Description
A focus of the subject is how economic agents behave or interact and how economies work.
ABSTRACT
Title:
ESSAYS IN COMPARATIVE INSTITUTIONAL ECONOMICS Peter Grajzl, Doctor of Philosophy, 2005
Directed By:
Professor Peter Murrell Department of Economics
This dissertation examines how decentralized institutional structures and organizational forms evolve and affect economic development under different politicoeconomic and legal arrangements. Organized legal professions are typically viewed by economists as rent-seeking interest groups – even though they have been central in institutional development in countries with the highest quality institutions. Chapter 1 develops a model that identifies the link between the role of organized legal professions and the quality of reform. Delaying institutional reform through deliberation, the profession's participation discounts the expected benefit from welfare-inferior reform proposal for rent-seeking interest groups. Professional review serves as a screening mechanism ameliorating the self-interested government's adverse selection problem. The model's predictions cast new light on the Glorious and the French revolutions, post-communist transition, why and when civil law and common law systems differ, and why post-independence institutions are of higher quality in settler than in extractive colonies. Although common, self-regulation as an alternative to direct government regulation has been little investigated. Chapter 2 uses a framework inspired by property rights
theory to address the allocation of regulatory authority. In a model of a regulatory process with bargaining, the authority to amend the enabling legislation can be either consolidated within the government, or extended to the producers in a self-regulatory regime. The chapter delineates the welfare implications of regulatory regime choice, and indicates whether the government's incentives to delegate or centralize regulatory authority lead to efficient institutional design. The model identifies those features of legal traditions that help to explain variation in regulatory arrangements across countries, illuminates the contrast in regulatory practice between the progressive era and the associational regime of the New Deal, and characterizes the mechanisms of intervention used in fascist economies. Chapter 3 discusses the channels through which civil society is expected to affect economic development. Utilizing the formal analysis laid out in Chapter 1, the chapter provides an introductory examination of the rationale for civil society aid and concludes with a conjectural interpretation of the determinants of the aid's effectiveness to bring about successful institutional change in post-communist countries.
ESSAYS IN COMPARATIVE INSTITUTIONAL ECONOMICS
By Peter Grajzl
Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2005
Advisory Committee: Professor Peter Murrell, Chair Professor John Wallis Professor Rachel Kranton Professor Martin Loeb Assistant Professor Rodrigo Soares
© Copyright by Peter Grajzl 2005
Preface
Institutions consist of formal rules and informal constraints structuring political, economic and social interaction (North 1990, 1991). Traditions, customs, moral values, religious beliefs and other informal social norms affecting economic development (e.g. Greif 1994, Barro and McCleary 2003) originate within societies themselves. In contrast, formal rules such as constitutions, statutory laws and government regulations are designed and enforced in a centralized manner by the state. An important part of the institutional infrastructure of societies is determined by a state-led top-down approach. However, elements of institutional structure other than social norms and culture nevertheless evolve in a decentralized manner, enabling a society to, at least partially, 'govern itself'. This dissertation examines how examples of decentralized institutional structures and organizational forms evolve and contribute to economic development under different politico-economic and legal arrangements. Chapter 1 examines the role of the organized legal professions' review of proposed legislation. Chapter 2 explores self-regulation as a general phenomenon of deliberate delegation of the state's law-making powers. Chapter 3 provides a preliminary investigation of civil society's contribution to institution building. Chapters 1 through 3 analyze different instances of institutional and organizational decentralization. The individual chapters' common methodological feature, however, is the focus on identification of the socially efficient institutional structure within the predetermined set of feasible institutional arrangements in a given politico-economic and legal setting, and prediction when the identified socially efficient
ii
institution is incentive-compatible with the non-benevolent government's agenda. That is, while organized legal professions' review of proposed legislation, cross-industry selfregulation, and civil society involvement may one by one be socially efficient, their implementation need not be desired by the ruling politicians.1 The ensuing paragraphs provide a brief motivation for each of the chapters as they stand as separate parts integrated under the common theme of institutional and organizational decentralization, in turn emphasizing each individual chapter's main contributions to the literature on comparative analysis of institutions and elucidating the key features of each chapter's analytical framework. Like much of a growing body of literature on comparative analysis of institutional arrangements that has raised, or freshly illuminated, many important questions in the last decade (see Djankov et al. 2003 for discussion), the arguments proposed in this dissertation rely on reduced form models, motivated by disparate historical examples and available data. Economists typically view organized legal professions as rent-seeking interest groups – even though they have been central in institutional development in countries with the highest quality institutions. The model developed in Chapter 1 identifies the link between the role of organized legal professions and the quality of policy reform. By delaying institutional reform through its deliberation, the profession's participation discounts the expected benefit from welfare-inferior reform proposals for rent-seeking interest groups. Professional review serves as a screening mechanism ameliorating the self-interested government's adverse selection problem. The model predicts how the role
1
As a consequence, an important question – the answer to which is beyond the scope of this dissertation – that may arise under each individual case of decentralization under investigation is how to align the government's incentives with the society's if the government's agenda subverts social efficiency.
iii
of organized legal professions varies with democracy, political stability and professional power. The analysis in Chapter 1 casts new light on, and identifies common elements of, such apparently widely different events as post-communist transition and the Glorious Revolution in Britain and the French revolution. The chapter also extends the recent discussion on the importance of legal and colonial origins for the quality of contemporary institutions (see e.g. LaPorta et al. 1997, Rajan and Zingales 2003, Acemoglu et al. 2001). The analysis suggests why and when civil and common law differ, and why postindependence institutions are of higher quality in settler than in extractive colonies. The political power of organized legal professions examined in Chapter 1 is manifested through their autonomy from the state ? their ability to self-regulate. Although self-regulation is a common phenomenon it has rarely been investigated as an alternative to direct government regulation. Chapter 2 uses a framework inspired by Grossman and Hart's (1986) property rights theory to address the allocation of regulatory authority. When regulatory contracts are incomplete, the distribution of regulatory rights critically shapes regulatory outcomes. The chapter develops a model of a regulatory process with bargaining, where the authority to amend the enabling legislation can be either consolidated within the government or extended to the producers in a selfregulatory regime. The model delineates the welfare implications of regulatory regime choice, and indicates whether the government’s incentives to delegate or centralize regulatory authority lead to efficient institutional design. Chapter 2 complements chapter 1 by suggesting another channel through which legal origin matters (see e.g. Beck at al. 2003): the analysis in Chapter 2 identifies those
iv
features of legal traditions that help to explain variation in regulatory arrangements across countries. The chapter also attempts to characterize the mechanisms of intervention used in fascist economies, and illuminate the contrast in regulatory practice between the progressive era and the associational regime of the New Deal, as put forward by Eisner (2000). Political independence and self-regulatory powers of organized legal professions investigated in Chapters 1 and 2, respectively, are considered "the backbone of an autonomous civil society" (Rueschemeyer 1997, p. 217) a notion in vogue in policy circles but lacking sound theoretical underpinnings. Chapter 3 first discusses the concept of civil society and the channels through which civil society is believed to affect development. It then focuses on the role of civil society as a bulwark against the state and discusses the experience with civil society aid in post-communist countries. Utilizing the formal analysis laid out in Chapter 1, the chapter provides an introductory examination of the rationale for civil society aid and concludes with a conjectural interpretation of the determinants of the aid's effectiveness to bring about successful institutional change.
v
Dedication
Atiju in mami: za vso pomo?, vzpodbudo in ljubezen.
vi
Acknowledgements
I am grateful to Peter Murrell for his inspiration, guidance, support and advice. I also thank Larry Ausubel, Chas Cadwell, Thomas Carothers, Valentina Dimitrova-Grajzl, Kathryn Hendley, Simon Johnson, Pamela Jordan, Philip Keefer, Rachel Kranton, Andrei Shleifer, Thomas Stratmann, John Wallis, Christopher Waters, and participants at seminars at the Federal Trade Commission, George Mason University, University of Maryland, Inter-University Graduate Student Conference at Yale University, Central European University, and CERGE-EI for many helpful comments.
vii
Table of Contents
Preface Dedication Acknowledgements Table of Contents Chapter 1: Lawyers and Politicians: The Impact of Organized Legal Professions on Institutional Reforms 1.1 The Rule of Law, Organized Professions, and Development 1.2 Organized Legal Professions: Nature and History 1.2.1 The Nature of Organized Legal Professions 1.2.2 Characterizing Legal Professions 1.2.3 A Brief History of Organized Legal Professions 1.3 A Model of Professions in the Reform Process 1.3.1 The Structure of Reform Programs 1.3.2 Interest Groups, Politicians, Society 1.3.3 Politician and the Organized Profession 1.3.4 The Timeline 1.4 When Do Politicians Use Powerless Professions 1.4.1 Baseline Equilibria: When the Profession Does Not Exist 1.4.2 Equilibria when the Profession Exists and the Politician's Self Interest Matters 1.4.3 Equilibria when the Profession Exists and the Politician's Aims Coincide with Society's 1.4.4 Professions Under Autocracy and Transitions to Democracy 1.5 When Do Politicians Use Professions that Have Political Power 1.5.1 Comparative Legal Development in England and France: 1688 and 1789 1.5.2 The Professions Under Civil and Common Law 1.5.3 Colonies and Institutions 1.6 A Postscript on Further Research Chapter 2: Allocating Law-Making Powers: Self-Regulation vs. Government Regulation 2.1 Introduction 2.2 A Simple Model 2.2.1 The Setup 2.2.2 Solving the Model 2.3 Comparing Regulatory Arrangements: Efficiency and Government's Incentives 2.4 Illustrating the Model 2.4.1 Two Modes of Attaining Regulatory Goals: The Progressive Era and the New Deal 2.4.2 The Fascist Economic System: Self-Regulation in a Totalitarian Regime ii vi vii viii 1 1 6 6 9 10 13 13 16 18 20 21 22 23 25 27 32 34 38 41 43 45 45 49 49 56 59 63 63 67
viii
2.4.3 Legal Origin and Regulatory Regime Choice 2.5 Conclusion Chapter 3: Fostering Civil Society: Towards an Analytical Framework 3.1 A Concept in Vogue 3.2 Delineating Civil Society 3.3 Promoting Civil Society: The Democracy Aid Project 3.3.1 Civil Society and Development 3.3.2 Civil Society and Institutional Change 3.4 Evaluating the Aid's Success: The Role of Initial Political Conditions 3.5 Aiding Whom and When: Towards an Analytical Framework 3.5.1 Model Background 3.5.2 When Does Aid Make A Difference? The Role of Political Conditions Revisited 3.5.3 Judging the Success of Civil Society Aid: A Conjectural Interpretation 3.6 Summing Up Appendix 1: Figures Figure 1: The Set of Reforms Figure 2: The Timeline Figure 3: The Politician's Welfare and Aggregate Welfare with and without a Powerless Profession Figure 4: Professional Political Power, Democracy, and Stability Figure 5: Political Power and the Use of the Profession Figure 6: The Political Cost of Removing the Profession Figure 7: Timeline of the Regulatory Process Figure 8: SR vs. R: Efficiency vs. Government's Incentives Figure 9: Legal Origin and Regulatory Regime Choice Figure 10: Effectiveness of Aid to Civil Society Appendix 2: Tables Table 1: Summary Statistics by Regulatory Arrangements: Means (and Standard Deviations) of Variables Table 2: The Relationship between Legal Origin and Regulatory Regime: Ordered Probit Results Appendix 3: Proofs of Propositions in Chapter 1 Appendix 4: Model with Reform 1 Appendix 5: Algebraic Derivations and Proofs of Results in Chapter 2 Appendix 6: Civil Society as Vehicle for Service Provision Appendix 7: Welfare Consequences of Delaying Reform Implementation when the Lobbying Group's Favored Reform Program Increases Social Welfare References
70 74 76 76 77 78 79 82 84 86 88 90 92 94 96 96 97 98 99 100 101 102 102 103 104 105 105 106 107 110 120 123 127 129
ix
Chapter 1 Lawyers and Politicians: The Impact of Organized Legal Professions on Institutional Reforms
1.1 The Rule Of Law, Organized Professions, and Development "Above all, it is now abundantly clear that the history of the legal profession is an important part of the history of the law…." Plucknett (1983, p. 332) "The history of the legal profession is to a great extent, and despite noisy and incessant protestation and apologetics, the history of efforts by all branches of the profession, including the professoriat and the judiciary, to secure a lustrous place in the financial and social-status sun." Posner (1995, p. 33) One of the most significant recent developments in economics has been the rediscovery of the central role of the rule of law in economic development (e.g., North 1990; La Porta et al.1998; Rodrik et al. 2002). Despite this new affection for law, economists have not been so eager to embrace lawyers, and even less so the organized legal professions (e.g., Datta and Nugent 1986; Murphy et al. 1991; Posner 1995). Yet, the rule of law is a human construct, not a natural endowment. Lawyers are the architects and engineers of that construction process, and in every developed country a strong, organized legal profession is central to lawyering. This paper presents a theory that establishes a role for an organized legal profession in economic development. It shows how a legal profession can improve the quality of institutional reforms. The argument does not rely on some newly identified capability of organized professions. Rather, it rests on that characteristic of the involvement of strong professions in reform processes that has elicited much negative commentary–the delay caused by deliberation. The model predicts how the role of the organized legal profession varies with levels of democracy and political stability, establishing the direct relevance of these predictions
1
to history and to development policy. The analysis casts new light on historical episodes that have been the object of considerable debate: the old regime, 1789, and Napoleon in France, the Glorious Revolution in England, the USSR and post-socialist transition, the contrasting nature of the English and French legal systems, and the comparative development of former colonies (North and Weingast 1989; La Porta et al. 1998; Rajan and Zingales 2003; Acemoglu et al. 2001). The paper elucidates the character of civil society, a popular notion in development policy, which lacks a sound theoretical underpinning (Carothers 1999). Economists typically view organized professions as rent-seeking interest groups, placing them in a negative light (e.g., Stigler 1971; Mueller 1989; Drazen 2000).2 However, with the organized legal profession cast as just another rent-seeker, there is a paradox to be resolved, encapsulated in the contrast between the epigraphs to this paper. Those countries adhering most strongly to the rule of law and with the highest quality legal rules (La Porta et al. 1998) have notoriously powerful legal professions. Already in medieval times the English legal profession was a force to be reckoned with, and has rarely been challenged since (Burrage 1997). There, the legal profession was a resolute companion of both the rise of democracy and the institutional revolution accompanying modern development: the legal professions have been builders of the liberal state and society (Halliday and Karpik 1997, p. 16). In contrast, in many former colonies and in transition countries, where law has played a much less significant role, organized legal professions have been virtually irrelevant (Johnson 1973; Waters 2004).
2
Becker (1983) shows how political competition can reduce the negative consequences of the activities of rent-seeking groups.
2
These facts suggest that organized legal professions play a role beyond that of rentseeking. As Hayek (1960, pp. 59-60) argues, successful development has often been the result of trial and error selecting superior institutions, even when the reason for that superiority is little understood. The tenacity of the organized legal profession, its central role in developed societies, and its coevolution with the rule of law constitute prima facie evidence that it has been one of the Hayekian survivors. This simple observation is a major stimulus for our analysis, which uses standard theories to isolate the way in which organized professions contribute to institutional development. We focus on the process of designing and implementing institutional reforms. In that process, the government faces the competing demands of different interests. Rentseeking interest groups have more information about the effects of proposed reforms than the politicians forming the government. Since interest groups submit self-serving reforms for consideration, the politician faces an informational problem because some reforms are damaging to the politician and others are not. One mechanism that ameliorates this problem is the government's use of the organized profession in a review process that stimulates debate about the reform. The prestige, status, and independence of the organized profession adds political legitimacy and credibility to the decision to use the profession in the face of powerful interest groups pressing for quick reforms. The review opens up the reform process to public debate and delays the adoption of reforms, which will be costly for interest groups, since the government might lose power before implementation occurs. This cost changes the incentives of the interest groups when submitting reforms; they drop initiatives with lower payoffs. Institutional reforms improve.
3
The politician is interested in more than the quality of institutional reforms. Interest groups support politicians, contributing funds when submitting reform proposals. A politician weighs these contributions against the pursuit of the general welfare, the latter being relatively more important when democratic constraints are tighter. Hence, the politician is more likely to use the organized profession when democracy is stronger. Political stability also matters. When government turnover is more likely, the politician's expected value from using the profession to obtain better, but later, reforms declines. Democracy and stability become substitutes. The politician uses professional review only at high levels of democracy in an unstable environment but at lower levels of democracy when there is stability. This trade-off is particularly important in transitions from autocracy, which often involve the paradox of rising democracy and reduced influence of the organized legal profession. The key to the paradox is that instability often rises during these transitions. We use these insights to interpret events in the postsocialist transition and in eighteenth century France. Once a professional group has acquired political power, the politician also has to consider the political costs of challenging the group's status. The power of the profession can change the politician's decision, leading to improved institutional reforms. This effect occurs only in the middle ranges of democracy. When it does, professional power substitutes for weak democracy. Matching this prediction to historical events, we examine the role of the English legal profession during the Glorious Revolution, and show how instability and a weak legal profession contributed to different results in France a century later.
4
Our focus on the organized professions can be used to interpret recent empirical work on the historical origins of institutional quality, allowing us to identify complementarities between different contributions to the literature. We show that the politician uses the organized legal profession more often in common law than in civil law countries, given that legal professions are naturally more powerful in the former. This choice leads on average to superior institutions in common law counties, but also to slower implementation, reconciling the views of La Porta et al. (1998) and Rajan and Zingales (2003). Our analysis provides a new perspective on mechanisms underlying the results of Acemoglu et al. (2001) on the colonial origins of institutional development. Since organized professions were stronger in those colonies with high levels of European migration, our analysis predicts that institutional quality is lower in the successor nation states of those colonies that were less hospitable to European settlement. This squares with the focus of Berkowitz et al. (2003) on local knowledge during legal transplantation, since the organized profession is the prime source of that knowledge. Lastly, given the correlation between professional power and legal origin and given that judicial independence is bolstered by a strong profession, our model is pertinent to the work that draws a connection between judicial independence and legal origin (La Porta et al. 2004). The chapter's argument proceeds by modulating between the model and related facts from history and from studies of professions. Section 1.2 examines the defining characteristics of organized professions and presents a brief history of the legal profession. Section 1.3 builds upon these facts in presenting the model. Section 1.4 develops predictions using the assumption that the profession is politically powerless, identifying the circumstances under which the profession contributes to the improvement
5
of institutional reforms. These predictions are used to interpret events in Louis XIV's France, in the Soviet Union, and in the transition from autocracy. Section 1.5 introduces the assumption that the profession has political power, showing that professional power and democracy are substitutes. The predictions cast light on the differing developments in England in 1688 and France in 1789, and the consequences of these developments. Lastly, we show how the model bears on the legal and colonial origins of institutional development. A concluding section considers implications for future research. 1.2 Organized Legal Professions: Nature and History This section summarizes the facts that suggest assumptions for the model and that provide context for interpretation of historical episodes. We examine the nature of organized legal professions, the special status they play in reviewing reform proposals, and the reasons why organized legal professions accumulate unusual amounts of political power. A brief overview of the history of organized legal professions summarizes differences across countries. 1.2.1 The Nature of Organized Legal Professions The practice of some occupations entails using a specialized body of knowledge that has been created by past practitioners. Practice requires costly training, which practitioners are best qualified to impart. These characteristics make an occupation a profession, in its broadest sense. But we use a narrower concept of a profession, adding the criterion that permission to practice depends upon demonstrated expertise in the body of knowledge.3 Moreover, the strongest organized professions usually are self-regulating
See, for example, Torstendahl and Burrage (1990) and Burrage, Jarausch and Siegrist (1990). Åmark (1990) draws the distinction between lobbying groups and professional lobbies, by noting that the former are open to the general public, while the latter are not. We do not examine why there are formal requirements. These requirements are nearly universal in the areas of professional activity that we
3
6
monopolies, administering the use of this body of knowledge by implementing a single set of policies, rules, and standards.4 Formalization of self-regulation confers quasigovernmental status. With the organized profession deemed to have monopoly expertise, its members are accorded special status in advising the government on reforms, especially in areas involving the rules and procedures that practitioners use. Already in the 1300's, legal professionals advised the English king on statutes (Rose 1998) and dominated a commission on legal reforms (Plucknett 1983, p. xix 335). The special advisory role might be so automatic that it becomes formalized. The British Commonwealth legal systems have gone far in this process by setting up Law Reform Commissions, dominated by professional lawyers and charged with reviewing all aspects of legal reform.5 The advisory role does not necessarily involve officially appointing the monopoly professional association to some formal body. The crucial issue is that the politician facilitates a process in which the voice of the independent profession fosters open debate. Governments can satisfy professional demands by consulting with prominent sub-groups
examine. But there are exceptions, which provide examples used later in the paper: Waters (2004, p. 95) notes that lawyering in Georgia was 'free for all' during the early 1990s. "...there can be no doubt that the profession's knowledge about the intricacies of professional practice, the official and unofficial procedures, and the opportunities for manoeuvre and circumvention that these provide, is rarely rivaled. This is an immense source of power since it means that any attempt to change or control professional behavior by instituting new rules and procedures has to be negotiated with residential procedural experts, the practitioners who will actually implement the change." (Burrage, Jarausch and Siegrist 1990, p. 210). For example, in the case of the Manitoba Law Reform Commission (2003): "The Commission's duties are to inquire into and consider any matter relating to law in Manitoba with a view to making recommendations for the improvement, modernization and reform of law…" The commission is "…composed of five members…At least one shall be a judge of the Court of Queen's Bench; at least one shall be a full-time member of the Faculty of Law, University of Manitoba; at least one shall be a barrister and solicitor entitled to practise in the province and who is not in the full-time employ of the Government of Manitoba or any agency thereof; and at least one shall not be a lawyer. One of the members shall be appointed President and he or she must be a lawyer."
5 4
7
or by using independent, universally respected, members of the profession.6 The process can be formal, as when the politician summons a reform commission consisting of prominent members of the legal profession, or it can be informal, as when the organized legal profession naturally emerges as the leader of a public debate that the politician has allowed to proceed. Whatever the nature of the process, the status and prestige of the organized profession affords political legitimacy and credibility to the politician's decision to use it in the face of importuning interest groups. Among the consequences of professional review, there is one that is universally agreed upon: a slowing of the reform process. The twenty-six-year, lawyer-dominated, gestation period of the first German civil code (Zimmerman 1996, p. 6) stands in contrast to the six months in which the Weimar politicians drafted Germany's first democratic constitution. While Napoleon's constitution of 1799 was created and ratified in two months, it took four years for a commission of old practitioners to craft his civil code, largely reviving the legal framework created under Louis XIV and Louis XV (David and de Vries 1958, p. 13). Examples from more recent times are the eight years for new bankruptcy laws in the US and the UK and the six years for new civil procedure codes in France and Japan, each process beginning with lengthy deliberations by commissions of professionals appointed by politicians (Carruthers and Halliday 2000; Cadiet 1999, p. 315; Hasebe 1999, p. 237). At the very least then, one may assume that professional review changes the pace of the reform process.
6
The role of the American Law Institute in the US provides an example.
8
1.2.2 Characterizing Legal Professions The foregoing assumes that a profession exists in the sense that there is an occupational group that is capable of self-organization whose members have sufficient expertise to be used by the politician in an advisory role. This assumption might not hold, as in China after the cultural revolution (Pei 2001, pp. 181-2). Thus, a first element of the characterization of a profession is whether it exists. Once it exists, the organized profession's status, particularly its relationship with government, can take on many different hues. In some cases the state creates a profession and gives it formal status; in others custom and political struggle are more relevant. Thus, the profession might be coterminous with a government department, as in early nineteenth century Prussia (Rueschemeyer 1997; Kocka 1990; Siegrist 1990), or it might be a ‘little commonwealth’ or ‘lesser government’ as was the English bar already by the middle ages, even though its autonomy was never formalized.7 Separation between the government and the profession involves several features, for example, the degree of self-regulatory authority or the strength of the opposition when the government tries to reduce the role of the profession. These are inter-related. Selfregulation is a political prize won by an independent profession, but self-regulation also protects independence, keeping a disciplining technology out of government control. Self-regulation enhances the profession's monopoly of specialized knowledge, which increases the need to involve it during reforms. Indeed, in the descriptive literature, the notions of independence, autonomy, and self-regulation are often used as synonyms.
7
Halliday and Karpik (1997, p. 354): "[In England] Barristers’ independence has its medieval origins in civil society, not the state. Indeed, barristers’ autonomy, and those of other ‘little commonwealths’ in civil society, are entrenched in the unwritten English constitution." Burrage (1988, p. 356): "Charles [II] sought to gain control of all ‘lesser governments’ in the Kingdom."
9
Therefore, in the model, we assume that a single variable, political strength, reflects all these notions. When strength is high, the politician incurs a large, direct cost in ignoring the profession during reforms. When strength is zero, there is no direct political cost of removing the profession from the reform process. 1.2.3 A Brief History of Organized Legal Professions Legal professions first emerged in England in the twelfth century (Brand 1992). Lawyers were recognized as having skills that were valuable to the government. They won privileges and status by virtue of their expertise. Serjeants, the most prominent members of the early English legal profession, were made officials of the crown. Parliament appointed lawyers as members of a commission inquiring into legal reforms. By 1300, judges were chosen from the bar rather than from royal officialdom (Plucknett 1983). The autonomous Inns of Court acquired monopoly control over education and admission to practice. The Inns' independence and power gradually became an element of the unwritten constitution. In the1680s, the crown attempted to challenge the power of the profession, but the result was the "the triumph of the Common Law and lawyers over the King, who had tried to put Prerogative above the law" (Trevelyan 1970, p. 71). Hence the power and independence of the English legal profession were won over a long period, in a process intertwined with the development of the common law. An attack on the status of the profession was not a viable political option after 1688 (Burrage 1997).8 The French legal profession developed to satisfy the specific needs of royal courts, parlements, which after their establishment in the late thirteenth century had full control over the profession (Karpik 1999). In the following centuries, the courts acted
Margaret Thatcher's government published a Green Paper that "anticipated, even celebrated, the end of sovereign, self-governing professions", but the government that had successfully implemented some of the most radical measures enacted in post-war Britain was forced to retreat quickly (Burrage 1997, p. 140).
8
10
partially as agents of the crown, while barristers became counsels of the crown. In contrast to England, the French barristers did not provide the pool for judicial appointments. The Paris Order of Barristers, France's first self-regulatory body of lawyers, was founded in 1660. The profession’s reputation stood high for the next century. In the 1760s, however, the French legal profession became a forum for battles between advocates of different views on the role of the monarchy, including whether the profession itself should be independent (Bell 1997). The internal dissensions, combined with an insecure power base, led to steadily diminishing authority in the period before the 1789 revolution. In contrast to England in 1688, the French legal profession was in no position to play the role of neutral arbiter in the country's decisive revolutionary event. The National Assembly abolished the Order of Barristers in 1790. Paradoxically, by pursuing broad political goals rather than narrow professional ones, French lawyers secured the demise of their own organized legal profession. Napoleon re-established the Order in 1810, as a creature of the state. In the volatile political environment that followed Napoleon's fall, the legal profession went through many vicissitudes, but never succeeded in achieving the prestige that it had acquired under Louis XIV (Karpik 1988). Only with the establishment of the Third Republic (1875) and the arrival of a stable democracy did the bar regain its old freedoms (Bell 1994). But these freedoms were bestowed by the state rather than won by the profession's political action. The German legal profession also had its origins in the state. Throughout the 18th century, Prussia reformed the size, composition, and education of the profession, in a
11
process that essentially converted lawyers into civil servants. Although the dominance of the government did come under attack in the 19th century, the central role of the state as patron of the profession was never in doubt. During German unification, professional autonomy increased but the formation of a country-wide association of lawyers in 1871 was in significant part a result of state action. In the Second Empire (1871?1918), the Ministry of Justice was the initiator of most legislation affecting the status of the profession (Rueschemeyer 1997). As the above indicates, the profession’s role varies across legal systems (Abel 1995). In a typical civil law country, the organized profession is a creature of the state.9 In a common law country, a unified profession usually won its own independence by dint of its technical expertise in a system of justice relying on decentralized courts and law. Historically, it has been largely autonomous from the state (Halliday and Karpik 1997, pp. 5, 354-5). Civil law professions are weaker, divided into sub-groups, and exert less influence on the judiciary. Colonial status also matters. In those British colonies where settlers were numerically dominant, such as Australia, legal professions followed the English path, attaining the same degree of autonomy and authority (Weisbrot 1988). In contrast where colonial settlement was less attractive, autonomous, independent professions failed to emerge, since the indigenous population was excluded from legal training. There was no basis for post-colonial professional power (Johnson 1973). In Kenya or in India, for example, no powerful independent profession ever emerged to exert a major influence (Ross 1992, Gandhi 1988).
For example, compare Belgium and Japan (Huyse 1988, p.182; Rokumoto 1988, p.128?9) to England and Australia (Burrage 1989; Weisbrot 1988).
9
12
Lastly, it is important to emphasize that while there are many examples of autocrats who oppose existing professions, powerful autocrats do not always take this stand. The French Order flourished under Louis XIV; Napoleon reestablished it after its demise in revolution. Lawyer groups had some (highly constrained) independence in Soviet times (Krause 1991). Moreover, transition from autocracy does not automatically translate into increased status for the profession, and might even lead to the reverse, as was the case in some post-Soviet republics (Waters 2004). 1.3 A Model of Professions in the Reform Process10 The model uses the phrasing of interest group politics in a democracy, but it could be recast to apply to autocracies. We use the terminology of competing politicians, elections, interest groups, and campaign contributions, but we could also have referred to an autocrat facing factions, receiving tribute, and deliberating on the possibility of being usurped. 1.3.1 The Structure of Reform Programs A fundamental institutional reform is under consideration. Reversal in the medium-term is impossible, and at most one reform program can be enacted.11 Politicians and interest groups compare reform programs to the status quo. Those directly affected fall into two interest groups, ? and ?, which can devote resources to influence the government’s decision.
10
Elements of the model are derived from Dewatripont and Maskin (1995) and Qian (1994), which focus on finance and on soft-budgets. To emphasize, our model is not relevant to short-term policy. Once the incumbent has had a chance to implement a reform, the country enters a new era with different economic circumstances and altered political configurations. Then, the question of which reforms to undertake is posed anew.
11
13
Reform programs comprise combinations of individual reforms. Reform j is completely characterized by B?j and B?j, the benefits to ? or ? from implementation. Classifying reforms according to whether ? or ? win or lose and whether the sum of payoffs is positive or negative, there are six distinct generic reform types. Figure 1 portrays these six, the line TT separating those with positive aggregate payoff from those with negative payoff. With reforms characterized by payoffs to ? or ?, all reform programs comprise from one to six reforms, at most one from each type.12 Hence, in the following, j = 1,…,6. The impetus for particular reforms comes from interest groups submitting programs. Hence, Reform 6, which is Pareto inferior to the status quo, can be dropped from the analysis because neither ? nor ? submits it. Given the assumptions below, Reform 1, which is Pareto improving, is always submitted and adopted. As its presence or absence does not alter the qualitative predictions of the model, reform 1 can also be omitted.13 Because reforms 4 and 5 play the same role for ? as 2 and 3 do for ?, the analysis begins by focusing on 2 and 3, whose payoff structure is summarized as follows: Reform 2 Reform 3 Interest group ? B?2>0 B?3>0 Interest group ? B?2<0 B?3<0 Aggregate (? and ?) B?2+B?2<0 B?3+B?3>0
Reform 2 is redistributive with a deadweight loss. Reform 3 also redistributes but with a welfare gain as a consequence of introducing more effective institutions. An example of reform 2 might be a change in corporate law that makes financial markets less efficient, aiding dominant owners by reducing protections for minority shareholders.
12 13
Two reforms of type j combined are equal to one reform of type j.
For further elaboration, see Section 1.4.2. A complete presentation of the model including reform 1 is presented in Appendix 4.
14
Reform 3 could be an improvement in contract law that facilitates the use of collateral, aiding new businesses, but reducing entry barriers that protect existing firms. Both measures could belong in an omnibus finance-oriented institutional reform program, underscoring the relevance of the model to situations where packages of related reforms are combined into one program. The structure of payoffs captures a fundamental aspect of institutional reform: rent-seeking interest groups favor reform programs with several sub-components, some contributing to general welfare and some reducing general welfare.14 The central problem of institutional reform is to foster adoption of the efficiency-enhancing components and reduce the likelihood of implementing the pure rent-seeking redistributions. We therefore introduce two assumptions that keep this problem at the center of considerations. Without these assumptions, a reform program can be dominated by one type of reform and would reflect only that type of reform. First, assume that reforms 2 and 3 have the same scale, in the sense that they involve the same amount of pure rent-seeking redistribution from ? to ?:15 Assumption A1. B?2=B?3. Since interest group submission fees must be related to the size of the reform, A1 fits naturally with our assumptions on those fees, which are discussed in the ensuing subsection. Second, assume that ? is, on balance, a rent-seeking interest group, in the sense that its favored program reduces the general welfare: Assumption A2. ?B?2+B?2?>B?3+B?3.16
14 15
This structure echoes the assumptions in Becker (1983). For brevity, the analogous assumptions for reforms 4 and 5 are omitted, but implicitly imposed.
15
A1 and A2 maintain the focus on the central problem of institutional reform by ensuring that A must blunt ?'s rent-seeking propensities if beneficial institutional reforms are to be enacted. 1.3.2 Interest Groups, Politicians, Society There are two rival politicians, A and B. The one forming the government has the power to implement reforms. The interest groups submit particular reform programs. The government accepts a submission only if the interest group pays a lobbying fee. Given A1, the fee does not vary across reforms: ?ik is the payment when the government of politician k accepts a submission of a single reform from interest group i.17 The politician is at an informational disadvantage and cannot recognize reform types when they are submitted. However, politicians do understand the model and recognize which interest group lobbies. Given the structure of payoffs, ? and ? are in opposition. Hence, it is natural to view enactment of reforms 2 or 3 as inconsistent with enactment of 4 or 5.18 One
Equivalently B?2+B?2+B?3+B?3 < 0. To understand the intuition of A2 note that if it does not hold, B?3 can be infinite. To see this, let B?3=B?2= ?c, 0<c<?. Since B?2+B?2<0, then B?2?(0,c). If ?B?2+B?2?<B?3+B?3, then B?3>?B?2+B?2??B?3= ?B?2+2c, so B?3=? is consistent with the model. With A2, B?3 is bounded by ?B?2+B?2??B?3= ?B?2+2c<?. Observe that if A2 does not hold, ?’s favorite reform program–although not first best for society? improves general welfare. The less ? is a rent-seeker, i.e. the greater is (B?3+B?3)??B?2+B?2?>0, the more the qualitative predictions differ from the ones presented here.
17 18
16
If interest group ? wants to submit both reforms 2 and 3 to A, for example, it must pay 2??A.
The assumption of direct conflict is common in the analysis of the effect of interest groups on policy. Examples from Grossman and Helpman (2001, sections 4.2.2, 8.4, and 9.3) are multiple lobbies when favored outcomes are on either side of the politician's ideal, when a business lobby wants lower tariffs and higher export subsidies than does a labor group, and when groups want to move in opposite directions from the status quo. Drazen (2000, Sections 3.7, 13.5) examines situations where redistribution is from one group to another and where special interest groups can block reforms in transition countries. Mueller (1989, pp. 108, 230, 245, 247, 279, and 453) offers many examples–liberal versus conservative reforms, rent-seeking for regulations, monopolies, tariffs, changes in the structure of regulatory processes, the supply of government output, macroeconomic policy, and agricultural programs.
16
politician cannot help both groups. We build this formally into the model's structure by assuming that each interest group is aligned with a different politician:19 Assumption A3: For all j such that B?j>0, ??A<B?j<??B. For all j such that B?j>0, ??B<B?j<??A. These assumptions fit the politics of longer-term institutional reform, where different visions of the future are in contention and the ideologies of politicians restrict their actions. Then, interest groups oppose each other in a manner that aligns with political ideologies: politicians could not credibly solicit support from all groups. The resultant structure of lobbying (? lobbies only A and ? only B) is consistent with empirical work on campaign contributions, which shows alignment between the objectives of donors and politicians (Poole and Roomer 1985; Poole, Romer, and Rosenthal 1987). This structure is also consistent with empirical work that shows that contributions influence politician's decisions (Stratmann 1991, 1995, 2002). Aggregate welfare is the sum of payoffs to the two interest groups: Assumption A4: Aggregate welfare from implementing reform j is B?j+B?j
19
A3 is formulated in strong terms to simplify the exposition. It is sufficient, not necessary. Its key implication is that the lobbying equilibrium within any time period results in only one of the following: (i) no lobbying or (ii) lobbying for 2 or 3 or both or (iii) lobbying for 4 or 5 or both. This implication is innocuous given that reforms 2 and 3 are contradictory with 4 and 5. Many different assumptions could lead to the same implication, for example, directly assuming that interest groups lobby against reforms they do not like or directly assuming that enactment of 2 or 3 involves repeal of 4 and 5. Assumption A3 can be interpreted as a strong alignment assumption since the lobbying fee charged by a politician to the non-aligned interest group exceeds the benefit from the reform to that interest group. The model’s qualitative features do not change with a weak alignment assumption, where the lobbying fee charged by a politician can be marginally smaller than the benefit from the reform to the interest group, that is, for all j such that B?j>0, ??B<??A<B?j<??A+? and for all j such that B?j>0, ??A<??B<B?j<??B+?, where ?>0 is arbitrarily small.
17
Note that the lobbying fee is a pure transfer and does not appear in aggregate welfare.20 Politician k’s payoff from accepting the proposal of reform j from group i and implementing it is a weighted average of aggregate welfare and lobbying income: Assumption A5: The politician's payoff from j is ?(B?j+B?j)+(1??)?ik, ??(0,1).21 ? is common knowledge. ? increases with either the politician's benevolence or with improvements in democracy that force the politician to pay greater attention to general welfare. If all politicians have equal degrees of benevolence, ? becomes a measure of democracy.22 Under dictatorship ? is closer to 0 and ?ik reflects the tribute the dictator extracts from subjects. 1.3.3 Politicians and the Organized Profession There exists an organized profession having the required expertise and authority to review, to comment on, and to stimulate public debate about proposed reform programs. Before interest groups submit proposals, the government decides whether such review is to be standard. By opening up the reform process to debate by an organized profession that values due process and openness, the review delays the passage of reform programs. In this paper, we ignore the negative role of the profession that is
20
This way of accounting for social welfare is consistent with Shleifer and Vishny (1994), Persson and Tabellini (2000, Chapter 7), Persson (1998), Grossman and Helpman (1994) and Grossman and Helpman (2001), Chapters 7 and 8.
? does not depend on the politician’s type, k. This specification of the politician’s payoff implicitly assumes that the politician does not attach any value to holding office per se. This form of the objective function is consistent with the political economy literature on campaign contributions; see Persson and Tabellini (2000, Chapter 7), Persson (1998) and Grossman and Helpman (2001, Chapters 7 and 8). If ?(B?j+B?j)+(1??)?ik, where ??(0,1), is rewritten as ?(B?j+B?j??ik)+?ik, then the politician’s objective function has the interpretation of the politician valuing a dollar in his hands more highly than a dollar in the hands of the public. Observe also that maximizing ?(B?j+B?j)+(1??)?ik, where ??(0,1), is equivalent to maximizing ?(B?j+B?j)+?ik, where ???/(1??)>0. See also Grossman and Helpman (1994) for a discussion of the politician’s objective functions. This interpretation of democracy exactly corresponds to Glaeser and Shleifer's (2002) interpretation of their parameter that captures the degree to which the preferences of the sovereign differ from those of the society.
22
21
18
usually emphasized, its interest in particular reforms, since our focus is on understanding whether there can be a positive aspect to professional activity. There are two central elements to our characterization of the government's use of professional review, first, that the government can build a credible, politically persuasive justification for review by using the organized profession and, second, that delay is the only effect. Section 1.2 presents the background on the first aspect: the politician can invoke the prestige and status of the profession, which flow from its monopoly expertise in a specialized body of knowledge. Since the interest group aligned with the incumbent politician is damaged by the delay, it is especially important that an aura of political legitimacy and credibility surround the decision to use the review process. This can only be provided by the use of the organized profession and the ensuing public debate: it could not be provided simply by qualified individuals personally chosen by the politician and it could not be provided by the politician simply deciding to delay consideration of the reform. Moreover, in those situations where the organized profession has political power, the politician's decision has a natural credibility. Professional involvement obviously does more than delay: the review process also generates technical information that might improve the quality of all reform proposals. But the generation of technical information does not require neutral and open review by the organized profession. It can be produced by qualified individuals who are trusted by the politician. Thus, the generation of technical information can be accomplished in a very different process from that examined here: it does not need the involvement of the organized profession. Hence, we assume that the review process does not lead to improvements in the quality of individual reforms. This assumption follows from the
19
objective of examining the effect of the organized profession, beyond the contribution of the technical skills possessed by individual members. The consequence of delay is that a particular government might run out of time before being able to implement a submitted reform program. The incumbent might lose an election and be replaced by the other politician.23 Then the interest groups have to incur lobbying costs again if they still want their reforms to be considered for implementation. 1.3.4 The Timeline There is a single reform process beginning when one politician, say A, takes the reins of power. The model covers two sub-periods, A's initial incumbency and the following time interval when either A or B forms the government. The decision on whether to use professional review is made at the beginning of A's incumbency and holds for the whole reform process. The irreversibility of this decision within the time-frame of the model reflects the view that once a single reform process has begun, it would be difficult for a politician to change that process. Let p?(0,1) be the known, exogenous probability that A stays in power in the second period. p is a measure of political stability.24 Values close to 0 correspond to chaotic, often revolutionary, environments, while values close to 1 reflect stagnation rather than stability. Hence, p is not an indicator of democracy: a p close to 1 occurs only under autocracy. A p close to 0 could occur under either democracy or autocracy.
23 24
We do not model the details of the process of change of power.
Our assumptions on p exactly match those in Persson and Tabellini's (2000, Section 13.3) analysis of political instability.
20
In both periods, each interest group decides whether or not to submit a proposal without knowing what the other is doing. The government then decides whether to accept the proposals. If it does, it receives the submission fee. If there is no professional review the measure passes immediately. If there is professional review, the measure is debated until the second period. Then, if A remains in office, the reforms submitted in period 1 are implemented. If B takes over, each interest group decides again whether to submit a proposal. If the reforms submitted in period 2 are the ones that already passed through professional review in period 1, they are implemented. If the reforms submitted in period 2 differ from the ones submitted in period 1, they are delayed beyond the end of this reform process.25 Figure 2 summarizes the timeline. 1.4 When Do Politicians Use Powerless Professions? We begin with the simpler form of the model, when a profession exists, but the politician suffers no special costs in ignoring it. This matches the circumstances of autocracy and the early phases of democracy. Section 1.5 introduces the political power of the profession, matching developments as democracy evolves. Reform 3 is always desirable for politician A.26 Reform 2 is undesirable for A if ?(B?2+B?2)+(1??)??A<0, or, equivalently, if ?>?, where 0 < ? = ??A/[?B?2+B?2?+??A] < 1. ? is therefore the lowest level of democracy at which the politician's consideration of the general welfare has any effect: for ?<?, A would knowingly accept aggregate-welfare-reducing reform 2. When ?>?, A has an
25
This assumption closes the model without an infinite horizon. It is equivalent to focusing on a single reform process that must take place within a certain time window, which is a common feature of the circumstances surrounding institutional reforms. That is, there does not exist a ??(0,1) such that ?(B?3+B?3)+(1??)??A<0.
26
21
informational problem. If there is no screening mechanism, A either approves all reforms or none. A approves all reforms if and only if [?(B?2+B?2)+(1??)??A]+[?(B?3+B?3)+(1??)??A]>0, or, equivalently, if ?<? ? where ?<? ? =??A/{??A+½[?B?2+B?2??(B?3+B?3)]}<1.27 ? ? is therefore the lowest level of democracy at which the general welfare dominates A's decisions. In sum,28 when ?<?, A likes reform 2, accepts all reforms, and does not face an ?, A accepts all submitted reform programs (unless a informational problem. When ?<?<? program is known to comprise 2 only), but faces an adverse selection problem. When ?>? ? A rejects all submitted reform programs (unless a program is known to comprise 3 only), but still faces an informational problem. The model's basic parameters are p and ?, the levels of stability and democracy. Figure 3 summarizes the analysis as it proceeds, depicting how the model's equilibria vary with (p, ?). 1.4.1 Baseline Equilibria: When the Profession Does Not Exist First consider equilibria in a baseline case, when an organized profession does not exist. Without the profession, with A in power, and with ?<? ?, ? never lobbies, while ? lobbies immediately for reforms 2 and 3 and A implements them both in the first period, ending the reform process. When ?>? ? neither ? nor ? lobby for reforms because A
27
Observe that ?<? ? implies that any politician in power accepts any reform program, submitted by ? or by ?. A might in principle (but not in equilibrium) receive reforms submitted by ? in period 1. If A accepts any reform submitted by ?, then A accepts any reform submitted from ?, since ?'s lobbying fee is higher and 4 and 5 are symmetric to 2 and 3. This summary holds only as long as the politician does not use a mechanism that ex-ante screens out the undesirable reform.
28
22
rejects all proposals.29 Denote social welfare relative to the status quo when there is no profession as W?. Hence, W? = (B?2+B?2)+(B?3+B?3) < 0 = 0 ? when ?<? ?. when ?>?
Similarly, let V? be politician A's payoff relative to the status quo when there is no profession: V? = ?[(B?2+B?2)+(B?3+B?3)]+(1??)2?11 > 0 = 0 when ?<? ? when ?>? ?.
Hence, ?<? ? provides the more interesting case, since the politician would then want to implement the proposed reform program, which would reduce aggregate economic ? , there is no such conflict between politician and society. welfare. In the case of ?>? 1.4.2 Equilibria When the Profession Exists and the Politician's Self Interest Matters ? , A is faced with an informational problem and approves all submitted When ?<? reforms, the standard adverse selection scenario.30 Then, the equilibrium of the reformlobbying game between ? and ? is specified by the following:31 ? , A is in power, and there is professional review. Proposition 1. Assume that ?<? ? never lobbies for reforms 4 and 5. In period 1, ? lobbies for 3 if and only if p???A/B?3 and also for 2 if and only if p???A/B?2. In period 2, ? does not lobby for any reform regardless of which politician is in power. Interest group ? never lobbies for reforms. Therefore, in equilibrium, ? remains inactive (i.e. never lobbies). This conclusion is crucially related to our assumption that all Pareto-improving reforms have already been
Given that at most one reform program is implemented and the symmetry between ? and A and ? and B, nothing will occur in the second period, whoever is in power.
30 31
29
See, for example, Dewatripont and Maskin (1995) and Qian (1994). Proofs of all propositions are in Appendix 3.
23
implemented. Had we included reform 1 in the analysis, ? and ? would behave strategically when lobbying for its implementation and ? would actively engage in lobbying in equilibrium. It is therefore appropriate to view ? as an interest group (rather than the general public, say).32 Let W+ be social welfare with professional review. Proposition 2 immediately follows: Proposition 2. When ?<? ?, then: 0 when p<??A/B?3, + W = p(B?3+B?3) when ??A/B?3?p<??A/B?2, when p???A/B?2 p(B?2+B?2+B?3+B?3) W+ > W? for all p: aggregate welfare is always higher with the profession than without it. The reason why the profession is advantageous to society varies with levels of stability. If p is low, professional review dissuades ? from lobbying because A's survival is too uncertain to repay lobbying costs. If p is high, review and delay reduce the probability of implementation of a welfare-reducing reform program (2 and 3). For midrange p's, the profession solves society's informational problem: ? lobbies only for reform 3. Therefore, the use of professional review makes its strongest contribution in this mid-range, suggesting patterns of variation across countries in the value-added from an organized profession. The contribution of the profession identified in Proposition 2 is only a potential one, contingent on whether the politician decides to implement professional review. ? , it would not be surprising to find situations where the politician's objectives When ?<?
32
Intuitively, ? lobbies for reform 1 in period 1 only, and ? lobbies for reform 1 in period 2 only if B is in power, resulting in an equilibrium where (i) ? actively engages in lobbying, and (ii) reform 1 is always implemented. See Appendix 2 for proofs.
24
conflict with those of the general population. Denote the politician’s payoff with professional review as V+. The following proposition relates the difference between V+ and V? to the values of p and ?: Proposition 3. Assume ?<? ?. Let p?(?)?1+[?(B?2+B?2)+(1??)??A]/[?(B?3+B?3)]. 1. With professional review, the politician's welfare is given by: 0 if p<??A/B?3 + V = ?p(B?3+B?3)+(1??)??A if ??A/B?3?p<??A/B?2 ?p(B?2+B?2+B?3+B?3)+(1??)2??A if p???A/B?2 2. The politician's gain or loss from professional review has the following properties: V+ < V? if p<??A/B?3 + ? if ??A/B?3?p<??A/B?2 and p< p?(?) V <V + ? V >V if ??A/B?3?p<??A/B?2 and p> p?(?) V+ > V? if p???A/B?2 ? 3. The line p (?) in Figure 3 is downward sloping, convex to the origin, intersects p = ??A/B?3 below ? ?, and intersects p = ??A/B?2 above ?. Proposition 3 establishes that the decision to use the profession is related to (p, ?) in a straightforward way, as shown in Figure 3. Before discussing this relationship, we find the remaining equilibria of the model by examining outcomes when ?>? ?. 1.4.3 Equilibria When the Profession Exists and the Politician's Aims Coincide with Society's With professional review and ?>? ?, if p<??A/B?3, ? does not lobby because the probability of A's keeping power is not high enough to repay lobbying costs.33 Therefore, if p<??A/B?3, W+ = V+ = 0 (= W? = V?). When p?[??A/B?3,??A/B?2), lobbying for only reform 3 is beneficial for ?, while lobbying for 2 and 3 is not. Since A knows ?’s incentives and reform 3 is desirable for A, it is approved. Therefore, when
33
We omit stating a modified version of Proposition 1 for brevity only. As in equilibria with ?<? ?, ? never lobbies for reforms.
25
p?[??A/B?3,??A/B?2), W+ = p(B?3+B?3) > W? = 0 and V+ = ?p(B?3+B?3)+(1??)??A > V? = 0. For p???A/B?2, A approves ?'s submission if ?p(B?2+B?2+B?3+B?3)+(1??)2??A>0, or equivalently, if ?<?C(p), where ?C(p) = ??A/{??A+½p[?B?2+B?2??(B?3+B?3)]}. ?C(p) is strictly decreasing and convex in p with ?C(0) = 1, ?C(1) = ? ?. Therefore, for p???A/B?2 and ?>?C(p), A rejects all submitted reform programs and consequently ? does not lobby for any: W+ = V+ = 0 (= W? = V?). For p???A/B?2 and ?<?C(p), however, ? lobbies for the reform program consisting of reforms 2 and 3, and A approves it, meaning that V+ > V? = 0 and W+ = p(B?2+B?2+B?3+B?3) < W? = 0. Figure 3 summarizes the central results and is used to motivate the applications of the model in the ensuing subsections. The profession, if used, promotes the general ? , and only detracts from general welfare in a small part of the region welfare for all ?<? ? . But it is only used if it is to the politician's advantage (V+?V?). At high where ?>? levels of either democracy or stability, the politician invokes professional review. As either democracy or stability fall from their highest levels, they become substitutes. At many levels of democracy, the profession is used only if political stability is high enough. It is because professional review solves the informational problem that there this trade-off between democracy and stability.
26
1.4.4 Professions Under Autocracy and Transitions to Democracy Raising the status of the profession is consistent with autocracy, so long as there is also a high level of stability. When democracy is low and stability is high, an absolute autocrat who brooks no independence would still use the profession (V+>V?). France under the old regime provides an outstanding example. During the century of domestic political stability initiated by Henri IV (1589-1610), the French legal profession rose in status and power. With Louis XIV (1643-1715) the very symbol of absolute monarchy, one would not associate his era with the development of decentralized institutions. His treatment of the Huguenots showed that independent views were not to be valued for their own sake. But France was very stable, and independent entities could provide a useful service to the monarch. The first bar association, the Order of Barristers, was formed in 1660 and won a monopoly over a range of legal activities while controlling recruitment and discipline. "In the second half of the 1660’s, Louis XIV issued his great ordinances on civil procedure and criminal law, and this overhaul of large parts of the French legal code put a premium both on legal expertise, and on the sort of back-breaking intellectual labor that barristers prided themselves on. Chancellor Lamoignon himself convened a special committee of barristers to work on the projects.…mere barristers sat at the same table with the chief magistrates of the parlement. After the project had been completed, the king himself thanked the Order for its services (Bell 1994, pp. 55?6)." Our model's predictions are consistent with the observation that the prestige of the French legal profession stood high in the early part of the 18th century. De Tocqueville (1955, pp. 115-6) commented about these times that: "For though as far as administration
27
and political institutions were concerned France had succumbed to absolutism, our judicial institutions were still those of a free people." Had revolution occurred at this time, when the legal profession had gained some power and was politically neutral, perhaps the evolution of the rule of law in France would have taken a different course than it did. But the strength of the legal profession declined over the 18th century. We postpone discussion of this until Section 1.5, when the introduction of professional political power facilitates comparison with English developments. The role of lawyers under Soviet rule and in the transition can also be elucidated with our predictions on the use of a politically powerless profession. While the communist revolution in 1917 destroyed the pre-revolutionary Russian legal profession (Shelley 1991, pp. 65?6), the return of stability under Stalin was accompanied by a recognition that lawyers would play a role in the Soviet state (Waters 2004, p. 41). Elements of the legal profession had a degree of independence not observed elsewhere in the Soviet system (Shapiro 1961). Stalin allowed the advokatura (defense lawyers) to form a "British barrister-style set of collegia or lawyers’ self-controlled practice groups, self-funding and attached to each major court….private, petit bourgeois 'mini-guilds' within the state structure" (Krause 1991, pp. 16?8), which were subject to far less state control than other types of collectives (Shapiro 1961). Even in the Stalinist state, "professional pride emerged among many lawyers, as they had a claim to expertise based on specialized knowledge" (Shelley 1991, p. 66) and their influence and autonomy increased over time (Krause 1991, p. 18). The advokatura provided a rare example of a profession that was formally outside all administrative structures and, within the narrow limits set by the Soviet state, self-governing (Huskey 1982).
28
The Soviet legal profession did affect policy in the manner depicted in our model. One well-documented case is that of criminal law and criminal legal-procedure reform in the years of destalinization (McCain 1982). In this process, legal scholars and practitioners were invited to debate the many proposals that surfaced after the Communist Party announced that new legal codes would be adopted. This resulted in exactly the process of delay envisaged in our model, a two-year "elaborate ritual of consultation with legal specialists" (McCain 1982, p. 6), during which lawyers played important roles in the process of drafting and re-drafting the proposed codes. "Because the political leadership needed the technical skills of lawyers in producing new criminal and procedural statutes, the drafting process was designed to involve Soviet jurists, and in many instances they predominated on the legislative committees." (McCain 1982, pp.1920). The special role played by sub-groups of the legal profession continued throughout the Soviet period, including involvement of the advokatura itself in the lengthy review of drafts of the 1979 Law on the Advokatura (Huskey 1982, p. 204). Although the role of the legal profession in Soviet society was minor even compared with its role in medieval England, the existence of any professional influence in a totalitarian state is surprising, but consistent with our prediction that an unchallenged dictator can find professions useful. The Soviet Union before Gorbachev could be placed at point S in Figure 3. The role of the legal professions was enhanced under Gorbachev, with lawyer groups influencing the development of criminal law, playing a more central role in legislative drafting, and gaining further tools of self-regulation (Jordan 1998). The Union of Advocates of the USSR was established in early 1989, the first independent union to
29
be formed since 1917 (Jordan 1998, p. 770). Because the early Gorbachev years were a time when democracy rose faster than instability, this is consistent with the model's predictions. In contrast, in early transition the continued rise of democracy was accompanied by increases in political instability. In Russia in the 1990s, the power of the advokatura declined: they sought the state's help in warding off new competition rather than increasing their own self-regulation (Jordan 1998). Their status still remained governed by a Brezhnev-era law. Thus, in the instability of the early Yeltsin years, the rise in democracy was not sufficient to encourage politicians to enhance the role of that part of the legal profession that had been most independent under Soviet rule.34 The developments were even more stark in the new republics, such as Georgia, where instability dominated. Legal groups lost power and lawyering became a free for all. Politicians paid little attention to the profession and it was scientists and literary figures, and not legal experts, who dominated the process of drafting a constitution (Waters 2004, pp. 57?8). This pattern of events is consistent with our model's predictions. While the Soviet regime found some use for the legal profession, the higher levels of democracy of the early 1990s did not translate into greater use of the profession, because they were more than countered by the increases in political instability. Early transition in Georgia, and perhaps even in Russia, is at G in Figure 3. In Poland, where much higher levels of democracy were quickly attained and instability declined, the legal profession was prominent. Law professors and legal
34
As is typical of civil law countries, the legal profession in Russia is divided into separate groups. Some elements of the legal profession did attain higher status under Yeltsin, for example judges in the arbitrazh courts.
