Political Corruption in Eastern Europe: Politics After Communism 9781626374157

Why has political corruption emerged as a major obstacle to successful democratic consolidation in Eastern Europe? Explo

178 70 3MB

English Pages 303 [317] Year 2012

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Political Corruption in Eastern Europe: Politics After Communism
 9781626374157

Citation preview

Political Corruption in Eastern Europe

Political Corruption in

Eastern Europe Politics After Communism

Tatiana Kostadinova

b o u l d e r l o n d o n

Published in the United States of America in 2012 by Lynne Rienner Publishers, Inc. 1800 30th Street, Boulder, Colorado 80301 www.rienner.com and in the United Kingdom by Lynne Rienner Publishers, Inc. 3 Henrietta Street, Covent Garden, London WC2E 8LU © 2012 by Lynne Rienner Publishers, Inc. All rights reserved Library of Congress Cataloging-in-Publication Data Kostadinova, Tatiana. Political corruption in Eastern Europe : politics after communism / Tatiana Kostadinova. p. cm. Includes bibliographical references and index. ISBN 978-1-58826-811-2 (hc : alk. paper) 1. Political corruption—Europe, Eastern. 2. Post-communism—Europe, Eastern. 3. Europe, Eastern—Politics and government—1989– I. Title. JN96.A56C6545 2012 364.1'3230947—dc23 2011046271 British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library. Printed and bound in the United States of America The paper used in this publication meets the requirements of the American National Standard for Permanence of Paper for Printed Library Materials Z39.48-1992. 5 4 3 2 1

To Vladimir and Ivan

Contents

ix xi

List of Tables and Figures Acknowledgments

1

Introduction

2

Corruption in the Context of Postcommunist Transition

23

3

Identifying Specific Factors: A Quantitative Investigation

45

4

Party and Election Finance

65

5

Networks of Corruption: “Friendly Circles” in Bulgaria

95

1

6

State Capacity: Is It Really the Problem?

125

7

EU Accession: Comparing Balkan Experiences

155

8

Implications for Democracy: Corruption and Public Trust

187

Implications for Democracy: Corruption Perceptions and Voting

211

Conclusion

233

9 10

Appendixes A: Description of Regression Analysis Variables and Sources B: Eastern European Participation in International Anticorruption Conventions C: Supporting Documentation for Empirical Tests Bibliography Index About the Book vii

249 255 257 263 293 303

Tables and Figures

Tables 3.1 3.2 4.1 4.2 4.3 6.1

7.1 7.2 7.3

8.1 8.2 8.3

Regression Analysis: The Causes of Corruption, 1997–2008 Regression Analysis: The Effects of PHARE on Corruption Levels, 1996–2003 Party Finance Regimes Corrupt Contributions to Political Parties by Private Companies Correlations of Corrupt Party Funding and Related Regulations The Effects of State and Nonstate Capacity on Corruption in Postcommunist Eastern Europe, 1997–2004 Progress in EU Accession in the Balkans Time Map of Adoption of Anticorruption Legislation The Effects of the EU Accession Process and Economic Liberalization on Corruption in the Balkans Levels of Institutional Confidence and Corruption Perceptions The Effect of Corruption on Public Confidence in Various Institutions, 1995–1998 Comparing the Effect of Corruption on Public Confidence in Political Institutions to Support for Democracy, 1995–1998 ix

56 58 77 82 83

144 166 171

180 196 201

207

x

9.1 9.2 9.3 9.4

Tables and Figures

Electoral Choices of Voters Who Perceived Corruption as Very Widespread Logit Coefficients for Predictors of Voting and Political Efficacy Direct and Indirect Effects of Corruption on Voting Recent Levels of Corruption Perceptions, Trust in National Parliament and Political Parties, and Voter Turnout

219 225 226

228

Figures 2.1 3.1 6.1 6.2 6.3 6.4 7.1 8.1 8.2 9.1 9.2

Corruption Perception Levels in Fifteen Eastern European Countries, 2005 Mean Corruption Perception Scores for Countries by Degree of Integration with the EU Corruption and State Material Capacity Corruption and State Institutional Capacity Corruption and State Technical Capacity Corruption and Civil Society Levels of Corruption in Three Groups of Balkan Countries, 1998–2008 Corruption Perceptions and Confidence in Parliament and in the Police Corruption Effects in Traditional and Secular Cultures A Recursive Model of Voting Affected by Corruption Perceptions Estimated Path Coefficients

29 51 133 135 137 139 174 197 205 221 226

Acknowledgments

The idea for this book developed about five years ago. My interest in the phenomenon of corruption was a by-product of earlier research in which I analyzed Eastern European elections, political parties, and voters in the context of democratic transition. Insights from those studies led me to believe that our understanding of postcommunist politics would remain incomplete if we failed to explain the persistence and spread of corrupt behavior. Bribes, kickbacks, vote buying, and various other dishonest practices seemed to have stymied fair elections and the provision of good public services. Fifteen years after the start of democratic transformation, malfeasance was widespread, especially in some Eastern European countries. Thus, I started to investigate the different manifestations of corruption after communism and to evaluate its potential to undermine the project of democracy. In this journey, I was fortunate to receive the support and advice of friends and colleagues. In the course of my work, I had the opportunity to present earlier versions of parts of the book at professional conferences and seminar series, and I benefited from invaluable professional advice. I am grateful to the participants in the research colloquium of the Department of Political Science at Florida International University (FIU), who provided criticism during the initial phase. In particular, I thank my colleagues Nicol Rae, for his insights and encouragement, and Sarah Poggione, for her suggestions on methodology. As I was making progress on the manuscript, I presented work at annual meetings of the American Political Science Association, the Midwest Political Science Association, and the Association for Slavic, Eastxi

xii

Acknowledgments

European, and Eurasian Studies (the former American Association for the Advancement of Slavic Studies). At those conferences, I received comments from many fellow panelists. I would like to acknowledge the role played by panel discussants Hilary Appel, Anna GrzymalaBusse, and Bryon Moraski, whose challenges and recommendations were extremely helpful. I gave research talks at the Center of European Studies at the University of Florida, the Miami European Union Center of Excellence (FIU–University of Miami), and the Peace Studies Program of Florida Atlantic University. I am deeply grateful for the thoughtful commentaries of Jeffrey Kopstein, Amie Kreppel, Noemi Marin, Conor O’Dwyer, Liza Prugl, and Renat Shaykhutdinov. I am also deeply grateful to both Mary Stegmaier and Joshua Tucker for their careful reading and criticism of an earlier version of Chapter 9 and for suggesting ways in which I could shape my arguments more convincingly. Data collection and writing were supported by a Faculty Summer Research Award granted by the College of Arts and Sciences at FIU in 2008 and by Miami EU Center of Excellence Faculty Research Awards in 2005 and 2009. I wish to thank Petia Kostadinova for sharing her data on EU preaccession funding for Eastern European states. Aline Machado and Lavinia Bucsa, graduate students at FIU, gathered and documented data for two of the chapters. My fieldwork in Bulgaria was facilitated by colleagues from the Center for the Study of Democracy in Sofia, who provided materials and offered professional advice. At the final stage of preparation, the manuscript benefited greatly from the competence and valuable advice of the editorial staff of Lynne Rienner Publishers. I am grateful to the late Marilyn Grobschmidt for supporting the idea of this project, to Jessica Gribble and Karen Williams for their professionalism and understanding, and to the anonymous reviewers whose comments and suggestions improved the quality of the book. My assistant Craig Lang worked diligently on the proofs and the index. Final thanks go to my family and especially to my husband and my son. I am greatly indebted to Vladimir and Ivan for their unfailing support, encouragement, and inspiration. I dedicate this book to them.

Political Corruption in Eastern Europe

1 Introduction

The very nature of post-communism encourages corruption. —Leslie Holmes, “Corruption and the Crisis of the Post-Communist State”

In the first decade of the twenty-first century, corruption has been recognized as one of the most serious problems facing the democratic transitions of Eastern European countries. Scholars and practitioners alike warn that a rise of covert illegal transactions in the public sector may reduce substantially the flow of foreign investment and the rates of economic growth in the region. Even worse, corrupt activities may have a damaging effect on the level of public trust in the government and consequently erode the legitimacy of the newly established democratic institutions.1 Therefore, there is an urgent need to understand and explain corruption in Eastern Europe. If its main source is the legacy of communism, then states in the region should be more or less equally corrupt. Yet the abuse of public office has been a much more serious threat in some postcommunist countries than in others. In this book, I explore the causes and consequences of corruption after 1989. First, I identify factors that facilitate rent seeking and other activities that undermine the existing mechanisms for constraint. Then, I analyze patterns of political behavior generated by the spread of corruption in the nascent postcommunist democracies. A better understanding of the post-1989 mechanisms of inappropriate exchange of favors and the context that maintains them 1

2

Political Corruption in Eastern Europe

will be beneficial in several ways. A thorough explanation of corruption after communism will offer insights about its scope, intensity, dynamics, penetration, and vulnerability. Thus, achieving this goal will enable us to suggest when illegal provision of public goods and services may become difficult to hide and costly to perform during democratic transition. By identifying the causes of post-1989 corrupt practices, we will be able to search for possible policy solutions to effectively target the actual roots of malfeasance. Learning more about the ways in which the lack of elite integrity affects citizens and their relation to politics is also especially valuable. Democracy is more fragile in places with young institutions of representation, immature civil society, and inexperienced voters such as the ones in Eastern Europe. A systematic study of the effects of corrupt government practices after communism will help assess the extent of popular disappointment with elites and its potential to demobilize and demoralize citizens.

What Have We Learned About Corruption in Eastern Europe? Corruption is not a new phenomenon; neither is it a new topic in the study of politics. Ink has not been spared to describe public office abuse and its forms and transformations. Most scholars, practitioners, and journalists have condemned it and only a few have not worried about its possible harmful effects. The world’s interest in understanding the roots of wrongdoing in the public sector increased exponentially to reach the highest levels ever in the late 1990s to early 2000s. 2 This timing coincided with the radical economic restructuring in the former communist countries of Eastern Europe that had just overthrown the political tyranny of totalitarianism. For good or for bad, one more dimension was added to the already complex transitional agenda—that of dealing with corruption. Reforms had to continue against the background of increasing cynicism among the population, generated by investigations of scandalous violations of public office ethics. Learning more about what politicians were doing behind the scenes has been hurtful and necessary at the same time, painful to comprehend but needed for the purposes of prevention and cure.

Introduction

3

Extant literature on corruption in postcommunist countries includes studies classifying various forms of corrupt practice in Eastern Europe, research aimed at revealing the magnitude of the problem during transition, and applied work (reports and working papers) of nongovernmental organizations (NGOs) investigating abuse of power in particular countries. One notable representative of the first cluster of research is Rasma Karklins’s The System Made Me Do It (2005). Her typology of postcommunist corruption is a comprehensive classification of various manifestations of improper exchanges in the public arena. Karklins differentiates between everyday official-citizen interactions, within-institutional interactions, and impacts on high-level political institutions. Her detailed list of corruption types creates a good basis for observation and description of popular attitudes and dispositions, societal relations, and possible anticorruption measures. Earlier research, from roughly the 1990s until the early 2000s, examines the significance of postcommunist corruption by exploring worldwide data made available through Transparency International and the World Bank. Some comparative studies document a stark difference in corruption levels between Western consolidated democracies and the Eastern European transitions (Dugan and Lechtman 1998; Earle 2000). Other scholarship, including Leslie Holmes (1993, 1997) and Andras Sajo (1998), establishes a rise in corruption after communism and proposes its structural characteristics. Noteworthy for its empirical richness is a series of articles on the emergence of Mafia-style networks, especially in Russia, involved in grand corruption schemes (Glinkina 1998; Volkov 2002). A third group of studies focus on particular types of corruption; for example, Beverley Earle’s (2000) work on bribery and the role played by Western states and international organizations to engage the Eastern European countries in the reduction of illegal payoffs. Overall, the research effort in the 1990s to early 2000s deals primarily with questions about the origin and spread of corruption and links those to the ongoing process of democratic and economic transformation (Kneen 2000; Volkov 2002; Philp 2002). The central conclusion reached by most authors is that new opportunities for embezzlement opened when privatization and economic restructuring were launched while older practices managed to survive and adapt (Schmidt 2007). Rich in accounts of various corrupt deals and stories of crony-

4

Political Corruption in Eastern Europe

ism, this research lacked conceptual vigor, but provided a basis for a future, more systematic approach to the study of corruption in Eastern Europe. What remained particularly difficult to tackle was one challenge facing the entire field and not just the study of postcommunist politics, namely how to define and measure corruption. The mid- and late 2000s were marked by an expansion of the thematic scope and by a number of methodological achievements. By that time, it became obvious that the long lists of anticorruption measures proposed in previous research were not working as expected. Privatization, recognized as a main source of unethical practices, was completed, but corruption continued to plague Eastern Europe. In a context of growing criticism against the neoliberal approach to economic reform, questions and doubts arose about the effectiveness of political competition and public engagement as factors containing corruption (Kotkin and Sajo 2002; Reed 2002). In particular, scholars were intrigued by the counterproductive role of exposing corruption by elites for the sole reason to attack political rivals (Krastev 2004). Moreover, experts on Eastern Europe proposed that the anticorruption rhetoric might have had a counterproductive effect there by undermining the much more important question of democratic stability (Smilov 2007). Civil society, praised by some for its awareness-raising and mobilizing role (Karklins 2005; Michael 2005), was put under scrutiny by others who cautioned against exaggerating its anticorruption potential (Muddle 2003; Mingiu-Pippidi 2010). Experts even warned about a “capture of the NGO sector” by special interests (Todorakov 2010, 16). Finally, concerns were expressed that the broadly perceived positive role of the European Union (EU) would weaken once the Eastern European candidate states achieve accession (Vachudova 2009). Because of a mixed record of anticorruption success, some studies turned to past legacies, including habits of informal connections and distrust of the state, in a search for explanations of the resilience of corrupt practices (Orlova 2005; Holmes 2006, 183). Research on corruption throughout the 2000s shifted the agenda toward a more systematic empirical exploration of the causes and effects of corruption. Scholars continued the public opinion survey series that had begun in the late 1990s (Miller, Grødeland, and Koshechkina 2001; Miller, Koshechkina, and Grødeland 1997;

Introduction

5

Mishler and Rose 2005) and accumulated more data on Eastern European mass attitudes about the spread of corruption. New and important steps have been made to address one thorny issue in the study of malfeasance—the problem of measurement. To the perception-based indicators started by Transparency International and the World Bank, Freedom House’s Nations in Transit project added the Corruption index and applied it to Eastern Europe. Coalition 2000, a group of Bulgarian NGOs, also launched a monitoring program that publishes survey-based indices measuring susceptibility to and actual experience of corruption in Southeastern Europe. Studies of public officials’ integrity in the postcommunist world employed these indicators (Earle 2000; Anderson and Gray 2006). Other index-based measures were constructed to measure anticorruption program intensity and to help assess the success of various policy instruments (Steves and Rousso 2003; Dorhoi 2006). In brief, the studies of corruption in post-1989 Eastern Europe reveal that the problem is complex, widespread, and pertinacious. It is much more serious and systemic than corruption is in consolidated democracies. We have also learned that the multifaceted character of the transition, political, economic, and in some cases statebuilding, offered even more opportunities for improper behavior and enrichment. While comparisons of the region to other groups of countries are empirically sound, our present understanding of why some postcommunist countries are less corrupt than others remains deficient. In particular, research has not been conclusive on the role of political competition, civil society organizations, and European integration. To the extent that existing scholarship establishes correlations between corruption and particular factors, the possibility of simultaneous effects has not been scrutinized. Thus, we are still unable to identify with a reasonable level of confidence the determinants of corruption within the group of postcommunist transitions. Even wider is the lacuna in our knowledge about the damaging impact of corrupt practices. Extant research focuses primarily on business-government relations and the harm caused to firms by bribery (Anderson and Gray 2006), yet we know little about how corruption shapes distinct patterns of political behavior. I attempt to fill in this gap by searching for answers to the questions about the reasons for and the consequences of corruption in postcommunist politics.

6

Political Corruption in Eastern Europe

Defining Corruption As mentioned above, and also admitted in the literature on numerous occasions, defining the concept of corruption is one of the challenges that anyone interested in studying it would face. In fact, the debate on what is the best way to define corruption is ongoing and it is beyond the purposes of this study to resolve the issue (Heidenheimer 1970; Philp 1997, 2002; Johnston 2005b, 10–11). The problem is not in the absence of definitions to draw on; quite the opposite is true—extant literature is abundant in definitions. In particular, there are too many opinions about the scope of the concept and the level of abstraction at which one should define and operationalize it. The difficulty in producing a universally accepted definition of corruption stems among other things from culturally distinct conceptions, from law ambiguity about what action (or nonaction) is forbidden, and from disagreement about the inclusion of nonpublic actor interactions (Johnston 2005b, 11; Holmes 2006, 17–18).3 Yet a core set of characteristics is identified in the literature around which agreement about immoral behavior in public office can be built. On a theoretical level, scholars seem to agree that the general meaning of political corruption includes “the inappropriate use of common power and authority for purposes of individual or group gain at common expense” (Warren 2004, 332). The most common and brief expression that one can find, especially in recent empirical studies of corruption, is “the misuse” or “the abuse” of public power “for private gain” or “for private advantage” (Holmes 2006, 20). Other researchers and international organizations, such as the World Bank and Transparency International after 2000, also concur and adopt variants of this conceptual specification in their studies (Johnston 2005b; Lambsdorff 2007).4 In Offe’s (2004, 78) terms, defined this way the concept implies political corruption since it clearly involves the public realm (i.e., at least one side of the “voluntary and deliberate illicit deal” must be a public officeholder). The simple definitions used in empirical research have been disputed by theoretical work connecting corruption to democratic theory and related normative issues. Mark Warren (2004), for example, summarizes the deficiencies of the modern conceptions as being normatively static, excessively state-centric, integrity neutral, and discounting a society’s potential for agreement. A deeper conception of

Introduction

7

corruption through a democratic norm of “empowered inclusion,” Warren proposes, would better encompass the various manifestations of the phenomenon and its harmful effects on democracy. This further conceptualization of the improper use of public authority in various domains of public life would help identify what might count as corrupt and how one can observe it. Leslie Holmes (2006, 30), concerned more about the inability of a short definition to grasp gray areas of corruption, proposes a set of five criteria that might be of a better use. These emphasize the involvement of “actual or aspiring” public officials, an office with some form of decisionmaking authority, personal interest in the exchange, a secretive (“clandestine”) way in which the deal is struck, and broad popular understanding that all of this is corruption. On a more practical level, different countries recognize corruption in their legislation in a variety of ways, some of them quite detailed. For example, according to the articles of the Russian Criminal Code used until late 1996, corruption is “the misuse of power or of an official position, exceeding one’s powers or official positions, the acceptance of bribes, involvement [lit. mediation— posrednichestvo] in bribery, official forgery” (Holmes 2006, 21–22). Chapter 1, Article 5(1), of Estonia’s Anti-Corruption Act of 1995 defines a corrupt act as “the use of official position for self-serving purposes by an official who makes undue or unlawful decisions or performs such acts, or fails to make lawful decisions or perform such acts.” Provisions apply to a long list of officials, including members of parliament, the president of the republic, ministers, mayors, judges, and county governors (Corruption and Anti-Corruption Policy in Estonia 2002). The way that I define corruption in this book is mostly determined by the goal to explain the distinct spread of the phenomenon in the Eastern European new democracies. Thus, I adopt a simple version of the definition—the misuse of public office for private gain. Or, to borrow from Joseph LaPalombara (1994, 336), it is public servants’ behavior “in deviation” from their “formally defined public roles” aimed at benefiting themselves or associated others. Throughout the book, I refer to corruption as distortions in the relationships between private and public actors that have been made intentionally and with the purpose of private enrichment and advantage. Since I am interested in the roots of systemic corruption and its

8

Political Corruption in Eastern Europe

impact on citizen behavior, the content of my definition includes the entire scope of improper exchange of wealth and power. This is a holistic approach that concurs with the understanding that one form of corruption cannot be studied in isolation because it is related to malfeasance in other sectors as well (Lambsdorff 2007, 28). At the center of my inquiry is grand corruption5 (rather than petty corruption), which provides the opportunity to examine the reasons for and the impact of office abuse at the highest levels of public authority. My inquiry is based on a conception of overall levels of corruption rather than particular forms of it. In such terms, I am interested in the exploration of the phenomenon as “a syndrome,” as Michael Johnston (2005b, 11–12) puts it, which indicates problems in the performance of the entire political system. While in the course of the investigation I refer to specific manifestations of wrongdoing, such as bribing, illegal fund-raising, and customs rules violations, it is the corruptness of the political system that I compare and analyze across space and time.

Observing and Measuring Corruption Along with the difficulties encountered when constructing a sound conceptual definition of corruption (and, perhaps, because of them), observing and measuring corrupt activities remain an extremely challenging enterprise. There are multiple reasons that make these tasks so difficult to deal with. In first place is the very nature of the phenomenon itself, its secrecy that hinders direct observation and quantification. Breaching the rules of public-private interaction requires hiding and concealment in order to avoid possible exposure and sanction. Even in social environments with high levels of social tolerance of bribing, connectedness, and cronyism, a sense prevails that these things are not right, that they should not exist in an ideal situation. Therefore, all actors who engage in corrupt activities try to ensure some degree of secretiveness, to escape prosecution and secure a clean social image for themselves. Because corruption is a game played in the dark, a valid assessment of all aspects of its study, spread, scope, and magnitude is difficult to obtain (Lambsdorff 2007, 236–237). Creative solutions, designed to capture violations of entrusted public authority, have been advanced in a number of stud-

Introduction

9

ies.6 Yet all of them face criticism for not meeting a threshold of desired validity or for failing to enable meaningful cross-national analyses (Sik 2002; Lambsdorff 2007). Indeed, the search for quality techniques is ongoing. There are two basic ways to observe and measure corruption: objective and subjective. For obvious reasons, the former are more valuable, but also more difficult to apply. Research can benefit from their directness (i.e., reflection of actual occurrences of the phenomenon) and smaller potential for bias (i.e., minimized risk for contamination from sources). Examples of such indicators include convictions of public officials for abuse of public office (Glaeser and Saks 2006), the incidence of misuse of privileged positions by diplomats (Fisman and Miguel 2006), gaps between government investment costs and the value of created public goods (Golden and Picci 2005), and national reports to the United Nations Office on Drugs and Crime of bribery and embezzlement. The main weakness of these measures is that they all fail to account for the inefficiency of national institutions. Another shortcoming is that some of them are not applicable in cross-national research because comparable data on multiple countries have not been collected. The gathering of subjective data on corruption is performed through opinion polls. Surveys are conducted with representatives of the political elites, the bureaucracy, national and international financial institutions, businessmen, and ordinary citizens. The most popular among these are the Transparency International Corruption Perceptions Index (TICPI), the World Bank Control of Corruption Index (WBI), the Nations in Transit Corruption Index (NIT), and the Business Environment and Enterprise Performance Survey (BEEPS) of the World Bank. With the exception of the latter, these indicators provide composite indirect measures of corruption based on aggregated perceptions of experts and analysts. The method of data collection is a source of criticism directed toward these index-based measures of corruption. Thus, the TICPI and the WBI have been questioned on the grounds of how reliable the survey-gathered information is (Golden and Picci 2005).7 There are advantages, however, offered by the perception indexes that should not be ignored. Conducted annually in a large number of countries, the TICPI (since 1995) and the WBI (since 1996) have offered opportunities for detecting cross-national patterns in the

10

Political Corruption in Eastern Europe

spread and development of corruption (Christopher J. Anderson and Tverdova 2003; Gerring and Thacker 2004; Mauro 1995; Persson and Tabellini 2005; Sandholtz and Koetzle 2000; Treisman 2000). Improvements have also been in order. To address some of the problems first identified in the TICPI method, the authors of the WBI started to assign specific weights to participant surveys on the basis of their believed reliability (Kaufmann, Kraay, and Zoido-Lobaton 1999). What also raises the level of confidence in these indirect measures is that they are highly correlated with each other, as argued by Arvind K. Jain (2001) and others.8 Subjective data on corruption can also be gathered through public opinion polls. Similar to the TICPI and the WBI expert surveys, this type of information is perception based, but relies primarily on popular assessments of the spread of corrupt practices in native countries. The World Values Survey, special editions of the EU Eurobarometer, and the Comparative Study of Electoral Systems (CSES) are a few examples of polls that have included questions measuring public perceptions of how widespread or serious a problem corruption is. 9 Although reflecting individual impressions, rather than actual corruption, these data can be aggregated and used for system-level comparisons. They are especially valuable for the study of particular corruption-related research questions. Because one of the purposes of this study was to assess corruption levels in a large number of countries over a period of time, I needed a measure that allows comparisons across space and over time. As discussed above, such features are offered by the TICPI and the WBI, which rank countries on the basis of survey assessments made by experts, country analysts, and polls.10 In making this choice, I considered arguments made by scholars who praised the availability of these indicators (Kaufmann, Kraay, and Mastruzzi 2005; Lancaster and Montinola 1997; Montinola and Jackman 2002; Persson 2002) as well as those made by others who criticized them for inflexibility and even bias (Sik 2002). Since direct observations on corruption levels are difficult (if not impossible) to make for all the postcommunist countries in this sample, for all years, and by the same methodology, I relied on the TICPI and the WBI for many of the cross-national empirical tests. In the analysis of party finance, I added indicators from the BEEPS as more efficient measures of the extent to which political parties operate through corrupt contributions. In a few

Introduction

11

instances, I reran the regression tests with yet another measure, the Nations in Transit Corruption Index (NIT),11 to check the robustness of the results and to produce more stable conclusions. In all these analyses, the working assumption was for a strong correlation between actual corruption and the perception of it and I discuss the results in terms of perceived corruption. For the study of the effects of corruption on political behavior, I employed public opinion data from the third wave of the World Values Survey and the CSES Module 2. Two important reasons stand behind this choice of data sources. First, attitudes toward political institutions and voting decisions form at the microlevel in the minds of individual citizens who experience the positives and the negatives of the political process. Survey data can offer such information on people’s confidence in the government and their sense of political efficacy and voter mobilization. Second, in this case perceptions of corruption appear even more relevant than actual corruption (which, in fact, might be lower). What is more important is not how corrupt a country is in reality, but how widespread individual citizens feel it is. After all, it is individuals’ perceptions that politicians are very corrupt that would undermine (if at all) their trust in national institutions and desire to participate in public life.

Other Methodological Issues Undertaking an empirical study of corruption is not an easy task. Overall, there are at least three sets of difficulties that researchers may face in a comparative cross-national study of corrupt behavior. The first group of problems is associated with the already discussed methods of observation and measurement. In brief, results from the empirical tests may be sensitive to the particular indicator used in the analysis. To avoid possible contamination, in all chapters of the book I applied more than one measure of corruption and reestimated the models with them. The results are quite stable across different specifications of the indicators, something that secures robust findings and strengthens the conclusions at the end. Second, possible endogeneity, inherent to this type of research, sometimes makes it difficult to establish causal directions with certainty (Goel and Nelson 2005). For example, is it that liberalization of trade creates fewer opportuni-

12

Political Corruption in Eastern Europe

ties for bribery or that states with more intensive corrupt activity choose to have more closed trade policies? To cope with this problem, I introduced lagged independent variables to model effects whose direction might be dubious. Third, corruption is a multifaceted problem, quite complex to dissect and to comprehend. My approach to this complexity was twofold. I dealt with it theoretically by examining both internal (domestic) and external (international) factors affecting corruptiveness. The central arguments in each chapter are made from a rational choice perspective, yet I do not turn a blind eye to alternative explanations including possible sociological and cultural sources of malfeasance. Methodologically, the analyses were designed at the macro- and at the microlevel, allowing for cross-national comparisons and for inferences made from empirical tests on individuals’ behavior. Furthermore, I used a variety of quantitative and qualitative methods of data analysis. These include descriptive statistics, measures of association, panel data and logistic multivariate regression, path analysis, and a case study revealing the mechanisms of corrupt networks in a single country. To my knowledge, this is the first study that investigated the experience with corruption of such a large number of former Soviet bloc nations. Geographically, the scope of the study included fifteen postcommunist countries: Albania, Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Macedonia, Poland, Romania, Russia, Slovakia, Slovenia, and Ukraine. The time span in most of the analysis was between 1996 (the first comparable crossnational data on corruption) and 2008. When I explored the impact of the EU on reducing corruption, I included one more Balkan country, Serbia, which together with Albania formed a group of late accession states. I excluded Montenegro, Bosnia and Herzegovina, Belarus, and the former Soviet Caucasian and Central Asian republics. The former two had to resolve stateness problems during most of the time under study here. The rest of the omitted countries either have unresolved border conflict problems or have had mixed success in establishing democratic freedoms. In a final note related to methodology, I would like to emphasize that any inquiry on corruption, especially a large-N cross-national study, is deemed to face the challenge of data quality. I already discussed at length the problems related to observing corrupt activities.

Introduction

13

In many instances, accurate measures of the factors that cause corruption are also hard to find for all countries and for the entire period of time under investigation. I tried to complete this task to the maximum possible degree by gathering information from the same sources and through the same methodological techniques. For example, much of the remaining data came from reputable institutions such as the World Development Indicators of the World Bank, the Heritage Foundation, Freedom House, the European Commission (EC), and the US Agency for International Development (USAID). Especially challenging was the task of collecting information for the investigation of informal networks in Bulgaria. All sources that I managed to find for the analysis in this case study are publicly available: newspaper articles, interviews with politicians and businessmen, and materials published by investigative journalists. I refrained from using information that was impossible to double-check and to verify.12

What the Book Does and What It Does Not Do It is important at this point to clearly delineate what I explain in this book and how I do it as well as what I do not claim as its domain. As elaborated earlier, there are dozens of studies investigating corruption that were published in the past ten to twenty years. These are scholarly works, expert reports released by international organizations, and working papers produced by nongovernmental or civil society groups. In this difficult-to-navigate sea of literature, one can find theoretical and empirical pieces, case studies of single countries, and analyses of nations from all over the globe as well as work authored by economists, political scientists, sociologists, and even psychologists. Moreover, researchers have tackled different tasks: some studied the causes of malfeasance and others the consequences of corrupt behavior. A vast majority of these studies advance long lists of policy prescriptions on how to deal with the problem. So, what new does this study offer? What does it do and what does it not attempt to do? The book is about the reasons why some postcommunist societies are more prone to corruption than others and about the effects of public office abuse on mass political behavior. My purpose is not to show

14

Political Corruption in Eastern Europe

how and why the Eastern European countries are more or less corrupt than other regions in the world; neither is it to explain that corrupt practices also existed during communism. These themes have been explored quite extensively in the literature. We do know today, with a reasonable degree of certainty, that Eastern Europe is more corrupt than many Western consolidated democracies and that public office abuse was not absent there before 1989. Instead, I analyze the reasons why political life has been freer from corruption in some places (e.g., in Estonia and Slovenia) than elsewhere in the region (especially in Russia and Albania). I also examine the damaging impact of the perceived lack of elite integrity on common people as citizens. Compared to previous published work, I do not advance a particular typology of corruption nor do I rank order particular corrupt acts or their perpetrators. Karklins’s (2005) attempt in this direction already provides an exhaustive list of possible forms based on level of interaction. Another, more parsimonious, typology was offered by the World Bank, which unbundles the phenomenon between state capture and administrative corruption (Hellman, Jones, and Kaufmann 2000). I concur with Holmes (2006, 42–43) that building a more sophisticated typology of corruption is warranted. This work, however, is directed in different avenues of inquiry. I hope that the findings in the empirical chapters of the book will inform the future enterprise of classifying corruption. Another important way in which this study differs from previous research is methodological in nature, but also bears theoretical and practical implications. To fulfill my primary aim, to explain the causes and consequences of corruption after communism, I rely primarily on multivariate analysis of data from a large number of states. By accounting for competing arguments about what causes corruption and how it shapes citizen participation, I was able to isolate the role of specific political, economic, and social configurations perceived as influential in democratic accountability practices. This has not been done before, at least not in studies on postcommunist corruption. From the viewpoint of practitioners, the most valuable insights that such an approach can produce are policy related. However, readers are discouraged from waiting for yet another list of anticorruption policies in the conclusions of the study. The diversity in the socioeconomic and competitive structure of Eastern European politics is large enough to require carefully crafted instead

Introduction

15

of automatic policy responses to the complex phenomenon of corruption. The knowledge derived from the analyses in this book will hopefully direct politicians, activists, and citizens to the optimum answers for their societies. To summarize, this is a scholarly book in which I attempt to make three main contributions to the study of corruption and democratic transition. First, I search for theoretical arguments made in extant literature to explain political corruption in the context of postcommunist democratic and market reforms. In the study, I identify domestic and international factors relevant to Eastern Europe that shape the formation of politicians’ preferences, open opportunities for engagement in corrupt acts, and constrain incentives for illegal enrichment. Second, I advance tentative answers to the substantial question about how to approach corruption. The findings of my study are policy relevant, although I propose no particular anticorruption strategy. Third, by focusing on the experience of postcommunist transitions, I build an understanding of how institutionally unstable and economically weaker countries are especially vulnerable to corruption. Moreover, the case studies of Balkan countries help us understand the conditions under which anticorruption policies may be effective when applied to societies that not only are postcommunist, but carry an even longer tradition of distrust of the central government and its institutions. The adoption of a cross-national empirical approach helps to advance the research program on political corruption in the postcommunist world.13 There is an urgent need for a comprehensive explanation of processes that have the potential to put at risk democratic reforms and achievements in the region. In these terms, such a book has been long overdue. The book is also distinct in its subject and scope. Using a systematic approach in this study, I compared incidents of corrupt activities and public perceptions of them across a large number of countries. At the same time, I also investigated in depth concrete mechanisms of fraudulent practices undermining the legitimacy of democratic governance. Finally, gathering comparable information for several countries over time is a formidable task that I carried out with caution regarding data validity and sources. As described earlier, I used a unique mix of data sets combining observations on different levels of aggregation, system and individual.

16

Political Corruption in Eastern Europe

The Subsequent Chapters Along with the introduction and the conclusion, the book includes three main parts. In the first part (Chapters 2 and 3), I draw on existing literature to identify reasons for the spread of corruptiveness among postcommunist political elites. In the second part (Chapters 4 through 7), I analyze particular sources of corruption and, in the third part (Chapters 8 and 9), I investigate how politicians’ corrupt behavior affects mass democratic behavior in Eastern Europe. In the next two chapters, I explore the validity of some previously developed arguments about what makes some countries more prone to corruption than others. I do this from a theoretical perspective in Chapter 2, constructing a framework that incorporates factors relevant to the postcommunist environment. The working assumption here is that participants in possible corrupt activities are rational players who decide whether to engage in such interactions after calculating the benefits and the costs from these. As readers will see, this calculus may be affected by considerations for domestic and international opportunities and constraints. While the rationalist vision is at the heart of my explanation of corruption, I cannot ignore its most popular challenger, the culturalist approach. In the rest of the second chapter, I discuss whether culture may be an effective predictor of the different levels of corruption in the former communist countries and propose ways in which a culturalist perspective may benefit the analyses in this book. In Chapter 3, I meet theory with reality in a first set of empirical tests. I start by operationalizing the internal and external factors identified earlier. The former include the electoral method as enhancing transparency and accountability, the regulatory burden on business operations as an opportunity for overtaxation, and the national economy as a facilitator of anticorruption policy enforcement. The latter group of factors involves openness to foreign trade, which raises the costs of malfeasance, and integration with the European Union as a motivating force to reduce corruption. Based on the results from the regression tests, I propose directions in which the analysis should expand in order to develop a clearer picture of why corruption levels are different across Eastern Europe. In the four chapters that follow, I look at various sources and deterrents of corrupt behavior in postcommunist politics. Chapter 4

Introduction

17

goes beyond the formal methods of allocation of mandates and adds the rules adopted in various countries to regulate campaign finance. I start my investigation with a review of the party finance regimes that evolved after 1989 with the development of the young party systems. My investigation continues with an evaluation of data revealing the presence and negative impact of private companies’ contributions in violation of the law. As I show in the final part of that chapter, some aspects of the finance regulation are correlated with the levels of corruption, a result that can have implications for the debate around the merit of separate policy instruments. What that analysis does not reveal is what the mechanisms of corrupt interactions between politicians and private interests are and the potentially harmful consequences of these connections. But in Chapter 5, I do exactly this by constructing and running a case study of the friendly circles that emerged around political parties in Bulgaria after 1989. By analyzing in depth the genesis and persistence of these powerful groupings, I seek to unveil their structure and forms of public-private interdependence. In Chapter 6 I expand the investigation by exploring the capacity, state and nonstate, of countries to tackle corruption. That chapter comes to address the issue of a lack of meaningful results even when political will seems to exist. One obvious answer, though still not systematically analyzed in the postcommunist context, is the capability of the political system to stand against and contain corruption. I argue that both a materially and institutionally stronger state and an enabling social environment are needed for anticorruption efforts to succeed. Interestingly, better-paid public officials and sustainable civil organizations emerge as key contributors to state capacity for fighting malfeasance. These answers are provocative, given the popular negative view in Eastern Europe about increasing bureaucrats’ salaries and the weak civil societies inherited from communism. My aim in Chapter 7 is to examine the role of international pressure on reducing corruption in the postcommunist states. I focus on the influence of the European Union, which has been different across space and over time due to variations in the progress of the EU accession process. The geographic scope of my study, involving many postcommunist countries, and the long range of years under investigation provide an opportunity to track down whether the EU has been effective in encouraging decisive measures against corruption. To

18

Political Corruption in Eastern Europe

account for possible interference of cultural diversity, I selected countries from the same subregion and with the same historical roots from the Balkans. As the results show, the EU has been able to create incentives for more active anticorruption engagement, but success has been slow and to some extent dependent on the stage of integration. Finally, the findings that I report in this chapter speak about the contribution of European integration to cleaner public affairs management as separate and independent from the process of liberalization and market reforms. In Chapters 8 and 9, I analyze the impact of corruption on how the Eastern Europeans feel about the government institutions, democracy, and voting. I thereby seek to develop a comprehensive picture of the less visible impacts of corrupt activities that, however, can be of crucial importance for preserving the legitimacy of the democratic order. The two chapters argue that perceptions of widespread corruption damage public confidence in institutions and affect voter decisions. As I show in Chapter 8, the harmful effect of corruption is much stronger in the case of institutional trust than in that of democracy as an ideal. Assessing the balance between abstention in elections and voting the corrupt rulers out, I also conclude Chapter 9 on a less than expected pessimistic note. Although the corrosive effects of corruption are confirmed in general, there are some signs that democracy in the region is not so fragile. From the beginning of this inquiry, I embraced a rational choice approach. And at the end, readers deserve to be given a fair assessment of what has been accomplished by using such a perspective. In Chapter 10, I summarize the findings and discuss them in the light of central debates in the literature about the institutional and cultural sources of corruption. My effort throughout the book to keep the analyses sensitive to alternative explanations pays off and, in that concluding chapter, I elaborate on its merit for the validity of the findings. The conclusions advanced in the final chapter assign credit to a set of institutional features allowing, preventing, or reducing corruption. Drawing on the analyses that expand beyond the effect of electoral systems, I highlight the relevance of particular party finance mechanisms, informal private-public networks, and administrative reform in Eastern Europe. My concluding discussion also focuses on the enabling impacts of the domestic and the external environment represented by civil society organiza-

Introduction

19

tions and the European Union. As it became clear toward the end of my project, institutions do play a vital role in shaping incentives for and against corruption in particular contexts. It is the magnitude of their effects that may vary by cultural distinctiveness, to the extent to which Eastern European societies differ with regard to their precommunist past.

Notes 1. Some earlier studies propose that the improper exchange of favors in the public sector may have a positive impact on development, mainly by improving bureaucratic efficiency (Leff 1964; Huntington 1968; Goldsmith 1999). Yet a consensus seems to be emerging that corruption has a predominantly negative effect, especially in unstable political environments. For example, empirical research suggests that corruption slows economic growth (Mauro 1995); impedes the inflow of investment (Knack and Keefer 1995; Tanzi and Davoodi 1997); increases social inequality (Rothstein and Uslaner 2005); favors the expansion of the unofficial economy (Johnson, Kaufmann, and Zoido-Lobaton 1999); depresses citizens’ sense of well-being (Tavits 2008); and threatens democratic legitimacy (Moran 2001). Even Huntington, who assumes a positive association of corruption with modernization, recognizes that societies where corruption is already pervasive may not benefit from it as a “lubricant easing the path to modernization” (1968, 69). As is argued elsewhere, the communist totalitarian systems from which the new Eastern European countries started to depart in 1989 were not free from graft and related organized crime (Simis 1982; Moran 2001; Holmes 2006). These arguments reinforce an expectation that corruption in those countries is more likely to have unfavorable consequences, rather than to increase efficiency or promote entrepreneurship. 2. For more detail on the chronology of subsequent waves of corruption and anticorruption scholarship and public debate in general, see Schmidt (2007). 3. Offe (2004, 77) offers a list of practices, broadly perceived as wrongdoings that are “neighboring” the domain of political corruption: fraud, embezzlement, theft, nepotism, cronyism, gifts, donations, lobbying, and others. Although associated with corruption, he argues, it remains unclear which of them would fall within the definitional boundaries determining the scope of corrupt phenomena. 4. Johnston’s (2005b, 12) exact definition is “the abuse of public roles or resources for private benefit.” He, however, cautions that the meaning and the clarity of these terms may vary across societies. 5. I employ the term grand here to distinguish corruption at the highest levels of political authority (executive, legislative, and the bureaucracy) from

20

Political Corruption in Eastern Europe

improper payments made during citizens’ everyday interactions with lowlevel bureaucrats and administrators. For a different use of the term grand corruption, see Jain (2001, 74) who classifies corruption in democratic societies as grand (of political elites), bureaucratic (of bureaucrats), and legislative (of parliamentarians). 6. For extensive reviews of existing measures of corruption, see RoseAckerman (1999), Jain (2001), and Lambsdorff (2006). 7. Galtung (2006, 109–123), for example, summarizes the “failings” of the TICPI in problem groups including country coverage and imprecise sources and definitions. 8. For a comprehensive review of the use of composite indexes as measures of corruption, see Knack (2007) and Treisman (2007). 9. Results from the World Values Survey are available at www.worldvaluessurvey.org, from the special Eurobarometers at http://ec .europa.eu/public_opinion/index_en.htm, and from the CSES at www .cses.org. 10. Using both TICPI and WBI data will ensure more robust results and confidence in the findings. The TICPI measures corruption in the public sector, using the opinions of resident and nonresident country experts and business leaders. Thus, it is “a survey of surveys” that ranges from 0 (most corrupt) to 10 (least corrupt). It was first constructed by Johann Lambsdorff (1999) to assist Transparency International in pursuing its objectives as a global coalition against corruption (for more details on the methodology behind the TICPI, see www.transparency.org/cpi). The WBI is also based on surveys, of individuals familiar with government operations, country analysts from international agencies such as the European Bank for Reconstruction and Development, expert assessments of the US State Department and the French Finance Ministry, nongovernmental organizations, and commercial business information providers such as the Economist Intelligence Unit (Kaufmann, Kraay, and Mastruzzi 2010, 6–7). The WBI uses year-to-year comparable sources and adjusts for in-unit assessment differences; it ranges from –2.5 (most corrupt) to 2.5 (least corrupt). WBI yearly assigned scores by country are available at http://info.worldbank.org/governance/wgi /index.asp. The Transparency International and the World Bank indexes are in agreement—they vary together well, as shown by the Pearson correlation parameter of .920 (significant at the 0.000 level) estimated for the data used in my analyses. 11. This index uses information on public opinion, officials’ business interests, and conflict of interest and financial disclosure legislation (Nations in Transit 2009). The NIT ranges on a scale from 1 (least corrupt) to 7 (most corrupt). This index relies on a rougher measurement and provides a shorter time series (starting in 2000) than the TICPI and the WBI. 12. In the perfect case scenario, information produced by participants and by direct observers would be most useful. However, it is difficult to find primary sources for research on elusive structures such as friendly circles in which politicians and private entrepreneurs engage. The participants not only prefer that these activities are hidden, but they do everything possible to keep

Introduction

21

such connections confidential. One rare exception is the public admissions of Ahmed Dogan, leader of a political party in Bulgaria, which offer direct evidence for the existence and operation of the circles. During my fieldwork in Bulgaria in the summer of 2009, I spoke with investigative journalists and NGO experts. They referred me to various publications, including their own, but were not willing to provide details—not surprising given the rise of organized crime and the recent instances of journalists being beaten. Thus, the sources used in Chapter 5 are mainly secondary (i.e., books and newspaper articles on the topic). 13. Dorhoi’s (2006) dissertation is also empirical and involves a large number of Eastern European transitional countries, but it is oriented more toward adopted anticorruption policies and the factors accounting for their intensity.

2 Corruption in the Context of Postcommunist Transition

Corruption . . . is a direct product of the rise of new groups with new resources and the efforts of these groups to make themselves effective within the political sphere. —Samuel P. Huntington, Political Order in Changing Societies

Corruption has become a salient issue in Eastern Europe that is discussed by journalists, condemned by the public, criticized by representatives of international lending institutions, and exploited by politicians to gain advantage over rivals in electoral campaigns. Illegal exchange networks and bribery, however, are not novel in the region. Considerable amounts of embezzlement were reported for the late 1980s when the communist authorities failed to enforce tight control in protection of the public interests (Millar 1985; Holmes 1997, 2003). Then, with the democratic transition unfolding, new representative institutions started to emerge in support of citizens’ rights and property ownership. The pursuit of more responsive and accountable policies has always been high on the transitional agenda, yet perceptions that corruption has increased prevail in Eastern Europe (Holmes 2003). These mass beliefs are accompanied by rising levels of disappointment with national institutions and the fairness of the electoral process. Incumbents have often been punished by voters for the failure to provide for a better life, and populist parties and leaders have won unexpectedly high levels of electoral support by promising more fair and moral politics. 23

24

Political Corruption in Eastern Europe

My goal in this chapter is to identify explanations for the corrupt behavior of public officeholders in Eastern Europe. As preliminary data show, the postcommunist countries as a group are more corrupt than their Western European counterparts and affected differently by corruption on an individual basis (Earle 2000; Transparency International Corruption Perceptions Index, various years). Admittedly, the factors driving this variation can be region specific and their impact may be conditioned by the uniqueness of the ancient regime and the postcommunist context.1 While the arguments in this chapter account for the distinctiveness of the region as a whole, the framework that I propose focuses primarily on the extent to which different countries have been plagued with corruption. In the following two sections, I identify specific attributes of the pre-1989 regime and the transitional period by drawing on earlier research. I also elaborate on some forms of corruption that flourished during these times. Then, I advance two groups of explanations, domestic and international, that unravel why some national elites in Eastern Europe have been less honest and less successful in fighting corruption. In the next section, I take on the issue of culture as an alternative explanation, which is often blamed for the overall high degree of corruptiveness of Eastern Europe. I conclude the chapter with a summary of the theoretical framework and some alternative explanations of malfeasance.

Corruption Before 1989 Extant research emphasizes the role of preceding regimes and colonial heritage as important contextual factors in shaping a nation’s ability to cope with corruption (Gerring and Thacker 2004). Previous studies also argue that transitional environments create more opportunities for corrupt acts due to a nascent, fragile political system that cannot cut ties to the past and incorporate new groups quickly enough (Huntington 1968, 61–65; Holmes 1997). The postcommunist transitions are no exception. They carry the legacy of communism with its lasting effects on societal relations marked by irresponsible attitudes toward public resources, antistate sentiments, and suspicion toward any form of cooperative action and volunteering for the common good. It not only is that certain totalitarian legacies created a fertile soil for corruption to flourish in the aftermath of the 1989 events, but the communist regimes themselves were corrupt.

Corruption in the Context of Postcommunist Transition

25

At the high levels of political power, patronage and clientelism existed as frequent practices before 1989 and survived thereafter. Relationships between members of the former communist nomenklatura were built on loyalty to party bosses in return for appointments to prestigious posts in the bureaucracy, promotions, and job security. Dependent in its work on the Communist Party, the legal system failed to expose and punish the racketeers and the extortionists, thus becoming by itself “a catalyst for corruption” (Simis 1982, 74). As socialism matured, these high-level bureaucrats, factory managers, and members of the apparatus turned extremely pragmatic and less ideological. Many of them, as Leslie Holmes (1997) observes, “lost faith” in communism and actively sought ways to siphon state assets and to position themselves so that they could transform into successful capitalists once a new system was set up. Corrupt exchanges between ordinary people and public officials, policemen, physicians, and even vendors were an everyday part of coping with life. The prevalent shortages of consumer goods, housing, automobiles, and food favored the “plunder of socialist property,” the mass use of bribery, and illegal trade (Kramer 1977, 215). Helping themselves to whatever other people had access to was not considered immoral; it was just a way to beat a system that had not succeeded in providing basic necessities. This phenomenon is often attributed in the literature to the so-called institutional blurring characteristic of communist states (Holmes 2006, 186). The lack of a clear boundary between public and private, and between state and party, had created confusion about what would count as stealing from society’s property and what would be abuse of public office. The major reasons that allowed for the corruptiveness of these highly centralized political systems were institutional in nature: individual initiative and responsibility were strongly discouraged; competition was suppressed; patronage thrived in the absence of transparency; and civil society as a potential anticorruption force was subdued (Holmes 2006, 183–187). To use Konstantin M. Simis’s terms (1982, 31), corruption became “a fact of life” in the Soviet Union. Members of the ruling elite, so brutal in persecuting other offenses, closed their eyes before corruption because they also participated in it at all levels. Additionally, there was a fear that exposing malfeasance would reveal just how dysfunctional the propagandapraised socialist order was.

26

Political Corruption in Eastern Europe

Despite the spread of corrupt exchange during communism, however, there were limits to the amount of irregular profits that a public official could realize (Glinkina 1998). The system was structured in a hierarchical order where corrupt officeholders could misbehave only within the domain of the assets under their control. Much greater opportunities for personal enrichment opened once economic reforms, especially privatization, were launched in a context of legal uncertainty and absent property rights protection. Consequently, corruption grew and obtaining privileged access to resources became more complicated than before.

Corruption and Transitioning from Communism The Eastern European transition project is broader in scope than previous democratic transformations, including those in Southern Europe and Latin America. Simultaneous political, economic, and social reforms take place in the absence of structures, knowledge, and experience to maintain and solidify the efforts (Philp 2002). The complex transformations, along with disputes over borders and land and interethnic conflict, made the postcommunist systems vulnerable to the spread of organized crime and improper enrichment (Holmes 1997; Kaldor and Vejvoda 1997; Xin and Rudel 2004). Deliberate delays in the adoption of relevant legislation and vagueness in the adopted administrative procedures allowed for malfeasance to flourish. This made it difficult to detect and punish corruption (Glinkina 1998). A legacy of “informality” inherited from the communist era enhanced the formation of exchange networks operating through contacts and privileges (Grodeland 2007).2 In this volatile environment, too many public officials preferred to stay loyal to “their colleagues and agency than to the state and society more generally” (Holmes 2006, 182). Because of the multifaceted character of postcommunist transition, numerous opportunities emerged for illicit payments, patronage, allocation of public contracts, black market interactions, and covert networks. These could spread and grow in the Eastern European societies, already suffering from endemic bribery and lack of elite integrity. In the late 1980s to early 1990s, state assets became the main direct source of enrichment (compared to the constrained barter-type exchanges during the communist era). Among various forms of such “transfers,” Glinkina (1998) observes for Russia

Corruption in the Context of Postcommunist Transition

27

state credits granted at low interest rates, illegal concessions to “friends,” improper renting of tax deductions, government subsidies for import goods, and privileged assignment of export quotas. In the absence of a national bourgeoisie privatization at an unprecedented scale was carried out, with members of the previous nomenklatura finding themselves in the advantageous position of transferring (and gaining) property rights through network connections (Holmes 1997). As experts on Latin America warn, marketization creates new opportunities for enrichment and corruption may increase in the absence of a system of sound rules governing the process of reforms (Manzetti and Blake 1996; Colazingari and RoseAckerman 1998).3 “Nomenklatura capitalism” in countries such as Russia suffocated every effort on the part of the more capable and accountable to seek public office (Kneen 2000). The practice of selling parliamentary votes in service of private interests is especially prevalent in places with poorly institutionalized party systems and weaker connections between constituents and their representatives. While in the past there was only one (wellfunded) party, multipartyism after communism functions through a fierce competition for money. Leaders of different parties are ready to violate existing finance regulations in desperate need of resources to compete for power. With the fall of the Berlin Wall and opening of the borders, the increased traffic of people and goods created yet new opportunities for abuse, and customs-related corruption started pouring money into party coffers. In the 1990s, the weakening of the traditional role of states worldwide has made it even more difficult for postcommunist governments to identify, recognize, and fight corruption. Along with the well-known ones, various new forms of extortion, bribery, and other illegal acts have emerged in Eastern Europe after 1989 (Coulloudon 2002). These include bribing of public officials during privatization; anonymous donations to parties running in elections in return for adoption of bills favoring special interests; nontransparent allocation of public contracts by members of parliament to friendly firms; courting of group interests through unfair and disproportional allocation of foreign aid, EU assistance funds, or state subsidies; incumbents’ transfer of public resources to support their own election campaigns; and bribing of parliamentary committee members who decide on broadcast frequency distribution. Due to regulatory loopholes and increased bureaucratic discretion, firms from

28

Political Corruption in Eastern Europe

the shadow economy bribe officials to avoid taxation (Eilat and Zinnes 2002). Entrenched patronage structures from the communist era provide fertile ground for exchange of votes against discriminatory distribution of individual benefits (Birch 1997). It is not surprising then that recent reports released by international organizations such as the World Bank, the European Commission of the European Union, and Transparency International emphasize that postcommunist societies remain significantly more corrupt than Western industrialized nations (Anticorruption in Transition 2000, xiv; Monitoring the EU Accession Process 2002, 16–19). One way to illustrate the relatively high levels of corruption is through comparison with other countries in Europe. The rankings produced by Transparency International show that, since 1995, the young Eastern European democracies have scored significantly lower in terms of transparent and clean policies than their Western counterparts. Data over time, however, reveal that by the early 2000s, a few of the best-performing economies in Eastern Europe also undertook successful steps in combating corruption. At present, Slovenia and Estonia are ranked by TICPI as less corrupt than some of the older EU member states such as Greece and Italy (Transparency International Corruption Perception Index, various years). Given the political, economic, and social similarities shared by the Eastern European states, especially the legacies of communism, one may be surprised by the unequal spread and varying extent of corruptness in the region. Figure 2.1 depicts 2005 levels of corruption-free politics in fifteen postcommunist countries. The histogram in the figure shows that Albania and Ukraine are much more corrupt than Slovenia and Estonia, Russia than the Czech Republic, and Macedonia than Hungary. Considering this variation, I next seek to identify factors that may explain the different intensity of illegal exchanges aimed at personal enrichment or reelection.

What Determines Corruption in Post-1989 Eastern Europe? Theoretically, a sound understanding of corruption can be developed only when one takes into account the factors that shape opportunities for corrupt behavior and constraints on office abuse. Which of these are factors in Eastern Europe? At this point in my inquiry, I start with

Corruption in the Context of Postcommunist Transition

29

Figure 2.1 Corruption Perception Levels in Fifteen Eastern European Countries, 2005

Source: Transparency International Corruption Perceptions Index, 2005, www.transparency.org. Note: Higher value = less corruption.

an exploration of the arguments advanced in previous studies on corruption. In doing so, I am able to identify what determinants of corruption are also relevant in the Eastern European context. I also seek insights about other possible impacts, unique for the region and for postcommunist politics. I proceed with a theoretical exploration of two groups of factors, domestic and international, that may have affected incentives for improper behavior or increased the costs of engaging in such activities. As previous studies claim, the collapse of the former totalitarian systems and the success of the transition itself have depended on both internal and external factors (Linz and Stepan 1996; Kaldor and Vejvoda 1997).4 Domestic Factors Electoral rules. A large body of research associates domestic institutions —electoral systems, power distribution among political players, cabi-

30

Political Corruption in Eastern Europe

net stability, and budgetary structures—with national fiscal performance (Persson and Tabellini 1999; Gerring and Thacker 2001; Annett 2002; Milesi-Ferretti et al. 2002). These institutional factors have been found to affect governments’ spending patterns and policy effectiveness through the decisionmaking process; they shape the structure of competition over the allocation of scarce resources. The relationship between institutions and corruption has been theorized mainly through the principal-agent model, where public officials abuse their power for personal gain, in violation of established norms and rules of conduct (Rose-Ackerman 1999; Persson, Tabellini, and Trebbi 2001; Kunicova and Rose-Ackerman 2005). Since officials receive their mandates in popular elections, the rules that govern the process of recruitment generate incentives to either violate or comply with the law. Elected officials are interested in reelection, which they would not achieve if exposed “for enriching themselves at public expense” (Geddes and Neto 1992, 657). Furthermore, to raise campaign money and resources, politicians would be tempted to fund their race by improper means. Scholars have proposed that institutions affect the likelihood of this happening and a system’s propensity to detect wrongdoing (Walecki 2005). Electoral rules differ by the way they shape the voter-representative link and, consequently, by the degree to which accountability can be achieved. Plurality single-member district (SMD) systems favor a direct link between members of parliament (MPs) and their constituents, where the latter hold the former individually (hence, directly) responsible for what they do or fail to do. Proportional representation (PR) closed list systems are associated with exactly the opposite, an indirect link of representation and voter inability to follow exactly who is responsible and to sanction the guilty. Thus, existing research concurs, corrupt activities may become more costly to perform in systems with higher possibility for malfeasance to be detected and punished (Kunicova and Rose-Ackerman 2005). The above argument is straightforward and easy to apply to contexts with political systems in equilibrium. However, transitional countries are characterized by weakly institutionalized parties, new political actors entering the race for power, and voters with changing voting preferences. Can we then expect that the electoral rules will be a relevant determinant of corruption in the uncertain environment of postcommunist transition? A positive answer to this question requires that we take into account the transitional dynamics in the region and

Corruption in the Context of Postcommunist Transition

31

those elements of the electoral systems that might matter. Research has already established that the election formulas adopted in Eastern Europe have affected the size of the young party systems, the fate of separate parties, the strategic behavior of party leadership in nominating candidates for MPs, and voter behavior (see, e.g., Gabel 1995; Moraski and Lowenberg 1999; Duch and Palmer 2002; Kostadinova 2002). It is plausible to assume that, by the late 1990s, the election rules had also started to shape incentives of engagement in or avoidance of exchanges breaching the public trust. The main mechanisms through which electoral rules may affect motivations for corrupt acts are arrangements related to (1) the way seats are allocated; and (2) the type of representation (i.e., the link between voters and their representatives). Scholars have agreed that, in plurality SMD elections, the personal vote is what helps the development of a direct connection: it is clear who receives the mandate and becomes responsible for the adopted policies. And as Arend Lijphart (1999, 289) proposes, “the greater clarity of responsibility in majoritarian elections may hinder corruption.”5 A careerconcerned representative would pay attention to what he or she does in order not to be detected and voted out (Myerson 1993; Persson, Tabellini, and Trebbi 2001). In party list elections, on the other hand, MPs are less concerned with how voters think of them individually. Individual representatives have weaker incentives to behave “appropriately” because group representation dilutes personal responsibility (i.e., accountability is collective and sanctioning individual MPs is less likely). Therefore, party list systems are expected to cause more corrupt practices than rules based on personal voting. More controversial conclusions have been reached on the role of district magnitude. Roger B. Myerson (1993) and others (Persson, Tabellini, and Trebbi 2001) argued that the use of multimember districts leads to less corruption due to a “barriers-to-entry” effect. The rationale is that it is easier for voters to remove from office parties that have become known as corrupt when the threshold for representation is low. Also, candidates from new parties have a better chance of getting elected in more permissive systems. Moreover, in SMD elections, voters may decide to vote strategically (prevent the least preferred candidate from winning) and yet cast a ballot for a corrupt politician. Another school of thought, however, questions the above arguments. This theory emphasizes that monitoring of illicit rent

32

Political Corruption in Eastern Europe

seeking is easier to carry out in single-member than in multimember district systems where corrupt activities can be detected and exposed more often (Kunicova and Rose-Ackerman 2005). While there is an obvious disagreement between these two arguments, they tell two stories that are connected. One relates to monitoring and revealing illegal enrichment, and the other to the ability to punish incumbents after they are found to be corrupt. Thus, in a multimember system, it is more difficult for voters to detect separate instances of malfeasance but, if they do, it is easier to get rid of corrupt parties by supporting newcomers (who claim higher morals). The interesting question is whether politicians perceive the threat of not being elected (i.e., be punished) as a feasible possibility in cases where the thresholds for representation are too low. Since PR systems in practice use (1) the same basic method of allocation of seats according to vote proportions; (2) multimember districting; and (3) party list competition, their net effect is to tolerate rather than depress incentives for rent seeking. At the engagement level, collective representation in party list elections produces an indirect link between voters and legislators where individual MPs are more concerned with relations to the party leadership than to their constituents. At the monitoring level, detecting and exposing corrupt incidents may not be a preferred strategy for opposition parties and rank-and-file representatives all of the time; considerations of future nominations and possible coalition partnerships can turn out to be stronger than the incentive to condemn violators. At the sanction level, all that matters is how elected officials assess the threat of electoral defeat. The prospect of voter coordination may make legislators in some single-member constituencies feel safe (at least in nonmarginal districts), but in multimember PR systems hope for representation is always strong because the minimum vote needed by parties to stay in parliament is low. Therefore, the net effect of PR multimember district party list rules on corruption should be one of increasing corruption (i.e., we can expect that in PR systems corrupt activities are more frequent, more difficult to monitor, and less likely to be punished than in majoritarian SMD systems). Regulatory burden. Empirical evidence from global studies on corruption reveals a clear pattern of relationship between the intensity of government regulation and the incidence of office abuse (Bardhan

Corruption in the Context of Postcommunist Transition

33

1997; Goel and Nelson 2005; Liu 2005). Based on these findings, we can anticipate that, where government intervention in the economy is more active and extensive, the new postcommunist systems will be more corrupt. Administrative barriers and regulations are known to breed inefficiency. When businessmen have to visit many agencies and pay various overlapping charges, and when licensing depends on submission of excessive documentation to bureaucrats, extortion and bribing escalate (Kaufmann and Siegelbaum 1997; Radaev 2002). Samuel P. Huntington (1968, 62) succinctly concludes that “the multiplication of laws thus multiplies the possibilities for corruption.” In Eastern Europe, the problem of regulatory overburdening of citizens is complex. It has roots in the past when individuals’ lives, careers, and well-being all depended on the mercy of officeholders who were loyal to the Communist Party. The image of a bureaucrat separated from the people by a window, or sitting in a chair in his office distanced from those whom he was supposed to serve, is hard to extinguish. The procedures of obtaining numerous approvals, permits, and validations were cumbersome in the past, which further amplified the role of the administration. Delays in civil service reform almost everywhere in Eastern Europe (except in Slovenia and Hungary) contributed to the survival, and even growth, of excessive regulation in the 1990s. Quite notably, bureaucrats desired regulatory overburden because it opened opportunities for extra gains from the privatization process and the establishment of new business entities. Complicated procedures to start a business are corruption prone everywhere, but it is reasonable to suspect that the scope and magnitude of this effect are much greater during the marketization of a centrally planned economy. Thus, the high demand for government services, a slowly changing bureaucratic apparatus, and a culture burdened by the past helped regulatory complexity to endure and made its influence particularly damaging. Economic capacity. The ability to prevent and punish corruption is further shaped by the strength of the national economy (Treisman 2000; Gerring and Thacker 2004; Goel and Nelson 2005; Xin and Rudel 2004). There are several reasons to consider economic capacity as a relevant explanation of the existence of corrupt practices. First, in order to be effective, even the most well-designed anticorruption policies need to be strictly implemented. Among many other things

34

Political Corruption in Eastern Europe

(well-trained personnel and adequate technical equipment, to name a couple), the administration’s material capacity increases a nation’s ability to effectively cope with corrupt practices. Second, scholars and policymakers alike have argued, and survey research confirms, that well-paid bureaucrats are less likely to accept bribes (Goel and Rich 1989; Hankiss 2002; Montinola and Jackman 2002; Xin and Rudel 2004).6 Moreover, citizens in more economically advanced regions are better educated and more likely to detect and reject corrupt behavior (Dininio and Orttung 2005). Finally, elites in poorer countries seem to be more likely to raise funding for party activities by using illegal means (Kolarska-Bobinska 2002). The region of Eastern Europe underwent a radical economic transformation toward a market economy after 1989, which led to disturbances in the rates of economic growth in all postcommunist countries. A common feature shared by all these states was a lack of sufficient resources in the national budget to adequately fund the administration, including the law enforcement structures (Kaldor and Vejvoda 1997). Yet countries recovered at different points in time during the 1990s, following economic restructuring, liberalization, and privatization. As various economic indicators reveal, the rate of economic success also varies across space, with a few nations performing almost as well as some of the Western European ones and the rest lagging behind. Thus, even among themselves, the new democracies differ when it comes to material capabilities. They have relied on different economic resources available for investment in anticorruption monitoring and prevention as well as for rewarding public officials sufficiently well. These varying capabilities should be held accountable when assessing the causes of corruption and the efforts to combat it. Democracy. Previous studies have hypothesized that democratic development may have a positive effect on the effort to contain corruption.7 This impact is exercised through practices of more transparent policies and equality before the law (Xin and Rudel 2004), years of uninterrupted democracy (Gerring and Thacker 2004), and improved citizen access to information (Boix 2003, 208–209). Political freedoms, such as a free press and freedom of assembly, deter corruption by exposing its menace and by mobilizing citizens to remove the guilty from power (Rose-Ackerman 1999; Goel and Nelson 2005). Moreover, some scholars specify a curvilinear rela-

Corruption in the Context of Postcommunist Transition

35

tionship over time, where corrupt activities initially intensify due to the numerous opportunities for enrichment and the immaturity of democratic institutions at the start of democratization (Montinola and Jackman 2002).8 Democratization in Eastern Europe advanced at a different rate after 1989; therefore, this might be a reason that some countries are less successful in reducing corruption. The new constitutions, especially the abolition of the leading role of the Communist Party, provided the institutional basis and guarantee for all people to be treated equally before the law. This principle was introduced in a swift manner throughout the region. The Eastern European countries, however, had different experiences with the enforcement of democratic minority rights and the introduction of access-to-information legislation. At certain times, some ethnic groups were less free (e.g., Russians in Latvia and Estonia, and Hungarians in Romania and Slovakia) and citizens were poorly informed on public affairs (especially in the early years of transition). Under such circumstances, policy transparency suffered and corrupt behavior could not be exposed. Finally, despite their common communist past, the Eastern European nations had different pretotalitarian experiences with democracy. Those might be less relevant, given the forty-five-year (or longer in the case of Russia) communist rule, but it is possible that previous democratic practices had a long-term impact enhancing a culture of intolerance toward malfeasance. International Factors The EU accession process. Immediately after the collapse of the communist regimes and the start of democratic transition, the elites of most of the Central and Eastern European countries declared their desire to join the European Union. The goal to “return back to Europe” became as important as the pursuit of democratic freedoms and the effort to transform the economy on a market principle (Grabbe and Hughes 1998, 7–8). Aspiring for membership, these states demonstrated readiness and remarkable commitment to introduce and to implement significant changes required by the EU. In a few policy areas such as environmental protection, legislation was adopted to establish standards even higher than those in practice in some member states. Could the struggle to become integrated into the EU also lead to cleaner policies?

