Judicial Power and Strategic Communication in Mexico

Judicial Power and Strategic Communication in Mexico Although they are not directly accountable to voters, constitutio...

Author: Jeffrey K. Staton

11 downloads 1095 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Judicial Power and Strategic Communication in Mexico

Although they are not directly accountable to voters, constitutional court judges around the world nevertheless communicate with the general public through the media. In Judicial Power and Strategic Communication in Mexico, Jeffrey K. Staton argues that constitutional courts develop public relations strategies to increase the transparency of judicial behavior and promote judicial legitimacy – two conditions that are favorable for the exercise of independent judicial power. Yet, in some political contexts, there can be a tension between transparency and legitimacy, and for this reason, courts cannot necessarily advance both conditions simultaneously. The argument is tested via an analysis of the Mexican Supreme Court during Mexico’s recent transition to democracy and also through a cross-national analysis of public perceptions of judicial legitimacy. The results demonstrate that judges can be active participants in the construction of their own power. More broadly, the study develops a positive political theory of institutions, which highlights the connections between democratization and the rule of law. Jeffrey K. Staton is Assistant Professor of Political Science at Emory University. His research has appeared in the American Journal of Political Science, the Journal of Politics, Comparative Politics, Political Research Quarterly, and International Studies Quarterly. Professor Staton was previously Assistant Professor of Political Science at Florida State University and postdoctoral Fellow at the Center for U.S. Mexican Studies at the University of California, San Diego, and at the New York University School of Law.

Judicial Power and Strategic Communication in Mexico

JEFFREY K. STATON Emory University

cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Dubai, Tokyo Cambridge University Press 32 Avenue of the Americas, New York, NY 10013-2473, USA www.cambridge.org Information on this title: www.cambridge.org/9780521195218 © Jeffrey K. Staton 2010 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2010 Printed in the United States of America A catalog record for this publication is available from the British Library. Library of Congress Cataloging in Publication data Staton, Jeffrey K. Judicial power and strategic communication in Mexico / Jeffrey K. Staton. p. cm. Includes bibliographical references and index. ISBN 978-0-521-19521-8 (hardback) 1. Judicial power – Mexico. 2. Constitutional law – Mexico. 3. Law reform – Mexico. 4. Strategic Communications. 5. Transparency in government – Mexico. I. Title. KGF3156.S73 2010 2009051395 347.72 012–dc22 ISBN 978-0-521-19521-8 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party Internet Web sites referred to in this publication and does not guarantee that any content on such Web sites is, or will remain, accurate or appropriate.

In memory of Nadeem Siddiqui

Contents

List of Tables List of Figures Acknowledgments

page ix xi xiii

part i judicial communication and judicial power 1

Introduction The Puzzle Judicial Power A Fundamental Problem of Judicial Policymaking Theoretical Solutions Argument Summary Approach of the Book Chapter Summaries

2

A Model of Constitutional Review and Case Promotion A Public Enforcement Mechanism Models of Constitutional Review Baseline Model Promotion Model Implications Appendix 2

3 5 8 9 11 14 17 19 22 22 26 27 33 42 44

part ii the politics of constitutional review in mexico 3

Public Relations on the Mexican Supreme Court Legitimacy and Transparency Problems Public Relations Tactics Transparency and Legitimacy

53 54 57 62 vii

Contents

viii

4

Decisions, Case Promotion, and Compliance Decisional and Case Promotion Data A Final Note on Research Design Public Authority Compliance The Supreme Court in Mexican Politics Appendix 4A Appendix 4B

part iii relationships between transparency and legitimacy 5 Constitutional Review and the Development of Judicial Legitimacy An Extended Model of Constitutional Review Modeling Awareness A Different Type of Partisan Court Conclusion Appendix 5A

6

A Cross-National Analysis of Judicial Legitimacy Hypotheses and Data Analysis A Tension between Transparency and Legitimacy The Development of Judicial Power

7

Democratic States and the Development of Judicial Power Democracy and the Public Enforcement Mechanism Limits on Judicial Power under Democracy Additional Research

References Index

65 68 79 90 100 101 114

127 132 141 148 152 153 167 167 172 186 190 196 197 199 202 205 219

List of Tables

1.1

Public Relations Summary for Constitutional Courts or High Courts with Constitutional Jurisdiction in Latin America page 5 2.1 Observable Implications by Model Type 41 3.1 Social Communication Analysis of Supreme Court Print Coverage, October 1999 60 4.1 Summary of Supreme Court Constitutional Docket, 1997–2002 69 4.2 Print Media Coverage of Supreme Court Constitutional Decisions (by prior coverage) 76 4.3 Determinants of Supreme Court Decisions to Strike Policy and Promote Cases 80 4.4 Predicted Probability of La Jornada Coverage 86 4.5 Predicted Probability of Press Release by Case Outcome and Prior Media Coverage 87 4.6 Type, Promotion, and Coverage of Incidents of Noncompliance 94 4.7 Complainants and Responsible Authorities in Incidents of Noncompliance and the Supreme Court’s Plenary Amparo Docket 95 4.8 Hazard Models of Compliance Time 97 4B.1 Validity Analysis 117 4B.2 Determinants of Strike and Press for Three Measures of Federal Policy Importance (Procedural Dismissals Excluded) 119

ix

x

List of Tables

4B.3 Determinants of Strike and Press for Three Measures of Federal Policy Importance (Procedural Dismissals Included) 5.1 The Relationship between Awareness and Legitimacy in Eighteen Countries 6.1 Determinants of Judicial Legitimacy Beliefs 6.2 Awareness Effects for Alternative Deﬁance Cost/Judicial Independence Measures 6.3 Effects of Deﬁance Costs and Judicial Independence on the Awareness Estimate 6.4 Differences in Deﬁance Costs 7.1 Judicial Independence in Democratic States, 1990–2004

120 128 173 180 184 186 200

List of Figures

2.1 2.2 2.3 2.4 4.1

Baseline Model page 27 Equilibria (Baseline Model) 30 Promotion Model 34 Equilibria (Promotion Model) 37 Predicted Probability of Policy Invalidation by Federal Policy Importance and Prior La Jornada Coverage 82 4.2 Marginal Effect of Coverage as Importance Varies (Merits Decisions) 83 4.3 Marginal Effect of Coverage as Importance Varies (All Decisions) 84 4A.1 Structure of Mexican Federal Judiciary 102 5.1 An Indirect Effect of Awareness on Diffuse Support 129 5.2 A Conditional Effect of Awareness on Diffuse Support 131 5.3 Equilibrium Behavior with Low Public Backlash Cost 137 5.4 Equilibrium Behavior with High Public Backlash Cost 140 5.5 Updated Beliefs in Impartiality by Awareness 144 5A.1 Updated Beliefs in Impartiality by the Partisanship Assumption 165 6.1 Marginal Effects of Awareness on Judicial Legitimacy by Deﬁance Cost Measure (Two-Item Models) 176 6.2 Marginal Effects of Awareness on Judicial Legitimacy by Deﬁance Cost Measure (Three-Item Models) 178 6.3 Marginal Effects of Awareness (World Bank and Contract Intensive Money) 182 6.4 Judicial Power Dynamics 190

xi

Acknowledgments

I am deeply grateful for the intellectual guidance provided by cherished mentors and colleagues during the process of completing this book. As a doctoral candidate at Washington University in St. Louis, I beneﬁted greatly from the patient direction of my thesis advisors: Lee Epstein, John Carey, and Randall Calvert. The origins of the project lie in a course on Law and Society instructed by James L. Gibson, and I would like to thank him for his inspiration. In my case, the journey from dissertation proposal to book manuscript took a number of detours, ultimately fruitful I believe, but detours nonetheless. On account of this indirect journey, my thinking about the project has changed considerably over the years, and I owe much of that change to an ongoing conversation about law and politics with Georg Vanberg, Clifford Carrubba, Gretchen Helmke, and James Rogers. During this time, I have also received helpful suggestions from many other colleagues, including William Berry, Kathleen Bruhn, Tom Clark, Wayne Cornelius, Barry Friedman, Jennifer Gandhi, Matt Golder, Sona Golder, Anna Harvey, Matthew Ingram, Diana Kapiszewski, Beatriz Magaloni, Will Moore, Christopher Reenock, Julio Ríos Figueroa, Matthew Taylor, and Steven Wuhs. I thank Judy Kaplan for her invaluable editorial suggestions. For the ﬁnal version of the book, I am greatly indebted to two anonymous reviewers, and to Eric Crahan at Cambridge University Press, for both challenging me to clarify and deepen my argument and encouraging me along the way. I could not have conducted my ﬁeldwork in Mexico without the assistance of Benito Nacif and Jeffrey Weldon, and I thank them for xiii

xiv

Acknowledgments

their wise counsel. For logistical support, I would also like to thank the political studies department at the Centro de Investigación y Docencia Económicas and the political science department at the Instituto Tecnológico Autónomo de México. I thank Joy Langston, Gabriel Negretto, Guillermo Trejo, Allison Rowland, and José Ramón Cossío for helping focus my early work on the project. For their research assistance, I would like to thank Javier Rojas and Cindy Rayo. To complete the data collection process and to identify interview subjects, I leaned heavily on a number of outstanding people at Mexico’s Supreme Court. I would like to thank Héctor Arturo Hermoso and Marco Antonio Valadéz for their assistance in negotiating the Court’s archives. I thank Mara Gómez for her perspective, her selﬂess assistance, and her friendship. For their loving support, travel companionship, and quite frankly, for their tolerance, I wish to thank Shannon, Isabel, and John Staton.

part i JUDICIAL COMMUNICATION AND JUDICIAL POWER

1 Introduction

On August 24, 2000, the Mexican Supreme Court resolved a constitutional conﬂict between opposition members of the lower chamber of Congress and President Ernesto Zedillo.1 The sentence granted a congressional committee access to a trust account previously housed in a failed bank, which the federal government had taken control of in the weeks preceding the 1994 peso crisis. The committee sought access to the trust’s records, because it believed that the records might reveal a scheme to fund Zedillo’s presidential campaign illegally. This was the ﬁrst time in modern Mexican history that the Supreme Court challenged the power of the presidency in a case of such magnitude, and the court was quick to highlight it. Its ministers gave press conferences and interviews with various media outlets in which they detailed what the decision required of Zedillo and described their jurisprudential rationale. Although the court’s primary public face was its president, Genaro Góngora Pimentel, the effort was collective. Practically every minister played a role. The court’s public communication campaign was coordinated and aggressive. The Supreme Court’s reaction is not uniquely Mexican. Constitutional judges around the world engage the public through the media.2 Nearly every high court maintains a Web site where it houses information on pending and completed cases, descriptions of its jurisdiction, and

1 Controversia Constitucional 26/99, Semanario Judicial de la Federación y su Gaceta,

Novena Época, Tomo XII, Agosto de 2000, pp. 575, 962–963, 966–967, y 980. 2 I will refer to judges who sit on high courts with constitutional jurisdiction (e.g., U.S.

Supreme Court) and European-style constitutional courts (e.g., Austrian Constitutional Court) as “constitutional judges.”

3

Judicial Communication and Judicial Power

4

biographical summaries of its membership. Of course, this is fairly passive communication. Like the members of the Mexican Supreme Court, constitutional judges are commonly more direct. Often they use the media to underline key jurisprudential points. For example, Colombia Constitutional Court President Jaime Córdova Triviño gave a series of interviews in May 2006 clarifying a decision striking down a law that had granted partial amnesty to paramilitary group leaders.3 Canadian Supreme Court Justice Ian Binnie presented a lecture in February 2004 in which he discussed whether the court had usurped legislative authority with its interpretation of the Canadian Charter of Rights. Judges also defend publicly the concept of the rule of law in the context of particular cases (Kommers 1997). Even more commonly, judges use public forums simply to request better coverage. Australia High Court Justice Michael Kirby has suggested that the failure of the Australian media to construct a High Court beat makes it difﬁcult to communicate its decisions properly (also see Badinter and Breyer 2004, 265–266). In order to help organize their public activities, constitutional courts often house public relations ofﬁces. Christian Neuwirth, press ofﬁcer for the Austrian Constitutional Court, provides a representative statement on its varied work. But let me express that the written press information is not a big part in my usual work. If there are cases to be decided, I try to prepare journalists [for] what they can expect. [I]f the decision is made, I try to explain what the case is about. This is a permanent dialogue – far more than a written press statement.4

Table 1.1 underscores the breadth of the phenomenon in Latin America, where courts have developed particularly aggressive public relations strategies. As the table suggests, all but one constitutional court or Supreme Court with constitutional jurisdiction in the region makes ﬁnal sentences directly available to the media via their Web sites, and 72 percent of these courts alert the media to their resolutions through press releases. Of the courts that issue press releases, 92 percent of them do so selectively. That is, they promote some but not all of their decisions. Because press releases are a simple and common form of political communication, these data are merely suggestive of the multiple ways by which courts communicate with the public. Still, they reﬂect a systematic effort 3 See Clara Isabel Vélez Rincón, “Ley 975: quedó la forma pero cambió el fondo,” El

Colombiano, May 20, 2006. 4 Personal communication with author, July 24, 2006.

Introduction

5

table 1.1. Public Relations Summary for Constitutional Courts or High Courts with Constitutional Jurisdiction in Latin America Make Decisions Available on Publicly Accessible Web Site

Argentina Bolivia Brazil Chile Colombia Costa Rica Dominican Republic Ecuador El Salvador Guatemala Honduras Mexico Nicaragua Panama Paraguay Peru Uruguay Venezuela

√ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √

Announce Decision through Press Release Selective Promotion √ √ √

Universal Promotion

√ √ √ √ √ √

√

√ √ √ √

Note: Summarizes public relations activities of constitutional courts or high courts with constitutional jurisdiction in Latin America. The selective and universal promotion columns indicate whether the court announces some or all decisions by issuing press release.

to inﬂuence the quality and quantity of information about constitutional tribunals. There can be no doubt that high court judges are trying to get their public relations right.

the puzzle For sure, good public relations are essential in politics. Articulating a policy agenda, defending a controversial policy failure, managing a campaign message, and, perhaps most importantly, explaining a personal indiscretion, all require effective strategies of public communication (Flowers, Haynes, and Crespin 2003; Hillygus and Jackman 2003; Kernell 1993; McGraw 1991). It is difﬁcult to think of a scenario in which political actors do not have an incentive to get their public relations

6

Judicial Communication and Judicial Power

right. The natural explanation, however, is that, in one way or another, the ballot box constitutes the primary incentive for democratic political action (e.g., Downs 1957; Mayhew 1974; Powell 2000). For this reason, the desire to communicate with the public is theoretically intuitive and normatively appealing. But high court judges do not depend directly on votes. Perhaps of greater concern, judicial legitimacy is thought to derive from a healthy separation of judges from the public, a separation that allows judicial deliberation to be perceived as principled, neutral, and guided by procedure (Gibson, Caldeira, and Baird 1998; Hibbing and Theiss-Morse 1995; Scheb and Lyons 2000). This concern has not been lost on the judiciary. Consider U.S. Supreme Court Chief Justice Harlan Stone’s rationale for declining Senator Styles Bridges’ invitation to a testimonial dinner: The Court, as you know, has of late suffered from overmuch publicity. After all, its only claim to public conﬁdence is the thoroughness and ﬁdelity with which it does its daily task, which is exacting enough to demand the undivided attention of all its members. The majority are new in their positions and not too familiar with the traditions of the Court which have stood it in such good stead during the 150 years of its history. The upshot of all this is that I am anxious to see the Court removed more from the public eye except on decision day, as soon as possible – to imbue its members by example and by precept with the idea that the big job placed on us by the Constitution is our single intent in life and that, for the present, public appearances and addresses by the judges and the attendant publicity ought to be avoided. (as quoted in Mason 1953)

Likewise, in 1948, Felix Frankfurter famously suggested that he suffered from “judicial lockjaw,” a condition of self-censorship in which judges refrain from extrajudicial conduct (Dubek 2007). And, there is at least anecdotal evidence that judges invite trouble though public engagement. In states as different as Russia and Germany, high court judges have been criticized for publicly arguing for compliance in the context of signiﬁcant resistance to their decisions (Hausmaninger 1995; Jackson and Tushnet 2006, 740). In many familiar ways, the judiciary differs from the elected branches of government, and for that reason, public, nonadjudicatory communication has been regarded skeptically. Nevertheless, high court judges communicate with the public directly and quite outside the structure of written legal opinions. The central puzzle of this book concerns why they would do so. Why do judges go public?

Introduction

7

In the broadest sense, I wish to suggest that, to understand why judges go public, we must ﬁrst turn our attention to the politics of constitutional review. We must consider a fundamental problem of judicial policymaking, one that undermines judicial independence and threatens constitutionalism. Once we develop a sense of the conditions under which this problem can be solved, an explanation of judicial public relations emerges: judges go public to construct conditions favorable to the exercise of independent judicial power.5 To foreshadow, communication strategies are designed to advance the transparency of the conﬂicts constitutional courts resolve and to promote a deep societal belief in judicial legitimacy, conditions that promote judicial power. I will demonstrate that public relations offers material, if ultimately limited, control over transparency and, by so doing, expands the boundaries of judicial power. I will also claim that although it is possible to advance judicial legitimacy merely through public relations, judicial behavior itself, and not just what courts communicate about it, affects legitimacy. A key implication of this argument is that there can be a tension between the goals of constructing transparency and legitimacy. Judges might like to maximize the public’s information about their work under some conditions, but might prefer public ignorance under others. Where courts are free to resolve conﬂicts sincerely, without concern for external political interference, complete transparency is highly useful. However, where courts have incentives to engage in prudent decision making, complete transparency can be problematic, because it can highlight the lack of impartiality necessary to negotiate difﬁcult political controversies and, by doing so, undermine legitimacy. Strategic public relations can address this problem; however, because the media is not an arm of the judiciary, courts cannot fully control what is reported about them. For that reason, the tension is not easily resolved, and courts under serious political constraints may confront a power trap: promote transparency and risk undermining legitimacy or do not promote transparency and risk political irrelevance. In the remainder of this chapter, I will deﬁne judicial power and state the problem of judicial policymaking around which the argument revolves. I discuss theoretical solutions to this problem, and in that context, I summarize the argument, introduce the research design, and describe the chapters that follow. 5 Baum (2006) presents an excellent analysis of the multiple audiences judges target. The

goal of this book is not to catalog these various audiences in an international context, but rather to place the phenomenon of nonadjudicatory judicial speech within a uniﬁed model of judicial power.

8

Judicial Communication and Judicial Power

judicial power What does it mean to call a court powerful? A common conceptualization of judicial power centers on the rules that shape a court’s jurisdiction, rules that are deﬁned explicitly by constitutions and statutes or developed by judges themselves in their jurisprudence. (Baker 1971; Barber 1993; Beard 1962; Billikopf 1973; Boudin 1962; Burgoa 1984, 1998; Fix-Zamudio 1987; González Casanova 1967; González Cosío 1985; Gunther 1991; Johnson 1996; Lasser 1988; Rabasa 1982; Tena Ramírez 1957). The power of a court, its de jure power, is a legal description of what a court may do, what kind of conﬂicts it may resolve, and what remedies it may propose. Courts exercise power when they have jurisdiction. Billikopf (1973, 205) delivers a representative statement on the traditional legal notion of judicial power as jurisdiction. He writes: No clear line can be drawn between the terms judicial power and jurisdiction. . . . The word power is generally used in reference to the means employed in carrying jurisdiction into execution; the term jurisdiction refers to the capacity of the court to exercise its powers.

This conceptualization is not limited to the formal institutional scholarship often found in comparative legal analysis. Indeed, Tate and Vallinder’s (1995) foundational volume drew attention to the increasing degree to which courts were entering political debates in the 1990s, debates traditionally thought to be outside of their jurisdiction (also see Baird 2004; Stone 1992; Stone Sweet 2000). Judiciaries were becoming politicized precisely because the jurisdictions of the world’s courts were being expanded, either by politicians or by courts themselves, in part with the aid of a network of rights activists who wanted more politically relevant courts (Epp 1998). Nearly a decade after the Tate and Vallinder project began, Ginsburg (2003, 7–8) provided systematic evidence for the general expansion of formal judicial power, demonstrating that just about every third-wave democracy established a kind of constitutional review. The de jure concept is useful in arguments about institutional design, especially where subsequent compliance is assumed. A court with highly limited jurisdiction clearly is unlikely to affect the majority of a state’s policy debates. Although the de jure power concept works nicely within a discussion of the rules that should regulate constitutional review, it does not offer a satisfactory conceptual structure for addressing public policy outcomes. Under the de jure deﬁnition, we can call a court powerful if its

Introduction

9

formal rules grant it signiﬁcant authority, whether or not that authority is exercised in practice. This risks overlooking weakness in two kinds of courts: those that are openly deﬁed and those that shy away from conﬂict in order to avoid open deﬁance. And as summarized above, we observe outcomes of this sort commonly enough. For this reason, because compliance cannot be assumed, I adopt a de facto concept of judicial power. By power, I mean that an actor can cause by its actions the outcome that it prefers.6 This deﬁnition is identical to Cameron’s (2002) judicial independence concept (also see Larkins 1996, 611), although I will generally use the term power. Because there are moments when it will be convenient to use independence, it is worth distinguishing the power concept from a common alternative. Judicial independence is also conceptualized as a state of the world in which judges are able to make decisions that are free from external inﬂuence, whether that inﬂuence comes from coordinate branches of government, the private sector, or even from within the judiciary.7 As Kornhauser (2002) suggests, independent judges in this second sense are the “authors of their own decisions.” Critically, however, a powerful court under the Cameron concept must be autonomous under the Kornhauser concept. If it were not, then it would be impossible for the court to cause by its actions the outcome it prefers. The difference between the concepts, at least insofar as I wish to use them, is that an autonomous court might lack power, because it is unable to induce compliance generally, whereas a powerful court is autonomous and it is obeyed.

a fundamental problem of judicial policymaking The problem on which I will focus follows immediately from Publius’s contention in the 78th Federalist that courts are inherently weak political institutions. Lacking physically or ﬁnancially coercive means of enforcement, judicial power ultimately turns on the choices of elected ofﬁcials to respect the authority of courts. On many accounts, the problem undermines judicial power through either overt noncompliance or judicial prudence. It is understood that constitutional courts are deﬁed on 6 Also see Dahl (1963) or Nagel (1975). 7 Kapiszewski and Taylor (2008) provide an excellent conceptual discussion of concepts of

judicial independence and power. On independence as autonomy (in a variety of ways), see Couso (2004), Brinks (2005), and Ríos-Figueroa (2006).

10

Judicial Communication and Judicial Power

occasion (Canon and Johnson 1999; Johnson 1967; Rosenberg 1991; Spriggs 1996; Staton 2004; Trochev 2002; Vanberg 2005; Volcansek 2000) and that they sometimes strategically avoid conﬂict (Clinton 1994; Epstein and Knight 1996; Fix-Fierro 1998b; Lasser 1988; Murphy 1964; Schwarz 1973; Volcansek 1991).8 The entire Pakistani Supreme Court was dismissed and jailed in 2007 by President Pervez Musharraf when it was suspected that it would nullify his election to a third term. To avoid such outcomes, we commonly observe prudential decision making. The president of the Venezuelan Supreme Court, Cecilia Sosa Gomez, resigned her post in 1999 over her court’s approval of a highly suspect judicial reform enacted by allies of Hugo Chavez. Sosa suggested that the court had “committed suicide to avoid being assassinated. But the result is the same. It is dead.”9 Helmke (2005) even suggests that instances in which judges seem to challenge powerful and potentially dangerous political ofﬁcials are designed to avoid being purged from the bench following a regime or government transition. Importantly, examples of this sort can be found even in states where we anticipate widespread adherence to the rule of law. Local governments in the United States and Germany have simply refused to implement critical constitutional decisions over equal protection and religious establishment (Rosenberg 1991; Vanberg 2005). In much democratic theory, the stakes of solving this problem are high. Powerful judiciaries are, in part, solutions to a core democratic dilemma: how can government be sufﬁciently energized to induce social cooperation yet sufﬁciently restrained from violating individual rights (e.g., Madison 1787)? The response to this problem in classic and modern political theory involves dividing sovereignty (e.g., Falaschetti and Miller 2000; Locke 1698; Montesquieu 1962, 152), and the judiciary is a crucial component of this division (North 1990; North and Weingast 1989). But the institutional hedge provided by the judiciary only works if judges are willing and able to constrain government choices. If they cannot, then the judiciary provides no solution. Indeed, on some accounts, a state only can be considered democratic if it contains a judiciary that 8 It is important to note that the Argentine judges in Helmke’s study are strategically invit-

ing conﬂict, rather than avoiding it. Nevertheless, in an important sense, this is only a matter of labeling. What is going on here is that judges are inviting conﬂict with a current government to avoid conﬂict with a future one. Although the temporal dynamic is illuminating, the behavior is still consistent with the general incentive to decide prudentially on occasion. 9 “Top Venezuelan judge resigns,” BBC News, online, http://news.bbc.co.uk, August 25, 1999.

Introduction

11

enforces the rule of law (Linz and Stepan 1996; O’Donnell 1999). In this sense, democracy depends on solving the fundamental problem of judicial policymaking.

theoretical solutions Under what conditions do courts successfully exercise their power? One simple, yet elegant answer is that courts exercise power when governments want them to. Governments willfully delegate power to courts to solve a variety of political dilemmas. There are at least four variants of this line of argument. The ﬁrst suggests that ruling political coalitions empower judiciaries in contexts of increasing political competition. Judicial review serves as a form of insurance against potential violations of the current majority’s interests or over changes to fundamental policy regimes in the event that it loses control over the state (Finkel 2008; Ginsburg 2003; Hirshl 2001). A second type of willful delegation argument suggests that powerful courts provide informational advantages to political majorities, weeding out policies adopted under uncertainty that turn out to be ill-designed or simply ineffectual, but that governments cannot unravel through the legislative process (Rogers 2001). Whittington (2007) develops a related argument in which governments use powerful courts to dismantle the policies adopted by past majorities, but that would be difﬁcult or impossible to reform through law or rule making. A fourth willful delegation model suggests that courts gain power when governments recognize that an unconstrained state renders sovereign promises to respect property rights meaningless. These “noncredible” commitments erode the incentives for economic growth and threaten state solvency (Barro 1997; North and Weingast 1989). Courts are given power to render promises to protect property rights credible (Moustafa 2007). Although there is a great deal to admire in each of these arguments, they either deal with the choice to empower courts formally rather than with the implementation problem, or they have trouble explaining why we do not see powerful courts in nearly all corners of the world. The information and credible commitment stories provide compelling rationales for creating formally strong courts, but they struggle to explain international variance in judicial power. Few states can thrive without investment, and consequently, nearly all states confront the time-inconsistency problem that drives the credible commitment argument. Likewise, it would seem that uncertainty affects policymaking in every state around the world, a problem that could be aided by ex post judicial review, as Rogers suggests.

12

Judicial Communication and Judicial Power

And one would think that governments in nearly all democracies have confronted past legislation that is difﬁcult to remove through ordinary legislative procedures. Yet not even close to all states around the world, not even all democracies, host powerful judiciaries. Finally, the political insurance argument provides a persuasive rationale for choices to delegate formal authority to judiciaries, but it does not directly address what happens after a transition in power and a hostile government confronts a newly empowered court. Why would the new ruling coalition comply with constraints imposed by the old? If noncompliance is a possibility, in what way can judicial review be considered insurance? In summary, willful delegation arguments all provide intriguing rationales for judicial reform, but they either do not directly explain why courts constrain after they have been granted the power to do so, or they cannot explain the many cases in which courts are simply not powerful. Other arguments about judicial–government relations provide considerable traction over variance in de facto judicial power following a de jure empowerment; and these arguments offer a way of understanding what judicial public communication has to do with power. Many positive political models of interbranch relations suggest that the fundamental problem of judicial policymaking is solved only by fortunate arrangements of competing political interests. Insofar as judges care about either the implementation of their decisions or the institutional structure of the judiciary, they have incentives on occasion to uphold policies to avoid conﬂict with powerful ofﬁcials, no matter how constitutionally suspect the policies may be (Carrubba 2005; Epstein and Knight 1998; Ferejohn 1999; Ferejohn and Shipan 1990; Ginsburg 2003; Helmke 2005; Herron and Randazzo 2003; Iaryczower, Spiller, and Tommasi 2002; Ramseyer and Rasmusen 2001; Rogers 2001; Smithey and Ishiyama 2002). This type of strategic judicial deference is especially likely when the political branches ﬁnd it relatively easy to coordinate on a response to unfavorable resolutions (Andrews and Montinola 2004; Helmke 2005; Rios-Figueroa 2006). When government is fragmented, however, coordination is difﬁcult, and space is created for independent judicial review. Thus, interbranch models characterize judicial power as bounded, and the boundaries are regulated by the fragmentation of government. Courts exercise their power when the state is divided in ways that shield the judiciary from political retribution. Despite the incentives for prudence induced by the compliance problem, we observe some courts exercising their authority in ways that suggest that the bounds of judicial power are relatively large, perhaps

Introduction

13

large enough to ignore. For example, even in the context of serious national security threats, the Israeli Supreme Court has capped the intensity of interrogation methods used by the General Security Service in terrorism interrogations (Jackson and Tushnet 2006, 844–852), and even where strong domestic pressures suggest protectionist economic policies, the European Court of Justice routinely constrains member states to the terms of the Treaty of Rome (Alter 1998). Scholars in the behavioral tradition have argued that the key to inducing compliance with judicial decisions is a deep belief among a population in the court’s legitimacy (Caldeira and Gibson 1995; Gibson, Caldeira, and Baird 1998; Milner 1971; Mondak 1991; Murphy and Tanenhaus 1968). Interbranch pressures are largely irrelevant if courts are legitimate. It is critical to stress the normative basis of this argument. It is not that people particularly like the policy implications of every judicial decision. On the contrary, there will be policies that are offensive to majorities. The key is a normative commitment to the appropriateness of the constitutional review mechanism. This normative commitment to law trumps policy interests. On this account, the fundamental problem is solved by public beliefs in judicial legitimacy. The focus on legitimacy in behavioral scholarship need not be divorced from the models of strategic behavior found in the positive political literature. If we continue to assume that public preferences constitute the primary incentive for political action in the elected branches, then we can conclude that the public will inﬂuence the choice to respect judicial decisions as much as it does other political choices. This much is implied by the behavioral judicial legitimacy literature. Where voters are willing to tolerate noncompliance or political efforts to discipline overly active courts, constitutional judges will lack the leverage necessary to exercise meaningful authority. In such environments, where courts are constrained, we ought to expect judicial decisions to be linked tightly to the preferences of the elected branches. In other words, under such conditions, courts will lack power in the Cameron sense. In contrast, where voters are unwilling to accept the deﬁance of judicial authority and where they are capable of coordinating sufﬁciently to constitute a threat, courts will ﬁnd the leverage necessary to inﬂuence policy outcomes effectively (Carrubba 2003; Clark 2008; Stephenson 2004; Weingast 1997). Under this approach, courts continue to confront the fundamental problem of judicial policymaking; however, the public serves as a baseline source of power. The problem can be resolved by the pressure people place on their representatives to respect judicial authority. I will call this

Judicial Communication and Judicial Power

14

a public enforcement mechanism for judicial power. The general argument unites two disparate but clearly related literatures; however, it is important to recognize that the mechanism identiﬁes conditions under which courts will constrain government. It does not stand for the principle that courts always constrain, even if they are widely supported. Vanberg (2005) summarizes the two conditions under which the mechanism can work. Condition 1: Courts must enjoy sufﬁcient public support. Condition 2: Information about judicial decisions must be transparent. To eliminate the incentives for strategic judicial deference, courts must enjoy sufﬁcient public support, and the public must be sufﬁciently informed about the decisions they are purportedly enforcing. Quite obviously, where courts are despised, where the public does not view constitutional review as a legitimate way to resolve political conﬂicts, courts will not constrain. But even where courts are greatly supported, the mechanism can fail. Where information is sparse or where people do not understand what is required by a judicial decision, again judges cannot count on public support to provide political cover. Judicial power continues to be bounded. If either condition fails, it is unclear how the public can be invoked as a solution to the fundamental problem. The critical point for my purposes is that linking the public to the interactions between courts and elected representatives suggests a way of understanding why judges go public.

argument summary The central argument of this book is that these two important conditions for the exercise of judicial power under a public enforcement mechanism are ultimately endogenous to judicial–government interactions, and, as a result, judges enjoy a measure of control over the boundaries of their power. Consider the transparency condition. If the public enforcement mechanism is to work, people need to be informed about what they are purportedly enforcing. But if the mechanism is correct, if judges have some rough understanding of it, and if we assume that judges value their authority, then we might reasonably anticipate that they will attempt to construct the transparency condition. And surely, transparency is something over which judges have a measure of control. From this perspective, judicial public relations can be understood, in part, as an effort to con-

Introduction

15

struct the transparency of the work courts conduct. In that way, judicial public relations address the fundamental problem of judicial policymaking through its impact on the transparency condition. In the ﬁrst part of this book, I develop a model of judicial decision making in which I consider what happens to the essential predictions of Vanberg’s public enforcement model when we relax the exogeneity assumption concerning transparency. I demonstrate that judges can solve the transparency problem in principle, but that there are limits to what judges can accomplish. Raw political interests continue to bound judicial power. There are simply some policies over which governments are willing to suffer a public backlash for deﬁance, and thus, there are some courts that will be deﬁed no matter how transparent the decision. Nevertheless, judges can expand the boundaries of their power. Turning to the legitimacy condition, there is ample evidence suggesting that constitutional judges perceive public relations to be a means for developing beliefs in judicial legitimacy. Judges explicitly say as much. If the judicial legitimacy literature is correct, there is a colorable for believing that judges may be correct. If what makes people support courts is the belief that they engage in principled decision making (where choices are guided by politically neutral procedures), then perhaps courts can help develop that belief by exposing people to their work. After all, judicial deliberation is undeniably technical, and it is often politically neutral on its face. Thus, it is plausible that courts can advance the legitimacy condition by rendering their decision-making process more transparent. Yet, a key element of the legitimacy belief, which is supposedly being developed, is that judicial decision making is impartial. And this is something that individuals can learn about by observing the outcomes of cases and not just the processes by which decisions are reached. If newspaper reporters can identify instances of partisanship or prudence, then it seems possible that their readers can, too. In the third part of the book, I explicitly consider what voters might learn about judicial legitimacy when they are exposed to the outcomes of constitutional cases. I will show that beliefs in judicial legitimacy are likely endogenous to interbranch interactions, whether or not legitimacy can be constructed merely by exposing people to judicial process. Also, I will demonstrate that, under some conditions, it is extremely useful for courts to ensure that the public is maximally familiar with their work, but that under other conditions, courts might prefer public ignorance. Four implications follow from this argument. First, the book highlights a general afﬁnity between democratization and judicial power.

16

Judicial Communication and Judicial Power

Although a public enforcement mechanism could work under authoritarian rules, it is more likely to function effectively in a democracy. In the simplest sense, the mechanism leverages a form of vertical accountability to provide horizontal accountability. By opening channels of communication through a free press, and by making relevant the general public to the policymaking process, democratization offers judges opportunities to construct their power. But as I noted earlier, these opportunities are limited. Judges do not gain power simply because states democratize; and there is no necessary reason for even a consolidated democracy to possess a fully unconstrained judiciary. Inasmuch as democracies depend on strong courts (Linz and Stepan 1996; O’Donnell 1999), courts beneﬁt from democracy. Still, it is important to recognize that democratic processes provide opportunities for judicial power rather than create strong courts themselves. Second, the book suggests that the boundaries of judicial power are neither fully exogenous to judicial choice nor only characterized by the fragmentation of government, as implied by extant positive political theories of interbranch relations (e.g., Epstein, Knight, and Shvetsova 2001). Instead they are endogenous to judicial behavior and inﬂuenced importantly by societal beliefs in judicial legitimacy. In this way, clever constitutional judges have some measure of control over the boundaries of their authority. They are not beholden to lucky arrangements of political interests. Although it is true that power is limited, the limits may be manipulated, and in this way, the extent to which the state can be constrained is subject to the choices justices make. Third, there is a potential tension between the transparency and legitimacy conditions in the public enforcement mechanism. If courts are sometimes guided by prudential concerns, then full transparency will undermine legitimacy when courts are constrained. Thus, while constitutional judges have incentives to maximize transparency, doing so may compromise their ability to establish and manage their legitimacy. Public relations can help manage this tension by giving judges a measure of inﬂuence over the quality and quantity of media coverage. However, as long as the media is not an arm of the judiciary, public relations can only provide judges with partial control over this tension. Critically, judges in democratizing states are especially likely to confront this problem, because these will be states where judicial legitimacy needs to be constructed, where courts likely have not received signiﬁcant attention from the media, and where judges have incentives to be prudent. Thus, precisely where judges have the strongest incentive to go public to

Introduction

17

create transparency and build legitimacy, they are likely to confront the transparency–legitimacy tension. Finally, the book suggests that judicial power may not be constructed via a strategy of prudence in the long run. Although courts might escape immediate retribution through prudence in the short run, if prudence undermines legitimacy, then it does not necessarily increase power. The experiences of the Venezuelan, Pakistani, and Argentine Supreme Courts are instructive. We might reasonably ask whether the Venezuelan Supreme Court is more powerful than the Pakistani Supreme Court for having carefully navigated the dangerous political waters of the early Chavez regime. Can it really be said that the legitimacy of the Supreme Court of Argentina has been promoted through the strategic defection logic Helmke highlights? Perhaps building strong courts requires boldness. Perhaps it even requires accepting overt acts of noncompliance. Of course, if courts undermine legitimacy via prudence, then high proﬁle, highly transparent legal claims against clearly unconstitutional governmental policies present a trade-off for judges: defer to government and undermine legitimacy or challenge government and risk interbranch conﬂict.

approach of the book The book’s theoretical approach is outcome oriented and outward looking. It posits that judges ultimately care about inﬂuencing public policy outcomes; and to do so, they must adopt a strategic approach to decision making. The strategic environment on which I focus is external to the court, and for this reason, I do not explicitly model the internal dynamics of collegial decision making (Maltzman, Spriggs, and Wahbeck 2000). Instead, I will treat constitutional courts as if they were unitary actors. Before going any further, then, it is useful to ask whether constitutional courts are unitary actors and always outward looking. It seems clear that the answer to both questions is that they are not; however, the issue is not whether they are actually unitary actors or always outwardly oriented, but whether it is useful to treat them as if they were for the purposes of the study (Laver and Schoﬁeld 1990, 14–15). The assumptions I make about constitutional courts, while inaccurate in a number of ways, are useful in the context of the particular research problem. The problem of judicial policymaking on which I focus is an outward-oriented problem. That is, it does not concern the process by which judges compromise, cajole, and browbeat their colleagues in search of rule or remedy, but

18

Judicial Communication and Judicial Power

rather, what happens after a decision has been reached and the power to implement it has been delegated to the elected branches. Moreover, if one of the sources of judicial legitimacy derives from public ignorance about the internal bargaining that takes place on constitutional courts (Hibbing and Theiss-Morse 1995, 59; but see Baird and Gangl 2006), then it is in the interests of all members of the court to advance an image that deemphasizes inevitable intracourt deliberation.10 In this sense, although the assumptions I make ignore much of the interesting dynamics of judicial decision making, they suit the problem of interest. The theoretical arguments I develop in Chapters 2 and 6 are not speciﬁc to regions of the world. They have implications for any high court that exercises constitutional review and lacks direct authority over the implementation of its sentences. However, much of the empirical analysis concerns the Mexican Supreme Court and its place within Mexican politics during the Ernesto Zedillo and Vicente Fox administrations. Mexico during this time of democratic transition represents an excellent opportunity to test the argument. The Supreme Court just had been granted new powers of constitutional review yet confronted a history of political interference (e.g., Domingo 2000; Fix-Fierro 1998b). Thus, there was reason to question whether challenging political authority would be costless. In addition, the court’s membership was absolutely stable during the short period under analysis so that assuming that the court’s preferences were stable during this period is less likely to confront arguments about long-run judicial preference change (e.g., Martin and Quinn 2007). Also, the fragmentation of federal government changed dramatically during the period under analysis. The years covered by this study include periods of fully divided and uniﬁed government across partisan lines. For this reason, it is possible to examine familiar partisan and institutional explanations of judicial decision making. Finally, the Zedillo and Fox years represent a signiﬁcant period of democratization in Mexico (Shirk 2005; Wuhs 2008). This time frame offers an opportunity to consider the challenges that newly empowered courts in democratizing states confront as they attempt to build their institutions. 10 There is some reason to believe that members of a collegial court might wish to undermine

the transparency of a constitutional decision, one that they disagree with, for example. Of course, by doing so, they undermine the power of the court as a whole and potentially induce their colleagues to reciprocate. Nevertheless, managing the power of a court may involve solving a variety of problems of social cooperation among individual members, and I do not consider the various ways by which these problems can be solved. I thank James Rogers for identifying this issue.

Introduction

19

Ultimately, I center the empirical analysis on a single country for theoretical and practical reasons. Proper tests of the model’s key theoretical propositions require detailed information on political institutions and the beliefs political actors have about each other and their respective political capital.11 Even the simplest tests require considerable information on features of constitutional decisions. Some of this information is available on the Internet; however, much of it required archival and interview-based ﬁeldwork. While designing the empirical research around one country limits the scope of the empirical conclusions, it does allow me to test the theoretical argument carefully in a number of ways.12 Having said this, it is critical to stress that the book draws on judicial experiences from around the globe and discusses a number of examples where appropriate. And importantly, the empirical tests of core model predictions about the development of beliefs in judicial legitimacy are conducted on cross-national data.

chapter summaries The book is organized as follows. Chapter 2 develops a public enforcement model of judicial policymaking in which I ﬁrst consider the problems for judicial power that emerge from the inability to control the way that decisions are presented to the public by the media. I then consider how courts might solve these problems when they have control over their coverage. I will suggest that public relations helps courts solve these problems, if only imperfectly. The model identiﬁes conditions under which judges can construct their own power by providing transparency. It also suggests conditions under which doing so would either be suboptimal or entirely useless. The model suggests that constitutional judges are not simply 11 On this point, see the discussion in Helmke (2005, 15–16). 12 Testing much of the argument in a single country allows for the kind of careful analysis

of background information necessary to test a game theory model. Also, judicial scholars have only recently begun to construct comparative data sets on constitutional decision making (Herron and Randazzo 2003; Tate et al. 2006). Unfortunately, these data sets are limited to countries that either publish in English or translate selectively into English. The former limits generalizability; the latter suggests selection bias, especially if the core argument of this book is correct. The alternatives are single-country data sets that share little agreement over coding rules and concepts (Helmke 2005; Iaryczower, Spiller, and Tommasi 2002; Magaloni and Sanchez 2001; Ríos-Figueroa 2006; Vanberg 2005). Finally, the data set I introduce later also happens to be the only one outside of the United States that includes information on media coverage prior to and following high court decisions.

20

Judicial Communication and Judicial Power

reacting to exogenously determined political conditions; instead, they are energetic participants in the construction of those conditions, and in that way they are helping to construct their own authority (see Burley and Mattli 1993, 63, for a similar point on supranational court judges). In Chapter 3, I turn to Mexico. I provide a brief introduction to the Mexican Supreme Court and its role in Mexican politics through the transitional period in the late 1990s. I then describe the Supreme Court’s public relations activities and summarize the court’s own explanation of what it is attempting to do by communicating directly with the public. In Chapter 4, I test the model’s implications for the court’s decisions on the merits of constitutional cases and its choices to promote those decisions by issuing press releases to national media outlets, and public authority compliance. I use two original data sets on Mexican constitutional review. The ﬁrst data set, which characterizes the court’s plenary constitutional docket, is the only one of which I am familiar that includes the full breadth of cases under the court’s constitutional jurisdiction. This allows me to study empirically the amparo suit for the ﬁrst time since the 1970s. I ﬁnd that the Supreme Court’s decisions are sensitive to the political importance of the policies it reviews as perceived by the federal government; however, it uses its case promotion strategies to ensure transparency when it is theoretically useful. The second data set describes public authority responses to Supreme Court efforts to ensure compliance with past decisions. This data set is unique as well and allows for the ﬁrst systematic empirical study of Mexico’s incident of noncompliance, a constitutional action through which the Supreme Court can remove an ofﬁcial from ofﬁce for defying the decision of a federal court. In Chapter 5, I address how judicial behavior might inﬂuence judicial legitimacy. I construct a model of constitutional review that explores how beliefs in judicial legitimacy develop when judicial decisions are subject to the kind of interbranch dynamics characterized by the model in Chapter 2. I ﬁnd that much can be learned from observing case outcomes, even when individuals have little information about constitutional courts. I also ﬁnd that, under some conditions, complete transparency is extremely useful as a means of supporting beliefs in judicial legitimacy, but that under other conditions, transparency can undermine legitimacy. In this sense, this chapter raises a possible tension between transparency and judicial legitimacy, the two necessary conditions for a robust public enforcement mechanism. I conclude the chapter by considering this tension and suggest that it provides an additional rationale for a carefully managed, strategic public relations strategy.

Introduction

21

In Chapter 6, I provide several tests of the central empirical implication of the judicial legitimacy model developed in Chapter 5. The empirical design in Chapter 6 involves estimating how individual-level beliefs in judicial legitimacy are affected by awareness and the information about judicial–government interactions to which respondents would have been exposed. I use a cross-national public opinion data set previously analyzed by Gibson, Caldeira, and Baird (1998). I ﬁnd that where courts are relatively unconstrained by external political pressures, familiarity with the courts breeds legitimacy. To know a court is to love a court. However, where courts are constrained, familiarity is either totally irrelevant or harmful to legitimacy. Chapter 7 offers a summary of the book’s central results, situates these results within the literature on democracy and the rule of law, and raises a number of questions for further analysis.

2 A Model of Constitutional Review and Case Promotion

Why do constitutional courts promote their decisions in the media? In this chapter, I describe how courts might construct their power by inﬂuencing information within a public enforcement model of constitutional review. This requires thinking about how case promotion affects the logic of existing models of constitutional review in which information is exogenously determined. The chapter is divided into three sections. I begin by describing the underlying public enforcement mechanism as it has emerged in the theoretical literature. I will identify problems that derive from such a mechanism and ask how judicial public relations might address them. In the second section, I develop two straightforward game theory models of judicial–government interaction, one of which allows the court to inﬂuence information while the other does not. I discuss the intuition behind the central results and identify observable implications, which I will test in subsequent chapters. In the ﬁnal section, I return to the political challenges judges face within a public enforcement mechanism when they lack control over information and consider the extent to which public relations can solve these problems.

a public enforcement mechanism Given the judiciary’s lack of enforcement powers, it seems clear enough that we should not treat judicial decisions as unconditionally binding, at least on coordinate branches of government. Still, it would strain empiricism beyond recognition to imagine that courts never control the outcomes to sensitive policy conﬂicts – that authorities can disregard their legal obligations always or even more often than not. Thus, it seems likely 22

A Model of Constitutional Review and Case Promotion

23

that we can identify conditions under which judicial decisions can be treated as if they were binding. Let us begin by supposing that the public serves as an indirect enforcement mechanism for judicial decisions, pressuring representatives for compliance. That the public would play such a role is consistent with standard assumptions about the behavior of elected ofﬁcials, that is, that they respond in some way to constituent desires. However, if we make this assumption, we necessarily ask whether the public ever responds to perceived threats to the rule of law. In a sense, we would not expect to observe many negative public reactions if politicians are generally forward looking and sensitive to constituent interests. Yet, whether because of myopia or uncertainty, there are examples. The experience of the U.S. Supreme Court during Roosevelt’s court-packing plan represents the most familiar example of a signiﬁcant, negative public response to a policy seemingly aimed at manipulating constitutional interpretation (Caldeira 1986). The Pakistani lawyers’ movement, which led a massive resistance against President Pervez Musharraf’s repeated assaults on the Supreme Court in 2007 and 2008, constitutes a recent and salient example of the potential negative consequences of attacking a court. If it is plausible to assume that, under some conditions, there are costs to attacking the authority of a court, whether through direct noncompliance or institutional manipulation, then we confront a second question. Why would the public defend unelected judges against the choices of its own representatives? Stephenson (2004) suggests that the public promotes compliance with judicial decisions when it believes that the judiciary is more likely to represent its interests than its elected representatives. In political systems in which politicians commonly engage in rent-seeking activities, it is certainly reasonable that voters will believe that politicians will pursue policies inconsistent with constituent interests. This is a special concern in policy areas where voters face informational challenges relative to their representatives, who may use their private information to pursue personal wealth or their own, independent ideological proclivities (Stokes 2001). Although plausible, this argument does not sit easily with the observation that publics do, at least on occasion, support courts in political conﬂicts where elected ofﬁcials seem to be representing community values accurately. As we know from the court-packing episode in the United States, Roosevelt’s efforts to shape the size of the Supreme Court in order to induce a constitutional jurisprudence more in line with dominant national interests were met with serious resistance from the public (Friedman 2000, 1037–1046). It is not only unclear how the

24

Judicial Communication and Judicial Power

Stephenson model would explain this outcome, but the behavior in question also raises a difﬁcult question in its own right. Why might people support a court against a president when the president is accurately reﬂecting their interests? In other words, why might people promote the rule of law universally? On Weingast’s (1997) familiar account, individuals promote the universal application of the rule of law in order to prevent the occasional state exploitation of their own rights. It is better to live in a world in which everyone’s rights are protected than in a world in which I sometimes beneﬁt from the exploitation of another group, and I sometimes have my own rights violated. The crucial point here is that the sovereign believes that individuals in society will successfully coordinate in response to state actions that violate rights. Weingast identiﬁes the coordination problem that must be resolved under any sort of public enforcement story; however, it is not clear how groups identify rights violations in the ﬁrst place. In the Weingast model, the problem is not whether individuals can observe the state violating limits on its authority. This is observed perfectly. The problem involves what to do once the state trespasses on rights. But interpreting a state action as outside the bounds of constitutional limits necessarily precedes the coordination dilemma over how to respond to a state that has surpassed its limits. Judicial legitimacy theorists do not address the coordination dilemma Weingast identiﬁes, but they do suggest how individuals identify violations in the ﬁrst place. The idea is not so much that people agree about whether policies cross the formal boundaries of state authority, but rather that people are willing to defend a court that says that they boundaries have been violated. Judicial legitimacy theorists suggest that individuals support courts even when it is personally costly to do so, but only when they endow them with legitimacy. Legitimacy is conceptualized as diffuse public support or a deep commitment to the institutional integrity of the judiciary (Caldeira and Gibson 1992, 638). The concept requires a commitment to defending the institutional structure and authority of the judiciary, even if particular decisions run against personal interests. Under this view, people support courts when they believe that doing so is the “right thing to do.” The empirical evidence presented in Gibson, Caldeira, and Baird’s study suggests that these beliefs likely are generated over a long period of time. Building on these insights, Carrubba (2009) suggests that, by carefully managing compliance through strategic constitutional review, courts can induce publics to believe that it would be in their interest to support the judiciary generally, even if this means

A Model of Constitutional Review and Case Promotion

25

punishing an elected ofﬁcial for noncompliance when the public gains from it, in the short run at least. Once individuals in society share this belief, judicial decisions can be treated as if they were fully binding on governments. I will return to the Carrubba point in Chapter 6. For now, it is sufﬁcient to note that these arguments suggest ways of understanding why individuals might support courts against elected governments, and by implication, why judicial orders might be treated as if they were binding when courts enjoy sufﬁcient public support. Still, although a supportive public that is able to coordinate on a response to state violations of the rule of law may be necessary for the exercise of judicial power, the public enforcement account also involves a monitoring problem (Vanberg 2005). In particular, people must be sufﬁciently informed about the cases that courts resolve in order to identify and deter potential noncompliance. It only makes sense to invoke the public as an enforcement mechanism for judicial decisions, indirectly or directly, if people are aware of what they are enforcing. Stated differently, the political incentives for noncompliance are strongest when it is unlikely that constituents will learn about it. Consequently, when the public is unlikely to be informed about the particulars of the conﬂicts courts are called upon to resolve, judicial decisions will not be binding, and the incentive for strategic deference returns. To summarize, scholars identify two conditions under which we can invoke the public as an indirect enforcement mechanism for judicial decisions: (1) courts must enjoy sufﬁcient legitimacy to make political deﬁance unattractive, and (2) people must be sufﬁciently informed about their activities in order to monitor potential noncompliance. The legitimacy condition sets a conceptual lower bound on the court’s support. The claim is not that only the most supported courts will exercise inﬂuence over policy outcomes, but that some level of support is necessary. Without any public support, elected ofﬁcials will anticipate no cost for defying judicial decisions, and the judicial incentive to resolve cases prudentially is strongest. The transparency condition suggests that publics must be informed about the results they are allegedly enforcing. If they are not, then noncompliance will go undetected. If noncompliance will go undetected, then the incentive to comply is considerably reduced, and the incentive to resolve cases on prudential grounds returns yet again. Three implications of these conditions bear directly on judicial public relations. If the argument is correct, then we ought to expect constitutional courts to be more likely to exercise their power when they expect the public to learn about and understand their decisions. Following immediately

26

Judicial Communication and Judicial Power

from this point, media inattention to constitutional decisions presents a serious problem for constitutional courts. Insofar as the media is the primary means by which individuals learn about political events (e.g., Graber 2005), media attention is crucial to the enforcement mechanism. If the media ignores judicial outputs, it is unclear how people will learn about the decisions they are supposed to enforce. Further, judges have an interest in developing and maintaining their legitimacy. Legitimacy is a precious resource. Without it, judicial power is highly dependent on institutional arrangements that constrain the elected branches of government (Andrews and Montinola 2004; Tsebelis 2002) or upon the choices of politicians to constrain themselves (Ginsburg 2003; Landes and Posner 1975). So far, media attention seems to be a solution to the transparency problem. But media attention presents a distinct problem of its own. Although the press frequently communicates judgments accurately (i.e., who wins and loses a case), reporters often misinterpret jurisprudential rationales, an outcome that judges themselves cite as frustrating (Davis 1994, 31). From a purely professional perspective, this is a problem. No judge can possibly enjoy reading an incorrect description of her reasoning. But perhaps there is a larger problem. An established literature in judicial politics suggests that judicial opinions themselves are a key element of the myth of judicial impartiality. Opinions support the idea that judicial deliberation is purely technical and impartial. If judges use opinions, in part, to help mold that image, and the media incorrectly summarizes their rationales, then it is immediately obvious that judges should take an interest in ensuring that their reasoning is properly communicated to the public through the press. Thus, the failure to control information presents both a transparency and a legitimacy problem for courts. Importantly, and in contrast to the transparency problem, the legitimacy problem is paramount when the media is expected to cover a case. In this way, it is possible that both media inattention and media attention are problematic for courts, albeit for different reasons. In the subsequent section, I seek to answer whether courts can solve the problems of media attention and inattention by promoting their cases. I ask whether the utility of case promotion is limited, and if so, what are the conditions that limit judicial inﬂuence?

models of constitutional review In this section, I develop two models of judicial–government interaction that explore the concerns I have just raised. The analytical approach involves comparing the results of two straightforward game theory

A Model of Constitutional Review and Case Promotion

27

models. A baseline model is designed to capture the key dynamics of the Vanberg story. It describes a public enforcement mechanism for judicial opinions in which transparency is exogenously determined. In the second model, I allow the court to inﬂuence transparency. I will compare the results of these two models and, by so doing, derive hypotheses concerning judicial public relations and judicial decision making that are consistent with an exogenous or endogenous treatment of case transparency.1 This exercise not only clariﬁes how judicial public relations might advance judicial power, it also suggests a way of distinguishing models that treat information as at least partially under the control of the judiciary from models that treat information as fully exogenous.

baseline model Figure 2.1 depicts an extensive form game of incomplete information played between a constitutional court and a national executive, although

ge covera Media

(p)

No me

dia co

verage

(1–p)

Court Uphold

Uphold

Status quo

Status quo Strike down

Executive Accept Alternative policy

Strike down

Executive Defy Status quo/ Public punishment

Accept Alternative policy

Defy Status quo

ﬁgure 2.1. Baseline Model 1 It is worth stressing that this theoretical exercise naturally extends the Vanberg argument.