30
professionals were regularly appointed to task forces that were charged with drafting new laws (Gostynski and Garfield 1993). The use of recognized legal experts to review proposed laws delayed economic reforms (Rich 1997). The higher levels of both stability and democracy would place Poland at point P in Figure 3. In Georgia, it was only in the late 1990s, with the appearance of greater political stability, that lawyers played a more important role. Lawyer groups were now chosen to carry out legislative drafting projects and were able to play a role in pressing the government to pay more attention to the rule of law (Waters 2004, pp. 100, 168?9). The interpretation of these historical developments provides insights into why the early years of transition to democracy are so important. Whether civil society groups, such as professional organizations, benefit from transition depends upon the balance between changes in democracy and instability. When the rise in democracy more than counters the increase in instability, as in Poland, politicians enhance professional status and the profession builds upon any prestige it held under the old regime. When the rise in instability dominates, as in Georgia, politicians ignore the profession, eroding any professional power inherited from the old regime. The early years of transition are therefore crucial: the maintenance of stability as democracy rises can lead to reinforcement of the role of the profession and improved institutional reforms. This increases the political power of the profession. Then, as the next section shows, the enhanced political power can encourage the politician to use the profession when the country is at a point such as G in Figure 3. These interactions increase the degree of path dependence in transition.
31
1.5 When Do Politicians Use Professions that Have Political Power? Often professions are entrenched and a politician bears significant extra political costs in removing a profession from its advisory role. These political costs increase with the level of democracy. A powerful autocrat can purge a profession at whim. For example, within one year of ascending to power, Hitler abolished the German lawyers association (Jarausch 1990, p. 118). In contrast, professions can vanquish even the most determined democratic politician. In the late 1980s, the Thatcher government retreated from an attempt to reform the English legal profession and reduce its autonomy (Burrage 1997, pp. 137?40). To capture the fact that professional power is more effective in a democracy, we assume that the political costs of removing the profession from its advisory role are c? (c>0), which is subtracted from the politician's payoff in A5. c quantifies the political strength of the profession: c is smaller where the profession is new or where it is a tool of the government. This simple functional form is chosen for ease of exposition. None of the results depend upon it.35 The politician would like to keep the profession in its advisory role if at the prevailing (p,?), V??V+ ? c? and would like to remove the profession when V??V+ > c?. For a given (p,c), define ?T(p,c) as the least value in [0,1] such that for all ?>?T(p,c), the politician keeps the profession. Then, since V??V+ is non-increasing in ? at any given p, the politician removes the profession if 0 < ? < ?T(p,c). The following proposition
35
For example, all of the results hold when the political costs are constant, independent of democracy, or when political costs are an increasing function of ? and the first derivative of political costs with respect to ? increase with the profession's political strength. Even these assumptions are far stronger than necessary.
32
establishes that a unique ?T(p,c) exists for each (p,c) and characterizes the effect of p on ?T: Proposition 4. For a given c?0, (i) ?T(p,c) is a function with points of discontinuity at p = ??A/B?2 and at p = ??A/B?3. (ii) ?T(p,c) is constant for all p?(0, ??A/B?3) and attains its maximum in that interval. (iii) ?T(p,c) is decreasing and convex in p for p?[??A/B?3,??A/B?2). (iv) ?T(p,c) = 0 for p???A/B?2. The next proposition shows how changes in the strength of the profession, c, affect outcomes: Proposition 5. If c1>c2?0, then (i) For p<??A/B?2, ?T(p, c2) > ?T(p, c1) > 0. For p<??A/B?3, ?T(p, c2) ? ?T(p, c1) > 0 does not vary with p. For p?[??A/B?3,??A/B?2), ?T(p, c2) ? ?T(p, c1) > 0 is decreasing in p. (ii) For p???A/B?2, ?T(p, c2) = ?T(p, c1) = 0. Figures 4 and 5 summarize these two propositions, highlighting the interactions between stability, democracy, and professional power. At high levels of stability, professional power does not change the politician's decisions (and consequently aggregate welfare): the profession is used even by autocrats facing a weak profession. As levels of stability decrease, democracy becomes more important: ? affects political decisions and there is a widening of the interval of ? in which professional power makes a difference. This result appears clearly in Figure 5, which is Figure 3 modified to show the set of (p, ?) for which professional political power makes a difference.36 In the shaded area in Figure 5, professional power increases the level of social welfare by encouraging the politician to keep an institution, professional review, that improves institutional reforms for society as a whole, but not for the politician. When the strength
36
Figure 5 has the c2 of proposition 5 equal to zero. Note that ?T(p,0) lies along the solid line in Figure 3.
33
of the profession is higher, the level of democracy at which the politician chooses the socially optimal outcome is lower. The strength of the profession substitutes for weak democracy.37 1.5.1 Comparative Legal Development in England and France: 1688 and 1789 These results provide insights into why the revolutions of 1789 and 1688 had such different consequences for the role of law in the two countries. The French legal profession's high status, achieved under Louis XIV, continued into the reign of Louis XV, when the Jansenist movement confronted the monarchy. As the Jansenists fought their battles in the courts, barristers played key roles by using factums, written briefs that escaped censorship (Bell 1997, pp. 74?8): as in our model, the profession leads to deliberation and delay. The regime chose to do no more than denounce the Order as "a sort of absolutely independent little republic at the heart of the state" (Bell 1994, p. 67). To illustrate using Figure 5, France could lie at F1 under Louis XIV, where the unchallenged autocrat uses a powerless profession, and at point F2 under Louis XV: a rise in instability alters the monarch's view of the profession but not the outcome, since the profession has acquired political power. Matters changed as the lawyers began to have wider political ambitions. In the 1750's, a long-standing conflict between the parlements and the crown degenerated into full-scale constitutional conflict: magistrates, advised by Jansenist barristers, demanded a greater role in the legislative process, while the king wanted them to play a purely judicial role (Bell 1997, p. 83). The earlier political success of the barrister pamphleteers
37
This conclusion does not depend at all on the assumption that the costs of removing the profession are c?. Exactly the same conclusion would apply if the costs of removing the profession were completely independent of levels of democracy.
34
attracted imitators who did not share the same political and professional views and chose to support the crown, attacking the idea of an organized legal profession. The Order was fatally weakened. Louis XV could now respond by attacking the legal profession in general. The Order’s monopoly and disciplinary powers were abolished in 1771. The old parlements were replaced with trusted royal appointees. Although Louis XVI recalled the old parlements, the organized legal profession did not recover its former status (Bell 1997, pp. 90?5). In contrast to earlier in the century, when Louis XV chose not to remove a troublesome profession, the profession's power was lower. In contrast to a century earlier, when Louis XIV used the profession and allowed it to attain power, political instability was higher making the profession less valuable. France under Louis XVI can still be viewed as lying at F2 in Figure 5, but since the profession weakened itself and lost its power, the solid upper line is now relevant, not the dashed lower one: the monarch could do what was too costly before, remove the profession from its role. When 1789 came, the Order of Barristers had no authority (Bell 1997, pp. 90?5). A year later, the National Assembly formally abolished the Order and the French legal profession virtually vanished (Burrage 1989, p. 330). Consistent with our model, this brief period of higher democracy did not result in increased use of the legal profession: politicians were neither forced to use a politically weak profession nor saw advantage in it at a time of great instability (e.g., point F3 in Figure 5). Napoleon used prerevolutionary judges and practitioners to draft his civil code and re-established the Orders in 1810, showing that a return to stability, even after a decline in democracy, could lead to a reinstatement of the profession (e.g., back to point F1). But he instituted a very
35
different legal profession than before. In contrast to earlier times, Napoleon's legal profession was his own creation and remained under strict state control (Bell 1997, p. 99; Karpik 1999, p. 120; Bell 1994, pp. 214?5). From that time on, any independence attained by the French legal profession was given by government, not won from the state. The English legal profession had substantial power already in the middle ages and asserted itself more strongly than ever in the early 1600's. In 1640, at the beginning of the Puritan revolution, Parliament established the supremacy of the common law courts over the crown's prerogative courts, bolstering the power of the legal profession. However, the Puritan revolutionaries who were able to remove a king's head were unwilling to go as far as pressing for popular legal reforms over the objections of the legal establishment: "…the power of the English lawyers, which confounded Cromwell, derived from a solidarity that transcended revolutionary enthusiasms…" (Burrage 1989, p. 356). Later, in the 1680's, the legal profession was prominent in the struggle against the crown over the independence of chartered bodies and other "lesser governments", essentially a civil war fought with law rather than the sword (Landon 1970, pp. 100?1). One of those lesser governments was the Inns of Court, the legal profession's stronghold. Because the Inns' independence was based on custom, not charter, the crown pressured the Inns by forcing royal appointees onto their governing bodies. The Inns resisted (Burrage 1997, p. 150). James II even faced a shortage of lawyers willing to fight his battles. In March 1686, the attorney general and the solicitor general both refused to implement royal measures that would have flouted common-law precedent (Landon 1970, pp.191). In 1688, in the trial of the Seven Bishops, often taken to signify the
36
beginning of the Glorious Revolution, the crown faced an almost united legal profession, including eminent Tory lawyers, and lost. At issue was whether the royal prerogative could negate the common law, as interpreted by the legal profession (Landon 1970, Chapter 6). In sum, England in the years preceding the Glorious Revolution could be thought of as being at E in Figure 5: the King had every reason to remove the legal profession from its role, but was unwilling to pay the political cost of doing so by extra-legal means. Of course, James' battle with the legal profession was only one element of many contributing to the final outcome. But James did not succeed in weakening or even dividing the legal profession: the Glorious Revolution entailed the triumph of commonlaw lawyers and their interpretation of the law (Landon 1970, p. 248). In the construction of the post-revolution settlement, lawyers played a fundamental role, even in the House of Lords where the Tory establishment was strong. Exactly as in our model, the legal profession was central in the debate because the settlement had to be made consistent with 500 years of common law. There could not be a greater contrast between the fate of the English legal profession in 1688 and that of the French profession in 1789. The English profession had been in existence for four centuries and the removal of its autonomous institutions and practices would have come at great political cost. The French profession weakened itself during the eighteenth century and was easily removed by a monarch who no longer regarded it as useful. The English legal profession played a major role in the post-1688 settlement, improving institutional reforms. The French profession was abolished a year
37
after the Bastille fell and played no role in institutional development. The consequences of these divergent histories are with us still. 1.5.2 The Professions Under Civil and Common Law A burgeoning empirical literature examines the historical roots of institutional quality, particularly focusing on legal and colonial origins (La Porta et al. 1998; Acemoglu, Johnson, and Robinson 2001). In this section and the following one, we show how an emphasis on organized legal professions can be complementary to both of these approaches. Moreover, as our model offers a new lens to interpret existing results, we shed light on empirical findings that have been difficult to explain within existing approaches. Although there is substantial evidence that legal origin affects institutional quality (Djankov et al. 2003), there is still uncertainty about the precise mechanisms that make legal origin matter (La Porta et al. 2004; Beck et al. 2003b). La Porta et al. (1998) and La Porta et al. (2004) suggest two complementary reasons: both the quality of legal rules and the strength of judicial checks and balances vary across legal systems. Rajan and Zingales (2003) challenge the evidence on quality of legal rules, showing that commonlaw countries had lower levels of financial development in the beginning of the twentieth century. They propose a different reason for the varying properties of different legal systems: private interests are more likely to see their agenda enacted in civil law countries because laws emanate from the center rather than evolving through legal process. New policies are enacted more quickly in those countries, sometimes leading to better outcomes, sometimes worse, depending on the configuration of interest groups.
38
By introducing a different causal mechanism, our model suggests a reconciliation of La Porta et al. (1998) and Rajan and Zingales (2003). A strong, independent legal profession coevolved with the common law system, exactly because decentralized and autonomous legal institutions are intrinsic in that system. In contrast, state-sponsored, state-controlled professions are consistent with a civil law system.38 Since the use of the profession improves the quality of institutional reforms and since the strength of the profession can be critical in determining its use by the politician, our model predicts that common law countries tend to have better legal rules than civil law countries, consistently with La Porta et al. (1998). Moreover, as delay is intrinsic in the use of the profession, reform will be quicker on average in civil law countries, and the government responds faster to interest groups, consistently with Rajan and Zingales (2003). La Porta et al. (2004) show that judicial independence is associated with economic and political freedom and that the higher level of judicial independence in common law countries accounts for some of the effects of legal origin on economic freedoms. Even though the current incarnation of our model does not focus directly on judicial independence and checks and balances, the mechanisms depicted in our model could well account for the results of La Porta et al. (2004). Civil law legal professions are weaker, more dependent on the state, and balkanized, with judges and private lawyers organized separately. Organized legal professions in common-law countries are more likely to have the incentive and the power to protect judicial independence. The professional review depicted in our model has some similarities to judicial checks and
The difference between the profession’s political power in common and civil law countries has been widely recognized in the sociological literature (Halliday and Karpik 1997, pp. 5, 354?5; Abel 1995, p. 3; Burrage, Jarausch and Siegrist 1990, pp. 219?20; Siegrist 1990, pp. 181?2; Collins 1990, p. 16; Cleaves 1987, p. 10).
38
39
balances, delaying the implementation of legislation through a process that examines its consistency with existing legal doctrines. If this analogy holds, then our model would predict that politicians would use judicial checks and balances more in common law countries and that this choice results in better institutions, consistently with the results of La Porta et al. (2004). Our model also identifies when legal origin matters most, and by doing so resolves some apparent empirical anomalies. At the lowest levels of democracy, neither common nor civil law countries adopt professional review (unless stability is very high). When the level of democracy is high, both common and civil law countries use the profession.39 Therefore, legal system matters most at middling levels of democracy, when the politician's decision hangs on the political cost of removing the profession from its advisory role. In this range of democracy, common-law politicians are more likely to use the legal profession to act as a filtering device, with a resultant increase in the quality of institutional reforms. Legal origin has no effect at either the highest or the lowest levels of democracy. One recurring empirical observation is that France itself and the German-legalorigin countries do not suffer as much from the disadvantages of the civil-law system as expected from the results on civil-law countries as a whole (Beck et al. 2003b). Our analysis provides a possible explanation: France has a high level of democracy relative to French legal-origin countries in general and all the German legal-origin countries have high levels of democracy or stability (often both). In these situations, the political power
Our conclusion that common and civil law systems converge at highest levels of democracy is therefore in accordance with Glaeser and Shleifer (2002) analysis of circumstances under which the decisions made under bright line rules, predominantly characteristic of civil codes, and independent juries overlap.
39
40
of the profession makes less difference and our model predicts similar outcomes for civiland common-law countries. Berkowitz et al. (2003) have challenged the importance of legal origin, arguing that the crucial issue is whether legal transplantation was receptive or unreceptive. A key aspect of receptive transplantation is that the country contemplates the consequences of adopting a foreign legal system. Unreceptive transplantation occurs when the legal profession is so underdeveloped that laws are implemented without considering their contents (Berkowitz et al. 2003, p. 180). Hence, receptive transplants are more likely to occur when an organized legal profession undertakes review of reform proposals. This is consistent with our approach, but our model adds an extra element, showing why receptive transplants are more likely to occur in common law countries, which use the professions more often. Hence, our approach shows why results on legal origin are less strong when form of transplantation is included in the same analysis (Berkowitz et al. 2003): mode of transplantation is a reflection of legal origin itself.40 1.5.3 Colonies and Institutions Acemoglu, Johnson, and Robinson (2001) argue that institutions supporting longterm growth were more likely to be established in those colonies where Europeans settled in larger numbers, with consequences persisting until today. Their reasoning focuses on the demand side of politics: dominant political groups demanded different institutions in settler colonies than in extractive colonies. The key variable driving their evidence is
Glaeser and Shleifer (2002) argue that transplantation might be particularly unsuccessful for civil law systems because of their reliance on bright-line rules, which need to be modified for a different setting. Given that the organized legal profession will be important in such modification, the two views are complementary.
40
41
hospitality of settlement, extractive colonies appearing where settlement was less attractive for Europeans. Our model has supply-side implications that are consistent with the evidence. In those British colonies where settlement was unattractive, the demand for legal services was met by temporary labor. The colonial administration became the patron of the professions and did not find it advantageous to develop indigenous talent (Johnson 1973). Under such circumstances, "…there was no viable basis for independent professional practice nor was there a secure foundation for the emergence of professionalism…Under these conditions the technically-based authority of the professions was subordinated to extra-professional sources of power." (Johnson 1973, p. 289) These characteristics persisted into the post-independence period. In Kenya, for example, the legal profession was controlled by non-African lawyers and was fragmented during and after the colonial period (Odenyo 1979; Ross 1992). With both democracy and the profession weak, post-colonial governments impeded the growth of an autonomous legal profession (Ross 1992). In India, the profession was a British tool with no indigenous roots. The colonial authorities, apprehensive that members of the legal profession might lead the struggle for independence, repressed autonomous Indian lawyers’ associations. Consequently, a strong, autonomous profession did not emerge in modern India (Gandhi 1988). Australia, a settler colony, provides a remarkable contrast. The legal profession descended directly from British roots and developed remarkable autonomy and power. Key elements of the modern professional structure were in place before colonial selfgovernment. Therefore in contrast to most African and Asian colonies, post-
42
independence Australian governments faced a legal profession with much power to control legal practice, including admission and discipline. Although under political attack several times in both the nineteenth and twentieth centuries, Australian lawyer associations have generally been the victors (Weisbrot 1988). The Australian Law Reform Commission (2003), on which the legal profession is dominant, matches the advisory role played by the profession in our model. The colonial heritage of professional political power had significant effects.41 Lastly, our model can account for interactions between the effects of legal and colonial origin. Beck et al. (2003a) identify two such interactions: first, the effect of settler mortality on financial development is stronger for English legal origin countries than for French legal-origin ones; and second, the effect of legal origin on financial development is greater in countries where settler mortality is low. Our model predicts the first result because of the different strengths of professions in the two sets of countries: higher levels of migration by professionals are more likely to translate into an influence of the organized profession on reforms in English legal-origin countries than in those with French origins. The second result follows because organized legal professions did not develop the countries where settlement was unattractive for Europeans, implying that the effect of legal origin would be diminished in such countries. 1.6 A Postscript on Further Research This chapter has argued that organized legal professions provide an example of an institution that has survived in a Hayekian trial-and-error development process. The
41
Note the consistency with Berkowitz et al. (2003), who argue that migration is a key to receptive transplantation in colonies.
43
argument should apply to other organized professions that have similar characteristics, but we leave it to future research to determine to which professions our analysis applies. We have stressed the importance of the review of proposed legislation by an organized profession. Since in our model this review occurs before the passage of legislation, we have chosen not to place judicial checks and balances as an element of the professional review. However, there are obvious similarities between the review of proposed legislation against a set of legal standards promulgated by an organized profession and review of the implementation of legislation by judges administering a body of law. Given these similarities, one obvious extension of our approach would be an examination of the processes of judicial checks and balances, which La Porta et al. (2004) have shown to be of substantial importance. The professions are one element in that set of arrangements that have come to be known as civil society (Bell 1997, p. 86). In policy circles, it is an article of faith that promotion of civil society contributes to economic development (Carothers 1999). Yet, fostering civil society sounds suspiciously like promoting interest group development, and there is much in past economic research to suggest that interest groups slow economic development. As another extension of this essay, an analytical examination of how civil society can be beneficial would offer guidance for the conduct of practical development policy.
44
Chapter 2 Allocating Law-Making Powers: Self-Regulation vs. Government Regulation
2.1 Introduction Governments employ a variety of institutional arrangements to regulate and intervene in the economy. Self-regulation is a result of "deliberate delegation of the state’s law-making powers to an agency, the membership of which wholly or mainly comprises representatives of the firms or individuals whose activities are being regulated" (Ogus 1999, p. 590). Many sectors and activities self-regulate.42 Moreover, there are episodes when self-regulatory policies are initiated on an economy-wide basis.43 This chapter examines when self-regulatory regimes arise, and when they are socially efficient relative to regimes that consolidate regulatory powers at the governmental level. Despite the ubiquity of self-regulation and the attention it attracts in government policy circles, formal analyses of the economics of self-regulation are scarce and, in the aggregate, inconclusive about its welfare properties.44 Leland (1979) and Shaked and
42
E.g., professional and financial services, sports, advertising, internet, education, insurance, and press. Likewise, implementation of regulatory goals in the context of environmental policy is often delegated to industries. Self-regulatory arrangements differ in the extent of the role played by the self-regulators, selfregulators' degree of autonomy from the government, and legal force of their rules. Voluntary selfregulation is exemplified by the proliferation of industry codes of conduct. Occupational licensing and certification are examples of statutory self-regulation. Under supervised self-regulation, a self-regulatory organization is subject to continuous oversight by a specialized government body, as when the SEC supervises the NASD. See Priest (1997), Baldwin and Cave (1999, Chapter 10), Ogus (1999, pp. 587?8) for a comprehensive discussion. We discuss two prominent historical examples, fascism and New Deal.
43 44
See, e.g., http://www.selfregulation.gov.au, http://www.selfregulation.info, http://www.easa-alliance.org, http://medt.strategypartner.ru/eng/selfreg.htm for recent policy-oriented discussions on self-regulation. The formal literature reviewed here typically focuses on a narrower question of functioning of selfregulated markets rather than analyzing economy-wide delegation of regulatory authority and institutional transformation, associated with a self-regulatory regime (Eisner 2000).
45
Sutton (1981) stress the undesirable anti-competitive effects of self-regulation.45 In contrast, Kranton (2003) and Lyon and Maxwell (2000) argue that self-regulation may improve social welfare.46 But foremost, this literature sidesteps explicitly comparing self-regulation with its alternative, direct government regulation. Two key questions for successful regulatory design are thereby left unanswered. First, under what circumstances, if any, does self-regulation welfare-dominate its alternative, government regulation? And second, what factors influence the government’s choice of centralization or delegation of regulatory authority?47 In the spirit of Gehrig and Jost (1995), we argue that the source of potential efficiency gains from implementing a self-regulatory regime is the producers’ superior knowledge about the regulatory issues at stake.48 However, we depart from the existing
45
Building on Akerlof’s work, Leland (1979) shows that licensing, by setting minimum quality standards, improves the welfare-properties of equilibria in markets with asymmetric information. Professional (or industry) self-regulation, by exerting monopoly-like behavior, however, is expected to result in these standards set too high. This holds at least when the opportunity costs of supplying quality increase with the latter. When the opportunity costs decrease with quality, however, Leland shows that licensing standards set by producers themselves may not be too restrictive. Shaked and Sutton (1981) similarly show that granting a profession monopoly rights is welfare reducing. By self-regulating, the profession shrinks to a size that is socially sub-optimal. Kranton (2003) illustrates that industry self-regulation, with correspondingly restricted competition, sanctioned by the government, may be necessary to preserve incentives for high-quality production and therefore higher social welfare. Maxwell and Lyon (2000) argue that industries voluntarily adopt environment-friendly processes to preempt government regulation. Doing so must increase firms' profits, and may increase consumers' welfare as well when consumers' costs of influencing governmental regulatory policy are taken into account.
46
47
Donabedian (1995) and Ashby et al. (2004) offer a rationale for the observed differences in regulatory arrangements, but they are silent on their welfare properties. Donabedian (1995) shows that allowing for professional self-control (i.e. self-regulation) as a means of solving the enforcement problem might be a politically feasible alternative to government-imposed punishment in occupations with high, rather than low, specificity of human capital. Ashby at al. (2004) argue that individual firm compliance generates both private and public (to the industry as a whole) benefits. Self-regulation is then sustainable in industries that are more likely to overcome the free-rider problem and where governmental tolerance towards noncompliance is sufficiently high.
48
Gehrig and Jost (1995) is to the best of our knowledge the only account in the literature that evaluates the welfare properties of producer self-regulation by explicitly contrasting it with direct government regulation. (In his analysis, Leland (1979) implicitly assumes that, when not set by the profession itself, the minimum quality standards are set by a benevolent regulatory agency.) Gehrig and Jost (1995) consider a set-up with asymmetric information between consumers and producers about product quality, and introduce mobility of
46
literature by investigating a set-up inspired by property rights theory (Grossman and Hart 1986). Although property rights theory has primarily focused on the optimal allocation of control rights within firms49, we argue that it provides a natural framework to address the question of the allocation of regulatory authority. A regulator’s residual right to set rules with the power of law confers authority much as ownership rights over assets do when contracts are incomplete. A policy-relevant theory of regulatory institutions has to acknowledge the difficulties in writing complete regulatory contracts. Mostly, these contracts are confined to assign authority over specific regulatory decisions, only roughly specifying the rights and obligations of the regulators (Goldberg 1976, Williamson 1976, Estache and Martimort 1999).50 The distribution of regulatory rights then critically shapes the regulatory outcomes. This chapter contrasts the implications of two different assignments of regulatory rights. We refer to government regulation as a regime where the authority over implementation of regulatory decisions is centralized at the governmental level, and regulatory powers assigned to governmental regulatory agencies or ministerial
consumers across markets, thereby allowing for reputation effects. In their model, producers have an incentive to self-regulate by jointly committing to high quality in environments where consumer mobility is high relative to industry monitoring costs and market fragmentation. Self-regulation, if viable, welfare dominates government regulation only if the firms' profits are assigned a sufficiently high weight in social welfare. Gehrig and Jost show that without the firms' informational advantage, government regulation could always mimic self-regulatory actions, at least when government's monitoring costs are ignored.
49
Luelfesmann (2000), an exception, applies the Grossman and Hart (1986) framework to investigate optimal governance structure in the context of fiscal federalism.
50
"At the risk of oversimplification, regulation may be described contractually as a highly incomplete form of long-term contracting…" (Williamson 1976, p. 91). Yet interestingly, the theory of incomplete contracts, first explored by Grossman and Hart (1986), has not had much impact on the study of regulation. See Lyon and Huang (2002) for an exception and discussion.
47
departments.51 Alternatively, the regulatory authority can be delegated to producers ? industries or professions, represented by self-regulatory organizations ? whose activities are to be regulated.52 We call such regimes self-regulatory. We focus on examining the implications of different allocations of regulatory authority, and therefore abstract from incentive and accountability problems due to the complexity of internal governmental organization, as well as potential agency problems between self-regulatory organizations and their members.53 Methodologically, this chapter falls within a growing body of literature on comparative analysis of institutional arrangements (see Djankov et al. 2003), which has raised, or freshly illuminated, many important questions in the economics of institutions in the last decade.54 Complex economy-wide institutional structures are not easily examined with analytical models derived from first principles. Therefore, this literature is characterized by the use of reduced form models, motivated by disparate historical examples and available data. The chapter is organized as follows. In Section 2.2, we develop a model of a regulatory process with bargaining where the authority to amend the enabling legislation
51
There are important differences in the way state exercises its regulatory powers through a central government department and a regulatory agency that we neglect in our analysis. As opposed to a ministerial department, regulatory agencies are by a parent statute typically independent from the central government, yet nevertheless under close political oversight. Through the power of key appointments and overall direction of regulatory agencies, the central government maintains a significant portion of its regulatory authority. See Chapter 5 in Baldwin and Cave (1999) for a thorough discussion. We ignore potentially important institutional details of various self-regulatory arrangements. See e.g. Ayres and Braithwaite (1992) for discussion.
52
53
See, e.g., Estache and Martimort (1999) and Tirole (1994) for discussion of organization of government. See Nunez (2001) and DeMarzo at al. (2001) for analysis of agency problems between an SRO and its members. See, for example, Shleifer and Vishny (1994), Glaeser and Shleifer (2002, 2003), Glaeser, Shleifer and Johnson (2001), Che and Qian (1998), Qian (1994), Greif (1993), Roland (2000), Acemoglu and Robinson (2000), Dewatripont and Maskin (1995), Kranton and Swamy (1999), and many others.
54
48
can be either consolidated within the government or extended to the producers in a selfregulatory regime. In Section 2.3 we delineate the welfare implications of regulatory regime choice, and indicate whether the government’s incentives to delegate or centralize regulatory authority lead to efficient institutional design. Section 2.4 illustrates the model with a range of examples: we motivate the changes in regulatory practice between the progressive era and the associational regime of the New Deal, characterize the mechanisms of intervention used in fascist economies, and identify those features of legal traditions that help to explain variation in regulatory arrangements across countries. Section 2.5 concludes. 2.2 A Simple Model 2.2.1 The Setup The government and regulators are implementing rules to regulate an economy consisting of producers and consumers (the latter standing for the general public). We assume some form of regulation is desirable on economic grounds. The final set of rules results from a process in which the enabling legislation for the particular regulatory arrangement is subject to refinement and interpretation as implementation proceeds. Additional, typically administrative, rules formulated during implementation will have the power of law. We assume that the legal rules adopted in the regulatory process can be characterized through their effect on a variable, L, defined on the real line. L could be, for example, an index of legal complexity, with higher values implying greater legal complexity.55 Under the status quo, L = 0. The desirability of regulation follows from
55
We implicitly assume there exists a mapping of legal rules (e.g. minimum educational requirements for professions) into values of the index of legal complexity, L.
49
assumptions given below, which ensure that some positive values of L are preferred by everyone in the economy to L = 0.56 At t = 1, when there is a degree of consensus that a problem needs to be solved, the government passes the skeletal (or enabling) regulatory legislation. We assume the latter results in L = L*. The enabling legislation will normally be rather vague about the specifics of attaining regulatory goals (Eisner 2000, pp. 13?5) and will be passed with an understanding that the sketchy legal rules will change during implementation.57 There are many reasons why such changes happen. Some occur because the very nature and complexity of the law-making process makes the outcome of legislation unpredictable.58 In addition, shifts in the politico-economic environment, perhaps causing institutional transformations, stimulate revisions and reinterpretations of existing laws.59 Whatever the specific reason for adjustments, they are an inevitable product of the fact that enabling legislation is formulated when there is much uncertainty about what its precise effects will be.
56
For example, lawyering in many republics of the former Soviet Union during the early transition was a free for all: legislation governing the newly established relationships between market participants was yet to be implemented (Waters 2004). This situation that was clearly undesirable for both lawyers and the general public. For example, the statutory framework for a profession might assert some minimum degree of training to be acquired, or environmental legislation might specify broad guidelines about adoption of environmentally friendly processes, or a law for financial markets might stress the need for some level of information disclosure. Since regulatory policies originate in legislation that defines policy goals in the broadest possible terms, it is the regulators that must give regulatory policies substantive content. An original aim might not be practical (e.g. new professional standards have to piggy-back on old elements of the existing educational system), or a previously hidden conflict with other legislation becomes apparent (e.g. financial disclosure requirements come into conflict with privacy legislation), or practical issues in measurement induce a different specification (e.g. the measurement of pollution targets depends on existing scientific equipment).
57
58
During the post-communist transition, massive institutional reforms were conducted, often with a trialand-error approach. The skeletal legislation necessarily had to be amended with supplementary rules. Implementation of rules regulating organized legal professions in Georgia, for example, involved a lengthy process of reformulations and adjustments of the newly established body of laws (Waters 2004).
59
50
We capture the passage from initial legislation to a set of regulatory rules in the simplest manner. If no further decisions were made in the regulatory process, the L implied by the enabling regulatory legislation would equal L = L*+? at t = 2, with random variable ? capturing all the uncertainties involved in institutional implementation. We assume that ? has mean 0 and variance ?2. While ? is unknown at t = 1, it is known by all for t ? 2. Therefore, the passage of the enabling legislation represented by L = L*, results inevitably, even if no other decisions are made, in a different L (= L*+?), which has different effects than would L = L*.60 The variance ?2 is a central parameter in our model. It is a measure of the degree of uncertainty of the effects of the enabling legislation at the time of passage of that legislation. It will be higher, for example, in those countries that are changing related institutions the quickest (where the structure of the future economy is harder to predict), in those sectors where regulation has not been undertaken before (where the enabling legislation is not simply the reform of existing institutions), and in those countries where law-making is decentralized and subject to interpretation (rather than centralized and under greater control of the designer). The effects of the uncertainty can be mitigated. At t = 3, regulators can choose to implement supplementary legal rules, which have the result of changing the effect that L = L*+? has on the economy. We call the regime where the government has the right to
60
Capturing this spirit of uncertainty and its resolution associated with the process of any institutional implementation, Dewatripont and Roland (1995), for example, analyze the outcomes of different reform strategies for economies in transition. In the context of the Georgian lawyers example examined by Waters (2004), L = L* could be interpreted as the legal complexity arising from provisions about some minimum level of quality of services specified by a governmental statute. As implementation of the enabling legislation proceeds, however, many previously unknown but relevant details pertaining to regulation of organized legal professions become apparent, altering the legal complexity implied by (and hence effects of) the skeletal regulatory legislation. At t = 2, therefore, L = L* results in L = L*+?.
51
set those supplementary legal rules government regulation (R). In contrast selfregulation (SR) is the regime where the authority over the supplementary legal rules is vested in bodies whose members are chosen by the producers themselves. As before for ease of modeling, we adopt a particularly simple form for this process: the supplementary legal rules directly modify L = L*+?, so that after their implementation L = L*+?+?L. ?L hence represents the change in L (from L*+? to L*+?+?L) resulting from adoption of supplementary legal rules.61 Any adjustment ?L of L from L = L*+? is costly to implement. These adjustment costs arise from the expense of establishing additional rules and institutions and complying with additional rules, as well as the implicit costs arising from the reallocation of resources. Aside from assigning authority for choosing ?L to different parties, government regulation and self-regulation differ precisely in the adjustment costs that they impose on the economy. Many reasons have been put forward why the cost of implementing additional legal rules, and adjusting existing ones, is greater under government regulation than under selfregulation.62 Compared to the government, the producers typically command a greater degree of expertise and technical knowledge of practices and innovatory possibilities within the relevant area. Self-regulatory arrangements are less formalized than public regulatory regimes and hence less rigid (public agencies often have to go through a long bureaucratic process before an old rule can be changed or a new one enacted).
For the Georgian lawyers example, once ? is known, at t = 3 supplementary legal rules, such as licensing and certification standards, and admission requirements, are chosen by either a governmental regulatory body or the self-governing profession. ?L is then the change in legal complexity from L = L*+? to L = L*+?+?L arising from implementation of these supplementary legal rules.
62
61
For a comprehensive discussion, see Baldwin and Cave (1999), Ogus (1999), and Priest (1997).
52
Information costs for the formulation and interpretation of new standards are therefore lower under self-regulation. Monitoring and enforcement costs are also reduced under self-regulation, as are the costs to the regulated of dealing with regulators.63 We build this into the model as follows. L2 is the cost to the economy (producers and consumers together) of implementing the enabling legislation amended by supplementary rules (L = L*+?+?L). In addition, amending enabling legislation per se is costly. The size of adjustment costs, however, differs under different regulatory arrangements. The adjustment ?L of L from L*+? to L*+?+?L costs the economy ?i(?L)2, i?{R,SR}, with ?R>?SR>0. ?i>0 then measures the rigidity of a regulatory regime. The total costs of L = L*+?+?L for the economy are therefore L2+?i(?L)2 (all incurred after ?L is determined).64 The producers' and consumers' preferred L's will typically not coincide: we assume that the groups differ only in how much they value L, not in terms of the costs that they incur through its implementation.65 For ease of modeling, we let the agents' net payoffs be quadratic. L = L*+?+?L then yields a payoff of pL?½[L2+?i(?L)2] to producers and cL?½[L2+?i(?L)2] to consumers. p > 0 and c > 0, respectively, represent
63
Unlike regulation governed by public authorities and dependent on tax income, self-regulation also does not cause a deadweight loss from taxation.
64
That all benefits and costs of a legal rule are realized at t = 4 only is merely a consequence of the review and deliberation to which new legal rules are subject in a normal political process. This implies that the producers who typically "pay" the adjustment and implementation costs can transfer ½ of these total costs onto consumers. This assumption certainly limits the scope of our analysis by abstracting from important effects of industrial structure. On the other hand, our analysis is primarily concerned with regulatory regimes (as opposed to more narrowly defined regulatory policies) that transcend the specific problem of a given industry (see, e.g., Eisner 2000, p. 3).
65
53
the producers' and consumers' preferred L's in the absence of legal rule adjustments.66 In general, p ? c.67 The difference between p and c captures the divergence of interests between producers and consumers. (p?c)2, for example, will tend to be high in a polarized economy with prosperous owners and a public resenting perceived or real monopoly power, but low when there is a high degree of consensus among the groups about the regulatory agenda. The government's payoff from L is a weighted average of producers' and consumers' payoffs, ?{cL?½[L2+?i(?L)2]}+(1??){pL?½[L2+?i(?L)2]} = AL?½[L2+?i(?L)2], where A ? ?c+(1??)p.68 ? is close to 0 (A is close to p) when the producers’ lobby has bought the government. ? is close to 1 (A is close to c) in the case of a populist government pandering to the economy’s consumers. Under R, it is the government that has the right to set ?L. Under SR, these lawmaking powers are delegated to the producers. But political bargaining is a pervasive feature of the relationship between the government and the industry (see e.g. Shleifer and Vishny 1998, Laffont 2002).69 Therefore, regardless of who has the authority to define supplementary legal rules, the government and producers may negotiate ?L. The
66
Observe also that we implicitly assume that aggregate uncertainty affects the benefits and the costs of producers and consumers in the same direction: increases in L increase both the value from, and the implementation costs of, L = L* for both groups.
67
A profession, for example, would want to limit the entry in the market more than is desirable by consumers (see e.g. Leland 1979), implying that the two groups prefer different L's. We assume that all government-specific costs of legal rule implementation are covered by taxes imposed on producers and consumers. In the post-communist Georgia, for example, the body of laws regulating organized legal professions was implemented through lengthy negotiations between the lawyers' associations and the government (Waters 2004). Political bargain between the government and the producers lobbies is, of course, a pervasive feature of politically less volatile environments as well.
68
69
54
consumers, as the general public, do not directly participate in negotiations (Olson 1971), but may exert pressure on the government (when ? is high). The government and the producers enter negotiations at t = 3. Since the government's and the producers' preferred ?L's implemented under the respective default scenarios differ (unless ? = 0), there are always gains from negotiating away from the default ?L. The bargaining process between the government and the producers results in a Nash-bargaining solution. The government and producers implement their jointly efficient ?L, irrespective of whether the authority over ?L lies with the government (R) or the producers (SR). Each party obtains its regime-dependent default payoff plus one half of the gains that arise from implementation of their jointly efficient ?L (rather than the respective default ?L).70 Figure 7 summarizes the timeline of events. Finally, we discuss our notion of social welfare. Producers and consumers count equally in this economy. We therefore define social welfare under regime i?{R,SR}, Wi, as the sum of the producers’ and consumers’ payoffs from the final realization of L under regime i: Wi = (p+c)Li?[Li +?i(?Li) ]. The transfers implied by the bargaining between the government and the producers are gains for one and losses for the other and hence not part of the social welfare calculation.71
2 2
70
For simplicity, we assume that the government and the industry have equal bargaining strengths. Renegotiation of the default ?L under any allocation of authority to implement amendments will involve exchange of transfers. Under R, for example, the initiative to renegotiate the default ?L will come from producers. In order to set a ?L different from the default ?L under R, however, the government will have to be compensated by producers. These transfers are payoff equivalents of direct monetary bribes or exchanges of political favors. Transfers per se therefore do not cause any resource misallocation due to rent-seeking, for example.
71
55
2.2.2 Solving the Model The model is solved straightforwardly which enables us to omit algebraic details in the presentation below.72 Working backwards, if no bargaining were to occur at t = 3, the default ?L under regime i?{SR,R}, ?Li,d, would equal ri ? L *i + ? 2 1 ? ? = argmax ?L ?ri L *i + ? + ?L ? L *i + ? + ?L + ? i (?L) 2 ? , ?L = 1+ ?i 2 ? ?
i,d
(
)
(
)
[(
)
]
where rSR = p and rR = A.73 Ceteris paribus, the more flexible the regulatory regime, the greater the adjustment ?L from L = L*+?. Since gains from renegotiation of ?L always exist, at t = 3 the government and the producers bargain to implement the jointly efficient ?L under regime i?{R,SR}: A+p ? L *i + ? 2 ?Li,b = 2 = argmax ?L (A + p) L *i + ? + ?L ? L *i + ? + ?L ? ? i (?L) 2 .74 1+ ?i
(
)
{
(
) (
)
}
Anticipating the bargaining outcome, at t = 1 (when ? is unknown) the government chooses L*i to maximize the expected value of the sum of its default payoff at t = 3 under regime i?{SR,R} and half of the gains from negotiation, implying L*i = A ? ?c+(1??)p. In our model, the (ex-ante) impact of the enabling legislation on the economy is independent of the regulatory regime in place, and fully reflects the government's
72 73
See Appendix 5 for derivations of all the expressions and results in the chapter.
The producers’ and the government’s t = 3 default payoffs for i?{R,SR} are then equal to p(L*+?+?Li,d)?½[(L*+?+?Li,d)2+?i(?Li,d)2] and A(L*+?+?Li,d)?½[(L*+?+?Li,d)2+?i(?Li,d)2], respectively.
74
The producers’ and the government’s t = 4 payoffs under bargaining for i?{R,SR}are then equal to p(L*+?+?Li,b)?½[(L*+?+?Li,b)2+?i(?Li,b)2] and A(L*+?+?Li,b)?½[(L*+?+?Li,b)2+?i(?Li,b)2], respectively.
56
weighting of the welfare of the economy's two groups.75 This is consistent with features of the regulatory implementation process. The enabling legislation is passed by the legislature, a fertile ground for the realization of the agendas of interest groups (Eisner 2000, p. 13). However, the skeletal enabling legislation merely opens the regulatory process, the effect of regulatory regime occurs when law is given its substantive content through implementation of additional legal rules. The outcome of the regulatory process is the final realization of L at t = 4. Under regime i?{SR,R}, the latter is a weighted average of A+?, the value of L implied by skeletal regulatory legislation as affected by uncertainty, and ½(A+p), the cooperatively desirable value of L for the government and producers: ? i ( ? + ?) + 1+ ?i ?+p 2 .
With R the more rigid of the two regulatory regimes, R places a relatively higher weight on A+?. Consequently, the expected social welfare76 under regime i?{SR,R}, Wi, equals
75
L*i equals the government's preferred legal rule (in the absence of amendments). The principal reason for this is that the government and the producers in our model care equally about the costs of legal rule adjustment (the ?i(?L)2 term) relative to the costs that arise purely from implementation of L (the L2 term). (This is in turn implied by the assumption that consumers and producers in our model differ only in how much they value L, but not in terms of the costs that they incur because of L.) To illustrate the intuition, take the no-bargaining scenario (or, equivalently, a scenario where one party possesses all the bargaining power). If the government cared about the adjustment (relative to pure implementation) costs more than the producers do, for example, then, under R, unable to predict ?, the government would have no incentive to set L* away from its preferred value (A). Under SR, in contrast, knowing that the producers have an incentive to adjust L*+? quite a bit toward p, the government would put a positive weight on p when setting L* (rather than simply setting L*=A). To make an analogy with the model of Grossman and Hart (1986), where different ownership structures lead to different distortions in ex-ante investments, in our model the selection of the ex-ante regulatory target embodied in the basic legal rule does not depend on the assignment of regulatory rights.
76
That is, social welfare as expected before the regulatory process begins (and hence before ? is realized) but after regulatory rights have been assigned. We refer to Wi as 'social welfare' in the remainder of the text for brevity. Same applies to the definition of Vi.
57
W i = A(p + c ? A) ?
?i 1 1 ?2 + ? ?(3? ? 2)(p ? c) 2 , 1+ ?i 1+ ?i 4
and the expected payoffs of the government under R and SR, respectively, are equal to
VR = 1 1 1 2 1 ?R A ? ? ?2 + ? ? 2 (p ? c) 2 2 1+ ?R 8 1+ ?R 2 1 2 1 ? SR 3 1 ?2 ? ? ? 2 (p ? c) 2 . A ? ? 2 2 1 + ? SR 8 1 + ? SR
V SR =
While aggregate uncertainty is always detrimental, greater flexibility of the regulatory arrangement in place moderates its impact. Before we examine these expressions we briefly discuss the inefficiencies that arise under the two competing regulatory arrangements. As a benchmark, define the first best (FB) final realization of L as that implemented by a social-welfare maximizing planner possessing all the expertise of the producers in the relevant area of regulation (?FB = ?SR) and able to resist negotiating with producers over ?L.77 Once observing ? at t = 3, the benevolent social planner amends the enabling legislation by choosing ?L. It can be easily shown that ?LFB = [½(p+c)?(L*FB+?)]/[1+?SR], L*FB = ½(p+c), and the final realization of L under first best equals ½(p+c)+[?SR/(1+?SR)]?. Comparing these with the expression for the final realization of L above, it is evident that R cannot implement the first best L primarily because of its rigidity in adapting legal rules, regardless of the value of ?. SR, on the other hand, fails at implementing the first best L because of the active participation of the regulated interests in the legal rule determination process. Explicitly
77
In contrast to the fictitious benevolent social planner, a self-interested government will not abstain from negotiations with the producers: since there are gains from renegotiating ?L, a self-interested government will always agree to renegotiate the default ?L in exchange for a transfer.
58
acknowledging that regulatory contracts are incomplete hence renders regulatory decision-making a second-best world.78 2.3 Comparing Regulatory Arrangements: Efficiency and Government's Incentives When does R attain higher social welfare than SR and vice versa? The following result summarizes the argument: Result 1: Social welfare is higher (lower) under SR than under R if ?2/(p?c)2+¾?2?½? >(<) 0. The socially efficient regulatory regime depends on the parameter values describing the economy. The less rigid SR yields higher social welfare than R when aggregate uncertainty is high and/or the divergence of interests between producers and consumers is low, and/or the government is populist. This corresponds to regions I and II in Figure 8.79 In contrast, when aggregate uncertainty is low, and/or the economy is polarized, and the producer lobby is strong (regions III and IV), social welfare is higher under R than under SR. To illustrate the intuition of this, recall that the final realization of L under regime i?{SR,R} equals [?i(A+?)+(p+A)/2]/[1+?i] and that the social-welfare maximizing value of L when no amendments of enabling legislation are needed, is ½(p+c). It is then readily seen that, given (p?c)2, when ? is high, SR on average keeps the final realization of L relatively closer to ½(p+c) than R for any realization of ?. When ? is lower, on the
78
Note, first, that if R possessed all the flexibility of SR, R would always yield the same social welfare as SR. This result is consistent with Gehrig and Jost (1995): they show that without the firms' informational advantage, government regulation could always mimic self-regulatory actions, at least when government's monitoring costs are ignored. Second, R would fall short of attaining the first best social welfare even if the government possessed all the expertise of producers (?R = ?SR) and weighed the welfare of the economy's groups equally (? = ½): since there are gains from renegotiating ?L, a self-interested government maximizing its payoff will always agree to renegotiate the default ?L in exchange for a transfer. Note that WSR = WR when ?2/(p?c)2+¾?2?½? = 0.
79
59
other hand, the more rigid R on average keeps the final legal rule closer to the socialwelfare maximizing value than SR for low, but not for high, values of ?2.80 While Result 1 indicates the circumstances under which each of the two regulatory regimes implements higher social welfare, it does not establish whether the regime yielding higher social welfare will in fact be chosen by the government at t = 0. We examine this question by investigating the government’s payoffs under R and SR: Result 2: The government’s payoff is greater (smaller) under SR than under R if ?2/(p?c)2?¼[(4+3?R+?SR)/(?R??SR)]?2 >(<) 0. The government's preferred regulatory regime, like the socially efficient regulatory arrangement, depends on the parameter values describing the economy. The minimum level of ?2/(p?c)2 necessary to render SR comparatively more attractive to the government increases with ?. The government would prefer to centralize regulatory authority in regions II and III in Figure 8, and delegate it in regions I and IV.81 To see the intuition, note that the government’s optimal L equals A. Given ?R, ?SR, and (p?c)2, the attractiveness of a given regulatory arrangement for the government depends on a trade-off between ?2 and ??[0,1]. When ? ? 0 (A ? p), the government would prefer to institute the regime that weighs (p+A)/2 ? p relatively more heavily than
When ? ? 1 (A ? c), for example, ½(p+A) ? ½(p+c), so the regime that implements the final legal rule closer to ½(A+p) rather than A+? yields higher social welfare for any value of ?2. When ? ? ? (A ? ?p+?c), ½(p+A) ? (5/6)p+(1/6)c, the more rigid R is more likely to keep the final realization of L relatively closer to ½(p+c) than SR as long as ?2 is small. A direct consequence of Result 1 is that which regulatory regime attains higher social welfare does not depend on the size of ?R??SR>0. Suppose that, in a given institutional setting, higher social welfare is attained under R and that the regulatory authority is (correctly) consolidated within the government. As regulatory practice is carried on, however, suppose further that R becomes more rigid (e.g., due to increasing bureaucratic complexity). Result 1 implies that this should not warrant a reversal in regulatory practice and re-allocation of regulatory authority to the more flexible producers.
81
80
Note that VSR = VR when ?2/(p?c)2?¼[(4+3?R+?SR)/(?R??SR)]?2 = 0.
60
A+? ? p+?: SR is favored even in the case of small aggregate uncertainty. For ? greater than 0 (A closer to c), on the other hand, (p+A)/2 is further away from A than A+? at least for small realizations of ?, thus making R the government's preferred choice. Observe also that the attractiveness of SR (or R) for the government naturally varies with the size of ?R??SR>0.82 Regions II and IV of Figure 8 highlight the conflict between the socially efficient regulatory regime and that compatible with the government’s incentives: in a given institutional setting, the welfare-superior regulatory regime will not be implemented when such implementation is incompatible with the government's incentives. In region IV, a government bought by the producer lobby bestows regulatory powers on its constituency, when direct government regulation would be socially efficient. In region II, in contrast, a populist government, eager to protect the interests of consumers, resorts to direct government regulation, despite the fact that delegation of regulatory powers to producers would increase social welfare. If the distribution of ?2/(p?c)2 across political systems was independent of ?, then the model predicts that a populist government will be more likely to mis-allocate law-making rights at the expense of the society than a government aligned with the producers because region II is larger than region IV.83
Given ?>0, ?SR and ?R, suppose ?2/(p?c)2 is just sufficiently high that the government prefers SR to R. That is, selecting the regime that places a substantially lower weight on A+? as a component of the final legal rule is favored by the government. Notice, though, that ½(A+p) ? A for any ?>0 and that?½(A+p)?A?>0 increases with ?>0. In contrast, A+? ? A for ? ? 0. With ?R sufficiently lower and everything else constant, then, tolerating the increased risk associated with selecting the regime that weighs A+? somewhat (but not substantially) more, pays off to the government: especially for higher values of ?, the final realization of L under lower ?R is more likely to be closer to A if the government centralizes rather than delegates the regulatory authority.
83
82
It can be easily shown that, in Figure 2, (a) the WR = WSR curve attains its maximum at ? = 1/3, and (b) the VR = VSR curve intersects the WR = WSR curve at ? < 1/3.
61
The model also shows that two parameters, both describing the state of the economy, yet of a widely different character, play exactly equivalent roles. Aggregate uncertainty about the effects of enabling legislation (?2) and the degree of homogeneity of affected interests (1/(p?c)2) influence the choices of the socially efficient and the politically desirable regulatory regimes in the same way. Ceteris paribus, this implies that two markedly different societies will make the same choices on regulatory regime, one society being beset by great aggregate uncertainty about the effects of outcomes of institutional changes (extraordinarily high ?2) and the other characterized by a great degree of consensus among affected interests about the regulatory agenda (very small (p?c)2). In both, delegation of regulatory powers to producers is socially efficient and preferred by the government. Given that the former society might be one undergoing revolutionary changes and that the latter could be a peaceable nation where there is little social conflict, the commonality of predictions for the two is certainly not a hypothesis that would be readily generated from intuition, showing the virtues of placing the analysis in a systematic model. The model also bears on Hayek's maxim that the certainty of the law is a chief attribute of good law, and a prerequisite for successful development (Hayek, 1960).84 Our model captures this notion, since uncertainty about the effects of the enabling legislation (?2) reduces social welfare regardless of the regulatory regime in place.85 Yet
84
"[A]…chief attribute which must be required of true laws is that they be known and certain. The importance which the certainty of the law has for the smooth and efficient running of a free society can hardly be exaggerated. There is probably no single factor which has contributed more to the prosperity of the West than the relative certainty of the law which has prevailed here. This is not altered by the fact that complete certainty of the law is an ideal which we must try to approach but which we can never perfectly attain" (Hayek, 1960, p.208). While both WR and WSR decrease in ?2, WR decreases faster.
85
62
the government is able to moderate the uncertainty about the effects of the final realization of L by delegating regulatory powers to the producers.86 The results derived above indicate that the equilibrium regulatory regime choice increases the predictability of the effects of final legal rules when the effects of the enabling legislation are very unpredictable. When effects of the enabling legislation are hardest to predict (when ?2 is large), we show that, in equilibrium, a self-regulatory regime will in fact be chosen by the government. 2.4 Illustrating the Model This section illustrates the applicability of the model, using a range of examples. We show that the model provides explanatory power in identifying factors that might explain differences in the regulatory practices of the progressive era and the New Deal. Our analysis helps in interpreting the choice of mechanisms of intervention used in fascist economies. It also identifies those features of legal traditions that help to explain variation in regulatory arrangements across countries. 2.4.1 Two Modes of Attaining Regulatory Goals: the Progressive Era and the New Deal Two episodes in American history, the progressive era and the New Deal, while both known for widespread regulation of the economy, provide an example of a stark contrast in allocation of regulatory powers. Our model helps to explain why the two periods were characterized by such markedly different regulatory regimes. To set the stage, we first
The variance of the final legal rule under regime i?{R,SR} equals (?i/(1+?i))?2. Because ?SR < ?R, (?SR/(1+?SR))?2 < (?R/(1+?R))?2.
86
63
turn to briefly discussing the regulatory practice under the progressive era and the New Deal.87 In the progressive era, direct government regulation proliferated. A large number of regulatory initiatives led to the creation of new administrative agencies addressing corporate organization and practices in a number of sectors, including rail transportation, finance, food, and pharmaceuticals.88 The progressive policy-makers were committed to preserving markets and a decentralized market structure. Reforms on a grand scale were not advocated: mainstream progressivism was "an attempt to reclaim the market-based sphere of individual liberty and opportunity" (Eisner 2000, p. 47).89 Where the structural features of an industry precluded an antitrust strategy (e.g., in transportation), Congress and the president initiated policies aiming to compensate for the lack of effective market mechanisms by regulating rate-making and other activities. The progressives' faith in expert knowledge and their desire to foster administrative neutrality and efficiency at all levels of government led to consolidation of governmental regulatory authority in the form of regulatory commissions.90 Under the New Deal, in great contrast to the progressive era, regulatory authority was formally extended to the interests being regulated in many sectors of the economy.
87
The sketchy depiction of regulation under the progressive era and New Deal below draws heavily on Eisner (2000).
88
The Sherman Antitrust Act established the foundation of US competition policy; the Hepburn Act dramatically enhanced the powers of the Interstate Commerce Commission; the Meat and Inspection Act and the Pure Food and Drug Act; the Federal Reserve Act creating the Federal Reserve Board and expanding national regulation of finance; the Federal Trade Commission and Clayton Acts strengthening antitrust enforcement. See Eisner (2000, p. 44). Wilson's vision of an order, for example, was based on markets and market-corrective regulation (Eisner 2000, p. 39). Although formally independent from the executive branch, these were unable to withstand the pressures of partisan and interest groups politics, and constraints imposed by the courts. See Chapter 3 in Eisner (2000) for a discussion.
89
90
64
A system of government-supervised self-regulation was established: "[Under the New Deal]…many private economic associations became quasi-public in nature, for they were given public authority and an important role in making and implementing regulatory policy" (Eisner 2000, p. 90). Trade associations and industry groups were authorized to establish codes of conduct that were exempt from antitrust laws.91 The Agricultural Adjustment Act decentralized regulatory decision-making in agriculture relying on farm associations for policy implementation. The Securities and Exchange Commission, established in 1934, played a critical role in giving the financial industry the necessary authority to self-regulate and facilitated the creation of the largest self-regulatory body in the country, the National Association of Securities Dealers. The divergent pattern of regulatory arrangements of the progressive era and the New Deal reflects underlying features of the respective eras that can be interpreted using our model. First, the early progressive era (late 1890s, see e.g. Hofstadter 1972) marked the ending of a period of rapid socio-economic change and the return of stability.92 The progressives did not advocate radical transformations of the economic system. They believed that market-corrective actions and reliance on scientific methods would yield positive outcomes. The New Deal was, in contrast, launched amidst the middle of unprecedented economic failure and widespread questioning of existing institutional
91
"The National Recovery Administration chose business executives…to serve as members of the Industrial Advisory Board. On this board, they assisted in the drafting of codes and the creation of code authorities, the bodies of business and trade association representatives which interpreted code provisions and granted exemptions" (Eisner 2000, p. 85).
In the latter half of the 19th century, the growth of tangible capital stock greatly outpaced the growth of the population. Tremendous waves of immigration accelerated urbanization of society. Economic growth was disrupted by a fairly deep depression that lasted from 1893 to 1896 (see Eisner, 2000). The return of tranquility is conventionally dated to 1897 and the inauguration of the Republican president McKinley: "The disturbing changes that Bryan [a Populist nominee] had promised…would not pass. The frightening prospect of a radical alliance of farmers and workers had collapsed. The emerging industrial order…seemed safe" (McGerr 2003, p. 3).
