Citizens against the State: The Riddle of High Impact, Low Functionality Courts in Brazil

The Brazilian federal judiciary offers an interesting riddle to scholars of judicial politics and policy change. While the courts have played a major policy role over the past two decades, constraining and altering federal policy across a range of subjects, the court system has simultaneously been labeled “dysfunctional.” This paper investigates this riddle: a system plagued by major systemic flaws in its dayto-day operations, which nonetheless still manages to exert a powerful influence on public policy in Brazil. I adopt a new institutional perspective, focusing on how the institutional and normative structure within which judges and other legal actors operate affects policy outcomes.

A wide variety of disputes and conflicts have arisen out of the twin economic and political transitions that followed the end of military rule in Brazil.In resolving these conflicts, Brazilian federal courts have repeatedly been thrust into the vortex of policy debate.Brazilian courts have taken an active, dynamic stance: they have shaped policy choices, influenced policy implementation, challenged many of the executive-driven reforms of the 1990s, and stalwartly defended the 1988 Constitution.
On numerous occasions, federal courts have been called upon to challenge decisions made by Congress or the President: social security reforms have been scuttled; monetary correction challenged; taxes declared unconstitutional; privatizations halted; and the impeachment of a president questioned and ultimately upheld, in a major coup for democratic principle.On a fair number of such occasions, courts effectively halted policy implementation and sent policymakers back to the drawing board.At the very least, courts have been important loci for policy debate, with ramifications for how the policy process is carried out.Most importantly, compliance has been assumed, and governments have closely abided by these major court decisions, despite significant costs to their policy agendas.
The riddle at the center of this paper arises, however, when one looks at how unwieldy Brazil's courts are at what might be termed the micro level.Thousands of quotidian interactions between government and civil society in the courts have, in an accretive fashion, played a role in how policy is legally contested by citizens.The sum of such suits has been as important to public policy as the more "macro" examples of the courts' policy impact above.But the sum total of these interactions has not been positive.The court system offers few prospects for quick, effective, lasting remedies by the common citizen against government policies, with the possible exception of small victories in particularistic claims whose consequences apply to a single plaintiff.Brazilians show enormous concern about the slowness and elitism of the courts, 2 and three-fifths claim to have little or no confidence in the judiciary, 3 factors which greatly reduce the courts' potential as a locus for citizens to influence policy.It is no accident that Brazil's judiciary has been described as Jurassic, dysfunctional, and encrusted with layers of complication. 4he riddle then, is a court system that is strong -able to check the government, reverse policy, and ensure compliance with its rulings -and its coexistence with a court system that is nearly inoperative, providing few credible guarantees to the average citizen of rapid legal recourse against the government's public policies.On the one hand, the courts have enabled peaceful resolution of important policy disputes and incorporated opposition groups, such as political parties and social organizations shut out of the executive and legislative branches, in democracy-enhancing deliberations on policy.But on the other, the manner in which the court system operates excludes a large share of the citizenry from consuming the commodity of justice, and may thus temper their views of the effectiveness of a major democratic institution.
This paper tries to understand this riddle by focusing on how the institutional architecture of the court system, combined with the prevailing norms and beliefs of legal practitioners, structures public policy debate and conflict resolution within the judiciary, and thus in Brazil as a whole.I seek to bring to bear a new institutionalist perspective to the courts.This task is addressed in the next section.

A NEW INSTITUTIONAL APPROACH TO BRAZILIAN COURTS
In the judiciary, no less than in presidencies and assemblies, politics is given shape by the institutional and normative structure within which judges and other legal practitioners operate.In researching the courts' effects on policy, the "black box" of the state has to be opened up, if we are to understand how societal interests are translated by institutions into policy, and from policy into policy outcomes. 5Bureaucratic structures, constitutional arrangements, legal instruments, and the professional norms of judges, lawyers and prosecutors influence the public policy debate within the judiciary, and thus affect public policy outcomes, by determining who has access to the courts, where in the court system that access is granted, and how and under what conditions courts make decisions.
Douglass North, the most influential voice of new institutional economics, has emphasized the role institutions play in reducing uncertainty in society by structuring human interaction.But at a broader level, North also recognizes that institutions "are not necessarily or even usually created to be socially efficient; rather they, or at least the formal rules, are created to serve the interests of those with the bargaining power to devise new rules." 6Moreover, bargaining strength matters where institutions impose transaction costs; institutions therefore affect long-run economic change by altering the distribution of power between different groups in society.This message of the possible inefficiency of a state's reflection of the demands or interests of society has resonated forcefully in the realm of political economy and political science.
The least common denominator of "new institutionalism" is that institutional norms and/or structures may affect political behavior by shaping the ability of political actors to advance their preferred course of conduct.Three competing, if occasionally overlapping, variants of new institutionalism have emerged in political science around this central theme. 7The closest to institutional economics is rational choice new institutionalism, with its emphasis on collective action and the strategic interaction between players.Sociological new institutionalists are critical of the rational choice model and its marginalization of cultural variables.In their view, institutions "frame" behavior through cognitive scripts that shape not only individuals' strategic calculations, but also their identity.While North's later work attempts to respond to this criticism by incorporating informal institutions, 8 critics have sought to better incorporate the social-cultural context into new institutionalism, emphasizing the fact that institutions may affect individual behavior and vice versa.Finally, historical new institutionalists fill a middle ground, emphasizing distinctions between national political outcomes, power asymmetries, and the unintended consequences of institutional arrangements, while often drawing on culture as a transmission belt for the habitualization and internalization of norms. 9Despite their considerable differences, however, these three different schools of new institutionalism share a common belief that institutions are not neutral in their effect on politics and society, and that the state has its own interests, which affect policy.
