Corruption, Development and the Environment
Lorenzo Pellegrini
Corruption, Development and the Environment
1 3
Dr. Lorenzo Pellegrini International Institute of Social Studies (ISS) Kortenaerkade 12 2518 AX The Hague The Netherlands
[email protected]
ISBN 978-94-007-0598-2â•…â•…â•…â•… e-ISBN 978-94-007-0599-9 DOI 10.1007/978-94-007-0599-9 Springer Dordrecht Heidelberg London New York Library of Congress Control Number: 2011920998 © Springer Science+Business Media B.V. 2011 No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Cover design: deblik, Berlin Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
Acknowledgements
The intellectual contribution of Reyer Gerlagh is gratefully acknowledged. Reyer has been a stimulating PhD supervisor and the co-author of four articles that—in revised versions—are included in this book.
v
Contents
1 I ntroduction ����������������������������������尓������������������������������������尓��������������������������� 1 1.1╅Background ����������������������������������尓������������������������������������尓�������������������� ╅ 1 1.2╅Economics of Corruption ����������������������������������尓����������������������������������� ╅ 3 1.2.1╅Corruption and the Environment ����������������������������������尓������������ ╅ 5 1.3╅Research Questions and Outline ����������������������������������尓������������������������ ╅ 6 References ����������������������������������尓������������������������������������尓������������������������������� ╇ 10 2 E conomic Analysis of Corruption ����������������������������������尓��������������������������� ╇ 13 2.1╅Introduction ����������������������������������尓������������������������������������尓�������������������� ╇ 13 2.2╅The Definition of Corruption ����������������������������������尓����������������������������� ╇ 14 2.3╅Corruption and the Private Sector ����������������������������������尓���������������������� ╇ 19 2.3.1╅Perceptions of Corruption ����������������������������������尓���������������������� ╇ 20 2.3.2╅Opportunities for Corruption and Anti-corruption Strategies ����������������������������������尓������������������������������������尓������������ ╇ 21 2.3.3╅Corruption and Market Failures ����������������������������������尓������������� ╇ 23 2.4╅The Measurement of Corruption ����������������������������������尓������������������������ ╇ 24 2.5╅Discussion ����������������������������������尓������������������������������������尓���������������������� ╇ 25 References ����������������������������������尓������������������������������������尓������������������������������� ╇ 26 3 C auses of Corruption: A Survey of Cross-Country Analyses and Extended Results ����������������������������������尓������������������������������������尓����������� ╇ 29 3.1╅Introduction ����������������������������������尓������������������������������������尓�������������������� ╇ 29 3.2╅Theories of Determinants of Corruption ����������������������������������尓������������ ╇ 33 3.2.1╅Historical Roots of Corruption ����������������������������������尓��������������� ╇ 34 3.2.2╅Contemporary Causes of Corruption ����������������������������������尓������ ╇ 35 3.3╅Data Sources on Corruption ����������������������������������尓������������������������������� ╇ 37 3.4╅Empirical Analysis ����������������������������������尓������������������������������������尓��������� ╇ 38 3.5╅Limits and Interpretation of Econometric Cross-Country Analysis ����������������������������������尓������������������������������������尓�������������������������� ╇ 45 3.6╅Discussion and Conclusions ����������������������������������尓������������������������������� ╇ 47 Appendix ����������������������������������尓������������������������������������尓��������������������������������� ╇ 48 References ����������������������������������尓������������������������������������尓������������������������������� ╇ 49 vii
viii
Contents
4 T he Effect of Corruption on Growth and Its Transmission Channels ����������������������������������尓������������������������������������尓������������������������������� ╅ 4.1╅Introduction ����������������������������������尓������������������������������������尓������������������ ╅ 4.2╅Cross-Country Growth Regressions ����������������������������������尓����������������� ╅ 4.3╅Transmission Channels for Corruption ����������������������������������尓������������ ╅ 4.3.1╅The Investment Transmission Channel ����������������������������������尓 ╅ 4.3.2╅The Schooling Transmission Channel ����������������������������������尓�� ╅ 4.3.3╅The Trade Openness Transmission Channel �������������������������� ╅ 4.3.4╅The Political Violence Transmission Channel ������������������������ ╅ 4.4╅Direct and Indirect Effects of Corruption ����������������������������������尓�������� ╅ 4.5╅The Long-Term Effect of Corruption on the Transmission Variables ����������������������������������尓������������������������������������尓����������������������� ╅ 4.6╅Conclusions ����������������������������������尓������������������������������������尓������������������ ╅ Appendix 1: Long-Term Income Effects ����������������������������������尓������������������ ╅ Appendix 2: Robustness Checks ����������������������������������尓������������������������������� ╅ Appendix 3: Data ����������������������������������尓������������������������������������尓������������������ ╅ References ����������������������������������尓������������������������������������尓����������������������������� ╅ 5 C orruption, Democracy, and Environmental Policy: An Empirical Contribution to the Debate ����������������������������������尓������������������ ╅ 5.1╅Introduction ����������������������������������尓������������������������������������尓������������������ ╅ 5.2╅Theoretical Hypotheses and Their Empirical Testing ������������������������ ╅ 5.2.1╅Democracy and the Environment ����������������������������������尓��������� ╅ 5.2.2╅Corruption and the Environment ����������������������������������尓���������� ╅ 5.2.3╅Institutions and the Environment ����������������������������������尓��������� ╅ 5.3╅The Data ����������������������������������尓������������������������������������尓����������������������� ╅ 5.4╅Empirical Results ����������������������������������尓������������������������������������尓��������� ╅ 5.4.1╅The Case for Interaction Between Democracy and Income Variables ����������������������������������尓���������������������������������� ╅ 5.5╅Implications for the Environmental Kuznets Curve ��������������������������� ╅ 5.6╅Conclusions ����������������������������������尓������������������������������������尓������������������ ╅ Appendix 1 ����������������������������������尓������������������������������������尓���������������������������� ╅ Appendix 2: Data ����������������������������������尓������������������������������������尓������������������ ╅ References ����������������������������������尓������������������������������������尓����������������������������� ╅
53 53 57 60 61 62 62 63 63 65 67 68 69 72 73 75 75 77 77 79 80 82 85 90 92 94 95 96 96
6 C orruption and Environmental Policies: What Are the Implications for the Enlarged EU? ����������������������������������尓����������������������� ╇ 101 6.1╅Introduction ����������������������������������尓������������������������������������尓������������������ ╇ 101 6.2╅Determinants of Environmental Policy Stringency ���������������������������� ╇ 105 6.2.1╅Cross-Country Evidence ����������������������������������尓���������������������� ╇ 106 6.3╅Environmental Policies and Institutions in the EU ���������������������������� ╇ 111 6.3.1╅The Accession Process and Its Review: Focus on Corruption ����������������������������������尓������������������������������������尓��� ╇ 111 6.3.2╅New and Old Member States of the EU: Environmental Policies and Corruption Levels ����������������������������������尓������������ ╇ 113
Contents
ix
6.4╅Conclusions ����������������������������������尓������������������������������������尓������������������ ╇ 115 Appendix: Data ����������������������������������尓������������������������������������尓��������������������� ╇ 116 References ����������������������������������尓������������������������������������尓����������������������������� ╇ 118 7 T he Rule of the Jungle in Pakistan: A Case Study on Corruption and Forest Management in Swat ����������������������������������尓������� ╇ 121 7.1╅Introduction ����������������������������������尓������������������������������������尓������������������ ╇ 122 7.2╅Institutions and Forest Management ����������������������������������尓���������������� ╇ 124 7.2.1╅Corruption and Forestry ����������������������������������尓����������������������� ╇ 124 7.2.2╅Methodological Framework ����������������������������������尓����������������� ╇ 127 7.3╅A Profile of Swat, Pakistan ����������������������������������尓������������������������������� ╇ 129 7.3.1╅Geography and Ecology ����������������������������������尓����������������������� ╇ 129 7.3.2╅History and Institutions ����������������������������������尓������������������������ ╇ 130 7.4╅Alternative Forest Management Regimes in Swat ����������������������������� ╇ 132 7.5╅Deforestation in Swat, Pakistan ����������������������������������尓����������������������� ╇ 133 7.5.1╅Incentives and Institutions over the Forests of Swat: Focus on Corruption ����������������������������������尓����������������������������� ╇ 134 7.6╅Strategies for Institutional Change ����������������������������������尓������������������� ╇ 136 7.6.1╅The Reform Process of the Early 2000s ��������������������������������� ╇ 137 7.6.2╅Institutional Reforms ����������������������������������尓���������������������������� ╇ 139 7.6.3╅Paths of Change ����������������������������������尓������������������������������������尓 ╇ 140 7.7╅Conclusions ����������������������������������尓������������������������������������尓������������������ ╇ 144 Appendix: Sources of Information in the Study ����������������������������������尓������� ╇ 145 References ����������������������������������尓������������������������������������尓����������������������������� ╇ 145 8 C onclusions and Future Research ����������������������������������尓������������������������� ╇ 149 8.1╅Main Conclusions ����������������������������������尓������������������������������������尓�������� ╇ 149 8.2╅Policy Conclusions ����������������������������������尓������������������������������������尓������� ╇ 151 8.3╅Future Research����������������������������������尓������������������������������������尓������������� ╇ 153 References����������������������������������尓������������������������������������尓������������������������������ ╇ 156 Author Index����������������������������������尓������������������������������������尓������������������������������ ╇ 157 Subject Index ����������������������������������尓������������������������������������尓����������������������������� ╇ 159
List of Figures
1.1↜渀̕Overview of the issues analysed in the Chaps.€3–6 (the chapters based on econometric analysis)����������������������������������尓��������� ╇╅ 7 3.1↜渀̕A higher score (a darker colour) indicates higher corruption perception. Our elaboration on data from Kaufmann et€al. (2005)����������� â•… 31 4.1↜渀̕Growth rate of per capita income, in the period 1980–2004, versus corruption, in the period 1980–1985. Growth rates are corrected for initial income effect as in regression (1)����������������������������� â•… 60 6.1↜渀̕Environmental Regulatory Regime Index (↜ERRI). A darker colour indicates more stringent environmental policies. Only countries that are EU members and for which the ERRI is available are included in the map����������������������������������尓��������������������������� ╇ 102 6.2↜渀̕GDP per capita (measured in 1997 in Euro adjusted for purchasing power parity). A darker colour indicates higher GDP levels. Only countries that are EU members and for which the ERRI is available are included in the map����������������������������������尓�������������� ╇ 104 6.3↜渀̕Scatter plot for corruption and the Environmental Regulatory Regime Index. The regression line is estimated to fit all countries in the sample, not only EU countries����������������������������������尓�������������� ╇ 109 6.4↜渀̕Corruption Perception Index 2001. A darker colour indicates higher perceived corruption. Only countries that are EU members and for which the ERRI is available are included in the map����������� ╇ 112 6.5↜渀̕Scatter plot with the ERRI (adjusted for income) on the y-axis and corruption on the x-axis����������������������������������尓����������������������������������� ╇ 114
xi
List of Tables
3.1↜渀̀ Regressions results����������������������������������尓������������������������������������尓������������ ╅ 39 3.2↜渀̀ Descriptive statistics����������������������������������尓������������������������������������尓��������� ╅ 48 4.1↜渀̀ Growth regressions as in Eq.€4.1����������������������������������尓�������������������������� ╅ 59 4.2↜渀̀ Indirect transmission channels as in Eq.€4.2����������������������������������尓��������� ╅ 62 4.3↜渀̀ Growth regressions as in Eq.€4.3����������������������������������尓�������������������������� ╅ 64 4.4↜渀̀ Relative importance of transmission channels, as in Eq.€4.3������������������ ╅ 65 4.5↜渀̀ȕLong-term effects of corruption on transmission variables, as in Eq.€4.4����������������������������������尓������������������������������������尓���������������������� ╅ 66 4.6↜渀̀ Indirect transmission channels calculated as in Eq.€4.5�������������������������� ╅ 67 4.7↜渀̀ Growth regressions as in Eq.€4.1����������������������������������尓�������������������������� ╅ 70 4.8↜渀̀ Indirect transmission channels as in Eq.€4.2����������������������������������尓��������� ╅ 71 4.9↜渀̀ Growth regression as in Eq.€4.3����������������������������������尓���������������������������� ╅ 71 4.10↜渀 Relative importance of transmission channels as in Eq.€4.3������������������� ╅ 72 5.1↜渀̀ Results from statistical analysis as in Eq.€5.1����������������������������������尓������� ╅ 86 5.2↜渀̀ Results from statistical analysis as in Eq.€5.1����������������������������������尓������� ╅ 89 5.3↜渀̀ Results from statistical analysis as in Eq. 5.1����������������������������������尓������� ╅ 91 5.4↜渀̀ Results from statistical analysis as in Eq.€5.2����������������������������������尓������� ╅ 95 6.1↜渀̀ Regressions as in Eq. 6.1����������������������������������尓������������������������������������尓�� ╇ 108 6.2↜渀̀ Correlations����������������������������������尓������������������������������������尓����������������������� ╇ 110 6.3↜渀̀ Descriptive statistics for EU countries����������������������������������尓����������������� ╇ 115 6.4↜渀̀ Data for EU countries����������������������������������尓������������������������������������尓������� ╇ 117 6.5↜渀̀ Descriptive statistics����������������������������������尓������������������������������������尓��������� ╇ 117
xiii
Chapter 1
Introduction
Abstract╇ The purpose of this book is to explore, in a broad perspective, how corruption affects economic growth and environmental protection. Other aspects of development, such as education and inequality in entitlements to natural resources, will also be considered within the context of the main objectives. The introductory chapter provides an overview of economic approaches to corruption, of the relationship of corruption with environmental policies, the research questions and the book’s outline. Keywords╇ Corruption in history • Pervasiveness • Economics
1.1 Background Moral condemnation of corruption has a long history. Take the case of King David and Bathsheba, as told in the Bible (1 Samuel 11:1–28). During the Israeli siege of the city of Rabbah, David—King of Israel—falls in love with Bathsheba, a married woman. Her husband, Uriah the Hittite, is a trustworthy member of the army. David sends Uriah to the general with a sealed letter carrying the message that Uriah is to be employed in the front line where the fighting is the fiercest to get him killed. After Bathsheba’s mourning for her husband is over, David brings her to his house to be his wife. The biblical story continues with a parable that makes clear how much the Lord was displeased by David’s abuse of power, and the child born out of the union of Bathsheba and David dies. David abused the rulers’ powers and his
L. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9_1, ©Â€Springer Science+Business Media B.V. 2011
1
2
1 Introduction
actions fit in a popular modern definition of corruption: “the misuse of entrusted power for private benefit” (from Transparency International).1,â•›2 In another historical text, corruption is depicted as a constant feature of government servants. In Kautilya’s Arthasastra (fourth century B.C.) it is stated that “just as it is impossible not to taste the honey or the poison that finds itself at the tip of the tongue, so it is impossible for a government servant not to eat up, at least, a bit of the king’s revenue. Just as fish moving under water cannot possibly be found out either as drinking or not drinking water, so government servants employed in the government work cannot be found out (while) taking money (for themselves)”. The gloomy message from the Arthasastra is that corruption is intrinsic in government’s activities and we cannot expect to be able to single out those civil servants that engage in it. Recently, corruption has been the focus of much scientific research for its undesired welfare effects. Since Myrdal’s influential critique of “the diplomacy of research” (1968, p.€939)—which in his opinion caused the subject to be overlooked— studies have scrutinized the sources of corruption and related many aspects of (under-) development with corruption. Overall, social sciences have dealt with the definition of corruption, its determinants and its effects. With respect to the definitions of corruption, a discussion is ongoing over the exact characterization of phenomena that can be called “corruption”. Dictionary definitions of corruption refer to moral decay and express incontrovertible condemnation. In the social sciences the exact meaning of corruption has been determined according to disciplinary lines and, quite remarkably, many economists have been abstaining themselves from the discussion. As a result, there are many economic analysis of corruption where the exact meaning of corruption is left unclear (Williams 1999; cf. Lambsdorff 2007). An associated issue is the measurement of corruption and its use in econometric analyses. Since the mid-1990s, making use of available statistical indexes of corruption, several authors have analysed econometrically the factors associated with corruption. The findings of these studies are not very conclusive as some found that most factors contributing to corruption can be explained by different historical trajectories (e.g. La Porta et€al. 1999; Treisman 2000), while others found variables related to contemporary institutions to be influential on corruption (e.g. Arikan 2004; Chowdhury 2004). Furthermore, economists associate the extent of corruption with many societal shortfalls: economic stagnation, inequality in the distribution of income, underin1╇ From Deuteronomy (5:14–21) we know that the King received his powers from God with the purpose of serving the Israeli Nation, and was himself subject to the law: [the King may] “not consider himself better than his brothers and turn from the law to the right or to the left” (Deuteronomy 5:20). It is apparent that a distinction between the public and the private sphere of rulers was already present in the Old Testament and that the use of power by the ruler was subject to restraints. Therefore, we can interpret this story as a case of corruption; see the discussion below on the definition of corruption. 2╇ See the definition given at http://www.transparency.org/news_room/faq/corruption_faq, 3 November 2009.
1.2 Economics of Corruption
3
vestment in human capital, inflation of defence budgets, and lack of protection for environmental resources, to mention just a few. In his econometric analysis Mauro (1995) found that corruption affects economic growth indirectly through its negative effect on investment. Since the publication of Mauro’s influential paper, researchers have tried to identify further channels through which corruption affects growth (e.g. Mo 2001); for example, finding that corruption detracts resources from sectors that promote economic development, such as education (Mauro 1998), towards sectors that do not promote welfare creation, such as military expenditures (Gupta et€al. 2001). Another aspect of public policies that has attracted attention is the effect of corruption on environmental policies. In fact, theoretical and empirical studies have found corruption to be associated with more lax environmental policies (e.g. Fredriksson and Millimet 2001; Damania et€al. 2003). Nonetheless, the condemnation of corruption among social scientists is not universal. Some authors have argued that, under certain conditions, corruption can be beneficial for economic development (Huntington 1968) and others have suggested that campaigns against corruption detract attention from more pressing problems and hinder democratic transitions (Naim 2005). The purpose of this book is to explore, in a broad perspective, how corruption affects economic growth and environmental protection. Other aspects of development, such as education and inequality in entitlements to natural resources, will also be considered within the context of the main objectives. An overview of economic approaches to corruption, of the relationship of corruption with environmental policies, the research questions and the book’s outline are provided in the following sections.
1.2 Economics of Corruption The social sciences have analysed the issue of corruption from several different perspectives. Here, we introduce some of the most prominent frameworks to study corruption (rent seeking, principal-agent, functionalist and New Institutionalist) and the economics related to each. Rent seeking—profit-seeking activities where resources are used to achieve redistribution—is an activity that depresses economic growth: economic agents deplete assets in activities that are not associated with the generation of welfare, but only with its transfer from one agent to the other (e.g. Krueger 1974; Murphy et€al. 1993). Corruption can be understood as a form of this activity: it is an exercise of public rent seeking. Through corruption in the public sector, resources are transferred from the private sector to public officials who abuse their power over the fortunes of private economic agents.3 Within the private sector, agents abuse their powers to gain “extraordinary” benefits often at the expense of the economic agent 3╇ Public rent seeking also includes activities that are not corruption strictu sensu, for example lobbying (Murphy et€al. 1993).
4
1 Introduction
or organization they work for. These exercises entail costs (transfer costs)—that are a net loss for the economy—and decrease the incentive to undertake productive activities.4,â•›5 Another economic approach that has been applied to the study of corruption is the principal−agent theory. In the case of bureaucratic corruption, the politician cannot perfectly monitor the actions of the bureaucrat (i.e. there is informational asymmetry). The bureaucrat (the agent) trespasses the rules and the objective set for her occupation in order to maximize her own private benefits. That is, the decision to engage in corruption is a part of utility maximization where monitoring efforts, penalties and incentives are decision factors. The utility function of the bureaucrat can be characterised by the existence of moral costs of corruption, risk aversion, and decreasing marginal utility of income.6 The same model can be applied to political corruption: in representative democracies, the principal would be the voters and the agent the politician. The voters have a limited control over the politician who can use her power to serve her own interests (e.g. Groenendijk 1997). Another approach—which holds currency among some political scientists—is the functionalist (relativist) one. According to functionalist theory, corruption (as any other social arrangement) arises from an institutional vacuum and serves a social function that would not be satisfied otherwise; hence it is not intrinsically condemnable (Roy 1970). As Huntington (1968, p.€69) famously put it: “the only thing worse than a society with a rigid, over centralized, dishonest bureaucracy is one with a rigid, over centralized honest bureaucracy”. Some economists who take a functionalist perspective interpret corruption as a method to introduce market mechanisms within the bureaucratic and political process (or to create a “Coasean bargaining process”, Bardhan 1997). This type of analysis is based on a second-best environment where corruption can be a way to circumvent unnecessary bureaucratic burdens and avoid delays: it provides incentives for bureaucrats that are otherwise alien to the private sector (e.g. Lui 1985). There are two important limitations to these models. First, the bureaucrat, in a highly corrupt system, has an incentive to impede the actions of private agents by creating additional bureaucratic regulations in order to extract more bribes and to apply existing ones as obtrusively as possible; in this case corruption creates bureaucratic delays rather than helping to overcome them.7 Second, the allocation of resources (such as bureaucrats’ time) through bribes is fundamentally different from a market process because there cannot be open bidding. As a result of the opaque nature of corruption agreements, apart from the size For a discussion of corruption in light of the rent-seeking literature, see Lambsdorff (2002). Bhagwati (1982) highlighted how diversion of resources from productive activities reduces social welfare, but also that in already distorted economies such activities might have an ambiguous effect. Some of the arguments put forward to substantiate the benefits of corruption (i.e. that in economies where corruption is common additional corruption can be beneficial) can be linked to this line of reasoning. 6╇ For supporting experimental evidence on intrinsic motivations (moral costs) and corruption, see Schulze and Frank (2003). 7╇ The point was made by several authors, e.g. Alam (1989); Lambsdorff (2002). 4╇ 5╇
1.2 Economics of Corruption
5
of the bribe, other factors (e.g. personal relations and trust between the parties) will influence the allocation of resources; hence efficiency cannot be achieved.8 Another approach that is gaining currency is the New Institutional economics approach to corruption (see Lambsdorff 2007).9 In the studies belonging to this school of thought the role of transaction costs in determining investment and trading opportunities is emphasized. Many corruption cases involve transactions, the clearest example being bribery where valuable goods are given (or extorted) in return for a specific action by a person in power. The parties in corruption-related transactions face the same problems in terms of possible non-compliance and opportunistic behaviour as any other party involved in a transaction. Notwithstanding these similarities, the uncertainties and risks associated with corrupt transactions are exacerbated by the fact that in most cases, corruption is a secret and illegal act and generally there is no possibility to involve a court for redress when a party does not perform according to the corruption deal. While in the standard New Institutional Economics literature transaction costs are seen as a barrier to efficiency gains and the institutional framework is meant to provide the rules for abating them, with respect to corruption the new institutional economics approach tries to identify manners in which corruption deals become more “costly” for the parties involved. The most common strategy is to manipulate the interest of the parties so that they are in contrast and opportunistic behaviour becomes more likely. In this way, mechanisms to increase the transaction costs of corrupt deals become tools in an anti-corruption strategy. Finally, the occurrence of corruption in many different settings, especially in already distorted economies or authoritarian regimes, implies that no univocal conclusion can be derived on the implications of corruption for each and every case. It might well be that in some special situations corruption has a positive effect on social welfare, or can be morally the best option. What we will be looking for in this book is a generalization of the influence of corruption on the economy and on environmental policies without the ambition of finding universal rules that hold without exception.
1.2.1 Corruption and the Environment Some authors point to the fact that the best solution to the problem of corruption is to reduce opportunities to engage in it. These authors, focusing solely on corruption in the public sector, found that the best antidote to corruption would be less involvement of the state in the economy (with less state ownership, less regulations, and less policies in general; see Lambsdorff 2007, pp.€4–6). However, 8╇ For an example applied to forestry, see Robbins (2000, p.€ 435), for a general discussion, see Bardhan (1997, p.€1322). 9╇ Some authors are also putting forward an institutional economics perspective on corruption that is based on the “traditional” institutionalist approach (see, Hodgson and Jiang 2007).
6
1 Introduction
state intervention (through regulation or other means) is important where market failures are present (Acemoglu and Verdier 2000; Banerjee 1997) and the environmental realm is a typical case: many environmental goods are public goods, or are characterized by ill-defined property rights and externalities. These features imply that environmental problems cannot be resolved solely through market forces, but that state intervention is necessary. In every instance in which the public interest is filtered by policy makers and bureaucrats (i.e. in any case of market failure correction), corruption is potentially an issue and environmental policy falls in this category. Therefore, the influence of corruption on the environmental regulatory systems deserves scrutiny. In the environmental sphere, political corruption can limit the scope of the state’s intervention to solve environmental problems, and bureaucratic corruption can limit the implementation of environmental policies (e.g. Damania 2002). Even beyond the market failure approach to environmental problems and policies, any type of collective action oriented towards the achievement of social objectives in general and environmental objectives in particular (e.g. the legislative process in elective democracies) can be obstructed by corruption. Economically and politically well-endowed interests can interfere with the ideal working of institutions mediating interests and organizing cooperation in society by capturing parts of the policy making process. A strand of literature deals with the design and implementation of environmental policies when corruption is present. For example, forestry policies in many developing countries are affected by the corruption of agents in charge of implementation coupled with the hierarchy of enforcers and controllers (e.g. Pellegrini 2011). Corruption can evolve in an institutional system where rights—in the management and distribution of benefits of natural resources—have to be surrendered in exchange for gifts, and the poor are penalised (e.g. Robbins 2000, p.€440). Finally, apart from its focus on the influence of corruption on environmental problems and solutions to them, the literature examines how natural resource abundant countries are characterised by higher corruption levels. Some authors suggest that the abundance of natural resources leads to engagement in rent-seeking and detracts resources from productive activities. Corruption is one of the tools used to appropriate rents (e.g. Leite and Weidmann 1999) and there is evidence suggesting that one of the main reasons for the little benefit that the poor derive from bounties associated with natural resource extraction is the influence of corruption in the distribution of rents (UNCTAD 2007, p.€152).
1.3 Research Questions and Outline The main aim of the book is to analyse the influence of corruption on economic growth, environmental protection and other aspects of development. To accomplish this aim we are going to look at corruption from different angles. We start a critical discussion of the concept of corruption and of its understanding in economics. First,
1.3 Research Questions and Outline
7
we show that the issues inherent in finding a coherent and universal definition of the concept at hand are non-trivial. We also highlight how specific definitions stir the discussion towards certain aspects of corruption: a misconception of corruption can lead to simplistic recipes of anti-corruption that have little relevance in reality. Furthermore, we find a working definition and we put our research in the context of the existing literature. Next, we continue with an analysis of the sources of corruption, i.e. first we scrutinize the variables associated with corruption incidence considering the latter as a dependent variable using econometric methods. Once the factors influencing corruption are analysed, we examine how corruption affects variables associated with human welfare, i.e. we consider corruption as an independent variable. First, we provide an estimate of the influence of corruption on economic growth and of the variables that work as transmission channels. Second, we scrutinize the effects of corruption on environmental policy stringency. Next, the study of corruption and environmental protection is put into a policy context and will explore the implications for the EU environmental harmonization process. The econometric analyses are followed by a case study of corruption in the forestry sector in Pakistan. Figure€1.1 is a graphical depiction of the issues studied in Chaps.€3 to 6 of this book, with an indication of the chapters addressing each subject. Below, there is an explanation of each of the research issues outlined here and an overview of how they are analysed in each chapter. Chapters€2 and 7—dealing respectively with the definition of corruption and with the role of corruption in forest management in Swat, Pakistan—inform the interpretation and complement the material presented in the other chapters. Often times, the word corruption is used to characterize a wide range of conducts and the concept becomes a synonym of “morally reproachable”. An opposite tendency is to circumscribe corruption to a set of behaviours—mostly referring to bribes—by agents belonging to the public sector. Overall, the issue of the definition of corruption is often avoided in the writings of economists, which leaves the impression that they are talking about something without being entirely clear on what the “something” really is. The definition of corruption is crucial because it determines what we are analysing (and measuring) and has a direct relation to the strategies we can adopt to reduce corruption.
Economic growth and transmission channels (Chapter 4) Sources of corruption (Chapter 3)
Fig. 1.1↜渀 Overview of the issues analysed in the Chaps.€3–6 (the chapters based on econometric analysis)
Corruption
Environmental policies (Chapter 5)
Application to the EU’s policy context (Chapter 6)
8
1 Introduction When the term corruption is used to epitomize very different phenomena and assumes a strongly morally charged connotation, corruption becomes an analytically unmanageable synonym of anything bad. On the other side of the spectrum, definitions of corruption that narrowly define the phenomenon as something occurring in the public sector run the risk of “solving” the problem of corruption simply by labelling problematic behaviours in alternative ways. How can we define corruption avoiding at the same time too broad and simplistic categorizations?
The definition of corruption is discussed in Chap.€2, where we examine different approaches to the definitional quest and discuss related concepts. Furthermore, we critically look at policy conclusions that emanate from partial definitions and discuss the measurement project. From accounts of corruption, such as the ones cited at the beginning of this chapter, it is possible to draw the conclusion that corruption is an ever-existing phenomenon and that it would be better to learn to live with it, rather than trying to address it. Nevertheless, we find that corruption affects societies to different degrees. For example, there is a large difference in the frequency with which bribes are disbursed to obtain a public service across different countries. Reliable information on illegal matters is hard to find, but survey methods and interviews help to uncover the extent of corruption. For example, interview surveys, from different countries show vast differences in the frequency and size of bribes and multinational firms devote dissimilar shares of their revenues to bribing in different countries. If corruption incidence varies so much across countries, then what are the determinants of the corruption level? What factors induce individuals in some societies to refrain from abusing their powers while in others such abuse is common?
This issue is addressed in Chap.€3, where we present the literature on the determinants of corruption levels across countries. Thanks to improved availability of data and a careful choice of variables, it is possible to test empirically several hypotheses in a large sample of countries. Previous findings that were already tested in comprehensive empirical models are re-tested jointly with more recent findings that have not received sufficient empirical scrutiny. The results are less disheartening than those from some previous studies, which found that corruption is mostly determined by variables related to the historical development of countries or to their endowments. The existence of corruption affects economic agents’ expectations and behaviours. In turn, economic agents jointly determine the economic performance of a country. Corruption has been found to directly affect the growth rate of the economy, but this relationship becomes murkier when additional variables are added to the econometric analysis. These variables can themselves be affected by corruption levels. For example, early studies (Mauro 1995) have found corruption to affect the economy through the investment channel: in an economy where corruption thrives the level of investment is depressed. Corruption, at least in the most affected countries, is a pervasive phenomenon capable of being a fundamental determinant of choices made by policy makers, bureaucrats and private economic agents. What are the transmission channels through which corruption indirectly affects the growth rate of the economy?
1.3 Research Questions and Outline
9
Chapter€ 4 produces estimates of the transmission channels through which corruption affects the growth rate. The econometric analysis focuses on investment, schooling, trade openness, and political violence. For the first time trade policies are included among the channels through which corruption affects growth and we support our analysis with a series of robustness checks and the use of instrumental variables. We also calculate the long-term effects of corruption on growth and on the transmission variables. A debate is going on over the determinants of environmental quality, where one prominent hypothesis states that the relationship between environmental quality and income would follow an inverted-U path. At low levels of income, an increase in income will lead to a decrease in environmental quality, while after a certain threshold of income, further increases in income result in higher environmental quality. This path has been observed in historical data of now-developed countries for some measures of environmental quality (Grossman and Krueger 1995; Cole 2004). This historical path has been termed the Environment Kuznets Curve, because of the similarly shaped relationship that Kuznets identified between inequality and economic growth. The hypothesis has already been challenged, and our results further strengthen existing criticisms because we find that corruption is a determinant of economic growth and of environmental policy stringency at the same time. What are the implications of high corruption levels for environmental protection? Once the effect of corruption is estimated alongside democracy (that is, in conjunction with an institutional characteristics that are correlated to corruption, but can assert independent influence on the way policies are formulated), does corruption retain its influence?
Chapter€5 analyses empirically the determinants of environmental policy stringency. First, it reviews the theoretical and empirical literature on the determinants of natural resource depletion, pollution, and environmental policy stringency. Then, it produces econometric tests on democracy and corruption as determinants of the degree of protection of the environment. The analysis presented is the first extensive study to include different aspects of governance quality (corruption and democracy) while studying their impact on general measures of environmental policy stringency. Once the influence of corruption on environmental policies is taken for granted a question arises as to the implications for policy prescriptions. The environment is a normal (or even a luxurious) good and the willingness to pay for it increases with income. Furthermore, we note that the costs of environmental policy interventions are different in different countries, depending on economic and environmental factors. Based on these considerations we might conclude that each country should be able to adjust its own environmental policies to its income levels, environmental conditions, and to the preference for environmental protection of its polity in order to obtain the socially optimal level of environmental protection. When social systems, such as the new EU members, are forced to adopt more stringent environmental policies, is there undisputed evidence that they will obtain levels of environmental protection that are above what is socially optimal, as it would appear from standard economic analysis? What are the implications of these impositions for countries marked by pervasive corruption?
1 Introduction
10
Chapter€ 6 builds on the evidence gathered in the previous chapter and puts the evidence of the effects of corruption into a policy context. The European Union enlargement process brought great opportunities, together with swift changes for “new” and “old” members. The analysis focuses on two features of new member states: low environmental protection coupled with high corruption levels (when compared to the old members). The study analyses how income and corruption contribute to the stringency of environmental policies and suggests that a new rationale can be found for the upward harmonization of environmental standards that is taking place through the accession process. Chapter€7 is a case study from Swat, Pakistan. The forest management regime in Swat is mired in corruption, which is facilitating illegal logging and the rapid depletion of forest stocks. This case study highlights the difficulties of using the standard countrywide recipes for reducing corruption, when the objective is to rapidly reform a relatively small sector and there is political unwillingness to undertake wholesale change. The analysis then turns to ways through which institutional change at the local level can reduce the incidence of corruption and improve management. When natural resources such as forest are being depleted because of the lax enforcement of the management regime and one of the mechanisms at play is corruption, what are the policy options that are available? What can we expect from the implementation of policies that increase the power of enforcement agents?
We analyse corruption against the backdrop of the reform options that are most often cited as possible solutions. As we show in this study, the “crime and punishment” (i.e. increased monitoring coupled with harsher penalties) approach is not feasible to implement if the overall institutional environment is weak. Since countrywide overhaul of corruption through sweeping reform programs, the other reform approach, is a difficult and lengthy task, there is a need for an alternative kind of reform. Chapter€8 concludes highlighting the policy relevance of the book’s findings and pertinent areas for future research.
References Acemoglu, D., & Verdier, T. (2000). The choice between market failures and corruption. American Economic Review, 90(1), 194–211. Alam, M. S. (1989). Anatomy of corruption: An approach to the political economy of underdevelopment. American Journal of Economics and Sociology, 48(4), 441–456. Arikan, G. G. (2004). Fiscal decentralization: A remedy for corruption? International Tax and Public Finance, 11(2), 175–195. Banerjee, A.-V. (1997). A theory of misgovernance. Quarterly Journal of Economics, 112(4), 1289–1332. Bardhan, P. (1997). Corruption and development: A review of issues. Journal of Economic Literature, 35(3), 1320–1346. Bhagwati, J. N. (1982). Directly unproductive, profit-seeking (DUP) activities. The Journal of Political Economy, 90(5), 988–1002. Chowdhury, S. K. (2004). The effect of democracy and press freedom on corruption: An empirical test. Economics Letters, 85(1), 93–101.
References
11
Cole, M. A. (2004). Trade, the pollution haven hypothesis and the environmental kuznets curve: Examining the linkages. Ecological Economics, 48(1), 71–81. Damania, R. (2002). Environmental controls with corrupt bureaucrats. Environment and Development Economics, 7, 407–427. Damania, R., Fredriksson, P. G., & List, J. A. (2003). Trade liberalization, corruption, and environmental policy formation: Theory and evidence. Journal of Environmental Economics and Management, 46(3), 490–512. Fredriksson, P. G., & Millimet, D. L. (2001). Bureaucratic corruption and environmental policy: Theory and evidence from the United States. Unpublished manuscript ftp://193.196.11.222/ pub/zew-docs/sw_2000/Papers/paper%20Per%20Fredriksson.pdf. Groenendijk, N. (1997). A principal-agent model of corruption. Crime, Law and Social Change, 27(3, 4), 207–229. Grossman, G. M., & Krueger, A. B. (1995). Economic-growth and the environment. Quarterly Journal of Economics, 110(2), 353–377. Gupta, S., de Mello, L., & Sharan, R. (2001). Corruption and military spending. European Journal of Political Economy, 17(4), 749–777. Hodgson, G. M., & Jiang, S. X. (2007). The economics of corruption and the corruption of economics: An institutionalist perspective. Journal of Economic Issues, 41(4), 1043–1061. Huntington, S. P. (1968). Political order in changing societies. New Haven: Yale University Press. Krueger, A. O. (1974). Political economy of rent-seeking society. American Economic Review, 64(3), 291–303. La Porta, R., Lopez-De-Silanes, F., Shleifer, A., et€al. (1999). The quality of government. Journal of Law Economics & Organization, 15(1), 222–279. Lambsdorff, J.-G. (2002). Corruption and rent-seeking. Public Choice, 113(1, 2), 97–125. Lambsdorff, J. G. (2007). The institutional economics of corruption and reform: Theory, evidence, and policy. Cambridge: Cambridge University Press. Leite, C., & Weidmann, J. (1999). Does mother nature corrupt? Natural resources, corruption, and economic growth. IMF working paper. Washington: International Monetary Fund. Lui, F. T. (1985). An equilibrium queuing model of bribery. Journal of Political Economy, 93(4), 760–781. Mauro, P. (1995). Corruption and growth. Quarterly Journal of Economics, 110(3), 681–712. Mauro, P. (1998). Corruption and the composition of government expenditure. Journal of Public Economics, 69(2), 263–279. Mo, P. H. (2001). Corruption and economic growth. Journal of Comparative Economics, 29(1), 66–79. Murphy, K.-M., Shleifer, A., & Vishny, R.-W. (1993). Why is rent-seeking so costly to growth?. American Economic Review, 83(2), 409–414. Myrdal, G. (1968). Asian drama: An inquiry into the poverty of nations. New York: Twentieth Century Fund. Naim, M. (2005). Bad medicine: The war on corruption is leaving the world worse than it found it. Foreign Policy, (147), 96–98. Pellegrini, L. (2011). Forest management and poverty in Bolivia, Honduras and Nicaragua: Reform failures? European Journal of Development Research. Robbins, P. (2000). The rotten institution: Corruption in natural resource management. Political Geography, 19(4), 423–443. Roy, E. V. (1970). On the theory of corruption. Economic Development and Cultural Change, 19(1), 86–110. Schulze, G.-G., & Frank, B. (2003). Deterrence versus intrinsic motivation: Experimental evidence on the determinants of corruptibility. Economics of Governance, 4(2), 143–160. Treisman, D. (2000). The causes of corruption: A cross-national study. Journal of Public Economics, 76(3), 399–457. UNCTAD. (2007). World investment report 2007: Transnational corporations, extractive industries and development. United Nations conference on trade and development. New York, Geneva. Williams, R. (1999). New concepts for old? Third World Quarterly, 20(3), 503–513.
Chapter 2
Economic Analysis of Corruption
Abstract╇ In this chapter, we contextualize and discuss the issue of defining corruption, endorsing a definition that focuses on the abuse of power. We also discuss how limiting the scope of corruption to the public sector contradicts the understanding that is common in the public and also in the economic science. Furthermore, we show how policies that are based on a concept of corruption that narrowly ascribes the phenomenon to the public sector, can reach simplistic conclusions such as that the extent of the public sector is invariably positively correlated with the occurrence of corruption. Such conclusions might drive policies that cannot satisfactorily address the problem of corruption, are conceptually flawed and are not backed by empirical analysis that enquired into the relation between the public sector and corruption. In this context, we also scrutinize the proposition that there is an evident trade-off between market failures and corruption. On the contrary, anticipating one of the main conclusions of this book, we argue that corruption is one of the causes of the persistence of market failures. The abuse of power for private gains in the case of environmental policies is an egregious example of how public institutions fail in dealing with issues that cannot be solved solely by market mechanisms because of corruption. Finally, the chapter also contains a discussion of basic concepts used throughout the book and of issues related to the measurement of corruption. JEL classification╇ O17 • P48 Keywords╇ Definition of corruption • Anti-corruption policies • Privatization • Private sector • Market failures
2.1 Introduction Corruption is an emotionally charged concept that inspires in most people immediate condemnation. These reactions can be channelled into supporting anti-corruption strategies to deal with the problem, but the specific strategies will depend on the definition and conceptualization of the phenomenon at hand. These definitions L. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9_2, ©Â€Springer Science+Business Media B.V. 2011
13
14
2 Economic Analysis of Corruption
can collide with what the public and international institutions have in mind when lamenting corruption and might also not fit with the analysis of some of the detrimental effects that corruption has on social welfare and with the formulation of sensible policies to deal with the problem. Here we will discuss the definition of corruption and the fallacies intrinsic in understandings of corruption that are based on one-sided definitions, focusing in particular to the questionable characterization of corruption as something that has to do only with the public sector. This specific definition is popular with economists and, in most cases covertly, it underpins the simplistic conclusion that to decrease corruption it is sufficient to shrink the public sector, privatize or that there is a dilemma between market failures and corruption (e.g. Acemoglu and Verdier 2000; cf. Hodgson and Jiang 2007). In general, definitions based on conceptual fallacies can drive policy makers to apply specific measures that would have little potential to affect corruption if differently defined.
2.2 The Definition of Corruption Corruption is a multifaceted concept that escapes monolithic characterizations. Corruption, as defined in the dictionary, epitomizes moral decay, is intrinsically bad and subject of unconditional condemnation: it is the “impairment of integrity, virtue, or moral principle”.1 In search for definitions fit for the purpose of social sciences, alternative—and arguably more morally neutral and less comprehensive—definitions have been developed. A social science approach will benefit from a definition of corruption that does not have a strong moral component because of the benefits of focusing on a set of behaviours (in this case corrupt behaviours) without being able to condemn them a priori. Morally charged definitions—for example—would make it very difficult to analyze the writings of those authors that argue that corruption satisfies societal needs and ultimately has beneficial effects on social welfare (e.g. Huntington 1968). A related issue is that the class of phenomena characterized as “corrupt” should be kept at an analytically manageable size. In many contexts morally defined corruption becomes a catch-all derogatory word and a definition of corruption that would include phenomena that are of very different natures would render such definition useless. In general, issues that are very diverse relate to different analytical tools in different ways and a definition of corruption that is too inclusive runs the risk of being meaningless. Overall, the task of identifying the definition of corruption is not an easy undertaking because of the emotions the concept inspires and the fact that several different definitions exist.2 As a result of these difficulties many social scientists “Corruption.” Meriam–Webster online, 6 November 2007. As Jain states, “one of the difficulties in studying corruption lies in defining it. While it may appear to be a semantic issue, how corruption is defined actually ends up determining what gets modelled and measured” (2001, p. 73).
1╇ 2╇
2.2 The Definition of Corruption
15
simply shy away from any explicit definition of corruption (see Williams 1999; Lambsdorff 2007, pp.€15–16) and it is often difficult to understand what exactly different authors have in mind when they discuss corruption. Many times it is only from their analysis and conclusions that is possible to appreciate what characterization of corruption they are using (e.g. see the discussion below on the public sector). The diversity of approaches to corruption underpins the diverse definitions; the earliest approach being the legalistic one (Williams 1999). According to this approach, corruption is simply defined as the breaching of legal codes—defining public duties—in order to obtain personal advantages. However, the usefulness of this definition is disputable once the prominence of power in defining the law and the indeterminacy of the legal codes is recognized. The point was forcefully made by the schools of critical legal studies and legal realism (see Hasnas 1995).3,4 The first school emphasized how powerful interests can influence the development of legal codes, use them to justify the status quo and further pursue their welfare. In light of these considerations, legal provisions are a dubious benchmark for defining corrupt acts, and the fact that certain behaviour is not defined as corruption in legal codes might be better interpreted as an indicator of the influence of powerful interests rather than that of lack of corruption. A telling example of similar issues is the case of the decree that the Italian government enacted in 1993 to decriminalize illicit financing of political parties (known in Italian as the “colpo di spugna”). The Premier Giuliano Amato rushed through a government decree decriminalizing one type of corruption in order to salvage the members of his political party from judiciary prosecution. Eventually, the measure had to be withdrawn because of the public outcry that followed its approval by the government.5 If the proponents of this legislation had succeeded, according to Italian criminal law bribing political parties in order to obtain personal advantages would not be classified as corruption 3╇ In any case, the occurrence of corruption is often sanctioned by the law; in other words, often the act of corruption is a crime. In economics, a large literature has developed starting from the seminal work of Becker (1968) on crime and punishment. The application of such literature to the case of corruption is limited on the one hand by the fact that not all corruption is illegal. On the other hand, the case of corruption is peculiar because there is evidence that in some countries enforcement agents—the institutions in charge of punishment—are the worst offenders with respect to corruption, which limits the straight application of crime models (e.g. Pellegrini 2007). The inclusion of corruption in models of economics of crime can produce predictions quite opposite to the standard prescription of increased punishment and monitoring to decrease crime (e.g. Kugler et€al. 2005). 4╇ Corruption itself can determine which and how many laws are passed and the roman historian Publius Cornelius Tacitus noted—as early as in the first century AD—that “now bills were passed, not only for national objects but for individual cases, and laws were most numerous when the commonwealth was most corrupt” (↜Annals, Book€3, Chap.€27). It is interesting to note that the current Italian prime minister—Silvio Berlusconi—has been charged of creating laws ad personam; i.e. he has been charged of creating specific laws to safeguard the specific interests of one individual. The beneficiaries of such legal measures include the prime minister himself and other close associates that now cannot be prosecuted in numerous corruption-related trials. 5╇ See Cinzia Sasso, “Borrelli e gli inediti di Craxi: Volevano frenare Mani pulite” Repubblica, 7 December 2007. http://ricerca.repubblica.it/repubblica/archivio/repubblica/2007/12/07/borrelligli-inediti-di-craxi-volevano-frenare.html, 25 May 2009.