92
65
arrangements. The necessity for, and the spirit of, institutional experimentation extended the government’s authority into places where it had not been before. The uncertainty about the outcomes of institutional changes, as captured by ?2 in our model, was therefore comparatively larger under the New Deal. Second, the power of corporations grew tremendously during the later decades of the nineteenth century. Yet the whole country did not share in the new wealth, prestige and optimism. The public's dissatisfaction with growing inequalities suggested a growing threat to property rights (Eisner 2000, p. 33). In contrast, during the New Deal, there was a sense of nation-wide consensus about the necessity of coordination and cooperation between the industry and the public. F.D.R. presented the system created by the National Recovery Act as one that would "sink selfish interests and present a solid front against a common peril" (Eisner 2000, p. 84). This implies that the degree of polarization in society, that is, (p?c)2, was comparatively smaller under the New Deal. We argue that the economy-wide ? was unlikely close to 1 or close to 0 in either historical episode.93 On the one hand, elements of the New Deal often openly promoted cartel agreement to insure industrial 'recovery and relief', and similarly there were many episodes of regulatory capture under the progressive era (implying ? could have been close to 0).94 On the other hand, the New Deal responded to the needs of the poor and the unemployed, while many measures of the progressive era, such as reforms of civil service, direct elections of senators, and recall of judges, were of very populist nature
Glaeser and Shleifer (2003) argue that events of the progressive era may be explained by both capture theory and public interest approach. Similarly, Wallis (1987, 1998) shows that the New Deal measures were motivated by both economic and political factors.
94
93
See, e.g. Eisner (2000), Glaeser and Shleifer (2003) and references therein.
66
(implying ? could have been close to 1).95 With the economy-wide ? being neither too high nor too low, Figure 8 predicts that the direct governmental regulation of progressive era (placed in area III) and self-regulatory associational regime of the New Deal (placed in area I) can be viewed as efficient institutional responses to the varying historical circumstances. 2.4.2 The Fascist Economic System: Self-Regulation in a Totalitarian Regime Analysis of the structure of the fascist economic system has drawn significantly less attention than the ideology that underpinned the historical events surrounding it.96 One reason for this is that fascism, as an economic system, does not fit any simple categorization, with its admixture of markets and controls, sometimes looking more capitalist, sometimes having elements of socialism (Milward 1976, p. 379). Our model can help to interpret the fascist system97, by viewing self-regulation as one practical mechanism to implement extensive governmental controls. We argue this by first turning to a brief characterization of the fascist economy. At the center of fascist philosophy was glorification of the nation state, in which the pursuit of economic power was highly instrumental. Maximizing production was crucial, and the individual, as consumer, was given little priority in welfare. Prominence was ascribed to industry and to the workers as producers: "...the interests of production are the interests of the Nation...The fascist State is primarily a State of 'producers'" (Florinsky 1938, p. 90). To pursue economic activity that would enhance the nation state, a high
95 96
See Glaeser and Shleifer (2003).
At their inception, however, economic claims of Italian Fascism and German Nazism did prove a subject of compelling interest (Maier 1987, p. 70).
97
While there clearly were differences in the fascist systems of Italy and Germany, Maier (1987), for example, argues that the many common traits of the two regimes warrant an encompassing analysis.
67
degree of intervention into the market was deemed necessary. What really distinguished fascism from free market capitalism was the fascist view that pervasive intervention was crucial in promoting the national good (Sternhell 1976, p. 357). The need for large-scale intervention accepted, there follows the question of what practical mode of implementation to use. In Italy, but also in Germany, the answer lay in the intimate involvement of representatives of businesses in governmental economic decisions. Private associations of businessmen were integrated into state administration and "…transformed into instruments of economic control. But whereas in both countries the titles of these organizations changed, the personnel at the head of them did not... The line between government and business was blurred because the personnel became in many cases one and the same" (Milward 1976, p. 396?7).98 In Italy, as one theorist of fascism declared, "Fascism...would now organize the production of wealth for national ends through the self-government of the categories under the high control of the State" (Field 1968, p. 185).99 Political, and particularly fascist party, control was imposed at the highest levels, but within these constraints business had a degree of autonomy to regulate itself, given its control over both information and micro-decisions (Milward 1976, p. 397; Lyttelton 1976, p. 141). In Italy, the corporations created by Mussolini, but dominated by business, were involved in detailed regulation of prices, production quotas, import suppliers, and myriad other economic variables (Field 1968, pp. 61, 181?95).100 The fascist economy
98 99
See also Lyttelton (1976). See also De Grand (1995, pp. 40?1).
A fascist corporation (la corporazione) was not a corporation in a legal sense of the English word, but a union of all elements involved in a particular economic activity (employers, experts, workers), together with representatives of general interests of the nation and the consuming public. Corporations were first
100
68
can therefore be characterized as one with massive governmental intervention into the market-place, in pursuit of glorification of the nation state, which was implemented through a process of self-regulation.101 Our model is suggestive of why the fascist system might not have been as incongruous as might seem on first inspection. The interpretation rests upon several basic facts about fascism and the world in which it emerged. First, capitalism was not rejected, but markets were to be subject to extensive intervention (i.e., either government regulation or self-regulation). Second, fascist regimes came to power in times of crisis.102 Moreover, they did so with the purpose of promoting truly revolutionary changes in the institutional structures of their countries (Linz 1976, p. 13). In terms of our model, ?2 was extraordinarily high. Third, as noted above, the fascist state emphasized production and the emphasis on production soon became an emphasis on helping producers.103 ? was therefore low. Finally, the state and the fascist party certainly exerted much influence on industry, but this occurred through bargaining with the self-regulators (Florinsky 1938, p. 97), as envisaged in our model.
authorized under the syndical legislation of 1926. A law on corporations was passed in February 1934, establishing 22 corporations with jurisdictions based on the notion of a product cycle. The corporations had the authority to issue rules for the ‘collective regulation of economic relations and for the unitary discipline of production’, subject to the consent of the Head of the Government and after a proposal by one syndical association. The first corporation to produce a regulation legally enforceable under the provisions of the law was that of Beets and Sugar. See Field (1968).
101 102
Mussolini himself used the term 'self-discipline' (Florinsky 1938, p. 97).
Both Italy and Germany were in search of a path out of the recession. Maier (1987, p. 104) argues that the Italian Fascists had to steer the economy through the worst shoals of the World Depression; the Nazis, however, were luckier and could ride the wave of recovery.
103
Even the confederation representing artists and liberal professions was requested by the Charter 'to promote the interests of art, science and letters with a view to improving production' (see Florinsky 1938, p. 90). One of the earliest acts of fascist regimes was to ban labor organizations. See De Grand (1995, pp. 40?1).
69
With extraordinarily high ?2 and low ?, fascism would therefore correspond to a point in area I in Figure 8. Self-regulation would have been an appropriate choice from a social welfare point of view.104 Hence, in contrast to contemporary observers, who saw great inefficiencies in the economic structures of fascism (e.g., Florinsky 1938, pp. 93?4), our model hints that they were not as ill designed as is it is easy to assume, suggesting why fascist economies had some staying power in a world beset by recession. 2.4.3 Legal Origin and Regulatory Regime Choice The differences between civil and common law systems have been a central area of concern in writing on institutions in the last decade (La Porta et al. 1997, Beck et al. 2003). The properties of these two systems have important implications for the constructs that are central in our model: in the predictability of institutional construction emanating from the law-making process and in the ease of decentralized adjustments in institutions. These are both factors that feed into the choice of regulatory regime. In the paragraphs below, we use our model to show how that choice is affected by the nature of the legal system. Then we review some facts about variations in regulatory regimes across countries, examining the fit between these facts and our model's predictions. The first important difference arises from the fact that under civil law regimes, law emanates directly from the center, with judges restricted to using the statutes themselves as primary input into interpretation. Under common law, by contrast, judges have a considerable degree of power to interpret, using case law, a large part of which cannot be modified by the writers of a statute creating an institution (Zweigert and Kötz
Observe that none of our arguments depend on the value of (p?c)2 in a fascist economy, as long as (p?c)2 was not extraordinarily high. The size of (p?c)2 may proxy for the nation-wide sentiments about fascist ideology. Generalizations about such sentiments would be controversial enough as to be immediately unproductive.
104
70
1992, pp. 273-278). Hence, ceteris paribus, the institutional outcomes resultant from a statute are harder for the institution-designer to predict under common law than under civil law. There is less uncertainty about the effects of the skeletal regulatory legislation, at the time of passage, under civil law. In the model’s notation, ?2e > ?2f where subscripts e and f denote common law and civil law, respectively (evoking England and France, which provide the archetypes of the two systems). Second, legal traditions differ in their ability to reshape rules to fit changing conditions and to facilitate adjustments in existing institutions. Legal systems that embrace case law and judicial discretion tend to exhibit a greater degree of adaptability (Zweigert and Kötz 1992).105 Rejecting jurisprudence and relying on changes in statutory law renders the adjustment of legal rules costlier under civil than under common law, regardless of the regulatory arrangement: ?i,f > ?i,e for all i?{R,SR}. Additionally, the inherently centralized character of civil law implies that centralized institutions, such as governmental regulation, are comparatively less costly for a country with that tradition: centralized institutions work comparatively better in a centralized system. Hence, centralization of regulatory authority naturally leads to a greater increase in the cost of legal rule adjustment under the inherently more decentralized common law: ?R,e ? ?SR,e ? ?R,f ? ?SR,f.106
Beck et al. (2003) provide an extended discussion of the relative flexibility of civil and common law systems. Two historical examples are suggestive. The increasing centralizations in seventeenth century England and France produced revolutions in the former and European dominance for the latter. In England, the attempts by Charles I and James II to centralize power in the monarchy led to their deposition and Oliver Cromwell's relatively centralized interregnum failed. In contrast, in France Louis XIV's reign was known for its absolutism and for the political and cultural dominance of France in Europe. Turning to later times, the economic performance of France was surprisingly strong in the period immediately following the Second World War, when the zeitgeist was infected with the centralizing tendencies of socialism and planning, while the UK's economic performance was disappointing (Shonfield 1974). By contrast, as
106
105
71
These intrinsic features of the legal traditions suggest two reasons why there might be a greater tendency for government to choose to centralize, rather than delegate, regulatory authority under civil law than under common law. (Figures 9(a) and 9(b) graphically illustrate Result 3.107) Result 3: (a) If ?2f < ?2e, then {?: VR(?2e) ? VSR(?2e)} ? {?: VR(?2f) ? VSR(?2f)}. (b) If ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0, then {(?,?2/(p?c)2): VR(?R,e) ? VSR(?SR,e)} ? {(?,?2/(p?c)2): VR(?R,f)? VSR(?SR,f)}.108 The essence of the predictions in Result 3 is that self-regulation will occur more often in common-law systems than under civil law. We have sought data to test this prediction but, given the lack of systematic study of regulation versus self-regulation, only snippets exist. At the very least, the conclusions of sociologists' studies of professions are consistent with Result 3: in civil-law countries, the state has played a much more important role in the licensing and regulation of professions. In common-law countries, professionalization has, almost invariably, been practitioner-led: their associations have sought and usually obtained state authority to self-regulate (Burrage and Torstendahl 1990).109
liberalization and privatization became popular in the 1980's and 1990's, the UK has grown faster than its continental neighbors.
107
For notational simplicity, we suppress all of the arguments of Vi except for those crucial to an understanding of the result.
?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0 is in fact a stronger condition than necessary. Note that ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0 if and only if ?SR,f??SR,e ? ?R,f??R,e > 0, which implies ?SR,f??SR,e > (?R,f??R,e)[(1+?SR,e)/(1+?R,e)] > 0. The latter is a necessary and sufficient condition for {(?,?2/(p?c)2): VR(?R,e) ? VSR(?SR,e)} ? {(?,?2/(p?c)2): VR(?R,f) ? VSR(?SR,f)}. See Appendix 5 for proof of this point. Scattered evidence can be found, primarily on the legal and accounting professions. According to a systematic cross-country study by Global Accounting Education (2004): "In general professional bodies in common law countries show a tendency towards professional self-regulation compared with a tendency towards government regulation in civil law countries. This is particularly true for common law countries with a historical UK background" (Global Accounting Education, 2004, http://www.eiasm.be/gae/default.asp?id=25, accessed 9/14/2004).
109
108
72
We have found one area where there has been systematic data collection on which types of regulatory arrangements have been implemented: alcohol beverage advertising. In 1996 the Centre for Information on Beverage Alcohol collected information on how 119 countries regulated alcohol beverage advertising (International Center for Alcohol Policies 2001). The categories were self-regulation, statutory legislation (that is, central regulation), and a combination of both.110 Table 1 matches the data on regulatory arrangements with information on legal origin obtained from La Porta et al. (1999), who distinguish between English, German, French, Scandinavian, and socialist legal traditions. As the Table indicates, there are indeed systematic differences among countries with different legal origins: consistent with Result 3, English legal origin countries have a greater tendency for self-regulation than do countries with a civil law tradition. To test the effect of legal origin, we used alcohol-advertising regulatory regime as the dependent variable in an ordered probit, with the lowest category being central regulation, followed by the combination of both types of regulation, and self-regulation being the highest category. Since it is likely that choice of regulatory regime varies with level of development, and since level of development is correlated with legal origin (see e.g. La Porta et al. 1998), we add log of GNI per capita to the equation. Since countries with larger populations exhibit different regulatory behavior (Mulligan and Shleifer 2004) we also added log of population (see Table 1). The results of the ordered probit appear in Table 2. In Table 2, the omitted legal origin dummy is the French one. Countries of English legal origin clearly use more self-
110
Other categories, such as a complete ban or no relevant policies, are irrelevant here.
73
regulation than countries from the civil law system.111 It must be emphasized that we only view these results as suggestive: the dependent variable only reflects a relatively specialized and narrow area of regulatory activity; there are undoubtedly omitted variables in the regression; and we have not paid any attention to the possible endogeneity of GNI per capita.112 Nevertheless, these suggestive results do offer clear evidence in favor of Result 3. 2.5 Conclusion Allocation of regulatory rights matters because regulatory contracts are incomplete. The source of potential welfare gains in delegating rule-making powers to producers is in their greater adaptability to changing institutional conditions. The source of potential welfare losses is in the active participation of the regulated in the regulatory process. Therefore, whether rule-making authority should be extended to producers, or consolidated within the government, will depend on country-specific factors. When either there is a large level of uncertainty surrounding the results of institutional construction, or the degree of divergence of producers' and the public’s interests is small, the benefits of delegating regulatory powers outweigh the costs. In the opposite case, the regulatory arrangement that yields higher social welfare depends on the extent to which the government's motives are populist or aligned with those of producers.
As an aside, note the negative and statistically significant coefficient on the dummy for socialist legal origin (which may be interpreted as a transition-country dummy): countries of socialist legal tradition resort even less to self-regulation, and even more to central regulation, than countries of French legal tradition. Given conventionally accepted facts characterizing the transition experience, our analytical model seems consistent with this empirical evidence. The populist politics of transition suggests that ? was not low. Highly polarized economies and unprecedented institutional changes imply that both ?2 and (p?c)2 were large during early transition. ?2/(p?c)2 was therefore neither too high nor too low. Our analytical model implies that government controlled regulation would have been a natural political choice, yet socially suboptimal (area II in Figure 1).
112
111
GNI per capita is lagged by five years to reduce this problem.
74
Our analysis highlights the conflict between the socially efficient regulatory arrangement and the arrangement compatible with the government’s incentives. Ultimately, the choice between centralization and delegation of regulatory authority is that of the ruling administration. A socially efficient regulatory regime will be implemented only if such implementation is in the government’s interest. In this light, we argue that the widely different institutional frameworks of regulatory practice during the progressive era and the New Deal, and the peculiar fascist system of organizing production by self-government of industries under control of the state, were not only compelling equilibria, but also relatively socially efficient. We show that regulatory regime choice is also influenced by differences in legal traditions. These differences have been a central area of concern in writing on institutions the last decade. We further this discussion by identifying those intrinsic features of common law and civil law that help to explain variation in regulatory arrangements across countries.
75
Chapter 3 Fostering Civil Society: Towards an Analytical Framework
3.1 A Concept in Vogue The notion of civil society has long been widely debated in political and sociological theory and philosophy (see e.g., Cohen and Arato 1994; Hall 1995; Ehrenberg 1999), culminating in an extraordinarily voluminous body of literature. The concept has more recently drawn much consideration in policy circles' debate on development.113 Many channels through which civil society involvement may foster social and economic progress have been discussed, as argued below. This chapter focuses on civil society as a bulwark against the state. As an associational realm separate from the state, civil society brings a society's concerns to broader public attention, thereby strengthening institutions for, and contributing to mechanisms of, good governance (Salomon et al. 2004). It is due to this proposition that fostering civil society per se – rather than guided by a specific goal of service provision, for example ? has become means of promoting good governance and economic development in post-communist and developing countries. Yet the quest for promoting democratic governance and development via fostering civil society in countries of Central and Eastern Europe (CEE) and the former Soviet Union (FSU) has proven elusive (see e.g. Carothers 1999; Ottaway and Carothers 2000). In this chapter, we attempt to illuminate the recent experience with civil society aid for promoting good governance in the context of a simple analytical model.
See, e.g., the Civil Society and Governance Programme at the Institute of Development Studies, University of Sussex, (http://www.ids.ac.uk/ids/civsoc/), the World Bank's Civil Society program (http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/CSO/0,,pagePK:220469~theSitePK:228717,00.html), the LSE's Centre for Civil Society (http://www.lse.ac.uk/collections/CCS/), the International Center for Not-for-Profit Law (http://www.icnl.org/).
113
76
The rest of this chapter is organized as follows. In section 3.2 we elaborate on the notion of civil society. In section 3.3 we discuss the reasoning that guided civil society aid as part of the democracy aid project. We also review the economics literature of pertinence to analyzing civil society. In section 3.4 we present the scant available evidence on success of civil society aid. In section 3.5 we attempt to illuminate the donors’ experience with a simple analytical model. Section 3.6 concludes. 3.2 Delineating Civil Society The boundaries of civil society are subject of an ongoing debate in social sciences (see e.g. Hall 1995). Most commonly, civil society has been confined to "an intermediate associational realm between state and family", consisting of organizations that are autonomous from the state, and are formed voluntarily by members of the society to protect or advance their interests or values (Ottaway and Carrothers 2000, p.9).114 A majority of the literature asserts that political parties and other groups that explicitly seek to gain control of the state are not part of civil society (Carothers and Ottaway 2000). Likewise, features such as existence of free press and individual forms of citizen action (such as voting and writing to legislators) sometimes enter the discussion on civil society, but are typically excluded from its definition (Salomon et al. 2004). A pragmatic characterization of civil society (see e.g. Salomon et al. 2004; Quigley 2000) divides organizations into those performing essentially expressive
A recent comprehensive study by The Johns Hopkins Institute for Policy Studies (Salomon et al. 2004) has defined civil society organizations (CSOs) as those that are (i) organized, which includes informal as well as registered groups, (ii) private, although they may receive support from governmental sources, (iii) not profit-distributing - while organizations may generate surplus, any such surplus must be reinvested in the objectives of organization, and (iv) selfgoverning. Such definition encompasses a multitude of formal and informal organizations, religious and secular, with paid staff and those staffed by volunteers (e.g., hospitals, universities, social clubs, professional organizations, day care centers, grass-root development organizations, health clinics, environmental groups, religious congregations, family counseling agencies, self-help groups, religious congregations, sports clubs, job training centers, human rights organizations, community associations etc.) See Salomon et al. (2004) for details.
114
77
functions (advocacy, community organizing, representation of interests, cultural expression, environmental protection, human rights, religion, and political expression), and those performing essentially service functions (e.g. provision of health, education and research, social services, development and housing).115 3.3 Promoting Civil Society: the Democracy Aid Project116 Civil society aid is best understood as a component of an ambitious project of "democracy promotion" ? broadly speaking, aid designated for promotion of good governance (see e.g. Carothers 2004), channeled to post-communist (as well as developing) countries. Civil society aid was not a major component of financial aid to CEE and FSU for democracy promotion in the early 1990s.117 Early on, the donors typically concentrated on providing support for administering elections. Democratic elections, however, were only a small step towards securing democratic governance and rule of law. In the second phase, the donors extended the scope of their assistance to pursue reform of major state institutions (especially legislature and judiciary). The task to reform state institutions, however, was much more daunting than anticipated. There was a lack of will to reform among both incumbent political elites, whose power was directly threatened, and the personnel at lower levels of bureaucracies, content with status quo. Likewise, the scale of the project was excessive given available funds.118 It is only in the third phase, in the mid 1990s, that donors' attention has shifted to strengthening civil society. Yet, if
115 116 117
Some CSOs may of course combine both functions. The following sketch draws heavily on Carothers and Ottaway 2000.
This aid is a narrow subset of Western aid to CSOs generally. Likewise, the aid's hoped for effects include only a fraction of the total contribution of CSOs in developing and transition countries.
118 As Carothers and Ottaway put it, "[A] comprehensive program of judicial reform …was far more than any donor could think of financing in any country" (Carothers and Ottaway 2000, p. 8).
78
strengthening civil society was to promote (economic and social) development, who exactly were the donors to aid, and for what purpose? 3.3.1 Civil Society and Development Three channels through which civil society may affect economic development have been proposed. First, civil society organizations serve as vehicles for service provision under poorly functioning states, often to alleviate poverty (see e.g., World Bank 2003). Second, through community building, civil society organizations create social capital (Putnam 1993) ? bonds of trust and reciprocity deemed crucial for institutions of market economy to function effectively. Third, as an associational realm, civil society brings society's concerns to broader public attention, thereby strengthening institutions for good governance. While all three channels provide for a fruitful research agenda in social sciences, including economics, we argue below that the donors’ reasoning for providing aid was most heavily influenced by one of the three. Civil society organizations as vehicles for service provision Organizations of civil society have been widely acknowledged as important vehicles for service provision in developing countries (World Bank 2003). Serviceoriented CSOs are large recipients of foreign aid. The economics literature has provided some guidance on desirability of service provision through CSOs, but much work remains to be done (see Appendix 6 for a review of this literature and discussion). Unsurprisingly, the capacity of directly promoting good (democratic) governance, however, was not seen as a comparative advantage of service-oriented CSOs. Rather, it was another subset of CSOs that donors identified as targets of aid for promoting good
79
(democratic) governance. Before we turn to that, however, we briefly mention another channel through which civil society influences development. Civil society as a breeding ground of social capital Robert Putnam's analysis of development of Italian regions (Putnam 1993) posed a connection between forms of social capital ? norms and networks within civil community, manifested through participation in groups and associations of civil society – and quality of governance and economic development.119 The economics profession has only in the last decade expressed greater interest in formation and effects of social capital. This line of research now includes contributions at both empirical and theoretical level of analysis (e.g., Knack and Keefer 1997; Temple and Johnson 1998; Glaeser et al. 2000; Guiso et al. 2004; Huang 2003). During the 1990s, when decisions about promoting civil society were made, however, the social capital building view of civil society did not play a role. As Ottaway and Carothers put it succinctly, "with a limited amount of money, the need to produce visible results in a short time, and a critical press and an even more critical [legislature] ever alert to denounce waste of taxpayer money, [the donors] could hardly get in the business of setting up bowling leagues…" (Carothers and Ottaway 2000, p.10).120 Civil society as a bulwark against the state Given the donors' aim of promoting rule of law and good governance, policyoriented CSOs (a subset of CSOs providing expressive functions rather than service
119 120
Another influential contribution is Fukuyama (1995).
The bowling leagues example comes from Putnam's article "Bowling Alone: America’s Declining Social Capital", published in the Journal of Democracy in 1995.
80
functions, according to our above taxonomy) emerged as natural candidates.121 Moreover, the donors chose to exclude political parties from consideration. The belief was that is it possible to obtain results without becoming interfering unduly in the domestic politics of another country (Carothers and Ottaway 2000). The donors hence decided to focus on voluntary associations with capacity to directly foster democracy and promote democratic consolidation.122 These are associations that specifically seek interaction with the state, whether to advocate interests of the citizens, to oppose nondemocratic behavior of the state, or to hold states accountable to citizens for their actions. These associations, however, do not explicitly compete for the office. Civil society aid was not seen as a panacea for the intricate task of good governance promotion. It was expected, however, that it would make a difference. In the context of post-communist transition, for example, civil society was "…deemed crucial to stimulating the public pressure and participation necessary to force poorly functioning state institutions to become more responsive and accountable…." A vibrant civil society was "…held out as necessary to cushion the effects of restructuring, to ensure public understanding and support for market reforms, to prevent privatization from lapsing into cronyism, to connect newly empowered local governments to citizens, and to create many of the other components necessary to the transition to market economies" (Carothers 1999, p.115).
121
Sports clubs, cultural associations, religious organizations and less formalized social networks were therefore absent from most civil society assistance programs. Media and trade unions were also excluded from the target group.
122
It must be emphasized that more recently, the donors have expanded the set of civil society aid recipients to include service-oriented NGOs as well. This is especially true of European Union's DemNet program. See Quigley (2000) for a discussion.
81
Before evaluating the aid’s success, we briefly summarize the economics literature that has put forward – but hardly scrutinized ? the notion of civil society congruent with the democracy aid project.123 3.3.2 Civil Society and Institutional Change A handful of economics scholars have attempted to link civil society and mechanisms of good governance. Roland (2004) focuses on factors influencing institutional change. He argues that understanding the interaction of slow-moving institutions (culture, values and norms) with fast-moving institutions (e.g. political institutions) is key to understanding institutional change. Transplanting 'best-practiceone-size-fits-all' fast-moving institutions usually does not work; the outsiders' knowledge of local conditions, including cultural factors (slow-moving institutions), is typically deficient. The local public's involvement in assessing the correct institutional mix is therefore crucial. A dialogue with the governments, however, is unlikely going to be fruitful, in part due to the governments' proneness to capture by elites with vested interests in status quo. Instead, Roland argues ? but does not deliberate on his proposal ? different components of civil society should participate in the process of institutional construction. Beck and Laeven (2005) empirically explore the determinants of economic performance and quality of institutions in transition countries. They argue institution building is best understood when the relative abundance of natural resources and length of time spent under socialism in individual countries are taken into account. Beck and Laeven (2005) believe that a strong civil society may be key to successful transition.
Interestingly, the literature below never attempted to directly relate its suppositions to the civil society aid project.
123
82
Civil society groups provide, according to them, a check on behavior of (transition) elites. In particular, the length of time spent under socialism matters for institution building because civil society groups were less likely to emerge in countries that spent lots of time under socialism. Without a strong civil society, a country’s transition elites are thus less likely to be counter-balanced and are thus better able to earn rents through resource extraction.124 Campos (1999) similarly empirically establishes a connection between the strength of civil society and measures of economic development for transition countries.125 Campos argues that a strong civil society, via participation in public affairs, increases transparency of policy making. Thereby, civil society involvement may complement, or even substitute for, other institutions of good governance, such as high quality of bureaucracy, public character of policy-making, and the executive’s accountability. Dethier at al. (1999) argue that the existence of a vibrant civil society at the start of the transition has the most explanatory power in their (empirical) analysis of determinants of successful economic liberalization in 25 post-communist countries of Central and Eastern Europe and the former Soviet Union.126 Hoff and Stiglitz (2004) appeal to civil society in their theoretical model of the evolution of rule of law. They argue civil society critically affects institution building: "A consequence of a prolonged experience under communism was the absence of civil
124 Beck and Laeven, however, do not test this hypothesis empirically. Had they included a measure of strength of civil society in early transition in their regression for institution-building, one would expect that the coefficient on Years under Socialism would become statistically insignificant if their story holds. 125
To measure the strength of civil society, Campos combines the Freedom House measures of civil liberties and political rights with the "civil society" indicator from Karatnycky et al. (1998).
126
They measure the strength of civil society with the Freedom House index of civil and political liberties.
83
society institutions…with countervailing power to hold the state to account" (Hoff and Stiglitz 2004, p. ?). In their model, the absence of a strong civil society creates pessimism about the emergence of rule of law, reducing the probability of rule of law being implemented given existing support for rule of law. In equilibrium, this makes emergence of rule of law less likely.127 The above contributions relate civil society to the process of institution building. In that process, civil society emerges as a force counterbalancing the government and the elites; through active involvement in public affairs, civil society involvement improves institutional reform. Ceteris paribus, promoting civil society's participation in public process should therefore have been social welfare improving. 3.4 Evaluating the Aid's Success: the Role of Initial Political Conditions128 As newly emerging organizations, civil society organizations were believed not to have vested interests in the old system. They were seen as a potential source of pressure for reform. An important part of their appeal was that they were also small enough that modest amounts of aid were significant to them. Qualitative, primarily anecdotal, evidence on the success of civil society aid for democracy promotion is scarce and emerging only now (Carothers 2004). It appears that the success of assistance to publicpolicy oriented CSOs in transition countries has, when seen in the aggregate, proven moderate, and extensively dependant on local conditions.
Similarly, they argue that the presence of agents with civic virtues - which they define as the agents who demand rule of law irrespective of their private interests - leads to an equilibrium increase in demand for rule of law. The following assessment is based on Quigley 2000 who focuses on Central and Eastern European countries. No analogous evaluation of the success of civil society aid exists for the countries of the former Soviet Union.
128
127
84
Civil society aid was provided mainly through means of technical assistance and grants. Many factors affected the success of the designated civil society aid. The knowledge and experience of the advising staff, for example, often proved inadequate for the local conditions. Likewise, many policy-oriented CSOs proved to have highly personal or political agendas.129 The overall success of the aid, however, seems to have been (most) critically dependant on the prevailing local political conditions. In the success stories of transition ? Poland, Slovenia, and Hungary, for example ? the governments were in general supportive of the CSOs. In these countries, policy-oriented CSOs have contributed to the public debate on crucial issues, particularly the pace and course of economic reform. In the less democratic and politically unstable group of transition countries – in particular in Romania, Bulgaria, Serbia and Slovakia ? on the other hand, the ruling governments hindered, and were opposed to strengthening, civil society. Thus, despite good faith and reasoning, civil society has since the collapse of communism not always been effective at promoting good governance. In bulk of the region – the countries marked by political instability and lower levels of democratic attainment ? the donors appear not to have (had) an impact on either civil society emancipation or institutional development. Ex post, the unsatisfactory experience has raised a debate, primarily in policy but also academic circles (Carothers 2004). Is fostering civil society at all worth spending a dollar on, or should all foreign aid instead be directed toward programs directly promoting economic reform? Was the focus on policy-oriented CSOs correct, or should
129
This in particular holds for policy-related research institutions or think tanks focusing on economic issues.
85
the donors have rather channeled funds to service-oriented CSOs very early on (even if service-oriented CSOs are less successful in contributing to democracy promotion in the short run)? 3.5 Aiding Whom and When: Towards an Analytical Framework Both the economics literature summarized in section 3.3.2 and the practitioners' reasoning seem to agree that a key feature of civil society involvement in the process of institution-building is to render the reform process open to public and subject to deliberation. At the very least, this should increase the transparency of the process. Organized legal professions' review of reform proposals, as detailed in chapter 1 of this dissertation, is an example of such involvement. An independent bar has long been seen as "the backbone of an autonomous civil society" (Rueschemeyer 1997, p. 217). Involvement of lawyers' associations "…is both an expression of the autonomy of civil society and the means of opposition to the state, as well as a force that reinforces this dualism. Bar action constitutes (with others) civil society" (Halliday and Karpik 1997, p. 56). Lawyers' associations are natural candidates for civil society aid due to their expertise about local institutional specifics. Indeed, the American Bar Association's Central European and Eurasian Law Initiative (CEELI) has recently launched a project to help strengthen the role of organized legal professions in emerging democracies.130 Lawyers associations, however, should not be mistaken as the only component of civil society with capacity to open up the policy-making process. In fact, the entire subset of policy-oriented CSOs as a group (CSOs performing mainly expressive functions
130
See http://www.abanet.org/ceeli/areas/legal_prof_index.html.
86
according to the taxonomy of Salomon et al. (2004)) may, through their active and vocal involvement, we believe, contribute to the success of the reform process. Prevention of quick and ad hoc implementation of reform proposals by ruling political elites and their allies is, we maintain, a key role of civil society in the process of institution-building. Orenstein's (2000) analysis of political economy of pension reform resonates with this. Orenstein provides illuminating case studies of design and implementation of fundamental pension reform programs in Hungary, Poland, and Kazakhstan. From all three post-communist countries, in the least democratic Kazakhstan, the deliberative process was significantly shorter than in the more democratic Hungary and Poland. In the authoritarian Kazakhstan, civil society organizations in Kazakhstan "face greater sanctions for voicing opposition" and have "fewer opportunities to access the policy process". Pension reform deliberations were performed in secrecy by a single governmental commission under the direct authority of the prime minister, with full approval of the president. In fact, "[c]ivil society groups did not even know about the progress of governmental pension reform proposals" (Orenstein 2000, p. 18). We therefore believe the initial focus of the donors on the aforementioned subset of CSOs was generally correct.131 The caveat, we believe, lied in correct evaluation of existing political conditions in transition countries. Indeed, we argue below that aiding civil society is wasteful when the attained levels of political stability and democratization are inadequately low. We now formalize our argument, relying heavily on, and extending the analysis in chapter 1.
131
Quigley (2000), for example, doubts that such focus was indeed correct.
87
3.5.1 Model Background The difference between the model as set up in chapter 1 and the model utilized in this section is interpretative. In chapter 1, the delay of reform implementation was due to review performed by organized legal profession. In this section, we view delay of reform implementation as arising via civil society's involvement in the public process, as stipulated among others by Roland (2004), Hoff and Stiglitz (2004), Beck and Laeven (2005), Campos (1999) and Dethier at al. (1999). Taking this into account, we now merely summarize the model's main qualitative features and refer the reader to chapter 1 for the mathematical description of the model. In the process of designing and implementing institutional reform, the government faces the competing demands of different interests. Rent-seeking interest groups ? some with objectives more congruent with the ruling government than others – possess better information about the effects of proposed reforms than the politicians. Interest groups support politicians, contributing funds when submitting reform proposals. A politician forming the government weighs these contributions against the pursuit of the general welfare, the latter being relatively more important when democratic constraints are tighter (i.e. when ? is greater). Since interest groups submit self-serving reforms for consideration, the politician faces an informational problem because some reforms are damaging to the politician and others are not. A government's conscious decision to slow down the reform process ameliorates the politicians' adverse selection problem. The government may, or may not, allow for active participation of non-partisan, policy-minded civil society groups that have the capacity to stimulate the debate about the reform. The autonomy of civil society adds
88
political legitimacy and credibility to the decision to deliberate on reform proposals and delay their implementation in the face of powerful interest groups pressing for quick reforms. Civil society's participation opens up the reform process to public debate and delays the adoption of reforms. Delaying reforms will be costly for interest groups, since the government might lose power (with probability 1?p) before implementation occurs. This cost changes the incentives of the interest groups when submitting reforms: initiatives with lower payoffs are dropped. Institutional reform may therefore improve (in line with Proposition 2 in chapter 1).132 Prevailing political conditions critically influence the government's decision to promote or suppress civil society's involvement (in line with Proposition 3 in chapter 1). When democracy is stronger (when ? is greater), the politician weighs general welfare more heavily than interest groups' contributions. Political stability also matters. When government turnover is more likely (when p is smaller), the politician's expected benefit from delaying a reform program to obtain better reforms, but later, declines. Democracy and political stability thereby become substitutes. The politician endorses civil society's involvement in an unstable environment only at high levels of democracy. When there is political stability, however, allowing for civil society's deliberation is attractive to the politician even at lower levels of democracy. This could be precisely the reason why Pérez-Diaz, for example, emphasizes that while civil society
Recall that the delay of a reform program improves the outcome of reform from society's standpoint because the preferred reform program of the interest group aligned with the ruling political option decreases social welfare (in notation of chapter 1: B?3+B?3+B?2+B?2<0). If, in contrast, the lobbying interest group's preferred reform program although not first best for society - increases social welfare (B?3+B?3+B?2+B?2>0), then it may be easily shown that delaying reform implementation is socially undesirable when political instability is either too high or too low, but may still be socially desirable for middling ranges of political instability. See Appendix 2 for analysis.
132
89
can fully emerge only in conjunction with a limited state that observes the rule of law, it nevertheless can emerge under authoritarian regimes (Bell 1997, p. 86). 3.5.2 When Does Aid Make a Difference? The Role of Political Conditions Revisited Aiding civil society may take on many hues, as argued in section 3.4. A more vibrant and autonomous civil society sector increases the transparency of the reform process, and effectively put a hold on the self-interested government's actions. We investigate the following question: In a setting where the civil society is easily encroached by the state, when does fostering civil society lead to improvement in social welfare?133 For convenience, let us focus on equilibria when the politician's self-interest matters. When ?<? ?, the politician in power is faced with an informational problem and approves the submitted reform program that is social welfare reducing.134 We assume the donor's decision to aid civil society does not bind the politician to endorse civil society: once civil society has been "strengthened", it is still up to the politician's discretion to decide whether to assure civil society's participation in the reform process or not.135 For simplicity, we also assume that the political cost of suppressing the civil society before the donors' have confirmed their participation equals 0. After the aid has been channeled, let the political cost of suppressing civil society be c? as in section 1.5 of chapter 1. c is now strictly positive.
133
The qualitative conclusions are not different if we instead analyze the effect of aid for the purpose of bringing civil society, capable of participating in public affairs, into existence (rather than strengthening an already existing civil society).
? the politician's aims coincide with society's aims: the politician rejects Recall that in the less interesting case of ?>? the submitted social welfare reducing reform program.
134 135
Effectively, we added a stage at the very beginning of the game as displayed in Figure 2 in chapter 1. This assumption certainly reflects the practice of civil society aid. See e.g. Ottaway and Carothers (2000) and Carothers (2004).
90
Let W denote social welfare and V the politician's payoff, as in chapter 1. Let superscript ? indicate no participation of civil society in the reform process, and let superscript + stand for involvement of civil society in the reform process. Define Unoaid as the social welfare when civil society exists but the political cost of preventing it to deliberate on the reform proposal equals zero (c = 0). Applying analysis from section 1.4 of chapter 1, Unoaid = W+ if V+ > V? and Unoaid = W? if V+ < V?. Let Uaid be the social welfare when the civil society has been aided, so c>0. Then, applying analysis from section 1.5 of chapter 1, Uaid = W+ if V+ > V? ? c? and Uaid = W? if V+ < V? ? c?. Accordingly, we have the following proposition: ? and c>0. Define Proposition.136 Let ?<? P = {(p,?): p<??A/B?3 and V? ? c? < V+ < V?} Q = {(p,?): p?[??A/B?3,??A/B?2) and V? ? c? < V+ < V?} R = {(p,?): p<??A/B?2 and V? ? c? < V+ < V?} S = {(p,?): p>??A/B?2 and V+ > V?} P, Q, R and S are (i) non-empty, (ii) pairwise mutually exclusive, (iii) P?Q?R?S = {(p,?): p?(0,1) and ??(0,? ?)} Define also ? ? Uaid ? Unoaid. Then: (iv) ? = 0 for all (p,?)?R?S, (v) ? = ?1 for all (p,?)?P, (vi) ? = ?2 for all (p,?)?Q, where ?2 > ?1 > 0. Figure 10 assists the interpretation of Proposition. Figure 10 shows that aiding civil society is superfluous when the attained levels of democracy and political stability are relatively low (region R); the ruling government is willing to incur the political cost of suppressing the civil society even once the latter has been strengthened. In contrast, when the attained levels of democracy and political stability are already high, there exists no need to aid civil society (region S).
136
This proposition is a corollary to Propositions 4 and 5 in chapter 1, so we omit its proof.
91
It is therefore in regions P and Q that political conditions are such that aiding civil society will improve social welfare. In region P, the attained levels of democracy are already considerable, but a country is still plagued with political instability. In region Q, both the attained levels of democracy and political stability are still relatively modest but have surpassed some minimum levels. Moreover, while aiding civil society improves social welfare in both regions P and Q, the aid is most effective in increasing social welfare when political instability has already been reduced (in region Q). 3.5.3 Judging the Success of Civil Society Aid: A Conjectural Interpretation The above Proposition confirms the importance of initial political conditions in the country where civil society is to be aided, as stipulated by Quigley (2000). In addition, it proposes a conjectural explanation for the observed variation in local governments' responses to the civil society aid project in countries undergoing postcommunist transition. Note first that, as a benchmark, the group of democratic and politically stable developed countries may be placed in Region S of Figure 10. In region S it is in the politicians' best interest to endorse, or at least not to impinge on, the existing civil society sector. Aiding civil society in developed countries is, according to the Proposition, superfluous. The transition countries during the early 1990s differed substantially in their political and economic conditions, and so did the experience with civil society aid. In the advanced transition countries (such as Poland, Slovenia, Czech Republic and Hungary), the levels of democracy and political stability were already considerable. Accordingly, this group of transition countries may be placed in regions Q and P. As predicted by the
92
model, the donors' efforts to foster civil society were successful, with the local governments actively endorsing civil society's participation in the reform process.137 The less democratic and politically less stable group of transition countries may be placed in region R. In this group, the ruling governments were often reluctant, or outright opposed, to promoting civil society (Quigley 2000), as the model predicts. The international donors' attempts to foster civil society were destined to fail. And they largely did, as Quigley (2000) argues about experiences of Romania, Bulgaria, Serbia and Slovakia, for example.138 The complex phenomenon of transition manifested through implementation of large-scale institutional reforms. The success of the latter was comparatively greater in the relatively more democratic Poland, Slovenia, Czech Republic and Hungary than in the less democratic and politically unstable group of transition countries. A conjectural interpretation of the proposition above also suggests that greater civil society participation in the former group of countries – in part assured by international aid – may have contributed to the quality of reform. The empirical evidence provided by Beck and Laeven (2005), Campos (1999) and Dethier at al. (1999) appears to confirm this supposition. The proposition above also provides support for initial phases of democracy project, when the donors aimed at directly fostering democracy (through ensuring impartial and democratic elections, for example) and pursued reform of major legislative
137
"In Poland and Slovenia, for example, the government was generally receptive to the [civil society] sector and gave it considerable space within which to operate" (Ottaway and Carothers 2000, p. 205).
138
"In Romania and Bulgaria, for example, the leaders…were opposed to strengthening civil society, and created roadblocks that delayed the establishment of foundations in their countries" (Ottaway and Carothers 2000, p. 204). Likewise, "[in] Slovakia and Serbia…the government provided no space or limited space for extragovernmental activities" (Ottaway and Carothers 2000, p.205).
93
and judicial institutions. In the light of the proposed model, earlier phases of democracy aid project may be interpreted as attempts to raise the initial levels of democracy (thereby increasing ?) and reduce political instability (thereby increasing p). It was only later on that civil society aid became means of promoting good governance and effective institutions. Interpretation of Proposition, however, suggests that aid to policy-oriented, non-partisan civil society organizations perhaps often came too early to countries undergoing post-communist transition. For the civil society sector to properly function as a bulwark against the state, the condition of minimum attained levels of democracy and political stability must already be fulfilled.139 3.6 Summing Up The many facets of civil society provide for a fruitful research agenda. Only very recently has the economics literature started illuminating the role of civil society for service provision and social capital accumulation. This chapter focused on the role of civil society for building institutions of good governance. Civil society's active deliberation on the proposed reform agenda restricts the set of viable proposals for special interests. A vibrant civil society may therefore be vital to adopting good institutions. This notion of civil society as a "bulwark against the state" underpinned the initial stages of civil society aid in early 1990s to post-communist countries as part of the greater democracy aid project. The scant available evidence on the aid's success reveals that aiding non-partisan, policy-oriented civil society organizations is no panacea for
139
In terms of Figure 1, a country must be placed in region Q rather than R for civil society aid to increase social welfare.
94
promoting good governance. Much is left to be understood.140 Our model predicts that aiding civil society improves social welfare as long as the political conditions in the recipient country are such that endorsing civil society's participation is in the politician's interest. Aiding civil society in an unstable authoritarian regime, for example, will be unproductive. A fundamental problem with aiding civil society lies in the bewildering array of organizations it encompasses. Accordingly, "[some] civil society groups may stand for 'higher' – that is, nonmaterial – principles and values, but much of civil society is preoccupied with the pursuit of private and frequently parochial and grubby ends" (Carothers 2004, p. 101). Aiding interest groups is hardly justifiable for economists (see, e.g., Olson 1982). Yet aiding a subset of civil society groups may nevertheless be beneficial even if civil society organizations do not pursue "higher goals". CSOs need not have a direct stake in the reform process.141 They may participate in the politics of fundamental institutional reforms perhaps to extract rents from prospective winners and losers from reform, or simply to raise their own prestige.142 By preventing hasty reform implementation, civil society's involvement may, as argued above, thus improve institutional reform.143
140
Thomas Carothers’ "…summation of ten years of Western civil-society aid to Eastern Europe and the former Soviet Union…boils down to a rather modest conclusion: They went to teach; they stayed to learn; they are learning still" (Carothers 2004, p.120).
141 142 143
In contrast, interest groups ? and ? in the model of chapter 1 do have a direct stake in reform. Lawyers associations are again an example of such civil society groups.
To a certain extent, however, our analysis still overstates the positive contribution of civil society since we do not account for costs associated with extracting rents from ? and ? in our social welfare calculation.
95
Appendix 1: Figures
Figure 1: The Set of Reforms
T
Benefit to ? (B?j) Reform 3 Reform 1 Reform 2 Benefit to ? (B?j) Reform 5 Reform 6 Reform 4 T
96
Figure 2: The Timeline
A stays in power with probability p Interest groups lobby for reforms Politician chooses to use review by profession Review
Implementation Interest groups lobby for reforms that have Interest groups lobby for reforms that have not been subject to review yet Interest groups choose not to lobby for f
Implementation
A is elected, p and ? are realized
Interest groups choose not to lobby for reforms
B comes to power with probability 1?p
Review
End of reform proces
Politician chooses not to use review by profession
Interest groups choose not to lobby for reforms
Interest groups lobby for reforms
Implementation
Time period 1
Time period 2
97
Figure 3: The Politician's Welfare and Aggregate Welfare with and without a Powerless Profession
democracy (?)
1
? ?
W+=W? V+=V?
W+=W? V+=V?
?C(p)
W+>W? V+>V?
?
W+<W? V+>V?
W >W V+<V?
??A/B?3
+
?
?G
?P
p?(?)
?S
0
??A/B?2
1
stability (p)
Notes: 1. W+ = aggregate welfare with professional review 2. W? = aggregate welfare without professional review 3. V+ = politician’s welfare with professional review 4. V? = politician’s welfare without professional review 5. Boundary between use of profession and non-use 7. S = The Soviet Union before Gorbachev 8. G = Georgia during early transition 9. P = Poland during early transition
98
Figure 4: Professional Political Power, Democracy, and Stability
democracy ?T(p,c)
? ? ?T(p,c2) ?T(p,c1)
?
??A/B?3
??A/B?2
stability (p)
Notes: 1. c1>c2?0 2. ?T(p,0) lies along the solid line in Figure 3.
99
Figure 5: Political Power and the Use of the Profession
democracy (?)
1
? ?
The profession directly aids the politician through its effect on institutional reforms
The profession's political power dictates its use by the politician
E? F3?
? F2
Politician does not use the profession
? F1
p*(?)
0
??A/B?3
??A/B?2
1
stability (p)
Boundary between use of profession and non-use when political power (c) is positive. Boundary between use of profession and non-use when political power (c) is zero. (This line is the same one as in Figure 3.) F1 = F2 = F3 = E = France under Louis XIV, Napoleon (c=0 and is pertinent). France early in the reign of Louis XV (c>0 and is pertinent). France late in the reign of Louis XV (c=0 and is pertinent). France soon after the revolution (c=0 and is pertinent). England under James II (c>0 and is pertinent).
100
Figure 6: The Political Cost of Removing the Profession
Cm(?, p)
2??A Cm(?,p1) c? ??A Cm(?,p2) Cm(?,p3) Cm(?,p4) p1<??A/B?3 p2=??A/B?3 p3, p4?[??A/B?3,??A/B?2) p1<p2<p3<p4
0 ? (p ,c) ? (p ,c) ? (p ,c)
T 2 T 4 T 3
? (p ,c)
T
1
? ?
?
101
Figure 7: Timeline of the Regulatory Process
t=0 Regulatory rights assigned
t=1 Gov’t passes enabling legislation: L = L*
t=2 ? realized: L = L*+?
t=3 If no renegotiations, default ?L implemented; however, renegotiations will occur
t=4 L = L*+?+?L determined, payoffs realized
Figure 8: SR vs. R: Efficiency vs. Government's Incentives
?2/(p?c)2 VR = VSR
4+3?R+?SR 4(?R??SR)
I
IV III
0
II
WR = WSR
1
?
102
Figure 9: Legal Origin and Regulatory Regime Choice
Figure 9(a): Legal Origin and Regulatory Regime Choice when ?2e > ?2f ?2/(p?c)2
Figure 9(b): Legal Origin and Regulatory Regime Choice when ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0 ?2/(p?c)2
4+3?R,f+?SR,f 4(?R,f??SR,f) VR(?R,f) = VSR(?SR,f) 4+3?R,e+?SR,e 4(?R,e??SR,e)
4+3?R+?SR 4(?R??SR) V =V
R SR
VR(?R,e) = VSR(?SR,e) ?2e (p?c)2 ?2f (p?c)2 0 {?: VR(?2e) ? VSR(?2e)} {?: VR(?2f) ? VSR(?2f)} 1 ? 0
A1
A2
1 ? A2 = {(?,?2/(p?c)2): VR(?R,e) ? VSR(?SR,e)} A1?A2 = {(?,?2/(p?c)2): VR(?R,f) ? VSR(?SR,f)}
103
Figure 10: Effectiveness of Aid to Civil Society
democracy (?)
? ?
(? = ?1>0)
P
Q
(? = 0) 0 ??A/B?3
(? = 0)
S
(? = ?2 > ?1)
R
??A/B?2
1
stability (p)
Notes: Refer to chapter 1, section 1.5 and Propositions 4-5, for definition of ?T(p,c). The thicker dashed line represent the ?T(p,0) function. The thicker solid line represents the ?T(p,c) function, with c>0. ? = Uaid ? Unoaid, as defined in Proposition in the text.
104
Appendix 2: Tables
Table 1: Summary Statistics by Regulatory Arrangements: Means (and Standard Deviations) of Variables Definition (and Source) All observations 0.39 (0.49) 0.17 (0.38) 0.31 (0.47) 0.07 (0.25) 0.07 (0.25) 8.76 (0.92) By regulatory arrangement Central SelfRegulation Combination Regulation 0.38 (0.49) 0.31 (0.47) 0.15 (0.37) 0.05 (0.22) 0.10 (0.31) 8.57 (0.87) 0.43 (0.51) 0.05 (0.22) 0.38 (0.50) 0.09 (0.30) 0.05 (0.22) 9.09 (0.99) 0.35 (0.49) 0 (0) 0.58 (0.51) 0.06 (0.24) 0 (0) 8.80 (0.90)
Variable French
= 1 if French legal origin (La Porta et al. 1999) = 1 if socialist legal socialist origin (La Porta et al. 1999) = 1 if English legal English origin (La Porta et al. 1999) = 1 if German legal German origin (La Porta et al. 1999) = 1 if Scandinavian Scandin legal origin (La Porta et al. 1999) lgnipc1991 Logarithm of GNI per capita, PPP current international $ (World Development Indicators) Logarithm of lpop1991 population total (World Development Indicators) Number of observations
16.28 (1.73)
16.26 (1.79)
16.48 (1.78)
16.09 (1.61)
77
39
21
17
105
Table 2: The Relationship between Legal Origin and Regulatory Regime: Ordered Probit Results Variable socialist English German Scandin lgnipc1991 lpop1991 threshold 1 threshold 2 Coefficient ?1.4818 0.8019 ?0.2047 ?1.3623 0.3807 ?0.0113 3.1226 4.0820 Standard Error* 0.5232 0.3349 0.5474 0.6230 0.1736 0.0835 2.3729 2.4082 z ?2.83 2.39 ?0.37 ?2.19 2.19 ?0.14 P>?z? 0.005 0.017 0.708 0.029 0.028 0.892
(Ancillary parameters)
Log-pseudo likelihood = ?65.7562 Pseudo R2 = 0.1728 Number of observations = 77
*
Robust standard errors.
Notes: 1. French origin is the omitted dummy. 2. Self-regulation is the highest category of the dependent variable.
106
Appendix 3: Proofs of Propositions in Chapter 1
? the politician accepts all reform proposals. Consider reform Proposition 1. When ?<? j?{2,3}. ? will never want to lobby for j?{2,3}, since B?j<0 for j?{2,3}. There are four possible sub-games at the beginning of period 2, depending on whether ? chose to lobby for reform j?{2,3} in period 1 or not and whether A or B is in power in period 2. 1. If ? lobbied for j?{2,3} in period 1 and A is in power in period 2, then the payoff to ? when it lobbies for reform j?{2,3} in period 2 is B?j???A and B?j when it does not lobby. Therefore, ? does not lobby. 2. If ? lobbied for j?{2,3} in period 1 and B is in power in period 2, then the payoff to ? when it lobbies for reform j?{2,3} in period 2 is B?j???B<0 and 0 when it does not lobby. Therefore, ? does not lobby. 3. If ? did not lobby for j?{2,3} in period 1 and politician A is in power in period 2, then the payoff to ? when it lobbies for reform j?{2,3} in period 2 is ???A<0 and 0 when it does not lobby. Therefore, ? does not lobby. 4. If ? did not lobby for j?{2,3} in period 1 and politician B is in power in period 2, then the payoff to interest group ? when it lobbies for reform j?{2,3} in period 2 is ???B<0 and 0 when it does not lobby. Therefore, ? does not lobby. If ? chooses to lobby for j?{2,3} in period 1, its payoff is pB?j???A. If ? chooses not to lobby for j?{2,3} in period 1, its payoff is 0. Therefore, ? chooses to lobby for j?{2,3} in period 1 if and only if p???A/B?j. Consider now reform j?{4,5}. ? will never lobby for j?{4,5}, since B?j<0 for j?{4,5}. There are four possible sub-games at the beginning of period 2, depending on whether ? chose to lobby for j?{4,5} in period 1 or not and whether politician A or politician B is in power in period 2. 1. If ? lobbied for j?{4,5} in period 1 and A is in power in period 2, then the payoff to ? when it lobbies for j?{4,5} in period 2 is B?j???A<0 and B?j when it does not lobby. Therefore, ? does not lobby. 2. If ? lobbied for j?{4,5} in period 1 and B is in power in period 2, then the payoff to ? when it lobbies for j?{4,5} in period 2 is B?j???B>0 and 0 when it does not lobby. Therefore, ? lobbies. 3. If ? did not lobby for j?{4,5} in period 1 and B is in power in period 2, then the payoff to ? when it lobbies for j?{4,5} in period 2 is ???B<0 and 0 when it does not lobby. Therefore, ? does not lobby. 4. If ? did not lobby for j?{4,5} in period 1 and politician A is in power in period 2, then the payoff to ? when it lobbies for j?{4,5} in period 2 is ???A<0 and 0 when it does not lobby. Therefore, ? does not lobby. If ? chooses to lobby for j?{4,5} in period 1, its payoff is pB?j+(1?p)(B?j???B)???A<0. If ? chooses not to lobby for j?{4,5} in period 1, its payoff is 0. Therefore, ? chooses not to lobby for reform j?{4,5} in period 1.