The new institutionalist approach has been widely applied to a number of institutions, but only recently to the judiciary.Most work on the courts that places itself self-consciously in the new institutionalist camp has focused on the U.S. case, and particularly on how the courts' decisions are shaped by institutional settings. 10Partly as a result of the choice of the U.S. as a case study, these approaches often are couched as a complement to attitudinal models of court action, with their emphasis on the importance of justices' personal policy preferences in determining case outcomes.
I draw on new institutionalism here in a manner that is intended to be complementary to both attitudinal approaches and more legalistic analysis.The new institutional view provides a complementary perspective that draws on both informal and formal factors to analyze how the judiciary hears cases, which cases it hears, and where it hears them, and hence, how, when and where the judiciary influences policy.Courts have less apparent freedom of action than the executive or legislative branches -they are largely reactive rather than proactive, and face more latent constraints in their need to follow the letter of the law -yet some cases are more likely to reach the courts than others.Understanding how and why they do is as important as understanding the general likelihood that a given judge will be inclined to vote in one direction or another on a given case.
In the next sections, I address history, culture and structure as they apply to the Brazilian federal judiciary.There is necessarily some repetition and overlap involved in applying these different lenses.In dividing my approach into three separate sections, I may temporarily undermine the richness of a more combined approach.But in the conclusion, I pull the disparate threads together in an effort to answer the riddle posed at the outset.

AN HISTORICAL APPROACH
The unique pattern of political evolution from the military regime, through the drafting of the 1988 Constitution, and into an era of neoliberal reform during the 1990s has placed the Brazilian judiciary in frequent tension with other institutions in society.The legacy of opposition to military rule played out in the institutional framework created by the 1988 Constitution, which guaranteed a high degree of judicial independence, greater public access to the high court, and the creation of new case types aimed at constraining executive arrogation of power.Legal groups such as the OAB saw their opposition to the military and defense of democracy rewarded with new forms of access to the courts.But as Steinmo, et al. suggested, institutions may shape not only political actors' strategies, but their very goals.The very important position of the judiciary and lawyers in drafting the extensive democratic rights incorporated in the Constitution, the related preferences these groups have strongly voiced in years since, and the resulting manner in which the judicial system is used, have placed the Brazilian judiciary on a collision course with the executive branch's reformist goals at key moments over the past fifteen years.
It has been noted that worldwide, courts are the least democratic of all democratic institutions.They are not representative, they are only mildly accountable, and they are the most elitist of institutions (barring perhaps central banks), employing a cadre of highly trained specialists responsible only to national constitutions and an amorphous concept often referred to as the "rule of law."Brazilian history has provided a further and more fundamentally anti-democratic bias: under military rule (1964-85), governments often sought to cloak their actions in a mask of legality, and although judicial independence was significantly reduced by the military, 11 courts continued to perform their function of preserving a variant of the rule of law, highly prized by the military government, throughout the authoritarian period. 12Courts entered the post-military period with few changes in procedure or personnel, despite the fact that the courts -like other democratic institutions -had been unable to uniformly uphold the democratic values often considered a sine qua non of the concept of the rule of law.
Largely as a result of this history, while the post-military 1988 Constitution kept intact many of the formal structures of preceding legal systems dating from the Constitution of 1891 -including the civil code and the federal structure of the courts -the constituent assembly sought to incorporate new democratic guarantees within the judiciary.In addition to an extensive catalog of individual rights, this institutional expansion guaranteed greater judicial independence, especially from the executive branch, and much broader access to the courts. 13he independence of judges under the new constitution was assured by meritbased recruitment, by guaranteed life employment until age seventy, by 'irreducible' judicial salaries, and by guarantees that judges can only be removed by a vote of their peers.Meanwhile, the courts themselves were constitutionally guaranteed "administrative and financial autonomy," as well as complete internal administrative oversight.Brazilian Journal of Political Economy 25 (4), 2005 Judiciário No Brasil." in Uma Introdução Ao Estudo Da Justiça, Editor Maria Tereza Sadek.São Paulo: Editora Sumaré, 1995.Verner describes a number of incidents of pre-military exercise of judicial power, and then of court challenges to the military itself, in Verner, Joel G. "The Independence of Supreme Courts in Latin America: A Review of the Literature."Journal of Latin American Studies 16, no. 2 (November 1984): 489-490.