16
2 Economic Analysis of Corruption
anymore. This case exemplifies how misleading the reference to legal codes can be: the enactment of such regulation rather than causing a decrease in corruption might better be classified as a display of institutionalized corruption. Furthermore, the school of legal realism emphasized how the interpretation of legal codes plays a crucial role in sentences and there is little possible objectivity in “the law” per se. Finally, a definition of corruption based on legal codes is problematic for comparative analysis when legal codes differ across countries and as a result we are comparing different phenomena. Just as an example of such differences in the legal characterization of corruption, in the United States, lobbying is a legal practice and enterprises can legitimately invest in it and these expenditures enter their balance sheet as any other. In most European countries the same practices fall under the rubric of corruption and are legally sanctioned. On the other hand, in the United States it is illegal for congress’ members to hire one’s spouse to help in the congressional work, while in many European parliaments that practice is common. Thus, it appears that the diversity of legal codes renders them a dubious benchmark for comparison across countries. In sum, the legal definition is seemingly clear-cut and hides the difficulties of identifying corrupt acts according to one unambiguous standard. We prefer to discuss and opt for other definitions of corruption that do not hide the ambiguities of the subject, but are clear enough to narrow down the classes of acts defined as corrupt if compared to the moralistic definitions that appear in the dictionaries as mentioned above. A simple definition of corruption as “the misuse of entrusted power for private gain” is adopted by international institutions (e.g. by the international NGO Transparency International and by the Danish Development Agency Danida).6 Such definition is a useful reference point, but the benefit of simplicity comes at the cost of some vagueness. This definition suffers from the ambiguity associated with the lack of examples and further classifications; we consider it as a good starting point to achieve a useful definition of corruption, but it needs to be complemented by a more articulated characterization and that is what we are turning to now. We will complement the basic definition with a classic one from Nye that is narrower than the moralistic one, seemingly less clear-cut than the legal one, but more operational for the purposes of economic analysis: “Corruption is behavior which deviates from the formal duties of a public role because of private-regarding (personal, close family, private clique) pecuniary or status gains; or violates rules against the exercise of certain types of private regarding influence. This includes such behaviour as bribery (use of a reward to pervert the judgment of a person in a position of trust); nepotism (bestowal of patronage by reason of ascriptive relationship rather than merit); and misappropriation (illegal appropriation of public resources for private-regarding uses)” (1967, p.€419).7 This definition offers lucid 6╇ http://www.transparency.org/news_room/faq/corruption_faq, 3 June 2009. For Danida’s definition see http://www.um.dk/en/menu/DevelopmentPolicy/AntiCorruption/, 3 June 2009. 7╇ Similar definitions are also available in the encyclopaedia: “Improper and usually unlawful conduct intended to secure a benefit for oneself or another. ‘Corruption’”. Britannica Concise Encyclopedia. 2007. Encyclopædia Britannica Online, 6 November 2007.
2.2 The Definition of Corruption
17
classifications of corrupt behaviour, but fails to include corruption behaviour by people who are bestowed with power that derives from roles in private organizations (see the discussion in the next section). Putting together the general definition mentioned above with a corrected version of Nye’s we propose the following definition: Corruption is the misuse of entrusted power for private gain; it is behaviour which deviates from the formal duties of a given role because of private-regarding (personal, close family, private clique) pecuniary or status gains; or violates rules against the exercise of certain types of private regarding influence. This includes such behaviour as bribery (use of a reward to pervert the judgment of a person in a position of trust); nepotism (bestowal of patronage by reason of ascriptive relationship rather than merit); and misappropriation (illegal appropriation of public resources for private-regarding uses). The objective of endorsing this definition is not to identify the right definition, but of choosing a useful one for the purpose at hand.8 While concepts and their definitions cannot be said to be right or wrong and can be the basis of infinite discussions, we take the more pragmatic approach of finding a working characterization and highlighting its limitations.9 One of the main drawbacks of this definition is that there are grey areas around the notions of misuse of power and formal duties; that is to say that the concepts underpinning this definition do not have entirely univocal meanings especially when we look at different understandings that can arise because of cultural diversity. Choosing one characterization of these concepts, presumably from a model society, and casting in stone a single model of misuse of power and moral conduct would entail an ethnocentric approach (i.e. a western stipulation) (see Philp 1997; Roy 1970).10 At the same time, in the converse relativistic approach, in which norms differ according to the cultural context, nothing can really be called corruption, and it is not possible to pass normative judgments on any social phenomenon (this is also referred to as the cultural approach; Bardhan 1997). Whereas the understanding of duties and roles depends, to a certain extent on cultural norms, in most countries—where the policy framework prescribes a modern bureaucracy—norms are standardized and the concern about abuse of power for private gain is more or less universal and not confined to western societies.11 In fact, in opinion polls of developing countries, corruption is often cited as the prime concern of respondents (Bardhan 1997, p.€1330). We find that—while a certain de8╇ It is also a definition that is coherent with the measures of corruption used in the empirical analysis in the rest of the thesis. 9╇ The point is made in Williams 1999, p. 511. 10╇ In this approach, standards developed by a particular type of society (i.e. the western society) are assumed to be the ideal and are employed to set the standards against which we can measure progress of any society. In the case of corruption, the model would be the working of western rich democracies and their understanding of duty and morality. 11╇ We highlight that the indexes of corruption that we use throughout the thesis are aggregations of indexes coming from various sources. These sources include citizens’ opinion polls and international experts’ surveys and the high correlation in the original indexes indicates that cultural ambiguities with respect to the concept of corruption are not a serious issue (Kaufmann and Kraay 2007, p. 22).
18
2 Economic Analysis of Corruption
gree of uncertainty is unavoidable—our definition benefits from the fact that these uncertainties are reflected in the language used (by avoiding reference to external and seemingly clear categories such as “the law”) and not focusing solely in the public sector fits in a social sciences’ concern with the analysis of corruption and its effects. Additionally, we recognize that there can be extreme circumstances where the standards used to discern corruption as they pertain to misuse of power, duty and public role do not hold. For example, during revolutionary periods loyalties other than to the state are prominent and the idea of the state itself might be the object of contention. In such situations the concept of duty and hence of corruption—as defined here—is inadequate. Now, we discuss some concepts related to corruption and its different forms. This discussion aims at clarifying notions used throughout the book and puts them in the context of the economic literature. Social capital is one of the concepts associated with corruption. Social capital, as defined by Putnam, “refers to networks and the norms of reciprocity and trust that arise from them” (2000, p.€19). Putnam cautions us that not all forms of social capital have a positive impact and critically distinguishes between “bonding” and “bridging” social networks; the former are more likely to have negative externalities. Corruption results from bonding networks and is mentioned as a negative manifestation of social capital (Putnam 2000, p.€22). The fact that corruption might arise from networks based on trust and—through repeated interactions in corrupt transactions—might increase social capital has been analyzed empirically. The findings of this line of research suggest that while corruption might bond and contribute to trust among interacting parties it decreases trust in society in general (Seligson 2002). Another concept related to corruption is bribery. This is the most obvious case of corruption where inducements are used in order to convince the bribed to change her course of action. The act can be ascribed to the will of the briber or to a request of the bribed. Corruption in the case of bribery is evident because there is a direct exchange where the action of the person in power is dependent on the payment made by the briber. We will use the concept of bribery—for ease of interpretation— in exemplifying different types of corruption.12 Shleifer and Vishny (1993) differentiate between corruption with theft and without theft. In corruption with theft the official demands a payment in order to offer a service that the briber should not get, or offers it for a price that is a fraction of the regulated price. Custom officials letting illegal goods (or legal goods without exacting import duties) enter the country in exchange for kickbacks represent examples of corruption with theft. The final cost of the operation for the briber might well be smaller than the cost without corruption and both parties might have an Familism and patronage are other forms of corruption where the duty is subverted not for immediate gain, but for allegiances to special networks. Of course, belonging to the network and contributing to it can ultimately lead to personal advantages.
12╇
2.3 Corruption and the Private Sector
19
interest in keeping the deal secret (i.e. they collude). The case of a custom official that requires an unofficial payment—over and above import duties—just in order to process paperwork is a case of corruption without theft. The latter type of transaction is characterized by divergent interests, because the briber would prefer to avoid the payment, hence there can be defection (i.e. the briber might denounce the bribed). The former type of corruption can be characterized as “collusive” and the latter “extortive”. Strategies to combat corruption have to take the differences between these types of corruption into account in order to elicit defection of some of the parties involved in the deal. Another useful distinction, based on the status of the bribed, is between political corruption and bureaucratic corruption.13 In the first instance, the bribed is a policy maker that influences policies in exchange for a side payment. In the second scenario, the corrupt actor is a bureaucrat that does not implement the regulations that were set by her superiors.14 The different actors—politicians or bureaucrats—are likely to have different incentive structures (e.g. voters’ perceptions can be more influential in the case of a politician than for a bureaucrat) and at times conflicting interests. Also in this case, these distinctions can inform anti-corruption policies. In the case of political corruption, opposition figures and their parties can play an important role in effectively constraining corruption opportunities of government and majority members. In the case of bureaucratic corruption, politicians in general can be active in combating corruption that might otherwise fuel discontent among their constituencies.
2.3 Corruption and the Private Sector Corruption can be either understood as a phenomenon affecting only the public sector, where the key element is the misuse of public power. Alternatively, corruption can happen also in the private sector and the power entrusted on and abused by the corrupt agent might also relate—for example—to corporations and NGOs (see Hodgson and Jiang 2007). Here we will present evidence of how the understanding of corruption by the general public includes the private sector, how economists who have dealt with the nature and the effects of corruption also at times included corruption in the private sector, and the way a biased understanding of corruption leads to fallacies in terms of policy prescriptions.
Similarly, a distinction can be made between grand and petty corruption. Grand corruption involves high level bureaucrat or politicians, for example it occurs when a sizeable payment is made to secure a large procurement contract (Rose-Ackerman 1999, p. 27). Petty corruption involves small payments to people at low level of hierarchies (see Lambsdorff 2005). 14╇ See also Bardhan (2006) who has similar and additional categorizations. 13╇
20
2 Economic Analysis of Corruption
2.3.1 Perceptions of Corruption In the first place, it is worth highlighting how corruption always involves multiple parties (typically the briber and bribee), and in most instances some of the parties will belong to the private sector. In the case of collusive corruption, the party from the private sector might even play a dominant role by soliciting the bribee to accept an unrequested payment in exchange for some undue advantage that the person in a position of power is capable of offering. These roles of the private sector do require tools to deal with the private agents involved in the transactions and the prominence of private sector is evident and relatively uncontested (see Shleifer and Vishny 1993). Here, by arguing for a characterization of corruption that includes cases where none of the parties exercises a public role, we mean simply that cases in which the entrusted power that is being abused is of a private nature should fall under the rubric of corruption. Furthermore, we want to highlight how this characterization of corruption undermines the simplistic conclusion that the expansion of the private sector vis-à-vis the public sector (e.g. via privatization) will unequivocally imply a decrease of corruption given the fact that the interface between the private and the public sector will still exist—offering ample opportunities for corruption—and because corruption can continue to characterize the behaviour of agents belonging to the private sector. It might well be the case that the locus and the actors involved in corruption change without changing the nature of the transactions taking place. The public at large, international organizations, and social scientists when analyzing corruption reveal that their understanding is informed by the fact that corruption is something that can affect also the private sector. This perception is exemplified by the findings of the Global Corruption Barometer 2009 (Transparency International 2009): in a survey of the population of 69 countries—based on 73,000 interviews—it was found that the private sector is perceived in 12 countries to be the most corrupt when compared to political parties, parliament and legislature, the media, public officials, and the judiciary. These opinions, and the underlying understanding of corruption, stand in clear contrast with characterizations of corruption that focus uniquely on the public sector. With respect to the way economists interpreted the problem of corruption, a long citation from Alfred Marshall is in order. In fact, Marshall already at the end of the nineteenth century noted that: Every one is aware of the tendency to an increase in the size of individual businesses, with the consequent transference of authority and responsibility from the owners of each business to its salaried managers and officials. This would have been impossible had there not been a great improvement in the morality and uprightness of the average man: for even as late as the seventeenth and eighteenth centuries we find the great trading companies breaking down largely in consequence of the corruption and selfishness of their officials. (Marshall 1897, p.€130).
Marshall’s concerns for the structure of firms, their hierarchies, and functioning shows how corruption within the private sector can affect the basic unit on which economic development is based. The role of firms and the rationale for their
2.3 Corruption and the Private Sector
21
existence was later re-emphasized by Ronald Coase who argued that to understand the nature of the firm we should look at alternative ways in which transactions could take place and compare the associated costs (Coase 1937). Re-phrasing Marshall, it seems apparent that transactions—and the associated costs—within firms were already a concern for him and that corruption might affect these transaction to the point of limiting the size of firms in corrupt environments. The size of firms is, in turn, associated with investment and productivity. This is an example of how economic analysis has shown that the abuse of power, entrusted in the private sector, can also lead to detrimental economic outcomes and that the concern with corruption in economic theory is not bound to the public sector. Similarly, other aspects of corruption and of the ways corruption has been analyzed can be the bases for an extension of the analysis to the private sector. One example of these instances is the explanation of the association between high corruption levels and low investment levels because corruption acts as a tax on investment (e.g. Mauro 1995; Wei 2000). In this framework the investor would face public officials that exact bribes in order to allow the investment to take place and to facilitate the operation of the firm. The investor would discount such expenditures before taking the decision to invest and the impact of corruption would be analogous to that of taxation. We can easily extend these considerations to the private sector. The effect on investors of corruption within their own organization would be of the same nature: the siphoning of part of the revenues of the enterprise by its employees could also be seen as a tax on investment and would have similar impacts on future revenues and on investment decisions. Transparency International itself—as mentioned above—endorses the operational definition of corruption as “the abuse of entrusted power for private benefit” that includes also the private sector. Nevertheless, its Corruption Perception Index uses sources that define corruption as the “the misuse of public power for private benefit”.15 Also the World Bank at times defines corruption as something pertaining only to the public sector, while it also discusses the problem of “corporate corruption”.16 These inconsistencies show the tension created by the fact that often the starting point of the analysis is corruption only in the public sector, but many ramifications and issues require the inclusion of corruption in the private sector. The definitional issues underlying this tension are often unexplored.
2.3.2 O pportunities for Corruption and Anti-corruption Strategies Now we will present three examples of how reforms of public sector activities, namely privatization, do not automatically achieve one of their objectives: the reCf. Lambsdorff 2008 and http://www.transparency.org/news_room/faq/corruption_faq, 3 June 2009. 16╇ Kaufmann 2004 and Kaufmann et€al. 2005. 15╇
22
2 Economic Analysis of Corruption
duction of corruption (see Estache et€al. 2009). This issue is discussed here because one rationale of privatization is based on the belief that the public sector is often marred by corruption, while the private sector is not. The first two examples below show how privatization creates new interfaces between the private and the public sectors and large stakes are redistributed in the process creating incentives for corruption. The last example shows how the set of behaviour termed “corrupt” can continue unaltered after a state company is privatized. The privatization process in the former USSR was marked by corruption and the sale of state assets offered an opportunity to grab and accumulate large fortunes by illegal means (e.g. Sachs 2005). This outcome is the apparent result of the mismanagement of the whole process and of the erroneous assumptions on which it was based (Black et€ al. 2000). The idea that substituting markets and private agents for the state would automatically enhance efficiency and solve the problem of corruption underlies the course of action (Boycko et€ al. 1996) and its failure. Many lasting problems were created: the whole privatization process, because of corruption, produced a class of tycoons whose fortunes are tainted by their illegal nature. To this day these entrepreneurs depend on complacency from the government to retain their holdings and this dependency has further fuelled the corruption that is still pervasive. Furthermore, many of the “new rich” have preferred to siphon their illegally obtained funds abroad, deepening the lack of resources for investment and damaging economic development prospects of the country. Among the many negative effects of carrying out the privatization program in such a fashion was the detrimental impact that corrupt privatization has had on the whole democratization process in Russia. Because of the nexus created between the “new rich” and the politicians and the fact that privatization and political corruption proceeded hand in hand there was been widespread disillusionment with many the changes associated with democracy.17 The second example is from the privatization of healthcare services in Italy. The state provision of health care in Italy has been renown because of its inefficiencies and corruption and the sector was at the centre of many of the judiciary cases in the anti-corruption trials known as “mani pulite” (clean hands) at the beginning of the 1990s. The trials—among many other facts—have shown how the national health service was buying medicines at inflated prices because of the collusion of employees of the Ministry of Health, of the minister himself, and pharmaceutical companies. Since then the presence of private health care providers has increased and many health services are now delivered by private companies that get reimbursed by the state. Unfortunately, recent judiciary cases have shown how the extension of private operations in the sector has not been able to stamp out corruption. For example, private health care providers have been able to get compensations for services that they had never provided or, even worse, they were providing un-
This is not to say that privatization of state enterprises should not have taken place, on the contrary we mean that the potential benefits of the privatization process were lost because of the way the process was mismanaged (Black et€al. 2000).
17╇
2.3 Corruption and the Private Sector
23
necessary services (including surgeries) to unknowingly patients in order to obtain compensations for them.18 The third example is the privatized electric utility of Nicaragua. The generation companies have been privatized from the mid-1990s and formerly state-owned distribution companies were bought by a multinational company (Union Fenosa). The privatization process—as in the Russian case—has not been accompanied by the implementation of regulations to ensure that the potential benefits of the process would be realized (CEPAL 2003). Interestingly, one practice that is still common for employees of the distribution company is to collect bribes from consumers in order to forge consumption data and the resulting bills. Arguing that these are not cases of corruption since the company is now private, would imply some taxonomic changes (from corruption to theft, or fraud) without really addressing the problems. The examples of privatization in Russia and Italy show how the size of the public sector might not matter for the spread of corruption and show how privatization might change the practice of corruption but not eliminate it. Overall, the examples show how the shrinking of the public sector might actually create opportunities for corruption (as in the Russian example), or change the locus of corruption to the public–private interface (as in the Italian example), or simply, if we were to define corruption as something that can happen only in the public sector, change the name of things without changing their nature (as in the Nicaragua example). These examples represented different types of corruption and of processes at play. These differences and the ensuing problems related to privatization are obscured by an approach to corruption that focuses solely on the public sector or equates privatization with diminished corruption. Indeed, reforms such as privatization—if they are to succeed in reducing corruption—should be accompanied by anti-corruption strategies rather than rely on presumptions of automatic decreases of corruption. In other words, if expectations of corruption reduction in conjunction to privatization programs are not carefully worked out, the success in combating corruption can be purely semantic rather than substantive.
2.3.3 Corruption and Market Failures It is worth noticing that—among social scientists—economists stand out as particularly critical of the public sector and this position is reflected by the fact that many economists agree with the notion that corruption is an issue belonging solely the public sector (Hodgson and Jiang 2007, p.€57). Simplistic characterizations of corruption simply see a choice between state intervention with its associated corruption and a private sector that can be marked by market failures but is intrinsically not corrupt (for an example see Acemoglu and Verdier 2000). The policy advice that For an overview of the corruption problems related to healthcare in Italy, see Paolo 2008; for some of the most recent cases see http://www.repubblica.it/2008/06/sezioni/cronaca/mediciarrestati/medici-arrestati/medici-arrestati.html, 14 June 2009.
18╇
24
2 Economic Analysis of Corruption
derives from this one-sided characterizations of corruption does not make sense in much the same way it would not make sense to suggest that the way to solve problems of fraud and theft in the private sector is to nationalize private companies. Another fallacy is the fact that corruption might be the reason why the market failures exist in the first place especially when policy formulation is influenced by political corruption. Powerful interests can unduly influence policy makers into inaction and make them unresponsive to requests to regulate a sector and deal with societal demands. One of the main points of this book is that the lack of policies and regulations in the environmental sphere might actually be a result of corruption, rather than a way to deal with it. In this case there is no dichotomy, nor is there a choice between market failures and corruption: corruption actually is the source of market failures. We will transform the famous “because government intervention transfers resources from one party to another, it creates room for corruption” (Acemoglu and Verdier 2000, p.€194) into “because government intervention and non intervention transfers resources from one party to another, it creates room for corruption”. The point here is that in the presence of social problems inaction inevitably favours the status quo if compared to a situation where the state responds to societal needs regulating certain activities; i.e. by taking away the “right to pollute” enjoyed by some agents. Finally, these considerations are backed up by the lack of consistent econometric evidence that corruption is associated in any systematic way to the size of the public sector in the economy (e.g. Lambsdorff 2007, p.€4–5; see the results in Chap.€3), nor of a relation between neo-liberal policies and corruption (Gerring and Thacker 2005).
2.4 The Measurement of Corruption Several governance indicators are now available and provide the possibility to quickly gain an impression of the quality of many aspects of the institutional environment of countries. Furthermore, these data provide the opportunity for econometric analyses of the determinants and the impacts of corruption. At the same time the availability of these indexes has also offered ample opportunities for their “abuse” (Williams and Siddique 2008). For example, many authors have been using data without fully understanding them and have incorrectly used them as proxy variables, the most famous cases being those authors that used the data of the “Corruption” variable from the International Country Risk Guide in econometric analysis investigating the effects of corruption without realizing that the variable is actually a measure of political risk associated with corruption rather than corruption per se (see the discussion in Lambsdorff, 1999). The assessment of corruption levels across countries is a formidable empirical challenge as corruption is difficult to define uniformly, being culturally determined, and even more difficult to measure, because illegality implies secrecy. The individual indexes aggregated in order to create the proxies of corruption (or, more
2.5 Discussion
25
precisely of its perception), range from Gallup’s opinion surveys—where a sample of the polity is asked how common corruption is and at what scale it operates—to surveys of company executives that estimate the share of their companies’ revenues that are spent in bribes.19 On the one hand, the high correlation of the aggregated indexes, originated from different sources, gives some confidence in the belief that they are correct proxies for corruption. On the other hand, it should be noted that the definition of corruption can be interpreted differently in different cultural contexts and that there can be “emotionally driven” answers to survey. Thus, when a corruption scandal is escalating, the interviewees will possibly overrate the level of corruption, just as when the economy is booming interviewees can have a more positive attitude about the government and the civil servants. While some of these concerns are taken care of by the way the polls are realized and aggregated, a margin of uncertainty is inevitable (see Kaufmann et€al. 2005). Critiques of these indexes have, among other things, focused on the fact that they measure perceptions rather than “real” corruption levels. However, as it has been shown extensively in the empirical literature, even though these indexes may delineate the perception of corruption rather than its existence, it appears that perceptions have an impact on the economy and that the indexes have a high explanatory power when used as independent variables in econometric analyses (for a review see Jain 2001). Moreover, some of the sources, such as surveys of company executives’ expenditure on bribes as a percentage of revenues, are more akin to estimates of corruption rather than simple polling of opinions.20,â•›21
2.5 Discussion In this chapter we have discussed issues related to the definition of corruption and to its measurement. The objective is twofold: first we contribute to the literature on the meaning of corruption by showing how different characterizations can have a bearing on the anti-corruption policy options of choice. Second, we discuss the proposition that corruption is a necessary evil when dealing with market failure, showing that the dichotomy of market failure or corruption is conceptually flawed. Corruption is actually a cause of market failures rather than an alternative. Overall, we find that the root of the corruption problem is the abuse of power, regardless of the
For a full description of all the sources and of the aggregation methodology used by the World Bank, see Kaufmann et€al. 2005; for sources and methodology of the Transparency International index, see Lambsdorff 2004. 20╇ One example of a survey—included among the sources of our corruption indexes— containing such a variable is the “World Business Environment Survey” of the World Bank, available at http:// info.worldbank.org/governance/wbes/. 21╇ For updated reviews of the use (and abuse) of governance indicators in economic studies see Kaufmann and Kraay 2007 and Williams and Siddique 2008. 19╇
26
2 Economic Analysis of Corruption
nature of that power. Policies should focus on the root causes of this abuse and the conditions facilitating corruption rather than the nature of the power being abused. To sum up, a definition of corruption that does not include corruption in the private sector (e.g. as in Nye’s original definition; Nye 1967) leads one to the simplistic conclusion that to decrease corruption it is sufficient to decrease the size of the public sector. This conclusion is misleading in various ways: it ignores important strands of economic thought, goes against the grain of the common understanding of corruption in the public, and directs public action towards actions that would simply characterize one type of damaging behaviour (corruption) with other types of behaviours in the public sectors whose final result that is not in any way an improvement on the original state (e.g. by substituting corruption for theft or fraud). The associated issue of the alternative between market failure and corruption is also based on a simplistic characterization of corruption where state intervention is always associated with corruption, while non-intervention might be associated with market failure. One of the main points of Chaps.€5 and 6 is that corruption can actually lead to non-intervention in terms of environmental policies; in other words corruption and market failures would go hand in hand rather than be alternatives. Finally, we have introduced concepts and measures of corruption that are used throughout the book. Ultimately clarifying what we mean by corruption and what the features and limitations of the indexes we have used informs the reader of the meaning of subsequent analyses and of the caveats thereof.
References Acemoglu, D., & Verdier, T. (2000). The choice between market failures and corruption. American Economic Review, 90(1), 194–211. Bardhan, P. (1997). Corruption and development: A review of issues. Journal of Economic Literature, 35(3), 1320–1346. Bardhan, P. (2006). The economist’s approach to the problem of corruption. World Development, 34(2), 341–348. Becker, G. S. (1968). Crime and Punishment—Economic Approach. Journal of Political Economy, 76(2), 169–217. Black, B. S., Kraakman, R., & Tarassova, A. (2000). Russian privatization and corporate governance: What went wrong?. Stanford Law Review, 52, 1731–1808. Boycko, M., Shleifer, A., & Vishny, R. W. (1996). A theory of privatisation. Economic Journal, 106(435), 309–319. CEPAL (2003). Evaluación de Diez Años de Reforma en la Industria Eléctrica del Istmo Centroamericano. Santiago: Comisión Económica Para América Latina y el Caribe (CEPAL). Coase, R. (1937). The nature of the firm. Economica, 4(16), 386–405. Estache, A., Goicoechea, A., & Trujillo, L. (2009). Utilities reforms and corruption in developing countries. Utilities Policy, 17(2), 191–202. Gerring, J., & Thacker, S. C. (2005). Do neoliberal policies deter political corruption?. International Organization, 59(1), 233–254. Hasnas, J. (1995). Back to the future: From critical legal studies forward to legal realism, or how not to miss the point of the indeterminacy argument. Duke Law Journal, 45(1), 84–132. Hodgson, G. M., & Jiang, S. X. (2007). The economics of corruption and the corruption of economics: An institutionalist perspective. Journal of Economic Issues, 41(4), 1043–1061.
References
27
Huntington, S. P. (1968). Political order in changing societies. New Haven: Yale University Press. Jain, A. K. (2001). Corruption: A review. Journal of Economic Surveys, 15(1), 71–121. Kaufmann, D. (2004). Corruption, governance and security: Challenges for the rich countries and the world. Global Competitiveness Report 2004–2005, World Economic Forum, Geneva. Kaufmann, D., & Kraay, A. (2007). Governance indicators: where are we, where should we be going? World Bank Policy Research Working Paper, World Bank, Washington. Kaufmann, D., Kraay, A., & Mastruzzi, M. (2005). Governance matters IV: Governance indicators for 1996–2004. World Bank Policy Research Working Paper Series, World Bank, Washington. Kugler, M., Verdier, T., & Zenou, Y. (2005). Organized crime, corruption and punishment. Journal of Public Economics, 89(9,10), 1639–1663. Lambsdorff, J. G. (1999). Corruption in empirical research: A review. Berlin: Transparency International. Lambsdorff, J. G. (2004). Framework document 2004. A background paper to the 2004 corruption perceptions index. Transparency International and Passau University. Lambsdorff, J. G. (2005). Between two evils—investors prefer grand corruption! Diskussionsbeitrag. Passau: Passau University. Lambsdorff, J. G. (2007). The institutional economics of corruption and reform: Theory, evidence, and policy. Cambridge: Cambridge University Press. Lambsdorff, J. G. (2008). The methodology of the TI corruption perceptions index 2008. Berlin: Transparency International and University of Passau. Marshall, A. (1897). The old generation of economists and the new. The Quarterly Journal of Economics, 11(2), 115–135. Mauro, P. (1995). Corruption and growth. Quarterly Journal of Economics, 110(3), 681–712. Nye, J. S. (1967). Corruption and political development: A cost-benefit analysis. The American Political Science Review, 61(2), 417–427. Paolo, C. F. (2008). La casta bianca. Viaggio nei mali della sanità. Milano: Mondadori. Pellegrini, L. (2007). The rule of the jungle in Pakistan: A case study on corruption and forest management in Swat. FEEM Working Paper, FEEM, Milano, Venezia. Philp, M. (1997). Defining political corruption. Political Studies, 45(3), 436–462. Putnam, R. D. (2000). Bowling alone: The collapse and revival of american community. New York: Simon & Schuster. Rose-Ackerman, S. (1999). Corruption and government: Causes, consequences, and reform. Cambridge: Cambridge University Press. Roy, E. V. (1970). On the theory of corruption. Economic Development and Cultural Change, 19(1), 86–110. Sachs, J. (2005). The end of poverty: How we can make it happen in our lifetime. London: Penguin. Seligson, M. A. (2002). The impact of corruption on regime legitimacy: A comparative study of four latin american countries. The Journal of Politics, 64(2), 408–433. Shleifer, A., & Vishny, R. W. (1993). Corruption. The Quarterly Journal of Economics, 108(3), 599–617. Transparency International (2009). Global corruption barometer. Berlin: Transparency International. Wei, S. J. (2000). How taxing is corruption on international investors?. Review of Economics and Statistics, 82(1), 1–11. Williams, R. (1999). New concepts for old?. Third World Quarterly, 20(3), 503–513. Williams, A., & Siddique, A. (2008). The use (and abuse) of governance indicators in economics: A review. Economics of Governance, 9(2), 131–175.
Chapter 3
Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
Abstract╇ We survey and assess the empirical literature on the sources of corruption. Thanks to the improved availability of data, we are able to produce a comprehensive cross-country econometric model to test well-established and more recent hypotheses jointly. We do not find that the common law system, or a past as a British colony predicts corruption. Our results support cultural theories on the causes of corruption, and suggest that a medium-long exposure to uninterrupted democracy is associated with lower corruption levels, while political turnover tends to raise corruption. The results also suggest that the diffusion of newspapers helps to lower corruption levels. JEL classification╇ D72 • H11 • H50 • K42 • O17 Keywords╇ Corruption • Culture • Ethnolinguistic fractionalization • Democracy • Political turnover
3.1 Introduction Corruption is a widespread phenomenon affecting all societies to different degrees at different times. As corruption scandals have repeatedly shown, bribes are common in all countries notwithstanding differences in income levels and legal systems. Recent scandals over corruption have shown that the top level of governments can be involved and that supposedly free-from-corruption societies are affected. The ELF scandal demonstrated that corruption was rampant in the management of the French state-owned enterprise and involved French and German heads of state of different political leaning.1 The year following the ELF case, a corruption charge against President Chirac could not be brought to court because he was shielded by This chapter is a slightly revised version of Pellegrini and Gerlagh (2008). 1╇
On the ELF scandal see The Washington Post, Wednesday, February 9, 2000; Page A21.
L. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9_3, ©Â€Springer Science+Business Media B.V. 2011
29
30
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
immunity as the head of state.2 Also in Germany, the CDU and former Chancellor Helmut Kohl were fined for receiving illegal campaign funding.3 Among Nordic countries (which rank always among the less corrupt in international comparisons), Swedish and Norwegian managers of state-owned companies have been found to be involved in bribe-taking.4 Corruption has its place even during humanitarian emergencies. According to Transparency International, an NGO which strives to expose corruption, relief efforts in the aftermath of the South East Asian Tsunami earthquake of 2004 were hampered by corruption.5 Still, different countries are marked by large differences in the extent of corruption. In some societies, many transactions cannot be finalized without corruption having an effect, while in other countries it is considered an exception and rarely tolerated. Hard evidence of corruption is intrinsically difficult to obtain, because of the secrecy surrounding illegal deals, but there are several ways to obtain proxies of the extent of corruption. One such source comes from the pool of international interviews commissioned for the Global Corruption Perception Barometer (Transparency International 2004). From the barometer, we can see that around 90% of Brazilians think that petty and grand corruption are “very big problems”, around 50% of Finnish interviewees think that petty corruption is “no problem at all” and around 35% think that also grand corruption is no problem. With respect to the personal experiences of individuals with bribery, we see that more than 50% of Cameroon’s interviewees admit that somebody in their household paid a bribe in the last 12 months, while among Austrian, Canadian, German and Irish interviewees only 1% mentioned exposure to corruption for their households in the last year. Evidence on private firms’ expenditures for bribes is available from the World Business Environment Survey 2000.6 From the executives’ interviews in the survey, we see, for example, that while more than 90% of Canadian firms declare that they did not pay any bribes in the last year, none of the Armenian and Azerbaijani firms interviewed declared that no such payments took place. Figure€3.1 presents a geo-referenced representation of corruption levels worldwide; the figure shows that corruption tends to be pervasive especially in developing countries. The evidence of the difference in corruption levels worldwide is matched by indication that corruption has adverse socio-economic implications and numerous studies have demonstrated the pernicious effects of corrupt practices on—among other things—economic growth, investment, human development and environmental policies. Once we consider the negative effects of corruption on welfare, an obvious research question arises: what is the reason for corruption to be common in some http://archives.cnn.com/2001/WORLD/europe/10/10/chirac.court/. http://www.businessweek.com/magazine/content/01_48/b3759151.htm. 4╇ Corruption Blackens Nordic Region’s Lily-White Image, Agence France Presse, December 4, 2003. 5╇ http://www.transparency.org/pressreleases_archive/2005/faq_tsunami.html#faqti1. 6╇ See http://info.worldbank.org/governance/wbes/. 2╇ 3╇
Fig. 3.1↜渀 A higher score (a darker colour) indicates higher corruption perception. Our elaboration on data from Kaufmann et€al. (2005)
3.1 Introduction 31
32
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
countries while other countries succeed in preventing corruption from hampering their welfare? While theoretical literature on this subject abounds, empirical studies are relatively scarce. Since several indexes of corruption perception have become available over the last few years, it is now possible to test statistically some of the ideas from the theoretical literature. To the best knowledge of the author, the most comprehensive and widely cited econometric analysis of the sources of corruption is a cross-sectional study from Treisman (2000).7 The study takes into consideration and tests empirically a wide range of theoretical explanations of corruption and finds mostly “fixed factors” as the determinants of corruption. That is, the significant explanatory variables are all persistent over time, i.e. impossible to change in the short and medium run. The most policy-amenable variable is the “exposure to democracy for more than 46 years”.8 As the sources of corruption are persistent, corruption itself will remain stable over time and can barely be affected by policies; corrupt countries should rather learn to live with it. Similarly, La Porta et€al. (1999) provide a comprehensive econometric analysis of the sources of corruption. In their paper, the dataset and the variables included in the analysis are different from Treisman’s, but the results also found the determinants of corruption to be predominantly historical factors (e.g. legal traditions, ethnic heterogeneity and religious affiliation). The studies by Treisman and La Porta et€al. will be the point of departure of our work.9 Previous and subsequent empirical studies—mostly cross-sectional—have focused more on single issues where the authors have tested a particular theory by inserting a proxy for a specific variable into multiple regressions. These studies most often produced results that confirmed the theories that were tested and, while they are valuable contributions for the identification of likely sources of corruption, they may have over-emphasized the importance of the variables analyzed because of omitted variables biases. Several studies focused on the role of democracy as a determinant of corruption (among the most recent: Sung 2004; Chowdhury 2004), some focused on regulatory burden and economic freedom (Chafuen and Guzmán 1999), others have focused on decentralization and federalism (e.g. Fisman and Gatti 2002; Arikan 2004), on natural resources prevalence as a source of income in the economy (Leite and Weidmann 1999), and on legal origins of a country as a determent of corruption (Glaeser and Shleifer 2002). We assess the different theories and calculate our own estimates of the sources of corruption in a comprehensive econometric model using indexes of corruption that have become available recently (Kaufmann et€al. 2005) and cover a large sample of countries. Furthermore, we make use of some independent variables that are improved proxies for the variables that theory would suggest. The empirical 7╇ This relatively recent paper has been already cited in 231 other works according to ISI Web of Science (checked on the 14th August, 2010). 8╇ Treisman (2000) selected exactly 46 years because his source of data on democracy was available from 1950 to 1995. 9╇ Lambsdorff (1999) provides a comprehensive survey of earlier empirical results.
3.2 Theories of Determinants of Corruption
33
contribution of this chapter—when compared to existing literature—is threefold. First, due to improved data availability, we can work with a larger data sample that gives more power to our statistical tests compared to earlier studies.10 Second, we have a large set of independent variables available that we can test jointly so that we do not suffer from omitted variable bias; a problem of many previous and partial studies that may have overemphasized the importance of the variables under consideration. This advantage is particularly valuable as it helps us to assess the importance of long-lived versus contemporary causes of corruption whereas these sources of corruption have mostly been assessed independently so far. Thirdly, a crucial advantage over previous authors is the availability of two alternative corruption indexes, which can be used for robustness checks of the findings. Indeed, our results differ from the ones of previous authors. We trace such differences—when we compare our results to other comprehensive empirical analyses—mostly on our larger sample size. The alternative use of the two indexes of corruption confirms the robustness of our findings. In this chapter we will first provide a systematic overview of different theories on the causes of corruption. In this survey, presented in the next section, we distinguish theories that focus on historical roots versus theories that give more attention to contemporary causes. Thereafter, Sect.€3.3 describes the data that we use for our own econometric analysis, Sect.€3.4 introduces and discusses the econometric estimates, and Sect.€3.5 concludes.
3.2 Theories of Determinants of Corruption Theories of the determinants of corruption (and more generally of the quality of government) abound. We will take as a starting point the theories on the sources of corruption that are mentioned in Treisman (2000) and La Porta et€al. (1999) as these studies are considered a benchmark in the literature and they provided a powerful battery of empirical tests. To these we will add the most recent findings of empirically backed literature in order to test and build upon their findings. The subsequent econometric estimates will follow a similar approach. The theories and variables we will analyse are divided into historical and contemporary determinants of corruption. Another distinction could be made between underlying and proximate causes of corruption, where the former refer to the driving forces determining corruption levels and the second to manifestations of these driving forces. In our analysis we recognize that it is difficult to identify when some variables might be only a manifestation of others and that the underlying causes are more related to the historical determinants of corruption. The long-run/historical determinants of corruption are thus those that can be described with more certainty Specifically, when we compare our study to Treisman’s, we find that our sample is larger than 100 countries in most regressions, while Treisman’s sample size is between 44 and 64 countries in the majority of regressions.
10╇
34
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
as underlying causes of corruption. Some of the more contemporary causes of corruption can be more closely related to proximate causes, as they might be manifestations of the driving forces.11
3.2.1 Historical Roots of Corruption Among the theories that relate the history of countries to their present day levels of corruption, legal theories suggest that the kind of legal codes that are in place in a country affect the quality of government, including the level of control of corruption. In many countries legal codes are influenced—to varying degrees—by the French civil law or by the British common law. The latter has been singled out because of the relative independence of law enforcers from other state powers. Some authors have elaborated historically motivated theories tracing the effort of property owners to limit the discretionary power of the monarch’s power as the origin of common law legal codes (Glaeser and Shleifer 2002). Furthermore, they suggest that the actions of an independent judicial system in countries that adopted the British legal code will be conductive to better governance, including lower levels of corruption (for an in depth discussion, see La Porta et€al. 1999). Another theory, also based on the historical roots of corruption, is one that sees former British colonies as having a better civil service code due to the influence of the British bureaucracy. In this system, the British civil servant is focused on procedural aspects of the law, which enhances the capability of subordinates and judges to challenge hierarchies in order to enforce the law (Treisman 2000). However, as a result of the method with which colonization was imposed upon countries, this positive effect of British colonization can be questioned. Quoting Macaulay (1843/2001) “the business of a servant of the [East India] Company was simply to wring out of the natives a hundred or two hundred thousand pounds as speedily as possible, that he might return home before his constitution had suffered from the heat, to marry a peer’s daughter, to buy rotten boroughs in Cornwall, and to give balls in St. James’s Square”.12 If British colonizers were extracting resources, in addition to establishing legal procedures, the effect of colonization on present corruption deserves to be tested. Another theory, which has been put under scrutiny in previous literature, is that Protestant religion, being relatively less hierarchical when compared to other churches and religions (such as the Orthodox and Catholic churches and Islam), is
The same point is valid also for the rest of the econometric analyses in this study. In Chap.€4 in particular, the difference between corruption and the transmission channels could be described as corruption being the underlying determinant and the transmission channels being the proximate variables influencing economic growth. 12╇ Macaulay was in India working for the Supreme Council of India and later became a member of the British Parliament. 11╇
3.2 Theories of Determinants of Corruption
35
less prone to tolerance towards power abuses and corruption.13 Furthermore, the Protestant church has traditionally been apart from the state and played the role of opposition to the abuses of the government (Treisman 2000). Thus, this theory suggests that Protestant countries will be less affected by corruption. Also pointing at long-living causes of corruption are those theories that suggest that more ethnically fractionalized countries tend to be more corrupt (Mauro 1995). One root of the link between ethnolinguistic fractionalization and corruption can be the existence of alternative affiliations and obedience with respect to the state. Thus, in ethnically divided societies civil servants and politicians would exploit their positions to favour members of their own ethnic group. Another possible explanation is the fact that divided societies tend to under-provide public goods and this, in turn, would augment the dependency on special bounds to obtain essential services from the state. Natural resources are a common source of high rents, available to those that have obtained the rights for their exploration and extraction. These rents promote activities aiming at influencing policymakers who have power in the distribution of exploitation rights, drawing away resources from other productive activities (Leite and Weidmann 1999). Thus, abundance of natural resources would be associated to higher corruption, though we cannot take this effect for granted since revenues from natural resources could also be used in order to increase the availability of stateprovided goods, decreasing the need to revert to bribing in order to access them.
3.2.2 Contemporary Causes of Corruption Another set of theories relates the level of corruption to institutions, economic structures, and the level of development. These theories provide more of an entrance for anti-corruption policy compared to the theories based on historic variables, as there would be some policy-amenable factors among the determinants of corruption. To start with, income levels may affect corruption in several ways. Richer countries can be expected to afford better institutions. Furthermore, many variables correlated with income, such as schooling levels, urbanization and access to mass media, are associated with higher development levels and decrease the tolerance of the polity towards corruption. It may need some development to recognize corruption as a violation of the border between the public and the private sphere. Thus, we expect a real income variable to be negatively correlated with corruption. The rent-seeking literature emphasizes the link between corruption and possibilities for economic agents to gain access to sources of higher-than-average rents, when state intervention prevents free entry (see Rose-Ackerman 1999). In this perspective, the fight against corruption is helped by a reduction of non-generic state Further distinctions could be made between different groups within these religions. For example, Sunni and other forms of Islam could be differentiated to verify whether they have a different impact on corruption.
13╇
36
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
regulation. Thus, corruption would be associated to the size of government activities (Chafuen and Guzmán 1999; Acemoglu and Verdier 2000). In a similar vein, increasing the supply of foreign products on the domestic market enhances competition, thereby reducing rents and corruption. Another strand of the literature considers the relation between political institutions and corruption, starting with democracy. A negative correlation between democracy and corruption is tautological when based on a substantial definition of democracy, since corruption favours the interests of the individual, or a minority, as opposed to the interests of the majority. Once we consider democracy from a procedural perspective (free elections and electoral competitions) the association is less straightforward. Most indexes of democracy are based on the procedural aspects of democracy, and previous empirical studies have found contrasting results. Many papers that focus on democracy and employ few control variables find contemporary democracy to decrease corruption levels (e.g. Hill 2003; Chowdhury 2004; Bohara et€al. 2004), while more comprehensive studies do not find such a correlation (Treisman 2000). The experience of some Latin American countries suggests that the transition to democracy did not produce much of a dividend in terms of a reduction in corruption. A telling example is Mexico. Up to 2000, the country was ruled for seven decades by the PRI, known for its cronies and patronage system. Then, for the first time, a president was elected that was not from the PRI. This was considered a turning point for Mexican democracy, but the following years were marked by corruption scandals that affected virtually all political parties, including the ruling party (the PAN, which has gained power on its platform promoting change and a fight against corruption),14 the former ruling party PRI, the other opposition party (the PRD)15 and even the marginal ecologist party.16,17 Also in Europe, cases of corruption in long-established democracies abound; the above-mentioned high-level scandals in France are examples. As another example, Italy scores very high on all indexes of corruption, despite having been a democracy with high levels of electoral competition for six decades. Looking at the political system from a dynamic point of view leads us to political turnover as another variable that may affect corruption levels. In politically stable administrations, politicians and bureaucrats face less chances of dismissal and have more opportunity for long-run advancement in their careers, which provides an incentive to build an open and honest reputation for a career development. On the other hand, a permanent position in power may help maintain “patronage President Fox has been accused of using illegal funds to finance his campaign, see http://news. bbc.co.uk/hi/spanish/latin_america/newsid_2802000/2802161.stm. 15╇ On the tape-scandals, involving the most important aides of Lopez Obrador, see http://news.bbc. co.uk/hi/spanish/latin_america/newsid_3531000/3531475.stm. 16╇ On the scandal involving the young leader of the Partido Verde Ecologista, see http://www. esmas.com/noticierostelevisa/mexico/345598.html. 17╇ We note that the effect of democracy on corruption could also work on a longer time frame; in this case the benefits of democratic changes in Mexico will be reaped in the future. Indeed, in the analysis below we find that exposure to democracy for a long period of time is associated with corruption levels, while contemporary democracy is not. 14╇
3.3 Data Sources on Corruption
37
and corruption” reputations and relations. There are therefore two alternative and contradictory hypotheses on whether corruption is discouraged or fostered by political stability (Treisman 2000). Apart from the duration of a job in the administration, the wages may also affect the vulnerability to corruption. Higher wages imply higher costs when a position in the civil service is lost, and a cost-benefit analysis suggests that higher wages thereby provide an incentive to refrain from corruption (Becker 1968; Treisman 2000). Furthermore, at higher income levels additional income—obtained through corruption—could be less tempting because of decreasing marginal utility of income (see also Schulze and Frank 2003). Finally, we look at a rather different institutional variable, recently studied in both theoretical and empirical work: newspaper circulation, where the press is supposed to act as a check on those that should represent the public interest (Brunetti and Weder 2003). The hypothesis claims that corruption scandals freely enquired and exposed by the mass media act as a deterrent for bureaucrats to engage in corruption activities.