36

Political Corruption in Eastern Europe

International integration may have a constraining effect on corruption. Wayne Sandholtz and Mark M. Gray (2003) propose that this effect is exercised through two main channels: economic and normative. The economic model assumes rational motivations on the part of political and economic actors who choose to avoid corrupt activity in order to maximize benefits. Since the costs of corruption rise in transnational interactions (from resulting relative price levels), the number of those who prefer to commit fraud in order to get things done will decrease in internationally integrated states. The normative model, on its part, assumes that engagement in international organizations plays a socializing effect by diffusing anticorruption ideas and norms. Thus, in participating countries, perceptions that corruption is wrong and not appropriate spread at a higher rate. Also, by embracing these standards of conduct, member states of international organizations are more effective in reducing and eliminating corrupt behavior. Applying this framework to the case of the European Union, one can argue that EU member states are less corrupt and empirical research so far points in this direction (Anderson and Tverdova 2003; Taagepera 2002). But until May 2004, none of the Eastern European postcommunist countries had obtained EU member status. Between 1991 and 1995 nine of them signed Association Agreements,9 and by 1996 ten had applied for accession. By signing those agreements, the postcommunist states made a commitment to comply with “the principles of freedom, respect for human rights and fundamental freedoms, and the rule of law” (Article 6 of the Treaty on European Union). There are good reasons for applicant status to also affect a country’s determination and success in reducing corruption. Similar to member states, while still aspiring for membership countries become both “norm-rational” and “utilityrational.”10 In important ways, the pursuit of accession may be an even stronger motivator for effective control over corrupt practices than membership itself. Let us first look at the possibility for a socialization effect of the EU candidate status. By signing the Association Agreements, the postcommunist states made an important step toward integration. For their elites and citizens, this was the beginning of a dream turning into reality—rejoining Europe to which they claimed to belong geographically, historically, and culturally. The principles on which the Western European societies had developed their political systems

Corruption in the Context of Postcommunist Transition

37

were readily accepted; values and beliefs were absorbed as right and fit to compensate for the forty-year period of communist rule. Consequently, understandings that corruption should be condemned penetrated the candidate countries at a higher rate than noncandidate states. The second mechanism through which the EU may affect anticorruption efforts in applicant countries is more economic in nature. In the long run, the postcommunist states expect that becoming part of the Union would make their nations more prosperous. In the short run, financial aid distributed by the EU to support the preparation for accession is conditioned by the progress of states in fulfilling the conditions for membership. The latter were specified at the June 1993 EU meeting in Copenhagen as falling in three categories: political, economic, and legal. Among the political criteria, stable institutions and respect for the rule of law were requirements indirectly related to the ability to avoid, detect, or sanction abuse of power. Only in July 1997 did the European Commission include in the political criteria section brief recommendations for applicant countries to reduce corruption (“Agenda 2000”). The European Commission used its leverage in the accession negotiations and, in the late 1990s, pushed candidate states to join the Council of Europe and the Organisation for Economic Co-operation and Development (OECD) conventions against corruption and bribery (Monitoring the EU Accession Process 2002, 18). While ratification of these documents by Eastern European applicants was perceived as important, the EU was lacking common standards in this policy area and not all member states signed the agreements.11 We can expect that postcommunist candidate countries aspiring to become EU members are more likely to reduce corruption than states not pursuing accession. The rationale for this distinct behavior is based on expected benefits through the preaccession aid programs and the achievement of full membership status. Involvement in international trade. The extent to which a country has liberalized its foreign trade is another external factor that might affect the level of corrupt exchange in a political system. Previous research has shown that the more active a country’s interactions are with other states, the higher are the costs of border bribery and extortion (Sandholtz and Gray 2003). Moreover, corruption has been found to grow in places where domestic producers are “shel-

38

Political Corruption in Eastern Europe

tered from foreign competition” (Ades and Di Tella 1999, 512). Experts contend that openness to international trade exposes a country to competitive pressures that penalize corrupt practices (Treisman 2000). The latter can thrive in closed economies where such pressures are much weaker. Thus, for countries tied to international networks of trade exchange, it is economically costly to ignore the bribery tax collected by officials. In brief, greater involvement in foreign trade changes the structure of available opportunities by increasing the costs of corrupt practices (Sandholtz and Koetzle 2000). How relevant is this explanation in the Eastern European post1989 context? One unique circumstance was the extreme restrictions imposed by the preceding regimes with regard to international interactions. In the past, trade was to a great extent limited within the framework of the Council for Mutual Economic Assistance (CMEA). In 1989, only a few communist states (i.e., Hungary, Poland, Romania, and Yugoslavia) were members of the General Agreement on Tariffs and Trade (GATT). After the collapse of communism, trade relations with the rest of the world exponentially increased. Until 2010, all except Russia had managed to enter the World Trade Organization (WTO) as a successor of GATT. Liberalization of trade connections became an important part of the economic transformation toward a market economy. The postcommunist states had to compensate for the loss of previous markets, especially in the former Soviet Union, and most of them quickly established active relations with Western Europe, the United States, and Asia. Consequently, with the increased exposure to competitive global markets, pressure to curb corrupt practices in the new postcommunist democracies also grew.

Other Explanations: Is Culture Relevant? The theoretical arguments developed in the preceding section are all made from a rational choice perspective (i.e., corruption persists where and when the costs of engaging in various forms of malfeasance do not exceed the gains expected from it). This implies that people commit corrupt acts, or refrain from doing so, for purely material reasons and that they are not culturally predisposed to bribery, favors, and fraud. However, claims that culture determines

Corruption in the Context of Postcommunist Transition

39

the degree of corruptiveness of a nation are not absent from the literature. Empirical studies, for example, have found that Protestant culture and British colonial heritage are significantly correlated with cleaner and more transparent policymaking (Lipset and Lenz 2000; Treisman 2000). Others claim that a connection between culture and corruption exists that is based on the extent to which different nations tolerate bribe giving as an expression of gratitude for well-delivered services or as a form of survival in times of scarcity and poor enforcement of the law (Kneen 2000; Holmes 2006, 186). A third group of scholars assigns breaches in elite integrity for the purpose of narrow gains to “sticky” cultural norms, thus questioning the effectiveness of any proposed anticorruption policies (Fisman and Miguel 2006). These cultures either emphasize achievement but impose limits on opportunities (the “means-ends schema”) or promote family ties and particularistic preferences (“amoral familism”) (Banfield 1958; Merton 1968; Lipset and Linz 2000, 116). In the particular case of Russia, Asian culture and communism shaped a value system emphasizing family links and friendship, believed by observers to favor corruption (Glinkina 1998). Similarly, experts on the Balkans suggest that a century-old culture of reliance on the family and on local strongmen enabled patronage networks and organized crime to thrive (Kanin 2003). Yet there is also research that fails to empirically ascertain that corruption depends on traditional norms. For example, apparent patterns of association between levels of malfeasance and culture break apart once controlled for the impact of other factors, including gross domestic product (GDP) per capita (La Porta et al. 1999). Also, the mechanism of the impact of colonialism as a cultural factor is not yet fully established (Lambsdorff 1999). The problem relevant to the goals of this study, regardless of the validity of the cultural argument in general, is: can culture help us to better understand postcommunist corruption? The answer to the above question may not be a simple one. We can certainly learn more about Eastern European practices of nepotism and bribery by merely linking those to the cultural patterns emerging from world survey data (e.g., from the waves of the World Values Survey). On the global cultural map, the postcommunist countries from the region cluster together, and separately from the Western consolidated and less corrupt democracies (see Figure 1 in

40

Political Corruption in Eastern Europe

Taagepera 2002). One may argue then that the Eastern European societies—rationally oriented and emphasizing survival—are more vulnerable to corruption than their Western self-expression–oriented counterparts (Taagepera 2002; Sandholtz and Taagepera 2005). As Ronald Inglehart and Christian Welzel (2005, 208) argue, nations poor in self-expression values are prone to “severe deficiencies in elite integrity.” Used in this way, the cultural explanation appears promising. Would this approach, however, help us when it comes to explaining cross-country differences in the region? Skeptics might quickly point to the fact that communism as a political system was introduced in Eastern Europe by Moscow in quite a uniform manner, including the enforced format of citizen socialization. Also, the pretotalitarian cultural differences should have been weakened by the communist regimes over the course of four decades. If, as argued, significant variation is absent, one can hardly expect that culture would provide us with an explanation of the puzzle of why some Eastern European states are more corrupt than others.12 To the extent to which distinctiveness was preserved, I expect that it may condition the effects of corruption on the way citizens relate to the state and its institutions.13 Thus, in my analyses culture is present more as a contextual characteristic and less as a factor that makes some postcommunist countries more corrupt than others.

Conclusion Corruption did not start in Eastern Europe in 1989; neither did it disappear with the democratization and marketization reforms there in the past twenty years. During the previous regime, members of the nomenklatura kept loyal to the Communist Party in exchange for prestigious jobs, promotions, and lucrative salaries and bonuses. Faithful service for a narrow party elite made bureaucrats disproportionately better off while the public suffered from administrators’ lack of competence and integrity. To survive and provide for their families, common people often became engaged in improper exchanges with public officials. The blurred boundary between state and party and between private and public hindered the development of a solid understanding of what people were entitled to and what they could not help themselves to.

Corruption in the Context of Postcommunist Transition

41

After 1989, corrupt practices changed in magnitude and form. While in the past making personal favors occurred within certain limits imposed by the centralized and hierarchical political structure, the fall of communism opened new and greater opportunities for stealing from the state. Privatization, in particular, was especially conducive to corrupt behavior: leaking of insider information on bids and taking huge bribes for privileged treatment at auctions for state assets were widely used across the region. Patronage and clientelism blossomed in variants not exploited before. In a volatile context with no clear or with quite vague rules and delayed reforms, corruption became one of the most serious problems in postcommunist Europe. Yet the degree to which it thrived has varied across space and time, thus raising the big question of why. My arguments in this chapter advance tentative answers that I drew from existing studies. To explain why some former communist countries are more corrupt than others, I focused on domestic and international factors that may have shaped distinct incentives for office abuse and may have affected the success of anticorruption measures. My theoretical framework identifies proportional electoral rules and excessive regulatory burdens as factors enabling corruption. At the same time, national wealth, progress in EU accession, and openness to international trade are expected to reduce the incentives for office abuse by raising the costs of malfeasance. Finally, I discussed the possible impact of culture as one that might be more helpful in understanding the causes of corruption in Eastern Europe from a global perspective and as less effective in explaining cross-national variation in the region. In the next chapter, I test these theoretical propositions with data from fifteen postcommunist states.

Notes 1. Some earlier research juxtaposes structural (geography, economy, and history) and transitional factors, and argues that the former carry more weight than the latter as determinants of corruption in Eastern Europe (Kanin 2003). 2. Studying the post-1989 transformation of the “second economy” in Hungary, Sik (1992) shows how the informal economy rapidly expanded to exceed in size the previous sector of activities left outside the control of the communist state.

42

Political Corruption in Eastern Europe

3. The question of whether privatization increased corruption in Eastern Europe has received different treatment in the literature. Some studies concur that the unprecedented scale of privatization after communism had a negative effect (i.e., it encouraged corruption) (Arikan 2008). Others argue that, in the absence of privatization, corruption might have occurred even more often (Kaufmann and Siegelbaum 1997). The latter propose that what matters is the mode of transfer of ownership, with faster reform through mass privatization and vouchers reducing the incidence of malfeasance. 4. Different studies provide distinct specifications of the domestic and foreign determinants of the postcommunist transitions. For Linz and Stepan (1996), the nature of the preceding nondemocratic regime is the determining force behind democratic consolidation. Kaldor and Vejvoda (1997) make a similar argument referring to past legacies as the “internal political energy” that shapes a “particular variant of democracy.” The external impact is seen as produced mainly by the “Gorbachev factor” (e.g., Linz and Stepan 1996) or by Western states and organizations offering expertise and financial aid against deepening of the democratic reforms (see, e.g., Kaldor and Vejvoda 1997; Vachudova 2005). 5. However, Lijphart’s (1999) empirical analysis does not conclusively confirm that there is a difference in corruption levels across PR and majoritarian systems. 6. There are, however, studies that produced poor or ambiguous empirical evidence for the positive effect of better-paid bureaucracies (see discussion in Lambsdorff 2007, 37–38; and also Treisman 2000). 7. But see also Ades and Di Tella (1997) and Fisman and Gatti (2002) who do not find empirical support for this impact. Mogan (2001, 386) even argues that democratization facilitates political corruption due to strong incentives to build political parties and to fund their struggle for power. 8. Empirical support for the proposed constraining effect of democracy has been mixed, however. Studies such as Ades and Di Tella (1997) and Fisman and Gatti (2002) failed to produce conclusive results on the constraining impact of democracy. 9. Hungary and Poland signed their Association Agreement in 1991; Bulgaria, the Czech Republic, Romania, and Slovakia in 1993; and Estonia, Latvia, and Lithuania in 1995. Only Slovenia signed in 1996. 10. Terms borrowed from Sandholtz and Gray (2003). 11. These international agreements include the Council of Europe Criminal Law Convention on Corruption; the Council of Europe Civil Law Convention on Corruption; the European Convention on Laundering, Search, Seizure and Confiscation of the Proceeds from Crime; and the OECD Convention on Combating Bribery of Foreign Public Officials. Because the European Commission has strictly followed their ratification in the accession process, by the summer of 2002 applicant states had a better record of joining the Council of Europe anticorruption regimes than EU member states (Monitoring the EU Accession Process 2002, 37).

Corruption in the Context of Postcommunist Transition

43

12. Two potential sources of such cultural distinctiveness are religion and colonial legacy, on which the Eastern European countries obviously vary. Cross-national studies that find evidence for the impact of culture, however, confirm the effects of only Protestantism and British colonial heritage (Treisman 2000). Both of these are not relevant in the Eastern European context. 13. I base my expectation on an insight from Husted’s (1999) work that cautions that anticorruption policies should take into account the cultural specifics of the place in which they are to be applied.

3 Identifying Specific Factors: A Quantitative Investigation

Depending on the socio-economic and political institutions of a given society, the incentives and opportunities for corruption differ. —Luigi Manzetti and Charles H. Blake, “Market Reforms and Corruption in Latin America: New Means for Old Ways”

How to effectively combat corruption during transition ultimately depends on a systematic investigation of the foundation on which malfeasance thrives. In the postcommunist world, past legacies and current reforms shaped an environment that sustained practices of and amplified appetites for illicit enrichment. The common legacy of communism and the market economy transformation present similar challenges to democratic consolidation throughout the region. Yet practices of office abuse penetrated some postcommunist states deeper than others (see Figure 2.1). Why is this the case? Why have some transitional contexts been more conducive to corruption, even though they share important structural features? In Chapter 2, I identified possible answers to this question by examining contributions made by previous studies of malfeasance in various parts of the world. To recall, my central arguments related corruption levels to party-centered political competition, regulatory burden, the pursuit of EU membership, and the economic strength of the state. In this chapter, I analyze cross-national variations in political corruption in Eastern Europe. While corruption may exist on differ-

45

46

Political Corruption in Eastern Europe

ent levels and in different forms, I describe the model that I designed to explain illegal activity as originating from the privileged position of public authority. In a general model of corrupt exchange (see also Philip 2002, 57), a public official violates the rules and norms of conduct to benefit a third party who gains access to goods and services. As a result of this, the official is rewarded for what he or she did (or failed to do) while the public interest is harmed. I propose that opportunities for illicit activities of public officials and the efforts made to limit those acts are shaped by political competition and economic structures domestically and by new allegiances internationally. My first hypothesis is based on an expectation that Eastern European states with party-centered multimember district electoral systems experience higher frequencies of office abuse. The effects of the rules that govern competition for power have already been explored in previous research (Persson, Tabellini, and Trebbi 2001; Kunicova and Rose-Ackerman 2005) but, to my knowledge, not yet in the Eastern European context of newly adopted election laws. Moreover, their proper assessment requires a systematic analysis that isolates the impact of electoral systems from other factors. My second hypothesis states that heavier regulation of business activities creates more opportunities for violations and personal enrichments of bureaucrats. This claim is based on World Bank studies of businessgovernment relations that report a strong association between economic freedom and good-quality governance (World Bank 2006). My third argument builds on an understanding that a country needs resources for effective monitoring and prevention of corrupt activities. That wealthier nations are less corrupt than poorer ones is a controversial proposition that needs to be subjected to an empirical test to check its validity in Eastern Europe. My fourth and fifth hypotheses reflect international influences over domestic levels of corruption. Thus, the pursuit of integration with the European Union is conceived as playing a positive role for containing corruption. While the literature on international integration and corruption finds a clear association between them (Sandholtz and Gray 2003), observers of the EU enlargement process note that combating illegal financing was not explicitly required for acquiring membership. Nevertheless, applicant countries undertook remarkable reforms in their legislation to harmonize policies with the Union, often introducing stricter standards than

Identifying Specific Factors

47

those existing in member states. Has reducing corruption also been embraced with vigor and commitment despite the absence of a specific anticorruption criterion for membership? I propose that the pursuit of EU membership made a difference. Finally, I also included exposure to international markets in the empirical tests as a factor that reduces corruption due to fewer barriers (i.e., opportunities for malfeasance) in foreign trade relations. Previous comparative research has examined corruption by analyzing cross-national worldwide samples of states (Ades and Di Tella 1999; Sandholtz and Gray 2003; Kunicova and Rose-Ackerman 2005). The results produced by that approach are sensitive to the particular time of observation. More importantly, the method also lacks the potential to account for dynamics in changing transitional environments. In my analysis, I used panel data: the units are countries that share a common history with communist governance and a political culture of purposive behavior. The cross-sectional time series design offers opportunities for comparisons across states while accounting for possible changes over time. This is an advantage, given the dynamic development of the simultaneous postcommunist political and economic transition. This chapter is organized as follows. First, I specify a general cross-national model that incorporates sources of political corruption, as recognized in extant literature and substantiated in Chapter 2. Then, I describe the statistical tests of the model using observations of fifteen postcommunist states between 1996 and 2008. I conclude with a discussion of the emerging factors responsible for the observed variation in corruption levels across the region. In this final part, I consider which theoretical claims have received solid empirical support and outline themes for more detailed investigation in subsequent chapters.

Modeling the Causes of Corruption Given the main goal of this chapter (i.e., to empirically identify factors explaining why corruption has been more pervasive in some countries than in others), I next build a basic model that incorporates the main theoretical arguments. Below, I operationalize the elements of this model and propose ways in which we can observe them and assess the expected relationships.1

48

Political Corruption in Eastern Europe

Corruption The challenge of observing and measuring corruption is one of the main obstacles that research of this kind needs to deal with. The TICPI and the WBI offer opportunities for a systematic comparison across countries, an advantage that enhances the test of each theoretical argument specified in the corruption model.2 The TICPI, which varies from 1 (most corrupt) to 10 (least corrupt), scored an average of 3.99 for the Eastern European sample of fifteen states (in Figure 2.1), between 1997 and 2008. Among these states, Estonia and Slovenia were given the highest score of 6.7 (in 2006 and 2008, respectively) and Ukraine received the lowest, 1.5 (in 2000). The WBI, which ranges from –2.5 (most corrupt) to 2.5 (least corrupt), produced an average score of 0.42 for the same group and period. This index identified Russia (–1.02) in 2000 as most corrupt and recognized Slovenia (1.10), again in 2000, as cleanest. Predictors of Corruption Proportional election rules. In Chapter 2, I argued that both domestic and international factors emerge as important for building a meaningful explanation of corruption in Eastern Europe. Among the former, party-centered political competition may be crucial; its structures shape motives and limit incentives in the game of (re)election, a main goal pursued by rational politicians. The proposed link between election rules and corruption is uniquely conditioned by the postcommunist transitional context. After 1989, the Eastern Europeans acquired opportunities to influence policies for good and to control policymakers through free and competitive elections. For voters, frustrated with previous practices of facade representation, the change meant electing responsive officials caring for the public good (Birch 2003, 79). To politicians, elections in times of reform meant acquiring access to an unprecedented process of redistribution of wealth from the public to the private sector. The rules of competition became crucial for the achievement of democratic representative policymaking in a transition from a command to a market economy. In this environment, new mechanisms of accountability emerged slowly; yet voters successfully removed incumbents from office in several countries. To model the effects of electoral systems, I used two specifications that reflect different aspects of the theoretical argument. First, I

Identifying Specific Factors

49

operationalized the concept of party-centered competition by the amount of representation based on party list elections. In pure PR systems, this amount is at a maximum, but in hybrid (mixed) systems it varies. To measure the effects of this factor, I introduced a continuous variable, PR share, which is the proportion of PR or multimember district seats converted into percentages. Second, I incorporated the notion of corruption monitoring ability through a district magnitude variable. Following (Persson 2002), I calculated a score for each country and system currently in use, by applying the formula 1 – (total districts)/(total seats). Higher values reflect larger average district magnitudes. In my sample, PR share had a mean of 78.3 and district magnitude of –0.74. Since the above two operationalizations are highly correlated with each other, I entered them separately in the regression tests. In all instances, I expected a negative relationship with corruption perceptions because more proportional systems are theorized to increase incentives for malfeasance. Regulatory burden. Another factor that also opens opportunities for office abuse is the amount of regulation imposed on the business in its interactions with public officials. I modeled excessive regulatory burden by including a variable that reflects the degree of government market-unfriendly policy measures. It is measured using the Business Freedom Index of the Heritage Foundation, ranging from 0 (least free) to 100 (freest). This composite measure captures various aspects of the “ability to start, operate, and close a business,” including procedures, time, cost, and capital (see a description of the index at www.heritage.org/index/business-freedom). Among the fifteen postcommunist countries, the Czech Republic received the highest business freedom score of 100 in 1996–1997 while Ukraine was the least free with scores of 43–44 in 2006–2008. Given this scale of measurement, I expected a positive correlation between business freedom and the TICPI and WBI indexes. Economic capacity. Third, I theorized national economic development as exercising a constraining effect on corruption. As described in Chapter 2, higher levels of national prosperity are associated with less illegal enrichment (Montinola and Jackman 2002; Liu 2005). This is because more abundant resources help to better fund the bureaucracy and the implementation of anticorruption measures. I operationalized

50

Political Corruption in Eastern Europe

the economic factor as national wealth, measured in GDP (thousands of US dollars) per capita. In the sample, it varied between 0.636 in Ukraine (1999) and 27.019 in Slovenia (2008). Following theory, we can predict a positive relationship between GDP per capita and the TICPI and WBI measures of elite corruptiveness. Democratization. My discussion in Chapter 2 suggested an impact produced by the overall levels of democratization on corruption. Equality of citizens before the law and information rights for all people, to name a few, are indeed important components of the emerging young democratic systems that were not achieved everywhere in the region at the same time. This variation can be legitimately considered as a reason for the different levels of corruption as well. More democratic systems conduct policymaking in a more transparent and accountable way; they are better equipped to deal with office abuse and to protect the interests of the public. For the empirical analysis, levels of democratization are measured using the Freedom House yearly ratings of countries on protection of human rights and political freedoms (Freedom in the World, various years). Since 2004, countries such as the Czech Republic, Hungary, and Poland were assigned a composite score of 2 (more free) while Russia received a score of 11 (least free in the sample). Democratization is expected to be negatively correlated with the TICPI and the WBI. EU integration. At the international level, the European Union is expected to have had a pronounced effect through the rising costs of ignoring anticorruption measures. Political and economic cooperation with the West was strongly desired in Central Eastern Europe from the very beginning of the transition. In the pursuit of closer integration, many there started to embrace the EU principles of government, including clean policies and commitment to combat malfeasance. As corruption began to emerge as a salient issue after the mid-1990s, the requirements for accession increasingly emphasized the need for transparent policies. Failure to meet the EU’s expectations for clean government might have turned into a threat to the prospect of integration, something to which national elites had to pay close attention. The above discussion points to two mechanisms of the EU’s impact, as reflected in the way extant scholarship analyzes the effectiveness of external factors. Thus, previous research operationalized the impact of international integration through membership (social-

Identifying Specific Factors

51

ization effect) and cost-benefit expectations (economic incentives) (Sandholtz and Gray 2003). I first operationalized the concept of EU integration as a state’s status of association with the Union, which generates incentives for socialization, including absorption of clean policies and norms. My preliminary analysis has shown that on average, Eastern European states that signed Association Agreements earlier in the 1990s have been more successful in containing improper exchange aimed to enrich public officeholders.3 Figure 3.1 illustrates that the gap in corruption perception levels between the early associFigure 3.1 Mean Corruption Perception Scores for Countries by Degree of Integration with the EU

Source: Transparency International Corruption Perceptions Index, various years, www.transparency.org. Note: Higher value = less corruption.

52

Political Corruption in Eastern Europe

ated countries on the one side and the late and nonassociated on the other has been maintained over time. Multivariate analysis helped to establish whether this comparison held after I controlled for other possible factors. For the purpose of this more rigorous test, I constructed a dummy variable in which countries that had signed the Association Agreement with the EU were assigned a value of one and those with no such agreement were assigned a value of zero. Preaccession funding. To model economic pressures encountered by postcommunist states in their interactions with the EU, I considered inflows of financial help from the Union to Eastern Europe. For nations with Association Agreements, one indicator of such assistance is the EU preaccession funding programs intended to improve states’ capabilities to implement reforms. This variable is measured through the amount of financial help per capita received through the program of community aid to the countries of Central and Eastern Europe—initially called Poland and Hungary: Assistance for Restructuring Their Economies (PHARE) and later extended to other candidate states—as reported in the country reports for progress in accession (Progress Reports of Candidate Countries, various years).4 Because PHARE was in force before accession, I included observations for the recipient countries until 2003 in the regression analysis that follows. For both specifications, a positive relationship with the dependent variable was anticipated (i.e., cleaner government is more likely in countries with Association Agreements and where PHARE financial assistance was more generous). Trade openness. Another factor incorporated in the model, which was expected to directly affect the rate of corruption incidents, is foreign trade. As I discussed in Chapter 2, scholars have argued that the more active a country’s interactions are with other states, the higher the cost of border bribery and extortion will be (Sandholtz and Koetzle 2000). This happens because of unfavorable changes in relative prices that hurt firms’ competitiveness on international markets. Empirical cross-national data support this proposition (Ades and Di Tella 1999; Sandholtz and Gray 2003; Liu 2005). I controlled for this possible effect in postcommunist states by introducing a variable for trade openness measured as imports and exports weighted by total GDP. In brief, I anticipated a positive correlation between trade openness and corruption-free government.

Identifying Specific Factors

53

Other Modeling Issues Before I present and discuss the results of the multivariate regression analysis, I address two potential problems that might produce contamination in the findings. First, democratization is strongly correlated with another independent variable in one of the tests of the basic model, EU integration/association (r = –0.824, significant at the 0.000 level). This high level of correlation is not surprising, given that democratization was a main attribute that the EU looked for when signing Association Agreements with the postcommunist states. Given my interest in the effects of European integration, unique in the case of the last wave of EU enlargement to postcommunist countries, and to make sure the results were robust, I chose to exclude the democracy factor from this part of the analysis.5 Democratization stayed in the analysis when the effect of PHARE was entered, as the levels of correlation between these two factors (r = –0.283, significant at the 0.000 level) are lower, and so is the threat of multicollinearity. Another issue is the possibility for endogenous relationships or self-selection effects in the data analysis. In other words, could there be a reversed direction of causality (i.e., elites were corrupt and that is why they adopted PR systems)? Is it that exactly the countries with cleaner policies were invited to sign EU Association Agreements? In my analysis, the problem of endogeneity was moderated by the following developments and means: first, recent research on constitutional engineering in postcommunist Europe does not recognize corrupt practices as a determinant of electoral system choice. In the very early years of transition, party organizational coherence, voter support, and ethnic composition were the main factors that shaped the choice of new electoral methods (Ishiyama 1996; Birch et al. 2002). In many instances, proportional rules were preferred as more democratic and were considered as an institutional means to promote the emergence of a multiparty system. Second, when the first ten EU Association Agreements were signed, corruption was not a salient issue and had not been considered as relevant. There is also no evidence that the EU used a corruption-free criterion to select candidates (Monitoring the EU Accession Process 2002). Third, in order to allow some time for the theorized effects to take place in the expected causal direction, I lagged several variables by one year, funding from PHARE, trade openness, and economic capacity.

54

Political Corruption in Eastern Europe

The final specification of the basic model is as follows: Corruption = a + b1(proportional rules) + b2(regulatory burden) + b3(economic capacity) + b4(democratization) + b5(EU integration) + b6(trade openness) + e

where proportional rules stands for either the share of PR seats in a system or for the size of the district, and EU integration is operationalized first as EU association status and then as preaccession funding through the PHARE program.

Analysis To assess the impact of the factors specified in the model above, I now turn to the data. As I reported in the description of operationalization and measurement of variables in the previous section, the Eastern European nations differ with regard to their domestic electoral institutions and economic potential. Also, some states signed EU Association Agreements earlier, others later, and the rest have not done so. I examined these variations to establish patterns of correlation with the prevalence of public office abuse. The basic model allows for comparisons of corruption levels across states (e.g., by election method) and over time (within same countries as they undergo institutional reform and experience economic growth). Thus, the spatial and dynamic components help to utilize the empirical data in their full richness. Two considerations guided me in determining which countries to include in the analysis. First, due to data limitations, I could not investigate the determinants of corruption in all postcommunist countries. Moreover, the time series for some countries have substantial gaps. Second, I had to exclude those states that in the mid1990s were still experiencing serious border disputes (Armenia, Azerbaijan, Georgia, Moldova, and Serbia and Montenegro) or had not made significant progress in the establishment of democratic institutions and competitive electoral practices (Belarus and the Central Asian former Soviet republics). Thus, after these exclusions, my data set contains information from fifteen countries in the period between 1996 and 2008. 6 The most complete time series include thirteen years (the Czech Republic, Hungary, Poland, and Russia)

Identifying Specific Factors

55

while the shortest ones are composed of six to seven years only (Croatia and Macedonia). In the first round of tests, I analyzed the hypothesized role of electoral rules, regulatory burden, economic capacity, trade openness, and EU association status. For the impact of proportionality of election arrangements, I tested the impact of the weight (PR share) of multimember districts compared to that of single-member districts and explored the role of larger versus smaller district size (district magnitude). To recall, evidence for the former indicator would confirm the importance of accountability in public officials’ deciding whether to engage in illicit exchanges. Evidence for the latter institutional trait would speak in support of the theorized easier monitoring and, subsequently, prevention of malfeasance in smaller electoral districts. Furthermore, empirical evidence for the importance of early conclusion of an EU Association Agreement (EU association) would support the argument for the role played by international institutions in combating corruption. Finally, if confirmed by the data, excessive regulation (regulatory burden) would come out as generating more corruption while national wealth (GDP per capita) and engagement in international transactions (trade openness) would be factors constraining corrupt behavior. Table 3.1 reports the results of the estimation. The results offer strong evidence for the importance of regulatory burden, economic capacity, and EU association status.7 I estimated correctly (positively) signed and statistically significant coefficients for these three factors in all tests and for both TICPI and WBI measures of corruption. The results for the effect of extensive government regulation were very consistent. The estimated parameters suggest that all else being the same, when procedures of business licensing are more complicated by one category, perceptions of public officials’ engagement in illicit enrichment increase by about 0.022 (TICPI models) and 0.008 (WBI models) points. Readers may recall that government regulation here is measured through business freedom scores; hence, the results reflect a positive rather than negative association. Another strong result offered by the tests reported in Table 3.1 is the positive effect of national economic capacity on transparent policymaking. Measured as GDP per capita, this factor is consistently correlated with the rates of corruption. The estimates allowed me to predict that an increase of US$1,000 in national GDP per capita would result on the average in a 0.10 (TICPI) or 0.04 (WBI) point

56

Political Corruption in Eastern Europe

Table 3.1 Regression Analysis: The Causes of Corruption, 1997–2008 TICPI Independent Variable PR share District magnitude Regulatory burden GDP per capita EU Association Trade openness Constant R2 N

WBI

1

2

3

4

.002 (.002) —



.001 (.001) —



.023*** (.007) .103*** (.016) .700*** (.239) .496* (.280) .698 (.462)

.061 (.227) .024*** (.007) .104*** (.016) .745*** (.236) .511* (.277) .740 (.459)

.713 163

.008*** (.003) .038*** (.013) .585*** (.117) .150 (.152) –1.399*** (.227)

.714 163

.049 (.108) .008*** (.003) .038*** (.013) .601*** (.118) .152 (.151) –1.376*** (.229)

.405 188

.405 188

Notes: TICPI = Transparency International Corruption Perceptions Index; WBI = World Bank Control of Corruption Index. Entries are ordinary least squares regression coefficients, common AR(1) parameter, pairwise estimation, panel corrected standard errors in parentheses. * p < .1; *** p < .01; two-tailed tests.

decrease in corruption. As expected, differences in the national economic capacities in Eastern Europe explain part of the variation in the way these countries manage to punish corrupt practices and award integrity in public office. In agreement with theory, EU integration appears to have had a pronounced impact on reducing corruption. The empirical data from Eastern Europe suggest that overall, corruption in politics is evaluated as between 0.6 (WBI) and 0.7 (TICPI) points stronger in countries that had not signed agreements for integration with the Union until 1996. This result is consistent across the two measures of corruption. The results also confirm that the mobilizing effect of association and probable EU membership is independent from domestic political and economic factors.

Identifying Specific Factors

57

Increased exposure to competitive pressures from international trade appears to have reduced corruption. The estimated parameters had the expected positive sign in all tests and reached statistical significance when corruption was measured through the TICPI. Although not as convincing as the findings for the three factors discussed above, the results for international trade offer a good amount of empirical support. In brief, there is some evidence that corruption is at lower rates when the postcommunist nations are more engaged in networks of trade with other countries. The results for proportional rules are much less certain. They are in the opposite direction and are not statistically significant. What this means is that the data from the fifteen postcommunist states do not provide an empirical confirmation for a damaging effect of the PR systems on transparency in politics. The data do not suggest the opposite either. From this first set of tests, it seems that election system type does not matter for achieving cleaner from malfeasance government.8 The second set of tests explored whether material benefits from compliance with EU norms also drive incentives for reducing corruption. Here, I focused on the role of preaccession finance programs in associated states. The results of this second round of regression analysis are shown in Table 3.2. As in the first part of the analysis, I ran the tests with TICPI and WBI data to ensure more robust results. Because the PHARE preaccession funds (i.e., their potential losing) are of central importance, I used only data from country recipients of such assistance. Similar to the previous findings, regulatory burden and economic capacity (GDP per capita) were positively correlated to cleaner from corruption policies and their impact was statistically significant. These results are robust across all tests, as are the ones for democratization. Solid empirical evidence suggests that where political freedoms and human rights are respected, corruption levels drop. At the same time, the parameters obtained for trade openness were weaker than in Table 3.1. The coefficients carry the correct sign, but all failed to reach significance at the commonly accepted levels. Again, little support is offered by the data for a negative effect of proportionality in election rules on elite integrity. In only one of the four tests, the estimate of this impact had the correct sign, but was not statistically significant. Now, I focus more closely on the results for the role of PHARE. The hypothesis of its constraining influence on corruption did not

58

Political Corruption in Eastern Europe

Table 3.2 Regression Analysis: The Effects of PHARE on Corruption Levels, 1996–2003 TICPI Independent Variable PR share District magnitude Regulatory burden GDP per capita Democratization PHARE Trade openness Constant R2 N

WBI

1

2

3

4

.001 (.004) —



.001 (.002) —



.034*** (.006) .106*** (.026) –.201*** (.052) –.006** (.003) .158 (.307) 1.876** (.808)

–.317 (.402) .034*** (.006) .108*** (.027) –.218*** (.050) –.006** (.003) .181 (.307) 2.206*** (.784)

.809 103

.009*** (.003) .105*** (.016) –.105*** (.021) .001 (.001) .150 (.155) –.754** (.298)

.810 103

.016 (.176) .009*** (.003) .106*** (.016) –.111*** (.020) .001 (.002) .153 (.157) –.656** (.280)

.645 128

.647 128

Notes: TICPI = Transparency International Corruption Perceptions Index; WBI = World Bank Control of Corruption Index. Entries are ordinary least squares regression coefficients, common AR(1) parameter, pairwise estimation, panel corrected standard errors in parentheses. The panel includes observations on Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, and Slovenia. ** p < .05; *** p < .01; two-tailed tests.

receive much support from the data. The estimated regression coefficients were extremely small in magnitude, not significant in the WBI tests, and significant but wrongly signed in the TICPI tests. In brief, the results were mixed and did not allow any definitive conclusions. The puzzling question is why EU association had a robust effect on reducing corruption while more material aspects of integration such as preaccession money did not appear to. In the next section, I discuss the findings, including the mixed evidence about EU integration and weak results for electoral rules as well as the need to explore them further.

Identifying Specific Factors

59

Discussion and Conclusion A cross-national analysis of corruption in Eastern Europe is important from at least two perspectives. Theoretically, we may be provided with new insights about the causes of public office abuse in postcommunist transitional environments. From a public policy point of view, findings about what favors corruption may offer answers to the question of what to do in order to prevent the spread of power abuse in young democracies and to reduce its damaging effects. The analysis thus far confirms the impact of some domestic and international factors on the incidence of corruption. The results suggest that excessive regulation, the economy, and the EU integration status have affected the rates of corruption after communism. At the same time, the role of election systems and financial aid from the Union remains uncertain. PR and Other Electoral Institutions The weak findings for the potential of PR to favor corrupt behavior require further investigation. Given the initial shift toward proportionality in Eastern Europe after 1989 and the continuing electoral reform, one cannot simply ignore the issue. The analysis in this chapter confirmed that democratization is an effective deterrent of corruption; the election method, however, seems not to matter. In other words, systems with more seats contested through party lists cannot be blamed for breeding more corruption. Yet popular frustration with PR, especially with its closed list version, grows in the region. Under such pressure, Bulgaria and Romania introduced a majoritarian element in their most recent parliamentary elections (2009 and 2008, respectively). Along with strengthening democracy, would electoral reform provide a remedy? It would be naïve, I think, to expect that simply switching from multimember to single-member districts would provide a magical solution to the problem of corruption. Moreover, a radical shift is less feasible than a moderate change. The record of electoral reform in the region up to now shows that, when countries depart from pure PR, they choose a mixed formula retaining a PR element.9 Proportional rules have been praised throughout the region for their capacity to facilitate the development of multiparty systems after communism and to ensure representation for minorities and

60

Political Corruption in Eastern Europe

women. Such arrangements will most likely remain in use and anticorruption policies will have to be adapted to address the very mechanisms hindering accountability under PR. One logical solution would be to introduce preferential voting (open list PR). Several Eastern European states have used it (the Czech Republic, Estonia, and Latvia, to name a few), but preliminary analyses (not reported here) did not gather sufficient empirical support for this possibility.10 To make the threat of sanctions against corrupt MPs real, active engagement of voters at the nomination stage is needed. More decentralized procedures for nomination of candidates for the PR party lists should be considered, including candid dialogues with local organizations and open meetings with supporters. Difficult to detect, acts of corrupt exchange of favors between politicians and organized interests remain closely connected to the process of electoral competition. Especially important in this regard seem to be existing legislative frameworks dealing with campaign finance. In particular, these laws establish the rules of public financing of political parties, funding and spending limits, and disclosure regulations and enforcement. By analyzing various models of election finance in addition to the election methods, we might be able to identify institutional traits facilitating transparent and accountable government. Furthermore, one should not ignore the emergence and evolution of interdependent informal political and economic networks in Eastern Europe. In those, political elites in power award particular businesses with public procurement contracts or favorable legislation while the companies, on their part, finance parties beyond what is envisioned by the law. A closer look at these informal networks will cast light over how election and campaign rules are ignored and manipulated. I explore these themes in more detail in the following two chapters. The EU Normative and Financial Pressures The conclusion that EU accession policies have had a positive impact on the postcommunist countries’ efforts to combat corruption bears implications for both Eastern Europe and other transitional democracies. Previous research has shown that membership in international organizations has the potential of inspiring reforms and raising the costs of bribing and extortion (Sandholtz and Gray 2003). In this chapter I suggest that pursuit of integration may also play an impor-

Identifying Specific Factors

61

tant role in constraining corrupt practices. Indeed, much has been done toward strengthening the democratic principles of government in Eastern Europe that would not have been in the absence of this new international allegiance. Apparently, aspirations to join the EU made a difference. To further explore this theme, it is worthwhile to disentangle the mechanisms of the EU effects and assess them separately as norm rational and utility rational drivers of change. My empirical results suggest that, when accession is desired and pursued as a national goal, political will to combat corruption at home is more stable even in the uncertain context of transition. This finding is supported by a 2006 World Bank report (Anderson and Gray 2006, 80–81) that also distinguishes between phases of integration. Focusing on government improvements in Central Eastern Europe, this World Bank report finds that anticorruption policies are actually most intense in countries closer to accession. This means that the EU effect may vary with the degree of integration (i.e., intensifying when membership is feasible and weakening after achievement of member status). Whether changes in the anticorruption policy effort follow the EU accession dynamics is a question that will be empirically addressed in Chapter 7. On the utility rational side, the empirical findings raise concerns that the EU preaccession programs may not be as effective as desired. At least in the area of anticorruption policies, private benefits from breaching the law make the loss of international aid meaningless to elites. Or, is PHARE turning into yet another source of funds for improper redistribution, instead of helping applicant countries’ transitional efforts? From a broader point of view, under what conditions can concerns for loss of economic benefits become a major factor in the calculus of national elites so that they decide to crack on corruption? The analyses in following chapters attempt to provide answers to these questions. State Capacity Finally, the empirical cross-national findings reported in Tables 3.1 and 3.2 offer solid evidence for the claim that containing and limiting corruption are a function of states’ economic capacity. We often hear excuses that bureaucrats and public officials take bribes as compensation for low salaries or that governments in transitional countries cannot do much in the fight against corruption because they lack

62

Political Corruption in Eastern Europe

resources to finance anticorruption campaigns. I explore this issue in Chapter 6, where I also assess the role of state vis-à-vis nonstate capacity. While a state’s material and institutional resources are important, there could also be societal factors that enable the fight against malfeasance. Therefore, my analyses account for how the Eastern European societies have been involved and to what extent they increased the national capability to deal with corrupt elites. In Chapter 4, I turn to the topic of campaign funding of political parties and corruption.

Notes 1. Readers can find a complete list of all variables used in this and other chapters, as well as their sources, in Appendix A. 2. Despite being criticized for their “perception-based” character, these indexes remain among the few best tools available for cross-national comparison. Inglehart and Welzel (2005, 193) characterize the WBI as “the most comprehensive and methodologically most sound measure of law-abiding and honest elite behavior, or ‘elite integrity.’” Introducing the TICPI in their cross-national analysis, Montinola and Jackman (2002) emphasize the sensitivity of the index to the issue of measurement bias and praise it for its reliability. For a more detailed discussion of these vis-à-vis other indicators of corruption, see Chapter 1. 3. This more EU-integrated group includes Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, and Slovenia. The second, less integrated, group consists of those that signed agreements after 2000 (Albania, Croatia, and Macedonia) and those that did not sign (Russia and Ukraine). 4. I would like to thank Petia Kostadinova for sharing some of her data on funding through PHARE and Aline Machado for collecting the rest of the data. This effort was partially supported by a 2005 summer research grant from the Miami-Florida European Union Center of Excellence for the project “The EU’s Commitment to Capacity Building in the New Europe.” 5. The results from tests run with “democratization” (“EU integration” excluded) are consistent with regard to findings on the importance of the remaining factors. 6. Albania, Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Macedonia, Poland, Romania, Russia, Slovakia, Slovenia, and Ukraine. 7. Given the cross-national time series dimensions of the panel data and the high probability of correlated errors across nations in the sample, I estimated the model using an ordinary least squares (OLS) technique with panel corrected standard errors (PCSE) (Beck and Katz 1995). The latter helps correct the standard errors produced by OLS with regard to two problems,

Identifying Specific Factors

63

groupwise heteroskedasticity and contemporaneous correlation. However, possible autocorrelation over time among the disturbances requires additional measures to ensure reliable estimates. Measuring the dependent variable through the TICPI makes the potential danger for timewise-dependent error terms real. As some scholars warn, the index may well become “self-referential” because it allows survey respondents to rank corruption levels on the basis of prior assigned values, rather than their own independent experiences (Golden and Picci 2005, 40). To model these dynamics, I employed a common AR(1) parameter that accounts for autocorrelation of the disturbances. Finally, the panel is unbalanced because data were not available for all fifteen countries for the time period under investigation. Therefore, I applied the Stata Data Analysis and Statistical Software pairwise specification for computation of the covariance. 8. This result is consistent with Lijphart’s (1999, 289) findings derived from an analysis of twenty-seven democracies. He found no significant correlation between corruption and regime type, consensus versus majoritarian democracy. 9. In fact, one reason for the inconclusive results on PR election rules in the regression analysis could be the absence of SMD-only systems during the time examined here. Thus, I made comparisons between pure PR and mixed system elections, both containing multimember districts. 10. Personalizing PR by allowing voters to identify favorite individual candidates on the ballot does not seem to have helped reduce corruption. The tests that I made with a variable distinguishing PR with preferential voting from closed list PR showed no difference with regard to impact on corruption. An alternative point of view is that open list provisions actually contribute to clientelism and particularistic distribution of public goods (Geddes and Neto 1992). Chang and Golden (2006) propose that open list PR systems may not be necessarily less corrupt, especially with large electoral districts. Similarly, Nyblade and Reed (2008) suggest that intraparty competition, characteristic for preferential voting, can in fact motivate cheating for electoral gain. Nevertheless, it might be early to dismiss the importance of clarity of responsibility, which is higher in open list systems when compared to the closed list version of PR.

4 Party and Election Finance

It is a common belief that the financing of political parties provides fertile ground for the development of corrupt societies. —Daniel Smilov, “Party Funding, Campaign Finance and Corruption in Eastern Europe”

In this part of the book I draw upon factors, identified in the concluding discussion in Chapter 3, and analyze in more depth particular sources of corruption. I explore possible linkages between party and election finance and illegal enrichment in this chapter. I survey the rules governing party financial affairs and the context within which they operate in Eastern Europe. In analyzing the institutional context, I distinguish between private and public funds, limited and unlimited amounts, and transparency favorable and accountability indifferent arrangements. My emphasis is on the selection of a finance model and how effective this choice has been in allowing parties to raise needed resources without jeopardizing the public interest. While I build on the achievements of a few previous studies dealing with parties and corruption in consolidated democracies, I keep the arguments sensitive to the peculiarities of the region and the challenges of the postcommunist complex socioeconomic transition. The working proposition, drawn from previous research, is that public funding contributes to cleaner policies while private funding makes parties more dependent on special interests and less responsive to their voters. In Eastern Europe, state subsidies were of crucial importance for the organizational development of new parties from the

65

66

Political Corruption in Eastern Europe

beginning of the transition. Other regulations, such as limits on donations, public disclosure, and sanctions, were introduced in response to the growing concerns linking parties to suspicious sources with hidden agendas. As I show in the analysis, politics were kept cleaner where the financial regimes included subventions, transparency requirements, and more stringent rules for punishment of violators. In the following two sections, I define the problem and explore its institutional, economic, and cultural dimensions. Then, I analyze the postcommunist party finance regimes and their consequences for the parties themselves, for the state, and for voters. To conclude, I draw implications about various forms of finance regulation as facilitators of clean politics.

Defining Corruption in Political Finance The starting point in an inquiry on the possible link between political finance and corruption should involve a definition of the nature and the scope of such an exchange. There are multiple ways in which corrupt funding of politicians has been characterized. Some rely on broad interpretations of political corruption along the lines of Donatella Della Porta and Susan Rose-Ackerman’s (2002, 9) conception of “exchange of money or material benefits in return for preferential treatment by a public agent.” Others cast more narrow specifications by detailing particular manifestations and their arenas. Examples of the latter approach include illegal financial exchanges to the benefit of “a political party, interest group, or of an individual candidate” (Walecki 2005, 2) and enrichment of politicians who betray their supporters to serve large contributors’ interests in the legislative field (Rose-Ackerman 1978). The most detailed description of corrupt political finance can be found in Rasma Karklins’s (2005, 25) typology of corrupt acts where illicit campaign funding is categorized through its effects as either “take over of state institutions,” “forming secret power networks,” “undermining elections and political competition,” or “misuse of legislative power.” Considerations are in order when deciding whether to use the common definition of corruption, as misuse of public office, when studying corrupt party funding. A broader definition would include all forms of political fund-raising. However, some of them are clearly not prone to malfeasance; rather, these maintain healthy links

Party and Election Finance

67

between parties and their supporters. In this regard, large amounts from membership fees and modest voter donations may generate a significant income for parties. Whether the latter spend the money irresponsibly is a matter different from a purposeful acceptance of bribes against delivery of returns for significant donors. Moreover, a corrupt deal involving party finance may be struck by politicians who are not in power; challengers may also accept money against a promise for future favors should they win the election. Another feature that makes political finance corruption distinct relates to the beneficiaries, who are not primarily individuals but most often political organizations (Pinto-Duschinsky 2002). One of the challenges in developing a clearer notion of the borderline between corrupt and only unauthorized political finance is that what is illegal might not necessarily be corrupt. For Marcin Walecki (2005), both legal contributions from disreputable sources and acceptance of money in return for favors are examples of irregular financing. While these behaviors are obviously illicit, only the ones linked to a repayment imply that a secret deal hurting the public interest was struck. For a corrupt exchange in party finance to take place, there has to be an expectation, a purpose, and a plan for provision of privileged access to public goods in return for irregular funding. In the framework of a principal-agent model, the party plays the role of an agent who misuses discretionary power delegated by voters to the benefit of a briber who pays private money to the party in order to receive a political rent (Della Porta and Vannucci 1999, 21). Forms of this, easy to classify as party finance–related corruption, include contributions to parties for favors, contracts, policy change, appointments, privileged access to information, and extortion by forcing the private sector to pay for favors or protection. At the same time, irregular fund-raising, improper documentation of expenditures, overspending, higher than normal membership fees from MPs, and contributions from government officials for appointments would not amount to corruption had there not been a prearrangement for material returns or preferential treatment. Before turning to particular indicators of malfeasance in party finance, one other trait of the phenomenon that is especially important in nascent democracies needs to be addressed. The private character of exchange has specific projections when it comes to party finance. Money obtained by politicians in corrupt ways may or may

68

Political Corruption in Eastern Europe

not be transferred to their campaigns; yet most of the time, funds can serve for sustaining a good standing in the electoral race. Examples from Ukraine and Poland point to a thin and often invisible boundary between candidates’ material and political capacities to compete (Walecki 2005). The issue becomes more complicated in environments marked by uninstitutionalized party systems and unstable voter support. There, in the absence of parties with strong roots in society, a business gets directly involved in the race for public office by engaging in friendly circles with the governing political elite and even by forming parties (as in Poland and Bulgaria) or interparty parliamentary factions (as in Ukraine and Russia). This type of systemic corruption, enhanced by an underdeveloped civil society, impotent judiciary, and weak party-voter links, is saturated with politicians who run the business of politics for their own gain. This happens regardless of society’s urgent needs, with businesspeople getting into politics to directly prey on the state by playing the game of politics. Existing cross-national studies on party finance offer long lists of categories of potentially corrupt funding (Pinto-Duschinsky 2002). These forms of politician-party-briber deals are marked by an element of secrecy and a profound ignorance of the public. Examples of such payments or other forms of exchange are contributions made against the existing law (because it prevents corruption), the use of bribe money for campaigning and other party operations, unauthorized use of state resources, and spending on banned activities such as vote-buying (Roma votes in Bulgaria). Drawing from the Eastern European experience with economic and political transformation in the 1990s, the list can be continued to include (mis)use of revenues from privatization and international aid for narrow party purposes, siphoning of state resources by the communist successor parties through offshore companies or unregulated bank loans, and enrichment through bribes or nontransparent procedures of privatization. Apparently, opportunities for improper party funding exist in abundance, as shown by the creative ways used to raise money. Practices of redeployment of national resources have been facilitated by weakened state infrastructures, the enormous tasks of a simultaneous economic and political transition, and institutional changes in process. The deepening process of globalization helped violators cover their crimes and transfer unpunished the gains to other parts of

Party and Election Finance

69

the world (Ganev 2007). Next, I analyze in more depth the incentives of parties and private interests to enter corrupt exchanges in the context of post-1989 economic opportunities, institutional constraints, and cultural preconditions.

What Drives Parties to Engage in Corrupt Financing Schemes? In order to function and to run in elections, parties need resources that they can raise from membership fees, business activities (rent, property management, and publication), donations, and public subsidies. If they fail to fully institutionalize into mass ideological organizations, parties cannot rely on monetary support from numerous like-minded supporters and on voluntary campaigning by loyal party members (Hopkin 1997). Moreover, mounting data from around the world suggest that the costs of electoral campaigning are increasing at unprecedented rates with the introduction of modern and more sophisticated means of mass communication (Heywood 1997; Pinto-Duschinsky 2002). We have many reasons to consider the Eastern European post-1989 parties as resembling more a business firm, professionalized type than the traditional organization known in the West. This observation points to an increased risk for politicians to resort to illegal sources and, consequently, become dependent on them due to a reduced base of social support. There is also evidence that parties in postcommunist Eastern Europe have significantly increased the money spent on billboard advertising and running of television ads in order to reach their potential voters and to alter the electoral outcome in their favor (Walecki 2003; Smilov and Toplak 2007). Interested in maximizing their profits, private contributors seek to achieve gains and to reduce costs. To attain competitive advantage vis-à-vis their rivals, firms look for preferential treatment from the politicians in government. Carolyn M. Warner (2005) characterizes this situation as one generating a typical collective action problem: firms contribute to parties in exchange for favorable policies, but others who did not incur that expense benefit as well. Therefore, private interests demand specific favors tailored to their business only, such as winning a government contract or a subsidy. To solve the collective action problem, politicians come up with selective incentives in the

70

Political Corruption in Eastern Europe

form of providing restricted access or offering patronage. The key point here is that donations may be promised and given, but election results are uncertain and politicians cannot guarantee the returns with absolute certainty. Donors cannot totally ignore and refuse to pay them because, if elected, the politicians will be in a position to cause problems for the firm. Because of the uncertainty, firms may not offer significant support, in which case parties remain underfunded and have to turn to multiple donors in the illegal market for funds. The above framework of incentives, developed in studies on consolidated democracies, may also be a useful tool in examining corrupt practices in party finance in Eastern Europe. What further shapes the choices made by postcommunist politicians in these countries is the unique transition from a one-party system to multipartyism and competitive politics and from a command toward a market economy.1 As Paul Heywood (1997) argues, variations in perceptions of corruption can be explained best if one focuses on the rational choices made by main players within the constraints of existing institutions and the opportunities offered by changing social environments. In Eastern Europe after 1989, new constitutions, electoral laws, and acts gradually established the forms and the limits of party finance for political parties. Many of those legislative acts have gone through changes or been replaced by new ones in the course of the past fifteen years. The opportunities for extra and corrupt party fund-raising also changed by type, from redistribution of state resources to enrichment from privatization deals and public procurement contracts. The Institutional Context To avoid illegal financing and reduce the uncertainty surrounding resource mobilization, regulations were introduced after the beginning of the transition and those rules reflected the interests of sitting incumbents. The parliamentary parties adopted laws establishing regimes that, to a large extent, shaped the emerging new party systems. In this policy area, postcommunist elites have aimed at creating such sets of regulations (contribution and spending caps, reporting, and sanctions for noncompliance) that would benefit their own political survival. In making these choices, parties in government were motivated by two main goals: (1) prevent newcomers from entering the arena of policy decisionmaking; and (2) prevent a possible withdrawal of voter support (in case voters get disappointed). The result-

Party and Election Finance

71

ing party finance regimes could contain four basic elements broadly believed to constrain corruption: public funding (direct state subsidies); limits on donations and expenditures; requirements for transparent finance through public disclosure; and punishment for violators. All of these are believed to increase the costs of illegal and corrupt fund-raising and to increase the benefits from clean and transparent finance. Direct state subsidies. Public funding, also strongly associated with proportional representation electoral systems, was endorsed in most postcommunist countries following foreign examples. In continental Western Europe, state aid has been popular with the political class and highly unpopular with the electors (Pinto-Duschinsky 2002). Throughout the world, subventions have been intended to make parties independent from large donor and corporate money; yet it remains uncertain if this policy tool can succeed in curing the problem (Pinto-Duschinsky 2002). Some of the most serious corruption scandals broke out in countries with the most generous subsidies (France, Germany, and Spain). For some experts, public subsidies will not stop parties from looking for more money to outperform opponents; in fact, state money may discourage parties from establishing stronger connections with voters, supporters, and activists. The amounts provided may also be insufficient or nonexistent, which, in practice, would make the provisions meaningless. Scholars maintain that rather than being substitutive, available public funding has a supplementary impact on unauthorized party funding (Van Biezen and Kopecky 2001). Empirical research in this area has produced contradictory findings. On the one hand, receivers of greater subventions also tend to be more attractive to private business support. On the other hand, evidence that providing state finance hurts competition or makes it difficult for newcomers has been quite elusive. One would expect that if fairly distributed across parties, subventions should be of great help in postcommunist countries for organizational development, functioning, and campaigning, especially for noncommunist parties with limited means of their own. Anecdotal evidence suggests that in the Czech Republic and Slovenia, the size of subventions has been high enough to form a substantial portion of party revenues, yet state support has been questioned as an effective deterrent to corrupt financing (Toplak 2007). It is worthwhile to undertake a more systematic investigation to establish whether the

72

Political Corruption in Eastern Europe

introduction of public finance has had any distinctive impact on the level of corruptiveness of these newly emerging democratic political systems. Caps on contributions and spending. Setting maximum amounts for individual donations is another tool often used to prevent parties from falling under dependence on big businesses or individuals with wealth of unclear origin. The importance of private donations is twofold: on the one hand, supporters need to be given a chance to help their party financially in pursuing a favorite cause (Scarrow 2003); on the other hand, extra-large donations are few and they always raise suspicions of behind-the-scenes deals. Regulating the amounts spent on campaigning does not seem as controversial as the caps on contributions. Here the logic is that the less parties are allowed to spend, the less likely it is they will raise money from large donors who demand special treatment. Certainly, there are no guarantees that even in receiving the allowed amounts, parties will not engage in corrupt finance schemes. Yet it seems reasonable to expect that caps on contributions and spending will reduce the raising and spending rates, including the incidence of illicit bargaining and extortion. Transparency and disclosure requirements. Arrangements for regular and full release of information by parties are control mechanisms to ensure a fair and open process of party finance. Since politicians do not have the incentives to set those in place, they are usually introduced and strengthened at a later point in time. Setting up an independent control agency, with authority to investigate the origin and amount of money, is another important element in making sure the system of transparency is effective. Release of information by parties and independent monitoring and control of campaign funding become effective tools in achieving transparency only when there is active communication with the citizens. In some countries such as Bulgaria, ineffective communications between state officials and society in exposing outrageous cases of political finance have hurt the anticorruption effort and weakened public trust in the system of regulation (Galabov 2003). Punishment for violators. Another important component of an effective party funding regulation is, as in any other public policy instrument, a system of sanctions for noncompliance. Most policy

Party and Election Finance

73

experts worldwide concur that, in order to work, a law needs teeth; the case of political finance regulation is no exception. What provisions for punishment of violations do is increase the costs for ignoring the law (i.e., if caught, parties and bribers should expect to pay the price of not obeying). These sanctions can vary in kind and severity, ranging from loss of registration to loss of eligibility to run in elections and financial penalties. Choosing a combination of these measures may be difficult in nascent democracies that have to find the right balance between political and economic equality, older and younger parties, and more active and less intrusive engagement of the state. These decisions may work out differently depending on other aspects of the socioeconomic environment. The Economic Context The economic arena imposes its own conditions on the way that parties raise money and on how regulations constrain corruptive exchanges. Like elsewhere in the world, in Eastern Europe the costs of electoral campaigns have increased dramatically. Under the pressure of ever-growing prices of television broadcasts, radio time, and billboard ads, politicians have been trying to find more creative ways of attracting donor money, even if of suspicious origin or given against a promise for future favors. The post-1989 party finance experience in Eastern Europe has gone through three main stages with regard to the economic opportunities for illicit enrichment of parties and candidates. These distinct periods were shaped much by the course of transition to a market economy. The first two are transition specific while the third resembles, in form and structure, corrupt party finance elsewhere. In the early 1990s, public resources were siphoned off to private groupings and the latter then supported the parties that let them prey on the state. This process was facilitated by the liberalization of prices, delays in the implementation of structural reforms, and the absence of stringent budgetary requirements (Pashev, Dyulgerov, and Kaschiev 2006). In the mid- to late 1990s, large-scale privatization created opportunities for more corrupt transactions and brought extra money to party coffers in return for privileged treatment of particular businesses at sell-offs. During this same period of time customs corruption flourished, pouring money “up to the chain of government all the way to party treasuries” (Pashev, Dyulgerov, and Kaschiev 2006, 12).

74

Political Corruption in Eastern Europe

The more recent phase is closer in structure to the classical rentseeking model. Party finance benefits the most from public procurement, a phenomenon that has taken various forms. What private contributors pay to parties in this exchange is an investment for the future (Ganev 2007) or a “membership fee for the club of those who stand close to the government” (Pashev, Dyulgerov, and Kaschiev 2006, 11); parties assure such donors with a government contract in return. At the third phase, corrupt party finance has become more risky because this form of exchange depends on election results and regulations of funding and disclosure have been tightened. The Cultural Context Aside from institutional and economic structures, can failures to contain corrupt public funding be attributed to cultural specifics as well? According to a culturalist approach, different national traditions and value systems establish distinct thresholds of tolerance of illegal finance (Treisman 2000). Thus, regardless of the complexity of formal regulations, corrupt party funding is lower where respect for the law and the “public spiritedness” of officials increase the moral costs of illicit enrichment (Della Porta and Vannucci 2005, 16). Within the group of the postcommunist countries, there also may be variations in the way people identify and condemn political corruption. If such distinct attitudes are proven to exist, can we expect to find their roots in history, religion, and ideology? All postcommunist countries in Eastern Europe have gone through a period of totalitarian government but in the more distant past some of them were part of the Ottoman Empire, others belonged to the Habsburg Empire, and a third group was part of Russia or the USSR. At present, some of these nations are predominantly Catholic, others are Eastern Orthodox, and still others include Muslim populations. If communism did not erase unique national value systems bred by history and religion, then it should be possible to find evidence for the importance of culture in the spread or containment of corrupt party finance practices. This possibility lessens when one takes into account extant research that confirms a different extent of corruptiveness only in the case of Protestantism and British colonialism (Treisman 2000). Yet some Eastern European observers note crossregional patterns in the intensity of corruption consistent with historical and religious differences (Earle 2000). Others suggest “grouping”

Party and Election Finance

75

by subregions with regard to patterns rather than levels of corruption, although campaign funding is not at the center of their analyses (James H. Anderson and Gray 2006, 67). At this point, I prefer to keep the issue of a possible impact of the cultural context on party finance open as an empirical question that deserves to be addressed with data.