In the original setup, transparency is treated as an exogenous variable, although Vanberg never makes a principled argument about why we should treat transparency in that way. The theoretical analysis here considers what we should expect to observe if we imagine that transparency is endogenous to judicial behavior. A further approach might imagine what happens if we allow the players to compete over how they are covered.

28

Judicial Communication and Judicial Power

the second player may be conceptualized as any government ofﬁcial. The game begins after the court has been asked to review the constitutionality of an action taken by the president. This approach to modeling judicial decision making suggests that judges essentially partition policies (sets of case facts, if you like) into bins that indicate acceptable and unacceptable choices rather than establish points on a line (as in Lax 2007). For the sake of argument, let us assume that some individual in society, affected by the president’s action, has standing to bring a constitutional action against the president. The court may uphold the status quo policy (i.e., the president’s action), or it may strike it down and substitute an alternative policy in its place. If we imagine that the status quo policy is something like detaining a citizen in a military prison, then the alternative policy might be conceptualized as setting the citizen free. The constitutional action, then, is best thought of as some sort of individual constitutional complaint (e.g., habeas corpus or amparo), although this limitation is not required. If the court strikes down the status quo policy, the executive chooses whether to accept the alternative. I conceptualize noncompliance broadly. Although noncompliance always involves the continued implementation of the status quo, noncompliance may also involve taking retributive actions against the court, the severity of which I parameterize in the court’s payoff function. For each terminal history of the game, Figure 2.1 includes a brief description of what policy obtains and whether the executive is punished for deﬁance. I model the transparency of the case with a simple device that incorporates the media. Prior to either player taking an action, there is a random draw from a set that determines the level of media attention and, by implication, public awareness of the case. Either the media will cover the resolution and accurately communicate the court’s judgment, or the media will ignore the decision and the public will go uninformed about the interaction. Denote the probability that the media covers the resolution p. The court does not observe the outcome of the draw, but it knows the distribution from which it is drawn (i.e., the court knows p).2 2 The court moves before it sees whether there is coverage of the decision. This seems

natural insofar as, strictly speaking, there cannot be coverage of a decision that has not been released. Yet, reporters often obtain information about case outcomes prior to a formal release date, and their editors sometimes run stories in advance. In such a case, the court would be making a decision under complete certainty. Although the model does not capture this scenario perfectly, it does allow for scenarios that are functionally equivalent. What is required is an evaluation of the model in which we set p arbitrarily close to one. The results that follow consider equilibria under all values of p, so that the

A Model of Constitutional Review and Case Promotion

29

Because the executive moves after the resolution, she observes whether the media covers the case. As a result, the executive is perfectly informed about the history of the game. Thus, a strategy for the court is a function sc : Ic → Ac , that assigns an action from Ac = {Uphold, Strike Down} to the court’s single information set Ic . A strategy for the executive is a function se : Ie → Ae that assigns an action from Ae = {Accept, Defy} to each of her two singleton information sets. In keeping with common assumptions in judicial politics, I assume that judicial choice is primarily a function of policy preferences. The value of the status quo policy is ﬁxed at 0 for both players. If the executive faithfully implements an alternative, the court receives ac ∈ + .3 The court pays a cost c > 0 if the executive fails to comply. Although I assume that the media accurately covers the judgment, consistent with the judicial belief that reporters often mischaracterize jurisprudential details, I assume that the court pays a cost e ∈ (0, ac ), which reﬂects the primacy of policy preferences. This cost represents the value the court places on reporters accurately communicating their decisional rationale. Note that the court only pays this cost if the media covers the resolution. The executive is assumed to prefer the status quo to any alternative and will pay a cost ae > 0 if she complies with an unfavorable decision. If the executive fails to comply and the media is covering the case, she pays a cost among her constituents for defying the court, b > 0. One natural interpretation of these parameters incorporates the concepts of speciﬁc public support (the public’s issue-related preferences) and diffuse public support. If we assume that the executive is generally responsive to public preferences, then ae may represent speciﬁc public support for the status quo and b may represent diffuse public support for the court. Interpreted in this way, the model allows executives to defy highly legitimate courts over valuable policies, but it also allows executives to implement decisions that are unpopular among their constituents, as long as diffuse public support is sufﬁciently large. This is a sensible result to allow for because the heart of the diffuse public support concept is that people are willing to accept unpopular decisions as long as the decisional source is legitimate. For any given values of the parameters, there is a unique pure strategy subgame perfect equilibrium. There are three cases in the baseline model.

model essentially captures a situation in which judges know for certain that a case will be covered. 3 The key results of the model are not affected by allowing a < 0. The implications of this c assumption are discussed in Staton (2006).

Judicial Communication and Judicial Power

30

c Case 1: Informational Deference For ae ≤ b & p < c+a c sc = Uphold the status quo. se = Defy if the media fails to cover the decision and Accept otherwise. c Case 2: Judicial Power For ae ≤ b & p ≥ c+a c sc = Strike down the status quo. se = Defy if the media fails to cover the decision and Accept otherwise. Case 3: Judicial Deference For ae > b & ∀p sc = Uphold the status quo. se = Defy all unfavorable verdicts. Figure 2.2 depicts the equilibria for varying judicial beliefs over ex post media coverage and for varying degrees of importance to which the executive assigns the status quo. The Informational Deference and Judicial Power equilibria reﬂect the core dynamics in the Vanberg argument. When c it is unlikely that the media will cover the resolution i.e., for p < c+a , c the court strategically avoids conﬂict with the executive. Because the executive will defy the decision if the media ignores it, the court upholds the status quo in the Informational Deference equilibrium because it cannot depend on the public to serve as its enforcement mechanism. Importantly,

Probability of media coverage (p)

p=

Judicial power

c c + aC

Judicial deference Informational deference

b Cost of alternative policy to government (ae)

ﬁgure 2.2. Equilibria (Baseline Model)

A Model of Constitutional Review and Case Promotion

31

the information environment in this case induces the court’s deference. Indeed, a court that is greatly supported may ﬁnd itself in this kind of equilibrium whenever the cases it resolves are difﬁcult to communicate to the public. For this reason, Vanberg (2005) ﬁnds that the German Federal Constitutional Court behaves strategically despite its overwhelming public support. Regardless of a court’s legitimacy, prudential decision making results from insufﬁcient information. In the Judicial Power equilibrium, the court challenges the executive’s policy, expecting the media to provide transparency through its coverage of the resolution. As is clear from the expression, the critical value of p, above which the court will strike down the status quo, increases in the cost of noncompliance (c) and decreases in the value the court assigns to the alternative policy (ac ).4 In other words, the more costly the noncompliance (or the less valuable the alternative), the more certain the court will need to be in the media’s ex post coverage before it strikes down the status quo. This much summarizes Vanberg’s transparency argument; however, there is a critical difference between the empirical prediction Vanberg suggests and what this model implies. Note that the dynamics I have just described completely disappear when the status quo is sufﬁciently important to the executive. In the Judicial Deference equilibrium, the government does not condition its response to the court on whether the media covers the resolution. Indeed, the government will defy an unfavorable resolution even if the public will become perfectly informed about the deﬁance. In such a case, transparency does not afford the court the necessary leverage to confront the government. Consequently, the court’s decisions are not a function of its beliefs about the likely transparency of the case. It simply defers to the executive’s authority. The baseline model suggests empirical implications concerning the choices of constitutional courts to strike down public policies and the choices of public ofﬁcials to respect those decisions. Baseline Implication 1: When courts do not inﬂuence the transparency of the cases they resolve, they should be less likely to strike down public policies as the importance of the policy to the government increases. Baseline Implication 2: When courts do not inﬂuence the transparency of the cases they resolve, the relationship between judicial beliefs 4 ∂p > 0 and ∂p < 0. ∂c ∂aC

32

Judicial Communication and Judicial Power about transparency and the choice to strike down a public policy is conditioned by the importance of the status quo to the government. When policy importance is relatively small, courts should be more likely to strike policies down as transparency increases; however, when importance is relatively large, there should be no relationship between transparency and the court’s decision.

These results reﬂect the common perception that strategic judicial behavior is increasingly likely as the importance of public policies to relevant political actors increases (e.g., Helmke 2002; Iaryczower, Spiller, and Tommasi 2002); however, it is important to stress two qualiﬁcations. First, the claim that judicial power decreases in policy salience must be conditioned by a ceteris paribus condition. Importantly, for sufﬁciently large public backlash costs, the power of a court may cover extremely salient policies. Second, the model also suggests that strategic behavior is observable in less salient conﬂicts, especially when the media is unlikely to cover a resolution. Most importantly, it suggests that the inﬂuence of transparency on judicial power is conditional – it depends on the importance of the underlying policy issue to the government being challenged. For many public policies, transparency will matter a great deal; however, for sufﬁciently important policies, transparency is irrelevant to the choice to strike down a public policy. If this implication is correct, then estimates of the effect of transparency on judicial decision making may be both overstated and understated if analysts do not allow this effect to be conditioned by policy importance. For highly important policies, unconditional estimates of transparency are likely larger than they should be; however, for less important policies, unconditional estimates are likely smaller than they should be. Finally, if we continue to assume that ae reﬂects the public’s speciﬁc support for the policy under review and that b reﬂects something like the public’s diffuse support for the court, then the result also places a reasonable upper bound on the role of judicial legitimacy in interbranch relations. There are likely some public policies that are important enough that publics will not support even the most legitimate courts in battles with their representatives. The baseline model also implies a hypothesis concerning compliance with unfavorable constitutional decisions, one that underscores the importance of considering the strategic environment in which constitutional courts operate.

A Model of Constitutional Review and Case Promotion

33

Baseline Implication 3: When courts cannot inﬂuence the transparency of the cases they resolve, ofﬁcials should be more likely to defy constitutional court decisions over relatively unimportant policies that go uncovered by the media. It is crucial to stress that noncompliance should be decreasingly likely as policy importance increases. Although it is true that the incentive to defy a decision increases as salience increases, salient policies are precisely the policies over which the court defers to the executive. Finally, the baseline model highlights the two informational problems judges may be able to resolve through public relations. The ﬁrst concerns transparency – the problem identiﬁed by the Vanberg model. It is uncertainty surrounding the public’s ability to monitor the interaction that produces strategic judicial behavior under the Informational Deference equilibrium. If the court were sure that the media would cover the resolution, it would have no incentive to uphold the status quo, precisely because these are decisions that the government would respect given sufﬁcient public information. Thus, if judges are able to induce media attention for their resolutions, they may be able to eliminate the necessity to defer strategically in situations characterized by the Informational Deference equilibrium. The second problem concerns the legitimacy condition. In all cases, the court expects to lose the cost of misinformation, discounted by the probability of media coverage (i.e., p ∗ e in equilibrium. This problem may be resolvable via public relations as well. Case promotion can give judges the ability to clarify their resolutions, improving on the reporters’ descriptions of the case. I now turn to a version of the model in which the court is able to produce accurate media coverage through case promotion.

promotion model The promotion model, depicted in Figure 2.3, is identical to the baseline in all ways but one. I allow the court to promote a resolution among the national media at a cost k > 0. This cost may be interpreted in one of two ways. We might imagine that it measures the administrative costs associated with promoting a case, in the sense that it is at least minimally costly to translate dense legal text into language that is appropriate for public consumption, and it may be personally or institutionally costly to contact editorial staffs or beat reporters with the purpose of lobbying for a particular kind of coverage. Alternatively, we can conceptualize k as the opportunity cost of promoting one decision, knowing that not all

34

e (p)

overag

c Media

Court Uphold

Status quo

Strike down and promote Executive Defy

Uphold

Status quo

Strike down

Accept

dia cove rage (1 –p)

Uphold and promote

Status quo

Executive

No me

Accept

Accept

Alternative Status quo/ Alternative Alternative Status quo/ Public punishment policy policy policy Public punishment

ﬁgure 2.3. Promotion Model

Status quo

Strike down

Strike down and promote

Executive Defy

Uphold and Promote

Executive Defy Status quo

Accept Alternative policy

Defy Status quo/ Public punishment

A Model of Constitutional Review and Case Promotion

35

decisions can be successfully promoted. The logic here is that the promotion of all cases largely renders it impossible for the media to recognize clearly why one case is more worthy of coverage than another. The key point is that promoting a case is costly in the model, although this cost may be arbitrarily small. Maintaining the assumption that constitutional politics is ultimately about policy, I assume ac > k.5 I further assume that, if the court promotes the resolution, the media accurately translates information about the case to the public. Clearly, this is a strong assumption, but it reduces the set of parameters, and relaxing it does not materially affect the implications of the model. What matters is that case promotion at least increases the chances of quality media coverage.6 I return to this issue after describing the results. The remaining structure of the model is identical to the baseline. A strategy for the court continues to assign a single action to its single information set; however, its action set now takes the following form: Ac = {Uphold; Uphold and Promote; Strike Down; and Strike Down and Promote}. An executive strategy assigns an action, Comply or Defy, to each of its four singleton information sets. Like the baseline model, there is a unique subgame perfect equilibrium for any given values of the parameters. There are four cases to consider. k Case 1: Judicial Deference For ae > b & p ≤ e sc = Uphold the status quo. se = Defy all unfavorable verdicts.

k e sc = Uphold the status quo and Promote the decision. se = Defy all unfavorable verdicts. c + ac − k Case 3: Public Judicial Power For ae ≤ b & p ≤ c + ac − e sc = Strike down the status quo and Promote the decision. Case 2: Public Judicial Deference For ae > b & p >

5 This assumption reduces the number of cases I consider. Allowing k > a does not affect c

the incentives to promote decisions when courts either uphold or strike down; however, it would allow for some equilibria in which courts strategically uphold the status quo in order to avoid having to pay the cost of promotion, a result that only obtains for small ac . Because policy is generally perceived to be paramount, the model assumes away this possibility. 6 Davis (1994) provides empirical support for this claim in the United States, and my own investigation of the Mexican data, which I brieﬂy discuss below, supports it as well.

36

Judicial Communication and Judicial Power se = Defy if the media fails to cover the decision and Accept otherwise. c + ac − k Case 4: Judicial Power For ae ≤ b & p > c + ac − e sc = Strike down the status quo. se = Defy if the media fails to cover the decision and Accept otherwise.

Figure 2.4 depicts the promotion model equilibria. The top panel (Figure 2.4a) shows the results when the cost of promoting a case is low relative to the cost of poor reporting, while the bottom panel (Figure 2.4b) shows the results under the opposite assumption. First, note that for sufﬁciently important policies (ae > b), judicial decision making looks nearly identical to the baseline model. In both the Judicial Deference and Public Judicial Deference cases (shaded dark in the ﬁgure), the executive will defy all unfavorable decisions, whether or not the public is informed about the deﬁance. Under such scenarios, the court defers to the government’s authority, recognizing that public support is insufﬁcient to induce compliance. This is true no matter how likely the court is to believe that the media will cover the decision. Still, the court’s beliefs about ex post media coverage distinguish between the two judicial weakness equilibria. As long as the cost of case promotion is sufﬁciently low, as it is on the top, the court only promotes its decision when the media is likely to cover the case (i.e., when p > ke ). It does so in order to resolve the problem associated with poor reporting of the decision’s rationale. Once it becomes sufﬁciently unlikely that the media will report on the case, the court simply upholds the status quo and avoids the cost of promotion. When the cost of case promotion is large relative to the errors the court anticipates in the media, as is true in the bottom panel, if the court upholds the status quo, it will never promote the case. This suggests that even when case promotion cannot solve the transparency problem, courts may use case promotion to advance legitimacy, but only if the administrative costs of doing so are not prohibitive. A key difference between the promotion and baseline models concerns the effect of transparency on decision making. In the promotion model, although it is still the case that the court will uphold policies that are highly important to the executive, it is no longer the case that it will strategically defer to the government when the probability of ex post c −k media coverage is too low. Instead, when p ≤ c+a c+ac −e , the court will strike down the status quo and create its own media coverage via case

A Model of Constitutional Review and Case Promotion (a) Equilibria with low promotion costs (k<e) pS* =

c + aC – k c + aC – e

Probability of media coverage (p)

pU* =

Public judicial deference

k e

Public judicial power Judicial deference

b Cost of alternative policy to government (ae) (b)

Equilibria with high promotion costs (k≥e) p* = U

k e

Probability of media coverage (p) c + aC – k * p = S c + aC – e

Judicial power

Judicial deference Public judicial power

b Cost of alternative policy to government (ae)

ﬁgure 2.4. Equilibria (Promotion Model)

37

38

Judicial Communication and Judicial Power

promotion. This Public Judicial Power case is depicted in left side of the c −k left panel. Alternatively, if p > c+a c+ac −e , the court strikes down the policy yet saves the cost of case promotion, relying on the media to provide the public with the information necessary to induce compliance. It is useful to note that when the court intends to strike down the status quo and the cost of case promotion is low (the left panel), the condition on the court’s beliefs about media coverage is met for all values of the parameters; however, when the cost of case promotion is relatively high, this condition is only met for sufﬁciently low probabilities of ex post media coverage. Substantively, when case promotion is relatively costless, the court promotes all decisions striking down the status quo; however, when case promotion is relatively costly, the court promotes decisions striking down the status quo only if the media is sufﬁciently unlikely to cover the resolution. I will return to this issue below. In summary, the key point is this: if the transparency of the cases the courts resolve is endogenous to judicial–government interaction, then judges have a limited degree of control over their own power. I am now in position to consider the empirical implications of the promotion model. Promotion Implication 1: When courts can inﬂuence the transparency of the cases they resolve, they should be less likely to strike down policies as the importance the government assigns to the policy increases. Promotion Implication 2: When courts can inﬂuence the transparency of the cases they resolve, there should be no relationship between beliefs about transparency and decision making. The ﬁrst implication is perfectly consistent with the baseline model. Whether or not judges can inﬂuence the transparency of their cases, judicial decision making is affected by the importance of the policy to the government. The second implication is distinct. When courts have control over transparency, there should be no relationship between beliefs in transparency as provided by the media and decision making. When policies are sufﬁciently important, these beliefs are irrelevant because transparency does not offer courts sufﬁcient leverage to exercise their authority. On the other hand, for less important policies, courts can create transparency when they need it. Consequently, their beliefs over whether the media will cover their resolutions should not be related to their decision-making process.

A Model of Constitutional Review and Case Promotion

39

When courts can inﬂuence the transparency of the cases they resolve, noncompliance depends on the cost of case promotion. Promotion Implication 3: If the cost of case promotion is trivial, ofﬁcials should never be observed in deﬁance of constitutional courts. If the cost of case promotion is more than trivial, ofﬁcials should be more likely to defy constitutional court decisions over relatively unimportant policies that are not covered. This implication has two parts. The ﬁrst concerns what we should observe if promoting a case is effectively costless. The second concerns what we should observe if case promotion is relatively costly. The logic is as follows. If the cost of case promotion is small relative to the errors judges perceive in media coverage, then the model suggests that the court will publicize all cases striking down the status quo. Of course, it is still the case that the court only strikes down policies that are insufﬁciently important, so that if the public is informed, the government will comply. Because the court provides transparency for its decisions when transparency is required, there should be no instances in which the government will defy the court in equilibrium. In contrast, if case promotion is relatively costly, then the promotion model suggests the same implication as the baseline model. So far I have considered the implications of the baseline and promotion models for constitutional decision making and compliance. Yet the core reason for constructing a model of case promotion is to explain promotion itself. This is something that existing theoretical models of constitutional review cannot do, because they do not consider the possibility. Promotion Implication 4: Constitutional courts should be more likely to promote decisions that strike down the status quo than those that uphold it. This relationship should be especially strong when the court believes that the case is unlikely to be transparent. The intuition behind this result is as follows. Using the results described under the equilibrium summary above for the promotion model, let pU = c+ac −k k S e and p = c+ac −e . The court will promote a resolution upholding the status quo if and only if p ≥ pU ; it will promote a resolution striking down the status quo if and only if p < pS . Whether the court upholds the status quo or strikes it down, it faces the costs imposed by imprecise media coverage (e); however, when the court strikes down a policy, it

40

Judicial Communication and Judicial Power

confronts the additional political problem of noncompliance. If the cost of promotion is sufﬁciently small, there will be cases in which the court promotes decisions upholding the status quo, but every time it is optimal to promote a decision upholding the status quo, it is optimal to promote a decision striking it down. If we continue to assume that k is small enough so that pU can be satisﬁed, at high values of p, the court will promote all kinds of decisions, resolving reporter inaccuracy; but at low values of p, the incentive to promote in order to address inaccuracy disappears because it is unlikely that there will be any inaccuracy to correct. At such values of p, only decisions striking down the status quo will be promoted in order to deal with potential noncompliance. Thus, constitutional courts should be more likely to promote cases striking down the status quo than those upholding the status quo, yet this relationship should be especially strong when the court is unlikely to believe that the media will cover the resolution. Formally, recognize that if there is a p that satisﬁes the pU condition (and the court promotes a decision upholding the status quo), then the same p satisﬁes the pS condition (and the court promotes a decision striking it down); however, the converse is not true. There are two possibilities. Either the cost of promotion is less than the cost of media inaccuracy, or it is not (i.e., k < e or k ≥ e). If k < e, then pU ∈ (0,1) and will be satisﬁed for sufﬁciently high p. If k < e, however, pS is satisﬁed for all c −k parameter values, because c+a c+ac −e > 1. Alternatively, if k ≥ e, the court will never promote a decision upholding the status quo, because pU ≥ 1, and no p is this large; however, k ≥ e, pS ∈ (0,1), and it will be satisﬁed for sufﬁciently low p. Summary of models and their implications Although the preceding arguments are simple, they suggest a number of observable implications across a range of substantive topics. The central difference between the two models concerns the degree to which judges can inﬂuence their media coverage. In the baseline model, media coverage is exogenous; in the promotion model, it is endogenous to the interaction. In some cases, these different approaches produce exactly the same observable implications; however, in other cases, they suggest hypotheses that are distinct. Table 2.1 summarizes the key hypotheses that I will examine in the second part of this book. Check marks are placed in the ﬁnal two columns in order to indicate the modeling approach that generates the hypotheses to the left.

A Model of Constitutional Review and Case Promotion

41

table 2.1. Observable Implications by Model Type Subject

Hypothesis

Case Promotion

A constitutional court is more likely to promote a decision striking down the status quo than a decision upholding the status quo A constitutional court is less likely to strike down a public policy as the importance of the policy to the government increases A constitutional court is more likely to strike down a public policy as likelihood of ex post media coverage increases, but only for insufﬁciently important policies The likelihood of ex post media coverage is unrelated to a constitutional court’s decisions to strike down public policies Governmental noncompliance should not be observed Governmental noncompliance should be more likely in relatively unimportant cases that are not covered by the media

Constitutional Decision Making

Constitutional Decision Making

Constitutional Decision Making

Compliance Compliance

Baseline Promotion Model Model √

√

√

√

√

√ √

√

∗

∗∗

Note: Summarizes the observable implications of the baseline and promotion models. √ Checks ( ) indicate the model that produces the hypothesis listed to the left. *If case promotion is costless. **If case promotion is costly.

Having reviewed the key empirical implications of the model, it is important to return to the assumption about the court’s control over transparency if it chooses to promote a case. A natural alternative modeling approach might allow the court to successfully promote its cases probabilistically. What are the consequences of such an assumption for the key results? For a sufﬁciently low probability of case promotion, the model collapses into the baseline and all of the predictions in that model apply. Above this threshold, the alternative collapses into the promotion model, and all of its predictions apply. Importantly, no matter how

42

Judicial Communication and Judicial Power

likely it is that a court can inﬂuence its coverage, there is no material effect on the relationship between case salience and the incentive to defer strategically. Similarly, a court is always more likely to promote a decision striking down a policy than upholding a policy, especially when the media is unlikely to cover the resolution.7 The essential consequence of relaxing the assumption of perfect promotion at a cost is that the promotion model would retain some features of the baseline model. The bottom line is that it might be more difﬁcult to ﬁnd support for the case promotion hypothesis if courts believe they can only partially control their coverage.

implications At the beginning of this chapter, I suggested that both media inattention and media attention might produce distinct problems for constitutional judges. The problem of media inattention is directly implied by existing theoretical models of judicial power in which the public is an enforcement mechanism for judicial resolutions. If individuals derive information about national politics from the media, and the media ignores the constitutional court, then monitoring will be impossible and the enforcement mechanism will fall apart no matter how beloved the court. Insofar as that is true, media inattention undermines the use of judicial power. The transparency problem Can judges solve this problem through public relations? If they can, then this suggests that judges can be active participants in the construction of their own power. Judges, on this account, do not merely respond to existing conditions, rather they attempt to inﬂuence those conditions whenever 7 There is one case, which only exists in a model in which promotion is probabilistically

successful. When the value of the policy to the government is insufﬁciently high, when the costs of promotion are sufﬁciently low, when the costs of noncompliance are sufﬁciently high, when the value of the policy is sufﬁciently low for the court, and when both the probability of ex post coverage and the probability of successfully promoting a case are neither too large nor too small, there can exist an equilibrium in which the court will strategically uphold a policy and promote its decision. The logic is that if the probability of successful promotion is too low and ex post media coverage is too unlikely, then it is likely that an effort to strike and promote will fail. The court wishes to avoid this. However, because the probability of ex post coverage is sufﬁciently high, the court wishes to avoid poor reporting of its rationale. Finally, because the probability of success is not too small, it is worth making the effort to try to inﬂuence its coverage, but again, only for decisions strategically upholding the status quo.

A Model of Constitutional Review and Case Promotion

43

possible. If the models I have just developed are correct, then the answer is that public relations can inﬂuence judicial power, but not under all conditions. In particular, the degree to which judicial public relations can increase judicial power depends on the importance of the status quo policy to the government relative to the cost the government must pay for deﬁance. When policy importance is low relative to the cost of deﬁance, the court has a great deal of leverage over the government, and it can use public relations to ensure its power. However, when the policy’s importance is high relative to the cost of deﬁance, the court will have far less leverage, and public relations as a tool of judicial power will be of less use. In other words, judicial public relations does not expand the boundaries of judicial power outward arbitrarily – its effect is limited. From the perspective of research design, this limitation reminds us that tests of the model should be conducted at the level of the constitutional conﬂict. Although there is absolutely nothing wrong with a research design that includes multiple countries, the interesting variance will be at the level of the conﬂict, not say at the country or even countryyear level. In summary, judges can play an active role in constructing the conditions for judicial power, yet that role is limited by the value of the status quo. The legitimacy problem The second problem I identiﬁed above concerns the problem induced by media attention. The idea was that the press frequently misinterprets jurisprudential rationales, which is both professionally unappealing and may even undermine judicial power. If written opinions are a primary means by which judges communicate impartiality, and if impartiality is a core element of judicial legitimacy, then it is immediately obvious that judges should take an interest in ensuring that their reasoning is properly communicated to the public through the press. In contrast to the transparency problem induced by media inattention, this problem is paramount when the media is expected to cover a case. Can public relations solve this problem as well? As before, the answer is a qualiﬁed yes; however, the limit to public relations’ success in solving this problem is uniquely related to the costs of case promotion. As long as the cost of promoting a case is sufﬁciently low, then courts can always resolve this problem via promotion. In general, this idea suggests that courts have an interest in lowering the costs of case promotion. One way courts might tackle this problem is

44

Judicial Communication and Judicial Power

by attempting to construct their own media coverage. As I discuss in subsequent chapters, the Mexican Supreme Court took an extremely active role in attempting to train its beat reporters to provide better coverage. In terms of the theoretical argument, this practice might appear useful inasmuch as it can lower the costs of case promotion. Indeed, it ought to be easier to communicate with a reporting corps as their level of familiarity with the law increases. This seems to have been Justice Kirby’s rationale for seeking a change in the way the national media covered the Australian High Court. Although this logic seems plausible, it is certainly possible that constructing a beat may increase these costs. Although a better-informed media is more likely to be able to understand a judge’s jurisprudential rationale, it is also possible that this better-informed media would be less willing to simply report a court’s description of its decisions and instead rely on its own interpretation. If this is true, then constructing a beat may have no inﬂuence on the net costs of case promotion. Although greater information may lower those costs by making any particular message from the bench easier to communicate, greater professionalization may decrease the probability that the media is willing to listen to a judge’s appeals. To summarize, the theoretical argument suggests that judicial public relations offers a solution to both the transparency and legitimacy problems, and as a result, judges have control over their own power, even if this control is bounded. In the subsequent chapters, I test a number of the implications that have emerged from the argument. Much of the empirical work takes place in Mexico, and it is to that setting that I now turn.

appendix 2 Baseline model The solution concept is subgame perfect equilibrium. If indifferent, I assume that the court upholds the policy and the executive accepts. Consider the executive’s choice in the baseline model ﬁrst. If ae > b, then the executive always deﬁes the court. If ae < b, then the executive complies if the media has not covered the case and deﬁes if the media has covered the case. Consider that the expected utility for the court of upholding the policy is –pe whatever the executive does. If the executive deﬁes decisions only if the media does not cover the case, then the court’s expected utility of striking down is p(c + ac − e) − c; however, if the

A Model of Constitutional Review and Case Promotion

45

executive deﬁes all decisions, then the court’s expected utility of striking down is −pe−c. Clearly, if ae > b and the executive deﬁes always, the court must uphold the policy. To do otherwise only adds an additional cost. This establishes Judicial Deference case. Now, if ae < b, then the court’s choice depends on whether p(c + ac − e) − c< −pe, which it is if c . In the event that this condition holds, then the court upholds; p < c+a c however, if it does not, the court strikes down. This simple condition on p establishes the Informational Deference and Judicial Power cases. Promotion model The executive’s equilibrium choices are identical to those in the baseline: defy all verdicts if ae > b; however, if ae < b, accept verdicts when the media covers the case and defy otherwise. If the court does not promote the decision, its expected utilities are identical to those in the baseline model. If the court upholds the status quo and promotes, it earns −k, the cost of promotion. If ae < b and the court strikes down the status quo and promotes it receives ac − k; however, if ae > b and the court promotes a decision striking down the status quo, it expects −c − k. Clearly, if ae > b, striking down is never better than upholding for the reason given in the baseline model. Similarly, if ae > b, then striking down and promoting, which yields −k − c,is never better than upholding and promoting, which yields −k. Thus, the only question is whether the court will promote a decision upholding the status quo or not. This it will do if and only if p > ke . If this condition does not hold, the court simply upholds the policy. These results establish the Judicial Deference and Public Judicial Deference cases. Finally, consider what happens if ae < b and the executive will comply if the media covers the case. Upholding and promoting can never be optimal, because it would yield only −k and the court could obtain ac − k by striking down and promoting, which is larger by the restriction on ac . Further, simply upholding is also never optimal because ac − k is clearly larger than –pe. Thus, the only question is whether the court will promote a decision striking down the status quo. This involves asking when ac − k < p(c + ac − e) − c, which it is if and c −k only if p > c+a c+ac −e . If this condition holds, then the court will strike down the status quo but fail to promote, relying on exogenously determined media coverage. If the condition fails, the court strikes down the status quo and creates its own coverage. This establishes Public Judicial Power and Judicial Power cases.

part ii THE POLITICS OF CONSTITUTIONAL REVIEW IN MEXICO

In many respects, July 2, 2000, the day on which Vicente Fox’s presidential campaign ended seventy-one years of Partido Revolucionario Institutional (PRI) rule, marks a watershed in Mexican political history. After July 2, it was impossible to claim that no meaningful opportunity existed for national representation outside the broad yet ultimately exclusive structure of the PRI. The election suggested that power could be transferred peacefully between parties via fair democratic procedures. Patterns of political participation also changed in 2000. Historically, the educated and politically aware participated at levels far below what standard turnout models predict. The PRI strategy of mobilizing poor and rural voters through clientelistic networks and corporatist state structures induced a participatory pattern reﬂective of broad cynicism in the electorate. In 2000, however, participation patterns reﬂected those of advanced democracies, indicating that Mexicans came to believe en masse that their votes would count (Klesner and Lawson 2001). Certainly, July 2 was a big day. Yet, although the Fox victory serves as a useful symbol of the transition, treating the date as a true critical juncture misses two important elements of the democratization process. Most scholars of Mexico perceive the transition to have been protracted, playing out in a variety of ways (e.g., development of signiﬁcant opposition parties, increased political competition at the state level, PRI congressional losses in 1997, and so on) roughly during the ﬁfteen years preceding the Fox election (Cornelius, Gentleman, and Smith 1989; Shirk 2005; Wuhs 2008). July 2 is probably best thought of as wholly consistent with a general process of democratization rather than the date on which Mexico became a democracy. But beyond the distinction between a gradual

48

The Politics of Constitutional Review in Mexico

and a punctuated change, the second element of the story that is blurred by focusing on the Fox election is a legal one. In the sense that democracy can be as much about broad adherence to the rule of law (e.g., Linz and Stepan 1996, 10) as it is about inclusive and competitive elections (e.g., Dahl 1972), the story of the Mexican transition should not be restricted to tales of increasing electoral competition. Rather it should also consider the sustained process of judicial reform aimed at constructing a more effective counterbalance to presidential or even legislative power. And that process began in earnest long before the 2000 federal elections. Responding to a growing public concern over injustice, insecurity, and judicial corruption, and in the wake of a series of political assassinations, Ernesto Zedillo, the PRI’s candidate for the presidency in 1994, developed a bold plan for the reformation of the Mexican justice system (González Oropeza 1995). And reforming the judiciary was Zedillo’s ﬁrst order of business after taking ofﬁce. The Zedillo reform reduced the size of the Supreme Court from twenty-six to eleven ministers (justices) and reduced the number of benches from four to two.1 As part of its transitory provisions, all current members of the Supreme Court were forced to resign, paving the way for the appointment of an entirely new court, which was to be staffed by judges more committed to the judicial career (Staton 2007). Zedillo also created the seven-member Federal Council of the Judiciary (Consejo de la Judicatura Federal [CJF]) to relieve the Supreme Court of much of its administrative responsibilities.2 Critically, the reform altered the Supreme Court’s constitutional jurisdiction. It created a new institution of abstract constitutional review, the action of unconstitutionality, and expanded the court’s authority under the constitutional controversy, a constitutional action largely designed to resolve federalism and separation of powers questions. Two years later, the Supreme Court was granted jurisdiction over cases challenging the constitutionality of electoral laws. Scholars have interpreted these reforms through the lens of two related, willful delegation models of judicial power. Finkel (2008) views the Zedillo reform as a form of political insurance, a delegation of presidential authority to hedge against possible losses of power in the ongoing process of democratization. The Supreme Court was granted new powers 1 The benches previous to the reform separately specialized in civil, penal, administrative,

and labor matters. Under the new conﬁguration, the ﬁrst bench hears penal and civil cases, while the second bench hears labor and administrative cases. 2 CPM, Artículo 100. For a review of judicial councils in Latin America see Héctor Fix Zamudio and Héctor Fix Fierro, El Consejo de la Judicatura (México: Instituto de Investigaciones Juridicas, UNAM, 1996).

The Politics of Constitutional Review in Mexico

49

of judicial review to ensure that core party interests would not be undermined in the event that the PRI lost a major election in the near future. Magaloni (2003) instead argues that the reform reﬂected an effort to manage political conﬂicts in a context of increasing heterogeneity of representation at the state level. Where once the president could resolve disputes within or across state boundaries, or between states and the republic, through his control over the national party (and by implication his control over local parties), as opposition parties came to win elections across the republic, this mechanism broke down. According to Magaloni, Zedillo empowered the Supreme Court to avoid constant political crises. Whatever the rationale for the reform, it is important to note that both arguments rest on an assumption around which this book revolves: the Mexican Supreme Court would exercise meaningful control over the conﬂicts it would be called upon to resolve. Absent de facto power, it is not clear how the court could have been conceptualized as “insurance” against electoral misfortunes or as a successful arbiter of future political conﬂicts. A cursory look at the institutional history of the Mexican federal judiciary suggests that the Zedillo appointees themselves would have been justiﬁed to question their authority. Generally speaking, major institutions governing the Supreme Court’s tenure and jurisdiction have been malleable. The size of the Mexican Supreme Court was changed ﬁve times between 1917 and 1994, and the ministers’ tenure changed seven times between independence and the Zedillo reforms. More broadly, the primary constitutional provisions concerning the federal judiciary were amended sixty-nine times during the twentieth century (Carranco Zúñiga 2000, 97). Scholars have suggested that judicial reforms in 1928 and in 1934 should be understood as political responses to undesirable judicial qualities. The 1928 reform increased the size of the Supreme Court from eleven to sixteen and altered the appointment process. Domingo writes: The 1928 reform was welcomed by the more progressive elements of the Revolution as a way of controlling what at the time was seen as a reactionary judiciary, that served the interests of the regional caudillos and landlords, to the detriment of the revolutionary reforms. (Domingo 2000, 713)

The 1934 reform, which eliminated life tenure, also appears to have been designed to control a judiciary out of step with the political philosophy of the day. On President Lázaro Cárdenas’s understanding of life tenure for judges, Baker writes: Such a concept of judicial tenure . . . was inconsistent with the revolutionary principle of limited terms without re-election and derogated from the sovereignty that

50

The Politics of Constitutional Review in Mexico

properly resided in the people. The practice of the courts and their legal views, their exaggerated reverence for tradition and established precedent, was only too well known, whereas the administration of justice in the new Mexico properly necessitated the ﬂexible, expedient, spontaneous approach of the Revolution. (Baker 1971, 57)

The 1958 reform to the Supreme Court’s internal structure was explicitly designed to control an overly active court. The reform amended Article 11 of the Organic Law of the Judicial Branch and required the Supreme Court to sit en banc in review of decisions questioning the constitutionality of a state or federal law. Previously, the court had been empowered to hear such claims in benches. The motivation of the reform appears to have been the Supreme Court’s administrative bench, which was striking down federal laws in amparo suits with too high a frequency (Baker 1971, 73). The reform seems to have had an effect, as Schwarz claims: The reluctance [of the Court] to void governmental enactments on constitutional grounds [during the 1970s] is primarily the result of an unfavorable congressional response in 1958 to this style of judicial activism: by granting monopoly power to the Plenary Court over all “constitutionality amparos,” the Congress ensured that it would be an awkward and rarely successful remedy. (Schwarz 1973, 313)

Events following the Zedillo reform suggest the reasonableness of assuming that the new members of the Mexican Supreme Court would have perceived a potential for political attacks on the judiciary designed to undermine its authority. Indeed, as recently as December 2004, members of the PRI called for the impeachment of two Supreme Court ministers for having accepted for review a constitutional action in which President Vicente Fox challenged the constitutionality of a congressional override of the federal budget.3 In light of its institutional history, it would not have been surprising if the Zedillo appointees wondered about the limits of their power in early 1995. It is for this reason that Mexico, during the period of democratization, offers an excellent opportunity to evaluate the predictions of the model developed in Chapter 2. Before subjecting explicit empirical predictions to empirical analysis with data on Mexican Supreme Court decision making, it is worth asking what the men and women charged with the task of providing political insurance or managing party disputes 3 It is interesting to note that Alberto Núñez Esteva of the Mexican Federation of Employers

publicly came out in opposition to impeachment, defending the institutional integrity of the federal judiciary. See “Lamentan empresarios ataques a Corte” [Businessmen Regret Attacks on the Court], El Universal, January 18, 2005.

The Politics of Constitutional Review in Mexico

51

set about doing in the early days of their tenure. What might we expect from a court with new powers of constitutional review in a context of a democratic transition and following a history of relative judicial subjugation? As it turns out, one activity involved the development of a coherent public relations strategy. This part of the book probes the empirical implications of the argument developed in Chapter 2 through an analysis of the Mexican Supreme Court at the turn of the twenty-ﬁrst century. Chapter 3 details the Supreme Court’s public relations work during the period of Mexico’s transition. Chapter 4 then provides an analysis of constitutional review, case promotion, and compliance on the Mexican Supreme Court. The primary implication of this research is that, although the democratization process opens opportunities for the creation of a powerful court, judicial power is not legislated. Courts are not powerful because politicians endow them with formal powers of constitutional review. Instead, judicial power emerges out of strategic interactions between courts and political ofﬁcials within the context of the normal politics of political representation. Public support for courts and the pressure constituents can place on their representatives for failing to respect the rule of law is essential to judicial power. For this reason, judges have strong incentives to get the public relations right. These observations speak not just to theories of judicial power, but also to theories of democratic transition and judicial reform. Periods of democratization represent opportunities for courts to develop their authority by engaging the public; however, public engagement will not turn weak courts into strong courts overnight. Judges in democratizing states still confront signiﬁcant political obstacles. Recognizing that judges can construct judicial power, but that power is sensitive to external political conditions, suggests that new willful delegation models of judicial reform, perhaps building on the accounts given by Finkel or Magaloni, might beneﬁt from rolling expectations about the future politics of constitutional review into the calculus of institutional design.

3 Public Relations on the Mexican Supreme Court

Roughly six weeks after Fox won the Mexican presidency, his transition team on justice and public security, led by Senator Francisco Molina Ruiz and José Luis Reyes, joined members of the Supreme Court for breakfast and a conversation about the president-elect’s judicial reform agenda. Fox proposed moving control over the federal agrarian, labor, and administrative courts from the executive branch to the administrative arm of the federal judiciary. More importantly, he proposed transforming the federal attorney general’s ofﬁce (PGR) into something like the U.S. Justice Department and creating a new cabinet position on security and justice services.1 The meeting garnered signiﬁcant national media attention; however, none of Fox’s proposals reﬂected the Supreme Court’s concerns.2 Still, Supreme Court President Genaro Góngora took advantage of the well-publicized opportunity and hosted a midmorning press conference. Góngora argued for a constitutional amendment guaranteeing the federal judiciary a yearly appropriation level between 3 and 6 percent of the federation’s total expenditures. Budget reform was a pressing institutional matter for the Supreme Court in light of crushing caseloads and a budget allocation that had dropped below 1 percent of total federal spending. 1 See Daniel Lizárraga, “Negocian hoy reformas Ministros y foxistas,” August 17, 2000,

Reforma, 2; Jesús Aranda, “El Poder Judicial y el equipo foxista ‘uniﬁcan criterios’ en enmiendas de ley,” August 18, 2000, La Jornada, 7; Mario Torres, “Hoy, reunión del Poder Judicial con equipo de VFQ,” August 17, 2000, El Universal, 7A. 2 Although Minister Román Palacios publicly attacked the plan in the weeks preceding the meeting, the full court publicly welcomed a serious investigation. See Daniel Lizárraga, “Descaliﬁcan plan judicial del panista,” August 11, 2000, Reforma, 2A.

53

54

The Politics of Constitutional Review in Mexico

Góngora presented the transition team with a 13-page comparative analysis of judicial budgets in Latin America3 and contrasted Mexico’s budget with those in countries like Costa Rica, Honduras, and Guatemala, all of which had constitutionally ﬁxed appropriation levels. Endeavoring to capitalize on the contemporary narrative of a genuine separation of powers that followed the transition from PRI rule, Góngora reminded his audience that, “the judiciary is the only branch of government that fails to possess direct mechanisms to safeguard its economic necessities, leaving it clearly disadvantaged with respect to the other branches.”4 Four major issues of policy reform led the following day’s coverage – Fox’s three structural initiatives and the Supreme Court’s proposal to guarantee the judiciary a ﬁxed proportion of federal spending. What was supposed to be a story about the president-elect’s intention to restructure the investigatory arm of the federal government became a story about the institutional impediments to judicial independence. With this success in mind, the current chapter describes the Mexican Supreme Court’s public relations activities during the transitional period. It reviews two key problems the court confronted following the Zedillo reforms and then summarizes the tactics the court used to address those problems.

legitimacy and transparency problems The Supreme Court judges Zedillo appointed, referred to as “ministers” in Mexico, desired a society characterized by the rule of law, and they believed that the Supreme Court’s chief function in such a society involved authoritatively shaping constitutional meaning.5 Unfortunately, from the 3 This brief study is entitled Debilidad constitucinal en el Presupuesto de Egresos del

Poder Judicial de la Federación. The document may be obtained from the judiciary’s Ofﬁce of Social Communication upon request (Comunicación Social, www.cjf.gob.mx/ comsocial/default.asp). 4 Mario Torres, “Demanda SCJN 6% del presupuesto federal,” August 18, 2000, El Universal, 7A. 5 Ministers express this general preference in publications and speeches. For example, Minister Aguirre Anguiano (Comunicado, 439), in a speech commemorating the ﬁftyyear anniversary of the creation of the Collegial Circuit Tribunals, urged his audience to recognize that the Supreme Court was already a constitutional court in the European sense of the term. He stated, “Of course [the Supreme Court] is a constitutional tribunal; it is a tribunal eminently constitutional because it resolves conﬂicts between organs of the states, resolutions which always effect the regulation of the constitution . . . . Whomever argues that a constitutional tribunal is needed in addition to the Supreme Court, does not much understand the range of the Court’s jurisdiction.” Minister Sánchez

Public Relations on the Mexican Supreme Court

55

perspective of a public enforcement mechanism of constitutional control, the ministers inherited a difﬁcult political landscape. The ﬁrst problem was society’s belief that the federal judiciary was likely subservient to the executive branch, a belief that undermined the Court’s legitimacy. This belief was more or less apparent to the ministers. On December 15, 1995, in his ﬁrst annual report on the state of the judiciary, then Supreme Court President José Vicente Aguinaco Alemán claimed, “Without hesitation, we have never avoided the responsibility to pursue the objectives that gave life to 1994 judicial reform. Our commitment with Mexico was, and will continue to be, recovering the lost conﬁdence in the federal courts” (Aguinaco Alemán 1998, 22–23). Minister Humberto Román Palacios, in an interview with the daily magazine Milenio, admitted, “I have worked in the federal judiciary for more than twenty-ﬁve years; and if I denied the perception that it has been subordinated to the executive, I would be denying a fact that was relatively notorious.” In a 1999 interview, Minister Guillermo Ortiz Mayagoitia contended, I cannot say whether the Court was previously subordinated to the President of the Republic. But, what I can say is that this Supreme Court has never been subordinated to either of the other two powers of government . . . . From the bottom of my heart, it would be contrary to my human nature to say that I am not thankful to the president for having thought of me as a candidate, but this does not inﬂuence my decisions. Our duty is to the Constitution and to the law.6

The ministers do not shy away from sensitive political language in the context of the subservience question. Indeed, they commonly argue that constitutional review is inherently political, but that it is political only in the sense of resolving major questions of public policy. They stress that the procedures they use to reach their decisions largely purge partisan interests Cordero delivered a similar speech in 1997 at a gathering commemorating the founding of the Universidad de la Salle (Comunicado, 18, Coordinación General de Comunicación Social, Suprema Corte de Justicia de la Nación). See also Aguinaco Alemán (1998) on the subject. Furthermore, many of the court’s Acuerdos on the federal judiciary’s internal administration are justiﬁed as means of ensuring that the Supreme Court can function as a true constitutional court. See, for example, Acuerdo 6/1999, whose exposition of motives states, “It is essential to permit the Supreme Court, as happens in other nations, to concentrate all of its efforts on the recognition and resolution of new issues or on those issues of such high importance that their resolution will inﬂuence the interpretation and application of the national judicial order.” Also see Acuerdos 6/2000 and 9/1999, which both advocate the Supreme Court’s position as a constitutional court. 6 Óscar Camacho Guzmán, “El presidente de México no tiene privelegios en la Suprema Corte,” October 11, 1999, Milenio Diario, 39.

56

The Politics of Constitutional Review in Mexico

from constitutional review. Responding to a question from Época magazine in 2000 about whether the court was “politicized,” Genaro Góngora responded, If by politicized we mean that [the court] forms a part of Mexican political life, then of course it must be politicized – it is one of the three political powers of the state. We should not be afraid of those words. Now, if by politicized we mean sharing the interests of a party, the answer is no. The day on which . . . the federal judiciary adopts a [political] party will be the day that [it loses its] reason for being. (Góngora Pimentel 2000a, 14)

Minister Juan Silva Meza’s remarks reﬂect this sense of politicization: The Court, I insist, is a political organ, it is a political branch of the state, and it is carrying out its function as the instrument of constitutional control, which makes it an instrument of political control. . . . But I reiterate, political in the purest sense of the word, not in the sense that is sometimes used in daily language. (Silva Meza 2000a, 197)

Minister Olga Sánchez Cordero has even presented an empirical analysis of the Court’s decisional patterns, which she contends demonstrate its independence (Sánchez Cordero 2000). She ﬁnds that the Supreme Court has been more likely to support amparo claims brought against federal executive ofﬁcials than against other kinds of authorities.7 A related, and likely more signiﬁcant, problem for the court involved transparency. In the prereform days, the media largely ignored the federal 7 Indeed, of the 291 times that the Supreme Court found on behalf of an amparo com-

plainant between January 1, 1997, and March 28, 1998, the responsible authority was federal in 181 cases (62.2 percent) and state or local in only 110 cases (37.8 percent). Sánchez Cordero infers that the court has been “clearly independent” of the federal executive (12). Unfortunately, these data do not warrant this inference. Although we know that the Supreme Court conceded more amparos brought against authorities associated with the federal executive than with other branches or levels of government, we do not know the total number of challenges brought against both types of authorities. Simply put, we do not have the appropriate denominator. Sánchez Cordero does provide aggregate data on the total number of amparo appeals resolved during the time period studied, however, she does not indicate which cases were brought against the various types of authorities about which she wants to draw inferences. Without this information, we cannot properly evaluate her claim. It may very well be the case that the court was more likely to rule against the federal executive than other authorities, but unless we know more about the total number of claims brought against federal and state authorities, we cannot draw this inference. In fact, the full court hears far more claims against federal authorities than state authorities. In light of this disparity, any inference drawn about the relative likelihood of the court to challenge the authority of different types of public ofﬁcials must be drawn from data that accounts for the difference in constitutional challenges made against those different ofﬁcials.