12 Although he does not explore the role of the courts in any great detail, Thomas Skidmore offers a compelling history of the military's schizophrenic commitment to legality: early on, the military government cloaked anti-democratic moves in legalism by using "institutional acts" to undermine the constitution; the first military president, Castelo Branco, was emphatically committed to avoiding typical Latin American "caudillismo," hoping to opt instead for the "legal path, instead of sliding toward a dictatorship," by avoiding any talk of an extended presidential term and by creating informal rules to prevent his successors from holding power indefinitely (52); the 1967 Constitution was aimed at generating strong government, with "a residual of representative democracy and the rule of law"(79); and "even at moments of arbitrary action, the Brazilian military persisted in believing a respectable legal rationale for their actions could be found"(96).Later in the regime, by the time of President Médici's first address as President in 1969, Skidmore notes that this cloak was no longer viable, but the promise of legality remained an important concept for the military.Skidmore, Thomas E. The Politics of Military Rule in Brazil, 1964-85.New York: Oxford University Press, 1988.Gonçalves Ferreira Filho notes that the courts used ações populares (a lawsuit that can be brought against injuries caused by public figures) to impose some control on the government at a time when the Congress was "domesticated."Gonçalves Ferreira Filho, Manoel."Poder Judiciário na Constituição de 1988: Judicialização da Política e Politicização da Justiça."Revista de Direito Administrativo  198 (October 1994-December 1994): 5. 13 Prillaman notes that independence and access were granted without any compensating constraints of responsibility.Prillaman, William C. The Judiciary and Democratic Decay in Latin America: Declining Confidence in the Rule of Law.Westport: Praeger, 2000.Castro de Faro notes that the enlarged judicial independence included in the 1988 Constitution was due to the the lobby of judges in the constitutional assembly.Castro, Marcus Faro de."The Courts, Law, and Democracy in Brazil."International Social Science Journal 49, no.152 (1997): 23.Access to the courts under the 1988 Constitution was expanded by creation of small claims courts (juizados especiais), as well as by broadening access to some legal instruments. 14The Constitution allowed for widespread challenges to the constitutionality of laws via the Ação Direta de Inconstitucionalidade (ADIN), or Direct Action of Unconstitutionality.The ADIN mechanism had existed under the military, but was expanded by the new Constitution, allowing the constitutionality of a law to be directly contested in the highest court, the Supreme Federal Tribunal, by a broadened set of institutional actors: the President, the Senate, the Chamber of Deputies, the state assemblies, state governors, the prosecutor general, the Federal Council of the national bar association (Ordem dos Advogados Brasileiros, OAB), political parties, and national "class or union confederations." 15In addition, new case types were created or modified to strengthen the ability of civil society groups to challenge the government directly, and the Constitution created the autonomous Ministério Público, a federal prosecutorial body with extensive powers and a degree of autonomy that have led some to call it the "fourth power" of government in Brazil.
All of these innovations substantially increased the number of players with an institutional stake in the court system, but in ways that were also a direct result of past historical developments.The OAB, for example, has been an influential player in the political system since before military rule, but its vocal and prominent opposition to the military government in the 1970s and 1980s gave it special prerogatives in the new system.These prerogatives both preserved the OAB's prominence, and gave it reasons and instruments to defend the new institutions forcefully, as it has done since the 1988 Constitution was enacted.
In sum, the new Constitution broadened access to the courts, and especially to the high courts, by creating or modifying legal instruments such as the ADIN to allow wider challenges in the high courts against the initiatives of the executive and legislative branches.The concrete effect of these changes, however, still restricted access to the high courts to a select (if broader) list of organized political actors.For ordinary citizens using the lower courts, new mechanisms such as the public civil suit (ação civil pública) and the popular suit (ação popular) were intended to offer similar recourse against government policies.But the complexity, cost, and duration of such cases have largely circumscribed their use by the average citizen.More quotidian court challenges by individual citizens seldom carried over into policy because of the extraordinary delays engendered by the judicial process, and the rather narrow effect of lower court decisions, discussed in the next two sections.

A SOCIOLOGICAL APPROACH
From a sociological new institutionalist perspective, the culture of the legal profession in Brazil plays an important role in how the law is applied and how the legal system fits into the overall political structure.Brazil was once known as the "República dos Bacharéis," referring to the fact that over half the presidents in the 1889-1930 Republic were drawn from the ranks of lawyers.Although presidents are no longer as likely to be lawyers, the formalism of legal training and legal practice, the high societal and professional status accorded to legal organizations such as the Order of Brazilian Lawyers (OAB) and the Association of Brazilian Magistrates (AMB), and the strong adherence of judges and lawyers to the long list of rights enumerated in the 1988 Constitution has an important effect on how legal decisions are made, how legal processes are structured, and how government policy initiatives are viewed.
Legal culture has been defined by Pérez-Perdomo and Friedman as "the cluster of attitudes, ideas, expectations, and values that people hold with regard to their legal system, legal institutions, and legal rules." 16The internal legal culture of the members of the Brazilian judiciary -lawyers, prosecutors, and judges -have together contributed to a new form of judicial activism and new types of conflict between the judiciary and the executive branch in the past 15 years.The breadth of the 1988 Constitution and its commitment to social justice has charged "judges with the task of protecting vulnerable social classes,"17 a responsibility that frequently plays out in the defense of particular rather than universal values.This ethos of protecting the vulnerable is communicated and implemented throughout the judicial branch by two key institutional factors: legal education and the independence of judges, both from the executive and from each other.