Party Finance Regimes in Postcommunist States The importance of the first studies on the role of party finance rules in Eastern Europe is in identifying the institutional choices made by postcommunist elites (van Biezen and Kopecki 2001; Ikstens, Smilov, and Walecki 2001). This research also begins to recognize politicians’ intentions to deal with the inequality in the supply of resources for political competition and the signs of abuse or unintended effects. For a more thorough investigation of the possible links between party finance rules and corrupt practices, one needs to (1) focus on the concrete mechanisms designed to limit and prevent irregular raising or spending of money; and (2) assess the resulting impact systematically across space and time. With the adoption of new election laws in the 1990s, finance regulation instruments were also considered in order to enhance proper functioning of the emerging Eastern European multiparty competitive systems. The urgency of passing such legislation, however, was contested due to controversial views about the expected effects of early regulation. On the one hand, a policy instrument for authorizing contribution amounts, their origin, and forms of spending and accounting sets legal parameters for party funding. By determining what is allowed and what is not, finance regimes were intended to facilitate the development of parties as viable and publicly responsible organizations. On the other hand, because of an unequal start in the post1989 political context, opposition parties expected to be hurt more by finance restrictions than did the communists. The latter had an enormous advantage in accessing state resources and financial security, especially at the start of the democratic transition (Walecki 2005). This circumstance explains to a great extent the delay in introducing campaign finance rules, at least in the Czech Republic and Hungary where the anticommunist forces won the first free elections (van Biezen and Kopecky 2001). The deliberate postponement of legisla-

76

Political Corruption in Eastern Europe

tion on party finances was supposed to give the opposition some time to breathe with no restrictions, mainly on foreign aid, and to enable it to compete against the former Communist Party. After the introduction of the first rules, multiple revisions were aimed at keeping campaign funding clean and at preventing parties from falling into dependency on large donors and criminal groups. Many countries chose to allocate certain amounts of state resources to parties as organizations directly involved in governance; in some cases, those were extended later to cover a few nonparliamentary organizations. Restrictions on foreign and anonymous donations and mandatory reporting of sources and amounts were put into place to ensure accountability and to protect the public interest. Assessments of the power of these regulations vary. Some authors believe that the party funding process is still corrupt and that the provisions should be tighter and punishments for violators more severe (Ikstens, Smilov, and Walecki 2002, 9–10; Walecki 2003, 87). According to others, the problem in new democracies is not the lack of or insufficient legislation, but the unwillingness of the elites to play by the rules (van Biezen and Kopecki 2001; Walecki 2005). This more skeptical view is reinforced by those who blame past legacies for their disregard of official rules or who suggest explanations on cultural grounds (Walecki 2003, 89). Table 4.1 summarizes some of the mechanisms available for party finance and control from the mid-1990s until 2005.2 It becomes clear that, when choosing between more libertarian (less involvement of the state) and more interventionist (active state regulation) models, most postcommunist countries followed the example of continental Western Europe and introduced systems combining public resources with public control (Pinto-Duschinsky 2002). State funding was initially introduced to allow access to public resources for noncommunist parties. While some countries opted for regular subventions, others chose to reimburse parties for campaign expenses incurred only in election years. Later on, subventions became a preferred policy tool intended to enable parties to maintain organizational life without falling into the money for favors trap. A vast majority of the fifteen countries included in Table 4.1 passed legislation for regular state subventions for parties in the 1990s; Russia adopted public funding more recently, while Latvia and Ukraine still do not provide for such. As Sarah Birch (2003, 34) observes, Central European states have chosen more generous public

Table 4.1 Party Finance Regimes Country

State Subsidies (% vote hurdle)

Albania 1997 Albania 2001 Bulgaria 1997 Bulgaria 2001 Croatia 1995 Croatia 2000 Czech Rep. 1995 Czech Rep. 2000 Estonia 1996 Estonia 2004 Hungary 1994– Latvia 1995– Lithuania 1999– Macedonia 1994 Poland 1997– Romania 1996 Romania 2003 Russia 1993 Russia 2001

Slovakia 1994 Slovenia 1994

Yes (2%) Yes (2.5%) Yes (4%) Yes (1.4%) Yes (5%) Yes (5%) Yes (3%) Yes (1.5%) Yes (5%) Yes (1.5%) Yes (1%) No Yes (3%) Yes (3%) Yes (3%) Yes (2%) Yes (2%) No Yesa (3%) Yes (3%) Yes (3.2%)

Donations

Spending

Public Disclosure

Sanctions (highest)

No

No

No

Fines

No

No

No

Fines

Yes

Yes

No

Confiscation

Yes

Yes

Yes

Loss of subsidy

No

No

Yes

Fines

No

No

Yes

Fines

No

No

Yes

Loss of subsidy

No

No

Yes

Loss of subsidy

No

No

Yes

Fines

No

No

Yes

Fines

No

No

Yes

Confiscation

No No

No Yes

Yes Yes

Confiscation Fines

Yes

No

No

Yes

Yes

Yes

Fines; loss of seats Loss of subsidy

Yes

No

No

Confiscation

Yes

Yes

Yes

Loss of seats

Yes Yes

Yes Yes

No Yes

No

Yesb

No

None Revoke candidate registration Fines

Yes

No

Noc

Loss of subsidy

(continues)

78

Political Corruption in Eastern Europe

Table 4.1 continued Country Slovenia 2000 Ukraine 1992 Ukraine 2001

State Subsidies (% vote hurdle) Yes (1%) No No

Donations

Spending

Public Disclosure

Sanctions (highest)

Yes

No

Yes

Loss of subsidy

No Yes

No Yes

Noc Noc

Confiscation Confiscation

Sources: Janis Ikstens, Daniel Smilov, and Marcin Walecki, “Campaign Finance in Central and Eastern Europe: Lessons Learned and Challenges Ahead” (International Foundation for Election Systems, 2002), www.ifes.org, accessed 30 November 2009; Steven D. Roper, “The Influence of Romanian Campaign Finance Laws on Party System Development and Corruption,” Party Politics 8, no. 2 (2002): 175–192; Sarah Birch, Electoral Systems and Political Transformation in Post-Communist Europe (London: Palgrave Macmillan, 2003); Monitoring Election Campaign Finance, Open Society Justice Initiative (New York: Open Society Institute, 2005), www.soros.org /initiatives/justice/articles_publications/publications/monitoring_20041123/Handbook _in_full.pdf; Alan Sikk, “Party Financing Regime in Estonia,” paper presented at the conference “Campaign Finance and Party System Development in Eastern Europe,” Riga, Latvia, October 2005; Ingrida Unikaite, “The Effect of Party and Campaign Finance on Lithuanian Party Development,” paper presented at the conference “Campaign Finance and Party System Development in Eastern Europe,” Riga, Latvia, October 2005; Michael Perottino, Marek Chmaj, Marian Walecki, Jaroslaw Zbieranek, Adrian Moraru, Elena Iorga, and Ihor Sheviliakov, Legislation and Control Mechanisms of Political Parties’ Funding: Czech Republic, Poland, Romania, and Ukraine (Bucharest: Institut Pentru Politici Publice, 2005); Daniel Smilov and Jurij Toplak, eds., Political Finance and Corruption in Eastern Europe: The Transition Period (Aldershot, UK: Ashgate, 2007). Notes: a. Started in 2004. b. Very low. c. Required, but not well specified.

subsidizing of party organizations while some former Soviet republics have stuck to more private-oriented finance models. We can also see that party access to regular state subsidizing varies across countries in Eastern Europe: in some places, only parliamentary parties are entitled to it; in others, lower thresholds of eligibility allow for smaller parties to also receive some public financial support. If state funding has a potential to prevent parties from corrupt fund-raising, then one would expect to see lower levels of malfeasance in countries that do provide regular subsidies and where support from the state is more generous.

Party and Election Finance

79

Other measures equalizing access to the electoral market space establish caps on fund-raising and expenditures. Again, because of the initial unequal distribution of resources among political parties, rationing of individual contributions and limits on spending during campaigns were imposed at various levels. Comparing across columns in Table 4.1, we can see that preferences for constraining caps have been less popular than the idea of enabling funding through public money. Seven of the fifteen countries included in the analysis introduced limitations on the amount of private contributions, and six chose to put a ceiling on campaign expenditures. Lithuania and Slovakia imposed a cap on the size of donations only while Macedonia constrained just spending. If finance limitations do contribute to clean party finances, as theory predicts, we should be able to see that countries that use such policy instruments also have lower levels of corruption. To ensure compliance with the law and effective implementation of any policy, a system of monitoring and sanctions for violations is of crucial importance. In Eastern Europe, the enabling mechanisms for party finance precede in time the adoption of precise control mechanisms; this gave the earlier established parties advantage over newcomers in building viable organizational structures; older parties financed their structures when restrictions had not yet been imposed. Across the region, finance regulation has gradually moved toward adoption of stricter requirements for regular annual and timely reporting on finances, origin of revenues and contributions, and ways of spending. To ensure transparency and increase public knowledge about politicians’ resources, countries (Albania, Bulgaria, Hungary, and Slovenia) have empowered auditing bodies to disseminate party financial information. If complete disclosure of party accounts to the public results in a more transparent environment where parties worry that they might be sanctioned for breaching the law, fear that voters might withdraw their support, and know that their opponents’ finances are also exposed, corruption levels should be much lower. Sanctions for noncompliance exist in much of the adopted legislation, but concerns about their effectiveness and about the entire monitoring process also mount. Punishments are considered to be important and they are included in almost any law on party organization and election campaigning. As public policy research has shown, however, these tools may meet the desired outcome only if carefully designed and successfully enforced. The experience of Poland helps

80

Political Corruption in Eastern Europe

to identify administrative fines and possible cuts in reimbursement of campaign expenses among the more effective sanctioning mechanisms (Walecki 2005). Other states have preferred bans on running in elections and on party registration, and even loss of state subsidies for a certain period of time. Thus, hypothetically, more stringent penalty mechanisms will result in weaker incentives for corrupt party finance. Finally, the control mechanisms still may be ineffective if fully independent enforcement bodies are missing. Independent monitoring turns into a key condition for the success of the entire funding system and its ability to prevent corruption in party finance. A financially and institutionally independent agency better serves the public interest because it is enabled to exercise fair and objective control, check and track down the flow of money to party coffers, and oversee financial management. While some postcommunist states have made progress in establishing relatively independent institutions for monitoring, others are either lacking such structures or have created enforcement bodies with no power to expose those who evade the rules. Next, I explore to what extent these enabling and controlling instruments have facilitated fairer and cleaner funding in practice.

Has Finance Regulation Been Effective? Two main goals of finance regulation are to assure parties’ (1) corruption-free fund-raising and spending; and (2) equal access to the electoral process. Anecdotal evidence from new democracies reveals that parties extract significant material support through illegal mechanisms such as kickbacks from suspicious businesses and payments from firms to party-linked organizations for undelivered services (Van Biezen and Kopecky 2001). Another widely practiced form of undocumented contributions, also difficult to track down, has been the coverage of campaign transport expenses incurred by parties (Aleksenia Dimitrova and Stefanov 2004).3 Scandals around privatization, patronage appointments, and public procurement are often associated with parties and their appetites for money (Appel 2001; Treneska 2007; Walecki 2007). Large groups of businesspeople, 42 percent in Bulgaria and 35 percent in Latvia, blame private contributions to parties as the source of their failure or success (Hellman, Jones, and Kaufmann 2000). At the same

Party and Election Finance

81

time, however, just 4 percent in Hungary and 6 percent in the Czech Republic attribute their performance to illegal contributions. What emerges then is a picture of presumably corrupt exchanges of funds for political favors, unevenly scattered across countries. Do the introduced regulations account for this variation? Have they played their expected roles? We need to relate the rules and the resulting practices in a systematic way in order to assess the impact proposed earlier with real-world data. Measuring corruption in party finance is as difficult as measuring any other type of corruption. The problem again is the secret character of the illegal exchange, which makes it unmanageable to observe and gauge its magnitude. What can be observed, however, is the frequency and the importance of instances in which private companies pay in order to be treated preferentially or even to receive services to which they are entitled by law. The Business Environment and Enterprise Performance Survey (BEEPS, 1999), a joint project of the World Bank and the Office of the Chief Economist at the European Bank for Reconstruction and Development (EBRD), offers such data. Among the questions asked in the survey is one that reflects the impact of corrupt contributions to political parties by private interests on the business performance of firms. Assuming that more illegal party enrichment results in stronger effects on companies’ profits, we can consider answers to this question as a valid measure of corrupt party finance. Table 4.2 offers a summary of the responses by country. Interestingly, the impact of party funding through agreed corrupt exchange or extortion varies across countries within quite a broad range. It had no impact (an indicator of cleaner practices) on almost 90 percent of the businesses in Hungary and the Czech Republic, but on as low as onethird of the businesses in Latvia. Also, illegal contributions influenced over one-fifth of the firms in Bulgaria while just about 2 percent were affected in Hungary. These differences are quite significant against the common socioeconomic and political background inherited from communism. Before I look at the proposed role of institutions, I seek to discern differences across countries that could be attributed to culture. If religion matters as a cultural factor generating distinct values, the main difference would be along the main religious divide in the region, Catholic (plus Lutheran) and other nations. Indeed, on average, corrupt party finance appears to be more serious in the latter group,

82

Political Corruption in Eastern Europe

Table 4.2 Corrupt Contributions to Political Parties by Private Companies Impact (%) Country

None

Minor

Significant

Very Significant

N

22.4 7.9 28.7 6.5 31.1 18.1 16.3 23.8 22 19.6

25.2 2.2 13 1.9 23.5 7.2 6.4 14.3 6.4 11.1

4.7 3.4 3.7 1.9 11.8 6 3.6 6 4.6 5.1

107 89 108 108 119 83 172 84 109 109

6.2 10.3 15.8 14.1 21.1 16.4 14

16.5 21.8 7.4 16.9 13 20.2 16

8.2 20.5 7.4 9.9 10.7 8.7 10.9

97 78 95 71 384 183 151

Catholic or Lutheran Croatia Czech Rep. Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Average

47.7 88.5 54.6 89.8 33.6 68.7 73.8 56 67 64.4

Eastern Orthodox or Muslim Albania Bulgaria Macedonia Romania Russia Ukraine Average

69.1 47.4 69.5 59.2 55.2 54.6 59.2

Source: 1999 Business Environment and Enterprise Performance Survey, World Bank, 2000, www.ebrd.com/pages/research/analysis/surveys/beeps.shtml. Note: Entries are answers of firms to the question: “To what extent have contributions by private interests to political parties and election campaigns had an impact on your business?” (Q68).

which includes predominantly Eastern Orthodox and Muslim (Albania) countries. These also share a common past of being part of the Ottoman or Russian empires and of less experience with parliamentary democracy before the communist governments took over. Which regulatory mechanisms, if any, have helped contain corrupt party finance in some countries in Eastern Europe? To recall from above, institutional designers argue that state subsidies, contribution and spending caps, public disclosure, and penalties for violation of rules may reduce the party leadership’s reliance on illegal sources. I began the analysis by calculating correlations of institutional traits with levels of corrupt party finance (BEEPS) and with

Party and Election Finance

83

Table 4.3 Correlations of Corrupt Party Funding and Related Regulations

Regulatory Tool Regular state subsidy Contribution cap Spending cap Full public disclosure Highest sanctionsa

Corrupt Private Contributions BEEPS, N = 14 (high = corrupt) –.251 .128 .060 –.413 –.468

Corruption TICPI, N = 119 (high = clean)

Corruption WBI, N = 165 (high = clean)

.450*** –.255*** –.239*** .586*** .269***

.393*** –.181** –.108 .516*** .321***

Notes: Values in italics are coefficients with correct sign. BEEPS = Business Environment and Enterprise Performance Survey; TICPI = Transparency International Corruption Perceptions Index; WBI = World Bank Index. Entries are bivariate correlations. ** p < .05; *** p < .01; two-tailed tests. a. Highest level of sanctions provided for noncompliance with rules is a continuous variable coded: 0 = no penalties, 1 = fines, 2 = confiscation of illegitimate donations, 3 = revoked candidate registration, 4 = loss of seats, and 5 = loss of subsidy.

levels of general corruption (TICPI and WBI) in the fifteen postcommunist states.4 The bivariate coefficients reported in Table 4.3 suggest that party finance is less corrupt in political systems that offer regular state subsidies.5 All entrees in the top row carry the correct sign and two of them (for TICPI and WBI) are statistically significant. Requirements for full public disclosure of party accounts and sources of funding are also associated with cleaner practices of fund-raising, as shown by the estimated coefficients reported in the fourth row. Moreover, the results are statistically significant for two corruption indicators (TICPI and WBI). Finally, the threat of sanctions against violations seems to work: the findings in the last row reveal that, in countries that provide more severe punishments for illegal financing of parties, corruption is less pronounced. These findings are stable: conclusive for the TICPI and the WBI indicators where more observations are available over time and approaching statistical significance for the BEEPS with just fourteen cases. The two regulations not supported by the data are those that set limits on expenditures and on donations. What this suggests is that such mechanisms may not be an effective remedy against malfeasance in the area of party finance.6 Limiting parties in their ability to

84

Political Corruption in Eastern Europe

raise or spend needed money would most likely not stop them in finding illegal ways to breach the amount allowed. Perhaps those caps are not set realistically (i.e., without paying attention to the rising costs of election campaigning). Another reason can be found in the competitive nature of politics itself. In other words, regardless of what is allowed, party leaders in Eastern Europe often feel tempted to break the rules and outfund and outspend their rivals.

Consequences of Corrupt Party Finance in Eastern Europe Even if some of the institutional mechanisms for party finance regulation limit incentives for corrupt exchange, as suggested above, they certainly do not have the capacity to root out malfeasance entirely. Concerns that parties continue to use illegal sources and amounts of funding in return for favors, appointments, and public contracts are growing stronger than ever in the postcommunist world. As imperfect as current measures of corrupt party finance may be, they point at the fact that malfeasance exists to a degree significant enough to cause serious repercussions for the democratic process in nascent democracies. A thorough analysis of party corruptiveness should not ignore the consequences of illegal political money, and I next explore some of these effects. Effects on Parties: Cartelization? The ways that politicians raise support for their organizational needs and election campaigns may first of all affect parties themselves. Accepting illegal contributions may disproportionately increase the role of certain party functionaries and make the nomination of candidates in elections nontransparent. Improper fund-raising may hurt the internal democratization of parties by subjecting the principle of clear and fair procedures to the need for resources. In addition, finance regulations can have serious consequences for the development of the party system. Serious concerns exist about the provision of state subsidies. There are two contradictory views on the role of public funding: On the one hand, subsidies from the state are expected to reduce the dependence of parties on corporations and self-interested individuals;

Party and Election Finance

85

on the other hand, access for some parties to state resources leads to cartelization of the system by providing subsidy recipients with an advantage, thus reducing competition. As shown in Table 4.1, most Eastern European countries made the decision to allocate regular state subsidies. And the results of analysis reported in Table 4.3 offer some evidence for cleaner policies in countries with state funding. A question remains: has the provision of public money led to the emergence of established parties who share mutual interest to sustain the group’s and limit others’ access to state resources? If this is the case, one may expect weaker competition among parties within the cartel and reduced opportunities for new parties to win elections. The experience of Eastern Europe reveals a mixed picture with some signs of more open competition in countries with no state funding and a freezing effect where subsidies are distributed. Among the few cases where state funding is not available, Latvia has shown different party composition of the Saeima with each new electoral cycle: parties formed at independence disappeared and others easily replaced them (Pabriks and Stokenberga 2006, 62). By comparison a survey of states providing public funding shows that in some instances it looks like state funding did exercise a freezing effect on the party system for a while, but then newcomers broke through by winning significant representation in parliament. In Poland the liberal-conservative Civic Platform won sixty-five parliamentary seats in 2001, in Estonia the conservative Res Publica unexpectedly received 35 percent of the seats in the Riigikogu in 2003, and in Bulgaria the populist National Movement Simeon II (NMSS) emerged as first power in the national assembly in 2001 by winning half of the mandates. While Res Publica drew on significant private resources to compensate for the inequality in state money distribution (Sikk 2003), the NMSS won its victory with limited funds (and a popular message for clean policies). At the same time, in Slovenia public subsidies have helped the same several parties win parliamentary seats since 1992 and the Liberal Democratic Party stay in power until 2004. Corruption levels in Slovenia are the lowest among all postcommunist states, but many worry that the process of etatization of parties, facilitated by state money, alienates citizens from politics (Toplak 2007, 187). Overall, subsidies have been desired by parties in Eastern Europe, but their potential to stop newcomers from winning remains under question. Subsidies are aspired to because the costs of election

86

Political Corruption in Eastern Europe

campaigning increase exponentially every year and parties with small or contracting memberships cannot rely on funding of their own. Statistics from many countries in the region show that the relative weight of public funding is considerable, reaching 90 percent in some party accounts (Jungerstam-Mulders 2006). Yet state subsidies allocated in postcommunist countries are not sufficient to prevent parties from turning to big business, often in violation of existing laws and by engaging in corrupt networks. Even in Slovenia where subsidies are the most generous, parties in power amend the law to support themselves even better (Toplak 2007, 186–187). High volatility in voter preferences allows for new parties to form and successfully challenge those blessed by public funding. Often, newcomers win by employing anticorruption rhetoric to run against the party establishment. Effects on the State: State Capture From a public policy point of view, perhaps the most malignant impact of corrupt party finance is the phenomenon of state capture. Previous research defines state capture as “shaping the formation of the basic rules of the game (i.e., laws, rules, decrees and regulations) through illicit and nontransparent private payments to public officials” (Hellman, Jones, and Kaufmann 2000, 2, emphasis in original). Identified in this way, the concept reflects most of all illicit contributions to those who hold, or are expected to hold, positions with the power to adopt and enforce laws. Since parties are the main actors in the competition for public office in modern democratic states, they are the most likely target (or initiator) of corrupt exchange. In this exchange parties in power, which are the recipients of extra and more generous nontransparent donations, deliver. They do this through policymaking decisions that are inefficient for the state and beneficial to particular private interests. Given the scope of economic and political transformation in Eastern Europe after 1989, one can only begin to comprehend the extreme variety of opportunities for parties to supply their donors with services. By the mid-1990s, rules and procedures related to privatization were the ones that parties could tailor to accommodate specific interests. Later, decisions over distribution of private contracts and EU funds became valuable currency for satisfying and attracting

Party and Election Finance

87

large party donors. By purchasing influence over legislation and policymaking, a small group of rent extractors could gain control of the state in a process that remains hidden from the broader public. Parties that are being secretly paid could fail their most important function, representing their constituents. Concrete examples of buying political influence from parties abound, despite the secrecy surrounding these deals. In 2002, a Polish media entrepreneur solicited a US$17.5 million bribe to lobby the Sejm for a decision in favor of Agora, a publishing house trying to buy the large television company Polsat (Walecki 2007). The bribe money was intended for the Democratic Left Alliance currently governing the country. In Russia and Ukraine, a semiofficial market of legislative services operates with mainstream media publishing the prices of various services. Newly elected politicians to the Ukrainian Rada are being offered between $20,000 and $50,000 to join a parliamentary faction; recently, the Hromada parliamentary group enlarged by ninety such members (Protsyk and Walecki 2007, 205). According to some accounts, prior to 2005, Russian members of the State Duma sold their votes on ordinary bills for $5,000 to $7,000 (Baker and Glasser 2005, 83) while voting on impeachment was as expensive as $10,000 and placing a written question to a minister as cheap as $5,000 (Gleisner 2007, 153). Duma seats could be purchased on all party ballots, according to a party official who did not feel uncomfortable talking about the “golden price list,” including $3 million to $4 million for a winnable spot on the Communist Party PR list and $1 million on the regional list of another party (Baker and Glasser 2005, 281). Other stunning cases of buying political influence can be found in the Balkans. In Croatia a written contract was signed by the chairman and the secretary of the Democratic Centre Party and a contributor, who was promised a mandate in the Sabor against a donation of 250,000 euros (Kregar, Gardasevic, and Gotovac 2007, 61). The public learned about the deal after the briber did not win the seat in the election and openly asked for his money back. Anecdotal evidence from Bulgaria points to the penetration of corporate influence in the formation of the governing cabinet in 2005. Tatiana Dontcheva, an outspoken Bulgarian Socialist Party (BSP) MP, admits in an interview following revelations about connections of high-level government officials to suspicious businesspeople sponsors of the party that

88

Political Corruption in Eastern Europe

the selection of a new minister is a prime minister’s job and, most of all, a question of “balance among the business groups which determined the composition of this government” (Neikov 2008). State capture is further manifested through party patronage, defined as “the use of public resources in particularistic and direct exchanges between clients and party politicians or party functionaries” (Muller 2006, 189). Parties pay the political rent by appointing representatives of private businesses, large contributors to party coffers, to high-level or (too) well-paid jobs in the public sector; by shifting privatization decisions to the advantage of their sponsors; by nontransparent distribution of government contracts; or by protection from prosecution. Examples of party hires are not difficult to find. Such appointments have become an established practice in Poland since the beginning of transition: as many as 200,000 positions of state-owned companies’ executive board members, department heads, and managers of municipal hospitals and heads of local service bureaus have been used by parties to reward loyalty and financial support (Szczerbiak 2006, 113). Party finance generated from public procurement has grown remarkably in size and variety after the completion of privatization in the 1990s (Pashev, Dyulgerov, and Kaschiev 2006, 13). These unfair payoffs affect the business sector throughout Eastern Europe: more in Bulgaria and Romania (43 percent and 42 percent of the firms, respectively), and less in Hungary and Slovenia (6 percent and 11 percent, respectively) (BEEPS 1999). Daily newspapers and other media report on a regular basis violations of government contracting to the advantage of companies helping the ruling parties. In Croatia, Miomir Zuzul, vice prime minister, was forced to resign after lobbying in favor of the Bechtel construction company, a sponsor of his party, to receive a public contract without competition (Kregar, Gardasevic, and Gotovac 2007, 61). In both Croatia and Macedonia, a well-established practice existed in the 1990s for parties to have their own companies that were rewarded large shares in the process of privatization. The Croatian Domovina d.d., closely associated with the Croatian Democratic Union (CDU), made huge profits and helped the CDU through illegal transfers of funds (Kregar, Gardasevic, and Gotovac 2007, 56–57). The ruling Internal Macedonian Revolutionary Organization (IMRO) in Macedonia bought state firms for more than 1.5 million euros through privatization and favored them by providing privileged positions in the emerging market. After

Party and Election Finance

89

the Constitutional Court ruled that the IMRO had violated the law, the party “sold” the stocks to loyal supporters who started donating back revenues from the business (Treneska 2007, 110). Finally, party protection of businesses with suspicious revenues may affect the operations of local government. In Bulgaria, Plamen Galev and Angel Hristov, the famous Galevi brothers and sponsors of the BSP, in practice rule the southwestern town of Dupnitza. In the words of a former mayor of the town, local public officials failed in the battle over who governs the municipality because the central leadership of the BSP represented by the minister of interior was holding an “umbrella” of political protection over their private sponsors (Simov 2008). The losses sustained by the state in the Eastern European transitional environment are far-reaching: Party patronage has decreased the quality of public service and, at the same time, led to the expansion of public spending without improving or extending available social programs. Effects on Voters: Alienation Despite the adoption of measures intended to increase transparency and punish violations in party finance, the Eastern European citizens are becoming increasingly aware that parties do not fully comply with the limitations for fund-raising. It has become public knowledge that party elites are supported financially by private corporate interests and that this support is often illegal. The threat is quite real now that voters may lose trust in parties when they learn about politicians’ failures to put society’s demands first. Citizens’ confidence in institutions in any democracy depends on the performance of these institutions, but the citizens of postcommunist states are especially sensitive to institutions. The latter are expected to provide what people were deprived of during communism: free expression and respect for the rule of law (Mishler and Rose 1997). In Eastern Europe, as in many other countries in the world, political parties rank as the most distrusted institutions. Abundant evidence for this claim comes from various sources, including the Transparency International Global Corruption Barometer and the Eurobarometer surveys.7 That illegal financing is driving popular distrust is confirmed by the fact that so many Eastern Europeans feel an urgency to clean up corruption in the way that parties operate. When asked “If you had a magic wand and you could eliminate corruption from one of the fol-

90

Political Corruption in Eastern Europe

lowing institutions, what would your first choice be?” the largest group (20.2 percent in Bulgaria, 27.2 percent in Poland, 24.3 percent in Romania, and 24.5 percent in Russia) of respondents to the Transparency International Global Corruption Barometer (2003) survey chose political parties. These highly unfavorable perceptions of political parties may not necessarily coincide with the extent of actual illegal financing, but the levels of frustration continue to rise and a gap widens between public expectations for change and political elites’ lack of will to enforce the rules. Requirements for finance disclosure may have had some anticorruption impacts, but these are far from being fully adhered to. Observers note that, where disclosure is required by law, many reports are submitted after the deadline, not all parties file such reports, or the accounting documentation is not complete (Smilov and Toplak 2007). The absence of reliable information and the resistance of leaders to disclose organizational finances add to the already existing high level of mistrust among the population toward political parties. Disappointment with party politics hurts the development of healthy connections with voters. Party membership rates decrease as they do in the West, but in Eastern Europe they are as low as 5 percent (Lithuania and Slovenia), 1.5 percent (Poland), and 2 percent (Hungary) of the electorate (Jungerstam-Mulders 2006). As party financial reports show, membership fees are of little importance for financing the organizational needs of parties. In a vicious circle, private contributions and state subsidies come to compensate for a shrinking membership, but they also weaken parties’ incentives to attract new members and push away supporters. Beliefs that the parties of the early 1990s betrayed the public have led to the election of populist parties and figures. Three Latvian parties, the People’s Movement for Latvia, Latvia’s Unity Party, and the Democratic Party Saimnieks won big in the 1995 parliamentary election by running campaigns based on rather unrealistic promises (Pabriks and Stokenberga 2006, 59). The populist Self Defense Party in Poland managed to enter the governing coalition in 2005; the Slovak Smer Social Democracy, led by the populist R. Fico, won a large victory in the 2006 parliamentary election, despite the incumbent Slovak Democratic and Christian Union’s strong record of economic reforms (Larrabee 2006). Sizable groups within the electorate are available for mobilization by those who claim they will carry on clean politics and bring a new morality.

Party and Election Finance

91

Conclusion Existing research on campaign finance predicts that the rules governing the way parties raise and spend money are linked to the corruptiveness of a political system (Pinto-Duschinsky 2002; Walecki 2005). In this chapter, I presented the impact of four institutional traits: state subsidies, caps on contributions and expenditures, disclosure requirements, and sanctions for noncompliance. My findings identified some party finance regulations in Eastern Europe as being especially important in cleaning up politics. In particular, my analysis suggests that, in countries using regular subventions, both the process of party fund-raising and corruption in general are less pronounced. This finding is especially interesting given the ongoing debate about whether subsidies have the potential to make parties independent from large self-interested contributors. The data that I analyzed in this chapter give a positive answer to this dilemma. While there are some troubling signs for cartelization of the party systems due to state subsidies, challengers can still succeed. The transitional context, mainly the volatility in voter preferences and the widely used proportional election rules, makes it possible for newcomers to break through and win without access to state funding. Another valuable finding of my analysis underscores the effects of public disclosure of parties’ financial accounts. Where such requirements exist in Eastern Europe, corrupt practices have been at lower rates. This aspect of finance regulation has a special meaning in countries where in the past the Communist Party never revealed its revenues and property to the public. Furthermore, making the release of financial documents mandatory would not have worked if there had not been punishments for violators. My analysis also establishes that what matters is not just the presence of sanctions, but their severity. More stringent regulations in some postcommunist countries have been associated with less corruption, a finding with clear policy implications for designers of party finance legislation. Despite the suggested relevance of legislation to set parameters for party finance and reporting, corrupt practices have not been eliminated and scandals related to political money abound. Previous research notes that the problem may not be so much one of strict provisions as one of insufficient enforcement (Walecki 2005; Warner 2005). Therefore, an analysis of the reasons for the spread of illicit financing should not ignore contextual factors such as the

92

Political Corruption in Eastern Europe

ability of elites to enforce the law and the will of society to outvote violators. At the beginning of the transition, regulation was vague and insufficient: legislation gaps and violations were deliberate to help opposition parties compete against the communists who had a clear advantage in mobilizing resources (van Biezen and Kopecky 2001). Later, the rules and regulations became more detailed and stringent, but evasion in party finance was not eliminated because opportunities for illegal enrichment mushroomed with privatization and procurement. All of these insights suggest that certain regulations can be effective anticorruption tools if properly enforced and that political will may be crucial for their implementation. As the final sections of my analysis reveal, regardless of the effects of institutions, buying of legislative votes to influence policy decisions continues and this may alienate voters from the political process. At stake is the core principle on which parties are formed (i.e., to represent people’s interests). What the mechanisms of corrupt exchange between parties and private interests are and how deep and harmful they may become are what this chapter has not revealed. In order to achieve this, one needs to design a case study and investigate in depth the genesis and growth of hidden, corrupt relationships between the public and the private sectors. In Chapter 5, I examine the emergence and evolution of the friendly circles in Bulgaria, loops of companies around political parties formed as interdependent informal networks.

Notes 1. But see Smilov (2007, 8–9) for a more differentiated approach with regard to the character of the transition across the region. He warns that as of the mid-2000s, elements of “something other than democracy” existed in Russia and Ukraine that were putting at risk the prospect for democratic consolidation. 2. The beginning of the time period covered in this chapter is the mid1990s because corruption indicator scores are not available for the early postcommunist transition years. 3. In this scheme, charges are put on credit cards issued by banks and the banks are later awarded with favorable policy decisions by the elites in power. 4. The BEEPS indicator for party finance is strongly correlated with the two indicators for overall corruption (–0.672 for TICPI and –0.695 for WBI, both significant at the 0.01 level), reflecting the match between perceptions

Party and Election Finance

93

of parties’ improper raising of money and the corruptiveness of the whole political system. 5. While the results reported in Table 4.3 attest to an association between corruption and party law regulations, they are based on correlation analysis and should not be interpreted as evidence for a causal link. More rigorous techniques, such as multivariate regression, confirmed the independent effect of two regulations: state subsidy and higher sanctions. 6. For particular examples of violations of these restrictions, see Ikstens, Smilov, and Walecki (2001) and Walecki (2005). 7. These studies use different scales to compare trust in institutions across countries. The results in all of them concur that parties are the most distrusted political institutions (Transparency International Global Corruption Barometer; Eurobarometer surveys).

5 Networks of Corruption: “Friendly Circles” in Bulgaria

People are easily fooled by talk about donations, private or corporate, that might be the case but not the whole truth, indeed. You see, everywhere from America to Japan, each party has, so to say, a ring of firms. . . . If you think that I am less influential than one banker, your idea of what a politician can achieve is very far from reality. In the last 15 years, perhaps half of all above-the-average businessmen are [what they are] . . . either with my blessing or at least a smile from me. —Ahmed Dogan, 2005

These words of Ahmed Dogan, the leader of the ethnic Turks’ party in Bulgaria, Movement for Rights and Freedoms (MRF), stirred a wave of disgust in the aftermath of the 2005 parliamentary election. It was not because of what Dogan admitted, since most Bulgarians had already developed an idea about the links between politics and business. What made many people angry were the cynicism and sense of impunity with which a party leader would talk before television cameras. Stories about firms and businessmen receiving lucrative procurement deals against generous financial support for party coffers circulated in the Bulgarian media for years. This time, it was the vulgar openness in Dogan’s words that pushed the public debate on corruption to a realization that the phenomenon was not limited to a few isolated cases, but instead widespread and systemic. If each significant party with a chance to come to power had built a network of supporting companies that relied on privileged access to state resources, the role of voters as participants in the 95

96

Political Corruption in Eastern Europe

political process should have been greatly undermined. Indeed, following the 2005 election, a tripartite coalition government was formed in which parties ignored their public campaign promises and shared power in a way that allegedly facilitated service for powerful corporate sponsors. Much of the current conception of how politicians and businesspeople get involved in vicious rings of interdependence during transition is based on the assumption about maximization of gains during times of uncertainty. Parties and firms have mutual interest in maintaining a relationship that enables the former to finance their organizations and expensive campaigns and ensures profits for the latter through public procurement and favorable regulation. In the context of ongoing reforms and incomplete institutionalization, such informal networks may turn out to be extremely beneficial and even may bring a sense of stability. When later challenged over his preelection statement, Dogan extended the metaphor to explain how the circle’s firms support a party: if even one of the rings that hold together a bucket is removed, the vessel falls apart (“Dogan Explained About the Ring” 2006). In this chapter, I analyze the emergence and evolution of interdependent informal political and economic networks in Bulgaria. Also called “friendly circles,” these function on a patron-client basis: political elites in power award particular businesses with public contracts or favorable legislation; the companies, on their part, finance parties beyond what is envisioned by the law. I utilized a case study to explore instances of such circles of companies that formed around Bulgarian political parties by researching press articles, interviews given by politicians, and journalist investigations. While direct evidence for the functioning of the friendly circles is difficult to gather, in the case of Bulgaria there is a rare instance of openly admitted practices of party-firm interdependence. Moreover, as I describe next, in Bulgaria the politics of partial reform of the early 1990s enhanced the formation and sustained the existence of specific business circles around each elite group in power. With the introduction of a functioning market economy, the monopoly of one super-influential group in the public realm broke and multiple corporate actors won access to the rulers. In each instance, the public interest was hurt. The Bulgarian case offers an opportunity to study the origins of these distinct systems of informal connections and the changing structure of competition over state resources after communism.

Networks of Corruption

97

Early Winners and Competitive Captors When it comes to the outcomes of reforms undertaken after communism, recent studies warn that the withdrawal of the state from the market is not a panacea. In fact, it can inflict damage if institutional changes and political representation are ignored. Joel S. Hellman (1998) first revealed how attention to the behavior of the losers from marketization in the early years of transition undermined the greater question of what the first winners did. He argues that, contrary to previous expectations that the beneficiaries from economic transformation push for more reform, it is exactly the immediate winners in the Eastern European transition who shortly after its start became interested in slowing down. In practice, they stalled the reformist effort in the first half of the 1990s. The group of the victorious—“enterprise insiders,” “commercial bankers,” “local officials,” and “Mafiosi” who enriched themselves overnight through privileged access to resources and connection networks from the communist regime—suddenly lost interest in furthering the liberalization agenda (Hellman 1998, 204). Now they expected to multiply profits in the context of delayed reform and postponed institutionalization of property rights, where huge gains could be realized due to market distortions and weakened monitoring capacity of the state. Thus, it was in the interest of the big winners of 1989–1992 to block active reform policies. They had acquired enormous financial and other kinds of potential to influence the state and maintain the status quo of incomplete transition. Whether these early beneficiaries could be constrained, and their impact on policymaking reduced, depended on the capacity of the system to get first-round losers involved in the political system of government. Building on Hellman’s work, Andrew Barnes (2007) suggests that the possible outcomes of the politics of postcommunist economic transition are not necessarily only two (i.e., completion of marketization and full capture of the state by large conglomerates). Barnes (2007, 73–74) distinguishes a third possible path to which he assigns Bulgaria, one where from partial reform a country enters a phase of “competitive capture equilibrium.” Under this scenario no single grouping dominates, but several private actors compete for rents. Full insulation of the state does not happen because multiple captors cultivate ties to different parties to ensure permanent rents regardless of who wins the elections. While there is competition among eco-

98

Political Corruption in Eastern Europe

nomic groups for prevailing in the battle over benefits from the public sector, this market space is not ruled by the principle of transparency and openness. As a result, both parties and their sponsors become increasingly cartelized while government policies are crafted mainly in the interest of a few private actors with an agenda hidden from the public. These two theoretical frameworks are complementary; the latter respecifies the former for certain cases in which Bulgaria may belong. There is no doubt that the country went through the first phase, partial reform equilibrium, which was marked by cronyism, dominant economic actors linked to the previous regime, and plundering of state assets at an unprecedented scale. As Venelin I. Ganev (2006) correctly notes, changes began with the Union of Democratic Forces (UDF) 1997–2001 administration, where there was a behavioral shift among the elite toward allocation of rents on a more competitive basis. This was not the end of corrupt practices, however; rather, state privileges began to be distributed on a higher-offer principle and to a wider range of interests. The public domain of corrupt practices also changed. While state assets available for plunder shrank after years of hidden privatization, the prospect for European integration opened new opportunities for parasitizing, this time over EU funds (Ganev 2006). In the literature, the exact timing of this shift from first-winner domination to a more competitive form of corrupt exchange in Bulgaria remains disputable: 1997 for Ganev (2007) and 2001 for Barnes (2007). It is clear, though, that the political scene and the illegal ties of politicians to the corporate world have changed in a direction toward more pluralism. In the following sections, I explore the main economic groupings that sprang into existence and flourished due to special links to Bulgaria’s government after 1989. To analyze the characteristic features of these friendly circles, I look at the origin of the groupings, their possible connections to the former secret service, the structure of relations and interdependence between the participants, and the fate of their leadership over time. Dividing the postcommunist years into two phases (1989–2001 and 2001 and beyond) is tentative and for the purposes of this analysis only. It is based on preliminary observations of single, easily recognizable, private actors with allegedly close links to the ruling elites until 2001, and of an overcrowded political market space with no dominant players afterward. Focusing on separate cases of informal networks will help to estab-

Networks of Corruption

99

lish the mechanisms of state capture, the extent to which a private group was able to use state agencies for narrow corporate ends, and the weakness of political agents in overcoming dependence through unaggressive efforts to clean up their own camps. Finally, my investigation relates the mode of group dominance, single or multiple, to the size of the governing coalitions. By interweaving characteristic features of economic and political actors, my inquiry into the specifics of the Bulgarian friendly circles suggests answers to important questions. Among these are: Why did the state captors manage to conquer the public sector? And how did their configuration evolve over time?

First- and Second-Round Winners in Ascendance The first winners of the economic transition in Bulgaria include two large conglomerates, Multigrup and Orion, closely linked to the former communist party and the secret service. Both groups possessed excessive economic power that translated into political influence once the factions they were tied to within the Bulgarian Socialist Party obtained control of the executive. When the opposition UDF finally won a majority in the National Assembly in 1997 and formed a cabinet, another single dominant player emerged that benefited disproportionately from its warm connections to the new ruling elite, the Olymp circle. Although similar in regard to their uncontested access to state privilege, the three networks are very different when it comes to origin, mode of corrupt enrichment, and exit from the public arena. Multigrup It is hard to name another group in postcommunist Bulgaria that penetrated state agencies and influenced government decisions as much as Multigrup did. Associated with its growth are the names of prominent politicians, including Andrei Lukanov, a high-level official in the Ministry of Foreign Trade during the previous regime, one of the initiators of the removal of Todor Zhivkov from power on 10 November 1989, and prime minister of Bulgaria from 1990 to 1991. Many people believe that Lukanov was the patron of the group, its brains and strategy maker. Among the other founders is Ognian Doinov, member of the Politbureau of the Communist Party (Barnes 2007). Yet for the public, the leader of Multigrup was Ilia Pavlov, a former wrestler who

100

Political Corruption in Eastern Europe

became a successful entrepreneur in the last years of the previous regime.1 The roots of Multigrup can be traced to the late 1980s, when the communist regime of Zhivkov introduced changes in the banking and foreign trade sectors that allowed well-connected cadres of the ruling party to siphon state resources. Lukanov and Doinov were active participants in these new regulations directed toward economic liberalization, carefully designed to channel gains into the hands of the party leadership and subsequently to ensure survival of both the elite and the party. 2 Lukanov’s party comrades used Multigrup to transfer money out of the country, hoping to be able to later invest it in Bulgaria as private capital (Roth 2008, 251). Through skillful exploitation of loopholes in the government regulations during the first years of transition (when Lukanov was prime minister), Multigrup grew to become the sole, dominant private player in the Bulgarian economy. By the mid-1990s, the conglomerate of firms, corporations, and holdings had accumulated unprecedented wealth through smuggling, money laundering, and trafficking (Barnes 2007). Multigrup managed to find and take advantage of various sources of enrichment in the public sector (Ganev 2007, 106–112).3 The greatest success of Multigrup was realized through its entry into the market of natural gas, a resource delivered to the Bulgarian market exclusively from Russia. Pavlov and Multigrup’s corporate organization exploited this monopolized supply of a much-needed energy resource to the best extent possible. In the quest for obtaining a share from the big pie, the most powerful economic group in the country became an important player in the foreign policy arena. In 1995, a joint venture named Topenergy was founded in which the Russian Gazprom and Bulgarian Bulgargaz each owned half of the shares. The chairman of the Bulgarian Committee on Energy allowed the private entity to purchase shares in Topenergy and appointed Lukanov 4 a member of Topenergy’s board of directors. Zhan Videnov’s BSP government tried to block these orders to prevent the entry of Multigrup into the gas trade, but Gazprom countered by appointing Lukanov representative of the Russian company in the negotiations with the Bulgarian government. The result was devastating for Bulgaria, which had to buy gas at the highest prices in Europe, but profitable for Multigrup, which made US$100 million in one year (Ganev 2007, 105).

Networks of Corruption

101

Retrospectively, Multigrup and Pavlov’s most successful years lasted until the mid-1990s when they lost the previous absolute hegemony to another group of friends. Multigrup never managed to restore its previous dominant position. Lukanov was assassinated in front of his home in 1996, a murder that remains unsolved. According to some, the former prime minister and then Topenergy representative was organizing a coup against Videnov’s cabinet and its favorites, Multigrup’s rivals in the competition for economic influence. Others are inclined to blame Pavlov who, by September 1996, was in conflict with Lukanov over the leadership position in the gas business with Russia.5 After 1997, the new UDF government undertook investigations into the past operations of Multigrup and its relationships with the political elite. Weakened by accusations of involvement in corruption, blackmail, tax evasion, and illegal trafficking, the former giant conglomerate made a final attempt for revival in the early 2000s that ended with Pavlov’s assassination. Orion The circle of friends Orion is associated with the socialist government of Videnov that came to power after winning an absolute majority of the seats in the National Assembly in December 1994. It came to the political scene after a three-year period of parliamentary fragmentation. This was also a time when the dominant Multigrup successfully blocked attempts for economic transformation. Orion was formed in 1994 by a group looking for access to high-level politics with a goal to enter big business (Ganev 2007, 78). They contacted BSP’s young leader to offer support in the upcoming parliamentary election, including arranging preelectoral meetings for him in Germany and Greece to boost the image of Videnov, who was less known abroad. The group consisted of companies named after stars in a famous constellation; hence, the heavenly label of the new economic ring closely tied to the government. The core members of Orion had warm connections to the BSP: in the past they held important positions in the communist secret service (Ganev 2007, 78). 6 In his desire to remain independent from the Lukanov-Multigrup wing within the BSP, Videnov trusted his well-connected friends in Orion and some of them became members of his administration (Barnes 2007). They received powerful posts: Mikhail Danov was appointed director of the

102

Political Corruption in Eastern Europe

Bulgarian Telephone Company and financial manager of BSP’s daily Duma. Krassimir Raidovski was named director of the press center of the Council of Ministers, which was in control of the information flow between the prime minister and the media. Lyubomir Kolarov, member of the BSP Supreme Council, became chairman of the Committee on Mail and Long-Distance Communications. Another influential Orion member, Zlatimir Orsov, was a BSP deputy who served on the parliamentary Committee on Agriculture where he played a key role in promoting policies favoring the economic interests of the circle. Perhaps the most elaborate scheme used by Orion to accumulate huge profits by exploiting its warm relations with the cabinet was the affair with the Bulgarian Agricultural and Industrial Bank (BAIB). The BAIB was founded by the group in October 1994 and the money necessary for its registration with the court was collected through loans from the State Savings Bank (the second-largest state-owned bank in the country). In 1995 after the Videnov cabinet was formed, the bank undertook an aggressive fund-raising campaign. The funds were supposed to be collected through an organization of the peasants who wanted to participate in collective farms, the Union of the Agricultural Cooperatives in Bulgaria (UACB). What appeared to be a grassroots movement in support of agricultural cooperatives was skillfully manipulated by the BAIB. Its leadership pretended to be representing the socialist government and called on all UACB members to deposit their money and privatization vouchers in the bank. The idea was to put all resources together into investment funds and start state enterprises for production of various agricultural commodities. Large amounts of money were gathered because the peasants thought that the plan was a good opportunity for them. Where doubts arose, threats were used to expropriate the money (Ganev 2007, 81).7 Despite all the work done, an investigation into the origin of BAIB’s start-up money found irregularities in the registration of the bank; the supreme court confirmed the findings; and arrests of top BAIB officials were ordered. It was too late, however; some of the most prominent members of Orion had emigrated, together with the money deposited by the peasants in the BAIB. But not only did the bank collapse; the entire Bulgarian economy was moving toward a catastrophe. By the fall of 1996, Bulgaria fell into a deep financial and economic crisis, and Videnov’s government proved incapable of revers-

Networks of Corruption

103

ing the rapid deterioration of living conditions in the country. Banks that in the early 1990s had created the millionaires and continued to give bad loans to inefficient state enterprises were closing down. The state was no longer in a position to help them out: inflation was skyrocketing and the threat of a hungry winter soon turned into a reality. Orion brought its own contribution to the crisis; taking advantage of its close relations to the government, its companies realized huge profits from export of wheat. Even when it became clear that the grain supply for the domestic market had been jeopardized, ministers from the government used loopholes in the regulations limiting agricultural exports in order to salvage the last opportunities for plunder for “their” firms.8 Orion’s Raidovski and Rumen Spassov were co-shareholders in Betelheise Commerce, one of the largest exporters of grain.9 Others owned shares in Popular Bank Lovech, Agrocommerce, and Rigel-Commerce. A complex but complete circle of relationships took shape: bank capital, control over the agricultural production, and political influence. Through the latter, Orion received from the state a subsidy for every ton of grain bought from farmers and 30 percent reimbursement for transportation costs. The friendly firms also received from the government cash for the storage of grain at warehouses, even though in reality most of the wheat left the country.10 Videnov and his cabinet, the political pillar of Orion, came under attack from multiple sides. Chief prosecutor Ivan Tatarchev, investigating possible violations of the law and illegal transactions in the banking sector, stated that “the Orion group is not a circle of friends, but rather a circle of bandits.”11 The opposition UDF initiated a parliamentary procedure for the formation of a committee to establish the reasons for the crisis within the executive branch and the bank system caused by Orion’s operations. The most serious accusation was that the prime minister had tied, in an unacceptable way, the government to the Orion circle through members of his party in the parliament, board members of the national bank, members of the cabinet, and his own adviser and press center chief. 12 Videnov was isolated even within his own party, with Lukanov openly confronting the prime minister on almost all policy decisions made by the administration (Karasimeonov 2006, 85–86). Ideologically, Videnov had prevailed over his intraparty rivals at the 1994 election, but failed to completely defeat Multigrup, whose resources were sufficient to weaken the prime minister’s favorites.

104

Political Corruption in Eastern Europe

Olymp The victory of the UDF in the parliamentary election of 1997 changed the party in government, and a new circle of friends emerged that allegedly benefited from close connections to the ruling elite. This time the new dominant player was not linked to the former communist party; rather, it was formed to support the opposition in its conquest for power in the last months of Videnov’s government. According to some observers, by that time leading politicians within the UDF were ready to create their own structures and benefit from redistribution of public resources (Zlatkov 2008, 194). In those terms, the politicians initiated an economic scheme quite similar to the ones used by the former nomenklatura. Others believe that the ascendance of the UDF-linked Olymp is a result of the privatization policy model launched by the new government that encouraged rent-seeking behavior. Most of the time privatization deals were concluded after direct negotiations rather than through auctions, which allowed for special treatment of particular participants (Barnes 2007). Therefore, despite the crackdown on Multigrup and Orion, the UDF government could not avoid public accusations of corruption and favoritism, especially for allowing the emergence and consolidation of another giant economic group. Overall, however, the pro-reform, pro-European UDF government of Ivan Kostov led the country to a long-awaited rebuilding and economic stabilization (Ganev 2006; Barnes 2007). The birth of the economic group Olymp is often associated with a meeting of UDF leaders and big-business representatives to discuss support for the UDF candidate in the presidential election in the fall of 1996. The gathering was organized by Mario Tagarinski, member of the UDF National Coordination Council and future minister of the state administration, and by Evgenii Bakurdjiev, head of the UDF election campaign committee and future vice prime minister. Tagarinski invited potential sponsors to a restaurant called Olymp, owned by Slavcho Hristov, a supporter of the “blue idea” (or the UDF), who had tried several business paths after 1989, but none had led him to the top (Zlatkov 2008, 194). During the time of the blue cabinet, Hristov managed to develop a business empire that consisted of restaurants, hotels in luxurious mountain and sea resorts, construction companies, and a bank. This success was enhanced by the defeat of Multigrup and Orion after the BSP electoral loss and the UDF-

Networks of Corruption

105

initiated investigations for fraud and embezzlement. A majority of the public, however, believed that decisive for Olymp’s gallop to wealth was its close relationship with some of the UDF leaders since that secret dinner at Olymp took place. Opportunities for fast expansion of Hristov’s business opened in two arenas, privatization and procurement, both controlled by the government. In 1998, his company Balkanstroi Ingenering won profitable public contracts for construction work at the Sofia Airport, despite serious conflict-of-interest concerns (Peev 2000; Zlatkov 2008, 206). Through the management-employee buyout privatization, Olymp’s boss became the owner of large parts of the Zlatni Pyasuzi resort on the Black Sea and of Borovetz in the Rila Mountains. News that the Olymp circle participated in corrupt schemes of privatization of two hydropower stations spread in the media.13 By late 2000, Hristov’s firms operated successfully in a wide range of business sectors: advertising, publishing, insurance, consultancy, trade, and construction. The greatest achievement of Olymp that gave the group advantages over other viable competitors was the purchase of the failed stateowned Commercial Bank (Stopanska Banka). In the spring of 2000, Hristov was selected as head of the supervisory board of BRIBank, where he was a shareholder of just 1.5 percent of the capital. Several months later, BRIBank made a move and offered to pay 80 million leva to the creditors of the failed Commercial Bank.14 The purchase was a big event in the banking sector; Olymp realized a “dream” that Multigrup could not fulfill—having its own bank (Zlatkov 2008, 208). BRIBank then changed its name to CIBank, which soon turned into the most successful privately owned bank in Bulgaria. Key state enterprises and agencies—including the Bulgarian Telecommunications Company (BTC), the National Electrical Company, the National Insurance Institute, Bulgarian State Railways, Bulgartabak, customs, and the state lottery—sought the services of the “new” CIBank (Peev 2000; Zlatkov 2008, 210). Although Prime Minister Kostov made personnel changes in his cabinet and tried to curb corruption within the administration, the UDF lost the 2001 election exactly on charges of malfeasance and embezzlement of state property. As former (1991–1992) UDF prime minister Filip Dimitrov explains, the achievements of Kostov’s government are indisputable; yet a sense of lack of morality settled among the public that the ideas, in the name of which the UDF was

106

Political Corruption in Eastern Europe

created in the first place, had been betrayed (Dimitrov 2003, 78). In the eyes of common Bulgarians, Hristov’s circle was just another Mafiosi group that politicians in power, this time the anticommunist UDF, came to partner with. A closer look at Olymp, however, reveals a profile somewhat different from that of the previous dominant groups. First, the patron of the group was not part of the former nomenklatura, nor is there evidence for direct connection to the communist secret service machine. Second, by the time the UDF came to power, Hristov was a mediocre businessman struggling to establish himself in the hospitality sector. He was a follower of the UDF at whose restaurant some of the blue politicians used to meet for dinner; this circumstance helped him come close to the future election winners. Third, and more importantly, the wealth of Olymp was accumulated through participation in the privatization of large assets in the tourist industry and in the banking sector. Perhaps Hristov was favored over other candidates by the politicians whom he hosted, but his enrichment came more in the course of furthering the liberalization of the economy (as nontransparent as it was) than from blocking the process of reforms. Thus, the functioning of Olymp resembles more a rent-seeking behavior whereby the privileged circle needed the state to develop and improve the business environment than the robbery of public assets made possible by weak institutions. By the end of the UDF’s term in office other groups also consolidated their positions and challenged the ascendancy of Olymp. The group around Rozexim, headed by banker Emil Kyulev, was one of them.15 Despite losses to Olymp, Kyulev had managed to survive and keep significant resources.16 Another group that competed for UDF protection, Obshtinska Banka (Municipal Bank)–Bulgarian Holding Company, formed around the minister of the economy and chief negotiator with the EU Alexander Bozhkov and Sofia mayor Stefan Sofiyanski. By the end of the Kostov government, at least three interest groups were in a fierce competition lobbying before different UDF figures for greater rents.

Coalition Governments and State Capture by Multiple Groups The eight years that followed were marked by pluralism in the political and economic arenas. In the executive branch, coalition cabinets

Networks of Corruption

107

had to accommodate to the needs of a wide range of interests. Multiple economic circles competed for establishing themselves as players with influence on the policymaking process. None of them prevailed, nor did they win exclusive access to the state. Depending on the relative strength of their political patron, various groups became more influential and extracted higher rents at different times. In the special case of Dogan’s ethnic party, “permanent access to corrupt transactions” guaranteed social peace and a stable flow of resources to MRF’s business friends (Ganev 2006, 85). The NMSS’s Promise for Clean Governance In 2001 the Bulgarian voters, angry at those on the political left (BSP) and right (UDF) for ravaging the state and letting corruption flourish, elected the former king’s movement, National Movement Simeon II Saxkoburggotski (NMSS), to power. From 2001 to 2005, Bulgaria was governed by a coalitional cabinet of two parties, the NMSS and the MRF. The NMSS came to power on the promise to break up the bipolar model of politics of the 1990s. This model was blamed for serving solely the needs and the appetite for profits of certain economic groups, each with ties to either the BSP or the UDF. Simeon Saxkoburggotski, with his past in exile, aristocratic appearance, and speeches about decency, honesty, and transparency, was seen as a guarantee that this time people’s hopes would not be betrayed. 17 Since the NMSS was short of only one seat for an absolute majority in the National Assembly, they chose to govern in coalition with the MRF. As it became clear later, both parties had friends to satisfy and in order to stay in office for the whole term the coalition partners did not hesitate to blackmail each other. At the beginning it was big business that made a move to ensure channels of influence (Avramov 2004). Just three months after the 2001 election, representatives of big business gathered to register a nonprofit organization called Bulgarian Business Club “Vuzrazhdane” (“Renaissance”) (BBCV). Among the founders were Emil Kyulev (a prominent baker), Iliya Pavlov (president of Multigrup, now renamed MG Corporation), Vassil Bozhkov (president of Nove Holding, with a monopoly over the gambling business), Dobromir Gushterov (chairman of Orel-G, a prominent insurance company), Petyo Bluskov (founder of Press Group 168 Hours), Radosvet Radev (executive director of Darik Radio, a large privately owned radio station), and Tosho Toshev (editor in chief of Trud, one of

108

Political Corruption in Eastern Europe

two daily newspapers with the largest circulation).18 The BBCV members shared similarities, the most important of which (close connections to the former secret service and communist nomenklatura) can explain the origin of their capital. Initially, these businesses were well placed strategically, but suffered under the 1997–2001 UDF government, especially Multigrup, which had to liquidate one part of its business and transfer another to offshore companies. Under the new NMSS-MRF government, they quickly made a step forward to request a special status, claiming that they were the “patriotic” and “nationally responsible” businesses in Bulgaria (Peeva 2002). The BBCV itself found an institutional form under which to exercise more influence over Saxkoburggotski’s cabinet than the rest of the entrepreneurs in the country. Through Dimitar Kalchev, minister of the state administration and member of Multigrup, the elite club submitted a proposal for the formation of a consultative body, the Council for Economic Growth (CEG), to advise the prime minister on economic matters. In this council, the BBCV was to have more representatives than the rest of the employer organizations all together. This spurred angry reactions from the parliamentary opposition parties, and even divisions within the cabinet, with some ministers insisting on broader representation for the business (Peeva 2002). The institutional structure that the BBCV envisioned was intended as a tool for direct impact on the executive branch, and for long-term entrenchment of the club in the politicoeconomic edifice (Avramov 2004). Its leaders realized that they had common interests and goals that none of them was strong enough to realize alone. Together they could assure for themselves privileged positions in the privatization of the remaining large state enterprises, such as Bulgartabak and the BTC, and monopolize the agenda on Bulgaria’s European integration by isolating other economic agents. To achieve its goals, the BBCV portrayed itself before the public as “nationally responsible” capitalists who were concerned about the well-being of the country, “socially” sensitive, competent, and prepared to help the cabinet with its professional knowledge and skills (Avramov 2004, 248). The construction of a favorable image was also intended to remove any stains of suspicion about how these club members became so rich in times of economic hardship. Members of the elite club were eager to participate in various social activities, invest in sport clubs, sponsor cultural events, patronize art and monuments of art, contribute to charity causes, and build churches. The

Networks of Corruption

109

main objective was to convince both the public and the NMSS-MRF government that they were the most capable and responsible and the most preferred. Through aggressive lobbying, the BBCV tried to keep the speed of reform under control. One result of this effort was blocking legislation on lobbying, a law much needed for the increase of transparency and enhancement of open economic competition (Avramov 2004). The BBCV was founded in an environment where competition for rents had intensified with the acceleration of market reforms in 1997–2001. It was an attempt to consolidate businesses with close ties to the former communist regime, which had lost some of their previous might while the UDF was in power. Domestic competition for influence in the public sector and growing pressure for market liberalization in the process of EU accession encouraged the formation of the BBCV, yet the club could not be sustained for long. Some of its leaders, Pavlov and Kyulev, were assassinated in 2003–2005.19 The strong appetite for direct involvement in politics made others run for public office.20 Both the structure of competition among economic players and the collective action problems within the BBCV became obstacles to its success. The club failed to earn for itself a monopolist position in an environment where other influential interests were making serious advances. If the image of the nationally responsible business served Saxkoburggotski’s idea to portray himself as an above-party-politics, nation-unifying ruler, other representatives of his cabinet had different commitments. The group of the “Bulgarian yuppies” (including Finance Minister Milen Velchev, Minister of Economy Nikolai Vassilev, Minister of Transportation Plamen Petrov, and NMSS Member of Parliament Miroslav Sevlievski) allegedly had connections to one of the most suspect Bulgarian businessmen of the transition, Ivan Todorov “Doctora” (The Doctor).21 In violation of border tax and other regulations, Todorov accumulated significant wealth from trafficking of cigarettes, drugs, and weapons to Yugoslavia during the embargo. He was also actively involved in illegal trade of Chinese goods and in transferring dirty money from Bulgarian to British banks (Lilov 2008a, 90). Although direct evidence for provision of political protection for Doctora never surfaced, suspicions remained that Velchev’s choice of the British Crown Agents for supervision of the Bulgarian customs system did not happen by chance. Critics believe that the nontransparent way in which the con-

110

Political Corruption in Eastern Europe

tract with this firm was negotiated was a sign that the minister was pursuing a hidden agenda.22 When testifying before the parliamentary committee on internal security in April 2008, a former director of the Interior Ministry department for combating organized crime described how information from the ministry was deliberately leaked and directed to then–former finance minister Velchev, who effectively stopped a special forces operation against a group of alcohol smugglers.23 While particular economic groups realized huge profits due to close connections to previous governments, during the NMSS-MRF term in office particular politicians also came up as big beneficiaries. NMSS finance minister Velchev, together with his brother, came to own some of the most luxurious hotels on the Black Sea coast and the largest plant for production of ceramic tiles in Bulgaria (Zlatkov 2008, 116). Saxkoburggotski also enriched himself using his unique position of prime minister and former king to claim ownership of hundreds of acres of land and forests as well as palaces in Bulgaria that, he maintains, were once his family’s property. At the end of his term, the cabinet decided to grant a thirty-five-year lucrative concession deal to a Bulgarian-Portuguese consortium, to which two of Saxkoburggotski’s sons were connected. 24 Finally, rumors spread about “fairy” estates possessed by Dogan, confirmed by the latter in campaign speeches but not reflected in his tax return forms (Roth 2008, 235). His lavish wedding ceremony was a demonstration of extreme splendor that astounded common people who were dealing with unemployment and poverty. As Dogan thinks of himself as “untouchable,” he quite bluntly admits that political power is the reason for his wealth: “I come from a small village and now am a rich man. I am a politician. I am a businessman” (Roth 2008, 231). The MRF’s Ring of Firms The MRF is the only Bulgarian party whose leader openly talks about the links between his organization and firms around it and about the mechanisms of their relationship. As a coalition partner in two governments (2001–2009), the MRF consolidated its positions in the executive branch and mobilized all power resources to support its sponsors. Through the parliamentary group, businesses close to the MRF influenced the legislation process. Serving the demands of the timber industry, the MRF deputies in the assembly led the oppo-

Networks of Corruption

111

sition to the bill on Black Sea coast preservation (Roth 2008, 233). Through its continued presence in the executive branch, Dogan’s party gained access to the EU prestructural funds and the allocation of public contracts.25 The MRF held ministerial positions in the sectors of the environment, agriculture, and natural disasters, and received fourteen deputy ministerships and high-level administrative positions in state agencies. These bureaucrats made sure that companies close to the party were assigned contracts and got larger shares of the EU funds. On their part, the firms from the MRF circle made contributions in cash for the party’s election campaigns. One MRF deputy was cited as having claimed in 2005: “We buy the votes of 2,500 Roma in Dulgopol, 1,800 in Dolen Chiflik, more than 500 in Kamenar and Ignatievo, and 100,000 in neighboring Turkey” (Roth 2008, 230). And to buy out these votes, the party needed money that could not be ensured only through membership fees and public subsidies. Two economic sectors in which MRF companies were unequivocally favored over others are the timber and tobacco industries. Keeping control over the Ministry of Agriculture and Forestry in 2001–2009, the party and its firms established almost total control over the export of timber. In many instances, it was illegal; exporting firms paid large bribes at the border under the protection of their political patrons (Roth 2008, 234). Many believe that the frequent fires in environmentally protected forest areas were not accidental and attribute them to the illegal production of timber for export. The MRF minister of the environment Djevdet Tchakurov was accused of not doing enough against the violators and even of changing the status of these areas in order to let in developers from friendly firms. When opportunities for rents existed, the environment was an attractive target for exploitation. In his own “honest” way, Dogan explains: “The environment is too important to leave it to the environmentalists. Environmental protection is a very effective instrument to make money” (Roth 2008, 232). Well-known to the public are his ideas for investment in ecotourism or health tourism.26 Examples are the luxurious spa complex Orfei in the Rodopi Mountains and Casa Domini in Ribaritza. The official investor in Orfei is one of the firms gravitating around the MRF, Bulgarian Eco Project of Nikolai Vulchanov, former vice president of Multigrup and a large MRF donor.27 In 2005 only, Vulchanov contributed to the MRF’s coffers a total of 80,000 leva, donated by different firms that he owned (Aneva

112

Political Corruption in Eastern Europe

2007). To maximize the profits of these private hotels and spas owned by friends, the MRF requested from its coalition partners in the cabinet that retreat events for government officials be organized at these luxury complexes. The republican budget poured hundreds of thousands of leva into enterprises favored by MRF leaders and ministers. Opportunities for the party and for its close business partners to enrich themselves at the expense of the state abound in the repertoire of the MRF. In this regard, Dogan’s party has done much better than any of its political rivals (Stantchev 2007). Selling political influence for cash by the MRF has been so successful due to favorable strategic positioning, enhanced rent extraction, and absolute control over its electorate. Firms from the MRF circle were awarded well-paid public contracts for construction of highways and waste treatment stations, granted privileged positions in the redistribution of land endowed with natural resources, and allowed to siphon fuel and grain out of state reserves and resell them for profit (Petrov 2007; Stantchev 2007; Kahramanova 2009). From its position of a broker in the ruling coalitions, the MRF could successfully direct state resources to finance local projects and industries in places populated by ethnic Turks. Environmental projects of overall minor importance, but related to ethnically mixed regions, were approved for funding by the Enterprise for Management of Environmental Protection Activities with the environmental ministry. When the cabinet changed in 2009, the new minister of the environment established that millions of leva had been directed by the previous administration to municipalities with MRF political protection, often relevant to the lives of only a few people.28 The notorious case of Bulgartabak is yet another example of how the Turkish party managed to hold a monopoly over the ethnic vote. By postponing and blocking the privatization of the largest state-owned tobacco enterprise, Dogan provided an ineffective industry with a constant flow of state subsidies intended to help the impoverished tobacco producers, most of whom are ethnic Turks. This way, votes were guaranteed and the firms’ sponsors reassured. In the above schemes of party-company interactions, Dogan saw his role as one of commander in chief, directing and coordinating the work of deputies and bureaucrats, undermining their role while emphasizing his own. He said in a campaign speech in 2009: “I am the instrument in power who distributes the portions of funding in the state. . . . How would a deputy solve the problem with the projects

Networks of Corruption

113

and their financing? It’s not possible! Power is in my hands! I want you to understand this!” (Minchev 2009). No doubt, this structure is a highly centralized system of operation and financing, at the top of which is the leader.29 Internal party democracy is not a favored principle within the party and neither is the issue of representation; therefore, winning seats in the National Assembly is just a means to gain better positions for influence over the executive. This is where the MRF was in a position to provide for its sponsors. Because of the ethnic character of the party and its centralized structure, Ognian Minchev (2009) defines Dogan’s project for the MRF and its networks as “an authoritarian ethnic corporation” that operates through informal relationships and exchange while imposing a monopoly over the life of the Bulgarian Muslims. All structures of the MRF at all levels, Minchev (2009) contends, operate through corruption: “Corruption is a virtual identity of MRF, through which Dogan controls the [ethnically] mixed regions, and through which [he] systematically erodes and degenerates the Bulgarian institutions.” The Parcelization of the BSP The BSP won a plurality of the parliamentary seats in 2005, promising its electorate to initiate judicial investigations against the previous government. The former communists did not have a majority in the assembly; therefore they formed a coalition cabinet, albeit with the NMSS and the MRF.30 Later on, the public learned that the three parties had agreed to share power by the 8:5:3 formula. Subsequently, this led to expansion of the government, creating new ministries and deputy minister positions to satisfy the ambitions for power of three coalition partners. Moreover, the BSP had to relinquish its preelection promise to review suspicious contracts concluded by the previous government. Exposing informal networks that had formed around the NMSS and the MRF was not a preferred strategy because the BSP needed the two parties to keep the ruling coalition alive. The socialists could not eliminate the favorites of the outgoing government; consequently, the BSP’s wealthy friends had to compete over state resources both among themselves and with previous unpunished rent seekers. Several circles of economic interests emerged around politically influential figures in the socialist party. Among the generous sponsors of President Georgi Parvanov, elected on a BSP ballot in 2001 and

114

Political Corruption in Eastern Europe

reelected in 2006, were businessmen investigated by the European Anti-Fraud Office (OLAF) for misuse of preaccession funds. A conglomerate of firms headed by Lyudmil Stoikov and Mario Nikolov was accused of organizing a criminal scheme and of tax and subsidy fraud for thousands of euros. 31 Not only did Stoikov finance Parvanov’s 2006 campaign, but he also organized a citizen group (of businessmen), “Coalition for Pernik,” that raised the candidacy of Parvanov for a second mandate (Carvajal and Castle 2008).32 When OLAF contacted the Bulgarian services in 2006 and requested that Stoikov be investigated, the Sofia prosecutor’s office was pressured “from above” to postpone any announcements and a possible arrest until after the presidential election (Zlatkov 2008, 88). Stoikov continued to enjoy high-level political protection. In mid-2008, President Parvanov inaugurated a new production line in the Pernik metallurgy plant owned by his sponsor, Stoikov (Zlatkov 2008, 82). The Stoikov-Nikolov group of fifty enterprises maintained more than one close connection to top socialists. While Stoikov was pouring money into Parvanov’s campaign, his business partner funded the BSP. The OLAF report carried evidence for tax and subsidy fraud and for illegal reexport of meat from China to Germany and France, not sanctioned by the Bulgarian authorities. When asked why his companies sponsored the BSP’s 2005 campaign, Nikolov confessed: “These are working companies and their management decided that certain people had contributed to the excellent financial results” (Krusteva 2008). Yet Prime Minister Stanishev, leader of the BSP, rejected the insinuations of corrupt exchanges and assured that the government had not opened a political umbrella of protection over anybody. These statements were soon rebuffed by the release of a video showing Nikolov giving Stanishev a tour around one of his meat factories just a few weeks before the contributions were made. According to documents presented by the opposition to the prosecutor’s office, Nikolov poured into the socialist party’s coffers over US$137,000 (Carvajal and Castle 2008).33 Informal networks based on political protection of economic interests also formed around two other prominent BSP figures, Roumen Ovcharov and Roumen Petkov. As head of the megaministry of the economy and energy, Ovcharov had enormous power. He enhanced the negotiation of contracts for natural gas supply between Bulgarian and Russian companies. Those agreements, according to

Networks of Corruption

115

the opposition, were not beneficial for Bulgaria (Petkova 2007). Ovcharov also tried to accelerate the many times postponed privatization of Bulgartabak, but met fierce opposition from Dogan and the MRF. Journalists’ investigations revealed that, in this conflict, key players among the ruling elite tried to establish control over the trade with cigarettes, providing millions of leva for the parties’ “black coffers” through nine large distributors. 34 The political effect of a Bulgartabak privatization would have caused financial exhaustion for the MRF and the subsequent isolation of Dogan from power. In a scandal that blew up in the spring of 2007, Ovcharov’s name was also linked to an ongoing embezzlement investigation of the former Sofia heating utilities’ head and to a big-business person accused of money laundering and suspicious participation in energy projects. 35 Ovcharov did not fully cooperate with the prosecution in order to protect friends and was later urged by the prime minister to resign.36 Petkov headed the Ministry of Internal Affairs in the BSPNMSS-MRF cabinet. Within the party, he was a political rival of Ovcharov with whom Petkov competed for influence over the BSP structures. He was in charge of the BSP fund-raising campaigns and allegedly secured the money through nontransparent deals with business circles (Roth 2008, 173). Moreover, observers note that under his tenure the Ministry of Interior interfered in the market competition among different firms through deliberate and selective leaking of information in return for sponsorship money (Roth 2008, 120). The scandal that threw light over Petkov’s connections to suspicious business circles linked him to two of the wealthiest post-1989 Bulgarians, the so-called Galev brothers. They became notorious for production and trade of amphetamines, activities that according to a former top officer in the security services were protected from prosecution directly by the interior minister.37 After evidence was released for improper meetings between the minister and the Galev brothers, Petkov tried to hinder the investigation. The opposition accused him of using public office to turn the ministry from an institution investigating criminal behavior into an agency that concealed crime (Petkova 2007). Under pressure from the West, Petkov was asked to submit his resignation. The removal of both Ovcharov and Petkov from the cabinet was considered a big victory for Prime Minister Stanishev, who used the circumstances to strengthen his position within the BSP. At last, the

116

Political Corruption in Eastern Europe

office of the prosecutor general initiated trials of bosses of economic groups protected by the two unfortunate ministers. However, this did not save the party from the grasp of special interests. It was just one moment in an ongoing competition, when some of the circles prevailed over others in a party parceled out by captor firms. Only after the electoral defeat of 2009 did prominent socialists speak in a loud voice about the suffocating effects of the friendly circles on organizational life.38

Conclusion: The Permanence of Friendly Circles The case of Bulgaria illustrates well the propensity of corruption to create “conspirational networks held together by mutual obligations, fueling nepotism, favoritism, and clientelism” (Inglehart and Welzel 2005, 193). Among the factors that favored the rapid growth of informal corrupt ties between big business conglomerates and politicians was the direct access of the groups’ leaders to the government. In the case of Multigrup, the channels for privileged contact with public officials from Zhivkov’s time remained open after 1989, thus facilitating a continued plundering of state assets and offering new excellent opportunities for enrichment (Barnes 2007). Former managers and other high-level employees of state agencies left the public sector in the early transition years to take positions in Multigrup that promised a better payoff. The list includes many, all of whom ended their service in the public sector by doing final huge favors for the conglomerate at the state’s expense (Ganev 2007, 102). This direct access to top government officials was instrumental at the phase of money accumulation and was maintained later in a more institutionalized form. For example, Multigrup assigned Stoyan Denchev, an MRF deputy and secretary of the Council of Ministers under Prime Minister Berov, the responsibility to ensure a permanent contact for the corporation with the state.39 These connection lines impact longevity and endurance, which explains how Multigrup revived from the ashes after the blows it received from the UDF government. The groups also survived because the Bulgarian political parties were in constant need of financial resources. At the beginning of the transition, only the BSP had the finances and the organizational capacity to hold a nationwide electoral campaign. Later, state subsi-

Networks of Corruption

117

dies became available for larger parties, but these amounts were far from sufficient to cover organizational and operational expenses (Kostadinova 2007). Money started to flow into party coffers in exchange for political protection and patronage. Pavlov was very direct when explaining to his associates the psychology of these deals: “It is important that they [politicians] think they deserved the money, that they are smart and talented, and not that we are buying them. Actually, they know it but prefer to be hypocritical.”40 Big business strived for protection and the parties needed the money for increasingly expensive campaigns. Given these reciprocal relationships, one cannot refrain from asking: Who was in charge in the mighty politicoeconomic networks that in effect governed the country and stripped it of its assets? Did the powerful multimillionaires prevail over the politicians who carry a mandate from voters to reform and democratize Bulgaria? Pavlov once reminded Lukanov directly: “Don’t forget that I drive the train!” (Lilov 2008b, 116). After this confrontation between Multigrup’s president and his patron in the fall of 1996, the latter lost his life. Yet Pavlov outlived Lukanov by only seven years, a testament to the fact that the former was just a “hired” billionaire and not entirely in charge of the mighty conglomerate. By the time the public learned something about the business of the group and its enormous influence over previous governments, Multigrup’s most prominent leaders had already been eliminated. As established by previous studies, corruption networks differ with regard to organizational form and activities (Heywood 1997). In Bulgaria, the informal networks varied by extent of political involvement and by structural composition. Usually politicians were not the business decisionmakers, but there were exceptions. In the case of Orion, more than in any other friendly circle, politicians in high government positions were also top businessmen. The first advisers of the prime minister, ministers, and deputies were activists in the business operations of the circle. As the transition unfolded, marketization progressed, and the rules of competition settled, economic and political actors in the informal networks sought more distinct forms of cooperation. With regard to composition, some of the groups managed to develop a classical structure composed of a business, a bank, and a media. Multigrup and Orion exemplify such an evolution. Olymp never owned a media, and neither have the multiple networks that emerged after 2001.

