Printer Friendly

Brazilian anti-corruption legislation and its enforcement: potential lessons for institutional design.

Introduction

Over the past few decades corruption has emerged as a major issue in the global development discourse as policymakers and academics have increasingly focused on the political economy factors that promote or hinder inclusive, sustainable growth. No longer dismissed as innocuous grease speeding the wheels of inefficient bureaucracies (Huntington, 1968), corruption is now recognized widely as a force that undermines economic expansion and equality, accountable and transparent governance, and social cohesion (Bardhan, 1999; Gray and Kaufman, 1998; Mauro, 1995). While earlier investigations into the causes of corruption focused on societal and cultural factors, more recently, the rational actor model and new institutional economics have emphasized the role of institutional incentives on corrupt behaviors (RoseAckerman, 1999), suggesting they may be curtailed or fostered through institutional arrangements and reforms.

Building on this framework, we investigate Brazil's struggle against corruption and the institutional lessons revealed therein. Since returning to a democratic system in 1985, enacting a new constitution in 1988, and holding direct elections in 1989, Brazil has been plagued by corruption scandals. While the country outperforms many of its regional and developmental peers on various corruption-related indicators, corruption remains a problem in many areas of public life, most notably in regional and state governments, political parties, and parliament, as well as public procurement at all levels of government (Melo, 2013). In addition, many of the metrics capturing corruption have remained relatively stable since the transition to democracy in the late 1980s, and persistent and repeated scandals reveal continued and widespread corruption in various public institutions (Carson and Prado, 2014).

In this paper, we examine those reforms and institutions that have, anecdotally and empirically, proven potent in combating corruption in Brazil. Specifically, there has been significant progress associated with the systems of oversight and investigation (Speck, 2011; Arantes, 2011) but very little progress associated with punishment (Avritzer, 2011; Filgueiras, 2011; Taylor, 2009). Highlighting the interrelationships among state accountability institutions in these arenas, we argue that the duplication of oversight and investigative functions among various governmental entities has strengthened their collective impact. After reviewing the literature condemning the inefficiencies conventionally associated with institutional multiplicity, we examine the theoretical and empirical advantages of functional overlaps in the context of corruption, in which public power is used to secure private benefits. We then explore the instances of institutional multiplicity in Brazil's systems of corruption oversight (National Audit Court, Office of Comptroller General, media, civil society) and investigation (Public Ministry, Federal Police Department, Comptroller General). We contrast the successes achieved in oversight and investigation by these competing institutions with the obstacles encountered at the punishment stage of enforcement in which a single institution--the judiciary--has authority.

We conclude by arguing that our analysis of the Brazilian experience reveals the advantages in pursuing alternative institutional avenues through institutional multiplicity, in developing strategies to reduce corruption across contexts. We emphasize how functional institutional overlaps allow for compensation, collaboration and competition among various governmental entities and have bolstered anti-corruption efforts in Brazil. We argue that the country's experience suggests that institutional multiplicity provides unique advantages in combatting a complex governance challenge like corruption.

Considering the complex and causally-bidirectional relationship between corruption and development (Lambsdorff, 2007; Rose-Ackerman, 1999; Treisman, 2007), we acknowledge that investments in institutional multiplicity in the corruption arena may not be desirable or possible in lower-income countries where resource constraints require prioritization of other developmental objectives. However, while the policy implications of this analysis may be particularly relevant for fast-growing economies that have reached some critical developmental threshold, our findings also highlight the potential for non-state actors, including the media and civil society, to provide beneficial institutional overlaps that may complement and strengthen governmental anti-corruption initiatives.