The dogmatism of legal education in the context of a civil law system has significant implications in terms of policy: judges are trained to focus on principles, rather than consequences; and the focus of judges is on the law as it is written, rather than on the intent of the legislators or on the consequences of any given decision in the nation at large.This tendency may be further exacerbated by the perception that legal decisions have no widespread consequences, in the absence of erga omnes effects or stare decisis, described in the next section. 18he education of judges imparts formalism and perhaps a degree of corporatism into the judiciary. 19As in most professions in Brazil, legal education is highly profession-specific, meaning that students have little exposure to disciplines outside their field which might speak to the broader implications of legal decisions.An above-average student graduating from high school, who does well on state-wide examinations, could expect to enter a government-financed law school at age eighteen.For ten semesters, such a student would study everything from juridical sociology to history of the law to constitutional law to civil process.At a top-tier school such as the University of São Paulo (USP) school of law, the curriculum averages around 25 hours of class per semester (although class burden is heaviest in the early years).But over the course of five years of law school, an USP student is required to take only 20-25 hours of non-law classes on economics, sociology or political science, mainly during the first two years of classes.
Meanwhile, the University of São Paulo and other government-financed schools are an island of excellence in an ever more popular field of study.The number of law schools tripled in the past decade to over seven hundred schools in 2003; the number of graduating law students taking the Education Ministry's "provão," a rough indicator of the number graduating from accredited law schools, almost doubled from 26,000 to 51,000 between 1996 and 2001. 20Partly as a lawyers, noting, "Brazilian legal education has been overwhelmingly formalistic....Generations of Brazilian law students have been taught that law is a science, and that the task of the legal scientist is to analyze and elaborate principles that can be derived from a careful study of positive legislation into a harmonious, systematic structure."Rosenn, Keith S. "Brazil's Legal Culture: The Jeito Revisited."Florida International Law Journal I, no. 1 (Fall 1984), 23.See also, Rosenn, Keith S. "The Jeito: Brazil's Institutional Bypass of the Formal Legal System and Its Developmental Implications."American Journal of Comparative Law 19 (1971): 514-49.A critical review, noting that the "jeito" is not a solely Brazilian phenomenon, can be found in Lacerda Teixeira, Egberto."O Jeito : Instituição Jurídica Brasileira Com Foros Internacionais ."Revista dos Tribunais 437 (March 1972): 473-75. 20The "provão" is an exam created by the Cardoso government and administered to all graduated college students as a test of their colleges' strength.The test evaluates knowledge of basic concepts across a variety of fields within the given course of study, and is aimed a creating quality benchmarks that are comparable across universities nationally.As in most Latin American countries, college education in Brazil does not follow a liberal arts model.High school graduates enter college already enrolled in their major field of study, such as law or medicine, and graduate 4 or 5 years later as lawyers and doctors (pending approval on a bar or medical association practitioners' exam)."Diretores da OAB SP e CAASP tomam posse solene hoje," 10 March 2004, Assessoria de Imprensa, OAB-São Paulo; accessed 10 March 2004, at http://www.oabsp.org.br/;"República dos bacharéis: Direito formou 50.933 bacharéis no ano passado," Revista Consultor Jurídico, 24 January 2002.
result, concern about the quality of legal education has been sharp, driving the OAB to create a new ethics exam, and as of 1994, a bar exam.The numbers approved on the bar exam are shockingly low, ranging from state to state, but remaining below 50% in both Rio de Janeiro and São Paulo.The result may well be that "[s]oon Brazil will be a country with many graduated law students but fewer lawyers." 21he funnel continues to narrow as one moves from law students, to lawyers, to judges, who form a highly qualified elite at the apex of the legal profession.Most judges are selected via a rigorous entrance examination after a few years practicing law, in their early twenties.Lateral entry into federal judgeships from other professions is virtually unheard-of, except in the top courts, whose judges are sometimes chosen from the ranks of the Justice Ministry, the Ministério Público, or private practice.The judge's focus is thus highly theoretical, and often devoid of real-world circumstance and experience, with important implications for how judges rule in the abstract.Nalini, a prominent judge himself, notes that for the judge trained in this manner: The best decision is one aimed at resurrecting the past, that is, one that manages to return the situation to the status quo ante.This results in the total impossibility of the judge facing the future, of exercising a proactive stance, of evaluating the consequences of his or her decision for the parties [to the dispute], for the community, for the nation. 22urthermore, formalistic procedures within the judiciary, despite the intent to have the law interpreted to the letter of the civil code, mean that different judges, acting under different circumstances, from different perspectives, are likely to come to different interpretations of the same combination of case and law.
This mix of formalism with decisions that have only particularistic effects is further embedded by the absence of vertical downward control in the judiciary: judges are tied together by a civil service mentality, rather than a hierarchical pattern of decision-making.Taken together, these factors suggest that despite an extensive catalogue of individual rights in the 1988 Constitution, the realities of the judiciary in Brazil mean that inevitably, protection of individual rights is almost the exception, rather than the rule.As Pereira notes, "The nonuniversal application of law gives rise to what might be called "elitist liberalism," a policy that justifies the granting of rights on a particularistic basis." 23he training of judges and the autonomy of the various component members of the judiciary have implications in terms of what is to be expected from the courts.The broad-based 1988 Constitution was written in the hopes of reversing the worst abuses of the military regime.But the impressive list of expanded rights laid out by the Constitution is undermined first by the inability of the state to afford the claims written into the Constitution, and second, in the court system, by the inefficient performance of the judiciary as an institution.This dual failure in turn increases the likelihood of particularistic appeals for justice.Judges are not only able to meet such particularistic claims, but in light of their training, loose hierarchy, and commitment to the letter of the Constitution, may be willing to do so in many cases.

