1017/psrm. 2015.64. ZEIT Online. (2019, June 16). Tote bei Zusammenstößen zwischen Polizei und Zivilisten. Zeit Online.

Index

A activism, 6, 8, 10, 30, 31, 33, 82, 86, 88, 89, 95, 99, 101, 103, 105, 106, 135, 136, 138, 146, 153, 157, 161, 164, 181–184 agency, 8, 11, 20, 80, 154, 157, 159, 180, 181 Alternative Citoyenne, 96 autocratic regimes, 2, 5, 9, 16, 17, 22, 24, 25, 30, 41–43, 45, 46, 64, 67, 83, 90, 97, 102, 106, 107, 112, 117, 143, 148, 156, 157, 164, 166, 182 electoral autocracies, 8, 182 military dictatorship, 67, 83, 100, 111, 116, 117 autogolpe, 4, 17, 25, 48, 104, 107, 166 Aylwin, Patricio, 69, 116, 117 B Bachelet, Michelle, 69, 99, 101 Benin, 1–4, 10, 11, 64–67, 70, 73, 75, 77–81, 84, 86, 92, 94–96,

102, 103, 111, 114, 115, 118, 122, 135, 136, 138, 141, 143, 144, 149, 151, 153, 154, 158, 160, 162, 163, 166, 176, 177, 180, 182, 183 renouveau démocratique, 2, 64, 66 National Conference, 1, 2, 66, 73, 75–77, 79–81, 88, 93, 94, 111–113, 136, 154, 158, 160, 166 Quota War, 75, 76, 111 Boni Yayi, Thomas, 2, 67, 95, 96, 177 Burkina Faso, 110, 111, 114 C Cape Verde, 10, 64, 86, 88, 89, 105, 122, 135, 138, 163, 177 causal mechanism, 7, 9–11, 16, 25, 27, 29, 30, 33, 60, 63, 64, 70, 75, 79, 86, 87, 89, 90, 102, 105, 107, 118, 122, 133, 135–139, 141, 142, 145, 146, 148–150, 153, 156, 160, 162, 163, 182, 185

© The Editor(s) (if applicable) and The Author(s), under exclusive licence to Springer Nature Switzerland AG 2020 D. Lambach et al., Nonviolent Resistance and Democratic Consolidation, https://doi.org/10.1007/978-3-030-39371-7

189

190

INDEX

Chile, 10, 11, 64, 67, 69, 70, 75, 81, 84–86, 92, 97–102, 111, 115–118, 122, 135, 136, 138, 143, 144, 149, 151, 153, 158, 177, 182, 183 transitional justice, 116, 117 citizens, 2, 4, 5, 8, 11, 15, 16, 20, 21, 24–27, 30, 32, 33, 46, 48, 68, 74, 77, 79, 84, 85, 90–92, 97, 99, 102, 105, 106, 116, 135–137, 140, 143, 144, 146–148, 150, 154, 155, 157, 158, 160, 163, 166, 178–181, 183 civil–military relations, 11, 15, 24, 26, 29, 33, 107, 108, 110, 111, 114, 115, 117, 118, 120, 121, 141, 155 civil liberties, 8, 17, 73, 93 civil society, 2, 4, 17, 25, 30–32, 57, 58, 60, 64, 68–70, 72, 74–76, 79–84, 86, 88–90, 94–96, 98–107, 110, 111, 114–116, 118, 122, 135, 137, 140, 143–146, 148, 149, 151, 153–155, 160–162, 165, 177, 180, 182–184 civil society organizations, 31, 57, 58, 72–76, 84, 85, 89, 96, 106, 116, 140, 154, 162, 183 civil war, 4, 5, 17, 29, 104, 158 collective memory, 90, 92, 94, 96, 141, 149, 162, 167, 176 Colorado Party, 86, 87, 107, 119, 138 comparative politics, 16, 83, 133, 145, 147, 150, 154, 157, 159, 165, 181, 182 transition research, 19, 28, 30, 147, 159 Concertación, 68, 69, 84, 98, 99, 101, 116, 118

contentious politics, 145, 146, 150, 153, 157, 164, 178 coup d’état, 1, 3, 22, 24, 46, 48, 65, 86, 97, 108, 109, 111, 115, 136, 137, 141, 147, 166, 184 critical junctures, 11, 15, 27, 106, 150, 166, 185