Public Relations on the Mexican Supreme Court

57

judiciary. Jesús Aranda, one of the ﬁrst Supreme Court beat reporters, described the situation as follows8 : Before [the reform] the Court was very closed-off. There was a social communications ofﬁce that would suddenly issue some press release [on some issue] but reporters did not go to the Court – in the end nothing was known or understood about the Court. This reﬂected the judiciary’s situation in the country. The judicial branch was always seen as an appendage of the executive – the judiciary did whatever the president wanted.9

Without any coverage, the ministers were obviously incapable of affecting their public image and of leveraging whatever public support they had for compliance pressure. This problem, too, was apparent to the ministers. Minister Sánchez Cordero writes: The relationship that unites the media with the judicial branch is transcendental, since in a society that demands information as a necessity and a right, the informers assume the role of intermediaries between those who carry out the mission of ensuring respect of individual liberties and the society. . . . For that reason, the information that you supply to your distinct media organizations should be transparent, punctual, reliable, as should be the actions of those who impart justice, for in transparency, as I have suggested, is independence.10

By viewing independence as a function of transparency, Sánchez Cordero articulates a core feature of the public enforcement mechanism. The court needed the media to render its conﬂicts transparent, but the court was off the political radar screen in 1995. Clearly, this was a major problem for a court that wanted to control the constitutional order.

public relations tactics The ministers charged their Ofﬁce of Social Communications (DCS) with changing this situation. Formally opened in 1988, the DCS did not begin to aggressively promote the court’s public image until after the Zedillo reforms. In fact, nothing much changed until 1996, when the court named 8 Aranda’s analysis was echoed by both Norma Jiménez of Milenio and Victor Fuentes of

Reforma in interviews with the author. That is to say, the court was largely ignored prior to 1995. Both Jiménez and Fuentes joined Aranda as the court’s ﬁrst daily correspondents in the ﬁrst years following the reform. At the time, Jiménez was writing for El Universal while Fuentes was with El Finaciero. 9 Jesús Aranda, personal communication, June 6, 2001, Mexico City, Mexico. 10 Olga Sanchez Cordero, “Discurso de la Señora Ministra, Olga Sanchez Cordero, Durante la Comida con los Reporteros de La Fuente, el 3 de Agosto de 2000, en la Suprema Corte de Justicia de la Nación,” August 14, 2000.

58

The Politics of Constitutional Review in Mexico

Gerardo Laveaga Rendón general director. Laveaga, in turn, brought on Álvaro Vizcaíno Zamora as his chief assistant, charging him with organizing the court’s daily media relations.11 The DCS’s ﬁrst substantive goal was to distance the Supreme Court from the executive branch. One high-ranking member of the DCS admitted, When we ﬁrst arrived, we would get requests from Gobernación [the Interior Ministry] asking to review our press releases before we sent them out. We had to repeatedly let them know that the federal judiciary was an independent branch of government, and if they wanted to see the press releases they could – but we were not going to change a thing based on what they had to say. After some time, they stopped calling.12

The addition of Laveaga and Vizcaíno, attorneys both, also addressed an important point of concern on the court. If ever the media began to cover the Supreme Court, in light of the technical language of its proceedings, it was not clear whether reporters with little knowledge or training in the law could efﬁciently or accurately summarize decisions. As Carranco Zúñiga (2000) suggests, a highly publicized yet ill-described case may actually do more harm than good to judicial independence. Poor reporting may result in strong yet inaccurate public beliefs about the proper resolution of a case. This problem is exacerbated if poor coverage is provided prior to a decision’s release. If the resolution cuts against public expectations, especially in cases in which a government action is being directly challenged, people are left to infer that the judge was unduly inﬂuenced by political concerns. According to Carranco Zúñiga, in order to combat this problem, the judiciary needs to provide the media with an ofﬁcial, well-constructed statement. He writes, “It is important to educate the public about such delicate issues, avoiding if possible the situation where some thoughtless statement made by one of the parties to the case is the only statement on the matter carried by the press” (2000, 294–295). Laveaga’s job was to act as a liaison between the worlds of law and popular culture, and because he was educated in the law, the thought was that he could better describe the court’s positions than a public relations director culled from the press itself, as had been done in the past. A core goal for the DCS, then, became communication of Supreme Court decisions to 11 Jiménez (2000) in particular highlights the important role played by Vizcaíno in terms

of turning around the DCS. 12 Interview DCS(B), with author, August 17, 2000.

Public Relations on the Mexican Supreme Court

59

the media to ensure that cases were covered and that they were covered accurately. In order to further clarify its role within the court, the Supreme Court’s Committee on Social Communications developed a policy statement in 1999 that required the DCS to build conﬁdence in the federal judiciary.13 The committee asked the DCS to create a collective consciousness among the members of the federal courts, highlighting their professionalism, their commitment to public service, and a spirit of excellence. Judges and clerks were encouraged to participate in academic activities and to cooperate with the media in order to portray a “good public image.” The committee also requested that the DCS promote important Supreme Court decisions, clarify the distinctions between the executive and judicial branches, and defend the judiciary’s independence. Laveaga was prohibited, however, from challenging aggressively unfavorable reports in the media and from attempting to purchase better coverage.14 In promoting the court, the DCS has emphasized the technical features of its decision-making process. It has published books and pamphlets summarizing the court’s most salient decisions, describing its internal structure, and highlighting its new role in Mexican politics (Aguinaco Alemán 1997a). It has taken out advertisements in major national newspapers describing the Court’s new powers of constitutional review. It has sponsored educational programs in the public schools, launched essay contests, and even produced a cartoon booklet and video about the federal judiciary. The DCS promoted the court’s institutional reform interests as well. It was extremely active in the days preceding the Fox transition team’s breakfast meeting on the president-elect’s reform proposals – especially in terms of negotiating the set of issues on which the participants would focus. On the day before the breakfast meeting, the DCS prepped Supreme Court beat reporters on the court’s own plan, an important step in paving the way for a more complete coverage.15 The DCS also kept track of the kind of coverage the Supreme Court received in the media. Each day, the ofﬁce distributed an informational packet to all members of the court. The packet contained a copy of every 13 This policy is published in Compromiso, órgano informativo del Poder Judicial de la

Federación, número 1, Julio-agosto 1999, pp. 21–22. 14 Buying coverage from Mexican reporters was a not-all-uncommon practice in twentieth-

century Mexico. See Lawson (2002) for a careful study on Mexican media politics and their change through the country’s transition to democracy. 15 I provide a fuller description of the court’s reform interests in Staton (2007).

60

The Politics of Constitutional Review in Mexico

table 3.1. Social Communication Analysis of Supreme Court Print Coverage, October 1999 Source

El Universal El Financiero Ovaciones La Prensa Excélsior El Heraldo La Jornada Reforma El Sol de México Diario de México Unomásuno El Día La Aﬁcíon Cine Mundial Novedades Crónica El Economista Magazines Proceso Siempre El Mundo del Abogado Milenio Epoca Etcétera Lex Impacto Totals (Row %)

Distribution Articles Positive (#) (#) 160,000 135,000 130,000 120,000 120,000 120,000 118,000 95,000 76,000 75,000 55,000 50,000 50,000

Neutral (#)

Negative Percentage (#) of Total Negative

43,000 40,000 35,000

61 21 23 16 47 25 56 54 24 22 26 9 10 6 26 9 11

2 0 0 1 2 0 0 1 2 1 0 0 0 0 0 0 0

59 21 23 15 43 25 56 53 21 21 26 9 10 6 25 7 11

0 0 0 0 2 0 0 0 1 0 0 0 0 0 1 2 0

0 0 0 0 28.6 0 0 0 14.3 0 0 0 0 0 14.3 28.6 0

130,000 85,000

4 1

0 0 0

4 0

0 1

0 14.3

40,000 25,000 18,000

0 2 0 0 4 0

0 0 0 0 0 0

0 2 0 0 4 0

0 0 0 0 0 0

0 0 0 0 0 0

15,000

457

9 (2%)

441 (96.5%)

7 (1.5%)

Source: Suprema Corte de Justicia, Dirección General de Comunicación Social.

article that even remotely mentioned any part of the federal judiciary. They coded each article according to whether or not it presented the court in a positive, negative, or neutral light. Table 3.1 displays a sample of the DCS’s analysis for October 1999. It is important to note that the court’s coverage was overwhelmingly neutral and that it was concentrated in the larger publications. There is nothing unrepresentative about this ﬁnding.

Public Relations on the Mexican Supreme Court

61

In a report entitled “Who Spoke Well and Who Spoke Poorly of the Supreme Court during the First Half of 1999,”16 the DCS reported that although they coded articles from seventeen distinct newspapers, more than half of the court’s coverage came from just ﬁve national papers: El Universal, Reforma, La Jornada, Excélsior, and El Sol de México. More importantly, the vast majority – 87.2 percent (2,812 articles) – of articles on the Supreme Court were what the DCS called neutral; that is, they merely presented facts without taking a position on the court’s actions. Only 5.6 percent (180 articles) were coded as positive, and 7.2 percent (230 articles) were coded negative. Furthermore, of the 230 articles that the DCS considered negative, 166 or 72.2 percent derived from quotations in the articles from dissatisﬁed litigants rather than from the opinion of the article’s author. The election of Genaro Góngora as president changed the DCS in some important ways. Góngora believed that the court should be more accessible to the media than his predecessor had believed it should be (Aranda, Jesús. Personal communication. June 6, 2001; Fuentes, Víctor. Personal communication. June 8, 2001). Although President Aguinaco allowed the DCS to publish speeches and books on the court, he gave few interviews. In contrast, Góngora was eager to meet the legal and popular press. By so doing, Góngora attempted to put a personal face on the Supreme Court. Góngora also created the new Social Communications Ofﬁce at the Federal Judicial Council (DCSJ). Góngora named Néstor Martínez, a long-time reporter for La Jornada, director of the DCSJ. Martínez, in turn, brought on Jorge Camargo, who had been a reporter with El Universal and Reforma, and who would shortly become the DCSJ’s director. Unlike Laveaga and Vizcaíno, Martínez and Camargo were newsmen, not attorneys, and they carried a better understanding of the press corps. Having experimented with Laveaga, the Supreme Court returned to press men, believing that a more news-savvy press ofﬁce would understand better how to nurture quality relationships with the correspondents.17 The DCSJ developed a media strategy for the entire federal judiciary, not just for the Supreme Court, which had been the primary focus of the DCS. Reporters were granted access to once-private information. In

16 A copy of “Quiénes hablaron bien y quiénes hablaron mal de la Suprema Corte de Justicia

de la Nación en el primer semestre de 1999?” may be requested from the Coordinación General de Comunicación Social de Poder Judicial de la Federación. 17 Víctor Fuentes, personal communication with author, June 8, 2001, Mexico City, Mexico.

62

The Politics of Constitutional Review in Mexico

particular, they could request information from the DCSJ on the status of cases being resolved below the Supreme Court level. In this way, the DCSJ offered better access to newsworthy cases at the subnational level, and it gave reporters the opportunity to develop background for stories that would eventually be resolved by the Supreme Court. The DCSJ also organized weekend retreats for reporters aimed at providing basic instruction on the judicial system and suggestions for how to interpret legal doctrine. The DCSJ developed, to put a ﬁne point on it, a more aggressive and more complete media strategy than did the DCS. For two years, both ofﬁces remained open and simultaneously organized the Supreme Court’s media relations – Laveaga and Vizcaino for the court and Martínez and Camargo for the Federal Judicial Counsel. The ﬁnal structural change in the court’s public relations ofﬁces occurred in the ﬁrst days of January 2001, when the ministers integrated the DCS and DCSJ. Góngora appointed Camargo director of the new DCS. Laveaga, in turn, was made director of the Supreme Court’s Ofﬁce on Publications of Judicial Culture, effectively removing the former director’s power by relegating him to a far less important administrative ofﬁce.

transparency and legitimacy Despite the absence of a direct link between voters and the ministers, and despite the possibility of appearing overtly political, the Mexican Supreme Court engaged the public through the media in a variety of ways in the years following the Zedillo reforms. According to the court itself, it used the media to advance its goal of becoming an effective constitutional tribunal. In order to accomplish this goal, the court required media coverage. It needed to place itself on the national political radar screen. By attracting coverage and by instructing reporters on how to report on cases – on what was important and what was not – the Supreme Court advanced the transparency condition for judicial power. But in addition to transparency, the court has sought public support. Genaro Góngora himself insisted that the Zedillo appointees “[arrive] with the spirit of legitimating the Court.”18 In a September 1999 speech before the swearing-in ceremony of new federal district court judges, he delivered perhaps the clearest statement of this goal: 18 Óscar Camacho Guzmán, “El presidente de México no tiene privelegios en la Suprema

Corte,” October 11, 1999, Milenio Diario, 39.

Public Relations on the Mexican Supreme Court

63

In the judicial branch we are conscious of the fact that the only asset that we have is public trust . . . . The judge does not have and cannot have the sword of the executive branch or the purse of the legislative branch . . . . Today more than ever, we must work to achieve social trust, convincing people with actions that display neutral judgment, equal treatment of parties and a lack of any personal interest in the cases we resolve.19

Minister Silva Meza (Silva Meza 2000b) repeated this sentiment in a May 13, 2000, speech entitled “Trust Is Won Day by Day,” in which he suggests that by exhibiting the clearest neutrality in their work, Mexican judges can help achieve the legitimacy the judiciary so dearly requires. Silva Meza reminds us that the goal of building legitimacy cannot be achieved merely through rhetoric, but that Mexican judges must be observed behaving professionally and independently. By viewing the court’s approach to developing legitimacy through this lens of demonstration, we are led to the possibility that the goal of creating transparency and the goal of building legitimacy may be linked, perhaps in a useful way, from the ministers’ perspective. One of the most provocative claims in law and politics scholarship is that as people become familiar with their courts, they are increasingly likely to believe in their legitimacy (Casey 1974; Gibson, Caldeira, and Baird 1998). As people attend to their courts, they become exposed to and convinced by judicially framed messages and symbols (e.g., opinions that cite past jurisprudence, formal dress, even hierarchical seating arrangements) that promote a mythical sense of adjudication (Gibson, Caldeira, and Baird 1998, 345). Judicial decision-making is technical and guided by principles. In principle, it does not involve the kind of ﬂuid bargaining that we observe commonly in the elected branches of government. Although we know that judges bargain over case outcomes (Epstein and Knight 1998; Maltzman, Spriggs, and Wahbeck 2000), the idea is that people do not typically observe that behavior. At least, relative to legislative debate, judicial deliberation appears fundamentally different, and this difference is crucial to developing judicial legitimacy (Hibbing and Theiss-Morse 1995). If this argument is correct, then exposing Mexicans to the work of the Supreme Court may simultaneously create transparency and construct legitimacy. 19 Genaro Góngora Pimentel, “Palabras Pronunciadas por el Ministro Presidente de la

Suprema Corte de Justicia de la Nación y de Consejo de la Judicatura Federeal, Genaro Góngora Pimentel, con Motivo de la Toma de Protesta de Jueces de Distrito, el 9 de Septiembre de 1999,” September 9, 1999.

64

The Politics of Constitutional Review in Mexico

This logic suggests that the goals of transparency and legitimacy may reinforce each other, but it is important to consider carefully Silva Meza’s point about demonstration. The court argues publicly that it is impartial and that its decisions are unaffected by politics. The problem is that although the court can easily publicize the technical aspects of judicial decision-making, it is more difﬁcult to demonstrate impartiality simply by saying so. This is especially difﬁcult if the Mexican Supreme Court is constrained in some ways by real political interests, especially if those constraints induce strategic decision-making, at least on occasion. We might wonder about the consequences of those constraints for a nonstrategic public relations strategy aimed at maximizing public awareness by making the Supreme Court’s work as transparent as possible. Before we can answer these questions, however, it is useful to know whether there is evidence of political inﬂuences on Supreme Court decision-making.

4 Decisions, Case Promotion, and Compliance

The conﬂict between Ernesto Zedillo and the Chamber of Deputies, which the Supreme Court resolved and publicly promoted in the summer of 2000, concerned one of the most politically controversial policies of the time. To save a collapsing ﬁnancial sector during the 1994 peso crisis, the federal government had assumed the debts of several failing banks under the Banking Fund for Savings Protection (FOBAPROA), a program originally designed to ensure depositor assets. By 1998, it had become clear that many of these obligations were unrecoverable, and Zedillo proposed transferring them to the general public account, this time to ensure the solvency of FOBAPROA itself. Opposition parties in the Chamber of Deputies viewed the proposal skeptically, largely because a number of the loans in question were granted under lax risk management standards and, in some cases, were simply illegal, made by bank managers to family members, friends, and in some cases, to themselves.1 In addition, the president of the bankrupt Banco Unión, Carlos Cabal Peniche, announced in a Miami Herald interview that he illegally channeled U.S. $25 million to 1994 PRI electoral campaigns via a trust account that was assumed by FOBAPROA. The trust had been established by Fernando Ortiz Arana y Carlos Enrique Sales Gutiérrez, then the president and ﬁnance secretary of the PRI; and, it had been funded by loans from Banco Unión to companies associated with Cabal Peniche himself.2 In support of his statement, Cabal Peniche produced what he claimed were copies of cancelled checks 1 See Mackey (1998), Section 7 on reportable transactions. 2 Andres Oppenheimer, “Banker tells of huge PRI donations: Fugitive says $25 million

campaign gift ‘normal’ in Mexico,” Miami Herald, May 29, 1999.

65

66

The Politics of Constitutional Review in Mexico

from the PRI in the amount of U.S. $4 million.3 Although Zedillo denied knowledge of these transactions, Cabal Peniche claimed that the president had participated directly in the scheme. Whether or not Cabal Peniche was lying, as alleged by the president, his allegations gave a degree of credibility to the growing congressional skepticism over FOBAPROA. The Chamber of Deputies requested detailed information on the trust in question, including the names of the ofﬁcials who carried out each transaction and notes on the decision-making process by which bank interventions were carried out.4 Zedillo denied all of these requests, which led the chamber to initiate a controversy in the Supreme Court demanding that the information be released. The court’s unanimous decision did just that. The opinion noted that Constitutional Articles 73 and 74 granted Congress the responsibility of approving the federal budget and overseeing actions taken by the president that affect the public debt. The court held that, although these powers do not ordinarily permit the legislature to access information on private trust accounts, “It must be concluded that the interest safeguarded by the concept of ﬁduciary secrecy must not obstruct the Congress’s faculties when private debt is converted into public debt.” When private debt is made public, the Congress, acting in the interests of all Mexican citizens, must be given access to information relevant to its responsibilities concerning public accounts. As we know from Chapter 1, the court aggressively promoted the outcome of this case, granting interviews with any media source interested in a statement. Although President Zedillo was in Central America at the time of the announcement, his response was immediate. Within hours of the announcement the president directed his Secretary of Finance to release the information. Coming less than two months after Fox’s victory, the case received a great deal of attention and was widely regarded in the popular press as a fundamental statement on the changing nature of the Mexican separation of powers system.5 On that account, the reforms of the mid-1990s had

3 Sam Dillon, “Donor Implicates Mexican Party in Scandal,” New York Times, June 7,

1999. 4 Mireya Cuéllar et al. “Documentación detallada sobre el Fobaproa, pide la oposición a

la SG,” La Jornada, July 22, 1998, http://www.jornada.unam.mx. 5 See, Agustín Ambríz, “Cómo Aplastó la Corte a la Presidencia,” Revista Proceso (Agosto

29, 2000): 16–20; Alberto Aziz Nassif, “Fin de sexenio, ﬁn de régimen,” La Jornada, Agosto 29, 2000; Federecio Reyes-Heroles, “Cría ultras y . . . ,” Reforma, Agosto 29, 2000; Carlos Sodi Serret, “Suprema Corte: El Poder Silencioso,” Excélsior, Agosto 30,

Decisions, Case Promotion, and Compliance

67

created a powerful new force in Mexican politics, as Finkel’s insurance account would suggest. The Zedillo decision simply conﬁrmed the court’s expanded authority. Despite this attractive rule-of-law narrative, it is not clearly the correct lesson to take away from the resolution. Instead, the Zedillo decision might simply reﬂect the Supreme Court’s effort to exercise its authority as effectively as possible in an uncertain political environment. In this regard, it is important to note that the resolution, authored by Minister Olga Sanchez, was ﬁnished nearly two months prior to its announcement, placing the original resolution date roughly one month prior to the presidential contest. Local media coverage reported an internal battle at the court over whether to announce the decision before the election,6 and the court itself issued a press release claiming that it would not be rushed, for to do so would “call its role as ﬁnal arbiter into question.”7 According to reporter Norma Jiménez, the motive for delaying the release was to avoid being perceived as supporting one political party over another. The majority of the ministers opted to delay the announcement in order to maintain the image of the federal judiciary as fundamentally impartial. Of course, waiting until after the election also permitted the Supreme Court to evaluate President Zedillo’s postelectoral political position. Would it be an Ernesto Zedillo whose preferred candidate had just won the presidential election, or would it be Ernesto Zedillo, the lame duck? Rather than demonstrate the court’s unconditioned power, it is possible that the decision provides a lens through which we can view the conditions under which that power is used, conditions that the court could have been inﬂuencing with its aggressive approach to public relations. This chapter provides an empirical basis to answer questions about the role of the Mexican Supreme Court in Mexican politics during the transitional period. It also suggests answers to more general questions about judicial public relations and power.

2000; Jesús Vergara Aceves, “Resolución clave de la Suprema Corte,” El Universal, Agosto 29, 2000; “Fallo positivo y trascendente,” El Universal, Agosto 26, 2000; “El fallo histórico equilibria poderes,” Revista Impacto, Agosto 27, 2000. 6 Norma Jiménez, “Zedillo, obligado a entregar información del Fobaproa,” Milenio Diario, August 24, 2000, . 7 Comunicado de Prensa Número 278, “Si la Suprema Corte de Justicia Se Sometiera a Tiempos Electorales, Pedería su Credibilidad Como Árbitro: Ministro Góngora Pimentel,” Suprema Corte de Justicia de la Nación, Junio 26, 2000.

68

The Politics of Constitutional Review in Mexico

decision and case promotion data The ﬁrst section of this chapter tests decision making and case promotion hypotheses. I present tests that distinguish between a public support model in which the transparency of constitutional conﬂicts is exogenously determined and a model in which judges themselves have control over transparency. With these goals in mind, I attempt to explain the Mexican Supreme Court’s choices to strike down public policies under its constitutional review authority and its choices to promote these decisions before the national media, behaviors that I have argued are linked theoretically. In addition, I present information on Supreme Court efforts to seek compliance with orders of the lower federal judiciary. I use an original data set on all constitutional decisions in review of state or federal laws or reviewable executive actions (e.g., decrees) resolved by the Supreme Court in plenary session between January 1, 1997 and December 31, 2002. The data set includes three kinds of constitutional actions: amparo appeals, constitutional controversies, and actions of unconstitutionality. Amparo is Mexico’s traditional constitutional action, developed in the nineteenth century. It is an individual constitutional complaint in which persons, organizations, and corporations may challenge governmental violations of individual rights in federal court. In constitutional controversies, the court resolves conﬂicts between levels of government (e.g., municipalities and states or states and the republic) and across powers of government (e.g., state legislature and governor). Constitutional questions in these cases deal with issues of federalism and the separation of powers. Finally, under actions of unconstitutionality, the court exercises abstract review over the constitutionality of state laws, federal laws, and international treaties.8 Appendix 4A describes the current structure of the Mexican judiciary, provides some technical information on these constitutional actions, and summarizes the process by which constitutional cases come to be resolved by the Supreme Court.9 As Table 4.1 indicates, there were 1,006 analytically distinct constitutional cases during this period.10 With one exception, the Supreme 8 On the differences between these actions, see Carranco Zúñiga (2000) or Cossío (2002). 9 The unit of analysis is thus the constitutional case. All replication material is available

upon request. Summaries of the Supreme Court opinions that are included in the sample are available at www.scjn.gob.mx. 10 Between 1995 and 2002, the Supreme Court decided 3,084 cases, or better put, it resolved 3,084 distinct case ﬁles. The data set restricts the number of analyzable cases

Decisions, Case Promotion, and Compliance

69

table 4.1. Summary of Supreme Court Constitutional Docket, 1997–2002

Indirect Amparo Appeal Direct Amparo Appeal Constitutional Controversy Action of Unconstitutionality total

N

%

658 161 100 87 1,006

65.4 16.0 9.9 8.7 100

Source: Supreme Court of the Nation (http://scjn.gob.mx). Note: As described in the text, the table reﬂects the number of analytically distinct cases resolved by the Supreme Court under its three core institutions of constitutional review. Appendix 4B describes case selection criteria.

Court’s jurisdiction in all three cases is mandatory.11 Thus, the data set includes all constitutional claims that met the court’s statutorily deﬁned jurisdiction. Restricting the analysis to the six-year period from the beginning of 1997 to the end of 2002 is done, in part, on pragmatic grounds. The only systematic, valid indicator of the Supreme Court’s case promotion choices, measured at the level of the constitutional case (described later) is easily coded, but only available beginning in January 1997. This explains the start date. And, the full court ceased hearing amparo appeals to 1,536 over this period, 1,006 of which were resolved between January 1997 and December 2002. The rationale behind the reduction is as follows. Many cases with distinct case ﬁle numbers address identical legal issues and result in identical resolutions resolved by an identical vote on the same day. For example, a major constitutional reform on indigenous rights reform that passed in the summer of 2001 generated 292 distinct constitutional controversies ﬁled by municipalities from across the country. Each of the 292 cases challenged the procedures Congress used to adopt the amendment. These cases make up nearly 10 percent of the court’s total constitutional caseload, yet they were all formally resolved on September 6, 2002, by the same vote and for the same reason. These cases are clearly not independent of each other. Moreover, because the theoretical model is concerned with constitutional court reactions to public policies, and each of these similar cases deals with the same policy, we do not want to treat them as separate observations. Appendix B describes the rules for case selection. 11 Pursuant to a 1988 constitutional amendment, the Supreme Court enjoys a limited power of discretionary review through what is called atracción (Mexican Constitution, Article 107, Section 8). If the court deems an issue raised in an appeal outside of its appellate jurisdiction fundamentally important to the law, it may exercise atracción. In addition, the federal attorney general may request that the court take up a case of national interest, and circuit courts may certify appeals for Supreme Court review.

70

The Politics of Constitutional Review in Mexico

under normal circumstances in 2003, which makes comparison across the traditional constitutional review actions impossible after December 2002.12 Importantly, there is nothing in the theoretical argument that demands a long time series even if it were possible to construct one, and the date restrictions have some useful analytical properties. The Supreme Court’s membership did not change between February 1995 and December 2003, when two of the original eleven ministers were replaced. This absolute stability in membership allows me to control for aggregate judicial ideology. Unless preferences are rapidly changing over the course of the period studied, the aggregate ideology of the court, whatever that may be, is controlled by design. Second, the court did not hear many constitutional controversies or actions of unconstitutionality prior to 1997, so the start date eliminates very little information. Third, the time period captures three distinct institutional settings from the perspective of a fragmented government theory of power. There is a period of fully uniﬁed national government under a PRI president, a period of divided government under a PRI president, and the beginning of the Fox administration, a period of truly divided national government. Fourth, as I demonstrated in the previous chapter, there is evidence suggesting that the Supreme Court’s media coverage was almost entirely neutral during the period around the Fox election, an implicit assumption in the theoretical model.13 The ﬁnal advantage of the Mexican case concerns judicial legitimacy. Diffuse support is controlled by design, unless it is rapidly changing over the short time period of the study. This result is highly unlikely given the slow way diffuse support has been shown to develop (Gibson, Caldeira, and Baird 1998, 353). The study does require that diffuse support be greater than zero. Unfortunately, there is no precise measure of diffuse public support for the Mexican Supreme Court during the period studied here.14 Still, available public opinion data suggests that diffuse

12 On April 16, 2002, the Supreme Court elected to remit all revision appeals in amparo

to the benches, save those dealing with governmental corruption. See Supreme Court Acuedo 4/2002. 13 Although there is no clear reason to suspect that the court’s coverage changed dramatically in the subsequent years, I cannot test this possibility. 14 Unfortunately, neither the World Values Survey nor the Latinobarómetro include items appropriate for measuring Supreme Court support. The World Values Survey asks about public trust in the “legal system” (Question V137, 1995 Wave, www. worldvaluessurvey.org) while the Latinobarómetro asks about trust in the “judiciary”

Decisions, Case Promotion, and Compliance

71

support may not have been completely absent. In a series of national opinion surveys conducted by the newspaper Reforma between December 2000 and March 2002, respondents were asked to evaluate the job of the Supreme Court. During this period, the percentage of respondents expressing favorable opinions varied between 40 percent and 50 percent, statistics that climb as high as 87 percent when excluding nonresponses and neutral ratings.15

Hypothesis Review The theoretical analysis suggests a number of hypotheses concerning a constitutional court’s likelihood of striking down public policies. The key is to identify implications that distinguish between the baseline model, in which the transparency of the case is exogenously given, and the promotion model, in which courts can control transparency. Both models propose that courts should be decreasingly likely to strike down public policies as the importance of the policy to government ofﬁcials with control over the institutions of the federal judiciary increases; however, the models differ over the effects of judicial beliefs over transparency. If the baseline model is correct, courts should be more likely to strike down public policies when they believe that their decisions will be observed and understood by the public; however, this effect should be strongest for the least important policies. As policy importance increases, there should be a point at which beliefs in transparency have no effect on the likelihood of striking down a public policy, precisely because the policy is important enough that diffuse public support is an insufﬁcient (Question V.2.32, http://www.latinobarometro.org). Given the high levels of public confusion in Mexico about what institutions in particular fall within the judiciary or the legal system, these are highly imprecise measures of institutional commitment to the Supreme Court itself. The Latinobarómetro provides some reason to perceive a moderate degree of support for the court, even under the broad measure that it uses. In 1995, only 24 percent of respondents expressed no trust at all in the judiciary. This percentage climbed as high as 44 percent in 1996 and 55 percent in 2003, but the simple average between 1995 and 2004 was only 34 percent. Clearly, when one-third of the country expresses no trust at all in the judiciary, there is a problem. That said, a sizable proportion of Mexico seems to have trusted the judiciary at least a little. And again, there is a difference between the judiciary in general and the Supreme Court. 15 During this period, the director of public opinion for Reforma was political scientist Alejandro Moreno, the country investigator for the 2000 wave of the World Values Survey. Information on the Reforma sampling design can be obtained at www.reforma.com/encuestas or by contacting the author. I have been unable to locate national public opinion results for the period between 1997 and 2000.

72

The Politics of Constitutional Review in Mexico

constraint on state action. In contrast, if the promotion model is correct, there should be no observable effect of beliefs on transparency, because either the policy is too important for public support to matter or because public support should matter, but courts can create their own transparency. As discussed in Chapter 2, because transparency is exogenous, the baseline model provides no prediction about when constitutional courts will promote their decisions. The promotion model provides a simple prediction. Courts should be more likely to promote decisions striking down public policies, but especially so when judges believe that their decisions are unlikely to be observed and understood by the public if they do nothing. In other words, when courts do not expect the media to cover a resolution, their choices to promote cases should be highly sensitive to the decision outcome. This is because the risk of the media improperly reporting the decision rationale (which exists however the court treats the status quo policy) is extremely small, yet the incentive to promote as a means of ensuring that the public may monitor compliance remains strong. Because the judicial interest in public monitoring only kicks in when the status quo has been disturbed, courts should condition their promotion choices on whether they strike down the policy under review. On the other hand, when courts expect ex post media coverage, the concern over ensuring public monitoring becomes less important and promotion concentrates on ensuring proper coverage of the decision rationale, which exists independently of whether the court strikes down the policy under review. This is not to say that the incentive to promote in order to ensure the capacity of the public to monitor disappears; however, given the expectation of ex post media coverage and the corresponding possibility of reporting errors, the effect of striking down the policy under review should attenuate, or at least not increase. Measurement and Estimation Two measures are required to test these expectations: an indicator of whether the Supreme Court strikes down the policy under review and an indicator of the court’s decision to promote the resolution. A measure of the former concept is easily obtained. The variable Strike indicates whether the Supreme Court invalidated a policy challenged through one of the three constitutional actions. Strike is coded 1 if the court supported at least one argument against the constitutionality of the policy and 0 otherwise.

Decisions, Case Promotion, and Compliance

73

A systematic, case-by-case measure of promotion is less obvious, because the court’s case promotion behavior takes a variety of forms. Ministers give press conferences and offer interviews with the mass media. They publish full-length manuscripts on constitutional doctrine, and they sponsor and participate in academic conferences. More informally, the ministers hold conversations in the halls of the Supreme Court building and in their ofﬁces with key members of the press. And, of course, the Supreme Court has its own public relations ofﬁce, which is responsible for its overall media strategy. Unfortunately, it is impossible to obtain observable, reliable, case-by-case indicators of the vast majority of public relations efforts carried out by the court. Yet, one readily available indicator provides a reasonable summary of the court’s case-speciﬁc promotion strategy. The second dependent variable, Press, is coded 1 if the Supreme Court issued a press release announcing the result and 0 otherwise. Press releases on case outcomes typically summarize the major components of the resolution, including the parties to the case, the constitutional question, and a brief statement of the rationale. Ministers and members of the press ofﬁce are frequently happy to ﬁll in over the phone or e-mail details not covered by the press releases. In this sense, it is better to think of the variable Press as a general indicator of the court’s promotion strategy rather than the only means by which the court communicates with the public.16 If the promotion model is correct, then Strike and Press are jointly determined. We can estimate their joint distribution and test the hypotheses speciﬁed above by the following recursive simultaneous equations model. Strike = [β1 (Importance) + β2 (Coverage) + β3 (Importance ∗ Coverage) + α1 X1 ] Press = [β4 (Strike) + β5 (Coverage) + β6 (Strike ∗ Coverage) + α2 X2 ], where Importance is a measure of political importance assigned to the challenged policy by Mexican federal ofﬁcials (the ofﬁcials with control over the Supreme Court’s institutional structure); Coverage measures the court’s expectations about subsequent media coverage in the event that 16 Although many press releases are drafted by DCS staff, the language is approved by

individual ministers, the DCS director, or his immediate assistant.

74

The Politics of Constitutional Review in Mexico

the case is not promoted; and the Xi are vectors of control variables for each equation. As recommended by Greene (1998), this model can be estimated via bivariate probit. Policy importance I measure federal policy importance by appealing to the kind of constitutional claim under review and the level of the public ofﬁcial against whom the claim was raised. Two assumptions underlie the measure. Federal ofﬁcials care more about federal policies than state policies, and they care more about the policies challenged through constitutional controversies and actions of unconstitutionality than they do about those challenged through amparo. Given these assumptions, I generate a scale of increasing federal importance, coded from 0 to 3, ranging from a state law challenged via an amparo suit to a federal law challenged under the action of unconstitutionality or constitutional controversy.17 The ﬁrst assumption would appear relatively uncontroversial. On average, the Federal Congress likely cares more about the federal penal code than it does about the penal code of Nuevo León. The second assumption deserves further justiﬁcation. Why would federal ofﬁcials care more about policies challenged under the constitutional controversy or action of unconstitutionality than policies challenged under amparo? Whereas resolutions to amparo suits settle only the immediate controversy being adjudicated,18 decisions in both constitutional controversies and actions of unconstitutionality have the potential of setting general effects. With respect to the measure, I argue that federal ofﬁcials ought to care more about resolutions that have the potential of setting precedent than about those that do not.19 17 The results are robust to an alternative coding scheme in which I combine the state

policies into one category and where I use a dummy variable distinguishing between state and federal policies. 18 The court may establish formal jurisprudential theses in amparo. A jurisprudential thesis is binding on all lower federal court judges; however, it is not clear that theses signiﬁcantly affect bureaucratic decisions to enforce laws. Legal principles may become part of the jurisprudence when the court establishes that it has invoked a similar principle in ﬁve consecutive cases (Ley de Amparo, Article 192). These theses may be abrogated by a coalition of eight ministers voting contrary to provision under analysis (Article 194). 19 Another element of salience that this coding procedure picks up is the average age of the statute under attack. Although the data source makes it impossible to know exactly how old the particular section being challenged was in the amparo cases, there is good reason to believe that these provisions were older, on average. In the ﬁrst place, actions of unconstitutionality must be moved within thirty days of ﬁnal passage and common constitutional controversies involve live political conﬂicts. None of this is necessarily true of the amparo cases.

Decisions, Case Promotion, and Compliance

75

Constitutional controversies and actions of unconstitutionality also deal with what we might understand as more signiﬁcant political issues. Typical constitutional controversies involve state-municipal conﬂicts over the autonomy of local governments, state-state boundary disputes, and federal interbranch conﬂicts over competing claims on power (Fix-Fierro 1998a, 1998b). Actions of unconstitutionality frequently involve political party challenges to the constitutionality of electoral laws, rules quite essential to political interests. These are cases that affect large numbers of people and large sums of money. In contrast, state entities have extremely limited standing in amparo, and political parties have none. Also, individuals are prohibited from challenging electoral codes through amparo. Finally, the court itself has argued that amparo suits are generically less important than constitutional controversies and actions of unconstitutionality.20 A more complete analysis of measurement validity is detailed in Appendix 4B.21 Beliefs in transparency A perfect measure of Supreme Court beliefs in the likelihood of ex post media coverage would get inside the heads of the ministers at the time of the decision. The case ﬁles do not include plausible measures of these beliefs. For example, I was unable to locate intracourt memos that discuss how the ministers understood media attention. It is also unclear how one would construct a case-by-case indicator over a six-year period simply by asking the ministers to recall what they were thinking for each of more than a thousand decisions rendered in the past. To obtain a measure of beliefs, I code prior national print media coverage of the case or the conﬂict that generated the Supreme Court’s resolution. I assume that the court estimates a higher probability of ex post coverage when the print media has already covered the case. As discussed in Chapter 3, each member of the Supreme Court receives a daily news brieﬁng with photocopied articles of every print story that even tangentially relates to the federal judiciary. Consequently, print coverage is a natural way for 20 See supra note 8. 21 Of course, some amparo cases are likely to be salient, perhaps more so than some consti-

tutional controversies. Although the appendix provides a number of robustness checks, it is not possible to remove all error from this indicator. Without a clear argument for what exactly makes a particular class of amparo appeals more or less salient (beyond the level of government), it is likely that the error is random. If this is the case, the consequence is that causal effects will be biased toward zero, making it more difﬁcult to uncover a signiﬁcant causal effect of importance.

The Politics of Constitutional Review in Mexico

76

the ministers to generate beliefs about the probability that the press will cover its resolutions. Speciﬁcally, I use the newspaper La Jornada, one of the major national papers with a dedicated and respected beat reporter, as a proxy for general media attention. Coverage is coded 1 if La Jornada ran an article on the controversy prior to the decision and 0 otherwise. Appendix 4B to this chapter describes how articles describing coverage of the Supreme Court cases were selected. Preliminary analysis within Mexico at research university libraries offered little conﬁdence in the reliability of electronic searches for articles concerning the Mexican judiciary.22 The best strategy appeared to be a daily, full text search. La Jornada maintains a publicly accessible Web site that contains past issues dating from March 29, 1996 to the present. In addition, La Jornada is one of a small number of widely circulated newspapers known for its independence (Lawson 2002, 69). Unsurprisingly, only 3 percent of the court’s docket during the period studied received prior coverage by La Jornada. On the other hand, prior coverage is an excellent predictor of whether the resolution will receive subsequent media attention. Consider Table 4.2. Although La Jornada covered only 3 percent of the ﬁnal resolutions to conﬂicts it had not been

table 4.2. Print Media Coverage of Supreme Court Constitutional Decisions (by prior coverage) La Jornada Coverage Prior to the Decision?

No (%) Yes (%)

La Jornada Coverage After the Decision? No

Yes

949 (97.1) 8 (27.6)

28 (2.9) 21 (72.4)

Total

977 (100) 29 (100)

Sources: La Jornada (http://www.jornada.unam.mx); Supreme Court of the Nation (http://scjn.gob.mx). Note: Shows coverage of Supreme Court constitutional decisions following the decision in La Jornada based on whether La Jornada covered the case or policy conﬂict that led to the case prior to the decision. Row percentages are in parentheses. Note that although La Jornada only covered 3 percent of the decisions it had not covered prior to the resolution date, it covered more than 72 percent of the decisions it had already been covering.

22 InfoLatina, Lexis-Nexis’s Spanish search engine and all utilities maintained by the main

national newspapers produced highly erratic results, especially for the mid- to late 1990s.

Decisions, Case Promotion, and Compliance

77

covering, it covered more than 72 percent of resolutions to cases it was already covering. In other words, La Jornada was about 2,400 percent more likely to cover a case outcome if it was already following the conﬂict than if it was not. Controls In the Strike equation, I control for three potential inﬂuences on judicial decision making. Following the basic insight of veto player theory (Tsebelis 2002), Iaryczower, Spiller, and Tommasi (2002) suggest that courts should be more deferential under uniﬁed government, because it is easier for presidents and legislators to coordinate a response to the judiciary’s activism. Ríos-Figueroa (2006) ﬁnds evidence of this veto player effect in an analysis of Mexican Supreme Court decision making in constitutional controversies and actions of unconstitutionality. To capture the three phases of uniﬁed-divided government in Mexico between 1997 and 2002, I include dummy variables for the period of truly divided national government under Vicente Fox (Divided), when the National Action Party (PAN) controlled the presidency but did not control a majority in the Chamber of Deputies. I include another dummy variable for the period of PRI government under Ernesto Zedillo, when the PRI controlled the presidency but enjoyed only plurality support in the legislature (Uniﬁed-Pl). Thus, the period of uniﬁed government under Zedillo, in which the PRI still maintained majority control over the Congress, is the base category.23 In order to address the possibility that Supreme Court decisions were affected by partisan factors (Magaloni and Sánchez 2001), I also include a dummy variable, PRI, coded 1 if authority responsible for the challenged policy is either a priísta executive or a majority priísta legislature and 0 otherwise.24 I also control for the identity of the plaintiff moving the constitutional action, capturing the notion that the quality of legal representation should vary according to plaintiff type, and using the assumption that political 23 The ﬁrst period ended on September 1, 1997, and the second on December 1, 2000.

Given this setup it is not possible also to control for regime type. On many accounts, Mexico was a democracy during this entire period. This seems to be the consensus among country experts, as reviewed at the beginning of chapter 3. Mexico scored above 6 on the Polity IV measure during the entire period. It was coded as an autocracy by the ACLP measure until 2000. Either regime type is controlled during the entire period or it is perfectly collinear with the dummy variable that picks up the Fox administration. 24 Partisanship data for the state legislatures and governors may be found at the Web site for the Mexican Senate (www.senado.gob.mx).

78

The Politics of Constitutional Review in Mexico

status is a reasonable proxy for litigant resources (McGuire 1995; Sheehan, Mishler, and Songer 1992). The issue of legal representation is especially relevant in Mexico where access to good counsel is limited and the disparities between the legal representation of various parties is often great (Rubio, Magaloni, and Jaime 1994, 119–134). Controlling for plaintiff identity involves estimating a series of dummy variables that account for a large number of plaintiff categories.25 The base category is the modal plaintiff, a private individual.26 In the Press equation, I control for a policy importance and an idiosyncratic, personality-based theory of case promotion. As we know, the DCS was charged with publicizing the results of cases that resolve intrinsically “important” issues in Mexican law.27 Unfortunately, the court did not deﬁne what it meant by an intrinsically important issue, and therefore, it does not appear to be a ready-made measure of this concept. Still, it would appear that Importance, as deﬁned above, might capture generally what makes a particular constitutional question more intrinsically important than another. The questions the court addresses ought to matter more 25 These categories include individual or private business, municipality, union, civil, or

religious association, small political party, corporation, large political party (PRI, PAN, PRD), university, state legislature, state executive, federal commission, federal congress, state judiciary, federal judiciary, and federal executive. The results are robust to an ordinal measure of complainant political status. 26 Once we include amparos, constitutional controversies, and actions of unconstitutionality in the same model, it is not possible to control for issue area without double-counting the concept. This is because the issues that arise under these actions are different. For example, you cannot challenge an electoral law under amparo. Of course, although this is disadvantageous in one sense, it provides a signiﬁcant advantage in another. In particular, this distinction is one of the motivations for treating policies challenged under actions of unconstitutionality and constitutional controversies as generically more important than those challenged under amparo. 27 It is possible that case promotion has nothing to do with the public support mechanism. Indeed, we might imagine that there is no underlying noncompliance problem. In such a world, the court nonetheless might be particularly likely to promote cases striking down public policies in the sense that they reﬂect potential changes in the law worth announcing. This assumption makes the case outcome results (which are themselves consistent with a model in which there is an underlying noncompliance problem) described below difﬁcult to interpret, but it is worth considering the argument setting those results aside for argument’s sake. It is not clear that the direction of the decision necessarily reﬂects an inherently more important legal outcome. Important legal principles are carried in cases that uphold the constitutionality of particular policies. For example, in 2008, the Supreme Court upheld a federal district law that legalized abortion during the ﬁrst twelve weeks of pregnancy. The court refused to deﬁne a fetus as a person in the ﬁrst trimester (Acción de inconstitucionalidad 146/2007). This was, quite frankly, a massive constitutional decision, and it came in the form of a decision upholding the constitutionality of a law.

Decisions, Case Promotion, and Compliance

79

as the level of government challenged and precedent-setting value of the decision increase. They are intrinsically important questions for exactly the same reason that federal political ofﬁcials ought to care more about the outcomes of these cases. Accordingly, we ought to expect the Supreme Court to be more likely to publicize its decisions as the federal importance of the policy increases. Finally, I control for the Genaro Góngora Pimentel regime. The Supreme Court selects its own president from among all current members. Each president serves a four-year, nonrenewable term. The court elected Góngora in 1999, and it is popularly assumed that the court’s public relations activities became increasingly aggressive and strategic following his election. In order to address this possibility, I include a dummy variable Góngora, which is coded 1 for all cases resolved during President Góngora’s tenure and 0 otherwise.

a final note on research design Whittington (2005) suggests that governments may, at times, wish a court to invalidate policies that they ﬁnd politically difﬁcult to change through typical democratic processes. This raises the possibility that a variable that indicates whether the government won a particular case might not provide a valid measure of judicial power. The Whittington argument is persuasive. There are likely occasions when governments wish to lose cases. A measure of governmental success is not necessarily a good measure of judicial independence. Although valid, the critique does not undermine the current design. The dependent variable does not indicate whether the national government was successful. Instead, it indicates whether the Supreme Court invalidated whatever public policy was before it. Some of the challenged policies, after all, are state laws. Moreover, the research design does not propose that “striking public policies” is a valid measure of judicial power. Rather, the logic behind the design is that if the theory of judicial power is correct, then the court’s choices to strike down policies should vary with theoretically relevant concepts (e.g., federal policy importance and transparency). Analysis Table 4.3 displays results from two bivariate probit estimations. Model 1 excludes those observations in which the Supreme Court failed to reach the substantive question underlying the case, and instead dismissed the

80

The Politics of Constitutional Review in Mexico

table 4.3. Determinants of Supreme Court Decisions to Strike Policy and Promote Cases Excludes Procedural Dismissals Model 1 β Strike Equation Coverage Importance Coverage* Importance Uniﬁed-Pl Divided Party Complainant Types Corporation Large Political Party (PRI, PAN, PRD) State Legislature Federal Executive Constant Press Equation Strike Coverage Strike* Coverage Importance Góngora Constant

All Decisions Model 2

RSE

β

*** *** ***

0.34 0.02 0.04

1.12 −0.131 −0.367

0.264 0.351 0.078

*

0.13 0.22 0.08

0.213 0.342 0.082

*

0.11 0.22 0.09

0.251 0.904

*** ***

0.07 0.13

0.292 0.741

*** ***

0.292 0.741

0.624 6.10

*** ***

0.17 0.31

0.494 7.06

***

0.17 0.58

−0.864

***

0.06

−0.910

***

0.04

2.13 1.80 −1.02 0.325 −0.106 −2.25

*** *** ***

0.29 0.29 0.17 0.05 0.11 0.27

1.88 2.01 −0.956 0.282 −0.121 −2.14

*** *** ** ***

0.33 0.21 0.37 0.07 0.15 0.34

1.56 −0.120 −0.603

N = 852 ρ = −0.723 χ 2 = 45.68 p < 0.001

***

RSE

***

***

0.81 0.02 0.27

N = 1, 005 ρ = −0.558 χ 2 = 6.35 p < 0.011

Source: Supreme Court of the Nation (http://scjn.gob.mx). Note: Results of bivariate probit model with robust standard errors correcting for clustering within constitutional action type. ***p < 0.001, **p < 0.01, *p < 0.05.

claim on procedural grounds. Model 2 includes all observations, including those in which the court dismissed the claim for procedural default. There is a fair reason to treat procedural dismissals as equivalent to denials on the merits, which is what Model 2 does. Mexican federal judges have

Decisions, Case Promotion, and Compliance

81

a great deal of freedom to ﬁnd procedural default in their cases, and scholars have argued that they manipulate these rules in order to reduce the judicial caseload (Rubio, Magaloni, and Jaime 1994). It is certainly possible that they do so to avoid political conﬂict or to beneﬁt political allies (Magaloni and Sanchez 2001). On the other hand, some dismissals would appear to be clearly required by law, suggesting that these cases are fundamentally different from those in which the court resolved the underlying constitutional issue.28 Consequently, I report models run on both sets of data.

Decisions to invalidate public policies Table 4.3 provides support for the baseline and promotion models. Consider the Strike equation. The Importance coefﬁcients are negative and strongly signiﬁcant in both models. As summarized above, both theoretical models anticipate this effect. More importantly, note that the Coverage coefﬁcient is positive in both models and that the interaction term (Coverage*Importance) is negative. As the baseline model suggests, it appears that the Supreme Court was more likely to strike down public policies in cases that were likely to be covered by the press than cases that were not; however, that effect would appear to attenuate as the importance of the policy under review to federal ofﬁcials increases. On the other hand, although the coefﬁcients are similarly signed in both models, they are smaller and lose statistical signiﬁcance in Model 2. It is important to note that Model 2 is estimated on more than 150 more observations than Model 1; so if anything, the Model 2 estimates should be more precise. Critically, it is impossible to evaluate the statistical signiﬁcance of the Coverage effect simply from reviewing the coefﬁcients and their standard errors in Table 4.3 (Brambor, Clark, and Golder 2006). The hypothesis tests summarized in Table 4.3 for the Coverage effect are useful in only one regard. We can discern from Model 1 that the effect of Coverage is positive and statistically signiﬁcant when Importance is equal to 0. Model 2, in contrast, suggests that the Coverage effect is statistically indistinguishable from zero at the identical value of Importance. In order to evaluate better the statistical and substantive effects of Coverage, we require additional analysis. 28 For example, in 2001, the Supreme Court dismissed a municipal challenge to President

Fox’s decision to expropriate a plot of communal land for the construction of a new airport after Fox reversed his decision and the cause of action ceased to exist.

The Politics of Constitutional Review in Mexico

82

〈Merits decisions (N = 852)〉

Probability of striking policy

1

.8

.6

.4 .2

0 0

1

2

3

Policy importance

Note. Shows predicted probability of policy invalidation across the range of policy importance Prior coverage

No prior coverage

ﬁgure 4.1. Predicted Probability of Policy Invalidation by Federal Policy Importance and Prior La Jornada Coverage

Figure 4.1 shows the predicted probability of the Supreme Court invalidating a public policy on constitutional grounds across the range of Importance. All other variables in the model are set to their mean levels. Estimates for cases that received prior coverage by La Jornada are in red; those that did not receive prior coverage are in black. Note that both lines slope downward, which reﬂects the negative effect of federal policy importance. For cases that received prior coverage, the predicted probability of a policy invalidation drops from a high of approximately 0.78 to a low of 0.14 across the range of Importance. For cases that did not receive prior coverage, the predicted probability of an invalidation drops from a high of 0.28 to 0.17. These estimates are consistent with both theoretical models. Of course, the empirical results that can distinguish between the exogenous and endogenous models of transparency concern the differences between the black and red lines in Figure 4.1. The distance between the black line and the red line is precisely the effect the Coverage on Strike, conditional on a value of Importance. Although Figure 4.1 suggests support for the exogenous model, once again we do not know whether the differences are statistically distinguishable from zero. Figure 4.2, which reports the effect of Coverage on Strike for the Model 1 estimation, provides exactly the information we require. The dark gray

Decisions, Case Promotion, and Compliance

83

Decisions on merits (N = 852)

Marginal effect of coverage

8 6 4 2 0 –2 –4 –6 –8 0

1

2

3

Policy importance Note. Displays changes in the predicted probability of a decision to strike a public policy when the media covers a case prior to the decision, for all values of the policy importance measure. Cases that were dismissed on procedural grounds are excluded. Dashed lines reflect the 95% confidence interval around the predicted change.