A RATIONAL CHOICE APPROACH
From a rational choice new institutional perspective, perhaps the most interesting phenomenon is that judicial decisions are tenuous in light of the absence of binding precedent, their largely case-specific (rather than universal) applicability, and the strong independence of judges from each other.The result is that judges' decisions are to a certain extent less important to the overall functioning of the courts than the management of the huge caseloads that enter at all levels of the court system.This creates interesting incentives for clients of the court system, privileging those who can bypass the congested middle courts, as well as incentives for using the lower courts to delay definitive policy implementation.
By virtue of Brazil's particular institutional arrangements, the structure of judicial decision-making is both more bureaucratic and far less centralized than in many other countries.By way of illustrative anecdote, Brazil's highest court is equipped with a drive-through window in the basement, where lawyers can drop off cases; most federal courts have a full-time binding operation which sorts, categorizes, and binds incoming cases in a color-coded system for justices to review; and most courts are equipped with a machine resembling an automatic teller machine (ATM), where lawyers can type in case numbers to track their cases wherever they may be in the judicial system.The incentives for justices and the implications of such a process-oriented system for public policy are far different from many other countries, with the timing of suits and access to high court hearings given considerable strategic importance by political actors.Individual judges are granted a good deal of discretion and power in individual cases; but ironically, the importance of any individual decision is correspondingly deflated in its impact on society overall.
The Brazilian judicial branch is hierarchically structured in a federal system made up of both federal and state courts.Under the 1988 Constitution, a group of courts sit at the apex of the court system, each with different functional responsibilities: the Supreme Federal Tribunal (Supremo Tribunal Federal, or STF) is the key court for constitutional matters, and matters involving conflict between members of the federation; the Superior Justice Tribunal (Superior Tribunal de Justiça, or STJ) acts as the court of last instance on non-constitutional matters; and the Superior Military Court, Superior Electoral Court, and Superior Labor Court act as the courts of last instance in their respective areas of expertise.
Despite this flat structure, in practice the Supreme Federal Tribunal (Supremo Tribunal Federal, or STF) is the highest court in the country, serving as the court of last instance for both state and federal cases involving federal constitutional law.Given the breadth of the 1988 Constitution -with its 245 articles and 73 transitional provisions -this constitutional distinction means little in practice, with many rather banal cases, which on the face of it should have little constitutional merit, appealed on constitutional grounds all the way up to the STF.One STF justice, on hearing a constitutional case about a goat that had been barbequed by the owner's neighbor, complained, "matters related to cows, butterflies and horses are all constitutional matters." 24he legal procedures employed by the courts also influence the use of the STF, with the breadth of constitutional review generating a substantial workload for STF ministers.Constitutional review is largely abstract, dealing with laws as they are written, rather than concrete review of actual cases between contending parties, in which the law is co-incidentally questioned.STF ministers thus have fewer opportunities to avoid constitutional interpretation than might be the case under concrete review.Meanwhile, the absence of legal mechanisms that might limit such review contributes to the workload.As Stepan noted, unlike the U.S. Supreme Court, the STF does not benefit from the writ of certiorari, which might allow it to reject some cases; its decisions in most cases are not erga omnes, or effective against everyone; its decisions for the most part do not impose stare decisis, or binding precedent; and there is no "political question doctrine," such as the U.S. Supreme Court uses to avoid entangling itself in political disputes that can be resolved in the Congress. 25In the absence of such rules, STF docket control is a question of drinking from the firehose, with no choice but to address all of the cases thrown before it.
Figure 1 provides a heuristic diagram of the hierarchical nature of the federal judicial system in Brazil.The courts of first instance are known as "varas."Cases from the varas move up on appeal to the federal regional tribunals (known as TRFs), each of which covers a region made up of several states.Appeals from the TRFs rise to the STJ, and from there frequently to the STF.A parallel judicial hierarchy exists for the Labor Courts (1,109 varas, 24 regional courts, and the Superior Labor Tribunal, TST), the Electoral Courts (2,884 electoral cartórios, 27 regional courts, and the Superior Electoral Tribunal, TSE), and the military courts (20 federal military "auditorias" and the Superior Military Tribunal).Each state also has its own judiciary, adding nearly 2,400 judicial districts and over 6,000 judges to the national total; cases from the state high courts can be appealed to the STJ, as illustrated in the figure.In addition to the hierarchy of the courts in legal matters, the administrative rules for the various courts follow the same vertical structure.The administration of the federal justice system is overseen by the Council of Federal Justice (Conselho da Justiça Federal, CJF), which is made up of the president and vice president of the STJ, two administrative staff, the presidents of the TRF regional courts, and other judges from the STJ and TRFs, appointed on a rotating basis.The CJF meets monthly to discuss the administration of the judiciary, and its decisions are binding for all federal courts.Although Brazil has an estimated 63,000 staff working in the federal judiciary, unlike many countries, it does not have a formal profession of court administrator.This puts judges in the ambiguous role of both hearing cases and negotiating contracts with local governments for the rental or purchase of properties and other administrative tasks.Many have noted the undesirable nature of this arrangement: not only is the staff not centrally trained to a uniform pattern of administration, but judges are frequently charged with carrying out administrative tasks which may place them in a conflict of interest. 26The budget woes of state and federal governments, meanwhile, have frequently led to lower than budgeted transfers to the judiciary, leading judges to complain about the unfair infringement of their "financial autonomy", and its effects on judicial independence.