118

Political Corruption in Eastern Europe

Relations with parties also changed over time. Orion and Olymp each were clearly linked to only one party, the BSP and the UDF respectively. At the beginning, Multigrup was connected solely to the socialist elite; however, after the BSP lost power in 1991 the group’s strategy was to target for support deputies from different parliamentary groups.41 During the 2001–2005 and 2005–2009 coalitional governments, influential economic groups allied with powerful wings within the BSP and the NMSS. To make sure that the coalitional cabinet stayed in power, some economic groups entwined around multiple patrons.42 Only in the case of the MRF is there a case of centralized management of party-sponsor relationships. In practice, the MRF was transformed by the dominant figure of Dogan into a center that had, gravitating around it, businesses whose fate was entirely in the hands of the party leader. The analysis of the Bulgarian post-1989 informal corrupt networks suggests that powerful groups emerged due to connections to the previous regime, but they managed to endure because parties needed resources that the big businesses were willing to supply. When institutions responsible for investigation and punishment of illegal behavior are in place, such networks should not prosper and corrode politics. In Bulgaria, this was obviously not the case. Investigators, customs officials, and prosecutors have been blamed for not working effectively, hindering and postponing trials, and even for being bribed. The usual excuse has been that they are not well paid and easily accept bribes or retract under threat. In the next chapter, I deal with the problem of state capacity and whether an improved flow of resources to departments and agencies can help combat corruption in Eastern Europe. Notes The epigraph to this chapter is from an interview with Ahmed Dogan on Slavi’s Show, broadcast by BTV, 23 June 2005, www.mediapool.bg/show /?storyid=106410, accessed 7 September 2010. 1. Pavlov’s first wife was the daughter of a general who was head of the military intelligence unit of the Committee of State Security in the 1980s. Pavlov’s academic adviser for his postgraduate studies in journalism was Dimitar Ivanov, a high-ranking officer at the Ministry of the Interior and chief of the notorious Sixth Department of State Security that was assigned to investigate dissident behavior among the Bulgarian intelligentsia (Lilov 2008b, 131).

Networks of Corruption

119

2. About the same time, in 1988, Pavlov started his private business by founding Multiart and then Multigrup, which later evolved into MG Corporation. Through his connections to the high echelons of the nomenklatura, he managed to receive lucrative deals, including the dissembling and preparation for scrap of old Soviet submarines that his firm acquired in Malta. 3. Ganev (2007) reveals in detail the nature of these interventions on the part of Multigrup and the innovation and aggressiveness with which its leadership imposed the conglomerate on three public enterprises. In the HimkoKremikovtzi affair, Multigrup essentially bought out the debt of two enormous state-owned plants (producing chemicals and steel) to the single big deliverer (also state owned) of gas, Bulgargaz. As owner of their previous financial responsibilities, Pavlov’s group collected payments from the two debtors (to the state) and never undertook the construction works that it had promised under the agreement. In a different affair, Multigrup gained access to the board of directors of another state enterprise, Digital InformationStoring Instruments in Stara Zagora, and set up a joint venture for delivery of production equipment with a Swiss firm close to Multigrup. Later on, the foreign company bought the entire compact disc production line and stripped the state enterprise of its most profitable asset. 4. At that time, Lukanov was a member of the Bulgarian parliament. 5. According to witnesses, there was a heated verbal exchange between Pavlov and Lukanov at a Topenergy meeting in Russia in 1996 (Lilov 2008b, 116–117; Petrova 2008, 69). 6. Roumen Spassov, owner of the first registered firm in the group, Orion Commerce, was employed at the Ministry of Interior in the 1980s. Krassimir Raidovski, a former officer in the state security intelligence department, led the Bulgarian counterintelligence effort in Greece. Mikhail Danov, also associated with the secret service in the past, worked for Bulgaria’s largest arms-exporting company, Kintex. 7. As head of the mail and long-distance committee, Kolarov urged the leadership of the state-owned Bulgarian Post Bank (BPB) to finance Orion’s BAIB with millions of leva and US dollars for purchase of new products and service of outstanding debt (Dimitrova 1995). Despite a maneuver that put the BPB under Kolarov’s direct supervision in August 1995, it managed to resist the pressure from Orion. 8. In mid-April 1995, the government decided to raise the grain export quota to 540,000 tons (from 350,000 tons) and to lower the export tax to US$5 per ton (from $25 per ton). The export companies, most of them close to the cabinet, made huge gains since they had already signed contracts on the basis of the old tax and the government did not collect the difference. In July, the quota was entirely removed without paying attention to the quality of the current crop of wheat; the export increased exponentially despite the increased export tax ($35 per ton). In August, the tax went up to $55 per ton, but Turkey dropped its import taxes and the export flow all moved to that country. In late September, after warnings from mill managers for insufficient deliveries, export of wheat was finally banned with the exception of

120

Political Corruption in Eastern Europe

seed. In late November, the cabinet stopped all grain exports, but the ban was not enforced until early December (Vladimirova 1995). 9. “Criminal Experts Surround Videnov’s Advisors,” Demokraziya, no. 282 (December 1995): 1–2. 10. “Pravitelstvoto Subsidiralo Priyatelski Firmi” (The Government Subsidized Friendly Firms), Demokraziya, no. 12 (January 1996). 11. “Litza ot Grupirovkata Orion Sa Sledstveni” (Members of the Orion Group Are Under Investigation,” Demokraziya, no. 28 (February 1996): 1. 12. “SDS Iska Anketa za Orion” (UDF Wants Parliamentary Investigation of Orion), Demokraziya, no. 56 (March 1996): 1, 2. 13. “Citizens Live Behind Bars Instead of Criminals,” interview with Bogomil Bonev, former minister of the interior in the UDF government, in Standart, 8 March 2001, www.standartnews.com/archive/2001/03/08 /English/interview/index.htm, accessed 2 July 2009. 14. Rumor had it that other offers were discounted and that the deal was facilitated by experts from the Bulgarian National Bank supervisory division who were close to the prime minister (Peev 2000; Zlatkov 2008, 208). 15. This group lost its privileged access to power after Kostov removed from office the ministers who maintained friendly relationships with the rich banker. Kyulev’s consortium was eliminated from the competition for Hebros Bank and lost to Olymp’s CIBank many of its previous big customers, including state-run enterprises and agencies. The latter loss was especially painful because, whatever the losses in the other branches, Rozexim would compensate with flowing fresh capital from the service of state enterprises (Peev 2000). 16. Kyulev was well connected in the political circles. Just two years after the fall of the UDF government, the new cabinet of Saxkoburggotski allowed Kyulev to “buy” the state insurance institute Durzhaven Zastrahovatelen Institut for a ridiculously low price. After he was assassinated in October 2005, his widow sold the company for an amount nine times higher (Zlatkov 2007, 62). 17. Saxkoburggotski was the firstborn son of the late Bulgarian czar Boris III, who died in 1943, when Simeon was just six years old. After the end of World War II, a referendum rejected the monarchy in Bulgaria and the country became a republic. Saxkoburggotski, along with his mother and sister, was deported from the country. The heir to the Bulgarian throne settled in Madrid where he lived until his return to Bulgaria, maintaining a good command of the language and keeping in touch with Bulgarian immigrants. 18. The first attempt to organize such an elite club was in 1993, when the same businessmen, with the exception of Radev and Toshev, participated in the foundation of a confederation of the large industrialists, also known as G13. The main goal of G-13 was to lobby before Berov’s government in favor of big business. It did not last for long because of conflicts that the members could not overcome. One reason for that might have been the fact that, at that time, Multigrup was disproportionately larger and more influential than the rest of the club and could rely on its own connections to the government. In fact, Reneta Indjova, prime minister of a caretaker cabinet in 1994, called the

Networks of Corruption

121

outgoing Berov government “the cabinet of ‘Multigrup’” (Avramov 2004, 250). 19. The names of their killers remain unknown. 20. Bluskov ran for the Sofia mayoral seat in 2003 and lost, and Gushterov was elected a member of parliament on the BSP ballot in 2005. Searching for opportunities to directly influence the distribution of significant amounts of EU structural funds, other businessmen from the competition also ran in elections. 21. These ties were confirmed by a set of photographs found in Todorov’s car after an unsuccessful assassination attempt against him in 2003. The photographs had been taken on a yacht in Monaco in 2001 and reveal Velchev, Petrov, Sevlievski, and their wives in the company of Doctora, playing cards. As Doctora later admitted, he hosted such parties for the Formula-1 races for friends and their friends. The politicians in the photographs claimed they did not know who their host was. Along with the photographs, the police discovered documents listing huge sums of money for distribution. Doctora was murdered in the streets of Sofia in early 2006. 22. The Bulgarian yuppies studied in universities in Great Britain and the United States; when they were recruited by Saxkoburggotski, they were working for financial institutions in London. There, they allegedly met Doctora and became friends with the sons of the former king. In the case of Crown Agents, the popular belief was that Velchev was either ensuring a contract for British friends or ensuring customs protection for Doctora. In both instances, he was expected to have gained. 23. Parliamentary Committee on Internal Security and Public Order, Hearing of General Vanyo Tanev (stenographic protocol, in Bulgarian), Kapital, no. 15 (Special Appendix, 17–18), 12–18 April 2008. 24. The concession was awarded for operation of the longest Bulgarian highway, Trakia. The scandalous part was that, without competition or tender, the government committed public funds through two state companies that participated in the consortium, along with three Portuguese private firms. Simeon’s sons’ presence in the latter raised questions of possible conflict of interest. The deal was quite unfavorable for the Bulgarian state. According to its provisions, the consortium would do the remaining construction work and maintenance and collect the road toll for thirty-five years. Losses from lower vehicle flow were to be covered by the Bulgarian government. For the construction of the highway, the cabinet relied on finances from the EU and the European Investment Bank, but the latter withdrew after learning that other bidders were not allowed. 25. In two governments, 2001–2005 and 2005–2009, the MRF played a crucial role in keeping the ruling coalition together by assuring it that it had parliamentary support. In both cases, splits within the NMSS threatened the survival of the cabinet, but the disciplined MRF kept it together. This circumstance also allowed Dogan to obtain the administrative positions he wanted, far more in number than the proportional weight of his party. The tripartite coalition (2005–2009) initially negotiated an 8:5:3 formula for distribution of posts among the BSP, the NMSS, and the MRF, respectively.

122

Political Corruption in Eastern Europe

However, over time the bargaining position of the NMSS weakened to the benefit of the MRF. 26. Dogan’s statements about investment are quite puzzling given that officially, he owned just one apartment in Sofia. And also, in his words, Dogan received only a small part of his salary as member of parliament because he rarely went to work (Obretenov 2007). 27. “Imperiyata, v Koyato Imperator Dogan ne Pritezhava Nishto” (The Empire Where Emperor Dogan Does Not Own Anything), Sega, 26 May 2008, www.segabg.com/online/new/article new.asp?issueid=2982§ionid =5&id=0001001, accessed 14 July 2009. 28. “Prez Ecoproekti Sa Nalivani Pari v Obshtini na DPS” (Money Has Been Poured into MRF Municipalities Through Environmental Projects), Mediapool, 3 August 2009, www.mediapool.bg/show/?storyid=155046& srcpos=5, accessed 3 August 2009. 29. In the words of Kerim Karaali, a former activist of the Turkish party, the organization turned into a “commercial party” dealing with trade. The headquarters of MRF worked as a business center that facilitated the deals. Dogan received businessmen who wanted to talk with him about their ideas—“from asparagus to renewable energy resources”—and seek his support. To meet with Dogan, a businessman had to pay an entry fee of 5,000 to 10,000 leva (Gospodinova 2008). 30. The BSP had to invite at least one other party for the formation of the cabinet. President Parvanov, a former leader of the BSP, insisted that the other party be the MRF, whose votes he was hoping to receive one year later when he was up for reelection. This plan failed because the two parties together were short of the necessary 50 percent–plus–one majority. The party with the second large parliamentary group, the NMSS, also failed to gather such support. At the third round, when the mandate for formation of the government was given to the MRF, the three parties came together and their proposal for a coalitional cabinet was approved by the assembly. 31. In this scheme, the Stoikov-Nikolov group worked with partners from the former East Germany. They applied for development aid from the EU Special Accession Program for Agriculture and Rural Development to buy new equipment, purchased old machines from Germany, and kept the difference in cash (Carvajal and Castle 2008). 32. President Parvanov was born and raised in the Pernik region. 33. Copies of receipts and even contract agreements were handed to the prosecutors by Boyko Borisov, then Sofia mayor and at the time of this writing prime minister and leader of Citizens for European Development of Bulgaria. 34. “Voinata vuv Vlastta Bila za 117 Mln. Leva” (The War in the Government Was over 117 Million Leva), Express Online, 3 May 2007, www.express.bg, accessed 3 February 2009. 35. The businessman’s name is Krassimir Georgiev, a friend of Ovcharov’s from the 1970s when both studied in Moscow and later worked for one of the district committees of the Communist Party in Sofia. Through active lobbying of various institutions, Georgiev’s company, Frontier, was

Networks of Corruption

123

able to win participation in the largest public energy and construction projects (Radmilova, Nikolov, and Dachkova 2007). 36. According to a report by the Center for the Study of Democracy, in Sofia the highest risk for use of nonregulated funds is in the energy sector. The reasons for this are the high percentage (30 percent) of contracts that are awarded through direct negotiation, rather than competition, and the insufficient subsequent control and weak mechanisms for punishment (Vasileva 2007). 37. In fact Plamen Galev and Angel Hristov, the Galevs, are former officers of the national service for combating organized crime. Through racketeering and blackmail, they managed to accumulate significant wealth and take total control over Dupnitza, a southern town in Bulgaria. For the citizens of Dupnitza, the two brothers are untouchable. When in 2006 the mayor of Dupnitza admitted that he was helpless to deal with the Mafia, the interior minister was called to intervene. Petkov went and left after spending just a few hours, explaining that “[the Ministry of Interior] cannot interfere in regional conflicts, let them figure it out on their own.” Using his “connections,” the minister met the Galevs in secret locations to urge them to stop the Mafia war at least until Bulgaria is admitted into the EU (Zlatkov 151, 164). 38. In an interview by the television channel BTV, former BSP deputy Mladen Chervenyakov described his party as currently being “parceled out by clans” and that, on some occasions, party structures had fallen in the hands of “particular financial interests” (Petkova 2009). 39. Interview of Stoyan Denchev by Marina Chertova, Vsekiden, 17 July 2009, www.vsekiden.com/?p=54821, accessed 17 July 2009. 40. According to a report issued by the Ministry of Interior, Pavlov promised Lukanov 1 million Bulgarian leva for a future campaign (Lilov 2008b, 136). 41. “Tzenite na Deputatite Padat” (The Prices of Deputy Votes Are Dropping), Kapital, no. 7, 17–23 February 1997, pp. 1, 8. 42. Tatiana Dontcheva, former BSP deputy, explains that at some point the rings of firms begin to interweave and form “common rings that alloy a common government.” These common rings, according to Dontcheva, are the reason why the BSP compromised so much and did not distance itself from the MRF, its coalition partner, earlier. “Dontcheva: BSP Nyama Kak da Izbira Predsedatel Sega” (Dontcheva: The BSP Is Not in a Position to Choose a Chairman Now), Vsekiden, 9 August 2009, www.vsekiden.com /?p=56075, accessed 9 August 2009.

6 State Capacity: Is It Really the Problem?

For a large number of officials, bribes have become the only stimulus for activity. —Svetlana P. Glinkina, “The Ominous Landscape of Russian Corruption”

This chapter expands my investigation to assess the role of state capacity (material, administrative, and legislative) in successfully constraining political corruption. Survey research reveals that significant majorities of the public in Hungary, Bulgaria, and Russia believe that the state should be responsible for tackling bribery, extortion, and other offenses (Holmes 2006, 211). The wisdom that “the fish smells from the head” has turned into a popular conviction that politicians and public officials, not ordinary people, are to blame for corruption and it therefore is the responsibility of the state to correct the behavior of its employees. Things are not that simple, however. Previous research argues that, even when there is political will, resources and expertise may hinder the effort. For example, Franklin Steves and Alan Rousso (2003) note that formal measures such as the adoption of anticorruption national strategies are needed to satisfy international funding agencies and investors, but there could be “substantial gaps” between such commitments and the government ability to implement them. Indeed, “extremely weak administrative capacity” has been proposed as one of the most important reasons for the ineffective fight against corruption in Albania and Macedonia (Vachudova 2009, 46).

125

126

Political Corruption in Eastern Europe

Is state capacity the main problem for the tremendous difficulties faced by the Eastern European transitions in developing transparent policymaking and implementation processes? It is clear that all nascent democracies cannot afford the resources that industrialized nations have at their disposal to combat corruption. Yet after 1989, the Eastern European countries experienced different rates of economic success that might have enabled some states more than others to prosecute malfeasance. Independent from material potential, some nations in the region might have an advantage in the fight against corruption because of specific characteristics in their social and cultural construct. These are the possibilities that I explore in the chapter. First, I examine state capacity in postcommunist Europe as resources needed to reward public officials for protecting the public interest and as potential to detect, constrain, and punish malfeasance. Next, I look at some nonstate factors that may create a context that reduces the occurrence of corruption and enables the fight against it. By isolating the impact of state capacity through multivariate analysis, I identify specific state and nonstate capacity-enabling factors that play crucial roles. I conclude the chapter with an assessment of the prospects for anticorruption capacity building in Eastern Europe.

The Problem of Capacity The capability of political systems to achieve policy objectives is now considered by many experts to be as important as the ability to recognize the problems and to propose solutions to them. Studying the success of national governments in implementing environmental protection policies, Martin Janicke and Helmut Weidner (1997, 1–2) conclude that “it is not only the kind of action, but also its structural characteristics which matter.” In a study on the effectiveness of postcommunist anticorruption measures, Steves and Rousso (2003) note that regardless of the “intensity” of the institutional changes initiated by elites to curb corruption, policy objectives may not be achieved in the absence of sufficient government capacity. In other words, there are facilitating conditions and objective limitations that affect public sector performance even when and where politicians’ willingness is strong. These factors may be cognitive (different levels of knowledge about the problem); political (institutional basis for action);

State Capacity

127

economic (financial resources, technology, and infrastructure); and societal (cultural norms, public support for policies, and civic activism). It is not surprising then that international organizations (e.g., the OECD, the EU, and the World Bank) have strongly emphasized that countries, especially those in democratic transition, need to further develop their administrative and legislative capabilities (OECD 1994). The contemporary understanding of the concept of national capacity for effective governance encompasses state and nonstate resources to enhance success and avoid failure. State capacity is usually referred to as the ability of the state to “implement policies” and “deliver services” (Polidano 2000). The scope of these abilities is broad, from extraction of resources to the development and maintenance of institutional infrastructure, both critically important for implementation of all sorts of policies. Public sector capacity, in particular, refers to the capabilities of the administrative machinery within the state to perform its functions, deliver services, and provide policy advice. As Charles Polidano (2000) specifies, a well-institutionalized bureaucracy is in a position to feed the policymaking process with advice and to enforce the rules of the land rigorously. States and their administrations do not function in a vacuum and the context within which they operate may also enhance or impede success. Polidano (2000) uses “enabling environment” as a summary term for the social and economic factors conditioning the work of state institutions. If present, ethnic divisions, civil society weaknesses, political instability, and economic crises can all undermine a state’s effort to act. Janicke and Weidner (1997) also pay particular attention to context when they analyze the factors enhancing environmental policy implementation. Their “cognitive-informational framework” includes NGOs and the media as nongovernmental proponents of environmental policy implementation as well as public awareness, “epistemic communities,” and the levels of mobilization and propensity for cooperation. Why should we expect capacity to also matter in the area of anticorruption? As in any other policy effort, the task of fighting corruption also requires a significant influx of resources for monitoring, decent (if not motivating) compensation of state employees, introduction of modern technologies, and so forth. As recognized in earlier

128

Political Corruption in Eastern Europe

research, success in managing corruption requires that the state “be a credible enforcer of laws” (LaPalombara 1994, 333). Laws enabling prevention, discouraging engagement in corrupt activities, and punishing violators also appear as necessary in empowering the state to exercise authority and to combat illegal behavior. More than in any other public sector policy area, rewarding bureaucrats for professionalism and integrity at the workplace seems a reasonable mechanism for higher-quality performance because it directly impacts personal income. In sum, it is logical to expect that the ability of a state to mobilize resources, to develop modern institutional infrastructures, and to improve the technological standards of public administration will increase the effectiveness of anticorruption policies. In some ways, the fight against corruption resembles the battle for a clean environment. Both transparent policymaking and environmental protection have broad social importance and, from a normative point of view, they are strongly desired as good for everyone. Their success, however, is not guaranteed and the effort of the state needs to be matched, if not preceded, by a strong push from civil society. Public awareness of the problem of bribery and rejection of it as wrong and unacceptable can become crucial factors in identifying and exposing corrupt officials. At the same time, divisions within society are sometimes reflected by allocation of political posts and policymaking in favor of particular groups. Such fragmentation carries the potential for distorted representation and results in a reduced ability of the state to serve the needs of all its citizens (Polidano 2000). Despite the wide range of scientifically established sources of policy capacity, experts and citizens do not agree on what slows down the successful implementation of the anticorruption measures already adopted in Eastern Europe. The most frequently given reason for the widespread corruption is low salaries of public officials. That salaries in the private sector are much higher than in state offices is a fact that does not need further validation. But is this the problem or there are contextual characteristics that further diminish the ability of the state to more effectively contain corruption? To put it differently, is society also responsible for the speed with which anticorruption reforms take root? After all, politicians and public servants are recruited from within this same postcommunist society that so severely criticizes the state for being insensitive and unfair to its citizens.

State Capacity

129

State Capacity After Communism One of the most unexpected lessons learned from the collapse of communism in Eastern Europe is that the communist state was much weaker than expected.1 The extreme centralization in party organization and operation, as well as the ultimate power vested in the Politbureau and in a single leader, had misled many to believe that the state was strong and coherent, functioning under the overarching Marxist-Leninist ideology. As it turned out, the “bloated and inefficient” communist state relied on a relatively small core administration, many times smaller than counterparts in Western consolidated democracies (Grzymala-Busse 2007, 135). This legacy, coupled with the postponement of administrative reform, contributed to an expansion of the state apparatus after 1989. The growth was used for narrow party interests, rather than strengthening the potential for policy implementation. In the absence of laws containing political appointments and establishing much-needed oversight procedures, discretionary hiring bloomed and new state institutions proliferated throughout the 1990s. State administrations grew in size, but not so much in professionalism. Insufficient state capacity to deal with the complex tasks of economic transformation and international integration was recognized by the EU as a major challenge facing postcommunist candidate states in assuming the duties of membership (Monitoring the EU Accession Process 2002, 57). The abilities of state administrations to perform their functions in countries such as Macedonia and Albania have been assessed as feeble, undermining any effort to deal with corruption (Vachudova 2009). Even in the more advanced transitions such as in Slovenia, municipal administrations suffered from understaffing and a lack of merit-based formulas for reward and promotion (Hacek and Baclija 2009). By 2008, the European Commission had identified poor administrative capacity in Bulgaria as a major obstacle to implementing the law and limiting corrupt practices. After 2001, the task of improving the administrative capabilities to implement the EU acquis was directly linked to the prospect of successfully fighting corruption (Monitoring the EU Accession Process 2002, 57). In its reports the European Commission identified two main areas of capacity building: (1) civil service aimed at modernization, professionalization, and depoliticization of the administration; and (2) enabling the judicial structures by making them more

130

Political Corruption in Eastern Europe

independent and better trained. In addition, problems of inadequate staffing and salaries were often identified as ones that needed particular attention in securing a transparent and effective allocation of the EU accession funds (World Bank 2006a). Delays in administrative restructuring due to disagreement of postcommunist elites over civil service reform weakened government capabilities on every level. The major Czech political parties, led by self-interest rather than concerns of effectiveness and functionality, postponed decentralization and thus deprived regions from competencies and financial instruments necessary for the delivery of public services (LaPlant et al. 2004). Lack of consensus over the form and speed of civil service reform in Bulgaria, Latvia, Slovakia, and Romania delayed capacity building and uncertainty exhausted the available resources (Antoaneta Dimitrova 2002). Yet where political competition was robust, modernization of the public sector started earlier and state exploitation by ruling elites was contained (Grzymala-Busse 2007). In the absence of healthy competition, however, political parties could postpone administrative reform and continue to exercise discretionary hiring. This exploitation impaired the state’s ability to deliver services to the broader public.

Indicators of State Capacity For the purpose of a more systematic exploration, I next focus on particular indicators, observe them across nations, and compare the rates of variation with those of corruption. These include material resources, legislative instruments, and technological advance available to the postcommunist states. Material Capacity Scholars have long argued that states need resources in order to successfully perform their functions (Schmitter 2005). They extract revenues in the form of taxes from citizens and firms and sometimes also receive foreign aid and grants that are then used to fund policies and pay state employees. The more revenues collected, the better a state is able to operate and deliver to its citizens. There are more financial resources to enforce anticorruption policies, including the purchase of modern technology for better monitoring and coordination. There

State Capacity

131

also is money for wages and other social benefits for bureaucrats, high enough to overpower temptations to accept bribes (Monitola and Jackman 2002). As Guillermo O’Donnell and Mario Vargas Llosa (1993, 52) emphasize, one important requirement for strengthening the role of the state to achieve effective democracy in Latin America has been “a public administrator class that can make a decent living so they do not resort to corruption.” Poverty is a traditional excuse for breaking the law and taking bribes for services. Therefore, increasing the amount of wages and compensation appears to be one logical way to help overcome the seductive power of lucrative offers. Pay raises and security of tenure for the public administration, however, have been discussed as controversial in the postcommunist context. Such an idea seems unfair given that potential recipients of greater benefits are the corrupt officials themselves. Therefore, some observers argue, the egalitarian Eastern European public would not feel enthusiastic about a more highly paid state bureaucracy in the face of serious economic difficulties (Monitoring the EU Accession Process 2002, 58). Indeed, about one-third of the public in the Czech Republic, Slovakia, and Ukraine have expressed clear antipathy to administrative reforms aimed at increasing public officials’ salaries (Miller, Grødeland, and Koshechkina 2001, 318). Concerned about further growth of government and taxes, some people feel more enthusiastic about a solution that would include increased pay but fewer bureaucrats. Public opinion polls have shown that significant groups of Eastern Europeans explain corruption among public officials with the fact that the latter are not paid well. About one-fourth of the public considered the fact that “the government does not pay officials properly” as the main cause of bribery (Miller, Grødeland, and Koshechkina 2001, 75). In Bulgaria in particular, far more people blamed the state for inadequate salaries than the officials for being too greedy. There, public resistance toward possible increase in bureaucrats’ salaries was measured by Miller and his associates as significantly lower than in the Czech Republic and Slovakia. Furthermore, officials in all countries included in the study pointed to poorly paying jobs and insufficient funds as explanations for lowquality service. About 40 percent of the bureaucrats believed that “it is impossible to live on their official salary” and over three-fourths claimed that the state does not provide the necessary funding to secure clients’

132

Political Corruption in Eastern Europe

legal rights (Miller, Grødeland, and Koshechkina 2001, 210, 217). Svetlana P. Glinkina (1998) attributes the frequency with which Russia’s bureaucrats engage in bribe-taking to a combination of insecure and poorly paid jobs.2 Bribing Russian state employees, who in the 1990s were paid on the average about US$200 to $300 a month, was inexpensive and thus it became rampant (Dugan and Lechtman 1998). All this suggests that the public may be more understanding than expected of the idea to secure public sector jobs and thus lessen the attractiveness of bribes as additional compensation. But if wealthier nations can afford to pay their state officials better, do they indeed experience lower levels of corruption? Is there empirical evidence for such a correlation in Eastern Europe? To observe national wealth, I chose to use two indicators, GDP per capita and spending on public officials’ salaries. The former is a proxy that reflects the financial potential of a government to pay for modern office equipment and monitoring and to spend on higher wages in the public sector. The latter is a more direct indicator of compensation of central government employees. I used data from the World Development Indicators of the World Bank and from the Government Finance Statistics Yearbook of the International Monetary Fund. Let us turn to the data now. State financial capacity in Eastern Europe varies within a rather broad range. It is strongly correlated with all measures of corruption; moreover, the regression analysis in Chapter 3 establishes an independent effect of GDP per capita on corruption rates. In brief, there is empirical evidence that more wealthy postcommunist countries are also less corrupt. Among other things that money helps them afford might be the ability to better compensate the work of administrators. This expectation also receives support from the data. The two indicators of material capacity, GDP per capita and wages in the public sector (see Appendix A for variable description), are strongly correlated (r = 0.735, significant at the 0.000 level).3 Bulgaria, the country that spent the least (in absolute terms) on its public sector employees, also had the lowest GDP (especially in 1997). In contrast, the wealthiest country, Slovenia, paid the highest amounts on administrative salaries. From a policy perspective, the important question is whether the above-described capacity to pay more for state employee compensation results in less corruption. Figure 6.1 shows a clear pattern of correlation between these two variables. Indeed, the data suggest that postcommunist countries that spend more money to reward the work

State Capacity

133

Figure 6.1 Corruption and State Material Capacity

Sources: Government Finance Statistics Yearbook (Washington, DC: International Monetary Fund, various years); Anna Grzymala-Busse, Rebuilding Leviathan: Party Competition and State Exploitation in Post-Communist Democracies (Cambridge: Cambridge University Press, 2007); World Bank, The Worldwide Governance Indicators Project, http://info.worldbank.org/governance/wgi/index.asp. Notes: Y axis is World Bank Control of Corruption Index; higher value = less corruption. X axis shows amount spent on public employee compensation weighted by administration size.

of their public officials are less corrupt. This supports the validity of a frequently made excuse of administrators that they would take a bribe because their salary is not sufficient for their household to make ends meet. The correlation statistics here suggest there is less engagement in illegal material exchange in systems with more generous compensation mechanisms. Institutional Capacity By institutional capacity, we usually understand the framework of arrangements established to govern interactions among state administrative branches and citizens (i.e., constitutions and specific legisla-

134

Political Corruption in Eastern Europe

tion). Focusing on the institutionalized rules aimed at changing incentive structures and generating corruption-free behavior, I studied the impact of policy-specific laws. In most postcommunist countries, major anticorruption legislation was adopted between 1997 and 2003 (Dorhoi 2005, 85). Latvia, Estonia, and Ukraine launched official strategies earlier; by 2003, all European postcommunist states with the exception of Russia followed. As the group of countries adopting outright anticorruption strategies grew, the diversity in the measures to which they were committing decreased. Governments copied legislation passed by others, either by adopting new rules or by improving existing provisions. The road to increased institutional capacity through legislation was not smooth. The most serious challenges were faced when considering arrangements for monitoring, conflict-of-interest disclosure, and immunity withdrawal (Dorhoi 2005, 89). Reducing the effect of politics on the work of the prosecution and the courts that were directly involved in investigating and punishing corruption was complicated from the beginning. Where it met severe domestic opposition, pressure from the EU became a decisive factor for administrative reform (Antoaneta Dimitrova 2002). In other places, such as Ukraine and Albania, reform was imposed from the top down and its early timing surprised observers. For a better comparison, I used composite measures of anticorruption legislation constructed and measured for 1995, 2002, and 2003 by Monica Dorhoi (2005). Dorhoi’s Executive and Civil Service Index (ECS) includes information on whether members of the executive should give explanations for their policy decisions; the scope of discretionary funding authority and duration of immunity; arrangements for submission of asset and income declarations by public officials; the scope and breadth of disclosure laws; and existence of provisions for political independence and professionalism of civil servants. For the fifteen postcommunist countries in my study, the ECS scores vary from least intense anticorruption legislation (0.14; Albania in the mid-1990s) to most intense (9.7; Estonia in 2003). Another relevant indicator of institutional capacity is Dorhoi’s Financial Audit and Control Index (FAC). This measure reflects whether there is a national audit agency or other body; its status and degree of independence vis-à-vis other institutions; appointments and professionalism of the staff; powers (coordinating, monitoring, and prosecuting); funding and working standards; and access for citizens

State Capacity

135

to the documentation and findings of this audit structure. Across countries, the index varied between 1.83 (Ukraine in the mid-1990s) and 11.4 (Slovenia in 2002–2003). Both measures of anticorruption legislation are statistically correlated with national rates of elite integrity. Yet the extent of association with FAC is much stronger than that with ECS.4 Figure 6.2 illustrates well how increased financial control corresponds to cleaner politics in the postcommunist world. This attests to the special importance of national audit agencies, their status, and their independence in reducing corruption. Where these institutions were vested with authority not only to coordinate and to verify, but also to prosecute improper behavior and violations of finance regulations, the fight against corrupt practices has been much more successful. In sum, strengthening

Figure 6.2 Corruption and State Institutional Capacity

Source: Monica Dorhoi, “Anti-Corruption Strategies and Fighting Corruption in Central and Eastern Europe,” PhD dissertation, Michigan State University, 2005; World Bank, The Worldwide Governance Indicators Project, http://info.worldbank.org/governance /wgi/index.asp. Notes: Y axis is World Bank Control of Corruption Index; higher value = less corruption. X axis is Dorhoi’s Financial Audit and Control Index of anticorruption legislation; low score = less intense legislation.

136

Political Corruption in Eastern Europe

the institutional capacity of the state by adoption and implementation of civil service reforms and financial control instruments is empirically linked to more transparent and cleaner policies. Technical Capacity Scholars and practitioners seem to agree that the more modernized that countries are, the more success they have in containing corruption. The argument is straightforward: industrialization, urbanization, education, and communication systems improve nations’ capabilities for high-quality government service (Schmitter 2005). These elements of the socioeconomic context contribute to the accumulation of greater state assets that can then be used to effectively detect malfeasance, expose corruption crimes to the public, recruit capable administrators, and provide adequate personnel training. When the overall levels of education are higher, more competent bureaucrats are likely to be appointed who would perform their duties in a professional manner free from political influence. Furthermore, urbanized and industrialized nations with developed systems of modern communications are in a better position to introduce e-government and other advanced forms of administration of public affairs. Consequently, these would save time, increase the quality of service, and cut off opportunities for bribery. One indicator of modernization is the density of a country’s telecommunications network, which is expected to enable effective provision of government services, fewer opportunities for abuse of power, and more direct access of citizens to information. Beverley Earle (2000) suggests that better communication systems facilitate the fight against corruption by allowing broad access to the Internet and, hence, increased transparency. The range of variation of this variable indicates that the Eastern European states differed dramatically in technological capabilities over the years of transition. The weakest in this regard is Albania, where there were 3 telephone lines per 100 citizens in 1997; this number did not rise higher than 11 by 2008. In contrast, the strongest in these terms is Slovenia, where access to the telephone network increased from 36 lines per 100 people in 1997 up to 50 in 2008. As mentioned earlier, these two countries also have, accordingly, the worst and the best records of corrupt practices. Testing for a possible connection to corrupt behavior, I plotted national numbers for telephone lines per 100 people against corrup-

State Capacity

137

tion levels. All three indexes (TICPI, WBI, and NIT) are correlated with the density of the telephone network, but for the purposes of illustration I show only one of the resulting plots in Figure 6.3. It reveals a strong and positive association, suggesting cleaner politics in places with higher density of phone lines. This last indicator of state capacity produces a pattern similar to the ones depicted by officials’ compensation and administrative legislation (i.e., state resource and institutional capabilities covary with corruption).

Indicators of Nonstate Capacity Along with theorizing about the importance of state capacities to achieve policy goals, existing scholarship suggests that nonstate char-

Figure 6.3 Corruption and State Technical Capacity

Source: World Development Indicators, various years (Washington, DC: World Bank); World Bank, The Worldwide Governance Indicators Project, http://info.worldbank.org /governance/wgi/index.asp. Notes: Y axis is World Bank Control of Corruption Index; higher value = less corruption. X axis is telephone lines per 100 people.

138

Political Corruption in Eastern Europe

acteristics of the political system may play a role. In their study of policy capacity, Janicke and Weidner (1997, 7) propose that “the utilization of the existing capacity” in environmental policymaking and enforcement depends heavily on the behavior of nongovernmental “proponents” and various “situative opportunities.” In formulating a similar argument that is applicable to all policy areas, Polidano (2000) emphasizes the importance of nonstate factors, which he dubs “the enabling environment.” As factors that are autonomous to the public sector but may influence its performance, I considered the strength of civil society, cultural predispositions, and ethnic fragmentation. Each one of these carries a potential to condition a state’s capability of dealing with malfeasance. Civil Society Active participation of nongovernmental groups in combating corruption is considered by many as a condition that enables successful implementation, monitoring, and prevention (Johnston 2005a). Theoretically, civil society organizations and the media could exercise oversight with regard to procurement and privatization of stateowned assets by requiring more information from the government and by stimulating openness (Diagnostic Surveys of Corruption in Romania 2001, 34). In other words, good-quality government can become possible only after “society as a whole” demands it (Earle 2000). In Lithuania, and to some extent also in Poland, Estonia, and Latvia, civil society along with all branches of government was actively involved in the design of comprehensive national- and areaspecific antifraud programs (Steves and Rousso 2003). In Bulgaria, Coalition 2000, a network of NGOs, launched a collaborative project to facilitate cooperation among government institutions, civic groups, and individuals for implementation of an anticorruption monitoring system (Coalition 2000, 2005). In contrast, experts on Ukraine and Russia note that domestic financial support for NGOs remained low and this undermined the ability of the civil sector to actively challenge grand corruption (Shelley 2005, 17).5 Reliable and cross-country comparable data on the strength and viability of nongovernmental organizations in Eastern Europe are available from the US Agency for International Development NGO Sustainability Index project.6 Low values of the composite index correspond to a network of more sustainable nongovernmental organiza-

State Capacity

139

tions (i.e., a stronger civil society). Similarly, higher values of the index reflect less organized and weaker civil societies. Among the postcommunist countries that I analyzed, Hungary stands out as a country with the strongest society mobilization (a score of 1.6 for 1998) and Albania as the most immature system in this regard (a score of 4.8 in 1999). Figure 6.4 depicts a pattern revealing that the more viable civil societies in Eastern Europe are also the ones that have lower rates of corruption. Culture Some believe that corruption is a cultural thing that is well rooted and widespread in places where people show more tolerance to the illegal exchange of money and favors in obtaining services (Husted 1999;

Figure 6.4 Corruption and Civil Society

Source: US Agency for International Development, Nongovernmental Organization Sustainability Index (Washington, DC: USAID, 2008); World Bank, The Worldwide Governance Indicators Project, http://info.worldbank.org/governance/wgi/index.asp. Notes: Y axis is World Bank Control of Corruption Index; higher value = less corruption. X axis is USAID NGO Sustainability Index; low score = more sustainable.

140

Political Corruption in Eastern Europe

Lipset and Lenz 2000; Fisman and Miguel 2006).7 This weaker social sensitivity to corrupt practices might be historically or economically predetermined. Regardless of the sources, public (in)tolerance of corruption contributes to the overall capability of a country to reduce malfeasance and improve the quality of public service. When bribing is perceived by many as “a way to get things done” or as “not such a big issue,” the chances to root it out significantly decrease because many people are potential participants in illegal exchanges with public officials. Moreover, weaker sensitivity to the practice of corruption and ignorance of the nature of the harm it causes hinder the active participation of the public as an agent pressing the government to more effectively combat corruption. Another body of research, including studies on Eastern European public attitudes, discounts the role of culture by shifting the attention to the specific post-1989 context of change and uncertainty. For example, William L. Miller, Åse B. Grødeland, and Tatyana Y. Koshechkina (2001, 282) report an average of 43 percent of the public who blame the “moral crisis in a period of transition” for the frequent buying of favors from officials while just 36 percent consider corruption as a “permanent part” of their country’s culture. Interestingly, a closer look at their survey results reveals opposing cross-country patterns. Czechs and Slovaks are more likely to attribute the improper use of gifts and contacts to influence officials to national culture, and Bulgarians and Ukrainians tend to think of public office abuse as a less permanent phenomenon, a feature of the transitional “season of corruption.” Miller’s team does conclude that economic differences across countries explain better than culture the willingness of street-level bureaucrats to engage in improper material exchange. If traditions, religion, and history do indeed shape environments with distinct levels of tolerance of corruption, culture should be accounted for when assessing the capacity of a political system to contain malfeasance. Among many possible indicators, I considered the index developed by Ronald Inglehart and Christian Welzel (2005), which rank orders nations along a secular-traditional continuum. 8 This measurement instrument (IW) assigns higher values to countries with more secular culture and lower values to those where particularistic norms favoring family ties and kin connections prevail. Thus, one would expect that secular cultures enable the state to deal

State Capacity

141

with corruption. In contrast, in traditional cultures citizens more readily justify the use of bribes and accept corruption as a normal state of affairs while politicians and public officials are less scrutinized for violations. When related to corruption, secular-traditional characteristics produce mixed results: the levels of association with the TICPI and the WBI are 0.273 (significant at the 0.001 level) and 0.141 (not significant). A few examples help illustrate why. Poland is an extreme case of a more socially conservative culture (I/W = –0.47), but it is not among the most corrupt countries within the postcommunist sample. Ukraine and Russia, on their part, combine more secularist cultures (IW = 0.84 and 0.87, respectively) and more corruption in the public sector. Estonia and Macedonia fit a little better the expectation for a capacity-enabling effect of culture, the former being ranked as both secular (IW = 1.27) and with cleaner policies and the latter ranked as both culturally traditional (IW = 0.31) and more corrupt. Ethnic Homogeneity Finally, divisions within society may create extra demands for service delivery to particular groups that are met under electoral pressures, but undermine the efforts for fair and honest government. Albeit controversial, some arguments in the literature relate social cleavages to fairness in the allocation of public goods. For example, scholars propose that the presence of ethnic and other minority groups may lead to specific forms of politicization in the public economy sector (Kitschelt and Wilkinson 2007, 39). When these groups get politically mobilized, they are targeted by politicians as potential sources of voter support. In return, such minorities may become preferred recipients of subsidies vis-à-vis the majority group. Ethnic parties, including the Movement for Rights and Freedoms in Bulgaria, acting as power brokers in the formation of coalition cabinets and as revealed in Chapter 5, may get involved in clientelistic networks and remain unpunished. By emphasizing social divisions (especially along ethnic and religious lines), minority elites justify their existence as a separate political actor. Furthermore, by distributing larger pieces of the pie, including public contracts and jobs, these elites keep the discipline among their electorate and secure electoral survival.

142

Political Corruption in Eastern Europe

On a more conceptual level, Herbert Kitshelt and Steven I. Wilkinson (2007, 34) describe a “dynamic of ‘deepening’ clientelism under conditions of intensifying inter-ethnic party competition.” Drawing on Arend Lijphart’s (1977) and Donald Horowitz’s (1985) work, they argue that ethnic pluralism creates fertile ground for the development of political constructs that rely on targeted particularistic favors for winning electoral support. Empirically, this proposition finds strong support in Henry E. Hale’s (2007, 245–246) analysis of Russia where ethnically defined regions emerge as plagued by distinctively higher rates of clientelism. At the level of ordinary citizens, individuals from ethnic minority groups have reported experiences of extortion from public officials at higher rates than representatives of the titular nationality. The suspicion of ethnic discrimination in service provision was described by a resident of Kurdjali, a Bulgarian town with an ethnically mixed population, as a feeling that Turks, Gypsies, and Jews are not served unless they “have money” (Miller, Grødeland, and Koshechkina 2001, 185). Yet not only minority representatives, but also some members of the ethnic majority, suspect officials of ethnic bias. The evidence for actual experience of discrimination is mixed, with minority poverty often negating extortion and titular nationalities registering significantly higher levels of satisfaction with treatment by officials (Miller, Grødeland, and Koshechkina 2001, 188–189). Therefore, the capacity of a country to reduce corruption may decrease in the presence of ethnic divisions. How valid is such an argument? To explore it further, I employed a measure of ethnic heterogeneity that reflects the presence of sizable communities in a country, also known as the effective number of ethnic groups (Ordeshook and Shvetsova 1994; Amorim Neto and Cox 1997).9 Among the fifteen postcommunist countries in my sample, Poland is the least (1.07) while Latvia is the most (2.37) fragmented. Their positioning at the two extreme ends of the fragmentation continuum notwithstanding, these two nations are not that different in terms of corruption. Moreover, the most ethnically divided, Latvia, has experienced average degrees of corruption (WBI mean for 1997–2008 of 0.23). Interestingly, one of the least ethnically divided nations, Albania (1.11), is among the most corrupt (WBI mean of –0.76). Overall, corruption and ethnic fragmentation are correlated in the expected direction, at –0.330 (significant at the 0.000 level) for WBI and –0.293 (significant at the 0.000 level) for the TICPI.

State Capacity

143

Which Capacity? To sum, the above analysis offers empirical support for the role played by material, institutional, and technological capacities of the state to contain corruption. The data also suggest that civil society may be an important enabling factor, a facilitator of education in democratic norms, and a promoter of anticorruption activism. The significant questions that remain unanswered are: (1) Would improvements in state capacity be sufficient to effectively fight corruption or does it all depend on the nonstate enabling environment? and (2) If state capacity makes a difference, which aspect of it matters the most? Intrigued by these questions, I constructed a regression model to test for the independent impact of each capability-building factor while keeping the rest constant. This approach presents an opportunity to isolate the effects of various forms of state capacity and, consequently, to focus on concrete anticorruption policy solutions. The regression model is specified as follows: Corruption = a + b1(wages) + b2(reform) + b3(telephones) + b4(NGOs) + b5(secular) + b6(fragmentation) + e

Corruption is measured yearly, using the TICPI or the WBI, which assign higher values to countries with more transparent and cleaner policies. Wages stands for state material capacity in terms of public employee compensation. 10 It is measured as spending on salaries in the central government sector in thousands of US dollars, controlling for the size of the administration. Reform accounts for the institutional capacity of states. Because Dorhoi’s (2005) indexes are available for three years only, I used a more crude binary measure that reflects whether civil service reform had been introduced. 11 Technical capacity is modeled through telephones, measured as number of telephone lines per 100 people. NGOs accounts for the strength of civil society, the first nonstate capacity factor, operationalized as NGO sustainability and measured through the USAID NGO Sustainability Index. Secular stands for culture, ranging from traditional/particularistic (lower values) to secular (higher values) on Inglehart and Welzel’s (2005) traditional/secular–rational value dimension. Social homogeneity as an environmentally enabling factor is modeled through fragmentation that reflects the degree of ethnic dividedness. This variable is measured as the effective number of eth-

144

Political Corruption in Eastern Europe

nic groups. Finally, a and e are the constant and the error terms, respectively. In summary, in the above model wages, reform, telephones, and secular are expected to be positively correlated while NGOs and fragmentation negatively correlate with corruption. (For a description of variables in the regression model and of their sources, see Appendix A.) The results from the regression analysis, reported in Table 6.1, show consistent support for three of the proposed effects of state and nonstate factors on corruption.12 First, there is empirical evidence that the amounts paid for wages in the public sector matter for preventing bribery and other temptations for illegal enrichment. This is indicated by the positive and statistically significant coefficient for wages in both tests, with TICPI and WBI data. The standard errors are small enough to allow high confidence in this result. Second, the theorized effect of public administration reform as a source of institutional

Table 6.1 The Effects of State and Nonstate Capacity on Corruption in Postcommunist Eastern Europe, 1997–2004 Independent Variable Wages in the public sector Reform of civil service Telephones per 100 people NGOs Secular culture Fragmentation Constant R2 N

TICPI

WBI

.037*** (.009) .516** (.211) .019 (.020) –.274** (.118) .314** (.134) –.784*** (.224) 4.483*** (.748)

.023*** (.003) .316*** (.056) .014*** (.005) –.252*** (.055) .042 (.060) –.147 (.108) .309 (.233)

.826 51

.689 56

Notes: TICPI = Transparency International Corruption Perceptions Index; WBI = World Bank Index. Entries are ordinary least squares regression coefficients, common AR(1) parameter, pairwise estimation, panel corrected standard errors in parentheses. The panel includes observations on Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia. ** p < .05; *** p < .01; two-tailed tests.

State Capacity

145

capacity finds support also in the two sets of data, as shown by the expected positive and significant coefficients. Moreover, the effects of public officials’ work compensation and civil service reform are independent from the influence of other factors shaping a state’s anticorruption capability.13 The third finding is that civil society, as represented by a strong and sustainable NGO network, increases the ability of a country to reduce corruption. This result is indicated by the negative and statistically significant estimate obtained for NGOs in the two models above. The results suggest that a strong and vibrant nongovernmental sector increases the potential of the postcommunist political systems to fight corruption regardless of their material or institutional capacities for which the models that I tested account. The remaining components of capacity received mixed support from the data. The parameters for telephones do indicate a positive effect of denser telecommunication networks on the ability of a state to implement policies in a more transparent way. While the results from the WBI test are more conclusive, the empirical support offered by the TICPI data is not sufficient in statistical terms. Among the nonstate factors, secularism and ethnic fragmentation were also not fully confirmed. The estimates for secular-traditional culture are correctly signed, but the standard error in the WBI model is larger than the coefficient itself. It might be that, after accounting for the viability of civil society, social conservatism loses its impact on the ability of a country to deal with corruption. An extreme case in this regard is Poland, which is the least secular in this sample of postcommunist states and, at the same time, has the most developed and active civil society. 14 Similarly, the role of social heterogeneity as a factor impairing the ability to contain corruption is supported to some extent. The estimated parameters for the impact of fragmentation are correctly signed in the two tests and only in the TICPI model does the coefficient reach statistical significance.

Prospects for Strengthening the National Anticorruption Capacity The results of the regression analysis suggest that when all other forms of capacity are the same, stronger state material and institutional capacity and a vibrant civil society would reduce corrupt practices. These findings are important in several ways. First of all, evi-

146

Political Corruption in Eastern Europe

dence is provided that states that can afford to pay more in employee salaries will benefit from increased levels of elite integrity. Such a conclusion weakens arguments about the prevalence of culture. Those place the emphasis on norms and traditions that take more time to change than the structure of material and other incentives. For the Eastern European states, the challenge is to find and invest resources in their administrations at times of complex transformation. Independent from this effort, strengthening the emerging nongovernmental organizations, societies, and communities would also improve countries’ abilities to identify, punish, and prevent corruption. Because communism suffocated the development of civil society for a long period of time, citizen activism and civic participation have been developing slowly since 1989. I discuss these implications in more detail below. Improving the Incentive Structure Empirical evidence here confirms that better paid administrators engage less often in corrupt behavior; however, salary increase for public officials has not been a simple solution to the problem of corruption in post-1989 Eastern Europe. The development of an incentive system to help mitigate corruption risks has remained problematic in the context of undergoing civil service reform. On the one hand, fiscal constraints made it difficult for many states to increase wages in the public sector and, consequently, to attract a talented and better-qualified workforce. Insufficient funds in the state budgets became the reason for a weak potential to adequately pay civil servants and to sustain bureaucrats’ salaries (Kaldor and Vejvoda 1997; Nunberg 1999). On the other hand, civil service reform itself was postponed in many places and reverse waves of politicization hindered the expected effects of the transition to a merit-based system. Even where employee compensations were raised, salaries remained low due to emphasis on seniority and job security (World Bank 2006a, 12). In general, average rates of compensation may remain low, but targeted financial rewards for professionalism and competence can still effectively prevent corruption. The idea of strengthening the incentive system through wage diversification was accepted in Eastern Europe with reluctance by ruling elites who were unwilling to stop political appointments and by public sector employees con-

State Capacity

147

cerned about fair implementation. As one team of World Bank (2006a, 37) experts found, even in the new EU member states the long-lived tradition of egalitarianism in Eastern Europe hinders the replacement of seniority-driven wage formation with a merit-based compensation model. Yet some Central European countries came up with interesting approaches in an attempt to award the better-performing civil servants and to discourage malfeasance (World Bank 2006a, 12–13). In order to diversify its pay system, Latvia allowed for up to 100 percent additional pay under “management contracts.” Lithuania designed a system for more marginal improvement in compensation, hoping to enhance the state’s competitive position as an employer and to increase its implementation capabilities. Among other reform measures of the customs service in the early 2000s, Hungary allowed for bonus rewards to officials for detecting violations of the customs procedures (Holmes 2006, 232–233). Poland and Slovakia relied on introducing more flexibility in pay levels and other additional stimuli. One unique aspect of reforms in Slovakia was the adoption of the “posts of superior importance” with rigid requirements for asset disclosure, but allowing for 50 to 100 percent additional pay.15 Despite these innovations of targeted employee payments, experts agree that the lack of transparency, clearly defined criteria of bonus distribution, and continued politicization continue to jeopardize the imposition of an incentive system that will effectively limit corruption. Finally, wage diversification based on merit might turn out to be more important than the total amounts that postcommunist states spend on employee salaries. One comparison across subregions shows that in relative terms, the Balkan states have outspent their Central European (including Estonia, Latvia, and Lithuania) and former Soviet (Russia and Ukraine) counterparts.16 What appears as important for reducing corruption is how high individual salaries are and whether an effective diversification scheme of payment is adopted. The size of compensation argument receives strong empirical support from the results of the regression analysis shown in Table 6.1. The few countries that made an effort to encourage competence and integrity through diversification in payment have also lowered their corruption levels. This is demonstrated by the previously discussed cases of Latvia (TICPI scores 2.7 in 1998 to 5.0 in 2008), Slovakia (TICPI scores 3.9 in 1998 to 5.0 in 2008), and Lithuania (TICPI scores 3.8 in 1999 to 4.6 in 2008).

148

Political Corruption in Eastern Europe

Perhaps the strongest challenge to the provision of adequate compensation to public employees comes from the broader public. Civil service employees are still considered in Eastern Europe as belonging to a branch that provides services for the elite rather than for the common people. Under communism, appointments and promotions in the state administration were entirely dependent on party loyalty and seniority. Following 1989, a bureaucratic culture based on loyalty to the state rather than to a particular party or politician remains missing (Holmes 2006, 182). Moreover, corruption scandals in the two decades after 1989 damaged the image of civil servants in the eyes of the public. As a consequence, ideas of raising officials’ salaries, and doing this in times of economic hardship, are difficult for many citizens to accept (Miller, Grødeland, and Koshechkina 2001, 317). Ironically, the latter often refer to the amount of compensation in the public sector as a reason for the spread of bribery and extortion. In a mid-2000s public opinion survey, the second most frequently cited factor leading to corruption was “low salaries,” surpassed only by unemployment (Bulgaria and Poland) and high rates of inflation (Hungary and Russia) (Holmes 2006, 205). Therefore, clear standards and transparency emerge as important conditions in overcoming popular unease about salary increases in a sector otherwise broadly considered as underpaid. Despite the Legacies of Culture Economic factors were important sources of state capacity in limiting corruption regardless of the cultural characteristics of the social environment. This finding is in agreement with Martin Paldam’s (2002) results based on a study of a cross section of nations from all over the world, suggesting that GDP is a far more influential determinant of corruption than culture. How does my analysis relate to the ongoing debate about the role of traditions in the Eastern European context? Leslie Holmes (2006, 185), for example, argues that one of the legacies of communism was that no “culture of compromise” existed to facilitate agreement over immediate reforms whose delay created opportunities for misuse of public power. Rein Taagepera (2002, 254) also blames communism for an “overemphasis on survival values,” which explains the higher rates of corruption in the Baltic former communist countries when compared to Nordic states with no such

State Capacity

149

past. Louise Shelley (2005, 18), who refers to Ukraine and Russia in particular, stresses that in these two nations “the culture of corruption remains predominant over the culture of civil society.” Miller, Grødeland, and Koshechkina (2001, 277) conclude differently, specifying that culture in Eastern Europe matters but not at the system level, where the economy is a stronger predictor of corruption.17 Moreover, there is strong empirical evidence that most Eastern Europeans condemn the use of money and gifts to influence officials (Miller, Grødeland, and Koshechkina 2001, 138) and consider bribe taking as unjustifiable (World Values Survey 2000). My analysis partially supports an argument about the relevance of secular values. Yet the evidence is not strong enough to conclude with confidence that culture predetermines levels of malfeasance and failures of anticorruption policies. By accounting for the impact of other sources of capacity on anticorruption action, I was able to isolate the effect of culture and found its contribution modest. It is worth emphasizing that a distinct Eastern European culture of public-private interactions may be a powerful explanation when comparisons are drawn with consolidated democracies. Secular-traditional cultural differences among the postcommunist states that I analyzed account for some of the variation in corruptiveness in only one of the two tests. Interestingly, regression results for self-expression and survival cultural orientations (Inglehart and Welzel 2005) become even weaker once one accounts for the role of the nongovernmental sector. And my analysis in this chapter provided consistent support for the anticorruption enabling role of sustainable civil society organizations. Therefore, strengthening civil society in Eastern Europe, where it had been suppressed by the communist totalitarian regimes for decades, should be placed higher on the anticorruption agenda. Expanding Civil Society Involvement Similar to other aspects of the postcommunist transition, the formation of citizen self-initiated groups for raising awareness about corruption has not remained unchallenged. Domestic NGOs committed to the cause of exposing corrupt practices and their damaging effect on democracy have been more successful in some places than others. Among the most active are Coalition 2000 in Bulgaria, the Against Corruption Program of the Stefan Batory Foundation in Poland, the

150

Political Corruption in Eastern Europe

Civil Society Development Foundation in Romania, the Hungarian KMonitor, and the local chapters of Transparency International. The Partnership for Social Development, a common project for Croatia and Bosnia, was launched to sharpen citizen sensitivity to corrupt behavior and to facilitate mass media campaigns in this regard (OECD 2002, 58–59). The role of NGOs as whistle-blowers has made a difference, as confirmed by observers (Coalition 2000, 2005) and by the data that I used for the analysis in this chapter. Transparency in government improved where nongovernmental groups were better equipped to communicate their message, to push for anticorruption legislation, and to survive with scarce resources. Yet critics note that the Eastern European anticorruption organizations have not managed to engage larger groups among the public and that, in some cases, they work too “closely with government agencies” (Holmes 2006, 261; Dionisie and Checcki 2008). Perhaps the main reason for these deficiencies, especially of NGOs in the Balkans, is financial viability. Foreign funding remains the primary source of revenue for civic organizations there while national government support is lower and membership fees are rare and miniscule (OECD 2002, 53, 64). The situation is much more troubling in Russia where restrictions on political freedoms and civil liberties, especially in the past decade, have hampered the work of the independent media and the NGOs (Shelley 2005). In my sample of fifteen Eastern European transitional democracies, Russia is the only country with a sharp and significant decline (from 3.4 to 4.4 on the USAID NGO Sustainability Index; USAID 2008) in the rate at which nongovernmental groups provide services, find adequate funding, reach out, and influence public policy. A recent case of pressure by the authorities on the civic sector was the experience of Novoe Vremya, an opposition magazine well-known for exposing high-level corruption. Its premises were stormed and searched by armed and masked police officers in relation to an “unspecified investigation” (“Russian Police Raid Opposition Magazine” 2010). Returning to my finding in the previous section about the capacity-enabling effect of civil society, I can note that in the case of Russia this impact is feeble and uncertain. In an environment of weakened citizen involvement, corruption flourishes despite the vows of Russia’s leaders to defeat it.