3.3 Data Sources on Corruption In this section we describe our dataset and the underpinning concept of corruption. The other variables will be described below as they are introduced in the econometric analysis. The variables’ sources are listed in the Appendix together with the descriptive statistics. We will use data that comply with the general definition of “abuse of power for personal gains”18 (similar to the one we introduced in Sect.€ 2.1) from the World Bank (Kaufmann et€al. 2005) and Transparency International. The two datasets are similar in the sense that they gather existing measures on the perception of corruption, and produce an aggregate index. In this chapter, we will use the data from the World Bank in the main analysis and the index from Transparency International as a robustness check. The dataset on corruption perception is very extensive and includes a number of very small countries with typical features (e.g. having an attractive investment climate) that can easily lead to selection bias in our results.19 For this reason, we omit
For an overview of complexities and the evolution in social sciences of the definition of corruption see Williams (1999). 19╇ One example of the special features that very small countries have is the more limited extent of ethnolinguistic fractionalization and the fact that they tend to be more open to trade (e.g. Knack and Azfar 2003). Since small countries are included in corruption surveys only when they are more interesting for investors, the over-representation of small countries with good investment climate and low corruption levels could easily introduce in the sample a spurious correlation between corruption, openness (negative) and ethnolinguistic fractionalization (positive). 18╇
38
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
from the empirical analysis countries that have less than 1,000,000 inhabitants.20 The total number of countries included in the analysis is around 105 in most regressions, depending on the availability of data. Lastly, from the descriptive statistics of Table€3.2 in the Appendix, note that the corruption index has a mean of 0 and a standard deviation of around 1.21 Thus, our results can be easily interpreted in standardized form.
3.4 Empirical Analysis In this section we provide cross-country estimates of the sources of corruption, highlighting where and why our data and results differ from previous studies. Following Treisman’s (2000) methods, we will firstly test the hypotheses—already summarized in the previous section—based on more stable variables and subsequently include variables that can change over time and that can more easily be influenced by public policies. Thus, our regressions in Table€3.1 proceed first with the inclusion of historical variables, then we include variables that are determined in the medium-short run (and we omit the historical variables that are not significant), and finally we include policy-amenable variables (omitting the insignificant variables). Our preferred estimation technique is weighted least squares regressions, where the weights of the estimates are the reciprocal of the margins of errors of the corruption index for each individual observation (as estimated by Kaufmann et€al. 2005). Thus, the estimates of corruption perception that are more uniform across sources are considered more reliable and get a higher weight in the regressions.22 In our regressions, we begin by including—as independent variables—a dummy variable for the common law system, a dummy for British colonies including the UK itself,23 a measure of the percentage of Protestants in the population, a measure of ethnolinguistic fractionalization, and a resource abundance measure based on the percentage of fuels, metals, and minerals in exports. These variables are predetermined by the history of the country or by natural characteristics and can be considered fixed over long time scales. The dependent variable is the corruption index from the World Bank for the year 2004. The results are reported in Table€3.1.
Depending on the regression, the number of observations that drop out ranges between 16 and zero. 21╇ The standard deviation is exactly equal to 1 in the complete sample, but because of missing data it changes slightly in each sample. 22╇ Our results hold also for ordinary least squares estimations, but as expected weighted least squares produce more “precise” estimates (i.e. slightly higher t statistics). 23╇ Being a former-British colony should affect the degree of corruption because of the lasting effect British occupation has on the organization of the civil service. The UK has that same civil service organization and this is the reason for including the UK together with its former colonies in the dummy. Furthermore, we find the dummy to be statistically insignificant and excluding the UK would further strengthen this result. 20╇
−0.13 (0.32) −0.15 (0.28) −2.06*** (0.35) 1.53*** (0.30) 0.82*** (0.26) −2.21*** (0.34) 1.36*** (0.22) 0.85*** (0.23) −0.60 (0.47)
(2)
−0.72*** (0.06)
−1.10*** (0.25) −0.13 (0.21) 0.41** (0.21)
(3)
−1.07*** (0.13) −0.08 (0.27)
−0.85*** (0.30) 0.03 (0.23) −0.45 (0.51)
(4)
0.03 (0.29)
−0.71*** (0.08)
−1.10*** (0.25) −0.11 (0.21) 0.41* (0.23)
(5)
−0.37** (0.17)
−0.64*** (0.08)
−0.91*** (0.24) −0.13 (0.21) 0.40* (0.21)
(6)
−0.35** (0.17) −0.12*** (0.05)
−0.55*** (0.09)
−0.73*** (0.25) −0.18 (0.20) 0.40* (0.22)
(7)
−0.39** (0.18) −0.15** (0.07) 0.03 (0.08) 1.29*** (0.43)
−0.53*** (0.09)
−0.60* (0.31) −0.16 (0.21) 0.44* (0.22)
(8)
−0.34* (0.19) −0.08 (0.10)
−0.67*** (0.11)
−0.82** (0.37) 0.08 (0.22) 0.52* (0.29)
(9)
1.66*** (0.45) −0.04 Government wage (0.03) R2 0.41 0.44 0.77 0.89 0.78 0.79 0.80 0.81 0.85 106 107 106 42 104 106 104 103 72 Number of countries WLS estimation with the Corruption Perception Index as dependent variable and the reciprocal of errors’ estimates as weights. The constants are included in the regressions, but the coefficients are omitted in the table. Robust standard errors are in parenthesis under the coefficients. The variance inflation factor is lower than 10 for efvery variable (indicating that collinerity is not a major problem) *10% significance; **5% significance; ***1% significance
Political turnover
Government intervention
Newspaper circulation
30 years of uninterrupted democracy
Contemporary democracy
Decentralization
Income
Imports
Fuels and minerals
Ethnolinguistic fractionalization
Protestants
British colony
(1)
Dependent variable: Corruption
Common law
Table 3.1↜渀 Regressions results 3.4 Empirical Analysis 39
40
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
In regression (1), we find results that throw some doubt on many conventional findings. Specifically, while as in most previous analyses we do find a strong correlation with some of the country-fixed factors, we find that neither the existence of a common law system or having been a British colony are aspects associated with a country’s corruption levels. This result challenges the hypothesis that having adopted the British law system or having a past as a British colony reduces corruption in the present day (e.g. Treisman 2000; Glaeser and Shleifer 2002). When we compare Treisman’s work, where he found a British colonial past to be an important determinant of present levels of perceived corruption, with the model and variables used in this study, we find that our dataset is largely expanded (there are up to 107 countries in our regressions, almost double the number of countries in the largest dataset from Treisman) and that the index of corruption we are using has a lower standard error associated with it.24 We ascribe our finding to the fact that our dataset includes countries that were not included in previous studies, are former British colonies, and are also perceived to be very corrupt.25 As a British colonial past tends to overlap with the common law system,26 collinearity is a risk in considering these highly correlated variables, and we also include them separately in the regressions.27 We still find the two variables to be individually non-significant. In our analysis, one factor that significantly affects the level of corruption in all regressions with sufficiently large sample size is the percentage of Protestants in the population, which is negatively associated with the level of perceived corruption. This finding relates to the theories of culture and the work of those (Weber 1904/2002 and Landes 1999 among others) that suggests that religion is fundamental in shaping culture. The influence of the presence of Protestantism on corruption is confirmed over the different specifications and is always statistically very significant. It must be noted that, from regression (2) to (9), the introduction of other independent variables more than halves the coefficient of Protestant religion prevalence. The conversion of 20% of the population from a non-protestant religion to the protestant religion (a one-standard deviation change) would be associated with a reduction in the corruption index by 0.2 points on an approximate 0–4 scale, that is one fifth of a standard deviation. We also find ethnolinguistic fractionalization to be associated with corruption (a result similar to La Porta et€al. 1999). Specifically, a one standard deviation increase
For a thorough comparison of the corruption perception index from Transparency International and the one from the World Bank, see Kaufmann et€al. (2005). 25╇ Our dataset includes former British colonies such as Myanmar and Sudan, which rank among the countries where corruption is perceived to be the highest in the world. Data on these countries has only recently become available. 26╇ The list of countries, in our dataset, that experienced British control, but did not adopt the British legal system are: Egypt, Iraq, Jordan, Kuwait, Myanmar, Mauritius, and Oman. Countries, included in the dataset, that adopted the British legal system without being colonies are: United Arab Emirates, Liberia, Namibia, Saudi Arabia, Somalia, and Thailand. 27╇ We also computed the variance inflator factors, for both variables, which were well below the conventional level of 10. 24╇
3.4 Empirical Analysis
41
in our ethnic fractionalization measure is associated with an increase of the corruption index by approximately half standard deviation. Associating the prevalence of fuels and minerals with corruption is not a new concept. In an influential paper, Leite and Weidmann (1999) found that corruption levels are statistically associated with natural resources and provide a rent-seeking explanation. When compared to their analysis, our substantially expanded dataset still finds the same association, though the relation is not as strong as in their results. Thus, in our model, a change of 25% in the value of natural resources on exports (i.e. a one standard deviation change) would be associated with a change of 0.2 points in our corruption perception index. In regression (2) we include, as a proxy for openness, the share of imports in GDP averaged over of a period of 10 years. We find it not to be a significant predictor of the corruption index. This result differs from Treisman who found imports over GDP to be associated with a decrease of the corruption index, but it corresponds with the results from Knack and Azfar (2003) who have disputed the link between trade intensity and corruption. The difference between results can be ascribed to the propensity of smaller datasets to be affected by selection bias. That is, earlier and less comprehensive surveys tended to include small economies only when they were of interest for international investors who are the main source of funding for the surveys. The selection thus favoured small countries that were open and not very corrupt. The inclusion of income as an independent variable in regression (3), causes some major changes in all coefficients.28 Most notably, the ethnic division proxy becomes insignificant (and remains insignificant after the inclusion of other explanatory variables). It appears that ethnic diversity is not correlated per se with corruption, but through lower income levels or through other variables correlated to income. A similar effect occurs for the coefficients of the Protestants variable and of natural resources in exports, though these coefficients remain statistically significant.29 At the same time, we must emphasize that the inclusion of income as a independent variable runs the risk of creating an endogeneity problem: since corruption, and institutions in general, have been found to affect the growth rate of income, and since institutional quality tends to be stable over time, it is possible that corruption determines income levels rather than the other way round.30 A possible strategy for tackling the issue of endogenity is the use of instrumental variables. We identify
The income variable refers to 2001. Indeed, for natural resources there is a large literature on the “resource curse” and the “Dutch disease”, which have shown the detrimental effect that natural resources have on economic growth (Stevens 2003). 30╇ On the direction of causality between institutions and income there is a large and growing empirical literature. Most of the authors seem to agree that causality runs from institutions to income, rather than the other way around (e.g. Rodrik et€al. 2004; Acemoglu et€al. 2001). For an example of an econometric study finding the opposite direction in the causality between growth and institutions, see Chong and Calderon (2000). 28╇ 29╇
42
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
latitude as a good instrumental variable31 (it is correlated with income at 60%). Latitude has been suggested as a determinant of income by the proponent of theories of geography as a determinant of economic development (e.g. Sachs 2001). Once latitude is inserted among the independent variables, acting as an instrument for income, the coefficients of the other independent variables tend to increase in magnitude and to become more statistically significant (not shown). Most notably, the variables describing democracy, newspapers circulation and political turnover (see below) increase their coefficients. But, equally important, we do not find any of the variables that we found not to be significant in the main analysis to become significant.32,33 Finally, the hypothesis of exogeneity cannot be rejected by the Hausman test at 5% confidence. While the income endogeneity issue can be taken care of via the use of instrumental variables, we prefer an alternative solution, using a lagged income variable. The reason is that we are not so much interested in the coefficient of income itself, which can be biased due to endogeneity, but we are more interested in the use of income as a control variable. Our aim is to filter out that part of the coefficients for other independent variables that may go through income. Summing up, choosing to present results from the WLS regressions—with income included among the independent variables—we favour type II errors, but the difference with the results with the instrumental variables would not alter substantially our conclusions. One of the main findings of Treisman (2000) that contradicts existing literature (e.g. Fisman and Gatti 2002) is the trend for federalism to increase corruption. We use, as a proxy of power decentralization, the share of expenses that are delegated to local authorities as compared to the central government.34 Once we include such a proxy in our regressions, we do not find that decentralization has a positive correlation with corruption. On the contrary, we find the proxy for decentralization to have a negative correlation with corruption, though the coefficient is not statistically significant. We interpret this result as a weak suggestion that federalism does
31╇ A good instrumental variable must be highly correlated with the variable to be instrumented and should not have additional explanatory power. 32╇ The only exception is contemporary democracy that becomes significant. In the analysis below, we prefer to use and discuss the measure of medium-term persistence of democracy, because it is significant even with the inclusion of income. 33╇ A note of caution is needed when we analyze the results with the instrumental variable, because theories that link geographical factors to institutions and through them to income levels have been developed (Hall and Jones 1999; Acemoglu et€al. 2001). If these theories are correct, latitude could be used as an instrumental variable for corruption as well, and the interpretation of the 2-stage results would become problematic. For our own dataset, we checked whether latitude could be used as an instrument for corruption in a regression on income and found indications that latitude would not be a valid instrument, because it retained explanatory power when added to corruption in the regression. 34╇ Unfortunately, our proxy for decentralization is available just for a small sample of countries. Once more statistics on government finance, uniform across countries, are available a more reliable empirical test of the link between decentralization and corruption will be possible.
3.4 Empirical Analysis
43
not increase corruption.35 One reason for the difference between our findings and Treisman’s (2000) finding can lie in the fact that Treisman uses a dummy variable to characterize a state as federalist or centralist based on an incomplete list of federal countries in Elazar (1995). Apart from its incompleteness, another problem of this list is that an officially federalist country may in fact be overly centralized. Mexico is a case in point.36 The evidence is less than conclusive, as it is based on a sample of only 42 countries, but deserves to be mentioned as a matter that merits further inquiry (see also Arikan 2004). In the remaining regressions, the decentralization proxy is omitted in order to keep a larger sample of countries. The inclusion of a contemporary democracy variable in regression (5) does not produce a significant coefficient. This result is different when compared to most of the literature that specifically tests the democracy hypothesis. We conjecture that the reason for this is that most other models work with fewer control variables (e.g. Hill 2003; Chowdhury 2004) leading to omitted variable bias with smaller samples (Sandholtz and Koetzle 2000). In the table, we introduced a proxy for democracy from the Polity IV dataset (measuring democratic levels of institutions as judged by a panel of experts, averaged over the period 1994–2003). As an alternative (not reported in the table), we used a proxy from the Polyarchy dataset (measuring participation and competition at the elections through a mathematical interpretation of elections results, also for the period 1994–2003). None of the two democracy’s proxies seem to affect the corruption level and only when we exclude the income variable do the democracy variables become significant. Thus, it is possible that there is an indirect effect of democracy levels through income on corruption (see Barro 1996; and the literature that has sprung from his work). Constructing a dummy variable for stable democracies37 (where there are no coups, or other major interruptions in the nature of the government, and the country has democratic institutions in terms both of elections and of administration of power), we find that a long exposure to democracy has a mitigating effect on corruption. Thus, a “30 years of uninterrupted democracy” variable (as reported in regression (6)) is significant at 5%. The absolute size and the statistical significance of the coefficient increase if more decades are included in the dummy. A dummy characterizing countries that experienced 50 years of uninterrupted democracy is To be sure, we also included, as a proxy of the size of the country, the natural logarithm of the population (as in Fisman and Gatti 2002), to account for the fact that countries with different size may have different “natural” centralization levels. Conforming with previous literature, we did not find the variable to be significant or to affect the coefficient of the decentralization variable. 36╇ In the United States of Mexico, central government spending exceeds the States and the local governments spending together by more than three times. While the Mexican constitution is of federal nature, political power is centralized in the country’s capital. “For most of the seven decades of rule by the Institutional Revolutionary Party (PRI), Mexico was a highly centralized one-party polity. State governors, and even many mayors, were named by the president and answered to him, even if they were duly elected, by fraud if need be”, see “Mexico’s truncated moves towards real federalism”, March 27th 2003, From The Economist print edition. 37╇ We set the cut off point at the level of seven on a 0–10 scale of democracy in the Polity IV variable. 35╇
44
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
significant at 1%. Our finding thereby presents an intermediate position where it is not the contemporary level of democracy that is a significant predictor of corruption (e.g. Chowdhury 2004), but it is also unnecessary to have 46 years of uninterrupted democracy as in Treisman’s (2000) analysis. According to our estimates, more than 10, but less than 46 years of a persistently democratic regime are sufficient to produce a significant dividend in terms of corruption reduction. An issue related to democracy that attracted some attention recently is the role of press. We include the variable of newspaper circulation in order to catch the possible effect that mass medias can play in reducing corruption. Our empirical finding (regression (7)) supports the hypothesis that countries where access to press is more wide spread will have less corruption.38 Finally, there are the theories related to government policies. Thus, we have included in regressions (8) and (9) an index of government intervention in the economy, a proxy of political turnover, and a measure of wages in the public sector. The proxy for government intervention in the economy is computed using government consumption as a percentage of the economy, government ownership of businesses and industries, the share of government revenues from state-owned enterprises, and government ownership of property and economic output produced by the government. We do not find support for the hypothesis that government intervention is associated with corruption. As to the interpretation of this result, we highlight that an increase in government activities in the economy can cause more opportunities for rents’ appropriation, but also increase the availability of state-provided services and decrease the need to resort to corruption to access them. The political turnover is obtained averaging the number of “veto players” changes in the political system each year. The variable we employ is a proxy of political turnover that recognizes the diffuse nature of political power in most countries. It computes the total number of veto players in the political structure and then evaluates the number of them that have changed in a year. The results show that political turnover is significantly associated with corruption, thus the shortening of the time horizons of politicians would affect corruption levels. While evidence of this relation is tentative and would deserve further testing, we highlight that this result is mostly due to the increase in sample size and the use of a better proxy for political turnover as compared to Treisman’s work.39 Finally, we add government wage as an independent variable. We add this government-related variable last, because of the fact that the variable is available only for a reduced sample (72 countries). The government wage is obtained by dividing the average wage in the public sector by GDP per capita. The government wage variable has the predicted negative sign, but is significant only at 12%. Since the We also checked whether the interaction term between contemporary democracy and newspapers circulation would be significant, controlling whether a widespread press circulation together with a democratic regime would have a special effect on corruption levels. In our regressions the interaction term was not significant. 39╇ Treisman employed a variable that simply stated the number of government leaders changes in each year. 38╇
3.5 Limits and Interpretation of Econometric Cross-Country Analysis
45
sample is relatively small, it is possible that with more data on government salaries and a better proxy for the opportunity costs faced by civil servants if they lose their jobs it would be possible to get a significant coefficient.40 As a robustness check of our findings, we have run ordinary least squares regressions with identical specifications, only with the Transparency International’s Corruption Perception Index for 2004 as a dependent variable.41 We find similar results, but with a slightly smaller dataset and decreased significance of most coefficients.42 There are two differences in the estimates. First, the share of fuels and minerals in exports tends to become statistically insignificant in most regressions, and second, the import share in GDP is significant when income is excluded as an independent variable. Subsequently, we use the corruption perception index of the World Bank for the years 1996, 1998, 2000, and 2002 (instead of 2004) as a dependent variable to check our results, with the independent variables also referring to earlier years than in our main analysis. Again, results are similar with two exceptions. Fuels and minerals in exports tend to be slightly more significant in some regressions, and newspapers circulation tends to be slightly less significant in some other regressions.
3.5 L imits and Interpretation of Econometric Cross-Country Analysis A number of general issues arise from the use of cross-country econometric techniques, such as the ones employed here. Here we present a brief discussion of the main ones that qualify the confidence in the results that can be obtained through such techniques and the policy relevant interpretation of these results. These qualifiers apply to the empirical analyses carried out throughout the book. The first “classic” issue is that of data quality, especially pertinent because of the use of data on corruption that are based on perceptions. Specifically for corruption, concerns with respect to political biases and “halo” effects are of special importance. Political biases can be problematic when the institutions that provide data which are the sources of the aggregated indexes used here are themselves biased and their rankings reflect ideological prejudice rather than governance quality. The issue has been investigated by Kufmann et€al. (2004, pp.€24–25) and it was 40╇ A better proxy would be the ratio of civil servants pay to service or manufacturing salaries (that are not influenced by the share of population employed in the agricultural sector). Van Rijckeghem and Weder (2001) use the ratio of government wages to manufacturing wages and find it to be a significant determinant of corruption levels. Their data sample, though, is limited to 31 countries and data limitations do not allow us to follow their data. 41╇ The estimates if Transparency International’s margins of error have been shown to be unduly large (Kaufmann et€al. 2005) and we preferred the ordinary least squares, as a regression technique, rather than the weighted least squares technique. 42╇ The sample size ranges between 98 and 67 countries.
46
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
found that only one organization tends to produce data that are ideologically biased (namely the Heritage Foundation which favours rightwing governments) but the bias has been found to be quantitatively small and not affecting significantly governance indexes. However, if these prejudices affect the business community as a whole—and the bulk of the interviewees used in the indexes do come from that community—the problem of ideological biases would still be present. The halo effect would take place when countries that are performing well according to development criteria (e.g. they are full-fledged democracies with high rates of economic growth) have a tendency to have their governance indicators overestimated. The underlying mechanism would be the fact that success in development would be interpreted as a symptom of institutional quality. While the possibility of such biases cannot be discounted there is evidence that halo effects should be extremely large to explain by themselves the positive correlation existing between income and governance (Kaufmann et€al. 2005, pp.€32–36). Another issue is that the focus on corruption—and other governance indicators—can be partially data driven. While the availability of high-quality indexes that cover a large set of countries do offer opportunities for empirical research they can also crowd out research from other relevant, but harder to estimate, governance features. A related issue is that studies of one specific aspects of governance—and this book is no exception—cannot easily disentangle related aspects of governance. In the case of studies focused on corruption other variables offer different explanations of the empirical results. For example, a variable such as rule of the law that is highly correlated with corruption since no country can have an effective rule of the law and be marked by high levels of corruption at the same time. It is important to recognize the possibility to have complementary analyses based on alternative indicators and to recognize that—continuing with our example—what we find about corruption might as well hold true for rule of the law (or related governance indicators). Furthermore, the test of alternative hypothesis concerning these closely related indicators is difficult because of the high degree of statistical and conceptual correlation across them. While country heterogeneity is a general concern for all cross-country analyses, in the case of corruption a special concern arises from the fact that it can be argued that in different countries corruption means different things and serves different functions, so that it is difficult to make generalizations. These arguments resonate in the position that corruption is a necessary stage that societies go through on their way towards modernity (e.g. Huntington 1968), but their general significance is challenged by the very fact that empirical regularities do appear to link corruption and other variables concerning human welfare. Finally, special caution should be used when interpreting the results of econometric analysis and suggesting policies based on empirical findings. For example, the finding in the analysis that the prevalence of Protestants in the population is associated with lower corruption should not lead to the conclusion that conversion to Protestantism would be the solution to corruption problems across the world. Rather, the finding should raise questions about the specific values and aspects of Protestantism that contribute to these results. Thus, we could single out the fact that
3.6 Discussion and Conclusions
47
cultures in which challenging authority is more accepted tend to be less tolerant with “abuses of power for private benefit”; this would be the key aspect of Protestant culture that we might endorse with specific policies. Overall, through econometric analysis we obtain leads rather than definitive answers especially in terms of policies. A general point is that quantitative methods should be complemented by other types of analyses (e.g. case studies) rather than see these approaches as alternatives. While in this book we combine different approaches to shed light on different aspects of corruption our effort is less than exhaustive.
3.6 Discussion and Conclusions Our exercise contributes to the existing literature as it questions some central findings of previous econometric studies. Most notably we do not find support for the belief that certain national historical characteristics are a cause of corruption. In our statistical analysis we do not find that the common law system or a past as a British colony (negatively) predicts corruption. Furthermore, we do not find any association between decentralization and corruption. Moreover, the link between ethnolinguistic fractionalization and corruption is diminished and becomes insignificant once income is included in the regressions. We do find systematic evidence that supports some cultural theories of the causes of corruption, in that the presence of Protestants in the population is found to be associated with lower corruption. We also find that richer countries are less corrupt. As mentioned above, caution is needed as there could be reverse causality from institutional quality to income, though the result is upheld using an instrumental variable. Another finding shows that a long exposure (30 years) to uninterrupted democracy is associated with lower corruption, that political turnover tends to raise corruption, and that the diffusion of newspapers is associated with lower corruption levels. Finally, we also find some evidence of an association of higher wages in the public sector with lower corruption. What distinguishes our study from the previous econometric works is, apart from some different variable choices, a larger sample, and we suggest that the most notable differences with earlier studies can be traced to the inclusion of new countries in the present study. We are aware of the limitations on the interpretation of econometric results. Econometrics has a bias towards theories that can easily be quantified. Case studies and more theoretical studies can act as a necessary complement of this type of work (e.g. Johnston 2005). For future research, we hope to follow three approaches. First, we want to investigate determinants of corruption that are amenable to policy changes. This search can be enhanced through the use of econometric analyses and experimental techniques that identify the sources of corruption at the micro level (e.g. Fisman and Miguel 2006; Miller 2006; Reinikka and Svensson 2006; Schulze and Frank 2003). Second, we consider another challenge to be the collection of sufficient data for a panel data approach over a number of decades. The combined cross-country—or cross-region—and intertemporal analyses pos-
48
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
sible with panel data can be a key to finding other sources of corruption. Finally, there are some sources of corruption that are more important for developing countries and would represent an interesting subject to study in that sub-set of countries. One of the sources of corruption that some authors have been found to be associated with corruption is aid (e.g. Knack 2001; Svensson 2000), but the issue has not yet received much empirical scrutiny.
Appendix Corruption is the Perceived Corruption Index 2004, from the World Bank. It is an aggregate indicator combing information—from a number of sources—measuring the incidence of corruption. The sources include surveys of experts, investors, and citizens’ opinion polls (see Kaufmann et€al. 2005) and the data are available at http:// www.worldbank.org/wbi/governance/data.html. For robustness checks the Corruption Perception Index from Transparency International was used (available at http:// www.icgg.org/). Data on corruption have been rescaled throughout the book so that an increase in the index has the intuitive meaning of increase in corruption. Protestants is the share of Protestants in the population (see La Porta et€al. 1999). Ethnolinguistic fractionalization is an average of five different indexes—based on linguistic groups—measuring the probability that two randomly selected individuals in the population would belong to the different groups and the percentage of population that does not speak the most common/official language (see La Porta et€al. 1999). Fuels and minerals equals to the share of fuels and minerals on exports, averaged over 1993–2002 (from the World Development Indicators, 2004). Income is the natural logarithm of GDP per capita in 2001 (from the World Development Indicators, 2004). Table 3.2↜渀 Descriptive statistics Variables Corruption Protestants Ethnolinguistic fractionalization Fuels and minerals Imports Income Decentralization Contemporary democracy Newspapers circulation Government intervention Political turnover Government wage
Obs 106 106 106 106 105 106 42 104 104 106 105 73
Mean 0.00 0.13 0.33 0.20 0.39 8.38 0.33 0.72 1.10 2.94 0.14 3.25
Std. Dev. 1.07 0.21 0.30 0.26 0.20 1.19 0.28 0.29 1.48 0.75 0.10 2.47
Min −2.47 0.00 0.00 0.00 0.09 6.20 0.02 0.05 0.00 1.50 0.00 0.60
Max 1.55 0.98 0.89 0.97 1.34 10.39 1.33 1.00 7.60 4.70 0.41 11.80
References
49
Decentralization is the expenses of state and local government divided by the central government averaged over 1993–2002 (from the “Government Finance Statistics 2004” of the International Monetary Fund). Contemporary Democracy is the average of the institutional democracy score for the years 1994–2003 from the Polity IV dataset (the “polity” variable in the original dataset). Democracy is measured along three lines: the first is the influence of citizens in the choice of leaders and policies, the second is the existence of constraints on the exercise of power by the executive, and the third is the guarantee of civil liberties to all citizens with respect to their daily lives and to political participation. The score is obtained as a weighted sum of the components and the scores are given by experts. The original indicator has been rescaled to a 0–1 scale (see http://www. cidcm.umd.edu/inscr/polity/). Newspapers Circulation is daily newspapers circulation for ten people (from the World Development Indicators 2004). Imports is a measure of the openness of the economy and equals to the share of imports over GDP, averaged over 1993–2002 (from the World Development Indicators, 2004). Government Intervention is an index for 2004 of the influence of government on the economy based on government consumption as a percentage of the economy, government ownership of businesses and industries, the share of government revenues from state-owned enterprises, and government ownership of property and economic output produced by the government (from The Index of Economic Freedom 2005). Political Turnover is the percentage of veto players in the political system that changed every year, averaged for 1991–2000 (see the Database of Political Institutions 2000 of the World Bank, Clarke et€al. 1999). Government Wage is the average government wage as a multiple of GDP per capita (see Schiavo-Campo 1998). British Colony is the dummy variable for countries that have been under British control (from Treisman 2000 augmented with information from Flags of the World Website http://flagspot.net/flags/gb-colon.html). Common Law is the dummy variable for countries that adopted the common law system in their commercial code (see La Porta et€al. 1999).
References Acemoglu, D., & Verdier, T. (2000). The choice between market failures and corruption. American Economic Review, 90(1), 194–211. Acemoglu, D., Johnson, S., & Robinson, J. A. (2001). The colonial origins of comparative development: An empirical investigation. American Economic Review, 91(5), 1369–1401. Arikan, G. G. (2004). Fiscal decentralization: A remedy for corruption? International tax and public finance, 11(2), 175–195. Barro, R. (1996). Democracy and growth. Journal of Economic Growth, 1(1), 1–27.
50
3 Causes of Corruption: A Survey of Cross-Country Analyses and Extended Results
Becker, G. (1968). Crime and punishment—Economic approach. Journal of Political Economy, 76(2), 169–217. Bohara, A. K., Mitchell, N. J., & Mittendorff, C. F. (2004). Compound democracy and the control of corruption: A cross-country investigation. Policy Studies Journal, 32(4), 481–499. Brunetti, A., & Weder, B. (2003). A free press is bad news for corruption. Journal of Public Economics, 87(7–8), 1801–1824. Chafuen, A. A., & Guzmán, E. (1999). Economic freedom and corruption. In G. P. O’Driscoll, K. R. Holmes, & M. Kirkpatrick (Eds.), 2000 index of economic freedom. Washington: Heritage Foundation. Chong, A., & Calderon, C. (2000). Causality and feedback between institutional measures and economic growth. Economics and Politics, 12(1), 69–81. Chowdhury, S. K. (2004). The effect of democracy and press freedom on corruption: An empirical test. Economics Letters, 85(1), 93–101. Clarke, G., Beck, T., Keefer, P., et€al. (1999). New tools and new tests in comparative political economy: The database of political institutions. Washington: World Bank. Elazar, D. J. (1995). From statism to federalism—a paradigm shift. Publius—the Journal of Federalism, 25(2), 5–18. Fisman, R., & Gatti, R. (2002). Decentralization and corruption: Evidence across countries. Journal of Public Economics, 83(3), 325–345. Fisman, R., & Miguel, E. (2006). Cultures of corruption: Evidence from diplomatic parking tickets. NBER working paper. National Bureau of Economic Research. Glaeser, E. L., & Shleifer, A. (2002). Legal origins. Quarterly Journal of Economics, 117(4), 1193–1229. Hall, R. E., & Jones, C. I. (1999). Why do some countries produce so much more output per worker than others? Quarterly Journal of Economics, 114(1), 83–116. Hill, K. Q. (2003). Democratization and corruption: Systematic evidence from the American states. American Politics Research, 31(6), 613–631. Huntington, S. P. (1968). Political order in changing societies. New Haven: Yale University Press. Johnston, M. (2005). Syndromes of corruption: Wealth, power, and democracy. Cambridge: Cambridge University Press. Kaufmann, D., Kraay, A., & Mastruzzi, M. (2004). Governance matters III: Governance indicators for 1996–2002. World Bank policy research working paper, World Bank, Washington DC. Kaufmann, D., Kraay, A., & Mastruzzi, M. (2005). Governance matters IV: Governance indicators for 1996–2004. World Bank policy research working paper series, World Bank, Washington DC. Knack, S. (2001). Aid dependence and the quality of governance: Cross-country empirical tests. Southern Economic Journal, 68(2), 310–329. Knack, S., & Azfar, O. (2003). Trade intensity, country size and corruption. Economics of Governance, 4(1), 1–18. La Porta, R., Lopez-De-Silanes, F., Shleifer, A., et€al. (1999). The quality of government. Journal of Law Economics & Organization, 15(1), 222–279. Lambsdorff, J. G. (1999). Corruption in empirical research: A review. Berlin: Transparency International. Landes, D. (1999). The wealth and poverty of nations: Why some are so rich and some so poor. New York: Norton. Leite, C., & Weidmann, J. (1999). Does mother nature corrupt? Natural resources, corruption, and economic growth. IMF working paper, International Monetary Fund, Washington DC. Macaulay, T. B. (1843/2001). Critical and historical essays: Vol.€ 1. Seattle: The World Wide School. Mauro, P. (1995). Corruption and growth. Quarterly Journal of Economics, 110(3), 681–712. Miller, W. L. (2006). Corruption and corruptibility. World Development, 34(2), 371–380. Pellegrini, L., & Gerlagh, R. (2008). Causes of corruption: A survey of cross-country analyses and extended results. Economics of Governance, 9(3), 245–263. Reinikka, R., & Svensson J. (2006). Using micro-surveys to measure and explain corruption. World Development, 34(2), 359–370.
References
51
Rodrik, D., Subramanian, A., & Trebbi, F. (2004). Institutions rule: The primacy of institutions over geography and integration in economic development. Journal of Economic Growth, 9(2), 131–165. Rose-Ackerman, S. (1999). Corruption and government: Causes, consequences, and reform. Cambridge: Cambridge University Press. Sachs, J. D. (2001). Tropical underdevelopment. NBER working paper. Sandholtz, W., & Koetzle, W. (2000). Accounting for corruption: Economic structure, democracy, and trade. International Studies Quarterly, 44(1), 31–50. Schiavo-Campo, S. (1998). Government employment and pay: The global and regional evidence. Public Administration and Development, 18, 457–478. Schulze, G.-G., & Frank, B. (2003). Deterrence versus intrinsic motivation: Experimental evidence on the determinants of corruptibility. Economics of Governance, 4(2), 143–160. Stevens, P. (2003). Resource impact—curse or blessing? A literature review. Journal of Energy Literature, 9(1), 3–42. Sung, H.-E. (2004). Democracy and political corruption: A cross-national comparison. Crime, Law and Social Change, 41, 179–194. Svensson, J. (2000). Foreign aid and rent-seeking. Journal of International Economics, 51(2), 437–461. Transparency International. (2004). Global corruption barometer. Berlin: Transparency International. Treisman, D. (2000). The causes of corruption: A cross-national study. Journal of Public Economics, 76(3), 399–457. Van Rijckeghem, C., & Weder, B. (2001). Bureaucratic corruption and the rate of temptation: Do wages in the civil service affect corruption, and by how much? Journal of Development Economics, 65(2), 307–331. Weber, M. (1904/2002). The Protestant ethic and the spirit of capitalism. Los Angeles, Roxbury. Williams, R. (1999). New concepts for old? Third World Quarterly, 20(3), 503–513.
Chapter 4
The Effect of Corruption on Growth and Its Transmission Channels
Abstract╇ Through growth regression analysis, we estimate the direct effect of corruption on economic growth and the indirect transmission channels, specifically, investments, trade policy, schooling, and political violence. We find that onestandard deviation increase in the corruption index is associated with a substantial decrease in economic growth. The most important channels through which corruption effects economic growth are investment, followed by schooling and trade openness. JEL classification╇ C31 • H50 • O17 Keywords╇ Corruption • Trade • Investment • Education • Political violence
4.1 Introduction In the Buon Governo frescos of Ambrogio Lorenzetti we see a striking early representation of the effects of good governance on society. The frescos—commissioned for the Palazzo Pubblico (Siena, Italy) and carried out in the second half of the 1330s—contain two contrasting allegories: in the Allegory of Good Government we can see the orderly life of thriving cities and countryside, while in the Allegory of Bad Government we see chaos and degeneration. These frescos are the first ones that during the Renaissance do not contain any reference to religious themes; they were commissioned to praise and sanctify the achievements of the Republic of Siena and acknowledge the role of enlightened governance in attaining social welfare. Also in economics the recognition of the role good governments—and especially good governance and good institution in the economic literature—dates back to This chapter is a substantially revised version of Pellegrini and Gerlagh 2004. The main difference between the two versions is that in the present version, we used more recent data and some different data sources. Most notably, in this version the GDP growth rate is for the years 1980–2004, while it was 1975–1996 in the version published in 2004. Pellegrini and Gerlagh 2004 remains a useful reference point in order to test the robustness of the results presented here. L. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9_4, ©Â€Springer Science+Business Media B.V. 2011
53
54
4 The Effect of Corruption on Growth and Its Transmission Channels
Adam Smith’s Wealth of Nations1, but it has recently been revived both by the new institutional economics studies (e.g. North 1990) and by econometric analyses that focused on measures of institutional quality. Among the econometric studies that delved into governance and economic development, corruption has figured prominently. The studies that dealt with corruption commonly found that corruption hinders economic growth (Jain 2001; Boycko et€ al. 1995, 1996; Gupta et€al. 2001; Kaufmann 1997; Mauro 1995, 1997, 1998, Murphy et€ al. 1991; Della Porta and Vannucci 1997; Tanzi and Davoodi 1997). In this chapter, we study empirically the direct and indirect transmission channels through which corruption affects growth levels. Specifically, we focus on the effect of corruption on investment, schooling, trade policy, and political violence, as well as estimate the contribution of the various channels to the overall negative effect on growth. There is a growing interest in the relation between economic growth and the institutional environment. The new interest is partly driven by new data that have become available over the past decades. The Freedom House Indexes of Political Freedoms and Civil Liberties, sometimes referred to by the name of their creator Raymond Gastil, and the indicators from Business Environmental Risk Intelligence are among the indexes that appeared in the early 1970s. Nowadays, we have a rich data set that includes sources that are also used by companies to evaluate investment opportunities in foreign countries. These data cover many aspects of the economic environment that are considered important by economic agents: risk of expropriation, definition of property rights, contract enforceability, infrastructure quality, working of markets, bureaucratic efficiency, political and institutional stability, repudiation of contracts by government, and so forth. Also, data on institutional features have become available for statistical use by international institutions and by policy advisors. The data employed in this chapter fall in the latter category. We use data on corruption, available since 1995, gathered by Transparency International, a non-governmental organisation providing free access to its information. Though nowadays it is common to assume that corruption negatively affects growth, the conclusion is not trivial. According to some earlier authors, corruption is like grease for the economy. Leff (1964) and Huntington (1968) underline two mechanisms through which corruption can foster economic growth. Bribes can help firms to avoid burdensome bureaucratic regulations and they can serve as an incentive to civil servants to accomplish their duties. Lui (1985) argues that agenda-setting and labour efforts of government official can be made more efficient through bribes. 1╇ Specifically, Smith mentions among the duties of the Sovereign “that of erecting and maintaining those public institutions and those public works, which, though they may be in the highest degree advantageous to a great society, are, however, of such a nature that the profit could never repay the expense to any individual or small number of individuals, and which it therefore cannot be expected that any individual or small number of individuals should erect or maintain” (↜An Inquiry Into The Nature and Causes of The Wealth of Nations by Adam Smith 1776, Book 5, Chap.€1).
4.1 Introduction
55
These arguments, in support for the hypothesis that corruption is beneficial for economic output, rely however on static efficiency arguments, and Kaufmann (1997) offers a comprehensive list of practical and theoretical shortcomings. Also, the recent theoretical and empirical literature suggests less optimistic growth scenarios for countries affected by pervasive corruption. North (1990) emphasizes the need of reliable institutions (incompatible with corruption) to defend property rights and reduce transaction costs; institutions are identified as “the underlying determinant of the long-run performance of economies” (1990, p.€ 107). Murphy et€al. (1991) assert that corrupt societies create incentives that stimulate the most talented people to earn their income through bribing rather than in more productive activities. Boycko et€al. (1995, 1996) argue that agreements produced by corrupt practices are inherently unenforceable and this produces an uncertainty that is disadvantageous to the economic process. Mauro (1995) finds, through empirical research, evidence of a negative relationship between corruption and investment, and through this channel, eventually, corruption lowers economic growth. As the basic mechanism behind this finding, Mauro (1998) claims that corruption, when understood as an institution that raises revenues for the administration, has more distortionary effects than taxation because of its illegal character. Economic agents spend substantial efforts to avoid detection and punishment. In addition to the distortionary effect of corruption on the private sector, different authors have argued that corruption also affects the ways policymakers take decisions. Krueger (1993a, b) argues that incompetent policies are not always the result of lack of knowledge, but rather the outcome of decision-makers’ efforts to capture personal rents. With high level of corruption, the allocation of government resources is influenced by bribe opportunities, and resources are allocated to activities with a high potential for bribes to be collected, as opposed to welfare enhancing activities (Della Porta and Vannucci 1997). Bardhan (1997), along the same lines, argues that “because different activities have different chances of detection for bribes, there will be some substitution effect following from corruption by which corrupt officials will try to induce investment and transactions in the direction of lower-detection activities” (p.€1326)2. Tanzi and Davoodi (1997) provide evidence that corrupt government officials direct public investment towards large projects, possibly at the expense of basic expenditures such as education and health. Tanzi and Davoodi put forward both theoretical arguments and empirical evidence supporting the idea that corruption negatively affects the quality of infrastructure. Gupta et€al. (2001) claim that the defence sector is very sensitive to bribes while it does not contribute much to economic growth, and indeed, in their empirical analysis they find a positive correlation between the level of corruption and defence expenses. 2╇ In Italy, court cases (in the “clean hands” trials) have shown that governments were devoting an unusual amount of resources to assist developing countries because of favourable bribe collection opportunities. Bribe collection was easier in this part of public expenditures because of the difficulties of Italian judges to undertake enquiries abroad, especially in developing countries (Bollini and Reich 1994).
56
4 The Effect of Corruption on Growth and Its Transmission Channels
In sum, recent studies on the impact of corruption on the economy indicate that the effects of corruption are negative and pervasive throughout the economy (Jain 2001, p.€72). Yet, we can only estimate empirically the effect of corruption on economic growth when corruption is not endogenous to the growth process. That is, we need to make sure that the causality does not run the other way around, from lowincome levels to corruption. It could be that low income would result in poor institutional settings that, in turn, create incentives for civil servants to collect bribes. However, empirical work suggests that the level of corruption is better explained by the quality of economic institutions, rather than by income. Indeed, Mauro (1995) finds high correlation levels between corruption and other institutional quality indexes. Furthermore, Acemoglu et€al. (2001) and Easterly and Levine (2003) have shown that institutions are very persistent over time and are fundamental determinants of economic growth. This, in turn, implies a high degree of persistence over time of corruption levels, so that we can consider corruption as an exogenous variable when used in regressions explaining recent growth rates. Still, to be sure, when carrying out regression analysis, we checked for the causality among corruption, investments, schooling, openness, political violence, and economic growth, using the share of fuels and minerals on merchandise exports and the share of Protestants in the population as instrumental variables for corruption. In the present chapter, we use a method similar to one developed by Mo (2000, 2001) and elaborated on in Pellegrini and Gerlagh (2004) and in Papyrakis and Gerlagh (2004), to empirically analyse the contribution of various direct and indirect channels through which corruption affects economic growth. Our findings suggest that corruption slows down economic growth, mainly through its effect on investments, schooling, and trade policies. The latter transmission channel has not been taken into account in previous empirical work, while trade openness has been shown to be of major importance for economic growth.3 Our study is also the first whose results are supported by the use of instrumental variables that are chosen according to the results as an extensive analysis (from the previous chapter) and include overidentification and endogeneity tests. Furthermore, we include calculations (with two alternative methods) of the long-run effect of corruption on growth and on the transmission variables. The chapter proceeds as follows. In the next section, basic cross-country regressions are presented with estimates of the direct effect of investments, schooling, trade openness, political violence, and corruption, on growth. In Sect.€4.3, the transmission channels through which corruption affects growth are studied, in Sect.€4.4 their relative importance is estimated and long-term effects on growth are calculated, in Sect.€4.5 estimates of long-run effects of corruption on the transmission variables are estimated. The last section concludes. Furthermore, Appendix€1 illustrates the methodology to derive long-term income effects, Appendix€2 provides robustness checks of our main results, and Appendix€3 provides a description of the variables. 3╇ For example, Sachs and Warner (1995), in their extensive analysis, find strong evidence of convergence among open economies and higher growth rates in economies after trade liberalisation programmes (see also Frankel and Romer 1999).
4.2 Cross-Country Growth Regressions
57
4.2 Cross-Country Growth Regressions In this section, we estimate basic growth regressions to quantify the effect of corruption on economic growth, both in a context with and without other independent variables. At this stage we will not produce an explicit estimate of the transmission channels. We start with the common regression equation in which the dependent variable G denotes the GDP growth rate per year in the period from t0â•›=â•›1980 to tTâ•›=â•›2004: Giâ•›=â•›(1/T)ln(↜YTi/Y0i), times 100%4. We include (the natural logarithm of) the level of initial income, ln(↜Y0i), as an independent variable and expect, according to the conditional convergence hypothesis, that this variable will have a negative coefficient, α1â•›<â•›0. That is, we expect the growth rate of income to be negatively associated with the level of income at the beginning of the period. As a second independent variable, we take corruption, Ci, the coefficient of which, α2, is the subject of the analysis, and as the other independent variables, denoted by the vector Zi, we take the common regression variables in the growth literature (e.g. Levine and Renelt 1992; Sachs and Warner 1995) that are possibly linked with corruption as discussed in the next section: investments, schooling, trade openness, and political violence.