D defective democracies, 182 democratic backsliding, 8, 57, 140, 148, 155, 161, 163 democratic breakdown, 9, 16, 18, 19, 22, 28, 48, 50, 56, 58, 104, 107, 111, 118, 134, 148, 152, 159, 161, 165, 166, 175, 176 democratic consolidation, 7, 9–11, 15–20, 22, 23, 26, 28, 29, 41, 42, 45–47, 51–53, 56, 57, 59, 60, 63–65, 67, 69, 70, 78, 84, 97, 107, 117, 122, 123, 133–135, 142, 146, 147, 149, 155–157, 159, 165–167, 177, 178, 183, 185 democratic equilibrium, 15, 20, 21, 23–25, 29, 33, 107 democratic quality, 8–11, 16–18, 23, 33, 41, 42, 45–47, 53–59, 103, 134, 135, 152, 155, 182 democratic regression, 88 democratic survival, 9, 10, 16, 18, 41, 42, 45–48, 50, 56, 59, 107, 134, 141, 147, 152, 154, 155, 165 democratic transition, 4, 8, 9, 11, 15–17, 19, 20, 24, 28, 31, 32, 41–46, 48, 50, 51, 53, 59, 64, 65, 67, 75, 80, 81, 84, 88–90, 92, 103–105, 107, 109–111, 115, 117, 119, 133–136, 140–144, 147–153, 156–159, 161, 163, 165, 182, 184

INDEX

distributive conflict transitions, 148, 149, 157 drivers of transition, 11, 28, 44, 66, 101, 119, 149, 158, 160

E elections, 2–4, 8, 16, 19, 23–26, 29, 30, 42, 51, 53, 58, 60, 66–69, 72, 73, 78, 82–85, 87–89, 92, 94, 96, 98, 99, 103, 104, 116, 117, 119, 120, 136, 138, 146, 158–161, 166, 175–177 elite-centred approaches, 19, 136, 139, 157, 180 elite-led transition, 10, 11, 16, 20, 29, 30, 32, 33, 41, 64, 83, 86, 88, 99, 102, 106–108, 119, 120, 122, 134, 135, 138, 143, 147–150, 153, 157, 159, 160, 162 elites, 4, 8, 11, 15, 19, 20, 23, 25, 28–33, 44, 48, 55, 64, 67, 68, 70, 77, 82–86, 90, 95, 102, 110, 115, 116, 135–137, 139, 143, 144, 147–150, 153, 155–161, 163, 166, 178, 180–183, 185 elite struggles, 28, 87, 113, 156 El Salvador, 10, 44, 64, 89, 104, 105, 119, 122, 135, 138, 177

F freedom of association, 11, 72, 74, 79, 92, 135 freedom of expression, 11, 53, 58, 72, 79, 135, 161 Frente Farabundo Martí para la Liberación Nacional (FMLN), 104, 105, 119, 138

191

G government, 2–5, 7, 9–11, 15, 17, 20, 22–26, 28–33, 43, 48, 66–70, 72–77, 79–85, 88, 90, 92–95, 97–99, 101, 106, 107, 110, 116–118, 120, 121, 137, 141, 143, 144, 146, 148, 149, 151, 155–162, 177, 180, 182–185 Guinea-Bissau, 2–4 I India, 145 K Kérékou, Mathieu, 1, 2, 64–67, 76, 92, 95, 103, 111, 112, 114, 136, 160, 163, 166 L labour unions, 1, 65, 67, 76, 81, 82, 93, 95, 106, 112, 113, 151 levelling the political playing field, 10, 30, 69, 74, 85, 86, 122, 135, 138, 146, 151, 165 Liberia, 2, 4 M Mali, 114, 144 Marzo Paraguayo, 106, 119 mobilization, 4, 9, 22, 31, 46, 66, 81, 82, 85, 86, 88, 95, 100–103, 105, 107, 116, 123, 135, 140, 143–145, 147–149, 151–154, 156, 162, 163, 177, 180–184 remobilization, 10, 96, 97, 103, 138, 162, 183 modernization, 45, 65, 81, 165, 179 modes of transition, 4, 9, 15, 16, 27–29, 57, 59, 106, 108, 119, 138, 147, 149, 152, 160, 165

192

INDEX

movement duration, 152 Movimento para a Democracia (MPD), 88, 89, 105, 138

N Namibia, 10, 64, 89, 103, 104, 119–122, 135, 138, 153, 162, 163, 177 Namibia Defence Force (NDF), 120, 121 Nepal, 145 Non-governmental Organization, 95, 96, 98, 106, 181, 184 nonviolence, 5, 6, 113, 139, 178 movements, 5, 9, 30, 41, 44, 51, 53, 56–59, 65, 71, 74, 75, 82, 85, 86, 90, 100, 102, 107, 115, 118, 122, 123, 138, 142, 143, 145, 157–161, 166, 182–184 strategies, 5, 6, 9, 10, 92, 98, 139, 159, 178