107
Proposition 2. If p<??A/B?3, ? chooses not to lobby, so W+ = 0>W?. If ??A/B?3?p<??A/B?2, ? lobbies for reform 3, so W+ = p(B?3+B?3)>W?. If p???A/B?2, ? lobbies for reforms 2 and 3, so W+ = p(B?2+B?2+B?3+B?3)>W?. Proposition 3. If p<??A/B?3, ? chooses not to lobby in period 1, so V+ = 0<V?. If p???A/B?2, ? lobbies for 2 and 3 in period 1, so V+ = ?p(B?2+B?2+B?3+B?3)+(1??)2??A>V?. If ??A/B?3?p<??A/B?2, ? lobbies for 3 in period 1, so V+=?p(B?3+B?3)+??A. Then, V+>(<)V??p>(<)p?(?)?1+[?(B?2+B?2)+(1??)??A]/[?(B?3+B?3)]. The following is useful for the purposes of constructing a diagram to examine the tradeoffs between ? and p. Observe that p?(?)<(>)1 if and only if ?>(<)?. It can be easily shown that ?p?(?)/??<0 and ?2p?(?)/??2>0. Define ?L = ??A/[(??A/B?2?1)(B?3+B?3)+?B?2+B?2?+??A] and ?H = ??A/[(??A/B?3?1)(B?3+B?3)+?B?2+B?2?+??A]. ¯, p*(?H)=??A/B?3 and p*(?L)=??A/B?2. Then ?<?L<?H<? Proposition 4. Define Cm(p,?) as the maximum political cost that the politician is willing to incur to remove the profession from its advisory role. If V+<V?, Cm(p,?) = V?(p,?) ? V+(p,?) and if V+?V?, Cm(p,?) = 0. More precisely, ? ?[(B?2+B?2)+(B?3+B?3)]+(1??)2??A, when p<??A/B?3 and ?<? ? C (p,?) = ?[(B?2+B?2)+(B?3+B?3)]?p?(B?3+B?3)+(1??)??A, when p<p (?) and p?[??A/B?3,??A/B?2) 0, elsewhere. The politician chooses not to use the profession for all values of ? and p such that Cm(p,?) > c?. Cm(p,?) has the following properties: ?, 1. When p<??A/B?3 and ?<? ?Cm(p,?)/?p = 0, ?Cm(p,?)/?? = [(B?2+B?2)+(B?3+B?3)]?2??A<0 2. When p?[??A/B?3,??A/B?2) and p<p?(?), ?Cm(p,?)/?p = ?(B?2+B?2)<0, ?Cm(p,?)/?? = [(B?2+B?2)+(B?3+B?3)]?p(B?3+B?3)???A<0, ?2Cm(p,?)/???p = ?(B?2+B?2)<0. 3. Elsewhere, Cm(p,?) = 0. 4. Two additional properties that enable us to construct Figure 6 are:
m
?C m (p, ? 0 )/?? C m (p, ? 0 )
p < ? ?A /B ?3
< ?C m (p, ? 0 )/??
p = ? ?A /B ?3
for all ? 0 < ? for all ? 0 < ?
p < ? ?A /B ?3
> C m (p, ? 0 )
p?[? ?A /B ?3 , ? ?A /B ?3 ) and p < p*(? 0 )
The properties of Cm(p,?) summarized above imply than for any pair (p,c), there exists a unique value of ?T(p,c)>0 such that Cm(p,?T(p,c)) = c?T(p,c) if V+<V?. On the other hand, Cm(p,?) = 0 if V+?V? and hence ?T(p,c) = 0. This proves that ?T(p,c) is a function. To
108
prove that ?T(p,c) is discontinuous at p = ??A/B?2 and at p = ??A/B?3, and points (ii) – (iv), note also the following: 1. When p<??A/B?3, ?T(p,c) is defined by c?T(p,c) = ?T(p,c)(B?2+B?2+B?3+B?3)+(1??T(p,c))2??A, so that ?T(p,c) = 2??A/[c?(B?2+B?2+B?3+B?3)+2??A]. Then, ??T(p,c)/?p = 0. 2. When p?[??A/B?3,??A/B?2), ?T(p,c) is defined by c?T(p,c) = ?T(p,c)(B?2+B?2+B?3+B?3)?p?T(p,c)(B?3+B?3)+(1??T(p,c))??A, so that ?T(p,c) = ??A/[c?(B?2+B?2+B?3+B?3)+p(B?3+B?3)+??A]. ??T(p,c)/?p = ?(B?3+B?3)?T(p,c)/[c?(B?2+B?2+B?3+B?3)+(B?3+B?3)p+??A] < 0 and ?2?T(p,c)/?p2 > 0. 3. When p???A/B?3, ?T(p,c) = 0. 4. ? T (p, c 0 ) > ? T (p, c 0 ) > ? T (p, c 0 ) = 0 for all c 0 > 0.
p < ? ?A /B ?3 p?[? ?A /B ?3 , ? ?A /B ?2 ) p ? ? ?A /B ?2
Proposition 5. Let c1>c2?0. Define ?T(p,c1,c2) = ?T(p,c2) ? ?T(p,c1). 1. When p<??A/B?3, ?T(p,c1,c2) = 2??A/[c2?(B?2+B?2+B?3+B?3)+2??A] ? 2??A/[c1?(B?2+B?2+B?3+B?3)+2??A] > 0 and ??T(p,c1,c2)/?p = 0. 2. When p?[??A/B?3,??A/B?2), ?T(p,c1,c2) = ??A/[c2?(B?2+B?2+B?3+B?3)+p(B?3+B?3)+??A] ? ??A/[c1?(B?2+B?2+B?3+B?3)+p(B?3+B?3)+??A] > 0 and ??T(p,c1,c2)/?p < 0. 3. When p???A/B?2, ?T(p,c2) = ?T(p,c1) = 0, so ?T(p,c1,c2) = 0. ? and when p?[??A/B?3,??A/B?2), Note that when p<??A/B?3, ?T(p,0) = ? T ?1 ? (p,0) = p* (?).
109
Appendix 4: Model with Reform 1
This appendix presents the model, as developed in sections 1.3 and 1.4 of the text, but with Pareto-improving reform 1 explicitly included in the analysis. No other features of the model have been changed. We show that (a) reform 1 will always be implemented and (b) the model’s qualitative features do not change when reform 1 is included in the analysis. The presentation here closely parallels that in the main text. The following table restates the payoff structure for reforms 1, 2 and 3: Reform type\Payoff Reform 1 Reform 2 Reform 3 Interest group ? B?1>0 B?2>0 B?3>0 Interest group ? B?1>0 B?2<0 B?3<0 Aggregate (? and ?) B?1+B?1>0 B?2+B?2<0 B?3+B?3>0
The sequence of events is as described in section 1.3 of the main text. There is a single reform process beginning when one politician, say A, takes the reins of power. The model covers two sub-periods, A’s initial incumbency and the following time interval when either A or B forms the government. The decision on whether to use professional review is made at the beginning of A's incumbency and holds for the whole reform process. Let p?(0,1) be the known, exogenous probability that A stays in power in the second period. In both periods, each interest group decides whether or not to submit a proposal without knowing what the other is doing. The government then decides whether to accept the proposal. If it does, it receives the submission fee and if there is no professional review the measure is immediately passed. If there is professional review, the measure is debated until the second period. Then, if A remains in office, the reforms submitted in period 1 are implemented. If B takes over, each interest group decides again whether to submit a proposal. If the reforms submitted in period 2 are the ones that already passed through professional review in period 1, they are 110
implemented. If the reforms submitted in period 2 differ from the ones submitted in period 1, they are delayed beyond the end of this reform process.144 We maintain the same assumptions as in the chapter itself: Assumption A1. B?2=B?3 Assumption A2. B?2+B?2+B?3+B?3<0 Assumption A3. For all j such that B?j>0, ??A<B?j<??B For all j such that B?j>0, ??B<B?j<??A Assumption A4. Aggregate welfare from implementing reform j is B?j+B?j. Assumption A5: The politician's payoff from j is ?(B?j+B?j)+(1??)?ik, ??(0,1).
? are defined as follows: ? = ??A/[?B?2+B?2?+??A] and Recall that ? and ¯ ¯= ??A/{??A+½[?B?2+B?2??(B?3+B?3)]}<1. Reforms 1 and 3 are always desirable for ? politician A. Reform 2 is undesirable for A if ?>? and desirable if ?<?. Thus, when ?>?, the politician faces an adverse selection problem. Politician A likes the average reform ¯. consisting of reforms 2 and 3 if and only if ?<? Without a screening mechanism, the politician either approves all reforms or none of them. A approves all reforms if and only if [?(B?1+B?1)+(1??)??A]+[?(B?2+B?2)+(1??)??A]+[?(B?3+B?3)+(1??)??A]>0 ??[B?1+B?1+B?2+B?2+B?3+B?3?3??A]>?3??A. ~ Define ?=??A/{??A?(?)[B?1+B?1+B?2+B?2+B?3+B?3]}. It is easy to show that if ~ ??(0,1) then the immediately preceding inequality is satisfied for all ??(0,1). Hence,
This assumption closes the model without an infinite horizon. It is equivalent to focusing on a single reform process that must take place within a certain time window, which is a common feature of the circumstances surrounding institutional reforms.
144
111
~ when there is no professional review A approves all reforms if ??(0,1), which is the case ~ if and only if B?1+B?1+B?2+B?2+B?3+B?3>0. When B?1+B?1+B?2+B?2+B?3+B?3<0, 1>?>¯ ? ~ and A approves all reforms if and only if ?<?. As in the main text, we split the analysis into two parts. The more interesting case ~ ~ (corresponding to ?<¯ ? in the paper) is when either ??(0,1) or 0<?<?<1 implying that A approves all reforms when there is no professional review. The other case ¯ in the paper) is when 0<~ (corresponding to ?>? ?<?<1 and A rejects all reforms when there is no professional review. Propositions A4.1?A4.3 below match the corresponding ~ ~ propositions in the text and hold when either ??(0,1) or 0<?<?<1. Figure A4.1 summarizes the conclusions graphically. ~ ~ Proposition A4.1. Assume that A is in power, either ??(0,1) or 0<?<?<1, and there is professional review. In period 1, ? lobbies for reform 1 for any p and for reforms 2 and 3 if and only if p???A/B?2 and p???A/B?3, respectively. In period 2, ? does not lobby for any reform regardless of which politician is in power. ? does not lobby for any of the reforms in period 1, and lobbies for reform 1 in period 2 if politician B is in power. Reform 1 will therefore always be implemented. (See the end of this appendix for the lengthy proof of Proposition A4.1.) ~ ~ Proposition A4.2. When either ??(0,1) or 0<?<?<1, social welfare is always higher with the profession than without it. Proof: Without the profession, W? = B?1+B?1+B?2+B?2+B?3+B?3. With the profession: For p<??A/B?3, W+ = B?1+B?1>W?. For ??A/B?3?p<??A/B?2, W+ = B?1+B?1+p(B?3+B?3)>W?. For p???A/B?2, W+= B?1+B?1+p(B?2+B?2+B?3+B?3)>W?.
?
112
~ ~ Proposition A4.3. Assume that either ??(0,1) or 0<?<?<1. For p<??A/B?3, the politician is worse off with the profession if and only if ?<¯ ? and better off with ¯ the profession if and only if ?>?. For p???A/B?2, the politician is better off with the profession. For p?[??A/B?3,??A/B?2), the politician is better off with the profession if and only if p>p?(?) and worse off with the profession if and only if p<p?(?), where p?(?)?1+[?(B?2+B?2)+(1??)??A]/[?(B?3+B?3)]. Proof: Without the profession, V?=?(B?1+B?1+B?2+B?2+B?3+B?3)+(1??)3??A. With the profession: For p<??A/B?3, V+=?(B?1+B?1)+(1??)??A. V+>(<)V??0>(<)?(B?2+B?2+B?3+B?3)+(1??)2??A??>(<) ¯ ?. For ??A/B?3?p<??A/B?2, V+=?(B?1+B?1) +?p(B?3+B?3)+(1??)2??A. V+>(<)V??p>(<)1+[?(B?2+B?2)+(1??)??A]/?(B?3+B?3)?p*(?) The following is useful for the purposes of constructing Figure 1, which examines the trade-offs between ? and p. Observe that p?(?)<(>)1 if and only if ?>(<)?. It can be easily shown that ?p?(?)/??<0 and ?2p?(?)/??2>0. Define ?L = ??A/[(??A/B?2?1)(B?3+B?3)+?B?2+B?2?+??A] and ?H = ??A/[(??A/B?3?1)(B?3+B?3)+?B?2+B?2?+??A]. Then ?<?L<?H<¯ ?, p*(?H)=??A/B?3, and p*(?L)=??A/B?2. For p???A/B?2, V+=?(B?1+B?1)+?p(B?2+B?2+B?3+B?3)+(1??)3??A. V+>V? because ?p(B?2+B?2+B?3+B?3)>?(B?2+B?2+B?3+B?3). ? ~ Finally, consider the outcomes when 0<?<?<1. The politician does not approve any reforms without a screening mechanism, so V? = W? = 0. With professional review, the reform-submission equilibria are the same as in Proposition 1 when ??A/B?2>p. For p<??A/B?3, reform 1 only is submitted and it is approved and implemented. Hence W+>W? and V+>V?. When p?[??A/B?3,??A/B?2), reforms 1 and 3 are submitted and approved, with reform 1 always implemented and reform 3 implemented if A stays in power in the second period. Again, W+>W? and V+>V?. In contrast, when p???A/B?2, the welfare analysis is less straightforward. We summarize it in the following proposition:
113
~ Proposition A4.4: Let 0<?<?<1 and p???A/B?2. Define: (i) ? = [B?1+B?1?3??A]/[?B?2+B?2??(B?3+B?3)] , (ii) ?=[B?1+B?1]/[?B?2+B?2??(B?3+B?3)], (iii) ?C(p)=??A/{??A?? [B?1+B?1+p(B?2+B?2+B?3+B?3)]} Then: (1) if p>? and ?>?C(p), W+=W? and V+=V? (2) if p<? or ?<?C(p), then V+>V? (3) if p<? or ?<?C(p), then W+>W? if p<? and W+< W? if p>? Proof: Given p???A/B?2, A approves ?’s submitted reform program consisting of reforms 1, 2, and 3 if and only if ?(B?1+B?1+p(B?2+B?2+B?3+B?3))+(1??)3??A>0 ??[B?1+B?1+p(B?2+B?2+B?3+B?3)?3??A]>?3??A. If B?1+B?1+p(B?2+B?2+B?3+B?3)?3??A>0?p<?, then A approves the submission ~ for all ??(?,1). In contrast, if B?1+B?1+p(B?2+B?2+B?3+B?3)?3??A<0?p>?, then A approves the submission if and only if ?<?C(p). Hence, A does not approve any reforms when p>? and ?>?C(p), which together ~ with 0<?<?<1 implies that W+=W?=0 and V+=V?=0. If p<? or ?<?C(p), then A approves submitted reforms implying that V+>V? and W+=B?1+B?1+p(B?2+B?2+B?3+B?3)>(<)0=W??p<(>)?. ? Proposition A4.4 facilitates the construction of Figures A4.2?A4.4. First, note ~ that the definitions of ? and ? imply ?>?, where 0<?<1 because 0<?<1, but ? need not be ~ positive. Also, to construct ?C(p) note that it is strictly decreasing in p>?, ?C(1)=?, ?C(?)=1, and ?C(p)?? as p?? from above. Three configurations of the parameters are possible: (1) ?<?<??A/B?2, (2) ?<??A/B?2<?, (3) ??A/B?2<?<?.
114
The mapping of the conclusions of Proposition A4.4 into each of these regions is straightforward, leading to Figures A4.2?A4.4.145 Figures A4.1?A4.4 graphically summarize the conclusions of this appendix and show that the model with reform 1 has all the same qualitative predictions as the model in the main text.
145
Note that in Figures A4.2 and A4.3 we let ?>0 without loss of generality.
115
Figure A4.1: B?1+B?1+B?2+B?2+B?3+B?3>0
?
1
¯ ?
W+>W? V+>V? W+>W? V+<V?
??A/B?3 ??A/B?2 1 p?(?)
?
0
p
Figure A4.2: B?1+B?1+B?2+B?2+B?3+B?3<0, ?<?<??A/B?2
?
1 ~ ? ¯ ?
W+=W? V+=V?
W >W V+>V? W+>W? V+<V?
? ? ??A/B?3
+
?
W <W V+>V?
+
?
?C(p)
?
p*(?)
0
??A/B?2
1
p
116
Figure A4.3: B?1+B?1+B?2+B?2+B?3+B?3<0, ?<??A/B?2<?
?
1 ~ ? ¯ ?
W+=W? V+=V? W+<W? V+>V?
?C(p)
W+>W? V+>V?
?
W+>W? V+<V?
? ? ??A/B?2
1
p*(?)
0
??A/B?3
p
Figure A4.4: B?1+B?1+B?2+B?2+B?3+B?3<0, ??A/B?2<?<?
?
1 ~ ? ¯ ?
W+=W? V+=V? W <W V+>V?
+ ?
?C(p)
W+>W? V+>V?
?
W+>W? V+<V?
? ? ??A/B?3 ??A/B?2
1
p*(?)
0
p
117
Proof of Proposition A4.1. See Appendix 3 for reforms 2, 3, 4 and 5. Reform 1 is interesting for both interest groups. They will therefore act strategically: assuming that interest groups cannot collude, either of them would prefer not to lobby for reform 1 given that the other interest group does. There are 8 possible situations (sub-games) at the beginning of period 2: 1. If only ? lobbied for reform 1 in period 1 and A is in power in period 2 then the following payoff matrix can be used to find the (Nash) equilibrium of the subgame: ? Lobby Lobby B?1???A, B?1???A ? Do not lobby B?1, B?1???A In the equilibrium of this sub-game neither ? nor ? lobbies. Do not lobby B?1???A, B?1 B?1, B?1
2. If only ? lobbied for reform 1 in period 1 and B is in power in period 2: ? Lobby Do not lobby Lobby B?1???B, B?1 B?1???B, B?1???B ? 0,0 Do not lobby B?1, B?1???B In the equilibrium of this sub-game, only ? lobbies. 3. If only ? lobbied for reform 1 in period 1 and A is in power in period 2: ? Lobby Do not lobby Lobby B?1???A, B?1???A B?1???A, B?1 ? Do not lobby B?1, B?1???A B?1, B?1 In the equilibrium of this sub-game neither ? nor ? lobbies. 4. If only ? lobbied for reform 1 in period 1 and B is in power in period 2: ? Lobby Do not lobby Lobby B?1???B, B?1 B?1???B, B?1???B ? Do not lobby 0,0 B?1, B?1???B In the equilibrium of this sub-game, only ? lobbies. 5. If both interest groups lobbied for reform 1 in period 1 and that A is in power in period 2: ? Lobby Do not lobby Lobby B?1???A, B?1???A B?1???A, B?1 ? Do not lobby B?1, B?1 B?1, B?1???A In the equilibrium of this sub-game neither ? nor ? lobbies.
118
6. If both interest groups lobbied for reform 1 in period 1 and B is in power in period 2: ? Lobby Do not lobby Lobby B?1???B, B?1???B B?1???B, B?1 ? Do not lobby 0,0 B?1, B?1???B In the equilibrium of this sub-game, only ? lobbies. 7. If none of the interest groups lobbied for reform 1 in period 1 and politician A is in power in period 2: Lobby Lobby ???A, ???A ? Do not lobby 0, ???A In the equilibrium of this sub-game neither ? nor ? lobbies. ? Do not lobby ???A, 0 0,0
8. If none of the interest groups lobbied for reform 1 in period 1 and politician B is in power in period 2: ? Lobby Do not lobby Lobby ???B, ???B ???B, 0 ? Do not lobby 0,0 0, ???B In the equilibrium of this sub-game neither ? nor ? lobbies. In period 1, the incumbent politician is politician A. The following payoff matrix can be used to find the (sub-game perfect Nash) equilibrium of the lobbying game for reform 1: ? Lobby B?1???A, pB?1+(1?p)(B?1???B)???A B?1, pB?1+(1?p)(B?1???B)???A Do not lobby B?1???A, pB?1+(1?p)(B?1???B) 0, 0
Lobby ? Do not lobby
Therefore, in the equilibrium, ? lobbies in period 1 and ? does not. In period 2, neither ? nor ? lobby if A stays in power; if B comes to power, ? does not lobby and ? does. That is, no matter who is in power in period 2, reform 1 is always implemented. ?
119
Appendix 5: Algebraic Derivations and Proofs of Results in Chapter 2
This appendix includes the derivation of expressions and results in the text. The derivations are often algebraically intensive, but conceptually easy to follow. In most cases, it is therefore sufficient to merely indicate the key steps involved. Derivation of L*i = A, i?{SR,R}. At t = 1, the government chooses L*i for i?{SR,R} to maximize E{A(L*i+?+?Li,d)?½[(L*i+?+?Li,d)2+?i(?Li,d)2] + A(L*i+?+?Li,b)?½[(L*i+?+?Li,b)2+?i(?Li,b)2] + p(L*i+?+?Li,b)?½[(L*i+?+?Li,b)2+?i(?Li,b)2] ? p(L*i+?+?Li,d)+½[(L*i+?+?Li,d)2+?i(?Li,d)2]}, where E is the expectations operator, ?Li,d = [ri?(L*i+?)]/(1+?i), with rSR = p, and rR = A, and ?Li,b = [½(A+p)?(L*i+?)]/(1+?i). Differentiating with respect to L*i through the expectations operator, taking expectations, simplifying, and setting equal to zero gives the first-order condition 2[?i/(1+?i)](A?L*i) = 0. Note that L*i = A is the unique maximizer since ?2[?i/(1+?i)]<0 for all L*i. Derivation of the final realization of L under regime i?{R,SR}. L*i+?+?Li,b = A+?+[½(A+p)?(A+?)]/(1+?i) = [?i(A+?)+½(A+p)]/(1+?i). Derivation of Wi and Vi, i?{SR,R}. The expected social welfare under regime i?(SR,R}, Wi = A(p+c?A)??i/(1+?i)?2+¼[1/(1+?i)]?(3??2)(p?c)2,
120
is obtained by simplifying Wi = E{(p+c)(L*i+?+?Li,b)?(L*i+?+?Li,b)2??i(?Li,b)2}. Similarly, the expected government's payoffs under R and SR, VR = ½A2?½[?R/(1+?R)]?2+?[1/(1+?R)]?2(p?c)2 and VSR = ½A2?½[?SR/(1+?SR)]?2??[1/(1+?SR)]?2(p?c)2, are obtained by simplifying: Vi = E{A(L*i+?+?Li,d)?½[(L*i+?+?Li,d)2+?i(?Li,d)2] + ½{A(L*i+?+?Li,b)?½[(L*i+?+?Li,b)2+?i(?Li,b)2] + p(L*i+?+?Li,b)?½[(L*i+?+?Li,b)2+?i(?Li,b)2] ?A(L*i+?+?Li,d)?½[(L*i+?+?Li,d)2+?i(?Li,d)2] ? p(L*i+?+?Li,d)+½[(L*i+?+?Li,d)2+?i(?Li,d)2]}}, where L*i, ?Li,b, and ?Li,d for i?{R,SR} are defined above. Proof of Results 1 and 2. Results 1 and 2 are obtained by a straightforward comparison of WSR with WR (Result 1), and VSR with VR (Result 2), so we omit the algebraic proof altogether. Proof of Result 3. Result 3(a) readily follows from Figure 9(a). To prove Result 3(b), recall that VR = VSR when ?2/(p?c)2?¼[(4+3?R+?SR)/(?R??SR)]?2 = 0. The VR = VSR line intersects the ?2/(p?c)2 axis at (4+3?R+?SR)/(?R??SR) when ? = 1. Proof of Result 3(b) proceeds by showing that a "change in legal origin" from common law to civil law, rotates the VR = VSR line upward, as shown in Figure 9(b). Define f(?R,?SR) ? (4+3?R+?SR)/(?R??SR). Then, the total differential of f(?R,?SR), ?f, equals ?[(1+?SR)/(?R??SR)2]??R+[(1+?R)/(?R??SR)2]??SR. Now suppose that the change in 121
the ?i's is induced by the "change in legal origin" from common law to civil law. In notation, ??SR ? ?SR,f??SR,e, ??R ? ?R,f??R,e. Both ??SR and ??SR are positive by assumption. Then, ?f > 0 if (and only if) ?SR,f??SR,e > (?R,f??R,e)[(1+?SR,e)/(1+?R,e)] > 0. The latter expression is, however, implied by ?SR,f??SR,e ? ?R,f??R,e > 0, or, equivalently, by ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0. That is, when ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0, as Figure 9(b) indicates, the VR(?i,o) = VSR(?i,o) line rotates upward when ?i,o increases from ?i,e to ?i,f, i?{SR,R}.
122
Appendix 6: Civil Society as Vehicle for Service Provision
This appendix reviews and discusses some of the literature on CSOs as vehicles of service provision. This section has been relegated to the appendix since it does not immediately underpin the argument on civil society aid in chapter 3. The review below argues that, while progress has undoubtedly been made, the service-provision facet of civil society is still largely understudied. Azam (2003) provides a theoretical underpinning for civil society's role in alleviating poverty. His model is about the rich and the poor. While all rich care about the wellbeing of the poor, the rich are of two types: 'absolutists' and 'relativists'. The two types differ in their views about poverty (although the utility of both types is affected negatively by the number of the poor): the marginal willingness of the relativists to help the poor is increasing in their own consumption level, while absolutists' one is independent of theirs. The political system is such that public redistribution in favor of the poor (taxes) is chosen by the largest number of the rich, assumed to be absolutists. Given tax, the rich may also contribute an additional voluntary transfer. Only the relativists will be interested in contributing voluntarily to alleviate poverty. Azam views the aggregate value of voluntary transfers from the rich to the poor as an indicator of development of civil society. In his model, the minority of the rich selecting the level of voluntary contributions (the relativists) is assumed to be organized as a general assembly of NGOs (Azam thereby also assumes NGOs have developed some sort of mechanisms to overcome free-riding). In a later part of the paper, Azam also assumes that NGOs – the minority of the rich contributing voluntary transfers ? may initiate the redistribution
123
process by lobbying the government: the 'relativist' rich offer side payments to the government (formed by the 'absolutist' rich). Gibbons and Rutten (2004) argue that like the state, institutions of civil society (families and firms, clubs and communities, networks and neighborhoods) can in principle enforce rights by providing members with incentives both to respect rights and punish those who do not. Yet which does a better job supplying a particular public good, the state or the community? Furthermore, for societies facing more than one collectiveaction problem, how does the use of one institution to solve one problem affect the performance of various institutions that could be used to solve other problems (e.g. does having the government to enforce contracts affect the ability of the community to provide public goods) – when and how are institutions complements, substitutes or independent? Gibbons and Rutten stress the equilibrium approach to institutions: when there is no disinterested third party to enforce society's rules, institutions must be self-enforcing. They cast their model in variations of an N-person infinitely repeated prisoners' dilemma. In one of the variations, each period, first, the state chooses an amount to expend on a public good (at this stage, funding is taken from state's treasury, assumed to be arbitrarily large). Second, each citizen receives two benefits from the public good, one denominated in utility, the other denominated in money (only money can be expropriated by the state). Third, for each citizen, the state chooses how large a tax to collect; each citizen observes her own tax, but not any other citizen's tax. Fourth, each citizen decides whether to sound alarm. If at least one citizen sounds alarm, all citizens hear it. Fifth the citizens decide (as a unified body) whether to retract state's power, in which the individual who had been the state returns to being a citizen and anarchy ensues for all future periods.
124
Sixth, each citizen decides how much of her remaining resources to consume. In their model, the distinction between the state and the civil society is tantamount to citizens themselves contributing toward the public good rather than the state expanding resources on it in stage one of the game. Both Azam (2003) and Gibbons and Rutten (2004) are undoubtedly valuable contributions towards formulating economic theory of civil society. Both, however, leave room for improvement and unanswered questions. In Azam's model, as explained above, civil society is identified by the rich making voluntary transfers to the poor. The transfers are assumed to be implemented through an organized assembly of NGOs. Although an appealing concept, in his model, the general assembly of NGOs is never modeled. Moreover, Azam does not explain why the rich would even want to form such organizations. (Why would the rich not instead elaborate on the existing taxation system?) While Gibbons and Rutten (2004) provide a stronger underpinning of formation of civil society than Azam (2003), much like the latter, they do not provide any guidance on how and when, if at all, to promote civil society. Given the appeal of CSOs, an important normative question in development practice is when should the responsibility for providing services (e.g. social welfare and development) be delegated to CSOs? Besley and Ghatak (2001) apply the incomplete contracts paradigm to consider ownership issues (state vs. nongovernmental organization) for public goods. They argue that the party with the highest valuation of the project should be the owner regardless of the relative importance of the investments or other aspects of the production technology. The welfare implications of their model, however, are not clear-cut, and are dependant on the political economy of the setting that they do
125
not model. Jack (2001) explicitly casts his analysis in a development context, where a poorly functioning state is often a reality. If direct government provision of services is ineffectual or very inefficient, when is contracting out to a CSO preferred to using a traditional for-profit firm? Building on Glaeser and Shleifer (1998), Jack argues that CSO provision is preferred if the gross benefits of higher quality outweigh the increased financing costs. A non-profit entrepreneur is constrained in using her revenues. She must hence be paid more for service provision to induce the same levels of effort as a forprofit entrepreneur. Furthermore, if CSOs are to serve as important vehicles for delivery of foreign assistance to poor countries, who should implement a project: an international CSO or a grass-root CSO? Building on Besley and Ghatak (2001), Jack concludes that when donor and government valuations of the service do not differ much, provision through an international CSO yields higher joint surplus than under grass roots provision. One limitation of Jack’s analysis is that he models the CSOs as passive transformers of inputs into output. The literature on service provision role of CSOs is also related to the literature on public-private partnerships. See e.g. Bentz et al. (2004) and references therein.
126
Appendix 7: Welfare Consequences of Delaying Reform Implementation When the Lobbying Group's Favored Reform Program Increases Social Welfare
This appendix analyzes the scenario when the preferred reform program of the interest group aligned with the ruling political option – although not first best for society – increases social welfare. Intuitively, in this case, one would expect that existence of civil society to constrain the elites' behavior is less warranted. Suppose A2 as stated in chapter 1 does not hold. That is, suppose ?B?2+B?2?<B?3+B?3, or equivalently, B?2+B?2+B?3+B?3>0. No other assumptions of the model have been changed. Observe, first, that the politician in power, A, now accepts the reform program consisting of reforms 2 and 3 for any value of ??(0,1): ?(B?2+B?2+B?3+B?3)+(1??)2??A>0 for all ??(0,1). Without civil society's involvement, A implements ?'s proposed reform program readily and immediately. Hence, W? = B?2+B?2+B?3+B?3>0 and V? = ?(B?2+B?2+B?3+B?3)+(1??)2??A. (Recall that W denotes social welfare and V politician's payoff. Superscripts ? and + indicate nonexistence and existence of civil society, respectively.) When civil society exists, its participation opens the reform debate to public and delays implementation of reform. Adopting the analysis of chapter 1, when p<??A/B?3, ? does not submit anything. V+ = 0 < V?, W+ = 0 < W?. When p???A/B?2, ? submits a reform program consisting of reforms 2 and 3 and A accepts it: V+ = p?(B?2+B?2+B?3+B?3)+(1??)2??A > 0, but clearly V+ < V?, and likewise, W+ = p(B?2+B?2+B?3+B?3) < W?. Finally, when ??A/B?3?p<??A/B?2, ?'s submitted reform program consists of reform 3 only. V+ = p?(B?3+B?3)+(1??)??A>0, so A accepts it. Hence, V+>V? if and only if
127
p?(B?3+B?3)+(1??)??A>?(B?2+B?2+B?3+B?3)+(1??)2??A, or equivalently, if and only if p>p*(?), where p*(?) is defined in Proposition 3 in chapter 1. Likewise, W+ = p(B?3+B?3), and W+ > W? if and only if p(B?3+B?3)>B?2+B?2+B?3+B?3, or equivalently, if and only if p>pC, where pC?1??B?2+B?2?/(B?3+B?3)?(0,1). Note that in this case we cannot compare social welfare with and without civil society's involvement in reform process without imposing additional assumptions on the size of pC. (If pC<??A/B?3, then W+ > W?. If pC?[??A/B?3,??A/B?2), then W+ > W? for p>pC and W+ < W? for p<pC. If pC???A/B?2, W+ < W?.) When society's net gains and losses from reforms 3 and 2 are roughly equal, for example, ?B?2+B?2? ? B?3+B?3, pC is close to 0, and civil society's involvement is desirable from society's standpoint.
128
References
Abel, Richard L. "Revisioning Lawyers." In Richard L. Abel and Philip S. C. Lewis ed., Lawyers in Society: An Overview, Berkley and Los Angeles: University of California Press, 1995. Acemoglu, Daron and James A. Robinson. 2000. "Why Did The West Extend The Franchise? Democracy, Inequality, And Growth In Historical Perspective." Quarterly Journal of Economics. 115:4, pp. 1167?99. Acemoglu, Daron, Simon Johnson and James A. Robinson. 2001. "The Colonial Origins of Comparative Development: An Empirical Investigation." American Economic Review. December, 91:5, pp. 1369?1401. Åmark, Klas. "Open Cartels and Social Closures: Professional Strategies in Sweden, 1860?1950." In Michael Burrage and Rolf Torstendahl (eds.) Professions in Theory and History: Rethinking the Study of the Professions, London, SAGE Publications, 1990. Ashby, Simon, Swee-Hoon Chuah, and Robert Hoffmann. 2004. "Industry SelfRegulation: A Game-Theoretic Typology of Strategic Voluntary Compliance." International Journal of the Economics of Business. February, 11:1, pp. 91-106(16). Australian Law Reform Commission, 2003. http://www.alrc.gov.au/about/index.htm Ayres, Ian and John Braithwaite. Responsive Regulation: Transcending the Deregulation Debate. Oxford University Press, 1995. Azam, Jean-Paul 2003 "A Theory of Poverty Aversion and Civil Society Development", Economics and Politics, 15:1, pp. 61-84. Baldwin, Robert and Martin Cave. Understanding Regulation: Theory, Strategy, and Practice. Oxford University Press, 1999. Barro, Robert J. and Rachel M. McLeary. 2003. "Religion and Economic Growth across Countries." American Sociological Review. Vol. 68, pp. 760-81. Beck, Thorsten and L. Laeven. 2005. “Institutions and Growth in Transition Economics.” Working paper. Beck, Thorsten, Asli Demirguc-Kunt, and Ross Levine, “Law, Endowments, and Finance,” Journal of Financial Economics, 70:137-181, 2003a. Beck, Thorsten, Asli Demirguc-Kunt, and Ross Levine. 2003b. "Law and Finance: Why Does Legal Origin Matter?" Journal of Comparative Economics. 31:4, pp. 653?75. Becker, Gary, “A Theory of Competition Among Pressure Groups for Political Influence,” Quarterly Journal of Economics. 98, 3:371-400, Aug. 1983.
129
Bell, David A. Lawyers and Citizens. The Making of a Political Elite in Old Regime France. New York: Oxford University Press, 1994. Bell, David A. "Barristers, Politics, and the Failure of Civil Society in Old Regime France." In Terence C. Halliday and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism: Europe and North America from the Eighteenth to Twentieth Century, Oxford: Clarendon Press, 1997. Bentz Andreas, Paul A. Grout and Maija L. Halonen. 2004. "What Should Governments Buy from the Private Sector – Assets or Services?" Working paper. Berkowitz, Daniel, Katharina Pistor, and Jean-Francois Richard, “Economic Development, Legality, and the Transplant Effect,” European Economic Review. 47, 1:165-195, Feb. 2003. Besley, Timothy and Maitreesh Ghatak. 2001. "Government and Public Provision of Public Goods". Quarterly Journal of Economics,, November. pp. 1343-72. Brand, Paul. The Origins of the English Legal Profession. Oxford: Blackwell Publishers, 1992. Burrage, Michael and Rolf Torstendahl (eds.) Professions in Theory and History. Rethinking the Study of the Professions, London: SAGE Publications, 1990. Burrage, Michael and Rolf Torstendahl (eds.) The Formation of Professions. Knowledge, State and Strategy, London: SAGE Publications, 1990. Burrage, Michael, Konrad Jarausch and Hannes Siegrist. "An Actor-Based Framework for the Study of the Professions." In Michael Burrage and Rolf Torstendahl (eds.) Professions in Theory and History: Rethinking the Study of the Professions, London: SAGE Publications, 1990. Burrage, Michael. "Mrs Thatcher Against the ‘Little Republics’: Ideology, Precedents, and Reactions." In Terence C. Halliday and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism: Europe and North America from the Eighteenth to Twentieth Century, Oxford: Clarendon Press, 1997. Burrage, Michael. "Revolution as a Starting Point for the Comparative Analysis of the French, American, and English Legal Profession." In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society, Volume Three, Comparative Theories, Berkley and Los Angeles: University of California Press, 1989. Cadiet, Loïc, "Civil Justice Reform: Access, Cost, and Delay. The French Perspective" In Adrian Zuckerman ed. Civil Justice in Crisis: Comparative Perspectives of Civil Procedure. Oxford, Oxford, 1999.
130
Campos, Nauro F. 1999. "Context is Everything: Measuring Institutional Change in Transition Economies." World Bank Working Paper No. 2269. Carothers, Thomas and Marina Ottaway. "The Burgeoning World of Civil Society Aid." In Marina Ottaway and Thomas Carothers (eds.). Funding Virtue. Civil Society Aid and Democracy Promotion. Washington, D.C.: Carnegie Endowment for International Peace, 2000. Carothers, Thomas. "Western Civil Society Aid to Eastern Europe and the Former Soviet Union." In Thomas Carothers. Critical Mission. Essays on Democracy Promotion, Washington, D.C.: Carnegie Endowment for International Peace, 2004. Carothers, Thomas. 1999. "Western Civil-Society Aid to Eastern Europe and the Former Soviet Union." East European Constitutional Review. Fall, 8:4. http://www.law.nyu.edu/eecr/vol8num4/special/western.html. Carothers, Thomas. Critical Mission. Essays on Democracy Promotion. Washington, D.C.: Carnegie Endowment for International Peace, 2004. Carruthers, Bruce, and Terence Halliday. "Professionals in Systemic Reform of Bankruptcy Law: The 1978 U.S. Bankruptcy Code and the English Insolvency Act 1986" American Bankruptcy Law Journal, Winter, 2000, 74, pp. 35-75. Cleaves, Peter S. Professions and the State: The Mexican Case. Tucson: The University of Arizona Press, 1987. Che, Jiahua and Yingyi Qian. 1998. "Insecure Property Rights and Government Ownership of Firms." Quarterly Journal of Economics. May, 113:2, pp. 467?96. Cohen, Jean L. and Andrew Arato. Civil Society and Political Theory. MIT Press, 1994. Collins, Randall. "Changing Conceptions in the Sociology of Professions." In Rolf Torstendahl and Michael Burrage (eds.) The Formation of Professions: Knowledge, State and Strategy, London: SAGE Publications, 1990. Datta, Samar K. and Jeffrey B. Nugent. 1986. "Adversary Activities and Per Capita Income Growth." World Development, 14. David, René and Henry de Vries, 1958. The French Legal System: An Introduction to Civil Law Systems. Oceana Publications, NY. De Grand, Alexander J. Fascist Italy and Nazi Germany. The 'fascist' style of rule. Routledge, London and New York, 1995. DeMarzo, Peter M., Michael J. Fishman, and Kathleen M. Hagerty. 2001. "Contracting and Enforcement with a Self-Regulatory Organization." http://papers.ssrn.com/sol3/papers.cfm?abstract_id=297302
131
Dethier, Jean Jacques, Hafez Ghanem and Edda Zoli. 1999. “Does Democracy Facilitate the Economic Transition? An Empirical Study of Central and Eastern Europe and the Former Soviet Union.” World Bank Working Paper No. 2194. Dewatripont, M. and E. Maskin. 1995. "Credit and Efficiency in Centralized and Decentralized Economies." Review of Economic Studies. October, 62:4, pp. 541?55. Dewatripont, Mathias and Gerard Roland. 1995. "The Design of Reform Packages under Uncertainty." American Economic Review, December, 85:5, pp. 1207?23. Djankov, S., E. Glaeser, R. La Porta, F. Lopez-de-Silanes and A. Shleifer. 2003. "The New Comparative Economics." Journal of Comparative Economics. December, 31:4, pp. 595?619. Donabedian, Bairj. 1995. "Self-Regulation and the Enforcement of Professional Codes." Public Choice. 85:1?2, pp. 107-18. Drazen, Allan. Political Economy in Macroeconomics. Princeton: Princeton University Press, 2000. Ehrenberg, John. Civil Society. The Critical History of an Idea. New York University Press, 1999. Eisner, Marc Allen. Regulatory Politics in Transition, Second Edition, Johns Hopkins University Press, Baltimore, 2000. Estache, Antonio and David Martimort. 1999. "Politics, Transacstion Costs, and the Design of Regulatory Institutions." World Bank working paper 2073. www.worldbank.org/html/dec/Publications/ Workpapers/wps2000series/wps2073. Field, G. Lowell. The Syndical and Corporative Institutions of Italian Fascism. Columbia University Studies in the Social Sciences 433, AMS Press, New York, 1968. Florinsky, Michael T. Fascism and National Socialism. A Study of the Economic and Social Policies of the Totalitarian State. The Macmillan Company, New York, 1938. Fukuyama, Francis. Trust: The Social Virtues and the Creation of Prosperity. Free Press, 1996. Gandhi, J.S. "Past and Present. A Sociological Portrait of the Indian Legal Profession." In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society, Volume One, The Common Law World, Berkley and Los Angeles: University of California Press, 1988. Gehrig, Thomas and Peter-J Jost. 1995. "Quacks, Lemons, and Self Regulation: A Welfare Analysis." Journal of Regulatory Economics. 7:3, pp. 309?25.
132
Gibbons, Robert and Andrew Rutten. 2004. "Institutional Interactions: An Equilibrium Approach to the State and Civil Society". Working paper. Glaeser, Edward L. and Andrei Shleifer. 1998. “Not-for-profit entrepreneurs.” NBER Working Paper No. 6810. Glaeser, Edward L. and Andrei Shleifer. 2002. "Legal Origins." Quarterly Journal of Economics. 117:4, pp. 1193?229. Glaeser, Edward L. and Andrei Shleifer. 2003. "The Rise of the Regulatory State." Journal of Economic Literature. June, 41:2, pp. 401?25. Glaeser, Edward L., David Laibson and Bruce Sacerdote. 2000. "The Economic Approach to Social Capital." NBER Working Paper 7728. Glaeser, Edward, Simon Johnson, and Andrei Shleifer. 2001. "Coase vs. Coasians." Quarterly Journal of Economics. August, 116:3, pp. 853?99. Goldberg, Victor P. 1976. "Regulation and Administered Contracts." Bell Journal of Economics. Autumn, 7:2, pp. 426?48. Gostynski, Zbigniew and Alan Garfield. "Taking the Other Road: Polish Legal Education During the Past Thirty Years." Temple International and Comparative Law Journal. Fall, 1993. Greif, Avner. 1993. "Contract Enforceability and Economic Institutions in Early Trade: The Maghribi Traders' Coalition." American Economic Review. June, 83:3, pp. 525?48. Greif, Avner. 1994. "Cultural Beliefs and the Organization of Society: A Historical and Theoretical Reflection on Collectivist and Individualist Societies." Journal of Political Economy, Vol. 102, No. 5, pp. 912-50. Grossman, Gene M. and Elhanan Helpman. 1994. "Protection for Sale." American Economic Review. September, 84:4, pp. 833?50. Grossman, Gene M. and Elhanan Helpman. Special Interest Politics. Cambridge: MIT Press, 2001. Grossman, Sanford and Oliver Hart. 1986. "The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration." Journal of Political Economy. 94:4, pp. 691?719. Guiso, Luigi, Paola Sapienza and Luigi Zingales. 2004. "The Role of Social Capital in Financial Development." American Economic Review. June, 94:3, pp. 526-56. Hall, John A. Civil Society: Theory, History, Comparison. Polity Press, 1995.
133
Halliday, Terence C. and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism. Europe and North America from the Eighteenth to Twentieth Century. Oxford: Clarendon Press, 1997. Halliday, Terence C. and Lucien Karpik. "Postscript: Lawyers, Political Liberalism, and Globalization." In Terence C. Halliday and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism: Europe and North America from the Eighteenth to Twentieth Century, Oxford: Clarendon Press, 1997. Hasebe, Yukiko, "Civil Justice Reform: Access, Cost, and Expedition. The Japanese Perspective" In Adrian Zuckerman ed. Civil Justice in Crisis: Comparative Perspectives of Civil Procedure. Oxford, Oxford, 1999. Hayek, F. A. The Constitution of Liberty. University of Chicago Press, 1960. Hoff, Karla and Joseph E. Stiglitz. 2004. “After Big Bang? Obstacles to the Emergence of the Rule of Law in Post-Communist Societies.” American EconomicReview. June, 94:3, pp. 753-63. Hofstadter, Richard. The Age of Reform: From Bryan to F.D.R. Vintage Books USA, 1956. Huang, Fali. 2003. "Social Trust, Cooperation, and Human Capital." Working paper. Huskey, Eugene. "The Limits to Institutional Autonomy in the Soviet Union: The Case of the Advokatura" Soviet Studies, Vol. 34, No. 2. Apr., 1982, pp. 200-227. Huyse, Luc. "Legal Experts in Belgium." In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society: An Overview, Berkley and Los Angeles: University of California Press, 1995. International Center for Alcohol Policies. 2001. ICAP Reports 9, January 2001. "SelfRegulation of Beverage Alcohol Advertising." http://www.icap.org/publications/report9.html accessed 9/14/2004. Jack, William. 2001. "Public Policy toward Non-Governmental Organizations in Developing Countries", World Bank Working Paper No. 2639. Jarausch, Konrad H. The Unfree Professions. German Lawyers, Teachers, and Engineers, 1900?1950. New York, Oxford: Oxford University Press, 1990. Johnson, Terence. "Imperialism and the Professions: Notes on the Development of Professional Occupations in Britain’s Colonies and the New States." In Paul Halmos (ed.) The Sociological Review Monograph 20, Professionalization and Social Change, Keele University, December 1973.
134
Jordan, Pamela. 1998. "The Russian Advokatura (Bar) and the State in the 1990s." Europe-Asia Studies. July, 50:5, pp, 765?91. Karatnycky, A., A. Motly and B. Shor. Nations in Transit 1997: Civil Society, Democracy and Markets in East Central Europe and the Newly Independent States, New Brunswick: Transaction Publishers, 1998. Karpik, Lucien. 1988. "Lawyers and Politics in France, 1814?1950: The State, the Market and the Public." Law & Social Inquiry. Fall, 13:4, pp. 707?36. Karpik, Lucien. French Lawyers. A Study in Collective Action 1274 to 1994. Oxford: Clarendon Press, 1999. Knack, Stephen and Phillip Keefer. 2000. "Does Social Capital Have an Economic Payoff: A Cross-Country Investigation." Quarterly Journal of Economics. November, 112:4, pp. 1251-88. Kocka, Jürgen. "‘Bürgertum’ and Professions in the 19th Century: Two Alternative Approaches." In Michael Burrage and Rolf Torstendahl (eds.) Professions in Theory and History: Rethinking the Study of the Professions, London: SAGE Publications, 1990. Kranton, Rachel E. 2003. "Competition and the Incentive to Produce High Quality." Economica. 70:279, pp. 385?404. Kranton, Rachel E. and Anand V. Swamy. 1999. "The Hazards of Piecemeal Reform: British Civil Courts and the Credit Market in Colonial India." Journal of Development Economics. 58:1, pp. 1?24. Krause, Elliott A. "Professions and the State in the Soviet Union and Eastern Europe: Theoretical Issues." In Anthony Jones (ed.) Professions and the State: Expertise and Autonomy in the Soviet Union and Eastern Europe, Philadelphia: Temple University Press, 1991. La Porta, R., F Lopez-de-Silanes, Pop-Eleches, C. and A. Shleifer “Judicial Checks and Balances” Journal of Political Economy, forthcoming 2004. La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. Vishny. 1999. "The Quality of Government." Journal of Law, Economics and Organization. 15:1, pp. 222?79. La Porta, R., F. Lopez-de-Silanes, A. Shleifer and R. Vishny. 1998. "Law and Finance." Journal of Political Economy. December, 106:6, pp. 1113?55. La Porta, R., F. Lopez-de-Silanes, A. Shleifer and R. Vishny. 1997. "Legal Determinants of External Finance." Journal of Finance. 52:3, pp. 1131?50.
135
Laffont, Jean-Jacques. Incentives and Political Economy. Clarendon Lectures in Economics, Oxford University Press, 2002. Landon, Michael. The Triumph of the lawyers: Their Role in English Politics, 1678?1689. University of Alabama Press, 1969. Leland, Hayne E. 1979. "Quacks, Lemons, and Licensing: A Theory of Minimum Quality Standards." Journal of Political Economy, December, 87:6, pp. 1328?46. Lülfesmann, Christoph. 2002. "Central Governance or Subsidiarity: A Property-Rights Approach to Federalism." European Economic Review. September, 46:8, pp. 1379?97. Lyon, Thomas and Haizhou Huang. 2002. "Legal Remedies for Breach of the Regulatory 'Contract'." Journal of Regulatory Economics. 22:2, pp. 107?32. Lyttelton, Adrian. "Italian Fascism." In Walter Laqueur (ed.) Fascism: A Reader’s Guide. University of California Press, 1976. Maier, Charles S. In Search of Stability: Explorations in Historical Political Economy. Cambridge Studies in Modern Political Economies, Cambridge University Press, 1987. Manitoba Law Reform Commission, 2003. http://www.gov.mb.ca/justice/mlrc Maxwell, John, Thomas Lyon, and Steven Hackett. 2000. "Self-Regulation and Social Welfare: The Political Economy of Corporate Environmentalism." Journal of Law and Economics. 43:2, pp. 583-617. McCain, Morris. "A Soviet Lawyers in the Reform Debate: Cohesion and Efficacy." Soviet Studies, Vol. 34, No. 1., Jan., 1982, pp. 3-22. McGerr, Michael. A Fierce Discontent : The Rise and Fall of the Progressive Movement in America, 1870-1920. Free Press, 2003. Milward, Alan S. "Fascism and the Economy." In Walter Laqueur (ed.) Fascism: A Reader’s Guide. University of California Press, 1976. Mueller, Dennis. Public Choice II. Cambridge University Press, 1989 Mulligan, Casey B. and Andrei Shleifer. "Population and Regulation." NBER Working Paper 10234, 2004. Murphy, Kevin M., Andrei Shleifer and Robert W. Vishny. 1991. "The Allocation of Talent: Implications for Growth." Quarterly Journal of Economics. 106 .
136
North, Douglass C. and Barry Weingast, "Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth-Century England." Journal of Economic History, December, 1989. North, Douglass C. 1991. "Institutions." Journal of Economic Perspectives. Vol. 5, No. 1., pp. 97-112. North, Douglass C., Institutions, Institutional Change, and Economic Performance, Cambridge University Press, 1990. Nunez, Javier. 2001. "A Model of Self-Regulation." Economics Letters. 74:1, pp. 91?7. Odenyo, Amos O. 1979. "Professionalization Amidst Change: The Case of Emerging Legal Profession in Kenya." African Studies Review. December, 22:3, pp. 33?44. Ogus, Anthony. "Self-Regulation." In Boudewijn Bouckaert and Gerrit De Geest (eds) Encyclopedia of Law and Economics, 1999. http://encyclo.findlaw.com. Olson, Mancur. Logic of Collective Action: Public Goods and the Theory of Groups. Revised edition. Harvard Economic Studies, Harvard University Press, 1971. Olson, Mancur. The Rise and Decline of Nations. Yale University Press, 1982. Orenstein, Mitchell A. 2000. "How Politics and Institutions Affect Pension Reform in Three Postcommunist Countries." World Bank Working Paper No. 2310. Ottaway, Marina and Thomas Carothers (eds.). Funding Virtue. Civil Society Aid and Democracy Promotion. Washington, D.C.: Carnegie Endowment for International Peace, 2000. Pei, Minxin. "Does Legal Reform Protect Economic Transactions? Commercial Disputes in China." In Peter Murrell (ed.) Assessing the Value of Law in Transition Economies, Ann Arbor: University of Michigan Press, 2001. Persson, T. 1998. "Economic Policy and Special Interest Politics." Economic Journal. March, 108, pp. 310?27. Persson, Torsten and Guido Tabellini. Political Economics: Explaining Economic Policy. MIT Press, Cambridge, Massachusetts, 2000. Plucknett, T. F. T. Studies in English Legal History. London: The Hambledon Press, 1983. Poole, Keith T. and Thomas Romer. 1985. "Patterns of Political Action Committee Contributions to the 1980 Campaigns for the U.S. House of Representatives." Public Choice. 47:1, pp. 63–112.
137
Poole, Keith T., Thomas Romer and Howard Rosenthal. 1987. "The Revealed Preferences of Political Action Committees." American Economic Review. May, 77:2, pp. 298–302. Posner, Richard A. Overcoming Law. Cambridge: Harvard University Press, 1995. Priest, Margot. 1997. "The Privatization of Regulation: Five Models of Self-Regulation." Ottawa Law Review. 29, pp. 236?314 Putnam, Robert. Making Democracy Work. Civic Traditions in Modern Italy. Princeton University Press, 1993. Qian, Yingyi. 1994. "A Theory of Shortage in Socialist Economies Based on the ‘Soft Budget Constraint’." American Economic Review. March, 84:1, pp. 145?56. Quigley, Kevin F.F. "Lofty Goals, Modest Results: Assisting Civil Society in Eastern Europe." In Marina Ottaway and Thomas Carothers (eds.). Funding Virtue. Civil Society Aid and Democracy Promotion. Washington, D.C.: Carnegie Endowment for International Peace, 2000. Rajan, Raghuram and Luigi Zingales. 2003. "The Great Reversals: The Politics of Financial Development in the 20th Century." Journal of Financial Economics. 69, pp. 5?50. Rich, William A. "Poland’s new collateral law." Law In Transition, Summer 1997. Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi "Institutions Rule: The Primacy Of Institutions Over Geography And Integration In Economic Development" NBER Working Paper, 2002. Rokumoto, Kahei. "The Present State of Japanese Practicing Attorneys: On the Way to Full Professionalization?" In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society: An Overview, Berkley and Los Angeles: University of California Press, 1995. Roland, Gerard. 2004. “Understanding Institutional Change: Fast-Moving and SlowMoving Institutions”. Studies in Comparative International Development. Roland, Gerard. Transition and Economics: Politics, Markets, and Firms. MIT Press, 2000. Rose, Jonathan. 1998. “The Legal Profession in Medieval England: A History of Regulation.” Syracuse Law Review. 48:1. Ross, Stanley D. 1992. "The Rule of Law and Lawyers in Kenya." The Journal of Modern African Studies. September, 30:3, pp. 421?42.
138
Rueschemeyer, Dietrich. "State, Capitalism, and the Organization of Legal Counsel: Examining an Extreme Case ? the Prussian Bar, 1700?1914." In Terence C. Halliday and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism: Europe and North America from the Eighteenth to Twentieth Century, Oxford: Clarendon Press, 1997. Salomon, Lester M., Wojciech Sokolowski and Associates. Global Civil Society, Dimensions of the Nonprofit Sector, Vol. 2. Kumarian Press, 2004. Shaked, Avner and John Sutton. 1981. "The Self-Regulating Profession." Review of Economic Studies. April, 48:2, pp. 217?34. Shapiro, Isaac "The Soviet Bar-Past and Present." Russian Review, Vol. 20, No. 2., Apr., 1961, pp. 143-150. Shelley, Louise I. "Lawyers in Soviet Union." In Anthony Jones (ed.) Professions and the State: Expertise and Autonomy in the Soviet Union and Eastern Europe, Philadelphia: Temple University Press, 1991. Shleifer, Andrei and Robert W. Vishny. 1994. "Politicians and Firms." Quarterly Journal of Economics. November, 109:4, pp. 995?1025. Shleifer, Andrei and Robert W. Vishny. The Grabbing Hand: Government Pathologies and Their Cures. Harvard University Press, 1998. Shonfield, Andrew. Modern Capitalism. The Changing Balance of Public and Private Power. Oxford University Press, 1974. Siegrist, Hannes. "Professionalization as a Process: Patterns, Progression and Discontinuity." In Michael Burrage and Rolf Torstendahl (eds.) Professions in Theory and History: Rethinking the Study of the Professions, London, SAGE Publications, 1990. Sternhell, Zeev."Fascist Ideology." In Walter Laqueur (ed.) Fascism: A Reader’s Guide. University of California Press, 1976. Stigler, George. "The Theory of Economic Regulation." Bell Journal. Spring 1971, 2, 137-146. Stratmann, Thomas. 1991. "What Do Campaign Contributions Buy? Deciphering Causal Effects of Money and Votes." Southern Economic Journal. January, 57:3, pp. 606–20. Stratmann, Thomas. 1995. "Campaign Contributions and Congressional Voting: Does the Timing of Contributions Matter?" Review of Economics and Statistics. February, 77:1, pp. 127–36.
139
Stratmann, Thomas. 2002. "Can Special Interests Buy Congressional Votes? Evidence from Financial Services Legislation." Journal of Law and Economics, forthcoming. Temple, Jonathan and Paul A. Johnson. 1998. "Social Capability and Economic Growth." Quarterly Journal of Economics. August: 113:3, pp. 965-90. Tirole, Jean. 1994. "The Internal Organization of Government." Oxford Economic Papers, January, 46:1, pp. 1?29. Tocqueville, Alexis De. The Old Regime and the French Revolution. Garden City, New York: Doubleday Anchor Books, 1955. Trevelyan, G.M. The English Revolution, 1688?1689. New York: Oxford University Press, 1967. Wallis, John Joseph. 1987. "Employment, Politics, and Economic Recovery During the Great Depression." Review of Economics and Statistics, August, 69, pp. 516?20. Wallis, John Joseph. 1998. "The Political Economy of New Deal Spending Revisited, Again: With and Without Nevada." Explorations in Economic History. April, 35, pp. 140?70. Waters, Christopher P.M. Counsel in the Caucasus: Professionalization and Law in Georgia. The Hague/London/New York: Martinus Nijhoff Publishers, forthcoming, 2004. Weisbrot, David. "The Australian Legal Profession. From Provincial Family Firms to Multinationals." In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society, Volume One, The Common Law World, Berkley and Los Angeles: University of California Press, California, 1988. Williamson, Oliver E. 1976. "Franchise Bidding for Natural Monopolies-in General and with Respect to CATV." Bell Journal of Economics. Spring, 7:1, pp. 73?104. Working Together. World Bank-Civil Society Relations. World Bank 2003. Zimmerman, Richard. 1996. "An Introduction to German Legal Culture." In Werner F. Ebke and Matthew W. Finkin, eds., Introduction to German Law. Kluwer, The Hague. Zweigert, Konrad and Hein Kötz. Introduction to Comparative Law. Second revised edition. Oxford University Press, 1992.
140
doc_840689689.pdf
A focus of the subject is how economic agents behave or interact and how economies work.