State Capacity

151

Conclusion In this chapter, I explored various capacity-building mechanisms expected to increase the potential of postcommunist countries to reduce corruption. Both state and nonstate factors contributed to the performance of the new Eastern European political systems in their effort to provide the population with honest and fair services. Material and financial resources allowed some states to award elite integrity more generously. Where vibrant, civil organizations kept the issue of corruption high on the public agenda and encouraged citizens to expose violations and to insist on transparency in the provision of public services. As I showed in this chapter, culture is not a dominant force that entirely disables Eastern Europe in the fight against corruption. Some have warned about the power of cultural legacies from the past that create an environment conducive to malfeasance and tolerant of abuse of power. By employing multivariate regression techniques, I was able to isolate and assess the influence of culture from the impact of other sources of state and nonstate capacity. The results of my analysis suggest that, within the region, traditional values are not stronger impediments to increasing transparency in government than scarce financial resources and weak civil societies. In order to be more successful in containing corruption, the Eastern European countries need to mobilize resources and to strengthen the nongovernmental sector. Both of these tasks face serious challenges. Insufficient resources and perseverance of the seniority principle postpone the adoption of an incentive system promoting professionalism and integrity among officials. For nongovernmental organizations and the media, preserving coherence and independence is crucial for fulfilling their mission of educating the public and intensifying the anticorruption pressure. Corruption levels declined where the NGOs effectively raised funds, maintained a positive image, and enhanced connections with citizens and the business sector. In this chapter I focused on domestic factors that affected the anticorruption capacities of Eastern European nations to fight corruption. Yet the results of the analysis in Chapter 3 suggested that the international context may also impact countries’ success in reducing malfeasance. I next explore the importance of the process of European integration for adopting and sustaining effective anticorruption measures in the Balkans.

152

Political Corruption in Eastern Europe

Notes 1. As Solnick (1998, 34–36) explains, the Soviet state weakened especially in the last years of its existence when Gorbachev’s reforms needed it to be most effective. Local bureaucrats, empowered by limited decentralization, saw an opportunity and quickly claimed state assets. 2. Glinkina (1998) cites Anders Aslund’s estimates, according to which about 70 percent of Russian bureaucrats took bribes in the 1990s. Private actors paid up to 30 percent to 50 percent of their profits to “socialize” with corrupt government officials. Yet, Glinkina argues, the resulting profitability was sometimes larger than what they would have earned from normal commercial activity. 3. To control for country/administration size, the compensation amounts listed by the International Monetary Fund are weighted by the number of employees in the state administration, as reported by Grzymala-Busse (2007, 241). Since she included only nine of the fifteen states in this study and only until 2004, the number of observations in this part of the analysis is reduced to 69. These countries are Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia. 4. The bivariate correlation coefficients for Financial Audit and Control and Executive and Civil Service legislation with the WBI are 0.502 (significant at the 0.001 level) and 0.390 (significant at the 0.011 level), respectively. 5. Because domestic financial support was poor, foreign aid for Russian and Ukrainian NGOs became very welcome. However, the activities that it spurred were more “a market response to the availability of funding” than a result of incentives originating in society itself (Shelley 2005, 5). Moreover, unfriendly treatment by the authorities has put at risk the survival and operation of groups such as the Soldiers’ Mothers’ Committee and Moscow Helsinki Group. 6. For more detail on the specifics and utility of this index, see Petrova (2007). 7. For a detailed account of the scholarly achievements in linking culture to corruption, see Lambsdorff (2006). 8. To develop the traditional/secular index, Inglehart and Welzel (2000) use World Values Survey items such as the importance of religion in respondents’ life, church attendance, and attitudes toward work, parent duties, divorce, good and evil, the family, and women’s rights. 9. The effective number of ethnic groups is calculated as an index of fragmentation by the formula 1/gi, where g is the proportion of group i from the total population. Unlike the raw number of groups, this measure gives weight to contexts with large ethnic communities. 10. The other indicator of material capacity, GDP per capita, is strongly correlated with the amounts spent on public employees’ salaries (r = 0.735, significant at the 0.000 level). This is not surprising because wealthier countries can usually afford to pay their public servants better. For reasons of multicollinearity, these two variables cannot be entered in the model simultaneously. Between the two, I prefer to include spending on salaries for a more

State Capacity

153

direct test of the proposition that public officials resort to improper deals when their income is low. 11. This variable is coded with a one-year lag to allow time for the start of implementation. 12. To ensure robustness of the results, I considered a control variable of democratization (political freedoms and civil liberties, combined rankings of Freedom House, www.freedomhouse.org) and test for possible presence of multicollinearity. For most of the independent variables, the generated Pearson correlation coefficients are not of a magnitude that would raise serious concerns. However, the control variable democratization was strongly correlated with NGOs (r = 0.745, significant at the 0.000 level) and I excluded it from the analysis. In a separate set of tests, I controlled for democratization and dropped the civil society variable; the central findings remained unaffected. 13. The World Bank’s (2001) report on Romania and Rauch and Evans’s (2000) research on thirty-five less developed (non-European) countries found no association between public officials’ salaries and bureaucratic performance. What are the reasons for the discrepancy between these and my results? I argue that there are important research design differences: different levels of analysis, samples, and measures. The World Bank study worked at the microlevel, only on Romania, and with a measure of compensation based on bureaucrats’ self-assessments. Rauch and Evans’s analysis was at the system level, cross-national and non-European, and used the proportion of payments from total revenues to measure compensation. Two additional points need to be kept in mind. First, it is important that the proposed impact of material capacity of the state in general (for salaries, equipment, training, etc.) is strongly supported by the data in this chapter. Taking into account other factors, the data still confirm that the amount of resources spent on compensation matters. Future research needs to expand and include other postcommunist countries and update the measure of payment at the system level. Second, both studies cited in this note underscore the role of meritocratic recruitment and transparency for reducing corruption and enhancing bureaucratic performance. I discuss this issue in combination with attempted diversified payment in some Eastern European countries in the following section. 14. To test the robustness of this result and its sensitivity to a single outlier such as Poland, I used another indicator of culture that reflects to what extent national culture is self-expression versus survival oriented. As is argued in the literature (Sandholtz and Taagepera 2005), the economic hardships of the last decade of communist rule made many people uncertain about the well-being of their households, thus increasing the importance of personal and economic security in popular thinking. The self-expression values indicator does not change the results estimated with secular culture and reported in Table 6.1 (i.e., officials’ compensations, administrative reform, and civil society remain significant determinants of corruption levels while culture lacks empirical support in both tests). 15. These posts include the head of the State Treasury, the head of the Debt Management Agency, the chief economist at the Ministry of Finance,

154

Political Corruption in Eastern Europe

the head of the Anti-Corruption Unit at the Government Office, and the head of programming of structural funds at the Ministry of Labor, Social Affairs and Family (World Bank 2006a, 13). 16. What Albania, Bulgaria, Croatia, and Romania spent on public employee wages constitutes an average of 16 percent of their central government expense and 24 percent of their central government revenue. The respective numbers for Russia and Ukraine are 14 percent and 18 percent; and for the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia, 13 percent and 20 percent (my calculations based on statistics from IMF, Government Finance Statistics Yearbook, various years). 17. Miller, Grødeland, and Koshechkina (2001, 261) found that the “bargaining power of officials vis-à-vis their clients” is the kind of culture that determines opportunities and practices of corrupt exchanges in their sample of Eastern European states, including the Czech Republic, Slovakia, Bulgaria, and Ukraine.

7 EU Accession: Comparing Balkan Experiences

Indeed, the desire to meet European standards is a motivator for governance improvements far beyond the borders of the EU. —James H. Anderson and Cheryl W. Gray, Anticorruption in Transition 3: Who Is Succeeding . . . and Why?

The integration of postcommunist states into the European Union is considered by many as an influential factor with potential to promote democratic government in places heavily burdened by a totalitarian past. Proponents of this thesis believe that hopes in Eastern Europe to join the club of wealthy democracies encouraged reforms and inspired nations to sacrifice in the face of severe economic hardships. Skeptics, however, caution that the EU’s leverage is not that certain at all; they warn that, once membership is achieved, new member states have little or no incentive to comply. One of the most important areas in which European integration has been called on is transparent and fair government. While reducing corruption was not an explicit requirement for accession in the past, more recently it has emerged as a central issue in the process of Eastern European enlargement. As clean government becomes important for new EU entrants, integrationists expect that the more countries get integrated, the more they will enforce anticorruption policies and comply with promises made to the Union. But skeptics question the rigor of such commitments.

155

156

Political Corruption in Eastern Europe

In this chapter, I subject to yet another test the proposition that European integration has had a positive impact on the quality of governance after communism. I explored my preliminary empirical results from Chapter 3 in greater depth in order to determine the impact of EU integration at different phases of the accession process. First, I analyzed the mechanisms through which international organizations and the EU may affect national anticorruption policies. Then, I developed a research design that allows comparison of levels of corruption and anticorruption effort in three groups of postcommunist countries, all from a geographic area with a history of corrupt practices—the Balkans. These countries are similar in many ways, but differ with regard to their progress toward EU membership: Serbia and Albania (not yet candidates), Macedonia and Croatia (candidates), and Bulgaria and Romania (EU members since 2007). My analysis compares success in legislative reform and changes in corruption levels across space (between more and less integrated countries) and over time (at different phases of the accession). I conclude by assessing the effectiveness of the EU’s leverage in promoting clean and transparent government.

International Integration and Transparent Government Since the mid-1990s when corruption emerged as a prominent issue on the political agenda worldwide, international institutions have played a significant role in promoting policy transparency and in raising public awareness. External pressure on national governments to pay attention to conflict of interest issues, improve financial control, and relieve the regulatory burden on private firms urged elites to consider more stringent anticorruption instruments. In transitional democracies actively seeking integration in international economic and defense structures, external influence fostered adoption of legislative and administrative reforms aimed at overcoming the legacies of authoritarian leadership and clientelistic networks (Anderson and Gray 2006, xviii). The desire to become part of the world’s family of democracies and to enjoy the benefits from international cooperation motivated countries to set clear boundaries for the influence of powerful private interests that took advantage of the uncertainties of transition.

EU Accession

157

Among the international actors who first exercised direct pressure on national governments to fight corruption are the United Nations, the Council of Europe, the Organization for Economic Cooperation and Development, and the World Bank.1 They initiated negotiations and facilitated the conclusion of agreements among member states to commit to the fight against corruption. Recognizing bribery as a phenomenon with huge economic and social costs on societies all over the globe, in 1998 the OECD reached an agreement for a treaty that required countries to adopt national antibribery legislation. The OECD Convention on Combating Bribery of Foreign Public Officials in International Business Transactions was signed by all member states and five nonmembers, and entered into force in 1999.2 At about the same time the Council of Europe, an organization of forty-one European states, launched its own anticorruption initiatives. In 1998, the Council established the Group of States against Corruption (GRECO), which was conceived to help fight and reduce corruption through peer monitoring and pressure (Earle 2000). One of its first achievements was the passing of the Criminal Law Convention on Corruption, signed by twenty-nine European countries in 1999. On a more global scale, the World Bank emphasized the importance of ethics and equal opportunities for all market participants and created in 1998 the Oversight Committee on Fraud and Corruption and the Sanctions Committee to monitor and punish contractors in violation of antifraud and antibribery rules. These steps provided more adequate instruments of financial control in the Eastern European context where access to the Bank’s funds was conditioned on evidence for increased public oversight and administrative reform (Earle 2000). Finally, the UN Convention Against Corruption of 2003 was aimed at promoting integrity in public affairs, facilitating interstate cooperation, and supporting technical assistance to reduce and combat corruption. Signatories to the convention committed to passing legislation that criminalizes bribing of public officials and trading in influence and to cooperating with peers in the implementation of such policies.3 The postcommunist states joined these initiatives and, in many instances, they were among the founding parties of the international anticorruption treaties. The record of ratification of such agreements reveals overall high rates of promise to eradicate bribery and other forms of malfeasance (see Appendix B). Especially interesting is the

158

Political Corruption in Eastern Europe

case of Eastern European participation in the OECD convention, one of the first conventions designed to motivate countries to cooperate in combating corruption. Although only the Czech Republic, Hungary, and Poland were formally members of the OECD, Bulgaria and Slovakia volunteered and also ratified this organization’s antibribery agreement. While the postcommunist governments have clearly demonstrated readiness to join the global fight against corruption, it remains uncertain how successful they have been in meeting their promises. The general literature on compliance with international treaties recognizes that only declaring good intentions for cooperation is not sufficient for achieving the goals of the negotiating parties. On the domestic level, enacting and implementing policies in compliance with treaty requirements may involve high political costs for incumbents (Tallberg 2002; Schimmelfennig 2008). On the international level, weak mechanisms for monitoring and inadequate instruments for rewarding compliance and punishing deterrence may put at risk the attainment of goals targeted through cooperation (Keohane and Martin 1995; Downs, Rocke, and Barsoom 1996). The extent to which international cooperation has helped the young transitional democracies combat corruption is therefore partially dependent on the conditions of engagement, including institutional traits facilitating transparency and encouraging enforcement. Studying the specifics of some international incentives, scholars note that organizations such as the OECD, the Council of Europe, and the UN played an active role in making more countries recognize the threat of corruption, but were less effective in forcing them to clean up their government. Beverley Earle (2000, 490), for example, writes that the OECD is “designed as a forum for discussion, not action.” While she admits that promoting values such as pluralism, respect for human rights, and transparency is important for encouraging the Eastern European countries to pass anticorruption legislation, the question about actual implementation remains open. Similarly, the Council of Europe’s GRECO provides for mutual evaluation through on-site visits, recommendations, and peer pressure, but falls short of including sanctions for noncompliance. 4 Interestingly the World Bank, which was initially blamed for distributing resources with no adequate controls on whose pocket the money ended up in, developed in transitional countries a critical

EU Accession

159

leverage in the fight against corruption (Earle 2000). Facing a credible threat of losing much-needed funds for growth, national governments become motivated to accept the Bank’s country-specific programs designed to make corruption a high-risk activity that may jeopardize the inflow of external aid. The European Union is another organization that can use powerful positive incentives to encourage honest policymaking in Eastern Europe. To facilitate compliance, the Union can rely on a combination of “enforcement” and “management” mechanisms (Tallberg 2002). Since the very beginning of transition, EU membership has been a highly desired goal in the foreign policies of the postcommunist countries. The expected benefits from preaccession and structural funds as well as eventual membership can be used effectively by the EU as rewards for fulfilled requirements to reduce corruption. At the same time, the threat of being left out may generate strong incentives in candidate states to clean up government because failure to become integrated would incur significant losses in transitional economies that are striving to grow. To use Earle’s (2000) terms but in a different order, the EU is more a forum “for action” than “a forum for discussion.” Unlike the OECD or the Council of Europe, it is in a position to use a credible stick-and-carrot strategy. I discuss next how the EU shaped its approach to inspire rigorous anticorruption policies through the leverage of Eastern European enlargement.

Mechanisms of EU Influence By making democracy and the rule of law necessary conditions for future membership, the EU has been able to promote clean government policies in countries that are candidates for accession. It is the attractiveness of the prospect for accession that can make the impact of the Union stronger than any other external factor. The promise for membership is a powerful driving force that can help overcome domestic opposition to unpopular reforms (Vachudova 2009). By making membership feasible, the Union can motivate incumbents to adopt and implement more stringent policies in endorsing honest governance (Schimmelfennig 2008). Two theoretical frameworks attempt to explain the effects of EU conditionality. The first is an incentive-based argument, formulated

160

Political Corruption in Eastern Europe

from a rational choice perspective. It proposes that the EU conditionality policy will be effective if: (1) the offer of membership is credible; and (2) the domestic costs for compliance with the Union’s requirements are not considered too high by the ruling elite. Also known as the conditionality hypothesis, this proposition turns the attention to what makes an offer credible and which costs are dangerously high for a government to manage (Epstein and Sedelmeier 2008, 796). Scholars have proposed that a credible payoff structure should be in place in order for conditionality to work; that is, rewards should be distributed to encourage further progress and the prospects for applying sanctions should be real (Noutcheva and Bechev 2008). Another argument is made by constructivists who advance the socalled socialization hypothesis. According to the latter, the success of accession can be explained through a process of persuasion and social learning, rather than calculations for membership benefits (Sedelmeier 2008). Constructivists also attribute compliance with EU requirements to the legitimacy of the international organization and to the social context in the target country. How uniform has the impact of EU accession conditionality been on postcommunist politics? Extant literature offers tentative answers to the question of possible temporal, spatial, and issue area variation in the effectiveness of the EU’s accession policy. There seems to be agreement that the transformative power of the EU is not uniform throughout the process of accession. Milada Anna Vachudova (2009) argues that the Union’s leverage is stronger at “earlier junctures” such as signing association agreements and opening accession negotiations. Tim Haughton (2007) is more precise when specifying that the conditionality policy is most successful at the “decision phase” or when the EU decides on opening negotiations for accession with particular countries. On his part, Frank Schimmelfennig (2008) expects that only at the final stage of negotiations would governments pursuing integration be able to overcome domestic costs threatening their survival and to fulfill the EU’s recommendations. When it comes to cross-national differences, research has shown varying success in the integration process across the groups of “early reformers” (e.g., Hungary, Poland, and Slovenia) and “late reformers” (e.g., Bulgaria and Romania) (Vachudova 2005). Thus, concerns that domestic political and sociocultural contexts shape different outcomes have determined the EU’s decision to assess separately each country’s progress on the road to accession. Finally, the degree of compliance may vary by policy area, as suggested by

EU Accession

161

Haughton (2007), with more success registered when demands are clearly defined and rewards are tied to particular achievements. Given all we currently know about EU political accession conditionality, how effectively can the Union promote the adoption of anticorruption legislation and motivate governments to fight corruption? Where and when does the EU make more of a difference when it comes to the difficult issue of transparency and clean government? If the rationalist argument is valid, one would expect that the influence of conditionality is at its peak when the decision for the opening of accession negotiations is made; that is, at the point in the integration trajectory when accession becomes irreversible (Haughton 2007). Since the greatest reward is membership, compliance may weaken after accession is secured. Sanctions for not fulfilling the EU’s policies do exist for member states, but those are not as much of a threat as is delaying or denying membership (Sedelmeier 2008). Therefore, we can expect decreasing effectiveness of EU conditionality on the national anticorruption effort after accession negotiations begin and especially after their conclusion. Empirical studies on postenlargement compliance have produced mixed results. Some researchers have found unexpected durability in new members’ fulfillment of Union requirements in certain policy areas (Sedelmeier 2008).5 If those results are confirmed by future data, this would mean that, in the absence of the credible membership incentive (from the preaccession period), countries live up to their promises because they have already developed an institutional capacity or a habit of compliant behavior. Furthermore, a postenlargement success would support the constructivist proposition for the importance of persuasion and social learning. Studying compliance with EU policies in the newer member states Bulgaria and Romania, however, Vachudova (2009) warns that success is not certain at all and recommends that leverage be exercised “well before accession.” Whichever argument holds in the future, it appears that scholars now agree that the new norms promoted by the EU and adopted in Eastern Europe through short-term cost-benefit calculations will not endure if the standards do not penetrate societies to cause behavioral change (Epstein and Sedelmeier 2008; Sasse 2008; Sedelmeier 2008). Have there been clear demands and credible conditionality in the EU accession approach when it comes to reducing corruption? The Union did not have explicit anticorruption policy standards for member states and did not include such in the 1993 Copenhagen criteria

162

Political Corruption in Eastern Europe

for countries aspiring for membership. Later on, as the enlargement wave spread to Eastern Europe, the European Commission of the European Union recognized the magnitude of the corruption problem and its possible implications for the Union in the future. The new candidates had additional problems that required more specific demands for curbing corrupt behavior and different tools to induce compliance. Consensus began to emerge within the EU to insist on decisive anticorruption measures in candidate states (Vachudova 2009). Especially in the case of Bulgaria and Romania, the preparations for final granting of membership status were dominated by pressure from the Union on the two candidates to intensify the anticorruption effort.6 It is included as an integral part of the requirements for accession of any new applicants. The set of tools designed by the EU to influence membership candidates, especially the punishments for failure to meet the Union’s demands, became more diverse. From negative assessments in progress reports, through delaying association and accession negotiations, to withholding EU funds, these instruments were aimed at sending clear signals that noncompliance had a cost.7 As Gergana Noutcheva and Dimitar Bechev (2008) observe, in the past few years the EU has more effectively communicated its commitment to punish, if necessary, failures to fulfill promises. An empirical evaluation of the above theoretical constructs would help to establish their validity. If the rationalist argument for incentive-based behavior is correct, the data should show an increase in compliance with EU demands for transparent government as membership becomes more feasible. This is because applicants for membership are more committed to adoption and implementation of anticorruption policies when they receive credible signals that progress is rewarded fairly and defection is harshly punished. Once accession is achieved, the EU leverage in sustaining the anticorruption effort weakens. If the socialization argument is correct, the intensity of the fight against corruption should be preserved after EU accession.

EU Conditionality and the Fight Against Corruption in the Balkans When choosing an appropriate research design for the empirical analysis, I faced the imminent question of case selection. Focusing on

EU Accession

163

a few cases, instead of opting for a statistical cross-national evaluation of a larger set of countries, allowed for a more detailed inquiry into the dynamics of conditionality and its impact. Two needs emerged as demanding special attention here: (1) choosing states that were at different junctures in the EU integration; and (2) control for characteristics of the context in which the conditionality mechanism was applied. For the purposes of this analysis, I selected six Balkan postcommunist states: Albania, Bulgaria, Croatia, Macedonia, Romania, and Serbia. They share many similarities in regard to history and culture. At the same time, these countries are at different stages in the process of EU accession: Bulgaria and Romania, the most integrated so far, have already attained full membership; Croatia and Macedonia recently advanced to a candidate status; and Albania and Serbia, association treaty signatories, are at an earlier phase of accession. Controlling for other possible influences made it possible to better isolate the expected effects, if any, of the EU conditionality mechanism. Common Experiences We often start with history when in search of similarities and differences across countries, and this study is no exception. If there is a place on Earth where history is never forgotten, it is in this corner of Europe. The Balkan nations share a common past both of being part of the Ottoman Empire for about five centuries and of spending another forty-five years under communism. These two experiences left marks on the way people perceive the state, always as a foreign body created to oppress and to extract taxes rather than to serve the needs of the population. This alienation from the state accounts for a widespread and deep mistrust of institutions, a disbelief that officeholders can and should be held accountable, and a conviction that stealing from the oppressive state is justifiable. After centuries of existence in the peripheries of despotic political systems with weak bureaucracies, the levels of corruption in the Balkans have remained higher than in Central Eastern Europe. In the 1990s, the twin legacies of communism and ethnic wars linked together some politicians, military officers, and private actors in corrupt semilegal networks (Srdoc and Samy 2009). Corruption is widely perceived as unacceptable in principle, yet many are ready to make compromises in order to solve everyday problems (Coalition 2000,

164

Political Corruption in Eastern Europe

2001, 7–8). Some experts believe that the bakshish culture from the past is still alive, tainting the Balkan states’ political systems and threatening their democratic transformation.8 A significant body of scholarly literature shows that democratic transition developed at a different pace and with distinct outcomes compared to the group of Central European postcommunist states. For example, electoral system reform occurred in Central Europe in the first year after the communist collapse while it took much longer for Macedonia, Croatia, Bosnia and Herzegovina, and Serbia to change the majoritarian rules inherited from communism (Kostadinova 2000; Birch 2003, 20).9 Also, unlike the Hungarian and Slovenian ruling elites who passed legislation for civil service reform in the first half of the 1990s, transforming public administration did not begin in Southeastern Europe before 1997 (GrzymalaBusse 2007). In the economic arena, macrostructural stabilization and privatization started early in the transition and at an accelerated speed in Central Europe; the Southern European postcommunist states kept postponing the start of transformation toward a market economy. Even when finally given a green light, change continued to be slow. There are certainly good reasons for the delay in reforms and in democratization in general. Although also coming out of a significant period of communist government, Southeastern Europe had lower social and economic indicators than its immediate neighbors to the west. In the past, ethnic heterogeneity and its challenges were dealt with by the help of ideology. After the fall of the Berlin Wall and the disintegration of the Soviet bloc, tensions could no longer be artificially subdued; in the former Yugoslavia, they were even used by incumbents to solidify power. The totalitarian regimes in Southeastern Europe were much stiffer and coherent and civil society was weaker and poorly organized; nationalism there reemerged stronger and more destructive than in Central Europe. In the absence of rigorous political competition in the early 1990s, ruling elites could entrench in power, postpone the rebuilding of state institutions, and secure further exploitation of public resources through private informal groupings (Grzymala-Busse 2007, 132). The combination of “concentrated rule” and “partial economic reform” inevitably led to the capture of the state by powerful private interests (Vachudova 2009). The intensity of these trends varies

EU Accession

165

across the Balkan region but together, as a group, these states differ in some ways from their Central European and Baltic counterparts. One indicator of the different outcomes of transition is their slow fulfillment of EU accession requirements (political democratization and marketization) and, subsequently, a belated schedule of EU accession. Even the most successful in the Balkans, Bulgaria and Romania, are often referred to as “the laggards” in the latest EU enlargement wave in Eastern Europe (Noutcheva and Bechev 2008). The social and political context of the delayed transition in the Western Balkans has recently raised questions about the ability of the EU to induce compliance in countries that have aspired to membership (Schimmelfennig 2008; Vachudova 2009). Ethnic tension and its resolution appear to be the most difficult obstacle to effective application of the conditionality mechanism in countries such as Macedonia, Croatia, and Serbia. Since this issue may significantly increase the domestic costs of compliance with the EU policy requirements, incumbent elites perceive it as a threat to their political survival and yield to the pressure. Balkan ruling elites are pressured by nationalistic circles to secure the integrity of the state despite promises made to the EU to respect the human and civil rights of smaller ethnic groups. A similar outcome is possible in fighting corruption. Calls from the EU to fulfill obligations to curb corruption and sanction malfeasance are overpowered by influential domestic rent seekers. If Schimmelfennig (2008) is right, chances for the Western Balkan countries to break this dependence and reduce the domestic costs of compliance with the EU exist only in the final decisive stage of accession when the payoff is significant. In brief, because of the distinctive social, economic, and political specifics of the region, one can expect that the effects of the EU on reducing corruption in Southeastern Europe are weaker than in Central Europe and the Baltic states. Historical legacies make the Balkan societies more corrupt in the first place and the problem that governments have to deal with is more severe and widespread than in Central Europe. Because the domestic costs of compliance are higher in the Western Balkans, the success of anticorruption policies is weaker than in Romania and Bulgaria at similar phases of accession. But even in such difficult circumstances, the membership incentive may motivate governments to limit corruption on the eve of accession negotiations.

166

Political Corruption in Eastern Europe

EU Integration Dynamics in the Balkans All six Balkan transition countries have aspired to EU membership after the fall of communism, but their accession has followed a very different schedule (see Table 7.1). A closer look at the timing of important junctures in the EU enlargement process reveals three groups or, in Schimmelfennig’s (2008) terms, “eligibility waves” that roughly correspond to the overall success in democratization. The more successful, Bulgaria and Romania, signed Association Agreements and submitted formal applications by the mid-1990s; they have been full members of the Union since January 2007. Croatia and Macedonia followed by signing Stabilization and Association Agreements in 2001, filing membership applications in 2003–2004, and receiving a candidate status in 2004–2005. Croatia has made significant progress so far by closing a number of negotiation chapters to harmonize its policies with the EU; Macedonia has not yet started actual negotiations.10 The third group, Albania and Serbia, signed Stabilization and Association Agreements in 2006–2008 and submitted formal membership applications in 2009, but neither country has yet been granted a candidacy status. In sum, the first wave moved from association to negotiations (the phase when the conditionality mechanism is expected to be more effective) between 1993 and 2000, the second between 2001 and 2005, and the third started in 2006 but has not yet reached the critical juncture of negotiations. The above record shows that, despite talks about enlargement fatigue, the European Commission has been quite consistent in its

Table 7.1 Progress in EU Accession in the Balkans Country

Association/ SA Agreement

Membership Application

Candidacy Granted

Accession Agreement

Bulgaria Romania Croatia Macedonia Albania Serbia

March 1993 February 1993 October 2001 April 2001 June 2006 April 2008

December 1995 June 1995 February 2003 March 2004 April 2009 December 2009

December 1999 December 1999 June 2004 December 2005 — —

January 2007 January 2007 — — — —

Source: European Commission, “Enlargement,” n.d., http://ec.europa.eu/enlargement, accessed 19 January 2010. Note: SA = Stabilization and Association.

EU Accession

167

policies toward the postcommunist applicants for membership and the Southeastern European countries in particular. Moreover, fears of discriminatory exclusion of the troublesome Balkan region did not materialize (Schimmelfennig 2008). The six countries in my analysis have been offered the same membership incentives as the rest of the former communist bloc. To facilitate progress toward integration, the Union provided each country with access to financial assistance programs: PHARE, Special Accession Programme for Agriculture and Rural Development, and Instrument for Structural Policies for PreAccession for Bulgaria and Romania and the Community Assistance for Reconstruction, Development and Stability in the Balkans (CARDS) for the Western Balkans. The funds were distributed to strengthen the new democratic institutions, rebuild public administrations, and sustain the effort to preserve the rule of law and promote civil society. 11 Focusing on institution building, these programs offered assistance for professionalization of state bureaucracies and depoliticization of the judiciary, both being of crucial importance in the fight against corruption. The positive incentives of Association Agreements and negotiations, as well as material assistance, were balanced by requirements for commitment to ongoing reform and democratic stabilization. It has been a learning process for both membership aspirants and the EU. Over time, the EU diversified the conditionality instruments to facilitate long-term incentives for cooperation even in the more sensitive and challenging policy areas such as promoting policies of minority rights and ethnic reconciliation and curbing corruption (Vachudova 2008). At the same time, the European Commission has shown that failure to comply with its requirements would become the subject of sanctions, including delay of negotiations. For example, association with Serbia and beginning accession negotiations with Croatia were made conditional on their governments’ full cooperation with the International Criminal Tribunal. Thus, Croatia’s negotiations were postponed until General Ante Gotovina was finally arrested and sent to The Hague in December 2005, an unpopular measure in the eyes of ordinary Croats. Similarly, concluding the Stabilization and Association Agreement with Serbia had been suspended until the government in Belgrade detained a high-ranking military officer sought by the International Tribunal (Schimmelfennig 2008). For Macedonia, ratification of the Stabilization and Association

168

Political Corruption in Eastern Europe

Agreement with the EU did not happen until a peaceful resolution of the conflict between the Albanian ethnic minority and the state occurred in late 2001 (Risteska 2007).12 At the last phase of their accession, Romania and Bulgaria had to face unusually strict conditions issued by the European Commission; for the first time in the history of enlargement, membership was linked to success in the fight against corruption. By 2005, it was clear that the only two Balkan countries with early signed Association Agreements had been experiencing the negative consequences of state capture by informal networks in collaboration with organized crime (Noutcheva and Bechev 2008; Vachudova 2009). It appears that association and financial aid through PHARE had not effectively persuaded the governments in Sofia and Bucharest to fully enforce norms of transparency and integrity in the public sector. The EU chose new tools to exercise its leverage and push for anticorruption measures. Accession Agreements with Bulgaria and Romania were signed in 2005, but a “cooperation and verification mechanism” was imposed to facilitate the entry and also to safeguard the Union itself (“Report from the Commission to the European Parliament” 2007). Unusually “detailed” progress reports were issued to reveal the sources of trouble and provide guidance for government action; teams were sent to monitor and assess the situation (Vachudova 2009). Because corruption became recognized as a widespread phenomenon threatening the democratic process in both countries, a postponement clause was included in the accession treaties allowing for a one-year delay of full membership if the governments fail to implement the recommended changes. Although the clause was not enforced and Bulgaria and Romania did formally join the Union in January 2007, the European Commission reserved the right to monitor progress in the reformation of the judiciary and in the fight against corruption after accession. This “unprecedented measure” (Noutcheva and Bechev 2008) and the resolute manner in which it was introduced in Bulgaria and Romania sent a signal to the Western Balkan candidate states: the EU was determined to impose preconditions that link accession to fulfillment of obligations, including combating corrupt behavior. How effective these measures have been in promoting clean government is an empirical question to which I turn next.

EU Accession

169

Anticorruption Policymaking: Legislative Reform Adoption of anticorruption policies was not explicitly specified in the EU enlargement strategy of enforcement of European Community standards, also known as the acquis communautaire (Vachudova 2009). One reason for this was the absence of such established legislative standards for member states. Yet to satisfy the Copenhagen 1993 political criteria for admission eligibility, the postcommunist aspirants for membership were required to undertake measures, including legislative ones, to enhance transparency in democratic policymaking. Also, under EU pressure they joined international agreements and committed to passing anticorruption laws and regulations, something that not all older member states had done (Monitoring the EU Accession Process 2002, 52). A significant amount of anticorruption legislation was adopted in the decade between 1998 and 2008, although with different intensity and priority areas. Among the factors causing this variety emerges the status of integration with the EU (Dorhoi 2005). As an applicant for EU membership, Bulgaria, a non-OECD country, became a founding party to the 1998 Convention on Combating Bribery of Foreign Public Officials in International Business Transactions. Most of the states that were granted candidacy status in 1999 became founding parties to other international anticorruption treaties (Appendix B). But it took eight years for Russia, a member of the Council of Europe but with no prospects for EU integration, to join the Group of States against Corruption. Overall major anticorruption legislation in Central Eastern Europe was passed between 1997 and 2003, and the pathway of institutional reform was different across the region. Experts argue that over time, the policy instruments of governments to combat corruption improved and that countries “imitated” each other in developing the much needed legislative tools (Dorhoi 2005, 86). This line of research also establishes that states that had become Council of Europe members before 1995 and gained EU candidate status by 2002 also passed more comprehensive anticorruption programs. Franklin Steves and Alan Rousso (2003) stress that not only did the European Commission exert pressure on accession countries to pass anticorruption legislation, but it also provided them with assistance to craft more thorough and stringent policies.

170

Political Corruption in Eastern Europe

How do the Balkan states compare on the adoption of anticorruption legislation? If the integrationist argument is valid, Bulgaria and Romania should be expected to have earlier and more comprehensive legislation passed as they gained EU candidate status before 2003. Albania, Croatia, Macedonia, and Serbia, less advanced in the accession process, could be expected to have been slower and less rigorous in the making of anticorruption policies.13 Table 7.2 presents a time map of the adoption of seven institutional mechanisms aimed at facilitating public accountability; namely, legislation on prevention of money laundering, on the citizen right of access to information, on assets disclosure of public officials, on public procurement, and on conflict of interests as well as national strategies to combat corruption. Those were recognized by World Bank and Council of Europe experts and by the European Commission as important tools enabling countries to make governance more transparent, reduce the occurrence of financial misconduct, and prevent exploitation of public resources by officeholders with positions in the private sector (Steves and Rousso 2003). Omnibus reform, such as the introduction of a national strategy for anticorruption activities, was also strongly encouraged by international organizations as a way to strengthen the institutional capacity of the state to deal with the complex problem of power abuse. The time map of anticorruption policymaking in the Balkans reveals no clear patterns. As expected, Bulgaria and Romania began earlier, with laws on punishment of money laundering and assets disclosure adopted in 1996.14 They became more active in adopting anticorruption legislation between 1999 and 2002 (i.e., in the years preceding the EU decision on inviting candidate states to join the Union). In both countries, enacting conflict of interest legislation was postponed even after accession.15 The Bulgarian parliament passed such a law in October 2008, after a large corruption scandal at the national transportation agency blew up earlier that year. The investigation revealed shocking cases of misuse of EU funds and triggered the freezing of millions of euros in subsidies for Bulgaria. The European Commission promised to resume payments only after Bulgaria secures transparent distribution of resources in the future. The adoption of the Law on Prevention and Disclosure of Conflicts of Interest was an urgent response to the unprecedented sanction exercised by the EU on Bulgaria, a mem-

EU Accession

171

Table 7.2 Time Map of Adoption of Anticorruption Legislation Countries Year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

BG

RM

ML

DA

PP DA NS

ML

CR

MC

AL

SB

DA

AI, PP, NS

AI ML NS, PP AI, CI, (DA)

AI

AI PP, CI

(CI) CI

DA NS ML

NS

CI PP

PP CI, AI, (DA) ML, NS

ML

Sources: Public Accountability Mechanisms (Washington, DC: World Bank), www.agidata.org/pam/Methodology.aspx; LegislationOnline (Warsaw: Office of Democratic Institutions and Human Rights, Organization for Security and Cooperation in Europe, n.d.), http://legislationonline.org; The World Law Guide (Lexadin, n.d.), www.lexadin.nl/wlg/legis /nofr/legis.php. Notes: Countries: AL = Albania, BG = Bulgaria, CR = Croatia, MC = Macedonia, RM = Romania, SB = Serbia. Legislation: AI = Law on the Right to Access to Public Information, CI = Law on Conflict of Interests, DA = Law on Declaration of Assets and Income by Public Officials, ML = Law on Prevention of Money Laundering, NS = National Strategy to Prevent and Fight Corruption, PP = Law on Public Procurement. Parentheses indicate no separate law; provisions included in other legislation.

ber state that had failed to prevent the abuse of EU money allocated to help the national economy. In Romania, no such separate law was adopted. Instead, particular provisions in the Integrity Agency Law, the Penal Code, and regional government legislation regulate the area of conflict of interest. The entries in Table 7.2 also indicate that, where EU accession was not perceived as approaching, institution building started late. Looking closer at the timing of reform there, however, one cannot distinguish between the group that applied for membership in 2003–2004 and became candidates in 2005 (Croatia and Macedonia) and those who applied in 2009 (Albania and Serbia) and are not yet candidates. This somewhat contradicts what James H. Anderson and

172

Political Corruption in Eastern Europe

Cheryl W. Gray (2006, 81) propose (i.e., that government anticorruption activity is more intense when accession is felt probable, but not certain and less prominent prior to achieving candidacy status). What we can see is evidence for active legislating of anticorruption laws in Macedonia and Croatia after 2001, an even more aggressive threeyear effort in Serbia, and a consistent record of passing anticorruption laws in Albania over time (1995–2008). In sum, the data do not exhibit a pattern of time difference across these two groups. Until 2001, Serbia did not have a “coherent anticorruption policy” (Fatic 2001, 8). The country became motivated and undertook important steps to improve its governance only after Slobodan Milosevic was removed from power. After the fall of his regime, EU membership became more feasible and institutional building accelerated. Albania, which did not go through the same turmoil as Serbia, adopted its first law on disclosure of assets by public officials as early as 1995. Despite its lumpy road of transition, the country continued to pass antifraud legislation, including the Internal Audit of the Public Sector Act and a new asset disclosure law in 2003. Albania had developed one of the most comprehensive anticorruption legal frameworks prior to joining the process toward a Stabilization and Association Agreement. Shortly after, in an attempt to align its legislation with the EU acquis, the Albanian government approved additional provisions to the 2006 Public Procurement Act (Bushati 2008). In the above discussion, I identified instances of anticorruption institutional initiation that are difficult to predict through a standard argument about the impact of EU conditionality. These include the belated passing of conflict of interest arrangements in Bulgaria and Romania, and the early adoptions of disclosure and access to information legislation in Albania. The former point to the possibility for postaccession leverage of the European Commission, contrary to warnings in the literature that the conditionality principle is effective before accession (Vachudova 2005, 2009). The newest members of the EU, Bulgaria and Romania, raised the importance of tackling corruption within the EU, and its institutions finally responded with tougher measures. It is too early to form conclusions, but the new tools for pressure, including achievement of specific benchmarks and freezing of EU funds, might have had some influence after membership was granted—at least in the area of legislative reform. In the Western Balkans, where state capture and

EU Accession

173

organized crime pose even greater threats, leverage on adoption of anticorruption laws was exercised from the very beginning of the drive to accession. While belated institutional initiation can be explained through the increased salience of the corruption issue after 2005, the early adoption of assets disclosure and laws against money laundering cannot be accounted for by external pressure. Similar to the minority rights area, the problem of corruption had not been identified prior to the last wave of enlargement as one on which the EU should have concrete requirements for accession (Vachudova 2009). As a result, there was no corruption-fighting acquis. This leaves us with the proposition that, in the absence of direct pressure from the EU as in the general case of early institutional adoption (Grzymala-Busse 2007, 91), the passing of anticorruption legislation must have been shaped by domestic factors. Of course, this does not exclude the possibility that the postcommunist governments adopted measures aimed at developing European standards of good governance. They might have. Initiated voluntarily or under external pressure, however, those institutional arrangements would not automatically reduce corruption if there was no political will and capacity to enforce them. In other words, a leader in legal reform may still be ineffective in fighting corruption in practice (Kajsiu 2004). A thorough analysis of progress made toward clean governance over time can help us better comprehend the impact of reform under EU leverage.

Comparing Levels of Corruption by Integration Status and over Time If the legislative reforms promoted by the EU and other international organizations had an effect in making postcommunist politics more transparent and accountable, we should be able to see the lowering of national corruption levels with the deepening of integration. However, observers in the region warn that reform implementation has been hindered by insufficient political will to effectively investigate, expose, and prosecute impropriety and law violations (Monitoring the EU Accession Process 2002, 70). Has advancement to EU accession helped overcome this deficiency by raising the importance of the problem and conditioning its management on the prospect of membership? Figure 7.1 offers some clues to this ques-

174

Political Corruption in Eastern Europe

tion by tracking down corruption levels in the three groups of Balkan countries over a ten-year period (1998–2008). In the case of Bulgaria and Romania, the two newest members of the EU, a clear pattern of decline in corruption started right after they gained a candidacy status. Over the course of negotiations with the European Commission, their governments made significant steps in reforming the state administration on the principles of professionalization and depoliticization of the civil service. By late 2005, the time when accession agreements were signed, corruption in Bulgaria was reduced by 26 percent (from 4.75 to 3.50 on the NIT scale) and in Romania by 16 percent (from 4.75 to 4.0). The EU provided significant financial aid through its own programs and individual member state donations.16 Once the decision for membership had been made, however, the Bulgarian and Romanian

Figure 7.1 Levels of Corruption in Three Groups of Balkan Countries, 1998–2008

(continues)

EU Accession

175

authorities became less determined in fighting corruption. Romania managed to keep the 2006 level for the next two years, although there were signs that things were regressing. Shortly after membership was officially granted, one of its strongest anticorruption advocates, Justice Minister Monica Macovei, was dismissed from her post (Mungiu-Pippidi 2008). In Bulgaria, this trend was more pronounced. The BSP-MRF-NMSS coalitional government made little effort to ensure full implementation of existing policies, masking “reluctance and incapability for coherent action” behind broadly advertised police operations with no real consequences (On the Eve of EU Accession 2006, 6). After January 2007, the tripartite coalition partners became preoccupied with awarding their own rings of friendly firms with exclusive access to EU subsidies freely flowing from the EU. Although corruption among lower-ranked public administrators

Figure 7.1 continued

(continues)

176

Political Corruption in Eastern Europe

Figure 7.1 continued

Source: Freedom House, Nations in Transit (Washington, DC: Freedom House, 2003–2009), www.freedomhouse.org. Note: Nations in Transit Corruption Index; higher value = more corruption.

declined somewhat, as some business opinion polls reveal, special interests had captured the state in both Romania and Bulgaria (Krastev 2006; Mungiu-Pippidi 2006a).17 The graph representing Croatia’s and Macedonia’s corruption levels shows quite an interesting picture. Their signing of Stabilization and Association Agreements did not initiate a decline in corruption. Croatia had already started reducing malfeasance after the fall of Franjo Tudjman’s regime; in Macedonia, corrupt practices intensified with the escalation of ethnic conflict in 2000–2001. Since 2002, and especially following formal application for EU membership, however, Macedonia has experienced a clear trend (from 5.50 to 4.25, or 23 percent) toward more effective prevention of illicit enrichment in the public sector. The data do not point to such a consistent decline in Croatia.

EU Accession

177

Although the latter made some progress after accession negotiations started in 2005, recent violence against journalists investigating corruption and organized crime casts doubts about Croatia’s anticorruption performance.18 The third group, Albania and Serbia, allows comparison between the phase of no integration and Stabilization and Association Agreement signatory status. In both countries corruption was much more widespread than in the rest of the Balkans in 1998 and reduced to some extent in the 2000s.19 Overall, Serbia experienced a more consistent pattern of corruption decline over the course of seven years following the collapse of Milosevic’s regime, from 6.25 to 4.50 (28 percent). Albania saw a somewhat uncertain reduction, with state capture on the rise in 2002 compared to 1999, slightly contained on the eve of joining the Stabilization and Association Agreement process (Kajsiu 2004). Despite its good record in passing anticorruption laws and regulations, the country has not been able to make significant progress in coordinating its institutions, and acts of malfeasance remained unsanctioned. Receiving aid through the EU CARDS program does not appear to have produced tangible results.

The Impact of the EU and the Role of Economic Liberalization To analyze in greater depth the hypothesis that the EU had an impact on reducing corruption by applying the conditionality mechanism, I next tested this proposition against other possible explanations. Indeed, observers propose that progress made by postcommunist countries in curbing corruption might not be as much a result of what authorities do to contain abuse of power, but a consequence of the general process of economic liberalization. In other words, the transition to a market economy, the strengthening of the private property institution, and the liberalization of foreign trade had a stronger impact on the fight against corruption than government policies (Markovich 2008). It is important to keep in mind that the Eastern European postcommunist states chose a path of simultaneous political reform and transition to a market economy. Advancement in the EU accession process was also conditioned by progress in building strong democratic institutions as well as liberalization of the economy. In this

178

Political Corruption in Eastern Europe

sense, both the political and economic changes required by the Copenhagen criteria for acquiring membership in the EU might have had an impact on the fight against corruption. On the one hand, reduction of corruption can be a result of particular government policies in response to direct EU pressure to curb practices of abuse of power. On the other hand, corruption may be declining with the expansion of the private sector and the promotion of fair competition in the economic arena, also encouraged by the Union and the international financial institutions. Can these seemingly intertwined effects be untangled? Two differences are noteworthy when considering the EU’s role in these policy areas. Although reform was favored by the Union in both the economic and political spheres, the pressure for action has been more direct in the case of fighting corruption. The attainment of a functioning market economy status was a precondition for getting on the track for admission. Yet it was the set of political criteria that at the end was more important in the pursuit of a candidate status. The second distinction is in the mechanisms used by the EU in exercising its leverage. No country has been punished so far for failure to implement market reform while various tools, from monitoring through safeguard clauses to freezing of funding, have been designed and used to ensure compliance. There are good reasons to suppose that the EU pressure on candidate states to undertake active anticorruption measures had an effect separate from that of the process of liberalization of economic relations. For a deeper empirical exploration of these propositions, I constructed a regression model that incorporates the effects of integration status and trade liberalization on corruption in the six Balkan countries. If the EU has had a direct impact on reducing the misuse of power by promoting comprehensive legislation and implementation of national action programs, this effect should be demonstrated by the data after controlling for the possible role of economic reform. If those who believe that any success in the fight against corruption is an artifact of economic liberalization, rather than government policies in response to pressure from Brussels, the data should support only the market liberalization thesis. Since there are only Balkan countries in the data set, the model automatically controls for other historical, cultural, or traditional factors that might have affected corruption levels.20 Finally, the model accounts for differences in

EU Accession

179

material capacity that may affect the formation of incentives for illegal enrichment: Corruption = a + b1(integration) + b2(economic liberalization) + b3(economic capacity) + e

The operationalization of the variables from the model specified above is straightforward. To maximize the robustness of results, I estimated the regression equation with three different measures of the dependent variable. Thus, corruption was measured through the TICPI, the WBI, and the NIT. While the former two assign higher values to less corrupt countries, the NIT gives higher scores to more corrupt societies. Advancement toward integration was represented by a 5-category variable, awarding 4 (achieved EU membership), 3 (candidate status), 2 (submitted application for accession), 1 (signed an Association Agreement), and 0 (no form of association/integration). As described in the previous sections, over the period between 1996 and 2008, each of the six countries in the analysis went through different phases: Bulgaria and Romania from application status to membership (2 through 4 on the scale), Croatia and Macedonia from association to candidacy (1 through 3), and Albania and Serbia from no association to application (0 through 2). Given the hypothesized impact of the EU conditionality mechanism on corruption, the integration status variable was expected to be positively correlated with TICPI and WBI and negatively with the NIT. Economic liberalization is defined as eased conditions for business and trade that reduce opportunities for bribing and was measured in this analysis by the amount of imports as a proportion of a country’s gross domestic product. It varied between 24 percent (Serbia in 1997) and 85 percent (Bulgaria in 2007). Economic capacity was introduced to account for the enabling effect of financial resources, confirmed in my empirical analyses in Chapters 3 and 6. In economic terms, the Balkan states are less well off as a group than the Central Eastern European countries and the Baltic republics. Their per capita GDP figures for 2007 (current dollars) vary from 3,458 for Albania to 13,201 for Croatia (World Development Indicators). Both imports and GDP were lagging by one year and expected to be positively correlated with the CPI and the WBI while negatively with the NIT measure.

180

Political Corruption in Eastern Europe

The evidence offered by the data and reported in Table 7.3 confirms the expectation for a constraining effect of the EU accession process on corruption.21 That deeper integration has caused reduction of corrupt practices is suggested by the signs of the estimates for integration status, which are in agreement with what proponents of EU integration have believed. This effect was independent from the influence of economic liberalization and financial capacity. Moreover, the results for the role of the EU accession process remained statistically significant in all three tests after accounting for the spatial and temporal structure of the observations. A climate of liberalized economic relations also matters, but support from the data for this proposition was not as strong. In all three tests, the estimated parameters of this impact carried the right sign but only for the NIT model is the coefficient significant. The results also indicate that economic capacity, measured through GDP, was clearly related to corruptibility in the Balkans. What these findings reveal is that the impact of the EU conditionality mechanism on reducing corruption in the Balkans has intensified with advancement toward accession. This effect has not been

Table 7.3 The Effects of the EU Accession Process and Economic Liberalization on Corruption in the Balkans Corruption Levels Factors Integration status Economic liberalization Economic capacity Constant N R2

TICPI

WBI

NIT

.116* (.067) .009 (.006) .097*** (.033) 2.168*** (.306)

.108*** (.028) .002 (.003) .037** (.015) –.785*** (.129)

–.186*** (.068) –.012** (.005) –.011*** (.025) 5.726*** (.247)

64

76 .583

66 .306

.873

Notes: TICPI = Transparency International Corruption Perceptions Index; WBI = World Bank Control of Corruption Index; NIT = Nations in Transit Corruption Index. Entries are unstandardized coefficients from linear regression estimations with a common AR(1) parameter, panel corrected standard errors in parentheses. More corrupt countries receive lower scores from the CPI and the WBI, and higher scores from the NIT. * p < 0.10; ** p < 0.05; *** p < 0.01.

EU Accession

181

masked by progress in economic liberalization and the opening of national economies to international trade in particular. Marketization, privatization, and restructuring might have helped, although the amount of anticorruption success would have remained limited. By offering much desired membership and applying new tools to force compliance, the EU could sustain the pressure on Balkan states to introduce anticorruption legislation and to improve its implementation after completion of market reforms. Although initially the Union did not have specific policies in this area and did not include explicit requirements for candidates, by 2005 it emerged as the most effective external factor pushing the postcommunist governments to curb corrupt practices.

Conclusion In the mid-1990s, a number of international organizations, including the European Union, launched ambitious projects of bringing together countries to fight abuse of power in public office. Their effectiveness, however, has often been questioned. The analysis that I reported in this chapter leads to at least three conclusions about the role of the EU in promoting good governance after communism. First, even without specific anticorruption policy requirements of its own at the time, the Union encouraged the postcommunist states to cooperate in the international arena and join treaties initiated by the Council of Europe, OECD, and UN. The transitional Eastern European countries developed a better record of commitment to international agreements on cooperation in fighting money laundering, illegal trafficking, and bribing of foreign officials than did older EU member states. Although compliance with these treaties is nonbinding, becoming a party to them helped identify the problem as significant and increased public awareness of the threat of corruption. Second, the EU has facilitated the adoption of anticorruption legislation but, in the Balkans, this success has not been consistent over time and across space. Most legislative acts developing policies of transparency and accountability were passed during the accession negotiations phase. Yet Albania and Serbia adopted those acts even before they signed their Stabilization and Association Agreements. In terms of passing legislation, countries demonstrated determination to institutionalize European standards and norms even when the

182

Political Corruption in Eastern Europe

prospects for accession were not certain. The effectiveness of these policies suffered from poor implementation and lack of coordination among institutions, but the institutional blocks were put together one after another. Concepts such as conflict of interest and disclosure of public information, so new to the publics in these postcommunist states, were introduced in the EU-sponsored legislation as principal elements of state-society relations. Finally, the EU conditionality mechanism has created incentives for more active anticorruption engagement of governments. While observers do not rush to declare the Union’s policies effective, especially postaccession, positive changes have followed despite the odds (Noutcheva and Bechev 2008; Vachudova 2009). My analysis suggests that when Balkan states become more integrated, they manage their public affairs in cleaner ways. As Mungiu-Pippidi (2006b) notes, the main factor driving the success of democratic reforms in Bulgaria and Romania was external conditionality on the road to joining the EU. Concerns raised by some observers that progress in reducing corruption was an artifact of marketization, rather than of an EU impact, prove invalid. EU integration contributed independently to the challenging task of protecting the public sector from abuse. While the market mechanism began to operate once set, the Union had to constantly come up with new tools designed to enhance transparency and to promote accountability. Where domestic political will was deficient, postponement of accession, external monitoring, and cuts in funding seem to have helped. As the cases of the six Balkan states show, success is difficult but it can be sustained through active EU involvement and persistence. The empirical data that I analyzed thus far suggest an overall positive effect of the process of European integration, yet on average, corruption in the postcommunist new members remains higher than in the rest of the Union. The persistence of fraud, embezzlement, and nontransparent use of state resources, albeit reduced to some extent under integrationist pressure, may have negative effects not only on the economy but also on how the public feels about politics. Patience is an exhaustive resource and governments that fail to limit and punish corrupt practices may have to face an angry public that has suffered extreme hardships in the name of democracy. In the next chapter, I explore the questions of whether corruption has had any impact on popular feelings toward the new democratic institutions and if it threatens the legitimacy of the young political systems in Eastern Europe.

EU Accession

183

Notes 1. These are organizations with a direct influence on the postcommunist nascent democracies. Other international initiatives to combat corruption include the Inter-American Convention Against Corruption, passed by the Organization of American States in 1996, and the revised Rules of Conduct to Combat Extortion and Bribery, passed by the International Chamber of Commerce in 1999. 2. See OECD, Anti-Bribery Convention (21 November 1997), country ratification (as of March 2009), www.oecd.org/dataoecd/59/13/40272933 .pdf, accessed 6 January 2010. 3. See United Nations Office on Drugs and Crime, United Nations Convention Against Corruption (2004), www.unodc.org/unodc/en/treaties /CAC/index.html#textofthe, accessed 6 January 2010. 4. See Council of Europe, Group of States Against Corruption (1 May 1999), www.coe.int/t/dghl/monitoring/greco/general/2.%20Historical%20 Background_en.asp, accessed 29 December 2009. 5. Sedelmeir’s (2008) study does not include anticorruption policies. It focuses on transposition rates and infringements of EU law in member states. 6. The absence of clearly defined criteria became apparent in the accession of Bulgaria and Romania, where corruption had been of a more severe form. 7. In a letter to the Bulgarian government, Michael Lee, European Commission director for enlargement, states that payments from EU funds would be suspended until the country undertook the “correct actions.” These measures are identified for four areas: increasing the powers of the national authorizing officer; reinforcing the functioning of systems to detect violations; increasing the number and qualifications of the personnel responsible for control and monitoring; and improving and increasing reporting to the European Commission (“Michael Lee’s Letter” 2008). 8. In fact, Kanin (2003) argues that structural factors, such as geography and history, have contributed more to the spread of corruption and organized crime in the Balkans than transition itself. 9. With a few exceptions: Poland in the Central European group, which switched to proportional representation two and one-half years after the 1989 semicompetitive election; and Bulgaria and Romania in the Balkans, which reformed their electoral systems on the eve of their first multiparty elections in 1990. 10. One remaining issue that Macedonia must solve in order to begin negotiations is to achieve international recognition of its name, which is currently disputed by another EU member state—Greece. 11. “Financial Assistance,” European Commission, Enlargement, How Does It Work, http://ec.europa.eu/enlargement/how-does-it-work/financial -assistance/index_en.htm, accessed 19 January 2010. 12. Some critics question the way in which political conditionality has been applied to the Western Balkans. They believe that, by mixing normative and security concerns, the EU generated confusion and doubt among local

184

Political Corruption in Eastern Europe

elites. Emerson (2008) proposes a sequence of “functional conditionality” followed by “pure political conditionality,” thus leaving conditions for quality of governance for the final stage of accession. 13. Also, Bulgaria and Romania became members of the Council of Europe earlier (1992–1993) than Albania, Croatia, Macedonia, and Serbia (1995–2003). 14. Enforcement, however, remained low until the late 1990s and experts question whether any congruence existed between reforms and genuine desire to embrace them. As a senior official put it, “In Romania we have brought in a whole series of laws. Laws are not the problem. The point is their observance. People do not consider themselves to be under the law or above the law—they are beyond the law” (Lloyd 1999, 30). 15. This is not a deviation from a common Eastern European trend of late and weak regulation in the areas of identifying and sanctioning conflict of interest situations (James H. Anderson and Gray 2006, xviii). 16. For the period between 2004 and 2008, Bulgaria received through the PHARE program 182 million euros to strengthen and maintain the institutions directly involved in the fight against fraud and corruption. Both Bulgaria and Romania used EU funds to improve financial audit mechanisms and to secure adoption and implementation of anticorruption measures (Emerson 2008). 17. It is not certain whether or for how long the postaccession weakening of the anticorruption effort in Bulgaria and Romania will continue. It is possible that the institutionalized anticorruption norms might prevail after all. Another source of cautious optimism is that the EU preserved the right to punish future noncompliance. The freezing of EU funds for Bulgaria had some effects and the 2009 Transparency International measure shows a slight improvement toward cleaner policies in Sofia. 18. In 2008, prominent journalists were severely beaten and a famous publisher was killed by a car bomb in revenge for exposing cases of embezzlement and illegal seizure of property. The perpetrators were never brought to justice (Srdoc and Samy 2009). 19. The Milosevic regime in Serbia favored a “deeply entrenched synergism between the state apparatus and organized crime” where theft and embezzlement were widespread methods of operating (Fatic 2001, 7). In Albania, former cabinet members from the two major parties became involved in improper involvement in real estate business deals and smuggling. Pyramid schemes, such as VEFA, made significant investments in parties’ election campaigns (Carlson and Betts 2001, 66). 20. Other studies emphasize the role of democratization (Monitoring the EU Accession Process 2002, 43; Gerring and Thacker 2004). Democratization (measured by the Freedom House political freedoms and civil liberties rankings) is another factor that might be relevant, but I excluded it for reasons of multicollinearity as well as previous democratic experience (years democratic since 1900). Both specifications were highly correlated with integration status. Finally, popular tolerance of the exchange of goods or money for public services was not significantly different across the region. A 2001–2002 study

EU Accession

185

by Vitosha Research, conducted under the Southeast European Legal Development Initiative, measured citizen acceptability of the existence of corrupt practices in the six Balkan countries at levels between 1.4 and 2.5 on a scale of 0 to 10 (Corruption Indexes 2002). 21. The tests were conducted with panel data from Albania, Bulgaria, Croatia, Macedonia, Romania, and Serbia in the period between 1996 and 2008. To deal with possible heteroskedasticity and autocorrelation, common problems in pooled data analysis, I employed panel corrected standard errors and first-order AR parameter.

8 Implications for Democracy: Corruption and Public Trust

Corruption is insidious. It makes people lose faith. —Huguette Labelle, “Transparency: One in Four Pays Bribes”

Having analyzed the main reasons for the uneven spread of corruption across the former communist countries, I turn now to the second important question: What are the consequences of corrupt behavior by elites in Eastern Europe? Most of the previous literature on corruption has identified the forms of malfeasance and proposed policy measures for elimination of the problem. Interestingly, the consequences of public officials’ engagement in illegal transactions have received less scholarly attention. More recently, research on the impacts of corruption has accelerated by focusing on some negative macrolevel effects such as delayed economic development, decline in foreign direct investment, deterioration in the business climate, and heightened consumer prices (Mauro 1995; Rose-Ackerman 1999; Monitola and Jackman 2002). Other lines of scholarship have analyzed attempts by influential businesses to “capture the state” (Hellman, Jones, and Kaufmann 2000) and explored a possible link between citizen views about the state and corruption (Mishler and Rose 2001; Christopher J. Anderson and Tverdova 2003). Among many areas of life in which malfeasance may cause damage, mass attitudes and motivations seem the most vulnerable, especially in nascent democracies. In this chapter I analyze, through the prism of corruption, how the Eastern European citizens feel about their political elites and institutions. The question of public confi187

188

Political Corruption in Eastern Europe

dence is especially important because it relates to the issue of legitimacy, which is crucial for the survival of any new democracy. Trust in national institutions helps governments in making policy decisions for the long-term benefit of the public and in committing resources for their implementation (Mishler and Rose 1997). Trust also facilitates citizens’ engagement in collective organizations and, thus, strengthens civil society. For a normally performing democratic system, a certain dose of criticism is healthier than only blind trust. In difficult times of transformation, however, low levels of confidence in the newly established institutions are far more troubling.1 For example, Gabriela Catterberg and Alejandro Moreno (2006) propose that citizens in young democracies distrust institutions not because they look at them with a critical eye, but for reasons of “disillusionment and disaffection.” When transitional governments are unfair and dishonest, people may easily lose faith that the new system is well suited to respond to their needs. If corruption practices of public officials erode political trust, as experts worry, what is the magnitude of this effect and is it uniform across countries? Does the impact of corrupt government threaten the legitimacy of the Eastern European democratic systems by generating mass skepticism, disengagement, and alienation? My purpose in this chapter is to offer answers to these questions so closely linked to the greater problem of how fragile democracy is after communism. The rest of the chapter is organized as follows. I begin with a brief recapitulation of what we already know about the corruptiontrust link and of how this knowledge will be improved by my analyses. Then, I discuss levels of popular confidence in post-1989 representative institutions and how these match the widespread belief that Eastern European societies carry a legacy of distrust in the state. In the next two sections, I build on prior research to develop an explanation of how corruption in particular relates to people’s confidence in government and look into possible patterns of association between institutional distrust and corruption perceptions. Subsequently, I test the argument against alternative sources of trust and democratic support. The findings from the regression analyses are conclusive that public perceptions of corruption are the strongest predictor of institutional confidence, but do not have the same corrosive effect on support for democracy. In the subsequent section, I explore cross-national variations in the strength of the corruption effect. The results begin to draw a picture where the extent to which confidence is influenced

Implications: Corruption and Public Trust

189

by government performance may depend on the broader cultural environment. To conclude, I explore answers to the big question of whether citizen concerns about corruption threaten or strengthen democratic government in Eastern Europe.

Expanding the Current Knowledge of the Corruption-Trust Nexus Although quite young, the research program on the impact of corruption on public opinion in Eastern European has already realized some accomplishments. Previous studies have identified ruling elite dishonesty as a possible determinant of institutional trust at the microlevel. Analyzing New Democracies Barometer data, William Mishler and Richard Rose (1997, 2001, 2005) conclude that government performance should be at the heart of any explanation of citizen approval of democracy. Perceptions of a fair government, they argue, strengthen popular support for the nascent political system and for its structures. Moreover, Mishler and Rose (2001) establish that the political predictors of institutional trust in postcommunist countries are clearly superior to the cultural and historical determinants found relevant by research on other societies.2 This finding is consistent with Geoffrey Evans and Stephen Whitefield’s (1995) earlier argument that support for democracy after communism can be better understood by applying political rather than economic frameworks. Another group of scholars employed macrolevel measures to establish that corruption has a damaging effect in both developed and transitional democracies. The results of their analyses, based on data from the TICPI, reveal that corruption reduces people’s trust in civil servants and support for the system as a whole (Christopher J. Anderson and Tverdova 2003), sustains inequality and undermines institutional confidence (Rothstein and Uslaner 2005), and merely makes people feel unhappy (Tavits 2008). As the works of Richard Kohl (1995), Clare McManus-Czubinska et al. (2004), and John R. Hibbing and Samuel C. Patterson (1994) recognize, these effects are especially pronounced in the young postcommunist democracies where frustration with greedy politicians in times of economic hardship may erode the legitimacy of government itself. Drawing on progress made by extant scholarship, in this chapter I attempt to advance the study of the effects of corruption in several

190

Political Corruption in Eastern Europe

ways. First, I use a microlevel indicator of corruption that reflects individuals’ assessments of how corrupt the country where they live is. This measure is more direct because it gauges the perceived magnitude of the phenomenon at present while most previous studies relied on a comparative to the past assessment. It is possible that by asking respondents to rate present corruption to that during communism, the New Democracies Barometer survey might have introduced noise in the responses by forcing all who reject communism to dismiss corruption. Second, existing studies have either used an index combining scores on trust in several institutions (Mishler and Rose 2001) or analyzed confidence in one particular institution (Christopher J. Anderson and Tverdova 2003). My analysis includes effects of corruption on public trust in six institutional structures, three political and three administrative. This variety allows us to compare the magnitude of the impact across different institutions and to determine which of them are more vulnerable to the corrosive power of fraud and bribery. Third, much of the work done on institutional trust emphasizes the superior performance of political factors and remains inconclusive about the role of sociocultural explanations. To address this gap, I searched for evidence of a conditional effect of culture on the corruption-trust link at the system level. If the strength of this link is associated with culture at the aggregate level, then we can argue that the cultural orientations of a nation as a whole establish limits within which the effect of corruption operates. Finally, my study adds more countries to the analysis. I include three postcommunist nations currently aspiring to EU membership (Albania, Croatia, and Macedonia), which to the best of my knowledge are absent in all previous studies.