O Operation Hornkranz, 121 opposition, 1, 3, 9, 15, 17, 20–24, 26, 28–31, 33, 66, 68–72, 75, 76, 82–85, 88, 103, 112–115, 137, 146, 153, 158, 159, 161, 177, 182–184

P Paraguay, 10, 64, 86, 87, 106, 107, 119, 122, 135, 138, 177 Partido Africano da Independência de Cabo Verde (PAICV), 88, 89, 105, 138, 163 path dependence, 11, 15, 16, 27, 59, 85, 102, 122, 133, 134, 140, 143, 148, 150, 154

people power, 91, 102, 147, 184 Philippines, 91, 145, 146 Pinochet, Augusto, 67–69, 82, 84–86, 97, 99, 100, 102, 115, 117, 149, 183 Poland, 31, 44, 57, 90, 140, 142, 143, 158 political attitudes, 2, 25, 30, 31, 33, 55, 90–92, 97–99, 106, 110, 164, 166 political culture, 10, 19–22, 25, 30, 31, 33, 82, 87, 90, 92, 97, 102–105, 107, 116, 122, 135, 138–140, 145, 146, 149, 150, 153, 154, 160, 164, 165, 167, 178, 180 political efficacy, 4, 31, 81, 90, 93, 102, 106, 161, 164, 166 political institutions, 2, 8, 10, 15, 17–20, 22, 27, 28, 30, 33, 57, 58, 74, 76, 78, 83, 84, 90, 93, 95, 106, 116, 117, 122, 135, 138–141, 148, 154, 159, 161, 162, 164–166, 179, 183 political parties, 1–3, 15, 19, 23, 26, 29, 30, 33, 42, 65, 67–69, 71, 72, 76, 80, 83–89, 92, 96, 98, 99, 102–107, 113, 116–122, 136, 138, 145, 146, 151, 160, 163, 165, 176, 178, 181, 183 populism, 68, 176, 180, 181 Portugal, 44, 140, 145, 160 post-conflict, 5, 151 praetorian problem, 11, 30, 32, 33, 107, 119, 122, 136, 141, 160

R radical flanks, 112 rationalist approaches, 147, 149 regime, 16, 18, 20, 21, 43, 45, 50, 52, 106, 165, 181

INDEX

relational approach, 15, 20, 26, 33, 137, 145, 147–149, 154, 155, 181 horizontal relations, 16, 21, 69, 70, 135, 137, 155 vertical relations, 16, 21, 25, 69, 70, 135, 137, 155 repression, 9, 23, 25, 26, 32, 67, 69, 72, 74, 78, 82, 85, 87, 93, 100–102, 116, 154, 156, 184 resistance movements, 5, 8, 11, 32, 43–46, 59, 89, 93, 103, 107, 111, 122, 140, 141, 143, 150, 151, 153, 156, 161–163, 166, 184 revolution, 2, 4, 5, 89–92, 96, 112, 119, 134–136, 139, 144, 146, 147, 150, 154, 156, 162, 165, 181, 182, 184 riots, 100

S security forces, 9, 11, 15, 21, 23, 24, 26, 32, 33, 48, 65, 77, 102, 107, 110, 111, 113, 115–121, 136, 137, 141, 155 army, 4, 32, 67, 76, 107, 110–113, 115, 120, 121, 144, 147, 177 police, 24, 32, 66, 99, 120, 177 social movements, 20, 67, 82, 102, 133, 135, 141, 145–147, 150, 154, 155, 157, 164, 182 Soglo, Nicéphore, 2, 67, 76, 78, 81, 111 South African Defence Force (SADF), 120 South West African Territorial Force (SWATF), 120

193

South West Africa People’s Organisation (SWAPOI), 89, 103, 104, 120, 121, 138, 154, 162, 163 spaces of resistance, 1, 4, 46, 66, 76, 139, 146, 166 Spain, 43, 140, 145 Special Field Force (SFF), 120, 121 state–society relations, 15, 19, 29, 181 strikes, 1, 66, 68, 79, 82, 96, 112, 146 student protest, 81, 99, 101, 102 Sudan, 119, 182, 184 T Talon, Patrice, 67, 96, 176, 177 third wave of democratization, 8, 28, 142 touche pas à ma constitution, 2, 95, 96, 136, 162 Tunisia, 57, 158, 162, 165 turnover of power, 2, 10, 16, 18, 19, 23, 41, 47, 50, 52, 56, 57, 59, 67, 78, 87, 134, 166, 184 U utopian enactment, 163 V veto players, 4, 29, 32, 69, 83, 89, 101, 111, 115, 118, 136, 139, 158, 165, 182, 185 violent resistance, 2–4, 9, 16, 22, 23, 27, 28, 30, 44, 48, 104, 105, 139, 144, 146, 160, 166 Z Zimbabwe, 104