ABSTRACT
Title:
ESSAYS IN COMPARATIVE INSTITUTIONAL ECONOMICS Peter Grajzl, Doctor of Philosophy, 2005
Directed By:
Professor Peter Murrell Department of Economics
This dissertation examines how decentralized institutional structures and organizational forms evolve and affect economic development under different politicoeconomic and legal arrangements. Organized legal professions are typically viewed by economists as rent-seeking interest groups – even though they have been central in institutional development in countries with the highest quality institutions. Chapter 1 develops a model that identifies the link between the role of organized legal professions and the quality of reform. Delaying institutional reform through deliberation, the profession's participation discounts the expected benefit from welfare-inferior reform proposal for rent-seeking interest groups. Professional review serves as a screening mechanism ameliorating the self-interested government's adverse selection problem. The model's predictions cast new light on the Glorious and the French revolutions, post-communist transition, why and when civil law and common law systems differ, and why post-independence institutions are of higher quality in settler than in extractive colonies. Although common, self-regulation as an alternative to direct government regulation has been little investigated. Chapter 2 uses a framework inspired by property rights
theory to address the allocation of regulatory authority. In a model of a regulatory process with bargaining, the authority to amend the enabling legislation can be either consolidated within the government, or extended to the producers in a self-regulatory regime. The chapter delineates the welfare implications of regulatory regime choice, and indicates whether the government's incentives to delegate or centralize regulatory authority lead to efficient institutional design. The model identifies those features of legal traditions that help to explain variation in regulatory arrangements across countries, illuminates the contrast in regulatory practice between the progressive era and the associational regime of the New Deal, and characterizes the mechanisms of intervention used in fascist economies. Chapter 3 discusses the channels through which civil society is expected to affect economic development. Utilizing the formal analysis laid out in Chapter 1, the chapter provides an introductory examination of the rationale for civil society aid and concludes with a conjectural interpretation of the determinants of the aid's effectiveness to bring about successful institutional change in post-communist countries.
ESSAYS IN COMPARATIVE INSTITUTIONAL ECONOMICS
By Peter Grajzl
Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2005
Advisory Committee: Professor Peter Murrell, Chair Professor John Wallis Professor Rachel Kranton Professor Martin Loeb Assistant Professor Rodrigo Soares
© Copyright by Peter Grajzl 2005
Preface
Institutions consist of formal rules and informal constraints structuring political, economic and social interaction (North 1990, 1991). Traditions, customs, moral values, religious beliefs and other informal social norms affecting economic development (e.g. Greif 1994, Barro and McCleary 2003) originate within societies themselves. In contrast, formal rules such as constitutions, statutory laws and government regulations are designed and enforced in a centralized manner by the state. An important part of the institutional infrastructure of societies is determined by a state-led top-down approach. However, elements of institutional structure other than social norms and culture nevertheless evolve in a decentralized manner, enabling a society to, at least partially, 'govern itself'. This dissertation examines how examples of decentralized institutional structures and organizational forms evolve and contribute to economic development under different politico-economic and legal arrangements. Chapter 1 examines the role of the organized legal professions' review of proposed legislation. Chapter 2 explores self-regulation as a general phenomenon of deliberate delegation of the state's law-making powers. Chapter 3 provides a preliminary investigation of civil society's contribution to institution building. Chapters 1 through 3 analyze different instances of institutional and organizational decentralization. The individual chapters' common methodological feature, however, is the focus on identification of the socially efficient institutional structure within the predetermined set of feasible institutional arrangements in a given politico-economic and legal setting, and prediction when the identified socially efficient
ii
institution is incentive-compatible with the non-benevolent government's agenda. That is, while organized legal professions' review of proposed legislation, cross-industry selfregulation, and civil society involvement may one by one be socially efficient, their implementation need not be desired by the ruling politicians.1 The ensuing paragraphs provide a brief motivation for each of the chapters as they stand as separate parts integrated under the common theme of institutional and organizational decentralization, in turn emphasizing each individual chapter's main contributions to the literature on comparative analysis of institutions and elucidating the key features of each chapter's analytical framework. Like much of a growing body of literature on comparative analysis of institutional arrangements that has raised, or freshly illuminated, many important questions in the last decade (see Djankov et al. 2003 for discussion), the arguments proposed in this dissertation rely on reduced form models, motivated by disparate historical examples and available data. Economists typically view organized legal professions as rent-seeking interest groups – even though they have been central in institutional development in countries with the highest quality institutions. The model developed in Chapter 1 identifies the link between the role of organized legal professions and the quality of policy reform. By delaying institutional reform through its deliberation, the profession's participation discounts the expected benefit from welfare-inferior reform proposals for rent-seeking interest groups. Professional review serves as a screening mechanism ameliorating the self-interested government's adverse selection problem. The model predicts how the role
1
As a consequence, an important question – the answer to which is beyond the scope of this dissertation – that may arise under each individual case of decentralization under investigation is how to align the government's incentives with the society's if the government's agenda subverts social efficiency.
iii
of organized legal professions varies with democracy, political stability and professional power. The analysis in Chapter 1 casts new light on, and identifies common elements of, such apparently widely different events as post-communist transition and the Glorious Revolution in Britain and the French revolution. The chapter also extends the recent discussion on the importance of legal and colonial origins for the quality of contemporary institutions (see e.g. LaPorta et al. 1997, Rajan and Zingales 2003, Acemoglu et al. 2001). The analysis suggests why and when civil and common law differ, and why postindependence institutions are of higher quality in settler than in extractive colonies. The political power of organized legal professions examined in Chapter 1 is manifested through their autonomy from the state ? their ability to self-regulate. Although self-regulation is a common phenomenon it has rarely been investigated as an alternative to direct government regulation. Chapter 2 uses a framework inspired by Grossman and Hart's (1986) property rights theory to address the allocation of regulatory authority. When regulatory contracts are incomplete, the distribution of regulatory rights critically shapes regulatory outcomes. The chapter develops a model of a regulatory process with bargaining, where the authority to amend the enabling legislation can be either consolidated within the government or extended to the producers in a selfregulatory regime. The model delineates the welfare implications of regulatory regime choice, and indicates whether the government’s incentives to delegate or centralize regulatory authority lead to efficient institutional design. Chapter 2 complements chapter 1 by suggesting another channel through which legal origin matters (see e.g. Beck at al. 2003): the analysis in Chapter 2 identifies those
iv
features of legal traditions that help to explain variation in regulatory arrangements across countries. The chapter also attempts to characterize the mechanisms of intervention used in fascist economies, and illuminate the contrast in regulatory practice between the progressive era and the associational regime of the New Deal, as put forward by Eisner (2000). Political independence and self-regulatory powers of organized legal professions investigated in Chapters 1 and 2, respectively, are considered "the backbone of an autonomous civil society" (Rueschemeyer 1997, p. 217) a notion in vogue in policy circles but lacking sound theoretical underpinnings. Chapter 3 first discusses the concept of civil society and the channels through which civil society is believed to affect development. It then focuses on the role of civil society as a bulwark against the state and discusses the experience with civil society aid in post-communist countries. Utilizing the formal analysis laid out in Chapter 1, the chapter provides an introductory examination of the rationale for civil society aid and concludes with a conjectural interpretation of the determinants of the aid's effectiveness to bring about successful institutional change.
v
Dedication
Atiju in mami: za vso pomo?, vzpodbudo in ljubezen.
vi
Acknowledgements
I am grateful to Peter Murrell for his inspiration, guidance, support and advice. I also thank Larry Ausubel, Chas Cadwell, Thomas Carothers, Valentina Dimitrova-Grajzl, Kathryn Hendley, Simon Johnson, Pamela Jordan, Philip Keefer, Rachel Kranton, Andrei Shleifer, Thomas Stratmann, John Wallis, Christopher Waters, and participants at seminars at the Federal Trade Commission, George Mason University, University of Maryland, Inter-University Graduate Student Conference at Yale University, Central European University, and CERGE-EI for many helpful comments.
vii
Table of Contents
Preface Dedication Acknowledgements Table of Contents Chapter 1: Lawyers and Politicians: The Impact of Organized Legal Professions on Institutional Reforms 1.1 The Rule of Law, Organized Professions, and Development 1.2 Organized Legal Professions: Nature and History 1.2.1 The Nature of Organized Legal Professions 1.2.2 Characterizing Legal Professions 1.2.3 A Brief History of Organized Legal Professions 1.3 A Model of Professions in the Reform Process 1.3.1 The Structure of Reform Programs 1.3.2 Interest Groups, Politicians, Society 1.3.3 Politician and the Organized Profession 1.3.4 The Timeline 1.4 When Do Politicians Use Powerless Professions 1.4.1 Baseline Equilibria: When the Profession Does Not Exist 1.4.2 Equilibria when the Profession Exists and the Politician's Self Interest Matters 1.4.3 Equilibria when the Profession Exists and the Politician's Aims Coincide with Society's 1.4.4 Professions Under Autocracy and Transitions to Democracy 1.5 When Do Politicians Use Professions that Have Political Power 1.5.1 Comparative Legal Development in England and France: 1688 and 1789 1.5.2 The Professions Under Civil and Common Law 1.5.3 Colonies and Institutions 1.6 A Postscript on Further Research Chapter 2: Allocating Law-Making Powers: Self-Regulation vs. Government Regulation 2.1 Introduction 2.2 A Simple Model 2.2.1 The Setup 2.2.2 Solving the Model 2.3 Comparing Regulatory Arrangements: Efficiency and Government's Incentives 2.4 Illustrating the Model 2.4.1 Two Modes of Attaining Regulatory Goals: The Progressive Era and the New Deal 2.4.2 The Fascist Economic System: Self-Regulation in a Totalitarian Regime ii vi vii viii 1 1 6 6 9 10 13 13 16 18 20 21 22 23 25 27 32 34 38 41 43 45 45 49 49 56 59 63 63 67
viii
2.4.3 Legal Origin and Regulatory Regime Choice 2.5 Conclusion Chapter 3: Fostering Civil Society: Towards an Analytical Framework 3.1 A Concept in Vogue 3.2 Delineating Civil Society 3.3 Promoting Civil Society: The Democracy Aid Project 3.3.1 Civil Society and Development 3.3.2 Civil Society and Institutional Change 3.4 Evaluating the Aid's Success: The Role of Initial Political Conditions 3.5 Aiding Whom and When: Towards an Analytical Framework 3.5.1 Model Background 3.5.2 When Does Aid Make A Difference? The Role of Political Conditions Revisited 3.5.3 Judging the Success of Civil Society Aid: A Conjectural Interpretation 3.6 Summing Up Appendix 1: Figures Figure 1: The Set of Reforms Figure 2: The Timeline Figure 3: The Politician's Welfare and Aggregate Welfare with and without a Powerless Profession Figure 4: Professional Political Power, Democracy, and Stability Figure 5: Political Power and the Use of the Profession Figure 6: The Political Cost of Removing the Profession Figure 7: Timeline of the Regulatory Process Figure 8: SR vs. R: Efficiency vs. Government's Incentives Figure 9: Legal Origin and Regulatory Regime Choice Figure 10: Effectiveness of Aid to Civil Society Appendix 2: Tables Table 1: Summary Statistics by Regulatory Arrangements: Means (and Standard Deviations) of Variables Table 2: The Relationship between Legal Origin and Regulatory Regime: Ordered Probit Results Appendix 3: Proofs of Propositions in Chapter 1 Appendix 4: Model with Reform 1 Appendix 5: Algebraic Derivations and Proofs of Results in Chapter 2 Appendix 6: Civil Society as Vehicle for Service Provision Appendix 7: Welfare Consequences of Delaying Reform Implementation when the Lobbying Group's Favored Reform Program Increases Social Welfare References
70 74 76 76 77 78 79 82 84 86 88 90 92 94 96 96 97 98 99 100 101 102 102 103 104 105 105 106 107 110 120 123 127 129
ix
Chapter 1 Lawyers and Politicians: The Impact of Organized Legal Professions on Institutional Reforms
1.1 The Rule Of Law, Organized Professions, and Development "Above all, it is now abundantly clear that the history of the legal profession is an important part of the history of the law…." Plucknett (1983, p. 332) "The history of the legal profession is to a great extent, and despite noisy and incessant protestation and apologetics, the history of efforts by all branches of the profession, including the professoriat and the judiciary, to secure a lustrous place in the financial and social-status sun." Posner (1995, p. 33) One of the most significant recent developments in economics has been the rediscovery of the central role of the rule of law in economic development (e.g., North 1990; La Porta et al.1998; Rodrik et al. 2002). Despite this new affection for law, economists have not been so eager to embrace lawyers, and even less so the organized legal professions (e.g., Datta and Nugent 1986; Murphy et al. 1991; Posner 1995). Yet, the rule of law is a human construct, not a natural endowment. Lawyers are the architects and engineers of that construction process, and in every developed country a strong, organized legal profession is central to lawyering. This paper presents a theory that establishes a role for an organized legal profession in economic development. It shows how a legal profession can improve the quality of institutional reforms. The argument does not rely on some newly identified capability of organized professions. Rather, it rests on that characteristic of the involvement of strong professions in reform processes that has elicited much negative commentary–the delay caused by deliberation. The model predicts how the role of the organized legal profession varies with levels of democracy and political stability, establishing the direct relevance of these predictions
1
to history and to development policy. The analysis casts new light on historical episodes that have been the object of considerable debate: the old regime, 1789, and Napoleon in France, the Glorious Revolution in England, the USSR and post-socialist transition, the contrasting nature of the English and French legal systems, and the comparative development of former colonies (North and Weingast 1989; La Porta et al. 1998; Rajan and Zingales 2003; Acemoglu et al. 2001). The paper elucidates the character of civil society, a popular notion in development policy, which lacks a sound theoretical underpinning (Carothers 1999). Economists typically view organized professions as rent-seeking interest groups, placing them in a negative light (e.g., Stigler 1971; Mueller 1989; Drazen 2000).2 However, with the organized legal profession cast as just another rent-seeker, there is a paradox to be resolved, encapsulated in the contrast between the epigraphs to this paper. Those countries adhering most strongly to the rule of law and with the highest quality legal rules (La Porta et al. 1998) have notoriously powerful legal professions. Already in medieval times the English legal profession was a force to be reckoned with, and has rarely been challenged since (Burrage 1997). There, the legal profession was a resolute companion of both the rise of democracy and the institutional revolution accompanying modern development: the legal professions have been builders of the liberal state and society (Halliday and Karpik 1997, p. 16). In contrast, in many former colonies and in transition countries, where law has played a much less significant role, organized legal professions have been virtually irrelevant (Johnson 1973; Waters 2004).
2
Becker (1983) shows how political competition can reduce the negative consequences of the activities of rent-seeking groups.
2
These facts suggest that organized legal professions play a role beyond that of rentseeking. As Hayek (1960, pp. 59-60) argues, successful development has often been the result of trial and error selecting superior institutions, even when the reason for that superiority is little understood. The tenacity of the organized legal profession, its central role in developed societies, and its coevolution with the rule of law constitute prima facie evidence that it has been one of the Hayekian survivors. This simple observation is a major stimulus for our analysis, which uses standard theories to isolate the way in which organized professions contribute to institutional development. We focus on the process of designing and implementing institutional reforms. In that process, the government faces the competing demands of different interests. Rentseeking interest groups have more information about the effects of proposed reforms than the politicians forming the government. Since interest groups submit self-serving reforms for consideration, the politician faces an informational problem because some reforms are damaging to the politician and others are not. One mechanism that ameliorates this problem is the government's use of the organized profession in a review process that stimulates debate about the reform. The prestige, status, and independence of the organized profession adds political legitimacy and credibility to the decision to use the profession in the face of powerful interest groups pressing for quick reforms. The review opens up the reform process to public debate and delays the adoption of reforms, which will be costly for interest groups, since the government might lose power before implementation occurs. This cost changes the incentives of the interest groups when submitting reforms; they drop initiatives with lower payoffs. Institutional reforms improve.
3
The politician is interested in more than the quality of institutional reforms. Interest groups support politicians, contributing funds when submitting reform proposals. A politician weighs these contributions against the pursuit of the general welfare, the latter being relatively more important when democratic constraints are tighter. Hence, the politician is more likely to use the organized profession when democracy is stronger. Political stability also matters. When government turnover is more likely, the politician's expected value from using the profession to obtain better, but later, reforms declines. Democracy and stability become substitutes. The politician uses professional review only at high levels of democracy in an unstable environment but at lower levels of democracy when there is stability. This trade-off is particularly important in transitions from autocracy, which often involve the paradox of rising democracy and reduced influence of the organized legal profession. The key to the paradox is that instability often rises during these transitions. We use these insights to interpret events in the postsocialist transition and in eighteenth century France. Once a professional group has acquired political power, the politician also has to consider the political costs of challenging the group's status. The power of the profession can change the politician's decision, leading to improved institutional reforms. This effect occurs only in the middle ranges of democracy. When it does, professional power substitutes for weak democracy. Matching this prediction to historical events, we examine the role of the English legal profession during the Glorious Revolution, and show how instability and a weak legal profession contributed to different results in France a century later.
4
Our focus on the organized professions can be used to interpret recent empirical work on the historical origins of institutional quality, allowing us to identify complementarities between different contributions to the literature. We show that the politician uses the organized legal profession more often in common law than in civil law countries, given that legal professions are naturally more powerful in the former. This choice leads on average to superior institutions in common law counties, but also to slower implementation, reconciling the views of La Porta et al. (1998) and Rajan and Zingales (2003). Our analysis provides a new perspective on mechanisms underlying the results of Acemoglu et al. (2001) on the colonial origins of institutional development. Since organized professions were stronger in those colonies with high levels of European migration, our analysis predicts that institutional quality is lower in the successor nation states of those colonies that were less hospitable to European settlement. This squares with the focus of Berkowitz et al. (2003) on local knowledge during legal transplantation, since the organized profession is the prime source of that knowledge. Lastly, given the correlation between professional power and legal origin and given that judicial independence is bolstered by a strong profession, our model is pertinent to the work that draws a connection between judicial independence and legal origin (La Porta et al. 2004). The chapter's argument proceeds by modulating between the model and related facts from history and from studies of professions. Section 1.2 examines the defining characteristics of organized professions and presents a brief history of the legal profession. Section 1.3 builds upon these facts in presenting the model. Section 1.4 develops predictions using the assumption that the profession is politically powerless, identifying the circumstances under which the profession contributes to the improvement
5
of institutional reforms. These predictions are used to interpret events in Louis XIV's France, in the Soviet Union, and in the transition from autocracy. Section 1.5 introduces the assumption that the profession has political power, showing that professional power and democracy are substitutes. The predictions cast light on the differing developments in England in 1688 and France in 1789, and the consequences of these developments. Lastly, we show how the model bears on the legal and colonial origins of institutional development. A concluding section considers implications for future research. 1.2 Organized Legal Professions: Nature and History This section summarizes the facts that suggest assumptions for the model and that provide context for interpretation of historical episodes. We examine the nature of organized legal professions, the special status they play in reviewing reform proposals, and the reasons why organized legal professions accumulate unusual amounts of political power. A brief overview of the history of organized legal professions summarizes differences across countries. 1.2.1 The Nature of Organized Legal Professions The practice of some occupations entails using a specialized body of knowledge that has been created by past practitioners. Practice requires costly training, which practitioners are best qualified to impart. These characteristics make an occupation a profession, in its broadest sense. But we use a narrower concept of a profession, adding the criterion that permission to practice depends upon demonstrated expertise in the body of knowledge.3 Moreover, the strongest organized professions usually are self-regulating
See, for example, Torstendahl and Burrage (1990) and Burrage, Jarausch and Siegrist (1990). Åmark (1990) draws the distinction between lobbying groups and professional lobbies, by noting that the former are open to the general public, while the latter are not. We do not examine why there are formal requirements. These requirements are nearly universal in the areas of professional activity that we
3
6
monopolies, administering the use of this body of knowledge by implementing a single set of policies, rules, and standards.4 Formalization of self-regulation confers quasigovernmental status. With the organized profession deemed to have monopoly expertise, its members are accorded special status in advising the government on reforms, especially in areas involving the rules and procedures that practitioners use. Already in the 1300's, legal professionals advised the English king on statutes (Rose 1998) and dominated a commission on legal reforms (Plucknett 1983, p. xix 335). The special advisory role might be so automatic that it becomes formalized. The British Commonwealth legal systems have gone far in this process by setting up Law Reform Commissions, dominated by professional lawyers and charged with reviewing all aspects of legal reform.5 The advisory role does not necessarily involve officially appointing the monopoly professional association to some formal body. The crucial issue is that the politician facilitates a process in which the voice of the independent profession fosters open debate. Governments can satisfy professional demands by consulting with prominent sub-groups
examine. But there are exceptions, which provide examples used later in the paper: Waters (2004, p. 95) notes that lawyering in Georgia was 'free for all' during the early 1990s. "...there can be no doubt that the profession's knowledge about the intricacies of professional practice, the official and unofficial procedures, and the opportunities for manoeuvre and circumvention that these provide, is rarely rivaled. This is an immense source of power since it means that any attempt to change or control professional behavior by instituting new rules and procedures has to be negotiated with residential procedural experts, the practitioners who will actually implement the change." (Burrage, Jarausch and Siegrist 1990, p. 210). For example, in the case of the Manitoba Law Reform Commission (2003): "The Commission's duties are to inquire into and consider any matter relating to law in Manitoba with a view to making recommendations for the improvement, modernization and reform of law…" The commission is "…composed of five members…At least one shall be a judge of the Court of Queen's Bench; at least one shall be a full-time member of the Faculty of Law, University of Manitoba; at least one shall be a barrister and solicitor entitled to practise in the province and who is not in the full-time employ of the Government of Manitoba or any agency thereof; and at least one shall not be a lawyer. One of the members shall be appointed President and he or she must be a lawyer."
5 4
7
or by using independent, universally respected, members of the profession.6 The process can be formal, as when the politician summons a reform commission consisting of prominent members of the legal profession, or it can be informal, as when the organized legal profession naturally emerges as the leader of a public debate that the politician has allowed to proceed. Whatever the nature of the process, the status and prestige of the organized profession affords political legitimacy and credibility to the politician's decision to use it in the face of importuning interest groups. Among the consequences of professional review, there is one that is universally agreed upon: a slowing of the reform process. The twenty-six-year, lawyer-dominated, gestation period of the first German civil code (Zimmerman 1996, p. 6) stands in contrast to the six months in which the Weimar politicians drafted Germany's first democratic constitution. While Napoleon's constitution of 1799 was created and ratified in two months, it took four years for a commission of old practitioners to craft his civil code, largely reviving the legal framework created under Louis XIV and Louis XV (David and de Vries 1958, p. 13). Examples from more recent times are the eight years for new bankruptcy laws in the US and the UK and the six years for new civil procedure codes in France and Japan, each process beginning with lengthy deliberations by commissions of professionals appointed by politicians (Carruthers and Halliday 2000; Cadiet 1999, p. 315; Hasebe 1999, p. 237). At the very least then, one may assume that professional review changes the pace of the reform process.
6
The role of the American Law Institute in the US provides an example.
8
1.2.2 Characterizing Legal Professions The foregoing assumes that a profession exists in the sense that there is an occupational group that is capable of self-organization whose members have sufficient expertise to be used by the politician in an advisory role. This assumption might not hold, as in China after the cultural revolution (Pei 2001, pp. 181-2). Thus, a first element of the characterization of a profession is whether it exists. Once it exists, the organized profession's status, particularly its relationship with government, can take on many different hues. In some cases the state creates a profession and gives it formal status; in others custom and political struggle are more relevant. Thus, the profession might be coterminous with a government department, as in early nineteenth century Prussia (Rueschemeyer 1997; Kocka 1990; Siegrist 1990), or it might be a ‘little commonwealth’ or ‘lesser government’ as was the English bar already by the middle ages, even though its autonomy was never formalized.7 Separation between the government and the profession involves several features, for example, the degree of self-regulatory authority or the strength of the opposition when the government tries to reduce the role of the profession. These are inter-related. Selfregulation is a political prize won by an independent profession, but self-regulation also protects independence, keeping a disciplining technology out of government control. Self-regulation enhances the profession's monopoly of specialized knowledge, which increases the need to involve it during reforms. Indeed, in the descriptive literature, the notions of independence, autonomy, and self-regulation are often used as synonyms.
7
Halliday and Karpik (1997, p. 354): "[In England] Barristers’ independence has its medieval origins in civil society, not the state. Indeed, barristers’ autonomy, and those of other ‘little commonwealths’ in civil society, are entrenched in the unwritten English constitution." Burrage (1988, p. 356): "Charles [II] sought to gain control of all ‘lesser governments’ in the Kingdom."
9
Therefore, in the model, we assume that a single variable, political strength, reflects all these notions. When strength is high, the politician incurs a large, direct cost in ignoring the profession during reforms. When strength is zero, there is no direct political cost of removing the profession from the reform process. 1.2.3 A Brief History of Organized Legal Professions Legal professions first emerged in England in the twelfth century (Brand 1992). Lawyers were recognized as having skills that were valuable to the government. They won privileges and status by virtue of their expertise. Serjeants, the most prominent members of the early English legal profession, were made officials of the crown. Parliament appointed lawyers as members of a commission inquiring into legal reforms. By 1300, judges were chosen from the bar rather than from royal officialdom (Plucknett 1983). The autonomous Inns of Court acquired monopoly control over education and admission to practice. The Inns' independence and power gradually became an element of the unwritten constitution. In the1680s, the crown attempted to challenge the power of the profession, but the result was the "the triumph of the Common Law and lawyers over the King, who had tried to put Prerogative above the law" (Trevelyan 1970, p. 71). Hence the power and independence of the English legal profession were won over a long period, in a process intertwined with the development of the common law. An attack on the status of the profession was not a viable political option after 1688 (Burrage 1997).8 The French legal profession developed to satisfy the specific needs of royal courts, parlements, which after their establishment in the late thirteenth century had full control over the profession (Karpik 1999). In the following centuries, the courts acted
Margaret Thatcher's government published a Green Paper that "anticipated, even celebrated, the end of sovereign, self-governing professions", but the government that had successfully implemented some of the most radical measures enacted in post-war Britain was forced to retreat quickly (Burrage 1997, p. 140).
8
10
partially as agents of the crown, while barristers became counsels of the crown. In contrast to England, the French barristers did not provide the pool for judicial appointments. The Paris Order of Barristers, France's first self-regulatory body of lawyers, was founded in 1660. The profession’s reputation stood high for the next century. In the 1760s, however, the French legal profession became a forum for battles between advocates of different views on the role of the monarchy, including whether the profession itself should be independent (Bell 1997). The internal dissensions, combined with an insecure power base, led to steadily diminishing authority in the period before the 1789 revolution. In contrast to England in 1688, the French legal profession was in no position to play the role of neutral arbiter in the country's decisive revolutionary event. The National Assembly abolished the Order of Barristers in 1790. Paradoxically, by pursuing broad political goals rather than narrow professional ones, French lawyers secured the demise of their own organized legal profession. Napoleon re-established the Order in 1810, as a creature of the state. In the volatile political environment that followed Napoleon's fall, the legal profession went through many vicissitudes, but never succeeded in achieving the prestige that it had acquired under Louis XIV (Karpik 1988). Only with the establishment of the Third Republic (1875) and the arrival of a stable democracy did the bar regain its old freedoms (Bell 1994). But these freedoms were bestowed by the state rather than won by the profession's political action. The German legal profession also had its origins in the state. Throughout the 18th century, Prussia reformed the size, composition, and education of the profession, in a
11
process that essentially converted lawyers into civil servants. Although the dominance of the government did come under attack in the 19th century, the central role of the state as patron of the profession was never in doubt. During German unification, professional autonomy increased but the formation of a country-wide association of lawyers in 1871 was in significant part a result of state action. In the Second Empire (1871?1918), the Ministry of Justice was the initiator of most legislation affecting the status of the profession (Rueschemeyer 1997). As the above indicates, the profession’s role varies across legal systems (Abel 1995). In a typical civil law country, the organized profession is a creature of the state.9 In a common law country, a unified profession usually won its own independence by dint of its technical expertise in a system of justice relying on decentralized courts and law. Historically, it has been largely autonomous from the state (Halliday and Karpik 1997, pp. 5, 354-5). Civil law professions are weaker, divided into sub-groups, and exert less influence on the judiciary. Colonial status also matters. In those British colonies where settlers were numerically dominant, such as Australia, legal professions followed the English path, attaining the same degree of autonomy and authority (Weisbrot 1988). In contrast where colonial settlement was less attractive, autonomous, independent professions failed to emerge, since the indigenous population was excluded from legal training. There was no basis for post-colonial professional power (Johnson 1973). In Kenya or in India, for example, no powerful independent profession ever emerged to exert a major influence (Ross 1992, Gandhi 1988).
For example, compare Belgium and Japan (Huyse 1988, p.182; Rokumoto 1988, p.128?9) to England and Australia (Burrage 1989; Weisbrot 1988).
9
12
Lastly, it is important to emphasize that while there are many examples of autocrats who oppose existing professions, powerful autocrats do not always take this stand. The French Order flourished under Louis XIV; Napoleon reestablished it after its demise in revolution. Lawyer groups had some (highly constrained) independence in Soviet times (Krause 1991). Moreover, transition from autocracy does not automatically translate into increased status for the profession, and might even lead to the reverse, as was the case in some post-Soviet republics (Waters 2004). 1.3 A Model of Professions in the Reform Process10 The model uses the phrasing of interest group politics in a democracy, but it could be recast to apply to autocracies. We use the terminology of competing politicians, elections, interest groups, and campaign contributions, but we could also have referred to an autocrat facing factions, receiving tribute, and deliberating on the possibility of being usurped. 1.3.1 The Structure of Reform Programs A fundamental institutional reform is under consideration. Reversal in the medium-term is impossible, and at most one reform program can be enacted.11 Politicians and interest groups compare reform programs to the status quo. Those directly affected fall into two interest groups, ? and ?, which can devote resources to influence the government’s decision.
10
Elements of the model are derived from Dewatripont and Maskin (1995) and Qian (1994), which focus on finance and on soft-budgets. To emphasize, our model is not relevant to short-term policy. Once the incumbent has had a chance to implement a reform, the country enters a new era with different economic circumstances and altered political configurations. Then, the question of which reforms to undertake is posed anew.
11
13
Reform programs comprise combinations of individual reforms. Reform j is completely characterized by B?j and B?j, the benefits to ? or ? from implementation. Classifying reforms according to whether ? or ? win or lose and whether the sum of payoffs is positive or negative, there are six distinct generic reform types. Figure 1 portrays these six, the line TT separating those with positive aggregate payoff from those with negative payoff. With reforms characterized by payoffs to ? or ?, all reform programs comprise from one to six reforms, at most one from each type.12 Hence, in the following, j = 1,…,6. The impetus for particular reforms comes from interest groups submitting programs. Hence, Reform 6, which is Pareto inferior to the status quo, can be dropped from the analysis because neither ? nor ? submits it. Given the assumptions below, Reform 1, which is Pareto improving, is always submitted and adopted. As its presence or absence does not alter the qualitative predictions of the model, reform 1 can also be omitted.13 Because reforms 4 and 5 play the same role for ? as 2 and 3 do for ?, the analysis begins by focusing on 2 and 3, whose payoff structure is summarized as follows: Reform 2 Reform 3 Interest group ? B?2>0 B?3>0 Interest group ? B?2<0 B?3<0 Aggregate (? and ?) B?2+B?2<0 B?3+B?3>0
Reform 2 is redistributive with a deadweight loss. Reform 3 also redistributes but with a welfare gain as a consequence of introducing more effective institutions. An example of reform 2 might be a change in corporate law that makes financial markets less efficient, aiding dominant owners by reducing protections for minority shareholders.
12 13
Two reforms of type j combined are equal to one reform of type j.
For further elaboration, see Section 1.4.2. A complete presentation of the model including reform 1 is presented in Appendix 4.
14
Reform 3 could be an improvement in contract law that facilitates the use of collateral, aiding new businesses, but reducing entry barriers that protect existing firms. Both measures could belong in an omnibus finance-oriented institutional reform program, underscoring the relevance of the model to situations where packages of related reforms are combined into one program. The structure of payoffs captures a fundamental aspect of institutional reform: rent-seeking interest groups favor reform programs with several sub-components, some contributing to general welfare and some reducing general welfare.14 The central problem of institutional reform is to foster adoption of the efficiency-enhancing components and reduce the likelihood of implementing the pure rent-seeking redistributions. We therefore introduce two assumptions that keep this problem at the center of considerations. Without these assumptions, a reform program can be dominated by one type of reform and would reflect only that type of reform. First, assume that reforms 2 and 3 have the same scale, in the sense that they involve the same amount of pure rent-seeking redistribution from ? to ?:15 Assumption A1. B?2=B?3. Since interest group submission fees must be related to the size of the reform, A1 fits naturally with our assumptions on those fees, which are discussed in the ensuing subsection. Second, assume that ? is, on balance, a rent-seeking interest group, in the sense that its favored program reduces the general welfare: Assumption A2. ?B?2+B?2?>B?3+B?3.16
14 15
This structure echoes the assumptions in Becker (1983). For brevity, the analogous assumptions for reforms 4 and 5 are omitted, but implicitly imposed.
15
A1 and A2 maintain the focus on the central problem of institutional reform by ensuring that A must blunt ?'s rent-seeking propensities if beneficial institutional reforms are to be enacted. 1.3.2 Interest Groups, Politicians, Society There are two rival politicians, A and B. The one forming the government has the power to implement reforms. The interest groups submit particular reform programs. The government accepts a submission only if the interest group pays a lobbying fee. Given A1, the fee does not vary across reforms: ?ik is the payment when the government of politician k accepts a submission of a single reform from interest group i.17 The politician is at an informational disadvantage and cannot recognize reform types when they are submitted. However, politicians do understand the model and recognize which interest group lobbies. Given the structure of payoffs, ? and ? are in opposition. Hence, it is natural to view enactment of reforms 2 or 3 as inconsistent with enactment of 4 or 5.18 One
Equivalently B?2+B?2+B?3+B?3 < 0. To understand the intuition of A2 note that if it does not hold, B?3 can be infinite. To see this, let B?3=B?2= ?c, 0<c<?. Since B?2+B?2<0, then B?2?(0,c). If ?B?2+B?2?<B?3+B?3, then B?3>?B?2+B?2??B?3= ?B?2+2c, so B?3=? is consistent with the model. With A2, B?3 is bounded by ?B?2+B?2??B?3= ?B?2+2c<?. Observe that if A2 does not hold, ?’s favorite reform program–although not first best for society? improves general welfare. The less ? is a rent-seeker, i.e. the greater is (B?3+B?3)??B?2+B?2?>0, the more the qualitative predictions differ from the ones presented here.
17 18
16
If interest group ? wants to submit both reforms 2 and 3 to A, for example, it must pay 2??A.
The assumption of direct conflict is common in the analysis of the effect of interest groups on policy. Examples from Grossman and Helpman (2001, sections 4.2.2, 8.4, and 9.3) are multiple lobbies when favored outcomes are on either side of the politician's ideal, when a business lobby wants lower tariffs and higher export subsidies than does a labor group, and when groups want to move in opposite directions from the status quo. Drazen (2000, Sections 3.7, 13.5) examines situations where redistribution is from one group to another and where special interest groups can block reforms in transition countries. Mueller (1989, pp. 108, 230, 245, 247, 279, and 453) offers many examples–liberal versus conservative reforms, rent-seeking for regulations, monopolies, tariffs, changes in the structure of regulatory processes, the supply of government output, macroeconomic policy, and agricultural programs.
16
politician cannot help both groups. We build this formally into the model's structure by assuming that each interest group is aligned with a different politician:19 Assumption A3: For all j such that B?j>0, ??A<B?j<??B. For all j such that B?j>0, ??B<B?j<??A. These assumptions fit the politics of longer-term institutional reform, where different visions of the future are in contention and the ideologies of politicians restrict their actions. Then, interest groups oppose each other in a manner that aligns with political ideologies: politicians could not credibly solicit support from all groups. The resultant structure of lobbying (? lobbies only A and ? only B) is consistent with empirical work on campaign contributions, which shows alignment between the objectives of donors and politicians (Poole and Roomer 1985; Poole, Romer, and Rosenthal 1987). This structure is also consistent with empirical work that shows that contributions influence politician's decisions (Stratmann 1991, 1995, 2002). Aggregate welfare is the sum of payoffs to the two interest groups: Assumption A4: Aggregate welfare from implementing reform j is B?j+B?j
19
A3 is formulated in strong terms to simplify the exposition. It is sufficient, not necessary. Its key implication is that the lobbying equilibrium within any time period results in only one of the following: (i) no lobbying or (ii) lobbying for 2 or 3 or both or (iii) lobbying for 4 or 5 or both. This implication is innocuous given that reforms 2 and 3 are contradictory with 4 and 5. Many different assumptions could lead to the same implication, for example, directly assuming that interest groups lobby against reforms they do not like or directly assuming that enactment of 2 or 3 involves repeal of 4 and 5. Assumption A3 can be interpreted as a strong alignment assumption since the lobbying fee charged by a politician to the non-aligned interest group exceeds the benefit from the reform to that interest group. The model’s qualitative features do not change with a weak alignment assumption, where the lobbying fee charged by a politician can be marginally smaller than the benefit from the reform to the interest group, that is, for all j such that B?j>0, ??B<??A<B?j<??A+? and for all j such that B?j>0, ??A<??B<B?j<??B+?, where ?>0 is arbitrarily small.
17
Note that the lobbying fee is a pure transfer and does not appear in aggregate welfare.20 Politician k’s payoff from accepting the proposal of reform j from group i and implementing it is a weighted average of aggregate welfare and lobbying income: Assumption A5: The politician's payoff from j is ?(B?j+B?j)+(1??)?ik, ??(0,1).21 ? is common knowledge. ? increases with either the politician's benevolence or with improvements in democracy that force the politician to pay greater attention to general welfare. If all politicians have equal degrees of benevolence, ? becomes a measure of democracy.22 Under dictatorship ? is closer to 0 and ?ik reflects the tribute the dictator extracts from subjects. 1.3.3 Politicians and the Organized Profession There exists an organized profession having the required expertise and authority to review, to comment on, and to stimulate public debate about proposed reform programs. Before interest groups submit proposals, the government decides whether such review is to be standard. By opening up the reform process to debate by an organized profession that values due process and openness, the review delays the passage of reform programs. In this paper, we ignore the negative role of the profession that is
20
This way of accounting for social welfare is consistent with Shleifer and Vishny (1994), Persson and Tabellini (2000, Chapter 7), Persson (1998), Grossman and Helpman (1994) and Grossman and Helpman (2001), Chapters 7 and 8.
? does not depend on the politician’s type, k. This specification of the politician’s payoff implicitly assumes that the politician does not attach any value to holding office per se. This form of the objective function is consistent with the political economy literature on campaign contributions; see Persson and Tabellini (2000, Chapter 7), Persson (1998) and Grossman and Helpman (2001, Chapters 7 and 8). If ?(B?j+B?j)+(1??)?ik, where ??(0,1), is rewritten as ?(B?j+B?j??ik)+?ik, then the politician’s objective function has the interpretation of the politician valuing a dollar in his hands more highly than a dollar in the hands of the public. Observe also that maximizing ?(B?j+B?j)+(1??)?ik, where ??(0,1), is equivalent to maximizing ?(B?j+B?j)+?ik, where ???/(1??)>0. See also Grossman and Helpman (1994) for a discussion of the politician’s objective functions. This interpretation of democracy exactly corresponds to Glaeser and Shleifer's (2002) interpretation of their parameter that captures the degree to which the preferences of the sovereign differ from those of the society.
22
21
18
usually emphasized, its interest in particular reforms, since our focus is on understanding whether there can be a positive aspect to professional activity. There are two central elements to our characterization of the government's use of professional review, first, that the government can build a credible, politically persuasive justification for review by using the organized profession and, second, that delay is the only effect. Section 1.2 presents the background on the first aspect: the politician can invoke the prestige and status of the profession, which flow from its monopoly expertise in a specialized body of knowledge. Since the interest group aligned with the incumbent politician is damaged by the delay, it is especially important that an aura of political legitimacy and credibility surround the decision to use the review process. This can only be provided by the use of the organized profession and the ensuing public debate: it could not be provided simply by qualified individuals personally chosen by the politician and it could not be provided by the politician simply deciding to delay consideration of the reform. Moreover, in those situations where the organized profession has political power, the politician's decision has a natural credibility. Professional involvement obviously does more than delay: the review process also generates technical information that might improve the quality of all reform proposals. But the generation of technical information does not require neutral and open review by the organized profession. It can be produced by qualified individuals who are trusted by the politician. Thus, the generation of technical information can be accomplished in a very different process from that examined here: it does not need the involvement of the organized profession. Hence, we assume that the review process does not lead to improvements in the quality of individual reforms. This assumption follows from the
19
objective of examining the effect of the organized profession, beyond the contribution of the technical skills possessed by individual members. The consequence of delay is that a particular government might run out of time before being able to implement a submitted reform program. The incumbent might lose an election and be replaced by the other politician.23 Then the interest groups have to incur lobbying costs again if they still want their reforms to be considered for implementation. 1.3.4 The Timeline There is a single reform process beginning when one politician, say A, takes the reins of power. The model covers two sub-periods, A's initial incumbency and the following time interval when either A or B forms the government. The decision on whether to use professional review is made at the beginning of A's incumbency and holds for the whole reform process. The irreversibility of this decision within the time-frame of the model reflects the view that once a single reform process has begun, it would be difficult for a politician to change that process. Let p?(0,1) be the known, exogenous probability that A stays in power in the second period. p is a measure of political stability.24 Values close to 0 correspond to chaotic, often revolutionary, environments, while values close to 1 reflect stagnation rather than stability. Hence, p is not an indicator of democracy: a p close to 1 occurs only under autocracy. A p close to 0 could occur under either democracy or autocracy.
23 24
We do not model the details of the process of change of power.
Our assumptions on p exactly match those in Persson and Tabellini's (2000, Section 13.3) analysis of political instability.
20
In both periods, each interest group decides whether or not to submit a proposal without knowing what the other is doing. The government then decides whether to accept the proposals. If it does, it receives the submission fee. If there is no professional review the measure passes immediately. If there is professional review, the measure is debated until the second period. Then, if A remains in office, the reforms submitted in period 1 are implemented. If B takes over, each interest group decides again whether to submit a proposal. If the reforms submitted in period 2 are the ones that already passed through professional review in period 1, they are implemented. If the reforms submitted in period 2 differ from the ones submitted in period 1, they are delayed beyond the end of this reform process.25 Figure 2 summarizes the timeline. 1.4 When Do Politicians Use Powerless Professions? We begin with the simpler form of the model, when a profession exists, but the politician suffers no special costs in ignoring it. This matches the circumstances of autocracy and the early phases of democracy. Section 1.5 introduces the political power of the profession, matching developments as democracy evolves. Reform 3 is always desirable for politician A.26 Reform 2 is undesirable for A if ?(B?2+B?2)+(1??)??A<0, or, equivalently, if ?>?, where 0 < ? = ??A/[?B?2+B?2?+??A] < 1. ? is therefore the lowest level of democracy at which the politician's consideration of the general welfare has any effect: for ?<?, A would knowingly accept aggregate-welfare-reducing reform 2. When ?>?, A has an
25
This assumption closes the model without an infinite horizon. It is equivalent to focusing on a single reform process that must take place within a certain time window, which is a common feature of the circumstances surrounding institutional reforms. That is, there does not exist a ??(0,1) such that ?(B?3+B?3)+(1??)??A<0.
26
21
informational problem. If there is no screening mechanism, A either approves all reforms or none. A approves all reforms if and only if [?(B?2+B?2)+(1??)??A]+[?(B?3+B?3)+(1??)??A]>0, or, equivalently, if ?<? ? where ?<? ? =??A/{??A+½[?B?2+B?2??(B?3+B?3)]}<1.27 ? ? is therefore the lowest level of democracy at which the general welfare dominates A's decisions. In sum,28 when ?<?, A likes reform 2, accepts all reforms, and does not face an ?, A accepts all submitted reform programs (unless a informational problem. When ?<?<? program is known to comprise 2 only), but faces an adverse selection problem. When ?>? ? A rejects all submitted reform programs (unless a program is known to comprise 3 only), but still faces an informational problem. The model's basic parameters are p and ?, the levels of stability and democracy. Figure 3 summarizes the analysis as it proceeds, depicting how the model's equilibria vary with (p, ?). 1.4.1 Baseline Equilibria: When the Profession Does Not Exist First consider equilibria in a baseline case, when an organized profession does not exist. Without the profession, with A in power, and with ?<? ?, ? never lobbies, while ? lobbies immediately for reforms 2 and 3 and A implements them both in the first period, ending the reform process. When ?>? ? neither ? nor ? lobby for reforms because A
27
Observe that ?<? ? implies that any politician in power accepts any reform program, submitted by ? or by ?. A might in principle (but not in equilibrium) receive reforms submitted by ? in period 1. If A accepts any reform submitted by ?, then A accepts any reform submitted from ?, since ?'s lobbying fee is higher and 4 and 5 are symmetric to 2 and 3. This summary holds only as long as the politician does not use a mechanism that ex-ante screens out the undesirable reform.
28
22
rejects all proposals.29 Denote social welfare relative to the status quo when there is no profession as W?. Hence, W? = (B?2+B?2)+(B?3+B?3) < 0 = 0 ? when ?<? ?. when ?>?
Similarly, let V? be politician A's payoff relative to the status quo when there is no profession: V? = ?[(B?2+B?2)+(B?3+B?3)]+(1??)2?11 > 0 = 0 when ?<? ? when ?>? ?.
Hence, ?<? ? provides the more interesting case, since the politician would then want to implement the proposed reform program, which would reduce aggregate economic ? , there is no such conflict between politician and society. welfare. In the case of ?>? 1.4.2 Equilibria When the Profession Exists and the Politician's Self Interest Matters ? , A is faced with an informational problem and approves all submitted When ?<? reforms, the standard adverse selection scenario.30 Then, the equilibrium of the reformlobbying game between ? and ? is specified by the following:31 ? , A is in power, and there is professional review. Proposition 1. Assume that ?<? ? never lobbies for reforms 4 and 5. In period 1, ? lobbies for 3 if and only if p???A/B?3 and also for 2 if and only if p???A/B?2. In period 2, ? does not lobby for any reform regardless of which politician is in power. Interest group ? never lobbies for reforms. Therefore, in equilibrium, ? remains inactive (i.e. never lobbies). This conclusion is crucially related to our assumption that all Pareto-improving reforms have already been
Given that at most one reform program is implemented and the symmetry between ? and A and ? and B, nothing will occur in the second period, whoever is in power.
30 31
29
See, for example, Dewatripont and Maskin (1995) and Qian (1994). Proofs of all propositions are in Appendix 3.
23
implemented. Had we included reform 1 in the analysis, ? and ? would behave strategically when lobbying for its implementation and ? would actively engage in lobbying in equilibrium. It is therefore appropriate to view ? as an interest group (rather than the general public, say).32 Let W+ be social welfare with professional review. Proposition 2 immediately follows: Proposition 2. When ?<? ?, then: 0 when p<??A/B?3, + W = p(B?3+B?3) when ??A/B?3?p<??A/B?2, when p???A/B?2 p(B?2+B?2+B?3+B?3) W+ > W? for all p: aggregate welfare is always higher with the profession than without it. The reason why the profession is advantageous to society varies with levels of stability. If p is low, professional review dissuades ? from lobbying because A's survival is too uncertain to repay lobbying costs. If p is high, review and delay reduce the probability of implementation of a welfare-reducing reform program (2 and 3). For midrange p's, the profession solves society's informational problem: ? lobbies only for reform 3. Therefore, the use of professional review makes its strongest contribution in this mid-range, suggesting patterns of variation across countries in the value-added from an organized profession. The contribution of the profession identified in Proposition 2 is only a potential one, contingent on whether the politician decides to implement professional review. ? , it would not be surprising to find situations where the politician's objectives When ?<?
32
Intuitively, ? lobbies for reform 1 in period 1 only, and ? lobbies for reform 1 in period 2 only if B is in power, resulting in an equilibrium where (i) ? actively engages in lobbying, and (ii) reform 1 is always implemented. See Appendix 2 for proofs.
24
conflict with those of the general population. Denote the politician’s payoff with professional review as V+. The following proposition relates the difference between V+ and V? to the values of p and ?: Proposition 3. Assume ?<? ?. Let p?(?)?1+[?(B?2+B?2)+(1??)??A]/[?(B?3+B?3)]. 1. With professional review, the politician's welfare is given by: 0 if p<??A/B?3 + V = ?p(B?3+B?3)+(1??)??A if ??A/B?3?p<??A/B?2 ?p(B?2+B?2+B?3+B?3)+(1??)2??A if p???A/B?2 2. The politician's gain or loss from professional review has the following properties: V+ < V? if p<??A/B?3 + ? if ??A/B?3?p<??A/B?2 and p< p?(?) V <V + ? V >V if ??A/B?3?p<??A/B?2 and p> p?(?) V+ > V? if p???A/B?2 ? 3. The line p (?) in Figure 3 is downward sloping, convex to the origin, intersects p = ??A/B?3 below ? ?, and intersects p = ??A/B?2 above ?. Proposition 3 establishes that the decision to use the profession is related to (p, ?) in a straightforward way, as shown in Figure 3. Before discussing this relationship, we find the remaining equilibria of the model by examining outcomes when ?>? ?. 1.4.3 Equilibria When the Profession Exists and the Politician's Aims Coincide with Society's With professional review and ?>? ?, if p<??A/B?3, ? does not lobby because the probability of A's keeping power is not high enough to repay lobbying costs.33 Therefore, if p<??A/B?3, W+ = V+ = 0 (= W? = V?). When p?[??A/B?3,??A/B?2), lobbying for only reform 3 is beneficial for ?, while lobbying for 2 and 3 is not. Since A knows ?’s incentives and reform 3 is desirable for A, it is approved. Therefore, when
33
We omit stating a modified version of Proposition 1 for brevity only. As in equilibria with ?<? ?, ? never lobbies for reforms.
25
p?[??A/B?3,??A/B?2), W+ = p(B?3+B?3) > W? = 0 and V+ = ?p(B?3+B?3)+(1??)??A > V? = 0. For p???A/B?2, A approves ?'s submission if ?p(B?2+B?2+B?3+B?3)+(1??)2??A>0, or equivalently, if ?<?C(p), where ?C(p) = ??A/{??A+½p[?B?2+B?2??(B?3+B?3)]}. ?C(p) is strictly decreasing and convex in p with ?C(0) = 1, ?C(1) = ? ?. Therefore, for p???A/B?2 and ?>?C(p), A rejects all submitted reform programs and consequently ? does not lobby for any: W+ = V+ = 0 (= W? = V?). For p???A/B?2 and ?<?C(p), however, ? lobbies for the reform program consisting of reforms 2 and 3, and A approves it, meaning that V+ > V? = 0 and W+ = p(B?2+B?2+B?3+B?3) < W? = 0. Figure 3 summarizes the central results and is used to motivate the applications of the model in the ensuing subsections. The profession, if used, promotes the general ? , and only detracts from general welfare in a small part of the region welfare for all ?<? ? . But it is only used if it is to the politician's advantage (V+?V?). At high where ?>? levels of either democracy or stability, the politician invokes professional review. As either democracy or stability fall from their highest levels, they become substitutes. At many levels of democracy, the profession is used only if political stability is high enough. It is because professional review solves the informational problem that there this trade-off between democracy and stability.
26
1.4.4 Professions Under Autocracy and Transitions to Democracy Raising the status of the profession is consistent with autocracy, so long as there is also a high level of stability. When democracy is low and stability is high, an absolute autocrat who brooks no independence would still use the profession (V+>V?). France under the old regime provides an outstanding example. During the century of domestic political stability initiated by Henri IV (1589-1610), the French legal profession rose in status and power. With Louis XIV (1643-1715) the very symbol of absolute monarchy, one would not associate his era with the development of decentralized institutions. His treatment of the Huguenots showed that independent views were not to be valued for their own sake. But France was very stable, and independent entities could provide a useful service to the monarch. The first bar association, the Order of Barristers, was formed in 1660 and won a monopoly over a range of legal activities while controlling recruitment and discipline. "In the second half of the 1660’s, Louis XIV issued his great ordinances on civil procedure and criminal law, and this overhaul of large parts of the French legal code put a premium both on legal expertise, and on the sort of back-breaking intellectual labor that barristers prided themselves on. Chancellor Lamoignon himself convened a special committee of barristers to work on the projects.…mere barristers sat at the same table with the chief magistrates of the parlement. After the project had been completed, the king himself thanked the Order for its services (Bell 1994, pp. 55?6)." Our model's predictions are consistent with the observation that the prestige of the French legal profession stood high in the early part of the 18th century. De Tocqueville (1955, pp. 115-6) commented about these times that: "For though as far as administration
27
and political institutions were concerned France had succumbed to absolutism, our judicial institutions were still those of a free people." Had revolution occurred at this time, when the legal profession had gained some power and was politically neutral, perhaps the evolution of the rule of law in France would have taken a different course than it did. But the strength of the legal profession declined over the 18th century. We postpone discussion of this until Section 1.5, when the introduction of professional political power facilitates comparison with English developments. The role of lawyers under Soviet rule and in the transition can also be elucidated with our predictions on the use of a politically powerless profession. While the communist revolution in 1917 destroyed the pre-revolutionary Russian legal profession (Shelley 1991, pp. 65?6), the return of stability under Stalin was accompanied by a recognition that lawyers would play a role in the Soviet state (Waters 2004, p. 41). Elements of the legal profession had a degree of independence not observed elsewhere in the Soviet system (Shapiro 1961). Stalin allowed the advokatura (defense lawyers) to form a "British barrister-style set of collegia or lawyers’ self-controlled practice groups, self-funding and attached to each major court….private, petit bourgeois 'mini-guilds' within the state structure" (Krause 1991, pp. 16?8), which were subject to far less state control than other types of collectives (Shapiro 1961). Even in the Stalinist state, "professional pride emerged among many lawyers, as they had a claim to expertise based on specialized knowledge" (Shelley 1991, p. 66) and their influence and autonomy increased over time (Krause 1991, p. 18). The advokatura provided a rare example of a profession that was formally outside all administrative structures and, within the narrow limits set by the Soviet state, self-governing (Huskey 1982).
28
The Soviet legal profession did affect policy in the manner depicted in our model. One well-documented case is that of criminal law and criminal legal-procedure reform in the years of destalinization (McCain 1982). In this process, legal scholars and practitioners were invited to debate the many proposals that surfaced after the Communist Party announced that new legal codes would be adopted. This resulted in exactly the process of delay envisaged in our model, a two-year "elaborate ritual of consultation with legal specialists" (McCain 1982, p. 6), during which lawyers played important roles in the process of drafting and re-drafting the proposed codes. "Because the political leadership needed the technical skills of lawyers in producing new criminal and procedural statutes, the drafting process was designed to involve Soviet jurists, and in many instances they predominated on the legislative committees." (McCain 1982, pp.1920). The special role played by sub-groups of the legal profession continued throughout the Soviet period, including involvement of the advokatura itself in the lengthy review of drafts of the 1979 Law on the Advokatura (Huskey 1982, p. 204). Although the role of the legal profession in Soviet society was minor even compared with its role in medieval England, the existence of any professional influence in a totalitarian state is surprising, but consistent with our prediction that an unchallenged dictator can find professions useful. The Soviet Union before Gorbachev could be placed at point S in Figure 3. The role of the legal professions was enhanced under Gorbachev, with lawyer groups influencing the development of criminal law, playing a more central role in legislative drafting, and gaining further tools of self-regulation (Jordan 1998). The Union of Advocates of the USSR was established in early 1989, the first independent union to
29
be formed since 1917 (Jordan 1998, p. 770). Because the early Gorbachev years were a time when democracy rose faster than instability, this is consistent with the model's predictions. In contrast, in early transition the continued rise of democracy was accompanied by increases in political instability. In Russia in the 1990s, the power of the advokatura declined: they sought the state's help in warding off new competition rather than increasing their own self-regulation (Jordan 1998). Their status still remained governed by a Brezhnev-era law. Thus, in the instability of the early Yeltsin years, the rise in democracy was not sufficient to encourage politicians to enhance the role of that part of the legal profession that had been most independent under Soviet rule.34 The developments were even more stark in the new republics, such as Georgia, where instability dominated. Legal groups lost power and lawyering became a free for all. Politicians paid little attention to the profession and it was scientists and literary figures, and not legal experts, who dominated the process of drafting a constitution (Waters 2004, pp. 57?8). This pattern of events is consistent with our model's predictions. While the Soviet regime found some use for the legal profession, the higher levels of democracy of the early 1990s did not translate into greater use of the profession, because they were more than countered by the increases in political instability. Early transition in Georgia, and perhaps even in Russia, is at G in Figure 3. In Poland, where much higher levels of democracy were quickly attained and instability declined, the legal profession was prominent. Law professors and legal
34
As is typical of civil law countries, the legal profession in Russia is divided into separate groups. Some elements of the legal profession did attain higher status under Yeltsin, for example judges in the arbitrazh courts.