Confidence in Government Institutions During the Postcommunist Transition How trustful of their government have the Eastern Europeans been after 1989? After a long period of communist rule, new constitutions set the beginning of a political system aimed to fulfill the principles of representation, accountability, and transparency. In the aftermath of the “revolutions,” the Eastern Europeans quickly embraced the rights to speak their minds, freely criticize the government, and open-

Implications: Corruption and Public Trust

191

ly demand fair and equal treatment before the law. While people wholeheartedly enjoyed these new achievements, they soon came to suspect that things were not improving fast enough. Tremendous financial hardship was required of them while a few among the elite managed to accumulate too much wealth from their positions in public office. As news about new cases of embezzlement mounted, the number of disappointed citizens steadily grew. If approval of government institutions is deficient, what can this be attributed to? Democratization and marketization, the main goals of the transition, rely on the development of civic norms of rationality, trust, and activism. This might be the case in Eastern Europe, but Ronald Inglehart and Wayne E. Baker (2000) warn about traces left by history that are difficult to overcome swiftly. In other words, a cultural change from cynicism toward the ideological institutions of communism to confidence in the offices of democratic government is possible, but also path dependent. Such a shift is constrained by preexisting dispositions inherited from long years of foreign domination and totalitarianism. It is reasonable to expect that, with the progress of political and economic transformation, public trust will start to build but be fragile to sustain; it will certainly remain significantly lower than in consolidated democracies. Indeed, extant research has reported low overall levels of institutional trust across the entire postcommunist world (Rose-Ackerman 2001). A popular lack of confidence in the democratic structures of government was documented by studies of Eastern European public opinion in the early 1990s, which warned that trust was declining (Hibbing and Patterson 1994) and that political alienation was on the rise (Kohl 1995). Survey data from that period show that only 60 percent approved of the new system of free elections and less than half of the public felt confident in parties, parliaments, and civil servants (Kohl 1995). Later, Mishler and Rose (1997, 2001) found that the median attitude toward institutions in 1994 was one of skepticism, which by the late 1990s grew into a feeling of distrust among a large majority of the citizenry. Moreover, while public disaffection has become a problem in developed democracies as well (Pharr and Putnam 2000), confidence in national parliaments is estimated as 20 percent lower in the East than in the West (Mishler and Rose 1997). In the second decade after the fall of communism, many Eastern Europeans continue to distrust government institutions. In 2008, for example, Eastern European public confidence in the national parlia-

192

Political Corruption in Eastern Europe

ments was within a 13 percent (Bulgaria) to 38 percent (Estonia) range (Eurobarometer 69, 2008). What drives the skepticism and alienation of so many individual citizens is a question that I turn to next. At the core of my argument is a rational-based understanding that citizens lose, or fail to develop, faith in the structures of power when public officials betray popular expectations for fair and just government.

How Does Corruption Hinder Trust? Institutional trust, or the lack of it, has been explained in the literature through the efficiency of representation, the model of (re)distribution, and the quality of government service (Alesina and Wacziarg 2000). Two dimensions of policy quality are especially important, the capability of those trusted with power to govern and the honesty with which they perform. As Russell Hardin (2000) observes, competence and motivation are key qualities that affect citizen approval (or disapproval) of those in public office and ultimately build confidence (or mistrust) in public institutions. Incompetent civil servants are not efficient and effective in their work, something that can spur disappointment among citizens who receive few and badly provided public goods. While insufficient competence of bureaucrats in a transitional democracy may be partly forgiven due to incomplete administrative reforms, the absence of honesty frustrates the common people and generates a sense of betrayal. When those trusted to make and implement policies are thought of as lacking integrity to handle public affairs in a fair way, confidence in the institutional makeup of the system suffers. Citizens tend to associate politicians’ motivation for public service with the results of government performance. If they see candidates running for public office only as people driven by the desire to enrich themselves, suspicions arise that state agencies provide ambitious partisans with jobs, but fail in their main mission—to deliver the public good to society as a whole. Data from the third wave of the World Values Survey (World Values Survey 2000) provide some support for such an expectation. These data reveal that in the region of Eastern Europe, a majority of 86 percent among those who believed their country was run by a few big interests instead of “for all the people’s benefit” were fairly or very dissatisfied with “the people in

Implications: Corruption and Public Trust

193

public office.” Frustration with politicians who, once elected, serve only the top wealthy few to extract personal profits, evolves into disapproval of the office itself. Resulting poverty, unemployment, and absent opportunities are all blamed on greedy administrators tempted by bribes and indifferent to the needs of people. It should be underscored that in a transitional context the two aspects of trust, competence and commitment to honest service, may not have equal weight as determinants of confidence. First, because capable legislators and bureaucrats were hard to find among the cadres of the young parties, especially in the early years of transition, it was the former communists who could recruit from their ranks experienced and relatively knowledgeable administrators. The public realized this and did not mind the presence of writers, artists, and actors from the ranks of the opposition in the first post-1989 assemblies. In brief, capable lawmakers and other officeholders were welcomed, but the reform forces did not have many of those. Second, since the mid-1990s, corruption became such a sensitive issue that, in the eyes of the public, competent but corrupt was much worse than incompetent but clean. Consequently, electing unknown (and presumably less qualified) people to office became a preferred choice to casting votes for notoriously corrupt politicians. In other words, many citizens were ready to trade competence for integrity or even for the lack of information about someone being corrupt. The value of honesty in public office increased exponentially after the problem of corruption began to gain salience in the mid1990s. Alexandru Grigorescu (2006) observes an “eruption” of corruption in Central Eastern Europe since then that is much higher than in other places of the world. As he succinctly describes, the increased salience of corrupt politics in the postcommunist world has been a result of an intensified anticorruption rhetoric in the mass media, fueled by scandals and encouraged by intergovernmental organizations. The effort to promote awareness and keep the issue on the agenda might have widened the gap between real and perceived corruption, but both are potential consequences. If actual corruption has negative consequences on the economy, as Grigorescu (2006) argues, popular beliefs that most officeholders are corrupt may have serious political effects on civic culture. Trust in government is not a capital that postcommunist societies could inherit from the past, but something that had to be earned by their post-1989 administrations (Mishler and Rose 1997,

194

Political Corruption in Eastern Europe

2001). Confidence in the newly established institutions could be developed mainly through hard work of the bureaucracies to allocate goods and distribute services in a transparent and fair way. A failure to deliver better-quality governance was to be seen by many as the old well-known story of ruling in favor of the privileged few, behind the backs of the people and despite promises for fair competition. Disappointment with dishonest government during democratic transition can be much stronger and more harmful than the bitter feeling during communism that the nomenklatura enriches itself at the expense of the nation. The promise of the elite to build a political system based on the notion of governance by the people implied a commitment to making and implementing policies with integrity and openness. The performance of those trusted to govern became a measure of how well the new elite stuck to those objectives and transformed the institutions into representative and accountable bodies. In sum, we can expect that citizens disappointed with the performance of government will also be less trustful of political institutions. As a key aspect of government quality, corruption is an indicator to which the public is especially sensitive when it comes to how well the new institutions perform. Therefore, perceptions that corrupt practices of public officials are widespread hinder the formation of confidence in political institutions. This correlation should be stable across countries and independent from other sources of distrust. We can anticipate, though, that the strength of the proposed effect of corruption on confidence is conditioned by the underlying national culture and traditions.

Levels of Corruption Perceptions and Confidence in Institutions An empirical assessment of the proposed association between perceived corruption and distrust in government requires observations on (1) people’s assessments of officeholders’ honesty and integrity, what Hardin (2000) calls “motivation”; and (2) how much faith citizens have in the institutions represented by state officials. Data on these two dimensions are available from the third wave of the World Values Survey (2000) conducted in the second half of the 1990s. First, I seek

Implications: Corruption and Public Trust

195

to establish the distribution of assessments and confidence levels across groups and the degree of association between the two sets of opinions. By doing this, I set the stage for a more sophisticated statistical analysis later when I tested the suggested link of trust to corruption against other possible sources of popular disaffection. The microlevel is where one can expect to obtain the most valid empirical results for the impact of corruption on attitudes about the government. Thus, individual citizens from Eastern Europe with their unique set of attitudes and beliefs are the units of analysis. Relying on the World Values Survey data, I measured levels of corruption perceptions through individual responses to a question about how widespread they think corruption is in their country.3 Political confidence is gauged by answers to questions about how much confidence someone holds in various institutions.4 I looked at the opinions provided by respondents from the fifteen postcommunist states: Albania, Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Macedonia, Poland, Romania, Russia, Slovakia, Slovenia, and Bulgaria. Looking at Table 8.1, we can discern a uniform pattern of popular confidence in various political and administrative institutions. For each of the six institutional types, citizens who were not very trustful formed the modal category (between 39 percent and 46 percent). The second large group (ranging from 26 percent to 34 percent) for all institutions except “political parties” was those who have “quite a lot of confidence” in the structures of government. Indicative of the rate of popular trust was the large difference between the two clusters with extremely opposing attitudes, listed in the second and fifth columns of the table. While those who were completely distrustful were between 14 percent and 29 percent of all respondents, the ones who trusted the new government institutions “a great deal” were between only 2 percent and 8 percent. Overall, parliaments and political parties emerged as the least trusted institutions, as shown by the proportions of respondents who either did not trust them very much or not at all (63.1 percent and 74.7 percent, respectively). This is not surprising, given that parties compete for and win seats in the national assemblies and their leaderships have a strong impact on the agenda and the work of the parliaments. Among the administrative structures, the police force was the most distrusted, although at a lower rate than the political institutions.

196

Political Corruption in Eastern Europe

Table 8.1 Levels of Institutional Confidence and Corruption Perceptions (percentage) Confidence in:

Not at all

Not very much

Quite a lot

A great deal

Parliament Government Parties Police Civil service Justice system

21.7 19.1 29.1 17.1 13.7 14.9

41.4 39.0 45.6 38.4 40.3 40.2

26.3 31.3 15.0 33.8 32.3 32.7

3.8 5.4 1.6 7.6 3.9 7.5

N 17,344 17,664 17,039 18,020 16,795 17,773

Public officials engaged in corruption Almost all are

Most are

A few are

Almost no one is

N

32.2

45.7

19.6

2.5

16,646

Sources: European and World Values Surveys for Wave Integrated Data File, 1981–2004, v. 20060423, 2006; the European Values Study Foundation (www.europeanvalues.nl) and World Values Survey Association (www.worldvalues survey.org). The sample includes surveys conducted in Albania, Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Macedonia, Poland, Romania, Russia, Slovakia, Slovenia, and Ukraine. Note: Percentages are calculated on the basis of data from the third wave (1996–1998) of the World Values Survey.

Turning next to the measures of corruption perceptions, I assessed to what extent the Eastern European public believed that officeholders in their countries were engaged in bribery and corruption. Close to one-half (about 46 percent) of those who expressed an opinion in the World Values Survey believed that most officials were in some way corrupt. Another 32 percent even thought that almost all people in public office abuse their position and access to state resources. Only 22 percent were not concerned that corruption was widespread in their country. These numbers reveal variation in the popular assessment of politicians’ integrity, yet the distribution of opinions is skewed toward the groups that suspected less wrongdoing.5 Next, I tried to establish whether the distributions of attitudes toward institutions and toward officeholders followed a pattern. Recall that my primary goal was to examine how corruption perceptions and institutional trust correlate. Figure 8.1 shows the distribution of popular confidence in the national parliament and in the police across those who perceived corruption as widespread and those who

Implications: Corruption and Public Trust

197

did not. The bars represent percentages of citizens feeling confident in the respective institution, clustered by country. The graph reveals that, in all Eastern European countries included in the analysis, citizens who thought that none or just a few public officials were engaged in bribery also were more trustful of the government. Vice versa, those who thought of corruption as widespread or very much existing demonstrated lower levels of confidence in the assembly and in the police.6 The gap between the two groups varied somewhat across countries. For example, it was narrower in Romania and Albania for trust in both the parliament and the police. Quite substantial were the differences across corruption perception groups in Croatia and Lithuania for confidence in the legislature and in Ukraine and Slovakia for confidence in the police. Regardless of the size of

Figure 8.1 Corruption Perceptions and Confidence in Parliament and in the Police

(continues)

198

Political Corruption in Eastern Europe

Figure 8.1 continued

Sources: World Values Survey, Official Data File v. 20090914 (World Values Survey Association, 2000), www.worldvaluessurvey.org, Aggregate File Producer ASEP/JDS, Madrid.

the gap, the pattern is clear: beliefs that corruption is extensive appear to be associated with weaker confidence in national institutions. Since other factors may also exert influence over how people feel about the structures of government, I proceeded to build and test a regression model that incorporates a set of different determinants of institutional trust.

Isolating the Effects of Corruption Perceptions An unbiased study of the impact of corruption on public attitudes toward institutions has to take into account other possible determi-

Implications: Corruption and Public Trust

199

nants of trust. Extant research stresses the importance of various individual-level factors that, along with corruption perceptions, shape citizen approval of national institutions (Newton and Norris 2000; Mishler and Rose 2001). I now present several potential microlevel determinants of institutional confidence, drawn from arguments about government performance, political dispositions, socialization, and demography. One dimension of government performance, different from corruption and relevant to the process of formation of institutional trust, is the economy. In the difficult times of transformation from central planning toward market economy, evaluations of economic performance may have a significant impact on how people feel about institutions (Mishler and Rose 1997). The overall levels of literacy and sophistication in postcommunist societies, a necessary condition for understanding the challenges of transition, are relatively high; yet by the mid-1990s, the first signs of impatience and fatigue became apparent in the populace. One would expect that those who evaluate their financial situation as bad would blame the government for not being able to provide favorable conditions for economic growth. Therefore, in the following analysis I considered the possibility that individuals’ satisfaction with the financial situation of their household strengthens trust. Furthermore, existing research on citizen confidence in institutions confirms the effective role of political factors (Newton and Norris 2000). Studies focusing on public opinion in the United States have stressed that trust reflects individuals’ political experiences and values more and their personal characteristics less (Levi and Stoker 2000). Thus, people interested in political matters are more likely to be among those confident in the constitutional arrangements of the system. Also, in forming attitudes toward institutions, citizens are affected by their ideological beliefs. In particular, voters with a leftist orientation are more critical of and less trustful of government structures (Newton and Norris 2000). Accordingly, I took into account the possible effects of political interest and ideology. Another alternative explanation focuses on the specific social conditions and experiences in people’s lives as determinants of confidence. According to this perspective, the mode of socialization and individuals’ position in society shape patterns of affection for the institutional structures of the political system (Inglehart 1990). To isolate these possible social structure effects, I accounted for a possible impact of education and social class. For both factors, the antici-

200

Political Corruption in Eastern Europe

pated relationship was positive (i.e., higher education and higher social status integrated people in a way that made them feel more confident in the government). Finally, attributes such as gender and age may cause distinct patterns of institutional trust. Scholars have found that on average, women and older people are more likely to trust government structures. These relationships, despite being weak compared to institutional and political sources of confidence, have proved consistent in existing empirical research (Mishler and Rose 1997; Newton and Norris 2000). Therefore, I incorporated their possible impact in the model.

Analysis In this section, I analyze more rigorously the hypothesized effects of corruption perceptions on institutional confidence. Recall that the graphs presented in Figure 8.1 show some preliminary evidence for disparity in the levels of trust between those who believed corruption is widespread and the rest who did not think so. As the data reveal, by the late 1990s such a gap had occurred in all fifteen Eastern European countries. Given all other possible explanations, can we still argue that perceived corruption contributes to low trust? Below, I constructed a principal model that includes the key independent variable, corruption perceptions, and the group of control variables described in the preceding section. The functional form of this model is as follows: Confidence = a + b1(corruption) + b2(financial) + b3(interest) + b4(ideology) + b5(education) + b6(social class) + b7(gender) + b8(age) + e

Confidence is individual confidence in political institutions, measured in categories from least to most confident. Corruption is individuals’ perceptions of the spread of corruption in their country in categories ranging from “none at all” to “a great deal.” Financial is personal evaluations of the financial situation of the family, interest stands for interest in politics, ideology reflects individuals’ ideological views, education stands for level of educational attainment, social class measures belonging to social groups in an ascending order, gen-

Implications: Corruption and Public Trust

201

der specifies gender type, age is individuals’ years of age, a is the intercept, and e is the error term.7 For the tests of the above model, I again used survey data from the third wave of the World Values Survey (1995–2000). 8 I ran a series of ordinary least squares regressions of the principal model with different specifications of the dependent variable, confidence in the parliament, the national government, the political parties, the police, the civil service, and the judicial system.9 The results, reported in Table 8.2, provide strong support from the data for the hypothesized effect of corruption: individuals who believed that the majority of public officials were corrupt were less likely to trust the political institutions in their country. This result confirms Christopher J. Anderson and Yuliya V. Tverdova’s (2003) finding about the damaging effects of corruption on trust in civil servants in both developed and new democracies.10 Moreover, my analysis offers empirical support for a claim that public confidence in other institutions also suffers from beliefs that many officeholders are corrupt. This finding is in agreement with another piece of extant research that, using different data, found a correlation between trust

Table 8.2 The Effect of Corruption on Public Confidence in Various Institutions, 1995–1998 Parliament

Government

Parties

Police

Civil Service

Corruption Financial situation Ideology Interest in politics Education Social class Gender Age

–.18***

–.17**

–.13***

–.16***

–.17***

–.17***

R2 N

Variable

Justice System

.08*** .05***

.09*** .04***

.08*** .04***

.06*** .01

.07*** .00

.08*** .00

.09*** –.03** –.01 –.02 .02

.05*** –.04*** .01 –.02** .03***

.18*** –.06*** .00 –.03*** .00

.02 –.05*** –.02 –.03*** .04***

.04*** –.03*** .00 –.04*** .01

.01 –.06*** .00 –.02** .02

.110 10,460

.105 10,558

.082 10,362

.116 10,649

.082 10,172

.089 10,626

Notes: Dependent variables: degree of confidence in parliament, government, parties, police, civil service, and the justice system. Entries are standardized beta coefficients, using ordinary least squares regression models. ** p < .05; *** p < .01. Full models are listed in Appendix C.

202

Political Corruption in Eastern Europe

in political institutions and convictions that corruption has increased since communism (Mishler and Rose 2001).11 In Table 8.2, the erosion of trust in various government structures, political and administrative, is indicated by the negative signs of the coefficients for Corruption that are also statistically significant for all variants of the principal model. Furthermore, the impact of corruption perceptions on institutional trust is independent from the effects of other sources of disaffection, including economic performance. The findings suggest that the way the Eastern Europeans evaluate their financial situation is also related to their approval of various public offices. As Kenneth Newton and Pippa Norris (2000, 61) argue, “all citizens are exposed to government actions” and those from households affected negatively by failed economic policies easily lose trust in national institutions. In addition, my tests confirm previous findings about the impact of political factors (ideology and interest in politics) on confidence in the three political institutions; those interested in politics also trusted the civil service in their country. Among the social variables in the models, education and gender had a statistically significant, albeit marginal, effect on trust. According to the social profile of trustful citizens, suggested by the estimated parameters, older people and women in Eastern Europe tended to be more trusting in the structures of government. The results for social class and age were inconclusive. When comparing the beta coefficients for the factors in all tests of the principal model, perceived corruption can be identified as the most powerful determinant of institutional confidence. Its magnitude varied between –0.13 (police) and –0.18 (parliament). Interestingly, the estimates for corruption perceptions outplayed by magnitude even those of the economic indicator of government performance, household financial situation. In other words, in keeping confidence in the new institutions, people were more troubled by thoughts of corrupt officials than by concerns about a worsening economy. Only one of the tests suggests a stronger impact of a factor different from corruption perceptions. Interest in politics was the strongest predictor only in the model of trust in political parties, as indicated by a beta coefficient of 0.18 (compared to –0.13 for perceived corruption). All in all, the World Values Survey data offer strong support for the argued influence of unfair and nontransparent policymaking and implementation on citizen trust in institutions. This impact is consis-

Implications: Corruption and Public Trust

203

tent across political and administrative structures and is also more powerful than the effect of other sources of distrust. However, is this impact consistent across various postcommunist nations? In the next section, I attempt to answer this question by taking the analysis to the national level.

Is the Corruption Effect Equally Damaging Everywhere? In order to explore possible differences in the way corruption affects popular confidence in institutions, I first tested the regression model from the previous section on World Values Survey data for each of the fifteen countries separately. For parsimony, I construct a composite index of confidence to measure individuals’ trust in all institutions. Thus, the dependent variable is a simple additive index that sums up for every individual confidence in the parliament, the government, the parties, the police, the civil service, and the judicial system. The regression tests generated fifteen beta coefficients for the impact of elite integrity, or the lack of it, on institutional trust. These estimates serve as measures of the strength by which corruption perceptions erode public confidence in various country contexts. The obtained coefficients attest to a considerable cross-national variation in the effects of corruption. For example, citizen trust in institutions in Estonia suffered the most (beta = 0.384) from perceived corruption and in Romania such an impact was almost nonexistent (beta = 0.001). The average magnitude of the effect of elite integrity assessments on confidence was reflected by a beta of 0.241. What can explain these differences? The literature is inconclusive on the importance of system-level factors for individuals’ trust in institutions. Mishler and Rose (1997) argue that microlevel determinants are the ones that matter. However, Newton and Norris (2000) find that confidence and social trust are associated at the macrolevel and not at the microlevel. Given the system-level variation in the effects of corruption that I estimated, it is worthwhile to explore the reasons by searching for cross-national differences. As mentioned before, the Eastern European societies share a common experience with communism that itself “produced a culture of corruption” (Sandholtz and Taagepera 2005, 127). Another shared similarity is the simultaneous political and economic transition, including uncertainty

204

Political Corruption in Eastern Europe

but also opportunities. What remains different, then, are the precommunist past and those of its features that were not totally eliminated by the Marxist ideology and the centrally planned economy. Studying patterns of modernization across the world, Inglehart and Baker (2000) find that certain “distinctive cultural traditions” may persist despite the advance of modernization. Following their logic, the Eastern European societies should have preserved some sociocultural traditions that withstood the impact of communist mass education, urbanization, and industrialization. Indeed, Wayne Sandholtz and Rein Taagepera (2005) suggest that traditional authority–oriented countries are also more corrupt than nations with a secular culture. No doubt, communism contributed significantly to the secularization of the entire Eastern European region, but countries there might differ in the extent to which spiritual concerns continue to occupy the minds of people. Empirical data reveal that after 1989, large groups within the formerly communist societies started to attach more importance to God’s role in their lives (Inglehart and Baker 2000). But how would a nation’s historical-cultural heritage condition the effects of corruption perceptions on institutional confidence? Sandholtz and Taagepera’s (2005) research points to a direct relationship between traditionalism and malfeasance (i.e., social conservatism favors corruption). Thus, favors reserved for individuals from the inner circle of family or other closed networks in kinship-based societies are considered improper and illegal in settings characteristic of rational-impersonal values. Therefore, in societies with more particularistic values, people are more tolerant of instances of corruption. In other words, where personal favors are the way things are run, the (mis)use of public office to deliver them is often belittled. Accordingly, knowing that corruption exists may have a weaker impact on people’s feelings about government in more traditional than in more secular societies. To explore this possibility further, I plotted the country-specific measures of corruption perception effects against Inglehart and Welzel’s (2005) measure of traditional versus secular cultures. 12 Looking at Figure 8.2, we can see a clear positive pattern of relationship. The magnitude of corruption effects on institutional confidence grew with the increase in secularism in the cultural setting of a country. One outlier stands out, Romania, for which the model predicted almost no effect of elite integrity on institutional trust. This case

Implications: Corruption and Public Trust

205

Figure 8.2 Corruption Effects in Traditional and Secular Cultures

Sources: Author’s calculations of strength of corruption effects and Inglehart and Welzel’s Cultural Index, www.worldvaluessurvey.org/wvs/articles/folder_published /article_base_54.

clearly needs further investigation with regard to the near-absence of a corruption impact. Turning now to the rest of the countries, Poland presents an interesting case of a postcommunist state with a culture most loaded by religiosity and a slightly lower than average effect of corruption. Slovenia and the Czech Republic, much more secular societies, experienced patterns of impact similar in magnitude to that of Poland. Estonia and Latvia, with the least traditional value systems, are also places where corruption perceptions had stronger effects. Overall, it appears that a nation’s historical-cultural heritage conditioned the extent to which skepticism, resulting from dishonest policymaking, grew in Eastern Europe. The findings produced by my analysis so far are troublesome because they suggest that the Eastern Europeans who think of officeholders as corrupt are inclined to lose confidence in the institutions where these officials work. This pattern is more pronounced in states with an underlying culture that is more traditional than secular, a

206

Political Corruption in Eastern Europe

legacy related not to the common experience with communism but to more distant historical and social circumstances. In any event, at present blame is effectively being transferred from elected and appointed officials to entire government structures. If those who suspect corruption is widespread stop to trust institutions, they may not believe that other politicians could come to replace the crooks and clean up government. As other research has shown, political efficacy is also affected by the sense that dirty politics prevails (Appel 2001). Moreover, individuals in countries with higher overall corruption levels are found to evaluate the performance of their democratic system in negative terms (Anderson and Tverdova 2003). Assuming that people distinguish between democracy in practice and democracy as a model of government, one would want to know how volatile the idea of democracy is in Eastern Europe given the salience of the corruption issue.

Corruption Perceptions, Confidence, and Support for Democracy To answer the important question of whether corruption also erodes support for democracy, I performed additional tests. In this part of the analysis, I attempted to compare the levels of impact of corruption on trust to those on citizen support for democracy. To measure support for democracy, I used answers to the World Values Survey question about how people feel about democracy in general.13 Confidence in (all) institutions and support for democracy were then regressed on the independent variables in the principal model. What is most striking about the results shown in Table 8.3 is the statistically significant, but very weak, relationship between corruption perceptions and support for democracy. The standardized coefficient was negative and quite small (–0.05), suggesting that agreement that democracy is better than other forms of government was only marginally higher among those who were not concerned about corruption. At the same time, confidence in institutions appears to have been markedly affected by perceived corruption. The magnitude of this impact was much stronger (beta = –0.25) than that of support for democracy. As there was not a large difference between the goodnessof-fit statistics for the two models, there must be other factors that outperform corruption. Looking at the rest of the beta estimates, we see

Implications: Corruption and Public Trust

207

Table 8.3 Comparing the Effect of Corruption on Public Confidence in Political Institutions to Support for Democracy, 1995–1998 Variable Corruption Financial situation Ideology Interest in politics Education Social class Gender Age R2 N

Confidence in Institutions

Support for Democracy

–.23*** .11*** .04*** .10*** –.06*** .00 –.04*** .03***

–.05*** .04*** .11*** .05*** .09*** .01 .00 .02**

.133 9,532

.130 9,838

Notes: Dependent variables: degree of confidence in political institutions (index combining confidence measures for all six institutions in Table 8.1), and degree of agreement with the statement “Democracy may have problems but it's better than any other form of government.” Entries are beta standardized coefficients, using ordinary least squares regression models. ** p < .05; *** p < .01. Full models are listed in Appendix C.

that ideology (beta = 0.11) was a more powerful determinant of support for democracy than perceptions of widespread corruption. The closer Eastern Europeans identified themselves with the political right, the more supportive they were of the idea of democracy. However, ideology did not have as strong an impact on institutional confidence where thoughts that politicians are corrupt dominated the process of trust formation. The opposite pattern was revealed for the effects of financial situation and interest in politics: they were more important for developing institutional confidence than faith in democracy.

Conclusion In this chapter, I attempted to develop a better understanding of how citizens’ perceptions of corruption may shape feelings about national institutions in Eastern Europe. I examined World Values Survey data from fifteen countries to establish both levels of views about the spread of corruption and levels of confidence in political and administrative structures of government. Through regression analysis, I demonstrated that not only was the impact of elite integrity on public

208

Political Corruption in Eastern Europe

trust robust, but also that it outperformed all other factors, including personal economic well-being. It appears, however, that this effect varied across national historical-cultural settings. Interestingly, despite sharing a common communist past, the Eastern European societies differed in how strongly corruption affected citizens’ political attitudes. The effect was much weaker in countries with more traditionalist culture where people tended to tolerate relationships based on personal networks and favoritism. What do the findings suggest for the stability of the new order in the postcommunist young democracies? There is both bad news and good news. The bad news is that, when people perceived of corruption as a widespread phenomenon, they also tended to distrust public institutions. The fact that so many citizens believed that officeholders misuse delegated power for personal enrichment is worrisome for the prospects for survival of the entire political system. The good news is that, after all, democracy as a model of government was not so fragile in Eastern Europe. At least, as the results here suggest, the effect of corruption on support for democracy was not as destructive as it was on institutional trust. It is clear from Table 8.3 that perceived dishonesty in public policymaking was only weakly associated with approval of democracy as a better system despite its problems. In other words, support for representative government was fairly evenly distributed across groups with different assessments of the integrity of public officials. In sum, the threat of corruption was more relevant to approval of institutions and their current performance than to democracy as an ideal form of government in Eastern Europe. Yet, the erosive effects of corruption on institutional trust pose serious concerns about the ways in which Eastern Europeans relate to present day politics. When confidence in the institutions of representative government is lost, frustration with corrupt politicians grows, as do doubts that the democratic process can effect change. As we will see in the next chapter, perceived corruption has the potential to shape distinct patterns of voter behavior by generating completely different incentives. Notes The epigraph to the chapter is from Huguette Labelle, chair of Transparency International, quoted in Martin Kuebler, “Transparency: One in Four Pays Bribes,” Deutsche Welle, 9 December 2010, www.dw-world.de.

Implications: Corruption and Public Trust

209

1. In this chapter, I use the terms trust and confidence in institutions interchangeably. 2. But see Luhiste (2006) who contends that cultural and performance theories (including perceptions that officials are corrupt) offer equally effective explanations of institutional trust in Estonia, Latvia, and Lithuania. 3. The wording of World Values Survey question E196 is: “How widespread do you think bribe taking and corruption is in this country?” Respondents chose from these responses: “Almost no public officials engage in it,” “A few are,” “Most are,” and “Almost all public officials are engaged in it.” 4. Respondents in the World Values Survey were asked: “I am going to name a number of organisations. For each one, could you tell me how much confidence you have in them: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none at all?” I analyzed here confidence in the parliament, the national government, the political parties, the police, the civil servants, and the justice system (questions E074, E075, E076, E079, E080, and E085, respectively). 5. The Pew Research Center 2007 Global Attitudes Survey (http://pewglobal.org/reports/pdf/257.pdf) identified analogous trends. Its findings indicate that about 78 percent of the public in the Czech Republic, 66 percent in Slovakia, and 61 percent in Poland recognized corrupt leaders as the top national problem in 2007. 6. Similar patterns of distribution are produced by graphs (not shown here) constructed to gauge confidence in other institutions. 7. For description of variables in the regression model and their sources, see Appendix A. 8. The survey instrument used for the fourth wave (2000–2005) does not include such a question and we cannot run the analysis with more recent World Values Survey data. I considered another potential source, the Comparative Study of Electoral Systems (CSES), Module 2 (2001–2005), which included a question about how widespread corruption is in eight Eastern European postcommunist countries. However, the CSES did not include a question about how confident respondents felt about national political institutions. 9. To deal with potential heteroskedasticity driven by nonconstant variance due to unit effects, I added dummy variables for all countries except Ukraine. I also checked for excessive levels of multicollinearity by examining bivariate correlations between all variables on the right side of the regression equation. The highest among them was between social class and education (0.373, significant at the 0.000 level), which is of a magnitude that did not pose a serious threat to the robustness of the tests. To further explore the multicollinearity threat, I dropped social class, which did not change either the signs or the statistical significance of the regression parameters. I also checked the second largest correlation coefficient, between social class and financial (0.350, significant at the 0.000 level); dropping either of the two variables did not produce any disturbance in the results. 10. In their study, Christopher J. Anderson and Yuliya V. Tverdova (2003) used 1996 survey data gathered by the Institutional Social Survey

210

Political Corruption in Eastern Europe

Program. They included five postcommunist countries in their analysis: the Czech Republic, Hungary, Latvia, Russia, and Slovenia. 11. Mishler and Rose (2001) used New Democracies Barometer V (1998) survey data to identify sources of public trust. Their analysis employed a composite measure of trust in six institutions (parties, parliament, president/prime minister, courts, police, and the military). 12. See Chapter 6 for more information on Inglehart and Welzel’s (2005) traditional/secular index measure. 13. See Appendix A for more details.

9 Implications for Democracy: Corruption Perceptions and Voting

The strains of transition—individual insecurity and uncertainty, growing unemployment and social inequality, . . . the egoistic enterprise culture which affects administration as well as everyday life—have all contributed to a rapid disenchantment with politics, expressed in low voter turnout and a tendency to vote against whoever is in power. —Mary Kaldor and Ivan Vejvoda, “Democratization in Central and East European Countries”

Democracy as a form of government is a pluralist competitive political system where voters hold elected officials accountable for the quality of the provided public service. Citizens can use their ballot to punish irresponsible rulers or to reelect policymakers who remained sensitive to people’s needs. In democratic systems, the utility of elections as a control instrument is realized through the active engagement of informed citizens in politics; elections “link” the public and policymakers (Powell 2000, 5). This mechanism is less certain in nonconsolidated transitional democracies, such as the former communist states, where political identities are volatile and parties are weakly rooted in society. The collapse of the previous regime and the following market reforms opened abundant opportunities for corrupt enrichment while democratic institutions and process were still feeble. As I described in Chapter 8, mass disappointment grew deep with any politician who came to power just to plunder the state. Election records document cases of incumbents who were removed from power by angry voters rebelling against unfair governance 211

212

Political Corruption in Eastern Europe

(Roberts 2010, 70). At the same time, other citizens no longer bother to vote. In this chapter, I explore the effects of government quality on the behavior of Eastern European voters. More specifically, I address the question of how corruption affects citizens from former communist states in making the decision of whether to participate in elections. Previous research does not provide much guidance for exploration of this theme. Only a few empirical studies address the impact of trust in government on political behavior (Franklin 2004; Rose 2004). One study, by William Mishler and Richard Rose (2005), focuses on the political consequences of trust on regime support in Russia and suggests that government corruption perceptions indirectly discourage civic involvement. Overall, however, how corruption shapes electoral participation is a question that remains largely unexplored. Given the increasing prominence of the problem of nontransparent policymaking after communism, this type of research emerges as being much needed. In this chapter, I treat corruption as a factor with direct as well as indirect effects on the decision to cast a ballot. Drawing upon the literature on elections as instruments through which voters reward or sanction politicians, I identify a mobilizing impact of corruption that generates incentives to kick corrupt officials out of office. This effect is clear and straightforward. There is yet another, more complicated path of influence where corruption undermines the faith of voters in the democratic process and, consequently, weakens their desire to participate in politics. In the context of postcommunist transition, the public carries the heavy burden of market reforms for years while a few among the elite enrich themselves overnight. For some citizens, the disappointment turns into futility that may then lead to disengagement. For others, sheer anger produces a drive to sanction the guilty. In this chapter I analyze empirical data to establish the relative strength of the two effects, mobilization and abstention. In the following section, I discuss trends in the dynamics of voter participation in postcommunist Eastern Europe after 1989. I describe a decline in mobilization and identify some of the possible reasons for this. Next, drawing from Susan Rose-Ackerman’s (2007) concern about the cost of neglected transparency in policymaking at the beginning of transition, I develop a model of electoral participation. My theoretical framework builds an understanding of the behavior of voters who perceive corruption as a serious threat in their country. I

Implications: Corruption Perceptions and Voting

213

then specify the proposed causal links in a recursive model of the direct and indirect effects of corruption. Next, I assess the results of a path analysis that uses survey data from the Comparative Study of Electoral Systems (2007). The data offer evidence supporting the claim that corruption perceptions may mobilize the vote but may also corrode people’s beliefs in their ability to influence policies. Moreover, the two opposing incentives balance each other in this set of data to produce a net effect that slightly favors the mobilization argument. In the last two sections, I discuss these findings with reference to the most recent data on the high levels of public distrust in institutions and the increasing concerns about political corruption. I conclude with a summary of the results and their implications for Eastern European politics and suggest avenues for future research.

Trends in Voter Turnout After 1989 Since the fall of communism in Eastern Europe, voter participation rates have changed significantly. The initial surge of enthusiasm and excitement with the newly gained freedoms and opportunities weakened over the course of three to four electoral cycles (Kostadinova 2003; Kostadinova and Power 2007). Current levels of mobilization do not differ substantially from those of other countries in transition, and are lower than those reported in many of the consolidated democracies. Participation in elections for local government officials and representatives to the European Parliament has decreased even more. Voting used to be the most active form of political activism in postcommunist countries where civil society is weak and citizens are not accustomed to community organization. In the aftermath of 1989 the new right to abstain was appreciated, but the opportunity to choose among a host of parties was valued even more. Over time, Eastern Europeans have become increasingly indifferent to voting. The common trend of voter withdrawal from politics in Eastern Europe is well illustrated by the declining aggregate rates of participation recorded since transition began. While voter mobilization averaged 82 percent in the founding elections of 1989–1991 (Kostadinova and Power 2007), only about 59 percent of voters cast ballots between 2004 and 2008.1 Just 40.6 percent of Polish voters participated in the 2005 election of representatives to the Sejm. The

214

Political Corruption in Eastern Europe

electorate in Lithuania has become especially apathetic where fewer than half (48.6 percent) of those registered to vote turned out in the October 2008 elections, after a record low participation rate of 46.1 percent in 2004. Even in Albania, with voter participation rates of 90 percent or more until 1996, turnout dropped to 49.2 percent in 2005. The Eastern Europeans show less interest in voting in elections at all levels of government. In many countries in the region, voting in local elections has been broadly ignored. Participation in the elections for the European Parliament have been traditionally low among old member states and voter turnout reached record lows in Slovakia and Poland in 2004, 17 percent and 21 percent, respectively. There could be multiple reasons for the decline in voter mobilization. Over the course of a difficult transition to political pluralism and a market economy, people faced numerous challenges. One explanation could be a natural withdrawal of voters from the electoral process after an initial euphoria over the collapse of the totalitarian regime. With the passage of time, elections became more mundane and fascination with the freedom to cast a secret ballot gradually waned. Another reason for the consistent decline in turnout is associated with fatigue from the hardships of reform (Kostadinova 2003). Once the new democracies embarked on the economic reform process, citizens had to deal with unemployment, inflation, and declining living standards. Encountering serious economic problems in their everyday lives, many lost interest in elections. As it became more difficult to find ways to support a family, concern over making ends meet replaced citizens’ interest in voting. A third reason for the overall drop in participation might be that over time, the Eastern Europeans have learned about the importance of different elections and selectively voted in only those that they feel are worth the effort (Pacek, Pop-Eleches, and Tucker 2009). More recently, voices of concern were raised that popular disengagement from political life may be a result of the way transition progressed. Rose-Ackerman (2007), for example, notes that in the first years after 1989, elites were preoccupied with producing a flurry of legislation and did not pay enough attention to the quality of democratic governance. The neglect of transparency in policymaking and the disregard for broader public involvement may have affected the young Eastern European societies more than could have been envisioned. Regardless of how quickly the legislative basis is being transformed, the lack of sufficiently effective mechanisms to keep govern-

Implications: Corruption Perceptions and Voting

215

ment officials responsible to those who delegated power to them may facilitate a process of disengagement from politics. Disappointment with overlooked public interests and uncontrolled government carries the potential of weakening popular participation in politics and eroding voter turnout. I explore this proposition further by focusing on the incentives for voters generated by the phenomenon of corruption, an extreme type of nontransparent and irresponsible governance. Corruption has emerged as a major problem in the former communist countries and it may well affect the behavior of all participants in the electoral process, including voters.

The Direct Impact of Corruption on Voting Previous studies on democratic governance have proposed a direct connection between elected officials’ performance and voter choices (Roberts 2010, 33). This link is crucial in any democratic system; it is an indicator of the operation of the governance-by-the-people principle. Regardless of the institutional model that might be more or less representative of societal interests, voters in competitive political systems have the opportunity to express their opinion in elections. In the context of political conflict over allocation of state resources, the public delegates policymaking authority to a set of politicians (Przeworski, Stokes, and Manin 1999). The politicians may then pursue goals different from what they promised in the campaign or they might simply resort to rent appropriation to better themselves (Boix 2003, 207). When those delegated with power fail to meet their policy responsibilities with integrity, overspending and corruption become possible outcomes. The mechanism through which voters can discipline irresponsible state elites and hold them accountable is regular elections. As G. Bingham Powell (2000, 51) argues, if citizens can identify who formulates and implements policies, they can use elections as an effective means to control government officials. We have learned from previous research that the likelihood that mobilized voters will kick out bad politicians is clearly correlated with clean and efficient government (Adsera, Boix, and Payne 2000). When citizens are satisfied with government policies, they support their representatives and reelect them to office. However, if things are not going well, voters may reconsider their support. Among various measures of government performance, policy transparency and deci-

216

Political Corruption in Eastern Europe

sionmaking congruent with the public interest emerge as important determinants of voter approval. Suspicion of unfair distribution of goods by officials breeds outrage and frustration. Assuming that electors vote retrospectively, dissatisfaction and anger with mismanaged public affairs can turn into a powerful threat to those in office. Corruption may affect voter behavior by mobilizing citizens to bring to power politicians known for their integrity. Revelations of secret deals struck by government officials to enrich themselves or party coffers create incentives for some voters, who would otherwise choose to abstain, to vote the betrayers of public trust out of power. When doubts arise that public resource allocation and contract assignment are not transparent, angry voters rebel and vote in large numbers to empower the opposition. In this context, citizens get mobilized to sanction disloyal politicians. This type of response to corrupt behavior is possible because voters know they can punish the guilty at the ballot box. In brief, angry citizens may choose to rebel and oust irresponsible elected officials. Ample evidence from Eastern Europe demonstrates that voters have learned to use the mechanism of electoral defeat to remove ruling parties from power. In Poland, the Law and Justice/Self-Defense ruling coalition lost the election in 2007 after allegations of corruption and the revelation of related classified information outraged the public. Polish voters turned out in the election at the highest rate since the start of democratic reforms in 1989. In Bulgaria three of the post-1989 parliamentary elections ended with sound victories for the opposition, thus leading to the alternation of parties in the government. While voter participation had been declining over the 1990s, Simeon Saxkoburggotski’s political movement ran a vigorous anticorruption campaign and won by inspiring many people to return to the polling stations in 2001.

Corruption, Political Efficacy, and Voting There is another possible causal track, however, where corruption indirectly affects voters’ choices by eroding their confidence in the ability of the democratic process to produce clean government. Here, disappointment with corrupt politicians may generate withdrawal and lower turnout at the polls. Frequent scandals around public procurement deals, politically protected criminals, and alleged illegal party

Implications: Corruption Perceptions and Voting

217

financing reinforce convictions that corruption has penetrated the state and cannot be fought. A feeling that elections cannot change things mounts after several failed attempts to clean up government by electing another party. The let-new-people-come philosophy is then replaced by disbelief and apathy. In the above-described scenario, corruption corrodes one of the key conditions for civic engagement, the belief that one can influence policies. This deficiency in political efficacy, understood as declining faith in the ability of electors to make a difference through their vote, may have distorting effects on political participation: if one cannot influence the outcome, why bother to vote?2 Empirical studies have established a link between corruption and public attitudes toward politics and the role of individuals in it. Perceptions of corruption have been found to damage people’s approval of national institutions, mostly the parliament, and the democratic process of elite recruitment. Survey research on Poland has shown that citizen perceptions of high-level corruption correlate with distrust in members of parliament and in elections as a mechanism of representation of voters’ interests (McManus-Czubinska et al. 2004). Other scholars claim that the expansion of the black economy in Central and Eastern Europe after 1989 left a negative mark on mass perception of public life and undermined citizen confidence in government (Wallace and Latcheva 2006). The empirical results of this study that I reported in Chapter 8 underscore the extremely powerful depressing impact of corruption perceptions on institutional trust. Given the consistent findings that link corruption in Eastern Europe to the formation of negative perceptions about the performance of the political system, we can expect informal practices to also have important consequences for the way that citizens feel about their role in politics. If policymaking is perceived as a hidden process through which politicians exploit the state, newly developed attitudes about civic duty and political engagement may not endure. Existing scholarship on voter behavior offers mixed evidence for the effect of alienation on participation, but is more conclusive on how political efficacy shapes voting decisions. Analyzing elections in established democracies since 1945, Mark N. Franklin (2004, 215) concludes that political disaffection did not cause a decline in voter mobilization. Dalton and Weldon (2005), however, suggest that distrust in political parties weakened interest in voting and campaigning

218

Political Corruption in Eastern Europe

in the West. Among students of postcommunist politics, William Mishler and Richard Rose (2005) find that corruption perceptions have a weak indirect effect on voting in Russia through eroding institutional trust. Pooling data from both consolidated and transitional democracies, Richard Rose (2004) also concludes that the lack of trust in parties and national governments contributed to lower voter turnout in elections to the European Parliament. When it comes to political efficacy, the connection between beliefs about the meaningfulness of voting and participation is demonstrated in several studies (for further discussion, see Dimitrova-Grajzl, Simon, and Fischer 2010). Moreover, it is internal efficacy that has been shown to exert a stronger positive impact on political behavior. Transitional democracies are especially susceptible to the corrosive effects of corruption because citizens who suffer the pains of economic transformation feel betrayed by politicians who shamelessly steal from the state. Some people start to question not just one politician or one political party, which could be removed from power should citizens choose to show up and cast ballots. To disengaged voters, the entire political system seems hopeless. From a citizen’s right to choose the most competent to govern, voting turns into a chore, one with the sole purpose of empowering greedy politicians. In the post-1989 era, voters have learned that they can stay at home rather than participate in seemingly meaningless elections.

Electoral Participation When Corruption Is High To summarize, when the scale of corruption increases, some disgusted voters are likely to mobilize to elect more responsible politicians; in contrast, others lose faith that the public can make a difference, and abstain. Thus, for some citizens, the desire to oust the thieves and to clean up the government becomes a mobilizing force. Widespread corruption, however, may also demobilize significant groups who do not want to sacrifice personal time to empower another set of rascals. How have these two possibilities played out in the Eastern European context? Table 9.1 summarizes the electoral choices made by voters in eight postcommunist states who thought that corruption was very widespread in their country. The first striking observation is the size of the group convinced that corruption had penetrated their societies—half of all

Implications: Corruption Perceptions and Voting

219

Table 9.1 Electoral Choices of Voters Who Perceived Corruption as Very Widespread Countrya

Year

Vote for Opposition (%)b

Albania Bulgaria Czech Rep. Hungary Poland Romania Russia Slovenia

2005 2001 2002 2002 2001 2004 2004 2004

78.4 72.8 53.5 62.4 58.7 61.5 n/a 63.9

8.8 19.6 29.1 17.1 39.7 19.6 30.5 24.0

81.2 52.7 60.7 26.2 55.3 57.5 26.6 37.1

64.5

23.5

49.7

Average

Abstain (%)

All Respondents (%)

Source: The Comparative Study of Electoral Systems, www.cses.org, CSES Module 2, 27 June 2007 version. Notes: The remaining categories perceive corruption as quite widespread, not very widespread, and it hardly happens at all. “n/a” indicates not applicable. a. Albania, Bulgaria, the Czech Republic, Hungary, Poland, Romania, and Slovenia: parliamentary elections; Russia: presidential election. b. Vote for parties in opposition to the Socialist Party in Albania, the Union of Democratic Forces in Bulgaria, the Czech Social Democratic Party in the Czech Republic, Fidesz and the Hungarian Democratic Forum in Hungary, Electoral Action Solidarity of the Right in Poland, National Alliance Social Democratic Party and Humanist Party in Romania, and the Liberal Democracy of Slovenia in Slovenia.

respondents interviewed in the survey. The percentage is especially large in Albania and the Czech Republic. Of special interest are the entries in the third and fourth columns, representing the proportion of those who chose to vote for the opposition (possible direct effect) and those who chose not to vote (possible indirect effect through loss of faith that they can influence policies). A significant majority within the sample (65 percent) made a trip to the polls to support opposition parties.3 This drive was the strongest in Albania, where the vote for the incumbent socialists dropped from 42 percent in 2001 to 9 percent in 2005. In Bulgaria, those most disturbed by the spread of corruption voted overwhelmingly against the ruling Union of Democratic Forces. In all countries in Table 9.1, abstention was lower than the opposition vote, ranging from 9 percent to 40 percent. In the Czech Republic and Poland, where fewer respondents chose to support the opposition, those most sensitive to the spread of corruption abstained at a higher rate.

220

Political Corruption in Eastern Europe

To appreciate the importance of these statistics, I next explored whether association exists between perceptions of corruption and voting choices. As expected, abstention appeared to be more frequent as one moves from “none” to “very widespread” corruption perceptions (r = 0.021, significant at the 0.05 level). Similarly, support for opposition parties was associated with stronger beliefs of widespread corruption (r = 0.120, significant at the 0.01 level). While these results are in the anticipated direction, it was too early for any further conclusive thoughts at this stage of the analysis. First, the reported statistics were measures of association from which one cannot infer causal relationships. Second, there were other factors that also affect voter turnout that had not been isolated thus far to avoid spuriousness. Finally, as the theoretical argument suggests, the impact of corruption may be more complex than a simple unidirectional effect and therefore needed to be modeled accordingly. Next, I addressed these concerns by building a recursive model of the influence of corruption on voter behavior.

Modeling the Impact of Corruption Perceptions Since the demobilizing effect of malfeasance is the opposite of voting as a way to get revenge, I designed a model that incorporates the two tracks of causal effects. In Figure 9.1, corruption perceptions, C, were expected to exert a direct effect on voting, V, by increasing the likelihood that a citizen would go to the polls in an effort to clean up government. C, however, also affects another determinant of V, political efficacy, E, by weakening people’s belief that who they vote for makes a difference. Thus, through E, C may discourage voting among citizens who have lost hope that they can change the way things are done in their country. In this recursive system, C is the exogenous factor while V and E are endogenous and depend on C and other exogenous variables. The latter are denoted RV and RE in Figure 9.1. As in any recursive model, an equation for each endogenous factor has to be estimated with data to obtain coefficients for the effects of interest. Here, we were most interested in the parameters of the direct and the indirect effects of corruption (C) on voting (V). To compute precise estimates of these impacts, I controlled for the specifics of the electoral context and various socioeconomic charac-

Implications: Corruption Perceptions and Voting

221

Figure 9.1 A Recursive Model of Voting Affected by Corruption Perceptions

teristics.4 The following two equations specify the proposed relationships in Figure 9.1: Voting = a + b1(corruption) + c2E + ∑dnRV + e(1) Efficacy = a + b2(corruption) + ∑dmRE + e(2)

In the above specifications voting and efficacy are the two endogenous variables (V and E) both affected by corruption (C), the key exogenous variable in the model. R V and R E stand for other exogenous determinants of voting and political efficacy. Next, I describe the data, variables, and measurement procedures used for this analysis.

Variables and Measurement To test the hypothesized direct and indirect effects of corruption on voting behavior in contemporary Eastern Europe, I used survey data from eight former communist states gathered through the Comparative Study of Electoral Systems (CSES) project. The Eastern European polls included in the CSES Module 2 were taken in Albania, Bulgaria, the Czech Republic, Hungary, Poland, Romania, Russia, and Slovenia with relation to elections held

222

Political Corruption in Eastern Europe

between 2001 and 2005 (Comparative Study of Electoral Systems 2007). The election in Russia was for president, in Romania for both president and parliament, and in the rest of the countries only for parliament. The timing of the surveys is important. All were conducted at least twelve years after the start of the democratic transition and the initial wave of enthusiasm about competitive politics. By the early 2000s corruption had already been recognized as a big problem in postcommunist politics and much of the public’s interest in elections had dissipated (Grigorescu 2006; Kostadinova and Power 2007). If voting and malfeasance are indeed correlated, 2001–2005 data would be able to provide some confirmation for such a link. Before I describe the analysis, a discussion of variable indicators and measures is in order. Voting The dependent variable, the choice to participate or abstain, was measured through responses to the question of whether the respondent cast a vote. In postelection face-to-face surveys there is always the possibility of overreporting, with some of those who abstained claiming they voted. This is shown by discrepancies between the rates of participation documented in the surveys in the eight countries and the actual voter turnout. Despite this shortcoming, the survey data provide the best measure of voting at the individual level. The variable takes the value of 1 if the respondent indicated participation and 0 for abstention. (See Appendix A for descriptions of all variables and sources.) Corruption One reason for the absence of systematic cross-national studies on the effects of corruption, including its impact on voter behavior, is the secret nature of this phenomenon. As previously discussed, illegal exchange of favors and cash is difficult to observe directly and its magnitude is hard to gauge. For the purposes of this analysis, I considered perceptions as a better measure than the experience of bribe taking or giving. If there is an effect of corruption on what people choose to do on election day, this would be reflected more in what they think about the integrity of public officials and less in the actual level of corruption (which might even be benign in reality). Therefore, I measured

Implications: Corruption Perceptions and Voting

223

corruption by how widespread it was as perceived by individual voters. The variable is coded in four categories, with higher values given to perceptions of very widespread corruption. Political Efficacy I operationalized the concept of political efficacy as individual belief that elections matter. It is best captured by answers to the CSES question of whether “who people vote for makes a difference” (Comparative Study of Electoral Systems 2007). The variable is coded 1 for those who believe that the choice of candidates matters and 0 for those who do not believe this. Electoral Context Comparative studies of voter turnout have emphasized the importance of institutional and political process factors. G. Bingham Powell (1986) and Robert W. Jackman and Ross A. Miller (1987) argue that proportional representation systems with large district magnitudes favor voter participation because parties launch nationwide campaigns; single-member district systems do not generate such incentives and mobilization is not as broad based. Here, I measured the mobilization potential of electoral systems by the average district magnitude, logged for a more normal distribution. Multipartyism was the second institutional variable that I found to depress voting because too many parties may confuse and alienate voters (Jackman 1987; Kostadinova 2003). It was measured by Markku Laakso and Rein Taagepera’s (1979) effective number of parties, based on votes, in the lower house or single chamber of the national legislature. The third institutional factor was concurrently held presidential and parliamentary elections, with concurrence expected to raise the importance of participation (Franklin 1996). This variable is binary, scored as 1 for respondents from Romania, the only country in the data set with parliamentary and presidential elections held on the same day. Beyond institutional factors, the domestic political environment also shapes voting choices. Following Andre Blais (2000), I controlled for closeness of the race: the more competitive the election, the more voters feel that their ballots may turn the outcome. Closeness of the race is indicated by the distance in vote shares

224

Political Corruption in Eastern Europe

between the top two parties. Finally, democratization strengthens people’s perceptions of the meaning of voting. Despite sharing a common communist past, the Eastern European nations differ in levels of pluralism and political freedom. I relied on data from Freedom House and combined the separate political rights and civil liberties scales into a composite scale for each country, ranging from 2 (most democratic) to 14 (least democratic) (see Appendix A). Moving to demographic, socioeconomic, and political individual characteristics, I identified six factors that may stimulate turnout. Gender was one of them, with men more likely to vote and to recognize the importance of voting choices than women. I constructed a dummy variable by assigning 1 to male voters. Age, measured in years, was expected to show higher mobilization among older voters. School was a factor enhancing citizen knowledge about the political system and stimulating participation; I coded it on an eight-category scale from 1 (none) to 8 (completed university degree). Wealth, measured as household income in categories from low to high, and residence, coded 1 for residents of urban areas, were anticipated to favor turnout through exposure to diverse sources of information and to national politics. Finally, party identification has been found as a powerful factor driving people to vote; I coded it as 1 if the respondent revealed closeness to a party.

Analysis In choosing an appropriate method for estimation of the path coefficients in Figure 9.1, I first acknowledged the binary nature of the two endogenous variables. This led to the selection of a logistic regression technique that produced parameters for the probability of efficacy and the occurrence of voting. 5 The relative importance of the factors affecting the dependent variable can best be assessed by comparing the “effects coefficients” produced by the path analysis (Lewis-Beck 1977). To enable a thorough interpretation of the direct and indirect effects of corruption, I standardized the logit coefficients from the regression analysis.6 The results of the logistic regression analysis offered empirical evidence for the hypothesized effects of corruption perceptions. Malfeasance increased the likelihood of voting, as suggested by the positive sign of the respective coefficient produced by the estimation

Implications: Corruption Perceptions and Voting

225

Table 9.2 Logit Coefficients for Predictors of Voting and Political Efficacy Predictors Corruption Political efficacy Constant –2LL Pseudo R2 N

Voting .079 (.062) .999 (.113) –1.631 (.930) –2544.493 .162 5,889

Political Efficacy –.272 (.131) — 2.199 (.845) –3891.147 .028 8,140

Notes: Entries are logit coefficients; standard errors in parentheses are robust, clustered by country. The estimates reported in the table are only for key variables, corruption and political efficacy. Full results are listed in Appendix C.

of equation 1 (see Table 9.2). This result is indicative since it was generated after controlling for other individual-level sociopolitical factors and for contextual determinants of voter turnout established by previous scholarship. The same set of regression results also confirms that, when Eastern Europeans believe that voting matters, they are more likely to cast ballots in elections. Next, I estimated how that belief was affected by corruption. As expected, the impact of perceptions of misconduct appears to have eroded citizens’ understanding of elections as an instrument through which they could influence policymaking. This is indicated by the negative and statistically significant coefficient of corruption in the regression results for equation 2. Most of the control variables performed in the expected manner, producing the expected sign and reaching statistical significance. (See Appendix C for detailed results.) If perceptions of dirty politics mobilize voters but also weaken their sense of efficacy, what do the data suggest about the net impact of corruption? A more complete picture of the importance of corruption perceptions for voting can be revealed by examining the estimated parameters in terms of direct and indirect effects. Evaluating the direct impact is straightforward; it is indicated by the standardized coefficients produced by computing equation 1 (see Table 9.3). Both corruption and political efficacy were found to directly influence voting decisions: their standardized coefficients showed that the direct impact of the latter was much stronger than that of the former (com-

226

Political Corruption in Eastern Europe

Table 9.3 Direct and Indirect Effects of Corruption on Voting Predictor Corruption Political efficacy

Direct Effects

Indirect Effects

Total Effects

.03 .21

–.02 —

.01 .21

Note: “Indirect effects” is the product of the effect of corruption on political efficacy and the effect of political efficacy on voting. Entries are based on standardized coefficients (see Appendix C).

pare 0.21 to 0.03). Obviously, we would like to determine whether the effect of efficacy prevails over perceptions of malfeasance when the indirect impact of corruption is also considered. The indirect effect was obtained by multiplying the standardized coefficients on the path from C to V via E (Figure 9.2), which resulted in a negative number (–0.02) as expected. Comparing the absolute values of the two effects’ coefficients, one cannot help but note that the direct effect slightly outperformed the indirect effect. In sum, the data offered evidence for two opposite incentives generated by corruption perceptions, which balance each other, but the drive to vote slightly dominates (as shown by the net effect coefficient of 0.01). The conclusions, derived from the effects’ coefficients in Table 9.3, are in agreement with the results reported from the logit models. Yet estimating the indirect effects of corruption helped to compare the amount of apathy to that of voter mobilization. Clearly, these data do not provide evidence for a substantially stronger mobi-

Figure 9.2 Estimated Path Coefficients

Source: Author’s calculation based on regression analysis in Table 9.3.

Implications: Corruption Perceptions and Voting

227

lization effect. By the same token, one might argue that disengagement among voters, disappointed with politicians’ corrupt practices, was weaker than observers have feared.

Discussion Key ingredients of democracy, accountability and citizenship, play a central role in the way citizens react to malfeasance. During the democratic transition, however, public engagement in politics is more sensitive to failures in the administration of transparent policies. The survey data from Eastern Europe confirmed that corruption is becoming a major problem, one that threatens government accountability and challenges voters. The results of the path analysis above showed that the spread of corruption may mobilize some and push away others from electoral participation. At the aggregate level, either these opposite incentives will cancel each other out or the effect of one will prevail. Under what circumstances the latter may happen is a question of vital importance for the success of democratic governance. My analysis reveals that in eight Eastern European elections held between 2001 and 2005, a slightly more popular choice among those frustrated with the misuse of power by elites was to vote. Yet the indirect effect of perceived corruption was also in force, undermining citizens’ faith in the meaning of democratic elections. Perceptions of widespread misuse of power damaged the fragile belief in the principle that who people vote for should make a difference. If both types of citizens existed in 2001–2005, what are the most recent trends in participation? Toward what choices are Eastern European voters moving today, twenty years after transition started? The answer may not be especially promising. Recent surveys and elections from the region contain troubling signs that the negative effects of corruption on how people feel about politics might be increasing. The Special Eurobarometer 291 survey on Europeans’ attitudes toward corruption, conducted in November–December 2007, revealed extremely high levels of concern in the newly accepted EU member states (Attitudes of Europeans Towards Corruption 2008) (see Table 9.4). The view that there is corruption in national institutions was maintained by large majorities of Eastern Europeans; the highest rates of disappointment

228

Political Corruption in Eastern Europe

Table 9.4 Recent Levels of Corruption Perceptions, Trust in National Parliament and Political Parties, and Voter Turnout (percentage)

Country Bulgaria Croatia Czech Rep. Estonia Hungary Latvia Lithuania Macedonia Poland Romania Slovakia Slovenia

Political Institutions Leaders Are Are Corrupt Very Corrupt 85 n/a 93 75 89 88 91 n/a 83 88 91 90

64 n/a 78 n/a n/a n/a n/a n/a 61 n/a 66 n/a

Trust Parliament Parties (2008) (2008) 12 21 16 36 15 12 12 24 16 22 34 31

9 9 11 15 9 5 7 16 7 18 11 13

Voter Turnout in Recent Parliamentary Election 55.8 (2005) 59.6 (2007) 64.5 (2006) 61.9 (2007) 64.4 (2006) 61.0 (2006) 48.6 (2008) 58.0 (2008) 53.9 (2007) 58.5 (2004) 54.7 (2006) 63.1 (2008)

Sources: The Attitudes of Europeans Towards Corruption, Special Eurobarometer no. 291 (Brussels: European Commission, April 2008), http://ec.europa.eu/public _opinion/archives/ebs/ebs_291_sum_en.pdf; Global Opinion Trends, 2002–2007 (Washington, DC: Pew Research Center, July 2007), http://pewglobal.org/reports /pdf/257.pdf; Standard Eurobarometer 69: Public Opinion in the European Union (Brussels: European Commission, June 2008), http://ec.europa.eu/public_opinion /archives/eb/eb69/eb69_en.htm; International Institute for Democracy and Electoral Assistance Voter Turnout Project (Stockholm: International IDEA), www.idea.int/vt /index.cfm. Note: “n/a” indicates not available.

existed in the Czech Republic (93 percent) and Lithuania and Slovakia (91 percent each). This commonly shared opinion translates into declining levels of trust in national institutions, especially in legislatures and political parties. On average, only 21 percent of the citizens in postcommunist Europe trusted their assemblies, compared to 36 percent in the rest of the EU. Between 2005 and 2008, the Eastern European parliaments lost the trust of 13 percent (from 25 percent to 12 percent) of the citizens in Latvia, 11 percent (from 26 percent to 15 percent) in Hungary, and 10 percent (from 46 percent to 36 percent) in Estonia. Even lower were the relative numbers of those who thought of political parties as trustworthy. On average, only 11 percent of the Eastern Europeans (compared to 18 percent in the rest of the EU)

Implications: Corruption Perceptions and Voting

229

trusted parties as being representative of people’s interests in the spring of 2008. Their approval was extremely low in Latvia (5 percent) and Lithuania and Poland (7 percent each). The data suggest a pattern: where trust in parliaments and parties was low, voter participation in most recent elections was also low (e.g., Lithuania and Poland). The perceptions of political corruption damaged public trust in democratic institutions and fueled popular disillusionment with politics.

Conclusion Acknowledging previously confirmed determinants of voter participation, I sought in this chapter to establish the impact of political corruption on electoral behavior in postcommunist transitional democracies. Eastern Europeans are extremely frustrated with practices of unfair government and abuse of power by politicians, and their anger breeds skepticism and disillusionment. Instances of corrupt behavior among public officials may not have a crucial impact on voters’ choices to cast a ballot in mature democracies with established democratic processes and institutionalized party systems. In transitional environments, however, parties are not well rooted in society and the public lacks experience with democratic procedures to hold the ruling elite accountable (Rose-Ackerman 2007). As a result, citizens disillusioned by the irresponsible behavior of elites have good reasons to abstain. In postcommunist states, where abstention is a newly obtained right, such a choice can be attractive and costless for citizens. My findings confirm that Eastern European electoral behavior has been shaped by perceptions of how clean government policies have been and how trustworthy those competing for power are. The central argument that I advanced here is that the effects of corruption on voting are more complex than merely to say that malfeasance depresses voter turnout. While abstention is certainly one choice that disillusioned voters may consider, I also propose that outrage toward corrupt politicians may bring people to the polling booths. These relationships and outcomes were incorporated into a recursive model whose utility is in the inclusion of two opposite choices: voting (direct effect of mobilization) and abstention (indirect

230

Political Corruption in Eastern Europe

effect through erosion of beliefs that voter choices matter). This modeling strategy enables a better understanding of two different causal mechanisms through which corruption perceptions may influence the decision to vote. Furthermore, through path analysis, I was able to estimate and compare the magnitude of corruption’s direct and indirect effects. The most important finding is that perceptions of systemic office abuse in transitional settings may both encourage and discourage voter participation, all else kept the same. Widespread belief that government is corrupt makes some indignant citizens go to the polls to throw out the rascals. To others, in times when difficult reforms require sacrifice, casting ballots to empower yet another group of greedy politicians seems to be asking for too much. The relative size of the clusters of active and demobilized voters may vary across countries and over time, depending on the level of democratization and the scale of corruption. In settings with insufficient institutionalization of the party system and unconsolidated democratic processes, the group of apathetic voters may become significant and decisive in the final outcome of the election. Most recent survey data have revealed that large majorities among the Eastern European publics consider corruption the major problem facing their countries (Attitudes of Europeans Towards Corruption 2008). Opinion polls also have shown a further increase in the rates of public disapproval of the behavior of political parties and members of parliament (Standard Eurobarometer 69). Whether this will translate into an increased desire to participate in public life will depend on the success of democratization, including the socialization of voters. More than two decades after the start of democratic transition, important questions about the impact of corruption on political behavior remain unanswered. Is it possible that the structure of incentives, generated by corruption perceptions, changes over time? Under what conditions would socially weak groups trade their votes for money and favors, and how could this affect voter turnout? Aside from voting, how do corruption perceptions affect citizens’ choices to join campaign activities, sign petitions, or organize protests and demonstrations? Corruption has the potential to shape public behavior in ways much wider in range than those that I analyzed in this and the previous chapter.