ﬁgure 4.2. Marginal Effect of Coverage as Importance Varies (Merits Decisions)

line in the ﬁgure represents the change in the predicted probability of a policy invalidation associated with ﬂipping the Coverage variable from zero to one. The dashed lines surrounding the solid gray line indicate the 95 percent conﬁdence interval around that change. Figure 4.2 indicates strong support for the baseline model. Indeed, the effect of Coverage is positive and signiﬁcant for relatively low levels of policy importance, but it becomes statistically insigniﬁcant as Importance increases. Interestingly, and consistent with the impression given by Table 4.3, Figure 4.3 offers no support for the baseline model. Although the estimated Coverage effects look similar to those in Figure 4.2, the 95 percent conﬁdence interval straddles the zero line across the entire range of Importance. Thus, Figure 4.3 is consistent with the promotion model, which anticipated that prior media coverage would be unrelated to the court’s constitutional decisions. Having summarized Figures 4.2 and 4.3, we ﬁnd ourselves with mixed empirical results, which offer support for both models. Both models predicted the negative effect of Importance on Strike, but while Model 1 provides empirical support for the baseline model’s predictions concerning the effects of Coverage, Model 2 is consistent with the promotion model. How we evaluate the evidence, then, depends in part on whether we conceptualize procedural dismissals as a form of denial, as Magaloni

The Politics of Constitutional Review in Mexico

84

All decisions (N = 1005)

Marginal effect of coverage

8 6 4 2 0 –2 –4 –6 –8 0

1

2

3

Policy importance Note. Displays changes in the predicted probability of a decision to strike a public policy when the media covers a case prior to the decision, for all values of the policy importance measure. Cases that were dismissed on procedural grounds are included. Dashed lines reflect the 95% confidence interval around the predicted change.

ﬁgure 4.3. Marginal Effect of Coverage as Importance Varies (All Decisions)

and Sanchez (2001) and Rubio, Magaloni, and Jaime (1994) appear to do, or whether we conceptualize procedural dismissals as exogenously determined by objective criteria over which Supreme Court ministers have little control. If we adopt the former conceptualization, then the results support the promotion model. If we take the latter, then the results are consistent with the baseline model. There is an additional possibility. Recall that the promotion model assumes that a court has complete control over the transparency of its conﬂicts. As the court’s control weakens, the promotion model increasingly looks like the baseline model. In fact, in the limit, when the court has no control over subsequent media attention, the promotion model completely collapses into the baseline model. What if courts cannot render some cases fully transparent? This could happen in one of two ways. It is possible that the remedies for some constitutional violations lack transparency inherently. For example, a 2006 decision of the Israeli Supreme Court held that the targeted killings of Palestinian militants do not necessarily violate international law; however, the court established a set of clear guidelines that the Israeli Defense Forces (IDF) must follow in order to ensure the procedure’s legality. Two years later, the Haaretz

Decisions, Case Promotion, and Compliance

85

newspaper reported that the IDF had largely ignored the guidelines.29 One of the alleged violations was that the IDF ordered a killing in a situation in which innocent bystanders were likely to be killed. No matter how clear the court was in its initial decision, it is probably impossible to articulate perfectly what probability of collateral damage is sufﬁcient to restrict an assassination order. Insofar as this is true, noncompliance is not necessarily observable. In this sense, we might suspect that some decisions can never be rendered fully transparent. Critically, however, this cannot explain the result in Figure 4.2. If the lack of transparency is being driven by an inherent murkiness in some remedies, then we should not ﬁnd that prior media coverage matters. This is because the media will suffer from the court’s inability to clarify as well. So, if the inability to control transparency perfectly explains what we are observing, it must derive from some other feature of the process. The second possibility, which can be investigated in these data, is that courts cannot control perfectly the coverage they receive. If control is incomplete, then prior media coverage might continue to inﬂuence decision making. Supreme Court effects on media coverage The media’s desire for access should provide leverage to the court over its coverage (Cook 1996), but it is reasonable to believe that newsworthiness is in some ways independent of judicial appeals. This is increasingly true in Mexico where press independence has grown substantially over the past two decades (Díaz Barriga and Kleiber 1996; Lawson 2002). Whereas government agencies once routinely paid newspapers to publish government-produced stories (gacetillas), the modern-day Mexican press largely runs newsworthy articles as determined by professional journalistic norms (Lawson 2002, 84–88). To examine the court’s inﬂuence over subsequent media coverage, I estimate the effect of a press release on the probability that La Jornada will cover the case after it has been resolved. The dependent variable, Coverage After, is coded 1 if La Jornada ran a story on the case following the resolution and 0 otherwise. Because La Jornada’s beat reporter was particularly independent, the test is biased against ﬁnding a relationship. Jesus Aranda was far more likely than the average reporter to be familiar with cases worth covering, independent of the court’s prodding (Aranda 2001). The key causal variable is Press, again coded 1 if the court issued a 29 See URL: http://www.haaretz.com/hasen/spages/1041160.html.

86

The Politics of Constitutional Review in Mexico

table 4.4. Predicted Probability of La Jornada Coverage

Predicted probability

Press Release Issued (95% C.I.)

No Press Release Issued (95% C.I.)

Difference (95% C.I.)

0.16 (0.12, 0.20)

0.02 (0.01, 0.03)

0.14* (0.11, 0.17)

Source: Supreme Court of the Nation (http://scjn.gob.mx). Note: Table 4.4 displays predicted probabilities of La Jornada coverage of a Supreme Court constitutional decision by whether the Supreme Court issued a press release, holding Coverage at its mean value. The result is nearly identical when holding Coverage at its median value.

press release announcing the case and 0 otherwise. The most straightforward indicator of newsworthiness derives from the media’s own coverage of the conﬂict being adjudicated. In other words, the press itself should have the best sense of what kinds of cases are worthy of coverage. Thus, I control for whether La Jornada had covered the case prior to the Supreme Court resolution. As described in Chapter 4, Coverage takes a value of 1 if there was coverage and 0 otherwise. Table 4.4 displays predicted probabilities of La Jornada coverage by whether the Supreme Court issued a press release. It also contains 95 percent conﬁdence intervals around the predictions. The court appears to have a signiﬁcant, although ultimately limited inﬂuence over its coverage. Without a press release, Supreme Court decisions are extremely unlikely to be covered by the press. The predicted probability is 0.02. Indeed, the lower bound on the 95 percent conﬁdence interval for this estimate is effectively zero. By sending a press release, the court increases the probability of coverage by 700 percent. Still, it is clear that the court does not simply induce coverage with a press release. The results suggest that, although it is fair to treat transparency as endogenous to judicial behavior, it is unrealistic to believe that transparency is fully endogenous. There are limits to the court’s ability to inﬂuence its coverage. In this sense, the mixed results summarized above are probably best explained by the court’s imperfect control over its coverage. For this reason, beliefs about ex post coverage are more likely to inﬂuence decision making than they would if control were complete. Decisions to promote The results of the bivariate probit model are encouraging, but we have yet to consider the second equation. Why does the Supreme Court promote

Decisions, Case Promotion, and Compliance

87

table 4.5. Predicted Probability of Press Release by Case Outcome and Prior Media Coverage Merits Decisions (N = 852)

All Decisions (N = 1, 005)

No Prior Coverage

Prior Coverage

No Prior Coverage

Prior Coverage

0.77 (0.52, 0.93)

0.43 (0.09, 0.89)

0.78 (0.52, 0.94)

0.38 (0.11, 0.71)

0.02 (0.007, 0.04)

0.47 (0.14, 0.82)

0.39* (0.04, 0.69)

0.41* (0.08, 0.80)

0.31 (-0.12, 0.71)

Policy Struck Down 0.50 (95% Conﬁdence (0.28, 0.72) Interval) Policy Upheld 0.02 (95% Conﬁdence (0.005, 0.04) Interval) Difference 0.48* (95% Conﬁdence (0.26, 0.69) Interval)

Source: Supreme Court of the Nation (http://scjn.gob.mx). Note: Shows predicted probability of Supreme Court issuing a press release announcing its decision by whether it struck down the public policy challenged in the case and the prior media coverage. *p < 0.05.

some cases and not others? The baseline model offers no prediction, whereas the promotion model suggests that the court does so to address poor reporting of decision rationales and to ensure the public that the public may monitor the interaction if noncompliance is a potential problem that can be solved in part through transparency. Empirically, if the promotion model is correct, we should estimate a positive relationship between Strike and Press, especially so for resolutions unlikely to be covered by La Jornada. Table 4.3 provides preliminary support for this hypothesis across Models 1 and 2. The Strike coefﬁcient is large and strongly signiﬁcant in both models, and the interaction term (Strike*Coverage) is negative and signiﬁcant in both models as well. Table 4.5 shows the predicted probability of the court issuing a press release announcing its decision by whether it struck down the policy under review and whether the case had received prior media coverage. All other variables in the model are set to their mean levels. To evaluate the case promotion hypothesis, we must consider the difference between the predicted probability of a press release when the court upholds and strikes down the policy under review. The ﬁrst column indicates a massive effect of the court’s merits decision on the choice to promote. When La Jornada has not covered the case prior to the decision, the probability that

88

The Politics of Constitutional Review in Mexico

the court issues a press release announcing the case result is only 0.02. This probability increases to 0.50 simply by the court striking down the policy. As the last row of the table demonstrates, this difference, 0.48, is statistically signiﬁcant. When La Jornada has already provided coverage, the court is more likely to issue a press release, but the effect of the decision outcome is smaller – 0.39. Turning to the estimates from Model 2, we observe roughly the same relationship; however, the results are even more strongly consistent with the promotion model. In Model 2, the effect of Strike on Press is positive and signiﬁcant (0.41) when La Jornada has provided no coverage. But not only does the effect attenuate for cases in which La Jornada has provided coverage (0.31), it is statistically indistinguishable from zero. In other words, Model 2 suggests that when coverage is expected, there is no relationship between the merits decision and the probability of promoting, a result absolutely consistent with the argument that the court has two incentives to engage the media, one of which has to do with ensuring coverage at all and the other with ensuring that the coverage is accurate. In summary, exactly when public monitoring is needed and least likely to be available for lack of media coverage, the court promotes its cases. There is always an incentive to publicize in order to ensure accurate coverage when the media is likely to cover the resolution. However, when the media is not likely to cover the result, the case outcome is a powerful predictor of whether the court publicizes its decisions, precisely because these are the decisions for which public support is most helpful. Controls The empirical results provide support for some of the control variables suggested by the literature. In the ﬁrst equation, the identity of the complainant (plaintiff) clearly appears to inﬂuence decision making. Corporations, large political parties, state legislatures, and the federal executive were all more successful than individuals. These results provide support for the legal capacity argument (McGuire 1995; Rubio, Magaloni, and Jaime 1994; Sheehan, Mishler, and Songer 1992). Mexican organizations with greater resources are more successful before the Supreme Court. In addition, the measures of uniﬁed- divided government are both positive, though only Uniﬁed-Pl reaches statistical signiﬁcance.30 30 It should be noted Divided very nearly reaches conventional levels of statistical

signiﬁcance in the two-tailed test (p < 0.12).

Decisions, Case Promotion, and Compliance

89

This result provides support for the Iaryczower, Spiller, and Tommasi (2002) account and is consistent with results in Ríos-Figueroa (2006). The Supreme Court was less aggressive during the period of uniﬁed PRI government than it was after the PRI lost control of the Congress. Interestingly, Party is not signiﬁcant in either model, undermining the notion that the court was in some way especially attached to the PRI (Magaloni and Sanchez 2001).31 It is, of course, possible that the relationship between many of these variables and Strike is highly interactive, though not obviously implied by the particular theoretical argument I have advanced. For example, it is certainly possible that the Supreme Court was most sensitive to federal policy importance during periods of uniﬁed PRI governance. Turning to the second equation, the probability of a press release being issued increased based on the federal importance of the policy. This result is consistent with the court’s own account of its press releases. The Supreme Court is more likely to publicize cases as the importance of the case increases. Finally, it does not appear that Genaro Góngora was more likely than his predecessor to push the DCS to issue a press release.

Judicial Public Relations and Judicial Power The empirical tests summarized in this chapter are largely consistent with the promotion model of constitutional review, where the transparency of constitutional conﬂicts is endogenous to the conﬂicts themselves. There is evidence that the Supreme Court’s constitutional decisions are inﬂuenced by the federal importance of the policies it reviews. On the other hand, the results concerning the effects of beliefs in ex post media coverage are mixed. When estimated on all observations, the evidence suggests that the court’s choices to strike down public policies are not related to its beliefs about whether the media will cover its decisions. This much is consistent with the endogenous view of transparency. Yet, when estimated on only 31 Although the PRI result is tangential to the core argument in this chapter, it is interesting

to note the difference between the estimate here and in Magaloni and Sanchez. That said, I have no particular reason to doubt the partisan effects in Magaloni and Sanchez because there are a number of explanations. First and foremost, Magaloni and Sanchez do not include amparo suits in their analysis. More importantly, it is not clear how the authors have dealt with the issue of multiple case ﬁles, which I address in Appendix B. As a consequence of these different case selection rules, the results are not directly comparable. Of course, it is comforting to note that the uniﬁed government results are roughly identical in my results and those of Magaloni and Sanchez and Rios-Figueroa.

90

The Politics of Constitutional Review in Mexico

those cases in which the court reached the underlying constitutional question, the evidence suggests that the court is more likely to strike a policy when the media is paying attention, but only for insufﬁciently important policies, a result consistent with the exogenous account of transparency. Less ambiguous are the results on case promotion itself. I ﬁnd consistent evidence for the promotion model in the analysis of the court’s choices to issue press releases announcing case results. The court is more likely to do so when it strikes down public policies, especially if the case is unlikely to be covered if it does nothing. These results are robust to case selection. In summary, although the results are broadly consistent with the promotion model of constitutional review, there are pieces of evidence suggesting that the baseline approach, in which transparency is exogenous to judicial efforts, is the appropriate theoretical model. A theoretical middle ground is plausible. In it, constitutional judges are imperfectly capable of inﬂuencing the transparency of their conﬂicts. As a consequence, they must be sensitive to the media, because it cannot be controlled completely.

public authority compliance So far, the empirical analysis has centered on constitutional adjudication and the choices to promote resolutions. Yet the theoretical models make predictions about compliance behavior. As summarized in Table 2.1, if case promotion is relatively costless, then the promotion model suggests that we should not observe noncompliance; however, if case promotion is relatively costly, then both the baseline and case promotion models make the same prediction. Noncompliance is especially likely in relatively unimportant cases that go unreported by the media. The logic of this argument is straightforward. Courts should be extremely careful with salient cases, recognizing that government ofﬁcials might be willing to defy orders publicly when the policies matter enough. In this way, strategic deference reduces the opportunities for noncompliance in salient cases. As case salience declines, courts should be less deferential, and thus, the opportunity for noncompliance arises. Importantly, it is in these cases, when the media fails to cover the decision, that ofﬁcials are best able to ignore judicial decisions without cost. Now, if case promotion is costless for courts, they will always solve this transparency problem, and transparency will ensure compliance in the relatively low salience cases, just as strategic deference eliminates the possibility for noncompliance in high salience cases. For this reason, if promotion is costless, we should not

Decisions, Case Promotion, and Compliance

91

observe noncompliance. However, if promotion is relatively costly, then there will be sufﬁcient opportunities for noncompliance when courts elect not to signal to the media that a case is newsworthy. The central challenge in evaluating compliance with the data set I have used in this chapter concerns how to measure compliance with more than 1,000 Supreme Court decisions. The case ﬁles did not contain this information, and I could identify no individual at the Supreme Court who could tell me precisely what happened to each case after the decision was released.32 To evaluate the compliance hypotheses, then, I turn to another data set. I will evaluate public authority responses to the Supreme Court and its noncompliance incidents (incidente de inejecución), cases in which the court investigates claims that public ofﬁcials have deﬁed a federal judicial order in an amparo case. The advantage of this strategy is that the Supreme Court archives each incident, and the record carefully details the history of the deﬁant behavior under investigation. Although every case in the sample ultimately ends in compliance, there is massive variance in the time it takes to induce a public authority to comply. In many cases, the issue is not resolved until the responsible authority is replaced; in some cases, multiple replacements are necessary. Indeed, if we take the date on which amparo relief was ﬁrst granted by a federal district court as the starting point for the period of deﬁance, the longest period in the data set spans nineteen years! Thus, although decisions are eventually implemented, the court is deﬁed considerably in many cases. For this reason, I will conceptualize noncompliance temporally and code the time it takes between the date on which the Supreme Court formally admits an incident of noncompliance until the action that was required of the public ofﬁcial in the original decision is carried out. Noncompliance incidents The noncompliance incident gives force to Constitutional Articles 105 (III) and 107 (XVI), which grant the Supreme Court the power to remove a public ofﬁcial from ofﬁce for failing to implement a decision of a federal judge in a constitutional case. Noncompliance incidents may follow from actions of unconstitutionality, constitutional controversies, and amparo 32 Creyke and McMillan (2004), in their compelling analysis of administrative responses to

decisions of the Federal Court of Australia, obtained compliance measures by contacting the attorneys of record in each conﬂict. This proved impossible in Mexico because I was unable to gain permission at the Supreme Court to generate a contact list from the case ﬁles.

92

The Politics of Constitutional Review in Mexico

suits.33 In the event that a federal judge believes that an amparo decision is being ignored, she is required to contact the responsible authority and demand implementation. Typical communication is in writing. If this process fails, the judge may send the case to the Supreme Court. Federal district courts differ substantially in the time they spend seeking compliance with their orders. In the sample I describe below, the mean number of days spent seeking compliance prior to sending the case ﬁle to the Supreme Court is 404, with a standard deviation of 805. In actions of unconstitutionality and constitutional controversies, where the Supreme Court hears the conﬂict in the ﬁrst instance, the president of the Supreme Court makes the initial determination regarding noncompliance and then assigns the case to a minister. If the responsible authority ignores the court’s attempt to ensure compliance, the court is authorized to remove the authority from ofﬁce and submit him or her to the nearest federal penal court for prosecution. Thus, in principle, the incident of noncompliance grants the court a useful tool for ensuring judicial power. Of course, if de jure institutions were all that mattered, it is unlikely that we would have uncovered evidence of strategic behavior in the ﬁrst part of this chapter. The de jure threat of removal should eliminate the incentive for strategic deference, yet it does not appear to do so. The Supreme Court has explicitly exercised this power against a high-ranking federal ofﬁcial one time in its history. Unsurprisingly, the court did not issue its threat quietly. In the late 1990s, the Supreme Court was asked to address the failure of the federal Secretary of Agrarian Reform (SRA) to remove members of an indigenous commune in Ensenada, Baja California, from a valuable piece of beachfront property that they had gained illegally in the 1970s. The community had been renting parcels of land, largely to American retirees, despite a long-standing court order recognizing the title of the original owners. The political problems were clear. First and foremost, the community had consistently demonstrated an intent to resist an ejectment. Further, it is important to recognize that the lessees were almost entirely American, inserting a foreign relations issue into the conﬂict. Fearing the potential for violent conﬂict, federal authorities showed little enthusiasm for removing them by force. The court attempted to 33 The statutory implementation of these constitutional requirements is found in the

Amparo Law, Article 105, and the Federal Regulatory Law for Constitutional Article 105 and Article 47 for amparo, constitutional controversies, and actions of unconstitutionality, respectively.

Decisions, Case Promotion, and Compliance

93

resolve the dispute ﬁrst by negotiating informally with the SRA. Eventually, it became clear that no action would be taken, and so on October 23, 2000, the court set a short period within which the SRA was required to ensure the physical removal of the community members and their lessees. Failure to do so would result in removal. The important point is that when the court exercised its removal authority, it did so in an extremely public fashion. All major national and Baja news outlets were notiﬁed, and the court stated its requirements as clearly as possible. In this very public atmosphere, and in the wake of Mexico’s ﬁrst undeniably democratic election, the SRA complied. This case is reﬂective of the judicial concern for publicity when compliance is a problem. But, of course, we observe many of cases in which the Supreme Court’s requests for compliance are largely ignored. The key empirical questions are these: Can we say anything about the information environment in these cases? How much media coverage do these cases receive? Can we characterize the importance of the policies under review? Are these highly salient policies or not? Analysis The data set includes a 30 percent random sample of fully resolved incidents of noncompliance between 1997 and 2002, resulting in 113 analyzable cases. To code these cases, my research assistant (a law student at the Instituto Tecnológico Autónomo de México, ITAM) and I read every case ﬁle. We recorded the factual basis for the incident and the court’s own history of the case. Aside from formal case characteristics (e.g., case number, year, and court of ﬁrst instance), each observation indicates the dates of key decisions in the process, from the date on which the case was admitted in the ﬁrst instance to the date on which the Supreme Court ofﬁcially declared the case resolved. Table 4.6, which summarizes some simple characteristics of the sample, provides three essential pieces of information. Every case in the sample derives from an amparo suit. Indeed, I did not encounter an incident of noncompliance derived from a constitutional controversy or action of unconstitutionality in the full list of cases detailed in the Supreme Court archives. Second, I did not locate an article in La Jornada concerning any of the cases in the sample, prior to or following any of the court’s decisions. Third, I have similarly failed to locate a press release issued by the court over any of these cases in the sample. Now, the SRA case described above was not selected by the sampling procedure, and the court aggressively promoted

The Politics of Constitutional Review in Mexico

94

table 4.6. Type, Promotion, and Coverage of Incidents of Noncompliance Type of Case

Amparo Constitutional Controversy Action of Unconstitutionality

Press Release Issued?

Prior Media Coverage?

Media Coverage After?

Yes

No

Yes

No

Yes

No

0 0 0

113 0 0

0 0 0

113 0 0

0 0 0

113 0 0

Sources: La Jornada (http://www.jornada.unam.mx); Supreme Court of the Nation (http://scjn.gob.mx).

its decision in that case.34 It is important to recognize that the court exercised its constitutional removal authority in the SRA case, whereas it has simply pursued compliance through informal channels in the remaining incidents. Critically, all of the incidents of noncompliance in the sample occurred in what we have been thinking of as relatively unimportant cases, and the cases that enjoyed absolutely no attention from the national media. It would be impossible to characterize the policy environments in these cases as transparent, and it would be a stretch to call these policies salient. Table 4.7, which summarizes the type of complainants in these cases and the type of public authority challenged, provides additional evidence of the relative lack of case salience in the sample. Eighty-six percent of the noncompliance incidents on the Supreme Court’s docket were brought by individuals rather than the many corporate entities that have standing in amparo. In contrast, the percentage of individuals moving cases on the Supreme Court’s amparo docket was only 50. Thus, the status of the complainant in the incidents of noncompliance is relatively low on 34 The court has issued press releases over other two other incidents that were not selected.

Incident 493/2001 involved the failure of a local ofﬁcial in the federal district to pay damages for the unlawful taking of a piece of private property. The court, meeting in plenary session, was expected to remove the ofﬁcial from ofﬁce; however, the minister assigned to the case, Juan Díaz Romero, learned immediately prior to the meeting that at least half of the damages were deposited with district court managing the case. Incident 81/91 also involved the federal expropriation of a piece of property in the federal district. This case had become the subject of local media attention, but the complainant had long ago opted for the payment of money damages in lieu of seeking the return of the property. The press release merely informed the media that the court had never declared the president of the republic to be in deﬁance of an amparo resolution.

Decisions, Case Promotion, and Compliance

95

table 4.7. Complainants and Responsible Authorities in Incidents of Noncompliance and the Supreme Court’s Plenary Amparo Docket Type of Case

Incidents of Noncompliance Supreme Court’s Plenary Amparo Docket

Percentage of Complainants That Are Individuals

Percentage of Responsible Authorities That Are State or Local

86 50

85 41

Source: Supreme Court of the Nation (http://scjn.gob.mx).

average. Also, 85 percent of the public authorities whose deﬁant behavior the court investigated were state or local ofﬁcials, only 41 percent of the ofﬁcials challenged in the cases on the court’s plenary amparo docket were state or local. Again, the policies under review in the noncompliance cases appear considerably less salient, by the logic of the importance measure developed above. The simple statistics in Tables 4.6 and 4.7 are quite consistent with the theoretical argument, especially one in which case promotion is costly. Given the ﬁndings in Tables 4.6 and 4.7, it is tempting to stop at this point, because the ﬁndings are consistent with the compliance hypotheses derived in Chapter 2. Nevertheless, it is useful to push the data just a little further, largely because the information suggests a number of interesting features of the noncompliance in amparo. As I suggested above, the data set records the dates of important features of the process, and for this reason, a measure of the time it takes the Supreme Court to induce compliance can be derived easily. Speciﬁcally, we can count the number of days between the date on which the Supreme Court formally admitted the incident (the moment at which the court ofﬁcially recognizes the deﬁant behavior) until the date on which the original decision is fully carried out. For example, imagine a case (a common case as it turns out) in which Mexico City’s Public Security secretary fails to rehire a police ofﬁcer whom he ﬁred without cause and who successfully obtained an amparo requiring restitution. After sending ten letters requesting compliance, an administrative judge for the federal district remits the case to the Supreme Court, which admits it for review on January 1, 1999. On January 17, 1999, the court receives a letter from the secretary of Public Security reporting that the ofﬁcer was ofﬁcially rehired on January 10, 1999. The compliance measure, Compliance Time, would be coded 9,

96

The Politics of Constitutional Review in Mexico

the difference in days between January 10 and January 1. Note that this does not indicate the full length of deﬁance with the original order, which could have lasted for years. Conceptually, Compliance Time measures resistance to the Supreme Court, which only begins once the court has agreed to hear the case. Inefﬁciency in the federal system The mean Compliance Time is 111, with a standard deviation of 273; however, the most striking feature of the measure is the range: [−732, 1,372]. To be clear, the minimum of Compliance Time is a negative number, which is puzzling on its face. What is going on in such cases is that the public ofﬁcial complied with the original decision before the Supreme Court admitted the case. Amazingly, 25 percent of the cases in the sample involve compliance prior to admission. It is true that over half of the cases fall within one month of the date the lower court remitted the ﬁle to the Supreme Court, which suggests that many public ofﬁcials wait to comply until they receive notice of pending Supreme Court review. Still, nearly one-quarter of the negative Compliance Time cases involve compliance that occurred nearly one year prior to the Supreme Court admitting the case. The case ﬁles reveal a potential explanation. In each case, pursuant to the statutory requirement, the lower court judge attempted to uncover information about compliance by sending formal letters to the authority of record. The average number of letters sent in the sample is six, but the maximum is forty-one. In most cases, it is clear from the record that the public authority simply did not respond, and although it seems hard to believe, it appears that the lower court did not attempt to communicate with the ofﬁcial by telephone. It seems that, in many cases, the amparo was in fact implemented, yet the ofﬁcial did not bother to inform the court and continued to ignore its correspondence. Effectively, this element of the data reveals a serious inefﬁciency in the process by which amparos are implemented. This is especially true in cases where it is clear that the public authority implemented the decision long before the case ﬁle was remitted to the Supreme Court. What explains deﬁance in nonsalient cases? But what of the cases where Compliance Time is positive and the Supreme Court genuinely had to seek compliance actively? Conversations with

Decisions, Case Promotion, and Compliance

97

table 4.8. Hazard Models of Compliance Time Model 1 Draft Resolution Re-hire Hearing Transfer Return Land

Model 2

Model 3

Model 4

0.781 (0.25) 1.04 (0.79) 1.84 (1.91) 0.756 (0.25) 0.337* (0.17)

0.778 (0.25) 0.906 (0.65) 1.40 (1.45) 0.410∗∗ (0.14) 0.651 (0.340) 1.004∗∗∗ (0.0005)

0.700 (0.23) 0.916 (0.69) 1.37 (1.42) 0.381∗∗ (0.13) 0.404 (0.25) 1.004∗∗∗ (0.0006) 1.88 (0.81)

1.00∗∗∗ (0.08) 79 (79) −121.19 5.94

1.43∗∗∗ (0.13) 79 (79) −96.76 54.79∗∗∗

1.46∗∗∗ (0.13) 79 (79) −95.74 56.81∗∗∗

0.783 (0.26) 0.919 (0.70) 1.40 (1.46) 0.370∗∗ (0.13) 0.637 (0.45) 1.003∗∗∗ (0.0006) 3.23* (1.75) 0.389 (0.24) 1.49∗∗∗ (0.13) 79 (79) −94.48 59.36∗∗∗

Experience Federal Authority Agrarian Case p N (# failures) Log-likelihood χ2

Source: Supreme Court of the Nation (http://scjn.gob.mx). Note: Hazard models of Compliance Time assuming that the hazard function follows a Weibull distribution. Coefﬁcients are hazard ratios.

secretarios de estudio y cuenta (Supreme Court clerks), who are primarily responsible for managing these cases, suggested that the core problem in obtaining compliance arises in agrarian cases, where it is difﬁcult to get a public ofﬁcial to return land to an individual. Consistent with this story, the longest compliance time in the sample, 1,372 days, is a case involving the return of land that was illegally expropriated by the federal government. Table 4.8 summarizes the results of several parametric event history models of Compliance Time. It is worth stressing that, although the cases here largely involve administrative decisions, these are not merely “administrative law” cases. Each policy under review was deemed unconstitutional, typically in violation of the equal protection or due process clauses. Covariates in Model 1 are a series of dichotomous variables designed to measure the difﬁculty of implementing the order for the public ofﬁcial.

98

The Politics of Constitutional Review in Mexico

Reﬂecting the general lack of national policy salience, it turned out to be relatively easy to classify the deﬁant behaviors. The base category captures a scenario in which the ofﬁcial needed only provide the complainant with information. In most cases, this involved no more than sending a letter to a complainant indicating that she had been terminated. Each of the remaining covariates describes activities that are no less difﬁcult than simply providing information. Draft resolution indicates a requirement for the public ofﬁcial to produce a written document putting in writing a decision that it had already issued. Re-Hire indicates that the ofﬁcial needed to return a former employee to her position. Hearing indicates that the ofﬁcial needed to hold an ofﬁcial hearing. Transfer indicates that the ofﬁcial needed to transfer money to the complainant. This typically involved the payment of back salary, but the measure also captures scenarios in which the payment involves compensation associated with an illegally expropriated piece of property. Finally, Return Land indicates that the ofﬁcial needed to return a piece of real property to the complainant. These categories perfectly partition the set of deﬁant behaviors under review. Model 1 provides support for the general sense at the court that agrarian cases are problematic. Indeed, the only variable that appears to matter is Return Land, which reduces the hazard rate (i.e., increases the time of deﬁance). Substantively, the results suggest that cases in which the return of land is required tend to be more difﬁcult to resolve. That said, the overall model ﬁt is not good at all. Although this is primarily an exploratory exercise, it is still important to consider alternative explanations of deﬁance in nonsalient cases. I consider two. Model 2 includes a variable to account for the possibility that the court, which we must recall was relatively new when the sample begins, got better at solving these compliance problems over time. Experience records the date at which the case was admitted at the court, which will indicate an over-time trend.35 Model 3 adds a variable, Federal Authority. The variable again captures variance in policy salience in these relatively low salience cases. At the margins, if the argument is correct, we should observe increased compliance times with state or local

35 Recognize that this does not account for temporal variance in the underlying hazard rate,

which is already estimated in the model, and which is positive in all models. In other words, the hazard rate increases monotonically in survival time. In contrast, Experience simply captures the extent to which the time of admittance inﬂuences the hazard rate at a particular moment.

Decisions, Case Promotion, and Compliance

99

ofﬁcials. An interesting empirical pattern emerges in these models. Most striking, the Return Land effect disappears once the date of admission is controlled. It appears that these land return cases occurred relatively early in the new court’s tenure, and the court seems to have gotten better at pursuing compliance over time. Thus, while the worst case of noncompliance involves land, on average, the effect is negligible. In contrast, what seems to matter is Transfer. In addition, the models suggest that federal ofﬁcials are signiﬁcantly more likely to comply faster than state and local ofﬁcials. Given the results, it is possible to ask whether the Supreme Court clerks were wrong about the general trend in noncompliance cases, that the problem lies in agrarian law. Models 2 and 3 suggest that returning land is not really the problem. The variance in compliance times seems to be explained by structural factors, and if anything, the case characteristic that matters is money. Of course, agrarian cases often involve the cutting of checks rather than the return of land. So perhaps Transfer is picking up the effect of agrarian cases, but only insofar as they involve the transfer of money. To consider this possibility, Model 4 includes one additional covariate, Agrarian, which indicates whether the case involved a land issue in any way. Agrarian does seem to decrease the hazard rate; however, the result is not statistically signiﬁcant. On average at least, what matters in these relatively nonsalient cases is money, whether or not the money at stake derives from an agrarian conﬂict. This section has provided additional, indirect support for the theoretical model. Although the analysis does not constitute a precise test of the argument, it does appear that noncompliance with federal judicial orders seems to concentrate in cases of relatively low salience, and these also happen to be cases that are ignored by the media. It is instructive that there are no constitutional controversies or actions of unconstitutionality in the noncompliance sample. And, in the one case in which the Supreme Court exercised its power to remove a public ofﬁcial from ofﬁce, it did so in a highly public fashion. In a sense, this was the one incident of noncompliance in which the court made a deﬁnite public policy choice. It is reassuring that, in such a context, the court seemed concerned with transparency. Without a measure of compliance linked directly to the court’s plenary constitutional docket, it is impossible to test the model precisely. Nevertheless, the data on the incidents of noncompliance are highly suggestive of a pattern of strategic behavior that leaves noncompliance largely observable over policies that are relatively unimportant.

100

The Politics of Constitutional Review in Mexico

the supreme court in mexican politics This chapter suggests a few answers to questions raised in the introduction about the Supreme Court’s role in Mexican politics. The court’s remarkable decision in the Zedillo case represents the ﬁrst instance in the history of the Mexican republic in which the Supreme Court directly challenged the power of the Mexican president through the constitutional controversy. Viewed through the broader lens of the Mexican transition to democracy it is easy to interpret the case as marking another, equally important transition – namely, a transition to the rule of law. In a sense, the contemporaneous mass coverage of the Zedillo case, which highlighted the new and powerful political role of the Supreme Court, reﬂects this interpretation. Quite obviously, if this is the correct interpretation, it is a development of unquestioned importance. The results of this chapter suggest another, more sober yet perhaps more interesting view of the Zedillo case. The empirical results of this chapter are highly consistent with an important, yet fundamentally constrained political role for the Mexican Supreme Court. The evidence on the court’s constitutional decision making suggests that the Supreme Court has been politically aware and generally careful in the post-Zedillo constitutional reform era. Its decisions appear to respond to political concerns identiﬁed by generations of scholarship in judicial politics (Epstein and Knight 1998; Maltzman, Spriggs, and Walbeck 2000; Murphy 1964) and by recent comparative analysis of constitutional review (Helmke 2005; Iaryczower, Spiller, and Tommasi 2002; Vanberg 2005). Both in its decisions in constitutional cases and in its choices to publicize those cases, the court is sensitive to the political importance of the policies it reviews; but importantly, it behaves as if the public plays an important role in ensuring its ability to control the constitutional order. Similarly, the compliance data are quite consistent with a strategic model of constitutional review in which transparency is critical but costly. Many of the results summarized in this chapter are broadly consistent with other recent work on the Supreme Court (Magaloni and Sanchez 2001; Ríos-Figueroa 2007). The court has been more aggressive as Mexican government has fragmented. Still, the modern Mexican Supreme Court does not appear fundamentally different from the Supreme Court studied by Schwartz (1973), which also seemingly exercised an important, if limited control of the political authorities it was called upon to constrain.

Decisions, Case Promotion, and Compliance

101

All of this suggests that the Zedillo case itself should be viewed through a political theory of constitutional review. The court’s choice to hold the decision until after the election provided critical political space. By the time the decision was released, the PRI had lost its stranglehold on the executive branch, and Ernesto Zedillo was already engaged in a quite public effort to champion a peaceful, legal transition. Importantly, the Zedillo case highlights how courts can control the outcomes to highly salient policies, even if power, in general, decreases in salience. The critical point is that, although the policy was salient, the public costs of deﬁance were enormous for Zedillo, who was attempting to build a legacy as the PRI president who guided Mexico into a new era of democracy and the rule of law (Staton 2004). Further, the case was widely covered by the media – there could have been almost no possibility of missing the resolution. But even if these factors were insufﬁcient to induce an aggressive Supreme Court decision, recall that the complainant in this action was a minority of the federal Chamber of Deputies, whose candidate had just won the presidency and whose assent would have been required to attack the court for undue activism. The court resolved the Zedillo case under precisely the conditions conducive to judicial independence under a political theory of constitutional review.

appendix 4a The Supreme Court and the Mexican Federal Judiciary Figure 4A.1 represents the basic structure of the most important elements of the federal judiciary.36 The Supreme Court, which sits at the top of the 36 The Federal Judicial Council (Consejo de la Judicatura Federal) manages the judiciary’s

administrative matters, excluding those pertaining to the Supreme Court and the TEPJF. The council is made up of seven members named by the Supreme Court, the president of the republic, and the Senate. It is responsible for selecting and training future judges and for ensuring the general integrity of the judicial career. The council creates and eliminates federal courts and judgeships, develops the judiciary’s yearly budget, promotes its political interests, and investigates allegations of judicial corruption or incompetence. Excluded from Figure 4.1 and from my review below are the federal citizen juries, which may be formed by district court judges. Juries, which are not typically a part of the Mexican justice system, are composed of seven members selected at random from a district jury pool. The only requirements are that jurors must be literate Mexican citizens residing within the district containing the jury to which they are called. The procedures concerning jury selection and duty are deﬁned in the Federal Code of Criminal Procedure, Title IX, Chapter II, Articles 308–350.

102

The Politics of Constitutional Review in Mexico Supreme Court of the nation

Second Bench Administrative and Labor Law

First Bench Criminal and Civil Law

Federal Electoral Tribunal

Judicial Council

186 Collegial Circuit courts

70 Unitary Circuit Courts Five Regional Tribunals 314 District Courts

ﬁgure 4A.1. Structure of Mexican Federal Judiciary. Source: Consejo de la Judicatura Federal (http://www.cjf.gob.mx)

judicial hierarchy, resolves cases en banc and in two benches, specialized by subject matter.37 The major exception to the Supreme Court’s primacy among judicial tribunals concerns electoral law, where jurisdiction, with two exceptions, is exclusively granted to a special set of electoral courts (see Eisenstadt 2004). The highest court in this system is the Federal Electoral Tribunal (Tribunal Electoral Federal del Poder Judicial or TEPJF), which reviews decisions of ﬁve regional electoral tribunals. The decisions of the TEPJF are ﬁnal, except when it issues jurisprudential theses that directly interpret the constitution or when parties directly challenge the constitutionality of an electoral law.38 Below the Supreme Court, the republic is divided into 29 circuits, whose geographic jurisdictions are deﬁned according to state boundaries, with a few notable exceptions.39 Three hundred fourteen federal district courts spread across the circuits currently serve as the federation’s courts 37 The First Bench resolves cases concerning criminal and civil law, while the Second Bench

hears cases involving administrative and labor law. 38 Constitution of the United Mexican States (CPM), Article 99, §9 and Article 105, §2. 39 The exceptions to this rule include the states of Colima and Jalisco (3rd Circuit), Puebla

and Tlaxcala (6th Circuit), Campéche and Yucatán (14th Circuit), Hidalgo and Querétaro (22nd Circuit), and Aguascalientes and Zacatecas (23rd Circuit). Also, although it is rare, there are some municipalities that belong to circuits other than those that cover their own states. For example, the municipality General Simón Bolívar is part of the state of Durango. However, the municipality falls within the 8th Circuit, which covers

Decisions, Case Promotion, and Compliance

103

of ﬁrst instance. Unitary circuit tribunals, collegial circuit tribunals, and the Supreme Court directly may review their decisions, depending on the subject of the case and type of decision. Key Constitutional Actions In this section, I describe key features of the most important actions of Mexican constitutional review: the amparo suit, the constitutional controversy, and the action of unconstitutionality. It is useful to begin with amparo, because it is the oldest institution and provides a baseline against which the newer actions can be fruitfully compared. Amparo Constitutional articles 103 and 107 answer a number of questions about the nature of amparo. They tell us that laws may not be challenged in amparo until they are fully enacted, that only the federal courts are competent to hear amparo claims (Article 103), and that decisions in amparo only have speciﬁc effects (Article 107, §2). Beyond these basic constraints, the vast majority of rules governing amparo are regulated by federal law.40 All Mexican citizens of majority status may seek amparo relief, although aliens have standing under most circumstances.41 In addition, the state of Coahuila. (For the complete list of such irregularities, see Acuerdo General 23/2001, Pleno Consejo de la Consejo Federal.) 40 The Ley de Amparo contains 234 detailed articles on procedure, and there are countless jurisprudential theses on the subject as well. For the interested reader, Burgoa (1998) remains the deﬁnitive work on the subject; however, Ferrer Mac-Gregor’s (2000) is somewhat more accessible and provides a comparison between Mexican and Spanish amparo. As stated before, Baker (1971), although dated now, is the best book-length, English source with which I am familiar. 41 The president has exclusive authority over deportation, and his decisions are not reviewable by the courts. The relevant portion of Article 33 reads, “[T]he Executive of the Union will have the exclusive capacity to expel from the national territory, immediately and without prior necessity of a trial, all foreigners whose stay is deemed objectionable.” As discussed by Burgoa (1998, 355–377), standing in amparo may be broken into two related by distinct legal concepts: personality and capacity. Personality is a concept speciﬁc to the judicial proceeding and is a description of the subject who participates in the procedural aspects of the case. Personality in amparo may be original, as when a person seeks relief on his own behalf, or derivative, as when a person is properly contracted by the claimant to seek relief on her behalf. In contrast, capacity is an abstract concept, a general description of the entity’s nature. Capacity implies both (1) personal legal status, that is, whether or not the entity is subject to the rights and obligations afforded natural or artiﬁcial persons as provided for generally by the law; and, (2) whether or not the entity possesses the procedural faculty, as deﬁned by law, to exercise his rights (355–356).

104

The Politics of Constitutional Review in Mexico

many civic organizations, all autonomous indigenous communities, and even the government when acting as a private moral person may bring amparo suits.42 In this sense, standing in amparo is relatively broad. A number of artiﬁcial persons, both private and public, may have access to amparo as long as they state an individual rights violation. Yet the cases in which public moral persons possess the proper capacity to move amparo claims are very limited, and thus, amparo is generally considered a means for private individuals or organizations to seek constitutional protection. The 1917 constitution originally divided amparo jurisdiction between the Supreme Court and the federal district courts, with the former exercising jurisdiction in what is called direct amparo and the latter in indirect amparo. Although the Amparo Law has been reformed substantially since 1919, the essential difference between direct and indirect amparo remains. Indirect amparo claims are heard in the ﬁrst instance by the federal district courts or, in limited cases, by the unitary circuit courts in response to the following actions: (1) the publication of laws that by their mere promulgation prejudice the claimant’s liberties, (2) acts and decisions not arising out of judicial, administrative, or labor tribunals, (3) judicial, administrative, or labor tribunal decisions executed outside the bounds of the trial or after its conclusion, (4) acts within a trial whose executive would cause irreparable damage, and (5) decisions within a trial that affect parties outside the trial (Carranco Zúñiga 2000, 317). Thus, in indirect amparo, the Supreme Court exercises appellate jurisdiction. Direct amparos involve appeals of ﬁnal judgments in criminal or civil cases. As Baker (1971, 184) notes, these are legal outcomes Which resolve the principal point or points at issue in a suit and against which the law affords no ordinary recourse by which they may be modiﬁed or reviewed. This would apply to suits for which the law provided trial in single instance only, and to those in which recourses such as appeal have been exhausted. (Baker, 184)43 42 Private moral persons, as deﬁned by Article 25 of the Civil Code include (1) civil and mer-

cantile societies, (2) unions and professional associations as provided for by Article 123 of the Constitution, (3) cooperatives, and (4) legal, political, artistic, and scientiﬁc interest groups. Foreign companies are provided access to amparo as long as they are properly registered with the Mexican government. (See, Tesis, Sociedads Mercantiles Extranjeras, Modo de Acreditar su Existencia Juridíca, Semanario Judicial de la Federación, Tercera Sala, Volumen XXVI, Cuarta Parte, 193. On moral persons generally, see Tesis Personales Morales Privadas, Semanario Judicial de la Federación, Tomo XIII, 962.) The difference between an ejido and a communal is that in the case of an ejido, the community was granted communal property rights over some set of land by governmental degree, whereas communales are communities that have always enjoyed rights to their properties. (See Ferrer Mac-Gregor 2000, 239.) 43 See Article 46, Ley de Amparo.

Decisions, Case Promotion, and Compliance

105

Whereas the Supreme Court used to try direct amparos in the ﬁrst instance, they are now heard in the Collegial Circuit Courts (i.e., colegiados). These decisions are not appealable unless a colegiado declares a law or international treaty unconstitutional. In such cases, the Supreme Court may hear appeals when the ministers determine the issue is signiﬁcant.44 Although constitutional article 103 supplies amparo relief against laws, it is not clear whether it is the law itself or the execution of that law that invites amparo. Consequently, there is some ambiguity over the concreteness of the act necessary to trigger review. Article 114 of the Amparo Law clariﬁes this ambiguity, stating that laws may be challenged in the absence of a speciﬁc application only when the law’s mere enactment causes an unconstitutional prejudice of individual rights upon an identiﬁable person, that is, when the law is self-executing (auto-aplicado).45 Insofar as amparo relief may be sought against a law or regulation absent a speciﬁc application, we can consider amparo an abstract form of judicial review. On the other hand, the doctrine on self-executing laws makes clear that claimants must show an actual and immediate prejudice in the particular case arising out of the adoption of the law under attack. In that sense, the controversies adjudicated in amparo are concrete. Amparo primarily protects individual constitutional rights, but the Supreme Court has interpreted Mexico’s due process clause broadly enough to extend amparo’s coverage to nearly the entire constitution. Article 14 states that “no person shall be deprived of life, liberty, property, possessions, or rights without a trial by a duly created court in which the essential formalities of procedure are observed, and according to laws issued prior to the act.” The last clause extends amparo relief to all constitutional provisions, not just the individual rights guarantees. The argument hinges on the interpretation of the word law and emphasizes the difference between the way law is used in Article 14 and Article 103, the amparo provision. In the latter, law describes any public act of undetermined validity. This we know because the federal courts are given

44 See Acuerdo 5/1999 of the Pleno of the Supreme Court of the Nation for details on this

procedure. 45 See Tesis Leyes Autoaplicados, Amparo, Semenario Judicial de la Federación, 103–108, pag. 167. Primera Parte, 7a . Epocha, Pleno. A number of conditions must be met in order

to show that a law is self-executing. They are the following as summarized by Baker: (1) the provisions of the statue must identify clearly and unmistakably, by establishing explicit classes, those persons to whom it is applicable, (2) the persons so identiﬁed must be subjected, ipso jure, to an obligation, (3) compliance with the obligation would necessitate performing an act not previously required or abstaining from an act formerly permissible, and (4) such compliance would result in a prejudicial modiﬁcation of the rights (168).

106

The Politics of Constitutional Review in Mexico

authority over the interpretation of these laws. In contrast, law in Article 14 presupposes the act’s validity. On Rabasa’s (1982) account: The Constitution, when it speaks of laws in the sense of norms (for instance as in article 14), cannot be considered as referring to any legislative act whatsoever but only to those which in origin and content fall within its purview and which in conformity with it are valid. An unconstitutional law cannot be law in the language of the Constitution unless we admit that [that document] depreciates and nulliﬁes itself by its own terms.46

In this sense, a law that violates any part of the constitution may violate an important individual liberty – namely, the liberty from being bound to follow unconstitutional laws as granted by Article 14 (see Baker 1971, 119–123). Further highlighting its expansive nature, amparo implies the constitutionalization of statutory interpretation. The ﬁrst paragraph of Article 16 states, “No one shall be disturbed in his person, family, domicile, papers, or possessions except on the authority of a written order issued by the competent authority stating the legal basis and justiﬁcation of the action taken.”47 Article 16 demands that public ofﬁcials show a proper legal basis for their authority. Public acts must derive from and be in conformity with statutory provisions (Baker, 126).48 To enforce this right, federal courts must interpret statutory meaning. In this way, amparo requires that federal courts engage in statutory interpretation as a means of enforcing Article 16 rights. Formally, this is done through constitutional review. Constitutional controversies and actions of unconstitutionality The judicial reforms of the 1990s ampliﬁed the nature of the constitutional controversy, which existed largely unused throughout the twentieth century. They also introduced the action of unconstitutionality. The constitutional controversy essentially provides entities of the state an 46 As quoted in Baker 1971, 121. 47 Article 16’s competent authority provision implies another method by which amparo has

been extended beyond the individual guarantees in constitutional decisions. As discussed in Burgoa (1998, 257–259) and reviewed by Baker (1971, 125–126), all laws and acts emanating from public ofﬁcials who exercise authority not properly granted to them by the constitution are deemed unconstitutional via Article 16. Thus, for example, any law passed by a state that violates the sphere of authority granted to the federation is facially unconstitutional. 48 See Burgoa (1998, 265).

Decisions, Case Promotion, and Compliance

107

avenue to challenge the validity of acts that invade their sphere of competency. As detailed in Article 105, constitutional controversies may arise between the following entities: 1. The Federation and a state or the Federal District; 2. The Federation and a municipality; 3. The Federal Executive and the Congress or the Permanent Commission; 4. Two states; 5. A state and the Federal District; 6. The Federal District and a municipality; 7. Two municipalities of distinct states; 8. Two branches of the same state over the constitutionality of state acts or laws; 9. A state and one of its municipalities over the constitutionality of state acts or laws; 10. A state and a municipality of another state over the constitutionality of state acts or laws; and 11. Two parts of the Federal District government over the constitutionality of district acts or laws. The Supreme Court enjoys exclusive jurisdiction over constitutional controversies and hears only concrete claims through a posteriori review.49 Effects in constitutional controversies may be general or speciﬁc as detailed in the last two paragraphs of Article 105. When the federation moves against the a state or municipality, decisions supporting the federation’s claim by a majority of eight ministers set general effects over the entire federation. On the other hand, if a state moves against the federation and the court supports the state position, the law is only invalid in the state from which the claim was moved. In this sense, the effects are speciﬁc to the case. In contrast to the constitutional controversy, the action of unconstitutionality grants the Supreme Court the exclusive power of abstract review over the constitutionality of state and federal laws. Parties only may move an action of unconstitutionality after the law’s enactment and must do so within thirty days of ﬁnal passage.50 The court’s decisions in these cases always set general effects whenever eight or more ministers agree on the 49 Constitutional controversies must be moved within thirty days of the adoption of a

challenged law and sixty days following the establishment of a territorial boundary. 50 See Article 105, §2.

108

The Politics of Constitutional Review in Mexico

resolution. Like the constitutional controversy, individuals do not have standing in actions of unconstitutionality. Instead, Article 105, §2 details that the following entities will have standing: 1. Thirty-three percent (33%) of the Chamber of Deputies against a federal or Federal District laws adopted by the Congress; 2. Thirty-three percent (33%) of the Senate against federal or Federal District laws adopted by the Congress, or international treaties celebrated by the Republic; 3. The Attorney General of the Republic against federal, state or Federal District laws, or international treaties celebrated by the Republic; 4. Thirty-three percent (33%) of the members of the state legislatures against the laws of that state; 5. Thirty-three percent (33%) of the Representative Assembly of the Federal District against Federal District law adopted by the Assembly itself; and 6. Political parties properly registered with the Federal Elections Institute via their national representative against federal laws or representatives registered at the local level via their representatives against their own state laws. 7. The National Commission of Human Rights against federal laws, state or local laws and international treaties. State commissions of human rights against state laws and the Commission of Human Rights for the Federal District against laws enacted by the Legislative Assembly of the Federal District. As should be clear from these provisions, the action of unconstitutionality is a mechanism for ensuring that political minorities enjoy some access to constitutional review; however, not every minority is granted standing. To review, the Supreme Court enjoys exclusive jurisdiction over constitutional controversies and actions of unconstitutionality, while the vast majority of amparo suits are handled by the lower federal courts. Indeed, the Supreme Court only hears the most signiﬁcant claims against the constitutionality of federal laws. All forms of judicial review are conducted a posteriori. Standing in amparo is complicated. Both natural and moral (both private and public) persons are granted standing; however, the state also may bring an amparo claim when it subjects itself to the law, as would a private entity. The state, in its sovereign capacity, is prevented from seeking amparo relief. In contrast, standing in constitutional con-

Decisions, Case Promotion, and Compliance

109

troversies and actions of unconstitutionality is entirely straightforward as provided for in the constitution. Only corporate entities, not individuals, are granted access to these forms of review. In terms of the type of review, both amparo and constitutional controversies arise out of concrete acts, while the action of unconstitutionality may be used to challenge a possible violation of the constitution, absent any immediate controversy. Similarly, the basis of review in amparo and the constitutional controversy may be either legal or constitutional, while the basis of review in the action of unconstitutionality is always constitutional. Of course, it should be noted that the Supreme Court does not currently hear legality amparos. Finally, the effects in amparo, as deﬁned originally in Otero’s formula, are always speciﬁc to the case under review. In contrast, the effects in the action of unconstitutionality are always general. Effects in the constitutional controversy may be either general or speciﬁc, depending on the parties to the case.