30
professionals were regularly appointed to task forces that were charged with drafting new laws (Gostynski and Garfield 1993). The use of recognized legal experts to review proposed laws delayed economic reforms (Rich 1997). The higher levels of both stability and democracy would place Poland at point P in Figure 3. In Georgia, it was only in the late 1990s, with the appearance of greater political stability, that lawyers played a more important role. Lawyer groups were now chosen to carry out legislative drafting projects and were able to play a role in pressing the government to pay more attention to the rule of law (Waters 2004, pp. 100, 168?9). The interpretation of these historical developments provides insights into why the early years of transition to democracy are so important. Whether civil society groups, such as professional organizations, benefit from transition depends upon the balance between changes in democracy and instability. When the rise in democracy more than counters the increase in instability, as in Poland, politicians enhance professional status and the profession builds upon any prestige it held under the old regime. When the rise in instability dominates, as in Georgia, politicians ignore the profession, eroding any professional power inherited from the old regime. The early years of transition are therefore crucial: the maintenance of stability as democracy rises can lead to reinforcement of the role of the profession and improved institutional reforms. This increases the political power of the profession. Then, as the next section shows, the enhanced political power can encourage the politician to use the profession when the country is at a point such as G in Figure 3. These interactions increase the degree of path dependence in transition.
31
1.5 When Do Politicians Use Professions that Have Political Power? Often professions are entrenched and a politician bears significant extra political costs in removing a profession from its advisory role. These political costs increase with the level of democracy. A powerful autocrat can purge a profession at whim. For example, within one year of ascending to power, Hitler abolished the German lawyers association (Jarausch 1990, p. 118). In contrast, professions can vanquish even the most determined democratic politician. In the late 1980s, the Thatcher government retreated from an attempt to reform the English legal profession and reduce its autonomy (Burrage 1997, pp. 137?40). To capture the fact that professional power is more effective in a democracy, we assume that the political costs of removing the profession from its advisory role are c? (c>0), which is subtracted from the politician's payoff in A5. c quantifies the political strength of the profession: c is smaller where the profession is new or where it is a tool of the government. This simple functional form is chosen for ease of exposition. None of the results depend upon it.35 The politician would like to keep the profession in its advisory role if at the prevailing (p,?), V??V+ ? c? and would like to remove the profession when V??V+ > c?. For a given (p,c), define ?T(p,c) as the least value in [0,1] such that for all ?>?T(p,c), the politician keeps the profession. Then, since V??V+ is non-increasing in ? at any given p, the politician removes the profession if 0 < ? < ?T(p,c). The following proposition
35
For example, all of the results hold when the political costs are constant, independent of democracy, or when political costs are an increasing function of ? and the first derivative of political costs with respect to ? increase with the profession's political strength. Even these assumptions are far stronger than necessary.
32
establishes that a unique ?T(p,c) exists for each (p,c) and characterizes the effect of p on ?T: Proposition 4. For a given c?0, (i) ?T(p,c) is a function with points of discontinuity at p = ??A/B?2 and at p = ??A/B?3. (ii) ?T(p,c) is constant for all p?(0, ??A/B?3) and attains its maximum in that interval. (iii) ?T(p,c) is decreasing and convex in p for p?[??A/B?3,??A/B?2). (iv) ?T(p,c) = 0 for p???A/B?2. The next proposition shows how changes in the strength of the profession, c, affect outcomes: Proposition 5. If c1>c2?0, then (i) For p<??A/B?2, ?T(p, c2) > ?T(p, c1) > 0. For p<??A/B?3, ?T(p, c2) ? ?T(p, c1) > 0 does not vary with p. For p?[??A/B?3,??A/B?2), ?T(p, c2) ? ?T(p, c1) > 0 is decreasing in p. (ii) For p???A/B?2, ?T(p, c2) = ?T(p, c1) = 0. Figures 4 and 5 summarize these two propositions, highlighting the interactions between stability, democracy, and professional power. At high levels of stability, professional power does not change the politician's decisions (and consequently aggregate welfare): the profession is used even by autocrats facing a weak profession. As levels of stability decrease, democracy becomes more important: ? affects political decisions and there is a widening of the interval of ? in which professional power makes a difference. This result appears clearly in Figure 5, which is Figure 3 modified to show the set of (p, ?) for which professional political power makes a difference.36 In the shaded area in Figure 5, professional power increases the level of social welfare by encouraging the politician to keep an institution, professional review, that improves institutional reforms for society as a whole, but not for the politician. When the strength
36
Figure 5 has the c2 of proposition 5 equal to zero. Note that ?T(p,0) lies along the solid line in Figure 3.
33
of the profession is higher, the level of democracy at which the politician chooses the socially optimal outcome is lower. The strength of the profession substitutes for weak democracy.37 1.5.1 Comparative Legal Development in England and France: 1688 and 1789 These results provide insights into why the revolutions of 1789 and 1688 had such different consequences for the role of law in the two countries. The French legal profession's high status, achieved under Louis XIV, continued into the reign of Louis XV, when the Jansenist movement confronted the monarchy. As the Jansenists fought their battles in the courts, barristers played key roles by using factums, written briefs that escaped censorship (Bell 1997, pp. 74?8): as in our model, the profession leads to deliberation and delay. The regime chose to do no more than denounce the Order as "a sort of absolutely independent little republic at the heart of the state" (Bell 1994, p. 67). To illustrate using Figure 5, France could lie at F1 under Louis XIV, where the unchallenged autocrat uses a powerless profession, and at point F2 under Louis XV: a rise in instability alters the monarch's view of the profession but not the outcome, since the profession has acquired political power. Matters changed as the lawyers began to have wider political ambitions. In the 1750's, a long-standing conflict between the parlements and the crown degenerated into full-scale constitutional conflict: magistrates, advised by Jansenist barristers, demanded a greater role in the legislative process, while the king wanted them to play a purely judicial role (Bell 1997, p. 83). The earlier political success of the barrister pamphleteers
37
This conclusion does not depend at all on the assumption that the costs of removing the profession are c?. Exactly the same conclusion would apply if the costs of removing the profession were completely independent of levels of democracy.
34
attracted imitators who did not share the same political and professional views and chose to support the crown, attacking the idea of an organized legal profession. The Order was fatally weakened. Louis XV could now respond by attacking the legal profession in general. The Order’s monopoly and disciplinary powers were abolished in 1771. The old parlements were replaced with trusted royal appointees. Although Louis XVI recalled the old parlements, the organized legal profession did not recover its former status (Bell 1997, pp. 90?5). In contrast to earlier in the century, when Louis XV chose not to remove a troublesome profession, the profession's power was lower. In contrast to a century earlier, when Louis XIV used the profession and allowed it to attain power, political instability was higher making the profession less valuable. France under Louis XVI can still be viewed as lying at F2 in Figure 5, but since the profession weakened itself and lost its power, the solid upper line is now relevant, not the dashed lower one: the monarch could do what was too costly before, remove the profession from its role. When 1789 came, the Order of Barristers had no authority (Bell 1997, pp. 90?5). A year later, the National Assembly formally abolished the Order and the French legal profession virtually vanished (Burrage 1989, p. 330). Consistent with our model, this brief period of higher democracy did not result in increased use of the legal profession: politicians were neither forced to use a politically weak profession nor saw advantage in it at a time of great instability (e.g., point F3 in Figure 5). Napoleon used prerevolutionary judges and practitioners to draft his civil code and re-established the Orders in 1810, showing that a return to stability, even after a decline in democracy, could lead to a reinstatement of the profession (e.g., back to point F1). But he instituted a very
35
different legal profession than before. In contrast to earlier times, Napoleon's legal profession was his own creation and remained under strict state control (Bell 1997, p. 99; Karpik 1999, p. 120; Bell 1994, pp. 214?5). From that time on, any independence attained by the French legal profession was given by government, not won from the state. The English legal profession had substantial power already in the middle ages and asserted itself more strongly than ever in the early 1600's. In 1640, at the beginning of the Puritan revolution, Parliament established the supremacy of the common law courts over the crown's prerogative courts, bolstering the power of the legal profession. However, the Puritan revolutionaries who were able to remove a king's head were unwilling to go as far as pressing for popular legal reforms over the objections of the legal establishment: "…the power of the English lawyers, which confounded Cromwell, derived from a solidarity that transcended revolutionary enthusiasms…" (Burrage 1989, p. 356). Later, in the 1680's, the legal profession was prominent in the struggle against the crown over the independence of chartered bodies and other "lesser governments", essentially a civil war fought with law rather than the sword (Landon 1970, pp. 100?1). One of those lesser governments was the Inns of Court, the legal profession's stronghold. Because the Inns' independence was based on custom, not charter, the crown pressured the Inns by forcing royal appointees onto their governing bodies. The Inns resisted (Burrage 1997, p. 150). James II even faced a shortage of lawyers willing to fight his battles. In March 1686, the attorney general and the solicitor general both refused to implement royal measures that would have flouted common-law precedent (Landon 1970, pp.191). In 1688, in the trial of the Seven Bishops, often taken to signify the
36
beginning of the Glorious Revolution, the crown faced an almost united legal profession, including eminent Tory lawyers, and lost. At issue was whether the royal prerogative could negate the common law, as interpreted by the legal profession (Landon 1970, Chapter 6). In sum, England in the years preceding the Glorious Revolution could be thought of as being at E in Figure 5: the King had every reason to remove the legal profession from its role, but was unwilling to pay the political cost of doing so by extra-legal means. Of course, James' battle with the legal profession was only one element of many contributing to the final outcome. But James did not succeed in weakening or even dividing the legal profession: the Glorious Revolution entailed the triumph of commonlaw lawyers and their interpretation of the law (Landon 1970, p. 248). In the construction of the post-revolution settlement, lawyers played a fundamental role, even in the House of Lords where the Tory establishment was strong. Exactly as in our model, the legal profession was central in the debate because the settlement had to be made consistent with 500 years of common law. There could not be a greater contrast between the fate of the English legal profession in 1688 and that of the French profession in 1789. The English profession had been in existence for four centuries and the removal of its autonomous institutions and practices would have come at great political cost. The French profession weakened itself during the eighteenth century and was easily removed by a monarch who no longer regarded it as useful. The English legal profession played a major role in the post-1688 settlement, improving institutional reforms. The French profession was abolished a year
37
after the Bastille fell and played no role in institutional development. The consequences of these divergent histories are with us still. 1.5.2 The Professions Under Civil and Common Law A burgeoning empirical literature examines the historical roots of institutional quality, particularly focusing on legal and colonial origins (La Porta et al. 1998; Acemoglu, Johnson, and Robinson 2001). In this section and the following one, we show how an emphasis on organized legal professions can be complementary to both of these approaches. Moreover, as our model offers a new lens to interpret existing results, we shed light on empirical findings that have been difficult to explain within existing approaches. Although there is substantial evidence that legal origin affects institutional quality (Djankov et al. 2003), there is still uncertainty about the precise mechanisms that make legal origin matter (La Porta et al. 2004; Beck et al. 2003b). La Porta et al. (1998) and La Porta et al. (2004) suggest two complementary reasons: both the quality of legal rules and the strength of judicial checks and balances vary across legal systems. Rajan and Zingales (2003) challenge the evidence on quality of legal rules, showing that commonlaw countries had lower levels of financial development in the beginning of the twentieth century. They propose a different reason for the varying properties of different legal systems: private interests are more likely to see their agenda enacted in civil law countries because laws emanate from the center rather than evolving through legal process. New policies are enacted more quickly in those countries, sometimes leading to better outcomes, sometimes worse, depending on the configuration of interest groups.
38
By introducing a different causal mechanism, our model suggests a reconciliation of La Porta et al. (1998) and Rajan and Zingales (2003). A strong, independent legal profession coevolved with the common law system, exactly because decentralized and autonomous legal institutions are intrinsic in that system. In contrast, state-sponsored, state-controlled professions are consistent with a civil law system.38 Since the use of the profession improves the quality of institutional reforms and since the strength of the profession can be critical in determining its use by the politician, our model predicts that common law countries tend to have better legal rules than civil law countries, consistently with La Porta et al. (1998). Moreover, as delay is intrinsic in the use of the profession, reform will be quicker on average in civil law countries, and the government responds faster to interest groups, consistently with Rajan and Zingales (2003). La Porta et al. (2004) show that judicial independence is associated with economic and political freedom and that the higher level of judicial independence in common law countries accounts for some of the effects of legal origin on economic freedoms. Even though the current incarnation of our model does not focus directly on judicial independence and checks and balances, the mechanisms depicted in our model could well account for the results of La Porta et al. (2004). Civil law legal professions are weaker, more dependent on the state, and balkanized, with judges and private lawyers organized separately. Organized legal professions in common-law countries are more likely to have the incentive and the power to protect judicial independence. The professional review depicted in our model has some similarities to judicial checks and
The difference between the profession’s political power in common and civil law countries has been widely recognized in the sociological literature (Halliday and Karpik 1997, pp. 5, 354?5; Abel 1995, p. 3; Burrage, Jarausch and Siegrist 1990, pp. 219?20; Siegrist 1990, pp. 181?2; Collins 1990, p. 16; Cleaves 1987, p. 10).
38
39
balances, delaying the implementation of legislation through a process that examines its consistency with existing legal doctrines. If this analogy holds, then our model would predict that politicians would use judicial checks and balances more in common law countries and that this choice results in better institutions, consistently with the results of La Porta et al. (2004). Our model also identifies when legal origin matters most, and by doing so resolves some apparent empirical anomalies. At the lowest levels of democracy, neither common nor civil law countries adopt professional review (unless stability is very high). When the level of democracy is high, both common and civil law countries use the profession.39 Therefore, legal system matters most at middling levels of democracy, when the politician's decision hangs on the political cost of removing the profession from its advisory role. In this range of democracy, common-law politicians are more likely to use the legal profession to act as a filtering device, with a resultant increase in the quality of institutional reforms. Legal origin has no effect at either the highest or the lowest levels of democracy. One recurring empirical observation is that France itself and the German-legalorigin countries do not suffer as much from the disadvantages of the civil-law system as expected from the results on civil-law countries as a whole (Beck et al. 2003b). Our analysis provides a possible explanation: France has a high level of democracy relative to French legal-origin countries in general and all the German legal-origin countries have high levels of democracy or stability (often both). In these situations, the political power
Our conclusion that common and civil law systems converge at highest levels of democracy is therefore in accordance with Glaeser and Shleifer (2002) analysis of circumstances under which the decisions made under bright line rules, predominantly characteristic of civil codes, and independent juries overlap.
39
40
of the profession makes less difference and our model predicts similar outcomes for civiland common-law countries. Berkowitz et al. (2003) have challenged the importance of legal origin, arguing that the crucial issue is whether legal transplantation was receptive or unreceptive. A key aspect of receptive transplantation is that the country contemplates the consequences of adopting a foreign legal system. Unreceptive transplantation occurs when the legal profession is so underdeveloped that laws are implemented without considering their contents (Berkowitz et al. 2003, p. 180). Hence, receptive transplants are more likely to occur when an organized legal profession undertakes review of reform proposals. This is consistent with our approach, but our model adds an extra element, showing why receptive transplants are more likely to occur in common law countries, which use the professions more often. Hence, our approach shows why results on legal origin are less strong when form of transplantation is included in the same analysis (Berkowitz et al. 2003): mode of transplantation is a reflection of legal origin itself.40 1.5.3 Colonies and Institutions Acemoglu, Johnson, and Robinson (2001) argue that institutions supporting longterm growth were more likely to be established in those colonies where Europeans settled in larger numbers, with consequences persisting until today. Their reasoning focuses on the demand side of politics: dominant political groups demanded different institutions in settler colonies than in extractive colonies. The key variable driving their evidence is
Glaeser and Shleifer (2002) argue that transplantation might be particularly unsuccessful for civil law systems because of their reliance on bright-line rules, which need to be modified for a different setting. Given that the organized legal profession will be important in such modification, the two views are complementary.
40
41
hospitality of settlement, extractive colonies appearing where settlement was less attractive for Europeans. Our model has supply-side implications that are consistent with the evidence. In those British colonies where settlement was unattractive, the demand for legal services was met by temporary labor. The colonial administration became the patron of the professions and did not find it advantageous to develop indigenous talent (Johnson 1973). Under such circumstances, "…there was no viable basis for independent professional practice nor was there a secure foundation for the emergence of professionalism…Under these conditions the technically-based authority of the professions was subordinated to extra-professional sources of power." (Johnson 1973, p. 289) These characteristics persisted into the post-independence period. In Kenya, for example, the legal profession was controlled by non-African lawyers and was fragmented during and after the colonial period (Odenyo 1979; Ross 1992). With both democracy and the profession weak, post-colonial governments impeded the growth of an autonomous legal profession (Ross 1992). In India, the profession was a British tool with no indigenous roots. The colonial authorities, apprehensive that members of the legal profession might lead the struggle for independence, repressed autonomous Indian lawyers’ associations. Consequently, a strong, autonomous profession did not emerge in modern India (Gandhi 1988). Australia, a settler colony, provides a remarkable contrast. The legal profession descended directly from British roots and developed remarkable autonomy and power. Key elements of the modern professional structure were in place before colonial selfgovernment. Therefore in contrast to most African and Asian colonies, post-
42
independence Australian governments faced a legal profession with much power to control legal practice, including admission and discipline. Although under political attack several times in both the nineteenth and twentieth centuries, Australian lawyer associations have generally been the victors (Weisbrot 1988). The Australian Law Reform Commission (2003), on which the legal profession is dominant, matches the advisory role played by the profession in our model. The colonial heritage of professional political power had significant effects.41 Lastly, our model can account for interactions between the effects of legal and colonial origin. Beck et al. (2003a) identify two such interactions: first, the effect of settler mortality on financial development is stronger for English legal origin countries than for French legal-origin ones; and second, the effect of legal origin on financial development is greater in countries where settler mortality is low. Our model predicts the first result because of the different strengths of professions in the two sets of countries: higher levels of migration by professionals are more likely to translate into an influence of the organized profession on reforms in English legal-origin countries than in those with French origins. The second result follows because organized legal professions did not develop the countries where settlement was unattractive for Europeans, implying that the effect of legal origin would be diminished in such countries. 1.6 A Postscript on Further Research This chapter has argued that organized legal professions provide an example of an institution that has survived in a Hayekian trial-and-error development process. The
41
Note the consistency with Berkowitz et al. (2003), who argue that migration is a key to receptive transplantation in colonies.
43
argument should apply to other organized professions that have similar characteristics, but we leave it to future research to determine to which professions our analysis applies. We have stressed the importance of the review of proposed legislation by an organized profession. Since in our model this review occurs before the passage of legislation, we have chosen not to place judicial checks and balances as an element of the professional review. However, there are obvious similarities between the review of proposed legislation against a set of legal standards promulgated by an organized profession and review of the implementation of legislation by judges administering a body of law. Given these similarities, one obvious extension of our approach would be an examination of the processes of judicial checks and balances, which La Porta et al. (2004) have shown to be of substantial importance. The professions are one element in that set of arrangements that have come to be known as civil society (Bell 1997, p. 86). In policy circles, it is an article of faith that promotion of civil society contributes to economic development (Carothers 1999). Yet, fostering civil society sounds suspiciously like promoting interest group development, and there is much in past economic research to suggest that interest groups slow economic development. As another extension of this essay, an analytical examination of how civil society can be beneficial would offer guidance for the conduct of practical development policy.
44
Chapter 2 Allocating Law-Making Powers: Self-Regulation vs. Government Regulation
2.1 Introduction Governments employ a variety of institutional arrangements to regulate and intervene in the economy. Self-regulation is a result of "deliberate delegation of the state’s law-making powers to an agency, the membership of which wholly or mainly comprises representatives of the firms or individuals whose activities are being regulated" (Ogus 1999, p. 590). Many sectors and activities self-regulate.42 Moreover, there are episodes when self-regulatory policies are initiated on an economy-wide basis.43 This chapter examines when self-regulatory regimes arise, and when they are socially efficient relative to regimes that consolidate regulatory powers at the governmental level. Despite the ubiquity of self-regulation and the attention it attracts in government policy circles, formal analyses of the economics of self-regulation are scarce and, in the aggregate, inconclusive about its welfare properties.44 Leland (1979) and Shaked and
42
E.g., professional and financial services, sports, advertising, internet, education, insurance, and press. Likewise, implementation of regulatory goals in the context of environmental policy is often delegated to industries. Self-regulatory arrangements differ in the extent of the role played by the self-regulators, selfregulators' degree of autonomy from the government, and legal force of their rules. Voluntary selfregulation is exemplified by the proliferation of industry codes of conduct. Occupational licensing and certification are examples of statutory self-regulation. Under supervised self-regulation, a self-regulatory organization is subject to continuous oversight by a specialized government body, as when the SEC supervises the NASD. See Priest (1997), Baldwin and Cave (1999, Chapter 10), Ogus (1999, pp. 587?8) for a comprehensive discussion. We discuss two prominent historical examples, fascism and New Deal.
43 44
See, e.g., http://www.selfregulation.gov.au, http://www.selfregulation.info, http://www.easa-alliance.org, http://medt.strategypartner.ru/eng/selfreg.htm for recent policy-oriented discussions on self-regulation. The formal literature reviewed here typically focuses on a narrower question of functioning of selfregulated markets rather than analyzing economy-wide delegation of regulatory authority and institutional transformation, associated with a self-regulatory regime (Eisner 2000).
45
Sutton (1981) stress the undesirable anti-competitive effects of self-regulation.45 In contrast, Kranton (2003) and Lyon and Maxwell (2000) argue that self-regulation may improve social welfare.46 But foremost, this literature sidesteps explicitly comparing self-regulation with its alternative, direct government regulation. Two key questions for successful regulatory design are thereby left unanswered. First, under what circumstances, if any, does self-regulation welfare-dominate its alternative, government regulation? And second, what factors influence the government’s choice of centralization or delegation of regulatory authority?47 In the spirit of Gehrig and Jost (1995), we argue that the source of potential efficiency gains from implementing a self-regulatory regime is the producers’ superior knowledge about the regulatory issues at stake.48 However, we depart from the existing
45
Building on Akerlof’s work, Leland (1979) shows that licensing, by setting minimum quality standards, improves the welfare-properties of equilibria in markets with asymmetric information. Professional (or industry) self-regulation, by exerting monopoly-like behavior, however, is expected to result in these standards set too high. This holds at least when the opportunity costs of supplying quality increase with the latter. When the opportunity costs decrease with quality, however, Leland shows that licensing standards set by producers themselves may not be too restrictive. Shaked and Sutton (1981) similarly show that granting a profession monopoly rights is welfare reducing. By self-regulating, the profession shrinks to a size that is socially sub-optimal. Kranton (2003) illustrates that industry self-regulation, with correspondingly restricted competition, sanctioned by the government, may be necessary to preserve incentives for high-quality production and therefore higher social welfare. Maxwell and Lyon (2000) argue that industries voluntarily adopt environment-friendly processes to preempt government regulation. Doing so must increase firms' profits, and may increase consumers' welfare as well when consumers' costs of influencing governmental regulatory policy are taken into account.
46
47
Donabedian (1995) and Ashby et al. (2004) offer a rationale for the observed differences in regulatory arrangements, but they are silent on their welfare properties. Donabedian (1995) shows that allowing for professional self-control (i.e. self-regulation) as a means of solving the enforcement problem might be a politically feasible alternative to government-imposed punishment in occupations with high, rather than low, specificity of human capital. Ashby at al. (2004) argue that individual firm compliance generates both private and public (to the industry as a whole) benefits. Self-regulation is then sustainable in industries that are more likely to overcome the free-rider problem and where governmental tolerance towards noncompliance is sufficiently high.
48
Gehrig and Jost (1995) is to the best of our knowledge the only account in the literature that evaluates the welfare properties of producer self-regulation by explicitly contrasting it with direct government regulation. (In his analysis, Leland (1979) implicitly assumes that, when not set by the profession itself, the minimum quality standards are set by a benevolent regulatory agency.) Gehrig and Jost (1995) consider a set-up with asymmetric information between consumers and producers about product quality, and introduce mobility of
46
literature by investigating a set-up inspired by property rights theory (Grossman and Hart 1986). Although property rights theory has primarily focused on the optimal allocation of control rights within firms49, we argue that it provides a natural framework to address the question of the allocation of regulatory authority. A regulator’s residual right to set rules with the power of law confers authority much as ownership rights over assets do when contracts are incomplete. A policy-relevant theory of regulatory institutions has to acknowledge the difficulties in writing complete regulatory contracts. Mostly, these contracts are confined to assign authority over specific regulatory decisions, only roughly specifying the rights and obligations of the regulators (Goldberg 1976, Williamson 1976, Estache and Martimort 1999).50 The distribution of regulatory rights then critically shapes the regulatory outcomes. This chapter contrasts the implications of two different assignments of regulatory rights. We refer to government regulation as a regime where the authority over implementation of regulatory decisions is centralized at the governmental level, and regulatory powers assigned to governmental regulatory agencies or ministerial
consumers across markets, thereby allowing for reputation effects. In their model, producers have an incentive to self-regulate by jointly committing to high quality in environments where consumer mobility is high relative to industry monitoring costs and market fragmentation. Self-regulation, if viable, welfare dominates government regulation only if the firms' profits are assigned a sufficiently high weight in social welfare. Gehrig and Jost show that without the firms' informational advantage, government regulation could always mimic self-regulatory actions, at least when government's monitoring costs are ignored.
49
Luelfesmann (2000), an exception, applies the Grossman and Hart (1986) framework to investigate optimal governance structure in the context of fiscal federalism.
50
"At the risk of oversimplification, regulation may be described contractually as a highly incomplete form of long-term contracting…" (Williamson 1976, p. 91). Yet interestingly, the theory of incomplete contracts, first explored by Grossman and Hart (1986), has not had much impact on the study of regulation. See Lyon and Huang (2002) for an exception and discussion.
47
departments.51 Alternatively, the regulatory authority can be delegated to producers ? industries or professions, represented by self-regulatory organizations ? whose activities are to be regulated.52 We call such regimes self-regulatory. We focus on examining the implications of different allocations of regulatory authority, and therefore abstract from incentive and accountability problems due to the complexity of internal governmental organization, as well as potential agency problems between self-regulatory organizations and their members.53 Methodologically, this chapter falls within a growing body of literature on comparative analysis of institutional arrangements (see Djankov et al. 2003), which has raised, or freshly illuminated, many important questions in the economics of institutions in the last decade.54 Complex economy-wide institutional structures are not easily examined with analytical models derived from first principles. Therefore, this literature is characterized by the use of reduced form models, motivated by disparate historical examples and available data. The chapter is organized as follows. In Section 2.2, we develop a model of a regulatory process with bargaining where the authority to amend the enabling legislation
51
There are important differences in the way state exercises its regulatory powers through a central government department and a regulatory agency that we neglect in our analysis. As opposed to a ministerial department, regulatory agencies are by a parent statute typically independent from the central government, yet nevertheless under close political oversight. Through the power of key appointments and overall direction of regulatory agencies, the central government maintains a significant portion of its regulatory authority. See Chapter 5 in Baldwin and Cave (1999) for a thorough discussion. We ignore potentially important institutional details of various self-regulatory arrangements. See e.g. Ayres and Braithwaite (1992) for discussion.
52
53
See, e.g., Estache and Martimort (1999) and Tirole (1994) for discussion of organization of government. See Nunez (2001) and DeMarzo at al. (2001) for analysis of agency problems between an SRO and its members. See, for example, Shleifer and Vishny (1994), Glaeser and Shleifer (2002, 2003), Glaeser, Shleifer and Johnson (2001), Che and Qian (1998), Qian (1994), Greif (1993), Roland (2000), Acemoglu and Robinson (2000), Dewatripont and Maskin (1995), Kranton and Swamy (1999), and many others.
54
48
can be either consolidated within the government or extended to the producers in a selfregulatory regime. In Section 2.3 we delineate the welfare implications of regulatory regime choice, and indicate whether the government’s incentives to delegate or centralize regulatory authority lead to efficient institutional design. Section 2.4 illustrates the model with a range of examples: we motivate the changes in regulatory practice between the progressive era and the associational regime of the New Deal, characterize the mechanisms of intervention used in fascist economies, and identify those features of legal traditions that help to explain variation in regulatory arrangements across countries. Section 2.5 concludes. 2.2 A Simple Model 2.2.1 The Setup The government and regulators are implementing rules to regulate an economy consisting of producers and consumers (the latter standing for the general public). We assume some form of regulation is desirable on economic grounds. The final set of rules results from a process in which the enabling legislation for the particular regulatory arrangement is subject to refinement and interpretation as implementation proceeds. Additional, typically administrative, rules formulated during implementation will have the power of law. We assume that the legal rules adopted in the regulatory process can be characterized through their effect on a variable, L, defined on the real line. L could be, for example, an index of legal complexity, with higher values implying greater legal complexity.55 Under the status quo, L = 0. The desirability of regulation follows from
55
We implicitly assume there exists a mapping of legal rules (e.g. minimum educational requirements for professions) into values of the index of legal complexity, L.
49
assumptions given below, which ensure that some positive values of L are preferred by everyone in the economy to L = 0.56 At t = 1, when there is a degree of consensus that a problem needs to be solved, the government passes the skeletal (or enabling) regulatory legislation. We assume the latter results in L = L*. The enabling legislation will normally be rather vague about the specifics of attaining regulatory goals (Eisner 2000, pp. 13?5) and will be passed with an understanding that the sketchy legal rules will change during implementation.57 There are many reasons why such changes happen. Some occur because the very nature and complexity of the law-making process makes the outcome of legislation unpredictable.58 In addition, shifts in the politico-economic environment, perhaps causing institutional transformations, stimulate revisions and reinterpretations of existing laws.59 Whatever the specific reason for adjustments, they are an inevitable product of the fact that enabling legislation is formulated when there is much uncertainty about what its precise effects will be.
56
For example, lawyering in many republics of the former Soviet Union during the early transition was a free for all: legislation governing the newly established relationships between market participants was yet to be implemented (Waters 2004). This situation that was clearly undesirable for both lawyers and the general public. For example, the statutory framework for a profession might assert some minimum degree of training to be acquired, or environmental legislation might specify broad guidelines about adoption of environmentally friendly processes, or a law for financial markets might stress the need for some level of information disclosure. Since regulatory policies originate in legislation that defines policy goals in the broadest possible terms, it is the regulators that must give regulatory policies substantive content. An original aim might not be practical (e.g. new professional standards have to piggy-back on old elements of the existing educational system), or a previously hidden conflict with other legislation becomes apparent (e.g. financial disclosure requirements come into conflict with privacy legislation), or practical issues in measurement induce a different specification (e.g. the measurement of pollution targets depends on existing scientific equipment).
57
58
During the post-communist transition, massive institutional reforms were conducted, often with a trialand-error approach. The skeletal legislation necessarily had to be amended with supplementary rules. Implementation of rules regulating organized legal professions in Georgia, for example, involved a lengthy process of reformulations and adjustments of the newly established body of laws (Waters 2004).
59
50
We capture the passage from initial legislation to a set of regulatory rules in the simplest manner. If no further decisions were made in the regulatory process, the L implied by the enabling regulatory legislation would equal L = L*+? at t = 2, with random variable ? capturing all the uncertainties involved in institutional implementation. We assume that ? has mean 0 and variance ?2. While ? is unknown at t = 1, it is known by all for t ? 2. Therefore, the passage of the enabling legislation represented by L = L*, results inevitably, even if no other decisions are made, in a different L (= L*+?), which has different effects than would L = L*.60 The variance ?2 is a central parameter in our model. It is a measure of the degree of uncertainty of the effects of the enabling legislation at the time of passage of that legislation. It will be higher, for example, in those countries that are changing related institutions the quickest (where the structure of the future economy is harder to predict), in those sectors where regulation has not been undertaken before (where the enabling legislation is not simply the reform of existing institutions), and in those countries where law-making is decentralized and subject to interpretation (rather than centralized and under greater control of the designer). The effects of the uncertainty can be mitigated. At t = 3, regulators can choose to implement supplementary legal rules, which have the result of changing the effect that L = L*+? has on the economy. We call the regime where the government has the right to
60
Capturing this spirit of uncertainty and its resolution associated with the process of any institutional implementation, Dewatripont and Roland (1995), for example, analyze the outcomes of different reform strategies for economies in transition. In the context of the Georgian lawyers example examined by Waters (2004), L = L* could be interpreted as the legal complexity arising from provisions about some minimum level of quality of services specified by a governmental statute. As implementation of the enabling legislation proceeds, however, many previously unknown but relevant details pertaining to regulation of organized legal professions become apparent, altering the legal complexity implied by (and hence effects of) the skeletal regulatory legislation. At t = 2, therefore, L = L* results in L = L*+?.
51
set those supplementary legal rules government regulation (R). In contrast selfregulation (SR) is the regime where the authority over the supplementary legal rules is vested in bodies whose members are chosen by the producers themselves. As before for ease of modeling, we adopt a particularly simple form for this process: the supplementary legal rules directly modify L = L*+?, so that after their implementation L = L*+?+?L. ?L hence represents the change in L (from L*+? to L*+?+?L) resulting from adoption of supplementary legal rules.61 Any adjustment ?L of L from L = L*+? is costly to implement. These adjustment costs arise from the expense of establishing additional rules and institutions and complying with additional rules, as well as the implicit costs arising from the reallocation of resources. Aside from assigning authority for choosing ?L to different parties, government regulation and self-regulation differ precisely in the adjustment costs that they impose on the economy. Many reasons have been put forward why the cost of implementing additional legal rules, and adjusting existing ones, is greater under government regulation than under selfregulation.62 Compared to the government, the producers typically command a greater degree of expertise and technical knowledge of practices and innovatory possibilities within the relevant area. Self-regulatory arrangements are less formalized than public regulatory regimes and hence less rigid (public agencies often have to go through a long bureaucratic process before an old rule can be changed or a new one enacted).
For the Georgian lawyers example, once ? is known, at t = 3 supplementary legal rules, such as licensing and certification standards, and admission requirements, are chosen by either a governmental regulatory body or the self-governing profession. ?L is then the change in legal complexity from L = L*+? to L = L*+?+?L arising from implementation of these supplementary legal rules.
62
61
For a comprehensive discussion, see Baldwin and Cave (1999), Ogus (1999), and Priest (1997).
52
Information costs for the formulation and interpretation of new standards are therefore lower under self-regulation. Monitoring and enforcement costs are also reduced under self-regulation, as are the costs to the regulated of dealing with regulators.63 We build this into the model as follows. L2 is the cost to the economy (producers and consumers together) of implementing the enabling legislation amended by supplementary rules (L = L*+?+?L). In addition, amending enabling legislation per se is costly. The size of adjustment costs, however, differs under different regulatory arrangements. The adjustment ?L of L from L*+? to L*+?+?L costs the economy ?i(?L)2, i?{R,SR}, with ?R>?SR>0. ?i>0 then measures the rigidity of a regulatory regime. The total costs of L = L*+?+?L for the economy are therefore L2+?i(?L)2 (all incurred after ?L is determined).64 The producers' and consumers' preferred L's will typically not coincide: we assume that the groups differ only in how much they value L, not in terms of the costs that they incur through its implementation.65 For ease of modeling, we let the agents' net payoffs be quadratic. L = L*+?+?L then yields a payoff of pL?½[L2+?i(?L)2] to producers and cL?½[L2+?i(?L)2] to consumers. p > 0 and c > 0, respectively, represent
63
Unlike regulation governed by public authorities and dependent on tax income, self-regulation also does not cause a deadweight loss from taxation.
64
That all benefits and costs of a legal rule are realized at t = 4 only is merely a consequence of the review and deliberation to which new legal rules are subject in a normal political process. This implies that the producers who typically "pay" the adjustment and implementation costs can transfer ½ of these total costs onto consumers. This assumption certainly limits the scope of our analysis by abstracting from important effects of industrial structure. On the other hand, our analysis is primarily concerned with regulatory regimes (as opposed to more narrowly defined regulatory policies) that transcend the specific problem of a given industry (see, e.g., Eisner 2000, p. 3).
65
53
the producers' and consumers' preferred L's in the absence of legal rule adjustments.66 In general, p ? c.67 The difference between p and c captures the divergence of interests between producers and consumers. (p?c)2, for example, will tend to be high in a polarized economy with prosperous owners and a public resenting perceived or real monopoly power, but low when there is a high degree of consensus among the groups about the regulatory agenda. The government's payoff from L is a weighted average of producers' and consumers' payoffs, ?{cL?½[L2+?i(?L)2]}+(1??){pL?½[L2+?i(?L)2]} = AL?½[L2+?i(?L)2], where A ? ?c+(1??)p.68 ? is close to 0 (A is close to p) when the producers’ lobby has bought the government. ? is close to 1 (A is close to c) in the case of a populist government pandering to the economy’s consumers. Under R, it is the government that has the right to set ?L. Under SR, these lawmaking powers are delegated to the producers. But political bargaining is a pervasive feature of the relationship between the government and the industry (see e.g. Shleifer and Vishny 1998, Laffont 2002).69 Therefore, regardless of who has the authority to define supplementary legal rules, the government and producers may negotiate ?L. The
66
Observe also that we implicitly assume that aggregate uncertainty affects the benefits and the costs of producers and consumers in the same direction: increases in L increase both the value from, and the implementation costs of, L = L* for both groups.
67
A profession, for example, would want to limit the entry in the market more than is desirable by consumers (see e.g. Leland 1979), implying that the two groups prefer different L's. We assume that all government-specific costs of legal rule implementation are covered by taxes imposed on producers and consumers. In the post-communist Georgia, for example, the body of laws regulating organized legal professions was implemented through lengthy negotiations between the lawyers' associations and the government (Waters 2004). Political bargain between the government and the producers lobbies is, of course, a pervasive feature of politically less volatile environments as well.
68
69
54
consumers, as the general public, do not directly participate in negotiations (Olson 1971), but may exert pressure on the government (when ? is high). The government and the producers enter negotiations at t = 3. Since the government's and the producers' preferred ?L's implemented under the respective default scenarios differ (unless ? = 0), there are always gains from negotiating away from the default ?L. The bargaining process between the government and the producers results in a Nash-bargaining solution. The government and producers implement their jointly efficient ?L, irrespective of whether the authority over ?L lies with the government (R) or the producers (SR). Each party obtains its regime-dependent default payoff plus one half of the gains that arise from implementation of their jointly efficient ?L (rather than the respective default ?L).70 Figure 7 summarizes the timeline of events. Finally, we discuss our notion of social welfare. Producers and consumers count equally in this economy. We therefore define social welfare under regime i?{R,SR}, Wi, as the sum of the producers’ and consumers’ payoffs from the final realization of L under regime i: Wi = (p+c)Li?[Li +?i(?Li) ]. The transfers implied by the bargaining between the government and the producers are gains for one and losses for the other and hence not part of the social welfare calculation.71
2 2
70
For simplicity, we assume that the government and the industry have equal bargaining strengths. Renegotiation of the default ?L under any allocation of authority to implement amendments will involve exchange of transfers. Under R, for example, the initiative to renegotiate the default ?L will come from producers. In order to set a ?L different from the default ?L under R, however, the government will have to be compensated by producers. These transfers are payoff equivalents of direct monetary bribes or exchanges of political favors. Transfers per se therefore do not cause any resource misallocation due to rent-seeking, for example.
71
55
2.2.2 Solving the Model The model is solved straightforwardly which enables us to omit algebraic details in the presentation below.72 Working backwards, if no bargaining were to occur at t = 3, the default ?L under regime i?{SR,R}, ?Li,d, would equal ri ? L *i + ? 2 1 ? ? = argmax ?L ?ri L *i + ? + ?L ? L *i + ? + ?L + ? i (?L) 2 ? , ?L = 1+ ?i 2 ? ?
i,d
(
)
(
)
[(
)
]
where rSR = p and rR = A.73 Ceteris paribus, the more flexible the regulatory regime, the greater the adjustment ?L from L = L*+?. Since gains from renegotiation of ?L always exist, at t = 3 the government and the producers bargain to implement the jointly efficient ?L under regime i?{R,SR}: A+p ? L *i + ? 2 ?Li,b = 2 = argmax ?L (A + p) L *i + ? + ?L ? L *i + ? + ?L ? ? i (?L) 2 .74 1+ ?i
(
)
{
(
) (
)
}
Anticipating the bargaining outcome, at t = 1 (when ? is unknown) the government chooses L*i to maximize the expected value of the sum of its default payoff at t = 3 under regime i?{SR,R} and half of the gains from negotiation, implying L*i = A ? ?c+(1??)p. In our model, the (ex-ante) impact of the enabling legislation on the economy is independent of the regulatory regime in place, and fully reflects the government's
72 73
See Appendix 5 for derivations of all the expressions and results in the chapter.
The producers’ and the government’s t = 3 default payoffs for i?{R,SR} are then equal to p(L*+?+?Li,d)?½[(L*+?+?Li,d)2+?i(?Li,d)2] and A(L*+?+?Li,d)?½[(L*+?+?Li,d)2+?i(?Li,d)2], respectively.
74
The producers’ and the government’s t = 4 payoffs under bargaining for i?{R,SR}are then equal to p(L*+?+?Li,b)?½[(L*+?+?Li,b)2+?i(?Li,b)2] and A(L*+?+?Li,b)?½[(L*+?+?Li,b)2+?i(?Li,b)2], respectively.
56
weighting of the welfare of the economy's two groups.75 This is consistent with features of the regulatory implementation process. The enabling legislation is passed by the legislature, a fertile ground for the realization of the agendas of interest groups (Eisner 2000, p. 13). However, the skeletal enabling legislation merely opens the regulatory process, the effect of regulatory regime occurs when law is given its substantive content through implementation of additional legal rules. The outcome of the regulatory process is the final realization of L at t = 4. Under regime i?{SR,R}, the latter is a weighted average of A+?, the value of L implied by skeletal regulatory legislation as affected by uncertainty, and ½(A+p), the cooperatively desirable value of L for the government and producers: ? i ( ? + ?) + 1+ ?i ?+p 2 .
With R the more rigid of the two regulatory regimes, R places a relatively higher weight on A+?. Consequently, the expected social welfare76 under regime i?{SR,R}, Wi, equals
75
L*i equals the government's preferred legal rule (in the absence of amendments). The principal reason for this is that the government and the producers in our model care equally about the costs of legal rule adjustment (the ?i(?L)2 term) relative to the costs that arise purely from implementation of L (the L2 term). (This is in turn implied by the assumption that consumers and producers in our model differ only in how much they value L, but not in terms of the costs that they incur because of L.) To illustrate the intuition, take the no-bargaining scenario (or, equivalently, a scenario where one party possesses all the bargaining power). If the government cared about the adjustment (relative to pure implementation) costs more than the producers do, for example, then, under R, unable to predict ?, the government would have no incentive to set L* away from its preferred value (A). Under SR, in contrast, knowing that the producers have an incentive to adjust L*+? quite a bit toward p, the government would put a positive weight on p when setting L* (rather than simply setting L*=A). To make an analogy with the model of Grossman and Hart (1986), where different ownership structures lead to different distortions in ex-ante investments, in our model the selection of the ex-ante regulatory target embodied in the basic legal rule does not depend on the assignment of regulatory rights.
76
That is, social welfare as expected before the regulatory process begins (and hence before ? is realized) but after regulatory rights have been assigned. We refer to Wi as 'social welfare' in the remainder of the text for brevity. Same applies to the definition of Vi.
57
W i = A(p + c ? A) ?
?i 1 1 ?2 + ? ?(3? ? 2)(p ? c) 2 , 1+ ?i 1+ ?i 4
and the expected payoffs of the government under R and SR, respectively, are equal to
VR = 1 1 1 2 1 ?R A ? ? ?2 + ? ? 2 (p ? c) 2 2 1+ ?R 8 1+ ?R 2 1 2 1 ? SR 3 1 ?2 ? ? ? 2 (p ? c) 2 . A ? ? 2 2 1 + ? SR 8 1 + ? SR
V SR =
While aggregate uncertainty is always detrimental, greater flexibility of the regulatory arrangement in place moderates its impact. Before we examine these expressions we briefly discuss the inefficiencies that arise under the two competing regulatory arrangements. As a benchmark, define the first best (FB) final realization of L as that implemented by a social-welfare maximizing planner possessing all the expertise of the producers in the relevant area of regulation (?FB = ?SR) and able to resist negotiating with producers over ?L.77 Once observing ? at t = 3, the benevolent social planner amends the enabling legislation by choosing ?L. It can be easily shown that ?LFB = [½(p+c)?(L*FB+?)]/[1+?SR], L*FB = ½(p+c), and the final realization of L under first best equals ½(p+c)+[?SR/(1+?SR)]?. Comparing these with the expression for the final realization of L above, it is evident that R cannot implement the first best L primarily because of its rigidity in adapting legal rules, regardless of the value of ?. SR, on the other hand, fails at implementing the first best L because of the active participation of the regulated interests in the legal rule determination process. Explicitly
77
In contrast to the fictitious benevolent social planner, a self-interested government will not abstain from negotiations with the producers: since there are gains from renegotiating ?L, a self-interested government will always agree to renegotiate the default ?L in exchange for a transfer.
58
acknowledging that regulatory contracts are incomplete hence renders regulatory decision-making a second-best world.78 2.3 Comparing Regulatory Arrangements: Efficiency and Government's Incentives When does R attain higher social welfare than SR and vice versa? The following result summarizes the argument: Result 1: Social welfare is higher (lower) under SR than under R if ?2/(p?c)2+¾?2?½? >(<) 0. The socially efficient regulatory regime depends on the parameter values describing the economy. The less rigid SR yields higher social welfare than R when aggregate uncertainty is high and/or the divergence of interests between producers and consumers is low, and/or the government is populist. This corresponds to regions I and II in Figure 8.79 In contrast, when aggregate uncertainty is low, and/or the economy is polarized, and the producer lobby is strong (regions III and IV), social welfare is higher under R than under SR. To illustrate the intuition of this, recall that the final realization of L under regime i?{SR,R} equals [?i(A+?)+(p+A)/2]/[1+?i] and that the social-welfare maximizing value of L when no amendments of enabling legislation are needed, is ½(p+c). It is then readily seen that, given (p?c)2, when ? is high, SR on average keeps the final realization of L relatively closer to ½(p+c) than R for any realization of ?. When ? is lower, on the
78
Note, first, that if R possessed all the flexibility of SR, R would always yield the same social welfare as SR. This result is consistent with Gehrig and Jost (1995): they show that without the firms' informational advantage, government regulation could always mimic self-regulatory actions, at least when government's monitoring costs are ignored. Second, R would fall short of attaining the first best social welfare even if the government possessed all the expertise of producers (?R = ?SR) and weighed the welfare of the economy's groups equally (? = ½): since there are gains from renegotiating ?L, a self-interested government maximizing its payoff will always agree to renegotiate the default ?L in exchange for a transfer. Note that WSR = WR when ?2/(p?c)2+¾?2?½? = 0.
79
59
other hand, the more rigid R on average keeps the final legal rule closer to the socialwelfare maximizing value than SR for low, but not for high, values of ?2.80 While Result 1 indicates the circumstances under which each of the two regulatory regimes implements higher social welfare, it does not establish whether the regime yielding higher social welfare will in fact be chosen by the government at t = 0. We examine this question by investigating the government’s payoffs under R and SR: Result 2: The government’s payoff is greater (smaller) under SR than under R if ?2/(p?c)2?¼[(4+3?R+?SR)/(?R??SR)]?2 >(<) 0. The government's preferred regulatory regime, like the socially efficient regulatory arrangement, depends on the parameter values describing the economy. The minimum level of ?2/(p?c)2 necessary to render SR comparatively more attractive to the government increases with ?. The government would prefer to centralize regulatory authority in regions II and III in Figure 8, and delegate it in regions I and IV.81 To see the intuition, note that the government’s optimal L equals A. Given ?R, ?SR, and (p?c)2, the attractiveness of a given regulatory arrangement for the government depends on a trade-off between ?2 and ??[0,1]. When ? ? 0 (A ? p), the government would prefer to institute the regime that weighs (p+A)/2 ? p relatively more heavily than
When ? ? 1 (A ? c), for example, ½(p+A) ? ½(p+c), so the regime that implements the final legal rule closer to ½(A+p) rather than A+? yields higher social welfare for any value of ?2. When ? ? ? (A ? ?p+?c), ½(p+A) ? (5/6)p+(1/6)c, the more rigid R is more likely to keep the final realization of L relatively closer to ½(p+c) than SR as long as ?2 is small. A direct consequence of Result 1 is that which regulatory regime attains higher social welfare does not depend on the size of ?R??SR>0. Suppose that, in a given institutional setting, higher social welfare is attained under R and that the regulatory authority is (correctly) consolidated within the government. As regulatory practice is carried on, however, suppose further that R becomes more rigid (e.g., due to increasing bureaucratic complexity). Result 1 implies that this should not warrant a reversal in regulatory practice and re-allocation of regulatory authority to the more flexible producers.
81
80
Note that VSR = VR when ?2/(p?c)2?¼[(4+3?R+?SR)/(?R??SR)]?2 = 0.
60
A+? ? p+?: SR is favored even in the case of small aggregate uncertainty. For ? greater than 0 (A closer to c), on the other hand, (p+A)/2 is further away from A than A+? at least for small realizations of ?, thus making R the government's preferred choice. Observe also that the attractiveness of SR (or R) for the government naturally varies with the size of ?R??SR>0.82 Regions II and IV of Figure 8 highlight the conflict between the socially efficient regulatory regime and that compatible with the government’s incentives: in a given institutional setting, the welfare-superior regulatory regime will not be implemented when such implementation is incompatible with the government's incentives. In region IV, a government bought by the producer lobby bestows regulatory powers on its constituency, when direct government regulation would be socially efficient. In region II, in contrast, a populist government, eager to protect the interests of consumers, resorts to direct government regulation, despite the fact that delegation of regulatory powers to producers would increase social welfare. If the distribution of ?2/(p?c)2 across political systems was independent of ?, then the model predicts that a populist government will be more likely to mis-allocate law-making rights at the expense of the society than a government aligned with the producers because region II is larger than region IV.83
Given ?>0, ?SR and ?R, suppose ?2/(p?c)2 is just sufficiently high that the government prefers SR to R. That is, selecting the regime that places a substantially lower weight on A+? as a component of the final legal rule is favored by the government. Notice, though, that ½(A+p) ? A for any ?>0 and that?½(A+p)?A?>0 increases with ?>0. In contrast, A+? ? A for ? ? 0. With ?R sufficiently lower and everything else constant, then, tolerating the increased risk associated with selecting the regime that weighs A+? somewhat (but not substantially) more, pays off to the government: especially for higher values of ?, the final realization of L under lower ?R is more likely to be closer to A if the government centralizes rather than delegates the regulatory authority.
83
82
It can be easily shown that, in Figure 2, (a) the WR = WSR curve attains its maximum at ? = 1/3, and (b) the VR = VSR curve intersects the WR = WSR curve at ? < 1/3.
61
The model also shows that two parameters, both describing the state of the economy, yet of a widely different character, play exactly equivalent roles. Aggregate uncertainty about the effects of enabling legislation (?2) and the degree of homogeneity of affected interests (1/(p?c)2) influence the choices of the socially efficient and the politically desirable regulatory regimes in the same way. Ceteris paribus, this implies that two markedly different societies will make the same choices on regulatory regime, one society being beset by great aggregate uncertainty about the effects of outcomes of institutional changes (extraordinarily high ?2) and the other characterized by a great degree of consensus among affected interests about the regulatory agenda (very small (p?c)2). In both, delegation of regulatory powers to producers is socially efficient and preferred by the government. Given that the former society might be one undergoing revolutionary changes and that the latter could be a peaceable nation where there is little social conflict, the commonality of predictions for the two is certainly not a hypothesis that would be readily generated from intuition, showing the virtues of placing the analysis in a systematic model. The model also bears on Hayek's maxim that the certainty of the law is a chief attribute of good law, and a prerequisite for successful development (Hayek, 1960).84 Our model captures this notion, since uncertainty about the effects of the enabling legislation (?2) reduces social welfare regardless of the regulatory regime in place.85 Yet
84
"[A]…chief attribute which must be required of true laws is that they be known and certain. The importance which the certainty of the law has for the smooth and efficient running of a free society can hardly be exaggerated. There is probably no single factor which has contributed more to the prosperity of the West than the relative certainty of the law which has prevailed here. This is not altered by the fact that complete certainty of the law is an ideal which we must try to approach but which we can never perfectly attain" (Hayek, 1960, p.208). While both WR and WSR decrease in ?2, WR decreases faster.
85
62
the government is able to moderate the uncertainty about the effects of the final realization of L by delegating regulatory powers to the producers.86 The results derived above indicate that the equilibrium regulatory regime choice increases the predictability of the effects of final legal rules when the effects of the enabling legislation are very unpredictable. When effects of the enabling legislation are hardest to predict (when ?2 is large), we show that, in equilibrium, a self-regulatory regime will in fact be chosen by the government. 2.4 Illustrating the Model This section illustrates the applicability of the model, using a range of examples. We show that the model provides explanatory power in identifying factors that might explain differences in the regulatory practices of the progressive era and the New Deal. Our analysis helps in interpreting the choice of mechanisms of intervention used in fascist economies. It also identifies those features of legal traditions that help to explain variation in regulatory arrangements across countries. 2.4.1 Two Modes of Attaining Regulatory Goals: the Progressive Era and the New Deal Two episodes in American history, the progressive era and the New Deal, while both known for widespread regulation of the economy, provide an example of a stark contrast in allocation of regulatory powers. Our model helps to explain why the two periods were characterized by such markedly different regulatory regimes. To set the stage, we first
The variance of the final legal rule under regime i?{R,SR} equals (?i/(1+?i))?2. Because ?SR < ?R, (?SR/(1+?SR))?2 < (?R/(1+?R))?2.
86
63
turn to briefly discussing the regulatory practice under the progressive era and the New Deal.87 In the progressive era, direct government regulation proliferated. A large number of regulatory initiatives led to the creation of new administrative agencies addressing corporate organization and practices in a number of sectors, including rail transportation, finance, food, and pharmaceuticals.88 The progressive policy-makers were committed to preserving markets and a decentralized market structure. Reforms on a grand scale were not advocated: mainstream progressivism was "an attempt to reclaim the market-based sphere of individual liberty and opportunity" (Eisner 2000, p. 47).89 Where the structural features of an industry precluded an antitrust strategy (e.g., in transportation), Congress and the president initiated policies aiming to compensate for the lack of effective market mechanisms by regulating rate-making and other activities. The progressives' faith in expert knowledge and their desire to foster administrative neutrality and efficiency at all levels of government led to consolidation of governmental regulatory authority in the form of regulatory commissions.90 Under the New Deal, in great contrast to the progressive era, regulatory authority was formally extended to the interests being regulated in many sectors of the economy.
87
The sketchy depiction of regulation under the progressive era and New Deal below draws heavily on Eisner (2000).
88
The Sherman Antitrust Act established the foundation of US competition policy; the Hepburn Act dramatically enhanced the powers of the Interstate Commerce Commission; the Meat and Inspection Act and the Pure Food and Drug Act; the Federal Reserve Act creating the Federal Reserve Board and expanding national regulation of finance; the Federal Trade Commission and Clayton Acts strengthening antitrust enforcement. See Eisner (2000, p. 44). Wilson's vision of an order, for example, was based on markets and market-corrective regulation (Eisner 2000, p. 39). Although formally independent from the executive branch, these were unable to withstand the pressures of partisan and interest groups politics, and constraints imposed by the courts. See Chapter 3 in Eisner (2000) for a discussion.
89
90
64
A system of government-supervised self-regulation was established: "[Under the New Deal]…many private economic associations became quasi-public in nature, for they were given public authority and an important role in making and implementing regulatory policy" (Eisner 2000, p. 90). Trade associations and industry groups were authorized to establish codes of conduct that were exempt from antitrust laws.91 The Agricultural Adjustment Act decentralized regulatory decision-making in agriculture relying on farm associations for policy implementation. The Securities and Exchange Commission, established in 1934, played a critical role in giving the financial industry the necessary authority to self-regulate and facilitated the creation of the largest self-regulatory body in the country, the National Association of Securities Dealers. The divergent pattern of regulatory arrangements of the progressive era and the New Deal reflects underlying features of the respective eras that can be interpreted using our model. First, the early progressive era (late 1890s, see e.g. Hofstadter 1972) marked the ending of a period of rapid socio-economic change and the return of stability.92 The progressives did not advocate radical transformations of the economic system. They believed that market-corrective actions and reliance on scientific methods would yield positive outcomes. The New Deal was, in contrast, launched amidst the middle of unprecedented economic failure and widespread questioning of existing institutional
91
"The National Recovery Administration chose business executives…to serve as members of the Industrial Advisory Board. On this board, they assisted in the drafting of codes and the creation of code authorities, the bodies of business and trade association representatives which interpreted code provisions and granted exemptions" (Eisner 2000, p. 85).
In the latter half of the 19th century, the growth of tangible capital stock greatly outpaced the growth of the population. Tremendous waves of immigration accelerated urbanization of society. Economic growth was disrupted by a fairly deep depression that lasted from 1893 to 1896 (see Eisner, 2000). The return of tranquility is conventionally dated to 1897 and the inauguration of the Republican president McKinley: "The disturbing changes that Bryan [a Populist nominee] had promised…would not pass. The frightening prospect of a radical alliance of farmers and workers had collapsed. The emerging industrial order…seemed safe" (McGerr 2003, p. 3).