1. Incentive Structures as Anti-Corruption Strategies

By the 1990s, the economic efficiency theories of corruption (Huntington, 1968; Leff, 1964; Lui, 1985) had been largely dismissed, and researchers and policymakers increasingly embraced political economy theory to explain and combat corruption in developing countries (Filgueiras 2008). Underscoring the role institutions play in creating and perpetuating incentives that encourage or deter corrupt behaviors, political economy theory suggests that efforts to curtail corruption should focus on changing these institutional incentives. Incentive-based reforms can be divided into two types: eliminating opportunities to engage in corruption (ex-ante mechanisms), and increasing the risk of being punished (ex-post mechanisms). The former have prompted reform proposals that focused on reducing opportunities to engage in corruption, while the latter have focused on mechanisms to increase accountability.

The design and implementation of effective ex-ante mechanisms to deter government officials from engaging in corrupt acts must take account of the specific incentives faced by different categories of actors. The two core public sector audiences for anti-corruption initiatives are elected officials and civil servants. With regards to the former, reforms have focused on restructuring systems of campaign finance (Rose-Ackerman, 1999); increasing voter awareness of corruption (Brollo, 2009; Ferraz and Finan, 2008); and implementing financial and conflict disclosure regimes (Djankov, et al., 2010). Strategies to tackle corruption among bureaucrats generally target organizational structures and procedures and include meritocratic recruitment (Rauch and Evans, 2000); staff rotations (Abbink, 2004); higher wages (Abbink and Serra, 2012, pp. 94-95; Van Rijickeghem and Weder, 2001); bureaucratic competition (Ades and Di Tella, 1999; Rose-Ackerman, 1999) and the reduction of the discretion exercised by public officials through the simplification of bureaucratic regulations and procedures (Rose-Ackerman and Truex, 2012). However, while the theoretical arguments behind these and similar anticorruption initiatives may be strong, assessment of their efficacy is complicated by inherent measurement challenges (Carson and Prado, 2014; Johnson and Mason, 2013); in addition, much of the evidence is inconsistent, signifying that the impact of reforms varies significantly across contexts. For example, in a laboratory experiment with German students, Abbink (2004) found that more frequent rotation led to fewer corrupt behaviors among participants, while, in a series of randomized control trials using police officers in Rajasthan, India, Banerjee et al. (2012) discovered that reducing transfers improved the satisfaction of the public and crime victims with police performance, including on measures of corruption.

Ex-post measures focus on accountability mechanisms to ensure that corruption is detected, investigated, and punished. Within the context of the legal system, accountability first requires that states determine what constitutes impermissible corruption within the context of their specific cultural and political norms; while certain behaviors, such as bribery, are now considered indisputably corrupt across cultures (Noonan, 1984, p. 702), the status of other conduct, such as lobbying, gift-giving to public officials, and preferential hiring, varies from society to society, legal regime to legal regime. Many international organizations, NGOs and transnational initiatives have tried to promote the enactment of anti-corruption legislation as well as transnational policy convergence around corruption issues through international mechanisms, such as the OECD Convention against Foreign Bribery. Such mechanisms often encourage punishing private sector parties who engage in corrupt activities, as well as corporations and other legal persons whose employees, officers, or agents participate in corruption (OECD Convention, 1997, Art. II).

Once a legal rule or standard is in place, the proposals to increase accountability generally revolve around the creation or strengthening of mechanisms to enforce those legal provisions. In this regard, O'Donnell (1998) has usefully distinguished systems of horizontal accountability, i.e. government agencies that perform checks and balances, versus systems of vertical accountability, i.e. non-governmental entities, such as civil society and the media that exercise an active role in controlling those in power. For the institutions of horizontal accountability, reform efforts have included the establishment of independent judiciaries and autonomous anti-corruption agencies with strong investigative powers (World Bank, 2000). More recent literature has argued that the effectiveness of accountability systems depends on the "web" of institutions engaged in three core functions: oversight, investigation and punishment (Taylor and Buranelli, 2007; Power and Taylor, 2011). This paper focuses primarily on how these "webs" of oversight, investigation, and punishment institutions affect ex-post mechanisms of horizontal accountability in the context of corruption, focusing on the example provided by Brazil.