Courts at the lowest levels have a good deal of flexibility, both in how they spend their budgets, and in the rules they establish for their own operation.This flexibility has had curious effects, contributing to a division between higher court and lower court judges, especially on wage questions, where the lower court judges have vociferously pushed for higher wages against the public (and occasionally legal) opposition of the higher courts.Perhaps not surprisingly, Brazil has an almost even ratio between high court and low court judges' salaries, with a ratio of 1.17 to 1. 27 On occasion, the loose hierarchy may also have contributed to corrupt practices. 28Revelations of corruption have contributed to a perception of impunity accompanied by little transparency, expressed most recently by President Luis Inácio da Silva, who lashed out against the "black box" of the judiciary.The corruption scandals of the late 1990s have contributed to a marked decline in public confidence in the Brazilian judiciary: according to polls conducted by Latinobarometro, the percentage of Brazilians who had "a lot" or "some" confidence in the judiciary fell from 41% in 1996 to 32.5% in 2002.Although Brazilian courts are still more trusted than most of their counterparts in the rest of Latin America, by 2002, 63.9% of Brazilians said they had "little" or "no" confidence in the judiciary." 29onetheless, judges have been eager to avoid controls on their budgets and operations, credibly citing the potential threat to their independence, in the wake of the authoritarian encroachments of the recent past.As in all discussions of judicial independence in Brazil, whether the topic is budgeting or legal decisions, there are two meanings to "independence": the first is independence from interference by the President or Congress; the second is the independence of the individual judge from interference by higher courts or even from his or her own peers on the bench.Lower court judges have traditionally supported both definitions of independence, and have resisted incursions from within the judiciary as well as from outside.
Another long-term institutional feature of the Brazilian judiciary is the unique variant of the civil code generated by the mix of legal tradition and institutional framework.In the most simplistic shorthand, the existence of the civil code suggests a formalistic rigidity to the court system: in most civil code systems, the judge is seen as an enlightened clerk, applying rather than discovering the law.Given the inquisitorial process of the civil law, its codified nature, and the importance of the judge as interpreter of statute, the courts in civil law systems in theory have little procedural influence in creating the law.
Oliver Wendell Holmes said of the common law that the "life of the law has not been logic; it has been experience."30At least in theory, this is not the case with civil law systems, where logic should at least nominally override the experience and prejudices of judges.A civil law system may make the judiciary procedurally less democratic than it might otherwise be: juries are used only in murder cases, there is little "discovery" of legal foundations, and the law is applied, at least theoretically, in a formalistic procedural manner in which judges are given little latitude to shape the law.But this procedural rigidity is counterbalanced in Brazil by the relative autonomy of individual judges.Although jurisprudence has gained an important foothold in Brazilian law and reform is on the way, as of this writing there is still no system of binding precedent under civil law. 31The absence of precedent-setting by higher courts gives Brazilian judges a de facto right of discovery at almost every level in the judicial system, in addition to encouraging constant appeals to the top of the judicial hierarchy.
The formal opposition to precedent-setting argues that judicial decisions are not meant to create rules of law -under a civil code system, that job should be left to legislators or presidents.As one STF minister framed this very common assertion in his comments during a constitutional case, "...the constitutional review of normative acts by the Judiciary only allows the Judiciary to act as a 'negative legislator."32That is, the courts cannot change the original intent of the legislator by striking down select passages and changing the intent of the law; they can only rule that a given law is constitutional or not.
Cynics have also suggested another reason judges, particularly in the lower courts, have fought the introduction of binding precedent in the Brazilian context: judges want the freedom to grant individual rights to those who frequent their courtrooms, especially since they cannot guarantee collective rights to all. 33Lawfinding, in a civil system, is about the application of legal propositions.But even here, the immense body of laws adopted in Brazil -some 27,000 by one count 34 -suggests that the civil code can at times prove to be a large and unwieldy collection of often confusing, sometimes mutually contradictory, legal rules.At the very least, the combination of the absence of binding precedent, the individual freedom of judges, and an extensive body of laws has contributed to the flexible application of laws, as well as delays in ascertaining definitively how such laws apply.
The weakness of binding precedent also undermines the ostensible goals of formalism within the Brazilian civil law system.The number of laws, combined with a lack of stare decisis, means that in any given case, chances are that the individual interpretation of the law may vary from judge to judge.Similar cases may thus be legitimately decided, within institutional boundaries, in a multitude of different ways.This creates a setting in which particularism is not only possible, but likely, and in which the judge is far from an "expert clerk," and more of an interpreter.It also results, systemically, in a glut of cases arising from the variety of possible interpretations at various levels of the court system.

POLICY REPERCUSSIONS
On the case front, the number of cases in the system has increased significantly since the end of the military regime.The number of cases filed in the federal court system (excluding labor, military, and electoral courts) has mushroomed, from 350 thousand cases filed in 1989 to 1.84 million in 2001. 35On average, within the judiciary as a whole, this is not an especially large number in comparison with the more than 12 million cases currently working their way through state courts.However, the case overload has taken on especially shocking scale at the higher reaches of the federal judiciary, with the Brazilian high court (STF) receiving an impressive 11% of the volume of cases heard in the courts of first instance.A comparison may be illuminating: there were 10,070 filed per Brazilian STF minister in 2002, against 872 cases filed per U.S. Supreme Court justice (and fewer than 100 of these were actually accepted for debate by all U.S. Supreme Court justices).In the Brazilian STF, even though only a minority of cases actually goes to a hearing before the full STF, all cases must be judged by at least one minister, and the decision approved and signed by that minister.This is very different from the U.S., where a decision not to grant certiorari means that the pre-appeal decision stands, and Supreme Court judges are thus freed of any further responsibilities.