Implications: Corruption Perceptions and Voting

231

Notes A previous version of this chapter was published in the special issue “Elections, Parties, and Voters in the New European Democracies: 20 Years After the Fall of Communism” (ed. Mary Stegmaier), Politics and Policy 37, no. 4 (2009). This version benefited greatly from feedback provided by Mary Stegmaier and Joshua Tucker. Special thanks also to Sarah Poggione for valuable advice. 1. My calculations were based on data from seventeen Eastern European states in the period between 2004 and 2008, International Institute for Democracy and Electoral Assistance (IDEA) Voter Turnout project, Stockholm: International IDEA (www.idea.int/vt/index.cfm). 2. Scholars emphasize that the concept of political efficacy has two components: (1) external efficacy, which refers to individual assessments of the responsiveness of the government; and (2) internal efficacy, which is defined as an individual’s belief that he or she can understand and effectively take part in the political process (Karp and Banducci 2008; DimitrovaGrajzl, Simon, and Fischer 2010). In this study, my focus was on internal efficacy, which reflects doubts about citizens’ political aptitude. 3. The same statistic for the group that thought there was hardly any corruption was just 48 percent. Another 24 percent chose to abstain. 4. Previous research suggests that the decision to vote may be affected by individual socioeconomic and political factors (Pacek and Radcliff 1995; Norris 2004). Thus, men, older citizens, the better educated, the wealthier, urban residents, and those who identify themselves with a political party were found to be more likely to vote. Students of voter turnout also argue that larger electoral districts, multipartyism, competitive races, and concurrent presidential elections cause higher mobilization for participation (Powell 1986; Jackman 1987; Norris 2004; Blais 2006). Similar to trust in the political elite, efficacy can also be affected by socioeconomic status and by the level of democratization and political freedom in the country (Mishler and Rose 1997, 2001). 5. Another methodological concern is heteroskedasticity, a potential source of bias given that I used cross-sectional survey data from eight Eastern European states. I addressed this through computation of robust standard errors clustered by country. Finally, the data that I used for this analysis can be represented as a multilevel structure in which level 1 units are individual respondents and level 2 units are the Eastern European countries where CSES surveys were conducted. Some authors suggest the use of hierarchical modeling (Achen 2005), yet others argue in favor of a more pragmatic approach (Primo, Jacobsmeier, and Milyo 2007). Between the two possible methodological approaches for analysis, traditional models and multilevel models, I chose the former for several reasons. First, the individual-level variables are measured in exactly the same way across all surveys and that eliminates the issue of disparate measurement, as Duch and Stevenson (2005) suggest. More importantly, studies have emphasized that most of the

232

Political Corruption in Eastern Europe

damage from the use of traditional models occurs in the estimates for country-level factors (Christopher J. Anderson and Tverdova 2003; Karp and Banducci 2007). Since my main goal was to assess the impact of two individual-level factors, corruption perceptions and political efficacy, the threat of underestimated standard errors was not as great as it might be in other cases. Finally, my model did not include interaction terms. As Primo, Jacobsmeier, and Milyo (2007) argue, regression corrected for clustering on units is often a better approach than the trade-offs of hierarchical modeling, especially when there are no cross-level interactions and the data sets are large. 6. This procedure in logit models was not as straightforward as in OLS regression. For that, I chose an approach proposed by Menard (2004) and standardized on both the predictors and the dependent variable. This technique employs (1) the sample standard deviation of the predictor; and (2) π/√3 = 1.814, the standard deviation of the standard logistic distribution, as the estimate for the standard deviation of the dependent variable. The procedure gives the (logit standardized) coefficient B = (b)(sdX)/( π/√3).

10 Conclusion

Incomplete democratization often puts into place incentives for corruption, while lacking those elements of democracy that would enable those harmed to fight back. —Mark E. Warren, “What Does Corruption Mean in a Democracy?”

Konstantin Simis (1982, 211) concluded that the Soviet government “will never rid itself of corruption as long as it remains Soviet.” The USSR and the Soviet bloc are now long gone, as are the totalitarian regime and its apparatus, which Simis blamed for “infecting” society with the virus of power abuse and misappropriation of state property. Yet the empirical data offered by many studies, including mine, have shown that corruption has not been rooted out after the collapse of the previous system. In fact, corrupt practices have plagued the lives of the transforming Eastern European societies in a variety of forms after the demise of communism. Corruption there is more pervasive than in the group of consolidated democracies and some states in the region are in much worse shape than others when it comes to abuse of public authority. Why is this the case if all postcommunist countries share the common legacies of Soviet totalitarianism and the challenges of democratic institution building and marketization? And has malfeasance in Eastern Europe put at risk the formation of a majority of citizens to support democracy and engage in politics? These intriguing puzzles drove my interest in corruption after communism. I believe that the analyses in this book provide some answers to these questions. 233

234

Political Corruption in Eastern Europe

Overview of the Central Findings It is time now to step back and comprehend the results of my inquiry from the perspective of the research agenda on corruption in general. In the preceding chapters, I drew on previously made arguments to identify the sources of corruption in Eastern Europe and expanded the research in areas relevant to the postcommunist transition. The exploration of the causes of the uneven spread of malfeasance and of its damaging effects produced findings worth yet another round of elaboration. In the remainder of this chapter, I recapitulate the empirical results, address their implications for broader theoretical debates, and discuss areas of research that need urgent scholarly attention. Election Rules Whether and how electoral rules affect the incentives for corrupt behavior in the postcommunist countries remain questions requiring further consideration. My empirical analysis offered weak evidence for the impact of the specific method of seat allocation, previously confirmed in some studies on larger sets of countries (Persson, Tabellini, and Trebbi 2001; Kunicova and Rose-Ackerman 2005). In Chapter 3, I explored this possibility by operationalizing election formulas as degree of proportionality (i.e., as whether they were partyor candidate-centered systems) and as district size. There could be at least two reasons for the lack of empirical support. First, in the years that I analyzed, the Eastern European countries used either pure PR or mixed systems, but not entirely (for all parliamentary seats) majoritarian. Thus, the range of electoral methods included in this analysis is not as wide as one would desire. The results show that the presence of individual representative districts within mixed systems has not reduced the spread of corrupt practices. The second reason is based in controversies in extant scholarship about the mechanism of the election system—corruption causal effect. As I recognized in Chapter 2, previous research on the institutional sources of corruption has rationalized expectations for the PR systems in different, and sometimes opposing, ways. Moreover, a recent study by Herbert Kitschelt and Steven I. Wilkinson (2007, 43) proposes that formal electoral institutions might not be used as successfully as previously thought, at least directly, to explain “the precise operational techniques politicians employ to build their favorite

Conclusion

235

linkage patterns.” My results presented in this book can thus be understood through Kitschelt and Wilkinson’s notion that the influence of formal rules on democratic linkage practices is conditioned by other social, economic, and competitive structures. At the same time, another aspect of electoral rules appeared to be particularly relevant to the spread of dishonest politics in Eastern Europe. In my cross-national analysis, campaign funding regulations were significant barriers to suspicious money exchanges for political influence. No doubt, politicians may continue to find a way to work around caps and restrictions imposed by the law and regulations may be poorly enforced. Yet these findings based on empirical data from the postcommunist countries speak to the importance of introducing regimes that limit irresponsible fund-raising and provide the public with mechanisms to monitor, investigate, and sanction wrongdoing. My case study of Bulgarian informal political-private networks (see Chapter 5) provides more detail about how illegal financing of parties may put the latter in a dependent position and how flaws in the legislation exacerbate the situation further. The mechanisms through which friendly circles can form, siphon public resources, and order favorable laws are shaped both by the rising costs of modern election campaigning and by the active presence of former state security officers with close connections to party leaders. While in the Western democracies improper political funding is also a function of demand and supply, in the postcommunist world money largely comes from private interests with ties to the former regime. This explains to some extent the unique persistence of the informal networks in the post-1989 period. Economic Development Another important factor that I found to explain cross-national variation in corruption in post-1989 Eastern Europe was economic prosperity. The role of national wealth, broadly defined, has been discussed in previous research from different perspectives and I needed to specify its impact in clearer terms. Most of all, wealthier nations may be less corrupt because their citizens will not resort to corrupt practices to satisfy the daily needs of their families. Empirical support for this claim is offered by my analysis in Chapter 3 where economic development is indicated by individual well-being (GDP per capita). This finding is congruent with the excuse given by common

236

Political Corruption in Eastern Europe

people and bureaucrats in some countries that they engage in bribery because their income is so low (Miller, Koshechkina, and Grødeland 1997). It seems that, where the average citizen is better off, corruption is not as pervasive. The above argument can be contested, however, mainly on the grounds that the average Eastern European might actually be poorer than GDP per capita suggests. Democratization after 1989 reached significant achievements in the political arena, but the economic transformation to a market economy led to the impoverishment of significant strata of society. If skewed distribution of wealth leads to more corruption, as Eric M. Uslaner (2008, 122) argues, then Eastern Europe would inevitably fall into the “inequality trap” given that power abuse and low social trust are already prevalent. Indeed, political reforms “did not turn Romania or Russia into Sweden” and the poor are angered by the dishonesty of those who unfairly enriched themselves. Common people are frustrated not only there, but also in the better-performing, and less corrupt, Estonia and Lithuania. What make the Eastern European context distinct in dealing with corruption are the shared rates of relative social equality during communism and departure from a planned economy to capitalism since 1989. Within the region, postcommunist transformation has been similar in direction across space, yet some countries have become more successful in becoming market democracies than others. My results suggest that, by 2008, the wealthier Eastern European states had also managed to better contain corruption. There is another implication of the confirmed effect of the economy on reducing corruption. In brief, economic development strengthens the material capacity of a state to prevent, detect, and sanction misuse by officeholders. As the data reveal, corruption levels have been lower where the state has resources to pay bureaucrats better. To the best of my knowledge, these are among the first empirical findings confirming the importance of a merit-based, reasonably well-paid civil service for combating corrupt behavior within postcommunist administrations. Achieving such a reward system is more complicated than it looks, however. The problem is not only in how to secure funds for higher salaries, but also in how to rally support for this among society. It is difficult to convince those who became poorer because the state was robbed by greedy politicians and bureaucrats that public officials should be paid better. In the predominantly egalitarian Eastern

Conclusion

237

Europe, this is not an easy task. Even when a critical mass of approval exists in favor of decent pay for administrators, the state can finance salaries only to the extent that the economy permits. Data from nine postcommunist countries confirmed the relevance of making such funds available. When public sector employees were paid better to maintain a more decent living, the spread of corruption was reduced. Civil Society As Chapter 6 shows, politics were cleaner in countries where nongovernmental organizations managed to sustain financially and strengthen structurally over the course of the Eastern European transition. This third sector has emerged as a central player in the battle against corruption by taking active positions against officials’ ethical violations and by keeping the anticorruption discourse alive. The policy implication of this is that invigorating the work of the civil organizations and strengthening their mobilization effort may produce significant positive results. When applied to Eastern Europe, the argument that a vibrant civil society makes it difficult for corruption to flourish should account for certain specifics. The political pressure that organized societies can produce has already been recognized as an effective force everywhere (Earle 2000). The problem in transitional countries is that they inherit civil societies that are disabled and weak at the beginning of the transition when their role could be crucial. This is true especially for the postcommunist nascent democracies where the totalitarian regime had almost suffocated attempts at contemplation independent from government initiatives (Linz and Stepan 1996). Multipartyism was not allowed and the church, with small exceptions, was muted as an autonomous agent of socialization. In 1989, Poland and Czechoslovakia had their Solidarity and Civic Forum, but human rights and environmental groups in the other countries were feeble. Growing and maturing were tasks that the awakening civil societies of Eastern Europe had to achieve early in order to successfully serve their democratic purposes. Where nongovernmental organizations became strong enough to mobilize and push for transparency, the probability of containing corruption significantly increased despite the opening of extraordinary opportunities for misappropriation.

238

Political Corruption in Eastern Europe

On a less optimistic note, recent developments in some countries have raised concerns about the future of civil societies and their potential to put pressure on the governing elite to abide by the code of office integrity. In Russia, observers note a serious “roll back” during President Vladimir Putin’s second term in office, with a trend to treat civil society groups as “facilitators” of authorities in implementing government policies (Jeanne Wilson 2008). A new law was adopted in 2006 that obliged independent groups to file reports about their activities on a regular basis; consequently, several groups were closed down as they were considered “a threat to the state” (Harding 2008). The opportunity to criticize and express independent opinions was inhibited and so was the capacity of various NGOs and the media to expose embezzlement and corruption. If Russia has been sliding down this slope for some time now, Hungary surprised all experts with Prime Minister Viktor Orban’s new media law. The latter imposes central supervision and significant fines on Hungarian journalists’ publications, thus in practice weakening their voices and incentive for investigative work (“Charlemagne” 2011). Because healthy civil societies enable states to fight corruption, limiting the independence of third sector actors will inevitably hurt the anticorruption effort and reduce democratic quality. International Pressures My findings in this book speak to the ongoing debate about the effectiveness of international organizations in fighting corruption domestically. In particular, I present the role played by foreign actors in promoting public awareness of malfeasance and its various forms and in encouraging anticorruption policies in countries in transition. Concerns about the spread of corruption worldwide were raised by international actors after the Cold War and reached a peak in the mid-1990s. Interestingly, the increase in global awareness of this phenomenon coincided with the postcommunist transition. Amidst the intensified global campaign against corruption, the first scandals about improper privatization broke out in Eastern Europe (Grigorescu 2006). Thus, the overlap in time helped the postcommunist societies recognize the seriousness of the threat and the importance of fighting it. Under pressure from international organizations the Eastern European governments joined multilateral treaties for cooperation

Conclusion

239

and adopted national laws enhancing transparency and accountability in public affairs. These were formal commitments, however, with fulfillment heavily dependent on enforcement, resources, and political will. As my analysis of the experiences of the Balkan countries showed, the desire to join the European Union was a much more effective driving force for implementation of anticorruption policies. This comes to suggest that normative pressures are not sufficient for real results because they lead only to passing laws that might never be put into action. My findings in Chapter 7, derived from a differently designed analysis, confirm insights from previous studies about the temporal dynamics of the EU accession. What worked in Eastern Europe, at least in the EU-associated countries, was the credible promise of membership in the foreseeable future (James H. Anderson and Gray 2006).1 Moreover, with the effects of EU pressure isolated from those of marketization, my empirical results challenge the argument that economic liberalization is what actually drives the reduction of corruption (Markovich 2008). Finally, I would like to address comments made by observers about the appropriateness of Western criticisms and pressures on postcommunist countries to eradicate corruption. Some have argued that the West does not have the moral right to accuse Eastern Europe of corrupt practices because Western countries themselves are plagued by malfeasance (Krastev 1998; Sajo 1998). In Andras Sajo’s view, the Western accusation that countries in transition are corrupt is just “a therapeutic means of preserving self-esteem, of maintaining a sense of moral superiority” (Sajo 1998, 42). It is the “social construction” of the problem imposed from outside that is reflected by corruption perceptions, not real corruption (Sajo 2003). Other scholars go on to suggest that corruption levels in postcommunist politics are strongly exaggerated; therefore, the threat is not as great as portrayed by international organizations interested in strengthening their leverage over Eastern Europe (Holmes 1999; Raubisko 2001). Politicians in power concur by questioning the way the EU and other international bodies determine corruptiveness levels and by blaming foreign experts for being biased toward new democracies.2 Whether the troubling levels of corruption reported for the region by foreign agencies are an external construction or not is of little significance if the Eastern European countries think of the problem as one of vital importance for the health of their democratic systems. As the recent shift in the EU enlargement policy shows, reducing corrup-

240

Political Corruption in Eastern Europe

tion is now a requirement for the much-desired European integration; elites simply cannot ignore it. After all, combating corruption should not be perceived as a strange vagary meant to fight nonexistent problems. My findings presented in this book reveal that international pressures produced behavioral changes in the expected direction, results that did not occur in the absence of such a push. Media coverage of corruption scandals and their (mis)use in political struggles might have been overabundant, but one can sense a broadly shared understanding among the public that the EU’s requests for cleaner policies are not groundless. Ironically, many people in the admittedly more corrupt Romania and Bulgaria think that only the Union can save them from corrupt politicians. Taking a Toll on Political Trust and Engagement Political corruption may have especially dangerous effects in democracies because it undermines the core idea of citizen equality before the law (Heywood 1997). A central finding of my study is that in Eastern Europe, corruption has severely damaged people’s confidence in public institutions and, in many instances, has led to mass disengagement from politics. This is not unusual as these effects are well documented in countries in other parts of the world. Consolidated democracies are no exception; they also suffer from the harmful impact of malfeasance on citizen participation in public life. In the postcommunist world, however, the magnitude of this effect is much stronger because of the fragility of the nascent institutions and the unprecedented opportunities for office abuse opened by the privatization process. Both modernization of the public administration and adoption of labor and consumer laws were postponed, and the burden of transition was shifted to a vast majority of people who were left with no social protection (Haynes and Husan 2002). The insider nomenklatura privatization of the 1990s generated bitterness, frustration, and disappointment with the unfair and dishonest methods of redistribution of state property. Increasing public trust and confidence in the political process thus appears to be a necessary prerequisite for protecting the young postcommunist democratic project from losing legitimacy due to corruption. As recognized in other sociopolitical contexts, elite integrity and politicians’ responsiveness to the needs of people are the main pillars of the democratic government. By embodying these principles

Conclusion

241

in law and practice, states can facilitate an environment in which civic culture and involvement can thrive, which would increase their own policy effectiveness (Denhardt 2002). To use Mark E. Warren’s (2004, 341) terms of democratic conceptualizing, “empowering” common Eastern Europeans to reveal and resist corrupt behavior would be the most powerful protection of democracy against predatory elites. This is not easy to achieve, however, given past cynicism toward the state and disbelief in collective action. Yet inclusion in, rather than exclusion from, politics seems to be the strategy that would facilitate quality in democracy. My finding in Chapter 9 that abstention from electoral participation and voting to punish incumbents balance each other illustrates well how challenging the repair of the erosion effects of corruption will be in Eastern Europe. The high rates of voter turnout from the first years of transition are long gone. Important reasons for this decline are the citizens’ exhaustion with the painful reforms and the slow speed of improvement of their everyday life. Moreover, many people in Eastern Europe have developed bitter feelings toward teams of politicians who, after gaining power, became extremely rich and helped circles of others to do the same behind the scenes. Across the region, angry voters punished rulers who, in reality or allegedly, lost the public trust. Andrew Roberts (2010, 37) observes that the Eastern Europeans have effectively used the mechanism of electoral accountability to “control” politicians. What is bothersome, however, is that new winners come to power on platforms of big promises to save the people from corrupt incumbents, only to lose the next election for exactly the same reason. A sense that “they all are rascals” has spread that keeps many people from voting because they feel that nothing can be changed. Which of the two forms of citizen response to corruption prevails in the future is important for it will determine the quality of democracy in Eastern Europe. Then is the threat of exaggeration, about which some have warned, especially serious when it comes to how people perceive corruption in their country and what support they are ready to provide for their government? As I underscored earlier, sharpening awareness of corruption and keeping the issue high on the agenda of democratic transformation were prescribed by international organizations as crucial to the health of democracy. It is also possible, however, that the media fuel the debate for purely commercial reasons and that elites take advantage of this in the context of political competition. The

242

Political Corruption in Eastern Europe

puzzle is: has excessive talk about corruption created disbelief and disaffection in measures dangerous for democratic legitimacy? There is room for hope. I demonstrated in Chapters 8 and 9 that corruption, as perceived by the Eastern European people, has a far more corrosive effect on trust in national parliaments and parties than on approval of democracy in general. Furthermore, political efficacy is being eroded by perception of widespread corruption but, along with apathy toward politics, there also are significant groups of active voters ready to sanction irresponsible officials.

Institutional, Rational Choice, and Cultural Explanations Compared Based on the results that I reported in this book, it would be fair to conclude that some institutions mattered for the levels of corruption reached in the postcommunist states. In particular, certain elements of party finance legislation and the introduction of civil service reform definitely facilitated the exposure of and discouraged engagement in corrupt practices. Despite concerns that campaign finance laws in Eastern Europe are not sufficiently rigorous in ensuring transparency and sanctioning violations, my analysis suggests that some institutional traits have made a difference. Administrative reform also started to reshape the structure of incentives by introducing a merit-based system of recruitment and promotion and depoliticizing the bureaucracy. Similarly, although the transformation was hampered by poor implementation and backsliding, corruption was reduced after institutional change was launched. At the same time, conclusions about a possible impact of electoral rules are much more difficult to make. Not certain, at least in some places, is also the effect of adoption of anticorruption legislation under international pressure. Albania is a stark example of this. What are the possible reasons for the unclear role played by institutions in combating corruption in Eastern Europe? Why have some institutions made more of a difference than others? If laws structure the rules of the game and generate incentives for particular types of behavior, why have political actors not responded to all arrangements as expected? Have they acted irrationally? I do not believe so. There is plenty of evidence in this book, and in other research, that the postcommunist elites have been quite rational in both designing new

Conclusion

243

institutions and choosing strategies for maximizing access to power and policymaking. My analyses show that, where sanctions for illegal party finance and failure of disclosure are instituted, corrupt exchanges are at a much lower rate. Also, the better-paid Eastern European bureaucracies are less corrupt because the gains from official salaries outweigh those from violating the code of conduct. Finally, when the prospects of EU accession are real, the anticorruption effort intensifies because the threat of losing the gains of membership is greater than any booty acquired through malfeasance. Another alternative explanation for the failure of some institutions to play the role that is usually attributed to them could be found in culture. As others (Amorim Neto and Cox 1997; Taagepera 2002) note, a rivalry exists between institutionalist- and culturalist-based theories that explain political phenomena. I have already discussed this issue in Chapter 2, albeit from a different perspective. The argument I started with is that the Eastern European countries share past communist legacies of tolerance of stealing from the state and bribing public officials. To the extent to which these are in force, they favor the persistence of office abuse and account for why the Eastern European societies are more corrupt than those of established democracies. However, the communist past cannot explain corruption distinctiveness because it is a common, shared legacy. Even the citizens of the Baltic republics, who proudly self-affiliate with the Nordic countries, are significantly closer to the other Eastern Europeans than to the Scandinavians due to their predominantly survival-oriented value system (Taagepera 2002). When accounted for in my crossnational analysis presented in Chapter 6, the cultural factor received partial support for impacting corruption. There could be, however, a different causal link that connects culture to institutions. In Chapter 4, I include an observation that the Catholic and Lutheran societies of Central Europe produced cleaner party finance practices than the predominantly Eastern Orthodox and Muslim nations of the Balkans, Russia, and Ukraine. It could be that the former group introduced more stringent anticorruption measures, as Monica Dorhoi’s (2006) indices also show, because of more secular (less corruption tolerant) cultural values compared to the latter, more traditionally oriented group. If this proposition is correct, then culture might matter, but not as a direct determinant of the different levels of corruption observed in the postcommunist world. Rather, differences in the remaining precom-

244

Political Corruption in Eastern Europe

munist value orientations might have been reflected in some of the institutions that developed after 1989. The importance of the secular-traditional dimension of cultural specifics in Eastern Europe is also demonstrated by the different magnitudes of the impact of corruption perceptions on public trust. Despite the fact that secularism was officially imposed on all nations through the Marxist ideology, traditional values prioritizing family links proved more resilient among the Eastern Orthodox and Muslim group. That is why in exactly these countries, which are more tolerant of favoritism, the eroding effect of corruption on confidence in political institutions is much weaker.

Directions for Future Research Based on the empirical results of my study, we can begin to trace avenues worthy of future investigation. One such area is suggested by the weak results for the importance of electoral systems in promoting honest and fair government. Why does the distinctiveness of the electoral institutional setting not explain more conclusively the variation in corruption rates among the Eastern European countries? Recalling the main rationale behind this proposition, it is the nature of the voter-representative link that allows different opportunities for monitoring, exposure, and punishment of wrongdoing. Deeper exploration of the nature of this connection during transition would help to establish with greater certainty whether certain constituency types have been more permissive of corruption than others. I am inclined to think that the channels through which accountability usually operates are not all open yet in Eastern Europe. In this situation, the mechanism through which corruption is expected to be contained is either absent or seriously hindered. This state of affairs is most likely time sensitive; thus, paying attention to the dynamics of change in whether and how voters and their representatives interact will be especially fruitful. This line of future research will certainly draw on existing theoretical work and empirical studies, but it should not be misguided by conclusions about a strong correlation between corruption and election results in Eastern Europe. It is not only electoral accountability, narrowly defined as voters’ punishing or rewarding politicians in elections, that may matter. The electoral arrangement, proportional or

Conclusion

245

majoritarian, can make a difference because it determines how clearly citizens can see who is responsible for policies (Powell 2000, 253). Yet supporting evidence from Eastern Europe about the role played by the electoral rules is not as solid as one would wish. Voters have demonstrated readiness to punish incumbents, but whether politicians in power feel more threatened by the prospect of being exposed and sanctioned for corruption in majoritarian than in proportional races is a separate concern. Or, they may think they would be removed no matter what, assuming that voters automatically consider them corrupt. The answers to these questions and the reasons for them deserve to be studied in more detail should we aim at a better explanation of the anticorruption incentives generated by the electoral method. In addition, my analyses of the elections-corruption link will remain incomplete if we ignore the issue of electoral reform. The likely endogeneity makes such an inquiry even more interesting. Politicians might not be interested in changing the rules under which they were elected and adopting a new, presumably more restrictive and exposable system. In other words, because institutions draw the boundaries within which politicians behave, the latter would try to influence the design of rules in a favorable way. We know this from extant research on institutional reform—once established, institutions tend to be sticky. Among other things, convenience in breaching the ethical code of honest service makes politicians reluctant to change the rules that allow this. Institutional reform is yet possible if the fortunes of elites change or if a strong grassroots movement demands it. The Eastern European citizens have already started to blame the proportional closed list system (in which they cannot influence nominations and rankings) for the systemic corruption in their countries. In Bulgaria and Romania, and even in the less corrupt Slovenia, new election laws with some majoritarian elements were debated in parliament. The two former countries recently switched to a mixed system including single-member districts. Both the striving for institutional change generated by public disgust and the consequences of electoral system redesign need to be included in the future research agenda on corruption. The connection between corruption and the process of articulation and aggregation of societal interests is a topic that remains understudied. Samuel P. Huntington (1968, 71) writes that “corruption is more prevalent in states which lack effective political parties, in societies where the interests of the individual, the family, the

246

Political Corruption in Eastern Europe

clique, or the clan predominate.” He also notes that, where public resources are spent on party building, the occurrence of corruption is lower than in countries with weak party organizations. Applied to the Eastern European context, these arguments raise concerns. A consensus seems to be emerging that the nascent party systems in the region have not yet fully replaced the importance of connections and networking. To strengthen the organizational capacity of parties, some countries have opted for state funding in order to reduce the dependence of politics on private interests. The results of my analyses in Chapter 4 offer some empirical confirmation of the effectiveness of this instrument, despite it being criticized as a source of cartelization. It is important to continue this line of research by exploring the mechanisms through which party institutionalization relates to corruption. This can be done in carefully designed analyses of the social roots of the Eastern European parties, including the former communist parties, and the ways they connect to their supporters and other groups. Research on this has already begun, but mainly with regard to intraparty democracy. I propose that we now focus closer on the implications of the link between parties and society as well as the transparency in their interactions. A significantly less studied theme, relevant to postcommunist countries, is the possible connection between corruption and populism. In some of these states (Bulgaria, Russia, and Slovakia, to name a few) new parties and movements with no clear policy stands, but with strong anticorruption appeals, won big in parliamentary elections and even formed governments. It is not surprising that they gained representation, given the widespread citizen disappointment with corrupt politicians. There are clear signs that corruption can help newcomers using populist slogans; with relatively few resources, these groups can penetrate the legislative and the executive branch. While the reasons for the rise of populism are more certain and can be established through survey and other data, there is an urgent need for theoretical and empirical work on the integrity of government emerging as a result of populism. Such studies will yield considerable knowledge about a complex and dynamic phenomenon, elements of which have already been witnessed in Eastern Europe. Only then will we be able to assess the actual potential of charismatic nonprogrammatic leadership to fulfill its main promise: to save the public from the menace of corruption.

Conclusion

247

A more general area that may benefit from conceptual and empirical improvements in the research agenda on corruption is democratic quality. The recent explosion of interest in the study of democratic performance was itself inspired by flaws observed in the way the new 1989 democracies function, including broadly recognized dishonesty among politicians (Roberts 2010, 187). Many shortcomings overshadowing the governance-by-the-people principle, such as party funding hidden from the public and broken campaign promises, present topics for further investigation. Clarifying the mechanisms through which civic and political participation is harmed by allpervasive corruption would cast light on deficiencies in the transitional environment and how these can be minimized. Especially troubling in this regard is the phenomenon of vote buying, with the flawed justification that it increases voter participation. But by nature vote selling is a manifestation of “the corruption of the poor” (Huntington 1968, 61) and it raises concerns, not less serious than those evoked by bribed politicians, about the health of any young democracy. Further investigation of these new forms, magnitudes, and dynamics of corruption will offer insights into the success of the postcommunist transitions. Will they lead to outcomes distinct from those in today’s consolidated democracies? The intense expectation of answers to the question of government responsiveness confirms the worthiness of studying democratic quality in Eastern Europe through the prism of corruption.

Notes 1. A slightly different interpretation of the role of the EU is proposed by McManus-Chubinska et al. (2004, 107). They maintain that the EU has simultaneously helped reduce corruption and affected how people perceive malfeasance. By encouraging a “culture that can resist corruption,” the EU has boosted perceptions and allegations of public office abuse. 2. At a press conference in Sofia, responding to an Austrian journalist’s comment that Bulgaria is the most corrupt country in Europe, Bulgarian prime minister Boyko Borissov asked with irritation: “Who determines which one is the most corrupt state in the EU, by what criterion? If we choose so, our media could also ask many questions related to Austrian firms and others, and problems would pop up” (Lalov 2011).

Appendix A Description of Regression Analysis Variables and Sources

I. Macrolevel Regression Analysis Corruption TICPI (Chapters 3, 4, 6, 7): Transparency International Corruption Perceptions Index; lower scores indicate more corruption. Source: www.transparency.org, accessed 20 May 2010. WBI (Chapters 3, 4, 6, 7): World Bank Control of Corruption Index; lower scores indicate more corruption. Source: http://info.world bank.org/governance/wgi/index.asp, accessed 20 May 2010. NIT (Chapter 7): Nations in Transit Corruption Index; higher scores indicate more corruption. Source: http://freedomhouse.org /template.cfm?page=17. Proportional Rules PR share (Chapter 3): Amount of proportional representation (percentage PR) seats in national assemblies. Sources: Birch (2003); Interparliamentary Union, www.ipu.org. District magnitude (Chapter 3): Based on representatives per district. Measurement follows Persson’s (2002) formula: 1 – (total districts)/(total seats). Higher values reflect larger average district magnitudes. Sources: Birch (2003); Interparliamentary Union, www.ipu.org. Regulatory burden (Chapter 3): Business ability to operate as affected by excessive government regulation. Ranges from 0 (most bur-

249

250

Appendix A

dened) to 100 (least burdened). Source: Business Freedom Index of the Heritage Foundation, www.heritage.org/research /features/index. Trade openness (Chapter 3): The amount of foreign trade, calculated as percentage of imports and exports from the country’s gross domestic product (GDP). Source: Annual Totals Trade, International Merchandise Trade Statistics, Common Data Set, UN Statistics Division, http://unstats.un.org/unsd/trade/imts /annual totals.htm. EU association (Chapter 3): Association Agreement status with the European Union (EU), coded as 1 for countries with association agreements signed by 1995, and 0 for the rest. PHARE (Poland and Hungary: Assistance for Restructuring Their Economies) (Chapter 3): Amount of financial help from the EU, measured in euros per capita. Source: Progress Reports of Candidate Countries, http://ec.europa.eu/enlargement /how-does-it-work/progress_reports/index_en.htm. Economic capacity (Chapters 3, 6, 7): Per capita gross domestic product, measured in thousands of current US dollars (Chapter 3) or 2005 international US dollars (Chapters 6, 7). Source: World Bank, World Development Indicators, http://data.worldbank .org/indicator. Democratization (Chapters 3, 6): Political Freedoms and Civil Liberties, Freedom House country yearly rankings. Source: Freedom House, www.freedomhouse.org. Wages in the public sector (Chapter 6): Measured as thousands of US dollars spent on public employee compensation weighted by the size of the administration. Sources: Government Finance Statistics Yearbook, various years; Grzymala-Busse (2007, Table B5). Reform of civil service (Chapter 6): Takes the value of 1 if civil service reform had been introduced, and 0 in the absence of reform. Source: Country reports of the United Nations Public Administration Network (UNPAN), www.unpan.org. Telephones (Chapter 6): Telephone lines per 100 people. Source: World Bank, World Development Indicators, http://data .worldbank.org/indicator. NGOs (Chapter 6): Sustainability of nongovernmental organizations (NGOs). Source: USAID Nongovernmental Organization Sustainability Index, www.usaid.gov/locations/europe_eurasia /dem_gov/ngoindex.

Appendix A

251

Secular culture (Chapter 6): Secular-Traditional Values Index. Source: Inglehart-Welzel Cultural Map of the World, www.world valuessurvey.org/wvs/articles/folder_published/article_base_54. Fragmentation (Chapter 6): Index of ethnic group fragmentation, calculated by using the formula 1/gi, where g is the proportion of group i from the total population. Source: Author’s calculations using data on countries’ ethnic composition from Central Intelligence Agency, World Factbook, www.cia.gov/library /publications/the-world-factbook. Economic liberalization (Chapter 7): Calculated as percentage of imports from the country’s GDP. Source: Annual Totals Trade, International Merchandise Trade Statistics, Common Data Set, UN Statistics Division, http://unstats.un.org/unsd/trade/imts /annual totals.htm. Integration status (Chapter 7): Progress in the integration of a country with the EU. A 5-category variable: 4 = EU membership, 3 = candidate status, 2 = submitted application for accession, 1 = with Association Agreement, and 0 = no integration. Source: Coded by author from European Commission of the European Union, http://ec.europa.eu/enlargement/index_en.htm.

II. Microlevel Regression Analysis Corruption (Chapter 8): Spread of corruption in country: “How widespread do you think bribe taking and corruption is in this country?” Respondents were asked to choose from: Almost no public officials engage in it, A few are, Most are, and Almost all public officials are engaged in it. Source: World Values Survey, Question E196, www.worldvaluessurvey.org. Confidence in institutions (Chapter 8): Confidence in parliament, national government, political parties, police, the civil service, and the justice system: “I am going to name a number of organizations. For each one, could you tell me how much confidence you have in them: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none at all?” Source: World Values Survey, Questions E074, E075, E076, E079, E080, and E085, www.worldvaluessurvey.org. Financial situation (Chapter 8): Individual evaluation of financial situation: “How satisfied are you with the financial situation of

252

Appendix A

your household?” Respondents were asked to rank their level of satisfaction on a scale of 1 (dissatisfied) to 10 (completely satisfied). Source: World Values Survey, Question C006, www.world valuessurvey.org. Interest in politics (Chapter 8): Individual interest in politics: “How interested would you say you are in politics?” Responses are coded on a scale of 1 (not at all interested) to 4 (very interested). Source: World Values Survey, Question E023, www.world valuessurvey.org. Ideology (Chapter 8): Self-reported information on a 10-point, leftright scale of 1 (left) to 10 (right). Source: World Values Survey, Question E033, www.worldvaluessurvey.org. Education (Chapter 8): Level of education from 1 (some elementary education) to 8 (university with degree/higher education). Source: World Values Survey, Question X025, www.world valuessurvey.org. Social class (Chapter 8): Self-described social level, including 1 (lower), 2 (working), 3 (lower middle), 4 (upper middle), and 5 (upper). Source: World Values Survey, Question X045, www.worldvaluessurvey.org. Gender (Chapter 8): A dichotomous variable for gender, where 1 is assigned to males, and 0 to females; thus, the anticipated relationship is negative. Source: World Values Survey, Question X001, www.worldvaluessurvey.org. Age (Chapter 8): Respondent’s age in years. Source: World Values Survey, Question X003, www.worldvaluessurvey.org. Support for democracy (Chapter 8): Degree of agreement with statement: “Democracy may have problems but it’s better than any other form of government.” Answers vary from “strongly disagree,” “disagree,” “agree,” to “agree strongly.” Source: World Values Survey, Question E123, www.worldvalues survey.org. Corruption (Chapter 9): The degree to which corruption, such as bribe taking among politicians, is widespread in a country. A 4scale categorical variable, including 1 (it hardly happens at all), 2 (not very well spread), 3 (quite widespread), and 4 (very widespread). Source: The Comparative Study of Electoral Systems (2007), Question B3044. Voting (Chapter 9): The choice to vote or abstain. Dummy variable, scored as 1 when respondent indicated they had cast a ballot in

Appendix A

253

the current election. Source: World Values Survey, Question B3004_1, www.worldvaluessurvey.org. Political efficacy (Chapter 9): A belief that individuals can influence policymaking. Dummy variable, coded as 1 for individuals who believe that the candidate for whom people vote makes a difference. Source: World Values Survey, Question B3014, www .worldvaluessurvey.org. Age (Chapter 9): Voter’s age in years. Source: World Values Survey, Question B2001, www.worldvaluessurvey.org. Gender (Chapter 9): Dummy variable, coded as 1 for males. Source: World Values Survey, Question B2002, www.worldvalues survey.org. School (Chapter 9): Education on an 8-category scale, including 1 (none), 2 (incomplete primary), 3 (primary completed), 4 (incomplete secondary), 5 (secondary completed), 6 (post-secondary trade/vocational school), 7 (university incomplete), and 8 (university completed). Source: World Values Survey, Question B2003, www.worldvaluessurvey.org. Wealth (Chapter 9): Household income in 5 categories, from lowest to highest quintile, as appropriate to the respondent. Source: World Values Survey, Question B2020, www.worldvalues survey.org. Residence (Chapter 9): Dummy variable, scored as 1 for individuals residing in large cities and towns or in suburbs of large cities, and 0 for individuals residing in small towns or in rural area/village. Source: World Values Survey, Question B2030, www.world valuessurvey.org. Party ID (Chapter 9): Dummy variable for party identification. Coded as 1 for those feeling close to a particular political party. Source: World Values Survey, Question B3028, www.world valuessurvey.org. District magnitude (Chapter 9): Average district magnitude, logged. Sources: Birch (2003); author’s calculations. Competition (Chapter 9): Electoral competition is measured as the difference in vote shares between the first- and second-place parties. Source: Parties and Elections 2009, www.parties-and-elections.de. Multipartyism (Chapter 9): Laakso and Taagepera’s (1979) effective number of parties in lower house or unicameral legislature; author’s calculations. Source: Parties and Elections 2009, www.parties-and-elections.de.

254

Appendix A

Concurrent (Chapter 9): Dummy variable, scored as 1 when presidential and legislative elections were held on the same date. Source: Interparliamentary Union, www.ipu.org. Democratization (Chapter 9): A composite index representing the level of democracy, coded by summing up Freedom House’s Political Freedoms and Civil Liberties Indexes, country yearly rankings. Low values correspond to high freedom and high values to low freedom. Source: Freedom House, www.freedom house.org/ratings.

Appendix B Eastern European Participation in International Anticorruption Conventions

Country Albania Bulgaria Croatia Czech Rep. Estonia Hungary Latvia Lithuania Macedonia Poland Romania Russia Serbia Slovakia Slovenia Ukraine

OECDa (12.22.1998) (F) (01.21.2000) (11.23.2004) (A) (12.04.1998) (F)

(09.08.2000)

(09.24.2001) (09.06.2001) (A)

Council of Europeb

United Nationsc

(04.27.2001) (05.01.1999) (F) (12.02.2000) (02.09.2002) (05.01.1999) (F) (07.09.1999) (F) (07.27.2000) (05.01.1999) (F) (10.07.2000) (05.20.1999) (F) (05.01.1999) (F) (02.01.2007) (04.01.2003) (05.01.1999) (F) (05.01.1999) (F) (01.01.2006)

(12.18.2003) (F) (12.10.2003) (F) (12.10.2003) (F) (04.22.2005) (04.12.2010) (A) (12.10.2003) (F) (05.19.2005) (12.10.2003) (F) (08.18.2005) (12.10.2003) (F) (12.09.2003) (F) (12.09.2003) (F) (12.11.2003) (F) (12.09.2003) (F) (04.01.2008) (A) (12.11.2003) (F)

Notes: Entries are dates of joining; (F) denotes founding party; (A) denotes accession. a. Convention on Combating Bribery of Foreign Public Officials in International Business Transactions, initiated by the Organisation for Economic Co-operation and Development (OECD) in 1998, www.oecd.org/document/21 /0,3343,en_2649_34859_2017813_1_1_1_1,00.html. b. Group of States Against Corruption (GRECO), initiated by the Council of Europe in 1999, www.coe.int/t/dghl/monitoring/greco/default_en.asp. c. United Nations Convention Against Corruption, adopted in 2003, www.unodc.org/unodc/en/treaties/CAC/signatories.html.

255

Appendix C Supporting Documentation for Empirical Tests

Full Results for Table 8.2 Variable Corruption

Financial situation Ideology

Interest in politics Education

Social class

Gender

Age

Parliament –.18*** (.01) –.18 .027*** (.00) .08 .02*** (.00) .05 .09*** (.01) .09 –.01** (.00) –.03 –.01 (.01) –.01 –.02 (.02) –.02 .00 (.00) .02

Government

Parties

Police

–.17*** (.01) –.17 .04*** (.00) .09 .01*** (.00) .04 .05*** (.01) .05 –.02*** (.00) –.04 .01 (.01) .01 –.04** (.02) –.02 .00*** (.00) .03

–.12*** (.01) –.13 .03*** (.00) .08 .02*** (.00) .04 .16*** (.01) .18 –.02*** (.00) –.06 .00 (.01) .00 –.04*** (.01) –.03 .00 (.00) .00

–.17*** (.01) –.16 .02*** (.00) .06 .00 (.00) .01 .02 (.01) .02 –.20*** (.00) –.05 –.02 (.01) –.02 –.06*** (.02) –.03 .00*** (.00) .04

Civil Service –.16*** (.01) –.17 .02*** (.00) .07 .00 (.00) .00 .04*** (.01) .04 –.01*** (.00) –.03 .00 (.01) .00 –.06*** (.02) –.04 .00 (.00) .01

Justice System –.18*** (.01) –.17 .03*** (.00) .08 .00 (.00) .00 .01 (.01) .01 –.02*** (.00) –.06 .00 (.01) .00 –.04** (.02) –.02 .00 (.00) .02 (continues)

257

258

Appendix C

Full Results for Table 8.2 continued Albania

Bulgaria

Czech Rep.

Estonia

Hungary

Latvia

Lithuania

Macedonia

Poland

Romania

Russia

Slovakia

Slovenia

Constant R2 N

.29*** (.04) .09 .19*** (.04) .06 –.34*** (.03) –.12 .10*** (.07) .03 .07 (.04) .02 –.19*** (.03) –.07 .05 (.04) .01 –.29*** (.04) –.08 –.01 (.04) –.00 –.28*** (.04) –.09 –.21*** (.03) –.08 –.10*** (.03) –.03 –.25*** (.03) –.07 2.34*** (.06) .110 10,460

–.02 (.04) –.01 .33*** (.04) .09 –.29*** (.04) –.10 .08** (.04) .03 –.04 (.04) –.01 –.10*** (.04) –.03 .07 (.04) .02 –.38*** (.04) –.10 –.08* (.04) –.02 –.41*** (.04) –.13 –.33*** (.03) –.12 –.03 (.04) –.01 –.13*** (.04) –.04 2.48*** (.06) .105 10,558

.08** (.04) .03 .16*** (.04) .05 –.12*** (.03) –.05 .04 (.03) .01 –.04 (.04) –.01 –.22*** (.03) –.08 .15*** (.04) .05 –.09** (.04) –.03 –.11*** (.04) –.04 –.14*** (.03) –.05 .03 (.03) .01 .07** (.03) .03 –.19*** (.04) –.06 1.81*** (.06) .082 10,362

.66*** (.04) .20 .30*** (.04) .08 .17*** (.04) .06 .24*** (.04) .07 .36*** (.04) .09 –.03 (.04) –.01 –.08** (.04) –.01 –.03 (.04) –.01 .32*** (.04) .09 .06 (.04) .02 –.06 (.03) –.02 .13*** (.04) .04 .17*** (.04) .05 2.60*** (.06) .116 10,649

–.39*** (.04) –.12 .11*** (.04) .04 –.14*** (.03) –.05 .26*** (.04) .09 .18*** (.04) .05 .01 (.03) .01 .08** (.04) .02 –.31*** (.04) –.09 –.14*** (.04) –.05 –.26*** (.03) –.09 .11*** (.03) .04 –.06 (.03) –.02 –.27*** (.04) –.08 2.72*** (.06) .082 10,172

.18*** (.04) .06 –.17*** (.04) –.05 –.33*** (.04) –.11 .25*** (.04) .08 .10** (.04) .02 –.09*** (.04) –.03 –.23*** (.04) –.07 –.22*** (.04) –.06 .14*** (.04) .04 .01 (.04) .00 –.06 (.04) –.02 –.04 (.04) –.02 –.23*** (.04) –.07 2.84*** (.06) .089 10,626 (continues)

Appendix C

259

Full Results for Table 8.2 continued Sources: European and World Values Surveys for Wave Integrated Data File, 1981–2004, v. 20060423. The European Values Study Foundation, www .europeanvalues.nl, and World Values Survey Association, www.worldvalues survey.org. Notes: Dependent variables: Degree of confidence in parliament, government, parties, police, civil service, and the justice system. Cells by row contain OLS unstandardized coefficients, standard errors in parentheses, and standardized beta coefficients; ** p < .05; *** p < .01.

260

Appendix C

Full Results for Table 8.3 Variable Corruption

Financial situation Ideology

Interest in politics Education

Social class

Gender

Age

Albania

Bulgaria

Czech Rep.

Estonia

Hungary

Latvia

Confidence in Institutions

Support for Democracy

–.99*** (.04) –.23 .17*** (.02) .11 .06*** (.02) .04 .39*** (.04) .10 –.10*** (.02) –.06 –.01 (.05) .00 –.28*** (.07) –.04 .01*** (.00) .03 .75*** (.18) .05 1.03*** (.17) .07 –.98*** (.15) –.08 1.01*** (.16) .08 .65*** (.18) .04 –.59*** (.15) –.05

–.05*** (.01) –.05 .01*** (.00) .04 .04*** (.00) .11 .04*** (.01) .05 .03*** (.00) .09 .01 (.01) .01 .00 (.01) .00 .00** (.00) .02 .73*** (.04) .26 .22*** (.04) .07 .25*** (.03) .10 .36*** (.03) .13 .33*** (.04) .10 .06 (.03) .02 (continues)

Appendix C

261

Full Results for Table 8.3 continued Lithuania

Macedonia

Poland

Romania

Russia

Slovakia

Slovenia

Constant R2 N

.06 (.17) .00 –1.20*** (.17) –.08 .27 (.17) .02 –1.05*** (.16) –.08 –.45*** (.15) –.04 .05 (.15) .00 –.82*** (.17) –.06 14.68*** (.27) .133 9,532

.17*** (.04) .05 .01 (.04) .00 .24*** (.04) .08 .48*** (.03) .17 –.22*** (.03) –.08 .26*** (.03) .10 .16*** (.04) .05 2.49*** (.06) .130 9,838

Sources: European and World Values Surveys for Wave Integrated Data File, 1981–2004, v. 20060423. The European Values Study Foundation, www .europeanvalues.nl, and World Values Survey Association, www.worldvalues survey.org. Notes: Dependent variables: Degree of confidence in political institutions (index combining confidence measures for all six institutions in Table 8.1), and degree of agreement with the statement “Democracy may have problems but it’s better than any other form of government.” Cells by row contain OLS unstandardized coefficients, standard errors in parentheses, and standardized beta coefficients; ** p < .05; *** p < .01.

262

Appendix C

Full Results for Table 9.2 Voting Predictor

SE

[95%

CI]

B

SE

.079

.062

[–.043

.201]

–.272

.131

[–.529 –.016]

.999 .062 .023 .147 .130 –.026 1.237

.113 .066 .004 .047 .048 .108 .242

[.778 1.219] [–.066 .191] [.016 .030] [.056 .238] [.037 .223] [–.239 .186] [.763 1.710]

— –.003 .005 .074 .108 –.001 —

.067 .003 .035 .038 .111 —

[–.135 [–.001 [.004 [.034 [–.220

–.538 –.025 .149 –.281 —

.178 .014 .015 .260 —

[–.887 [–.052 [.120 [–.790

— — — — –.150

— — — — .052

[–.252 –.048]

–1.631 –2544.493 .162 5,889

.930

[–3.454

.192] 2.199 –3891.147 .028 8,140

.845

[.542 3.856]

Corruption Political efficacy Gender Age School Wealth Residence Party ID District magnitude Competition Multipartyism Concurrent Democratization Constant –2LL Pseudo R2 N

Political Efficacy

B

–.189] .002] .177] .229]

[95%

CI]

.128] .011] .143] .181] .217]

Note: Standard errors are robust, clustered by country.

Standardization of Logit Coefficients from Table 9.2 Predictor

Standard Deviation

Political efficacy CorruptionVoting CorruptionEfficacy

.389868 .6634899 .6634899

Unstandardized Coefficient .9985509 .0792734 –.2724767

Note: The formula in use is B = (b)(sdX)/(π/√3).

Standardized Coefficient .214610276 .028995093 –.099661266

Bibliography

Achen, Christopher H. 2005. “Two-Step Hierarchical Estimation: Beyond Regression Analysis.” Political Analysis 13 (April): 447–456. Ades, Alberto, and Rafael Di Tella. 1997. “The New Economics of Corruption: A Survey and Some New Results.” Political Studies 45, no. 3 (special issue): 496–515. Ades, Alberto, and Rafael Di Tella. 1999. “Rents, Competition, and Corruption.” American Economic Review 89, no. 4 (September): 982–993. Adsera, Alicia, Carles Boix, and Mark Payne. 2000. “Are You Being Served? Political Accountability and Quality of Government.” Inter-American Development Bank Research Department Working Paper No. 438. Washington, DC: Inter-American Development Bank. Alesina, Alberto, and Romain Wacziarg. 2000. “The Economics of Civic Trust.” In Susan J. Pharr and Robert D. Putnam, eds., Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton: Princeton University Press, 149–170. Amorim Neto, Octavio, and Gary W. Cox. 1997. “Electoral Institutions, Cleavage Structures, and the Number of Parties.” American Journal of Political Science 41, no. 1 (January): 149–174. Anderson, Christopher J., and Yuliya V. Tverdova. 2003. “Corruption, Political Allegiances, and Attitudes Towards Government in Contemporary Democracies.” American Journal of Political Science 47, no. 1 (January): 91–109. 263

264

Bibliography

Anderson, James H., and Cheryl W. Gray. 2006. Anticorruption in Transition 3: Who Is Succeeding . . . and Why? Washington, DC: World Bank. Aneva, Nadelina. 2007. “Kandidatite za Post Daryavali DPS Avansovo” (The Candidates for Positions Donated to the MRF in Advance). Sega, 28 May. www.segabg.com/online/new/articlenew.asp?issu id=2655& sectionid=2&id=0000201, accessed 14 July 2009. Annett, Anthony. 2002. “Politics, Government Size, and Fiscal Adjustment in Industrial Countries.” IMF Working Paper No. 162. Washington, DC: International Monetary Fund. Anticorruption in Transition: A Contribution to the Policy Debate. 2000. Washington, DC: World Bank. Appel, Hilary. 2001. “Corruption and the Collapse of the Czech Transition Miracle.” East European Politics and Societies 15, no. 3: 528–553. Arikan, G. Gulsun. 2008. “How Privatizations Affect the Level of Perceived Corruption.” Public Finance Review 36, no. 6 (November): 706–727. Attitudes of Europeans Towards Corruption. 2008. Special Eurobarometer no. 291, April. Brussels: European Commission. http://ec.europa.eu/public_opinion/archives/ebs/ebs_291_sum_en. pdf. Avramov, Kiril Alioshev. 2004. “The Bulgarian ‘Nationally Responsible Capital’ on the Eve of EU Integration.” In Gyorgy Lengyel, ed., Cohesion and Division of Economic Elites in Central and Eastern Europe. Budapest: Corvinus University of Budapest. www.bke.hu/szoc/doc/lengyel/lengyel_cohesion_division .pdf, accessed 30 July 2009. Baker, Peter, and Susan Glasser. 2005. Kremlin Rising: Vladimir Putin’s Russia and the End of Revolution. New York: A Lisa Drew Book/Scribner. Banfield, Edward. 1958. The Moral Basis of a Backward Society. Chicago: Free Press. Bardhan, Pranab. 1997. “Corruption and Development: A Review of Issues.” Journal of Economic Literature 35, no. 3 (September): 1320–1346. Barnes, Andrew. 2007. “Extricating the State: The Move to Competitive Capture in Post-Communist Bulgaria.” Europe-Asia Studies 59, no. 1 (January): 71–95.

Bibliography

265

Beck, Nathaniel, and Jonathan N. Katz. 1995. “What to Do (and Not to Do) with Time-Series Cross-Section Data in Comparative Politics.” American Political Science Review 89 (September): 634–647. Beck, Nathaniel, and Jonathan N. Katz. 1996. “Nuisance vs. Substance: Specifying and Estimating Time-Series-Cross-Section Models.” Political Analysis 6: 1–36. Birch, Sarah. 1997. “Nomenklatura Democratization: Electoral Clientelism in Post-Soviet Ukraine.” Democratization 4, no. 4 (Winter): 40–62. Birch, Sarah. 2003. Electoral Systems and Political Transformation in Post-Communist Europe. London: Palgrave Macmillan. Birch, Sarah, Frances Millard, Kieran Williams, and Marina Popescu. 2002. Embodying Democracy: Electoral System Design in PostCommunist Europe. London: Palgrave Macmillan. Blais, Andre. 2000. To Vote or Not to Vote: The Merits and Limits of Rational Choice Theory. Pittsburgh, PA: University of Pittsburgh Press. Blais, Andre. 2006. “What Affects Voter Turnout?” Annual Review of Political Science 9: 111–125. Boix, Carles. 2003. Democracy and Redistribution. Cambridge: Cambridge University Press. Bushati, Ditmir. 2008. Country Report: Albania. In Nations in Transit. Washington, DC: Freedom House. www.freedomhouse.org, accessed 16 January 2010. Business Environment and Enterprise Performance Survey. 1999. Washington, DC: World Bank. www.ebrd.com/pages/research/ analysis /surveys/beeps.shtml. Carlson, Scott, and Wendy Betts. 2001. Albania. In Nations in Transit. Washington, DC: Freedom House. www.freedomhouse .org, accessed 15 January 2010. Carvajal, Doreen, and Stephen Castle. 2008. “Mob Muscles Its Way into Politics in Bulgaria.” New York Times, 15 October, p. A1. Catterberg, Gabriela, and Alejandro Moreno. 2006. “The Individual Bases of Political Trust: Trends in New and Established Democracies.” International Journal of Public Opinion Research 18, no. 1 (Spring): 31–48. Central Intelligence Agency. Various years. World Factbook. www.cia.gov/library/publications/the-world-factbook/index.html, accessed 12 December 2009.

266

Bibliography

Chang, Eric C. C., and Miriam A. Golden. 2006. “Electoral Systems, District Magnitude and Corruption.” British Journal of Political Science 37, no. 1 (January): 115–137. “Charlemagne: Hungary’s Other Deficit.” 2011, 8 January. The Economist, p. 53. Coalition 2000. 2005. Antikorupzionnite Reformi v Bulgaria (The Anti-Corruption Reforms in Bulgaria). Sofia: Center for the Study of Democracy. www.csd.bg/files /bg_full.pdf, accessed 17 May 2010. Colazingari, Silvia, and Susan Rose-Ackerman. 1998. “Corruption in Paternalistic Democracy: Lessons from Italy for Latin America.” Political Science Quarterly 113, no. 3: 447–470. Comparative Study of Electoral Systems. 2007. CSES Module 2 Full Release [data set], 27 June 2007 version. www.cses.org. Corruption and Anti-Corruption Policy in Estonia. 2002. Open Society Institute. http://resources.transparency-bg.org/download.html?id=591, accessed 4 October 2010. Corruption Indexes: Regional Monitoring of Corruption in Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Macedonia, Romania, and Yugoslavia. Southeast European Legal Development Initiative. 2002, April. www.seldi.net/SELDI_fin _e2.PDF, accessed 19 February 2010. Coulloudon, Virginie. 2002. “Russia’s Distorted Anticorruption Campaigns.” In Stephen Kotkin and András Sajó, eds., Political Corruption in Transition. Budapest: Central European University Press, 187–205. Della Porta, Donatella, and Susan Rose-Ackerman. 2002. “Introduction.” In Corrupt Exchanges: Empirical Themes in the Politics and Political Economy of Corruption. Baden-Baden: Nomos Verlagsgesellschaft 8-19. Della Porta, Donatella, and Alberto Vannucci. 1999. Corrupt Exchanges: Actors, Resources, and Mechanisms of Political Corruption. New York: Aldine de Gruyter. Della Porta, Donatella, and Alberto Vannucci. 2005. “Corruption as a Normative System.” Paper presented at the International Conference in Political Life and the Quality of Democracy: A Comparative Perspective Europe–Latin America. CIES-ISCTE, May 2005. http://home.iscte-iul .pt/~ansmd/CC-DellaPorta.pdf.

Bibliography

267

Denhardt, Robert B. 2002. “Trust as Capacity: The Role of Integrity and Responsiveness.” Public Organization Review 2: 65–76. Diagnostic Surveys of Corruption in Romania. 2001. Washington, DC: World Bank. http://siteresources.worldbank.org/INTWBIGOVANT COR/Resources/RomaniaSurveyReport.pdf, accessed 11 June 2010. Dimitrov, Filip. 2003. Mitovete na Bulgarskiya Prehod (The Myths of the Bulgarian Transition). Sofia: Ciela. Dimitrova, Aleksenia, and Ognian Stefanov. 2004. “Partii, Fondazii, Pari, Vlast” (Parties, Foundations, Money, Power). Media Development Center. www.mediacenterbg.org/Library/Catalogues /Default.aspx?evntid=WM%2bVOf%2fK9Zk%3d, accessed 8 December 2010. Dimitrova, Antoaneta. 2002. “Enlargement, Institution-Building and the EU’s Administrative Capacity Requirement.” West European Politics 25, no. 4 (October): 171–190. Dimitrova, Yuliana. 1995. “Poshtenska Banka Mozhe da Stane Orionbank” (Mail Bank May Become Orion Bank). Demokraziya, no. 290 (December). Dimitrova-Grajzl, Valentina P., Eszter Simon, and Alex Fischer. 2010. “Political Efficacy of Emerging Elites in Post-Socialist Countries: The Impact of Disciplinary Culture and Political Opportunities.” Transition Studies Review 14, no. 4: 807–821. Dininio, Phyllis, and Robert Orttung. 2005. “Explaining Patterns of Corruption in the Russian Regions.” World Politics 57 (July): 500–529. “Dogan Explained About the Ring.” 2006. Information Agency Focus Online, 2 April. www.focus-news.net/?id=h1020. Dionisie, Dan, and Francesco Checcki. 2008. “Corruption and AntiCorruption Agencies in Eastern Europe and the CIS: A Practitioners’ Experience.” Paper presented at the Research Network of Anti-Corruption Agencies (ANCORAGE-NET) workshop. http://ancorage-net.org/content/documents/dionisie -checchi-corruption_in_ee.pdf, accessed 5 September 2010. Dorhoi, Monica. 2005. “Anti-Corruption Strategies and Fighting Corruption in Central and Eastern Europe.” PhD dissertation, Michigan State University. Doric, Petar. 2007. Country Report: Croatia. In Nations in Transit. Washington, DC: Freedom House. www.freedomhouse.org, accessed 16 January 2010.

268

Bibliography

Downs, George W., David M. Rocke, and Peter N. Barsoom. 1996. “Is the Good News About Compliance Good News About Cooperation?” International Organization 50 (Summer): 379–406. Duch, Raymond M., and Harvey D. Palmer. 2002. “Strategic Voting in Post-Communist Democracy?” British Journal of Political Science 32: 63–91. Duch, Raymond M., and Randy Stevenson. 2005. “Context and the Economic Vote: A Multilevel Analysis.” Political Analysis 13 (April): 387–409. Dugan, Christopher F., and Vladimir Lechtman. 1998. “The FCPA in Russia and Other Former Communist Countries.” American Journal of International Law 91, no. 2 (April): 378–388. Earle, Beverley. 2000. “Bribery and Corruption in Eastern Europe, the Baltic States, and the Commonwealth of Independent States: What Is to Be Done?” Cornell International Law Journal 33, no. 1 (June): 483–513. Eilat, Yair, and Clifford Zinnes. 2002. “The Shadow Economy in Transition Countries: Friend or Foe? A Policy Perspective.” World Development 30, no. 7: 1233–1254. Emerson, Michael. 2008. “Recalibrating EU Policy Towards the Western Balkans.” Centre for European Policy Studies Policy Brief No. 175, October. http://ssrn.com/abstract=1333499, accessed 20 February 2010. Epstein, Rachel A., and Ulrich Sedelmeier. 2008. “Beyond Conditionality: International Institutions in Post-Communist Europe After Enlargement.” Journal of European Public Policy 15, no. 6 (September): 795–805. EU Monitoring and Advocacy Program. Corruption and AntiCorruption Policy. www.eumap.org/topics/corruption, accessed 22 December 2009. Eurobarometer 69: Public Opinion in the European Union, Spring 2008. 2008. http://ec.europa.eu/public_opinion/archives/eb/eb69/eb69 _en.htm, accessed 14 September 2009. Eurobarometer Survey. Various years. Brussels: European Commission. http://ec.europa.eu/public_opinion/index_en.htm. European Marketing Data and Statistics. 1997–2007. London: Euromonitor International. Evans, Geoffrey, and Stephen Whitefield. 1995. “The Politics and Economics of Democratic Commitment: Support for Democracy in

Bibliography

269

Transition Societies.” British Journal of Political Science 25: 485–514. Fatic, Aleksandar. 2001. “Corruption and Anti-Corruption Policy in Serbia.” GRECO Paper. Belgrade: Transparency International Serbia. www.transparentnost.org.yu/publikacije/corr_and_anti-corr _pol _in_s.pdf, accessed 17 February 2010. Fisman, Raymond, and Roberta Gatti. 2002. “Decentralization and Corruption: Evidence Across Countries.” Journal of Public Economics 83, no. 3 (March): 325–345. Fisman, Raymond, and Edward Miguel. 2006. “Cultures of Corruption: Evidence from Diplomatic Parking Tickets.” NBER Working Paper No. 12312. www.usc.edu/schools/business/FBE/seminars/papers /AE_4-28-06_FISMAN-parking.pdf, accessed 13 January 2010. Franklin, Mark N. 1996. “Electoral Participation.” In Lawrence LeDuc, Richard G. Niemi, and Pippa Norris, eds., Comparative Democracies: Elections and Voting in Global Perspective. Thousand Oaks, CA: Sage, 214–233. Franklin, Mark N. 2004. Voter Turnout and the Dynamics of Electoral Competition in Established Democracies Since 1945. Cambridge: Cambridge University Press. Freedom House. 2003–2009. Nations in Transit. Washington, DC: Freedom House. www.freedomhouse.org, accessed 16 January 2010. Freedom in the World. Various years. Washington, DC: Freedom House. www.freedomhouse.org, accessed 18 December 2009. Gabel, Matthew J. 1995. “The Political Consequences of Electoral Laws in the 1990 Hungarian Elections.” Comparative Politics 27, no. 2 (January): 205–214. Galabov, Antonii. 2003. Transparency Index for Political Party Finance in Bulgaria. In Political Party and Election Campaign Financing in Southeastern Europe: Avoiding Corruption and Strengthening Financial Control. Sofia: Transparency International Bulgaria, 87–101. www.transparency-bg.org, accessed 5 October 2009. Galtung, Fredrik. 2006. “Measuring the Immeasurable: Boundaries and Functions of (Macro)Corruption Indices.” In Charles Samford, Arthur Shamlock, Carmel Connors, and Fredrik Galtung, eds., Measuring Corruption. Aldershot: Ashgate, 101–130. Ganev, Venelin I. 2006. “Ballots, Bribes, and State Building in Bulgaria.” Journal of Democracy 17, no. 1 (January): 75–89.

270

Bibliography

Ganev, Venelin I. 2007. Preying on the State: The Transformation of Bulgaria After 1989. Ithaca: Cornell University Press. Geddes, Barbara, and Artur Ribeiro Neto. 1992. “Institutional Sources of Corruption in Brazil.” Third World Quarterly 13, no. 4: 641–661. Gerring, John, and Strom C. Thacker. 2004. “Political Institutions and Corruption: The Role of Unitarism and Parliamentarism.” British Journal of Political Science 34, no. 2 (April): 295–330. Glaeser, Edward L., and Raven E. Saks. 2006. “Corruption in America.” Journal of Public Economics 90, nos. 6–7 (August): 1053–1072. Gleisner, Jeff. 2007. “Party Funding in Russia.” In Daniel Smilov and Jurij Toplak, eds., Political Finance and Corruption in Eastern Europe: The Transition Period. Aldershot: Ashgate, 143–158. Glinkina, Svetlana P. 1998. “The Ominous Landscape of Russian Corruption.” Transitions 5, no. 3 (March): 16–23. Global Corruption Barometer. 2003–2007. Transparency International. www.transparency.org/policy_research/surveys_indices/gcb. Global Opinion Trends. 2002–2007. Washington, DC: Pew Research Center, July 2007. http://pewglobal.org/reports/pdf/257.pdf. Goel, Rajeev K., and Michael A. Nelson. 2005. “Economic Freedom Versus Political Freedom: Cross-Country Influences on Corruption.” Australian Economic Papers 44, no. 2 (June): 121–133. Goel, Rajeev K., and Daniel P. Rich. 1989. “On the Economic Incentives of Taking Bribes.” Public Choice 61, no. 3 (June): 269–275. Golden, Miriam A., and Lucio Picci. 2005. “Proposal for a New Measure of Corruption, Illustrated with Italian Data.” Economics and Politics 17, no. 1 (March): 37–75. Goldsmith, Arthur A. 1999. “Slapping the Grasping Hand: Correlates of Political Corruption in Emerging Markets.” American Journal of Economics and Sociology 58, no. 4 (October): 865–883. Gospodinova, Velina. 2008, 21 November. “Nikoi Veche Ne Iska Dogan” (Nobody Wants Dogan Any More). Kapital, no. 47. www.capital .bg/printversion.php?storyid=586866, accessed 29 July 2009. Government Finance Statistics Yearbook. Various years. Washington, DC: International Monetary Fund. Grabbe, Heather, and Kirsty Hughes. 1998. Enlarging the EU Eastwards. Chatham House Papers. London: Royal Institute of International Affairs.