From the lower courts to the Supreme Court Whereas the Supreme Court hears constitutional controversies and actions of unconstitutionality under its original jurisdiction, the court entertains amparo suits through the recurso de revisión or revision appeal in essentially two ways.51 Nearly every amparo decision arrives under the court’s mandatory jurisdiction. Parties in amparo suits may seek revision appeals against unfavorable lower court decisions on the merits or against dismissals for procedural default in both indirect and direct amparo. In indirect amparo, if the original complaint impugned the constitutionality of a federal or state law or international treaty, or if the district court’s decision directly interpreted a constitutional precept, the Supreme Court must hear the appeal.52 In direct amparo, circuit court decisions 51 There are two other appeals in amparo: the queja and the reclamación. Although a

complete explanation of these appeals is well beyond the scope of this book, they differ from revisión in important ways. In short, queja appeals are ﬁled in order to address possible instances of noncompliance or in order to challenge the admission of a case that is obviously inadmissible. In contrast, the reclamación appeal is the proper recourse to challenge procedural decisions of the president of the Supreme Court, presidents of the Supreme Court’s benches, or presidents of the Collegial Circuit Courts. For example, one could challenge a decision of the Supreme Court president remitting a case to one of the Collegial Circuit Tribunals, arguing that it should be properly heard by the Supreme Court itself. Such appeals are then resolved by the full court. See Burgoa (1998) for a thorough review of these procedures. 52 Ley de Amparo (LA), Article 84, §1a.

110

The Politics of Constitutional Review in Mexico

may be appealed if they directly rule on the constitutionality of a law or international treaty.53 A constitutional amendment in 1988 granted the Supreme Court a limited, yet potentially important form of discretionary amparo jurisdiction (atracción). If a case falls outside of the court’s appellate jurisdiction but the court deems some element of the case to be fundamentally important to Mexican law, it may exercise atracción and take up the matter itself.54 In 1996, the court agreed to review an amparo suit initiated by Manuel Camacho Solís, the former regent of Mexico City. In 1997, the Mexico City regency was transformed from an appointed position to an elected mayorship; however, the reform prevented any person who had previously served as the Mexico City regent from running for ofﬁce.55 This rule was widely understood as a PRI effort to ensure that Camacho, who had lost favor with the current leadership, but maintained a signiﬁcant degree of popular support, could not run. In his complaint, Camacho claimed that the internal congressional procedures used to debate the proposed amendment were unconstitutional, and that as a result, the law itself was unconstitutional. A district court judge ruled the suit inadmissible on the grounds that internal legislative rules were not open to amparo challenge. The conﬂict clearly presented an important constitutional question: are Congress’s internal rules subject to judicial review via amparo? However, because the district court judge did not address the constitutionality of a federal law, state law, or international treaty, nor did the original claim impugn the constitutionality of a federal law, state law, or international treaty, the Supreme Court was not competent to hear the appeal. Thus, the only way for the Supreme Court to take up the matter was through atracción. In the end, the court decided in a 6–5 decision to require the district court judge to admit the suit and rule on the merits. Unfortunately for Camacho, despite getting his amparo suit admitted, the claim was denied, and he was ultimately incapable of

53 The court is similarly required to review revision appeals in amparos brought to challenge

federal laws or acts that infringe on the sovereignty of the states or state laws or acts that infringe on the sovereignty of the federation. Amparos brought under Article 103, §§2 and 3 are relatively rare as discussed in chapter 3. 54 In addition, the federal attorney general may ask the court to consider a case allegedly of national interest, and the colegiados may certify appeals for Supreme Court review. Given the court’s relatively large caseload, it does not exercise attraction frequently. Constitución Política de los Estados Unidos Mexicanos, Article 107, §8. 55 CPM, Article 122, §1.

Decisions, Case Promotion, and Compliance

111

running in the 1997 Mexico City mayoral election.56 That said, atracción allowed the court to establish an important precedent – that legislative procedures were not immune from judicial inspection under amparo.

Case assignment All constitutional claims submitted for review by the Supreme Court are registered in Oﬁcialía de Partes (OP), an administrative branch of the presidency directly responsible to the Secretarios de Acuerdo.57 Clerks in the OP date the cases, assign case numbers, and send them either to the Subsecretario General of Acuerdos for the full court or to one of the Secretarios de Acuerdos for the two benches. Insofar as constitutional controversies and actions of unconstitutionality are only resolved en banc, they are sent to the Subsecretario General of Acuerdos. His staff, in turn, passes the cases on to the Unit on Constitutional Controversies (or Actions of Unconstitutionality). Secretarios de Estudio y Cuenta, clerks directly responsible to the ministers, staff both units.58 Upon receipt, clerks assign cases to one of ten ministers (ministro instructor) so that an initial opinion may be developed. Whenever possible, clerks make opinion assignments with an eye toward spreading the workload across the court’s membership.59 Ministers then examine the case and determine whether or not to recommend admission and whether it should be resolved in plenary session or by one of the benches. If an amparo suit is to be resolved by the full court, it is sent to the Subsecretario de Acuerdos for the Supreme Court. Otherwise, it is assigned to the Secretario de Acuerdos for the relevant bench. In either case, clerks develop recommendations regarding the admissibility of the suit, which they submit to the president of the Supreme Court (or of the relevant 56 See Amparos en Revisión 2996/96 and 1334/98, Suprema Corte de Justicia de la Nación. 57 Secretaries of Acuerdos are the president’s agents responsible for ensuring the proper

assignment and progression of all matters before the court. 58 Each minister has eight secretaries (Muredo 2000), who in many respects resemble the

clerks of the U.S. Supreme Court. However, in contrast to clerks of the U.S. Supreme Court, Secretarios de Estudio y Cuenta are following the judicial career path, having been appointed by the National Judicial Council upon successful completion of a series of entrance examinations. They also tend to have many years of experience in the judiciary. In fact, their position is the sixth step in the judicial career path – a position from which the majority of new federal district court judges are chosen. 59 The president does not develop projects.

112

The Politics of Constitutional Review in Mexico

bench). Case assignment in amparo is a fairly simple process. The Secretario de Acuerdos maintains a list of the ministers, organized by date of entry into the federal judiciary and then by age. The clerk responsible for assigning the case simply sends cases to successive ministers on the list.60 Once ten cases are assigned, the clerk returns to the top of the list and repeats the process.61 Written opinions: Amparo Once a case is assigned to the full court, the minister in charge of developing the opinion (ministro ponente) delegates control over the initial process to a clerk (secretario de estudio y cuenta). The clerks research the record and return a preliminary resolution to their minister. Upon revision, ministers present their projects in plenary session, at which point the membership debates its merits. Following debate, a vote is taken on the minister’s proposed resolution. If the resolution receives majority but not unanimous support, any member of the minority may issue a formal dissent (voto particular). If the president of the Supreme Court receives the dissent within ﬁve days of the vote, he will append it to the ﬁnal resolution.62 In the event that the resolution is rejected, the president assigns the case to one of the members of the majority in order to develop an alternative project, which the court will take up when completed. If the vote results in a tie, the president tables the matter and brings it up again in the following session. This is done in the hope that previously absent ministers will return and participate. Because no minister may abstain, a vote by the full court must generate a majority.63 If there is no majority in the second session, the president’s vote resolves the suit. Written opinions: Constitutional controversies and actions of unconstitutionality The basic procedures for developing sentences in constitutional controversies and actions of unconstitutionality are similar to those for amparo 60 If two distinct cases challenge the same governmental act, both cases are assigned to the

same minister (Interview with Froylán Borges Aranda, August 2000). 61 Interview Borges Aranda, August 5, 2000. Ley Orgánica del Poder Judicial de la Feder-

acíon (LOPJ), Article 6 grants the court the ability to close its sessions when it deems it necessary. 62 LOPJ, Article 7. 63 Ministers may recuse themselves from the resolution of a reasonable set of cases (LOPJ, Article 7).

Decisions, Case Promotion, and Compliance

113

appeals in the sense that individual ministers develop preliminary projects with the assistance of their clerks; and those projects set the debate agenda in the court’s merits conference. Still, the court’s procedural rules concerning these cases differ precisely because amparo appeals are handled in writing whereas the constitutional controversy and action of unconstitutionality appeals allow for public evidentiary hearings. In contrast to an amparo appeal, constitutional controversies or actions of unconstitutionality are assigned by the president directly to a particular, who in turn determines its admissibility. The constitutional controversy is inadmissible against decisions of the Supreme Court, electoral laws, laws that are currently under review in a pending constitutional controversy, laws whose effects have expired, and in cases in which the party moving the controversy has not exhausted all other means of resolving the dispute.64 Rules over admissibility are identical for the action of unconstitutionality with two exceptions. First, electoral laws may be challenged. Second, the action must be moved within thirty days of publication of the law or treaty being challenged. When the minister assigned to the case recommends admitting a constitutional controversy, he will request an answer from the challenged party, which must be ﬁled within thirty days. Having received the answer, the minister sets a hearing for the presentation of evidence. Although the hearing should be held within thirty days, ministers may extend this period if necessary.65 As is the case with amparo hearings, all types of evidence admissible in ordinary trials are admissible here, including verbal and expert testimony.66 Written evidence, however, must be provided in advance of the hearing. Following the hearing, the minister develops a preliminary resolution to the controversy and presents it to the full court. Procedure is relatively similar for actions of unconstitutionality; however,

64 The temporal requirements for the constitutional controversy are as follows: If the con-

troversy is brought against a law, the complaint must be ﬁled within thirty days of its publication or on the day following its ﬁrst application. If the controversy is over a congressional determination on the territorial limits of a state, the complaint must be issued within 100 days of the determination. Finally, if the controversy challenges an act, like an executive degree or regulation, the complaint must be ﬁled within thirty days of the day following the execution of the challenged act. Ley Reglamentaria de las Fracciones I y II del Articulo 105 de la Constitución Política de los Estados Unidos Mexicanos (LRA105), Article 19. For the action of unconstitutionality, the complaint must be ﬁled within thirty days of the law or treaty’s publication (LRA105, Article 60). 65 LRA105, Article 29. 66 Ministers may also assign experts to the case in order to develop an independent evaluation of the facts. LRA105, Article 32.

114

The Politics of Constitutional Review in Mexico

public hearings are not necessary. In lieu of a hearing, the minister may simply request information from the parties as he or she sees ﬁt.67 Voting rules on the merits are identical for constitutional controversies, actions of unconstitutionality and amparo; however, the consequences (or effects) of each decision depend on the type of constitutional review. As reviewed in Chapter 3, amparo decisions are always restricted to the parties involved in the suit. This is not always true in constitutional controversies or actions of unconstitutionality. When the federation moves a controversy against the laws or acts of a state or municipality, decisions supporting the federation’s claim, when made by a super-majority of eight ministers, set general effects over the entire federation. In all other cases, the court’s decisions only affect the immediate parties to the case. Similarly, for the action of unconstitutionality, the effects of the decision are general if the court resolves the case by a super-majority of eight votes. Otherwise, the decision only affects the immediate parties to the case.

appendix 4b Case selection As identiﬁed by the Supreme Court’s own summaries, the court resolved 3,084 distinct constitutional actions (in amparo, controversia constitucional, and acción de inconstitucionalidad) between January 1, 1995 and December 31, 2002. The data set restricts the number of distinct analyzable cases to 1,536, of which 1,006 were resolved between 1997 and 2002. Here I describe how I arrived at the ﬁnal number of cases. The reduction in cases reﬂects an attempt to resolve the problem that distinct constitutional actions were clearly not independent observations for the purpose of statistical analysis. The problem results from the fact that, when multiple individuals challenge identical public policies, the Supreme Court rarely consolidates the cases. Thus, what might be considered one constitutional case results in tens, sometimes hundreds, of distinct case ﬁles. Sometimes these cases raise different issues. Insofar as they do, I treat them as different observations. However, these cases frequently attack the same legal provision for identical reasons. For example, one federal-municipal conﬂict generated 10 percent of the court’s 67 RA105, Article 68.

Decisions, Case Promotion, and Compliance

115

total constitutional actions. All were ﬁled by different municipalities challenging the same congressional procedure. The court resolved all of these actions on the same day for precisely the same reason. These are clearly not identical observations for the purpose of analyzing Supreme Court decision making. In order to address this problem, I considered distinct actions to be identical observations according to the following four rules. The rules are designed to ensure that the distinct constitutional actions were identical and ought to be treated as such. Policy Rule: The challenged law, regulation, administrative decision, or decree must be identical across case ﬁles. Resolution Date Rule: The resolution date must be identical across case ﬁles. Voting Rule: The individual votes of the court’s members must be identical across case ﬁles. Rationale Rule: The rationale for the decision must be identical across the case ﬁles. The data set includes the ﬁrst action in a string of identical actions as determined by the rules above. Cases are assigned case numbers according to the date and time at which they arrive at the court’s Parties Ofﬁce. As a result, the order of cases is generated by random process induced by the postal service.

La Jornada article search Articles from La Jornada were selected via a two-step procedure. First, I searched the sections Primera Plana [Front Page], Contraportada [Back Page], Pólitica [Politics], Economía [Economy], Sociedad y Justicia [Society and Justice], and Estados [The States] on La Jornada’s Web site (www.jornada.unam.mx). Second, I pulled an article for possible inclusion if the headline or subheading contained any of the following Supreme Court minister names, words, or abbreviations: José Vicente Aguinaco Alemán Guillermo Iberio Ortiz Mayagoitia Humberto Román Palacios Juan Díaz Romero José de Jesús Gudiño Pelayo Juan N. Silva Meza

116

The Politics of Constitutional Review in Mexico

Juventino Víctor Castro y Castro Mariano Azuela Güitrón Sergio Salvador Aguirre Anguiano Olga María del Carmen Sánchez Cordero Dávila de García Villegas José Ramón Cossío Díaz Suprema Corte Suprema Corte de Justicia SCJN SC Controversia Constitucional Controversia Acción de inconstitucionalidad Acción Amparo Amparo en revision Amparo Directo Tribunal Corte With these articles in hand, I then read each article to determine whether or not it concerned a pending or resolved constitutional action. Validity of federal policy importance measure Face validity I discuss the face validity of the measure in the text on pages 74–75. To review, the simple idea is that federal ofﬁcials should care more about federal policies than state policies and more about those policies challenged under the constitutional controversy or the action of unconstitutionality than those challenged under amparo, because the decisions in the former cases can set precedent. Construct validity Other scholars provide support for the assumptions that underlie my measure of policy importance. Helmke (2002) and Iaryczower, Spiller, and Tommasi (2002) test their models on “important” cases, and they limit their analyses to national policies. Also, the other Mexicanists who have studied the court’s decisions of late (Magaloni and Sánchez 2001; RíosFigueroa 2007) focus only on constitutional controversies or actions of unconstitutionality, arguing that they wish to select only “important”

Decisions, Case Promotion, and Compliance

117

table 4B.1. Validity Analysis (Dependent Variable Is Strike) Model 1 β (RSE) State Amparo State Controversy or Action Federal Amparo Federal Controversy or Action N

0.46∗∗ (0.18) 0.547∗∗∗ (14) 0.184 (0.19) – 1005

Model 2 β (RSE) 0.267∗∗∗ (0.03) 0.363∗∗∗ (0.10) – −0.184 (0.19) 1005

Model 3 β (RSE) −0.08 (0.08) – −0.363∗∗∗ (0.11) −0.547 (0.14) 1005

Model 4 β (RSE) – 0.08 (0.08) −0.277∗∗∗ (0.03) −0.46∗ (0.18) 1005

Note: Probit estimates displayed with robust standard errors clustered on the type of constitutional action. I have controlled for, but do not report, all of the variables in the Strike equation in Table 4.3. ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001.

cases. Because I wish importance to vary, the decision to include state policies and amparo suits is consistent with the literature. But is there empirical evidence that my measure has been validly constructed? A validly constructed ordinal measure, at the very least, should stack correctly. One way to test this is to rerun the Strike equation models with dummy variables that distinguish between categories. Table 4B.1 shows the results of these models. For ease of interpretation, I only include the estimates for the Importance dummies, but the remaining results do not change. As you read across Table 4B.1, note that I treat a different category of Importance as the base category for each of the four models. This allows me to test whether each category is different than every other category. The results suggest that the variable stacks reasonably well. Take Model 1, where the base category contains the most important policies. All of the dummies are positive, suggesting that the court was more likely to strike down each of the ﬁrst three categories than the last. Similarly Model 2, where the penultimate category (federal amparos) is now the base category, shows positive estimates for the ﬁrst two categories and a negative estimate for the highest category. Still, there are two results that suggest further analysis may be required. • The difference between the two federal case types is in the correct direction, but it is not statistically signiﬁcant at the 0.10 level.

118

The Politics of Constitutional Review in Mexico

• The relationship between the two state case types is in the wrong direction, but it is small and not signiﬁcantly different from zero. Although generally encouraging, this analysis suggests that alternative coding schemes may be in order. In the next section, I examine the robustness of the results to these alternative coding choices. Robustness analysis I have reestimated the full analysis with two different coding schemes for the Importance measure, both suggested by the results in Table 4B.2. • Importance2 (Combine State Policies) I ﬁrst recoded Importance by combining the two state policy categories, creating a three-category ordinal measure. This addresses the possibility that federal ofﬁcials do not distinguish between state policies, whatever constitutional action they are challenged under. • Importance3 (Federal vs. State Policies) I then created a binary variable, which is coded 1 for federal policies and 0 for state policies. This addresses the possibility that the only real difference is between federal and state policies. I report the results of models with these alternative measures in Tables 4B.2 and 4B.3 later. The full model speciﬁcations are identical to those in Table 4.3 of this chapter. To conserve space, I report only the estimates for Importance, Coverage, and Strike in both equations (the key theoretical variables), but I should note that the results for the control variables are robust to these changes. Table 4B.2 shows the results for the restricted data set, where I discard procedural dismissals. The ﬁrst model is the original model with the four-category ordinal measure. The second model uses the three-category ordinal measure (Importance2), and the third uses the binary measure (Importance3). As is clear from the table, the results are absolutely robust to the choice of Importance measure. Table 4B.3 shows the results for the full set of cases, where I include procedural dismissals. As above, the results here are similarly robust to the different measures of Importance. There are two differences between the results that I wish to highlight. If there is any substantive difference across the three models, it is that the Coverage result in the ﬁrst equation is less signiﬁcant with alternative measures. This provides greater support for my endogenous story. Second,

119

table 4B.2. Determinants of Strike and Press for Three Measures of Federal Policy Importance (Procedural Dismissals Excluded) Original Model β

RSE

Combined State Policies Model P

Strike Equation Importance Importance2 Importance3 Imp*Coverage Imp2*Coverage Imp3*Coverage Coverage

−0.120 – – −0.367 – – 1.12

0.02 – – 0.27 – – 0.81

<0.001 – – 0.178 – – 0.164

Press Equation Strike Coverage Strike*Coverage Importance Importance2 Importance3

1.88 2.01 −0.956 0.282 – –

0.33 0.21 0.37 0.07 – –

<0.001 <0.001 0.010 <0.001 – –

β – −0.178 – – −0.551

RSE – 0.009 – – 0.05

Federal vs. State Binary Model P – <0.001 – – <0.001

β

0.886

0.31

0.005

– – −0.164 – – −1.10 1.09

– 0.421 –

– 0.13 –

– <0.001 –

– – 0.374

RSE

P

– – 0.02 – – 0.51 0.50

– – <0.001 – – 0.029 0.030

– – 0.16

– – 0.018

Note: Bivariate results displayed with robust standard errors reported clustered on the type of constitutional action. The speciﬁcation for each model is identical to that in Table 4.3 of this chapter. Only the measure for federal policy importance changes. The ﬁrst column contains the original results. The second model uses a three-category measure of importance, where I combine state policies. The third model uses a dummy variable for federal vs. state policies. The results are robust to these coding choices.

120

table 4B.3. Determinants of Strike and Press for Three Measures of Federal Policy Importance (Procedural Dismissals Included) Original Model β

RSE

Combined State Policies Model∗ P

β

RSE

Federal vs. State Binary Model∗ P

β

RSE

P

Strike Equation Importance Importance2 Importance3 Imp*Coverage Imp2*Coverage Imp3*Coverage Coverage

−0.120 – – −0.367 – – 1.12

0.02 – – 0.27 – – 0.81

<0.001 – – 0.178 – – 0.164

– −0.254 – – −0.306 – 0.857

– 0.03 – – 0.34 – 0.58

– <0.001 – – 0.365 – 0.140

– – −0.270 – – −0.532 0.847

– – 0.05 – – 0.61 0.63

– – <0.001 – – 0.389 0.175

Press Equation Strike Coverage Strike*Coverage Importance Importance2

2.13 1.80 −1.02 0.325 –

0.14 0.29 0.17 0.05 –

<0.001 <0.001 <0.001 <0.001 –

0.964 2.37 −0.780 – 0.210

0.17 0.32 0.28 – 0.21

<0.001 <0.001 <0.001 – 0.319

0.952 2.46 −0.851 – –

0.18 0.32 0.31 – –

<0.001 <0.001 <0.001 – –

Note: Bivariate results displayed with robust standard errors reported clustered on the type of constitutional action. The speciﬁcation for each model is identical to that in Table 4.3 of this chapter. Only the measure for federal policy importance changes. The ﬁrst column contains the original results. The second model uses a three-category measure of importance, where I combine state policies. The third model uses a dummy variable for federal vs. state policies. The results are robust to these coding choices.

Decisions, Case Promotion, and Compliance

121

the reduction of categories produced a signiﬁcant degree of collinearity in the ﬁrst equations, and Stata was unable to estimate the correlation between the errors of the two equations properly. Accordingly, I was forced to estimate the system via two separate univariate probit models, which is equivalent to assuming ρ = 0 (Greene 1998). Although not ideal, the results are consistent across many of the models.

part iii RELATIONSHIPS BETWEEN TRANSPARENCY AND LEGITIMACY

The preceding chapters have addressed consequences of assuming that judicial power depends ultimately on the public. The immediate implications of this assumption are that high court judges have reason to concern themselves with judicial legitimacy and with the transparency of the decisions they render. I have suggested that judicial public relations should be understood as a means of ensuring transparency and promoting legitimacy, and as such, it is a means of developing power. In the last section of the book, I wish to expand the inquiry in three ways. First, transparency and legitimacy have been treated as if they were independent; however, the judicial legitimacy scholarship reviewed brieﬂy in Chapter 3 suggests that they may reinforce each other. That literature suggests a straightforward approach for developing judicial legitimacy: increase public exposure to courts while emphasizing procedure. As we know, high courts around the world have done exactly this over the past decade. They have courted media coverage and promoted impartial procedures by educating their publics on legal institutions and jurisprudence. Yet, we also know that courts have not necessarily sought to maximize transparency. Information is promoted selectively. If transparency promotes legitimacy, then this tactic may be mistaken. If more information is better, then providing less information is worse. But what if existing accounts of the link between transparency and legitimacy are incomplete or inaccurate? What if transparency does not always increase beliefs in judicial legitimacy? In this part of the book, I will consider how a general model of constitutional review, like that developed in the ﬁrst half of the book, might inform our understanding of the link between transparency and legitimacy. A number of key points

124

Relationships between Transparency and Legitimacy

emerge. If our interbranch conﬂict models are capturing something real about constitutional politics, then individuals should learn about judicial impartiality in a variety of ways from the substantive outcomes of cases. Importantly, some of the inferences people might draw do not suggest impartial judging. What this result suggests is that although transparency may be a necessary condition for the exercise of judicial power, too much transparency can undermine the other fundamental condition for power: judicial legitimacy. In short, there may be a very general tension between transparency and legitimacy. The basic logic, as I will develop in what follows, is that increased transparency can mean increased public exposure to political scenarios in which strategic judicial deference is required to protect a court’s institutional interests. For some observers, these kinds of decisions will not be consistent with an image of judicial impartiality, a key component of legitimacy. For this reason, constitutional judges might prefer that the media largely ignore these kinds of political conﬂicts. The problem, of course, is that by constructing a media coverage that is likely to render judicial decision making transparent in general, judges may also construct precisely the kind of coverage that will inform the public about decisions that are inconsistent with judicial impartiality. Second, to build this argument, it is necessary to expand the model developed in Chapter 2 in a way that allows us to consider how voters might learn about legitimacy if the interbranch dynamics of that model are roughly accurate. Empirically, it will be critical to leave the empirical context of Mexico behind at this stage. The reason is that important dynamics of the argument developed in this chapter involve a concept that varies cross-nationally. For this reason, the empirical tests require cross-national data. As it turns out, the most appropriate data set to test the argument does not include Mexico. Third, I will consider the implications of an argument linking transparency and legitimacy for a familiar theory about the construction of judicial power. Scholars commonly suggest that judicial prudence, the choice to strategically avoid conﬂict under unfavorable political conditions, offers a path toward developing a powerful high court (Carrubba 2009; Couso 2005; Ginsburg 2003; Shapiro 2005; Vanberg 2005). By deferring to the political branches in situations when challenging them would be dangerous, courts “live to ﬁght another day.” This strategy is especially effective when judges establish jurisprudential principles that have no immediate practical effect, but instead lay the groundwork for activism in the future. Marbury v. Madison is the paradigmatic example of this strategy, but there are many examples from around the world

Relationships between Transparency and Legitimacy

125

that ﬁt the pattern (see especially Couso 2005; Ginsburg 2003; Moustafa 2007). By managing conﬂicts in this way, courts induce a compliance norm, which ultimately ensures their power. If this argument is correct, then judges advance a long-term interest in judicial power by engaging in prudential decision making in the short run. I will consider the logic of prudence as a key to the development of judicial power, if we believe that judicial power derives from judicial legitimacy. I will ask the following question: might prudence undermine legitimacy under particular conditions and by so doing undermine power? The remainder of the book is divided as follows. In Chapter 5, I present a model of constitutional review in which voter beliefs in judicial legitimacy are endogenously determined. I ask how the information voters can observe in common press reports about cases might inﬂuence their beliefs about the impartiality of high courts. The model suggests that voters can learn a good deal about impartiality even if they know almost nothing about their court. More importantly, I ﬁnd that familiarity with high courts and the political environments those courts confront interactively affect beliefs in judicial impartiality. I test implications of this model in Chapter 6. I then consider both the tension between transparency and legitimacy and what that tension means for theories of judicial power construction that invoke the logic of strategic deference. In Chapter 7, I discuss what the book’s ﬁndings suggest about the connection between democratization and the rule of law, a connection that is immediately raised under a public enforcement mechanism. The experiences of the Mexican Supreme Court during the period of democratic transition, coupled with the results of the cross-national design, highlight the opportunities for and limits of developing powerful courts in democratizing states.

5 Constitutional Review and the Development of Judicial Legitimacy

If literatures on legitimacy in law and courts are correct, then it is likely that transparency reinforces beliefs in judicial legitimacy. The received wisdom in that literature is that legitimacy ultimately derives from perceptions of procedural fairness (e.g., Tyler 1990; 2009), and that these perceptions are likely to emerge with repeated exposure to the technical nature of judging. Essentially, the information people receive about judging (e.g., technical language and formal procedure) is biased toward creating positive affect, such that increasing exposure to courts increases support for courts on average (Caldeira 1986, 1222; Gibson and Caldeira 2009). The positivity bias hypothesis has received considerable empirical support, especially in studies conducted in the United States (Caldeira 1986; Casey 1974; Murphy and Tanenhaus 1968). In a number of contexts, using a variety of measures, scholars have found that greater awareness with the U.S. Supreme Court is associated with higher levels of diffuse public support (i.e., legitimacy). Comparative research has produced similar ﬁndings. Table 5.1 reviews key results from Gibson, Caldeira, and Baird’s (1998, 353, hereafter GCB) study of high court diffuse support in the United States and seventeen European countries, the only systematic comparative analysis of the relationship between awareness and high court legitimacy. GCB provide two empirical tests of the argument at the individual level. For each country, they regress individual beliefs in the legitimacy of high courts on the self-reported level of familiarity with those courts. As Table 5.1 summarizes, they ﬁnd a positive relationship in 15 out of 18 states. Second, they reestimate the models adding a single control variable that measures the degree to which a respondent is 127

Relationships between Transparency and Legitimacy

128

table 5.1. The Relationship between Awareness and Legitimacy in Eighteen Countries One-Variable Model

Two-Variable Model (Controlling for Speciﬁc Support)

Positive and Signiﬁcant Estimate

Insigniﬁcant Estimate

Positive and Signiﬁcant Estimate

Insigniﬁcant Estimate

Belgium Denmark Francea Germany (East) Germany (West) Greece Great Britain Hungary Ireland Italy Netherlands Poland Portugal Spaina United States

Bulgariab Luxembourgb Russia

Belgium Francea Great Britain Italy Netherlands Poland Spaina United States

Bulgariab Denmark Germany (East) Germany (West) Greece Hungary Ireland Luxembourgb Portugal Russia

Note: Summary of Gibson, Caldeira, and Baird’s (1998) results. a Models run on surveys conducted in both 1993 and 1995. b Coefﬁcient estimate is negative, although not signiﬁcant.

favorably disposed toward her court’s output, his or her “speciﬁc support.” The rationale for the speciﬁcation is that there is reason to believe that a person who enjoys short-term satisfaction with the job of her high court is likely to pay closer attention to the court; and speciﬁc support is a key cause of diffuse support (Caldeira and Gibson 1992). The second half of Table 5.1 shows that even with speciﬁc support controlled, the relationship between awareness and diffuse support is positive and signiﬁcant in eight cases. The evidence for the positivity bias account is persuasive in a number of ways, and the argument suggests important implications for the development of judicial legitimacy and, by implication, judicial power. On this account, what matters is procedural information. Substance is largely irrelevant. If this is correct, the key implication for high court judges is quite positive. It suggests that even a court that faces signiﬁcant

Constitutional Review and the Development of Judicial Legitimacy 129 constraints on its power can produce beliefs in its legitimacy by exposing individuals to its work. In terms of the argument of this book, high court judges might simultaneously advance the transparency and legitimacy conditions through public relations strategies that maximize public information about the judiciary. Insofar as exposure to the activities of courts is biased toward perceptions of procedural fairness, judges do well to maximize coverage, no matter what substantive information might emerge. Like the rule of law narrative of the Mexican Supreme Court in the wake of the Zedillo decision, this story is highly appealing, because it suggests a clear path toward building judicial power that can be applied to any court in the world. But just like the Zedillo case, it is worth considering Table 5.1 again, this time from a slightly different perspective. Once speciﬁc support is controlled, the relationship between awareness and legitimacy is insigniﬁcant in ten of the eighteen states. Because the messages and symbols that reinforce judicial impartiality are essentially constant across the world’s high courts (Stone Sweet 2000), it is not immediately clear how a procedural fairness theory can explain these results. This is not to suggest that the positivity bias story is wrong. Every research program collects empirical anomalies and null results. Nevertheless, within a procedurally oriented legitimacy account, the results are puzzling. The literature suggests two possible explanations. The ﬁrst is that, although awareness may increase beliefs in judicial impartiality, beliefs in impartiality have nothing to do with judicial legitimacy (Mondak 1993). Although possible, this argument does not sit well with the results in either column of Table 5.1, where there are a number of states in which awareness is associated with diffuse support. Second, although the literature on judicial legitimacy has evolved toward procedural explanations over the last decade, including work by Gibson and Caldeira (2008), GCB offer a substantive account in the original article, summarized in Figure 5.1. In particular, they suggest that increasing awareness likely

Awareness

Specific support

Diffuse support

ﬁgure 5.1. An Indirect Effect of Awareness on Diffuse Support

130

Relationships between Transparency and Legitimacy

inﬂates diffuse public support for the court; however, the effect of awareness is largely indirect and runs through speciﬁc support. The idea is that awareness increases speciﬁc support for the court through the positivity bias, and then speciﬁc support is translated into diffuse support. Again, this argument suggests a substantive mechanism for legitimacy, which is inconsistent with current thinking in the legitimacy literature (e.g., Tyler 2009), but it is a plausible explanation. If the effect of awareness on diffuse support is mostly indirect, then once we control for speciﬁc support, the estimated impact of awareness will likely attenuate. Thus, on the GCB account, what appears to be a puzzle is not really puzzling after all. Still, this explanation raises another question about the results in Table 5.1, which is not resolved by an appeal to indirect effects. Why do the respective magnitudes of the indirect and direct effects differ across states? Without the entire GCB data set, it is impossible to test precisely whether there are differences in magnitude across states, yet the published evidence is suggestive. If the model in Figure 5.1 is correct, we can partition the total effect of awareness on judicial legitimacy into its direct and indirect effects. Although we cannot explicitly estimate the system of equations this model implies without the full data set, we can imagine what those effects would look like if we could, at least in relative terms. When we see a statistically signiﬁcant awareness estimate, we might infer that the direct effect of awareness is stronger in that state than in a state where the awareness estimate is no different from zero. Likewise, we might conclude that the indirect effect of awareness is stronger in states where the awareness estimate is insigniﬁcant relative to those in which the effect is signiﬁcant. Clearly, we cannot draw these inferences precisely, but the results summarized in Table 5.1 are consistent with varying magnitudes across states. Importantly, the argument summarized in Figure 5.1 does not suggest that we should observe different direct and indirect effects across states. For example, why would the direct portion of the awareness effect be stronger in the United States than it seems to be in Russia? For this reason, we still confront a puzzle. In the remainder of this chapter, I construct a model of judicial legitimacy beliefs, which is substantive in nature. The goal of the model will be to consider what a voter might learn about the impartiality of her high court by observing case outcomes that emerge in a strategic model of constitutional review, much like the one summarized in Chapter 2. I will argue that people are more likely to perceive a court as legitimate if they perceive its judges as impartial; however, it is not always the case that increasing awareness induces beliefs in judicial impartiality. Rather,

Constitutional Review and the Development of Judicial Legitimacy 131 Information

Awareness

Diffuse support

ﬁgure 5.2. A Conditional Effect of Awareness on Diffuse Support

the kind of information to which individuals are exposed will affect their sense of judicial impartiality. The relationship between awareness and impartiality should be conditioned by a person’s information. Figure 5.2 summarizes this alternative. This argument suggests an entirely different implication for the relationship between transparency and legitimacy. Transparency should not always increase legitimacy. In some cases, high courts might advance legitimacy best if voters have less information about their activities. Consequently, a public relations strategy that manages transparency or that can inﬂuence the coverage it receives may be preferred to one that maximizes coverage. Consider the reaction to the Venezuela’s Supreme Court (Tribunal Supremo de Justicia) 2001 decision resolving a complaint against President Hugo Chavez that was ﬁled by Elias Santana, the director of Queremos Elegir, a civic association. Chavez had heavily criticized Santana on his program Alo Presidente, and Santana sought an injunction requiring equal time to respond. The court dismissed the complaint, suggesting that Santana had his own media outlets and could respond through those avenues. In resolving the case, the court developed a broader principle, holding that the constitution’s guarantee of equal media access does not apply to members of the media or to those individuals with close connections to media.1 This decision was severely criticized within Venezuela among opponents of Chavez and internationally among advocates of press freedom. Associated Press coverage of the event suggested, “Some Venezuelans consider the ruling a sign that President Hugo Chavez is cracking down on a media he often accuses of conspiring to destabilize his government. The Supreme Court magistrates were hand-picked by the government-controlled congress.”

1 “Venezuela: Journalists protest against Supreme Court ruling,” BBC Worldwide Moni-

toring, June 28, 2001.

132

Relationships between Transparency and Legitimacy

During the course of these events, the Supreme Court took out advertisements in national newspapers arguing that it was not “submissive” to Chavez.2 This is precisely the kind of case that a court might like to resolve without media fanfare, or at least the kind of case for which good media relations are essential. It is also precisely the type of case likely to undermine beliefs in impartiality, largely because the court made this decision in the context of considerable pressure on the judiciary and following the appointment of an entirely new high court. Formal legal procedures aside, political context can be critical, and as I discuss below, awareness of context should inform the way we learn about judicial legitimacy.

an extended model of constitutional review The model addresses two questions. Can a person learn anything about judicial impartiality by observing a constitutional court’s interaction with the government? If so, how does awareness affect the inferences one draws about impartiality? The answers to these questions unfold in a series of steps. I begin by deﬁning judicial impartiality. I then describe a game model and characterize its equilibria, highlighting how an individual learns about impartiality in each equilibrium. I then deﬁne awareness and show how awareness affects the learning process. Finally, I discuss how political context alters the players’ equilibrium behavior, thus linking the political environment to the process by which individuals learn about impartiality. Judicial impartiality Spaeth et al. (1972, 119) deﬁne judicial impartiality as a method of decision making under which justice is “situationally determined; that is, the personal attributes of the litigants [are] never relevant to the case.” Following this conceptualization, I deﬁne a constitutional court as impartial if it renders decisions according to its sincere evaluation of the policy’s constitutionality. If presented with identical constitutional questions, an impartial court’s resolutions to those questions must be invariant to changing litigant identities, partisan or otherwise, or changing extrajudicial pressures. To be clear, the concept places no constraints on the etiology of the court’s sincere evaluation (i.e., its constitutional 2 “Venezuela’s state lawyer resigns in protest over press freedoms ruling,” Associated Press

Worldstream, July 30, 2001.

Constitutional Review and the Development of Judicial Legitimacy 133 philosophy), which may derive from a recognized legal interpretive theory (e.g., originalism or purposivism) or from raw political ideology, as long as the court systematically resolves similar cases by consistently applying its philosophy. This concept is more general than a common alternative, in which impartial judges put aside their “values” and render purely “legal” decisions (Gibson, Caldeira, and Baird 1998, 345; Scheb and Lyons 2000, 929), and thus it is appropriate to consider this choice. I adopt the less restrictive concept of impartiality for two reasons. First, Casey (1974, 394) ﬁnds no signiﬁcant support for the notion that people believe in fundamental legal truths to which judges might appeal, and I have yet to ﬁnd convincing empirical results that unambiguously identify a public belief under which judicial decisions are produced by appealing to a set of knowable, legal truths.3 Second, if a reader chooses to do so, the model can be interpreted as if there were such legal truths and judicial impartiality means adhering to them. Under a purely legal interpretation of the model, the court will learn the legal “truth,” which tells it how to evaluate the policy sincerely.4 Sequence of play The game is played between a constitutional court, a president, and a voter. I conceptualize the voter as the polity’s median, though this need not be the case. All that needs to be true is that the executive is inﬂuenced in some way by the voter’s support for the policy and the court. Prior to the interaction analyzed here, imagine that the executive implements a suite of policies desired by the voter; however, only a proportion β ∈ (0, 1) of these polices are personally salient to the president. The remaining policies are enacted merely to satisfy a campaign promise. Subsequently, a member of society challenges the constitutionality of one policy in the suite. The court is then called upon to review the constitutionality of the policy. If the court vetoes the policy, the president chooses to accept the resolution or defy the court, as conceptualized in the model in Chapter 2. 3 Although Scheb and Lyons (2000, 934) claim to provide evidence for a belief among

many Americans that Supreme Court decisions are independent of political or ideological concerns, their results are consistent with a belief in which Supreme Court decisions are strongly affected by both ideology and legal models of interpretation. 4 Even if judicial impartiality is interpreted in this way, it is important to note that the model predicts the same empirical implications as it does under the broader deﬁnition I adopt.

134

Relationships between Transparency and Legitimacy

The voter moves last, after observing one of three events. If the court upholds the policy, the voter may raise his speciﬁc support for the policy or not. Symmetrically, if the president accepts the court’s veto, the voter may lower his speciﬁc support for the policy or not. Finally, if the president deﬁes the court, the voter may punish the president’s recalcitrance. The challenge for the voter will be to use both what he believes about the court and the president and what he observes during the game to draw inferences about the court’s impartiality. These inferences will determine how he should respond to the president. What do the players know? The voter confronts three sources of uncertainty. He is initially unsure about the salience of the policy to the president. His belief that the policy is salient is β. He is also uncertain about how the court sincerely evaluates the policy. There are two possibilities. With probability α ∈ (0, 1), the court believes that the policy is consistent with constitutional terms; it also believes that the policy is unconstitutional with probability 1 − α. The third piece of uncertainty concerns the nature of the court, the critical piece of information the voter is trying to discover. Let us imagine that there are three court types, although I will consider an alternative type below. With probability π ∈ (0, 1), the court is “impartial.” Following the deﬁnition above, an impartial court is one whose decisions are fully determined by its sincere evaluation of the record. Once the court reads the record and develops its sincere opinion, there is nothing left to do but announce it. Unfortunately for the voter, there are two ways in which the court might issue decisions that do not reﬂect its sincere view of the record. With probability π ∈ (0, 1), the court is “strategic.” A strategic court type may prefer to veto a policy associated with the president; however, it is reluctant to challenge the president if the institutional consequences are severe enough. That is, strategic courts are willing to veto policies under particular conditions, conditions derived below. Finally, with probability 1−π −π , the court is “partisan,” in the sense that it will never strike down a policy associated with this president no matter how egregious the behavior under review and no matter how it might evaluate a similar policy associated with another president. Although the strategic court type is not a presidential rubber stamp like the partisan court, it may behave as if it were when the stakes are high enough. Similarly, the strategic court may behave as if it were impartial when the stakes are low

Constitutional Review and the Development of Judicial Legitimacy 135 enough. The question is whether the voter can distinguish between these types given his prior beliefs and the behavior he observes. Naturally, I assume that the court learns its type and how it sincerely evaluates the policy. I will assume that it remains uncertain about the true salience of the policy to the president. For simplicity, I assume that the president learns the state of the world perfectly, though this is immaterial to the key results.5

Preferences Adding an uncertain voter requires some additional parameters, which I did not use in Chapter 2; however, the preferences described here are completely reﬂective of those in the game developed earlier. The president cares about policy outcomes and the public’s reaction to potential deﬁance. The court cares about policy outcomes and the president’s reaction to its decision. Speciﬁcally, the president gains one unit of utility for continuing to implement the status quo policy, if she perceives it to be salient. This is true whether the court upholds the policy or if she deﬁes a veto. If the voter raises (lowers) his speciﬁc support for a salient policy, the president gains (loses) r > 0. Finally, if the voter punishes the president’s deﬁance, the president pays b + r if the policy is salient and only b if it is not (b > 0). As before, b reﬂects the cost of deﬁance incurred by the executive, while r reﬂects the policy-related gains and losses from the voter’s speciﬁc support for the policy.6 The court gains one unit of utility under the following circumstances: (1) if it is impartial and issues a resolution that is consistent with its sincere evaluation, (2) if it is strategic and issues a resolution that is consistent with its sincere evaluation, assuming the president accepts a veto, and (3) if it is partisan and upholds the president’s policy. This means that if the strategic court strikes down a policy it believes to be unconstitutional and the president ignores the resolution, the strategic court gains 5 This follows directly from the president’s preferences as deﬁned in the text that follows.

The value the president places on the policy and the cost she pays for ignoring the court are independent of the court’s type and how the court sincerely evaluates the policy. Thus, these issues do not matter to the president, and they drop out of any expected utility calculus in versions of the model where I assume the president shares the voter’s uncertainty. 6 The U.S. Supreme Court’s ability to inﬂuence support in this way is analyzed by Dahl (1957), Franklin and Kosaki (1989), Gibson (1989), and Johnson and Martin (1998).

136

Relationships between Transparency and Legitimacy

no policy-related utility.7 Strategic and partisan courts absorb a cost c > 0 if the president deﬁes a judicial veto. The voter wishes to raise (lower) his support for the policy or punish the president for deﬁance only if he perceives the court to be impartial. In particular, the voter gains a unit of utility if he (1) raises (lowers) his opinion of a policy upheld (vetoed) by an impartial court, (2) punishes deﬁance of an impartial court, and (3) does nothing following any other type of decision. Because the voter supports all policies in the president’s policy suite (which is why the president produced them in the ﬁrst place), the payoff structure assumes that the voter is willing to give up a policy he supports in order to ensure respect for an impartial court.8

Results I discuss the results informally, framed around Figures 5.3 and 5.4. The formal analysis is in the appendix. There are three types of perfect Bayesian equilibria, which we can divide into three cases. In the ﬁrst two cases, the cost of a public backlash is small enough that the president will defy a judicial veto of a salient policy (b < 1). She will, however, accept a veto of a nonsalient policy. In the third case, the cost of a public backlash 7 This clariﬁes an important distinction between the impartial and strategic courts. Even if

we assume that the court’s sincere evaluation of the policy is derived from values and not the law, the impartial court is not an attitudinal court in the sense of Segal and Spaeth (1993, 64–73). Recognize that the strategic court’s preferences vary over the full range of outcomes in ways required by the attitudinal model, while the impartial court merely cares about getting the legal question right. The attitudinal model applies to the U.S. Supreme Court not because the justices do not care about policy outcomes or having their jurisdiction stripped – it applies because attitudinalists assume that the justices do not have to worry about enforcement or serious political pressure. Accordingly, we can interpret the model as capturing an “attitudinal” court if the court evaluates the costs of deﬁance as terriﬁcally low; however, this is just to say that while an “attitudinal” court may be observed in equilibrium, it is not a court type per se. 8 Theoretically, this assumption captures the essence of the judicial legitimacy concept, which suggests that people may be willing to defend decisions with which they disagree in order to preserve a legitimate institutional arrangement. Practically, explicitly modeling the policy implications for the voter of the president accepting a judicial veto (which would naturally induce a policy loss) would not add much. In fact, any case in which the voter would be unwilling to punish presidential deﬁance, because he values the policy too much, would be identical to cases in which the president’s cost of deﬁance is insufﬁcient to induce her to accept the resolution (i.e., for sufﬁciently low b). These cases, and the beliefs that the voter derives from the interaction, are described in Cases 1 and 2 below. Alternatively, because it is the voter that imposes b, we might also simply interpret the model such that b varies with the importance of the policy to the voter.

Constitutional Review and the Development of Judicial Legitimacy 137 is high enough that the executive will accept all vetoes, no matter the salience of the policy.

Case 1: Low salience and low public backlash cost

Probability court is impartial (p)

This case is depicted in the left, light-grey region of Figure 5.3. The horizontal axis in the ﬁgure represents the probability that the policy is salient to the executive (β); the vertical axis reﬂects the voter’s prior belief about the court’s impartiality (π). Given the executive’s expected course of action in the ﬁrst two cases, the court expects to be deﬁed if the policy turns out to be salient. Still, the strategic court behaves as if it were impartial because it is unlikely that the policy is salient. As the court’s own cost for being deﬁed (c) decreases, the light region expands to the right across Figure 5.3. Indeed, as the court comes to value making a sincere decision more and more over the potential consequences of noncompliance, the light region spans across the space, reﬂecting the commonsense idea that we should expect sincere constitutional review when the costs of doing so are perceived to be negligible. The voter’s equilibrium strategy is listed in Figure 5.3. As we move from top to bottom, it is clear that the voter is less and less likely to be inﬂuenced by the court’s opinion. In the top-left region, where the voter believes strongly in the court’s impartiality, he increases (decreases) his support for the policy if the court upholds (vetoes) it. As the cut-point

1–p 1 1+a 1 + p9a + p9 1+a

Voter raises, lowers, and punishes

Voter lowers and punishes

Voter raises, lowers, and punishes

Voter lowers and punishes

p9 Voter does nothing

Case 1: Strategic and impartial Case 2: Strategic and 1 courts pool 1 + c impartial courts separate Probability policy is salient (b)

ﬁgure 5.3. Equilibrium Behavior with Low Public Backlash Cost

1

138

Relationships between Transparency and Legitimacy

identifying the region suggests, this subcase is less and less likely as a voter is more likely to believe that the court is truly strategic. As the prior probability that the court is impartial decreases, the voter’s behavior is less sensitive to the court’s decision. For very low beliefs in judicial impartiality (in the bottom left of the ﬁgure), the voter simply does not respond to any combination of judicial and executive behavior. This is because it is too likely that the court is either partisan or strategic. This case highlights important pieces of the learning process identiﬁed by the model. In particular, a judicial veto is always informative for the voter, and it will never decrease beliefs in the court’s impartiality. This is because the voter knows that a partisan court will never veto a presidential policy and that no court type will veto a policy that it sincerely believes to be constitutional. Doing otherwise does not advance the court’s interests and only risks inducing deﬁance. Accordingly, if the voter observes the court veto a policy, he knows that either a strategic or impartial court type struck down a policy it sincerely believed was in violation of the constitution. Although the updated values of π and π do not change relative to one another, the veto eliminates the possibility that the court is purely partisan. As long as the voter believed it was more likely that the court was impartial than strategic (π > π ) prior to the interaction, the voter will lower his opinion of the policy and punish deﬁance. Although a veto is never predicted to decrease beliefs in judicial impartiality, every time a court upholds a presidential policy, beliefs in its impartiality are reduced.9 There is evidence of this result in the empirical literature. Baird and Gangl (2006) ﬁnd that perceptions of fairness associated with Supreme Court decisions are negatively correlated with whether the court upheld an antiterrorism statute perceived to be salient to the sitting administration, controlling for other relevant factors (e.g., ideology, perceptions of procedural fairness, and gender). As I shortly demonstrate, awareness affects how far beliefs in impartiality fall when the voter observes the court uphold a policy. Case 2: High salience and low constituent cost for deﬁance Case 2 is represented in the dark-gray region of Figure 5.3. The cost of public punishment continues to be insufﬁcient to induce acceptance of salient policy vetoes; however, it is now fairly likely that the policy is 9 See appendix for proof.

Constitutional Review and the Development of Judicial Legitimacy 139 salient. Accordingly, the strategic court will separate from the impartial court and behave as if it were partisan. The case is divided into two subcases. In the upper region, the voter is inﬂuenced by all types of decisions, because the probability that the court is impartial is sufﬁciently high. As the cut-point suggests, this region decreases as the probability that the policy is constitutional decreases. The reason is that as that probability decreases, it becomes decreasingly likely that the court upholding the policy is truly impartial. In the lower region, the results remind us that a judicial veto is extremely informative in this case. This is because only a truly impartial court would challenge the president over such a salient policy. This result suggests that politically charged cases offer ideal opportunities to learn about judicial impartiality. Unfortunately, voters will be able to take advantage of this only if the court is truly impartial. If the court happens to be strategic, voters will never observe such a highly informative decision in equilibrium, because such courts will defer to the president over such policies.10

Case 3: High constituent cost for deﬁance Figure 5.4 summarizes the results in this case. In this case, the cost of a public backlash is sufﬁciently large to induce the president to accept a veto of a salient policy if she anticipates that the voter will punish. Insofar as the president is expected to accept the veto, the strategic court type pools with the impartial type no matter how salient the policy is believed to be. Consequently, the entire region is light grey, reﬂecting the fact that the strategic court behaves as if it were impartial, no matter its beliefs about policy salience. In this case, the voter does not learn about judicial impartiality by observing deﬁance, precisely because he will not observe deﬁance in equilibrium. Nevertheless, the voter will learn about judicial impartiality from every judicial decision in this equilibrium. In fact, the learning process is identical to that described in Case 1. Importantly, the cut-points that deﬁne distinct voter behaviors in equilibrium are the same as they were in Case 1, precisely because the behavior of the court is 10 The sole difference between equilibria in the two regions of Case 2 is that in the upper-

right region, where beliefs in judicial impartiality are relatively high, the voter will be inﬂuenced by a decision upholding the policy. As these beliefs decrease, he will not. Again, because only an impartial court will strike down a policy in the dark region of Figure 5.2A, the voter learns the court’s true type with certainty and rationally punishes deﬁance.

Relationships between Transparency and Legitimacy

Probability court is impartial (p)

140 1–p

Voter raises, lowers and punishes 1 + p9a + p9 1+a

Voter lowers and punishes p9

Voter punishes

1 1+c Case 3: Strategic and impartial courts pool

1

Probability policy is salient (b)

ﬁgure 5.4. Equilibrium Behavior with High Public Backlash Cost

identical to that in Case 1.11 The key now is to consider how awareness inﬂuences this process.

11 For extremely low prior beliefs in judicial impartiality (π < π ), this result depends

on assuming that, if the voter observed deﬁance (which he shouldn’t given the strategy proﬁle), his updated beliefs over judicial impartiality would be greater than his updated beliefs in the court being strategic. The solution concept places no restriction on these “off-path” beliefs (Osborne and Rubenstein 1994, 233); however, it is fair to wonder about the consequences of this assumption. What would happen to the equilibria if we assumed that the voter does not update beliefs if he observes out-of-equilibrium behavior? For the ﬁrst two regions of the ﬁgure, the results are fully robust to this alternative, because the voter will punish a veto, because he knows for sure that the court is not partisan and π > π in those regions. That said, there would be no pure strategy PBE in the lower-right region. The reason is that the voter’s off-path behavior would not involve punishing noncompliance. Consequently, the president would not accept vetoes of salient policies. For sufﬁciently high probabilities of salience (in the lower-right region), the strategic and impartial courts would separate given the president’s behavior, and in such cases, the voter would observe deﬁance of only impartial courts. Importantly, his beliefs would be deﬁned via Bayes rule in this case, and he would rationally punish. But because b is sufﬁciently large by assumption, we know that the president would not defy the court if the voter will punish, which brings us back to where we began. For the lower-left region, this assumption would produce an equilibrium like that described in the lower-left region of Figure 5.3. Essentially, the president would defy vetoes of salient policies, because he would know that the voter would not punish deﬁance if the strategic and impartial courts pool. They will pool because the probability that the policy is salient is sufﬁciently small in the lower-left region. In this situation, deﬁance is again on the equilibrium path, and the voter will rationally fail to punish. In any event, because the key claims I make about how awareness affects the learning process ultimately depend on learning that takes place at information sets that are reached in equilibrium, the assumption about off-path beliefs does not meaningfully affect the analysis.