2. Institutional Design and Interdependence in Anti-Corruption Strategies

The academic and policy debates regarding anti-corruption strategies are often focused on the tradeoffs involved in specific institutional design choices. Do anti-corruption organizations perform better when they serve a discrete, circumscribed role, or when their mandates extend to cover a range of related functions? For example, do anti-corruption agencies, which are primarily investigative bodies, function more or less effectively if also empowered with some oversight and/or punishment authority (Bolongaita, 2010)?

On the one hand, concentrating multiple functions into one institution may reduce its dependence on other institutions. For instance, if functions are dispersed across various entities, a flawed system of monitoring may undermine the effectiveness of a system of investigation, regardless of how powerful and well equipped those in charge of investigations are. On the other hand, having separate institutions focusing on one particular function is likely to improve the gains of specialization, which may make an institution better capable of performing its particular function. However, it creates the problem of interdependence, which is often a central issue for a well functioning system (Pope, 2000; Mainwaring, 2003). The idea of interdependence has been used, for instance, to explain the effectiveness (or lack thereof) of the web of accountability institutions in Brazil (Taylor and Buranelli, 2007; Power and Taylor, 2011). Such interdependence has also been identified as the most important factor in determining the performance of autonomous audit agencies in Latin America (Santiso, 2007).

Power and Taylor (2011, pp. 13-14) present two arguments to support the idea that interdependencies are an important factor in an effective accountability system. First, the completion of each one of the steps in the accountability process is largely dependent on the preceding and succeeding steps. Thus, without oversight, there is no investigation, and without investigation there is no effective punishment. Less intuitively, they argue that the interdependence goes in the other direction as well: if punishment is unlikely to happen, this reduces incentives for effective oversight and investigation.

The second argument regarding interdependence concerns the potential advantages of having multiple institutions able to perform the same effective function. For instance, most accountability systems contain a multitude of punishments that reinforce each other: "at least four types of overlapping sanctions may contribute to curbing political corruption: electoral sanctions, such as failures to win reelection; political but non-electoral sanctions, such as congressional censure or removal from office; reputational sanctions, such as negative media coverage; and legal sanctions, such as criminal or civil judgments" (Power and Taylor 2011, p. 14). In other words, the fact that one type of sanction is effective increases the likelihood of other sanctions being effective as well.

An important policy implication of the interdependence argument is the recognition that corruption is a systemic problem that cannot be solved by a single governmental agency or unit, no matter how powerful. For example, over the past few decades, more than 60 countries have established anticorruption agencies (ACAs), "separate, permanent agencies whose primary function is to provide centralized leadership in core areas of anti-corruption activity" (Meagher, 2005, p. 70). However, the performance of these ACAs has been mixed; while those in countries with strong governance institutions and relatively low levels of corruption such as Hong Kong, Singapore, Chile, and Australia have achieved success, those in developing countries characterized by weaker governance, such as Kenya, Malawi, Nigeria, Sierra Leone, Tanzania, and Uganda, have not proven effective (Huther and Shah, 2000; Shah, 2007).

While some of the disparities in results may be attributed to resource constraints, measurement challenges, or lack of political will (Doig, Watt, and Williams, 2005; Johnson, Hechler, De Sousa, and Mathison, 2011), they also reflect the reality that even autonomous, well-resourced ACAs with strong popular and internal support must interact with other public entities. Thus, the effectiveness of an ACA will depend, to a large degree, on the performance of complementary governance institutions such as the judiciary, prosecutors, ombudsman, and auditing body, as well as ministries in areas in which corruption is particularly prevalent, e.g., revenue and public works (Meagher, 2004; UNDP, 2005). In sum, regardless of how centralized the accountability mechanisms are, some degree of institutional interdependence is unavoidable.