92
65
arrangements. The necessity for, and the spirit of, institutional experimentation extended the government’s authority into places where it had not been before. The uncertainty about the outcomes of institutional changes, as captured by ?2 in our model, was therefore comparatively larger under the New Deal. Second, the power of corporations grew tremendously during the later decades of the nineteenth century. Yet the whole country did not share in the new wealth, prestige and optimism. The public's dissatisfaction with growing inequalities suggested a growing threat to property rights (Eisner 2000, p. 33). In contrast, during the New Deal, there was a sense of nation-wide consensus about the necessity of coordination and cooperation between the industry and the public. F.D.R. presented the system created by the National Recovery Act as one that would "sink selfish interests and present a solid front against a common peril" (Eisner 2000, p. 84). This implies that the degree of polarization in society, that is, (p?c)2, was comparatively smaller under the New Deal. We argue that the economy-wide ? was unlikely close to 1 or close to 0 in either historical episode.93 On the one hand, elements of the New Deal often openly promoted cartel agreement to insure industrial 'recovery and relief', and similarly there were many episodes of regulatory capture under the progressive era (implying ? could have been close to 0).94 On the other hand, the New Deal responded to the needs of the poor and the unemployed, while many measures of the progressive era, such as reforms of civil service, direct elections of senators, and recall of judges, were of very populist nature
Glaeser and Shleifer (2003) argue that events of the progressive era may be explained by both capture theory and public interest approach. Similarly, Wallis (1987, 1998) shows that the New Deal measures were motivated by both economic and political factors.
94
93
See, e.g. Eisner (2000), Glaeser and Shleifer (2003) and references therein.
66
(implying ? could have been close to 1).95 With the economy-wide ? being neither too high nor too low, Figure 8 predicts that the direct governmental regulation of progressive era (placed in area III) and self-regulatory associational regime of the New Deal (placed in area I) can be viewed as efficient institutional responses to the varying historical circumstances. 2.4.2 The Fascist Economic System: Self-Regulation in a Totalitarian Regime Analysis of the structure of the fascist economic system has drawn significantly less attention than the ideology that underpinned the historical events surrounding it.96 One reason for this is that fascism, as an economic system, does not fit any simple categorization, with its admixture of markets and controls, sometimes looking more capitalist, sometimes having elements of socialism (Milward 1976, p. 379). Our model can help to interpret the fascist system97, by viewing self-regulation as one practical mechanism to implement extensive governmental controls. We argue this by first turning to a brief characterization of the fascist economy. At the center of fascist philosophy was glorification of the nation state, in which the pursuit of economic power was highly instrumental. Maximizing production was crucial, and the individual, as consumer, was given little priority in welfare. Prominence was ascribed to industry and to the workers as producers: "...the interests of production are the interests of the Nation...The fascist State is primarily a State of 'producers'" (Florinsky 1938, p. 90). To pursue economic activity that would enhance the nation state, a high
95 96
See Glaeser and Shleifer (2003).
At their inception, however, economic claims of Italian Fascism and German Nazism did prove a subject of compelling interest (Maier 1987, p. 70).
97
While there clearly were differences in the fascist systems of Italy and Germany, Maier (1987), for example, argues that the many common traits of the two regimes warrant an encompassing analysis.
67
degree of intervention into the market was deemed necessary. What really distinguished fascism from free market capitalism was the fascist view that pervasive intervention was crucial in promoting the national good (Sternhell 1976, p. 357). The need for large-scale intervention accepted, there follows the question of what practical mode of implementation to use. In Italy, but also in Germany, the answer lay in the intimate involvement of representatives of businesses in governmental economic decisions. Private associations of businessmen were integrated into state administration and "…transformed into instruments of economic control. But whereas in both countries the titles of these organizations changed, the personnel at the head of them did not... The line between government and business was blurred because the personnel became in many cases one and the same" (Milward 1976, p. 396?7).98 In Italy, as one theorist of fascism declared, "Fascism...would now organize the production of wealth for national ends through the self-government of the categories under the high control of the State" (Field 1968, p. 185).99 Political, and particularly fascist party, control was imposed at the highest levels, but within these constraints business had a degree of autonomy to regulate itself, given its control over both information and micro-decisions (Milward 1976, p. 397; Lyttelton 1976, p. 141). In Italy, the corporations created by Mussolini, but dominated by business, were involved in detailed regulation of prices, production quotas, import suppliers, and myriad other economic variables (Field 1968, pp. 61, 181?95).100 The fascist economy
98 99
See also Lyttelton (1976). See also De Grand (1995, pp. 40?1).
A fascist corporation (la corporazione) was not a corporation in a legal sense of the English word, but a union of all elements involved in a particular economic activity (employers, experts, workers), together with representatives of general interests of the nation and the consuming public. Corporations were first
100
68
can therefore be characterized as one with massive governmental intervention into the market-place, in pursuit of glorification of the nation state, which was implemented through a process of self-regulation.101 Our model is suggestive of why the fascist system might not have been as incongruous as might seem on first inspection. The interpretation rests upon several basic facts about fascism and the world in which it emerged. First, capitalism was not rejected, but markets were to be subject to extensive intervention (i.e., either government regulation or self-regulation). Second, fascist regimes came to power in times of crisis.102 Moreover, they did so with the purpose of promoting truly revolutionary changes in the institutional structures of their countries (Linz 1976, p. 13). In terms of our model, ?2 was extraordinarily high. Third, as noted above, the fascist state emphasized production and the emphasis on production soon became an emphasis on helping producers.103 ? was therefore low. Finally, the state and the fascist party certainly exerted much influence on industry, but this occurred through bargaining with the self-regulators (Florinsky 1938, p. 97), as envisaged in our model.
authorized under the syndical legislation of 1926. A law on corporations was passed in February 1934, establishing 22 corporations with jurisdictions based on the notion of a product cycle. The corporations had the authority to issue rules for the ‘collective regulation of economic relations and for the unitary discipline of production’, subject to the consent of the Head of the Government and after a proposal by one syndical association. The first corporation to produce a regulation legally enforceable under the provisions of the law was that of Beets and Sugar. See Field (1968).
101 102
Mussolini himself used the term 'self-discipline' (Florinsky 1938, p. 97).
Both Italy and Germany were in search of a path out of the recession. Maier (1987, p. 104) argues that the Italian Fascists had to steer the economy through the worst shoals of the World Depression; the Nazis, however, were luckier and could ride the wave of recovery.
103
Even the confederation representing artists and liberal professions was requested by the Charter 'to promote the interests of art, science and letters with a view to improving production' (see Florinsky 1938, p. 90). One of the earliest acts of fascist regimes was to ban labor organizations. See De Grand (1995, pp. 40?1).
69
With extraordinarily high ?2 and low ?, fascism would therefore correspond to a point in area I in Figure 8. Self-regulation would have been an appropriate choice from a social welfare point of view.104 Hence, in contrast to contemporary observers, who saw great inefficiencies in the economic structures of fascism (e.g., Florinsky 1938, pp. 93?4), our model hints that they were not as ill designed as is it is easy to assume, suggesting why fascist economies had some staying power in a world beset by recession. 2.4.3 Legal Origin and Regulatory Regime Choice The differences between civil and common law systems have been a central area of concern in writing on institutions in the last decade (La Porta et al. 1997, Beck et al. 2003). The properties of these two systems have important implications for the constructs that are central in our model: in the predictability of institutional construction emanating from the law-making process and in the ease of decentralized adjustments in institutions. These are both factors that feed into the choice of regulatory regime. In the paragraphs below, we use our model to show how that choice is affected by the nature of the legal system. Then we review some facts about variations in regulatory regimes across countries, examining the fit between these facts and our model's predictions. The first important difference arises from the fact that under civil law regimes, law emanates directly from the center, with judges restricted to using the statutes themselves as primary input into interpretation. Under common law, by contrast, judges have a considerable degree of power to interpret, using case law, a large part of which cannot be modified by the writers of a statute creating an institution (Zweigert and Kötz
Observe that none of our arguments depend on the value of (p?c)2 in a fascist economy, as long as (p?c)2 was not extraordinarily high. The size of (p?c)2 may proxy for the nation-wide sentiments about fascist ideology. Generalizations about such sentiments would be controversial enough as to be immediately unproductive.
104
70
1992, pp. 273-278). Hence, ceteris paribus, the institutional outcomes resultant from a statute are harder for the institution-designer to predict under common law than under civil law. There is less uncertainty about the effects of the skeletal regulatory legislation, at the time of passage, under civil law. In the model’s notation, ?2e > ?2f where subscripts e and f denote common law and civil law, respectively (evoking England and France, which provide the archetypes of the two systems). Second, legal traditions differ in their ability to reshape rules to fit changing conditions and to facilitate adjustments in existing institutions. Legal systems that embrace case law and judicial discretion tend to exhibit a greater degree of adaptability (Zweigert and Kötz 1992).105 Rejecting jurisprudence and relying on changes in statutory law renders the adjustment of legal rules costlier under civil than under common law, regardless of the regulatory arrangement: ?i,f > ?i,e for all i?{R,SR}. Additionally, the inherently centralized character of civil law implies that centralized institutions, such as governmental regulation, are comparatively less costly for a country with that tradition: centralized institutions work comparatively better in a centralized system. Hence, centralization of regulatory authority naturally leads to a greater increase in the cost of legal rule adjustment under the inherently more decentralized common law: ?R,e ? ?SR,e ? ?R,f ? ?SR,f.106
Beck et al. (2003) provide an extended discussion of the relative flexibility of civil and common law systems. Two historical examples are suggestive. The increasing centralizations in seventeenth century England and France produced revolutions in the former and European dominance for the latter. In England, the attempts by Charles I and James II to centralize power in the monarchy led to their deposition and Oliver Cromwell's relatively centralized interregnum failed. In contrast, in France Louis XIV's reign was known for its absolutism and for the political and cultural dominance of France in Europe. Turning to later times, the economic performance of France was surprisingly strong in the period immediately following the Second World War, when the zeitgeist was infected with the centralizing tendencies of socialism and planning, while the UK's economic performance was disappointing (Shonfield 1974). By contrast, as
106
105
71
These intrinsic features of the legal traditions suggest two reasons why there might be a greater tendency for government to choose to centralize, rather than delegate, regulatory authority under civil law than under common law. (Figures 9(a) and 9(b) graphically illustrate Result 3.107) Result 3: (a) If ?2f < ?2e, then {?: VR(?2e) ? VSR(?2e)} ? {?: VR(?2f) ? VSR(?2f)}. (b) If ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0, then {(?,?2/(p?c)2): VR(?R,e) ? VSR(?SR,e)} ? {(?,?2/(p?c)2): VR(?R,f)? VSR(?SR,f)}.108 The essence of the predictions in Result 3 is that self-regulation will occur more often in common-law systems than under civil law. We have sought data to test this prediction but, given the lack of systematic study of regulation versus self-regulation, only snippets exist. At the very least, the conclusions of sociologists' studies of professions are consistent with Result 3: in civil-law countries, the state has played a much more important role in the licensing and regulation of professions. In common-law countries, professionalization has, almost invariably, been practitioner-led: their associations have sought and usually obtained state authority to self-regulate (Burrage and Torstendahl 1990).109
liberalization and privatization became popular in the 1980's and 1990's, the UK has grown faster than its continental neighbors.
107
For notational simplicity, we suppress all of the arguments of Vi except for those crucial to an understanding of the result.
?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0 is in fact a stronger condition than necessary. Note that ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0 if and only if ?SR,f??SR,e ? ?R,f??R,e > 0, which implies ?SR,f??SR,e > (?R,f??R,e)[(1+?SR,e)/(1+?R,e)] > 0. The latter is a necessary and sufficient condition for {(?,?2/(p?c)2): VR(?R,e) ? VSR(?SR,e)} ? {(?,?2/(p?c)2): VR(?R,f) ? VSR(?SR,f)}. See Appendix 5 for proof of this point. Scattered evidence can be found, primarily on the legal and accounting professions. According to a systematic cross-country study by Global Accounting Education (2004): "In general professional bodies in common law countries show a tendency towards professional self-regulation compared with a tendency towards government regulation in civil law countries. This is particularly true for common law countries with a historical UK background" (Global Accounting Education, 2004, http://www.eiasm.be/gae/default.asp?id=25, accessed 9/14/2004).
109
108
72
We have found one area where there has been systematic data collection on which types of regulatory arrangements have been implemented: alcohol beverage advertising. In 1996 the Centre for Information on Beverage Alcohol collected information on how 119 countries regulated alcohol beverage advertising (International Center for Alcohol Policies 2001). The categories were self-regulation, statutory legislation (that is, central regulation), and a combination of both.110 Table 1 matches the data on regulatory arrangements with information on legal origin obtained from La Porta et al. (1999), who distinguish between English, German, French, Scandinavian, and socialist legal traditions. As the Table indicates, there are indeed systematic differences among countries with different legal origins: consistent with Result 3, English legal origin countries have a greater tendency for self-regulation than do countries with a civil law tradition. To test the effect of legal origin, we used alcohol-advertising regulatory regime as the dependent variable in an ordered probit, with the lowest category being central regulation, followed by the combination of both types of regulation, and self-regulation being the highest category. Since it is likely that choice of regulatory regime varies with level of development, and since level of development is correlated with legal origin (see e.g. La Porta et al. 1998), we add log of GNI per capita to the equation. Since countries with larger populations exhibit different regulatory behavior (Mulligan and Shleifer 2004) we also added log of population (see Table 1). The results of the ordered probit appear in Table 2. In Table 2, the omitted legal origin dummy is the French one. Countries of English legal origin clearly use more self-
110
Other categories, such as a complete ban or no relevant policies, are irrelevant here.
73
regulation than countries from the civil law system.111 It must be emphasized that we only view these results as suggestive: the dependent variable only reflects a relatively specialized and narrow area of regulatory activity; there are undoubtedly omitted variables in the regression; and we have not paid any attention to the possible endogeneity of GNI per capita.112 Nevertheless, these suggestive results do offer clear evidence in favor of Result 3. 2.5 Conclusion Allocation of regulatory rights matters because regulatory contracts are incomplete. The source of potential welfare gains in delegating rule-making powers to producers is in their greater adaptability to changing institutional conditions. The source of potential welfare losses is in the active participation of the regulated in the regulatory process. Therefore, whether rule-making authority should be extended to producers, or consolidated within the government, will depend on country-specific factors. When either there is a large level of uncertainty surrounding the results of institutional construction, or the degree of divergence of producers' and the public’s interests is small, the benefits of delegating regulatory powers outweigh the costs. In the opposite case, the regulatory arrangement that yields higher social welfare depends on the extent to which the government's motives are populist or aligned with those of producers.
As an aside, note the negative and statistically significant coefficient on the dummy for socialist legal origin (which may be interpreted as a transition-country dummy): countries of socialist legal tradition resort even less to self-regulation, and even more to central regulation, than countries of French legal tradition. Given conventionally accepted facts characterizing the transition experience, our analytical model seems consistent with this empirical evidence. The populist politics of transition suggests that ? was not low. Highly polarized economies and unprecedented institutional changes imply that both ?2 and (p?c)2 were large during early transition. ?2/(p?c)2 was therefore neither too high nor too low. Our analytical model implies that government controlled regulation would have been a natural political choice, yet socially suboptimal (area II in Figure 1).
112
111
GNI per capita is lagged by five years to reduce this problem.
74
Our analysis highlights the conflict between the socially efficient regulatory arrangement and the arrangement compatible with the government’s incentives. Ultimately, the choice between centralization and delegation of regulatory authority is that of the ruling administration. A socially efficient regulatory regime will be implemented only if such implementation is in the government’s interest. In this light, we argue that the widely different institutional frameworks of regulatory practice during the progressive era and the New Deal, and the peculiar fascist system of organizing production by self-government of industries under control of the state, were not only compelling equilibria, but also relatively socially efficient. We show that regulatory regime choice is also influenced by differences in legal traditions. These differences have been a central area of concern in writing on institutions the last decade. We further this discussion by identifying those intrinsic features of common law and civil law that help to explain variation in regulatory arrangements across countries.
75
Chapter 3 Fostering Civil Society: Towards an Analytical Framework
3.1 A Concept in Vogue The notion of civil society has long been widely debated in political and sociological theory and philosophy (see e.g., Cohen and Arato 1994; Hall 1995; Ehrenberg 1999), culminating in an extraordinarily voluminous body of literature. The concept has more recently drawn much consideration in policy circles' debate on development.113 Many channels through which civil society involvement may foster social and economic progress have been discussed, as argued below. This chapter focuses on civil society as a bulwark against the state. As an associational realm separate from the state, civil society brings a society's concerns to broader public attention, thereby strengthening institutions for, and contributing to mechanisms of, good governance (Salomon et al. 2004). It is due to this proposition that fostering civil society per se – rather than guided by a specific goal of service provision, for example ? has become means of promoting good governance and economic development in post-communist and developing countries. Yet the quest for promoting democratic governance and development via fostering civil society in countries of Central and Eastern Europe (CEE) and the former Soviet Union (FSU) has proven elusive (see e.g. Carothers 1999; Ottaway and Carothers 2000). In this chapter, we attempt to illuminate the recent experience with civil society aid for promoting good governance in the context of a simple analytical model.
See, e.g., the Civil Society and Governance Programme at the Institute of Development Studies, University of Sussex, (http://www.ids.ac.uk/ids/civsoc/), the World Bank's Civil Society program (http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/CSO/0,,pagePK:220469~theSitePK:228717,00.html), the LSE's Centre for Civil Society (http://www.lse.ac.uk/collections/CCS/), the International Center for Not-for-Profit Law (http://www.icnl.org/).
113
76
The rest of this chapter is organized as follows. In section 3.2 we elaborate on the notion of civil society. In section 3.3 we discuss the reasoning that guided civil society aid as part of the democracy aid project. We also review the economics literature of pertinence to analyzing civil society. In section 3.4 we present the scant available evidence on success of civil society aid. In section 3.5 we attempt to illuminate the donors’ experience with a simple analytical model. Section 3.6 concludes. 3.2 Delineating Civil Society The boundaries of civil society are subject of an ongoing debate in social sciences (see e.g. Hall 1995). Most commonly, civil society has been confined to "an intermediate associational realm between state and family", consisting of organizations that are autonomous from the state, and are formed voluntarily by members of the society to protect or advance their interests or values (Ottaway and Carrothers 2000, p.9).114 A majority of the literature asserts that political parties and other groups that explicitly seek to gain control of the state are not part of civil society (Carothers and Ottaway 2000). Likewise, features such as existence of free press and individual forms of citizen action (such as voting and writing to legislators) sometimes enter the discussion on civil society, but are typically excluded from its definition (Salomon et al. 2004). A pragmatic characterization of civil society (see e.g. Salomon et al. 2004; Quigley 2000) divides organizations into those performing essentially expressive
A recent comprehensive study by The Johns Hopkins Institute for Policy Studies (Salomon et al. 2004) has defined civil society organizations (CSOs) as those that are (i) organized, which includes informal as well as registered groups, (ii) private, although they may receive support from governmental sources, (iii) not profit-distributing - while organizations may generate surplus, any such surplus must be reinvested in the objectives of organization, and (iv) selfgoverning. Such definition encompasses a multitude of formal and informal organizations, religious and secular, with paid staff and those staffed by volunteers (e.g., hospitals, universities, social clubs, professional organizations, day care centers, grass-root development organizations, health clinics, environmental groups, religious congregations, family counseling agencies, self-help groups, religious congregations, sports clubs, job training centers, human rights organizations, community associations etc.) See Salomon et al. (2004) for details.
114
77
functions (advocacy, community organizing, representation of interests, cultural expression, environmental protection, human rights, religion, and political expression), and those performing essentially service functions (e.g. provision of health, education and research, social services, development and housing).115 3.3 Promoting Civil Society: the Democracy Aid Project116 Civil society aid is best understood as a component of an ambitious project of "democracy promotion" ? broadly speaking, aid designated for promotion of good governance (see e.g. Carothers 2004), channeled to post-communist (as well as developing) countries. Civil society aid was not a major component of financial aid to CEE and FSU for democracy promotion in the early 1990s.117 Early on, the donors typically concentrated on providing support for administering elections. Democratic elections, however, were only a small step towards securing democratic governance and rule of law. In the second phase, the donors extended the scope of their assistance to pursue reform of major state institutions (especially legislature and judiciary). The task to reform state institutions, however, was much more daunting than anticipated. There was a lack of will to reform among both incumbent political elites, whose power was directly threatened, and the personnel at lower levels of bureaucracies, content with status quo. Likewise, the scale of the project was excessive given available funds.118 It is only in the third phase, in the mid 1990s, that donors' attention has shifted to strengthening civil society. Yet, if
115 116 117
Some CSOs may of course combine both functions. The following sketch draws heavily on Carothers and Ottaway 2000.
This aid is a narrow subset of Western aid to CSOs generally. Likewise, the aid's hoped for effects include only a fraction of the total contribution of CSOs in developing and transition countries.
118 As Carothers and Ottaway put it, "[A] comprehensive program of judicial reform …was far more than any donor could think of financing in any country" (Carothers and Ottaway 2000, p. 8).
78
strengthening civil society was to promote (economic and social) development, who exactly were the donors to aid, and for what purpose? 3.3.1 Civil Society and Development Three channels through which civil society may affect economic development have been proposed. First, civil society organizations serve as vehicles for service provision under poorly functioning states, often to alleviate poverty (see e.g., World Bank 2003). Second, through community building, civil society organizations create social capital (Putnam 1993) ? bonds of trust and reciprocity deemed crucial for institutions of market economy to function effectively. Third, as an associational realm, civil society brings society's concerns to broader public attention, thereby strengthening institutions for good governance. While all three channels provide for a fruitful research agenda in social sciences, including economics, we argue below that the donors’ reasoning for providing aid was most heavily influenced by one of the three. Civil society organizations as vehicles for service provision Organizations of civil society have been widely acknowledged as important vehicles for service provision in developing countries (World Bank 2003). Serviceoriented CSOs are large recipients of foreign aid. The economics literature has provided some guidance on desirability of service provision through CSOs, but much work remains to be done (see Appendix 6 for a review of this literature and discussion). Unsurprisingly, the capacity of directly promoting good (democratic) governance, however, was not seen as a comparative advantage of service-oriented CSOs. Rather, it was another subset of CSOs that donors identified as targets of aid for promoting good
79
(democratic) governance. Before we turn to that, however, we briefly mention another channel through which civil society influences development. Civil society as a breeding ground of social capital Robert Putnam's analysis of development of Italian regions (Putnam 1993) posed a connection between forms of social capital ? norms and networks within civil community, manifested through participation in groups and associations of civil society – and quality of governance and economic development.119 The economics profession has only in the last decade expressed greater interest in formation and effects of social capital. This line of research now includes contributions at both empirical and theoretical level of analysis (e.g., Knack and Keefer 1997; Temple and Johnson 1998; Glaeser et al. 2000; Guiso et al. 2004; Huang 2003). During the 1990s, when decisions about promoting civil society were made, however, the social capital building view of civil society did not play a role. As Ottaway and Carothers put it succinctly, "with a limited amount of money, the need to produce visible results in a short time, and a critical press and an even more critical [legislature] ever alert to denounce waste of taxpayer money, [the donors] could hardly get in the business of setting up bowling leagues…" (Carothers and Ottaway 2000, p.10).120 Civil society as a bulwark against the state Given the donors' aim of promoting rule of law and good governance, policyoriented CSOs (a subset of CSOs providing expressive functions rather than service
119 120
Another influential contribution is Fukuyama (1995).
The bowling leagues example comes from Putnam's article "Bowling Alone: America’s Declining Social Capital", published in the Journal of Democracy in 1995.
80
functions, according to our above taxonomy) emerged as natural candidates.121 Moreover, the donors chose to exclude political parties from consideration. The belief was that is it possible to obtain results without becoming interfering unduly in the domestic politics of another country (Carothers and Ottaway 2000). The donors hence decided to focus on voluntary associations with capacity to directly foster democracy and promote democratic consolidation.122 These are associations that specifically seek interaction with the state, whether to advocate interests of the citizens, to oppose nondemocratic behavior of the state, or to hold states accountable to citizens for their actions. These associations, however, do not explicitly compete for the office. Civil society aid was not seen as a panacea for the intricate task of good governance promotion. It was expected, however, that it would make a difference. In the context of post-communist transition, for example, civil society was "…deemed crucial to stimulating the public pressure and participation necessary to force poorly functioning state institutions to become more responsive and accountable…." A vibrant civil society was "…held out as necessary to cushion the effects of restructuring, to ensure public understanding and support for market reforms, to prevent privatization from lapsing into cronyism, to connect newly empowered local governments to citizens, and to create many of the other components necessary to the transition to market economies" (Carothers 1999, p.115).
121
Sports clubs, cultural associations, religious organizations and less formalized social networks were therefore absent from most civil society assistance programs. Media and trade unions were also excluded from the target group.
122
It must be emphasized that more recently, the donors have expanded the set of civil society aid recipients to include service-oriented NGOs as well. This is especially true of European Union's DemNet program. See Quigley (2000) for a discussion.
81
Before evaluating the aid’s success, we briefly summarize the economics literature that has put forward – but hardly scrutinized ? the notion of civil society congruent with the democracy aid project.123 3.3.2 Civil Society and Institutional Change A handful of economics scholars have attempted to link civil society and mechanisms of good governance. Roland (2004) focuses on factors influencing institutional change. He argues that understanding the interaction of slow-moving institutions (culture, values and norms) with fast-moving institutions (e.g. political institutions) is key to understanding institutional change. Transplanting 'best-practiceone-size-fits-all' fast-moving institutions usually does not work; the outsiders' knowledge of local conditions, including cultural factors (slow-moving institutions), is typically deficient. The local public's involvement in assessing the correct institutional mix is therefore crucial. A dialogue with the governments, however, is unlikely going to be fruitful, in part due to the governments' proneness to capture by elites with vested interests in status quo. Instead, Roland argues ? but does not deliberate on his proposal ? different components of civil society should participate in the process of institutional construction. Beck and Laeven (2005) empirically explore the determinants of economic performance and quality of institutions in transition countries. They argue institution building is best understood when the relative abundance of natural resources and length of time spent under socialism in individual countries are taken into account. Beck and Laeven (2005) believe that a strong civil society may be key to successful transition.
Interestingly, the literature below never attempted to directly relate its suppositions to the civil society aid project.
123
82
Civil society groups provide, according to them, a check on behavior of (transition) elites. In particular, the length of time spent under socialism matters for institution building because civil society groups were less likely to emerge in countries that spent lots of time under socialism. Without a strong civil society, a country’s transition elites are thus less likely to be counter-balanced and are thus better able to earn rents through resource extraction.124 Campos (1999) similarly empirically establishes a connection between the strength of civil society and measures of economic development for transition countries.125 Campos argues that a strong civil society, via participation in public affairs, increases transparency of policy making. Thereby, civil society involvement may complement, or even substitute for, other institutions of good governance, such as high quality of bureaucracy, public character of policy-making, and the executive’s accountability. Dethier at al. (1999) argue that the existence of a vibrant civil society at the start of the transition has the most explanatory power in their (empirical) analysis of determinants of successful economic liberalization in 25 post-communist countries of Central and Eastern Europe and the former Soviet Union.126 Hoff and Stiglitz (2004) appeal to civil society in their theoretical model of the evolution of rule of law. They argue civil society critically affects institution building: "A consequence of a prolonged experience under communism was the absence of civil
124 Beck and Laeven, however, do not test this hypothesis empirically. Had they included a measure of strength of civil society in early transition in their regression for institution-building, one would expect that the coefficient on Years under Socialism would become statistically insignificant if their story holds. 125
To measure the strength of civil society, Campos combines the Freedom House measures of civil liberties and political rights with the "civil society" indicator from Karatnycky et al. (1998).
126
They measure the strength of civil society with the Freedom House index of civil and political liberties.
83
society institutions…with countervailing power to hold the state to account" (Hoff and Stiglitz 2004, p. ?). In their model, the absence of a strong civil society creates pessimism about the emergence of rule of law, reducing the probability of rule of law being implemented given existing support for rule of law. In equilibrium, this makes emergence of rule of law less likely.127 The above contributions relate civil society to the process of institution building. In that process, civil society emerges as a force counterbalancing the government and the elites; through active involvement in public affairs, civil society involvement improves institutional reform. Ceteris paribus, promoting civil society's participation in public process should therefore have been social welfare improving. 3.4 Evaluating the Aid's Success: the Role of Initial Political Conditions128 As newly emerging organizations, civil society organizations were believed not to have vested interests in the old system. They were seen as a potential source of pressure for reform. An important part of their appeal was that they were also small enough that modest amounts of aid were significant to them. Qualitative, primarily anecdotal, evidence on the success of civil society aid for democracy promotion is scarce and emerging only now (Carothers 2004). It appears that the success of assistance to publicpolicy oriented CSOs in transition countries has, when seen in the aggregate, proven moderate, and extensively dependant on local conditions.
Similarly, they argue that the presence of agents with civic virtues - which they define as the agents who demand rule of law irrespective of their private interests - leads to an equilibrium increase in demand for rule of law. The following assessment is based on Quigley 2000 who focuses on Central and Eastern European countries. No analogous evaluation of the success of civil society aid exists for the countries of the former Soviet Union.
128
127
84
Civil society aid was provided mainly through means of technical assistance and grants. Many factors affected the success of the designated civil society aid. The knowledge and experience of the advising staff, for example, often proved inadequate for the local conditions. Likewise, many policy-oriented CSOs proved to have highly personal or political agendas.129 The overall success of the aid, however, seems to have been (most) critically dependant on the prevailing local political conditions. In the success stories of transition ? Poland, Slovenia, and Hungary, for example ? the governments were in general supportive of the CSOs. In these countries, policy-oriented CSOs have contributed to the public debate on crucial issues, particularly the pace and course of economic reform. In the less democratic and politically unstable group of transition countries – in particular in Romania, Bulgaria, Serbia and Slovakia ? on the other hand, the ruling governments hindered, and were opposed to strengthening, civil society. Thus, despite good faith and reasoning, civil society has since the collapse of communism not always been effective at promoting good governance. In bulk of the region – the countries marked by political instability and lower levels of democratic attainment ? the donors appear not to have (had) an impact on either civil society emancipation or institutional development. Ex post, the unsatisfactory experience has raised a debate, primarily in policy but also academic circles (Carothers 2004). Is fostering civil society at all worth spending a dollar on, or should all foreign aid instead be directed toward programs directly promoting economic reform? Was the focus on policy-oriented CSOs correct, or should
129
This in particular holds for policy-related research institutions or think tanks focusing on economic issues.
85
the donors have rather channeled funds to service-oriented CSOs very early on (even if service-oriented CSOs are less successful in contributing to democracy promotion in the short run)? 3.5 Aiding Whom and When: Towards an Analytical Framework Both the economics literature summarized in section 3.3.2 and the practitioners' reasoning seem to agree that a key feature of civil society involvement in the process of institution-building is to render the reform process open to public and subject to deliberation. At the very least, this should increase the transparency of the process. Organized legal professions' review of reform proposals, as detailed in chapter 1 of this dissertation, is an example of such involvement. An independent bar has long been seen as "the backbone of an autonomous civil society" (Rueschemeyer 1997, p. 217). Involvement of lawyers' associations "…is both an expression of the autonomy of civil society and the means of opposition to the state, as well as a force that reinforces this dualism. Bar action constitutes (with others) civil society" (Halliday and Karpik 1997, p. 56). Lawyers' associations are natural candidates for civil society aid due to their expertise about local institutional specifics. Indeed, the American Bar Association's Central European and Eurasian Law Initiative (CEELI) has recently launched a project to help strengthen the role of organized legal professions in emerging democracies.130 Lawyers associations, however, should not be mistaken as the only component of civil society with capacity to open up the policy-making process. In fact, the entire subset of policy-oriented CSOs as a group (CSOs performing mainly expressive functions
130
See http://www.abanet.org/ceeli/areas/legal_prof_index.html.
86
according to the taxonomy of Salomon et al. (2004)) may, through their active and vocal involvement, we believe, contribute to the success of the reform process. Prevention of quick and ad hoc implementation of reform proposals by ruling political elites and their allies is, we maintain, a key role of civil society in the process of institution-building. Orenstein's (2000) analysis of political economy of pension reform resonates with this. Orenstein provides illuminating case studies of design and implementation of fundamental pension reform programs in Hungary, Poland, and Kazakhstan. From all three post-communist countries, in the least democratic Kazakhstan, the deliberative process was significantly shorter than in the more democratic Hungary and Poland. In the authoritarian Kazakhstan, civil society organizations in Kazakhstan "face greater sanctions for voicing opposition" and have "fewer opportunities to access the policy process". Pension reform deliberations were performed in secrecy by a single governmental commission under the direct authority of the prime minister, with full approval of the president. In fact, "[c]ivil society groups did not even know about the progress of governmental pension reform proposals" (Orenstein 2000, p. 18). We therefore believe the initial focus of the donors on the aforementioned subset of CSOs was generally correct.131 The caveat, we believe, lied in correct evaluation of existing political conditions in transition countries. Indeed, we argue below that aiding civil society is wasteful when the attained levels of political stability and democratization are inadequately low. We now formalize our argument, relying heavily on, and extending the analysis in chapter 1.
131
Quigley (2000), for example, doubts that such focus was indeed correct.
87
3.5.1 Model Background The difference between the model as set up in chapter 1 and the model utilized in this section is interpretative. In chapter 1, the delay of reform implementation was due to review performed by organized legal profession. In this section, we view delay of reform implementation as arising via civil society's involvement in the public process, as stipulated among others by Roland (2004), Hoff and Stiglitz (2004), Beck and Laeven (2005), Campos (1999) and Dethier at al. (1999). Taking this into account, we now merely summarize the model's main qualitative features and refer the reader to chapter 1 for the mathematical description of the model. In the process of designing and implementing institutional reform, the government faces the competing demands of different interests. Rent-seeking interest groups ? some with objectives more congruent with the ruling government than others – possess better information about the effects of proposed reforms than the politicians. Interest groups support politicians, contributing funds when submitting reform proposals. A politician forming the government weighs these contributions against the pursuit of the general welfare, the latter being relatively more important when democratic constraints are tighter (i.e. when ? is greater). Since interest groups submit self-serving reforms for consideration, the politician faces an informational problem because some reforms are damaging to the politician and others are not. A government's conscious decision to slow down the reform process ameliorates the politicians' adverse selection problem. The government may, or may not, allow for active participation of non-partisan, policy-minded civil society groups that have the capacity to stimulate the debate about the reform. The autonomy of civil society adds
88
political legitimacy and credibility to the decision to deliberate on reform proposals and delay their implementation in the face of powerful interest groups pressing for quick reforms. Civil society's participation opens up the reform process to public debate and delays the adoption of reforms. Delaying reforms will be costly for interest groups, since the government might lose power (with probability 1?p) before implementation occurs. This cost changes the incentives of the interest groups when submitting reforms: initiatives with lower payoffs are dropped. Institutional reform may therefore improve (in line with Proposition 2 in chapter 1).132 Prevailing political conditions critically influence the government's decision to promote or suppress civil society's involvement (in line with Proposition 3 in chapter 1). When democracy is stronger (when ? is greater), the politician weighs general welfare more heavily than interest groups' contributions. Political stability also matters. When government turnover is more likely (when p is smaller), the politician's expected benefit from delaying a reform program to obtain better reforms, but later, declines. Democracy and political stability thereby become substitutes. The politician endorses civil society's involvement in an unstable environment only at high levels of democracy. When there is political stability, however, allowing for civil society's deliberation is attractive to the politician even at lower levels of democracy. This could be precisely the reason why Pérez-Diaz, for example, emphasizes that while civil society
Recall that the delay of a reform program improves the outcome of reform from society's standpoint because the preferred reform program of the interest group aligned with the ruling political option decreases social welfare (in notation of chapter 1: B?3+B?3+B?2+B?2<0). If, in contrast, the lobbying interest group's preferred reform program although not first best for society - increases social welfare (B?3+B?3+B?2+B?2>0), then it may be easily shown that delaying reform implementation is socially undesirable when political instability is either too high or too low, but may still be socially desirable for middling ranges of political instability. See Appendix 2 for analysis.
132
89
can fully emerge only in conjunction with a limited state that observes the rule of law, it nevertheless can emerge under authoritarian regimes (Bell 1997, p. 86). 3.5.2 When Does Aid Make a Difference? The Role of Political Conditions Revisited Aiding civil society may take on many hues, as argued in section 3.4. A more vibrant and autonomous civil society sector increases the transparency of the reform process, and effectively put a hold on the self-interested government's actions. We investigate the following question: In a setting where the civil society is easily encroached by the state, when does fostering civil society lead to improvement in social welfare?133 For convenience, let us focus on equilibria when the politician's self-interest matters. When ?<? ?, the politician in power is faced with an informational problem and approves the submitted reform program that is social welfare reducing.134 We assume the donor's decision to aid civil society does not bind the politician to endorse civil society: once civil society has been "strengthened", it is still up to the politician's discretion to decide whether to assure civil society's participation in the reform process or not.135 For simplicity, we also assume that the political cost of suppressing the civil society before the donors' have confirmed their participation equals 0. After the aid has been channeled, let the political cost of suppressing civil society be c? as in section 1.5 of chapter 1. c is now strictly positive.
133
The qualitative conclusions are not different if we instead analyze the effect of aid for the purpose of bringing civil society, capable of participating in public affairs, into existence (rather than strengthening an already existing civil society).
? the politician's aims coincide with society's aims: the politician rejects Recall that in the less interesting case of ?>? the submitted social welfare reducing reform program.
134 135
Effectively, we added a stage at the very beginning of the game as displayed in Figure 2 in chapter 1. This assumption certainly reflects the practice of civil society aid. See e.g. Ottaway and Carothers (2000) and Carothers (2004).
90
Let W denote social welfare and V the politician's payoff, as in chapter 1. Let superscript ? indicate no participation of civil society in the reform process, and let superscript + stand for involvement of civil society in the reform process. Define Unoaid as the social welfare when civil society exists but the political cost of preventing it to deliberate on the reform proposal equals zero (c = 0). Applying analysis from section 1.4 of chapter 1, Unoaid = W+ if V+ > V? and Unoaid = W? if V+ < V?. Let Uaid be the social welfare when the civil society has been aided, so c>0. Then, applying analysis from section 1.5 of chapter 1, Uaid = W+ if V+ > V? ? c? and Uaid = W? if V+ < V? ? c?. Accordingly, we have the following proposition: ? and c>0. Define Proposition.136 Let ?<? P = {(p,?): p<??A/B?3 and V? ? c? < V+ < V?} Q = {(p,?): p?[??A/B?3,??A/B?2) and V? ? c? < V+ < V?} R = {(p,?): p<??A/B?2 and V? ? c? < V+ < V?} S = {(p,?): p>??A/B?2 and V+ > V?} P, Q, R and S are (i) non-empty, (ii) pairwise mutually exclusive, (iii) P?Q?R?S = {(p,?): p?(0,1) and ??(0,? ?)} Define also ? ? Uaid ? Unoaid. Then: (iv) ? = 0 for all (p,?)?R?S, (v) ? = ?1 for all (p,?)?P, (vi) ? = ?2 for all (p,?)?Q, where ?2 > ?1 > 0. Figure 10 assists the interpretation of Proposition. Figure 10 shows that aiding civil society is superfluous when the attained levels of democracy and political stability are relatively low (region R); the ruling government is willing to incur the political cost of suppressing the civil society even once the latter has been strengthened. In contrast, when the attained levels of democracy and political stability are already high, there exists no need to aid civil society (region S).
136
This proposition is a corollary to Propositions 4 and 5 in chapter 1, so we omit its proof.
91
It is therefore in regions P and Q that political conditions are such that aiding civil society will improve social welfare. In region P, the attained levels of democracy are already considerable, but a country is still plagued with political instability. In region Q, both the attained levels of democracy and political stability are still relatively modest but have surpassed some minimum levels. Moreover, while aiding civil society improves social welfare in both regions P and Q, the aid is most effective in increasing social welfare when political instability has already been reduced (in region Q). 3.5.3 Judging the Success of Civil Society Aid: A Conjectural Interpretation The above Proposition confirms the importance of initial political conditions in the country where civil society is to be aided, as stipulated by Quigley (2000). In addition, it proposes a conjectural explanation for the observed variation in local governments' responses to the civil society aid project in countries undergoing postcommunist transition. Note first that, as a benchmark, the group of democratic and politically stable developed countries may be placed in Region S of Figure 10. In region S it is in the politicians' best interest to endorse, or at least not to impinge on, the existing civil society sector. Aiding civil society in developed countries is, according to the Proposition, superfluous. The transition countries during the early 1990s differed substantially in their political and economic conditions, and so did the experience with civil society aid. In the advanced transition countries (such as Poland, Slovenia, Czech Republic and Hungary), the levels of democracy and political stability were already considerable. Accordingly, this group of transition countries may be placed in regions Q and P. As predicted by the
92
model, the donors' efforts to foster civil society were successful, with the local governments actively endorsing civil society's participation in the reform process.137 The less democratic and politically less stable group of transition countries may be placed in region R. In this group, the ruling governments were often reluctant, or outright opposed, to promoting civil society (Quigley 2000), as the model predicts. The international donors' attempts to foster civil society were destined to fail. And they largely did, as Quigley (2000) argues about experiences of Romania, Bulgaria, Serbia and Slovakia, for example.138 The complex phenomenon of transition manifested through implementation of large-scale institutional reforms. The success of the latter was comparatively greater in the relatively more democratic Poland, Slovenia, Czech Republic and Hungary than in the less democratic and politically unstable group of transition countries. A conjectural interpretation of the proposition above also suggests that greater civil society participation in the former group of countries – in part assured by international aid – may have contributed to the quality of reform. The empirical evidence provided by Beck and Laeven (2005), Campos (1999) and Dethier at al. (1999) appears to confirm this supposition. The proposition above also provides support for initial phases of democracy project, when the donors aimed at directly fostering democracy (through ensuring impartial and democratic elections, for example) and pursued reform of major legislative
137
"In Poland and Slovenia, for example, the government was generally receptive to the [civil society] sector and gave it considerable space within which to operate" (Ottaway and Carothers 2000, p. 205).
138
"In Romania and Bulgaria, for example, the leaders…were opposed to strengthening civil society, and created roadblocks that delayed the establishment of foundations in their countries" (Ottaway and Carothers 2000, p. 204). Likewise, "[in] Slovakia and Serbia…the government provided no space or limited space for extragovernmental activities" (Ottaway and Carothers 2000, p.205).
93
and judicial institutions. In the light of the proposed model, earlier phases of democracy aid project may be interpreted as attempts to raise the initial levels of democracy (thereby increasing ?) and reduce political instability (thereby increasing p). It was only later on that civil society aid became means of promoting good governance and effective institutions. Interpretation of Proposition, however, suggests that aid to policy-oriented, non-partisan civil society organizations perhaps often came too early to countries undergoing post-communist transition. For the civil society sector to properly function as a bulwark against the state, the condition of minimum attained levels of democracy and political stability must already be fulfilled.139 3.6 Summing Up The many facets of civil society provide for a fruitful research agenda. Only very recently has the economics literature started illuminating the role of civil society for service provision and social capital accumulation. This chapter focused on the role of civil society for building institutions of good governance. Civil society's active deliberation on the proposed reform agenda restricts the set of viable proposals for special interests. A vibrant civil society may therefore be vital to adopting good institutions. This notion of civil society as a "bulwark against the state" underpinned the initial stages of civil society aid in early 1990s to post-communist countries as part of the greater democracy aid project. The scant available evidence on the aid's success reveals that aiding non-partisan, policy-oriented civil society organizations is no panacea for
139
In terms of Figure 1, a country must be placed in region Q rather than R for civil society aid to increase social welfare.
94
promoting good governance. Much is left to be understood.140 Our model predicts that aiding civil society improves social welfare as long as the political conditions in the recipient country are such that endorsing civil society's participation is in the politician's interest. Aiding civil society in an unstable authoritarian regime, for example, will be unproductive. A fundamental problem with aiding civil society lies in the bewildering array of organizations it encompasses. Accordingly, "[some] civil society groups may stand for 'higher' – that is, nonmaterial – principles and values, but much of civil society is preoccupied with the pursuit of private and frequently parochial and grubby ends" (Carothers 2004, p. 101). Aiding interest groups is hardly justifiable for economists (see, e.g., Olson 1982). Yet aiding a subset of civil society groups may nevertheless be beneficial even if civil society organizations do not pursue "higher goals". CSOs need not have a direct stake in the reform process.141 They may participate in the politics of fundamental institutional reforms perhaps to extract rents from prospective winners and losers from reform, or simply to raise their own prestige.142 By preventing hasty reform implementation, civil society's involvement may, as argued above, thus improve institutional reform.143
140
Thomas Carothers’ "…summation of ten years of Western civil-society aid to Eastern Europe and the former Soviet Union…boils down to a rather modest conclusion: They went to teach; they stayed to learn; they are learning still" (Carothers 2004, p.120).
141 142 143
In contrast, interest groups ? and ? in the model of chapter 1 do have a direct stake in reform. Lawyers associations are again an example of such civil society groups.
To a certain extent, however, our analysis still overstates the positive contribution of civil society since we do not account for costs associated with extracting rents from ? and ? in our social welfare calculation.
95
Appendix 1: Figures
Figure 1: The Set of Reforms
T
Benefit to ? (B?j) Reform 3 Reform 1 Reform 2 Benefit to ? (B?j) Reform 5 Reform 6 Reform 4 T
96
Figure 2: The Timeline
A stays in power with probability p Interest groups lobby for reforms Politician chooses to use review by profession Review
Implementation Interest groups lobby for reforms that have Interest groups lobby for reforms that have not been subject to review yet Interest groups choose not to lobby for f
Implementation
A is elected, p and ? are realized
Interest groups choose not to lobby for reforms
B comes to power with probability 1?p
Review
End of reform proces
Politician chooses not to use review by profession
Interest groups choose not to lobby for reforms
Interest groups lobby for reforms
Implementation
Time period 1
Time period 2
97
Figure 3: The Politician's Welfare and Aggregate Welfare with and without a Powerless Profession
democracy (?)
1
? ?
W+=W? V+=V?
W+=W? V+=V?
?C(p)
W+>W? V+>V?
?
W+<W? V+>V?
W >W V+<V?
??A/B?3
+
?
?G
?P
p?(?)
?S
0
??A/B?2
1
stability (p)
Notes: 1. W+ = aggregate welfare with professional review 2. W? = aggregate welfare without professional review 3. V+ = politician’s welfare with professional review 4. V? = politician’s welfare without professional review 5. Boundary between use of profession and non-use 7. S = The Soviet Union before Gorbachev 8. G = Georgia during early transition 9. P = Poland during early transition
98
Figure 4: Professional Political Power, Democracy, and Stability
democracy ?T(p,c)
? ? ?T(p,c2) ?T(p,c1)
?
??A/B?3
??A/B?2
stability (p)
Notes: 1. c1>c2?0 2. ?T(p,0) lies along the solid line in Figure 3.
99
Figure 5: Political Power and the Use of the Profession
democracy (?)
1
? ?
The profession directly aids the politician through its effect on institutional reforms
The profession's political power dictates its use by the politician
E? F3?
? F2
Politician does not use the profession
? F1
p*(?)
0
??A/B?3
??A/B?2
1
stability (p)
Boundary between use of profession and non-use when political power (c) is positive. Boundary between use of profession and non-use when political power (c) is zero. (This line is the same one as in Figure 3.) F1 = F2 = F3 = E = France under Louis XIV, Napoleon (c=0 and is pertinent). France early in the reign of Louis XV (c>0 and is pertinent). France late in the reign of Louis XV (c=0 and is pertinent). France soon after the revolution (c=0 and is pertinent). England under James II (c>0 and is pertinent).
100
Figure 6: The Political Cost of Removing the Profession
Cm(?, p)
2??A Cm(?,p1) c? ??A Cm(?,p2) Cm(?,p3) Cm(?,p4) p1<??A/B?3 p2=??A/B?3 p3, p4?[??A/B?3,??A/B?2) p1<p2<p3<p4
0 ? (p ,c) ? (p ,c) ? (p ,c)
T 2 T 4 T 3
? (p ,c)
T
1
? ?
?
101
Figure 7: Timeline of the Regulatory Process
t=0 Regulatory rights assigned
t=1 Gov’t passes enabling legislation: L = L*
t=2 ? realized: L = L*+?
t=3 If no renegotiations, default ?L implemented; however, renegotiations will occur
t=4 L = L*+?+?L determined, payoffs realized
Figure 8: SR vs. R: Efficiency vs. Government's Incentives
?2/(p?c)2 VR = VSR
4+3?R+?SR 4(?R??SR)
I
IV III
0
II
WR = WSR
1
?
102
Figure 9: Legal Origin and Regulatory Regime Choice
Figure 9(a): Legal Origin and Regulatory Regime Choice when ?2e > ?2f ?2/(p?c)2
Figure 9(b): Legal Origin and Regulatory Regime Choice when ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0 ?2/(p?c)2
4+3?R,f+?SR,f 4(?R,f??SR,f) VR(?R,f) = VSR(?SR,f) 4+3?R,e+?SR,e 4(?R,e??SR,e)
4+3?R+?SR 4(?R??SR) V =V
R SR
VR(?R,e) = VSR(?SR,e) ?2e (p?c)2 ?2f (p?c)2 0 {?: VR(?2e) ? VSR(?2e)} {?: VR(?2f) ? VSR(?2f)} 1 ? 0
A1
A2
1 ? A2 = {(?,?2/(p?c)2): VR(?R,e) ? VSR(?SR,e)} A1?A2 = {(?,?2/(p?c)2): VR(?R,f) ? VSR(?SR,f)}
103
Figure 10: Effectiveness of Aid to Civil Society
democracy (?)
? ?
(? = ?1>0)
P
Q
(? = 0) 0 ??A/B?3
(? = 0)
S
(? = ?2 > ?1)
R
??A/B?2
1
stability (p)
Notes: Refer to chapter 1, section 1.5 and Propositions 4-5, for definition of ?T(p,c). The thicker dashed line represent the ?T(p,0) function. The thicker solid line represents the ?T(p,c) function, with c>0. ? = Uaid ? Unoaid, as defined in Proposition in the text.
104
Appendix 2: Tables
Table 1: Summary Statistics by Regulatory Arrangements: Means (and Standard Deviations) of Variables Definition (and Source) All observations 0.39 (0.49) 0.17 (0.38) 0.31 (0.47) 0.07 (0.25) 0.07 (0.25) 8.76 (0.92) By regulatory arrangement Central SelfRegulation Combination Regulation 0.38 (0.49) 0.31 (0.47) 0.15 (0.37) 0.05 (0.22) 0.10 (0.31) 8.57 (0.87) 0.43 (0.51) 0.05 (0.22) 0.38 (0.50) 0.09 (0.30) 0.05 (0.22) 9.09 (0.99) 0.35 (0.49) 0 (0) 0.58 (0.51) 0.06 (0.24) 0 (0) 8.80 (0.90)
Variable French
= 1 if French legal origin (La Porta et al. 1999) = 1 if socialist legal socialist origin (La Porta et al. 1999) = 1 if English legal English origin (La Porta et al. 1999) = 1 if German legal German origin (La Porta et al. 1999) = 1 if Scandinavian Scandin legal origin (La Porta et al. 1999) lgnipc1991 Logarithm of GNI per capita, PPP current international $ (World Development Indicators) Logarithm of lpop1991 population total (World Development Indicators) Number of observations
16.28 (1.73)
16.26 (1.79)
16.48 (1.78)
16.09 (1.61)
77
39
21
17
105
Table 2: The Relationship between Legal Origin and Regulatory Regime: Ordered Probit Results Variable socialist English German Scandin lgnipc1991 lpop1991 threshold 1 threshold 2 Coefficient ?1.4818 0.8019 ?0.2047 ?1.3623 0.3807 ?0.0113 3.1226 4.0820 Standard Error* 0.5232 0.3349 0.5474 0.6230 0.1736 0.0835 2.3729 2.4082 z ?2.83 2.39 ?0.37 ?2.19 2.19 ?0.14 P>?z? 0.005 0.017 0.708 0.029 0.028 0.892
(Ancillary parameters)
Log-pseudo likelihood = ?65.7562 Pseudo R2 = 0.1728 Number of observations = 77
*
Robust standard errors.
Notes: 1. French origin is the omitted dummy. 2. Self-regulation is the highest category of the dependent variable.
106
Appendix 3: Proofs of Propositions in Chapter 1
? the politician accepts all reform proposals. Consider reform Proposition 1. When ?<? j?{2,3}. ? will never want to lobby for j?{2,3}, since B?j<0 for j?{2,3}. There are four possible sub-games at the beginning of period 2, depending on whether ? chose to lobby for reform j?{2,3} in period 1 or not and whether A or B is in power in period 2. 1. If ? lobbied for j?{2,3} in period 1 and A is in power in period 2, then the payoff to ? when it lobbies for reform j?{2,3} in period 2 is B?j???A and B?j when it does not lobby. Therefore, ? does not lobby. 2. If ? lobbied for j?{2,3} in period 1 and B is in power in period 2, then the payoff to ? when it lobbies for reform j?{2,3} in period 2 is B?j???B<0 and 0 when it does not lobby. Therefore, ? does not lobby. 3. If ? did not lobby for j?{2,3} in period 1 and politician A is in power in period 2, then the payoff to ? when it lobbies for reform j?{2,3} in period 2 is ???A<0 and 0 when it does not lobby. Therefore, ? does not lobby. 4. If ? did not lobby for j?{2,3} in period 1 and politician B is in power in period 2, then the payoff to interest group ? when it lobbies for reform j?{2,3} in period 2 is ???B<0 and 0 when it does not lobby. Therefore, ? does not lobby. If ? chooses to lobby for j?{2,3} in period 1, its payoff is pB?j???A. If ? chooses not to lobby for j?{2,3} in period 1, its payoff is 0. Therefore, ? chooses to lobby for j?{2,3} in period 1 if and only if p???A/B?j. Consider now reform j?{4,5}. ? will never lobby for j?{4,5}, since B?j<0 for j?{4,5}. There are four possible sub-games at the beginning of period 2, depending on whether ? chose to lobby for j?{4,5} in period 1 or not and whether politician A or politician B is in power in period 2. 1. If ? lobbied for j?{4,5} in period 1 and A is in power in period 2, then the payoff to ? when it lobbies for j?{4,5} in period 2 is B?j???A<0 and B?j when it does not lobby. Therefore, ? does not lobby. 2. If ? lobbied for j?{4,5} in period 1 and B is in power in period 2, then the payoff to ? when it lobbies for j?{4,5} in period 2 is B?j???B>0 and 0 when it does not lobby. Therefore, ? lobbies. 3. If ? did not lobby for j?{4,5} in period 1 and B is in power in period 2, then the payoff to ? when it lobbies for j?{4,5} in period 2 is ???B<0 and 0 when it does not lobby. Therefore, ? does not lobby. 4. If ? did not lobby for j?{4,5} in period 1 and politician A is in power in period 2, then the payoff to ? when it lobbies for j?{4,5} in period 2 is ???A<0 and 0 when it does not lobby. Therefore, ? does not lobby. If ? chooses to lobby for j?{4,5} in period 1, its payoff is pB?j+(1?p)(B?j???B)???A<0. If ? chooses not to lobby for j?{4,5} in period 1, its payoff is 0. Therefore, ? chooses not to lobby for reform j?{4,5} in period 1.