Constitutional Review and the Development of Judicial Legitimacy 141

modeling awareness As originally formulated, the judicial awareness concept captured a person’s court-speciﬁc knowledge. The behavioral literature offers a number of measures of awareness, including respondent self-evaluations and exposure to media coverage about a court.12 Better measures of awareness code the ability of respondents to identify the U.S. Supreme Court’s constitutional role correctly (Murphy and Tanenhaus 1968, 365) or, better still, the ability to recall particular decisions correctly. Conﬁrming the utility of this approach, Casey (1974, 405) found that judicial knowledge is cumulative. People who could recall obscure decisions were likely to recall landmark decisions, and people with no recollection of landmark decisions were highly unlikely to recall obscure decisions. In order to tap into the conception of awareness as a form of case-speciﬁc knowledge, I model it through the voter’s belief in how the court will sincerely evaluate the policy: α. Let αi be any individual i’s belief that the court sincerely believes the policy is constitutional. Deﬁnition. Individual i is considered more aware than individual j, i = j, if and only if ai ≥ aj when the court truly believes the policy is constitutional and ai ≤ aj when the court truly believes the policy is unconstitutional.

In other words, awareness increases the accuracy of voter beliefs about the way the court sincerely evaluates the policy. Note that this deﬁnition places no restriction on the absolute accuracy of any two voters’ beliefs. The more-aware individual, i, could believe that the court is sure to evaluate the policy as constitutional when the court sees the policy as unconstitutional. It is merely the case that whatever are i’s beliefs, they are at least as accurate as j’s.13 How does awareness inﬂuence voter inferences about the impartiality of the court?14 Answering this question requires asking how the voter 12 For examples of these measures in order of textual reference, see Scheb and Lyons (2000,

934) and Caldeira (1986, 1216). 13 The awareness deﬁnition does rule out the possibility that less-aware voters might acci-

dentally have more accurate beliefs than more-aware voters. Ultimately, whether or not awareness increases the accuracy with which people can predict sincere judicial preferences is an empirical question, one I have yet to see fully analyzed. Still, if this conception of awareness is not suitable, the underlying result that it reﬂects remains. Prior beliefs about the court’s preferences are going to inﬂuence the way people learn about impartiality in ways that are conditioned by the political environment. 14 If the court is truly “partisan” or “impartial,” awareness will continue to increase to the voter’s sensitivity to the type, but it is considerably harder to learn, and more theoretically interesting, if the court is truly strategic.

142

Relationships between Transparency and Legitimacy

evaluates the court’s impartiality whenever he has to make a choice, and there are three events that the voter could have observed. The court could have upheld the policy, the president could have accepted a veto, or the president could have deﬁed a veto. Of course, there are many histories (or prior paths of play) that are consistent with these events. For example, in the aftermath of the 2007 purge of the Pakistani Supreme Court, the newly seated court rejected several appeals to Pervez Musharraf’s presidential reelection. This decision could have been made by a truly impartial court in review of a nonsalient policy that it believed to be constitutional, or it could have been made by a strategic court in review of a highly salient policy that it believed to be unconstitutional. In principle, there are many other possibilities, as well. The challenge for the voter is to use her prior beliefs and equilibrium expectations to infer the correct history. How might a voter distinguish between these very different histories? Let Iuphold represent the information set after the voter has observed the court uphold the policy. From the voter’s perspective, this information set contains all of the possible histories that could have preceded a world in which the court upholds the policy, within a particular equilibrium. Let Iaccept represent the information set after the voter observes the president accept a judicial veto. It contains all possible histories that could have preceded a world in which the court strikes down a policy and the veto is accepted. Finally, let Idefy represent the information set after the voter observes the president defy a veto. Unsurprisingly, the voter’s beliefs in the court’s impartiality depend on whether the strategic court behaves in equilibrium as if it were truly impartial or as if it were partisan. There are four possible beliefs to consider. At information sets Iaccept and Idefy , the voter’s beliefs are as follows: Pr(Impartial|Iaccept or Idefy and strategic court behaves as if it were partisan) = 1 Pr(Impartial|Iaccept or Idefy and strategic court behaves as π if it were impartial) = (π + π )

(1)

(2)

At Iuphold , his beliefs are Pr(Impartial|Iuphold and strategic court behaves as πα if it were partisan) = 1 − π(1 − α)

(3)

Constitutional Review and the Development of Judicial Legitimacy 143 Pr(Impartial|Iuphold and strategic court behaves as πα if it were impartial) = 1 − (π + π )(1 − α)

(4)

These results suggest a number of propositions about the relationship between awareness and beliefs in judicial impartiality. Proposition 1. The inferences voters draw about judicial impartiality are independent of judicial awareness when the court vetoes the policy.

In any equilibrium, if the court strikes down a presidential policy, prior beliefs over how the court sincerely evaluates the policy, α, do not inﬂuence the voter’s inferences. This is immediately apparent in Expressions 1 and 2, where α is conspicuously absent. A veto teaches the voter precisely what the court sincerely believed about the policy. This is because no court type will veto a policy it believes to be constitutional in equilibrium. This result implies that even the very unaware can learn a good deal about impartiality from the outcome of a case. As is clear from comparing Expressions 1 and 2 with Expressions 3 and 4, if awareness matters, it only matters when the court upholds the policy. Proposition 2a. When a court upholds a policy that it sincerely believes to be constitutional, beliefs in judicial impartiality among more-aware voters are reduced no more than those of less-aware voters. If there is a difference, more-aware voters will reduce their beliefs in impartiality less. Proposition 2b. When a court upholds a policy that it sincerely believes to be unconstitutional, beliefs in judicial impartiality among more-aware voters are reduced by at least as much as those of less-aware voters. If there is a difference, more-aware voters will reduce their beliefs in impartiality more.

In other words, there are conditions under which increasing awareness should increase judicial impartiality, yet there are also conditions under which increasing awareness should decrease beliefs in impartiality. Figure 5.5, which displays the relationship between awareness (α) and updated beliefs in the judicial impartiality, clariﬁes these propositions.15 Consider two voters, i and j. Assume that both share identical beliefs about the court’s true type. Let’s say π = π = 0.33. The only difference between i and j is their degree of awareness of the court’s jurisprudential history. Let i be more aware than j. Now consider two policies, x and y, 15 This is the graph of Expression 4, although the graph of Expression 3 makes precisely

the same point.

Relationships between Transparency and Legitimacy

144

Updated belief court is impartial (p*)

0.6

0.5

0.4

Voter i (Policy x)

p

Voter i (Policy y)

Voter j (Policies x,y)

0.3

0.2

0.1

0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Prior voter belief policy is constitutional (a)

ﬁgure 5.5. Updated Beliefs in Impartiality by Awareness

and assume that the court sincerely believes that policy x is unconstitutional while policy y is constitutional. Without loss of generality, assume that j’s prior beliefs about the court’s sincere evaluation of the policies y are identical: αjx = αj = 0.5. Similarly, assume that i’s beliefs about the y

policies’ constitutionality are αix = 0.2 and αi = 0.8. The arrows in Figure 5.5 represent how voters i and j update their beliefs after observing the court uphold policies x and y. The difference between the inferences that i and j draw clearly depend on i’s awareness advantage; however, awareness is not universally translated into stronger beliefs in impartiality. When voter i observes the court uphold policy y, which the court sincerely believes is constitutional, she reduces her belief y in the court’s impartiality much less than voter j. If we let αj approach 1, such a resolution is precisely consistent with voter i’s expectations, and there is effectively no change in his beliefs. In contrast, when voter i observes the court uphold policy x, which the court sincerely believes to be unconstitutional, her updated beliefs are reduced by a far greater margin than voter j’s. For voter i, the court’s validation of policy x comes as a surprise. In such a context, i is more likely than j to infer that the court has either avoided conﬂict with the president by strategically misrepresenting its preferences or, worse, that the court is a mere partisan tool. It is voter

Constitutional Review and the Development of Judicial Legitimacy 145 i’s superior familiarity with the court’s jurisprudential philosophy that translates such a decision into a signiﬁcantly reduced belief in the court’s impartiality. This result suggests that the positivity bias account of the relationship between awareness and legitimacy may be missing a crucial conditional relationship, but the analysis is not yet complete. This is because there is no reason to believe that voters will ever observe the court uphold a policy it sincerely believes to be unconstitutional in equilibrium. In order to address this possibility, we must consider whether or not there is a pattern to learning under particular types of equilibria. Consider any equilibrium in the light regions of Figures 5.3 and 5.4, where the strategic court behaves as if it were the impartial court. In all of these regions, the strategic court is free to resolve constitutional questions consistent with its sincere evaluation of the policy’s constitutionality. As such, its behavior is indistinguishable from the impartial court. In such a world, the voter will not observe the court upholding a policy it sincerely believes to be unconstitutional. Consequently, more-aware voters will be more likely to believe in judicial impartiality than less-aware voters. Although they may not be able to fully distinguish a strategic court from an impartial court, more-aware voters will reduce their beliefs in judicial legitimacy less than less-aware voters when they observe policies being upheld, precisely because there will be a consistency between their expectations and judicial outcomes. In a sense, this result reinforces the logic of the positivity bias account. Perhaps the reason why the awareness effect is so strong in worlds characterized by the light regions in Figure 5.3 is that there is a match of sorts between the symbols and rhetoric of judicial analysis and judicial behavior that is in fact sincere. As Mexican Supreme Court Minister Silva Meza suggested, it is easier to believe in judicial impartiality when judges tell you that they are impartial and then actually behave as if they were. The question, however, is whether this relationship holds in the dark shaded regions. When the strategic court behaves as if it were partisan, the relationship between awareness and impartiality is not as straightforward. In the darker regions of Figure 5.3, the relationship depends on the nature of the policy the court reviews. If the court reviews a policy it sincerely believes to be constitutional (e.g., policy y in Figure 5.5), more-aware voters will be more likely to accept the impartiality of the decision than less-aware voters. The intuition is as follows: Even though the policy is salient enough to induce the strategic court type to behave as if it were partisan, the more-aware voter will be more likely than a less-aware voter

146

Relationships between Transparency and Legitimacy

to attribute a decision upholding the policy to the court’s jurisprudence rather than to the political pressure or raw partisanship. Although beliefs in impartiality are reduced for more-aware voters, they are reduced less than for the less-aware voters, as depicted in Figure 5.5. On the other hand, if the court reviews a policy that it sincerely believes to be unconstitutional (e.g., policy x in Figure 5.5), greater awareness will induce a greater reduction in the updated beliefs of judicial impartiality. This is because more-aware voters do not expect an impartial court to uphold such a policy. More-aware voters will be more likely to attribute a decision upholding a policy like x to some sort of extrajudicial explanation. The court is either partisan or it is upholding the policy to protect its institutional interests. Because it is likely that courts will eventually be presented with policies they sincerely believe violate the constitution, it is likely that the relationship between awareness and impartiality in states characterized by the dark region will be negative. That said, if courts possess discretionary jurisdiction, they may select out those cases that present the highest risk, and the political branches may simply remove jurisdiction over highly sensitive matters. If either of these possibilities holds, it is possible that courts in political contexts that should incentivize strategic behavior may never have to behave strategically. If this is true, then the relationship between awareness and impartiality may indeed be positive in the dark region as well. In other words, any test of the negative relationship predicted by the model is biased against ﬁnding signiﬁcant results, because there are multiple mechanisms that make it less likely that courts will be confronted with the incentive to strategically avoid conﬂict. This analysis suggests the following proposition: Proposition 3. The relationship between awareness and judicial impartiality is positive in all equilibria in which the strategic court pools with the impartial court. When the strategic court separates from the impartial court, the relationship is likely to be negative but may be positive, depending on the mix of cases the court hears.16

To summarize, there are worlds in which people will expect courts to strategically uphold policies they believe to be unconstitutional, and it 16 Strictly speaking, the negative predictions are “weakly” negative, insofar as the observed

relationship should not be positive. Similarly, the positive predictions should be qualiﬁed as “weakly” positive – observed relationships should not be negative.

Constitutional Review and the Development of Judicial Legitimacy 147 is in those worlds that the positive relationship between awareness and impartiality may break down. Only one concern remains. Can the model give us conditions under which we should expect sincere or strategic constitutional review?

The political environment The model’s key implication is that where sincere judicial behavior is likely to emerge, we should observe a positive relationship between awareness and legitimacy; however, where courts are expected to strategically avoid conﬂict, the relationship is likely to be negative. So when should we expect sincere judicial behavior? As Figure 6.3 suggests, such behavior becomes increasingly likely as the judicially born costs of presidential deﬁance decrease.17 Speciﬁcally, as c → 0, the light region spans to the right across the space, and the strategic court type will pool with the impartial court type for higher and higher probabilities that the policy is salient. What might explain why courts perceive smaller or larger values of c? Although it may not be possible to measure the subjective costs of presidential deﬁance precisely, it seems likely that the political environment in which judges operate will affect the perceived costs of deﬁance. In worlds where judicial vetoes of salient policies are met with serious challenges to constitutional jurisdiction or even the removal of judges from the bench (Helmke 2005), we might expect judges to perceive relatively high costs of deﬁance. In contrast, where vetoes are ordinarily respected, and deﬁance manifests itself as simply failure to fully implement decisions rather than severe attacks on the judiciary, we might expect judges to perceive a relatively low deﬁance cost. In this sense, different political contexts induce different kinds of constitutional politics, which in turn produce different opportunities for learning about judicial impartiality. If this is true, then we should expect the public learning process about judicial impartiality (and by implication judicial legitimacy) to vary across political contexts. 17 Of course, the equilibria in the ﬁgure all involve the strategic court pooling with the

impartial court, even for relatively high values of c. Thus, it is also true that courts should be decreasingly likely to engage in strategic behavior when they simply expect presidential acceptance, no matter the salience of the policy and no matter the cost of deﬁance. Fortunately, it would appear that the external political environment should affect the values of c and simple expectations of compliance in similar ways. That is, the places where judges are generally insulated from serious external intrusion should be precisely the same places where judges simply expect acceptance.

148

Relationships between Transparency and Legitimacy

a different type of partisan court The model suggests that individuals should learn a considerable amount about judicial impartiality from the outcomes of cases. Although their beliefs may ultimately be incorrect in an absolute sense, the point is that substantive outcomes inform belief change. Where courts are free from signiﬁcant external pressure, greater awareness should produce stronger beliefs in judicial impartiality, and by implication, legitimacy. In contrast, environments that incentivize strategic behavior may produce an attenuated relationship between awareness and impartiality. In fact, the relationship can be negative. Also, the model makes very precise statements about the way that awareness should inﬂuence belief change, and even suggests that awareness may not matter at all. In particular, awareness should be unrelated to learning following judicial vetoes, because a veto communicates the court’s sincere evaluation of the policy. This follows because a partisan court does not challenge the government and the strategic court type never vetoes a policy it perceives to be constitutional. The implication is that prior doctrinal knowledge is only useful when confronted with a decision that could be characterized as strategically deferential. But this result likely depends on how we conceptualize a partisan court. Speciﬁcally, a partisan court in the model is associated with the current government. This assumption matches nicely the Mexican Supreme Court’s situation in the period immediately following the Zedillo reforms, and it is a common concern for newly constructed courts, in which judges have been appointed by sitting governments. Yet, this is not the only political scenario in which voters learn about their courts. After the transition from PRI to PAN presidential rule, the partisan nature of the Supreme Court’s position in Mexican politics changed. Where once there was a salient partisan concern about whether the Supreme Court would ever challenge a sitting PRI executive, after the transition, the partisan concern was about whether the Supreme Court would allow the new PAN government to rule. The early period of the Fox administration could have raised questions about the court’s partisanship. In September 2001, the court declared unconstitutional a Fox executive order establishing daylight savings time for the Federal District on the grounds that only the Congress was empowered to establish ofﬁcial times.18 And in April 2002, the court invalidated a major Fox electoral reform that had granted private companies the right to sell excess 18 Controversia constitucional 5/2001 and 8/2001.

Constitutional Review and the Development of Judicial Legitimacy 149 electrical capacity to the state for redistribution to the Mexican public (see Finkel 2008, 100).19 Both of these decisions could have suggested that the Supreme Court was merely a partisan tool of the past government.20 The Chilean Supreme Court confronted a similar scenario in the 1990s, in which it was accused of protecting the interests of the former military regime. In the Chanfreau case, the court granted jurisdiction over an infamous disappearance incident to a military court, ruling that the alleged crimes took place during a period of domestic conﬂict (Hilbink 2007, 192). Importantly, Hilbink argues that public shock over the decision derived from the fact that the court “had (again) changed its opinion regarding the state of war in 1973–1974” (193). Knowledge of past jurisprudence was critical to story. Likewise, the political environment the Supreme Court of Argentina confronted in the years following the Menem presidency reﬂected this alternative view of partisanship. As has been argued by many scholars, Carlos Menem ﬁlled the Supreme Court with political allies, creating a so-called automatic majority (Chavez 2004), and according to numerous accounts, these judges were instrumental in releasing Menem from house arrest in 2001 while he was being investigated on corruption charges. Just as the nature of the partisan concern regarding the Mexican Supreme Court changed after the Zedillo–Fox transition, what it meant for the Supreme Court of Argentina to be a partisan changed once Menem left ofﬁce. There is no question that the popular interpretation of what it means to be a partisan court may differ substantially from what I have assumed above, and the obvious, alternative concept of partisanship is related nicely to the insurance models of judicial reform (e.g., Finkel 2008; Ginsburg 2003). In fact, if the insurance model is correct, then there will be a number of cases in which what it means to be partisan will conform more to the scenario represented by Mexico under Fox or Chile under Aylwin than to Mexico under Zedillo or Venezuela during the early years of the Chavez regime. For this reason, it is worth considering how sensitive the results are to an alternative concept of partisanship that is consistent with this distinct scenario. In order to consider this possibility, I make a single change to the model. The formal results are reviewed in the appendix. I respecify the partisan court’s preferences such that it is incentivized to strike down the policy under review (e.g., Menem’s 19 Controversia constitucional 22/2001. 20 Magaloni and Sanchez (2003) have found evidence for a PRI bias in that period, although

the results in Chapter 4 do not.

150

Relationships between Transparency and Legitimacy

arrest or Fox’s electoral reform), as a court associated with a party out of power would be likely to do.21 Because the strategic court’s preferences have not changed, the three classes of equilibria remain. For the purposes of understanding how voter beliefs respond to case outcomes, there are still only two relevant scenarios. In one, the strategic court behaves as if it were the impartial court. In the second, the strategic court defers to the president, upholding all policies no matter how constitutionally suspect. Which results change? The logic in support of the claim that awareness does not inﬂuence beliefs about judicial impartiality when courts strike down policies depends clearly on the original partisanship assumption. When a partisan court is assumed to be associated with a party out of power, awareness can inﬂuence belief change, even when the court vetoes the policy. As in the original model, the way in which this works depends on the equilibrium type. Importantly however, this alternative version of the model mirrors the results in the original. Speciﬁcally, if the strategic court is expected to behave as if it were impartial, then a decision vetoing a policy is analytically identical to a decision upholding a policy under the original partisanship assumption. This is because all types of courts might have vetoed the policy just as all court types might have upheld the decision under the original assumption. For this reason, the voter cannot deﬁnitely rule out any particular type of court. She cannot know for sure whether the decision was issued by a partisan court or by either a strategic or impartial court that sincerely believed that the policy was unconstitutional. Just as before, awareness helps a voter distinguish court types, and because the strategic court is only vetoing policies it believes are unconstitutional, the more-aware voters are more likely to attribute a decision striking down a policy to legitimate doctrinal reasons than the less-aware voters.22 This mirroring result is demonstrated visually in the appendix to this chapter. 21 Although it would be possible to imagine the model with all four distinct court types,

this scenario would only add additional complication to the model without undermining the underlying dynamics. Dividing the model into two distinct scenarios clariﬁes exactly how belief change is affected by the alternative concepts of partisanship. Adding them to the same model would not inﬂuence the main empirical claim about how the equilibrium types inﬂuence learning, and it would muddy the exposition considerably. 22 Of course, this result depends on the court not being truly partisan, as deﬁned by the model (i.e., there is no policy that the court would strike down in the original version of the model or uphold in the current version). As I discuss earlier, a court is not strictly

Constitutional Review and the Development of Judicial Legitimacy 151 This addresses decisions vetoing policies, but we might also ask what happens if the voter observes a decision upholding a policy under the current partisanship assumption? As in the prior version of the model, awareness does not inﬂuence belief formation. The reason is that, in such a case, because the voter knows that a partisan court never upholds the policy and because the other courts gain nothing by striking down a valid policy, the voter knows for certain that the court is either impartial or strategic, whatever his prior beliefs were about the policy’s constitutionality. Further, we can conclude from this model that when the strategic court is expected to strategically defer to the president, then awareness inﬂuences voter inferences no matter what outcome is observed. If the policy is struck down, awareness helps the voter distinguish between a partisan and an impartial court. Of course, although the voter might learn a great deal in such a case, she will not observe this type of decision, because strategic courts will defer. Finally, if the policy is upheld, awareness helps distinguish between a strategic and an impartial court, as I describe later. The bottom line is that awareness is relevant to the impartiality beliefs in a larger set of scenarios under the alternative partisanship assumption.

Which results remain the same? The conditions establishing the equilibrium choices of the players are nearly identical, as are the equilibrium strategies for the players. As noted earlier, the equilibrium types remain the same. We still observe the basic interbranch dynamics that create worlds in which courts successfully constrain governments (when policies are not too salient or when the expected public backlash is signiﬁcant) and worlds in which courts do not constrain (when policies are highly salient or when the expected backlash is trivial). And voter beliefs respond to these different scenarios. Yet the particular ways in which voters learn have changed. So the key question is this: do these changes alter the essential predictions we have made across the political environments in which courts operate? Generally, the answer is no. partisan if it shares a worldview with a party in or out of power. What cannot be true is that there are no scenarios in which it would challenge this party. That said, if the court is truly partisan, increasing awareness will reduce beliefs in impartiality, even under a scenario in which strategic courts are expected to behave as if they were impartial. This hypothesis is not supported by the data I present in Chapter 6.

152

Relationships between Transparency and Legitimacy

First and foremost, substance matters. Much can be learned from the substantive outcomes of cases if judicial–government relations are described well by strategic models of constitutional review. More importantly, political environments that incentivize sincere judicial behavior still should induce a positive relationship between awareness and legitimacy. The rationale is as follows23 : In such a scenario, more-aware voters are more likely to observe consistency between past decisions and current decisions. Whether policies are upheld or struck down, more-aware voters will be more likely to attribute the decisions to doctrinal reasons than to political ones. On the other hand, when environments incentivize strategic judicial behavior, what voters infer depends on the type of policies the court reviews. If the court reviews a policy that it believes to be constitutional, then the voter will not observe a veto. As before, more-aware voters will attribute a decision upholding the policy to valid jurisprudential reasons. Consequently, their beliefs in judicial impartiality will be stronger than the less-aware voters. On the other hand, if the court reviews a policy that it believes to be unconstitutional, the voter still will not observe a veto, because the strategic court will defer to the president; however, the more-aware voter will attribute the decision upholding the policy to political factors more than the less-aware voter. Thus, as before, when the strategic court behaves as if it were impartial, the relationship between awareness and impartiality is positive, and when the strategic court is prudent, the relationship may be either positive or negative, depending on the mix of cases or its capacity to avoid delicate political decisions.

conclusion In this chapter, I considered what an individual might learn about judicial legitimacy by observing the interactions between a constitutional court and a government if the core interbranch dynamics so familiar to positive political models of constitutional review are picking up on something meaningful about the politics of constitutional review. The theoretical results suggest that there is much to learn about judicial legitimacy by observing patterns of high court decision making. Importantly, and consistent with familiar arguments in public law, the model suggests that increasing an individual’s familiarity with a court is likely to increase that individual’s beliefs in judicial legitimacy. Yet this result is limited to situations in which constitutional courts are not incentivized to strategically avoid conﬂict with government. In worlds in which judges are relatively 23 Again, I assume that courts are strategic.

Constitutional Review and the Development of Judicial Legitimacy 153 free from external constraints, judicial decisions will correspond very well with the images of impartiality suggested by the formal trappings of constitutional review. On the other hand, when courts are under pressure, when they face signiﬁcant constraints from government sources, the positive relationship between awareness and judicial legitimacy is likely to break down. The relationship between awareness and judicial legitimacy is likely conditioned by the kind of information to which people are exposed as they become familiar with their high courts. These results help address the questions around which the chapter was framed. Is there a relationship between the transparency and legitimacy conditions? How does strategic prudence construct judicial power if power depends on legitimacy? The model suggests that the relationship between transparency and legitimacy may not be straightforward, the strong empirical results in the United States notwithstanding. And, they suggest that strategic prudence may not necessarily create greater power. But it is important to recognize that these are only theoretical results. And they are derived from a model that makes obvious, simplifying assumptions. For this reason, it is far better to develop the implications of the argument once the model has been tested. In a subsequent chapter, I provide a number of tests of the major propositions summarized earlier.

appendix 5a This appendix contains the formal analysis for the model described in Chapter 5. I ﬁrst present results for the main model. I then present the model under the alternative partisan court assumption. I solve for perfect Bayesian equilibria (PBE). A PBE contains a sequentially rational strategy proﬁle and a proﬁle of beliefs that are consistent with this proﬁle and determined by Bayes rule whenever possible (Osborne and Rubenstein 1994, 233). Beliefs are undeﬁned at information sets that are not reached according to the equilibrium strategy proﬁle. I assume that no player takes a strictly dominated action on or off the equilibrium path, and I assume that, if indifferent, the court upholds, the president accepts, and the voter does not raise, lower, or punish.

case 1: low salience and low public backlash costs The PBE strategies in Case 1 are listed below. I wish to establish that these strategies and the subsequent beliefs constitute a PBE when b < 1, 1 and π satisﬁes the conditions below. β < 1+c

Relationships between Transparency and Legitimacy

154

Court: Partisan type upholds all policies, strategic and impartial types uphold policies they believe are constitutional and veto those they believe are unconstitutional. President: Accepts vetoes of nonsalient policies, but deﬁes vetoes of salient policies. α−π Voter: If π > max π , 1+π1+α , Raise opinion, Lower opinion, and Punishdeﬁance. α−π , Raise opinion, Do not lower opinion, Do not If π ∈ π , 1+π1+α punish24 (this holds whether or not b < 1). deﬁance α−π , π , Do not raise opinion, Lower opinion, and If π ∈ 1+π1+α Punish deﬁance. α−π , Do not raise opinion, Do not lower If π ≤ min π , 1+π1+α opinion, and Do not punish deﬁance (this holds whether or not b < 1). For the president, beliefs are trivial, because she is perfectly informed. The court’s beliefs are Pr(Salient) = β and Pr(Not Salient|) = 1 − β in all cases. The voter’s beliefs are as follows. Let (X,Y,Z) denote a state of the world, where X is the set of court types, Y is the set of sincere judicial evaluations, and Z is the set of presidential evaluations of salience. In particular, X = {g, s, p}, where g is the impartial type, s is the strategic type, and p is the partisan type. Let Y = {a, ∼a}, where a and ∼a represent policies the court sincerely believes to be constitutional and unconstitutional, respectively. Finally, let Z = {v, ∼v}, where v and ∼v represent policies the president perceives to be salient and not salient, respectively. Given the court and president strategies, when the voter observes the court uphold the policy, his beliefs are as follows: Pr(g, a, v|Iuphold ) =

παβ 1 − (π + π )(1 − α)

Pr(g, a, ∼v|Iuphold ) =

πα(1 − β) 1 − (π + π )(1 − α)

Pr(g, ∼a, |Iuphold ) = Pr(s, −∼a|Iuphold ) = 0 24 The precise conditions here depend on π , which I have assumed without loss of generality to be less than ½. All that changes when π is larger is the upper and lower bounds on

the interval.

Constitutional Review and the Development of Judicial Legitimacy 155 Pr(s, a, v|Iuphold ) =

π αβ 1 − (π + π )(1 − α)

Pr(s, a, ∼v|Iuphold ) = Pr(p, a, v|Iuphold ) =

π α(1 − β) 1 − (π + π )(1 − α)

(1 − π − π )αβ 1 − (π + π )(1 − α)

Pr(p, ∼a, v|Iuphold ) =

(1 − π − π )(β − αβ) 1 − (π + π )(1 − α)

Pr(p, a, ∼v|Iuphold ) =

(1 − π − π )(α − αβ) 1 − (π + π )(1 − α)

Pr(p, ∼a, ∼v|Iuphold ) =

(1 − π − π )(1 − α)(1 − β) 1 − (π + π )(1 − α)

When the voter observes the president accept a veto, his beliefs are Pr(g, ∼a, ∼v|Iaccept ) =

π π + π

Pr(s, ∼a, ∼v|Iaccept ) =

π π + π

Once Iaccept has been reached, the probability of being at any other history is 0. Finally, when the voter observes the president defy, his beliefs are Pr(g, a, v|Idefy ) =

π π + π

Pr(s, ∼a, v|Idefy ) =

π π + π

At Idefy , the probability of being at any other history is 0. By assumption, the president accepts all vetoes of nonsalient policies. Thus, the only issue is what will the president do if a salient policy is vetoed. Assume that the voter’s strategy is (Raise, Lower, and Punish). Given this strategy, the president expects −r if she accepts the veto and 1 − r − b if she deﬁes. Accordingly, the president’s strategy is sequentially rational if and only if b < 1. By assumption, we know the actions of the partisan and impartial courts. The question concerns what the strategic court will do when it reviews a policy it believes to be unconstitutional. Because the president will defy if the policy is salient, the court expects 1−β(1+c) if it vetoes and 0 if it upholds. Accordingly, the court’s strategy is sequentially rational if 1 . and only if β < 1+c

Relationships between Transparency and Legitimacy

156

Finally, consider the voter. Given the voter’s beliefs at Iuphold , he πα 1−π−π +π α expects 1−(π +π )(1−α) if he raises his opinion and 1−(π+π )(1−α) if he does not. Accordingly, the voter will raise its support if and only if α−π π . At the voter’s two other information sets, he expects π +π π > 1+π1+α

π for either lowering his support or punishing and π +π for doing nothing. Clearly the voter lowers and punishes deﬁance if and only if π > π . If π ≤ π , then the voter neither punishes nor lowers. Thus, (Raise opinion, Lower opinion, and Deﬁance) is Punish 1+π α−π sequentially rational if and only if π > max π , 1+α . If π ∈ α−π , then (Do not raise opinion, Lower opinion, and Punπ , 1+π1+α α−π ish deﬁance) is sequentially rational. Finally, if π ≤ min π , 1+π1+α ,

(Do not raise opinion, Do not lower opinion, and Do not punish deﬁance) is sequentially rational.25

case 2: high salience and low public backlash costs The PBE strategies for Case 2 are listed below. I wish to establish that these strategies, and the subsequent beliefs, constitute a PBE when b < 1, 1 , and when π satisﬁes the conditions below. β ≥ 1+c Court: Partisan and strategic types uphold all policies. Impartial type upholds a policy believed to be constitutional yet vetoes a policy believed to be unconstitutional. President: Accepts vetoes of nonsalient policies, but deﬁes vetoes of salient policies. 1 Voter: If π > 1+α , Raise opinion, Lower opinion, and Punish deﬁance. 1 If π ≤ 1+α , Do not raise opinion, Lower opinion, and Punish deﬁance. Because the strategic and impartial courts separate, the voter’s beliefs at Iuphold are different: Pr(g, a, v|Iuphold ) =

παβ 1 − π(1 − α)

25 In the event that the voter does not punish, the president’s strategy is no longer

constrained by b.

Constitutional Review and the Development of Judicial Legitimacy 157 πα(1 − β) 1 − π(1 − α)

Pr(g, a, ∼v|Iuphold ) = Pr(g, ∼a, |Iuphold ) = 0 Pr(s, a, v) =

π αβ 1 − π(1 − α)

Pr(s, a, ∼v|Iuphold ) =

π α(1 − β) 1 − π(1 − α)

Pr(s, ∼a, v|Iuphold ) =

π (1 − α)β 1 − π(1 − α)

Pr(s, ∼a, ∼v|Iuphold ) = Pr(p, a, v|Iuphold ) =

π (1 − α)(1 − β) 1 − (π + π )(1 − α)

(1 − π − π )αβ 1 − π(1 − α)

Pr(p, ∼a, v|Iuphold ) =

(1 − π − π )(β − αβ) 1 − (π + π )(1 − α)

Pr(p, a, ∼v|Iuphold ) =

(1 − π − π )(α − αβ) 1 − π(1 − α)

Pr(p, ∼a, ∼v|Iuphold ) =

(1 − π − π )(1 − α)(1 − β) 1 − π(1 − α)

When the voter observes the president accept an unfavorable resolution, his beliefs are Pr(i, ∼a, ∼v|Iaccept ) = 1, and the probability of being at other history is 0. Finally, when the voter observes deﬁance, his beliefs are Pr(i, ∼a, v|Iaccept ) = 1, and once Iaccept is reached, the probability of being at any history is 0. Because the president is still expected to defy and the courts beliefs are as they were in Case 1, the equilibrium conditions for the court and 1 , the strategic court the president are the same. Thus, as long as β ≥ 1+c will uphold, no matter what it sincerely believes. Similarly, because the voter is expected to punish deﬁance, the president’s strategy to defy the court’s veto of a salient policy is still optimal as long as b < 1. Given the voter’s beliefs, it is clearly optimal for him to punish deﬁance and lower his opinion if ever he observes the court veto a policy. This is because only an impartial court would have vetoed a policy it sincerely believed to violate

Relationships between Transparency and Legitimacy

158

πα 1−π(1−α)

if he

1−π and 1−π(1−α) if he does not. Accordingly, he will 1 1 . If π ≤ 1+α , he will not raise. only if π > 1+α

raise

the constitution in this equilibrium. At Iuphold , he expects raises his opinion his opinion if and

case 3: high constituent cost for defiance The PBE strategies in Case 3 are listed below. I wish to establish that these strategies, and the subsequent beliefs, constitute a PBE for b ≥ 1, all values of β, and when π satisﬁes the conditions below. Court: Partisan type upholds all policies, strategic and impartial types uphold policies believed to be constitutional and veto policies believed to be unconstitutional. President: Accepts all vetoes. α−π Voter: If π > max π , 1+π1+α , Raise opinion, Lower opinion, and Punish deﬁance. α−π , Do not raise opinion, Lower opinion, and If π ∈ π , 1+π1+α Punish deﬁance. α−π , Do not raise opinion, Do not lower If π ≤ min π , 1+π1+α opinion, and Punish deﬁance. Because the president accepts all vetoes, the strategic court will clearly pool with the impartial court for all values of β. Similarly, because the voter is expected to punish, and b ≥ 1, the president will accept all vetoes. With these strategies, the voter’s beliefs at Iuphold are the same as they were in Case 1, so the equilibrium condition on π is identical to what it was in Case 1. Also, because the beliefs are identical at Iaccept , in this case, as they were in Case 1, again the equilibrium condition on π above which the voter will lower his opinion of the policy is identical to that in Case 1. Finally, the voter’s beliefs at Iveto are undeﬁned via Bayes Rule, so they may be anything that induces a rational choice. There are many PBEs in this case, but for simplicity, I assume that Pr(g, ∼a, v|Iveto ) = 1. See footnote 7 in Chapter 6 for a discussion of how sensitive this result is to the assumption I make about beliefs off the equilibrium path. Indeed, assuming that players do not update does not have any effect on the equilibrium until π is very low. 1 Now, for β < 1+c , b ≥ 1, and π < π , there is an additional PBE where the strategic court pools with the impartial court, the president accepts only vetoes of nonsalient policies, and the voter refuses to raise or

Constitutional Review and the Development of Judicial Legitimacy 159 lower her opinion or punish deﬁance. Beliefs in this ﬁnal case are as they are in Case 1, and because the president deﬁes some vetoes, the voter’s beliefs at Iveto are deﬁned as they were in Case 1. With those beliefs and given the low priority on π, this equilibrium is more intuitively appealing than what is described in Case 3. Still, this equilibrium falls in the light region, so the predictions regarding the relationship between awareness and impartiality still hold.

beliefs in judicial impartiality decrease when court upholds Assume that there exists some allowable set of parameter values such that Pr(g) < Pr(g|Iaccept ). This implies that π<

πα . 1 − (π + π )(1 − α)

Multiplying both sides of the inequality by

1−(π +π )(1−α) π

yields

1 − π − π + πα + π α < α, which can be expressed as 1 − π − π < α(1 − π − π ), which is true only when α > 1. This can never be the case because α ∈(0,1).

analysis under the alternative partisan court assumption What if the partisan court is assumed to be associated with an out-ofpower coalition, such that it always strikes down presidential proposals? As described in the text, I have changed the model in only one way. Now the preferences of the partisan court are such that it obtains one unit of utility for striking down the policy and 0 otherwise. Thus, it has a dominant strategy to strike down policies. As is clear, the equilibria are nearly identical. The relevant conditions change in trivial ways.

case 1: low salience and low public backlash costs 1 Let b < 1 and β < 1+c . The condition on b ensures that the president will defy a salient veto no matter what the voter does; the condition on β

160

Relationships between Transparency and Legitimacy

ensures that the strategic court will strike down if invalid. What we need to know are the conditions under which the voter’s strategy is optimal. Consider the voter’s beliefs at Iuphold . They are Pr(g, a, v|Iuphold ) =

πβ π + π

Pr(g, a, ∼v|Iuphold ) =

π(1 − β) π + π

Pr(s, a, v|Iuphold ) =

π β π + π

Pr(s, a, ∼v|Iuphold ) =

π (1 − β) . π + π

The probability of being at any other history is 0 at Iuphold . Now consider voter beliefs at Iaccept . They are Pr(g, ∼a, ∼v|Iaccept ) =

π(1 − α) 1 − (π + π )α

Pr(s, ∼a, ∼v|Iaccept ) =

π (1 − α) 1 − (π + π )α

Pr(p, ∼a, ∼v|Iaccept ) =

(1 − π − π )(1 − α) 1 − (π + π )α

Pr(p, a, ∼v|Iaccept ) =

(1 − π − π )α . 1 − (π + π )α

The probability of being at any other history is 0 at Iaccept . Finally, voter beliefs at Idefy are Pr(i, ∼a, v|Iaccept ) =

π(1 − α) 1 − (π + π )α

Pr(s, ∼a, v|Iaccept ) =

π (1 − α) 1 − (π + π )α

Pr(p, ∼a, v|Iaccept ) =

(1 − π − π )(1 − α) 1 − (π + π )α

Pr(p, a, v|Iaccept ) =

(1 − π − π )α . 1 − (π + π )α

The probability of being at any other history is 0 at Iaccept . The PBE strategies for Case 1 are listed below. Court: Partisan type strikes all policies, strategic and impartial types uphold policies they believe are constitutional and veto those they believe are unconstitutional.

Constitutional Review and the Development of Judicial Legitimacy 161 President: Accepts vetoes of nonsalient policies but deﬁes vetoes of salient policies. α Voter: If π > max π , 1−π 2−α , Raise opinion, Lower opinion, and Punish deﬁance. α If π ∈ π , 1−π 2−α , Raise opinion, Do not lower opinion, and Do not punish deﬁance

(this holds whether or not b < 1). α , Do not raise opinion, Lower opinion, and If π ∈ 1−π , π 2−α Punish deﬁance. α If π ≤ min π , 1−π 2−α , Do not raise opinion, Do not lower opinion, and Do not punish deﬁance.

case 2: high salience and low public backlash costs 1 Let b < 1 and β ≥ 1+c . The condition on b ensures that the president will defy a salient veto, no matter what the voter does; the condition on β ensures that the strategic court will never veto a policy. We now need the conditions that shape the voter’s strategies. Voter beliefs at Iuphold are

Pr(i, a, v|Iuphold ) =

παβ πα + π

Pr(i, a, ∼v|Iuphold ) = Pr(s, a, v|Iuphold ) =

πα(1 − β) πα + π

π αβ πα + π

Pr(s, a, ∼v|Iuphold ) =

π α(1 − β) πα + π

Pr(s, ∼a, v|Iuphold ) =

π (1 − α)β πα + π

Pr(s, ∼a, v|Iuphold ) =

π (1 − α)(1 − β) πα + π

The probability of being at any other history is 0 at Iuphold . Voter beliefs at Iaccept are Pr(g, ∼a, ∼v|Iaccept ) =

π(1 − α) , 1 − π − πα

162

Relationships between Transparency and Legitimacy Pr(p, ∼a, ∼v|Iaccept ) = Pr(p, a, ∼v|Iaccept ) =

(1 − π − π )(1 − α) , 1 − π − πα

(1 − π − π )α . 1 − π − πα

The probability of being at any other history is 0 at Iaccept . Voter beliefs at Idefy are Pr(g, ∼a, v|Idefy ) =

π(1 − α) 1 − π − πα

Pr(p, ∼a, v|Idefy ) =

(1 − π − π )(1 − α) 1 − π − πα

Pr(p, a, v|Idefy ) =

(1 − π − π )α . 1 − π − πα

The probability of being at any other history is 0 at Idefy . The PBE strategies for Case 2 are as follows. Court: Partisan type strikes all policies, strategic and impartial types uphold policies they believe are constitutional and veto those they believe are unconstitutional. President: Accepts vetoes of nonsalient policies but deﬁes vetoes of salient policies. Voter: If π > max πα , 1−π 2−α , Raise opinion, Lower opinion, and Punishdeﬁance. If π ∈ πα , 1−π 2−α , Raise opinion, Do not lower opinion, and Do not

punishdeﬁance (this holds whether or not b < 1). π If π ∈ 1−π , Do not raise opinion, Lower opinion, and Punish , 2−α α deﬁance. If π ≤ min πα 1−π 2−α , Do not raise opinion, Do not lower opinion, and Do not punish deﬁance.

case 3: high constituent cost for defiance Now assume that b ≥ 1 and so the president will comply if the voter punishes. If the voter punishes, then for all values of β, the strategic court should pool with the impartial court. If the strategic court pools with the

Constitutional Review and the Development of Judicial Legitimacy 163 impartial court and the president accepts all vetoes, then we need to know voter beliefs at each information set. Voter beliefs at Iuphold are πβ π + π

Pr(g, a, v|Iuphold ) =

π(1 − β) π + π

Pr(g, a, ∼v|Iuphold ) = Pr(s, a, v|Iuphold ) =

π β π + π

Pr(s, a, ∼v|Iuphold ) =

π (1 − β) π + π

The probability of being at any other history is 0 at Iuphold . Voter beliefs at Iaccept are Pr(g, ∼a, v|Iaccept ) =

π(1 − α)β 1 − πα − π α

Pr(g, ∼a, ∼v|Iaccept ) = Pr(s, ∼a, v|Iaccept ) =

π (1 − α)β 1 − πα − π α

Pr(s, ∼a, ∼v|Iaccept ) = Pr(p, ∼a, v|Iaccept ) =

π (1 − α)(1 − β) 1 − πα − π α

(1 − π − π )(1 − α)β 1 − πα − π α

Pr(p, ∼a, ∼v|Iaccept ) = Pr(p, a, v|Iaccept ) =

π(1 − α)(1 − β) 1 − πα − π α

(1 − π − π )(1 − α)(1 − β) 1 − πα − π α

(1 − π − π )αβ 1 − πα − π α

Pr(p, a, ∼v|Iaccept ) =

(1 − π − π )(1 − β)α . 1 − πα − π α

Beliefs are undeﬁned at Idefy . As before, I assume that voter beliefs offpath are sufﬁciently high to induce punishment. The following are the PBE strategies for the conditions on b and β and for each condition on π described below.

164

Relationships between Transparency and Legitimacy

Court: Partisan type strikes all policies, strategic and impartial types uphold policies they believe are constitutional and veto those they believe are unconstitutional. President: Accepts all vetoes. α Voter: If π > max π , 1−π 2−α , Raise opinion, Lower opinion, and Punishdeﬁance. α If π ∈ π , 1−π 2−α , Raise opinion, Do not lower opinion, and Punish deﬁance.

α If π ∈ 1−π 2−α , π , Do not raise opinion, Lower opinion, and Punish deﬁance. α If π ≤ min π , 1−π 2−α , Do not raise opinion, Do not lower opinion, and Punish deﬁance. The beliefs about judicial impartiality of the voter at each information set, in each equilibrium, are as follows. Pr(Impartial|Idefy and strategic court behaves as if it were impartial) =

π(1 − α) 1 − (π + π )α

Pr(Impartial|Iaccept and strategic court behaves as if it were impartial) =

π(1 − α) 1 − (π + π )α

Pr(Impartial|Iuphold and strategic court behaves as if it were π impartial) = π + π Pr(Impartial|Idefy and strategic court behaves strategically) =

π(1 − α) 1 − π − πα

Pr(Impartial|Iaccept and strategic court behaves strategically) =

π(1 − α) 1 − π − πα

Pr(Impartial|Iuphold and strategic court behaves πα strategically) = . πα + π

Constitutional Review and the Development of Judicial Legitimacy 165 A simple inspection of these expressions demonstrates that awareness inﬂuences learning even when the court strikes down policies. More importantly, the mirroring result described in the text is clear. Note that when the impartial and strategic courts pool, awareness is irrelevant to impartiality beliefs, but only if the court upholds the policy. In the original model, this was true only if the court strikes down the policy. Figure 5A.1 further summarizes the key implication of the new partisanship assumption. The function plotted in blue, which reproduces the result in Figure 5A.1, reﬂects updated beliefs in impartiality under the ﬁrst model (Model 1), when the voter has observed the court uphold the policy and the strategic court behaves as if it were impartial in equilibrium. The light-colored function represents beliefs in impartiality under the second partisanship assumption, but when the voter observes the president defy an order striking down a policy and the strategic court pools with the impartial court.26 There are two critical points. First, note that just as a decision upholding the policy under the original model weakly reduces beliefs

Pr (Impartial | Iuphold, Original Model) 0.30 0.25 0.20 0.15 0.10 0.05

Pr (Impartial | IDefy, Alternative Model) 0.2 0.4 0.6 0.8 Prior belief that policy is constitutional (a)

1.0

ﬁgure 5A.1. Updated Beliefs in Impartiality by the Partisanship Assumption. Note: Figure shows updated beliefs about the impartiality of the court under the original and alternative partisanship assumptions, and in equilibria in which the strategic court pools with the impartial court. The blue curve reﬂects beliefs in the original model following a decision upholding the policy. The red curve reﬂects beliefs in the alternative model following a decision striking that policy down 26 The result is identical for cases in which the president accepts a veto.

166

Relationships between Transparency and Legitimacy

in impartiality, a decision striking down a policy weakly reduces beliefs in impartiality under the second model. Second, because the strategic court only strikes down policies it sincerely believes violate the constitution, more-aware voters will reduce their beliefs in impartiality less than less-aware voters after observing the court strike down a policy. For this reason, when the strategic court pools with the impartial court, the relationship between awareness and impartiality is weakly positive. The argument for why this relationship depends on the type of case the court hears is covered in the text, but is identical to that in the original model.

6 A Cross-National Analysis of Judicial Legitimacy

This chapter tests the empirical implications developed in Chapter 5. This requires cross-national data on beliefs in high court legitimacy. To date, the GCB study represents the only systematic cross-national analysis in which scholars use a precise measure of the judicial legitimacy concept targeted at the highest appellate court of a state. The regression analysis that follows makes use of the publicly available survey information from their study. I conclude the chapter by discussing implications for understanding the relationship between transparency and legitimacy and for the role of strategic deference in the construction of judicial power.

hypotheses and data The available, individual level measures of legitimacy beliefs derive from public opinion surveys conducted in Bulgaria, Poland, Russia, Spain, France, Hungary, and the United States in 1995. This accounts for 40 percent of the states in the original analysis (see Cohn, White, and Sanders 2000 for details). Although limited, the data establish a difﬁcult test for the argument. Even the full data set only includes European and North American states. With the possible exception of Russia, these are not political contexts we typically associate with imposing severe costs on judges for overly active behavior. There is no Argentina here, no Venezuela, no Zimbabwe. Empirically, the upper bound of any indicator of political constraints on judging may not be very high – we may not capture the cases that will allow us to uncover the conditionality anticipated by the model. For this reason, a ﬁnding that the effect of awareness on legitimacy 167

168

Relationships between Transparency and Legitimacy

is not constant across these states, and that it varies in ways anticipated by the argument, would be encouraging. Properly testing the argument is complicated further by the nature of the surveys. The most precise test requires measuring features of particular judicial–government interactions and knowing that each respondent was exposed to those interactions. Unfortunately, there is no measure in the surveys that indicates the types of decisions each respondent has observed; however, there are two ways to access this information. First, it is possible to measure the conditions that could have induced strategic judicial behavior in these countries by measuring the costs of deﬁance (c). Standard measures of the rule of law should capture judicial expectations about the costs of political deﬁance.1 Where deﬁance at most manifests as the failure to fully implement a decision, costs should be low; where deﬁance manifests as impeachment, or worse, some other metaconstitutional removal, costs should be high. With respect to a measure of deﬁance costs, the idea is that there is a considerable difference between massive judicial purges in the cases of Argentina (Helmke 2005, 68–71 and 86) and Pakistan (Masood 2007a, 2007b, 2007c, 2007d), substantial jurisdiction stripping in Chile (Hilbink 2007, 109–112), and the court-curbing that has characterized the history of the U.S. Supreme Court (Clark Forthcoming, 2010). The ﬁrst three countries have relatively low rule-of-law scores during those periods, while the United States’ score is relatively high. As Clark suggests about the U.S. case, “Most [court-curbing] bills never make it out of committee; indeed, most never receive a hearing in committee.” Of course, there is considerable evidence that the U.S. Supreme Court behaves strategically on occasion (e.g., Epstein and Knight 1998; Lasser 1988; Murphy 1962), so it is possible that we do not see severe attacks on the court precisely because it is prudent. Still, in order for U.S. court strategy to account fully for why Argentina, Chile, and Pakistan have looked so different, you would have to believe that American judges are somehow more strategic than judges outside the United States. I use two measures of these costs.2 The International Country Risk Guide’s (ICRG) Law and Order measure is scaled from 0 to 6, with higher scores indicating higher levels of the rule of law. The second measure, the Freedom House Civil Liberties index, is scaled in reverse, from 1 to 7, 1 Given the variance in rule-of-law concepts, I have attempted to select measures that cap-

ture the piece of the rule of law that captures the extent to which government ofﬁcials tend to be constrained in practice by legal obligations. 2 Another possibility is that low values of these measures indicate states whose courts may be considered “partisan,” as deﬁned in the original formulation of the model. This possibility is difﬁcult to rule out entirely, because a nonpartisan, yet strategic court may appear

A Cross-National Analysis of Judicial Legitimacy

169

with lower numbers reﬂecting increased protections for civil liberties. A key component of the Freedom House index indicates the extent to which the government is committed to the rule of law (Freedom House 2004). To be fair, both measures are yearly scores, and as such they are both probably best thought of as producing an average measure of the costs of deﬁance across multiple constitutional cases. It is also possible to conceptualize the Freedom House score as a measure of democracy, not the rule of law. Scholars have used the Civil Liberties index as one component of a democracy measure, though it is typically combined with the Freedom House’s Political Rights index. The Civil Liberties index was designed to tap into “liberty” not democracy, and in this regard, it is important to remember that, when used as a democracy measure, the scores measure a substantive concept of democracy (Clark, Golder, and Golder 2008, chapter 5). Critically, the substantive democracy concept that the Civil Liberties index most validly measures includes numerous features of the rule of law, which is what I hope to measure. Even with indicators of deﬁance costs, we require additional assumptions to test the model. It is not necessarily the case that respondents developed their beliefs in judicial legitimacy from events in 1995 only, the year the surveys were administered. To address this possibility, in addition to the 1995 score, I also estimate models using the ﬁve- and tenyear averages of these scores (i.e., from 1990 to 1995 and from 1985 to 1995). The idea is to average across varying political contexts over the recent history of the court, the history respondents would be most likely to know. Because the Russian Constitutional Court was essentially created anew in 1993 and the Bulgarian Constitutional Court in 1991, I only to observers to be a partisan court when the political conditions incentivize prudential decision making. One way to evaluate this is to consider whether courts in states with low rule-of-law scores have ever ruled against their appointers. The Russian constitutional court under Valery Zorkin is the most well-publicized example. As it turns out, that court’s ﬁrst decision invalidated Yeltsin’s order creating a uniﬁed police and national security ministry (Sharlet 1993 as cited in Jackson and Tushnet 2006, 744), so it is unclear that we want to consider the Russian Constitutional Court under Zorkin to be partisan. Furthermore, although Helmke demonstrates that Argentina’s Supreme Court was biased toward the military regime and toward both Alfonsín and Menem, it is simply not true that it never ruled against them (Helmke 2005, 102). The same is true of the Chilean Supreme Court, which issued pro-individual rights rulings during the Pinochet era (Hilbink 2007, 94). It is further instructive that eight of the thirteen Chilean Supreme Court justices from 1973 to 1980 were appointed by either Allende or Frei, which suggests that even a bias toward the regime cannot be explained easily by a simple partisan story (Hilbink 2007, 103). Finally, if we take the alternative view of partisanship, as reﬂected in the second version of the model, it is unclear how the rule-of-law measures could possibly measure that concept.