3. Institutional Multiplicity in Anti-Corruption Strategies

Building on Taylor and Power's second point about interdependence, we want to investigate the importance of institutional multiplicity in increasing the effectiveness of the "web" of accountability institutions in the context of corruption. By institutional multiplicity we refer to the idea that more than one institution is charged with performing a certain function. While Taylor and Powell's example contrasts legal sanction with political, electoral and social sanctions, we focus more specifically on institutional multiplicity within the larger legal system. For instance, administrative, civil and criminal sanctions can be defined as competing sanctions that may be interrelated to each other or not, and applied and enforced by the same institution or not. With this legal focus, we are interested in investigating the effect of having institutions performing the same function(s) in each of the steps of the accountability process: monitoring, investigation and punishment. Our question is to what extent the existence of institutional multiplicity in each of these stages helps or hinders efforts to combat corruption.

While the idea of institutional multiplicity has yet to be considered vis-avis ex-post mechanisms and the web of accountability institutions, it is certainly present in corruption literature regarding ex-ante mechanisms. The argument against monopolies in the provision of services or performance of functions has been discussed as an ex-ante mechanism to reduce the opportunities for corrupt behavior in bureaucracies. Susan Rose-Ackerman (1978, 1994) argues that giving officials competing jurisdictions provides an alternative to the private party who does not want to engage in corruption. There is however the risk of unintended consequences: while competing jurisdiction may decrease bribes, it can increase the amount of theft from the government (Shleifer and Vishny, 1993). Increased competition can also create more opportunities for corruption: if I want to obtain a license that I do not qualify to obtain, having two officials to approach with a bribe, rather than one, may increase the chances that I will be successful. Also, the implementation of an effective system of institutional multiplicity depends on the possibility of establishing competing jurisdictions which may not be possible due to limited resources or the type of service delivered. Moreover, it requires the creation and cultivation of an institutional structure in which competition creates incentives to improve performance; if institutional multiplicity merely facilitates shirking by one or more employees or agencies, it will be ineffective in helping to curtail corruption and may simply waste resources (Bardhan, 1997).

The idea of institutional multiplicity is also present in the broader rule of law and development literature. For instance, Thomas Heller (2003) has recommended institutional multiplicity that fosters competition as a strategy to overcome obstacles to rule of law reforms in developing countries. Effective change in established organizations, Heller argues, can only come about when they are motivated by incentives that come with competition, and he notes that competition between legal institutions already occurs within the developed world. Examples include the "horizontal" competition that arises when actors within a federal system are able to select between the various regulators established by sub-national governments, and the transnational dispute resolution mechanisms of the kind established by NAFTA, which permit litigants to review administrative decisions in Canada, the United States and Mexico. Heller argues that in cases such as these, where multiple legal organizations have non-exclusive jurisdiction, the ability of the incumbent institution to resist change is reduced.

In addition to the incentives generated by competition, another potential advantage of institutional multiplicity is compensation: if one of the institutions fails to perform its functions, another one is equipped to fill the resulting gap. This could reduce the risk of failures in each step of the accountability process. A third potential advantage of institutional multiplicity comes out of collaboration which may be advantageous simply because there are more human, financial, and other resources available for the performance of a single task. A fourth advantage of multiplicity is complementarity, which may be especially advantageous due to specialization (two different institutions may contribute different skills to perform a particular task, and these different sets of skills complement each other).

There are, however, potential disadvantages associated with institutional multiplicity. First, the creation of institutional overlaps may be interpreted as an inefficient allocation of resources, particularly in low-income developing countries where scarce fiscal resources already struggle to provide adequate coverage for other societal needs, such as education and health. Second, the competition engendered through institutional multiplicity may be destructive, creating unproductive tensions between two institutions performing the same function. An extreme example would be a situation in which one institution attempts to promote itself by undermining the efforts of another institution. Third, if accountability institutions themselves fall prey to corruption, institutional multiplicity may create more opportunities for corruption; for example, if authorities from multiple (corrupt) investigation institutions are able to extract bribes by threatening innocent citizens with false charges, the overall risk and incidence of corruption may increase.

We acknowledge these limitations, which should be considered in a careful cost-benefit analysis on a case-by-case basis, taking into consideration the resources, capacities, and policy needs within individual countries or societies. In this paper, we simply argue that institutional multiplicity offers a promising approach for those societies that have identified anti-corruption as a policy priority.