Cases on the STF docket ballooned over the past decade, from 20,000 in 1987 to a remarkable 160,000 in 2002.As a result of the use of appeals to the highest level, the STF frequently ran a "deficit" -i.e., cases that it simply does not have time to decide upon, whether for dismissal or judgment -in the years running up to 2003.This situation was dramatically improved in 2003, partly as a result of the Cardoso government's efforts to reduce government use of the courts. 36The most important measure in this regard was a 2001 government decision to force the Caixa Econômica Federal, a major federal bank, to file motions in the STF giving up on most cases relating to monetary correction in which it was a defendant; the number of such motions for dismissal by the government rose from 51 in 2001 to over 50,000 in 2003.Simultaneously, the STF was implementing major administrative improvements aimed at speeding up the administrative processing of cases, in some cases reducing the time between receipt and judgment in the STF from months to days.Despite these improvements, however, the figures from 2003 show that there is still an enormous workload in the STF: more than 87,000 cases were received and nearly 108,000 were judged.
The glut of cases in the STF and STJ includes not just appeals from lower courts, but also appeals of decisions already taken in the upper courts.According to a former minister, almost 80 percent of cases in the STF are repeat cases. 37epeat cases generated by special appeals are more and more the norm, rather than the exception: the STJ's case figures in 2002 likewise showed that 78.8% of its cases were a result of two such appeals.As one STF minister noted, the STF does not judge new cases, it judges repeat cases: "Equal sentences were approved in equivalent cases, in which the [STF] ministers acted merely as paper stampers for cases that had already been judged." 38s a result of high caseloads, repeated appeals, and the glut of cases engendered by government appeals of lower-court challenges to policy, the judicial process itself is slow, with the snail's pace of justice popularly seen as a fundamental flaw of the Brazilian judiciary.A recent student protest against the STF's failure to hear a case against a state governor featured a giant papier-mache turtle dressed in a judge's toga; in another case, a plaintiff rented a billboard to make a public plea for the courts to hear his case four years after he had filed it.Cases not resolved through accord are estimated -on the basis of polls of businessmento take 38 months on average to get through the common state justice system, and 46 months in the federal courts. 39A retired STF minister recently guessed that a typical case that started at the bottom of the judiciary and worked its way through appeal up to the STF would average between eight and ten years from start to finish.Sadder still, in this judge's view, in most cases, the high court upholds the decision made at the outset. 40iven these case load and case duration data, it is not surprising that actors within the political system often attempt to use the court's performance to their advantage.Some may bypass the courts entirely, turning to alternate forms of dispute resolution.This notion is not tested here, given the focus on public policy, but the incredible growth in the use of small claims courts at the federal and state level, the signing of contracts enforceable in foreign courts, and evidence of nonjudicial dispute settlement outside the formal court system suggest it is a distinct possibility.But within the formal court system, plaintiffs and defendants using the federal courts choose strategies that meet their own policy needs.
The federal government, particularly, has often sought to delay decisions against it.Partly as a result of this, and partly because of the reformist tack of the 1990s, the glut in the upper reaches of the Brazilian judiciary can be largely attributed to public sector litigation, with state companies and governments at all three levels (federal, state, and municipal) counting for 79% of litigation in the STF during the period 1999-2003.The federal government alone accounts for 64.8% of this case load, with most of this burden due to appeals by the National Social Security Institute (INSS), the Caixa Econômica Federal bank, and the Central Bank, of cases related to economic stabilization plans. 41he federal government's choice to engage in such delays is highly rational, despite the costs to the broader public.The costs of the government defending itself are quite small in comparison to the potential costs of some of the suits against the government: as of 2004, there are several major fiscal "skeletons" lurking in the courts, of which the three largest are cases related to the inflation readjustment of home mortgages under the 1990 Collor Plan (estimated potential cost, R$87bn); losses incurred by sugar farmers and mill operators in the wake of the 1986 Cruzado plan (R$40bn); and losses to the airlines under the same plan (R$7bn). 42Meanwhile, by putting off legal challenges for as long as possible, the government is able to proceed on given policy paths until they are a fait accompli (or the presidential administration has changed).
The history of Brazilian hyperinflation has also meant that there is a "judicial Tanzi effect": the longer cases remain in hearings, the more likely the real value of the underlying debt will be eroded.The result is that the proportion of cases involving the federal government increases the higher one looks in the judiciary.Given this endless appeals process, it is not unusual to hear the government described as a "bad faith" litigant, or to hear of government defenses in intermediate appeals that are, in the Brazilian phrase, "para inglês ver" (roughly translated, "for show").While the Cardoso government reduced the use of some such appeals, the practice continues, and the federal government continues to exert an important burden, especially at the upper reaches of the court system.