170

Relationships between Transparency and Legitimacy

take averages from 1993 to 1995 and 1991 to 1995 for those countries, respectively. Results are robust to this restriction. With these data limitations, it is important to be clear about empirical expectations. The hypothesis I wish to test is that the effect of awareness on judicial legitimacy is positive, but only if deﬁance costs are low.3 Where these costs are high, the effect of awareness is likely negative, but it may be positive, as well, depending on the mix of cases individuals observe, a mix we cannot see. As I noted in Chapter 5, in order for the relationship to be positive where deﬁance costs are high, the high court would have to have avoided all situations in which strategic deference should emerge in equilibrium. This is possible through strategic case selection strategies or simply because the court’s jurisdiction has been limited to relatively uncontroversial areas. Given the inability of the current research design to identify precisely what kind of interactions the respondents have observed, the most likely expectation is that where the costs of deﬁance are high, the positive effect of awareness on legitimacy should be attenuated. Despite this ambiguity, we can state deﬁnitely two expectations. First, we should not estimate a negative awareness effect where deﬁance costs are low. Second, if the effect of awareness is attenuated, it should be strongest and positive where deﬁance costs are low and weaken as these costs increase.

Measures of Public Support and Awareness GCB measure judicial legitimacy with a three-item summated index. Item 1 asks whether the respondent would support doing away with her highest court if it started making unpopular decisions. Item 2 asks whether the respondent would support restricting the court’s jurisdiction under the same scenario. These items directly tap into the core element of the “diffuse support” concept of legitimacy: institutional commitment (Gibson, Caldeira, and Baird 1998, 348). The authors then add a third item, which asks whether the respondent trusts the court to make decisions that are good for the country.4 Following GCB, the dependent variable, Judicial 3 Strictly speaking, the relationship is predicted to be non-negative, but only when deﬁance

costs are low. 4 The institutional commitment question was, “If the [HIGHEST COURT of YOUR

COUNTRY] started making a lot of decisions that most people disagree with, it might be better to do away with the [HIGHEST COURT OF YOUR COUNTRY] altogether” (strongly agree to strongly disagree). The jurisdiction question was, “The right of the [HIGHEST COURT OF YOUR COUNTRY] to decide certain types of controversial issues should be reduced” (strongly agree to strongly disagree). Finally, the trust measure was “The [HIGHEST COURT OF YOUR COUNTRY] can usually be trusted to make decisions that are right for the country as a whole” (strongly agree to strongly disagree).

A Cross-National Analysis of Judicial Legitimacy

171

Legitimacy, is the average of the responses to these three items, normalized on the [0,100] interval. Although judicial trust is related to the legitimacy concept, it is reasonable to believe that judicial trust and diffuse support are distinct concepts. Conceptually, I may trust a court because I believe it will produce policies that beneﬁt me; but if those policies changed, I would support reforming its constitutional role. In other words, trust does not imply necessarily institutional commitment. Empirically, the trust measure is much less strongly correlated with the other two measures of legitimacy than they are with each other, and Gibson and Caldeira (1998) themselves propose an alternative legitimacy measure in which they exclude the trust item.5 Because the alternative measure is conceptually clear, and because Gibson and Caldeira suggest it, I run the analysis on both the three- and twovariable indices. If the trust item should not be included in a legitimacy measure, then the likely effect is increased random measurement error. There is no reason to anticipate a difference in the direction or strength of results across the indices, but it is certainly possible that estimates in the three-item models will be less precise. The independent variables in the Gibson, Caldeira, and Baird study are as follows: Awareness is the respondent’s self-reported familiarity with her country’s high court, and Satisfaction measures the respondent’s speciﬁc support for her high court’s policy output.6 I have replicated the GCB results, ﬁnding signiﬁcant effects of awareness only in the United States, Spain, Poland, and France. The effort here will be to reestimate their models using a pooled data set and test whether political context conditions the effect of awareness. In order to remain faithful to the original speciﬁcation, I estimate the following multiplicative interaction model: Legitimacy = β1 (Awareness) + β2 (Deﬁance Costs) + β3 (Awareness * Deﬁance Costs) + β4 (Satisfaction) + ε

5 The correlation between institutional commitment and jurisdictional commitment is .34

(p < 0.001). The correlation between trust and each of the other measures is .11 (p < 0.001) and .15 (p < 0.001), respectively. 6 The awareness question was, “Now a few questions about the [HIGHEST COURT OF YOUR COUNTRY]. Would you say you are very aware, somewhat aware, not very aware or have never heard of [HIGHEST COURT OF YOUR COUNTRY]?” Like Gibson, Caldeira, and Baird, I excluded from the analysis all respondents who had never heard of the court.

172

Relationships between Transparency and Legitimacy

The only speciﬁcation change involves the Deﬁance Costs measure and its interaction with Awareness. The expectation is that, where deﬁance costs are low, the effect of awareness on legitimacy should be positive; however, where these costs are high, the effect is likely negative. The quantity of interest is the marginal effect of Awareness: β1 + β3 * Deﬁance Costs. Given the direction of the ICRG measure, this effect should be positive when the ICRG measure is high but decrease as ICRG decreases. Given the direction of the Freedom House measure, the marginal effect of Awareness should be positive when the civil liberties index is low and decrease as the civil liberties index increases.

analysis Table 6.1 displays the results from ordinary least-squares regressions with White-Huber standard errors, clustered within states. The top half of the table summarizes models in which the ICRG index measures Deﬁance Costs. The bottom half contains the same information substituting the Freedom House Civil Liberties index. The left half of the table displays results for the three-item judicial legitimacy index used in the GCB study, while the right half of the table shows results from models in which I used the two-item legitimacy measure, as suggested by Gibson and Caldeira (1998). Finally, for each dependent variable, I estimated the model based on the ten-year average, ﬁve-year average, and the 1995 value for the two measures of Deﬁance Costs. The results are generally consistent with expectations, although the two-item model results are stronger. Consider the top right section of the table. Note that the Awareness coefﬁcient, which gives the marginal effect of awareness on judicial legitimacy when the ICRG index is minimized (i.e., for states characterized by the lowest level of the rule of law), is negative in every model and that the interaction term is always positive. This suggests that awareness reduces legitimacy at the highest level of deﬁance costs, but that, as deﬁance costs decrease and individuals are increasingly exposed to the kind of information that emerges when strategic courts behave as if they were impartial, the negative effect of awareness attenuates and may even become positive. The results in the bottom right of the table are even stronger. Recall that, in contrast to the ICRG measure, under the Freedom House Civil Liberties index, low numbers reﬂect a higher degree of respect for the rule of law. With this in mind, the signs on the key coefﬁcients are exactly reversed from what we observed at the top right of the table. Now the Awareness coefﬁcient

173

table 6.1. Determinants of Judicial Legitimacy Beliefs Three-Item Legitimacy Measure

ICRG Rule of Law Index Awareness Deﬁance Costs Awareness * Deﬁance Costs Satisfaction Constant N R2 Freedom House Civil Liberties Index Awareness Deﬁance Costs Awareness * Deﬁance Costs Satisfaction Constant N R2

Two-Item Legitimacy Measure

Average 1985–1995

Average 1990–1995

1995

Average 1985–1995

Average 1990–1995

1995

−5.87 (5.04) −2.14# (0.955) 1.70 (0.970) 6.14∗∗∗ (0.796) 53.43∗∗∗ (5.11) 4594 0.13

−6.04 (6.34) −1.55∗ (0.623) 1.64 (1.17) 6.11∗∗∗ (0.745) 50.96∗∗∗ (3.29) 4594 0.13

−4.50 (6.31) −0.901 (0.895) 1.28 (1.11) 6.07∗∗∗ (0.704) 47.99∗∗∗ (5.33) 4594 0.13

−15.46∗ (6.12) −1.90 (1.53) 3.86∗ (1.12) 5.78∗∗∗ (0.90) 48.0∗∗∗ (7.61) 4607 0.10

−17.46 (9.39) −1.13 (0.993) 4.06∗ (1.67) 5.69∗∗∗ (0.826) 44.63∗∗∗ (5.47) 4607 0.10

−12.52 (9.51) −0.704 (0.95) 3.01 (1.65) 5.75∗∗∗ (0.88) 42.44∗∗∗ (0.80) 4607 0.10

4.07∗ (1.59) 0.363 (0.950) −706 (0.46) 6.01∗∗∗ (0.677) ∗∗∗ 42.09 (2.80) 4594 0.13

4.33# (1.97) −0.549 (1.18) −0.886 (0.93) 6.04∗∗∗ (0.688) 41.92∗∗∗ (23.17) 4594 0.13

4.87∗ (2.03) 1.32 (0.835) −1.13 (1.03) 6.10∗∗∗ (0.771) 40.2∗∗∗ (3.02) 4594 0.13

8.36∗∗ (1.75) −0.17 (1.02) −1.74∗∗ (0.43) 5.46∗∗∗ (0.71) 39.84∗∗∗ (2.90) 4607 0.10

9.05∗∗ (1.99) −0.17 (1.59) −2.68∗ (0.96) 5.58∗∗∗ (0.73) 39.4∗∗∗ (3.4) 4607 0.10

10.11∗∗ (2.20) 0.73 (1.54) −3.19∗ (1.16) 5.71∗∗∗ (0.87) 37.24∗∗∗ (4.41) 4607 0.10

Note: Displays the OLS estimates with White-Huber standard errors, clustered by country. *p < 0.05; ** p < 0.01; ***p < 0.001; #p < 0.10.

174

Relationships between Transparency and Legitimacy

is positive and strongly signiﬁcant, and the interaction term is negative, again suggesting that the marginal effect of awareness on judicial legitimacy is positive under conditions in which the rule of law is strongly respected but decreases as the rule of law breaks down. The results on the left side of the table resemble those on the right, but the coefﬁcients are weaker. Although the signs of the coefﬁcients in the upper left section of the table (i.e., the ICRG models) are correctly signed, they are not statistically signiﬁcant in two-tailed tests. The results in the Freedom House models are stronger, though again, the interaction terms are not signiﬁcant in any of the models. In summary, Table 6.2 suggests that the effect of awareness on legitimacy is likely positive in contexts in which deﬁance costs are low and attenuates as deﬁance costs increase. Yet the statistically signiﬁcant results seem to be limited largely to the legitimacy models that exclude the trust item. Although suggestive, Table 6.1 can take us only so far, because in the interactive model context, the true marginal effects and their standard errors are not described in the table (Friedrich 1982). What we want to know exactly is the marginal effect of Awareness across the full range of the costs of deﬁance. Figures 6.1 and 6.2 provide that information. The solid lines in Panel A of each ﬁgure indicate the marginal effect of Awareness on judicial legitimacy across the range of the ICRG measure. Panel B shows the same information for the Freedom House measure. The dashed lines represent 95 percent conﬁdence intervals for each marginal effect described by the solid line. Figure 6.1 summarizes the effects for the two-item legitimacy measure. As Panel A demonstrates, the marginal effect is negative and signiﬁcant for extremely low values of the ten-year ICRG average (where deﬁance costs are high), but this effect increases as we move to the right. Indeed, for extremely high values of the ICRG measure, the effect is positive and statistically signiﬁcant, reﬂecting the result in the literature on the positivity bias. The results are similar for the ﬁve-year average and 1995 score; however, in those models, the upper bound on the 95 percent conﬁdence interval is positive for the effects at the low end of the ICRG scale, suggesting that a null effect is likely when deﬁance costs are high. Notably, even in the graphs where this is true, the effect of awareness grows as we move to the right across the ICRG scale. Panel B suggests the identical conclusion, but again, because deﬁance costs increase to the right with the Freedom House Civil Liberties measure, we observe a positive effect of Awareness for extremely low values of the index (i.e., on the far left). This effect decreases as we move to the right, as deﬁance costs increase. Indeed, for extremely high values of the index, the

A Cross-National Analysis of Judicial Legitimacy

175

effect is negative. The story is roughly the same in Figure 6.2; however, we observe only null effects at high deﬁance costs. In both ﬁgures, it is clear that the effect of awareness grows and becomes substantial as deﬁance costs decrease. Additional tests Table 6.2 provides information on the robustness of the ﬁndings summarized so far. Each row displays the Awareness coefﬁcient and its interaction with another measure of deﬁance costs. I have suppressed the other coefﬁcients in the model, though the speciﬁcation is the same as described earlier. The White-Huber standard error associated with each coefﬁcient is listed below and p-values for two-tailed hypothesis tests are in brackets. The ﬁrst row shows the results for the World Bank Rule of Law score (Kaufmann, Kraay, and Mastruzzi 2007). Clearly, this could be an alternative measure of deﬁance costs given the measurement logic described earlier, but the scores became available only in 1996 (the year I use in the model), which limits this measure’s use in the current study. The last three rows show results for measures of judicial independence. Conceptually, a measure of judicial independence is unlikely to indicate the courts’ perceived deﬁance costs, but it can provide direct information on the kind of judicial behavior to which respondents have been exposed.7 The scores I include all have been used to measure the judicial power concept of independence, as articulated by Cameron (2002), so that each measure seeks information not just on the autonomy of the judicial decision-making process, but also on compliance with judicial decisions as well.8 With respect to the theoretical model in Chapter 5, high values of these scores correspond to the light regions of the parameter space. Low values, in contrast, correspond to the dark regions. Speciﬁcally, both the Tate and Keith (2007) and Cingranelli and Richards (2008) measures are derived from U.S. State Department country reports. They are three category ordinal measures (0 = no independence; 1 = partial independence; 2 = full independence). The third measure, the 7 Insofar as the rule-of-law measures, in part, pick up judicial independence, the same may

be true of those indicators. 8 The Cingranelli and Richards and the Tate and Keith measures are designed explicitly

to capture behavioral information about judicial independence; however, the Cingranelli and Richards score mixes institutional with behavioral information. See Ríos-Figueroa and Staton (2008) for validity analysis on each measure.

Relationships between Transparency and Legitimacy

176

Panel A: ICRG models Dependent variable: Judicial legitimacy (two item) 15 10 5 0 –5 Marginal Effect of Awareness 95% Confidence Interval

–10 –15 2

4 5 3 ICRG rule of law average (1985–1995)

6

Dependent variable: Judicial legitimacy (two item) 25 20 15 10 5 0 –5 –10 –15 –20 –25

Marginal Effect of Awareness 95% Confidence Interval 2

3 4 5 ICRG rule of law average (1990–1995)

6

Dependent variable: Judicial legitimacy (two item) 20 15 10 5 0 –5 –10 Marginal Effect of Awareness 95% Confidence Interval

–15 –20 2

3

4

5

6

ICRG rule of law Score (1995) Note. Displays marginal effects of Awareness on the two item legitimacy score for different values of the ICRG law and order measure.

ﬁgure 6.1. (Continued)

A Cross-National Analysis of Judicial Legitimacy

177

Panel B: Freedom house models Dependent variable: Judicial legitimacy (two item) 10

5

0

–5

Marginal Effect of Awareness 95% Confidence Interval

–10 1

2 3 4 5 6 Freedom house civil liberty average (1985–1995)

7

Dependent variable: Judicial legitimacy (two item) 20 15 10 5 0 –5 –10 Marginal Effect of Awareness 95% Confidence Interval

–15 –20 1

2

3 4 5 6 Freedom house civil liberty average (1990–1995)

7

Dependent variable: Judicial legitimacy (two item) 25 20 15 10 5 0 –5 –10 –15

Marginal Effect of Awareness 95% Confidence Interval

–20 –25 1

2

3 4 5 6 Freedom house civil liberty score (1995)

7

Note. Displays marginal effects of Awareness on two-item legitimacy score for different values of the Freedom House Civil Liberty measure.

ﬁgure 6.1. Marginal Effects of Awareness on Judicial Legitimacy by Deﬁance Cost Measure (Two-Item Models)

Relationships between Transparency and Legitimacy

178

Panel A: ICRG models Dependent variable: Judicial legitimacy (three item) 15 10 5 0 –5 –10

Marginal effect of awareness 95% confidence interval

–15 2

3 4 5 ICRG rule of law average (1985–1995)

6

Dependent variable: Judicial legitimacy (three item) 15 10 5 0 –5 –10

Marginal effect of awareness 95% confidence interval

–15 2

3 4 5 ICRG rule of law average (1990–1995)

6

Dependent variable: Judicial legitimacy (three item) 15 10 5 0 –5 –10

Marginal effect of awareness 95% confidence interval

–15 2

3 4 5 ICRG rule of law average (1995)

Note. Displays marginal effects of Awareness on the three-item legitimacy score for different values of the ICRG law and order measure.

ﬁgure 6.2. (Continued)

6

A Cross-National Analysis of Judicial Legitimacy

179

Panel B: Freedom house models Dependent variable: Judicial legitimacy (three item) 10 5 0 –5 Maginal effect of awareness 95% confidence interval –10 1

2 3 4 5 6 Freedom house civil liberity average (1985–1995)

7

Dependent variable: Judicial legitimacy (three item) 20

10 0 –10 Maginal effect of awareness 95% confidence interval –20 1

2 3 4 5 6 Freedom house civil liberity average (1990–1995)

7

Dependent variable: Judicial legitimacy (three item) 15 10 5 0 –5 –10

Maginal effect of awareness 95% confidence interval

–15 1

6 2 3 4 5 Freedom house civil liberity average (1990–1995)

7

Note. Displays marginal effects of Awareness on the three-item legitimacy score for different values of the Freedom House Civil Liberty measure.

ﬁgure 6.2. Marginal Effects of Awareness on Judicial Legitimacy by Deﬁance Cost Measure (Three-Item Models)

180

table 6.2. Awareness Effects for Alternative Deﬁance Cost/Judicial Independence Measures Dependent Variable: Judicial Legitimacy (Three Item) Deﬁance Cost/Judicial Independence Measure 1985–1995 Coef. (RSE)

p

1990–1995 Coef. (RSE)

p

World Bank Rule of Law Awareness Awareness * World Bank Contract Intensive Money Awareness Awareness * CIM Tate & Keith J.I. Awareness Awareness * Tate & Keith Cingranelli & Richards J.I. Awareness Awareness * CIRI

−14.98 (4.21) 20.55 (4.62) 0.403 (2.17) 0.711 (1.04) −1.25 (3.83) 2.27 (2.09)

0.012 0.004 0.862 0.533 0.754 0.320

−17.94 (9.33) 23.82 (10.29) −0.646 (2.86) 1.80 (1.57) −1.75 (4.11) 2.68 (2.23)

0.113 0.069 0.832 0.314 0.687 0.282

Dependent Variable: Judicial Legitimacy (Two Item) Deﬁance Cost/Judicial Independence Measure

1995

1985–1995

Coef. (RSE)

p

1.06 (1.00) 1.55 (0.935) −14.18 (8.01) 19.46 (9.15) −0.13 (2.34) 1.45 (1.18) −273.0 (2.59) 1.78 (1.29)

0.330

Coef. (RSE)

p

1990–1995 Coef. (RSE)

p

0.148 0.127 0.077 0.958 0.272 0.921 0.239

−30.78 (8.82) 40.64 (9.91) −1.26 (2.86) 1.89 (1.42) −6.52 (5.73) 6.16 (3.08)

0.013 0.006 0.681 0.254 0.299 0.092

−41.13 (12.86) 52.22 (14.44) −2.47 (3.68) 3.14 (1.98) −7.17 (6.13) 6.69 (3.26)

0.024 0.015 0.539 0.189 0.295 0.096

1995 Coef. (RSE)

p

0.407 (1.00) 3.66 (0.890) −36.30 (12.11) 46.84 (14.33) −1.27 (2.90) 2.39 (0.869) −1.28 (3.01) 2.51 (1.50)

0.698 0.006 0.024 0.017 0.679 0.154 0.690 0.170

Note: Table displays OLS estimates for Awareness and its interaction with the alternative measures. White-Huber standard errors clustered within states are reported below each estimate; p-values for two-tailed tests lie to the right.

A Cross-National Analysis of Judicial Legitimacy

181

Clague et al. Contract Intensive Money score, reports the proportion of a state’s currency that is held in banks (Clague et al. 1999, 188). Originally developed as an endogenous measure of property rights protections, it has been used as a general measure of judicial independence (Powell and Staton 2009). The conceptual logic behind the measure is that individuals are generally more comfortable holding their currency in banks when the set of institutions necessary for protecting their assets provide genuine protection against state predation. The judiciary provides the foundation for this institutional structure. The chief advantage of the measure is that it does not rely on idiosyncratic evaluations of country experts, but instead provides information on the mass behavior we would expect if courts were, in fact, independent. The results strongly reﬂect the ﬁndings uncovered earlier. Every single interaction term is positive, suggesting that the effect of awareness grows as deﬁance costs decrease (or judicial independence increases). For the two-item legitimacy measure, the interaction term is statistically significant for each model that uses the World Bank and Contract Intensive Money measures and for two of the Cingranelli and Richards models. In the three-item models, the interaction term is signiﬁcant in all of the Contract Intensive Money models at the .10 level or below. It is signiﬁcant in a one-tailed test for the World Bank measure. Importantly, the signs on all of the coefﬁcients are in the expected direction. Figure 6.3, which provides the necessary information to evaluate the signiﬁcance of the marginal effects across the range of deﬁance costs or judicial independence, conﬁrms the results described in Table 6.2. Indeed, for the two-item measure, awareness is negative and signiﬁcant for low values of the World Bank and Contract Intensive Money measures and positive and signiﬁcant for high values of these scores. For the three-item measure, we continue to see the signiﬁcant and positive relationship for high values of these scores and an attenuation toward zero as deﬁance costs decrease (judicial independence increases). In addition to evaluating the robustness of the results to alternative measures, it is important to consider another way of modeling these data. In particular, the data have an explicitly hierarchical structure. Individuals are clustered within states, and the hypotheses anticipate that state-level factors should condition individual-level relationships. The multiplicative interaction models described earlier are not necessarily biased, but they may signiﬁcantly underestimate the error associated with the cross-level interactions. To address this possibility, I use the two-stage model described in Lewis and Linzer (2005, 353–357), which despite its

Relationships between Transparency and Legitimacy

182

simplicity, turns out to be an extremely robust estimator. The ﬁrst step in the model estimates the relationship between Awareness and Judicial Legitimacy at the individual level in each state sample via ordinary least squares (OLS). In the second step, the OLS coefﬁcients from the ﬁrst step are then regressed on state-level factors of interest, using White-Huber standard errors. Table 6.3 shows the coefﬁcients estimated in the second step of the model, with standard errors and p-values (two-tailed) for each of the deﬁance costs and judicial independence measures described earlier. Again, I include the ten-year and ﬁve-year averages, along with the 1995 score where appropriate. If political context inﬂuences the relationship between awareness and legitimacy in the way the theoretical model anticipates, then we should observe positive coefﬁcients for each estimate in the

Panel A: (World bank models) Dependent variable: Judicial legitimacy (two Item) 15 10 5 0 –5 Marginal effect of awareness

–10

95% confidence interval

–15 –2

0

–1

1

2

World bank rule of law (1996) Dependent variable: Judicial legitimacy (three Item) 15 10 5 0 –5 Marginal effect of awareness

–10

95% confidence interval

–15 –2

–1

0 1 World bank rule of law (1996)

2

Note. Displays marginal effect of Awareness on the two-item legitimacy score for different values of the World Bank Rule of Law measure.

ﬁgure 6.3. (Continued)

A Cross-National Analysis of Judicial Legitimacy

183

Panel A: (Contract intensive money models) Dependent variable: Judicial legitimacy (two item) 70 60 50 40 30 20 10 0 –10 –20 –30 –40 –50 –60 –70

Marginal effect of awareness 95% confidence interval 0

.25 .5 .75 Contract intensive money (1995)

1

Dependent variable: Judicial legitimacy (three item) 40 30 20 10 0 –10 –20 Marginal effect of awareness 95% confidence interval

–30 –40 0

.25

.5 .75 Contract intensive money (1995)

1

Note. Displays marginal effect of awareness on legitimacy scores for different values of the Contract Intensive Money measure.

ﬁgure 6.3. Marginal Effects of Awareness (World Bank and Contract Intensive Money)

model’s second step, with the exception of the coefﬁcient for the Freedom House measure, which should be negative. A positive ICRG estimate indicates that, as the ICRG measure increases, the estimated relationship between awareness and legitimacy increases as well. In contrast, a negative Freedom House estimate suggests that, as the Freedom House score increases (and deﬁance costs increase), the positive effect of awareness on legitimacy decreases. This is precisely what we observe in Table 6.3. Every coefﬁcient is in the expected direction. In fact, the estimates are statistically signiﬁcant for the two-item legitimacy measures in all but the Contract Intensive Money and Tate and Keith models, although the Tate and Keith estimates are signiﬁcant in a one-tailed test. Again, the results are somewhat weaker

184

table 6.3. Effects of Deﬁance Costs and Judicial Independence on the Awareness Estimate Dependent Variable: OLS Estimate of Awareness Effect (Three Item) Average 1985–1995

ICRG Rule of Law RSE Freedom House Civil Liberties RSE World Bank Rule of Law RSE Contract Intensive Money RSE Take & Keith RSE Cingranelli and Richards RSE

Average 1990–1995

Dependent Variable: OLS Estimate of Awareness Effect (Two Item)

Average 1995

Average 1985–1995

Coef. (RSE)

p

Coef. (RSE)

p

Coef. (RSE)

p

Coef. (RSE)

0.603 (0.562) −0.831 (0.499)

0.332

0.905 (0.617) −0.750 (0.431)

0.203

0.989 (0.635) −0.394 (0.337) 1.63 (0.969) 13.47 (14.44) 1.18 (0.722) 1.12 (0.913)

0.180

1.54 (0.596) −1.42 (0.515)

10.49 (12.81) 0.785 (0.639) 1.96 (1.23)

0.157

0.450 0.308 0.173

14.6 (15.69) 1.78 (1.22) 2.18 (1.45)

0.142

0.402 0.240 0.206

0.295

Average 1990–1995

p

Coef. (RSE)

0.049

1.78 (0.718) −1.56 (0.493)

0.039

p

Coef. (RSE)

p

0.056

1.78 (0.783) −1.25 (0.278) 2.75 (1.01) 24.97 (15.29) 2.02 (0.894) 1.92 (1.13)

0.072

0.025

0.154 0.394 0.177 0.307

18.22 (15.88) 1.74 (0.867) 3.77 (1.54)

0.303 0.138 0.058

29.14 (16.08) 2.79 (1.55) 3.97 (1.77)

Average 1995

0.114 0.169 0.088

0.006 0.042 0.163 0.086 0.187

Note: Displays the OLS estimates of the second stage estimated dependent variable regression. Coefﬁcients reﬂect the change in the individual level awareness – legitimacy relationship associated with a unit change in deﬁance costs or judicial independence, with White-Huber errors in parentheses. p-values for two-tailed tests lie to the right of the coefﬁcients.

A Cross-National Analysis of Judicial Legitimacy

185

for the three-item legitimacy measures, although many of the coefﬁcients approach statistical signiﬁcance in a one-tailed test. Moreover, it is critical to recognize that these second-step models are being estimated on seven observations. Thus, in addition to the ways that strategic case selection or jurisdiction stripping might undermine our ability to observe attenuation effects, these tests may simply lack the power necessary to pick up the true relationships.

Expanding the sample The results described so far are consistent with the theoretical model developed in this chapter. Greater awareness can be translated into greater legitimacy, but only when the activities to which individuals are exposed are likely to match the nonpolitical image of judges promoted by judges themselves. When high courts confront signiﬁcant constraints, which incentivize them to behave strategically, greater awareness only exposes a person to potential jurisprudential contradictions. Unfortunately, the available data, which restrict the tests to individuals in seven states covered by the original GCB analysis, are limiting. There are only seven contexts into which individuals might fall, and the lack of data constrains the power of the statistical tests, which in turn biases the tests toward the null. Even if the full set of data is absent, it is possible to design a simple, if blunt, test using the full set of countries. If deﬁance costs inﬂuence the kind of information to which individuals are exposed, then we should observe a difference in deﬁance costs across the states in which the awareness effect was signiﬁcant in the original study and those in which it was not. Table 6.4 divides the eighteen states in the Gibson, Caldeira, and Baird study into two categories: states in which the effect of awareness on judicial legitimacy was originally estimated to be positive and signiﬁcant and states in which the effect was insigniﬁcant. The table displays the mean value of the deﬁance costs across these two categories of states from 1985 to 1995 and from 1990 to 1995 for both measures. Simple t-tests of the difference of means suggest moderate support for the hypothesis that the two samples are derived from fundamentally different types of states with respect to the rule of law. Two of the four tests show signiﬁcant differences at the 10 level. The remaining two tests produce borderline results. In light of the extremely small sample size and the fact that the

186

Relationships between Transparency and Legitimacy

table 6.4. Differences in Deﬁance Costs

ICRG Rule of Law 1985–1995 Average 1990–1995 Average Freedom House Civil Liberties 1985–1995 Average 1990–1995 Average

Awareness Signiﬁcant (N = 8)

Awareness Insigniﬁcant (N = 10)

Difference p-value

5.35 5.62

4.93 5.16

−0.41 −0.47

0.145 0.065

1.60 1.54

2.70 2.00

1.10 0.45

0.064 0.129

Source: Freedom House (2007); International Country Risk Guide (2004). Note: Table displays means of deﬁance cost scores across the ICRG and Freedom House measures and the two time periods (1985–1995 and 1990–1995). The p-values are for onetailed hypothesis tests for each measure. Given the directions of the two scales, I anticipate a negative difference for the ICRG measure and positive difference for the Freedom House measure.

countries in both samples are European or North American, the results are suggestive.

a tension between transparency and legitimacy The literature on judicial legitimacy in law and psychology and political science, which highlights procedural fairness and the positivity bias, suggests that high court judges ought to pursue greater transparency. By maximizing exposure to the public, judges cultivate beliefs in judicial impartiality, which is essential for the development of a deep normative commitment to the judiciary. For new courts, especially those born in a context of signiﬁcant public skepticism about legal institutions, these implications are highly appealing. They suggest a direct path to public support. More broadly, from the perspective of a public enforcement mechanism for judicial power, the implications also are appealing. Existing results in the literature imply that judges can promote the transparency and legitimacy conditions simultaneously – the worst-case scenario is that promoting a court will do it no harm. If this is all correct, the majority of courts around the world with constitutional jurisdiction may be using their resources inefﬁciently. Rather than maximizing public exposure, most courts seem to present selective information to the public. Indeed, even a majority of the world’s high courts selectively promote their case results. It is worth wondering, in light of actual judicial behavior, whether

A Cross-National Analysis of Judicial Legitimacy

187

there might be something more to the relationship between transparency and legitimacy. Anecdotally at least, there are reasons to suspect that keeping a low proﬁle is sometimes preferable. In 1995, Germany’s federal constitutional court voided a provision of a Bavarian primary education ordinance that required the display of a cruciﬁx on every classroom wall. As Vanberg (2005) notes, the decision met with massive resistance from the public in heavily Catholic Bavaria, and the Bavarian parliament received over 70,000 signatures advocating deﬁance. Amid the controversy, Justice Dieter Grimm published an open letter to the Bavarian public in the Frankfurter Allgemeine Zeitung. Grimm wrote: In a system that sets forth its political and social order in a constitution [sic] and establishes a constitutional court to protect that document, political and social conﬂicts are bound to arise in the form of constitutional disputes. Unlike the political arena, the court is unable to sidestep such disputes by refusing to decide them. It has to decide the conﬂicts, yet not on its own initiative [sic] and according to the justices’ individual preferences or the supposed wishes of a popular majority, but according to the preestablished provisions of the Basic Law. Not everyone will be satisﬁed with the court’s decision, but that is in the nature of the judicial resolution of the conﬂicts; and, at times, the majority will be the disappointed party. This is what constitutionalism is all about; its purpose is to safeguard the rights of minorities against encroachment by the majority. Under these circumstances [sic] constitutional Court decisions cannot always be greeted with universal approval. Criticism of such decisions is normal and in the interest of the court’s own reﬂections about its role as ﬁnal arbiter of the Constitution; [sic] indeed, such criticism is necessary. Disagreement with a decision, however, does not relieve the critic of the duty to comply with it. This is the basic premise of the entire system of constitutional governance . . . If in light of the Cruciﬁx case, state or church ofﬁcials create the impression that this is not so, they threaten to disrupt the Federal Republic’s generally stable history of postwar constitutional governance and are likely to shake the foundations of social peace. (as quoted in Kommers 1997)

Political ﬁgures on the left of the German political spectrum echoed the sentiment, but ultimately, the court backed away from the most radical interpretation of the decision – that cruciﬁxes would have to be removed even in the absence of a speciﬁc parental complaint (Kommers 1997, 484). Nevertheless, Jackson and Tushent report that Grimm’s letter itself produced a “controversy over the propriety of a sitting judge’s public nonadjudicatory speech” (Jackson and Tushnet 2006, 1408). In other words, promoting the adherence to the rule of law in a particular instance seemed to heighten the backlash against the court.

188

Relationships between Transparency and Legitimacy

Valery Zorkin, the chief justice of the Constitutional Court of Russia, also used the national media to advocate adherence to the rule of law. In 1992, the court was called upon to review the constitutionality of a proposed referendum in the Tartar Autonomous Republic. The referendum asked Tartars to consider becoming a separate, autonomous state in the Russia federation. The unconstitutionality of the proposal seemed clear enough to the court, and Zorkin went on national television describing the decision banning the referendum. Like the cruciﬁx case in Bavaria, the referendum case in Tartarstan met massive resistance. In this context, Zorkin predicted publicly that failure to comply would result in a situation “a hundred times worse than Yugoslavia” (Jackson and Tushnet 2006, 740). The Tartars simply ignored the court and went ahead with the referendum nonetheless. Commenting on the Zorkin experience, Hausmaninger (1995) suggests that chief justice’s public activity was exceptionally harmful, suggesting that the court was identiﬁed as essentially powerless by the Tartars’ deﬁance. Although there is no systematic evidence for a negative inﬂuence of public rule of law speech, the anecdotal evidence is suggestive. It is perhaps harmful for courts to publicly advocate compliance, in part because the activity itself may be viewed as improper, but it may be exceptionally harmful if the publicity serves to highlight a case of massive deﬁance. The German and Russian experiences suggest two rationales for judges to be careful about the kind of information they release to the public. In one sense, it is simply unbecoming for a judge to engage in nonadjudicatory appeals normally reserved to the politicians. Even a direct request for compliance can be interpreted as improper. Second, public communication, especially insofar as it highlights noncompliance, can undermine the judicial image. Indeed, Hausmaninger writes, “Given the lack of enforcement of the Court’s opinions, the decision to refrain from publishing a great number of ineffectual decisions may have been wise . . . . one could certainly argue that the Court should have made its legal points more guardedly and selectively” (as quoted in Jackson and Tushnet 2006, 752). This appeal for judicial prudence suggests that perhaps strategic deference alone can advance legitimacy. There is a reasonable argument in defense of that position. Yet to consider whether prudence resolves the problem of undermining legitimacy through too much publicity, it is worth considering the implications of prudence that derives from the theoretical model in Chapter 5. In the extreme, if transparency were complete, it would mean that a court’s audience (public and private actors alike) would be exposed maximally to the patterns of its decision making

A Cross-National Analysis of Judicial Legitimacy

189

and, thus, be as capable as possible of anticipating decisions in new cases. It is worth considering how strategic deference would advance legitimacy in such a setting. The model suggests that one consequence is to lower beliefs in judicial impartiality.9 What’s more, as a person becomes increasingly familiar with a court, as a person becomes better able to see strategic deference in action, the strategy of prudence becomes increasingly problematic. For a highly informed person, strategic deference can drive beliefs in judicial impartiality down considerably. Although the empirical evidence in this chapter is not perfect, there is nevertheless evidence across a number of tests that suggests that awareness decreases judicial legitimacy where courts are likely to face considerable external pressures. If this characterization of belief formation is accurate, then we confront a tension between transparency and legitimacy that cannot be resolved by strategic decision making; rather, strategic decision making raises the problem itself. In this sense, judges should not always prefer maximal transparency, precisely because it exposes the realities of judicial politics to public scrutiny. It is one thing for judicial deliberation to appear, on occasion, unguided by principles. It is quite another for a court to be observed lying down in the face of signiﬁcant governmental pressure. When the Venezuelan Supreme Court “committed suicide” to avoid being killed, it explicitly exposed itself as either highly strategic, or worse, a court entirely prepared to do the bidding of Hugo Chavez. The choice to declare constitutional an obviously unconstitutional referendum was highly informative and could have undermined the court’s legitimacy. When the court later removed a constitutional right to equal airtime for members of the media, it likely reinforced the view that the court was a mere partisan tool of the Chavez government. Where courts are free to resolve cases sincerely, there is not much of a need to control information, but where courts confront signiﬁcant political challenges, it may be better to inﬂuence strategically the type of information people are exposed to. Managing to do this well is a difﬁcult problem. Importantly, the need for a serious, professionally managed public relations strategy follows from this concern. As Dubek (2007) suggests, it is useful for courts to have surrogates in cases where its authority is 9 This is especially true under the original partisanship assumption. In the alternative model,

this result depends on the type of case the policy the court evaluates. However, if ever the court hears a policy that is widely believed to be unconstitutional, the result of strategic deference is to lower beliefs in impartiality.

Relationships between Transparency and Legitimacy

190

threatened. The German constitutional court was aided, in part, by party leaders, as was Zorkin in Russia. But a simpler, more direct strategy involves constructing an internal public relations ofﬁce whose job it is to publicly defend the court itself. Such an ofﬁce can help coordinate responses, from the court and from its allies. This is what we observe among many constitutional courts around the world. Nevertheless, it is impossible to control completely the kind of information that will emerge about a major political institution. Courts will sometimes be covered when they would rather remain obscure. Thus, it is important to consider further what strategic deference means for the development of power if legitimacy is a core source of judicial authority.

the development of judicial power In light of the possible tension between transparency and legitimacy it is useful to reconsider how it is that judges might go about building the power of their courts. Figure 6.4 represents two versions of a process commonly described in theoretical analyses of judicial power. The x-axis represents time; the y-axis reﬂects the power of a court. I discuss the ﬁgure through the work of Ginsburg (2003) and Carrubba (2009), who provide particularly careful expositions. Both authors address how judicial power can expand over time as a function of the simultaneous choices made by judges and the political ﬁgures whose behavior they review. Importantly, this is a story told about the development of judicial power in many parts

Ginsburg (2003)

Judicial power

High

Carrubba (2009) Low

Institutional infancy

Institutional maturity Time

ﬁgure 6.4. Judicial Power Dynamics

A Cross-National Analysis of Judicial Legitimacy

191

of the world. The argument emerges in Shapiro (2004), Couso (2004), Moustafa (2007), and many other places. The central feature of the arguments summarized in Figure 6.4 is that by prudential adjudication, courts can avoid the kinds of confrontations that are likely to result in deﬁance, or worse, institutional retribution. This process, however, evolves over time. Thus, in a court’s institutional infancy, decision making is careful, and power is relatively low; however, over time, courts are able to expand their authority, and the need for prudence is reduced. The central issue concerns how this is accomplished. Although the arguments share the core feature of careful review producing power, the proposed mechanisms are distinct. For Ginsburg, the shift from low to high power results from a Marshallian strategy exempliﬁed by the Marbury decision. When young courts assert judicial authority, they must do so where the implications of their decisions impose negligible shortrun costs on political principals. By not challenging ofﬁcials over salient policies, or in ways that would call for immediate policy change, courts ensure compliance. This process of strategic judicial deference constructs a sequence of compliant responses to decisions that challenge authority only in ways that require little if any policy adjustment. This sequence of compliance produces a norm of compliance. As long as a court is perceived to resolve cases fairly, social actors will increasingly view the court as a reasonable place to resolve conﬂicts; and given the norm of compliance, the court will begin to expand its authority to policy areas it was once unable to inﬂuence. The belief that parties will comply with judicial decisions is crucial here. Ginsburg (2003, 73) writes, “If the court develops a record of adjudicating fairly and ﬁnding at least some of the time for either side of a political cleavage, then the court gives the parties a stake in continuing to play by the constitutional rules.” This logic hinges on shared expectations about compliance, because these expectations are what ensure that a loser will comply in anticipation of future victories. If, alternatively, parties anticipated that future losers would fail to comply, compliance with a loss in the present only imposes a dead weight loss. In short, the logic unravels without a shared belief in future compliance. The question about the Ginsburg model concerns how the compliance belief is generated. The key issue is: Why would repeated instances of compliance over either nonsalient policies or where no policy adjustment is required convince a party that the state is likely to implement a signiﬁcant policy adjustment over a salient policy in the future? In other words, why is the norm of compliance over nonsalient policies transferable to salient policies? A plausible alternative is that when courts are highly strategic

192

Relationships between Transparency and Legitimacy

during their institutional infancy, they induce a belief in society that they are irrelevant to serious political outcomes – that if they attempted to enforce a decision against the state on a matter of consequence, it would be acceptable to defy. Ginsburg’s discussion of Taiwan’s Council of the Grand Justices is suggestive. In the 1950s, the Kuomintang (KMT) regime faced a dilemma. It was constitutionally committed to holding an election, but on account of its failure to retake the Chinese mainland, it could only hold an election on Taiwan and the island of Quemoy, locations where the KMT enjoyed little political support. For this reason, the party faced the real possibility of losing power in a fair election; yet if it suspended the election, it would undermine its effort to distinguish itself in the eyes of the West from the antidemocratic communists on the mainland. Upon a request by the Executive Yuan for a constitutional interpretation concerning the constitutionality of suspending elections, the Grand Justices held that as long as the mainland was occupied by communists, elections could be constitutionally suspended (129). Ginsburg writes: In molding the constitutional order to the needs of the government in the ongoing national emergency, the council was acting as a straightforward instrument of the KMT. The legitimacy of the council suffered accordingly. It is not surprising that the image of the justices as KMT partisan lingered four decades later. (130)

On this account, the justices were partisans. An alternative view, though generally consistent with the Ginsburg model, is that the justices were doing the best they could under the circumstances and resolved the conﬂict prudently. Either way, it is not clear that this sort of careful decision making should produce a norm that the court should be obeyed in future cases. It might just as well suggest that the court is either an arm of the regime or totally irrelevant to serious policy choices. Neither of these inferences seems to advance expectations of future compliance over salient policies, which are so important to the Ginsburg story. Carrubba develops a public enforcement mechanism for judicial decisions, yet tackles directly the critical theoretical question in such an argument. Imagine that public enforcement can induce public authority compliance. Imagine further that the state imperfectly reﬂects the public’s policy interests, but that it reﬂects these preferences better than a court empowered to review its choices, on average at least. Why would the public come to the defense of such a court against its democratic representatives? Carrubba argues that by allowing states to violate the law when it is net beneﬁcial for the public, and by only challenging states when

A Cross-National Analysis of Judicial Legitimacy

193

doing so is both net beneﬁcial to the public and not sufﬁciently costly to the state (in order to ensure compliance), courts can induce a belief among the public that allowing the court to enforce legal terms systematically as it sees ﬁt is preferable to permitting the state to ignore the rule of law. When the public comes to believe this, the court’s power expands considerably. Carrubba provides a precise mechanism for compliance, but the argument completely ignores the concept of legitimacy. In the Carrubba model, people do not support courts because they believe that they are impartial and trustworthy, but rather because the systematic application of the rule of law provides a greater expected value over the long run than allowing the state to violate the rule of law on occasion. There is nothing inherently wrong with the claim, of course. The argument provides a tight description of endogenous institutional change without resorting to a concept like legitimacy; however, if legitimacy is a meaningful component of a compliance mechanism, it is worth considering how the legitimacy concept would affect the argument. We can do so by considering dynamic implications of the model in Chapter 5. Clearly, the model does not speciﬁcally address repeated interactions, and doing so would be analytically challenging, given the information structure. Nevertheless, it is possible to consider its logic informally in a dynamic setting. The general claim I want to advance is that strategic deference can develop beliefs in judicial legitimacy, but only if the distribution of societal expectations about the nature of the court under analysis is appropriately structured, and only if information about what courts do is sufﬁciently muddied. The best-case scenario for a strategic court is that the distribution of beliefs about its true type is bimodal, where people believe that the court is either likely to be a simple extension of the regime (i.e., a partisan court) or entirely impartial, placing little weight on the possibility that the court is in fact strategic. Under the ﬁrst concept of partisanship, a decision upholding a policy (a strategically deferential decision or a sincere one) strictly decreases beliefs in impartiality (also see Baird and Gangl 2006). The result of a decision striking down a policy is quite different. Given the initial, bimodal distribution of beliefs, individuals are highly likely to perceive the court as impartial after the decision. I have argued elsewhere (Staton 2002) that this may be exactly what happened in the wake of the Mexican Supreme Court’s decision in the Zedillo case. A highly strategic court came to be perceived (in the short run at least) as impartial, largely because the preexisting frame for thinking about the Supreme Court only allowed the court to be either partisan or impartial. The inference derives from the

194

Relationships between Transparency and Legitimacy

following simple logic: because a partisan court would never have challenged Zedillo, it must have been an impartial court. This kind of dynamic can explain signiﬁcant increases in judicial legitimacy as a result of strategic decision making. Of course, we need to consider what happens after such a decision. Once this sort of decision has been issued, the court may again face the incentives for strategic deference. And again, each time it defers, beliefs in impartiality will be reduced. Importantly, however, if the information environment is sufﬁciently poor, such that people are genuinely uncertain about the court’s sincere evaluation of the case, then it is possible that this new distribution of beliefs about the court can be effectively locked in. Why? Because it will become impossible to distinguish between a truly impartial court and a strategic court, if nobody has any sense of the court’s sincere preferences, which is certainly possible in a less-than-fully transparent environment. Insofar as you cannot detect strategic behavior, you cannot learn anything else about the court. In this way, and under these circumstances, it can be quite valuable for a court to control the kind of information people have about their patterns of decision making. This is critical because, if ever a strategic court confronts a case where the result seems obvious and against the government’s interests, judicial deference will signiﬁcantly lower beliefs in impartiality. In such a context, transparency may constitute a real problem. If legitimacy is key to judicial power, and if courts are forced to act prudentially on occasion, then we confront a situation in which transparency is both a necessary condition for courts to take advantage of legitimacy (per the model in Chapter 2) and a potential problem courts must solve when they wish to strategically defer (per the model in Chapter 5). What this suggests about the development of judicial power is that strategic deference does not obviously build power through the construction of legitimacy absent the control of information. Under complete transparency, deferential strategies can undermine it. For this reason, judges have additional incentives to manage public information. The trick might be to ensure transparency about what is required for compliance while molding substantive coverage in ways that make it difﬁcult to recognize an obvious act of strategic deference. Fortunately, legal doctrine, with its inherent vagueness and ﬂexibility, is well suited for the job. Perhaps by inﬂuencing the way that the judgment and the rationale are reported, judges can ensure the kind of transparency conducive to solving the problems addressed in Chapter 2 without fully advancing the kind of transparency that can undermine legitimacy as discussed in Chapter 5.

A Cross-National Analysis of Judicial Legitimacy

195

The second key implication is that, if courts cannot fully control the information to which individuals are exposed, it is certainly possible that judicial power, at least insofar as it is constructed through beliefs in legitimacy, derives from precisely the opposite strategy suggested by Carrubba and Ginsburg. This is to say that judicial legitimacy and the corresponding power that derives from it under a public enforcement mechanism might be best advanced by decisions that are clearly costly. In the equilibria described in Chapter 5, a strategic court does not make such decisions, and courts do not do so in the Ginsburg or Carrubba accounts either. In other words, a strategic court never challenges government over a policy that is perceived to be sufﬁciently salient, so salient that an individual would be compelled to infer that any court that actually challenged the government in such a case must be impartial. A court that pursued such a strategy could strongly inﬂuence beliefs in judicial impartiality. Yet such a court would also risk a purge, which would imply a pyrrhic victory. Analyzing this possibility is beyond the scope of the current study, but the line of analysis suggests that we might revisit seriously our sense of how courts build power over time.

7 Democratic States and the Development of Judicial Power

Courts that constrain arbitrary government are believed to promote a number of appealing outcomes. They are associated with economic growth and development (Barro 1997; North 1990), the maintenance of civil order (North, Summerhill, and Weingast 2000), the protection of political rights (Howard and Carey 2004), and even with the prevention of torture and other cruel violations of personal integrity (Apodaca 2004: Hathaway 2007; Powell and Staton 2009). An intriguing development in legal research has been the explicit linking of democratic processes to judicial independence and, more broadly, to the rule of law.1 Although there are rationales for constraining the state through legal institutions under autocratic rule (e.g., Barros 2002; Moustafa 2007), scholars and activists nevertheless remain skeptical on the whole. Examples like Singapore are more than offset by examples like Cuba and Myanmar. Even Egypt, the case around which Moustafa’s persuasive analysis revolves, ultimately reigned in its constitutional court when it became overly active in the area of human rights (Moustafa 2007, 198–202). Reﬂecting a common understanding of the issue, Carothers (2007, 16) writes, “The core point is that rule-of-law development and autocracy . . . go poorly together. Key elements of the rule of law directly threaten autocratic rule. An independent judiciary will be a source of power and authority beyond the executive’s reach.” In general, autocrats just do not create powerful judiciaries. Even if autocrats do not generally delegate authority to judges, we cannot infer necessarily that democrats will. Yet common theoretical

1 Helmke and Rosenbluth (2009) provide an excellent review of this development.

196

Democratic States and the Development of Judicial Power

197

arguments for the emergence of powerful courts do in fact rely on democratic logics. The insurance model hangs critically on the political uncertainty created by increasing electoral competition (e.g., Finkel 2008). Without the risk of a power transfer in the future, the rationale to tie hands in the present is unclear. Whittington’s (2005) explanation for why courts are able to exercise judicial review against sitting governments depends on a modern understanding of policymaking under democratic institutions. Courts can do so because governments depend on them to eliminate policies that they cannot reform themselves. Judicial review allows governments to overcome veto structures created by federalism and fractious coalitions, and it allows cross-pressured politicians a means of implicit policymaking (i.e., through judicial review) when explicit policymaking, even position-taking, would be electorally infeasible. This logic is somewhat strained under autocratic rules, however, where many of these problems are solved by the concentration of power. Similarly, Rogers’s (2001) account of informative judicial review does not hold together well if governments are able to alter unsuccessful policies in the future. The lack of democratic constraints would largely remove Rogers’s rationale for a strong court. Finally, the well-known veto player account of judicial power, which emerges in the voluminous separationof-powers literature (e.g., Ferejohn and Weingast 1992), and which has been developed further in comparative politics (e.g., Ríos-Figueroa 2006) clearly operates on a democratic logic. What grants courts power under this account is the polarization (or fragmentation) of electoral politics, such that a space is created for independent constitutional review. When political ofﬁcials cannot coordinate on a response to active judiciaries, judicial power expands. In the context of a single veto player autocracy, where coordination is immediate, this argument predicts weak courts. For all these reasons, the third wave of democracy not only raised hopes about the possibility for building tighter connections between governments and the preferences of their citizens, but also about the creation of limited government itself. The arguments suggest a clear rationale for the promotion of democracy around the world. Through a blunt policy instrument, regime type, institutional designers establish incentives that ultimately promote economic growth, civil order, and liberty.

democracy and the public enforcement mechanism The afﬁnity between democratic processes and judicial power emerges in this book as well. I have argued that a theory of judicial power, in

198

Relationships between Transparency and Legitimacy

which the public ultimately binds its leaders to legal outcomes, implies that judges should care about public information. In order for a public enforcement mechanism to work, courts must enjoy some threshold level of legitimacy and the conﬂicts they resolve must be sufﬁciently transparent (Vanberg 2005). Judges go public to inﬂuence legitimacy and transparency, because those are the conditions necessary for power. The public enforcement mechanism, though not entirely bound to democratic institutions,2 operates by essentially leveraging a mechanism of vertical accountability to provide horizontal accountability, and likely this is easier to accomplish under democratic rules. There are two reasons for this. If Bueno de Mesquita et al. (2003) are basically right about regime dynamics, then the size of the coalition that supports an autocratic leader (i.e., the winning coalition) will be smaller than that which supports a democratic leader; and, for this reason, members of a winning autocratic coalition are more likely to beneﬁt from private goods that are tied to the survival of particular autocrats than members of a winning democratic coalition. A public enforcement mechanism in an autocracy would ask a considerable amount of the members of the winning coalition relative to what it asks in democracies. It essentially requires them to risk losing access to the private goods the current regime is providing in order to hold that regime accountable to a normative commitment to the rule of law. In a democracy, the mechanism still requires a commitment to rule-of-law values, potentially at the expense of particular policy outcomes, but because democratic governments are less likely to reward supporters with private goods, the risk of losing out on a beneﬁt that cannot be captured under another government is considerably attenuated. Second, and more simply, the transparency condition depends importantly on the availability of a free press. Indeed, absent free-ﬂowing, large-scale communication, it is unclear that the public enforcement mechanism could work. Although the logic of the public enforcement account is clearest under democratic institutions, the particular argument I develop raises a number of issues that help ﬂesh out the precise link between democracy and the rule of law. First, judicial power under a public enforcement mechanism does not depend entirely on fragmented or polarized politics. For sure, multiple veto points are likely to reduce the possibility of noncompliance in its various forms, but the conditions for independent judicial rule can be met even in a single veto player scenario. All that matters is that there 2 Moustafa’s (2007, 44–45) study of judicial defense networks under authoritarianism

suggests that public pressure can be leveraged even if elections are not fully free.