Gi = α0 + α1 ln(Y0i ) + α2 C i + α3 Z i + εi
(4.1)
where the superscript i denotes each country in the sample. C is the variable measuring the extent of the perception of corruption, over the period 1980–1985 (see Sect.€3.3). It covers 49 countries in our sample. More recent indexes cover a larger sample of countries and are highly correlated with the old data, but to make sure that the corruption variable is not endogenous, we preferred to use the earliest data on corruption levels. We consider the corruption data for the period 1980–1985 exogenous on a theoretical basis because of the inertia in formal and informal institutions (North 1990; Acemoglu et€al. 2001; Rodrik et€al. 2004). Apart from theoretical considerations and previous empirical work, we find that the corruption perception indexes are highly correlated over time5 and in our analysis we used instrumental variable techniques to check for endogeneity. As instrumental variables for corruption we used the share of fuels and minerals on merchandise exports and the share of Protestants in the population (that is, the historical determinants of corruption identified in Chap.€3). The use of two instrumental variables allowed us to perform the Sargan overidentification test. The test was never rejected at 10% confidence in the regressions that we used for the calculations of the transmission channels of corruption and for the total effect on growth indicating that the instruments are valid.6 Furthermore, the exogeneity hypothesis could not be rejected by the Hausman 4╇ Only for Uganda the growth rate refers to the period 1982–2004, because of missing values for the years 1980–1981. 5╇ The Corruption Perception Index for 1980–1985 and the one for 2001 have a correlation of 88%. 6╇ Only in the 2 Stage Least Squares equivalent of regression (1), Table€4.1 the p value is slightly above 3%.
58
4 The Effect of Corruption on Growth and Its Transmission Channels
test at 5% confidence, indicating that the use of instrumental variable techniques is not required (see the discussion in Wooldridge 2002, pp.€101, 118). Our data on corruption are from the Corruption Perception Index by Transparency International, the dataset that contains data for the earliest periods of time (if compared to the institutional indicators of the World Bank, for example). The Z vector includes investment, schooling, trade openness, and political violence. The investment variable is the percentage of gross investment (public and private) on GDP in the period 1980–2004. The proper time frame for the investment variable in growth regressions is subject to discussion. We preferred to have an average for the whole period as the most reliable measure of the size of investment in the economy, keeping in mind that there could be an endogeneity problem between the growth rate and investment. For a comprehensive treatment thereof, we refer to Temple (1999). The schooling variable measures the average years of schooling in the population over 25 in the year 1980; this variable is considered an approximation for the investments in human capital. The variable trade openness measures the number of years in which the country has been open for trade according to Sachs and Warner (1995) criteria over the period 1965–1990. The variable political violence measures the averaged sum of revolutions and the number of assassinations per million people per year in the period 1970–1985. A comprehensive description of sources and variables is presented in Appendix€3 of this chapter. Apart from the variables that act as transmission channels for corruption and for initial income, in the appendix we report the same regressions as in the main text, including regional dummy variables and an index of democracy. The regional dummies are included because we do not want our results to be driven just by a particular region and because the proxy we use for trade openness has been criticized for losing explanatory power in growth regressions once a dummy for Africa is included in the regressions (see Rodriguez et€al. 2001). We include a Sub-Saharan Africa and a Latin America dummy, since these are the dummies that produced significant coefficient in most regressions and they characterize the regions that had a marked slowdown in economic growth in the period considered in the analysis. Moreover, the democracy index is included in order not to emphasize too much the effect of corruption as a proxy for institutional quality (see the extensive discussion in Chap.€5). Regression (1) in Table€4.1 presents the results of estimating Eq.€4.1 including only the initial income level and corruption as independent variables. The coefficients have the predicted signs and are statistically significant at 1%: the coefficient on initial income is negative and equal to 1.04 and the one on corruption is negative and equal to 0.38. Corruption has substantial impact on economic growth and income. A decrease in the corruption level of one-standard deviation—e.g. from the position of the Philippines to the one of India, or from the position of Turkey to the one of Spain—increases economic growth by approximately 1% per year,7 and
7╇ From Table€ 4.1, we multiple the standard deviation for corruption (2.76) with its coefficient (0.38), and find 2.76â•›×â•›0.38â•›=â•›1.05.
59
4.2 Cross-Country Growth Regressions Table 4.1↜渀 Growth regressions as in Eq.€4.1 Dependent variable: (1) G1980–2004 Constant lnY1980 (1.03) Investment (4.68) Schooling (2.86) Openness (0.45) Political violence (0.12) Corruption (2.76)
12.863 −1.037 (3.05)***
−0.380 (2.94)***
(2)
(3)
6.30 −1.211 (3.66)*** 0.192 (4.53)*** 0.262 (2.50)** 0.975 (1.86)* −1.646 (1.00)
8.511 −1.348 (3.92)*** 0.190 (4.53)*** 0.227 (2.12)** 0.766 (1.41) −0.710 (0.40) −0.152 (1.32)
0.18 0.55 0.57 R2 Number of countries 49 48 48 OLS estimation with average annual GDP per capita growth rate as dependent variable. Standard deviations are in parenthesis under the independent variables (they are for the regression (2)–(3) sample), absolute t values are in parenthesis under the coefficients *10% significance; **5% significance; ***1% significance
increases the long-term income level by about 175%.8 However, the R2 of the first regression is only 0.18. The relation between corruption and the growth rate can be seen in Fig.€4.1. Regression (2) in Table€4.1 presents the results of regressing growth on all the variables included in the Zi vector (investments, schooling, openness, and political violence), but excluding corruption. The results are consistent with Solow’s growth model and with common findings of the empirical literature. The R2 is equal to 0.55 and most coefficients are significant. The coefficient of lnY1980 is negative; its value of −1.21 is consistent with the conditional convergence hypothesis. The coefficient of investment is positive and equal to 0.19. An increase in investments of 4.68% point, the standard deviation, increases growth rates by 0.90%, and increases the long-term income level by 52%. Schooling has also a positive coefficient, 0.26, and is significant at the 5% level. After investments, the human capital variable explains most of the growth differences; an increase of the variable of one-standard deviation implies an increase in growth of 0.75% per year, and an increase in long-term income of 46%, confirming the hypothesis that the accumulation of human capital spurs economic growth. The coefficient for openness is positive, 0.98 and a one-standard deviation in openness would increase long-run 8╇ From Table€ 4.1, we multiple the standard deviation for corruption (2.76) with its coefficient (0.38), and divide this by the coefficient for initial income (1.04) to calculate the effect on the log of long-term income. The change in income is now calculated as exp(2.76â•›×â•›0.38/1.04)–1â•›=â•›1.75. See Appendix€1 for a derivation and justification of this procedure.
4 The Effect of Corruption on Growth and Its Transmission Channels
Fig. 4.1↜渀 Growth rate of per capita income, in the period 1980–2004, versus corruption, in the period 1980–1985. Growth rates are corrected for initial income effect as in regression (1)
6
e (gdp growth | X)
60
4 2 0
–2 –4
–2
0 2 e (Corruption | X) coef = – .38016269, se = .12928363, t = – 2.94
4
income by 30%. Political violence has a negative coefficient, −1.65, but it is not statistically significant. The third regression (3) in Table€4.1 includes both corruption and the other independent variables. The R2 remains roughly unchanged, relative to the previous regression, while all coefficients, except for initial income, show a decrease in absolute value. Also statistical significance is decreased for most coefficients. The increase in (the absolute value of) the coefficient of lnY1980 can be interpreted as an improvement of the identification of the steady state path of the economy. Relative to the first regression, the coefficient for corruption has dropped to less than a half, and has become insignificant. A change of one-standard deviation in the corruption variable increases growth by 0.41%. The direct effect on the growth rate of a reduction in the corruption index is less substantial, when compared to the contribution of some of the other independent variables. However, this result is misleading. The suggested relative insignificance of corruption is due to the fact that large part of the effect of corruption on growth is transmitted through the other variables, investments, schooling, openness, and political violence, and their coefficients partly reflect the indirect effects of corruption on growth. In the next section, we isolate the indirect effects for each transmission channel.
4.3 Transmission Channels for Corruption Regression (3) only accounted for the direct effects of corruption on growth9. We suggest, in line with the literature, that corruption is a pervasive phenomenon that negatively affects the working of the economy in several ways. In this section we 9╇ It must be noted that the “direct” effect of corruption on growth can be interpreted as that part of the total effect whose transmission channels have not been identified yet.
4.3 Transmission Channels for Corruption
61
explore the transmission channels through which corruption can affect growth as captured by the other variables: investment, schooling, openness, and political violence. We estimate the dependence of the variables in Zi on corruption, according to the following equation:
Z i = β0 + β1 ln Y0i + β2 C i + µi
(4.2)
where β0, β1, and β2 are four-dimensional vectors of coefficients; β1 describes the effect of income at the beginning of the period, the latter describes the effect of corruption on the vector of dependent variables Zi, and μi is the vector of residuals. Before presenting the results, we note that we have to pay due attention to the problem of causality among the various variables. It is not obvious from the outset that causality runs from corruption to the transmission variables. The problem of causality is pressing, since one can easily imagine variables such as openness and political violence to affect corruption, as well as the other way around. We controlled for endogeneity of the corruption variable by use of the share of fuels and minerals on merchandise exports and the share of Protestant as instrument variables for corruption. The Hausman test does not reject the hypothesis of exogeneity of corruption in any of our regressions and the Sargan tests suggest that the chosen instruments are valid10. Furthermore, we checked robustness of our results by including in the regression analysis various other independent variables such as a democracy index and regional dummies. In Appendix€2, we present a detailed report on our robustness checks.
4.3.1 The Investment Transmission Channel As a first transmission channel we consider investments, which are affected by corruption because it adds uncertainty to the returns on investment activities, in two ways. First, bribes bring costs with them, when detected, and second, agreements based on bribes are unenforceable (Boycko et€al. 1995, 1996). Another effect on investment appears when one considers bribes as an additional tax on investment (Wei 2000). The fourth regression, presented in Table€4.2, shows that, indeed, corruption has a negative effect on the investment level. A one-standard deviation decrease in corruption levels increases investments by 1.71% point, which in turn, increases economic growth by 0.32% per year.11 The effect of the investment transmission channel by itself is equal to three-fourths of the direct corruption effect of 0.41% growth per year (calculated above).
10╇ 11╇
The choice of our instrumental variables is in line with the findings of Chap.€3. From Table€4.1 regression (3), we multiple the investment coefficient (0.19) with 1.71.
62
4 The Effect of Corruption on Growth and Its Transmission Channels
Table 4.2↜渀 Indirect transmission channels as in Eq.€4.2 Dependent variable (4) (5) Investment Schooling Constant 37.272 −8.931 −1.310 1.792 lnY1980 (1.30) (4.91)*** (1.03) Corruption −0.618 −0.219 (2.76) (1.61) (1.60)
(6) Openness −0.443 0.146 (1.86)* −0.053 (1.79)*
(7) Political violence −0.070 0.005 (0.20) 0.024 (2.73)***
R2 0.05 0.68 0.40 0.27 Number of countries 49 48 49 49 OLS estimation. Standard deviations are in parenthesis under the independent variables (they are for the regression (5) sample), absolute t values are in parenthesis under the coefficients *10% significance; **5% significance; ***1% significance
4.3.2 The Schooling Transmission Channel Second, the literature emphasizes the effects of corruption on education. Tanzi and Davoodi’s (1997) empirical analysis suggests that corruption lowers the ability of the government to raise revenues, while it increases the amount of public investment in large infrastructure projects. Thus, corruption decreases the availability of public funds for education. Mauro (1998) confirms this relationship by direct estimation, as he finds that government expenditure on education is negatively and significantly associated with corruption. From the coefficients of regression (5), Table€4.2, we find that a one-standard deviation decrease in corruption is associated with an increase of over half a year of schooling of the people above 25 years. This in turn, results in an increase of growth of 0.14% per year.
4.3.3 The Trade Openness Transmission Channel The third transmissions channel deals with the effects on trade policies of corrupt practices in governments. The literature in support of the general argument that corruption distorts policymaking has been summarized in the introduction. The influential paper by Krueger (1974) shows the mechanisms through which import restriction becomes a substantial government-induced source of rents, leading to rent seeking activities. Also, Southgate et€al. (2000) describe the relation between corruption and the allocation of trade licenses, import quotas and the implementation of other trade limitations. The openness variable we employ in our analysis is defined as the share of years the country has been open, according to Sachs and Warner’s (1995) definition, in the period 1970–1989. Indeed, to define countries as open or closed economies, Sachs and Warner used criteria that measure the possibility to capture trade rents and include: the size of black market exchange premiums,
4.4 Direct and Indirect Effects of Corruption
63
import quotas, and tariffs.12 Our estimation for the openness transmission channel thus describes the tendency in corrupt societies to raise trade barriers by regulation, thus creating a potential source of influence and income for policymakers, and bribe income for the customs bureaucracy. Regression (6), Table€4.2, shows that, indeed, a one-standard deviation decrease in corruption is associated with an increase in the openness of a country of 0.15, in turn associated with an increase in growth of 0.11% per year. This transmission channel has a smaller effect when compared to the investment and schooling transmission channels, but it still accounts for almost 30% of the direct effect of corruption.
4.3.4 The Political Violence Transmission Channel The last transmission channel we analyse runs through political violence. It is argued that corruption “challenges the popular legitimacy of democratic institutions, and it feeds political instability and the violence that can flow from it” (Mulloy 1999). Bardhan (1997) mentions the fact that in opinion polls made in developing countries, corruption is usually the most important problem referred to by respondents. Political discontent is fuelled by the perception that corrupt practices are widespread among governments and civil servants, and this discontent creates room for political violence. Mo (2001) argues that corruption bolsters political instability through its effects on income polarisation. Regression (7), Table€4.2 shows that corruption has a positive effect on political violence. A one-standard deviation decrease in corruption is associated with a decrease in the political violence index of 0.07, which in turn, increases growth by 0.05% per year.
4.4 Direct and Indirect Effects of Corruption Now that we have estimated the effect of corruption on the four variables in Z in Eq.€4.1, we can single out the direct and indirect effects of corruption on economic growth, and the effect of the independent variables in Z that are not explained by corruption. Formally, after substitution of Eq.€4.2 into 4.1, we obtain:
Gi = (α0 + α3 β0 ) + (α1 + α3 β1 ) ln Y0i + (α2 + α3 β2 ) C i + α3 µi + εi
(4.3)
where α2 is the direct effect of corruption on growth and α3â•›β2 captures the summed indirect effects of corruption on economic growth, and μi are the residuals of Eq.€4.2. The coefficient estimates of this regression are reported in Table€4.3.
12╇ For a full description of the variables and of the method used see Sachs and Warner, 1995, pp.€64–67.
64 Table 4.3↜渀 Growth regressions as in Eq.€4.3
4 The Effect of Corruption on Growth and Its Transmission Channels Dependent variable: G1980–2004
(8)
Constant lnY1980 (1.04) μ1 (Investment) (4.55) μ2 (Schooling) (1.62) μ3 (Openness) (0.35) μ4 (Political violence) (0.10) Corruption (2.76)
13.272 −1.082 (4.15)*** 0.190 (4.53)*** 0.227 (2.12)** 0.766 (1.41) −0.710 (0.40) −0.377 (3.84)***
R2 0.57 Number of countries 48 Standard deviations are in parenthesis under the independent variables, absolute t values are in parenthesis under the coefficients *10% significance; **5% significance; ***1% significance
Comparing the results of regression (8) with (3), there are two aspects that stand out. First, the coefficient for corruption has become highly significant, and secondly, it has increased by more than a factor 2. When taking into account the transmission channels, corruption is an important variable explaining growth. A one-standard deviation decrease in corruption leads to an increase in growth of 1% per year, for given initial income level. The long-term income level increases by 162%. These results are consistent with regression (1), as reported in Table€4.1, and certainly provide evidences of an outstanding effect of corruption on growth. For any other independent variable, apart from initial income, a one-standard deviation change has less effect. It should be noticed that we have omitted other institutional variables to avoid multi-collinearity, and since corruption is positively correlated with those variables, the corruption coefficient could also capture the effect of other institutional implicit variables. Comparing our results with earlier estimates on corruption and growth, we mention Mauro (1995) who finds that a one-standard deviation decrease of the corruption index increases the annual growth rate of GDP per capita by 0.8 percentage point13, but after controlling for investment the coefficient of corruption is halved and becomes statistically insignificant. Mauro (1995) does not study explicitly the transmission channels. Our results are qualitatively similar to Mo’s (2001) results, who also found the transmission channels to account for more than 80% of the total effect of corruption on growth. However, our results are different in various respects. Mo studied another set of transmission channels and specifically did not The analysis takes into consideration the time frame 1960–1985 and GDP per capita growth rate is regressed on corruption, GDP in 1960, secondary education in 1960 and population growth.
13╇
4.5 The Long-Term Effect of Corruption on the Transmission Variables
65
Table 4.4↜渀 Relative importance of transmission channels, as in Eq.€4.3 Transmission Contribution to α3 (Table€4.1) β2 (Table€4.2) channels α2+α3β2
Relative contribution (%)
Corruption Investment Schooling Openness Political violence Total
39 33 13 11 4 100
0.19 0.23 0.77 −0.71
−0.67 −0.23 −0.06 0.02
−0.15 −0.13 −0.05 −0.04 −0.02 −0.39
include the trade policy transmission channel. Also, Mo studied a shorter time span, based on data from earlier sources, and he seems to have used initial income, instead of the commonly used logarithm of initial income, as independent variable to account for conditional convergence. For these reasons we consider our quantitative results more reliable. We summarize the contributions of the transmission channels in Table€4.4. The direct effect of corruption on growth is captured through the coefficient α2, so that the contribution of the direct effect relative to the total effect is α2/(↜α2+α3â•›β2). The transmission channels are captured through the vector multiplication α3â•›β2. The relative contribution of the direct impact of corruption is 39%, the relative contribution of the investment channel to the total effect is 33%, the relative contribution of the schooling channel to the total effect is 13%, the relative contribution of the openness channel to the total effect is 11%, and the relative contribution of the political violence channel to the total effect is 4%. Thus, the main effect of corruption on economic growth is transmitted through the investment variable. This result is consistent with the empirical literature that underlined the importance of corruption on investment (e.g. Mauro 1995). The effect of corruption on schooling and trade openness are the second most important ones. We emphasize the latter finding as, even though the effect of corruption on trade openness has been noted in the theoretical literature, it has been overlooked by previous empirical analyses. Jointly, the transmission channels studied here account for 61% cent of the total negative effect of corruption on growth.
4.5 T he Long-Term Effect of Corruption on the Transmission Variables In this section we estimate the long-term effects of corruption on the transmission variables, investment, schooling, openness, and political violence. These long-term effects can help us to appreciate the pervasive effect of corruption on growth. There are two approaches through which we can estimate these long-term effects. First, we can use a technique similar to the one we used to calculate the long-term effects
66
4 The Effect of Corruption on Growth and Its Transmission Channels
of corruption on growth. That is, we use the results from the previous section to calculate the direct effect of corruption on the transmission variables as presented in Table€4.2 and add the indirect effect through the income channel based on data presented in Table€4.1 and Table€4.2. Second, we run regressions of the transmission channel variables assuming institutions to be the only determinants of the transmission variables (i.e. we exclude initial income as an explanatory variable). Comparing the results from both approaches provides a robustness check of our findings. The long-term effect of corruption on the transmission variables, as they were estimated in the previous session, can be calculated as:
Z∞ /C = β2 + (−(α2 + α3 β2 )/(α1 + α3 β1 ))β1 .
(4.4)
The first term, on the right hand side, is the direct effect of corruption on the transmission variables (from Eq.€4.2). It measures the effect of corruption on the transmission variables abstracting from income effects. The second term, on the right hand side, multiplies the long-term income effect of corruption (based on coefficients as expressed in Eq.€4.3)14 by the effect of income on the transmission variables (from Eq.€4.2). The second term can be interpreted as the long-term effect of corruption on the transmission variables operating through the income variable. The results are summarized in Table€4.5. A one-standard deviation decrease in corruption levels increases the long-term level of investment by 0.43% points. Furthermore, it increases schooling by 2.34 years (for the population more than 25 years old), increases the openness index by 0.29, and decreases the political violence index by 0.06. These results reveal the substantial long-term effect of corruption on those variables that are drivers of economic growth. As a check for these estimates, we also directly estimate the dependence of the transmission channel variables on the corruption level. We recall from the introduction that the recent literature, backed by various empirical studies, argues that institutions are persistent over time and it suggests that they are the main determinants of the long-term economic performance. Therefore, we may estimate the long-term effects of corruption on the transmission variables omitting the initial income variable as an explanatory variable. The specification of Eq.€4.2 becomes:
Z i = γ0 + γ1 C i + ν i
(4.5)
Table 4.5↜渀 Long-term effects of corruption on transmission variables, as in Eq.€4.4 Investment Schooling Openness Political violence Direct effect −0.62 −0.22 −0.05 0.02 Indirect income effect 0.46 −0.63 −0.05 −0.00 Long-term effect −0.16 −0.85 −0.10 0.02
14╇
See Appendix€1 for a derivation of the long-term income effects.
67
4.6 Conclusions Table 4.6↜渀 Indirect transmission channels calculated as in Eq.€4.5 Dependent variable (9) (10) (11) Investment Schooling Openness Constant 23.898 9.318 1.045 Corruption −0.229 −0.737 −0.097 (2.76) (0.95) (6.87)*** (5.09)***
(12) Political violence −0.023 0.022 (4.17)***
Adjusted R2 0.02 0.51 0.36 0.27 Number of countries 48 48 49 49 Absolute t values for coefficients in parentheses, standard deviation in parenthesis under the independent variable ***1% significance
where the variable ln Y1980 has been omitted as independent variable, γ0 and γ1 are four-dimensional vectors of coefficients; the latter describes the effect of corruption on the vector of dependent variables Zi, and νi is the vector of residuals. The coefficients are presented in Table€4.6. Comparing the results of Table€4.5 associated with Eq.€4.4 and Table€4.6 associated with Eq.€4.5, we observe a close resemblance between the values found for the long-term effect of corruption on the transmission variables. The small gap between the two set of values strongly supports the robustness of our results and the relative persistence over time of corruption levels.
4.6 Conclusions In this chapter, we have studied the effect of corruption on economic growth, directly and through its impact on investment, schooling, trade openness, and political violence. The outcome confirms a negative effect of corruption on growth, a conclusion accepted by most of the literature. The indirect impact of corruption is substantial, and our results justify the recent emphasis that international organisations put on corruption in particular and institutional soundness in general as important determinants of economic development (Meier 2001). Now that a consensus is emerging on the importance of institutional characteristics, the obvious search is for strategies that can be used to attack corruption. This will not prove simple, as the past has shown corruption to be persistent over time. Based on the empirical results, we can search for a better understanding of the transmission channels through which corruption affects economic performance. While combating corruption is a long-term task, an understanding of the transmission channels through which corruption affects the economy may suggest ways to limit the negative, but indirect, effects on growth. Our results suggest that the most important (for their effect on growth) variables that are likely to be distorted by the presence of corruption are investment, schooling, and trade openness. Once we are aware of the effects of corruption on these transmission channels, policy interventions in countries characterized by high levels of corruption should
68
4 The Effect of Corruption on Growth and Its Transmission Channels
take these effects into account. Thus, policies intended to deal with corruption should also consider its effects. For example, in a corrupt country, we can expect that policies to reduce corruption should be complemented by strategies to increase schooling. Furthermore, our estimates highlight how reducing corruption can also have a positive effect on transmission channels that have a value per se, apart from their effect on income. Thus, if we consider that education is something valuable in itself, the detrimental effect of corruption on social welfare should not only be measured by the decreased income of corrupt societies, but also by the decrease in education. In other words, the analysis of the transmission channels can help to draw a more complete picture of the negative social effects of corruption.
Appendix 1: Long-Term Income Effects In this appendix, we derive the long-term income effects as in Eq.€4.11. To be sure, let us recall that the conditional convergence hypothesis assumes that all countries, when the independent variables remain constant, converge to a steady state with the same positive economic growth rate but with different levels of income. Each country converges to its own growth trajectory. The long-term income effects we calculate thus measure the persistent difference in income levels, without assuming that income levels themselves converge. We assume that economic growth G for country i depends on its initial income Y0, corruption C, and a vector of other explanatory variables Z, as described in Eq.€4.1. Since Gi represents income growth over a period of T years, we can re-write Eq.€4.1 as: i (4.6) ln YT − ln Y0i /T = α0 + α1 ln Y0i + α2 C i + α3 Z i + ε0i and after rearranging terms, we derive income for country i at the end of the period (year T). (4.7) ln YTi = α0 T + (α1 T + 1)ln Y0i + α2 C i T + α3 Z i T + ε0i T We are interested in the expected difference in income between two scenarios labelled i and j, each with its own characteristics (↜Câ•›i; Zâ•›i) and (↜Câ•›j; Zâ•›j), and thus we eliminate the error terms:
E( ln (YT )) = (α1 T + 1) ln (Y0 ) + α2 T C + α3 T Z j
(4.8)
where ln (YT ) = ln (YT ) − ln (YTi ), C = C j − C i , and Z = Z j − Z i . To assess the long-term effects of C and Z on income, we assume ΔC and ΔZ constant over time, and we study propagation of income differences over time. After two periods of T years, income differences are equal to:
Appendix 2: Robustness Checks
69
E( ln (Y2T )) = (α1 T + 1)2 ln (Y0 ) + (α1 T + 2)(α2 T C + α3 T Z). (4.9)
After three periods, we have
E( ln (Y3T )) = (α1 T + 1)3 ln (Y0 ) + 1 + (α1 T + 1) + (α1 T + 1)2 × (α2 T C + α3 T Z).
(4.10)
For regressions (1)–(3), we have 0â•›<â•›α1Tâ•›+â•›1â•›<â•›1 (e.g. regression (3) has α1â•›=â•›−0.0135, Tâ•›=â•›24), so that for t→∞, the first term at the right-hand side vanishes and the other terms reduce to
E( ln (Y∞ )) = − (α2 /α1 ) C − (α3 /α1 ) Z.
(4.11)
The ratio −(↜α2/α1) captures the long-term income effect of changes in the corruption index. Similarly, the ratio −(↜α3/α1) captures the long-term impact of changes in other explanatory variables. Taking exponentials we can rewrite the equation and calculate the relative long-term income effect as:
E( Y∞ /Y∞ ) = exp[−(α2 /α1 ) C − (α3 /α1 ) Z] − 1.
(4.12)
For small values of (↜α2/α1)ΔC and (↜α3/α1)ΔZ, we can use the approximation
E( Y∞ /Y∞ ) ≈ −(α2 /α1 ) C − (α3 /α1 ) Z.
(4.13)
Appendix 2: Robustness Checks In this appendix, we present the analysis of causality and robustness of the results. A common way to test for causality is the use of instrumental variables in 2 Stage Least Squares (2SLS) regressions. A good instrumental variable is highly correlated with the instrumented variable and should not affect the dependent variable apart from its effect on the corruption variable. For our analysis of transmission channels, we should find an instrument variable that is highly correlated with corruption, but that is otherwise not related to the transmission variables. In the literature, many variables have been used as instruments for corruption: for example, ethnolinguistic fractionalisation (e.g. Mauro 1995) and legal origins (e.g. Fredriksson and Svensson 2003). Following the results of Chap.€3, we used the share of fuels and minerals on merchandise exports and the share of Protestants in the population as instruments for corruption. The use of multiple instrumental variables has the advantage that the validity of the instruments can be checked with an overidentification test. We run the Sargan overidentification test for all the regressions with instrumental variables. The hypothesis of validity of the instrument was never rejected at 10% confidence in the regressions that we used for the calculations of the transmission channels and for the total effect on growth, indicating
70
4 The Effect of Corruption on Growth and Its Transmission Channels
that the instruments are valid.15 Having verified the validity of the instruments, we checked the exogeneity hypothesis. The Hausman test could not be rejected at 5% confidence, indicating that our OLS regressions are not biased.16 After controlling for causality using the instruments, we provide checks for robustness adding a variable for political freedom and a set of regional dummies. The variable we use for democracy is an average for the years 1970–1980 of the democracy score from the Polity IV dataset (see Plümper and Martin (2003) for a description of the variable and for a recent survey of the literature on democracy and growth; see the appendix of Chap.€3 for a description and sources). Regional dummies have been found significant in many recent empirical analyses and adding them to our analysis makes sure that our results are not driven by geographical factors or by any particular group of countries (e.g. Rodriguez et€al. 2001). In Table€4.7, Table€4.8, Table€4.9, and Table€4.10 we present our results including the control variables. The overall effect of corruption holds surprisingly well: Table€4.7↜渀 Growth regressions as in Eq.€4.1 Dependent variable: (13) G1980–2004 Constant lnY1980 (1.04) Investment (4.66) Schooling (2.89) Openness (0.45) Political violence (0.12) Democracy (3.61) Latin America Sub-Saharan Africa Corruption (2.77)
16.843 −1.373 (4.73)***
−0.062 (0.85) −1.683 (3.39)*** −3.479 (5.47)*** −0.386 (3.48)***
(14)
(15)
6.881 −1.220 (3.72)*** 0.182 (3.51)*** 0.165 (1.56) 0.448 (0.83) −0.257 (0.13) 0.115 (1.38) −0.765 (1.40) −1.302 (1.66)
10.685 −1.393 (4.12)*** 0.155 (2.89)*** 0.123 (1.15) 0.237 (0.43) 0.178 (0.09) 0.062 (0.71) −0.860 (1.60) −1.846 (2.20)** −0.199 (1.62)
R2 0.56 0.63 0.65 Number of countries 48 47 47 OLS estimation with average annual GDP per capita growth rate as dependent variable. Standard deviations are in parenthesis under the independent variables, absolute t values are in parenthesis under the coefficients *10% significance; **5% significance; ***1% significance 15╇ Only in the 2 Stage Least Squares equivalent of regression (1), Table€4.1 the p value is slightly above 3%. 16╇ Hence, the use of instrumental variable techniques is not required and would not produce results that are more reliable than the standard OLS because of the inefficiency of the 2 stage least squares estimator (see the discussion in Wooldridge 2002, pp.€101, 118)
Appendix 2: Robustness Checks
71
Table 4.8↜渀 Indirect transmission channels as in Eq.€4.2 Dependent variable (16) (17) Investment Schooling Constant 45.946 −7.708 −1.349 1.582 lnY1980 (1.53) (3.87)*** (1.04) Democracy −0.818 0.091 (3.61) (3.71)*** (0.88) Latin America −4.500 −0.557 (2.99)*** (0.80) Sub-Saharan Africa −8.503 −1.082 (4.41)*** (1.12) Corruption −0.986 −0.163 (2.77) (2.94)*** (1.03)
(18) Openness −0.113 0.125 (1.48) −0.013 (0.60) −0.289 (2.00)* −0.388 (2.10)** −0.051 (1.57)
(19) Political violence 0.004 0.008 (0.31) −0.010 (1.59) 0.036 (0.82) −0.025 (0.44) 0.015 (1.50)
R2 0.46 0.70 0.48 0.33 Number of countries 48 47 48 48 OLS estimation. Standard deviations are in parenthesis under the independent variables, absolute t values are in parenthesis under the coefficients *10% significance; **5% significance; ***1% significance
Table 4.9↜渀 Growth regression as in Eq.€4.3
Dependent Variable G1980–2004
(20)
Constant lnY1980 (1.04) μ1 (Investment) (3.42) μ2 (Schooling) (1.58) μ3 (Openness)s (0.33) μ4 (Political violence) (0.10) Democracy (3.61) Latin America
16.818 −1.376 (5.05)*** 0.155 (2.89)*** 0.123 (1.15) 0.237 (0.43) 0.178 (0.09) −0.058 (0.84) −1.686 (3.62)*** −3.390 (5.25)*** −0.381 (3.62)***
Sub-Saharan Africa Corruption (2.77)
R2 0.65 Number of countries 47 OLS estimation. Standard deviations are in parenthesis under the independent variables, absolute t values are in parenthesis under the coefficients *10% significance; **5% significance; ***1% significance
72
4 The Effect of Corruption on Growth and Its Transmission Channels
Table 4.10↜渀 Relative importance of transmission channels as in Eq.€4.3 Transmission Contribution to α3 (Table€4.7) β2 (Table€4.8) α2â•›+â•›α3β2 channels
Relative contribution
Corruption Investment Schooling Openness Political violence Total
0.52 0.40 0.05 0.03 −0.01 100%
0.155 0.123 0.237 0.178
−0.986 −0.163 −0.051 0.015
−0.199 −0.153 −0.020 −0.012 0.003 −0.381
the coefficient becomes −0.38 (was −0.39). As for the transmission channels, and their relative importance, we find that the direct effect of corruption is increased and we still find that investment is the most important transmission channel. The main difference is that now the direct effect and the effect through investment together account for more than 90% of the total effect of corruption on economic growth. When we compare our results to the ones of Pellegrini and Gerlagh (2004) we find that the estimate of the total effect of corruption is almost identical. Now the direct effect of corruption and the effect through investment are larger and, conversely, the other transmission channels have smaller effects. Overall, our main results are supported by the use of the instrumental variable, by the use of control variables, and by the comparison with a similar study that used different data sources and time frames. The evidence presented in this section support the main results we reported above.
Appendix 3: Data Corruption is the Corruption Perceptions Index (CPI) from the Transparency International database. The CPI is a composite index available from the period 1980– 1985 and is based on interviews of “credible” sources, see Sect.€3.3. The scores of the index range between 0 and 10 (available at http://www.icgg.org/). Y1980, G1980–2004, are GDP per capita in 1980 and the growth rate of GDP per capita between 1980 and 2004, respectively (from the World Development Indicators, 2004). Investment is the average in the period 1980–2004 of the share of Real Gross Domestic Investment (private and public) on Real GDP per capita (from the World Development Indicators 2004). Openness is the proportion of years in which the country has been open (according to Sachs and Warner (1995) definition) in the period 1965–1990 from “Natural Resource Abundance and Economic Growth” by Jeffrey D Sachs and Andrew M. Warner (from http://www.cid.harvard.edu/ciddata/ciddata.html). Political violence is an average of the number of assassinations per million of inhabitants and revolutions in the period 1970–1985, from “Natural Resource Abun-
References
73
dance and Economic Growth” by Jeffrey D Sachs and Andrew M. Warner (from http://www.cid.harvard.edu/ciddata/ciddata.html). Schooling is the average number of schooling years in the population over the age of 25 in 1980 from: Barro-Lee Data Set for a Panel of 138 Countries Robert J. Barro and Lee Jong-Wha (as updated in April 2000). (http://www.cid.harvard.edu/ ciddata/ciddata.html). Democracy is the average of the institutional democracy score for the years 1994–2003 from the Polity IV dataset (the “polity” variable in the original dataset). The original indicator has been rescaled to a 0–1 scale. (see http://www.cidcm.umd. edu/inscr/polity/).
References Acemoglu, D., Johnson S., & Robinson, J. A. (2001). The colonial origins of comparative development: An empirical investigation. American Economic Review, 91(5), 1369–1401. Bardhan, P. (1997). Corruption and development: A review of issues. Journal of Economic Literature, 35(3), 1320–1346. Bollini, P., & Reich, M. R. (1994). The Italian fight against world hunger: A critical analysis of Italian aid for development in the 1980s. Social Science & Medicine, 39(5), 607–620. Boycko, M., Shleifer, A., Vishny, R. W., et€al. (1995). Privatizing Russia (p.€ix, 165). Cambridge: MIT. Boycko, M., Shleifer, A., & Vishny, R. W. (1996). A theory of privatisation. Economic Journal, 106(435), 309–319. Della Porta, D., & Vannucci, A. (1997). The “perverse effects” of political corruption. Political Studies, 45(3), 516–538. Easterly, W., & Levine, R. (2003). Tropics, germs, and crops: How endowments influence economic development. Journal of Monetary Economics, 50(1), 3–39. Frankel, J.-A., & Romer, D. (1999). Does trade cause growth? American-Economic-Review, 89(3), 379–399. Fredriksson, P. G., & Svensson, J. (2003). Political instability, corruption and policy formation: The case of environmental policy. Journal of Public Economics, 87(7–8), 1383–1405. Gupta, S., de Mello, L. & Sharan, R. (2001). Corruption and military spending. European Journal of Political Economy, 17(4), 749–777. Huntington, S. P. (1968). Political order in changing societies. New Haven: Yale University Press. Jain, A. K. (2001). Corruption: A review. Journal of Economic Surveys, 15(1), 71–121. Kaufmann, D. (1997). Corruption: The facts. Foreign Policy, 107, 114–131. Krueger, A. O. (1974). Political economy of rent-seeking society. American Economic Review, 64(3), 291–303. Krueger, A. O. (1993a). Political economy of policy reform in developing countries. Cambridge: MIT. Krueger, A. O. (1993b). Virtuous and vicious circles in economic-development. American Economic Review, 83(2), 351–355. Leff, N. H. (1964). Economic development through bureaucratic corruption. The American Behavior Scientist, 8(2), 8–14. Levine, R., & Renelt, D. (1992). A sensitivity analysis of cross-country growth regressions. American Economic Review, 82(4), 942–963. Lui, F. T. (1985). An equilibrium queuing model of bribery. Journal of Political Economy, 93(4), 760–781. Mauro, P. (1995). Corruption and Growth. Quarterly Journal of Economics, 110(3), 681–712.
74
4 The Effect of Corruption on Growth and Its Transmission Channels
Mauro, P. (1997). The effects of corruption on growth, investment and government expenditure: A cross country analysis. In K. A. Elliot (Ed.), Corruption and the global economy (pp.€83–107). Washington: Institute for International Economics. Mauro, P. (1998). Corruption and the composition of government expenditure. Journal of Public Economics, 69(2), 263–279. Meier, G. M. (2001). The old generation of development economists and the new. In G. M. Meier & J. E. Stiglitz (Eds.), Frontiers of development economics: The future in perspective (13–50). Washington: World Bank; Oxford: Oxford University Press. Mo, P. H. (2000). Income inequality and economic growth. Kyklos, 53(3), 293–315. Mo, P. H. (2001). Corruption and economic growth. Journal of Comparative Economics, 29(1), 66–79. Mulloy, P. (1999, November 10). Statement on Corruption Delivered by Assistant Secretary of Commerce, Patrick Mulloy, to the OSCE Review Conference, Istanbul. Murphy, K. M., Shleifer, A., & Vishny, R. W. (1991). The allocation of talent: Implications for Growth. Quarterly Journal of Economics, 106(2), 503–530. North, D. C. (1990). Institutions, institutional change, and economic performance. Cambridge: Cambridge University Press. Papyrakis, E., & Gerlagh, R. (2004). The resource curse hypothesis and its transmission channels. Journal of Comparative Economics, 32(1), 181–193. Pellegrini, L., & Gerlagh, R. (2004). Corruption’s effect on growth and its transmission channels. Kyklos, 57(3), 429–456. Plumper, T., & Martin, C. W. (2003). Democracy, government spending, and economic growth: A political-economic explanation of the Barro-effect. Public Choice, 117(1–2), 27–50. Rodriguez, F., Rodrik, D., Hseh C. T., et€al. (2001). Trade policy and economic growth: A skeptic’s guide to the cross-national evidence. Nber Macroeconomics Annual, 15, 261–324. Rodrik, D., Subramanian, A., & Trebbi, F. (2004). Institutions rule: The primacy of institutions over geography and integration in economic development. Journal of Economic Growth, 9(2), 131–165. Sachs, J. D., & Warner, A. M. (1995). Natural resource abundance and economic growth. National Bureau of Economic Research (NBER). Southgate, D., Salazar-Canelos, P., Camacho-Saa, C., et€al. (2000). Markets, institutions, and forestry: The consequences of timber trade liberalization in Ecuador. World Development, 28(11) 2005–2012. Tanzi, V., & Davoodi, H. (1997). Corruption, public investment, and growth. International Monetary Fund (IMF). Temple, J. (1999). The new growth evidence. Journal of Economic Literature, 37(1), 112–156. Wei, S. J. (2000). How taxing is corruption on international investors? Review of Economics and Statistics, 82(1), 1–11. Wooldridge, J. M. (2002). Econometric analysis of cross section and panel data. Cambridge: MIT.
Chapter 5
Corruption, Democracy, and Environmental Policy: An Empirical Contribution to the Debate
Abstract╇ Theoretical and empirical studies have shown that democracy and corruption influence environmental policies. In this chapter, we empirically analyse the relative importance of these determinants of environmental policy. When these variables are jointly included as explanatory variables in a multiple regression analysis, we find that corruption stands out as a substantial and significant determinant of environmental policies, while proxies for democracy have an insignificant impact. Nevertheless, democracy could affect environmental policy stringency given that countries with a history of democratic rule tend to be less corrupt (see Chap.€3). A discussion of our results in the context of the Environmental Kuznets Curve literature follows. We argue that improving environmental quality following increasing income is less probable in developing countries with institutional disarray. Finally, and more optimistically, when considering our results in the context of institutions and growth, we conclude that there is scope for reaping a double dividend: reductions in corruption induce higher economic growth rates and stricter environmental policies. JEL classification╇ H40 • D73 • Q56 • Q58 Keywords╇ Corruption • Democracy • Development • Environmental policy • Institutions
5.1 Introduction The role of democracy in bringing welfare to the people has a long history as a subject for analysis. Plato discussed the deficits of democracy in “The Republic”, and Mill (1859/1982) reflected extensively on majority governments and welfare. The rise of environmental concerns directed public attention to a specific link between democracy and welfare: what is the role of democracy in defining the way societies This chapter is a slightly modified version of Pellegrini and Gerlagh 2006. L. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9_5, ©Â€Springer Science+Business Media B.V. 2011
75
76
5 Corruption, Democracy, and Environmental Policy
deal with environmental issues? A first wave of literature—in the late 1960s and 1970s—was sceptical about the virtues of democracies with respect to environmental protection because of the link between democracy, market economies, and individual freedom (e.g. Ehrlich 1968; Heilbroner 1974).1 However, gaining support from the poor environmental performance of Soviet economies and dictatorships established in Latin America, Asia, and Africa, a new strand of literature identified democracy as a way to promote economic and environmental welfare (e.g. McCloskey 1983; Payne 1995). More recently, academic interest has shifted to the effects of corruption on environmental quality (Lopez and Mitra 2000; Fredriksson and Millimet 2001; Damania et€al. 2003). This interest followed the evident pillaging of natural resources taking place in developing countries dominated by corrupt regimes. It is also in accordance with the increasing awareness of the negative effects of corruption on the economy and on the production and provision of public goods. Thanks to the increasing availability of indexes of institutional qualities, there are now several empirical studies on the effects of democracy and corruption on environmental policy commitment and resource conservation. In general, these studies conclude that democracy is a significant positive and corruption a significant negative determinant of environmental protection. A problem with these studies, however, is that they focus on either democracy or corruption. This is troublesome because the two variables are highly correlated and, therefore, the individual estimation of their effects easily overemphasizes the importance of each variable. In other words, the coefficients suffer from an overestimation bias. The current study is the first to test extensively whether empirical evidence indicates both democracy and corruption to be fundamental determinants of environmental policy stringency when these two institutional variables are used simultaneously as explanatory variables. Alternatively, we may conclude that one of the two variables is the main channel through which environmental policies are affected, whereas the other variable is of secondary importance. Although most of the literature has been concerned with actual resource use— such as deforestation rates, ambient concentration of pollutants, and soil erosion— our focus is on political commitment to the environment. This slightly different scope, compared to the mainstream literature, has two advantages. First, actual resource use and environmental quality levels are affected by several factors (e.g. climate conditions) that are outside the control of policy makers. These external factors are important but difficult to control for econometrically. By studying the link between institutional indexes and environmental policies, our analysis aims to be less vulnerable to external disturbances. Second, the actual levels for environmental variables are often sluggish in adjusting to implemented policies. They are affected by a history of development and present concerns for the environment will 1╇ Similarly, Hardin (1968) argued—in his famous article on The Tragedy of the Commons—that coercion is the only way to limit human population growth and that “freedom to breed will bring ruin to all” (p.€1248). It must be noted though that the type of coercion proposed by Hardin is “mutual” and “mutually agreed upon” and that he argues that that is not a real loss of freedom.
5.2 Theoretical Hypotheses and Their Empirical Testing
77
only have an effect on future environmental quality. Focusing on environmental commitment as an intermediate variable can reveal more direct impacts of contemporary institutional quality. Also Putnam (1993, pp.€65–66), argues that to evaluate the performance of political institutions it is important to focus on measures of output rather than on outcomes. As an example he suggests to focus on environmental policies rather than air quality.2 The chapter proceeds as follows. The next section presents a review of theoretical and empirical findings in previous studies on democracy, corruption, and environmental protection. The following section introduces the data used in the analysis. The subsequent section presents and discusses the main results, the section on implications connects our study to the literature on economic development and the environment, and the last section concludes. Appendix€ 1 presents robustness checks of the main findings.
5.2 Theoretical Hypotheses and Their Empirical Testing Two different strands of literature have addressed the impact of democracy and corruption on environmental policy. These strands seem to have proceeded in parallel, independent of each other. As a result, it is possible that each strand of literature has been overemphasising, theoretically and empirically, the importance of each separate variable. In this section, we review the theoretical and empirical literature on democracy, corruption, and environmental policies. Subsequently, we identify a set of hypotheses that deserve empirical testing.
5.2.1 Democracy and the Environment Conservationist authors writing in the 1960s and 1970s have frequently called for a Hobbesian approach to environmental issues. To them, freedom needs to be constrained for the conservation of common goods in general, and for the environment specifically (e.g. Ehrlich 1968; Hardin 1968; and Heilbroner 1974). A different view emerged in the 1980s, after which most of the literature emphasized the positive ef2╇ van Beers and van den Bergh (1997, 2003) suggest a distinction between input- and outputoriented indicators, where the former “are based on input efforts devoted to environmental protection” and the latter on “concrete results of environmental regulations” (pp.€33–34). The authors prefer output indicators because they reflect countervailing policy interventions (e.g. when subsidies are given to a sector characterized by environmental regulations that increase costs) (van Beers and van den Bergh 2003, p.€17). The indexes we use do not have this particular shortcoming because they are comprehensive indexes that would reflect countervailing policies (e.g. the environmental regulatory regime index includes a variable describing the existence of subsidies that encourage the inefficient use of energy or materials). In any case, the alternative indicators used by the authors were available for much smaller sample sizes than the ones used here.