Bibliography

271

Grigorescu, Alexandru. 2006. “The Corruption Eruption in East-Central Europe: The Increased Salience of Corruption and the Role of Intergovernmental Organizations.” East European Politics and Societies 20, no. 3: 516–549. Grødeland, Åse Berit. 2007. “‘Red Mobs,’ ‘Yuppies,’ ‘Lamb Heads’ and Others: Contacts, Informal Networks and Politics in the Czech Republic, Slovenia, Bulgaria and Romania.” Europe-Asia Studies 59, no. 2 (March): 217–252. Grzymala-Busse, Anna. 2007. Rebuilding Leviathan: Party Competition and State Exploitation in Post-Communist Democracies. Cambridge: Cambridge University Press. Hacek, Miro, and Irena Baclija. 2009. “The Administrative Capacity of Slovenian Municipalities.” Journal of Local Self-Government 7, no. 3 (July): 307–327. Hale, Henry E. 2007. “Correlates of Clientelism: Political Economy, Politicized Ethnicity, and Post-Communist Transition.” In Herbert Kitschelt and Steven I. Wilkinson, eds., Patrons, Clients, and Politics: Patterns of Democratic Accountability and Political Competition. Cambridge: Cambridge University Press, 227–250. Hankiss, Elemer. 2002. “Games of Corruption: East-Central Europe, 1945–1999.” In Stephen Kotkin and Andras Sajo, eds., Political Corruption in Transition. Budapest: Central European University Press, 244–259. Hardin, Russell. 2000. “The Public Trust.” In Susan J. Pharr and Robert D. Putnam, eds., Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton: Princeton University Press, 31–51. Harding, Luke. 2008. “Putin Rolling Back Civil Rights, Warns Amnesty.” The Guardian, 27 February. www.guardian.co.uk /world/2008/feb/27/russia.humanrights, accessed 28 January 2011. Haughton, Tim. 2007. “When Does the EU Make a Difference? Conditionality and the Accession Process in Central and Eastern Europe.” Political Studies Review 5, no. 2 (May): 233–246. Haynes, Michael, and Rumy Husan. 2002. “Market Failure, State Failure, Institutions, and Historical Constraints in the East European Transition.” Journal of European Area Studies 10, no. 1: 105–129. Heidenheimer, Arnold J. 1970. “The Context of Analysis.” In Arnold J. Heidenheimer, ed., Political Corruption: Readings in Comparative Analysis. New Brunswick, NJ: Transaction Books, 3–28.

272

Bibliography

Hellman, Joel S. 1998. “Winners Take All: The Politics of Partial Reform in Postcommunist Transitions.” World Politics 50, no. 2 (January): 203–234. Hellman, Joel S., Geraint Jones, and Daniel Kaufmann. 2000. “Seize the State, Seize the Day.” World Bank Policy Research Working Paper No. 2444. Washington, DC: World Bank. Heritage Foundation/Wall Street Journal. Regulation Index. www.heritage.org/research/features/index, accessed 21 August 2010. Heywood, Paul. 1997. “Political Corruption: Problems and Perspectives.” Political Studies 45: 417–435. Hibbing, John R., and Samuel C. Patterson. 1994. “Public Trust in the New Parliaments of Central and Eastern Europe.” Political Studies 42, no. 4 (December): 570–592. Hicken, Allen D. 2007. “How Do Rules and Institutions Encourage Vote Buying?” In Frederic Charles Schaffer, ed., Elections for Sale: The Causes and Consequences of Vote Buying. Boulder: Lynne Rienner, 47–60. Holmes, Leslie. 1993. The End of Communist Power: Anti-Corruption Campaigns and Legitimation Crisis. New York: Oxford University Press. Holmes, Leslie. 1997. “Corruption and the Crisis of the PostCommunist State.” Crime, Law and Social Change 27, nos. 3–4: 275–297. Holmes, Leslie. 1999. “Corruption, Weak States and Economic Rationalism in Central and Eastern Europe.” Paper presented at the International Anti-Corruption Conference, Durban, South Africa, October. http://9iacc.org/papers/day1/ws2/dnld/d1ws2_lholmes.pdf, accessed 24 January 2011. Holmes, Leslie. 2003. “Political Corruption in Central and Eastern Europe.” In Martin J. Bull and James L. Newell, eds., Corruption in Contemporary Politics. Houndmills: Palgrave Macmillan, 193–206. Holmes, Leslie. 2006. Rotten States: Corruption, Post-Communism, and Neoliberalism. Durham, NC: Duke University Press. Hopkin, Jonathan. 1997. “Political Parties, Political Corruption, and the Economic Theory of Democracy.” Crime, Law and Social Change 27: 255–274.

Bibliography

273

Horowitz, Donald. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press. Huntington, Samuel P. 1968. Political Order in Changing Societies. New Haven: Yale University Press. Husted, Bryan W. 1999. “Wealth, Culture, and Corruption.” Journal of International Business Studies 30, no. 2 (June): 339–359. Ikstens, Janis, Daniel Smilov, and Marcin Walecki. 2001. “Party and Campaign Funding in Eastern Europe: A Study of 18 Member Countries of the ACEEEO.” Paper presented at the conference “Transparent Election Campaign Financing in the 21st Century,” Brijuni, Croatia, October 2001. Ikstens, Janis, Daniel Smilov, and Marcin Walecki. 2002. “Campaign Finance in Central and Eastern Europe: Lessons Learned and Challenges Ahead.” Washington, DC: International Foundation for Election Systems. www.ifes.org, accessed 30 November 2009. Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Society. Princeton: Princeton University Press. Inglehart, Ronald, and Wayne E. Baker. 2000. “Modernization, Cultural Change, and the Persistence of Traditional Values.” American Sociological Review 65, no. 1 (February): 19–51. Inglehart, Ronald, and Christian Welzel. 2005. Modernization, Cultural Change, and Democracy. New York: Cambridge University Press. Ishiyama, John T. 1996. “Electoral System Experimentation in the New Eastern Europe: The Single Transferable Vote and the Additional Member System in Estonia and Hungary.” East European Quarterly 29, no. 4: 487–507. Jackman, Robert. W. 1987. “Political Institutions and Voter Turnout in the Industrial Democracies.” American Political Science Review 81: 405–423. Jackman, Robert W., and Ross A. Miller. 1987. “Voter Turnout in the Industrialized Democracies During the 1980s.” Comparative Political Studies 27: 467–492. Jain, Arvind K. 2001. “Corruption: A Review.” Journal of Economic Surveys 15, no. 1 (February): 71–121. Janicke, Martin, and Helmut Weidner, eds. 1997. National Environmental Policies: A Comparative Study of Capacity Building. Berlin: Springer.

274

Bibliography

Johnson, Simon, Daniel Kaufmann, and Pablo Zoido-Lobaton. 1999. “Corruption, Public Finances, and the Unofficial Economy.” World Bank Policy Research Working Paper No. 2169. Washington, DC: World Bank. Johnston, Michael, ed. 2005a. Civil Society and Corruption: Mobilizing for Reform. Lanham, MD: University Press of America. Johnston, Michael. 2005b. Syndromes of Corruption: Wealth, Power, and Democracy. Cambridge: Cambridge University Press. Jungerstam-Mulders, Susanne, ed. 2006. Post-Communist EU Member States: Parties and Party Systems. Aldershot: Ashgate. Kahramanova, Darina. 2009. “Firmi ot Obrucha na DPS se Vklinyavat v Stroezha na ‘Lyulin’” (Firms from the MRF Ring Are Included in the Construction of “Lyulin”). Sega, 8 January. www.segabg.com /online/new/articlenew.asp?issueid=3242§ionId=16&id=0000 102, accessed 14 July 2009. Kajsiu, Blendi. 2004. Country Report: Albania. In Nations in Transit. Washington, DC: Freedom House. www.freedomhouse.org, accessed 16 January 2010. Kaldor, Mary, and Ivan Vejvoda. 1997. “Democratization in Central and East European Countries,” International Affairs 1: 59–82. Kanin, David B. 2003. “Big Men, Corruption, and Crime.” International Politics 40, no. 4: 491–526. Karasimeonov, Georgi. 2006. Partiinata Sistema v Bulgaria (The Party System in Bulgaria). Sofia: Goreks Press. Karatnycky, Adrian, Alexander Motyl, and Amanda Schnetzer. 2001. Nations in Transit. New Brunswick, NJ: Freedom House. Karklins, Rasma. 2005. The System Made Me Do It: Corruption in PostCommunist Societies. Armonk, NY: M. E. Sharpe. Karp, Jeffrey A., and Susan A. Banducci. 2007. “Political Efficacy and Participation in Twenty-Seven Democracies: How Electoral Systems Shape Political Behaviour.” British Journal of Political Science 38, no. 2 (April): 311–334. Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2005. “Governance Matters IV: Governance Indicators for 1996–2004.” World Bank Policy Research Working Paper No. 3630. Washington, DC: World Bank. www.worldbank.org/wbi/governance/pubs/gov matters4.html, accessed 27 September 2009. Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2010. “Worldwide Governance Indicators: Methodology and Analytical

Bibliography

275

Issues.” World Bank Policy Research Working Paper No. 5430. Washington, DC: World Bank. Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobaton. 1999. “Governance Matters.” World Bank Policy Research Working Paper No. 2196. Washington, DC: World Bank. Kaufmann, Daniel, and Paul Siegelbaum. 1997. “Privatization and Corruption in Transition Economies.” Journal of International Affairs 50, no. 2 (Winter): 419–458. Keohane, Robert O., and Lisa L. Martin. 1995. “The Promise of Institutionalist Theory.” International Security 20 (Summer): 39–51. Kitschelt, Herbert, and Steven I. Wilkinson. 2007. “Citizen-Politician Linkages: An Introduction.” In Herbert Kitschelt and Steven I. Wilkinson, eds., Patrons, Clients, and Politics: Patterns of Democratic Accountability and Political Competition. Cambridge: Cambridge University Press, 1–49. Knack, Stephen. 2007. “Measuring Corruption: A Critique of Indicators in Eastern Europe and Central Asia.” Journal of Public Policy 27, no. 3: 255–291. Knack, Stephen, and Phillip Keefer. 1995. “Institutions and Economic Performance: Cross-Country Tests Using Alternative Institutional Measures.” Economics and Politics 7, no. 3 (November): 207–227. Kneen, Peter. 2000. “Political Corruption in Russia and the Soviet Legacy.” Crime, Law and Social Change 34, no. 4 (June): 349–367. Kohl, Richard. 1995. “Sustaining the Transition: Are Social Contracts the Answer?” World Policy Journal 12 (Fall): 109–117. Kolarska-Bobinska, Lena. 2002. “The Impact of Corruption on Legitimacy of Authority in New Democracies.” In Stephen Kotkin and Andras Sajo, eds., Political Corruption in Transition. Budapest: Central European University Press, 314–325. Kostadinova, Tatiana. 2000. “Mixed Electoral Systems: Determinants and Political Consequences.” PhD dissertation, Florida State University. Kostadinova, Tatiana. 2002. “Do Mixed Electoral Systems Matter? A Cross-National Analysis of Their Effects in Eastern Europe.” Electoral Studies 21, no. 1 (March): 23–34. Kostadinova, Tatiana. 2003. “Voter Turnout Dynamics in PostCommunist Europe.” European Journal of Political Research 42: 741–759.

276

Bibliography

Kostadinova, Tatiana. 2007. “The Impact of Finance Regulations on Political Parties: The Case of Bulgaria.” Europe-Asia Studies 59, no. 5 (July): 807–827. Kostadinova, Tatiana, and Timothy Power. 2007. “Does Democratization Depress Participation?: Voter Turnout in the Latin American and Eastern European Democratic Transitions.” Political Research Quarterly 60, no. 3 (September): 363–377. Kotkin, Stephen, and Andras Sajo, eds. 2002. Political Corruption in Transition: A Sceptic’s Handbook. Budapest: CEU Press. Kramer, John M. 1977. “Political Corruption in the U.S.S.R.” Western Political Quarterly 30, no. 2 (June): 213–224. Krastev, Ivan. 1998. “Dancing with Anticorruption.” East European Constitutional Review 7, no. 3 (Summer): 56–58. Krastev, Ivan. 2004. Shifting Obsessions: Three Essays on the Politics of Anticorruption. Budapest: CEU Press. Krastev, Ivan. 2006. Country Report: Bulgaria. In Nations in Transit. Washington, DC: Freedom House. www.freedomhouse.org, accessed 16 January 2010. Kregar, Josip, Djordje Gardasevic, and Viktor Gotovac. 2007. “Party and Campaign Finance in Croatia.” In Daniel Smilov and Jurij Toplak, eds., Political Finance and Corruption in Eastern Europe: The Transition Period. Aldershot: Ashgate, 53–70. Krusteva, Sibina. 2008. “A’ga Yadohme Gushiya Drob” (When We Ate Duck Liver). Kapital no. 42, 17 October. www.capital.bg /printversion.php?storyid=566941, accessed 29 July 2009. Kunicova, Jana, and Susan Rose-Ackerman. 2005. “Electoral Rules and Constitutional Structures as Constraints on Corruption.” British Journal of Political Science 35, no. 4 (October): 573–606. Laakso, Markku, and Rein Taagepera. 1979. “Effective Number of Parties: A Measure with Application to Western Europe.” Comparative Political Studies 12: 13–27. Lalov, Nikola. 2011. “Boyko Borissov: Ne Sme Nai-korumpiranite v ES” (Boyko Borissov: We Are Not the Most Corrupt in the EU). Mediapool, 2 February. www.mediapool.bg/show/?storyid=175584, accessed 2 February 2011. Lambsdorff, Johann Graf. 1999. “Corruption in Empirical Research: A Review.” Paper presented at the International Anti-Corruption Conference, Durban, South Africa, October. http://9iacc.org

Bibliography

277

/papers/day2/ws1/d2ws1_jglambsdorgg.html, accessed 5 August 2010. Lambsdorff, Johann Graf. 2006. “Consequences and Causes of Corruption: What Do We Know from a Cross-Section of Countries?” In Susan Rose-Ackerman, ed., International Handbook on the Economics of Corruption. Cheltenham: Edward Elgar, 3–51. Lambsdorff, Johann Graf. 2007. The Institutional Economics of Corruption and Reform: Theory, Evidence, and Policy. Cambridge: Cambridge University Press. Lancaster, Thomas D., and Gabriella R. Montinola. 1997. “Toward a Methodology for the Comparative Study of Political Corruption.” Crime, Law and Social Change 27: 185–206. LaPalombara, Joseph. 1994. “Structural and Institutional Aspects of Corruption.” Social Research 61, no. 2 (September): 325–350. LaPlant, James T., Michael Baun, Jiri Lach, and Dan Marek. 2004. “Decentralization in the Czech Republic: The European Union, Political Parties, and the Creation of Regional Assemblies.” Journal of Federalism 34, no. 1 (Winter): 35–51. La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert Vishny. 1999. “The Quality of Government.” Journal of Law, Economics, and Organization 15, no. 1 (March): 222–279. Larrabee, F. Stephen. 2006. “Danger and Opportunity in Eastern Europe.” Foreign Affairs 85, no. 6 (November–December): 117–131. Leff, Nathaniel H. 1964. “Economic Development Through Bureaucratic Corruption.” American Behavioral Scientist 8, no. 3 (November): 8–14. Levi, Margaret, and Laura Stoker. 2000. “Political Trust and Trustworthiness.” Annual Review of Political Science 3: 475–507. Lewis-Beck, Michael S. 1977. “The Relative Importance of Socioeconomic and Political Variables for Public Policy.” American Political Science Review 71, no. 2 (June): 559–566. Lijphart, Arend. 1977. Democracy in Plural Societies: A Comparative Explanation. New Haven: Yale University Press. Lijphart, Arend. 1999. Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. New Haven: Yale University Press.

278

Bibliography

Lilov, Grigor. 2008a. Underground 1: Samokoveza, Doktora. Sofia: Kailas. Lilov, Grigor. 2008b. Underground 2: VIS, Multigrup. Sofia: Kailas. Linz, Juan J., and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe. Baltimore: Johns Hopkins University Press. Lipset, Seymour Martin, and Gabriel Salman Lenz. 2000. “Corruption, Culture, and Markets.” In Lawrence E. Harrison and Samuel P. Huntington, eds., Culture Matters: How Values Shape Human Progress. New York: Basic Books, 112–124. Liu, Zihua. 2005. “Understanding Corruption Through Cross-National Comparison.” Paper presented at the annual meeting of the American Political Science Association, Washington, DC, September. Lloyd, John. 1999. “Stranded Between Two Worlds.” New Statement, 6 September, pp. 29–30. Luhiste, Kadri. 2006. “Explaining Trust in Political Institutions: Some Illustrations from the Baltic States.” Communist and PostCommunist Studies 39, no. 4 (December): 475–496. Manzetti, Luigi, and Charles H. Blake. 1996. “Market Reforms and Corruption in Latin America: New Means for Old Ways.” Review of International Political Economy 3, no. 4 (Winter): 662–697. Markovich, Slobodan. 2008. Country Report: Serbia. In Nations in Transit. Washington, DC: Freedom House. www.freedomhouse .org, accessed 16 January 2010. Mauro, Paolo. 1995. “Corruption and Growth.” Quarterly Journal of Economics 110 (August): 681–712. McManus-Czubinska, Clare, William L. Miller, Radoslaw Markowski, and Jacek Wasilewski. 2004. “Why Is Corruption in Poland ‘A Serious Cause for Concern’?” Crime, Law and Social Change 41: 107–132. Menard, Scott. 2004. “Six Approaches to Calculating Standardized Logistic Regression Coefficients.” American Statistician 58, no. 3 (August): 218–223. Merton, Robert. 1968. Social Theory and Social Structure. New York: Free Press. Michael, Bryane. 2005. “Overview of Anti-Corruption Training in Central and Eastern Europe.” In Bryane Michael, ed., Anti-

Bibliography

279

Corruption Training Programmes in Central and Eastern Europe. Strasbourg: Council of Europe, 9–19. “Michael Lee’s Letter” (Pismoto na Michael Lee). 2008. Mediapool, 13 May. www.mediapool.bg/show/?storyid=138865, accessed 14 May 2009. Michev, Yordan. 2008. “Batzil” (Bacillus). Tema, no. 25, 5 July, pp. 12–18. Milesi-Ferretti, Gian Maria, Roberto Perotti, and Massimo Rostagno. 2002. “Electoral Systems and Public Spending.” Quarterly Journal of Economics 117, no. 2 (May): 609–657. Millar, James R. 1985. “The Little Deal: Brezhnev’s Contribution to Acquisitive Socialism.” Slavic Review 44, no. 4 (Winter): 694–706. Miller, William L., Åse B. Grødeland, and Tatyana Y. Koshechkina. 2001. A Culture of Corruption? Coping with Government in PostCommunist Europe. Budapest: CEU Press. Miller, William L., Tatyana Koshechkina, and Åse Grødeland. 1997. “How Citizens Cope with Postcommunist Officials: Evidence from Focus Group Discussions in Ukraine and the Czech Republic.” Political Studies 45, no. 3: 597–625. Minchev, Ognian. 2009. “Etno-Korporativniyat Model na Ahmed Dogan I Krizata na Bulgarskata Durzhava” (The Ethno-Corporate Model of Ahmed Dogan and the Crisis of the Bulgarian State). Mediapool, 1 July. www.mediapool.bg /show/?storyid=153924, accessed 3 August 2009. Mishler, William, and Richard Rose. 1997. “Trust, Distrust, and Skepticism: Popular Evaluations of Civil and Political Institutions in Post-Communist Societies.” Journal of Politics 59, no. 2 (May): 418–451. Mishler, William, and Richard Rose. 2001. “What Are the Origins of Political Trust? Testing Institutional and Cultural Theories in PostCommunist Societies.” Comparative Political Studies 34, no. 1 (February): 30–62. Mishler, William, and Richard Rose. 2005. “What Are the Political Consequences of Trust? A Test of Cultural and Institutional Theories in Russia.” Comparative Political Studies 20, no. 10: 1–29. Monitoring the EU Accession Process: Corruption and Anti-Corruption Policy. 2002. New York: Open Society Institute. www.soros .org/resources/articles_publications/publications/euaccess corruption_20020601, accessed 17 January 2011.

280

Bibliography

Montinola, Gabriella R., and Robert W. Jackman. 2002. “Sources of Corruption: A Cross-Country Study.” British Journal of Political Science 32, no. 1 (January): 147–170. Moran, J. 2001. “Democratic Transitions and Forms of Corruption.” Crime, Law and Social Change 36, no. 4 (December): 379–393. Moraski, Bryon, and Gerhard Lowenberg. 1999. “The Effect of Legal Thresholds on the Revival of Communist Parties in East-Central Europe.” Journal of Politics 61, no. 1 (February): 151–170. Muddle, Cas. 2003. “Civil Society in Post-Communist Europe: Lessons from the Dark Side.” In Petr Kopecky and Cas Muddle, eds., Uncivil Society: Contentious Politics in Post-Communist Europe. London: Routledge, 157–170. Muller, Wolfgang. 2006. “Party Patronage and Party Colonization of the State.” In Richard Katz and William Crotty, eds., Handbook of Party Politics. London: Sage, 189–195. Mungiu-Pippidi, Alina. 2006a. Country Report: Romania. In Nations in Transit. Washington, DC: Freedom House. www.freedomhouse.org, accessed 16 January 2010. Mungiu-Pippidi, Alina. 2006b. “Democratization Without Decommunization in the Balkans.” East European Democratization 50, no. 4 (Autumn): 641–655. Mungiu-Pippidi, Alina. 2008. Country Report: Romania. In Nations in Transit. Washington, DC: Freedom House. www.freedomhouse.org, accessed 16 January 2010. Mungiu-Pippidi, Alina. 2010. “The Experience of Civil Society as an Anti-Corruption Actor in East Central Europe.” Report prepared for the Civil Society Against Corruption project of the Romanian Academic Society. www.againstcorruption.eu/find-a-cure/, accessed 29 December 2010. Myerson, Roger B. 1993. “Effectiveness of Electoral Systems for Reducing Government Corruption: A Game-Theoretic Analysis.” Games and Economic Behavior 5: 118–132. Nations in Transit. 2009. Washington, DC: Freedom House. www.freedomhouse.org/template.cfm?page=17, accessed 15 October 2010. Neikov, Dian. 2008. “Tatiana Dontcheva: Support from the Galevi Brothers Is a ‘Bad Sign’ for the Authorities.” Mediapool, 31 March. www.mediapool.bg/show/?storyid=137685&srcpos=4.

Bibliography

281

Newton, Kenneth, and Pippa Norris. 2000. “Confidence in Public Institutions: Faith, Culture, or Performance?” In Susan J. Pharr and Robert D. Putnam, eds., Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton: Princeton University Press, 52–73. Norris, Pippa. 2004. Electoral Engineering: Voting Rules and Political Behavior. Cambridge: Cambridge University Press. Noutcheva, Gergana. 2007. “Fake, Partial and Imposed Compliance: The Limits of the EU’s Normative Power in the Western Balkans.” Working Document No. 274, July. Brussels: Centre for European Policy Studies. http://ssrn.com/abstract=1337980, accessed 20 February 2010. Noutcheva, Gergana, and Dimitar Bechev. 2008. “The Successful Laggards: Bulgaria and Romania’s Accession to the EU.” East European Politics and Societies 22, no. 1 (February): 114–144. Nunberg, Barbara. 1999. The State After Communism: Administrative Transitions in Central and Eastern Europe. Washington, DC: World Bank. Nyblade, Benjamin, and Steven R. Reed. 2008. “Who Cheats? Who Loots? Political Competition and Corruption in Japan.” American Journal of Political Science 52, no. 4 (October): 926–941. Obretenov, Lyuben. 2007. “Ahmed Dogan Prizna za Biznesa Si v Devin” (Ahmed Dogan Admitted His Business in Devin). Sega, 13 April. www.segabg.com/online/new/articlenew.asp?issueid=2614 §onId=16&id=0000101, accessed 14 July 2009. O’Donnell, Guillermo, and Mario Vargas Llosa. 1993. “Privatization Is Not Democratization: The Browning of Latin America.” New Perspectives Quarterly 10 (Fall): 50–53. Offe, Claus. 2004. “Political Corruption: Conceptual and Practical Issues.” In Janos Kornai and Susan Rose-Ackerman, eds., Building a Trustworthy State in Post-Socialist Transition. Houndmills: Palgrave Macmillan, 77–99. On the Eve of EU Accession: Anti-Corruption Reforms in Bulgaria. 2006. Sofia: Center for the Study of Democracy. Open Society Institute. 2002. Corruption and Anticorruption Policy in the EU Accession Process. Budapest: Open Society Institute. Ordeshook, Peter C., and Olga V. Shvetsova. 1994. “Ethnic Heterogeneity, District Magnitude, and the Number of Parties.”

282

Bibliography

American Journal of Political Science 38, no. 1 (February): 100–123. Organisation for Economic Co-operation and Development. 1994. Capacity Development in Environment. Paris: OECD. Organisation for Economic Co-operation and Development. 2002. AntiCorruption Measures in South Eastern Europe: Civil Society’s Involvement. www.oecd.org/dataoecd/49/14/37152511 .pdf, accessed 16 July 2010. Orlova, Alexandra V. 2005. “Organized Crime and the Rule of Law in the Russian Federation.” Essex Human Rights Review 2, no. 1: 23–37. Pabriks, Artis, and Aiga Stokenberga. 2006. “Political Parties and the Party System in Latvia.” In Susanne Jungerstam-Mulders, ed., PostCommunist EU Member States: Parties and Party Systems. Aldershot: Ashgate, 51–68. Pacek, Alexander C., Grigore Pop-Eleches, and Joshua A. Tucker. 2009. “Disenchanted or Discerning: Voter Turnout in Post-Communist Countries.” Journal of Politics 71, no. 2 (April): 473–491. Pacek, Alexander, and Benjamin Radcliff. 1995. “The Political Economy of Competitive Elections in the Developing World.” American Journal of Political Science 39, no. 3 (August): 745–759. Paldam, Martin. 2002. “The Cross-Country Pattern of Corruption: Economics, Culture, and the Seesaw Dynamics.” European Journal of Political Economy 18, no. 2 (June): 215–240. Parties and Elections: A Database About Parliamentary Elections and Political Parties in Europe. 2009. Wolfram Nordsieck. www.parties-and-elections.de, accessed 29 March 2010. Pashev, Konstantin, Assen Dyulgerov, and Georgi Kaschiev. 2006. Corruption in Public Procurement: Risks and Reform Policies. Sofia: Center for the Study of Democracy. Peev, Dimitar. 2000. “Siniyat ‘Olymp’ Razhda Nova Grupirovka” (The Blue “Olymp” Gives Birth to a New Grouping). Kapital, no. 14, 7 April. www.capital.bg/printversion.php?storyid =255186, accessed 7 July 2009. Peeva, Teodora. 2002. “Suvetut za Ikonomicheski Rastezh: Indulgenziya ili Vuzrazhdane na Bulgarskiya Biznes” (The Council for Economic Growth: Redemption or Renaissance of the Bulgarian Business). Sega, 29 January. www.segabg.com /online/new/articlenew.asp?issueId=559§ionid=5&id=0000, accessed 30 July 2009.

Bibliography

283

Perottino, Michael, Marek Chmaj, Marian Walecki, Jaroslaw Zbieranek, Adrian Moraru, Elena Iorga, and Ihor Sheviliakov. 2005. Legislation and Control Mechanisms of Political Parties’ Funding: Czech Republic, Poland, Romania, and Ukraine. Bucharest: Institut Pentru Politici Publice. Persson, Torsten. 2002. “Do Political Institutions Shape Economic Policy?” Econometrica 70, no. 3 (May): 883–905. Persson, Torsten, and Guido Tabellini. 1999. “The Size and Scope of Government: Comparative Politics with Rational Politicians.” European Economic Review 43: 699–735. Persson, Torsten, and Guido Tabellini. 2005. The Economic Effects of Constitutions. Cambridge, MA: MIT Press. Persson, Torsten, Guido Tabellini, and Francesco Trebbi. 2001. “Electoral Rules and Corruption.” Working Paper No. 8154. Cambridge: National Bureau of Economic Research. www.nber .org/papers/w8154, accessed 15 August 2006. Petkova, Tanya. 2007. “Stanishev: Nai-Veche Prokuraturata Ni Spasyava ot Predpazna Klauza” (Stanishev: It Is Mainly the Prosecutor’s Office That Spares Us the Safeguard Clause). Mediapool, 1 June. www.mediapool.bg/show/?storyid=129039, accessed 29 July 2009. Petkova, Tanya. 2009. “Politika na Nishtopravene, Arogantnost I Konsumaziya na Vlastta” (Politics of Doing-Nothing, Arogance, and Consumption of Power). Mediapool, 9 July. www.mediapool .bg/show/?storyid=154247&srcpos=20, accessed 10 July 2009. Petrov, Krustyo. 2007. “Trugva Nova Sudebna Saga Okolo Durzhavniya Rezerv” (A New Court Saga Begins with the State Reserve). Mediapool, 26 September. www.mediapool.bg /show/?storyid=132308, accessed 4 August 2009. Petrova, Velina P. 2007. “Civil Society in Post-Communist Eastern Europe and Eurasia: A Cross-National Analysis of Micro- and Macro-Factors.” World Development 35, no. 7: 1277–1305. Pharr, Susan J., and Robert D. Putnam. 2000. Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton: Princeton University Press. Philp, Mark. 1997. “Defining Political Corruption.” In Paul Heywood, ed., Political Corruption. Oxford: Blackwell, 20–46. Philp, Mark. 2002. “Political Corruption, Democratization, and Reform.” In Stephen Kotkin and Andras Sajo, eds., Political

284

Bibliography

Corruption in Transition: A Sceptic’s Handbook. Budapest: CEU Press, 57–80. Pinto-Duschinsky, Michael. 2002. “Financing Politics: A Global View.” Journal of Democracy 13, no. 4 (October): 69–86. Polidano, Charles. 2000. “Measuring Public Sector Capacity.” World Development 28, no. 5: 805–822. Powell, G. Bingham. 1986. “American Voter Turnout in Comparative Perspective.” American Political Science Review 80: 17–43. Powell, G. Bingham. 2000. Elections of Instruments of Democracy: Majoritarian and Proportional Visions. New Haven: Yale University Press. Primo, David M., Matthew L. Jacobsmeier, and Jeffrey Milyo. 2007. “Estimating the Impact of State Policies and Institutions with Mixed-Level Data.” State Politics and Policy Quarterly 7, no. 4: 446–459. Progress Reports of Candidate Countries. Various years. Brussels: European Commission. http://ec.europa.eu/enlargement/press _corner/key-documents/index_archive_en.htm. Protsyk, Oleh, and Marcin Walecki. 2007. “Party Funding in Ukraine.” In Daniel Smilov and Jurij Toplak, eds., Political Finance and Corruption in Eastern Europe: The Transition Period. Aldershot: Ashgate, 189–208. Przeworski, Adam, Susan C. Stokes, and Bernard Manin, eds. 1999. Democracy, Accountability, and Representation. New York: Cambridge University Press. Radaev, Vadim. 2002. “Corruption and Administrative Barriers for Russian Business.” In Stephen Kotkin and Andras Sajo, eds., Political Corruption in Transition. Budapest: Central European University Press, 287–311. Radmilova, Maria, Hristo Nikolov, and Doroteya Dachkova. 2007. “‘Frontier’: Prednata Linia na Golemite Pari” (“Frontier”: The Front Line of the Big Money). Sega, 31 May. www.segabg.com /online/new/articlenew.asp?issueid=2659§ionId=5&id=00014 01, accessed 14 July 2009. Raubisko, Ieva. 2001. “Letting in the Sunshine: Latvia Fights Public Perception of Corruption.” Central Europe Review 3, no. 15 (April), www.ce-review.org/01/15/raubisko15.html, accessed 20 January 2011.

Bibliography

285

Rauch, James E., and Peter B. Evans. 2000. “Bureaucratic Structure and Bureaucratic Performance in Less Developed Countries.” Journal of Public Economics 75: 49–71. Reed, Quentin. 2002. “Corruption in Czech Privatization: The Dangers of ‘Neo-Liberal’ Privatization.” In Stephen Kotkin and Andras Sajo, eds., Political Corruption in Transition: A Sceptic’s Handbook. Budapest: CEU Press, 261–286. “Report from the Commission to the European Parliament and the Council on Bulgaria’s Progress on Accompanying Measures Following Accession.” 2007. Brussels: Commission of the European Communities, 27 June. Risteska, Marija. 2007. “The Macedonian Accession to the European Union.” Position Paper. Prague: Institut pro Europskou Politiki EUROPEUM (Institute for European Policy). www.europeum.org /doc/pdf/860.pdf, accessed 19 January 2010. Roberts, Andrew. 2010. The Quality of Democracy in Eastern Europe: Public Preferences and Policy Reforms. Cambridge: Cambridge University Press. Roper, Steven D. 2002. “The Influence of Romanian Campaign Finance Laws on Party System Development and Corruption.” Party Politics 8, no. 2: 175–192. Rose, Richard. 2004. Europe Expands, Turnout Falls: The Significance of the 2004 European Parliament Election. Stockholm: International Institute for Democracy and Electoral Assistance. www.idea.int. Rose-Ackerman, Susan. 1978. Corruption: A Study in Political Economy. New York: Academic Press. Rose-Ackerman, Susan. 1999. Corruption and Government: Causes, Consequences, and Reform. Cambridge: Cambridge University Press. Rose-Ackerman, Susan. 2001. “Trust and Honesty in Post-Socialist Societies.” KYKLOS 54, nos. 2–3: 415–444. Rose-Ackerman, Susan. 2007. “From Elections to Democracy in Central Europe: Public Participation and the Role of Civil Society.” East European Politics and Societies 21, no. 1: 31–47. Roth, Jurgen. 2008. Novite Bulgarski Demoni (The New Bulgarian Demons). Sofia: Izdatelstvo Slunze. Rothstein, Bo, and Eric M. Uslaner. 2005. “All for All: Equality, Corruption, and Social Trust.” World Politics 58 (October): 41–72.

286

Bibliography

“Russian Police Raid Opposition Magazine.” 2010. Reuters, 2 September. http://in.reuters.com/article/idINIndia-51241820100902, accessed 2 September 2010. Sajo, Andras. 1998. “Corruption, Clientelism, and the Future of the Constitutional State in Eastern Europe.” East European Constitutional Review 7, no. 2 (Spring): 37–46. Sajo, Andras. 2003. “From Corruption to Extortion: Conceptualization of Post-Communist Corruption.” Crime, Law and Social Change 40: 171–194. Sandholtz, Wayne, and Mark M. Gray. 2003. “International Integration and National Corruption.” International Organization 57 (Fall): 761–800. Sandholtz, Wayne, and William Koetzle. 2000. “Accounting for Corruption: Economic Structure, Democracy, and Trade.” International Studies Quarterly 44, no. 1 (March): 31–50. Sandholtz, Wayne, and Rein Taagepera. 2005. “Corruption, Culture, and Communism.” International Review of Sociology 15, no. 1 (March): 109–131. Sasse, Gwendolyn. 2008. “The Politics of EU Conditionality: The Norm of Minority Protection During and Beyond EU Accession.” Journal of European Public Policy 16, no. 6 (September): 842–860. Scarrow, Susan E. 2003. “Party Finance Scandals and Their Consequences in the 2002 Election: Paying for Mistakes?” German Politics and Society 66, no. 1 (Spring): 119–137. Schimmelfennig, Frank. 2008. “EU Political Accession Conditionality After the 2004 Enlargement: Consistency and Effectiveness.” Journal of European Public Policy 15, no. 6 (September): 918–937. Schmidt, Diana. 2007. “Anti-Corruption: What Do We Know? Research on Preventing Corruption in the Post-Communist World.” Political Studies Review 5: 202–232. Schmitter, Philippe C. 2005. “Democratization and State Capacity.” Paper presented at X Congreso Internacional del CLAD Sobre la Reforma del Estado y de la Administracion Publica, Santiago, Chile, October. http://unpan1.un.org/intradoc/groups/public /documents/CLAD/clad0052201.pdf, accessed 26 April 2010. Sedelmeier, Ulrich. 2008. “After Conditionality: Post-Accession Compliance with EU Law in East Central Europe.” Journal of European Public Policy 15, no. 6 (September): 806–825.

Bibliography

287

Shelley, Louise. 2005. “Civil Society Mobilized Against Corruption: Russia and Ukraine.” In Michael Johnston, ed., Civil Society and Corruption: Mobilizing for Reform. Boulder, CO: University Press of America, 3–21. Sik, Endre. 1992. “From the Second to the Informal Economy.” Journal of Public Policy 12, no. 2 (April–June): 153–175. Sik, Endre. 2002. “The Bad, the Worse, and the Worst: Guesstimating the Level of Corruption.” In Stephen Kotkin and Andras Sajo, eds., Political Corruption in Transition: A Sceptic’s Handbook. Budapest: Central European University Press, 91–113. Sikk, Alan. 2003. “A Cartel Party System in a Post-Communist Country? The Case of Estonia.” Paper presented at the general conference of the European Consortium for Political Research, Marburg, Germany, September. Sikk, Alan. 2005. “Party Financing Regime in Estonia.” Paper presented at the conference “Campaign Finance and Party System Development in Eastern Europe,” Riga, Latvia, October. Simis, Konstantin M. 1982. USSR: Secrets of a Corrupt Society. London: J. M. Dent & Sons. Simov, Alexander. 2008, 29 March. “Roumen Petkov Tryabva da Podade Ostavka” (Roumen Petkov Must Resign). Duma. http://old.duma.bg/2008/0308/290308/obshtestvo/ob-5.html, accessed 5 January 2011. Smilov, Daniel. 2007. “Introduction: Party Funding, Campaign Finance and Corruption in Eastern Europe.” In Daniel Smilov and Jurij Toplak, eds., Political Finance and Corruption in Eastern Europe: The Transition Period. Aldershot: Ashgate, 1–32. Smilov, Daniel, and Jurij Toplak, eds. 2007. Political Finance and Corruption in Eastern Europe: The Transition Period. Aldershot: Ashgate. Solnick, Steven L. 1998. Stealing the State: Control and Collapse in Soviet Institutions. Cambridge: Harvard University Press. Special Eurobarometer 291: The Attitudes of Europeans Towards Corruption. 2008. Brussels: European Commission, April. http://ec.europa.eu/public_opinion/archives/ebs/ebs_291_en.pdf. Srdoc, Natasha, and Joel Anand Samy. 2009. “Corruption in Croatia: Despite the EU Assurance, Zagreb Is No Role Model for the Region.” Wall Street Journal, 19 May, p. A15.

288

Bibliography

Standard Eurobarometer 69: Public Opinion in the European Union. 2008. Brussels: European Commission, June. http://ec.europa.eu /public_opinion/archives/eb/eb69/eb69_en.htm. Stantchev, Krassen. 2007. “The Limitations on Voting May Have Economic Consequences.” Mediapool, 29 July. Steves, Franklin, and Alan Rousso. 2003. “Anti-Corruption Programs in Post-Communist Transition Countries and Changes in the Business Environment, 1999–2002.” Working Paper No. 85. London: European Bank for Reconstruction and Development. Szczerbiak, Aleks. 2006. “Power Without Love? Patterns of Party Politics in Post-1989 Poland.” In Susanne Jungerstam-Mulders, ed., Post-Communist EU Member States: Parties and Party Systems. Aldershot: Ashgate, 91–124. Taagepera, Rein. 2002. “Baltic Values and Corruption in Comparative Context.” Journal of Baltic Studies 33, no. 3 (Fall): 243–258. Tallberg, Jonas. 2002. “Paths to Compliance: Enforcement, Management, and the European Union.” International Organization 56 (Summer): 609–643. Tanzi, Vito, and Hamid Davoodi. 1997. “Corruption, Public Investment, and Growth.” Working Paper No. 97/139. Washington, DC: International Monetary Fund, Fiscal Affairs Department. www.imf.org/external /pubs/ft/wp/wp97139.pdf, accessed 18 December 2010. Tavits, Margit. 2008. “Representation, Corruption, and Subjective WellBeing.” Comparative Political Studies 41, no. 12 (December): 1607–1630. Todorakov, Lyubomir. 2010. “Bulgaria.” Report prepared for the Civil Society Against Corruption project of the Romanian Academic Society. www.againstcorruption.eu/find-a-cure/, accessed 29 December 2010. Toplak, Jurij. 2007. “Party Funding in Slovenia.” In Daniel Smilov and Jurij Toplak, eds., Political Finance and Corruption in Eastern Europe: The Transition Period. Aldershot: Ashgate, 171–188. Transparency International. 2006. “Corruption and the EU Accession Process.” Background Paper. www.transparency.org/news_room /in_focus/eu_accession, accessed 4 January 2010. Transparency International Corruption Perceptions Index. Various years. Transparency International. www.transparency.org. Transparency International Global Corruption Barometer. Various

Bibliography

289

years. Berlin: Transparency International. www.transparency.org /policy_research/surveys_indices/gcb, accessed 5 January 2010. Treisman, Daniel. 2000. “The Causes of Corruption: A Cross-National Study.” Journal of Public Economics 76, no. 3 (June): 399–457. Treisman, Daniel. 2007. “What Have We Learned About the Causes of Corruption from Ten Years of Cross-National Empirical Research?” Annual Review of Political Science 10 (June): 211–244. Treneska, Renata. 2007. “Party Funding and Campaign Finance in Macedonia.” In Daniel Smilov and Jurij Toplak, eds., Political Finance and Corruption in Eastern Europe: The Transition Period. Aldershot: Ashgate, 105–122. Unikaite, Ingrida. 2005. “The Effect of Party and Campaign Finance on Lithuanian Party Development.” Paper presented at the conference “Campaign Finance and Party System Development in Eastern Europe,” Riga, Latvia, October. US Agency for International Development. 2008. Nongovernmental Organization Sustainability Index. Washington, DC: USAID. www.usaid.gov/locations/europe_eurasia/dem_gov/ngoindex/, accessed 20 March 2010. Uslaner, Eric M. 2008. Corruption, Inequality, and the Rule of Law: The Bulging Pocket Makes the Easy Life. Cambridge: Cambridge University Press. Vachudova, Milada Anna. 2005. Europe Undivided: Democracy, Leverage, and Integration After Communism. Oxford: Oxford University Press. Vachudova, Milada Anna. 2008. “Tempered by the EU? Political Parties and Party Systems Before and After Accession.” Journal of European Public Policy 15, no. 6 (September): 861–879. Vachudova, Milada Anna. 2009. “Corruption and Compliance in the EU’s Post-Communist Members and Candidates.” Journal of Common Market Studies 47: 43–62. Van Biezen, Ingrid, and Petr Kopecky. 2001. “On the Predominance of State Money: Reassessing Party Financing in the New Democracies of Southern and Eastern Europe.” Perspectives on European Politics and Society 2, no. 3: 401–429. Vasileva, Maria. 2007. “Energiiniyat Sector u Nas E s Nai-visok Korupzionen Risk” (Our Energy Sector Carries the Highest Risk for Corruption). Mediapool, 16 May. www.mediapool.bg/show /?storyid =128642, accessed 20 July 2009.

290

Bibliography

Vladimirova, Petia. 1995. “Hronologiya na Zurnenata Kriza” (A Chronology of the Grain Crisis). Demokraziya, no. 294 (December 1995): 3. Volkov, Vadim. 2002. Violent Entrepreneurs: The Use of Force in the Making of Russian Capitalism. Ithaca: Cornell University Press. Walecki, Marcin. 2003. “Money and Politics in Central and Eastern Europe.” In Reginald Austin and Maja Tjernstrom, eds., Funding of Political Parties and Election Campaigns. Stockholm: International Institute for Democracy and Electoral Assistance, 71–93. Walecki, Marcin. 2005. “Political Money and Corruption.” International Foundation for Election Systems (IFES) Political Finance White Paper Series. Washington, DC: IFES. Walecki, Marcin. 2007. “Political Finance in Poland.” In Daniel Smilov and Jurij Toplak, eds., Political Finance and Corruption in Eastern Europe: The Transition Period. Aldershot: Ashgate, 123–142. Wallace, Claire, and Rossalina Latcheva. 2006. “Economic Transformation Outside the Law: Corruption, Trust in Public Institutions and the Informal Economy in Transition Countries of Central and Eastern Europe.” Europe-Asia Studies 58, no. 1 (January): 81–102. Warner, Carolyn M. 2005. “The Corruption of Campaign and Party Finance Laws.” Paper presented at the annual meeting of the American Political Science Association, Washington, DC, September. Warren, Mark E. 2004. “What Does Corruption Mean in a Democracy?” American Journal of Political Science 48, no. 2 (April): 328–343. Wilson, Jeanne. 2008. “Civil Society: A Russian Variant.” Open Democracy Russia, 17 June. www.opendemocracy.net/article /russia-theme/civil-society-a-russian-variant, accessed 28 January 2011. Wilson, Sven E., and Daniel M. Butler. 2007. “A Lot More to Do: The Sensitivity of Time-Series Cross-Sectional Analysis to Simple Alternative Specifications.” Political Analysis 15, no. 2 (Spring): 101–123. World Bank. 2001. Diagnostic Surveys of Corruption in Romania. Washington, DC: World Bank. http://siteresources.worldbank .org/intwbigovantcor/resources/roamniasurveyreroprt.pdf, accessed 27 March 2010.

Bibliography

291

World Bank. 2006a. “Administrative Capacity in the New Member States: The Limits of Innovation?” Report No. 36930-GLB. Washington, DC: World Bank, Poverty Reduction and Economic Management Unit, Europe and Central Asia. http://site resources.worldbank.org/INTECA/Resources/EU8_Admin Capacity_Dec06.pdf, accessed 8 April 2010. World Bank. 2006b. Doing Business in 2006: Creating Jobs. Washington, DC: World Bank. www.doingbusiness.org. World Development Indicators. Various years. Washington, DC: World Bank. World Values Survey. 2000. Official data file v. 20090914, World Values Survey Association. www.worldvaluessurvey.org. Aggregate File Producer: ASEP/JDS, Madrid. Xin, Xiaohui, and Thomas K. Rudel. 2004. “The Context for Political Corruption: A Cross-National Analysis.” Social Science Quarterly 85, no. 2 (June): 294–309. Zlatkov, Dimitar. 2007. BG Oligarchs: Secrets of the Richest Bulgarians. Sofia: New Media Group. Zlatkov, Dimitar. 2008. “BG Oligarsi 2: Nedosegaemite Miliarderi” (BG Oligarchs 2: The Untouchable Billionaires). Sofia: New Media Group.

Index

Accountability, 16, 182, 238; electoral, 60, 241, 244; government and, 55, 227; mechanisms of, 30–31, 48, 65, 170; party finance and, 76; principle of, 14, 182, 190, 227, 238 Administrative reform, 18, 129–131, 134–135, 154, 156–157, 192, 242 Albania, 12; anticorruption laws in, 172, 182, 242; civil society in, 139; corruption in, 14, 28, 125, 177, 219; ethnicity in, 142; EU integration, 156, 163, 166, 170, 172; modernization levels, 136; party finance in, 79; political trust in, 197; state administration, 129, 134; voter mobilization in, 214 Anticorruption, 7, 135, 154; laws, 169, 173, 177; policies, 33, 39, 60–61, 130, 149, 155–156, 162, 166, 169–170; strategies, 134

BEEPS. See Business Environment and Enterprise Performance Survey Black market, 26 Bribery, 23, 27, 125, 128, 196–197; causes of, 12, 131, 136, 148; consequences of, 190; culture and, 38–39, 128; during communism, 25–26; in definitions of corruption, 7; prevention from, 144, 157–158; previous research on, 3, 5, 37–38 BSP. See Bulgarian Socialist Party Bulgaria, 95–123; anticorruption laws, 158, 169–170, 172; civil service, 130–132, 148; economy, 102; elections in, 216, 219, 245; ethnicity in, 113; EU integration of, 108, 161–162, 166–168, 171, 174–175, 182; government, 87, 100; informal networks in, 13, 17, 116–118; media, 95; NGOs in, 5, 138, 150; Olymp, 99, 104–106, 118;

293

294

Index

Orion, 99, 101–104, 117, 118; parties in, 68, 116–117, 246; party finance, 72, 79–81, 85, 88; political protection, 89; political trust, 240; vote buying, 68 Bulgarian Socialist Party (BSP), 89, 99–100, 118; corruption and, 114; in a coalition cabinet, 113, 115–116, 176; organizational capacity, 117; Orion and, 101–104 Bureaucracy, 25, 49, 127, 131, 242 Business Environment and Enterprise Performance Survey (BEEPS), 9–10, 81–83, 88, 92 Business interests, 20 Capacity, 9, 17, 126, 130, 173; administrative, 125, 129; economic, 33–34, 49, 55, 179, 181; financial, 132, 180; institutional, 133–136, 145–146, 161, 170; material, 130, 132, 236; nonstate, 137–138, 150–151, 238; organizational, 117, 246; state, 61, 97, 127, 129, 143, 151; technical, 136, 145 Civic Platform (Poland), 85 Civil service, 175, 236; employees, 148; legislation, 164; reform, 33, 129–130, 136, 143, 145–146, 242; trust in, 202 Civil society, 2, 4–5, 25, 127–128; anticorruption role of, 138, 143, 145, 149–150, 237; culture of, 149; during

communism, 68, 146, 164; strength of, 143, 145–146, 213; trust and, 188 Clientelism, 25, 41, 63, 116, 142 Coalition 2000, 5, 8, 138, 150, 164 Communism, 40, 89, 146, 191; collapse of, 38, 41, 129; corrupt exchange during, 25; culture shaped by, 39, 148–149, 164, 194, 204; democracy after, 188–189; experience with, 203, 206, 236; legacy of, 17, 24, 28, 45, 81, 164; party loyalty during, 148 Communist party, 25, 33, 35, 40, 66, 76, 87, 91, 99, 104, 123 Conflict of interest, 20, 105, 171, 134, 156, 170–172, 182, 184 Control of Corruption Index (WBI), 9, 56, 58, 82, 133–139, 144, 180 Corruption, 1–5, 23–25, 40, 50, 59, 235, 241; actual/real, 10, 239; accusations of, 104; administrative, 14; awareness about, 149, 81, 241; causes of, 13–14, 30, 34; confidence and, 194, 204; containing, 46, 55, 61, 71, 105, 126, 136, 138, 143, 145; costs of, 36, 38; culture of, 39; customsrelated, 27, 73; definition of, 6–8, 66; democratic transition and, 15; difficulties to detect, 26; effects of, 14, 188–189, 203, 220, 227; explanation of, 16, 131; exposing, 4; fight against, 136, 143, 145, 157,

Index

159, 161–162, 167–168, 173, 178; foreign trade and, 52, 57; free press and, 34; grand, 3, 8, 138; indicators of, 83; institutions and, 30; international integration and, 36, 46, 56; legitimacy and, 240; levels of, 10, 16–17, 28, 46, 79, 163; managing, 128; measurement, 9–11, 81; organized crime and, 177; perceived, 11, 202, 241; perceptions of, 36, 51, 70, 188, 195–196, 203–205; petty, 8; political, 15, 66–67, 213, 240; practices of, 188; public debate on, 95; punishing, 134; regulatory burden and, 32; scandals, 71, 148, 240; spread of, 177, 187, 219–220, 237; systemic, 68, 245; tolerance to, 140; totalitarian legacies and, 24; trust and, 195, 201–202, 208; types of, 3, 24 Corruption Perception Index (TICPI), 9, 28–29, 51, 56, 58, 83, 144, 180 Croatia, 87–88, 150, 165, 172, 197; corruption levels in, 176–177; Democratic Centre Party, 87; EU integration of, 156, 166–168, 170 Croatian Democratic Union, 88 Culture, 16, 38–39, 81, 139–140, 146, 149; as a contextual characteristic, 40; bakshish, 164; bureaucratic, 148; civic, 193; impact/effect of, 41, 43, 149, 151, 190; importance of,

295

74, 243; legacy of, 148; of corruption, 35, 149, 203; political, 47; Protestant, 39; secular, 140–141, 143, 204; traditional, 140, 205, 208 Czech Republic, 49, 60, 131, 205, 219, 220, 227; party finance in, 75, 81 Democratic Centre Party (Croatia), 87 Democratic Left Alliance (Poland), 87 Democratic Party Saimnieks (Latvia), 90 Democratic transition, 15, 23, 35, 75, 127, 164, 194, 227 Democratization, 35, 53, 59, 166, 191, 223, 230, 236; measurement, 50; of parties, 84 Economic liberalization, 100, 177, 179–181, 239, 251 Elections, 30, 48, 66, 69, 73, 212, 227; candidates in, 84; competitive, 48; donations and, 27; founding, 213; free, 75, 191; interest in, 214, 222; local government, 213–214; parliamentary, 59, 216; participation in, 214, 218, 228; parties and, 85; party list, 31–32, 49, 223, 246; plurality (single-member district), 31; utility of, 211–212, 215 Electoral systems, 53, 223, 244; effect of, 18; elements of, 31, 46, 48; majoritarian rules, 32,

296

Index

59, 164, 234, 245; multimember district, 46; proportional representation (PR), 30, 71, 223; reform of, 164, 245 Elite integrity, 57, 135, 151, 203–204, 208, 240; breaches in, 39; deficiencies in, 40; lack of, 14, 26; levels of, 146 Elites, 24, 35, 70, 75, 92, 214, 229, 241; corrupt, 53, 76, 187; institutional change and, 126, 130, 245; EU integration and, 36, 50, 240; minority, 141; misuse of power by, 227; political, 9, 16, 60, 89–90, 96; ruling, 96, 98, 147, 164–165 Embezzlement, 23, 105, 115, 182, 191, 238 Estonia, 60, 138, 205; anticorruption laws, 7, 134; freer from corruption, 14, 28, 48, 236; institutional trust, 192, 203, 228; Res Publica, 85 Ethnic, 53, 112, 142, 164–165, 167–168; conflict, 26, 177; divisions, 127; fragmentation, 138, 142, 145, 166; groups, 35, 141–142, 144, 165; homogeneity, 141; party, 107, 113, 141–142; Turks, 95, 112 European Union (EU), 19, 36, 46, 169–170, 172, 181–183, 239; accession, 17, 35, 41, 46–47, 129, 160–162, 165–166, 239; aid, 27, 111, 114, 162, 175–176; as a factor, 37,

58–61, 155, 159, 181; association agreement status, 54–55; candidate status, 36, 170; conditionality, 160–163, 172, 181–182; corruption in, 36; leverage, 37, 162, 173; integration, 5, 58, 108, 155–156; principles of government in, 50; Extortion, 25, 27, 33, 37, 52, 60, 67, 72, 81, 125, 142, 148 France, 71, 114 Fraud, 15, 36, 38, 105, 114, 138, 157, 172, 182, 190 Freedom House, 5, 13, 50, 244 Friendly circles, 17, 68, 92, 95–96, 98–99, 116, 253 Georgia, 54 Germany, 71, 101, 114 Globalization, 68 GRECO. See Group of States Against Corruption Greece, 28, 101 Group of States Against Corruption (GRECO), 157–158 Hungary, 81, 88, 90, 139, 147, 160; economy, 148; media law, 238; political trust, 228 Illegal enrichment, 15, 32, 49, 65, 92, 144, 179 IMF. See International Monetary Fund IMRO. See Internal Macedonian Revolutionary Organization Institutional blurring, 25

Index

Institutionalization, 97, 230, 246 Institutional trust, 199–200, 202 Institutions, 2–3, 30, 182, 191–195, 240, 245; anticorruption, 135, 138; democratic, 35, 40, 54, 167, 178; design of, 245; electoral, 29, 59; international, 9, 55, 156, 173, 178; national, 9, 54,70, 129, 165, 188, 202, 227; performance of, 89, 194, 208; political, 11, 195; representative, 23, 188; role of, 81, 242–243; trust in, 11, 23, 89, 163, 194–195, 198–199, 203–205 Internal Macedonian Revolutionary Organization (IMRO), 88–89 International integration, 156, 181, 238 International Monetary Fund (IMF), 132–133, 154 Italy, 28 Latin America, 26–27, 131 Latvia, 35, 75, 85, 134, 142, 205; anticorruption laws, 12, 134; civil service, 130, 138, 147; Democratic Party Saimnieks, 90; elections in, 60, 138; ethnicity in, 35, 42; party finance, 80–81; political trust, 195–196, 288 Latvia’s Unity Party, 90 Law, 128, 169–173, 177, 235, 238, 240; compliance with the, 30, 79; election, 46, 70, 75, 245; enforcement, 34, 39,

297

86, 92, 235, 239; equality before the, 34, 240; on disclosure, 134, 170–172; on monetary laundering, 170–171, 173; rule of, 36–37, 50, 89, 159; violation of, 17, 61, 86, 89, 103, 133 Legitimacy, 15, 18, 160, 183, 188–189, 240, 242 Liberal Democratic Party (Slovenia), 85 Lithuania, 79, 90, 138, 147, 148, 197, 214, 227, 236; anticorruption laws, 12; civil service, 138, 147–148; economy, 236; elections in, 79, 82, 138; party finance, 79; political trust, 90, 197, 214, 227–228 Lobbying, 88, 106, 109 Macedonia, 28, 55, 79, 79, 141, 156, 163; anticorruption laws, 12, 125, 170, 177; economy, 129; ethnicity in, 165, 168; EU integration of, 156, 163, 165–166, 172, 179; parties in, 88; party finance, 77, 79, 82, 88; political trust, 164, 228 Market reform, 45, 109, 178, 181, 211–212 Measurement, 5, 11, 49, 54, 140, 221, 249 Moldova, 54 Movement for Rights and Freedoms (MRF), 95, 107–113, 115, 116, 118 Multigrup, 99–101, 103–105, 108, 111, 116–118

298

Index

National Movement Simeon II (NMSS), 85, 107–110, 113, 118, 176 Nations in Transit, 5; Corruption Index (NIT), 9, 11, 176, 180 Networks, 26, 37–39, 57, 204, 208; clientelistic, 141, 156; corrupt, 12, 86, 164, 168; informal, 60, 92, 235; Mafiastyle, 3; private-public, 18, 235; exchange, 23, 26. See also Bulgaria, informal networks in NIT. See Nations in Transit, Corruption Index NMSS. See National Movement Simeon II Nomenklatura, 25, 27, 40, 104, 106, 108, 194, 241 OECD. See Organisation for Economic Co-operation and Development Olymp (Bulgaria), 99, 104–106, 118 Organized crime, 19, 211, 26, 33, 110, 168, 173, 177 Organisation for Economic Co-operation and Development (OECD), 37, 127, 150, 157–159, 169, 181, 255 Orion (Bulgaria), 99, 101–104, 117, 118 Parties (political), 10, 31, 73, 80, 246; accountability of, 72; as business firms, 69; cartelization of, 84–85; contributions to, 67, 69, 74,

80–81; ethnic, 141; founding, 157, 169; in government, 70, 88, 216; institutionalization of, 30, 68, 211, 229; new, 31, 65; noncommunist, 71, 76; number of, 223; opposition, 32, 75, 219–220; parliamentary, 70, 78; successor, 68; trust in, 89, 191, 195–196, 202, 217–218, 227–228 Party finance, 10, 17–18, 66–77, 79–84, 86, 88–89, 91–92, 242–243; caps on contributions, 72, 235; culture of, 74–75, 81; disclosure/transparency requirements, 72; opportunities for, 73–74; public procurement and, 88; sanctions in, 72–73, 91; state subsidies, 71–72 Patronage, 25–26, 28, 39, 41, 70, 80, 88–89, 117 PHARE. See Poland and Hungary: Assistance for Restructuring Their Economies Pluralism, 98, 106, 142, 158, 214, 224 Poland, 38, 50, 52, 54, 141, 205, 219, 237; anticorruption laws, 158; Civic Platform, 85; civil service, 147–148; Democratic Left Alliance, 87; elections in, 68, 138, 214, 216–217, 219; ethnicity in, 142, 205; EU integration of, 52, 158, 160; NGOs in, 150; party finance, 79, 82, 85, 88;

Index

political trust, 90, 195, 228; Self Defense Party, 90 Poland and Hungary: Assistance for Restructuring Their Economies (PHARE), 52–54, 57–58, 61, 167–168, 250 Police, 25, 150, 176, 195–197, 201–203, 251 Policymaking, 86, 128, 138, 225, 243; anticorruption, 169–170; democratic/representative, 48, 169; dishonest, 202, 205, 208, 212; influence over, 87, 97; process, 107, 127, 217; transparent, 39, 50, 55, 126, 128, 159 Political competition, 4–5, 45–46, 48, 66, 75, 130, 164, 241 Political efficacy: 217–218, 220, 223, 242 Politicians, 17, 23, 30, 48, 189, 217, 235–236, 240; businesspeople and, 96, 116–117; campaign finance and, 67, 69, 73; changing the rules by, 245; corrupt, 11, 66–67, 193, 98, 216, 229; intentions, 75, 126, 192, 215; elections and, 48, 70, 84, 87; pubic perceptions of, 32, 207–208, 218, 246; voters and, 141, 212, 226, 241, 245 Politics, 56, 68, 95, 118, 135–137, 235; above-party, 109; attitudes toward, 217, 227; citizens and, 211; clean, 28, 66, 90; competitive, 70, 84, 222; disengagement from, 215, 227, 233, 240;

299

effect of, 134; interest in, 200–202, 207; participation in, 212, 215; postcommunist, 16, 29, 160, 173, 218; Power, 8, 29, 66–67, 99, 113, 129, 160, 148; abuse of, 3, 6, 30, 37, 136, 151, 170, 177, 181, 229, 233; access to, 243; delegated, 208, 215; political, 25, 110; public, 148 Private interests, 17, 27, 69, 81, 86, 92, 156, 165, 235, 246 Privatization: 27, 41, 86, 88, 103–105, 138, 164, 238; hidden, 98; large-scale, 73, 108, 112; opportunities and, 92, 240; postponed, 115; vouchers, 102 Proportional representation (PR), 30, 71, 223 Property rights, 26–27, 97 Public confidence, 188, 202, 240; definition of, 200; as institutional trust: 195, 197, 204, 208, 228 Public officials, 46, 49, 55, 72, 118, 125, 140, 216; abuse of power and, 30, 216; asset disclosure by, 170, 172; bribing of, 27–28, 38, 61, 243; compensation of, 17, 34, 126–128, 131–132, 145–146, 148; in corrupt exchanges, 25, 55, 87, 116, 140, 187; integrity of, 5, 151, 208, 222; irresponsible, 242, 216; loyalty of, 26; trust in, 188, 192, 194–197 Public opinion, 4, 10–11, 131, 148, 189, 191, 199

300

Index

Public procurement, 60, 70, 74, 80, 88, 96, 170, 172, 216 Public trust, 1, 31, 72, 187, 190–191, 208, 216, 228, 240–241, 246 Rational choice, 12, 18, 38, 70, 160, 242 Rent seeking, 1, 32, 104, 106, 165 Responsiveness, 240, 247 Res Publica (Estonia), 85 Right to information, 50, 72, 79, 90, 127, 136, 138, 193, 216, 224; legislation on the, 34–35, 170, 172, 182 Romania, 12, 35, 38, 59, 203, 236, 240; anticorruption laws, 157, 167–168; civil service, 130; economy, 236; elections in, 59, 219, 221, 223; ethnicity in, 141; EU integration of, 156, 161–163, 165–176, 179, 182; NGOs in, 138, 150; party finance, 77– 78, 82, 88; political trust, 90, 195–197, 203–204, 222, 245 Russia, 28, 35, 39, 48, 50, 54, 68, 147–150, 236, 243, 246; anticorruption laws, 134, 169; economy, 38.148; elections in, 68, 76–77, 82, 212, 218, 222; government, 132, 147; informal networks in, 100–101, 115; NGOs in, 138, 150, 238; party finance, 74, 76, 87; political trust, 90, 149, 151, 195; vote buying, 87

Scandals, 80, 91, 148, 193, 216, 238, 240 Self Defense Party (Poland), 90 Serbia, 155–183 Serbia and Montenegro, 54 Slovak Democratic and Christian Union, 90 Slovakia, 35, 90, 195, 197, 237, 246; civil service, 130–131; 147–148; elections in, 214; EU integration of, 158; party finance, 79, 82; political trust, 227; Smer Social Democracy, 90 Slovenia, 28, 33, 48, 90, 129, 136; civil service, 33, 129, 132, 164; economy, 50; elections in, 90, 222, 245; EU integration of, 160; Liberal Democratic Party, 85; party finance, 71, 79, 85–86; political trust, 90, 195, 205 Smer Social Democracy (Slovakia), 90 Smuggling, 100 Socialization, 36, 160, 162 Spain, 71 State capture, 14, 86, 88, 99, 106, 168, 173, 177 Surveys, 9–10, 89, 138, 222, 227 TICPI. See Transparency International, Corruption Perception Index Totalitarianism, 2, 191, 193 Transparency International, 3, 5–6, 28, 150; Corruption Perception Index (TICPI), 9,

Index

28–29, 51, 56, 58, 83, 144, 180; Global Corruption Barometer, 89–90 UDF. See Union of Democratic Forces Ukraine, 28, 48–50, 68, 76, 87, 141, 197, 243; civil service, 131, 134–135, 147, 149; economy, 50; NGOs in, 138; party finance, 68, 76, 243; political trust, 87 Union of Democratic Forces (UDF), 98–99, 103, 105, 219; government, 101, 108, 116; Olymp and, 104–106 United Nations Convention Against Corruption, 157 Voter abstention, 212, 216–218, 229, 241 Voter alienation, 89, 92

301

Voter mobilization, 212, 214, 216, 219, 229, 242 Voter turnout, 213–215, 222–223, 229–230, 241; decline in, 214–215, 218; factors of, 220, 223, 225; malfeasance and, 229 Voting impact of corruption, 211–230; measurement of, 221–226 World Bank, 3, 5–6, 9, 14, 28, 46, 81, 127, 157–158, 170; Control of Corruption Index (WBI), 9, 56, 58, 82, 133– 139, 144, 180; World Development Indicators, 13, 132 World Values Survey, 10–11, 39, 149, 192, 194–196, 201, 203, 206–207 Yugoslavia, 38, 109

About the Book

Why has political corruption emerged as a major obstacle to successful democratic consolidation in Eastern Europe? Exploring the origins, scope, and impact of political corruption in the region’s postcommunist states, Tatiana Kostadinova identifies the factors that favor illicit behavior and considers how the various forms of malfeasance are threatening democracy. Rich cross-national data and an in-depth study of Bulgaria reveal how parties and their leaders have exploited the transitional environment for private benefit, as well as how domestic and external forces—including the EU integration process—are constraining corruption. Kostadinova’s conclusions point to the policy-relevant implications of her study for Eastern Europe, and for transitional democracies around the world. Tatiana Kostadinova is associate professor of political science at Florida International University. She is author of Bulgaria 1879–1946: The Challenge of Choice.

303