107
Proposition 2. If p<??A/B?3, ? chooses not to lobby, so W+ = 0>W?. If ??A/B?3?p<??A/B?2, ? lobbies for reform 3, so W+ = p(B?3+B?3)>W?. If p???A/B?2, ? lobbies for reforms 2 and 3, so W+ = p(B?2+B?2+B?3+B?3)>W?. Proposition 3. If p<??A/B?3, ? chooses not to lobby in period 1, so V+ = 0<V?. If p???A/B?2, ? lobbies for 2 and 3 in period 1, so V+ = ?p(B?2+B?2+B?3+B?3)+(1??)2??A>V?. If ??A/B?3?p<??A/B?2, ? lobbies for 3 in period 1, so V+=?p(B?3+B?3)+??A. Then, V+>(<)V??p>(<)p?(?)?1+[?(B?2+B?2)+(1??)??A]/[?(B?3+B?3)]. The following is useful for the purposes of constructing a diagram to examine the tradeoffs between ? and p. Observe that p?(?)<(>)1 if and only if ?>(<)?. It can be easily shown that ?p?(?)/??<0 and ?2p?(?)/??2>0. Define ?L = ??A/[(??A/B?2?1)(B?3+B?3)+?B?2+B?2?+??A] and ?H = ??A/[(??A/B?3?1)(B?3+B?3)+?B?2+B?2?+??A]. ¯, p*(?H)=??A/B?3 and p*(?L)=??A/B?2. Then ?<?L<?H<? Proposition 4. Define Cm(p,?) as the maximum political cost that the politician is willing to incur to remove the profession from its advisory role. If V+<V?, Cm(p,?) = V?(p,?) ? V+(p,?) and if V+?V?, Cm(p,?) = 0. More precisely, ? ?[(B?2+B?2)+(B?3+B?3)]+(1??)2??A, when p<??A/B?3 and ?<? ? C (p,?) = ?[(B?2+B?2)+(B?3+B?3)]?p?(B?3+B?3)+(1??)??A, when p<p (?) and p?[??A/B?3,??A/B?2) 0, elsewhere. The politician chooses not to use the profession for all values of ? and p such that Cm(p,?) > c?. Cm(p,?) has the following properties: ?, 1. When p<??A/B?3 and ?<? ?Cm(p,?)/?p = 0, ?Cm(p,?)/?? = [(B?2+B?2)+(B?3+B?3)]?2??A<0 2. When p?[??A/B?3,??A/B?2) and p<p?(?), ?Cm(p,?)/?p = ?(B?2+B?2)<0, ?Cm(p,?)/?? = [(B?2+B?2)+(B?3+B?3)]?p(B?3+B?3)???A<0, ?2Cm(p,?)/???p = ?(B?2+B?2)<0. 3. Elsewhere, Cm(p,?) = 0. 4. Two additional properties that enable us to construct Figure 6 are:
m
?C m (p, ? 0 )/?? C m (p, ? 0 )
p < ? ?A /B ?3
< ?C m (p, ? 0 )/??
p = ? ?A /B ?3
for all ? 0 < ? for all ? 0 < ?
p < ? ?A /B ?3
> C m (p, ? 0 )
p?[? ?A /B ?3 , ? ?A /B ?3 ) and p < p*(? 0 )
The properties of Cm(p,?) summarized above imply than for any pair (p,c), there exists a unique value of ?T(p,c)>0 such that Cm(p,?T(p,c)) = c?T(p,c) if V+<V?. On the other hand, Cm(p,?) = 0 if V+?V? and hence ?T(p,c) = 0. This proves that ?T(p,c) is a function. To
108
prove that ?T(p,c) is discontinuous at p = ??A/B?2 and at p = ??A/B?3, and points (ii) – (iv), note also the following: 1. When p<??A/B?3, ?T(p,c) is defined by c?T(p,c) = ?T(p,c)(B?2+B?2+B?3+B?3)+(1??T(p,c))2??A, so that ?T(p,c) = 2??A/[c?(B?2+B?2+B?3+B?3)+2??A]. Then, ??T(p,c)/?p = 0. 2. When p?[??A/B?3,??A/B?2), ?T(p,c) is defined by c?T(p,c) = ?T(p,c)(B?2+B?2+B?3+B?3)?p?T(p,c)(B?3+B?3)+(1??T(p,c))??A, so that ?T(p,c) = ??A/[c?(B?2+B?2+B?3+B?3)+p(B?3+B?3)+??A]. ??T(p,c)/?p = ?(B?3+B?3)?T(p,c)/[c?(B?2+B?2+B?3+B?3)+(B?3+B?3)p+??A] < 0 and ?2?T(p,c)/?p2 > 0. 3. When p???A/B?3, ?T(p,c) = 0. 4. ? T (p, c 0 ) > ? T (p, c 0 ) > ? T (p, c 0 ) = 0 for all c 0 > 0.
p < ? ?A /B ?3 p?[? ?A /B ?3 , ? ?A /B ?2 ) p ? ? ?A /B ?2
Proposition 5. Let c1>c2?0. Define ?T(p,c1,c2) = ?T(p,c2) ? ?T(p,c1). 1. When p<??A/B?3, ?T(p,c1,c2) = 2??A/[c2?(B?2+B?2+B?3+B?3)+2??A] ? 2??A/[c1?(B?2+B?2+B?3+B?3)+2??A] > 0 and ??T(p,c1,c2)/?p = 0. 2. When p?[??A/B?3,??A/B?2), ?T(p,c1,c2) = ??A/[c2?(B?2+B?2+B?3+B?3)+p(B?3+B?3)+??A] ? ??A/[c1?(B?2+B?2+B?3+B?3)+p(B?3+B?3)+??A] > 0 and ??T(p,c1,c2)/?p < 0. 3. When p???A/B?2, ?T(p,c2) = ?T(p,c1) = 0, so ?T(p,c1,c2) = 0. ? and when p?[??A/B?3,??A/B?2), Note that when p<??A/B?3, ?T(p,0) = ? T ?1 ? (p,0) = p* (?).
109
Appendix 4: Model with Reform 1
This appendix presents the model, as developed in sections 1.3 and 1.4 of the text, but with Pareto-improving reform 1 explicitly included in the analysis. No other features of the model have been changed. We show that (a) reform 1 will always be implemented and (b) the model’s qualitative features do not change when reform 1 is included in the analysis. The presentation here closely parallels that in the main text. The following table restates the payoff structure for reforms 1, 2 and 3: Reform type\Payoff Reform 1 Reform 2 Reform 3 Interest group ? B?1>0 B?2>0 B?3>0 Interest group ? B?1>0 B?2<0 B?3<0 Aggregate (? and ?) B?1+B?1>0 B?2+B?2<0 B?3+B?3>0
The sequence of events is as described in section 1.3 of the main text. There is a single reform process beginning when one politician, say A, takes the reins of power. The model covers two sub-periods, A’s initial incumbency and the following time interval when either A or B forms the government. The decision on whether to use professional review is made at the beginning of A's incumbency and holds for the whole reform process. Let p?(0,1) be the known, exogenous probability that A stays in power in the second period. In both periods, each interest group decides whether or not to submit a proposal without knowing what the other is doing. The government then decides whether to accept the proposal. If it does, it receives the submission fee and if there is no professional review the measure is immediately passed. If there is professional review, the measure is debated until the second period. Then, if A remains in office, the reforms submitted in period 1 are implemented. If B takes over, each interest group decides again whether to submit a proposal. If the reforms submitted in period 2 are the ones that already passed through professional review in period 1, they are 110
implemented. If the reforms submitted in period 2 differ from the ones submitted in period 1, they are delayed beyond the end of this reform process.144 We maintain the same assumptions as in the chapter itself: Assumption A1. B?2=B?3 Assumption A2. B?2+B?2+B?3+B?3<0 Assumption A3. For all j such that B?j>0, ??A<B?j<??B For all j such that B?j>0, ??B<B?j<??A Assumption A4. Aggregate welfare from implementing reform j is B?j+B?j. Assumption A5: The politician's payoff from j is ?(B?j+B?j)+(1??)?ik, ??(0,1).
? are defined as follows: ? = ??A/[?B?2+B?2?+??A] and Recall that ? and ¯ ¯= ??A/{??A+½[?B?2+B?2??(B?3+B?3)]}<1. Reforms 1 and 3 are always desirable for ? politician A. Reform 2 is undesirable for A if ?>? and desirable if ?<?. Thus, when ?>?, the politician faces an adverse selection problem. Politician A likes the average reform ¯. consisting of reforms 2 and 3 if and only if ?<? Without a screening mechanism, the politician either approves all reforms or none of them. A approves all reforms if and only if [?(B?1+B?1)+(1??)??A]+[?(B?2+B?2)+(1??)??A]+[?(B?3+B?3)+(1??)??A]>0 ??[B?1+B?1+B?2+B?2+B?3+B?3?3??A]>?3??A. ~ Define ?=??A/{??A?(?)[B?1+B?1+B?2+B?2+B?3+B?3]}. It is easy to show that if ~ ??(0,1) then the immediately preceding inequality is satisfied for all ??(0,1). Hence,
This assumption closes the model without an infinite horizon. It is equivalent to focusing on a single reform process that must take place within a certain time window, which is a common feature of the circumstances surrounding institutional reforms.
144
111
~ when there is no professional review A approves all reforms if ??(0,1), which is the case ~ if and only if B?1+B?1+B?2+B?2+B?3+B?3>0. When B?1+B?1+B?2+B?2+B?3+B?3<0, 1>?>¯ ? ~ and A approves all reforms if and only if ?<?. As in the main text, we split the analysis into two parts. The more interesting case ~ ~ (corresponding to ?<¯ ? in the paper) is when either ??(0,1) or 0<?<?<1 implying that A approves all reforms when there is no professional review. The other case ¯ in the paper) is when 0<~ (corresponding to ?>? ?<?<1 and A rejects all reforms when there is no professional review. Propositions A4.1?A4.3 below match the corresponding ~ ~ propositions in the text and hold when either ??(0,1) or 0<?<?<1. Figure A4.1 summarizes the conclusions graphically. ~ ~ Proposition A4.1. Assume that A is in power, either ??(0,1) or 0<?<?<1, and there is professional review. In period 1, ? lobbies for reform 1 for any p and for reforms 2 and 3 if and only if p???A/B?2 and p???A/B?3, respectively. In period 2, ? does not lobby for any reform regardless of which politician is in power. ? does not lobby for any of the reforms in period 1, and lobbies for reform 1 in period 2 if politician B is in power. Reform 1 will therefore always be implemented. (See the end of this appendix for the lengthy proof of Proposition A4.1.) ~ ~ Proposition A4.2. When either ??(0,1) or 0<?<?<1, social welfare is always higher with the profession than without it. Proof: Without the profession, W? = B?1+B?1+B?2+B?2+B?3+B?3. With the profession: For p<??A/B?3, W+ = B?1+B?1>W?. For ??A/B?3?p<??A/B?2, W+ = B?1+B?1+p(B?3+B?3)>W?. For p???A/B?2, W+= B?1+B?1+p(B?2+B?2+B?3+B?3)>W?.
?
112
~ ~ Proposition A4.3. Assume that either ??(0,1) or 0<?<?<1. For p<??A/B?3, the politician is worse off with the profession if and only if ?<¯ ? and better off with ¯ the profession if and only if ?>?. For p???A/B?2, the politician is better off with the profession. For p?[??A/B?3,??A/B?2), the politician is better off with the profession if and only if p>p?(?) and worse off with the profession if and only if p<p?(?), where p?(?)?1+[?(B?2+B?2)+(1??)??A]/[?(B?3+B?3)]. Proof: Without the profession, V?=?(B?1+B?1+B?2+B?2+B?3+B?3)+(1??)3??A. With the profession: For p<??A/B?3, V+=?(B?1+B?1)+(1??)??A. V+>(<)V??0>(<)?(B?2+B?2+B?3+B?3)+(1??)2??A??>(<) ¯ ?. For ??A/B?3?p<??A/B?2, V+=?(B?1+B?1) +?p(B?3+B?3)+(1??)2??A. V+>(<)V??p>(<)1+[?(B?2+B?2)+(1??)??A]/?(B?3+B?3)?p*(?) The following is useful for the purposes of constructing Figure 1, which examines the trade-offs between ? and p. Observe that p?(?)<(>)1 if and only if ?>(<)?. It can be easily shown that ?p?(?)/??<0 and ?2p?(?)/??2>0. Define ?L = ??A/[(??A/B?2?1)(B?3+B?3)+?B?2+B?2?+??A] and ?H = ??A/[(??A/B?3?1)(B?3+B?3)+?B?2+B?2?+??A]. Then ?<?L<?H<¯ ?, p*(?H)=??A/B?3, and p*(?L)=??A/B?2. For p???A/B?2, V+=?(B?1+B?1)+?p(B?2+B?2+B?3+B?3)+(1??)3??A. V+>V? because ?p(B?2+B?2+B?3+B?3)>?(B?2+B?2+B?3+B?3). ? ~ Finally, consider the outcomes when 0<?<?<1. The politician does not approve any reforms without a screening mechanism, so V? = W? = 0. With professional review, the reform-submission equilibria are the same as in Proposition 1 when ??A/B?2>p. For p<??A/B?3, reform 1 only is submitted and it is approved and implemented. Hence W+>W? and V+>V?. When p?[??A/B?3,??A/B?2), reforms 1 and 3 are submitted and approved, with reform 1 always implemented and reform 3 implemented if A stays in power in the second period. Again, W+>W? and V+>V?. In contrast, when p???A/B?2, the welfare analysis is less straightforward. We summarize it in the following proposition:
113
~ Proposition A4.4: Let 0<?<?<1 and p???A/B?2. Define: (i) ? = [B?1+B?1?3??A]/[?B?2+B?2??(B?3+B?3)] , (ii) ?=[B?1+B?1]/[?B?2+B?2??(B?3+B?3)], (iii) ?C(p)=??A/{??A?? [B?1+B?1+p(B?2+B?2+B?3+B?3)]} Then: (1) if p>? and ?>?C(p), W+=W? and V+=V? (2) if p<? or ?<?C(p), then V+>V? (3) if p<? or ?<?C(p), then W+>W? if p<? and W+< W? if p>? Proof: Given p???A/B?2, A approves ?’s submitted reform program consisting of reforms 1, 2, and 3 if and only if ?(B?1+B?1+p(B?2+B?2+B?3+B?3))+(1??)3??A>0 ??[B?1+B?1+p(B?2+B?2+B?3+B?3)?3??A]>?3??A. If B?1+B?1+p(B?2+B?2+B?3+B?3)?3??A>0?p<?, then A approves the submission ~ for all ??(?,1). In contrast, if B?1+B?1+p(B?2+B?2+B?3+B?3)?3??A<0?p>?, then A approves the submission if and only if ?<?C(p). Hence, A does not approve any reforms when p>? and ?>?C(p), which together ~ with 0<?<?<1 implies that W+=W?=0 and V+=V?=0. If p<? or ?<?C(p), then A approves submitted reforms implying that V+>V? and W+=B?1+B?1+p(B?2+B?2+B?3+B?3)>(<)0=W??p<(>)?. ? Proposition A4.4 facilitates the construction of Figures A4.2?A4.4. First, note ~ that the definitions of ? and ? imply ?>?, where 0<?<1 because 0<?<1, but ? need not be ~ positive. Also, to construct ?C(p) note that it is strictly decreasing in p>?, ?C(1)=?, ?C(?)=1, and ?C(p)?? as p?? from above. Three configurations of the parameters are possible: (1) ?<?<??A/B?2, (2) ?<??A/B?2<?, (3) ??A/B?2<?<?.
114
The mapping of the conclusions of Proposition A4.4 into each of these regions is straightforward, leading to Figures A4.2?A4.4.145 Figures A4.1?A4.4 graphically summarize the conclusions of this appendix and show that the model with reform 1 has all the same qualitative predictions as the model in the main text.
145
Note that in Figures A4.2 and A4.3 we let ?>0 without loss of generality.
115
Figure A4.1: B?1+B?1+B?2+B?2+B?3+B?3>0
?
1
¯ ?
W+>W? V+>V? W+>W? V+<V?
??A/B?3 ??A/B?2 1 p?(?)
?
0
p
Figure A4.2: B?1+B?1+B?2+B?2+B?3+B?3<0, ?<?<??A/B?2
?
1 ~ ? ¯ ?
W+=W? V+=V?
W >W V+>V? W+>W? V+<V?
? ? ??A/B?3
+
?
W <W V+>V?
+
?
?C(p)
?
p*(?)
0
??A/B?2
1
p
116
Figure A4.3: B?1+B?1+B?2+B?2+B?3+B?3<0, ?<??A/B?2<?
?
1 ~ ? ¯ ?
W+=W? V+=V? W+<W? V+>V?
?C(p)
W+>W? V+>V?
?
W+>W? V+<V?
? ? ??A/B?2
1
p*(?)
0
??A/B?3
p
Figure A4.4: B?1+B?1+B?2+B?2+B?3+B?3<0, ??A/B?2<?<?
?
1 ~ ? ¯ ?
W+=W? V+=V? W <W V+>V?
+ ?
?C(p)
W+>W? V+>V?
?
W+>W? V+<V?
? ? ??A/B?3 ??A/B?2
1
p*(?)
0
p
117
Proof of Proposition A4.1. See Appendix 3 for reforms 2, 3, 4 and 5. Reform 1 is interesting for both interest groups. They will therefore act strategically: assuming that interest groups cannot collude, either of them would prefer not to lobby for reform 1 given that the other interest group does. There are 8 possible situations (sub-games) at the beginning of period 2: 1. If only ? lobbied for reform 1 in period 1 and A is in power in period 2 then the following payoff matrix can be used to find the (Nash) equilibrium of the subgame: ? Lobby Lobby B?1???A, B?1???A ? Do not lobby B?1, B?1???A In the equilibrium of this sub-game neither ? nor ? lobbies. Do not lobby B?1???A, B?1 B?1, B?1
2. If only ? lobbied for reform 1 in period 1 and B is in power in period 2: ? Lobby Do not lobby Lobby B?1???B, B?1 B?1???B, B?1???B ? 0,0 Do not lobby B?1, B?1???B In the equilibrium of this sub-game, only ? lobbies. 3. If only ? lobbied for reform 1 in period 1 and A is in power in period 2: ? Lobby Do not lobby Lobby B?1???A, B?1???A B?1???A, B?1 ? Do not lobby B?1, B?1???A B?1, B?1 In the equilibrium of this sub-game neither ? nor ? lobbies. 4. If only ? lobbied for reform 1 in period 1 and B is in power in period 2: ? Lobby Do not lobby Lobby B?1???B, B?1 B?1???B, B?1???B ? Do not lobby 0,0 B?1, B?1???B In the equilibrium of this sub-game, only ? lobbies. 5. If both interest groups lobbied for reform 1 in period 1 and that A is in power in period 2: ? Lobby Do not lobby Lobby B?1???A, B?1???A B?1???A, B?1 ? Do not lobby B?1, B?1 B?1, B?1???A In the equilibrium of this sub-game neither ? nor ? lobbies.
118
6. If both interest groups lobbied for reform 1 in period 1 and B is in power in period 2: ? Lobby Do not lobby Lobby B?1???B, B?1???B B?1???B, B?1 ? Do not lobby 0,0 B?1, B?1???B In the equilibrium of this sub-game, only ? lobbies. 7. If none of the interest groups lobbied for reform 1 in period 1 and politician A is in power in period 2: Lobby Lobby ???A, ???A ? Do not lobby 0, ???A In the equilibrium of this sub-game neither ? nor ? lobbies. ? Do not lobby ???A, 0 0,0
8. If none of the interest groups lobbied for reform 1 in period 1 and politician B is in power in period 2: ? Lobby Do not lobby Lobby ???B, ???B ???B, 0 ? Do not lobby 0,0 0, ???B In the equilibrium of this sub-game neither ? nor ? lobbies. In period 1, the incumbent politician is politician A. The following payoff matrix can be used to find the (sub-game perfect Nash) equilibrium of the lobbying game for reform 1: ? Lobby B?1???A, pB?1+(1?p)(B?1???B)???A B?1, pB?1+(1?p)(B?1???B)???A Do not lobby B?1???A, pB?1+(1?p)(B?1???B) 0, 0
Lobby ? Do not lobby
Therefore, in the equilibrium, ? lobbies in period 1 and ? does not. In period 2, neither ? nor ? lobby if A stays in power; if B comes to power, ? does not lobby and ? does. That is, no matter who is in power in period 2, reform 1 is always implemented. ?
119
Appendix 5: Algebraic Derivations and Proofs of Results in Chapter 2
This appendix includes the derivation of expressions and results in the text. The derivations are often algebraically intensive, but conceptually easy to follow. In most cases, it is therefore sufficient to merely indicate the key steps involved. Derivation of L*i = A, i?{SR,R}. At t = 1, the government chooses L*i for i?{SR,R} to maximize E{A(L*i+?+?Li,d)?½[(L*i+?+?Li,d)2+?i(?Li,d)2] + A(L*i+?+?Li,b)?½[(L*i+?+?Li,b)2+?i(?Li,b)2] + p(L*i+?+?Li,b)?½[(L*i+?+?Li,b)2+?i(?Li,b)2] ? p(L*i+?+?Li,d)+½[(L*i+?+?Li,d)2+?i(?Li,d)2]}, where E is the expectations operator, ?Li,d = [ri?(L*i+?)]/(1+?i), with rSR = p, and rR = A, and ?Li,b = [½(A+p)?(L*i+?)]/(1+?i). Differentiating with respect to L*i through the expectations operator, taking expectations, simplifying, and setting equal to zero gives the first-order condition 2[?i/(1+?i)](A?L*i) = 0. Note that L*i = A is the unique maximizer since ?2[?i/(1+?i)]<0 for all L*i. Derivation of the final realization of L under regime i?{R,SR}. L*i+?+?Li,b = A+?+[½(A+p)?(A+?)]/(1+?i) = [?i(A+?)+½(A+p)]/(1+?i). Derivation of Wi and Vi, i?{SR,R}. The expected social welfare under regime i?(SR,R}, Wi = A(p+c?A)??i/(1+?i)?2+¼[1/(1+?i)]?(3??2)(p?c)2,
120
is obtained by simplifying Wi = E{(p+c)(L*i+?+?Li,b)?(L*i+?+?Li,b)2??i(?Li,b)2}. Similarly, the expected government's payoffs under R and SR, VR = ½A2?½[?R/(1+?R)]?2+?[1/(1+?R)]?2(p?c)2 and VSR = ½A2?½[?SR/(1+?SR)]?2??[1/(1+?SR)]?2(p?c)2, are obtained by simplifying: Vi = E{A(L*i+?+?Li,d)?½[(L*i+?+?Li,d)2+?i(?Li,d)2] + ½{A(L*i+?+?Li,b)?½[(L*i+?+?Li,b)2+?i(?Li,b)2] + p(L*i+?+?Li,b)?½[(L*i+?+?Li,b)2+?i(?Li,b)2] ?A(L*i+?+?Li,d)?½[(L*i+?+?Li,d)2+?i(?Li,d)2] ? p(L*i+?+?Li,d)+½[(L*i+?+?Li,d)2+?i(?Li,d)2]}}, where L*i, ?Li,b, and ?Li,d for i?{R,SR} are defined above. Proof of Results 1 and 2. Results 1 and 2 are obtained by a straightforward comparison of WSR with WR (Result 1), and VSR with VR (Result 2), so we omit the algebraic proof altogether. Proof of Result 3. Result 3(a) readily follows from Figure 9(a). To prove Result 3(b), recall that VR = VSR when ?2/(p?c)2?¼[(4+3?R+?SR)/(?R??SR)]?2 = 0. The VR = VSR line intersects the ?2/(p?c)2 axis at (4+3?R+?SR)/(?R??SR) when ? = 1. Proof of Result 3(b) proceeds by showing that a "change in legal origin" from common law to civil law, rotates the VR = VSR line upward, as shown in Figure 9(b). Define f(?R,?SR) ? (4+3?R+?SR)/(?R??SR). Then, the total differential of f(?R,?SR), ?f, equals ?[(1+?SR)/(?R??SR)2]??R+[(1+?R)/(?R??SR)2]??SR. Now suppose that the change in 121
the ?i's is induced by the "change in legal origin" from common law to civil law. In notation, ??SR ? ?SR,f??SR,e, ??R ? ?R,f??R,e. Both ??SR and ??SR are positive by assumption. Then, ?f > 0 if (and only if) ?SR,f??SR,e > (?R,f??R,e)[(1+?SR,e)/(1+?R,e)] > 0. The latter expression is, however, implied by ?SR,f??SR,e ? ?R,f??R,e > 0, or, equivalently, by ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0. That is, when ?R,e ? ?SR,e ? ?R,f ? ?SR,f > 0, as Figure 9(b) indicates, the VR(?i,o) = VSR(?i,o) line rotates upward when ?i,o increases from ?i,e to ?i,f, i?{SR,R}.
122
Appendix 6: Civil Society as Vehicle for Service Provision
This appendix reviews and discusses some of the literature on CSOs as vehicles of service provision. This section has been relegated to the appendix since it does not immediately underpin the argument on civil society aid in chapter 3. The review below argues that, while progress has undoubtedly been made, the service-provision facet of civil society is still largely understudied. Azam (2003) provides a theoretical underpinning for civil society's role in alleviating poverty. His model is about the rich and the poor. While all rich care about the wellbeing of the poor, the rich are of two types: 'absolutists' and 'relativists'. The two types differ in their views about poverty (although the utility of both types is affected negatively by the number of the poor): the marginal willingness of the relativists to help the poor is increasing in their own consumption level, while absolutists' one is independent of theirs. The political system is such that public redistribution in favor of the poor (taxes) is chosen by the largest number of the rich, assumed to be absolutists. Given tax, the rich may also contribute an additional voluntary transfer. Only the relativists will be interested in contributing voluntarily to alleviate poverty. Azam views the aggregate value of voluntary transfers from the rich to the poor as an indicator of development of civil society. In his model, the minority of the rich selecting the level of voluntary contributions (the relativists) is assumed to be organized as a general assembly of NGOs (Azam thereby also assumes NGOs have developed some sort of mechanisms to overcome free-riding). In a later part of the paper, Azam also assumes that NGOs – the minority of the rich contributing voluntary transfers ? may initiate the redistribution
123
process by lobbying the government: the 'relativist' rich offer side payments to the government (formed by the 'absolutist' rich). Gibbons and Rutten (2004) argue that like the state, institutions of civil society (families and firms, clubs and communities, networks and neighborhoods) can in principle enforce rights by providing members with incentives both to respect rights and punish those who do not. Yet which does a better job supplying a particular public good, the state or the community? Furthermore, for societies facing more than one collectiveaction problem, how does the use of one institution to solve one problem affect the performance of various institutions that could be used to solve other problems (e.g. does having the government to enforce contracts affect the ability of the community to provide public goods) – when and how are institutions complements, substitutes or independent? Gibbons and Rutten stress the equilibrium approach to institutions: when there is no disinterested third party to enforce society's rules, institutions must be self-enforcing. They cast their model in variations of an N-person infinitely repeated prisoners' dilemma. In one of the variations, each period, first, the state chooses an amount to expend on a public good (at this stage, funding is taken from state's treasury, assumed to be arbitrarily large). Second, each citizen receives two benefits from the public good, one denominated in utility, the other denominated in money (only money can be expropriated by the state). Third, for each citizen, the state chooses how large a tax to collect; each citizen observes her own tax, but not any other citizen's tax. Fourth, each citizen decides whether to sound alarm. If at least one citizen sounds alarm, all citizens hear it. Fifth the citizens decide (as a unified body) whether to retract state's power, in which the individual who had been the state returns to being a citizen and anarchy ensues for all future periods.
124
Sixth, each citizen decides how much of her remaining resources to consume. In their model, the distinction between the state and the civil society is tantamount to citizens themselves contributing toward the public good rather than the state expanding resources on it in stage one of the game. Both Azam (2003) and Gibbons and Rutten (2004) are undoubtedly valuable contributions towards formulating economic theory of civil society. Both, however, leave room for improvement and unanswered questions. In Azam's model, as explained above, civil society is identified by the rich making voluntary transfers to the poor. The transfers are assumed to be implemented through an organized assembly of NGOs. Although an appealing concept, in his model, the general assembly of NGOs is never modeled. Moreover, Azam does not explain why the rich would even want to form such organizations. (Why would the rich not instead elaborate on the existing taxation system?) While Gibbons and Rutten (2004) provide a stronger underpinning of formation of civil society than Azam (2003), much like the latter, they do not provide any guidance on how and when, if at all, to promote civil society. Given the appeal of CSOs, an important normative question in development practice is when should the responsibility for providing services (e.g. social welfare and development) be delegated to CSOs? Besley and Ghatak (2001) apply the incomplete contracts paradigm to consider ownership issues (state vs. nongovernmental organization) for public goods. They argue that the party with the highest valuation of the project should be the owner regardless of the relative importance of the investments or other aspects of the production technology. The welfare implications of their model, however, are not clear-cut, and are dependant on the political economy of the setting that they do
125
not model. Jack (2001) explicitly casts his analysis in a development context, where a poorly functioning state is often a reality. If direct government provision of services is ineffectual or very inefficient, when is contracting out to a CSO preferred to using a traditional for-profit firm? Building on Glaeser and Shleifer (1998), Jack argues that CSO provision is preferred if the gross benefits of higher quality outweigh the increased financing costs. A non-profit entrepreneur is constrained in using her revenues. She must hence be paid more for service provision to induce the same levels of effort as a forprofit entrepreneur. Furthermore, if CSOs are to serve as important vehicles for delivery of foreign assistance to poor countries, who should implement a project: an international CSO or a grass-root CSO? Building on Besley and Ghatak (2001), Jack concludes that when donor and government valuations of the service do not differ much, provision through an international CSO yields higher joint surplus than under grass roots provision. One limitation of Jack’s analysis is that he models the CSOs as passive transformers of inputs into output. The literature on service provision role of CSOs is also related to the literature on public-private partnerships. See e.g. Bentz et al. (2004) and references therein.
126
Appendix 7: Welfare Consequences of Delaying Reform Implementation When the Lobbying Group's Favored Reform Program Increases Social Welfare
This appendix analyzes the scenario when the preferred reform program of the interest group aligned with the ruling political option – although not first best for society – increases social welfare. Intuitively, in this case, one would expect that existence of civil society to constrain the elites' behavior is less warranted. Suppose A2 as stated in chapter 1 does not hold. That is, suppose ?B?2+B?2?<B?3+B?3, or equivalently, B?2+B?2+B?3+B?3>0. No other assumptions of the model have been changed. Observe, first, that the politician in power, A, now accepts the reform program consisting of reforms 2 and 3 for any value of ??(0,1): ?(B?2+B?2+B?3+B?3)+(1??)2??A>0 for all ??(0,1). Without civil society's involvement, A implements ?'s proposed reform program readily and immediately. Hence, W? = B?2+B?2+B?3+B?3>0 and V? = ?(B?2+B?2+B?3+B?3)+(1??)2??A. (Recall that W denotes social welfare and V politician's payoff. Superscripts ? and + indicate nonexistence and existence of civil society, respectively.) When civil society exists, its participation opens the reform debate to public and delays implementation of reform. Adopting the analysis of chapter 1, when p<??A/B?3, ? does not submit anything. V+ = 0 < V?, W+ = 0 < W?. When p???A/B?2, ? submits a reform program consisting of reforms 2 and 3 and A accepts it: V+ = p?(B?2+B?2+B?3+B?3)+(1??)2??A > 0, but clearly V+ < V?, and likewise, W+ = p(B?2+B?2+B?3+B?3) < W?. Finally, when ??A/B?3?p<??A/B?2, ?'s submitted reform program consists of reform 3 only. V+ = p?(B?3+B?3)+(1??)??A>0, so A accepts it. Hence, V+>V? if and only if
127
p?(B?3+B?3)+(1??)??A>?(B?2+B?2+B?3+B?3)+(1??)2??A, or equivalently, if and only if p>p*(?), where p*(?) is defined in Proposition 3 in chapter 1. Likewise, W+ = p(B?3+B?3), and W+ > W? if and only if p(B?3+B?3)>B?2+B?2+B?3+B?3, or equivalently, if and only if p>pC, where pC?1??B?2+B?2?/(B?3+B?3)?(0,1). Note that in this case we cannot compare social welfare with and without civil society's involvement in reform process without imposing additional assumptions on the size of pC. (If pC<??A/B?3, then W+ > W?. If pC?[??A/B?3,??A/B?2), then W+ > W? for p>pC and W+ < W? for p<pC. If pC???A/B?2, W+ < W?.) When society's net gains and losses from reforms 3 and 2 are roughly equal, for example, ?B?2+B?2? ? B?3+B?3, pC is close to 0, and civil society's involvement is desirable from society's standpoint.
128
References
Abel, Richard L. "Revisioning Lawyers." In Richard L. Abel and Philip S. C. Lewis ed., Lawyers in Society: An Overview, Berkley and Los Angeles: University of California Press, 1995. Acemoglu, Daron and James A. Robinson. 2000. "Why Did The West Extend The Franchise? Democracy, Inequality, And Growth In Historical Perspective." Quarterly Journal of Economics. 115:4, pp. 1167?99. Acemoglu, Daron, Simon Johnson and James A. Robinson. 2001. "The Colonial Origins of Comparative Development: An Empirical Investigation." American Economic Review. December, 91:5, pp. 1369?1401. Åmark, Klas. "Open Cartels and Social Closures: Professional Strategies in Sweden, 1860?1950." In Michael Burrage and Rolf Torstendahl (eds.) Professions in Theory and History: Rethinking the Study of the Professions, London, SAGE Publications, 1990. Ashby, Simon, Swee-Hoon Chuah, and Robert Hoffmann. 2004. "Industry SelfRegulation: A Game-Theoretic Typology of Strategic Voluntary Compliance." International Journal of the Economics of Business. February, 11:1, pp. 91-106(16). Australian Law Reform Commission, 2003. http://www.alrc.gov.au/about/index.htm Ayres, Ian and John Braithwaite. Responsive Regulation: Transcending the Deregulation Debate. Oxford University Press, 1995. Azam, Jean-Paul 2003 "A Theory of Poverty Aversion and Civil Society Development", Economics and Politics, 15:1, pp. 61-84. Baldwin, Robert and Martin Cave. Understanding Regulation: Theory, Strategy, and Practice. Oxford University Press, 1999. Barro, Robert J. and Rachel M. McLeary. 2003. "Religion and Economic Growth across Countries." American Sociological Review. Vol. 68, pp. 760-81. Beck, Thorsten and L. Laeven. 2005. “Institutions and Growth in Transition Economics.” Working paper. Beck, Thorsten, Asli Demirguc-Kunt, and Ross Levine, “Law, Endowments, and Finance,” Journal of Financial Economics, 70:137-181, 2003a. Beck, Thorsten, Asli Demirguc-Kunt, and Ross Levine. 2003b. "Law and Finance: Why Does Legal Origin Matter?" Journal of Comparative Economics. 31:4, pp. 653?75. Becker, Gary, “A Theory of Competition Among Pressure Groups for Political Influence,” Quarterly Journal of Economics. 98, 3:371-400, Aug. 1983.
129
Bell, David A. Lawyers and Citizens. The Making of a Political Elite in Old Regime France. New York: Oxford University Press, 1994. Bell, David A. "Barristers, Politics, and the Failure of Civil Society in Old Regime France." In Terence C. Halliday and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism: Europe and North America from the Eighteenth to Twentieth Century, Oxford: Clarendon Press, 1997. Bentz Andreas, Paul A. Grout and Maija L. Halonen. 2004. "What Should Governments Buy from the Private Sector – Assets or Services?" Working paper. Berkowitz, Daniel, Katharina Pistor, and Jean-Francois Richard, “Economic Development, Legality, and the Transplant Effect,” European Economic Review. 47, 1:165-195, Feb. 2003. Besley, Timothy and Maitreesh Ghatak. 2001. "Government and Public Provision of Public Goods". Quarterly Journal of Economics,, November. pp. 1343-72. Brand, Paul. The Origins of the English Legal Profession. Oxford: Blackwell Publishers, 1992. Burrage, Michael and Rolf Torstendahl (eds.) Professions in Theory and History. Rethinking the Study of the Professions, London: SAGE Publications, 1990. Burrage, Michael and Rolf Torstendahl (eds.) The Formation of Professions. Knowledge, State and Strategy, London: SAGE Publications, 1990. Burrage, Michael, Konrad Jarausch and Hannes Siegrist. "An Actor-Based Framework for the Study of the Professions." In Michael Burrage and Rolf Torstendahl (eds.) Professions in Theory and History: Rethinking the Study of the Professions, London: SAGE Publications, 1990. Burrage, Michael. "Mrs Thatcher Against the ‘Little Republics’: Ideology, Precedents, and Reactions." In Terence C. Halliday and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism: Europe and North America from the Eighteenth to Twentieth Century, Oxford: Clarendon Press, 1997. Burrage, Michael. "Revolution as a Starting Point for the Comparative Analysis of the French, American, and English Legal Profession." In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society, Volume Three, Comparative Theories, Berkley and Los Angeles: University of California Press, 1989. Cadiet, Loïc, "Civil Justice Reform: Access, Cost, and Delay. The French Perspective" In Adrian Zuckerman ed. Civil Justice in Crisis: Comparative Perspectives of Civil Procedure. Oxford, Oxford, 1999.
130
Campos, Nauro F. 1999. "Context is Everything: Measuring Institutional Change in Transition Economies." World Bank Working Paper No. 2269. Carothers, Thomas and Marina Ottaway. "The Burgeoning World of Civil Society Aid." In Marina Ottaway and Thomas Carothers (eds.). Funding Virtue. Civil Society Aid and Democracy Promotion. Washington, D.C.: Carnegie Endowment for International Peace, 2000. Carothers, Thomas. "Western Civil Society Aid to Eastern Europe and the Former Soviet Union." In Thomas Carothers. Critical Mission. Essays on Democracy Promotion, Washington, D.C.: Carnegie Endowment for International Peace, 2004. Carothers, Thomas. 1999. "Western Civil-Society Aid to Eastern Europe and the Former Soviet Union." East European Constitutional Review. Fall, 8:4. http://www.law.nyu.edu/eecr/vol8num4/special/western.html. Carothers, Thomas. Critical Mission. Essays on Democracy Promotion. Washington, D.C.: Carnegie Endowment for International Peace, 2004. Carruthers, Bruce, and Terence Halliday. "Professionals in Systemic Reform of Bankruptcy Law: The 1978 U.S. Bankruptcy Code and the English Insolvency Act 1986" American Bankruptcy Law Journal, Winter, 2000, 74, pp. 35-75. Cleaves, Peter S. Professions and the State: The Mexican Case. Tucson: The University of Arizona Press, 1987. Che, Jiahua and Yingyi Qian. 1998. "Insecure Property Rights and Government Ownership of Firms." Quarterly Journal of Economics. May, 113:2, pp. 467?96. Cohen, Jean L. and Andrew Arato. Civil Society and Political Theory. MIT Press, 1994. Collins, Randall. "Changing Conceptions in the Sociology of Professions." In Rolf Torstendahl and Michael Burrage (eds.) The Formation of Professions: Knowledge, State and Strategy, London: SAGE Publications, 1990. Datta, Samar K. and Jeffrey B. Nugent. 1986. "Adversary Activities and Per Capita Income Growth." World Development, 14. David, René and Henry de Vries, 1958. The French Legal System: An Introduction to Civil Law Systems. Oceana Publications, NY. De Grand, Alexander J. Fascist Italy and Nazi Germany. The 'fascist' style of rule. Routledge, London and New York, 1995. DeMarzo, Peter M., Michael J. Fishman, and Kathleen M. Hagerty. 2001. "Contracting and Enforcement with a Self-Regulatory Organization." http://papers.ssrn.com/sol3/papers.cfm?abstract_id=297302
131
Dethier, Jean Jacques, Hafez Ghanem and Edda Zoli. 1999. “Does Democracy Facilitate the Economic Transition? An Empirical Study of Central and Eastern Europe and the Former Soviet Union.” World Bank Working Paper No. 2194. Dewatripont, M. and E. Maskin. 1995. "Credit and Efficiency in Centralized and Decentralized Economies." Review of Economic Studies. October, 62:4, pp. 541?55. Dewatripont, Mathias and Gerard Roland. 1995. "The Design of Reform Packages under Uncertainty." American Economic Review, December, 85:5, pp. 1207?23. Djankov, S., E. Glaeser, R. La Porta, F. Lopez-de-Silanes and A. Shleifer. 2003. "The New Comparative Economics." Journal of Comparative Economics. December, 31:4, pp. 595?619. Donabedian, Bairj. 1995. "Self-Regulation and the Enforcement of Professional Codes." Public Choice. 85:1?2, pp. 107-18. Drazen, Allan. Political Economy in Macroeconomics. Princeton: Princeton University Press, 2000. Ehrenberg, John. Civil Society. The Critical History of an Idea. New York University Press, 1999. Eisner, Marc Allen. Regulatory Politics in Transition, Second Edition, Johns Hopkins University Press, Baltimore, 2000. Estache, Antonio and David Martimort. 1999. "Politics, Transacstion Costs, and the Design of Regulatory Institutions." World Bank working paper 2073. www.worldbank.org/html/dec/Publications/ Workpapers/wps2000series/wps2073. Field, G. Lowell. The Syndical and Corporative Institutions of Italian Fascism. Columbia University Studies in the Social Sciences 433, AMS Press, New York, 1968. Florinsky, Michael T. Fascism and National Socialism. A Study of the Economic and Social Policies of the Totalitarian State. The Macmillan Company, New York, 1938. Fukuyama, Francis. Trust: The Social Virtues and the Creation of Prosperity. Free Press, 1996. Gandhi, J.S. "Past and Present. A Sociological Portrait of the Indian Legal Profession." In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society, Volume One, The Common Law World, Berkley and Los Angeles: University of California Press, 1988. Gehrig, Thomas and Peter-J Jost. 1995. "Quacks, Lemons, and Self Regulation: A Welfare Analysis." Journal of Regulatory Economics. 7:3, pp. 309?25.
132
Gibbons, Robert and Andrew Rutten. 2004. "Institutional Interactions: An Equilibrium Approach to the State and Civil Society". Working paper. Glaeser, Edward L. and Andrei Shleifer. 1998. “Not-for-profit entrepreneurs.” NBER Working Paper No. 6810. Glaeser, Edward L. and Andrei Shleifer. 2002. "Legal Origins." Quarterly Journal of Economics. 117:4, pp. 1193?229. Glaeser, Edward L. and Andrei Shleifer. 2003. "The Rise of the Regulatory State." Journal of Economic Literature. June, 41:2, pp. 401?25. Glaeser, Edward L., David Laibson and Bruce Sacerdote. 2000. "The Economic Approach to Social Capital." NBER Working Paper 7728. Glaeser, Edward, Simon Johnson, and Andrei Shleifer. 2001. "Coase vs. Coasians." Quarterly Journal of Economics. August, 116:3, pp. 853?99. Goldberg, Victor P. 1976. "Regulation and Administered Contracts." Bell Journal of Economics. Autumn, 7:2, pp. 426?48. Gostynski, Zbigniew and Alan Garfield. "Taking the Other Road: Polish Legal Education During the Past Thirty Years." Temple International and Comparative Law Journal. Fall, 1993. Greif, Avner. 1993. "Contract Enforceability and Economic Institutions in Early Trade: The Maghribi Traders' Coalition." American Economic Review. June, 83:3, pp. 525?48. Greif, Avner. 1994. "Cultural Beliefs and the Organization of Society: A Historical and Theoretical Reflection on Collectivist and Individualist Societies." Journal of Political Economy, Vol. 102, No. 5, pp. 912-50. Grossman, Gene M. and Elhanan Helpman. 1994. "Protection for Sale." American Economic Review. September, 84:4, pp. 833?50. Grossman, Gene M. and Elhanan Helpman. Special Interest Politics. Cambridge: MIT Press, 2001. Grossman, Sanford and Oliver Hart. 1986. "The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration." Journal of Political Economy. 94:4, pp. 691?719. Guiso, Luigi, Paola Sapienza and Luigi Zingales. 2004. "The Role of Social Capital in Financial Development." American Economic Review. June, 94:3, pp. 526-56. Hall, John A. Civil Society: Theory, History, Comparison. Polity Press, 1995.
133
Halliday, Terence C. and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism. Europe and North America from the Eighteenth to Twentieth Century. Oxford: Clarendon Press, 1997. Halliday, Terence C. and Lucien Karpik. "Postscript: Lawyers, Political Liberalism, and Globalization." In Terence C. Halliday and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism: Europe and North America from the Eighteenth to Twentieth Century, Oxford: Clarendon Press, 1997. Hasebe, Yukiko, "Civil Justice Reform: Access, Cost, and Expedition. The Japanese Perspective" In Adrian Zuckerman ed. Civil Justice in Crisis: Comparative Perspectives of Civil Procedure. Oxford, Oxford, 1999. Hayek, F. A. The Constitution of Liberty. University of Chicago Press, 1960. Hoff, Karla and Joseph E. Stiglitz. 2004. “After Big Bang? Obstacles to the Emergence of the Rule of Law in Post-Communist Societies.” American EconomicReview. June, 94:3, pp. 753-63. Hofstadter, Richard. The Age of Reform: From Bryan to F.D.R. Vintage Books USA, 1956. Huang, Fali. 2003. "Social Trust, Cooperation, and Human Capital." Working paper. Huskey, Eugene. "The Limits to Institutional Autonomy in the Soviet Union: The Case of the Advokatura" Soviet Studies, Vol. 34, No. 2. Apr., 1982, pp. 200-227. Huyse, Luc. "Legal Experts in Belgium." In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society: An Overview, Berkley and Los Angeles: University of California Press, 1995. International Center for Alcohol Policies. 2001. ICAP Reports 9, January 2001. "SelfRegulation of Beverage Alcohol Advertising." http://www.icap.org/publications/report9.html accessed 9/14/2004. Jack, William. 2001. "Public Policy toward Non-Governmental Organizations in Developing Countries", World Bank Working Paper No. 2639. Jarausch, Konrad H. The Unfree Professions. German Lawyers, Teachers, and Engineers, 1900?1950. New York, Oxford: Oxford University Press, 1990. Johnson, Terence. "Imperialism and the Professions: Notes on the Development of Professional Occupations in Britain’s Colonies and the New States." In Paul Halmos (ed.) The Sociological Review Monograph 20, Professionalization and Social Change, Keele University, December 1973.
134
Jordan, Pamela. 1998. "The Russian Advokatura (Bar) and the State in the 1990s." Europe-Asia Studies. July, 50:5, pp, 765?91. Karatnycky, A., A. Motly and B. Shor. Nations in Transit 1997: Civil Society, Democracy and Markets in East Central Europe and the Newly Independent States, New Brunswick: Transaction Publishers, 1998. Karpik, Lucien. 1988. "Lawyers and Politics in France, 1814?1950: The State, the Market and the Public." Law & Social Inquiry. Fall, 13:4, pp. 707?36. Karpik, Lucien. French Lawyers. A Study in Collective Action 1274 to 1994. Oxford: Clarendon Press, 1999. Knack, Stephen and Phillip Keefer. 2000. "Does Social Capital Have an Economic Payoff: A Cross-Country Investigation." Quarterly Journal of Economics. November, 112:4, pp. 1251-88. Kocka, Jürgen. "‘Bürgertum’ and Professions in the 19th Century: Two Alternative Approaches." In Michael Burrage and Rolf Torstendahl (eds.) Professions in Theory and History: Rethinking the Study of the Professions, London: SAGE Publications, 1990. Kranton, Rachel E. 2003. "Competition and the Incentive to Produce High Quality." Economica. 70:279, pp. 385?404. Kranton, Rachel E. and Anand V. Swamy. 1999. "The Hazards of Piecemeal Reform: British Civil Courts and the Credit Market in Colonial India." Journal of Development Economics. 58:1, pp. 1?24. Krause, Elliott A. "Professions and the State in the Soviet Union and Eastern Europe: Theoretical Issues." In Anthony Jones (ed.) Professions and the State: Expertise and Autonomy in the Soviet Union and Eastern Europe, Philadelphia: Temple University Press, 1991. La Porta, R., F Lopez-de-Silanes, Pop-Eleches, C. and A. Shleifer “Judicial Checks and Balances” Journal of Political Economy, forthcoming 2004. La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. Vishny. 1999. "The Quality of Government." Journal of Law, Economics and Organization. 15:1, pp. 222?79. La Porta, R., F. Lopez-de-Silanes, A. Shleifer and R. Vishny. 1998. "Law and Finance." Journal of Political Economy. December, 106:6, pp. 1113?55. La Porta, R., F. Lopez-de-Silanes, A. Shleifer and R. Vishny. 1997. "Legal Determinants of External Finance." Journal of Finance. 52:3, pp. 1131?50.
135
Laffont, Jean-Jacques. Incentives and Political Economy. Clarendon Lectures in Economics, Oxford University Press, 2002. Landon, Michael. The Triumph of the lawyers: Their Role in English Politics, 1678?1689. University of Alabama Press, 1969. Leland, Hayne E. 1979. "Quacks, Lemons, and Licensing: A Theory of Minimum Quality Standards." Journal of Political Economy, December, 87:6, pp. 1328?46. Lülfesmann, Christoph. 2002. "Central Governance or Subsidiarity: A Property-Rights Approach to Federalism." European Economic Review. September, 46:8, pp. 1379?97. Lyon, Thomas and Haizhou Huang. 2002. "Legal Remedies for Breach of the Regulatory 'Contract'." Journal of Regulatory Economics. 22:2, pp. 107?32. Lyttelton, Adrian. "Italian Fascism." In Walter Laqueur (ed.) Fascism: A Reader’s Guide. University of California Press, 1976. Maier, Charles S. In Search of Stability: Explorations in Historical Political Economy. Cambridge Studies in Modern Political Economies, Cambridge University Press, 1987. Manitoba Law Reform Commission, 2003. http://www.gov.mb.ca/justice/mlrc Maxwell, John, Thomas Lyon, and Steven Hackett. 2000. "Self-Regulation and Social Welfare: The Political Economy of Corporate Environmentalism." Journal of Law and Economics. 43:2, pp. 583-617. McCain, Morris. "A Soviet Lawyers in the Reform Debate: Cohesion and Efficacy." Soviet Studies, Vol. 34, No. 1., Jan., 1982, pp. 3-22. McGerr, Michael. A Fierce Discontent : The Rise and Fall of the Progressive Movement in America, 1870-1920. Free Press, 2003. Milward, Alan S. "Fascism and the Economy." In Walter Laqueur (ed.) Fascism: A Reader’s Guide. University of California Press, 1976. Mueller, Dennis. Public Choice II. Cambridge University Press, 1989 Mulligan, Casey B. and Andrei Shleifer. "Population and Regulation." NBER Working Paper 10234, 2004. Murphy, Kevin M., Andrei Shleifer and Robert W. Vishny. 1991. "The Allocation of Talent: Implications for Growth." Quarterly Journal of Economics. 106 .
136
North, Douglass C. and Barry Weingast, "Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth-Century England." Journal of Economic History, December, 1989. North, Douglass C. 1991. "Institutions." Journal of Economic Perspectives. Vol. 5, No. 1., pp. 97-112. North, Douglass C., Institutions, Institutional Change, and Economic Performance, Cambridge University Press, 1990. Nunez, Javier. 2001. "A Model of Self-Regulation." Economics Letters. 74:1, pp. 91?7. Odenyo, Amos O. 1979. "Professionalization Amidst Change: The Case of Emerging Legal Profession in Kenya." African Studies Review. December, 22:3, pp. 33?44. Ogus, Anthony. "Self-Regulation." In Boudewijn Bouckaert and Gerrit De Geest (eds) Encyclopedia of Law and Economics, 1999. http://encyclo.findlaw.com. Olson, Mancur. Logic of Collective Action: Public Goods and the Theory of Groups. Revised edition. Harvard Economic Studies, Harvard University Press, 1971. Olson, Mancur. The Rise and Decline of Nations. Yale University Press, 1982. Orenstein, Mitchell A. 2000. "How Politics and Institutions Affect Pension Reform in Three Postcommunist Countries." World Bank Working Paper No. 2310. Ottaway, Marina and Thomas Carothers (eds.). Funding Virtue. Civil Society Aid and Democracy Promotion. Washington, D.C.: Carnegie Endowment for International Peace, 2000. Pei, Minxin. "Does Legal Reform Protect Economic Transactions? Commercial Disputes in China." In Peter Murrell (ed.) Assessing the Value of Law in Transition Economies, Ann Arbor: University of Michigan Press, 2001. Persson, T. 1998. "Economic Policy and Special Interest Politics." Economic Journal. March, 108, pp. 310?27. Persson, Torsten and Guido Tabellini. Political Economics: Explaining Economic Policy. MIT Press, Cambridge, Massachusetts, 2000. Plucknett, T. F. T. Studies in English Legal History. London: The Hambledon Press, 1983. Poole, Keith T. and Thomas Romer. 1985. "Patterns of Political Action Committee Contributions to the 1980 Campaigns for the U.S. House of Representatives." Public Choice. 47:1, pp. 63–112.
137
Poole, Keith T., Thomas Romer and Howard Rosenthal. 1987. "The Revealed Preferences of Political Action Committees." American Economic Review. May, 77:2, pp. 298–302. Posner, Richard A. Overcoming Law. Cambridge: Harvard University Press, 1995. Priest, Margot. 1997. "The Privatization of Regulation: Five Models of Self-Regulation." Ottawa Law Review. 29, pp. 236?314 Putnam, Robert. Making Democracy Work. Civic Traditions in Modern Italy. Princeton University Press, 1993. Qian, Yingyi. 1994. "A Theory of Shortage in Socialist Economies Based on the ‘Soft Budget Constraint’." American Economic Review. March, 84:1, pp. 145?56. Quigley, Kevin F.F. "Lofty Goals, Modest Results: Assisting Civil Society in Eastern Europe." In Marina Ottaway and Thomas Carothers (eds.). Funding Virtue. Civil Society Aid and Democracy Promotion. Washington, D.C.: Carnegie Endowment for International Peace, 2000. Rajan, Raghuram and Luigi Zingales. 2003. "The Great Reversals: The Politics of Financial Development in the 20th Century." Journal of Financial Economics. 69, pp. 5?50. Rich, William A. "Poland’s new collateral law." Law In Transition, Summer 1997. Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi "Institutions Rule: The Primacy Of Institutions Over Geography And Integration In Economic Development" NBER Working Paper, 2002. Rokumoto, Kahei. "The Present State of Japanese Practicing Attorneys: On the Way to Full Professionalization?" In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society: An Overview, Berkley and Los Angeles: University of California Press, 1995. Roland, Gerard. 2004. “Understanding Institutional Change: Fast-Moving and SlowMoving Institutions”. Studies in Comparative International Development. Roland, Gerard. Transition and Economics: Politics, Markets, and Firms. MIT Press, 2000. Rose, Jonathan. 1998. “The Legal Profession in Medieval England: A History of Regulation.” Syracuse Law Review. 48:1. Ross, Stanley D. 1992. "The Rule of Law and Lawyers in Kenya." The Journal of Modern African Studies. September, 30:3, pp. 421?42.
138
Rueschemeyer, Dietrich. "State, Capitalism, and the Organization of Legal Counsel: Examining an Extreme Case ? the Prussian Bar, 1700?1914." In Terence C. Halliday and Lucien Karpik (eds.) Lawyers and the Rise of Western Political Liberalism: Europe and North America from the Eighteenth to Twentieth Century, Oxford: Clarendon Press, 1997. Salomon, Lester M., Wojciech Sokolowski and Associates. Global Civil Society, Dimensions of the Nonprofit Sector, Vol. 2. Kumarian Press, 2004. Shaked, Avner and John Sutton. 1981. "The Self-Regulating Profession." Review of Economic Studies. April, 48:2, pp. 217?34. Shapiro, Isaac "The Soviet Bar-Past and Present." Russian Review, Vol. 20, No. 2., Apr., 1961, pp. 143-150. Shelley, Louise I. "Lawyers in Soviet Union." In Anthony Jones (ed.) Professions and the State: Expertise and Autonomy in the Soviet Union and Eastern Europe, Philadelphia: Temple University Press, 1991. Shleifer, Andrei and Robert W. Vishny. 1994. "Politicians and Firms." Quarterly Journal of Economics. November, 109:4, pp. 995?1025. Shleifer, Andrei and Robert W. Vishny. The Grabbing Hand: Government Pathologies and Their Cures. Harvard University Press, 1998. Shonfield, Andrew. Modern Capitalism. The Changing Balance of Public and Private Power. Oxford University Press, 1974. Siegrist, Hannes. "Professionalization as a Process: Patterns, Progression and Discontinuity." In Michael Burrage and Rolf Torstendahl (eds.) Professions in Theory and History: Rethinking the Study of the Professions, London, SAGE Publications, 1990. Sternhell, Zeev."Fascist Ideology." In Walter Laqueur (ed.) Fascism: A Reader’s Guide. University of California Press, 1976. Stigler, George. "The Theory of Economic Regulation." Bell Journal. Spring 1971, 2, 137-146. Stratmann, Thomas. 1991. "What Do Campaign Contributions Buy? Deciphering Causal Effects of Money and Votes." Southern Economic Journal. January, 57:3, pp. 606–20. Stratmann, Thomas. 1995. "Campaign Contributions and Congressional Voting: Does the Timing of Contributions Matter?" Review of Economics and Statistics. February, 77:1, pp. 127–36.
139
Stratmann, Thomas. 2002. "Can Special Interests Buy Congressional Votes? Evidence from Financial Services Legislation." Journal of Law and Economics, forthcoming. Temple, Jonathan and Paul A. Johnson. 1998. "Social Capability and Economic Growth." Quarterly Journal of Economics. August: 113:3, pp. 965-90. Tirole, Jean. 1994. "The Internal Organization of Government." Oxford Economic Papers, January, 46:1, pp. 1?29. Tocqueville, Alexis De. The Old Regime and the French Revolution. Garden City, New York: Doubleday Anchor Books, 1955. Trevelyan, G.M. The English Revolution, 1688?1689. New York: Oxford University Press, 1967. Wallis, John Joseph. 1987. "Employment, Politics, and Economic Recovery During the Great Depression." Review of Economics and Statistics, August, 69, pp. 516?20. Wallis, John Joseph. 1998. "The Political Economy of New Deal Spending Revisited, Again: With and Without Nevada." Explorations in Economic History. April, 35, pp. 140?70. Waters, Christopher P.M. Counsel in the Caucasus: Professionalization and Law in Georgia. The Hague/London/New York: Martinus Nijhoff Publishers, forthcoming, 2004. Weisbrot, David. "The Australian Legal Profession. From Provincial Family Firms to Multinationals." In Richard L. Abel and Philip S. C. Lewis (eds.) Lawyers in Society, Volume One, The Common Law World, Berkley and Los Angeles: University of California Press, California, 1988. Williamson, Oliver E. 1976. "Franchise Bidding for Natural Monopolies-in General and with Respect to CATV." Bell Journal of Economics. Spring, 7:1, pp. 73?104. Working Together. World Bank-Civil Society Relations. World Bank 2003. Zimmerman, Richard. 1996. "An Introduction to German Legal Culture." In Werner F. Ebke and Matthew W. Finkin, eds., Introduction to German Law. Kluwer, The Hague. Zweigert, Konrad and Hein Kötz. Introduction to Comparative Law. Second revised edition. Oxford University Press, 1992.
140
doc_840689689.pdf