78
5 Corruption, Democracy, and Environmental Policy
fects of democracy. Mounting evidence of the poor environmental performance that characterized the Soviet Union, Eastern European countries, and the dictatorships of Latin America, Africa, and Asia was one of the factors behind this shift. After 1980, the arguments commonly cited in favour of democracy’s positive effect on the environment connect democracy with citizens’ freedom, the availability of information on environmental degradation, and the ability to protest against it. Moreover, the responsiveness of democracies to citizens’ demands, the propensity of democracies to engage in international cooperation, and the coincidence of market economic systems with democracies are stressed (e.g. McCloskey 1983; Payne 1995). Also, a recent case study of the Canary Islands finds that the main cause of environmentally disruptive decisions is the deterioration of democratic institutions (Aguilera-Klink and Sánchez-García 2005). As for the theory on democracy and environmental policy, an influential study has been conducted by McGuire and Olson (1996). They analyse the optimal behaviour of an autocrat in providing public goods. In their model, an increase in the size of the elite would bring about a more efficient solution with higher levels of public goods. The size of the ruling class, which could be considered a measure of democracy, would positively affect the provision of public goods such as environmental quality. Deacon (2003) presents an adaptation of McGuire and Olson’s model and provides empirical results that support the interpretation of environmental quality as a public good, and ascertain a positive effect of democracy on environmental quality. Torras and Boyce (1998) also find evidence of the positive effect democracy has on environmental quality when they estimate the Environmental Kuznets Curve (EKC) for sulphur dioxide, smoke, heavy particles, dissolved oxygen, faecal coliform, availability of safe water, and sanitation. Harbaugh et€al. (2002) provide similar results when they include a democracy index in their estimate of the EKC; they find a consistent negative relation between sulphur dioxide and democracy levels. Others have studied the link between democracy and environmental policy instead of targeting the environmental variables themselves. Congleton (1992) estimated the positive effect of democracy on the probability of signing the global convention on the reduction of emissions of ozone depleting substances.3 Neumayer (2002) presents statistical evidence of the positive effect of democracy on the degree of environmental commitment of countries. He uses the probability of signing multilateral environmental agreements, participating in environmental intergovernmental organizations, the area of a countries’ territory under protection, the presence of national councils on sustainable development, and the availability of environmentally relevant information as measures for environmental commitment. Fredriksson et€ al. (2005) present a theoretical model where democracy is interpreted in terms of increased participation of citizens in the electoral process 3╇ The analysis studied the probability that countries signed the agreement by 1985 or 1987. Since most countries, democracies as well as non-democracies, have signed the convention nowadays, the finding that democracies are more likely to sign the convention seems less valid. However, the evidence can indicate that democracies, if compared to autocracies, are faster to tackle environmental issues (see Neumayer 2002).
5.2 Theoretical Hypotheses and Their Empirical Testing
79
and competition for the support of the electorate. In this model, democracy induces the administration to better represent public preferences, leading to more stringent environmental policy. The authors support their model with an empirical analysis of maximum lead content in gasoline, as an inverse proxy for environmental policy stringency. Though most of the recent literature suggests a positive effect of democracy on environmental policy stringency, some authors reject the democracy–dictatorship dichotomy and claim that neither market-oriented democracy nor autocracy can solve environmental issues in a satisfactory manner. Dryzek (1987), for example, proposes radical decentralization as a political structure suitable for facing environmental problems.
5.2.2 Corruption and the Environment A large number of studies analyse the effect of corruption on environmental policy implementation. Carter (1997) studies the effects of crime and corruption on waste management and related health risks in the state of New York. The World Bank report on corruption and forestry emphasizes the detrimental influence of corruption on forest management and conservation (Callister 1999). A report on deforestation in Indonesia by the Environmental Investigation Agency and Telapak (2003, p.€1) explicitly asserts, “Forests are being destroyed because Indonesia is one of the most corrupt countries in the world.” In an earlier report on illegal logging of Ramin timber in Indonesia, the Environmental Investigation Agency (EIA) and Telapak (2001) find that formal rules are in place to protect the endangered tree specie and enforcement is relatively easy. The study finds that collusion between enforcement agencies and the “timber barons” is so strong that implementation is unattainable and the police follows the orders of the smugglers, rather than enforcing the law. EIA and Telapak (2001, p.€21) conclude, “at the core of the issue of illegal logging is corruption”. Robbins (2000) introduces a theoretical framework for the analysis of corruption and uses it to study the enforcement of protection for a nature reserve in Rajasthan, India. Robbins finds that the lack of enforcement is fuelled by corruption among foresters and that it leads to substantial habitat destruction. The view that corruption hinders environmental protection is stated plainly by the Environmental Public Prosecutor of Madrid: “the non-compliance with environmental laws has its roots in the corruption of the political system. […] Non-compliance with environmental laws is the best barometer of corruption in a political system” (Valerio 1998 op. cit. Aguilera-Klink and Sánchez-García 2005). Lopez and Mitra (2000) argued, in a theoretical paper, that corruption and environmental policy stringency are characterized by a monotonic (negative) relationship. They present a game theoretic model where the government has two objectives: to be re-elected and remain in power, and to receive direct transfers from lobby groups. Income transfers from lobby groups to the government are interpreted as a measure for the level of corruption. The model defines a Nash equilibrium game
80
5 Corruption, Democracy, and Environmental Policy
and a non-cooperative Stackelberg game with the representative firm as a leader, and the authors show that an increased sensitivity of the bureaucracy to bribes both increases the level of corruption and lowers the level of environmental protection. When possible, firms will bribe the government to tolerate overexploitation of the natural resource. Fredriksson and Millimet (2001) elaborate on this result, but claim that there is a non-monotonic correlation between corruption levels and environmental protection. In their model, an increase in the number of corrupt bureaucrats leads to a decrease in the income transfer per bribe. After a certain threshold of corruption, proxied by the number of corrupt bureaucrats, a further increase in the number of bribes is more than offset by a decrease in the effectiveness per bribe.4 Damania (2002) shows that environmental regulations are ineffective if bureaucrats are highly corrupt. He makes the case for complete deregulation if there is no possibility to reduce corruption. Welsch (2004) estimates the direct and indirect effects of corruption in an EKC framework. The author argues that corruption affects environmental quality because it hinders the formation and enforcement of environmental regulations (direct effect). Furthermore, corruption has an additional effect on the environment because it decreases economic growth. Thus, corruption would have a further impact because income levels affect environmental quality (indirect effect). The indirect effect of corruption can be either positive or negative, depending on income levels (in line with the EKC hypothesis). The total effect of corruption on environmental quality, according to the estimates of the author, is always negative.
5.2.3 Institutions and the Environment Some recent literature links environmental stringency to more complex institutional settings. Fredriksson and Svensson (2003) present a theoretical model where environmental policy making is influenced by political instability and corruption. In the model, political instability is interpreted as the replacement rate for the government administration in power. Political instability makes it less effective for polluting industries to bribe the administration, and thus increases the stringency of environmental policies. Specifically, the model predicts that, whereas in general corruption tends to decrease the stringency of environmental policies, its effect decreases with increasing political instability. The authors present econometric evidence supporting their predictions, based on the corruption proxy from the International Country
4╇ Fredriksson and Millimetet support their finding with econometric data at state level for the United States, using as a proxy for corruption the number of civil servants tried for crimes related to bribery as a share of the total number of public employees. Such a proxy has the obvious shortcoming that it can reflect judicial efficiency. Furthermore, trials are held only after some time following crimes, so there is a considerable time lag between the moment when the corruptionrelated crime is committed and its relevance in the index. Judiciary initiatives can be a symptom of anticorruption campaigns rather than corruption itself.
5.2 Theoretical Hypotheses and Their Empirical Testing
81
Risk Guide (ICRG).5 They also include a democracy dummy in most regressions, but as democracy does not enter the authors’ model, they do not discuss its coefficients. It must be noted that, in most of their regressions, the sign of the democracy variable is negative and significant (at 10% level).6 Finally, Wilson and Damania (2005) present a theoretical model that takes into account political competition for the electorate’s votes, corruption, and environmental policy. The authors show that political competition (a fundamental component of democracy) increases the costs of setting up suboptimal environmental policies (i.e. the incumbent party would lose votes). It follows that political competition enhances the stringency of environmental policy. But the authors also notice that one has to differentiate between grand and petty corruption, where the former is linked to policy making, and the latter to policy implementation. Since bureaucrats’ sensitivity to petty corruption is independent of political competition, corruption may still negatively affect the implementation of environmental rules notwithstanding high political competition. In short, both theoretical and empirical research support the conclusion that institutional settings affect the way policy makers respond to environmental concerns, and that democracy and corruption are two important variables in the process. However, both strands of literature that single out democracy and corruption potentially suffer from the problem of biased coefficients because of omitted variables.7 In our sample, the correlation between the corruption and democracy variable is −0.68, statistically significant at 1%; thus a high level of democracy corresponds to a low level of corruption. In the case that one of the two variables is the actual determinant of environmental policy stringency, a statistical analysis including only the other variable can easily produce a significant coefficient, because of omitted variable bias. It is worthwhile noting that there is little evidence of a causal relationship between democracy and corruption, if considered as contemporary variables. That is, contemporary levels of democracy seem not to affect corruption levels and viceversa. Only a long history (more than 40 years) of democracy is found to lead to a decrease in corruption levels (Treisman 2000), or 30 years from our results in Chap.€3.8 Therefore, even though correlated, these variables can be considered 5╇ The ICRG index, which has been used extensively in the empirical literature (beginning with Knack and Keefer (1995) who used it in an analysis of institutions on economic performance), measures political risk associated with corruption, rather than corruption itself. As such, the proxy for corruption chosen by the authors seems to be very much causally related to the other independent variable they use for political instability (see Lambsdorff 1999). 6╇ The authors do not produce a robustness analysis of the finding and their measure of democracy is from the Freedom House. This measure of democracy has been criticized because of opacity and ideological bias (see also footnote 106). 7╇ More technically, if the real relation among three variables is given by y = β1 + β2 X2 + β3 X3 + u€, the two independent variables are correlated (so that X3 = γ1 + γ2 X2 + ε ) and we omit variable X3; then in a regression we will find y = β1 + β3 γ1 + (β2 + β3 γ2 )X2 + ε + u . Therefore, the estimation bias of the coefficient of the included variable X2 will be equal to β3 γ2 . 8╇ Paldam (2002) also reaches conclusions similar to Treisman. In Chap.€3, we have checked their results on the recently available and much larger World Bank database (Kaufmann et€al. 2004) of
82
5 Corruption, Democracy, and Environmental Policy
independent in the analysis for the short and medium term. Also, while democracy and corruption are correlated, there are a number of countries that show high levels of corruption but are democracies and there are autocracies with low corruption levels. Italy acts as an example of the former: it has a history of more than 50 years of democracy but unusually high levels of corruption for its development stage. An example of the latter is Singapore—it is not a liberal democracy, but has very little corruption.9 Finally, it must be emphasized that, in conformity with the empirical literature, we use a narrow definition of democracy. That is, we define democracy on procedural grounds: democracies are characterized by factors such as competitive and free elections with a high level of participation and constraints on the use of power. A broader definition of democracy would be based on an estimate of the congruence of the political establishment’s action and people’s will.10 However, such a definition of democracy cannot be independent of a measure of corruption, as it seems incompatible for policy makers to exercise power in their own personal interest and in the interest of the people at the same time.11
5.3 The Data In this chapter, we use data on corruption perception, gathered by Transparency International (see description and sources in Sect.€3.3). Transparency International has data for the earlier period 1980–1985, but these data are only for 41 countries. We have augmented the sample by including 13 countries on which corruption data started to be available in 1997 and 1998.12 As expected, the corruption perception indexes for 1980–1985, 1997, and the 1998 are highly correlated (about 85%). Furcorruption perception and we find that only three decades of uninterrupted democracy correlate to lower contemporary levels of corruption. Sandholtz and Koetzle (2000), Sung (2004) and Chowdhury (2004), on the contrary, find that democracy tends to correlate to lower corruption, but they use a very limited set of explanatory variables in their econometric models. Overall, we find that there is a lack of empirical evidence for an effect of contemporary democracy on contemporary corruption levels. The possibility of causality in the longer run should still be kept in mind as it could provide to democracy a way to affect environmental policies indirectly. 9╇ The democracy score of Singapore in 2001 (from Polity IV) was 4 and the corruption perception index (from Transparency International) was 0.8 (both indexes are on a 0–0 scale with higher values indicating higher democracy and higher corruption, respectively). In the same year Italy had a democracy score of 10 and a corruption score of 4.8. 10╇ The word democracy derives from the two ancient Greek words: Demos (↜ής), which means “People”, and Kratos (Κάτς), which means “Rule”. 11╇ Also, it is important to recognize that substantive democracy is a primary goal per se, independently from its effects on other welfare variables. When we consider that democracy is a fundamental component of freedom, the point was convincingly made by Sen (1999). 12╇ Specifically, we used data of the corruption perception index 1997 for Uruguay, the index 1998 for Ghana, Iceland, Jamaica, Morocco, Malawi, Paraguay, Senegal, Tunisia, Tanzania, Zambia, and Zimbabwe and the 1980–1985 for the rest of the sample.
5.3 The Data
83
thermore, when we regressed the available data from 1980 to 1985 on the data from the 1997 and the 1998 samples, the constant and the coefficient are within the 95% confidence interval of 0 and 1, respectively. This finding confirms that the values of the corruption perception index from more recent surveys can be used, as an addition to the older set of indexes, without further transformation. We checked the robustness of our statistical results with respect to the sample, and found only a slight difference when the sample is restricted to the older data. For the regressions where we use the environmental regulatory regime index as a dependent variable, which refers to the year 2001, the corruption perception index refers to the year 1998.13 Our proxy for democracy levels is from the dataset Polity IV, produced by The Integrated Network for Societal Conflict Research (INSCR) of the University of Maryland (Jaggers and Gurr 1995, see the appendix of Chap.€3 for a description). In our main analysis, we averaged the annual values over the period 1980–1985 as we do with the corruption variable. When we used the environmental regulatory regime index as a dependent variable, we averaged the democracy variable over the period 1990–1995. Some authors, while running cross-country regressions, prefer the use of dummy variables for democracy that indicate low, medium, or high democracy (e.g. Hauge and Ellingsen 1998; and Neumayer 2002). The use of dummies, however, reduces the variability of the democracy index and, therefore, reduces the statistical significance of its coefficient. As a test for robustness, we also use another index of democracy. Thus, we will check if our results depend on the democracy measure we have chosen. While the Polity IV dataset accounts for institutionalized democracy and is compiled by experts’ evaluations, the index of democracy developed by Vanhanen (2000) scores countries according to elections results and participation. The index of democracy is made by the multiplication of two factors: one accounts for the level of political competition at elections and the other to the level of turnout. Thus, democracies are countries where competition among political parties is high and a large share of the electorate is active. We note that this index is, methodologically and conceptually, very different from the one used in the main analysis, and thus we consider it proper for a robustness analysis.14 The index has been averaged over two periods of time to be consistent with the other democracy index. We opt for measures of independent variables that refer to previous years, compared to the year of measurement for the dependent variable, because we assume that environmental policy and its implementation requires some time to adapt to the current socio-economic and political situation. In other words, if present conditions are favourable to more stringent environmental policies, we expect that it takes some years for these policies to come about and be implemented. 14╇ Other measures of democracy are available and have been widely used in the literature (e.g. the ones from the Freedom House). Here, for our robustness checks, we prefer to use Vanhanen’s index of democracy because of its methodological and conceptual differences with the Polity IV index. The main methodological difference is that the index of democracy, once its definition is taken for granted, is independent from experts opinions and as such it is a good test of possible biases in experts’ judgments. Conceptually, it is based on a minimalist definition of democracy (referring to Dahl 1971), while the Polity IV index is one of the most comprehensive indexes. Last, the Polity IV index has been criticized for underestimating the relevance of participation (e.g. Munck and Verkuilen 2002, p.€11), whereas this is a fundamental element in Vanhanen’s index of democracy 13╇
84
5 Corruption, Democracy, and Environmental Policy
We use data on environmental policy stringency based on the reports that were self-compiled by individual countries prior to the UN Earth Summit that took place in Rio in 1992. Dasgupta et€al. (1995) first developed an index of environmental policy stringency based on the questionnaires collected by the UN Environmental Program. Their country sample included 31 countries randomly chosen among the ones that participated in the conference. Eliste and Fredriksson (2002), using the same methodology, compiled the index for another 31 countries (also randomly selected). Together, these two sets make a sample of 62 countries with which it is possible to perform cross-country analysis. The index ranges from 1 to 250, with a lower value implying a less stringent policy. These data reflect several aspects of agricultural environmental policy: from policy formulation, to its implementation, to general awareness by the public of environmental issues. The base year for this index is 1990. The reports were completed by the governments, business representatives, and non governmental organizations (NGOs) of the countries concerned. The presence of NGOs in the process suggests a higher level of objectivity in the surveys and avoids the complacency typical of governmental self-reporting. An index of industry stringency of environmental regulations is also available, but only for 31 countries. The index for agriculture and the index for industry have a correlation of 0.96. As the indexes are so highly correlated, we can consider the agriculture index (which is available for a larger sample) as an indicator of overall environmental protection. Another index that quantifies the stringency of environmental policies is the Environmental Regulatory Regime Index compiled by Esty and Porter (2002), which is based on the Environmental Sustainability Index15 and the Global Competitiveness Report 2001–2002 survey of business and government leaders. The index includes the stringency of environmental pollution standards, sophistication of regulatory structure, quality of the environmental information available, extent of subsidization of natural resources, strictness of enforcement, and quality of environmental institutions. Esty and Porter (2002) have shown that the index they compiled is a statistically significant predictor of pollution levels (the authors used it as an explanatory variable for urban particulate concentration, urban SO2 concentration, and energy usage).16 We make use of this index as an alternative to the Environmental Protection Stringency Index in order to check the robustness of our results. It must be noted, however, that countries for which the Environmental Regulatory Regime Index is available tend to be more democratic than the world average.17 The participation. Furthermore, Freedom House’s indexes have been challenged for their opacity and ideological biases (Deacon 2003; Rosenblum and Salehyan 2004; and Landman 2004), thus, for our robustness checks, we prefer an alternative index that is not based on expert opinions. 15╇ The Environmental Sustainability Index is a joint project of the World Economic Forum, The Yale Center for Environmental Law and Policy, and the Columbia University Center for International Earth Science Information Network. See www.ciesin.columbia.edu/indicators/ESI/. 16╇ For a critical assessment of the ESI see Morse and Fraser (2005). 17╇ The mean of the democracy variable (when scaled 0–10) for the years 1980–1985 is 4.2, 5.9, and 6.4 for the complete sample, for the countries for which the Environmental Policy Stringency Index is available and for the countries for which the Environmental Regulatory Regime Index is
5.4 Empirical Results
85
presence of fewer autocracies in the sample could imply a sample bias. This bias could be explained by the fact that it is more difficult to collect the data needed to construct the index in dictatorships and, as a result, we would expect the democracy variable to have less explanatory power in this sample. Therefore, the results on the democracy variables, when the Environmental Regulatory Regime Index is used as a dependent variable, should be interpreted with caution. For data sources and for a description of the other data used in this chapter see Appendix€2.
5.4 Empirical Results As an estimation strategy, ordinary least squares (OLS) regressions on a cross-section of countries are produced. We begin with the most straightforward formulation: the dependent variable is a linear combination of the most fundamental independent variables. Subsequently, we add further explanatory variables and investigate in more detail the results that differ from previous literature. Finally, different dependent and independent variables are used to check the robustness of our results. We begin by estimating the following equation:
i EPSi = α0 + α1 ln(Y1980 ) + α2 C i + α3 D i + α4 Z i + ε i ,
(5.1)
where the superscript i denotes each country in the sample, EPS is the Environmental Protection Stringency Index, Y1980 is income per capita in 1980,18C is the Corruption Perception Index for the period 1980–1985, D is the index of democracy for the period 1980–1985. Finally, Z is a vector of control variables that are used to check the robustness of our findings. Before discussing the results, we note that institutional indexes are difficult to estimate with a high degree of precision and that their coefficients will presumably suffer a downward bias due to measurement error. The results of the regressions from Eq.€5.1 are presented in Table€5.1. The estimated coefficients are reported in standardized form for ease of interpretation (i.e. they can be interpreted in standard deviation terms). In other words, a coefficient of one implies that one standard deviation of the independent variable is associated with a one standard deviation of the dependent variable. When we use dependent variables that have different scale (i.e. the Environmental Protection Stringency Index and the Environmental Regulatory Regime Index) the standardized form makes available, respectively. The standard deviations of the same democracy variable (in the same order) are: 3.8, 3.9, and 3.6. Therefore, the sample for which the Environmental Regulatory Regime Index has been compiled appears to have a higher mean value of the democracy variable (and slightly lower variance). 18╇ We use the natural logarithm of income, in accordance with most of the empirical literature, in order to smooth the distribution of the income variable. The other variables are indexes; therefore this transformation is not in order.
86
5 Corruption, Democracy, and Environmental Policy
Table€5.1↜渀 Results from statistical analysis as in Eq.€5.1 (2) (3) Dependent variable: EPS (1) 0.56*** 0.37*** 0.29*** lnY1980 (5.23) (3.67) (2.75) Democracy 0.34*** 0.19* (3.19) (2.01) Corruption −0.56*** −0.49*** (5.50) (4.66) Urbanization Schooling Dummy for Latin America Dummy for OECD
(4) 0.64*** (4.32) 0.12 (1.28) −0.53*** (5.39) −0.38*** (3.13)
(5) 0.42** (2.62) 0.13 (1.44) −0.44*** (4.59) −0.38*** (3.46) 0.34*** (3.17)
Dummy for South Asia
(6) 0.29* (1.91) 0.05 (0.58) −0.34*** (4.02) −0.23** (2.16) 0.31*** (3.44) −0.10 (1.40) 0.25** (2.22) 0.11* (1.94)
0.69 0.76 0.78 0.81 0.84 0.89 Adjusted R2 Number of countries 54 54 54 54 51 51 OLS estimation with the Environmental Policy Stringency Index as dependent variable. Coefficients are standardized. t-statistics are in parentheses. The Ho of homoscedasticity, in the White test, was not rejected (i.e. pâ•›>â•›0.10) in any regression. To check for multicollinearity, we analysed the variance inflation factor (VIF). The VIF is never higher than the usual 10 threshold for the Corruption and Democracy variables *10% significance; **5% significance; ***1% significance
the coefficients more easily comparable.19 The first two models take into consideration the two institutional variables separately and produce the standard findings. Both variables appear to have an influence on the Environmental Policy Stringency Index: democracy a positive influence and corruption a negative one. In model (1), the democracy index has a positive coefficient of 0.34, significant at the 1% level. A one-standard deviation increase in the democracy variable induces an increase in the Environmental Protection Index by approximately one-third of its standard deviation. In model (2), the corruption index has a coefficient equal to −0.56 and is statistically significant at a 1% level of confidence. We find that an increase of one standard deviation in the corruption index reduces the environmental protection index by more than half its standard deviation. Though both variables are individually significant and contribute substantially to environmental policy stringency, when compared to each other corruption appears to be the most powerful explanatory variable. Also, the adjusted R2 increases from 0.69 to 0.76 when going from model (1) to model (2). In model (3), democracy and corruption are included jointly, and we see that the value of both the coefficients decreases and the statistical significance of the coef19╇
The standardization procedure does not affect the significance of the coefficients.
5.4 Empirical Results
87
ficient on democracy is reduced but still significant at the 10% level (just above the 5% threshold).20 This result again confirms the corruption index as the variable with higher explanatory power. As a note of caution, we mention again that the correlation between the corruption and the democracy variable is −0.68. The inclusion of highly correlated variables gives rise to multicollinearity concerns. The correlated variables will have unbiased coefficient estimates, but the standard errors will be inflated.21 Subsequently, we carried out a series of robustness checks. First, control variables have been added to the list of regressors in order to check for the robustness of the coefficients of the independent variables. The first control variable is the percentage of the population living in urban areas. Urbanization rates have been found to be associated with corruption (Hill 2003) and could also reflect different attitudes towards the environment.22 In model (4), we find that the inclusion of the urbanization variable further decreases the size and statistical significance of the democracy coefficient (now it is equal to 0.12). At the same time, the coefficient of corruption still retains economic and statistical significance. Moreover, urbanization is a significant determinant of environmental policy stringency and has a negative effect. That is, for given income, democracy, and corruption levels, the more people live in urban areas the less stringent environmental policy tends to be. This result is not very strong, as it is sensitive to the inclusion of additional control variables in the following models. The second control variable we added is schooling (model (5)). We find the coefficient on the democracy variable to remain insignificant. There is a decrease in the absolute value of the corruption coefficient to −0.44, though it remains significant at 1% confidence level. The schooling variable itself is highly significant. A onestandard deviation increase in schooling increases environmental policy stringency by approximately a third of one standard deviation. This is what we expected on theoretical grounds as improved education leads to an increased awareness of environmental problems (e.g. of health problems related to pollution). Last, note that the statistical significance of the coefficient on the income variable drops to 5%. There is an obvious multicollinearity and causality problem between schooling and income and it appears difficult to single out the effects of education from the effects
20╇ As a method to check for the possibility of clustering independent variables (maintaining their explanatory power) and discovering possible composed variables, we have tried to use the principal component method adding other institutional quality indexes. The results did not allow for any meaningful grouping of these variables and confirmed our impression on the independent role played by the corruption variable. 21╇ In order to check for multicollinearity problems, we used a Variance Inflation Factor (VIF). We did not find the VIF to be higher than the usual threshold of 10 for the variables of democracy and corruption in any of our regressions, except for the regressions where interaction factors were introduced. 22╇ Urbanization rates are typically included in EKC estimations. In our analysis, the coefficient on the urbanization variable can be interpreted as reflecting differences in preferences and political influence of urban and non-urban citizens.
88
5 Corruption, Democracy, and Environmental Policy
of income.23 Another control variable added to this model but not reported in the table is the share of the population employed in agriculture. We found this variable to have almost no effect on the other coefficients and the coefficient for agricultural employment itself was insignificant both statistically and in magnitude (this result holds even when the urbanization variable is omitted). The final set of control variables added were regional dummies (model (6)). These regional dummies have been found significant in many recent empirical analyses. Adding them to our analysis makes sure that our results are not driven by geographical factors or by a particular group of countries (e.g. see Rodriguez et€al. 2001). In model (6) the magnitude of the democracy variable is further decreased and remains non-significant. At the same time, the corruption coefficient is robust to the inclusion of regional dummies. The regional dummies for OECD and South Asian countries are significant (at 5% and 10% level of significance, respectively), which would indicate that there are some characteristics of OECD and South Asian countries that are affecting their level of environmental protection, and that are omitted from our model.24 Furthermore, in the literature it is sometimes suggested that the effect of corruption on environmental policy is non-monotonic (Fredriksson and Millimet 2001). In order to check for non-linearities, we have carried out the simple regression from Eq.€5.1, omitting the democracy index and the control variables, but including the corruption variable at three powers. The results (not reported in the table) provide no evidence for a non-monotonic relation. The coefficients for the first and the second power are both insignificant, and the one at the third power is significant just at 10%. Another robustness check we performed was an analysis of the outliers. The Environmental Policy Stringency Index of Turkey and Iceland lies outside the two standard deviations of the residual bound from their predicted values. To be sure that our results are not driven by these two cases we repeated the analysis excluding them from the sample. We found that the exclusion of these two countries would slightly strengthen our conclusions. Repeating regression (6) without outliers, we find that the coefficient of democracy remains insignificant. At the same time, the coefficient of the corruption index remains significant at 1% level and equals −0.36. As a third robustness check, we tested our results with respect to the specification of the democracy variable (results reported in Table€5.2). We use the index of democracy developed by Vanhanen (2000) as an alternative measure of democracy. As noted in Sect.€5.3, the methodology used for building these two indexes is very different and we consider such a different dependent variable as a strong test of our results. The results confirm the findings presented above. When only the democracy index is used as an independent variable (model (7)) it appears to be a fundamental determinant of environmental policy. After the inclusion of the corruption index 23╇ As some authors have argued that corruption levels affect public investment in education (e.g. Mauro 1998 and Pellegrini and Gerlagh 2004) the inclusion of the schooling variable can be considered an extreme test for the significance of the corruption coefficient, see Chap.€4. 24╇ The Latin America dummy is significant when it is the only regional dummy included, but not together with the other dummies. The omission of the Latin America dummy would not alter our results.
5.4 Empirical Results Table€5.2↜渀 Results from statistical analysis as in Eq.€5.1 (7) (8) Dependent variable: EPS 0.41*** 0.37*** lnY1980 (3.65) (3.67) Index of democracy 0.52*** (5.49) Corruption −0.56*** (5.50) Urbanization Schooling Dummy for Latin America Dummy for OECD Dummy for South Asia
89
(9) 0.21* (2.01) 0.35*** (3.52) −0.43*** (4.28)
(10) 0.28* (1.99) 0.15 (1.46) −0.33*** (4.19) −0.26** (2.65) 0.32*** (3.57) −0.10 (1.48) 0.18 (1.52) 0.10* (1.82)
0.74 0.76 0.81 0.89 Adjusted R2 Number of countries 54 54 54 51 OLS estimation, with the Environmental Policy Stringency Index as a dependent variable. Coefficients are standardized. t-statistics are in parentheses. The Ho of homoscedasticity, in the White test, was not rejected (i.e. pâ•›>â•›0.10) in any regression, apart from (7). Therefore the t-statistics in (7) are calculated with robust standard errors *10% significance; **5% significance; ***1% significance
(model (9)), the importance of democracy is markedly decreased. The effects of democracy disappear both statistically and in magnitude once other variables are included (model (10)). At the same time, size and statistical significance of the coefficient on the corruption index are only slightly affected by the change in the measure of democracy. Finally, the results presented above were confirmed by a fourth robustness check: the use of the Environmental Regulatory Regime Index replacing the Environmental Policy Stringency Index as the dependent variable (regression equation and models are presented in the appendixes). We find that corruption and income have high explanatory power when considered in isolation (not presented). However, the inclusion of income as a control variable makes the coefficient for democracy nonsignificant, both in regressions with and without corruption, and sometimes the coefficient has the wrong sign and is marginally significant (Table€5.4 in Appendix€1). In summary, our results suggest that institutional settings are important determinants of environmental policy stringency. We find robust evidence for a substantial effect of corruption on the stringency of environmental policy, while there is little evidence of democracy being an important determinant. As an explanation of this difference with previous econometric analyses, we suggest that earlier empirical works may have overemphasized the role of procedural democracy for environmental policies because of the omission of a corruption index as a control variable. In other words, our estimates suggest that increasing democratic standards has to
90
5 Corruption, Democracy, and Environmental Policy
be matched with low levels of corruption in order to induce stricter environmental policy and that democratic procedures per se are insufficient. We note a limitation of our analysis: as mentioned in Sect.€5.1 there are a number of variables related to environmental conditions that can influence environmental quality. We expect these variables to also influence the stringency of environmental policies, but it is extremely difficult to control for them econometrically. Among these variables are those that involve geographical/physical characteristics such as climate conditions (e.g. strength of dominant winds, rainfall), a country’s geographic position (landlocked, or island), and soil characteristics (e.g. prone to erosion, or desertification). Other characteristics will be related to economic conditions, such as dominant sectors (e.g. whether the country specialises in the production of energy intensive goods), and the timing of the development of the country. In any case the high R2 of our regressions indicate that these missing variables are not the main drivers of environmental policies. Finally, we note that some of the determinants of environmental policy stringency could be themselves influenced by corruption. In particular, the schooling variable was found in Chap.€3 to be one of the transmission channels of corruption on economic growth. The same type of analysis of direct and indirect effects of corruption could have been applied in this chapter and the importance of corruption as a determinant of environmental policy would have increased in our results.
5.4.1 T he Case for Interaction Between Democracy and Income Variables The results presented above do not suggest that democracy plays a significant role in enhancing environmental policy stringency. It is possible, however, that we need to take into account the interaction between democracy and income in order to correctly estimate the influence of democracy on environmental policy. It is possible that higher income raises demand for stricter environmental protection, and it is only then that a more responsive (democratic) government will affect environmental policy. If environmental policy depends on the interaction between income levels and institutional variables, introducing interaction terms should reveal this by producing sizable and statistically significant coefficients. Indeed, in our analysis, we find some weak evidence for an interaction effect between income and democracy. That is, environmental policies tend to be stricter at high-income levels when the country is a well-established democracy, or stated otherwise, the effect of democracy is more positive for higher-income levels. The results of the analysis are presented in Table€5.3. Caution is needed though, because in contrast with the main results presented above, the significance of the interaction variable does not pass the robustness tests. In model (11) the interaction term is significant at 5%. The use of the index of democracy as an alternative to the democracy variable from the Polity IV dataset, in model (12), produces statistically non-significant coefficients. Also the use of the Environmental Regulatory Index,
91
5.4 Empirical Results Table€5.3↜渀 Results from statistical analysis as in Eq. 5.1 Independent variables (11) (12) 0.14 0.28* lnY1980 (0.91) (2.00) Democracy −1.45** (2.21) Index of democracy −1.45 (1.49) Corruption −0.26*** −0.31*** (3.00) (3.88) Urbanization −0.21** −0.29*** (2.05) (2.96) Schooling 0.27*** 0.30*** (3.00) (3.38) Dummy for Latin America −0.11 −0.07 (1.55) (1.07) Dummy for OECD 0.12 0.16 (1.02) (1.39) Dummy for South Asia 0.11** 0.11** (2.12) (2.13) 1.78** Democracy × lnY (2.31) 1.67 Index of Democracy × lnY (1.65)
(13) 0.07 (0.22) −0.79 (1.00) −0.63*** (6.15) 0.10 (1.10) −0.09 (0.89) −0.09 (0.93) −0.03 (0.27) 0.06 (1.01) 0.93 (0.93)
(14) 0.10 (0.59) −1.86* (1.96) −0.61*** (6.01) 0.10 (1.19) −0.06 (0.58) −0.08 (0.97) −0.06 (0.58) 0.07 (1.19) 2.03* (1.94)
Adjusted R2 0.89 0.89 0.83 0.84 Number of countries 51 51 57 57 OLS estimation, models (11)–(12) have the Environmental Policy Stringency Index as a dependent variable, models (13)–(14) have the Environmental Regulatory Regime Index as a dependent variable. Coefficients are standardized. t-statistics are in parenthesis. The Ho of homoscedasticity, in the White test, was not rejected (i.e. pâ•›>â•›0.10) in any regression *10% significance; **5% significance; ***1% significance
as a dependent variable, weakly confirms the existence of an interaction between democracy and income levels, but not when the democracy variable is from the Polity IV dataset. In model (14), where the Environmental Regulatory Regime Index is used as a dependent variable and, the interaction term is significant at 10%. Furthermore, as variables have been standardized to a zero mean, the negative coefficient for democracy suggests that an increase in democracy negatively affects the environmental policy stringency for a country with average income level. It is only when income is above one standard deviation of the average income that democracy starts to have a positive effect.25 Thus, we find the interaction term between democracy and income to be statistically significant at 5% in just one of four specifications of our econometric model. As such we do not find the results to be particularly compelling and we report We also included an interaction factor of corruption and income, but found it not to be significant in any of the regressions.
25╇
92
5 Corruption, Democracy, and Environmental Policy
the tentative evidence as it stands as a possible qualification of our main results on the limited influence of democracy as a determinant of environmental policy stringency. One interpretation of our analysis of the interaction between democracy and income is that we are simply identifying one special group of countries. The income–democracy interaction variable considers the effect of income on environmental policies of the countries that have high-income levels and high democracy, but there can be specific groups of countries that do not have high-income levels and high democracy together. In particular, a large group of countries characterized by high-income levels and low scores on the democracy variable could be oil-rich countries (Ross 2001), but our dataset includes only a maximum of two OPEC country. Moreover, the OPEC country that is included in all the regressions is Venezuela, which has much higher degrees of democracy than the average OPEC country. An analysis on a more complete sample including a larger number of oilrich countries might allow us to disentangle the effect of that group of countries, but given our data limitations the question is left open.
5.5 Implications for the Environmental Kuznets Curve In this section, we will interpret our results in the context of the literature on economic development and the environment, and emphasize how our findings relate to the EKC and the leapfrogging hypothesis. Several authors have argued that, in the course of their development, currently developed countries first experienced a decrease in environmental quality due to increased production. Then, as their income further increased, these countries experienced an improvement in several indicators of ambient environmental quality (e.g. Grossman and Krueger 1995). In the later stage of development, cleaner production techniques and the decrease in the share of polluting production in total output allowed a decrease in pollution while there was an increase in the scale of the economy. Based on this evidence, it is hoped that currently developing countries will follow a similar path and attain increasing environmental quality together with increasing income levels. Even more optimistically, some authors argue that developing countries could “leapfrog” the pollution intensive part of their development path, i.e. they will tunnel through the EKC. According to this strand of literature, developing countries may benefit from experiences in developed countries, both in terms of clean technologies, but also in terms of effective and efficient environmental policies. Developing countries may skip the stage of “command and control” policies and go directly to more sophisticated and more efficient regulations (Perkins 2003). Our estimates of the effects of income on environmental policy provide some support for the positive effect of higher income on environmental protection. However, as a qualification of this optimistic perspective, let us first note that, even assuming that the EKC pattern can be reproduced in developing countries in a similar manner as the already developed countries, current developing countries will
5.5 Implications for the Environmental Kuznets Curve
93
still face declining environmental quality for several decades. Cole and Neumayer (2005) estimate that, on the basis of optimistic forecasts on both economic growth and the effects of income levels on the environment, ambient concentrations for a number of pollutants will continue to increase for the next century in many developing countries.26 Adding to the problem is the global shift in comparative advantages. Developed countries make use of cleaner production technologies but also move their production structure to sectors associated with lower pollution levels; thus developing countries will find their comparative advantage shifting to the higher polluting production sectors. Furthermore, Bruyn and Opschoor (1997) have observed how the inverted-U identified in the literature might not be persistent over time and a more correct characterization might be an N-shaped pattern in which, in addition to the common EKC path, a new increase in environmental degradation occurs at higher-income levels. Also, Stern (2004) highlights how recent developments in environmental policies of developing countries do not fit with the conventional EKC curve and casts doubt on its statistical foundations. We find that one important objection to using past EKCs to predict the future of developing countries is the effect of institutions on environmental policies, as studied in this chapter. In this context, our results portray a more pessimistic perspective on the future environmental quality in developing countries. One of the explanations for the EKC is an institutional policy response in reaction to increasing scarcity of the environmental goods and shifting preferences of the people (Grossman and Krueger 1995). Since lower quality institutions typically plague developing countries,27 we expect less of a policy response to changing preferences in these countries. Therefore, even when reaching similar income levels as the currently developed countries, developing countries can be expected to have more lax environmental policies. From this point of view, there is reason to believe that for developing countries the EKC’s apex shifts to higher-income levels, and will not resemble the developed countries’ EKCs. Consequently, tunnelling through the EKC seems improbable. The same empirical results also carry a more optimistic message. Improving a country’s institutional quality may render a double dividend in that it will be beneficial for environmental quality as well as for economic growth. Barro (1996) argues that at low levels of democracy an increase in democracy will foster economic development. At the same time, evidence strongly suggests that corruption has negative effects on economic development (Mauro 1995; Mo 2001). When a decrease in the level of corruption leads to higher growth rates, environmental policy will improve through both the direct channel (analysed in this study) and the indirect It must be noted that the turning points of the EKC used for the analysis of Cole and Neumayer (2005) are deemed to be overly optimistic in comparison to other studies (e.g. Harbaugh et€al. 2002; Stern 2004), which made use of more complete dataset and improved estimation techniques if compared to the benchmark works of Grossman and Krueger (1995) and to the more recent work of Cole (2003). 27╇ In our sample the average value of the corruption index was 7.14 for countries with income below US€$2000 in 1980, and 3.79 for the countries with income above US€$20,000. 26╇
94
5 Corruption, Democracy, and Environmental Policy
income channel. Our results therefore support the argument for improving the institutional infrastructures of developing countries (e.g. Meier 2001). Also, we recognize that countries can learn from each other when developing their environmental policies. Clean production techniques adopted in developed countries will diffuse and become available for developing countries. Also, it is not impossible that in the future increasing pressure will be put on governments to sign international environmental agreements and to act accordingly.28 The European Union provides a remarkable example of the impact that the international arena has on domestic environmental policies (see next chapter). It sets ambient quality standards for member countries and (more or less) forces its new members to take a shortcut through the EKC (see Chap.€6).
5.6 Conclusions Our analysis, in line with previous literature, shows statistically significant and sizeable coefficients when democracy and corruption are considered individually as explanatory variables for environmental policy stringency. The effect on the stringency of environmental policies is negative for corruption and positive for democracy. Contributing to the literature, we have shown that the inclusion of corruption and democracy together diminishes the significance and importance of the democracy variable. The further inclusion of additional control variables renders the democracy variable statistically insignificant and the magnitude of the coefficient is markedly further decreased. We deduct from our results that a large part of the positive effect of democracy indexes on environmental protection, as it is found in most of the previous studies, is due to the correlation between high levels of democracy and lower levels of corruption. Caution is warranted, however, since democracy and corruption are highly correlated; there may be a problem of multicollinearity, which may have decreased the statistical significance of the coefficient on democracy. We must also recognize the possible existence of a causal link between a history of democracy and corruption that could give further scope for democracy to influence environmental policy indirectly in the long run (see Chap.€3). Thus, if stable democracy is conducive to lower corruption, it could still have an effect in the medium-long run through the corruption variable. We conclude that we find no evidence of a direct sizeable and significant positive effect of democracy on environmental policy. We remind the reader that the use of indexes for democracy, corruption, and environmental policies (that is, all the Data from the Environmental Treaties and Resource Indicators (available at http://sedac.ciesin. columbia.edu/entri) confirm an increasing trend in the number of environmental treaties over time: 70 in the decade 1960–1970, 94 in the decade 1970–1980, 89 in the decade 1980–1990, and 110 for 1990–2000.
28╇
Appendix 1
95
fundamental variables in our analysis) implies that the results should be interpreted carefully. This is especially true for our measures of democracy; they are proxies for procedural democracy and should be read as such. This feature of the democracy index is not unique to our analysis, however, as the same indexes have been used in previous empirical analyses. Our results suggest that these previous findings are not entirely robust.
Appendix 1 In Table€5.4, as a robustness check, we reproduce the main results using alternative indexes for the stringency of the environmental policy and for the level of democracy. With respect to the former, we use the Environmental Regulatory Regime Index as an alternative to the Environmental Policy Stringency Index. With respect to the latter, we use the index of democracy instead of the democracy variable. Models (15)–(19) are as in the following equation:
i ERRI i = β0 + β1 ln(Y1990 ) + β2 C i + β3 Di + εi
(5.2)
where, ERRI is the environmental regulatory regime index, Y1990 is income in 1990, C is the corruption perception index in 1998, and D is the democracy index averaged for 1990–1995. Using these alternative variables, we confirm that corruption is an important determinant of environmental policies: there are consistently sizeable and statistically significant negative coefficients on the corruption variable in the various specifications. At the same time, we find limited support for the effect of democracy on environmental policies.
Table€5.4↜渀 Results from statistical analysis as in Eq.€5.2 (15) (16) Dependent variable: ERRI 0.93*** 0.26*** lnY1990 (9.65) (2.97) Democracy −0.19* (1.98) Index of democracy Corruption
−0.69*** (7.98)
(17) 0.35*** (3.47) −0.12* (1.72) −0.67*** (7.80)
(18) 0.85*** (6.92)
(19) 0.27** (2.42)
−0.06 (0.45)
−0.02 (0.22) −0.69*** (7.89)
Adjusted R2 0.67 0.83 0.83 0.65 0.83 Number of countries 62 62 62 62 62 OLS estimation, regressions with the Environmental Regulatory Regime Index as a dependent variable. Coefficients are standardized. t-statistics are in parentheses under the coefficients. The Ho of homoscedasticity, in the White test, was not rejected (i.e. pâ•›>â•›0.10) in any regression *10% significance; **5% significance; ***1% significance
96
5 Corruption, Democracy, and Environmental Policy
Appendix 2: Data EPS is the Environmental Policy Stringency index which measures policy formulation, implementation, and public awareness of environmental issues (see Eliste and Fredriksson 2002). ERRI is the Environmental Regulatory Regime Index which includes measures of the stringency of environmental standards, the quality of the regulatory regime, and the size of subsides favouring the consumption of natural resources (see Esty and Porter 2002). Y1980 and Y1990 are GDP per capita in 1980 and 1990, respectively. From the Summer and Heston database, specifically the Penn World Table€6.1 (available at http:// pwt.econ.upenn.edu/). Corruption is the Corruption Perception Index (CPI) from Transparency International, see Sect.€3.3. Data refer to the years 1980–1985 (augmented with data from 1997 and 1998) and to 1998 (available at http://www.icgg.org/). Democracy is the average of the institutional democracy score (for the years 1980–1985 and 1990–1995) from the Polity IV dataset (the “polity” variable in the original dataset). The original indicator has been rescaled to a 0–1 scale (see http:// www.cidcm.umd.edu/inscr/polity/). Index of democracy scores countries according to elections results and participation (Vanhanen 2000). Higher scores on democracy are given to countries where the election results suggest that political competition is high (i.e. where the largest party did not win the elections by an extremely large margin) and where a large share of the electorate went to the polls (see http://www.prio.no/). Urbanization is the percentage of the total population that lives in urban areas in 1999 (World Development Indicators 2004). Schooling is the number of years spent at school, on average, for the population above 25 years old in 1985 and 1995 (Barro–Lee “International data on educational attainment” dataset (version updated to April 2000) and are available at www.cid. harvard.edu/ciddata/ciddata.html).
References Aguilera-Klink, F., & Sánchez-García, J. (2005). Environmental degradation as a result of democratic disruption: The case of the Canary Islands. Lisbon: European Society for Ecological Economics Conference. Barro, R. (1996). Democracy and growth. Journal of Economic Growth, 1(1), 1–27. Callister, D. J. (1999). Corrupt and illegal activities in the forestry sector: Current understandings, and implications for World Bank Forest Policy. Forest Policy Implementation Review and Strategy Development: Analytical Studies. Washington: World Bank. Carter, T. S. (1997). The failure of environmental regulation in New York: The role of co-optation, corruption and a co-operative enforcement approach. Crime Law and Social Change, 26, 27–52. Chowdhury, S. K. (2004). The effect of democracy and press freedom on corruption: An empirical test. Economics Letters, 85(1), 93–101.