A second group of policy actors has drawn on the current institutional framework to develop strategic approaches that offer them greater comparative power in the courts.Two strategies, out of many potential legal strategies used by policy opponents, are of particular interest here. 43irst, policy actors given access to constitutional review mechanisms in the STF are able to leapfrog the queue that clogs the lower courts and obtain high court decisions on policy in an expedited fashion in the STF.While hearings on constitutional cases may take many months, and sometimes even years, to be decided, they are nonetheless quicker than resorting to ordinary cases that must make their way up the full judicial hierarchy.The ADIN mechanism, for example, has been used by opposition political parties, the OAB, and labor and industrial federations to overturn key elements of social security reform, agrarian reform, the 2001 electricity rationing measures, and other key policies.For those groups privileged with such access to the high court, the chances of overturning or obtaining an injunction temporarily halting a policy through use of the ADIN average about one in five. 44In cost-benefit terms, these are not long odds for a political actor considering the filing of an ADIN.This is especially true given the low costs, which are limited to legal fees and the potential of having a law's constitutionality upheld, and the potential benefits in comparison to running the gauntlet of the full judiciary.
A second strategy takes advantage of the particularism of individual judicial decisions in the lower courts.Opposition political parties, especially, have turned to the lower courts to fight privatization and electricity rationing, for example.In so doing, they use the "atomized" nature of the court system 45 to create greater political noise within the system, drawing attention to the potential flaws of the policy in question.These tactics use the independence of each judge as a means of increasing the odds that they will obtain temporary injunctions which delay policy implementation and place the finality of policy in question.They have been used particularly effectively against privatization, for example, where a host of injunction requests, often numbering in the hundreds, were systematically and simultaneously filed across a wide range of courts.The intent was to draw attention to the opposition's fight against policy, but also to take advantage of individual judges' potential opposition to privatization by obtaining injunctions that slow the progress of the sales of public companies.As Ballard noted, in the CVRD privatization of 1997, over 150 cases were filed, resulting in more than 20 injunctions.While the sale ultimately proceeded, by filing cases across a wide range of courts, the opposition was able to generate a few decisions in its favor that delayed the privatization, and embarrassed the government considerably.
The result of the system for individual citizens, however, is that the 1988 Constitution's promise of greater access has been scuttled.Individual challenges to policy result largely in case specific decisions, and only after considerable delays and monetary costs have been incurred.In this manner, the courts' institutional architecture tends to privilege organized policy actors, whether in government or opposition, over individual citizens.

CONCLUSION
This paper has offered some explanations for the riddle of the dichotomous performance of the federal judiciary -strong on major policy issues and dysfunctional for the average citizen -offered at the outset.I have aimed to place judges' behavior in an institutional context, explaining some of the problems of access and efficiency in a manner that does not deny the heroism of judges in confronting their herculean daily tasks, or their individual courage in confronting new violence that threatens their personal well-being.46I have also sought to illustrate how a new institutional approach to the courts may improve our understanding of the judiciary's impact on policy.Reiterating briefly: The post-military constituent assembly generated a constitution that includes an extensive catalog of rights, and considerable independence for judges, coupled with broad administrative autonomy, and many new venues of access to the courts.This was not combined, however, with measures to ensure the strict accountability of the courts to any external authority, be it abstract (legal consistency) or concrete (a judicial oversight panel, or the STF).Broad access has tended to favor those groups privileged in the Constitution with direct access to legal mechanisms that are heard in upper tier courts -such as political parties, corporatist associations, and the government -while recurrent appeals and the congestion provoked by heavy government use of the judiciary clog the courts and restrict effective access for the broader population.
Second, the pyramidal structure of the federal courts, in the absence of precedent-setting, means that the upper courts are likely to be swamped with appeals.Yet despite this overburdening, high courts have little downward legal or administrative control over lower courts, and they are further burdened by the broad avenues for repeated appeals throughout the upper tier of the federal court system.This means that policy issues are seldom definitively resolved in a universifiable manner until, or unless, they reach the high courts repeatedly.The one exception is the use of extraordinary case types such as the ADIN which offer direct access to the STF.Political actors -including the government, the Ministério Público, political parties and corporativistic professional groups -who are given the special privilege of using these case types have special policy influence.Ordinary citizens face an altogether more daunting legal process.
Third, under the Brazilian civil law system, the absence of a universal application of judicial decisions, as well as the enormous quantity of laws on the books, can lead to long delays in ascertaining the acceptability of contested laws, as well as endless appeals processes.It also tends to accentuate particularistic, non-universifiable solutions to legal problems, in contrast to the technocratic, neutral aims of ideal civil code systems.
Despite these signs of the important policy effects of the Brazilian federal judiciary's institutional composition, there are nonetheless hopeful signs of change in the air.The executive branch has increasingly withdrawn some of its opposition on major cases that clogged the system, leading to significant reductions in the case load figures during the first years of the new century.Growing pressure for adoption of some sort of súmula vinculante, or binding precedent, has led to progress on a major judicial reform amendment currently in Congress.And within the courts themselves, administrative change has streamlined case processing, while the effective use of small claims courts has helped outsource a large glut of cases arising out of past economic stabilization plans.It is too early to tell, but there are hopeful signs that together, these changes may alleviate some of the bipolar characteristics of the Brazilian judiciary, with its effective, democracyenhancing effect on policy in the high courts, and the delegitimating, particularistic nature with which citizen challenges to policy are treated lower in the judicial hierarchy.

Figure 1 :
Figure 1: Structure of the Federal Court System