Democratic States and the Development of Judicial Power

199

is a viable mechanism of vertical accountability, which can be leveraged against a potentially recalcitrant ofﬁcial. Second, and more importantly, judges on my account are critical players in the development of their own power. They do not simply accept the exogenous conditions that ultimately limit their authority, as they would under a veto player account. And they do not gain authority simply because leaders delegate formal powers to them in a process of judicial reform. The book began by suggesting that the conditions necessary for the exercise of judicial power under a public enforcement mechanism, transparency and legitimacy, are in some ways endogenous to judicial– government relations. The evidence summarized in Chapters 3 and 4 suggests that the Mexican Supreme Court, during a period of democratization, sought to ensure the transparency of its decisions by creating media coverage. Its public relations ofﬁce waged a battle for “good” coverage, attempting to develop a narrative about the court that is conducive to enhancing its legitimacy. It also promoted the outcomes of particular cases, especially when it appeared that the media was unlikely to cover a decision. More generally, high courts around the world maintain active relationships with their domestic media sources. Judges advocate publicly for compliance and argue for a broad public commitment to the rule of law. In this sense, the book suggests that democratization should be generally conducive to the development of powerful courts, which can promote the rule of law.

limits on judicial power under democracy Lest we get carried away with optimism about the normatively appealing consequences of democratization, it is useful to confront an alternative account. In the late 1990s, Fareed Zakaria (1997) suggested that the world was observing the rise of illiberal democracies, states with democratic electoral institutions but limited protections for civil and human rights. In fact, he placed the proportion of illiberal democratizing states at roughly one-half. Zakaria’s observation raises a fair question about the basic link between democracy and the rule of law. Helmke and Rosenbluth (2008) develop and sharpen the observation further, suggesting a number of reasons why we should not necessarily expect democracies to produce judicial independence. Taken to an extreme, the argument in this book might appear to be in tension with those more sobering accounts. If the solution to weak courts requires transparency and judicial legitimacy, and if judges control those

200

Relationships between Transparency and Legitimacy

table 7.1. Judicial Independence in Democratic States, 1990–2004

No independence Partial independence Full independence

Tate & Keith

Cingranelli & Richards

% 17 23 60

% 5 35 60

Source: Tate and Keith (2007); Cingranelli and Richards (2008); Cingranelli and Richards (2009).

conditions, then judicial power can be produced without trouble. Also, if democracy provides the opportunity to construct those conditions, then this would suggest that all democracies in the world should possess powerful courts. But as Helmke and Rosenbluth note, this is simply at odds with the empirical evidence. As the data in Table 7.1 show, between 1990 and 2004, roughly 40 percent of states with a Polity IV score above 6, the standard cutoff for democratic regimes, lacked a fully independent judiciary. In fact, the Tate and Keith score reports that 17 percent of democracies had a judiciary that was completely subjugated to political ofﬁcials. In 2004, the last available year in the data set, the percentage was 23. This group of states included the Dominican Republic, Honduras, Nicaragua, Panama, Venezuela, Ecuador, Paraguay, the Slovak Republic, Albania, Croatia, Bulgaria, Moldova, Romania, Russia, Ukraine, Georgia, Mali, Senegal, Benin, Ghana, Kenya, Mozambique, Madagascar, and the Philippines. In light of these data, it is useful to stress that the argument of the book does not suggest that judges will control perfectly the transparency and legitimacy conditions. Real political interests and the nature of independent reporting constrain what judges can do. Consequently, judicial power, though impacted by the choices judges make, is still fundamentally constrained. Critically, this is true even under democratic institutions. The book suggests that rather than turning courts into powerful political actors overnight, democratization presents judges with opportunities to construct their power. It presents new judicial allies and new methods for reaching and coordinating them. But these opportunities are limited in a number of ways. The empirical evidence from the Mexican case is suggestive. In the wake of the Zedillo reforms, the Supreme Court crafted a signiﬁcant, and generally neutral, coverage. Through constant contact and good management

Democratic States and the Development of Judicial Power

201

of access, it helped construct a Supreme Court beat, which it could use to increase information about its decisions. And through the careful review of its coverage, the court could recognize stories that threatened its image and was able to respond through its press ofﬁce. This was a signiﬁcant achievement. Still, we cannot say that the court ran La Jornada’s or Reforma’s editorial rooms from its public relations ofﬁce. It was not able to create coverage with certainty. Its decisions were reported, but its ability to determine fully which ones received attention was imperfect. Speciﬁc empirical results in the constitutional decision making and compliance sections of Chapter 4 paint a similar picture. In general, noncompliance incidents involve relatively nonsalient policies that are dealt with in political contexts that lack almost all transparency. In the only case in which the Supreme Court threatened to remove from ofﬁce a federal ofﬁcial, it promoted the outcome vigorously and received substantial media coverage. This is all consistent with a model of judicial power in which the transparency condition is under judicial control. On the other hand, prior media coverage is associated with the probability that the court strikes down insufﬁciently salient public policies, and this is what we would anticipate if the court can only control its coverage partially. Given the research design, it is important not to generalize signiﬁcantly beyond the Mexican case on this point. That said, Vanberg’s study of the German federal constitutional court also suggests that judicial beliefs about the transparency of case outcomes inﬂuence constitutional decisions. This result is consistent with a case promotion model in which media control is only partial. Although the research designs are different in some ways, it is nevertheless surprising and encouraging to ﬁnd such similar results in countries as different as Germany and Mexico. Even if high courts could control transparency perfectly, the model in Chapter 2 does not suggest that judges would be able to free themselves from political pressure entirely. Public support for courts is not always sufﬁcient to induce compliance and, by implication, avoid strategic deference. Again, the evidence in Chapter 4 is helpful. It is important to recognize that political salience inﬂuences Mexican Supreme Court decisions independently of the coverage surrounding the case and the fragmentation of the federal government. What is more, the third section of the book suggests that voters can learn about the impartiality of their high courts in ways that can create a tension between the transparency and legitimacy conditions. In political contexts in which courts are relatively free to resolve cases sincerely, increasing information about courts is likely associated with

202

Relationships between Transparency and Legitimacy

strong beliefs in judicial legitimacy. However, in political contexts in which courts are constrained, increasing information about them can undermine legitimacy. These results suggest a problem for high courts in unfriendly political contexts, especially if the media cannot be controlled completely. Although transparency is needed for the reasons discussed by Vanberg, producing transparency means constructing a judicial beat that will run stories independently at times. The consequence is that, although judicial public relations can advance the transparency condition, it may expose courts to public scrutiny in situations where obscurity would better advance legitimacy. This problem suggests a difﬁcult trade-off for courts under constraints: pursue transparency at the risk of reducing legitimacy or accept obscurity and a corresponding lack of inﬂuence. Strategic public relations, the pursuit of selective coverage, is consistent with a court negotiating this trade-off, but at some point, judges likely have to choose between certain irrelevance and a risk of damaging the judicial image. A related limitation on the ability of judges to create their own power under democratic conditions concerns the idea that new courts can build their authority over time through a strategy of strategically deferential decision making. If individuals learn from case outcomes, and if they can recognize instances of prudence (or at least if clever reporters identify these instances for them), then it is not clear that prudence builds power. If this is correct, then democratic transitions, in which courts confront considerable constraints, essentially present judges with a sort of power trap: pursue a strategy of strategic deference and undermine beliefs in judicial impartiality or boldly attempt to constrain a state that is not incentivized to comply and risk a purge. The bottom line is that the process of democratization provides opportunities for judges to construct their own power; however, judges are likely to confront signiﬁcant obstacles along the way.

additional research Future scholarship might address a number of questions raised by the argument in this book and left open by the empirical analysis. The international data in Chapter 1 suggest that many courts around the world selectively promote case outcomes, but others do not. Some promote nothing and others promote everything. It is not clear immediately why courts adopt these different general strategies. Still, it seems likely that the basic trade-offs regarding transparency, judicial legitimacy, and power could help frame an answer. For example, it may be that a court, such as the

Democratic States and the Development of Judicial Power

203

Supreme Court of Argentina, that is under consistent political pressure is served better by allowing its coverage to be developed by its national media than by trying to manipulate it through the strategic highlighting of particular resolutions. Indeed, the only goal that may be accomplished by heightening awareness of its decisions is further reinforcing the general sense that the Supreme Court is politicized. In contrast, the Sala IV of Costa Rica largely operates free from external interference (Wilson and Rodríguez 2006), but it resolves an overwhelming number of individual constitutional complaints each year, many of which affect bureaucratic budgets and could, in light of their quantity, invite occasional noncompliance. A court in such a situation might have a strong incentive to develop a politically sensitive public relations strategy. Of course, a fully developed model of exactly how judges should evaluate these trade-offs in particular cases is a necessary ﬁrst step toward explaining the variety of general public relations approaches observable around the world. Such a model could identify precisely the long-run consequences for legitimacy of a strategy of strategic deference. Although the model developed in Chapter 5 suggests a negative effect, this claim requires further analysis. Such an analysis would confront questions of the following sort: Do people remember cases of strategic deference, and if so, for how long? What happens to beliefs after a purge? How ﬂexible are these beliefs to membership change? For example, though the Pakistani Supreme Court was thoroughly purged in 2007, by March 2009, past members were returned to the bench, including Chief Justice Iﬁtkhar Muhammad Chaudhry. In such a case, would the boldness the court exempliﬁed prior to being purged translate into legitimacy even though the intervening court was considered a mere extension of the government? Finally, it would be useful to evaluate speciﬁcally whether courts are better off using strategic deference in combination with a strategic approach to public relations or allowing media coverage simply to develop on its own. No matter what the theoretical results, it is clear that our arguments about the development of judicial power, under democracy or autocracy, should be submitted to stronger tests than they are now. The standard approach in the literature involves a detailed historical account of decision patterns (e.g., Ginsburg 2003; Moustafa 2007). Although this type of analysis is highly suggestive, it is not able to distinguish fully between competing accounts. At a minimum, there is no careful way of estimating belief change in such a design. In this case, it seems that judicial politics could borrow from the law and psychology approach and turn to experimental design. In such a format, we could test directly whether strategic

204

Relationships between Transparency and Legitimacy

deference produces a norm of compliance and ultimately expands judicial power or whether it simply produces a belief that courts are not impartial and therefore illegitimate. Experiments would not solve all our empirical problems, and we would still wish information on actual political contexts, but they would provide new and needed information about the process by which courts build power. Finally, although there is systematic evidence in Mexico that the Supreme Court has had some impact on the cases the media covers, and although there is some evidence that the court has been able to inﬂuence the substance of its coverage, we are far from knowing precisely how well courts can inﬂuence substantive information in Mexico or beyond. Davis (1994) suggests that the U.S. Supreme Court successfully shaped its media coverage, especially in the 1960s and 1970s, but this is but one additional state. The account in this book is judge-centric, and we know that judges have allies outside of the judiciary, allies with political agendas that involve increasing judicial authority (Epp 1998; Moustafa 2007). Perhaps judges are the crucial actors in the process of promoting the judiciary, but it is possible that the real movers in promoting courts around the world are interest groups, domestic and international, which have the resources and time to present the media with information about courts, but also to publicize instances of noncompliance and to come to the defense of courts in the event of attacks. This would not suggest that judges are incapable of constructing their power, but it would provide additional information on precisely how they go about it.

References

Aguinaco Alemán, José Vicente. 1998. El Nuevo Poder Judicial de la Federación. México: Suprema Corte de la Nación. 1997a. La Toga y la Palabra. México: Dirección General de Comunicación Social. Suprema Corte de Jusiticia. 1997b. El Poder Judicial de la Federación Fortalece su Independencia. México: Dirección General de Comunicación Social, Suprema Corte de Jusiticia. 1997c. 75 Decisiones Relevantes de la Suprema Corte de Justicia de la Nacion (Novena Epoca). México: Dirección General de Comunicación Social, Suprema Corte de Jusiticia. Alter, Karen. 1998. “Who are the ‘Masters of the Treaty’: European Governments and the European Court of Justice.” International Organization 52: 121–47. Altheide, D. L., and Robert P. Snow. 1979. Media Logic. Beverly Hills, CA: Sage Publications. Alvarez, Mike, Jose Antonio Cheibub, Fernando Limongi, and Adam Przeworski. 1997. “Classifying Political Regimes.” Studies in Comparative International Development 31:236. Apodaca, Clair. 2004. “The Rule of Law and Human Rights.” Judicature 87: 292–9. Andrews, Josephine T., and Gabriella R. Montinola. 2004. “Veto Players and the Rule of Law in Emerging Democracies.” Comparative Political Studies 37:55–87. Arnold, Linda. 1996. Política y Justicia: La Suprema Corte Mexicana (1824– 1855). Trans. José Luis Soberatnes Fernández and Julián Bunster. México: Universidad Autónoma de México. 1988. Bureaucracy and Bureaucrats in Mexico City, 1742–1835. Tucson, AZ: University of Arizona Press. Badinter, Robert, and Stephen Breyer. 2004. Judges in Contemporary Democracy. New York: New York University Press.

205

206

References

Baird, Vanessa A. 2004. “The Effect of Politically Salient Decisions on the US Supreme Court’s Agenda.” Journal of Politics 66:755–72. Baird, Vanessa A., and Amy Gangl. 2006. “Shattering the Myth of Legality: The Impact of the Media’s Framing of Supreme Court Procedures on Perceptions of Fairness.” Political Psychology 27:597–614. Baker, Richard D. 1971. Judicial Review in Mexico. A Study of the Amparo Suit. Austin, TX: University of Texas Press. Barber, Sotirios A. 1993. The Constitution of Judicial Power. Baltimore, MD: Johns Hopkins University Press. Barro, Robert J. 1997. Determinants of Economic Growth: A Cross-Country Empirical Study. Cambridge, MA: MIT Press. Barros, Robert. 2002. Constitutionalism and Dictatorship: Pinochet, the Junta, and the 1980 Constitution. New York: Cambridge University Press. Baum, Lawrence. 2006. Judges and Their Audiences: A Perspective on Judicial Behavior. Princeton, NJ: Princeton University Press. Beard, Charles A. 1962. “The Supreme Court-Usurper or Grantee.” In The Courts: A Reader in the Judicial Process, ed. Robert G. Scigliano. Boston, MA: Little Brown. Billikopf, David Marshall. 1973. The Exercise of Judicial Power: 1789–1864. New York: Vantage Press. Boudin, Louis B. 1962. “Government by Judiciary.” In The Courts: A Reader in the Judicial Process, ed. Robert G. Scigliano. Boston, MA: Little Brown. Brambor, Thomas, William Roberts Clark, and Matt Golder. 2006. “Understanding Interaction Models: Improving Empirical Analyses.” Political Analysis 4:63–82. Brinks, Daniel. 2005. “Judicial Reform and Independence in Brazil and Argentina: The Beginning of a New Millennium?” Texas International Law Journal 40:595–622. Bueno de Mesquita, Bruce, Alastair Smith, Randolph M. Siverson, and James D. Morrow. 2003. The Logic of Political Survival. Cambridge, MA: The MIT Press. Burgoa, O. Ignaico. 1998. El Juicio de Amparo. México: Editorial Porrúa. 1984. El Juicio de Amparo, 21st ed. México: Porrúa. Burley, Anne-Marie, and Walter Mattli. 1993. “Europe Before the Court: A Political Theory of Legal Integration.” International Organization 47: 41–76. Cabrera Acevedo, Lucio. 1988. La Suprema Corte de Justicia. La República y el Imperio. México: Poder Judicial Federal. Caldeira, Gregory. 1986. “Neither the Purse Nor the Sword: Dynamics of Public Conﬁdence in the Supreme Court.” The American Political Science Review 80:1209–26. Caldeira, Gregory A., and James L. Gibson. 1995. “The Legitimacy of the Court of Justice in the European Union: Models of Institutional Support.” The American Political Science Review 89:356–76. 1992. “The Etiology of Public Support for the Supreme Court.” American Journal of Political Science 36:635–64.

References

207

Cameron, Charles M. 2002. “Judicial Independence: How Can You Tell It When You See It?” In Judicial Independence at the Crossroads: An Interdisciplinary Approach, eds. Stephen B. Burbank and Barry Friedman. Thousand Oaks, CA: Sage Publications. Canon, Bradley C., and Charles A. Johnson. 1999. Judicial Politics: Implementation and Impact. Washington, DC: CQ Press. Carothers, Thomas. 2007. “How Democracies Emerge: The “Sequencing” Fallacy.” Journal of Democracy 18:12–27. Carpizo, Jorge. 1995. “Reformas Constitucionales al Poder Judicial Federal y a la Jurisdicción Constitucional del 31 de diciembre de 1994.” Boletín Mexicano de Derecho Comparado 28. Carranco Zúñiga, Joel. 2000. Poder Judicial. México: Editorial Porrúa. Carrubba, Clifford J. 2009. “A Model of the Endogenous Development of Judicial Institutions in Federal and International Systems.” Journal of Politics 71: 1–15. 2005. “Courts and Compliance in International Regulatory Regimes.” Journal of Politics 67:669–89. 2003. “The European Court of Justice, Democracy and Enlargement.” European Union Politics 4:75–100. Casey, Gregory. 1974. “The Supreme Court and Myth: An Empirical Investigation.” Law and Society Review 8:385–419. Chávez Padrón, Martha. 1990. Evolución del Juicio de Amparo y del Poder Judicial Federal Mexicano. México: Porrúa. Cingranelli, David L., and David L. Richards. 2008. “The CingranelliRichards (CIRI) Human Rights Data Project Coding Manual.” Available at: http://ciri.binghamton.edu/documentation/ciri_coding_guide.pdf Clague, Christopher, Philip Keefer, Stephen Knack, and Mancur Olson. 1999. “Contract-Intensive Money: Contract Enforcement, Property Rights, and Economic Performance.” Journal of Economic Growth 4:185–211. Clark, Thomas. Forthcoming, 2010. “The Limits of Judicial Independence.” New York: Cambridge University Press. Clinton, Robert Lowry. 1994. “Game Theory, Legal History, and the Origins of Judicial Review: A Revisionist Analysis of Marbury v. Madison.” American Journal of Political Science 38: 285–302. Cohn, Ellen S., Susan O. White, and Joseph Sanders. 2000. “Distributive and Procedural Justice in Seven Nations.” Law and Human Behavior 24:553–80. Cook, Timothy E. 1996. “The Negotiation of Newsworthiness.” In The Psychology of Political Communication, ed. Ann N. Crigler. Ann Arbor, MI: University of Michigan Press. Couso, Javier A. 2004. “The Politics of Judicial Review in Chile in the Era of Democratic Transition.” In Democratization and the Judiciary: The Accountability Function of Courts in New Democracies, eds. Siri Gloppen, Roberto Gargarella, and Elin Skaar. London: Frank Cass Publishers. Cornelius, Wayne A., Judity Gentleman, and Peter H. Smith, eds. 1989. Mexico’s Alternative Political Futures. La Jolla, CA: Center for U.S.–Mexican Studies, University of California–San Diego.

208

References

Costeloe, Michael P. 1975. La Primera República Federal de México, 1825–1835: Un Estudio de los Partidos Políticos en el México Independiente. México: Fondo de Cultura Económica. Creyke, Robin, and John McMillan. 2004. “The Operation of Judicial Review in Australia.” In Judicial Review and Bureaucratic Impact: International and Interdisciplinary Perspectives, eds. Marc Hertogh and Simon Halliday. Cambridge: Cambridge University Press. Dahl, Robert A. 1957. “Decision-Making in a Democracy: The Supreme Court as a National Policy-Maker.” Journal of Public Law 6:279–95. 1963. Politics, Economics, and Welfare Planning. New York: Harper Collins. Davis, Richard. 1994. Decisions and Images: The Supreme Court and the Press. Englewood Cliffs, NJ: Prentice Hall. Díaz Barriga, Miguel, and Karen Kleiber. 1996. “The Press and Urban Conﬂict in Mexico City: A Case Study of Newspaper Reporting on Ecology and Urban Expansion in the Ajusco Region, 1982–1990.” Mexican Studies 12: 273–300. Domingo, Pilar. 2000. “Judicial Independence: The Politics of the Supreme Court in Mexico.” Journal of Latin American Studies 32:705–35. 1999a. “Judicial Independence and Judicial Reform in Latin America.” In The Self-Restraining State: Power and Accountability in New Democracies, eds. Andreas Schedler, Larry Diamond, and Marc F. Plattner. Boulder, CO: Lynne Rienner Publishers. 1999b. “Rule of Law, Citizenship and Access to Justice in Mexico.” Mexican Studies 15:151–91. Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper. Dubek, Leslie B. 2007. “Understanding ‘Judicial Lockjaw’: The Debate over Extrajudicial Activity.” New York University Law Review 82:569–601. Easterbrook, Frank H. 1987. “Allocating Antitrust Decision-making Tasks.” Georgetown Law Journal 76:305–20. Eisenstadt, Todd A. 2004. Courting Democracy in Mexico: Party Strategies and Electoral Institutions. Cambridge: Cambridge University Press. Epp, Charles R. 1998. The Rights Revolution: Lawyers, Activists and Supreme Courts in Comparative Perspective. Chicago, IL: University of Chicago Press. Epstein, Lee, and Jack Knight. 1998. The Choices Justices Make. Washington DC: CQ Press. 1996. “On the Struggle for Judicial Supremacy.” Law and Society Review 30:87–120. Falaschetti, Dino, and Gary Miller. 2000. “Constraining Leviathan: Moral Hazard and Credible Commitment in Constitutional Design.” Paper presented at the 2000 Meeting of the Midwest Political Science Association, Chicago, IL. Ferejohn, John. 1999. “Independent judges, dependent judiciary: Explaining judicial independence.” Southern California Law Review 72:353–84. Ferejohn, John, and Barry R. Weingast. 1992. “A positive theory of statutory interpretation.” International Review of Law and Economics 12:263–79. Ferejohn, John, and Charles Shipan. 1990. “Congressional Inﬂuence on Bureaucracy.” Journal of Law, Economics and Organization 6:1–21.

References

209

Ferrer Mac-Gregor, Eduardo. 2000. Acción Constitucional de Amparo en México y España. México: Editorial Porrúa. Finkel, Jodi S. 2008. Judicial Reform as Political Insurance: Argentina, Peru, and Mexico in the 1990s. South Bend, IN: University of Notre Dame Press. 2005. “Judicial Reform as Insurance Policy: Mexico in the 1990s.” Latin American Politics and Society 47: 87–113. Fiss, Owen. 1994. “The Right Degree of Independence.” In Transition to Democracy in Latin America: The Role of the Judiciary, ed. Irwin P. Stotzky. Boulder, CO: Westview Press. Fix-Fierro, Héctor. 1999. “Jueces y corrupcion.” Bien Común y Gobierno 55: 11–16. 1998a. “Judicial Reform and The Supreme Court of Mexico: The Trajectory of Three Years.” U.S.–Mexico Law Journal 6:1–21. 1998b. “Poder Judicial.” In Transiciones y Diseños Institucionales, ed. María Refugio González and Sergio López Ayllón. Distrito Federal: México. Fix-Zamudio, Héctor. 1987. “La Suprema Corte de Justicia Como Tribunal Constitutional,” In Las Nuevas Bases Constitucionales y Legales del Sistema Judicial Mexicano: La Reforma Judicial 1986–1987, ed. Jorge Madrazo. México: Porrúa. Fix-Zamudio, Héctor, and José Ramon Cossío Díaz. 1995. El Poder Judicial en el Ordenamiento Mexicano. México: Fondo de Cultura Económica. Flowers, Julianne, Audrey A. Haynes, and Michael H. Crespin. 2003. “The Media, the Campaign and the Message.” American Journal of Political Science 47:259–73. Franklin, Charles H., and Liane C. Kosaki. 1989. “Republican School Master: The U.S. Supreme Court, Public Opinion, and Abortion.” The American Political Science Review 83:751–71. Freedom House. Various years. Freedom in the World. New York: Freedom House. Friedman, Barry. 2000. “The History of the Countermajoritarion Difﬁculty, Part Four: Law’s Politics.” University of Pennsylvania Law Review 148: 974–1065. Friedrich, Robert J. 1982. “In Defense of Multiplicative Terms in Multiple Regression Equations.” American Journal of Political Science 26:797–833. Gibson, James L. 2008. “Challenges to the Impartiality of State Supreme Courts: Legitimacy Theory and ‘New-Style’ Judicial Campaigns.” American Political Science Review. Forthcoming. 1989. “Understandings of Justice: Institutional Legitimacy, Procedural Justice, and Political Tolerance.” Law and Society Review 3:469–96. Gibson, James L., and Gregory A. Caldeira. 2009. Citizens, Courts, and Conﬁrmations: Positivity Theory and the Judgments of the American People. Princeton, NJ: Princeton University Press. 1998. “Changes in the Legitimacy of the European Court of Justice: A PostMaastrict Analysis.” British Journal of Political Science 28:63–91. Gibson, James L., Gregory A. Caldeira, and Vanessa A. Baird. 1998. “On the Legitimacy of National High Courts.” American Political Science Review 92:343–58.

210

References

Ginsburg, Tom. 2003. Judicial Review in New Democracies: Constitutional Courts in Asian Cases. Cambridge: Cambridge University Press. Gongora Pimentel. 2000. “El Estado de Derecho como Promotor del Desarrollo Economico Nacional.” Remarks before the XI Convención del Mercado de Valores, October 17, 2000. 2000a. “Acerquemos la Justicia a la Sociedad.” In Once Voces, ed. Alberto Aragón Bolado. México: Suprema Corte de Justicia de la Nación. González Casanova, Pablo. 1967. La democracia en México. México: Era. González Cosío, Arturo. 1985. El Juicio de Amparo, 2nd ed. Distrito Federal: México. Gonzalez Oropeza, Manuel. 1995. “El Desafío de la Justicia, La Administración de Justicia y el Estado de Derecho en México.” Lex, Torreón, Coahuila: México. Graber, Doris A. 2005. Mass Media and American Politics, 7th ed. Washington DC: CQ Press. Greene, William H. 1998. “Gender Economics Courses in Liberal Arts Colleges: Further Results.” Journal of Economic Education 29:291–300. Gunther, Gerald. 1991. Constitutional Law, 12th ed. New York: Foundation Press, Inc. Hale, Charles A. 1968. Mexican Liberalism in the Age of Mora, 1821–1853. New Haven, CT: Yale University Press. Hall, Melina Gann. 1992. “Electoral Politics and Strategic Voting in State Supreme Courts.” The Journal of Politics 54:427–46. Harvey, Anna, and Barry Friedman. 2006. “Pulling Punches: Congressional Constraints on the Supreme Court’s Constitutional Rulings, 1987–2000.” Legislative Studies Quarterly 31:533. Hathaway, Oona. 2007. “Why Do Countries Commit to Human Rights Treaties?” Journal of Conﬂict Resolution 8:588–621. Hausmaninger, Herbert. 1995. “Toward a ‘New’ Russian Constitutional Court.” Cornell International Law Journal 28:305. Helmke, Gretchen. 2005. Courts Under Constraints: Judges, Generals, and Presidents in Argentina. New York: Cambridge University Press. 2002. “The Logic of Strategic Defection: Court-Executive Relations in Argentina Under Dictatorship and Democracy.” American Political Science Review 96:291–304. Helmke, Gretchen, and Frances Rosenbluth. 2008. “Regimes and the Rule of Law: Judicial Independence in Comparative Perspective.” Typescript. University of Rochester. Helmke, Gretchen and Frances Rosenbluth. 2009. “Regimes and the Rule of Law: Judicial Independence in Comparative Perspective.” Annual Review of Political Science 12:345–66. Herron, Erik S., and Kirk A. Randazzo. 2003. “The Relationship Between Independence and Judicial Review in Post-Communist Courts.” Journal of Politics 65:422–38. Hibbing, John R., and Elizabeth Theiss-Morse. 1995. Congress as Public Enemy: Public Attitudes Toward American Political Institutions. Cambridge: Cambridge University Press.

References

211

Hilbink, Lisa. 2007. “Judges Beyond Politics.” In Democracy and Dictatorship: Lessons from Chile. Cambridge: Cambridge University Press. Hillygus, D. Sunshine, and Simon Jackman. 2003. “Voter Decision Making in Election 2000: Campaign Effects, Partisan Activation, and the Clinton Legacy.” American Journal of Political Science 47:583–96. Hirshl, Ran. 2001. “The Political Origins of Judicial Empowerment Through Constitutionalization: Lessons from Israel’s Constitutional Revolution.” Comparative Politics 33:3. Howard, Richard M. and Henry F. Carey. 2004. “Is an Independent Judiciary Necessary for Democracy?” Judicature 87:284–90. Iaryczower, Matías, Pablo T. Spiller, and Mariano Tommasi. 2002. “Judicial Decision-Making in Unstable Environments, Argentina 1935–1998.” American Journal of Political Science 46:699–716. International Country Risk Guide. 2004. Political Risk Services, East Syracuse, NY. Jackson, Vicki C., and Mark Tushnet. 2006. Comparative Constitutional Law, 2nd ed. New York: Foundation Press. Johnson, Herbert Alan. 1996. The Chief Justiceship of John Marshall, 1801– 1835. Columbia: University of South Carolina Press. Johnson, Richard M. 1967. The Dynamics of Compliance: Supreme Court Decision-Making from a New Perspective. Evanston, IL: Northwestern University Press. Johnson, Timothy R., and Andrew D. Martin. 1998. “The Public’s Conditional Response to Supreme Court Decisions.” American Political Science Review 92:299–310. Kapiszewski, Diana and Matthew M. Talyor. 2008. “Doing Courts Justice? Studying Judicial Politics in Latin America.” Perspectives in Politics 6:741–67. Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2007. “Governance Matters VI: Governance Indicators for 1996–2006.” World Bank Policy Research Working Paper No. 4280. Kernell, Samuel. 1993. Going Public: New Strategies of Presidential Leadership, 2nd ed. Washington, DC: CQ Press. Klesner, Joseph L., and Chappell Lawson. 2001. “Adiós to the PRI? Voter Turnout in Mexico’s Political Transition.” Mexican Studies 17: 17–39. Kommers, Donald. 1997. The Constitutional Jurisprudence of the Federal Republic of Germany, 2nd ed. Durham, NC: Duke University Press. Landes, William M., and Richard A. Posner. 1975. “Independent Judiciary in an Interest-Group Perspective.” Journal of Law and Economics 18: 875–901. Lasser, William. 1988. The Limits of Judicial Power: The Supreme Court in American Politics. Chapel Hill, NC: University of North Carolina Press. Laver, Michael, and Norman Schoﬁeld. 1990. Multiparty Government: the Politics of Coalition in Europe. Oxford: Oxford University Press; repr. 1998, University of Michigan Press. Lawson, Chappell. 2002. Building the Fourth Estate: Democratization and the Rise of a Free Press in Mexico. Berkeley, CA: University of California Press.

212

References

Linz, Juan, and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe. Baltimore, MD: Johns Hopkins University Press. Locke, John. [1698] 1952. The Second Treatise of Government, ed. Thomas Peardon. Indianapolis, IN: Bobbs-Merrill. Mackey, Michael W. 1998. “Informe de Michael W. Mackey en la evaluación integral de las operaciones y funciones del Fondo Bancario de Proteccion al Ahorro, FOBAPROA y la calidad de supervisión de los programas del FOBAPROA de 1995 a 1998 (Julio de 1999).” Gaceta Parlamentaria, Cámara de Diputados. Macune, Charles W. 1978. El Estado de México y la Federación Mexicana, 1823– 1835. México: Fondo de Cultura Económica. Madison, James, Alexander Hamilton, and John Jay. [1787] 1961. The Federalist Papers. Introduction by Clinton Rossiter. New York: Mentor-New American Library. Magaloni, Beatriz, and Arianna Sánchez. 2003. “Empowering Courts as Constitutional Veto Players: Presidential Delegation and the New Mexican Supreme Court.” Paper presented at the 2001 Annual Meeting of the American Political Science Association, August 30–September 2, 2001, San Francisco. Maltzman, Forrest, James F. Spriggs, and Paul J. Walbeck. 2000. Crafting Law on the Supreme Court: The Collegial Game. Cambridge: Cambridge University Press. Marshall, Monty G., and Keith Jaggers. 2007. Polity IV Dataset Version 2006. Martin, Andrew D., and Kevin M. Quinn. 2007. “Assessing Preference Change on the U.S. Supreme Court.” Journal of Law, Economics and Organization 23:365–85. Mason, Alpheus Thomas. 1953. “Extra-Judicial Work for Judges: the Views of Chief Justice Stone.” Harvard Law Review 67:193–216. Masood, Salman. 2007a. “Musharraf Re-election Plan Disputed in Supreme Court.” New York Times. Available at: http://www.nytimes.com/2007/09/18/world/asia/18pakistan.html?partner= permalink&exprod=permalink. 2007b. “Suspension of Jurist Unleashes Furor against Musharraf.” New York Times. Available at: http://www.nytimes.com/2007/03/15/world/asia/15pakistan.html. 2007c. “World Brieﬁng: Asia: Pakistan: 6 Judges Resign to Protest Chief Justice’s Removal.” New York Times. Available at: http://query.nytimes.com/gst/fullpage.html?res=9D03E7DB1730F933A157 50C0A9619C8B63. 2007d. “World Brieﬁng: Asia: Pakistan: Judges Involuntarily Retired.” New York Times. Available at: http://query.nytimes.com/gst/fullpage.html?res=9503EFDD163DF935A357 51C1A9619C8B63. Mayhew, David R. 1974. Congress: The Electoral Connection. New Haven, CT: Yale University Press.

References

213

McGraw, Kathleen M. 1991. “Managing Blame: An Experimental Test of the Effects of Political Accounts.” American Political Science Review 85: 1133–58. McGuire, Kevin T. 1995. “Repeat Players in the Supreme Court: The Role of Experienced Lawyers in Litigation Success.” The Journal of Politics 57: 187–96. Meyer, Michael C., and William L. Sherman. 1991. The Course of Mexican History, 4th ed. New York: Oxford University Press. Middlebrook, Kevin. 1995. The Paradox of Revolution: Labor, the State and Authoritarianism in Mexico. Baltimore, MD: Johns Hopkins University Press. Milner, Neil A. 1971. The Court and Local Law Enforcement: The Impact of Miranda. Beverly Hills, CA: Sage Publications. Mondak, Jeffrey A. 1993. “Institutional Legitimacy and Procedural Justice: Reexamining the Question of Causality.” Law and Society Review 27: 599–608. 1991. “Substantive and Procedural Aspects of Supreme Court Decisions As Determinants of Institutional Approval.” American Politics Quarterly 19: 174–88. Montesquieu, Baron de. [1748] 1900. The Spirit of Laws. New York: Collier. Moustafa, Tamir. 2007. The Struggle for Constitutional Power: Law, Politics, and Economic Development in Egypt. New York: Cambridge University Press. Murphy, Walter F. 1964. Elements of Judicial Strategy. Chicago, IL: University of Chicago Press. Murphy, Walter F., and Joseph Tanenhaus. 1968. “Public Opinion and the United States Supreme Court.” Law and Society Review 2:357–82. Nagel, Jack H. 1975. The Descriptive Analysis of Power. New Haven, CT: Yale University Press. 1968. “Some Questions About the Concept of Power.” Behavioral Science 13:129–37. Niemeyer, E. V., Jr. 1974. Revolution at Querétaro: The Mexican Constitutional Convention of 1916–17. Austin, TX: University of Texas Press. North, Douglass. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press. North, Douglass C., and Barry R. Weingast. 1989. “Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in SeventeenthCentury England.” The Journal of Economic History 49: 803–32. North, Douglass C., William Summerhill, and Barry R. Weingast. 2000. “Order, Disorder and Economic Change: Latin America versus North America.” In Governing for Prosperity, ed. Bruce Bueno de Mesquita and Hilton Root. New Haven: Yale University Press. O’Donnell, Guillermo. 1999. “Horizontal Accountability in New Democracies.” In The Self-Restraining State: Power and Accountability in New Democracies, eds. Andreas Schedler, Larry Diamond, and Marc F. Plattner. Boulder, CO: Lynne Rienner Publishers.

214

References

Osborne, Martin, and Ariel Rubenstein. 1994. A Course in Game Theory. Cambridge, MA: MIT Press. Powell, G. Bingham. 2000. Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven, CT: Yale University Press. Powell, Emilia Justyna, and Jeffrey K. Staton. 2009. “Domestic Judicial Institutions and Human Rights Treaty Violation.” International Studies Quarterly 53:149–74. Rabasa, Oscar. 1982. El Derecho Angloamericano (Estudio Expositivo y Comparado del “Common Law”), 2nd ed. México: Porrúa. Ramseyer, J. Mark, and Eric Rasmusen. 2003. Measuring Judicial Independence. Chicago, IL: University of Chicago Press. 2001. “Why are Japanese Judges so Conservative in Politically Charged Cases.” American Political Science Review 2:331–44. Ríos-Figueroa, Julio. 2007. “The Emergence of an Effective Judiciary in Mexico, 1994–2002.” Latin American Politics and Society 49:31–57. 2006. Judicial Independence: Deﬁnition, Measurement, and Its Effects on Corruption: An Analysis of Latin America. PhD diss., Department of Politics, New York University. Ríos-Figueroa, Julio, and Jeffrey K. Staton. 2008. “Unpacking the Rule of Law: A Review of Judicial Independence Measures.” Committee on Concepts and Measures Working Paper Series, International Political Science Association. Available at: http://www.concepts-methods.org. Rogers, James R. 2001. “Information and Judicial Review: A Signaling Game of Legislative-Judicial Interaction.” American Journal of Political Science 45: 84–99. Rosenberg, Gerald. 1991. The Hollow Hope: Can Courts Bring About Social Change. Chicago, IL: University of Chicago Press. Rubio, Luis, Beatriz Magaloni, and Edna Jaime. 1994. A la puerta de la ley: El Estado de derecho en Mexico. Mexico DF: Cal y Arena. Salgado, Eréndia. 2001. Poderes en Conﬂicto. México: Suprema Corte de Justicia de la Nación. Sánchez Cordero de García Villegas, Olga. 2000. “La Independencia Judicial en México: Apuntes Sobre una Realidad Conquistada por los Jueces Mexicanos.” Available at: http://www.scjn.gob.mx/Ministros/oscgv/Conf/Conf-011.htm. 1997. “Hacia un Auténtico Sistema de Impartición de Justicia en México.” In La Toga y la palabra. México: Dirección General de Comunicación Social. Suprema Corte de Jusiticia. Scheb, II, John M., and William Lyons. 2000. “The Myth of Legality and Public Evaluation of the Supreme Court.” Social Science Quarterly 81:929–40. Schwarz, Carl. 1973. “Judges Under the Shadow: Judicial Independence in the United States and Mexico.” California Western International Law Journal 3:261–321. Segal, Jeffrey A. 1997. “Separations-of-Powers Games in the Positive Theory of Congress and Courts.” American Political Science Review 91:28–44. Segal, Jeffrey A., and Harold J. Spaeth. 2002. The Supreme Court and the Attitudinal Model Revisited. Cambridge: Cambridge University Press.

References

215

1993. The Supreme Court and the Attitudinal Model. Cambridge: Cambridge University Press. Shapiro, Martin. 2004. “Judicial Review in Developed Democracies.” In Democratization and the Judiciary: The Accountability Function of Courts in New Democracies, eds. Siri Gloppen, Roberto Gargarella, and Elin Skaar. London: Frank Cass Publishers. Sharlet, Robert. 1993. “Chief Justice as Judicial Politician.” East European Constitutional Review 2:32–37. Sheehan, Reginald S., William Mishler, and Donald R. Songer. 1992. “Ideology, Status, and the Differential Success of Direct Parties Before the Supreme Court.” The American Political Science Review 86:464–71. Shirk, David A. 2004. Mexico’s New Politics: The PAN and Democratic Change. Boulder, CO: Lynne Reinner Publishers. Silva Meza, Juan N. 2000a. “La Corte y la Defensa de la Constitución.” In Once Voces, ed. Alberto Aragón Bolado. México: Suprema Corte de Justicia de la Nación. 2000b. “La Conﬁanza Se Gana Día a Día.” Palabras pronuciadas por Ministro Juan N. Silva Meza, el 13 de enero de 2000, con motivo de la toma de protesta de jueces de Distrito y magistrados de Circuito. Colección Discursos, Número 11. Smithey, Shannon Ishiyama, and John Ishiyama. 2002. “Judicial Activism in Post-Communist Politics.” Law and Society Review 36:719–42. Soberantes Fernández, José Luis. 1992. “Notas Sobre el Origen del AmparoCasación en México.” Boletín Mexicano de Derecho Comparado. 63: 529–47. Sobocinski, Raymund A. 2006. “Adumbrations On Judicial Campaign Speech.” Idaho Law Review 43:193. Available at: SSRN: http://ssrn.com/abstract=892686. Spaeth, Harold J. 2004. “United States Supreme Court Judicial Database, 1953– 2003 Terms.” Available at: http://www.as.uky.edu/polisci/ulmerproject/. Spaeth, Harold J., David B. Meltz, Gregory J. Rathjen, and Michael V. Haselswerdt. 1972. “Is Justice Blind: An Empirical Investigation of a Normative Ideal.” Law and Society Review 7:119–38. Spriggs, James F. 1996. “The Supreme Court and Federal Administrative Agencies: A Resource-Based Theory and Analysis of Judicial Impact.” American Journal of Political Science 40:1122–51. Staton, Jeffrey K. 2007. “Lobbying for Judicial Reform: The Role of the Mexican Supreme Court in Institutional Selection.” In Reforming the Administration of Justice in Mexico, in eds. Wayne A. Cornelius and David A. Shirk. Norte Dame, IN: University of Notre Dame Press. 2006. “Constitutional Review and the Selective Promotion of Case Results.” American Journal of Political Science 50:98–112. 2004. “Judicial Policy Implementation in Mexico City and Mérida.” Comparative Politics 37:41–60. 2002. Judicial Activism and Public Authority Compliance. Unpublished PhD diss., Washington University in St. Louis.

216

References

Staton, Jeffrey K., and Georg Vanberg. 2008. “The Value of Vagueness: Delegation, Deﬁance, and Judicial Opinions.” American Journal of Political Science 58:504–19. Stephenson, Matthew C. 2004. “Court of Public Opinion: Government Accountability and Judicial Independence.” The Journal of Law, Economics, and Organization 20:379–99. Stokes, Susan C. 2001. Mandates and Democracy: Neoliberalism by Surprise in Latin America. New York: Cambridge University Press. Stone, Alec. 1992. The Birth of Judicial Politics in France: The Constitutional Council in Comparative Perspective. New York: Oxford University Press. Stone Sweet, Alec. 2000. Governing with Judges: Constitutional Politics in Europe. Oxford: Oxford University Press. Tate, C. Neal, and Linda Camp Keith. 2007. “Conceptualizing and Operationalizing Judicial Independence Globally.” Paper presented at the Annual Meeting of the American Political Science Association, August 30–September 1, 2007, Chicago. Tate, C. Neal, Stacia L. Haynie, Reginald Sheehan, and Donald R. Songer. 2006. “Collaborative Research: Extending a Multi-Country Database and Program of Research.” National Science Foundation, Law and Social Science Program, Basic Research Grant. Tate, C. Neal, and Torbjorn Vallinder. 1995. “The Global Expansion of Judicial Power: The Judicialization of Politics.” In The Global Expansion of Judicial Power, eds. C. Neal Tate and Torbjorn Vallinder. New York: New York University Press. Tena Ramírez, Felipe. 1957. Leyes Fundamentales de México, 1808–1957. México: Editorial Porrúa. Transparency International. 2007. “Global Corruption Report 2007: Corruption in Judicial Systems.” Transparency International, Berlin. Trochev, Alexei. 2002. “Implementing Russian Constitutional Court Decisions.” East European Constitutional Review 11:95–103. Tsebelis, George. 2002. Veto Players. How Political Institutions Work. New York: Russell Sage Foundation and Princeton University Press. Tyler, Tom R. 1990. Why People Follow the Law: Procedural Justice, Legitimacy, and Compliance. New Haven, CT: Yale University Press. 2009. “Law and Psychology.” In Oxford Handbook of Law and Politics, eds. Keith E. Whittington, R. Daniel Kelemen, and Gregory A. Caldeira. Oxford: Oxford University Press. Vanberg, Georg. 2005. The Politics of Constitutional Review in Germany. New York: Cambridge University Press. 2001. “Establishing Judicial Independence in West Germany: The Impact of Opinion Leadership and the Separation of Powers.” Comparative Politics 33:333–53. Volcansek, Mary L. 2000. Constitutional Politics in Italy: The Constitutional Court. New York: St. Martin’s Press. 1991. “Judicial Activism in Italy.” In Judicial Activism in Comparative Perspective, ed. Kenneth Holland. New York: St. Martin’s Press.

References

217

Workshop on Reporting of Court Proceedings by Media and Administration of Justice. Supreme Court of India. March 29–30, 2008, Vigyan Bhawan, New Delhi. Weingast, Barry. 1997. “The Political Foundations of Democracy and the Rule of Law.” American Political Science Review 91:245–63. Whittington, Keith E. 2005. “‘Interpose Your Friendly Hand’: Political Supports for the Exercise of Judicial Review by the United States Supreme Court.” American Political Science Review 99:583–96. Wilson, Bruce M., and Juan Carlos Rodríguez Cordero. 2006. “Legal Opportunity Structures and Social Movements: The Effects of Institutional Change on Costa Rican Politics.” Comparative Political Studies 39:325–51. Workshop on Reporting of Court Proceedings by Media and Administration of Justice. Supreme Court of India. March 29–30, 2008, Vigyan Bhawan, New Delhi. Wuhs, Steven T. 2008. Savage Democracy: Institutional Change and Party Development in Mexico. University Park, PA: Penn State University Press. Zakaria, Fareed. 1997. “The Rise of Illiberal Democracy.” Foreign Affairs 76: 22–43.

Index

Aguinaco Alemán, José Vicente 55, 61 Amparo, see Constitutional Review in Mexico Aranda, Jesús 57, 85 Argentine Supreme Court 149, 168, 203 Australia High Court 4 Austrian Constitutional Court 4 Awareness of Courts Measures of 141, 171 Relationship to judicial impartiality 129–131, 141–147, 148, 151–152 Relationship to judicial legitimacy 127–130, 147, 150, 153, 170, 172–186 Baird, Vanessa 24, 127–128, 170–171 Cabal Peniche, Carlos 65–66 Caldeira, Gregory 24, 127–128, 170–171 Cameron, Charles 9, 175 Canadian Supreme Court 4 Camacho Solís, Manuel 110 Carmargo, Jorge 61, 62 Carranco Zúñiga, Joel 58 Carrubba, Clifford J. 24, 190, 192–193 Case promotion 33–44 Costs of 36–41 Measures of 73 And non-compliance 90–91 See also public relations Casey, Gregory 133, 141 Chávez, Hugo 10, 131–132, 189 Chilean Supreme Court 149, 168 Columbia Constitutional Court 4 Compliance, see also Non-Compliance And case salience 191

Norm of compliance 191 And legitimacy 13 Constitutional Court of Russia 188 Constitutional review in Mexico Action of Unconstitutionality 48, 68–69, 74–75, 91–92, 106–114 Amparo 68–69, 74–75, 91–92, 93–96, 103–106, 108–111 Case assignment 111–112 Constitutional Controversy 48, 68–69, 74–75, 91–96, 106–107, 112–114 Written Opinions 112–114 DCS (see Ofﬁce of Social Communications) DCSJ (see Social Communications Ofﬁce at Federal Judicial Council) Deference, As strategic judicial behavior 124, 191 Deﬁance Costs 32, 101, 136, 139, 147 And awareness 170–186 And salience 137–139, 147 In theory 134–140 Measures of 168–169 Democracy Measures of 169 And the public enforcement mechanism 197–199 And judicial power 199–202 Democratization in Mexico 47–48 European Court of Justice 13 Federalist 78 9 Finkel, Jodi 48–49, 197 Frankfurter, Felix 6 Fox, Vicente 47, 53, 77, 148

219

220 German Federal Constitutional Court 10, 187, 190, 201 Gibson, James L. 24, 127–128, 170–171 Ginsburg, Tom 8, 190–192 Góngora Pimentel, Genaro, 53–54, 56, 61–62, 79, 89 Iaryczower, Matias 77, 89 Impartiality 26 Deﬁnition of 132–133 And awareness 129–131, 141–147, 151–152 And the concept of legitimacy 129, 153 In a separation-of-powers model 134–140 Israeli Supreme Court 13, 84–85 Jiménez, Norma 67 Judicial Independence Deﬁnition, see judicial power Measures of 175–181 And awareness 181 And democracy 196, 199 Judicial lockjaw 6 Judicial Power De jure v. de facto 8–9, 12 And judicial independence 9, 79 And democracy 10–11, 15–16, 51, 197, 199–202 Development of 190–195 And strategic judicial behavior 51 Judicial Reform In Mexico 48, 53 As political insurance 48–49, 149 As a barrier to political crises 49 La Jornada 76–77, 85–88 Legitimacy 6, 24, And transparency 7, 16, 123–124, 131, 186–190, 201 And public relations 7, 202 And awareness of courts 63, 128–131, 147, 170, 172–186 And compliance 13 And judicial trust 171, 174 Measures of 167, 170See also public support for courts Laveaga Rendón, Gerardo 58, 62 Magaloni, Beatriz 49 Marbury v. Madison 124, 191 Martínez, Néstor 61, 62 Media attention Quality of reporting in Mexico 58–59 And transparency 72 Measures of 75–77

Index And decisions to invalidate public policies 81–84, 89–90 And compliance/non-compliance 90 Mexican Federal Judiciary 101–114 Structure 101–103 See also Constitutional review in Mexico Mexican Supreme Court Budget 53–54 Legitimacy of 70–71 Jurisdiction 108–110 Media Coverage of 59–61, 66–67 Inﬂuence on media coverage 85–86 Likelihood of issuing press releases 87–88 Questions of Legitimacy 55–56, 63 Peso crisis 65 Public Relations 57–62, 201 Reforms to 48–50, 53–54 Removal authority 92–93, 94 Transparency 56–57, 62–64 Zedillo Case 3, 66–67, 100–101, 193–194 See also Constitutional Review in Mexico Moustafa, Tamir 196 Musharraf, Pervez 23, 142 National Action Party (PAN) 77, 148 Non-compliance Incidents of 9–10 And policy importance 33, 90, 94–95, 98–99, 201 And costs of case promotion 90 Measure of 91, 93–94 In amparo suits 93–96 Inefﬁciency of monitoring in the federal system 96 Agrarian cases 97–99 Ofﬁce of Social Communications (DCS) 57–62 Ortiz Mayagoitia, Guillermo 55 Pakistani Supreme Court 10, 23, 142, 168, 203 Partido Revolucionario Institutional (PRI) 47–48, 77, 89, 101 Partisan courts Alternative conceptualizations 148–150 Policy Importance Measures of 74 And decisions to invalidate public policy 81–84

Index And non-compliance 33, 90, 94–95, 98–99, 201 And strategic deference 32, 90 Positivity bias hypothesis 127–129, 145, 174, 186 Public enforcement mechanism 13–14, 22–26, In a democracy 198–199 Transparency 14–15, 198–199 Legitimacy 15, 198–199 Media attention 26 Coordination problem 24 Monitoring problem 25 Public Relations 7, 188–190, 202 And cost of case promotion 33–35 Press releases as a measure of 73 Public support for courts Diffuse support—see legitimacy Speciﬁc support And awareness 127–128 In Mexico 70–71 Ríos-Figueroa, Julio 77, 89 Román Palacios, Humberto 55 Rule of Law As a proxy for deﬁance costs 147, 168 Measures of 168–169 Salience, see Policy Importance Sánchez Cordero, Olga 56, 57, 67 Silva Meza, Juan 56, 63

221 Social Communications Ofﬁce at the Federal Judicial Council (DCSJ) 61–62 Spiller, Pablo 77, 89 Stephenson, Matthew 23–24 Stone, Harlan 6 Strategic judicial behavior Deference 124, 191 Taiwan Council of the Grand Justices 192 Tommasi, Mariano 77, 89 Transparency Tension with legitimacy 186–190 And judicial power 26–28, 31–32, 36–38, 71–72 United States Supreme Court 10 Court-packing 23 Court-curbing 168 Awareness of 141 Impact on media coverage 204 Vanberg, Georg 14, 27, 31, 187, 201 Venezuela Supreme Court 10, 131–132, 189 Vizcaíno Zamora, Álvaro 58, 62 Weingast, Barry 24, Whittington, Keith 79, 197 Zedillo, Ernesto 3, 48–49, 65–67, 77, 101