References
97
Cole, M. A. (2003). Development, trade, and the environment: How robust is the Environmental Kuznets Curve? Environment and Development Economics, 8, 557–580. Cole, M. A., & Neumayer, E. (2005). Economic growth and the environment in developing countries: What are the implications of the Environmental Kuznets Curve? In P. Dauvergne (Ed.), International handbook of environmental politics (pp.€298–318). Cheltenham: Edward Elgar. Congleton, R. D. (1992). Political-institutions and pollution-control. Review of Economics and Statistics, 74(3), 412–421. Dahl, R. A. (1971). Polyarchy: Participation and opposition. New Haven: Yale University Press. Damania, R. (2002). Environmental controls with corrupt bureaucrats. Environment and Development Economics, 7, 407–427. Damania, R., Fredriksson, P. G., & List, J. A. (2003). Trade liberalization, corruption, and environmental policy formation: Theory and evidence. Journal of Environmental Economics and Management, 46(3), 490–512. Dasgupta, S., Mody, A., Roy, S., et€ al. (1995). Environmental regulation and development: A cross-country empirical analysis. Policy Research Working Paper Series. Washington: The World Bank. de Bruyn, S. M., & Opschoor, J. B. (1997). Developments in the throughput-income relationship: Theoretical and empirical observations. Ecological Economics, 20(3), 255–268. Deacon, R. (2003). Dictatorship, democracy, and the provision of public goods. University of California, Santa Barbara CA, USA. Dryzek, J. S. (1987). Rational ecology: Environment and political economy. Oxford: Blackwell. Ehrlich, P. R. (1968). The population bomb. New York: Ballantine Books. Eliste, P., & Fredriksson, P. G. (2002). Environmental regulations, transfers, and trade: Theory and evidence. Journal of Environmental Economics and Management, 43(2), 234–250. Environmental Investigation Agency and Telapak (2001). Timber trafficking: Illegal logging in Indonesia, South East Asia and international consumption of illegally sourced timber. London: EIA Telapak. Environmental Investigation Agency and Telapak (2003). Above the law: Corruption, collusion, nepotism and the fate of Indonesia’s forests. London: EIA and Telepak Indonesia. Esty, D. C., & Porter, M. (2002). Ranking national environmental regulation and performance: A leading indicator of future competitiveness? In M. Porter, J. Sachs, & K. Schwab (Eds.), The Global Competitiveness Report 2001–2002 (pp.€78–100). New York: Oxford University Press. Fredriksson, P. G., &. Millimet, D. L. (2001). Bureaucratic corruption and environmental policy: Theory and evidence from the United States. Manuscript. Fredriksson, P. G., & Svensson, J. (2003). Political instability, corruption and policy formation: The case of environmental policy. Journal of Public Economics, 87(7–8), 1383–1405. Fredriksson, P. G., Neumayer, E., Damania, R., et€al. (2005). Environmentalism, democracy, and pollution control. Journal of Environmental Economics and Management, 49(2), 343–365. Grossman, G. M., & Krueger, A. B. (1995). Economic-growth and the environment. Quarterly Journal of Economics, 110(2), 353–377. Harbaugh, W. T., Levinson, A., & Wilson, D. M. (2002). Reexamining the empirical evidence for an environmental Kuznets curve. Review of Economics and Statistics, 84(3), 541–551. Hardin, G. (1968). Tragedy of the commons. Science, 162(3859), 1243–1248. Hauge, W., & Ellingsen, T. (1998). Beyond environmental scarcity: Causal pathways to conflict. Journal of Peace Research, 35(3), 299–317. Heilbroner, R. L. (1974). An inquiry into the human prospect. New York: Norton. Hill, K. Q. (2003). Democratization and corruption: Systematic evidence from the American states. American Politics Research, 31(6), 613–631. Jaggers, K., & Gurr, T. R. (1995). Transitions to democracy: Tracking democracy’s third wave with the Polity III data. Journal of Peace Research, 32, 469–482. Kaufmann, D., Kraay, A., & Mastruzzi, M. (2004). Governance matters III: Governance indicators for 1996–2002. World Bank Policy research working paper. Washington: World Bank. Knack, S., & Keefer, P. (1995). Institutions and economic performance: Cross-country tests using alternative institutional measures. Economics and Politics, 7(3), 207–227.
98
5 Corruption, Democracy, and Environmental Policy
Lambsdorff, J. G. (1999). Corruption in empirical research: A review. Berlin: Transparency International. Landman, T. (2004). Measuring human rights: Principle, practice, and policy. Human Rights Quarterly, 26(4), 906–931. Lopez, R., & Mitra, S. (2000). Corruption, pollution, and the Kuznets Environment Curve. Journal of Environmental Economics and Management, 40, 137–150. Mauro, P. (1995). Corruption and growth. Quarterly Journal of Economics, 110(3), 681–712. Mauro, P. (1998). Corruption and the composition of government expenditure. Journal of Public Economics, 69(2), 263–279. McCloskey, H. J. (1983). Ecological ethics and politics. Totowa: Rowman and Littlefield. McGuire, M. C., & Olson, M. (1996). The economics of autocracy and majority rule: The invisible hand and the use of force. Journal of Economic Literature, 34(1), 72–96. Meier, G. M. (2001). The old generation of development economists and the new. In G. M. Meier & J. E. Stiglitz (Eds.), Frontiers of development economics: The future in perspective (13–50). Washington: World Bank. Mill, J. S. (1859/1982). On liberty. New York: Viking Press. Mo, P. H. (2001). Corruption and economic growth. Journal of Comparative Economics, 29(1), 66–79. Morse, S., & Fraser, E. D. G. (2005). Making “dirty” nations look clean? The nation state and the problem of selecting and weighting indices as tools for measuring progress towards sustainability. Geoforum, 36(5), 625–640. Munck, G. L., & Verkuilen, J. (2002). Conceptualizing and measuring democracy: Evaluating alternative indices. Comparative Political Studies, 35(1), 5–34. Neumayer, E. (2002). Do democracies exhibit stronger international environmental commitment? A cross-country analysis. Journal of Peace Research, 39(2), 139–164. Paldam, M. (2002). The cross-country pattern of corruption: Economics, culture and the seesaw dynamics. European Journal of Political Economy, 18(2), 215–240. Payne, R. A. (1995). Freedom and the environment. Journal of Democracy, 6(3), 41–55. Pellegrini, L., & Gerlagh, R. (2004). Corruption’s effect on growth and its transmission channels. Kyklos, 57(3), 429–456. Pellegrini, L., & Gerlagh, R. (2006). Corruption, democracy, and environmental policy: An empirical contribution to the debate. The Journal of Environment & Development, 15(3), 332–354. Perkins, R. (2003). Environmental leapfrogging in developing countries: A critical assessment and reconstruction. Natural Resources Forum, 27(3), 177–188. Putnam, R. D. (1993). Making democracy work: Civic traditions in modern Italy. Princeton: Princeton University Press. Robbins, P. (2000). The rotten institution: Corruption in natural resource management. Political Geography, 19(4), 423–443. Rodriguez, F., Rodrik, D., Hseh C. T., et al. (2001). Trade policy and economic growth: A skeptic’s guide to the cross-national evidence. Nber Macroeconomics Annual 2000, 15, 261–324. Rosenblum, M.-R., & Salehyan, I. (2004). Norms and interests in US asylum enforcement. Journal of Peace Research, 41(6), 677–697. Ross, M. L. (2001). Does oil hinder democracy? World Politics, 53(3), 325–361. Sandholtz, W., & Koetzle, W. (2000). Accounting for corruption: Economic structure, democracy, and trade. International Studies Quarterly, 44(1), 31–50. Sen, A. (1999). Development as freedom. Oxford: Oxford University Press. Stern, D. I. (2004). The rise and fall of the environmental Kuznets curve. World Development, 32(8), 1419–1439. Sung, H.-E. (2004). Democracy and political corruption: A cross-national comparison. Crime, Law and Social Change, 41, 179–194. Torras, M., & Boyce, J. K. (1998). Income, inequality, and pollution: A reassessment of the environmental Kuznets Curve. Ecological Economics, 25(2), 147–160. Treisman, D. (2000). The causes of corruption: A cross-national study. Journal of Public Economics, 76(3), 399–457.
References
99
van Beers, C., & van den Bergh, J. C. J. M. (1997). An empirical multi-country analysis of the impact of environmental regulations on foreign trade flows. Kyklos, 50(1), 29–46. van Beers, C., & van den Bergh, J. C. J. M. (2003). Environmental regulation impacts on international trade: Aggregate and sectoral analyses with a bilateral trade flow model. International Journal of Global Environmental Issues, 3(1), 14–29. Vanhanen, T. (2000). A new dataset for measuring democracy, 1810–1998. Journal of Peace Research, 37(2), 251–265. Welsch, H. (2004). Corruption, growth, and the environment: A cross-country analysis. Environment and Development Economics, 9, 663–693. Wilson, J. K., & Damania, R. (2005). Corruption, political competition and environmental policy. Journal of Environmental Economics and Management, 49(3), 516–535.
Chapter 6
Corruption and Environmental Policies: What Are the Implications for the Enlarged EU?
Abstract╇ This chapter discusses the prescription of EU environmental regulations for new member states. It can be argued that these countries should be allowed looser directives as a way to take into consideration their lower-income levels and correspondingly different priorities. The chapter estimates the determinants of environmental policies’ stringency in the European enlargement context. We find that corruption levels are the most important factor in explaining the variance in environmental policies in the enlarged EU. Most notably differences in corruption levels across countries appear to be more important than income differences. Thus it is argued that lower environmental standards in new member states are not necessarily implied by lower-income levels, but more likely reflect low institutional quality. We argue that harmonization of environmental policies at the EU level can be a way to tackle this problem, and we provide a further rationale for new member states to adjust to existing EU environmental directives. JEL classification╇ H40 • D73 • Q56 • Q58 Keywords╇ Corruption • European Union • Environmental policy • New member states • EU enlargement
6.1 Introduction For decades the European Union (EU) has developed a growing body of environmental policies. The earliest European legislation on environmental issues dates from 1959, when a directive on radiation safety standards was passed into law (Tamara 1997). In 1972, the European Community instructed the Commission to draw up the First Environmental Action Programme (Andersen and Liefferink 1997), the first comprehensive environmental policy initiative. The attention on environmental protection is apparent, also from its explicit mentioning in recent treaties such as This chapter is a slightly modified version of Pellegrini and Gerlagh (2006a). L. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9_6, ©Â€Springer Science+Business Media B.V. 2011
101
102
6 Corruption and Environmental Policies
Fig. 6.1↜渀 Environmental Regulatory Regime Index (↜ERRI). A darker colour indicates more stringent environmental policies. Only countries that are EU members and for which the ERRI is available are included in the map
the “Consolidated Version of the Treaty Establishing the European Community” (emended in Maastricht 1992) and “Treaty on European Union” (also known as the Maastricht Treaty 1992). In the European Union’s draft Constitution—signed on the 29th of October 2004 but never ratified by the member states—sustainable development and “a high level of protection and improvement of the quality of the environment” are mentioned already in the first article among the main objectives of the Union. In 2004, ten new states entered the EU: Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia, and Slovenia. Relative to the pre2004 or “old” member states, these accession states have lower environmental standards, and some worry that it will be too demanding for these new EU members to fully comply with European environmental provisions.1 Nonetheless, a remarkable effort has been undertaken by the EU in order to secure compliance. In the Act of Accession of the ten new members, among the permanent provisions, there is a list of environmental issues on which there is the need of adopting and implementing the EU environmental acquis2 at the national level. Figure€6.1 offers a geographical 1╇ We note that less stringent environmental standards are not necessarily associated with lower environmental quality and that with respect to environmental pressures and quality the situation of new EU members is uneven depending on the issue (e.g. European Environment Agency 2003). 2╇ Acquis Communautaire is the expression used, in European Union law, to refer to all the regulations accumulated over time in the EU.
6.1 Introduction
103
representation of the stringency of environmental regulations across EU countries,3 and portrays the situation in which newer members of the EU have generally lower environmental policy standards. The case for EU’s environmental policies, however, cannot be taken for granted. The EU member states are heterogeneous and in the environmental sphere this has introduced fears of excessive regulations. These regulations could damage welfare levels as they detract resources that could be devoted to more urgent needs according to national priorities. As it is generally accepted that demand for environmental quality (as a normal good) increases with income, there is an argument that poorer countries prefer to opt for laxer environmental policies, avoiding the investment in environmental protection of an unduly high share of their income. This type of reasoning gains further relevance as the new member states have, on average, a lower level of income than older member states. Therefore, preferences among EU states will be further diversified—because of increasing differences in income levels—and some countries in the enlarged EU will be less sensitive towards environmental issues.4 Moreover, environmental issues and ecological conditions differ from country to country and a uniform approach to environmental policies can have disproportionately high costs for some countries without producing adequate benefits (Haigh 1992; Berglund et€ al. 2002). Figure€ 6.2 shows the geographical variance in income levels; it appears that new members are generally characterized by lower-income levels. At the same time, there are various arguments that explain the EU effort to protect environmental quality equally in old and new member states. The most fundamental one—which is cited most often in official documents (such as the above mentioned treaties)—is that the EU is more than a group of countries harmonizing their regulations in order to exploit the access to a larger common market. The EU is a political subject and the welfare of the EU citizens is at the centre of its concerns. It therefore has an active attitude towards countries that are lagging behind in defending the interests of their citizens and the political project of the EU, and these countries are to some extent forced to catch up with the European aquis. This stand is clearly present in the Maastricht Treaty. Another argument, non-specific to the EU as a political subject, is that many environmental problems have trans-boundary effects on neighbouring countries. In the case of these pollutants, the EU is an obvious forum for member countries to regulate these sorts of externality. On similar grounds, as the EU is an integrated market, the application of different environmental policies that result in cost differentials among countries would promote the transfer of polluting production activities from countries with more stringent policies to countries with looser environmental policies (Andersen and Liefferink 1997; Weale et€al. 2000, pp.€34–37). The introduction of differentiated policies in an integrated market would produce pollution leakage: the environment would not benefit optimally from the environmental 3╇ The figure is based on the Environmental Regulatory Regime Index, which is described below when we use it in our statistical analysis. The maps are from the “ESRI Data & Maps” CD-ROM (Environmental Systems Research Institute 1999). 4╇ See Tefertiller (2001) for similar arguments used in a U.S. context.
104
6 Corruption and Environmental Policies
Fig. 6.2↜渀 GDP per capita (measured in 1997 in Euro adjusted for purchasing power parity). A darker colour indicates higher GDP levels. Only countries that are EU members and for which the ERRI is available are included in the map
protective provisions and the most environmentally concerned countries would be economically harmed. There is ample evidence that income affects environmental policies. Pellegrini and Gerlagh (2006b) provide econometric estimates for the—expected—positive relationship between income and Environmental Policy Stringency (↜EPS, see Chap.€5). At the same time, the authors highlight the role of corruption in shaping the stringency of environmental policies. In this chapter, we reproduce some estimates of the determinants of EPS and we relate them to the realm of environmental policies in the EU. Through further statistical analysis, this chapter argues that applying these findings to the environmental arena in the EU underscores the rationale for the Union’s interventions in environmental policies, including the provision of higher environmental standards in the new member states. It also supports the EU’s active role in environmental policies, as citizens’ concerns are often better served by the EU effort to achieve an upward environmental harmonization, compared to the country level policy making where environmental protection is more often affected by domestic corruption. The chapter is organized as follows. The next section gives an account of the academic discussion on the determinants of EPS and it presents some econometric results; Sect.€6.3 puts in relation environmental policies with institutional settings
6.2 Determinants of Environmental Policy Stringency
105
in the EU; Sect.€6.4—drawing from the preceding analysis—discusses the implications of our findings for EU’s stand on environmental policies and concludes.
6.2 Determinants of Environmental Policy Stringency Economic theory suggests that the environment (or environmental quality) can be treated either as a normal or as a luxury good: its demand increases with income. Increased demand of environmental quality for high-income levels is one of the main explanations backing the Environment Kutznets Curve (EKC; Grossman and Krueger 1995), the commonly observed path along which environmental degradation is on the rise jointly with income growth for low levels of income, while after a turning point further increases in income correspond to a fall in pollution levels.5 One of the arguments explaining the inverted—U income—pollution relationship is the increased demand for environmental quality caused by increases in income, together with an assumed policy response (for a discussion see Roca 2003). Other literature strands have highlighted the effects of institutional settings on building environmental policies. For example, the linkages among democracy and the environment have been discussed many times (Payne 1995; Neumayer 2002). Likewise, the literature has also analysed the effects of corruption on the formulation and implementation of environmental policies (Lopez and Mitra 2000; Damania 2002). From the political science perspective the literature highlighted the complexity of societal requirements for addressing environmental issues (e.g. Weidner 2002).6 See Sect.€5.2 for a review of the literature on the determinants of environmental policy stringency. Recently, a strand of empirical literature has developed on the determinants of environmental quality (the above mentioned EKC is one example) and also estimates have been produced on the determinants of environmental policies (for a review see Pellegrini and Gerlagh 2006b). From a political economy perspective, it is no surprise to find that environmental policies are affected by the quality of governance structures. When environmental regulation harms economically endowed and concentrated interests, those affected negatively can easily raise funds to influence policy makers and bureaucrats in order to deter the approval and implementation of costly regulations. On the other hand, the benefits of most environmental policies are common goods affecting the polity at large, thus common citizens face a coordination problem when they would need to collect resources for buying influence 5╇ For methodological and statistical assessments of the EKC, see Spangenberg (2001) and Harbaugh et€al. (2002). 6╇ With the capacity-building approach, Weidner (2002) analyses 30 case studies from the ecological modernization point of view. The author is able to identify time cycles, influences of forerunners, institutional capacities, diffusion of environmental innovation (and its pace and pattern) as relevant determinant of ecological modernization.
106
6 Corruption and Environmental Policies
in order to have the environmental regulations enacted. There is thus a need for high-quality institutions that put the polity’s interest in focus, and that prevent selfinterested policy makers from maximizing their own benefits, which would make them relatively insensitive to polity’s demands for increases in the stringency of environmental policy. Also, when policy makers decide to set-up environmental regulation, but bureaucrats are susceptible to bribes, the implementation of policies and the achievement of environmental objectives become problematic. Thus, the argument applies both to the making of environmental regulation and the enforcement of written policy, and both grand and petty corruption would have a bearing on environmental policies.
6.2.1 Cross-Country Evidence First, we produce econometric estimates of the determinants of stringency of environmental policies in a cross-section of countries. Subsequently, we focus on the set of countries that are within the sphere of influence of EU’s environmental policies, i.e. old and new members, and candidate countries. Using two sets of indexes of the stringency of environmental policy, in Chap.€5 above we found, in two cross-sections of countries referring to two different time frames, that the main determinant of environmental policies is the country’s level of corruption (see also Pellegrini and Gerlagh 2006b). Firstly, we carried out regressions that identify the determinants of the index of EPS, which refers to the year 1991 and is based on data gathered for the UN summit in Rio de Janeiro in 1992 (see also Eliste and Fredriksson 2002). Secondly, we performed a similar econometric analysis for the Environmental Regulatory Regime Index (↜ERRI), which was compiled in 2002 and is based on a sub-set of the indexes forming the Environmental Sustainability Index, augmented by data from the competitiveness survey of the World Economic Forum (see also Esty and Porter 2002). Both these indexes refer to environmental policies in an extensive way (see Sect.€5.3 for a description). They measure the stringency of the stated objectives and use the available information on environmental qualities, and evaluate the presence of institutions implementing them and the quality of the regulatory framework. Moreover, the ERRI also looks at the actual share of environmental expenditures in the budget of firms in different countries, thus evaluating not only the stated objectives, but also the implementation of environmental policies. On the side of the independent variables, it must be noted that the effect on environmental policy attributed to corruption could partially include the effect of other institutional deficiencies of a country. One often cannot use multiple alternative measures of institutional quality as these are too correlated to disentangle each one’s effect on environmental policy. Thus, even though it is safe to say that the main driver of our statistical results is corruption, it is worth noting that some omitted variables, strongly correlated to corruption, can contribute to the result. Indeed, the presence of corruption will most likely imply, for example, a lack of rule of the
6.2 Determinants of Environmental Policy Stringency
107
law and a lack of efficient and independent judiciary power.7 At the same time, we should not take the argument too far in favour of a broad interpretation. Many institutional features (e.g. decentralization, veto points in the legislature, guarantees of environmental quality in the constitution, etc.) are not correlated with corruption. The impacts of these institutional variables (that do not correlate with corruption and are omitted from the regressions) on environmental policies are not captured in the coefficient of corruption. In this chapter, we present an econometric analysis on the determinants of the stringency of environmental policies making use of ERRI as a dependent variable. The choice of this indicator of the stringency of environmental policies is based on the fact that most European countries are included in the sample of countries for which the index is available. Furthermore, the base year for the EPS is 1991 and the former communist countries of eastern Europe have undergone dramatic institutional and environmental policy changes over the last decade. The ERRI, which is compiled for the year 2001, is more relevant for the actual environmental policy of Europe.8 Lastly, as already mentioned, we note that the ERRI has the advantage of also including measures of environmental policy implementation. For this index, and for the other variables, see the appendix for sources and descriptive statistics. Furthermore, we use, as a proxy of corruption levels, the Corruption Perception Index gathered by Transparency International (see Sect.€3.3 for a description, Lambsdorff 2001). In order to estimate the influences of income and corruption on the formulation and implementation of environmental policies, we estimate the following model:
i ERRI i = α0 + α1 ln Y1997 + α2 C i + α3 Z i + εi
(6.1)
where the superscript i denotes each country in the sample, ERRI is the Environmental Regulatory Regime Index, Y1997 is income per capita in 1997 and C is the Corruption Perception Index referring to 2001. Finally, Z is a vector of additional explanatory variables that are introduced in order to check the robustness of our findings.9 The results of the regressions are reported in Table€6.1, regression (1), showing the correlation between income and environmental policy, reproduces the finding 7╇ When we tried running regressions that would include additional explanatory variables from Kaufmann et€al. (2004) the result was that the variance inflation factor, for some variables including corruption, was up to more than 20 (with the conventional level of multicollinearity detection set at 10). 8╇ An extensive econometric analysis similar to the one undertaken here, but including EPS as a dependent variable, can be found in Pellegrini and Gerlagh (2006b). The authors used slightly different time frames and variables in their analysis. Most notably, they estimated also the effects of democracy on policy stringency, including a democracy index in their regressions. Here the democracy variable is omitted as there is little variation in the value of the democracy indexes within European countries and because the democracy variable—from the Polity IV project—was not significant in any of our regressions. 9╇ Our regression methodology is based on the assumption that causality is unidirectional: the effect of the dependent variable (↜ERRI) on the independent variables (including corruption) is negligible.
108 Table 6.1↜渀 Regressions as in Eq. 6.1 Dependent variable: ERRI (1) 2.51*** Income (12.65) Corruption Schooling Urbanization Old EU members
6 Corruption and Environmental Policies
(2) 0.51** (2.00) −0.80*** (8.46)
(3) 0.69** (2.23) −0.79*** (7.57) 0.06 (0.63) −1.40* (1.67)
(4) 0.47* (1.71) −0.75*** (6.56) 0.12 (1.06) −1.34* (1.69) 0.68 (1.59)
(5) 1.93** (2.82) −0.61*** (3.52) 0.32*** (2.99) 0.72 (0.33)
R2 0.70 0.86 0.87 0.87 0.88 69 66 59 59 21 Number of countries OLS estimation with the Environmental Regulatory Regime Index as dependent variable. Old EU members is a dummy variable for pre-2004 EU members. The constants are included in the regressions, but the coefficients are omitted from the table. t-statistics, based on robust standard errors, are in parenthesis under the coefficients *10% significance; **5% significance; ***1% significance
we expect from economic theory: richer countries tend to have more stringent environmental policies. The income variable has a statistically significant coefficient and a one standard deviation in the value of the income variable is associated with an increase of the environmental policy index by more than 0.8 standard deviations.10 In regression (2), once we include the corruption variable, we notice a drastic drop in the absolute value of the coefficient of the income variable: from 2.5 to 0.5. Also the statistical significance is reduced and the coefficient is significant only at 5%. At the same time, the coefficient on the corruption variable is sizeable in magnitude and is highly significant. Now, corruption turns out to be more important, as in this regression a one standard deviation change in corruption is associated with a 0.8 standard deviation change in the ERRI. A comparison of the two first regressions suggests that the coefficient of the income variable from regression (1) is inflated by an omitted variable bias. When the income and corruption proxies are included together in the regression, the effect of corruption appears to dominate the effect of income. Further evidence of the association between corruption and environmental policy is provided by the scatter plot in Fig.€6.3, were we plot the ERRI variable against the corruption perception index. In regression (3) schooling and urbanization are included as explanatory variables, following Pellegrini and Gerlagh (2006b). The schooling variable expresses the number of years spent at school, on average, for the population above 25 years old in 2000. The urbanization variable is the percentage of the total population that We interpret some of the results in standardized terms: we consider what change, in standard deviation terms, in the dependent variable is associated with a one standard deviation change in the independent variable.
10╇
6.2 Determinants of Environmental Policy Stringency FIN
10 Environmental Regulatory Regime Index
109
SWE NLD DNK
8
AUT GER FRA
GBR 6
BEL IRL
4
ESP
ITA EST HUN SVN CZE POL LVA PRT SVK LTU BGR GRC
2
ROM World
0 0
EU (old and new members) 2
4
6
8
10
Corruption Perception Index
Fig. 6.3↜渀 Scatter plot for corruption and the Environmental Regulatory Regime Index. The regression line is estimated to fit all countries in the sample, not only EU countries
lives in urban areas in 1999. In general, we would expect that the schooling variable would have a positive bearing on environmental policy stringency: the more educated the population, the more aware the citizens are about environmental problems. Moreover, a more educated polity will better be able to scrutinize measures that policy makers put in place to tackle environmental issues. The coefficient of the schooling variable is indeed positive, but it is very small in absolute terms and insignificant. The predicted effect of the urbanization variable is unclear. On the one hand, increased urbanization is associated with more concentrated population and urban citizens can more easily co-operate in order to push policy makers to undertake measures, such as setting environmental standards that satisfy their preferences. On the other hand, a highly urbanized population, more detached from nature, may be less interested in environmental protection. In our analysis, we find some weak evidence of the second effect dominating; the coefficient on urbanization is negative, but significant only at about 10% level (see also Pellegrini and Gerlagh 2006b and Chap.€5). A side effect of these additional variables is the increased size and statistical significance of the coefficient of income, whereas the corruption coefficient is not affected substantially. In regression (4), a dummy variable is included to verify whether there is a residual in EPS specific for the pre-2004 EU members. Indeed, the EU members seem to have slightly stricter environmental policies: the coefficient on the EU dummy is positive though it is only significant at just above 10%. An obvious argument explaining this higher stringency for the EU is the Union’s environmental policy
110
6 Corruption and Environmental Policies
that we outlined above. The EU has pushed environmental policy laggards to adopt stricter policies more in line with the forerunners. We notice that again the coefficient of the corruption variable is only slightly affected by the inclusion of additional explanatory variables. Overall, the econometric evidence presented here suggests, in line with previous findings, that corruption levels negatively affect the stringency of environmental policies. Our estimates suggest that, at a cross-country level, a one standard deviation decrease in the corruption variable is associated with a more than two-thirds improvement in the ERRI. This association appears to be statistically significant and robust. The income variable is associated with less variation of the ERRI; a one standard deviation increase in the income proxy is associated with half a standard deviation increase in the ERRI in regression (4), and the statistical significance ranges from 5% to 10%. It is important to highlight that many of our independent variables are highly correlated and this can cause multicollinearity. This is most obvious when in regression (3) we introduce the schooling variable and the urbanization variable. These variables are correlated between themselves and they are highly correlated with income levels (see Table€6.2). This results in an inflation of the standard error of the coefficients of these variables, and a decrease in their statistical significance. Given our sample size, this could be a serious problem when we try to disentangle the effects of the individual variables on environmental policy. It is important to note, however, that the purpose of this chapter is not to provide statistical evidence on the whole range of possible determinants of environmental policy, but to test the importance of the influence of corruption and to evaluate the impact of corruption and income within the EU countries. Stated positively, the fact that the corruption variable continues to be highly significant in all our regressions can be considered as an extreme test for the relevance of corruption for environmental policy stringency.11
Table 6.2↜渀 Correlations Income ERRI 1.00 0.86 ERRI 1.00 Income Corruption Schooling Urbanization Correlations. All the variables in the table are observations: 59
Corruption −0.92 −0.87 1.00
Schooling 0.73 0.78 −0.76 1.00
Urbanization 0.55 0.65 0.63 0.60 1.00 correlated at 1% level of significance. Number of
11╇ In any case, when we calculated the variance inflation factor, in regression (4), it was never higher than 6 (the conventional value for signaling serious multicollinearity is 10), and the variables income and corruption were the variables with the highest values.
6.3 Environmental Policies and Institutions in the EU
111
6.3 Environmental Policies and Institutions in the EU Now we turn to the implications of the previous analysis for environmental policies in the EU. From Fig.€6.3, we can see that European countries align on the global regression line.12 In this section, we will describe the efforts (and the shortcomings) of the EU to induce the new member and the candidate countries to tackle corruption before accession. We will also briefly touch upon the (lack of a coherent) effort of the EU on this issue with respect to older members. Furthermore, we will apply the results of the previous analysis in order to estimate the effect of corruption on the stringency of environmental policy for European countries.
6.3.1 T he Accession Process and Its Review: Focus on Corruption The progress made, by candidate and accession countries, towards the integration in the EU has been assessed through regular reports. The core of the criteria used for single country evaluations are the so-called “Copenhagen Criteria” (set in 1993€at the Copenhagen European Council).13 Part of the first criterion is the establishment of the “rule of the law”, which is not compatible with widespread corruption. The reports compiled by the Commission, which are at the basis for the decision “if and when” countries would be ready to access the Union, have always included corruption levels and trends thereof as central issues. Figure€6.4 offers a geo-referenced representation of corruption levels across Europe, where it is possible to see that eastern and southern countries are most affected by corruption. The EU’s assessments of corruption levels and trends have been criticized both methodologically and, even more seriously, for their content (Open Society Institute 2002). The assessment by the Commission is said to lack a coherent framework and the information used for the assessment of countries’ performances is derived from different sources and compiled with different methodologies. For example, in some country reports opinion polls have been used as evidence for assessing corruption levels, while in other country reports expert opinions, or even the number of convictions have been used. Furthermore, these sources have changed between years, and Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, Slovenia are the new EU members for which the ERRI is available. Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Netherlands, Portugal, Spain, Sweden, United Kingdom are the pre-2004 EU members for which the ERRI is available. 13╇ The criteria are usually broadly divided in three categories: the political criteria, the economic criteria, and the criteria of adoption of the acquis. The political criteria refers to the stability of institutions guaranteeing democracy, the rule of law, human rights, and respect for and protection of minorities; the economic criteria demands the existence of a functioning market economy as well as the capacity to cope with competitive pressure and market forces within the Union; the criteria of adoption of the acquis relates to the ability to take on the obligations of membership including adherence to the aims of political, economic, and monetary union. 12╇
112
6 Corruption and Environmental Policies
Fig. 6.4↜渀 Corruption Perception Index 2001. A darker colour indicates higher perceived corruption. Only countries that are EU members and for which the ERRI is available are included in the map
such changes do not support comparison of assessments over time (which appears to be a fundamental requirement for the quality of a yearly assessment of progress towards a goal). From a more substantial point of view, the lack of a clear benchmark and the weakness of the pressure to tackle corruption are apparently motivated by the fact that some pre-2004 member states would not be able to comply themselves with strict anticorruption frameworks (Open Society Institute 2002). According to most surveys, the least corrupt of the new members do better than some of EU’s founding members. Specifically, according to the Corruption Perception Index 2004, Malta, Estonia, Slovenia, and Cyprus are all affected by a lower degree of corruption than Italy and Greece. Already the Commission has been pushing new members and candidate countries to undertake initiatives to counteract corruption whereas member states have been reluctant to adopt these same regulations. A good example is the ratification of the “Criminal Law Convention on Corruption”, which the Commission has pressed applicant countries to sign and ratify. As of July 2004, all new members and candidates had ratified, while six out of 15 pre-2004 members still have not ratified it and Spain did not even sign.14 Greece and Italy, the member 14╇ http://conventions.coe.int/Treaty/Commun/ChercheSig.asp?NTâ•›=â•›173&CMâ•›=â•›7&DFâ•›=â•›06/07/04 &CL=ENG.
6.3 Environmental Policies and Institutions in the EU
113
countries that have the worst rating in the above-mentioned Corruption Perception Index 2004, are among the countries that did not ratify the convention. From this perspective it is easy to understand why the Commission cannot press the applicant countries too much to fight corruption. Notwithstanding these caveats, the EU’s requirements for accession have induced new and candidate members to undertake several initiatives in order to limit corruption. All the new member states have signed and ratified international conventions and modified their domestic legislation in order to fulfil the formal requirements of the EU. Nevertheless, it is at the implementation levels that many countries have failed to meet the standards of the EU. Most notably, in the case of Romania’s 2004 Regular Report (Commission of the European Communities 2004) the picture of progress made to improve corruption levels—which are rather high—was considered unsatisfactory. A passage deserves a long citation: “corruption remains a serious and widespread problem in Romania which affects almost all aspect of society. There has been no reduction in perceived levels of corruption and the number of successful prosecutions is low, particularly for high-level corruption. The fight against corruption is hampered by integrity problems even within institutions that are involved in law enforcement and the fight against corruption” (Commission of the European Communities 2004, p.€21). Furthermore, it must be noted that the report was published on the same day when the Commission confirmed that—thanks to its progress towards integration— Romania’s accession date was confirmed to be the 1st of January 2007. Looking at the overall corruption levels in new EU countries, the record of corruption that is depicted by the statistics of Transparency International is discouraging. From Table€6.5 in the Appendix we see that on average their score is equal to 5.7, while the older members of the EU score 2.5. This is equal to a difference of one standard deviation on the global scale.
6.3.2 N ew and Old Member States of the EU: Environmental Policies and Corruption Levels Regression (5) in Table€6.1 presents the statistical evidence on income, corruption, and EPS in the enlarged EU, where also the urbanization and schooling variables are included without affecting the significance of the corruption variable. Though regression results with such a small sample of countries as in regression (5) should be interpreted carefully, together with the global regressions, a robust pattern appears. When we compare the coefficients for the EU estimation with the world-wide cross-country evidence—that we presented above—we see that the dynamics of environmental policies within the EU reflect the global patterns, and specifically, we find that corruption is a (negative) determinant of EPS levels.15 From analysing the effect that corruption has on policy makers in the environmental field, it is also worth mentioning possible future strategic behaviour once the new member states’ policy
15╇
114
6 Corruption and Environmental Policies
10
ERRI (After controlling for Income)
FIN 8
SWE NLD DNK
6
AUT GER
FRA
GBR
BEL
IRL 4
LTU
ESP PRT
EU pre-2004
2 0
EST HUN SVN ITA
EU new members 2
4 Corruption Perception Index
POL
LVA
CZE BGR SVK
GRC 6
ROM
8
Fig. 6.5↜渀 Scatter plot with the ERRI (adjusted for income) on the y-axis and corruption on the x-axis
Figure€6.5 portrays this insight. For this figure, we adjusted the ERRI for income (using the coefficients from regression (5)), and plotted the adjusted ERRI values against corruption levels. The figure shows a strong correlation between the corruption variable and the EPS Index. Overall, the new member countries are grouped at the right-hand side characterized by higher levels of corruption and lower levels of environmental policy. To apply the figure to policy, we put in some numbers. For corruption, the gap between the average performance of the old members of the EU and the new members is 3.2 (5.7−2.5, see Table€6.3). This gap is substantially larger than the standard deviation of the corruption level within each of the groups of the old and new EU members, which is 1.6 and 0.9, respectively (Table€6.3). Thus, the variance between the groups exceeds the variance within the groups. If Poland were to improve its corruption index (5.9) to the level of Germany (2.6), on the basis of this change alone, we would expect its environmental performance as measured by the ERRI variable to improve by two points.16 If Poland would also increase its income level from makers enter more actively the European policy arena. On the one hand, negotiators that are more sensitive to kickbacks could try to stir the process in favour of lobbying interests. On the other hand, undersigning agreements in Brussels does not always entail undertaking the necessary steps to enforce them back home, in which case a more relaxed attitude is foreseeable. Moreover, some EU policies are also supported by endowments for covering their costs. Such funds can create an incentive for corrupt policy makers to accept those policies in an attempt to embezzle part of the funds. Thus, corruption could further complicate the operations in the EU environmental policy arena, but an in depth analysis of these scenarios is beyond the scope of this chapter. 16╇ We multiply 2.6−5.9â•›=â•›−3.3 with −0.61 from regression (5).
6.4 Conclusions
115
Table 6.3↜渀 Descriptive statistics for EU countries Mean
Standard Min Max deviation 6.86 2.04 2.54 10.00 ERRI (old members) 3.78 1.21 0.89 4.88 ERRI (new members) 2.54 1.58 0.10 5.80 Corruption (old members) 5.73 0.91 4.40 7.20 Corruption (new members) 9.62 0.18 9.22 9.85 Income (old members) 8.74 0.36 8.17 9.26 Income (new members) Descriptive statistics for old EU members (14 countries in our sample) and new members (10 countries in our sample), as used together with regression (5) for numerical illustrations in text
6,224€€17 per capita (in 1997) to 15,266€€ per capita (the average income of old EU members in 1997), this would increase the expected ERRI variable by 1.7.18 Thus, for Poland, catching up with EU low corruption levels can be expected to have more effect on its environmental policies than catching up with EU income levels. The effect would be even stronger when applied to the most corrupt country among the new members: Romania. If this country were to catch up and improve its corruption index to the average of the pre-2004 EU members (that is, reducing the corruption index from 7.2 to 2.5) it would improve its ERRI by 2.9 points.19 This alone would improve its position from the 24th position to the 20th in our sample of EU countries. It must be noted that these calculations are based on coefficients from regression (5) and they tend to be conservative; they will probably give too low a weight to the importance of corruption, compared to the effect of income increases. When we apply the coefficients from regression (4), which are based on the largest set of countries for which we have all data, we see that the role of corruption becomes more important relative to the role of income. In general, we consider regression (4) more reliable, because of the larger sample and of the reduced role for outliers, but we decided to use the coefficients from regression (5) to make a conservative calculation on the impact of corruption on environmental policy.
6.4 Conclusions The accession to the European Union of new member states represents a formidable challenge for the institutions of the EU. Not only has the enlargement created a more economically, environmentally, and socially diverse EU; the new countries— on average—are also affected by corruption to a higher degree than the pre-2004 members and their progress towards an improvement of their corruption records has produced mixed results. In this section, we highlight the effects of these differences on environmental policies concluding that the influence of corruption on EPS proFor convenience, we use the Euro as monetary unit. We could also use the European Currency Unit (ECU), which was commonly used before the 1st of January 1999, when it was converted into the Euro at a one-to-one exchange rate (i.e. 1€ECUâ•›=â•›1€€). 18╇ We multiply ln(6,224)−ln(15,266)â•›=â•›−0.9 with −1.93 from regression (5). 19╇ We multiply 7.2−2.5â•›=â•›−4.7 with −0.61 from regression (5). 17╇
116
6 Corruption and Environmental Policies
vides a further rationale for the formulation and implementation of environmental policies at the EU level. The EU enlargement is easily used as an argument to restrict the role of environmental policies because of the increase in variation in socio-economic and cultural conditions. The presence of different income levels provides impetus to those who argue for a reduced role of supranational environmental policies. Poorer countries should pursue economic objectives first and only when these objectives are met should they concentrate on environmental quality. Also, increasingly different preferences among EU’s citizens towards the environment are likely to arise. Added to income differences, the root of differences in preferences can also lie in variation in culture. It can be argued that each country should be allowed to pursue its own way in order to achieve higher welfare standards according to its own preferences. Finally, the enlargement process also implies an increase in environmental diversity. Thus, while some environmental measures are considered necessities in some countries, they may be superfluous in other countries that have a different environment.20 However, the arguments in favour of allowing diversity in environmental policy too easily neglect a major cause for this diversity, the difference in institutional quality among the countries. Given the numerical results presented above, it is likely that environmental policies in new EU member states are at a low level because of institutional failure. The EU environmental provisions could therefore be seen as possible corrections of national policy failures. Moreover, as corruption is a persistent phenomenon, it will take a long period for the new EU member states to catch up with average EU levels, and enforcement of higher environmental standards can then be understood as an early reaping of the fruits thereof. Indeed, the evidence presented here suggests that the enlarged EU can and should serve as a forum for the advancement and the diffusion of more progressive and stringent environmental policies among member countries (Andersen and Liefferink 1997). Having said this, we have to recognize that, given an increasing diversity in institutional quality among its members, for the EU to continue its role in improving environmental policy in the member states, it may be necessary to limit the role of purposive statements, and to shift emphasis to imposing measurable standards for implementation. Also capacity building and governance at the regulatory stage, when bureaucratic corruption can still play a role, should receive due attention.
Appendix: Data Tables 6.4 and 6.5 present data and descriptive statistics, respectively.
For an example on cross-country differences on opportunities and costs of paper recycling see Berglund et€al. (2002).
20╇
Appendix: Data
117
Table 6.4↜渀 Data for EU countries ERRI Income Old EU members ╇ 8.31 16,601 Austria Belgium ╇ 7.08 16,700 Denmark ╇ 7.66 18,940 Spain ╇ 5.24 12,339 Finland 10.00 15,802 France ╇ 7.86 15,680 United Kingdom ╇ 7.15 15,832 Germany ╇ 8.01 16,344 Greece ╇ 2.54 10,080 ╇ 5.52 15,536 Ireland Italy ╇ 5.39 15,961 Netherlands ╇ 8.58 16,929 Portugal ╇ 4.05 10,720 Sweden ╇ 8.65 16,257
Corruption
Schooling
Urbanization
2.20 3.40 0.50 3.00 0.10 3.30 1.70 2.60 5.80 2.50 4.50 1.20 3.70 1.00
╇ 8.80 ╇ 8.73 10.09 ╇ 7.25 10.14 ╇ 8.38 ╇ 9.35 ╇ 9.75 ╇ 8.52 ╇ 9.02 ╇ 7.00 ╇ 9.24 ╇ 4.91 11.36
0.65 0.97 0.85 0.77 0.67 0.75 0.89 0.87 0.60 0.59 0.67 0.89 0.63 0.83
New members Bulgaria Czech Republic Estonia Hungary Lithuania Latvia Poland Romania Slovakia Slovenia
6.10 6.10 4.40 4.70 5.20 6.60 5.90 7.20 6.30 4.80
9.74 9.46 – 8.81 – – 9.90 9.51 9.19 7.35
0.69 0.75 0.69 0.64 0.68 0.69 0.65 0.56 0.57 0.50
2.63 4.31 4.88 4.85 3.75 4.03 4.14 0.89 3.67 4.66
╇ 4,171 10,285 ╇ 6,292 ╇ 6,964 ╇ 5,217 ╇ 5,120 ╇ 6,224 ╇ 3,546 ╇ 8,070 10,539
Table 6.5↜渀 Descriptive statistics Mean
Standard Min deviation 4.2 2.5 0.0 ERRI 8.8 0.8 6.7 Income (in GDP/cap) 5.0 2.4 0.1 Corruption 7.7 2.4 2.4 Schooling 0.7 0.2 0.2 Urbanization Descriptive statistics for the 66 observations sample—as in regression (2)
Max 10.0 10.0 9.6 12.2 1.0
ERRI is the Environmental Regulatory Regime Index which includes measures of the stringency of environmental standards, the quality of the regulatory regime, and the size of subsides favouring the consumption of natural resources (see Esty and Porter 2002). Income is the natural logarithm of GDP measured in Euro, adjusted for purchasing power parity, and refers to the year 1997 (Summers and Heston database, version 6.1, available at http://pwt.econ.upenn.edu/; Euro the exchange rate from US dollar to Euro is 1.30801, from Eurostat, available at http://epp.eurostat.cec.eu.int/).
118
6 Corruption and Environmental Policies
Corruption is the Corruption Perception Index, see Sect.€3.3 (Transparency International, available at http://www.transparency.org/). Schooling is the number of years spent at school, on average, for the population above 25 years old in 2000 (Barro–Lee “International data on educational attainment” dataset (version updated to April 2000) and are available at www.cid.harvard. edu/ciddata/ciddata.html). Urbanization is the percentage of the total population that lives in urban areas in 1999 (World Development Indicators 2004).
References Andersen, M. S., & Liefferink, D. (1997). European environmental policy: The pioneers. Manchester: Manchester University Press. Berglund, C., Soderholm, P., & Nilsson, M. (2002). A note on inter-country differences in waste paper recovery and utilization. Resources Conservation and Recycling, 34(3), 175–191. Commission of the European Communities. (2004). 2004 regular report on Romania’s progress towards accession. Brussels: European Union. Damania, R. (2002). Environmental controls with corrupt bureaucrats. Environment and Development Economics, 7, 407–427. Eliste, P., & Fredriksson, P. G. (2002). Environmental regulations, transfers, and trade: Theory and evidence. Journal of Environmental Economics and Management, 43(2), 234–250. Esty, D. C., & Porter, M. (2002). Ranking national environmental regulation and performance: A leading indicator of future competitiveness? In M. Porter, J. Sachs, & K. Schwab (Eds.), The global competitiveness report 2001–2002 (pp.€78–100). New York: Oxford University Press. European Environment Agency. (2003). Europe’s environment: The third assessment. Luxembourg: Office for Official Publications of the European Communities. Grossman, G. M., & Krueger, A. B. (1995). Economic-growth and the environment. Quarterly Journal of Economics, 110(2), 353–377. Haigh, N. (1992). Manual of environmental policy: The EC and Britain. London: Institute for European Environmental Policy. Harbaugh, W. T., Levinson, A., & Wilson, D. M. (2002). Reexamining the empirical evidence for an Environmental Kuznets Curve. Review of Economics and Statistics, 84(3), 541–551. Kaufmann, D., Kraay, A., & Mastruzzi, M. (2004). Governance matters III: Governance indicators for 1996–2002. World Bank policy research working paper. Washington: World Bank. Lambsdorff, J. G. (2001). Framework document to the 2001 Corruption Perceptions Index. Background Paper. Berlin: Transparency International. Lopez, R., & Mitra, S. (2000). Corruption, pollution, and the Kuznets Environment Curve. Journal of Environmental Economics and Management, 40, 137–150. Neumayer, E. (2002). Do democracies exhibit stronger international environmental commitment? A cross-country analysis. Journal of Peace Research, 39(2), 139–164. Open Society Institute. (2002). Monitoring the EU accession process: Corruption and anti-corruption policy. Budapest: Open Society Institute. Payne, R. A. (1995). Freedom and the environment. Journal of Democracy, 6(3), 41–55. Pellegrini, L., & Gerlagh, R. (2006a). Corruption and environmental policies: What are the implications for the enlarged EU? European Environment, 16(3), 139–154. Pellegrini, L., & Gerlagh, R. (2006b). Corruption, democracy, and environmental policy: An empirical contribution to the debate. The Journal of Environment & Development, 15(3), 332–354. Roca, J. (2003). Do individual preferences explain the Environmental Kuznets Curve? Ecological Economics, 45(1), 3–10.
References
119
Spangenberg, J. H. (2001). The Environmental Kuznets Curve: A methodological artefact? Population and Environment, 23(2), 175–191. Tamara, L. J. (1997). The debate over environmental standards in the European Community: A race to the top rather than a race to the bottom? New York University Environmental Law Journal, 6, 162–199. Tefertiller, K. R. (2001). Government, environmental regulations, and income distribution: Where you stand depends on where you sit—statistical data included. Choices, 3(3), 15–19. Weale, A., Pridham, G., Cini, M., et al. (2000). Environmental governance in Europe: An ever closer ecological union? Oxford: Oxford University Press. Weidner, H. (2002). Capacity building for ecological modernization—lessons from cross-national research. American Behavioral Scientist, 45(9), 1340–1368.
Chapter 7
The Rule of the Jungle in Pakistan: A Case Study on Corruption and Forest Management in Swat
Abstract╇ Corruption in Swat, Pakistan is impairing the sustainable management of forest. In this study, we focus on the factors leading to corruption in forest management in Swat: we analyse corruption in a case study setting against the backdrop of the reform options that are most often cited as possible solutions. As we highlight in this study, the “crime and punishment” approach (Becker 1968), is not feasibly implemented if the overall institutional environment is weak. Since countrywide overhaul of corruption through sweeping reform programs, the other reform approach which we call the “holistic approach”, is a difficult and lengthy task, which is seldom accomplished (Kaufmann et€al. 2005), we find that there is a need for alternative/complementary kind of reform. In this study, we argue for reforms that are limited in geographical and institutional scope, and which can accompany or precede more sweeping reform programs. In the case of a corruption ridden centralized forest management regime, institutional reform should move away from enforcement of existing institutions and promote communal management of natural resources by locals. That is not to say that more comprehensive reforms should not be pursued: their implementation could go hand in hand with local reforms. We believe that the lack of implementation of wider reforms cannot be an excuse for inaction. JEL classification╇ D73 • Q23 • Q58 Keywords╇ Corruption • Forest management • Environmental policy • Institutional reform • South West Frontier Province The research is part of the Poverty Reduction and Environmental Management (PREM) program. An early version of this chapter was published as Pellegrini (2007). Since the fieldwork was conducted and the data gathered, the Taliban insurgency has reached the area and the Swat valley has witnessed confrontations between the Pakistani army and militant Islamic groups. The ensuing anomy has rendered the institutional setting volatile and this will have an impact on the situation described here, especially in terms of the engagement of state authority and state agents in forest management. While this change in the area makes the analysis somewhat obsolete, it also highlight its importance since the failures in forest management and the related problems can have contributed to the discontent fuelling the insurgency. L. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9_7, ©Â€Springer Science+Business Media B.V. 2011
121
122
7 The Rule of the Jungle in Pakistan
7.1 Introduction One of the tenets of this book is that developing countries are often characterized by long-standing and severe environmental problems, and lack institutional strength to tackle them effectively. We will discuss Pakistan and the forest management failures in the Swat as one case in point as the control of logging here is fraught with difficulties since formal institutions have little influence on the behaviour of individuals and organizations. In this chapter, we analyse, in a case study setting, how corruption affects the management of forests. While examining the factors that are conducive to corruption in forest management, we also highlight the difficulties of applying the standard policy prescriptions for dealing with corruption. Finally, we outline how strategies based on institutional change at the local level can help improving the management regime and controlling corruption. Pakistani state and society face challenges associated with the history of the country: political instability, secessionist movements, lack of rule of law, and shortage of basic public services, just to mention a few. Deforestation is one of these challenges with worrisome environmental, social, and economic consequences. Environmental services provided by the forest include biodiversity, watershed protection, microclimate regulation, amenity, and services at the global level (Asian Development Bank 2000; Brown and Durst 2003). Furthermore, the services provided by the forest are also important for agricultural activities where the forests, apart from their role in keeping watersheds, also prevent local climate change that can damage crops. The importance of forests for the country was recognized when the extensive floods of 1992 were imputed to deforestation and a commercial timber extraction ban was enacted in the same year (Rao and Marwat 2003). From the social point of view, another reason for interest in deforestation is that poor communities extract a substantial share of their revenues from the forest (Rafi Khan et€al. 2006). The administrative resources available to the forest sector are contended over by all the unsatisfied and pressing needs mentioned above.1 It is worth highlighting here that in most countries deforestation—defined as forest areas cleared permanently—is a land use change resulting mostly from the transformation of forest area to agriculture and pasture (see Bromley 1999; Chomitz and Buys 2007). In Swat deforestation is primarily driven by logging. The valley has been subject to such an extensive deforestation that there are few remaining stands at the foot of the valley and land use change already took place on the most valuable agricultural land. In the higher reaches of the valley, trees are being cut primarily in order to extract wood and the coniferous forest is cleared as a consequence.
1╇ We need to be cautious in this claims since the linkages between deforestation and many of the problems attributed to it are actually rather complex. Clarifying whether the hydrology of the region is really linking deforestation to phenomena such as local climate change, flooding, or siltation goes beyond the scope of this chapter, but these impacts need to be understood as “perceived” impacts and their connection with deforestation as “likely” rather than “certain” (see Rao and Marwat 2003; Chomitz and Buys 2007).
7.1 Introduction
123
In virtually all countries, formal rules have been adopted in order to prevent depletion of renewable natural resources. Where—like in Pakistan—deforestation is taking place, institutional failure is self-evident: the de facto management of the forest contradicts the de jure setting. In this case study, we analyse the causes of institutional failure in the protection of forests in the North West Frontier Province (NWFP) of Pakistan, namely in the Swat district. Statutory law rigorously constrains exploitation activities and the Forest Department is in charge of enforcement, but idiosyncrasies between formal provisions and actual institutions create open access of forests to loggers. During the study corruption emerged as a fundamental contributor to the current pattern of forest exploitation. Corruption is crucial in the deforestation process: at every step of illegal logging rules are abused and bribes are disbursed. These take place during the extraction of logs, their transport, and their marketing. As we will highlight in this study, the “crime and punishment” approach (Becker 1968), which prescribes increase of monitoring activities and harsher punishments in order to decrease corruption, is not easily implemented if the overall institutional environment is not supportive. The other popular approach to deal with corruption—countrywide overhaul based on sweeping reforms, is a difficult and lengthy task which is seldom accomplished (Kaufmann et€al. 2005). Analysing corruption in a case study setting leads us to look for institutional reforms that are limited to a specific sector and that do not require far-reaching institutional change. Corruption is endemic to many developing countries but Pakistan always features in the highest positions in the rankings that are produced by international organizations (e.g. Kaufmann et€al. 2005; Transparency International 2004). In Pakistani society the phenomenon of corruption is particularly pervasive and affects also the operations of law and enforcement agencies (police and the judiciary) (Transparency International 2002). Surveys given by international organizations with the intent of comparison across countries indicate that members of law enforcement agencies are among the main offenders in corruption cases. The abuse of power in order to serve private interest is widespread also in the forestry sector of Pakistan whose regulations are enforced by members of the public administration. The difficulty of monitoring civil servants operating in remote areas, and the general corruptibility of Pakistani public administration contribute to the prevalence of corruption in the forest sector. Also during our study, through interviews with informed agents, we collected witness evidence about employees of the Forest Department and of the Forest Development Corporation (that is, the agencies in charge of enforcement in the forest sector) being involved in corruption in the forestry sector (see also Rafi Khan et€al. 2006).2 Due to the fact that environmental problems require active public policies to be addressed, it is not a surprise to find that countries that are in institutional disarray are not successful in dealing with environmental issues. Many of the benefits produced by the forest are public goods whose production requires successful coordination of the efforts of several individuals (Ostrom and Gardner 1993). Nevertheless, institu2╇ These corruption issues were confirmed during our fieldwork by the employees of the Forest Department and of the Forest Development Corporation themselves.
124
7 The Rule of the Jungle in Pakistan
tional failures at the national level do not always lead to environmental disruption and unsustainable resource use. For example, Ostrom (1990) reports a list of cases where natural resources are managed sustainably, as commons, by local communities. Many of these success stories are from developing countries characterized by countrywide institutional disarray. It appears then that institutions managing specific natural resource, when members of local communities coordinate effectively, can be successful even when the national institutional setting is problematic. These findings suggest that local institutional settings could achieve sustainable development even though the country as a whole is characterized by institutional shortfalls. Even though corruption is often cited in reports and articles about illegal logging, we find that there is a lack of case studies on corruption that put it in the institutional context and use an economic framework for suggesting reform.3 In our study, corruption will be analysed in a New Institutional Economics framework, where the focus will be on transaction costs and enforcement costs of property rights. The chapter proceeds as follows: the next section presents a literature review of studies of corruption and forest management, it analyses levels of corruption in Pakistan, and highlights the conditions for successful communal management. Section€7.3 describes geography and institutions in the case study area. Section€7.4 describes the dynamics producing deforestation. Section€7.5 summarizes the ongoing reform process, underlines its shortfalls and possible improvements of the process and Sect.€7.6 concludes.
7.2 Institutions and Forest Management In this section, we review the literature on corruption and forest management, discuss the incidence of corruption in Pakistan, we highlight the theoretical underpinning of the study, and outline the conditions that enhance communal management of natural resources.
7.2.1 Corruption and Forestry Dealing with environmental degradation requires mechanisms of social coordination to be corrected, but successful involvement of the government is difficult to 3╇ Existing studies serve, for example, the purpose of highlighting cases of officials and individuals involved in illegal logging in order to push authorities to intervene (e.g. Environmental Investigation Agency and Telapak 2001). The lack of case studies can be caused by the difficulties of getting information on illegal operations. Already Myrdal (1968) mentioned that corruption is “almost taboo as a research topic […], revealing a general bias that we have characterized as diplomacy in research”. Even though research on corruption has largely expanded in recent years, case studies involving fieldwork can still be affected by diplomacy in research; for instance, Perry (1997) still finds corruption to be taboo in geography.
7.2 Institutions and Forest Management
125
obtain. Those who have vested interests in the current state of affairs will use their power to arrest institutional change (grand corruption) and enforcement at the lower level by civil servants will be ineffective (petty corruption). In this section, first we summarize existing evidence of the role that corruption has on forestry; subsequently, we outline corruption incidence in the agencies most directly involved in the implementation and enforcement of forestry policies in Pakistan: the police, the judiciary, and the Forest Department. 7.2.1.1 Overview of Corruption and Forestry In Chaps.€5 and 6, we have discussed evidence of the influence of corruption over the stringency of environmental policy; here we focus more specifically on forestry. The World Bank’s (Callister 1999) report on corruption and forestry emphasizes the detrimental influence of corruption on forests’ management and conservation. As for case studies, in a report on illegal logging of Ramin timber in Indonesia, the Environmental Investigation Agency (EIA) and Telapak (2001) find that formal rules are in place to protect the endangered tree species and enforcement is relatively easy. The study finds that collusion between enforcement agencies and the “timber barons” is so strong that implementation is unattainable and the police follows the orders of the smugglers, rather than enforcing the law. EIA and Telapak (2001, p.€21) conclude, “at the core of the issue of illegal logging is corruption”. Robbins (2000) introduces a theoretical framework for the analysis of corruption and then uses it to study the enforcement of protection for a natural reserve in Rajasthan, India. Robbins finds that the lack of enforcement is fuelled by corruption among foresters and that it leads to substantial habitat destruction. Finally, Rosenbaum (2005), after maintaining that corruption promotes illegal logging and trade, presents a list of lessons for reforming the sector which are drawn from Transparency International’s toolbox for fighting corruption. Transparency International promotes national integrity systems that are a series of interventions at the national level for decreasing corruption. Rosenbaum borrows some tools from the national integrity system for the forest sector in order to create a list of interventions for reducing corruption in the forest sector. These interventions include creating a knowledge centre, creating a forest sector corruption index, promoting awareness campaigns, encouraging forest certification, etc. Pellegrini (2011) discusses corruption in the forestry sector of Honduras and Nicaragua, finding that the implementation of centralized systems of enforcement is flawed because of corruption. Furthermore, the escalating stringency of forestry policy (evident with the logging ban in Nicaragua) and the toughening of the approach with respect to enforcement (as with the involvement of the army in both Nicaragua and Honduras) did not solve the either problems of corruption or those of enforcement. On the contrary, the process of military involvement in Nicaragua and Honduras is associated with anecdotal evidence of increasing corruption. In Honduras, in a paradoxical twist, several informants (including forest officials) claimed that the enforcement agents tend to be tougher on legal loggers rather than illegal ones, because often the legal loggers are reluctant to pay bribes.
126
7 The Rule of the Jungle in Pakistan
7.2.1.2 Evidence on the Extent of Corruption in Pakistan The public administration of Pakistan is plagued by corruption. Even though hard evidence of corruption’s incidence is difficult to obtain, surveys, news reports, case studies, historical accounts, and information collected during our study all indicate that corruption is pervasive throughout the public administration. Islam, in his study of the Pakistani public service, states that “Corruption in Pakistan is a way of life” (Islam 2004, p.€314). It is the National Accountability Bureau, a governmental agency in charge of drafting the National Anti-Corruption Strategy, that states, “Corruption has become a disease infecting every aspect of political, social and economic activity” (Hafiez 2002, p.€1). In order to be able to better place the forest sector in its institutional environment, in this sub-section we will focus on the evidence of corruption in the agencies most directly involved in the implementation and enforcement of forest policy and regulations: the judiciary, the police, and the Forest Department. The extent of corruption in Pakistani’s enforcement agencies is evident from the results of an extensive household survey conducted by Transparency International (2002). During the survey, 3,000 households from urban and semi-urban areas of Pakistan were interviewed. A high share of respondents declared that at least one member of the household had to deal with the police or the judiciary in the previous year (32% and 6%, respectively). Most strikingly, all of the interviewees who had had contact with the police reported corruption (that is, 960 respondents out of the 960 who had to deal with the police out of the total sample of 3,000 interviewees). According to the survey results, corruption—while in contact with the police—took several forms, including outright extortion perpetrated by police officers and payments required to be released from arbitrary imprisonment. Among those who dealt with the judiciary, 96% reported corruption. In most cases court employees and judges demanded bribes, but corruption in the law sector is so widespread that it was not uncommon to bribe also the opponent’s lawyer and the public persecutor. Corruption’s incidence, in the police department and in the judiciary, is all the more important to understand the effect of corruption on forest management. In fact, the success of enforcement tasks delegated to the Forest Department depends also on the collaboration of these enforcement agencies as they play an important role in cases where there is violence or complaints are taken to the courts. Remarkable examples of the difficult relationship between these government branches include Forest Department officials having to bribe in kind (i.e. with confiscated wood) judges in order to avoid the immediate release of wood smugglers. Essentially forest officials entered a bidding process and had to outdo timber smugglers. Reportedly in one case substantial bribes were necessary to convince judges to refrain from accepting a smaller bribe and set free a timber smuggler that—when caught red-handed—had physically attacked some forest officials. As for the Forest Department itself, anecdotal evidence of corruption abounds. During the fieldwork several interviews (with members of the local communities, with NGOs, but also with members of the Forest Department) provided evidence of corruption. Also Steimann (2004), in his detailed analysis of the structure of
7.2 Institutions and Forest Management
127
the Forest Department, finds that forest guards levy fines in a entirely arbitrary fashion, and in most case they result in a bribe. Moreover, it appears that the lower ranks of the Forest Department are completely oblivious of the Forest Ordinance (that should regulate many of their responsibilities, including fining) and they are not aware of the formal regime of fines (let alone being willing or able to properly enforce them). The incentive structure faced by the employees of the Forest Department and their basic needs determine largely their likelihood to collude. We must mention, among the factors that determine forest guards’ behaviour, that wages of the lower rank employees of the Forest Department are around or below the survival line for those who have to support a large family. Under these economic circumstances, turning a blind eye on illegal logging and collecting bribes looks like a necessary addition to a meagre salary (for a study of wages levels and corruption see Van Rijckeghem and Weder 2001).4 Another factor influencing the behaviour of enforcement agents is that conflicts in the area turn easily to violence and these conflicts have involved forest guards in the past. Duty-related incidents have resulted in serious injuries and, in a few cases, also the death of forest officials. Conflict makes the cooperation of the police and the judiciary even more essential, and the lack of cooperation weakens the position of the Forest Department in relation to timber smugglers.
7.2.2 Methodological Framework The theoretical framework utilized in this study of corruption and forest management is based on the New Institutional Economics approach. We will look at the way institutions constrain choices of economic agents (North 1990). In the context of formal institutions, corruption represents a weakening of constraints: disbursing bribes facilitates operations that are inconsistent with legal institutions. Institutions, apart from constraining behaviour, define also opportunities. In some circumstances, however, to exploit the prospects offered by the legal system bribes need to be paid (e.g. when locals have to pay inducements in order to extract logs for selfconsumption). In order to understand the failure of formal institutions to influence economic agents’ behaviour, we will focus on transaction costs and more specifically on enforcement costs of property rights.5,6
4╇ Forest guards need wood for heating purposes, but they are not given allowances to purchase it legally. This makes them dependent at least in part on the wood that is being logged illegally. 5╇ For a detailed account of the theoretical framework adopted during the research see Pellegrini and Kruseman (2005). 6╇ Transaction costs can be imputed in a variety of ways to different agents. Bromley (1989) has shown that the distribution of transaction costs influences distributional and environmental outcomes.
128
7 The Rule of the Jungle in Pakistan
Institutional settings affect transaction costs: the costs that occur when a bundle of rights (related to a good or service) is transferred across economic agents (Coase 1960) and one stage of activity ends and another begins (Williamson 1985). The reduction of transaction costs is one of the main purposes of institutions and success or failure of institutions can often be seen in terms of transaction costs (North 1990). When, as in Swat, de facto property regimes do not coincide with legal provisions even illegal costs—such as bribing—are part of the costs of the transaction. Enforcement costs are a sub-category of transaction costs and pertain to the expenses occurred in order to put into effect property rights. Trust built among officials and smugglers through repeated interactions reduce transaction costs of corruption (see the discussion on Putnam’s concept of bonding social capital in Sect.€2.2). Also, the small size and instantaneous nature of bribes that are collected from local people that extract timber to satisfy their basic needs facilitate corruption. These transactions also remind us of the difference between collusive and extortive corruption (see also Sect.€2.2). The former refers to the situation where the person in power allows the briber to gain access to a resource she would not be entitled according to the rules or for a price that is a fraction of the administrative or market price; while the latter refers to the situation where the briber has to pay to gain access to resources he is entitled to. Enforcement costs of de jure institutional settings are a key for understanding the evolution of natural resources management and its likelihood to fail. In the case of the forest, the enforcement of property rights entails guarding it and monitoring extractive activities. The neglect of enforcement costs by the legislature is underpinning the choice of a regime of profligate control by state agents. That is, the choice of a system with prohibitive enforcement costs coupled with the lack of resources to meet these costs unavoidably results in an institutional failure. Different property regimes entail different rights (and duties) for economic agents and have implications for the management of the resource.7 The formal property regime of most of the forest in Swat is state property, with some rights given to local communities and former owners and the enforcement is based on state agencies (primarily the Forest Department). In the rest of the chapter, we assess the working of existent management regime to diagnose the reasons of the failures. Furthermore, we also summarize theoretical and empirical underpinnings of an alternative management system that we explore: common property management. While reviewing the findings of studies completed over the last few decades, Ostrom (2000) summarizes the conditions that increase the probability of successful communal management. These conditions relate both to characteristics of the resource and of the appropriators. With respect to the resource, the likelihood of successful management as a common is enhanced if improvement is feasible, reliable indicators of the condition of the resource are available, the flow of resources is predictable, and forest boundaries are clear. 7╇ For a comprehensive overview of bundles of rights associated with property right regimes see Schlager and Ostrom (1992).
7.3 A Profile of Swat, Pakistan
129
Finally, as mentioned above, the transition to communal management can also be problematic and give way to a situation where the poorest members of the community continue being marginalized. It is important to note that local communities are not homogeneous. Individuals, within the community, can try to extract as much resources as possible, depending on their power (e.g. Garcia-Frapolli 2004) and initiatives encouraging self-organization can easily be hijacked by commanding agents (Rola and Coxhead 2005). Hence, the end results of policies promoting communal management should be carefully monitored (especially at the implementation stage) and they cannot be assumed a success.
7.3 A Profile of Swat, Pakistan For the understanding of dynamics of deforestation in Swat, it is important to describe the physical and human geography of the study area, and the institutional settings characterizing it.
7.3.1 Geography and Ecology Pakistani forest endowment is dwindling. In the decade 1990–2000, the deforestation rate for the whole country was 1.5% per year (Brown and Durst 2003). When we consider that only a minute part of Pakistan is forested (around 3% of the total area), the trend is even more compelling. The North West Frontier Province (one of four that compose Pakistan) contains the largest amount of forest (around 40% of the total, Government of Pakistan 1992). Our case study focused on an area that traditionally was mostly covered by forest and still contains significant tracts of forest, but is also marked by deforestation: the Swat district.8 The Swat valley is named after the river that runs through it and it includes several sub valleys. At the lower end of the valley is the district capital, Mingora, which is also the most important urban agglomerate. From there, moving northwards along the valley, lower plains with agriculture and fruit orchards are the dominant land use. Moving higher up in the valley with increasing altitude, forests, alpine pastures, and perennial glaciers take the place of agricultural activities. The settlements decrease in size going up the valley. Transport infrastructures become scarcer and most of the transport to the villages nearest to the glaciers is on foot. 8╇ The North West Frontier Province has a set up that was instituted during the British colonial rule. The Province is divided in divisions, themselves divided into districts. The Malakand division contains the Malakand, Swat, Dir, and Citral districts. Apart from the districts, The North West Frontier Province also contains Federally Administered Tribal Areas, where federal and provincial government have very limited powers (mostly limited to military incursions related to terrorism and drug activities).
130
7 The Rule of the Jungle in Pakistan
7.3.2 History and Institutions History is at the root of the status quo and it shapes expectations (claims and rights). The history of Swat has a bearing on current land tenure and forest management. Historical developments are at the origin of the right holding of heirs of the most important families allied to the local dictator. In this section, we outline the historical evolution leading to the modern management regime. Swat has a distinct history in relation to Pakistan and the rest of the Indian subcontinent. While the rest of the area was colonized by the British crown, Swat succeeded in conserving its autonomy and, once Pakistan became independent in 1947, it did not fully accede to Pakistan until 1969.9 7.3.2.1 Pre-accession The earlier history of Swat (from the end of the nineteenth century) was marked by the emergence of a political structure characterized by the extensive powers of the autocrat. The dictator, called Wali, was leading a rebellious country and needed the support of allies from the powerful families of Swat in order to defend the Princely State of Swat (as the state was officially called before the accession to Pakistan). Challenges to the state came from external powers (the British and the neighbouring Kalam state) and from internal ones (other local leaders trying to seize power) (Sultan-i-Rome 2005). As a result, the rulers were under continuous threat and they were used to build alliances with wealthy local lords. The history of the Wali is especially important for the management of the forest, because natural resources were valuable assets used for building and keeping alliances, thus their exploitation was crucial to the survival of the regime. The management system compounded harsh punishments to those who illegally encroached on the forest, while allowing favouritism and smuggling when such was in the Wali’s—or his allies’—benefit. The decision power, with respect to logging and forest exploitation, was centralized and rested in the hands of the Wali.10 7.3.2.2 Post-accession The accession to Pakistan of the Princely State of Swat swiftly changed the institutional setting. The national state took under its control the management and exploitation of forests. The law provided only for a minute disbursement of royalties (10% 9╇ For an account of Swat’s resistance to outsiders occupation and interference see Lindholm (1990). 10╇ For a detailed historical account of the autocratic rule of the Wali of Swat, the strictness of the application of the law to the commoners, the importance that forestry had under his rule, and the illegal cutting and favours for his allies (the Khans) see Sultan-i-Rome (2005). For an altogether different version of the Wali’s rule see his autobiography (Barth 1985).
7.3 A Profile of Swat, Pakistan
131
of the total net revenues) to be paid to the legal right holders and established some extraction rights for local communities to meet consumption needs of the residents. Pakistani state’s forest management was, from the very beginning, characterized by a command and control approach but the effective enforcement was beyond the state’s capability.11 The post-accession regime implied that most of the forests were declared reserved or protected, i.e. the management regimes strictly constrained exploitation rights.12 Ownership and management rested in the hands of the state, through the Forest Department. In the reserve forest, no cutting is allowed and members of the local community have limited rights for extracting dry wood and non-timber forest products. In protected forest, the Forest Department is in charge of the management plan and at the times when there are cuttings the Forest Development Corporation is in charge of the cutting of marked timber with shares of the net sales revenues are given to right holders (Steimann 2004). Moreover, a “local quota” of timber is reserved to local residents for their needs and there is an “emergency quota” that can be used under special circumstances, e.g. to rebuild a house after a fire.13 The shares of revenues from wood sales directed to right holders have been increasing over time, from 10% at the time of accession, to 60% nowadays. Those entitled to shares of the royalties of the forest are the landowners (mostly Khans families—the former allies of the Wali) that reside far away from the forest, or local communities, depending on the property situation of the forest at the time it was taken over by the state.14 The state has been unable to implement the management regime outlined above and face the large transaction costs associated with its implementation, and together with the dismantlement of the previous regime of property rights, this has led to an open access regime.15 In addition, the gap between de jure and de facto management of the forest created multiple bases for claiming rights on the forest: the statutory law, custom, and the de facto regime.16 Agents can then refer to competing
11╇ Pakistan was not the only state that tried to control the forest sector without the necessary resources to implement its own rules on the face of high transaction costs. For comparison see Ostrom (1990) on Nepal. 12╇ A third category of forests are the guzara forests. They are those whose property was left to the communities, even though relevant management responsibilities were taken over by the Forest Department. Guzara forest is an institutional setup that characterized almost no forest in our study area. 13╇ For a detailed description of legal status and management provisions in the forest of the whole of the North West Frontier Province see Steimann (2004). 14╇ Property rights over the forest are contested and contention over property rights is a source of endless litigation that is dealt with by Pakistan’s corrupt and inefficient judiciary system. 15╇ Dismantling existing institutions for the management of natural resources without succeeding in creating working alternatives has been a common procedure in many developed and developing countries over the last few decades (Ostrom 1990). 16╇ Customs and de facto regimes can differ because the former borrows legitimacy from the tradition, while the former is simply the current state of affairs.
132
7 The Rule of the Jungle in Pakistan
bases to justify their claims according to their interests (i.e. there is forum shopping) (Meinzen-Dick and Pradhan 2002). The open access regime on the forest implies that landowners and local residents exploiting the forest are only restrained by local power relations. An unabridged gap persists between the regulatory framework and the actual management practices. Encroachments and illegal cuttings continue. Furthermore, legal disputes over the cuttings’ royalties abound and court decisions settling them are delayed (at times for decades), and, in many cases, court decisions are still pending. The operations of the Forest Department have not become more functional to date.17 Corruption is widespread and it affects all the operations of the Forest Department and of the Forest Development Corporation (which is in charge of exploitative operations). Forest guards are paid salaries that cannot satisfy the needs of families (Steimann 2004) and they are lacking the support of the other enforcement agencies (which, as we saw above, are themselves inefficient and plagued by corruption). The overall enforcement system seems more geared to make ends meet for forest officials than to protect the forest. There are numerous checkpoints along the roads going down the Swat Valley, but these serve little purpose apart from extracting bribes in exchange for turning a blind eye on timber smugglers.18
7.4 Alternative Forest Management Regimes in Swat Now we turn to alternative forest management regimes that could exist in Swat, basing our discussion on Sect.€7.2 and on Ostrom (2000) and we will look into the feasibility of improvements on the forest stand, the availability of data on forest quality, the predictability of revenues from the forest, and forest boundaries. Parts of the forest of Swat are so degraded that improvements to its stock are feasible only over a medium to long period of time, but most of the other conditions are met: the state of the forest is directly observable, the regenerative potential is predictable, and communities’ boundaries are mostly undisputed. With respect to the appropriators, the success conditions are: resource saliency for the users, common understanding, low discount rate for the agents, trust and reciprocity in the community, autonomy (e.g. the appropriators are able to rule themselves), and prior organizational experience. These conditions are the ones that are the most difficult to meet in Swat. Adherence to these conditions is heterogeneous: some communities meet most of the conditions, while others are more deficient. Generally, saliency and shared understanding are the conditions that will be met more easily: income in rural communities is very low and resources extracted from the forest We collected information through focus group discussions and interviews during the fieldwork. Different sources, including employees of the Forest Department itself, confirmed the dismal state of the Forest Department operations. 18╇ This was confirmed by numerous interviews and by our own witnessing the ease of movement of truckloads of illegal logs in thorough the valley. 17╇
7.5 Deforestation in Swat, Pakistan
133
make up a large share of revenues (Rafi Khan et€al. 2006), the forest standing has a predictable evolution and communities’ members are aware of its functioning. Discount rates are not particularly low because incomes are meagre and there is migration. Trust, reciprocity, and previous organizational experience are contingent on each community, on its social cohesion and its history. Autonomy is absent in the present situation, but it can be achieved in case of institutional change towards communal management. Overall, the conditions mentioned above affect the individual agents’ cost benefit calculations of engaging in communal management (Ostrom 2000). Policy interventions that promote communal management should include instruments to promote these conditions in order to enhance the likelihood of success. It is worth reminding the reader that these conditions are not a list of necessary and sufficient conditions (Ostrom 2006). There is also an important role for individuals in the setting up of commons’ management mechanisms. The uncoordinated activities of individual members of local communities are partially unpredictable and we cannot be entirely certain of the outcome of their interaction. Furthermore, once we consider self-organization in a longer temporal framework, it is important to emphasize that social systems are dynamic and evolve over time, therefore though the local community is not self-organizing at one point in time it can develop in that direction at a later stage. Learning can make a community able to copy other communities’ solutions, or to adapt them to their own socio-environmental conditions. Summing up, the list of conditions mentioned above can be considered as a benchmark for estimating the likelihood to achieve successful coordination in a common over the short period, or objectives to be achieved as an intermediate step towards effective communal management.
7.5 Deforestation in Swat, Pakistan Deforestation in Swat is taking place because of institutional failure. Formal rules, if applied, could protect the forests, but de facto alternative institutions and lack of effective monitoring and implementations prevent them from being effective.19 At the level of economic agents, deforestation is caused by the desire to convert land for agriculture and/or for the extraction of the valuable logs (Bromley 1999). In the case of Swat, forestry per se can be a driver of deforestation since it is done with a clear-cut logging. In this section, we analyse how the most important incentives interact with the institutional framework determining the status quo. The process of deforestation is different in different sites according to differences in socio-economic and ecological conditions, and also local institutional settings are diverse in the case study area. The same formal institutions produce different environmental outcomes and causes different social agreements to arise depending In our study, we do not investigate the issue of underlying and proximate causes of deforestation. For an in depth discussion see Bromley (1999) and Contreras-Hermosilla (2000).
19╇
134
7 The Rule of the Jungle in Pakistan
on local conditions. Under current conditions, sustainable management of the forest is achieved only rarely. It is worth stressing that, as we will see, not only illegal logging, but also those cases of effective forest protection operations contradict the formal regulatory framework.
7.5.1 I ncentives and Institutions over the Forests of Swat: Focus on Corruption The occurrence of deforestation depends on local physical and social conditions. Incentives and institutions interact with geography because of geography’s effects on transaction costs. Below we provide a classification of institutional regimes that are in place in Swat. At the bottom of the valley, in the most fertile area, land conversion for agriculture has historically been the main cause of deforestation. Over time, forested areas have been cleared to make space for agricultural production and fruit orchards. Deforestation in the area, which is historically important, has lost significance. At the bottom of the valley, land conversion is almost complete and there are no large stretches of standing forest left. Property rights in this part of the valley are well established, especially since easy access diminishes enforcement costs. Historical patterns determined landownership to be concentrated in the hands of the Khans families (see Sect.€7.3, above). Agricultural activities at the bottom of the valley are the most valuable of the area and they are the most damaged by the ongoing deforestation taking place uphill. During our interviews, several stakeholders noted microclimatic change (especially increases of hail precipitation) and problems from the watershed of the Swat River (floods and erosion) that damage agricultural activities and linked them to the deforestation process. Landowners (the Khans) are responsible themselves for converting forest into agricultural land in the lower and more fertile parts of the valley. In the mid and upper valley, a lack of enforcement of property rights on forested areas determined a de facto open access situation in most of the forest. The state and the right holders—the former landowners20 (mostly the Khans residing at the bottom of the valley)—are formally entitled to the ownership and to shares of revenues arising from forest cuttings, but lack of enforcement allows members of local communities and timber smugglers to consider the forest as an open access resource and to exploit it accordingly. Formally, the Forest Department (representing the provincial government) and the landowners (that reside at the bottom of the valley) have control over the forest. Under the formal management scheme a quota of wood is reserved for the local community for its needs, and the net revenues that accrue when there are cuttings are shared among the right holders (i.e. the former The rightholders are the former landowners, because with the accession of Swat to Pakistan the formal property of forests passed to the state and the former owner enjoyed the right to a share of the revenues from logs extraction.
20╇
7.5 Deforestation in Swat, Pakistan
135
owners of the forest) and the Forest Department. The formal institutional setting, which remains unenforced, is not substituted by an alternative regime restricting access to the forest. The Forest Department and the Forest Development Corporation, at times when markings and cuttings are allowed, collude with loggers and allow more wood to be extracted than the amount planned. This illegal extraction damages right holders and the local community, depriving the former of royalties, of self-consumption opportunities the latter, and of environmental services both.21 Illegal logging takes place also when there is no marking and organized logging: locals log more than they are allowed because their quota are insufficient, or because of the lengthy process involved in obtaining their share of the local quota, or they extract logs for commercial purposes (mostly in connection with the timber mafia, whose members are reported to be the hard-to-trace string pullers). When timber is logged and transported, the chances of running into a control from forest officials is high and the unauthorized logging is punished with a fine or a bribe.22 The meagre results of the implementation of the logging ban in a weak institutional environment are predictable and in line with evidence from other countries (see Pellegrini 2009). In summary, the area is characterized by a failure to organize forms of control by the state or by the local community. In the mid-valley, the local community is not involved in monitoring and regulation of the exploitation activities. One of the factors contributing to this situation is the lack both of trust and of social capital in the communities.23 Competing stakeholders exploit the forest in an unsustainable manner. Every agent would benefit from a collective management of the forest, but cannot reap the benefits of its restraint. The dominant strategy is thus to harvest as much as possible resulting in a logging race typical of the so-called tragedy of the commons (Hardin 1968). Thus, neither the management practices of the Forest Department, nor the management practices of the local community point towards sustainable extraction. The elimination of trees from the steep slopes causes soil erosion and slides. As a result, problems of land degradation are evident in the whole area. Members of the local community, while depending on forest products Sources of information on illegal operations are: crosschecked confidential interviews with people belonging to different stakeholders groups (i.e. Forest Department, law scholars, members of the elders’ council), group discussions, and household surveys. 22╇ The regulatory framework that should guide the operations of forest officials is so confused that different officials interpret their duties differently. As a result, often the difference between a fine and a bribe becomes fuzzy. For example, when interviewed, different forest officials answer differently on whether fines should be paid on the spot, or fines should be reported first, and then paid at the office of the range officer; also the amount of the fine to be levied varies markedly depending on the respondent and it is unclear what should be done with confiscated timber (reportedly some forest officers keep it in order to smuggle it, or for their own benefit). For a detailed exposition, see Steimann (2004). 23╇ During our meetings, members of the community repeatedly asserted that they were not interested in the management of the forest, that they wanted these tasks to be supplied by the Forest Department and that all they want are shares of the revenues when marking and extraction takes place. 21╇
136
7 The Rule of the Jungle in Pakistan
to satisfy their household needs, contribute to deforestation since, without a credible coordination mechanism and a fair property regime, they would not appropriate any value of the natural resource in any case. At the high-end of the valley, next to the permanent glaciers, there is de facto communal management, where the local community has decided to entirely protect the forest stands. The de facto and the de jure situation are completely apart, but in this case, it is for the benefit of the forest. The forest plays a crucial role for the survival of the community: it impedes tracts of the glacier from sliding on the village. Furthermore, contributing to the successful management of the forest as a common is the fact that the villagers do not feel threatened in their control of the forest. Even though the forest in their village is nominally property of the state and under the management of the Forest Department, forest guards have not visited that part of the valley for more than a decade. The community has decided to follow an easy to enforce—even though probably not optimal—management strategy. Felling of trees is completely prohibited and the only wood that can be collected is from the trees that fell naturally, for example, because of the movements of the glacier. This simple rule guarantees that the forested area does not decrease in size and density, that the household needs are satisfied (because of the relative large size of the forest with respect to the local population), and that no conflict arises within the community (thanks to the ease of enforcement). Physically, a limited amount of sustainable wood extraction would be possible. This could lead to an inflow of much needed cash for the locals, but the costs and the risks of a more complex institutional arrangement may outweigh the benefits. In summary, in the Swat district, the idiosyncrasies between different regulatory frameworks create a situation of legal pluralism: there is a plethora of different rules that are contending in the governance of transactions. Each agent turns to the regulatory framework that is more convenient for her and neglect others’ claims (Meinzen-Dick and Pradhan 2002). In most of the area rather than the official management regime, centred on the control of the Forest Department, an open access situation is in place and corruption is pervasive as the enforcement agents abuse their powers for their own benefit within a weak institutional environment.
7.6 Strategies for Institutional Change Considering institutional arrangements as fixed and simply working from a “crime and punishment” perspective (Becker 1968) the most obvious solution to the enforcement problem would be to alter the cost-benefit outcomes for forest officials when entering into corruption deals and of loggers harvesting illegally. When applied to corruption in the Forest Department, this would imply increasing the penalties for employees of the forest service caught being involved in corruption as well as augmenting monitoring in order to identify more corruption cases. With respect to illegal loggers, this approach requires higher fines and increasing monitoring in the collection and transport of logs.
7.6 Strategies for Institutional Change
137
The successful application of the crime and punishment approach relies on the existence of enforcement agencies that operate efficiently and honestly. Recommendations based on this approach must take into account the institutional settings where they would be applied. However, in Pakistan the highly corrupted police and judiciary cannot patrol effectively the Forest Department. In turn the corrupt employees of the Forest Department cannot effectively monitor the behaviour of other agents. Furthermore, examples of sustainable management of the forest exist in the area, but they are not based on the state’s control of the forest, nor on better enforcement (or harsher rule) by the Forest Department. On the contrary, the only successful cases of sustainable management are based on collective action of local people and they exclude the official enforcement agency (i.e. the Forest Department) from the process. In light of these findings, in this section we will outline the ongoing reform process, underlining its idiosyncrasies and setbacks. Furthermore, we will suggest how the reform process could take a new spurt reinforcing collective action where it is already in place, and facilitating it where it is not happening.
7.6.1 The Reform Process of the Early 2000s The unsustainability of de facto management practices and the environmental, social, and economic shortcomings of current deforestation rates in most of the North West Frontier Province are apparent and cause discontent to be voiced in numerous ways (e.g. the press often laments the management regime and its effects). The social and environmental failures associated with the current management practices have motivated some reform attempts of the forestry sector. For the Swat district the most relevant reform effort was the Forest Sector Strategy, whose inception was supported by a loan of the Asian Development Bank and by technical assistance provided by the Dutch Royal Embassy. The reform process was conceived as a decentralization and participation program, but evolved into a “crime and punishment” one. The main task of this donor-driven program was to move the management of the forest from a state of command and control towards co-management. The donors hoped to strengthen the local capabilities and to move the emphasis of the Forest Department from punishment towards partnership with local communities. The program was not successful. Local committees were established in several villages, so that they could be an interlocutor of the Forest Department and, most importantly, contribute to the village land use plan.24 However, at the local level the most influential members of the community controlled the process as the committees over-represent the wealthy. Moreover, the Forest Department has been unwilling to give up part of its powers and enter into a real partnership with the locals. The committees are the Village Development Committee (composed of men) and the Women Organization (composed of women).
24╇
138
7 The Rule of the Jungle in Pakistan
Eventually, due to implementation slowness and especially because of the resistance at the provincial level to undertake the necessary steps to support the reform in due time, the support of donor agencies was withdrawn in 2004. This was a major setback, especially in the few places where the reform started to yield benefits. Now, trust has been lost in this type of interventions (see Suleri 2002).25 The Forest Ordinance of 2002 confirmed that the reform process was not proceeding smoothly. The inception (and the first draft) of the ordinance was in line with current thinking in the international development community and focused on decentralization and participation coupled by a pro-poor approach (see Pellegrini, 2011). It was geared towards co-management and partnership in a stable regulatory framework. However, under the insistence of more conservative powers in the Forest Department and the provincial government, the final version of the ordinance was very different from what expected. Once the nature of the ordinance became known, international donors along with local NGOs that from the beginning had been supportive of the reform process (such as Sungi, a Pakistani grass root NGO working for the empowerment of the most marginal members of the society), eventually rejected the Forest Ordinance. Of special importance in determining Sungi’s stance were the punitive powers that the ordinance gave to the Forest Department and that could be used against local communities. Since the donors’ support for the reform process was stopped, the whole process halted and has reversed. The latest initiatives strengthen the Forest Department control of the forest. In accordance to the 2002 Forest Ordinance, the Forest Force has been instituted.26 Now a 500 man-strong armed force will patrol the forest of the North West Frontier Province, including Swat. The Forest Department itself estimates that at least 3,000 would be required in order to patrol effectively the forest of the province.27 It appears that the insufficient number of men involved in the force will be in the future an excuse for another failure of the “crime and punishment” approach applied to illegal logging. It is also worth noting that if the existing force would be effective much less than 500 armed men would be necessary to control extraction activities in the area. Few roads run along the valley and functioning checkpoints should be sufficient to stop a large part of the smuggling. Nevertheless, the guards manning the checkpoints are colluding with the smugglers, and arming the very same guards will not change their incentive structure. Similarly, armed guards patrolling the forest, even more removed from detection possibilities, will easily become involved in illegal deals. Arming forest officials may not yield any environmental, economic, or social dividends.
The final result is similar to what Holling (2000) describes as a common pathology: “the ecological system loses resilience, the industries become dependent and inflexible, the management agencies become rigid and myopic, and the public loses trust in governance”. 26╇ See http://www.ttjonline.com/story.asp?scâ•›=â•›35057. Accessed 1 Jan 2005. 27╇ The expected cost of equipping such a force would be 60€million€Pakistani€Rupees (or around 825,000€EUR). See http://www.dawn.com/2004/06/16/local23.htm. 25╇
7.6 Strategies for Institutional Change
139
7.6.2 Institutional Reforms The “crime and punishment” approach that has been tried and failed in Pakistan is not the only approach put forward to deal with corruption. Another mainstream approach to reform is what we call here the “holistic approach”: a countrywide program of institutional change. For example, Transparency International—one of the most prominent proponents of this approach—called for a series of reforms (that altogether are labelled “National Integrity Systems”) that implemented together should reduce corruption throughout the Pakistani society. In the case of Pakistan, completing the checklist of the National Integrity System (including independent judiciary, free press, establishing an effective watchdog authority, etc.), would imply altering deeply the institutional framework of the country and the implementation of these reforms remains blocked by opposing powerful interests (e.g. Yusuf 2002). While the objective of decreasing corruption occurrence in the whole of Pakistan remains valid, the ecological and social problems caused by deforestation require viable initiatives to be adopted at a quick pace in the forestry sector.28 Moreover, obtaining improvements and reducing corruption in some spearhead sectors—such as the forestry—can help to create a momentum for wider initiatives against corruption.29 Since, as discussed above, corruption is widespread in Pakistan, the officials from the Forest Department and the Forest Development Corporation operate within a system where effective collaboration of complementary enforcement agencies is infeasible. Then institutional change should happen, given the still existing impetus for the reform of the forestry sector, mostly within the forestry sector. The forestry sector is characterized by pervasive powers given to state agents and chronic corruption. Institutional reform should reduce the state’s involvement in planning, management, and enforcement of forest policies. As mentioned above, institutional arrangements that do not coincide with the formal provisions can halt deforestation. The challenge is then to differentiate reform efforts between locations where present-day institutional arrangements prevent deforestation, and areas where deforestation is still rampant, or reforestation is needed because of the complete depletion of forests. Where property rights are well established (i.e. the bottom of the valley and the most marginal areas) the reform process should be very careful with altering the existing situation as it could lead to open access conditions. Where private property is undisputed, a system of market-based incentives would be supportive for the For an analysis of how social action coalitions can produce and sustain anti corruption efforts see Johnston and Kpundeh (2005). 29╇ The National Integrity System (NIS) has also been applied specifically to the forest sector (Rosenbaum 2005), but its application in Pakistan seems doubtful. Some of the most important tools of NIS involve the partnership of commercial forest interests and the public sector (e.g. forest integrity pacts and business principles for the forest sector). In Pakistan commercial logging is formally inexistent (because of the ban), therefore there is no legal basis for a partnership between the commercial sector and the state. 28╇
140
7 The Rule of the Jungle in Pakistan
internalization of the positive externalities arising from the forest. As we explained above, at the bottom of the valley, the avoidance of deforestation is not an issue as those areas have been already entirely cleared. Where communal efforts to protect the forest are in place—and deforestation is not occurring—a strengthening of the current institutional settings should take place, aligning the de jure to the existing situation. In this way the long-term management perspective of communities can be improved through stability. Where institutions are not well defined, the task of reform is more difficult. The impediments to the self-organization of the collective must be removed and social learning must take place, which can be a lengthy process. The broader institutional environment, especially the Forest Department, should operate as a facilitator in the process (Ostrom et€al. 1999). At the same time, it is important to recognize that local communities can also fail to manage the natural resource in an equitable and sustainable manner. Differentiation within communities, because of wealth, power, and/or ethnicity, can mean that the most influential members of the community can hijack the devolution process and the collective action process, turning it into their favour (Suleri 2002; Rola and Coxhead 2005). Social trust and self-organization capability are required and, if they are not present, time and good examples will be required to build them. Hence, special attention should be devoted to monitoring and implementation. Summing up, when thinking about institutional change for the management of the forest in order to overcome failures of the present system, it is important to accommodate local ecological conditions and social capabilities. Thus, institutional set ups cannot be uniform, but needs built-in flexibility and accurate monitoring of achievements and failures.
7.6.3 Paths of Change In the context of institutional change, a particular emphasis should be given to the building of coalitions that can support change. After highlighting the needed steps for change, in this section we outline possible patterns to achieve change, through an analysis of the incentives of the most important stakeholders. Finally, we will outline some of the auxiliary reforms and federal-level initiatives that could facilitate the reform process and the achievement of the final objective of sustainable forest management. 7.6.3.1 Actors and Incentives The main actors with stakes in the forestry sector are the federal and provincial governments, the Forest Department, the Forest Development Corporation, local communities (differentiated among themselves mostly because of right-holding issues), NGOs, and international donors. All these actors have some qualified interest
7.6 Strategies for Institutional Change
141
in forest conservation. In addition, there are the interests of those involved in illegal logging, the smugglers, the wood processing industry, and the economic agents involved in economic activities induced by timber trade. Also thanks to corruption, these actors can spread their influence and they are able to influence some of the other actors as well. Federal and provincial government have important stakes in forestry. The federal government is ultimately in the position to decide the overall legal framework and it should be able to allow for the internalization of the externalities created by the forest. The provincial government can also influence the legal framework (within the boundaries set by the federal government) and is the recipient of a share of the royalties produced when legal logging takes place. The situation is further complicated because there are idiosyncrasies between the government’s mandate and the interest of agents acting in the name of the government. From the perspective of the government as a whole, the conservation of forest resources would have the advantage of increasing welfare levels, ease the pressure on the issue exercised by mass media and political discontent, and attract donors. Nevertheless, from confidential interviews with high-rank members of the Forest Department and of NGOs, it was revealed that important figures from the government (mostly at the provincial level) have stakes in illegal logging. If change is offered as a package, the main losers could get a compensation for their losses. This compensation, that should be equitable, would ease opposition and the actual change would be more likely to happen. We identify as the main losers from the achievement of sustainable management the local notables who are right-holders and would lose part of these rights if management would be moved to the local community. Another potential loser is the Forest Department which would suffer in terms of power, routines and, for certain officials, relevant shares of their income (now provided by kickbacks). Last, we will look at those directly involved in the illegal timber trade. Especially because opposition from all these stakeholders would make any reform unattainable, compensation mechanisms and economic incentives could facilitate the reform process. Forest Department and Forest Development Corporation have as stated main objectives the management and protection of the forest cover, sustainable extraction, and fair repartition of royalties. Apart from their official missions, it is apparent that employees of these organizations have developed routines that are not oriented towards conservation of the forest, but towards appropriation of shares of revenues that are derived from illegal logging. Increasing the meagre salaries and training of the lower ranks of the Forest Department are preconditions for having the state officials working in the forestry sector playing a positive role in the field. Moreover, the Forest Department as a whole should be more involved in the reform process.30
30╇ At least the forest guards should be aware of it, on the contrary of what happened with previous efforts (as we have seen above). Going a step further, and actually involving the Forest Department in the reform process would give to its members a sense of ownership and the necessary information to increase compliance.
142
7 The Rule of the Jungle in Pakistan
Formal arrangements give landowners rights to part of the royalties from the net sale proceeds from log extraction operations. De facto, they obtain only a marginal share of what they are entitled to. Still, a move of the institutional framework towards co-management would further limit their ability to appropriate royalties from log extraction. If the opposition of such influential members of the community would stop the reform process, as seems likely, a scheme for compensation should be crafted. For example, incentives for the conservation and the re-afforestation in the lower part of the valley (where, as mentioned above, private property rights are well established) could also be included as a compensation for these losses. The fact that landowners are not getting much from their de jure rights and the fact that their agricultural operations would benefit of local positive externalities from increased forest cover, should contribute to their acceptance of reform. NGOs are willing to and should be partners in the whole reform process. Because of their interest in forest resources’ conservation and the information and experience that they have available, they are valuable partners of any reform process. For many NGOs, this type of activity is functional to the achievements of their goals and their involvement can increase transparency and checks on the implementation of reform programmes.31 International donors have already shown their interest and willingness to assist and facilitate the reform process in the forestry sector in Pakistan. Unfortunately, the lack of success in previous reform efforts may undermine future engagement. Still, international donors are active in Pakistan and projects involving the management of natural resources offer them the opportunity of benefits with respect to environmental and poverty reduction objectives. Furthermore, donors’ could exploit their experience with regulatory and implementation obstacles to reform that they met in the past. They could, for example, make their support conditional on ex-ante legal reforms and focus on their implementation (see also Suleri 2002). The additional funds that derive from donors’ involvement could be used for the activities to reduce resistance to reform and to guarantee equitable outcomes. Members of the local community that, to different degrees, are involved in illegal logging will suffer a setback in their economic opportunities. Very often, these members of the community are disadvantaged and in dire economic conditions, hence the reform package should be complemented with alternative forms of sustainable livelihoods. In other words, the intervention towards the protection of the natural resource should contain a social package to prevent its implementation and its social outcome to be regressive. It can be argued that the previous interventions of reform for the forestry sector were incomplete from this point of view. The “Timber Mafia” bases its operations on the high rents that can be captured because of the high profitability of illegal trading in wood. A strategy for reducing the potential rents from illegal extraction would be to liberalize timber trade, i.e. reducing prohibitive tariffs and red tape for imported wood, undercutting substantially the size of rents to be appropriated via illegal logging. Consequently, the level of resources available locally to facilitate illegal logging will be reduced. Given the 31╇
Such NGOs are, for example, IUCN, Sungi, or the WWF.
7.6 Strategies for Institutional Change
143
fact that it is difficult to devise policies to diffuse opposition from those who are gaining the most from illegal logging, their opposition should be taken into account and strategies to overcome it should be devised. 7.6.3.2 Auxiliary Initiatives and Federal-Level Reforms Finally, there are some initiatives that would facilitate the achievement of environmental and social targets, and some reforms that are necessary conditions for instituting sustainable management practices for the forest. In this sub-section, we provide an overview of the most important interventions that would facilitate the achievement of sustainable logging, but many other interventions related to institutional quality and environmental management in general could be mentioned (e.g. reforms to improve governance such as the ones proposed by Transparency International, population planning, land reform, etc). A key reform and a test of the will of the federal government to protect the Pakistani forest is the abolition of all trade barriers on imported wood.32 The domestic demand for wood cannot be met by the modest forest stands as it was recognized already in the Forestry Sector Master Plan (Government of Pakistan 1992). Allowing imports would simply be the logical consequence of government’s efforts to protect the forest. Interventions oriented towards the provision of sustainable livelihoods of local communities should focus both on those who use wood for self-consumption and those who extract it for commercial purposes. For the locals involved in logging, social programs providing alternative sources of income should be set up. For all households timber use could be reduced with the use of more efficient stoves and similar technical solutions (e.g. promoting the construction of micro-hydels).33 Payments for environmental services can be another way of financing the conservation of forests. Since Swat’s forests provide economic benefits (e.g. watershed protection of agricultural activities and hydropower generation), a quantification of the major benefits could lead to the implementation of a scheme where members of upstream local communities are paid in order to preserve the forest. In economic terms, the positive externality will be internalized. Some room could be available for the crime and punishment approach, when the state agencies in charge of enforcement are reliable and they are not likely to get involved in corruption. The one body of Pakistan’s security forces that has been praised for its honesty is the motorway police (Transparency International 2003). Opening the Pakistani borders to timber trade would imply increasing pressure on international markets and, possibly, also increasing illegally harvested timber in neighbouring countries. While we recognize issues related to unsustainable and illegal logging in worlds markets, reforms and interventions at the global level (which we strongly support) are beyond the scope of this study. 33╇ Pakistani and International NGOs operating in Pakistan have experiences with such programs and could be instrumental in implementing them within the forestry sector reform (e.g. see World Health Organization 2005). 32╇
144
7 The Rule of the Jungle in Pakistan
This force could enforce restrictions on the movement of logs on the main roads of Pakistan. Even though the motorway police can be considered a reliable agency, it is worth noticing that the courts that would judge complaints arising from the operations of this force do not share such a remarkable score on corruption.
7.7 Conclusions Our analysis shows that corruption is contributing to the failure of the management regime of Pakistani forest and highlights the shortcomings of the standard reform approaches to corruption. The most invoked reform options with respect to corruption are the “crime and punishment approach” and the “holistic approach”. The “crime and punishment” approach requires sound enforcement institutions, which are not available in Pakistan. In our study, we saw that the insistence of Pakistani legislators on the former approach is at the root of present failures. Furthermore, the sweeping reforms required to introduce transparency in the Pakistan administration, the “holistic approach”, are at odds with embedded routines and vested interests. These comprehensive reforms remain a valuable social objective, but cannot provide timely solutions for the issues analysed here. Corruption problems, when the overall institutional environment is weak, are best addressed by local institutional change. In a sector and in a region that are marked by low administrative capabilities to enforce state regulations, reforms should be aiming at reducing the coercive role of state agencies. Therefore, options where responsibilities in forest management are (at least partially) devolved to local communities should be at the centre stage of reform efforts. While the forest of Swat cannot wait for countrywide reforms that would solve the problem of corruption at all levels, they could benefit from decentralized reform efforts. At the same time, experiments at the local level might also inform and facilitate reforms at the national level. In contrast with what suggested here, the latest developments in the management of forest in Pakistan are increasing the enforcement effort and powers of state agencies. We find that the emphasis on state enforcement agencies is misplaced and is reproducing the conditions of present failures. A carefully crafted reform program should include the strengthening of property rights, provisions for communal management, and market-based incentives while taking into consideration the social, economic, and ecological characteristics of the different areas. These objectives should be endorsed by coalitions of actors empowered to enforce these propositions. State agencies should act as facilitators of the process rather than attempt to exert complete control over the forest. Ultimately, the relevant question is whether it is possible to save the remaining forest of Swat through reforms in the forest sector. The arguments presented in this chapter provide support for the existence of viable alternatives to unsustainable logging, but significant changes in the way the problem is dealt with by Pakistani authorities have to take place in order to achieve them.
References
145
Finally, we do not want to fall into simplistic and deterministic generalizations of anticorruption strategies: devolving powers and resources from the state to local actors is not going to generate automatic benefits. These benefits do not arise simply because the powers of government officials are being curbed (see the discussion in Chap.€2). Nor can we assume that indigenous communities are always homogenous groups that are going to achieve inclusive agreements resulting in benign policy outcomes (e.g. Véron et€al. 2006). Rather, we argue that these policies offer an opportunity for change and improved outcomes, but there is no guarantee that they will produce universally beneficial outcomes.
Appendix: Sources of Information in the Study This study is a part of the Poverty Reduction and Environmental Management (PREM) research program (http://www.prem-online.org/). During our study, we collected information from secondary sources, from individual interviews, from group discussions and from a household survey, we held during the fieldwork. The interviewees were informed agents: forest officials, a conservator, international donors, national and international NGOs, lawyers, students of history, councils of the elders, and landowners. Interviews were in some cases individual and in others in the form of focused group discussions. Further information was collected from 400 household surveys in the Swat district and from meetings in villages. Household information on economic and social conditions, as well as income dependence on natural resources was gathered. Village socio-economic profiles were drawn up after the visits. The fieldwork took place between January and August 2005. A summary of the findings and the methodology applied in the project is available in Rafi Khan et€al. 2006 and an outline of the conceptual framework used in the analysis is available in Pellegrini and Kruseman 2005.
References Asian Development Bank. (2000). Country Assistance Plan (2001–2003): Pakistan. Manila: Asian Development Bank. Barth, F. (1985). The last Wali of Swat. Bangkok: White Orchid. Becker, G. (1968). Crime and punishment—economic approach. Journal of Political Economy, 76(2), 169–217. Bromley, D. W. (1989). Economic interests and institutions: The conceptual foundations of public policy. Oxford: Blackwell. Bromley, D. W. (1999). Deforestation—institutional causes and solutions. In M. Palo & J. Usivouri (Eds.), World forests, society and environment 1. Amsterdam: Kluwer. Brown, C., & Durst, P. B. (2003). State of forestry in Asia and the Pacific–2003: Status, changes and trends. Bangkok: FAO. Callister, D. J. (1999). Corrupt and illegal activities in the forestry sector: Current understandings, and implications for World Bank Forest Policy. Forest policy implementation review and strategy development: Analytical studies. Washington: World Bank.
146
7 The Rule of the Jungle in Pakistan
Chomitz, K. M., & Buys, P. (2007). At loggerheads? Agricultural expansion, poverty reduction, and environment in the tropical forests. World Bank policy research report. Washington: World Bank. Coase, R. H. (1960). The problem of social cost. Journal of Law & Economics, 3(Oct), 1–44. Contreras-Hermosilla, A. (2000). The underlying causes of forest decline. Occasional paper. Jakarta: Center for International Forestry Research (CIFOR). Environmental Investigation Agency and Telapak. (2001). Timber trafficking: Illegal logging in Indonesia, South East Asia and international consumption of illegally sourced timber. London: EIA Telapak. Garcia-Frapolli, E. (2004). Decreeing the land as protected: The Outside-inside Perspective of Community-based Biodiversity Conservation in the Yucatan Peninsula. Montreal: ISEE Conference. Government of Pakistan. (1992). Forestry sector master plan. Islamabad: Government of Pakistan. Hafiez, M. (2002). National anti-corruption strategy. Islamabad: Government of Pakistan, National Accountability Bureau. Hardin, G. (1968). Tragedy of the commons. Science, 162(3859), 1243–1248. Holling, C. S. (2000). Theories for sustainable futures. Conservation Ecology, 4(2) (Art.€7). Islam, N. (2004). Sitarish, sycophants, power and collectivism: Administrative culture in Pakistan. International Review of Administrative Sciences, 70(2), 311–330. Johnston, M., & Kpundeh, S. J. (2005). Building a clean machine: Anti-corruption coalitions and sustainable reform. Policy Research Working Paper Series. Washington: The World Bank. econ.worldbank.org/files/41032_wps3466.pdf. Kaufmann, D., Kraay, A., & Mastruzzi, M. (2005). Governance matters IV: Governance indicators for 1996–2004. World Bank Policy Research Working Paper Series. Washington: World Bank. Lindholm, C. (1990). Leadership categories and social processes in Islam: The cases of Dir and Swat in Pakistan: The social sciences’ perspective. A. S. Ahmed. Karachi: Oxford University Press. Meinzen-Dick, R., & Pradhan, R. (2002). Legal pluralism and dynamic property rights. CAPRi Working Paper, no.€22. Washington: International Food Policy Research Institute (IFPRI). Myrdal, G. (1968). Asian drama: An inquiry into the poverty of nations. New York: Twentieth Century Fund. North, D. C. (1990). Institutions, institutional change, and economic performance. Cambridge: Cambridge University Press. Ostrom, E. (1990). Governing the commons: The evolution of institutions for collective action. New York: Cambridge University Press. Ostrom, E. (2000). Reformulating the commons. Schweizerische Zeitschrift für Politikwissenschaft, 6(1), 29–52. Ostrom, E. (2006). Monitoring makes a difference: Formal property rights to forests are ineffective unless officials or users invest in them. Kyoto: 3rd World Congress of Environmental and Resource Economists. Ostrom, E., & Gardner, R. (1993). Coping with asymmetries in the commons—self-governing irrigation systems can work. Journal of Economic Perspectives, 7(4), 93–112. Ostrom, E., Burger, J., Field, C. B., et€al. (1999). Revisiting the commons: Local lessons, global challenges. Science, 284, 278–282. Pellegrini, L. (2007). The rule of the jungle in Pakistan: A case study on corruption and forest management in Swat. FEEM Working Paper. Milano: FEEM. Pellegrini, L. (2009). Poverty reduction and forest management in Bolivia, Honduras and Nicaragua: Reform failures? The Hague: Institute for Social Studies. Pellegrini, L. (2011). “Forest management and poverty in Bolivia, Honduras and Nicaragua: Reform failures?” European Journal of Development Research. Pellegrini, L., & Kruseman, G. (2005). A conceptual framework for the study of institutions and institutional change in Pakistani forests management. Lisbon: European Society for Ecological Economics Conference.
References
147
Perry, P. J. (1997). Political corruption: The last geographical taboo? Comment. Political Geography, 16(3), 187–188. Rafi Khan, S., Yusuf, M., Haq, L. U., et€al. (2006). Assessing the poverty-environment nexus: Evidence from Swat. Amsterdam: Poverty Reduction and Environmental Management Program (PREM). Rao, A. L., & Marwat, A. H. (2003). Northern areas strategy for sustainable development: Background paper. Gilgit: IUCN Pakistan. Robbins, P. (2000). The rotten institution: Corruption in natural resource management. Political Geography, 19(4), 423–443. Rola, A. C., & Coxhead, I. (2005). Economic development and environmental policy in the uplands of Southeast Asia: Challenges for policy and institutional development. In D. Colman & N. Vink (Eds.), Reshaping agriculture’s contribution to society: Proceedings of the TwentyFifth International Conference of Agricultural Economists (pp.€243–256). Malden: Blackwell. Rosenbaum, K. L. (2005). Tools for civil society action to reduce forest corruption: Drawing lessons from transparency international. Washington: The World Bank. Schlager, E., & Ostrom, E. (1992). Property-rights regimes and natural-resources—a conceptual analysis. Land Economics, 68(3), 249–262. Steimann, B. (2004). Decentralization and participation in the forestry sector of NWFP, Pakistan— the role of the state. IP6 Working Paper. Swiss National Centre for Competence in Research. Islamabad. Suleri, A. Q. (2002). Regional study on forest policy and institutional reforms: Final report of the Pakistan case study. Manila: Asian Development Bank. Sultan-i-Rome (2005). Forestry in the princely state of Swat and Kalam (North-West Pakistan). IP6 Working Paper. Islamabad: Swiss National Centre for Competence in Research. Transparency International. (2002). Corruption in South Asia: Insights & benchmarks from citizen feedback surveys in five countries. Berlin: Transparency International. http://www.transparency.org/pressreleases_archive/2002/dnld/south_asia_report.pdf. Transparency International. (2003). National integrity systems: Pakistan 2003. TI country study report. Berlin: Transparency International. Transparency International. (2004). Corruption perceptions index 2004. Berlin: Transparency International. http://www.transparency.org/pressreleases_archive/2004/2004.10.20.cpi.en.html. Van Rijckeghem, C., & Weder, B. (2001). Bureaucratic corruption and the rate of temptation: Do wages in the civil service affect corruption, and by how much? Journal of Development Economics, 65(2), 307–331. Véron, R., Williams, G., Corbridge, S., et€al. (2006). Decentralized corruption or corrupt decentralization? Community monitoring of poverty-alleviation schemes in Eastern India. World Development, 34(11), 1922–1941. Williamson, O. E. (1985). The economic institutions of capitalism: Firms, markets, relational contracting. New York: Free Press, London: Macmillan. World Health Organization. (2005). Situation analysis of household energy use and indoor air pollution in Pakistan. Geneva: World Health Organization. Yusuf, Z. (2002). The powerlessness of the media in Pakistan. In Taking action against corruption in Asia and the Pacific: Conference Papers and Proceedings. Manila: Asian Development Bank.
Chapter 8
Conclusions and Future Research
The overall aim of the book is to contribute to our understanding of the influence of corruption on paths of sustainable development. At different stages, the study shows how corruption hinders economic growth and diminishes the stringency of environmental protection—two fundamental components of sustainable development. Furthermore, the study includes the related issue of the determinants of corruption levels and highlights the difficulties in dealing with corruption in weak institutional settings through a case study.
8.1 Main Conclusions The overall aim of the book is to contribute to our understanding of the influence of corruption on paths of sustainable development. At different stages, the study shows how corruption hinders economic growth and diminishes the stringency of environmental protection—two fundamental components of sustainable development. Furthermore, the study includes the related issue of the determinants of corruption levels and highlights the difficulties in dealing with corruption in weak institutional settings through a case study. Chapter€2 introduces the subject of corruption and engages with different ways to define it. First, a critical analysis of policies that derive from and depend on specific characterizations of corruption is provided. Together with a classification of different types of corruption and related concepts we discuss alternative definitions of corruption and highlight how definitions that are limited to the public sector cannot shed light on some phenomena as understood by the general public. We critically analyse the related notion that corruption is always linked to regulation and state intervention and that there is a trade-off between marked failures and corruption. Furthermore, we also reflected on the existence of economic studies which have focused on the negative impacts of corruption that are not related exclusively to the public sector. One of the main findings of this book is that corruption impedes the enactment and implementation of regulation in the environmental sphere. The trade-off between market failures and corruption is fictitious and based on oneL. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9_8, ©Â€Springer Science+Business Media B.V. 2011
149
150
8 Conclusions and Future Research
sided definitions of corruption and on a misconceived understanding of the relationship between corruption and regulation. Chapter€3 produces a cross-country analysis of the determinants of corruption perception. Thanks to improved data availability and the use of comprehensive econometric modelling, some central findings of previous econometric studies are questioned. The disheartening findings that historical patterns decisively determine present day corruption levels are mitigated. Specifically, links to a British past (because of common law adoption, or because of a past as a colonized country), and the degree of decentralization are questioned. After income levels are controlled for, ethnolinguistic fractionalization’s negative effect on corruption (which political economy links with more divided societies) becomes insignificant. Cultural theories of the determinants of corruption are confirmed (the presence of Protestants is associated with lower levels of perceived corruption). Higher-income levels are correlated to lower corruption and the inclusion of income levels lowers size and statistical significance of the coefficients of the other explanatory variables. As corruption is also found to be one of the determinants of income growth rates (see Chap.€4) caution is needed in interpreting the results. Thus, in the medium-to-long run it is corruption that affects income levels (via its cumulated effect on growth rates). Once the effect of corruption on income levels is discounted, the inclusion of income then serves as a test for the effect of the other independent variables. Nevertheless, the coefficient of the income variable is not reliable because of causality issues. The analysis also shows that other country features, which are policy amenable in the medium run, are associated with lower levels of corruption. It finds that a medium-long exposure (30 years) to uninterrupted moderate-to-high levels of democracy is associated with lower corruption. The frequency of changes in veto powers (a measure of the time horizon of political actors) is positively associated with corruption. Finally, some tentative evidence is provided on the negative correlation of newspaper readership, and higher wages in the public sector (when compared to average GDP) with respect to corruption. Chapter€4 presents an econometric analysis of the effect of corruption on economic growth. Direct and indirect transmission channels are considered. The indirect channels included in the analysis are: investment, schooling, trade openness, and political violence. The results highlight that the influence of corruption over GDP growth rates goes mostly through the transmission channels and that investment, education, and trade openness appear to bear most of the effect. Overall, the results suggest a strong negative influence of corruption on growth. Chapter€ 5 provides an empirical test of theories of the determinants of environmental policy stringency. The analysis finds that the corruption variable is a fundamental determinant of the level of environmental protection. The inclusion of an index of perceived corruption and a measure of procedural democracy finds that the effect of corruption dominates. As it was shown in Chap.€3, a medium-tolong exposure to democracy is associated with lower corruption, thus in the longer term democracy could still (positively) affect environmental protection through its mitigating effect on corruption. Furthermore, because of the high degree of correlation between democracy and corruption, multicollinearity could have inflated the
8.2 Policy Conclusions
151
standard error of democracy’s coefficient (i.e. it can have decreased the statistical significance of the coefficient). Our findings confirm the direct and large effect of corruption on the stringency of environmental policies and that they are robust to the inclusion of a proxy for democracy in the analysis. A similar direct effect of procedural democracy is questioned. Chapter€6 applies the analysis of environmental policy determinants to the case of the accession process to the EU. New EU members are characterized by markedly lower levels of environmental protection and—among the harmonization initiatives they have undergone in the accession process—they rapidly adopted (at least formally) measures to protect the environment. The study analyses the effects of the incidence of corruption in these countries and argues that high corruption levels in the new EU members—when compared to the old ones—provides a rationale for the harmonization process of environmental regulations. Chapter€ 7 takes a case study approach to highlight some facts of corruption in environmental sphere that are obscured by more macro-quantitative data. We looked at the issue of corruption and forest management analysing anti-corruption strategies in a case study setting: the enforcement of forestry policies in Swat, Pakistan. We find that increasing penalties and monitoring (the “crime and punishment approach”) might be counterproductive in weak institutional settings. Also the other popular approach—country-wide institutional overhauls (the “holistic approach”)—might be difficult to achieve in the short run and we turn to institutional change to a smaller scale as a potentially effective way to deal with corruption.
8.2 Policy Conclusions Different understandings of corruption have implications for the study of the effects of corruption and also for the analysis of ways to cope with corruption. Not all definitions in the literature, among policy makers, and in the public are coherent and mixing up different understandings might imply that to deal with the problem created by corruption—as defined in one sense—we actually deal with corruption defined in another sense. In particular, Chap.€2 delves in the differences between definitions of corruption referring only to the public sector and comprehensive definitions referring to public and private sector. It highlights how in the study of the effects of corruption often private corruption is included, and how policy responses such as privatization might have actually little to do with corruption defined comprehensively. While privatization might help to reduce corruption in certain cases, the remaining interface between the public and the private sector—that characterize most privatizations of crucial services such as health, transportation, etc.—offer ample opportunities for corruption to continue and even take more perverse forms. This result also highlights the limits of the commonly held conclusion that the solution to the problem of corruption is to simply to reduce corruption opportunities (decrease the space of regulatory power of the state, etc.). The absence of policies (as well as their lack of effectiveness) can be a result of corruption, rather than a
152
8 Conclusions and Future Research
solution to corruption problems. The advice to reduce state intervention remains valid when regulations do not serve any purpose apart from creating red tape and help bureaucrats to extract rents. However, when valuable purposes are being served by state intervention, we should be careful about advocating less state involvement and rather advocate a shift to policies and regulations that allow fewer opportunities for corruption. Moreover, the dilemma between market imperfections and corruption appears also to be misleading. While any type of regulation will be—most likely—associated with some degree of bureaucratic corruption, one of the main conclusions of this study is that corruption is one of the determinants of lack of environmental protection stringency. Hence turning the corruption versus regulations reasoning upside down, it appears that corruption can very well drive the lack of regulations resulting in environmental problems (and market failures in general), rather than be a necessary evil associated with policy responses. Economists have been largely concerned with economic incentives to be corrupt, while other social scientists have focused on cultural norms (Rose-Ackerman 1999); the findings of Chap.€ 3 provide ammunition to the arguments that some cultural norms can also influence corruption (e.g. those related to the Protestant religion). This, in turn, implies that cultural factors should not be downplayed when thinking about ways to deal with corruption.1 Corruption levels show little variation over time and strategies to reduce them require political will and can be lengthy. An examination of the transmission channels through which corruption affects economic growth can help the search for policies to ease the effects of corruption on the economy in the shorter run. For example, the knowledge that corruption in given countries affects some policies (such as investment in education), will suggest interventions to deal with that shortfall. Similarly, when we analyse environmental policies in countries characterized by a high degree of corruption we can expect to find them lax. Knowing the likely impacts of corruption on these policies can help to deal with these shortcomings. While the effects of corruption on economic growth and environmental protection can imply pessimistic perspectives for the welfare of developing countries, they suggest that the renewed emphasis on corruption in the academic and policy community could produce a double benefit: speeding economic growth and increasing the stringency of environmental protection. One of the outcomes of international efforts to tackle corruption is the 2003 UN Convention Against Corruption. For the convention to have real effect proper emphasis needs to be given to its implementation and to other agreements that aim to deter corruption should follow (e.g. limiting bank secrecy). Whereas the Environmental Kuznets Curve (EKC) has already been challenged in the literature, our findings cast further doubts on the ability of the EKC to predict the relationship between economic welfare and environmental quality. Developing countries are characterized by levels of corruption higher than developed ones and 1╇ One way to consider culture could be to include it in a standard incentive framework, where certain cultural features would be an incentive (disincentive) to engage in corruption. For examples of an economic study that explicitly considers how culture and informal norms affect corruptibility see Lambsdorff et€al. (2005).
8.3 Future Research
153
our findings suggest that corruption has an influence on environmental policy stringency. Therefore, even if the EKC were a correct characterization of the relation between pollutants and income for current rich countries (and we already have quite some evidence to the contrary), the possibility to replicate this development and generalize it as a norm that will be followed by all countries in the future is questioned. Our results put forward a new interpretation of the action of international bodies, such as the EU, that are pressing countries to tighten their environmental regulations. Countries where corruption is widespread may have less stringent environmental standards because of the increased influence of powerful lobbies on policy makers and because of the difficulty in the implementation of environmental regulations, rather than because of the preferences of their citizens. The pressure of the EU on accessing countries to adopt stringent environmental regulations may contribute to overcoming domestic institutional failures for the benefit of the citizens of these countries. On a different note, the pervasive effect of corruption on environmental policies might also suggest a shift in environmental regulations promoted by international bodies. In supporting regulation and other policy measures the focus should not be on the formal adherence to international standards, but on more enforceable measures. For example, rather than simply asking a government to ratify an agreement, the country could be asked to make progress towards environmental objectives and the citizens should be entitled to take legal action in case the government not complying. Finally, we looked at strategies to counteract corruption through a case study. The holistic approach (such as the toolbox proposed by Transparency International) can be used within certain contexts to hold institutions accountable and to indicate directions for improvement at the national and local levels. The lack of an effective national framework for dealing with corruption cannot become an excuse for avoiding action at the local level for specific and urgent issues. It is important to be mindful that decentralization is not a silver bullet against corruption (e.g. Jeffrey 2002) and legal pluralism is also a potential problem when decentralization is not carried out coherently. Nevertheless, we see that from the development and diffusion of approaches that favour decentralized management of natural resources one of the benefits is the possibility of creating new, or reinforce existing, institutional forms that might be less affected by corruption when compared to centralized agencies in pervasively corrupt countries. Clearly, reform at the local level should not be seen as an alternative to larger reforms and a two-pronged approach could be applied: country-wide institutional reform together with locally adapted institutional reforms to deal with corruption.
8.3 Future Research The economic literature on corruption is expanding in many directions. Here we mention some relatively unexplored but important issues that relate to both the causes and effects of corruption. When corruption indexes are geographically projected (as in Fig.€3.1), geographical patterns become apparent: neighbouring coun-
154
8 Conclusions and Future Research
tries tend to be affected by similar levels of corruption and border sharing countries are classified by similar ranking in corruption. Research questions arise on the estimation of these spatial correlations and their quantification (that can be analysed with the statistical tools associated with Geographical Information Systems (GIS)). Once these geographical patterns are tested statistically, theories (especially those of culture and geography) and further empirical analyses should provide an explanation. The collection of corruption data and the analysis of panel data can also serve the purpose of identifying determinants of corruption, the effects of corruption, and to provide statistical tests of precedence (e.g. through Granger causality testing). Data limitations have impaired such analysis so far, as the data sources with the standard definition of corruption and large samples date back only 10 years. Once these data sets are extended over time it would be possible to analyse country-fixed effects in order to identify the factors that are variant over time and that are associated with differences in corruption levels. The availability of data sets at firm level has allowed for investigating the determinants of corruption at the micro level. While previous analyses have focused on cross-sections of countries, the micro-econometric evidence has the advantage of allowing the investigation of variations in corruption incidence within the same macro environment. These analyses can complement the knowledge of dynamics of corruption that existing studies have produced and serve the purpose of evaluating the effectiveness of policies that operate at the sub-country level (e.g. Reinikka and Svensson 2006; Seligson 2006; Bertrand et€al. 2007). A related area of enquiry is the analyses of corruption incidence and socio-economic performance of countries at the regional level. Focusing research on different regions of one country, rather than on cross-sections of countries, has the drawback that the variability of corruption is decreased. However, such analysis has the advantage that the influence of unobserved variations are likely to be reduced. The Asian growth miracle represents a puzzle for researchers of corruption. It would appear that corruption is not having its hindering effects on the economic performance of Asian countries: some of them experienced growth miracles, but are also characterized by high levels of perceived corruption (e.g. China, India, and Thailand). While this is far from an unqualified truth and some of the countries that achieved higher-income levels in the region are ranked among the least corrupt in international comparison (e.g. Singapore), questions arise on how to reconcile the evidence of the effects of corruption, collected on a global scale, with the performance of Asian countries. Moreover, in-depth analysis of differences in income growth (and levels) in Asia and their relation with corruption also deserves further scrutiny. One example of this type of analysis—as mentioned above—is the focus on differences in corruption among regions of the same Asian countries that could help to identify detrimental effects of corruption within economies. These effects might be clouded by the focus on countries as a whole. Studies that include the use of instrumental variables for the research on institutions and growth has become quite common, especially after the influential papers from Hall and Jones (1999) (who used geographical variables and language as in-
8.3 Future Research
155
struments for institutions and policies) and Acemoglu et€al. (2001) (who used mortality rates of colonizers as instruments for current institutional quality of former colonies). Some authors have argued that good instrumental variables should just have some desired statistical properties and that instruments do not make a causal theory (Rodrik 2004), disputing that through the use of instrumental variables it is possible to discern the origins of development. The fact that a variety of instruments have been used for different explanatory variables lending support to many alternative hypotheses confirms the need of an assessment of the use of instrumental variables and the interpretation of results produced through them. There are also a number of analytical issues that deserve attention, but that might not allow for the application of econometric techniques. The importance of institutions for economic growth has received an increasing recognition leading to the booming interest in the methods of New Institutional Economics. These methods— focusing on transactions cost and on the constraints limiting the choice sets of economic agents—can be utilized for the study of corruption (Lambsdorff et€al. 2004). Corruption can be seen as a way to evade constraints imposed on economic agents, or as an additional cost incurred in order to enjoy one’s right. The role of transaction costs in affecting the working of market institutions and the need for policy makers to factor them has been recognized long ago (Coase 1960); however, the influence of corruption on the distribution and levels of transaction costs has been largely ignored. Studies on corruption and the environment range from cross-country empirical analyses to theoretical contributions, to reports whose main objective is to expose individuals responsible of corruption in some specific cases. The analysis of corruption and its influence on the design and implementation of environmental policies deserves more attention. The tendency of some institutional arrangements to be more conducive to corruption under certain conditions can inform the choice of institutions to regulate environmental resources exploitation. For example, the design of private, versus state, versus communal management of natural resources, should include information on corruption, especially in the case of countries that do not have strong institutions. Most anti-corruption strategies that emerge from the economics literature have a deductive genealogy: scholars take stock of existing or develop theoretical tools and derive conclusions about the effectiveness of anti-corruption strategies. Often times, this process is capable of producing novel insights (e.g. Lambsdorff et€al. 2005), but these deductive efforts could be combined with retroductive ones. In particular, we would find a study of successful anti-corruption strategies (at country and at micro level) as particularly interesting as it could shed light on the mechanisms and conditions that favour reductions in corruption. Unfortunately, the number of countries that in the last few decades have managed a transition from high levels of corruption to low levels is quite limited (e.g. Hong Kong, Italy in the first half of the 1990s, Singapore), but these experiences could be analysed in a comparative way to investigate regularities in these transitions. Furthermore, specific anti-corruption campaigns that have resulted in reduced corruption could be studies in detail to find the institutional conditions and changes that accompanied these success stories.
156
8 Conclusions and Future Research
Overall, this book investigated the influence of corruption on economic development and environmental protection. While there is a burgeoning literature on these subjects, further research is needed to disclose the factors contributing to the spread of corruption, to investigate the impact of corruption on societal welfare, and to provide inputs to inform public policies to effectively deal with these issues.
References Acemoglu, D., Johnson, S. & Robinson, J. A. (2001). The colonial origins of comparative development: An empirical investigation. American Economic Review, 91(5), 1369–1401. Bertrand, M., Djankov, S., Hanna, R., et€al. (2007). Obtaining a driver’s license in India: An experimental approach to studying corruption. Quarterly Journal of Economics, 122(4), 1639–1676. Coase, R. H. (1960). The problem of social cost. Journal of Law & Economics, 3(Oct), 1–44. Hall, R. E., & Jones C. I. (1999). Why do some countries produce so much more output per worker than others? Quarterly Journal of Economics, 114(1), 83–116. Jeffrey, C. (2002). Caste, class, and clientelism: A political economy of everyday corruption in rural North India. Economic Geography, 78(1), 21–41. Lambsdorff, J. G., Taube, M., & Schramm, M. (2004). The new institutional economics of corruption: Norms, trust, and reciprocity. London: Routledge. Lambsdorff, J. G., Taube, M., & Schramm, M. (2005). The new institutional economics of corruption: Norms, trust, and reciprocity. London: Routledge. Reinikka, R., & Svensson, J. (2006). Using micro-surveys to measure and explain corruption. World Development, 34(2), 359–370. Rodrik, D. (2004). Getting institutions right. CESifo DICE Report no.€2, pp.€10–15. Rose-Ackerman, S. (1999). Corruption and government: Causes, consequences, and reform. Cambridge: Cambridge University Press. Seligson, M. A. (2006). The measurement and impact of corruption victimization: Survey evidence from Latin America. World Development, 34(2), 381–404.
Author Index
Acemoglu D., 6, 14, 23, 24, 36, 56, 57, 155 Alam M. S., 4 Andersen M. S., 101, 103, 116 Arikan G. G., 2, 32, 43 Banerjee A.-V., 6 Bardhan P., 4, 17, 55, 63 Barro R., 43, 73, 93, 96, 117 Barth F., 130 Becker G., 37, 121, 123, 136 Berglund C., 103 Bertrand M., 154 Bhagwati J. N., 4 Black B. S., 22 Bohara A. K., 36 Bollini P., 55 Boycko M., 22, 54, 55, 61 Bromley D. W., 122, 133 Brown C., 122, 129 Brunetti A., 37 Carter T. S., 79 Chafuen A. A., 32, 36 Chong A., 41 Chowdhury S. K., 2, 32, 36, 43, 44 Coase R., 21, 128, 155 Cole M. A., 9, 93 Congleton R. D., 78 Dahl R. A., 83 Damania R., 3, 6, 76, 80, 81, 105 de Bruyn S. M., 93 Della Porta D., 54, 55 Dryzek J. S., 79 Easterly W., 56 Ehrlich P. R., 76, 77 Elazar D. J., 43 Eliste P., 84, 96, 106 Estache A., 22 Fisman R., 32, 42, 47 Frankel J.-A., 56
Fredriksson P. G., 3, 69, 76, 78, 80, 84, 88, 96, 106 Gerring J., 24 Glaeser E. L., 32, 34, 40 Groenendijk N., 4 Grossman G. M., 9, 92, 93, 105 Gupta S., 3, 54, 55 Hafiez M., 126 Haigh N., 103 Hall R. E., 42, 154 Harbaugh W. T., 78 Hardin G., 77, 135 Hasnas J., 15 Hauge W., 83 Heilbroner R. L., 76, 77 Hill K. Q., 36, 43, 87 Hodgson G. M., 14, 19, 23 Huntington S. P., 3, 4, 14, 46, 54 Islam N., 126 Jaggers K., 83 Jain A. K., 25, 54, 56 Jeffrey C., 153 Johnston M., 47, 139 Knack S., 41, 48, 81 Krueger A. O., 3, 9, 55, 62, 92, 93, 105 Kugler M., 15 La Porta R., 2, 32–34, 40, 48, 49 Lambsdorff J. G., 2, 5, 15, 24, 107, 155 Landes D., 40 Landman T., 84 Leff N. H., 54 Levine R., 56, 57 Lindholm C., 130 Lopez R., 76, 79, 105 Lui F. T., 4, 54 Macaulay T. B., 34 Marshall A., 20 Mauro P., 3, 8, 21, 35, 54–56, 62, 64, 65, 69, 93
L. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9, ©Â€Springer Science+Business Media B.V. 2011
157
158 McCloskey H. J., 76, 78 McGuire M. C., 78 Mill J. S., 75 Miller W. L., 47 Mo P. H., 3, 56, 63, 93 Morse S., 84 Munck G. L., 83 Murphy K. M., 3, 54, 55 Myrdal G., 2, 124 Neumayer E., 78, 83, 93, 105 North D. C., 54, 55, 57, 127, 128 Nye J. S., 26 Ostrom E., 123, 124, 128, 131–133, 140 Paldam M., 81 Paolo C. F., 23 Papyrakis E., 56 Payne R. A., 76, 78, 105 Pellegrini L., 6, 56, 72, 104–109, 125, 135, 138, 145 Perkins R., 92 Perry P. J., 124 Philp M., 17 Plumper T., 70 Putnam R. D., 18, 77 Rafi Khan S., 122, 123, 133, 145 Reinikka R., 47, 154 Robbins P., 6, 79, 125 Roca J., 105 Rodriguez F., 58, 70, 88 Rodrik D., 57, 155 Rose-Ackerman S., 35, 152 Rosenblum M.-R., 84 Ross M. L., 92 Roy E. V., 4, 17
Author Index Sachs J., 22, 42, 56–58, 62, 63, 72 Sandholtz W., 43 Schiavo-Campo S., 49 Schlager E., 128 Schulze G.-G., 37, 47 Seligson M. A., 18, 154 Sen A., 82 Shleifer A., 18, 20, 32, 34, 40 Southgate D., 62 Spangenberg J. H., 105 Stern D. I., 93 Stevens P., 41 Sung H.-E., 32, 82 Svensson J., 47, 48, 69, 80, 154 Tamara L. J., 101 Tefertiller K. R., 103 Temple J., 58 Torras M., 78 Treisman D., 2, 32–38, 40, 42–44, 49, 81 van Beers C., 77 Van Rijckeghem C., 45, 127 Vanhanen T., 83, 88, 96 Véron R., 145 Weale A., 103 Wei S. J., 21, 61 Weidner H., 105 Welsch H., 80 Williams A., 24, 25 Williams R., 2, 15, 24 Williamson O. E., 128 Wilson J. K., 81 Wooldridge J. M., 58
Subject Index
A Accession, 10, 102, 111, 113, 115, 130, 131, 151 B Bribery, 5, 16–18, 30 British colonies, 34, 38, 40 Bureaucratic corruption, 4, 6, 19, 116, 152 C Collusive corruption, 20 Contemporary determinants of corruption, 33 Corruption Perception Index, 21, 41, 45, 48, 57, 58, 82, 83, 85, 95, 96, 107, 108, 112, 113, 117 Corruption with theft, 18 Corruption without theft, 19 Crime and punishment, 10, 121, 123, 136–139, 143, 144, 151 D Decentralization, 32, 39, 42, 43, 47–49, 79, 107, 137, 138, 150, 153 Definition of corruption, 1, 2, 7, 8, 13–16, 21, 25, 26, 154 Democracy, 9, 22, 29, 32, 36, 39, 42–44, 47–49, 58, 61, 70, 71, 73, 75–79, 81–83, 85–96, 105, 150, 151 E Enforcement, 10, 79, 80, 84, 106, 113, 116, 121, 123–128, 131, 132, 134, 136, 137, 139, 143, 144, 151 Environmental Kuznets Curve (EKC), 75, 78, 92, 152
Environmental Policy Stringency, 7, 9, 75, 76, 79, 81, 84, 86–92, 94–96, 104, 105, 109, 110, 150, 153 Environmental Policy Stringency Index, 86, 88, 89, 95, 96 Environmental Regulatory Regime Index, 83–85, 89, 91, 95, 96, 102, 103, 106–109, 116 Ethnolinguistic fractionalization, 29, 35, 38–40, 47, 48, 150 Extortive corruption, 128 F Forest Department, 123, 125–128, 131, 132, 134–141 Forest management, 7, 10, 79, 121, 122, 124, 126, 127, 130–132, 140, 144, 151 Forestry, 6, 7, 79, 123–125, 133, 137, 143, 151 Freedom House, 54 Functionalist theory, 4 G Government intervention, 24, 39, 44, 48, 49 H Historical determinants of corruption, 33, 57 Holistic approach, 121, 139, 144, 151, 153 I Income, 2, 4, 9, 10, 29, 32, 35, 37, 39, 41–43, 45–48, 55–61, 63–66, 68, 69, 75, 79, 80, 85, 87–95, 101, 103–105,107–110, 113–117, 132, 133, 141, 143, 145, 150, 153, 154 Investment, 3, 5, 9, 21, 22, 30, 37, 53–67, 70–72, 103, 150, 152
L. Pellegrini, Corruption, Development and the Environment, DOI 10.1007/978-94-007-0599-9, ©Â€Springer Science+Business Media B.V. 2011
159
Subject Index
160 J Judiciary, 15, 20, 22, 107, 123, 125–127, 137, 139 L Legal theories, 34 Long-term effect of corruption, 9, 65–67 M Market failure, 6, 13, 14, 23–26, 149, 152 Measurement of Corruption, 2, 13, 24 N Natural resources, 3, 6, 10, 32, 35, 41, 76, 84, 96, 116, 121, 123, 124, 128, 130, 131, 142, 145, 153, 155 Newspapers Circulation, 42, 45, 48, 49 O Omitted variables, 32, 81, 106 P Police, 79, 123, 125–127, 137, 143, 144 Political corruption, 4, 6, 19, 22, 24 Political turnover, 29, 36, 39, 42, 44, 47–49 Political violence, 9, 53, 54, 56–67, 70–72, 150 Principal-agent theory, 4
Private sector, 3, 4, 13, 19–24, 26, 55, 151 Privatization, 13, 20–23, 151 Protestant religion, 34, 40, 152 Public sector, 3, 5, 7, 8, 13–15, 18–24, 26, 44, 47, 149–151 R Rent seeking, 3, 6, 35, 41, 62 S Schooling, 9, 35, 53, 54, 56–68, 70–73, 86, 87, 89–91, 96, 108–110, 113, 117, 150 Social capital, 18, 128, 135 Swat, 7, 10, 121–123, 128–130, 132–134, 136–138, 143–145, 151 T Trade policy, 53, 54, 65 Transparency International, 1, 2, 16, 20, 21, 30, 37, 45, 48, 54, 58, 72, 82, 96, 107, 113, 117, 123, 125, 126, 139, 143, 153 U Urbanization, 35, 86–89, 91, 96, 108–110, 113, 117, 118 W Wages, 37, 44, 47, 127, 150