4. Institutional Multiplicity in Brazil

Brazil possesses an extensive stock of anti-corruption legislation. Its anticorruption laws comply with international standards, criminalizing the six offenses considered core corruption-related misdeeds--bribery of public servants (active bribery); solicitation or acceptance of gifts by public servants (passive bribery); abuse of public position for personal gain; possession by a public servant of unexplained wealth; secret commissions made to or by an employee or agent (covering private sector corruption); and bribes and gifts to voters (Pope, 2000, p. 284; U4, 2010). Moreover, the country has taken legislative action across a range of other areas considered critical to combatting corruption, including laws addressing access to information, conflicts of interest, public procurement, freedom of the press and expression, and the powers and functions of the government Ombudsman (Pope, 2000, pp. 26970). In sum, "by almost all accounts, Brazil has an appropriate spectrum of laws in place to combat corruption" (Stocker, 2012).

Brazil also boasts a wealth of accountability institutions charged with monitoring, investigating and sanctioning those involved in corruption. While this section focuses on these institutions, it does not provide a complete list or a detailed description of the governance structures and competencies of these institutions as this work has been done elsewhere (Power and Taylor, 2011; Carson and Prado, 2014). Instead, we focus our analysis on the institutions engaged in each of the three accountability functions--monitoring, investigation and punishment--in order to resolve two questions: (i) is there some level of institutional multiplicity in each of these three accountability functions in Brazil?, and (ii) is there any evidence connecting the existence (or lack) of institutional multiplicity to the overall performance of the Brazilian accountability system?

Our main argument is that institutional multiplicity may explain, at least partially, why the Brazilian accountability system at the federal level, which has demonstrated strong vitality in monitoring and investigation, has proven so ineffective at punishing corruption.

4.1 Institutional Multiplicity in Oversight

At the federal level, there is a multitude of institutions performing constant monitoring of the government in Brazil, but our analysis focuses on only two: Tribunal de Contas da Uniao (Federal Accounting Tribunal or Federal Audit Court, TCU) and Controladoria-Geral da Uniao (Office of Comptroller General, CGU).

TCU is the external oversight body for the legislative and executive branches (Carson and Prado, 2014). Despite being formally touted as an archetypical auditing institution (Santiso 2007, Melo 2013), three characteristics of the TCU undermine its efficacy in fighting corruption. First, TCU officials spend most of their time on routine monitoring tasks such as the preparation of the audit statement of the government's annual accounts, review of annual financial reports from all units of government managing public funds, and approval of policies related to civil servants' employment and retirement, rather than detecting irregularities or analyzing systematically areas in which there is greater risk of corruption (Speck, 2008; Melo, 2013). Second, the TCU's governance structure is perceived to be dysfunctional. Despite being independent from Congress, and relying on a cadre of highly qualified and professional civil servants, the top echelon of politically-appointed and organizationally-powerful ministers has strong incentives to block politically sensitive issues and topics (Melo, 2013; Speck, 2008). Thus, as illustrated by the labor court case (TRT case) described below, the TCU is vulnerable to political capture (Santiso, 2007; Rocha, 2003). Third, the liberal accessibility of judicial appeals limits the ability of the TCU to punish effectively those involved in wrongdoing. While, as described below, the TCU may impose administrative and civil penalties, regular courts often strike down such sanctions or take so long to decide on these cases that they end up being closed due to the statute of limitations (Santiso, 2007; Speck, 2011; Melo, 2013).

While these three features impose severe obstacles to the effective performance of the TCU, there seems to be anecdotal evidence to support the idea that institutional multiplicity has helped, to a certain extent overcome these hurdles.

Starting with the emphasis on routine procedures, the formalistic and ossified auditing processes of the TCU failed to catch the municipal ambulance kickback scheme later uncovered through Operation Bloodsucker. In the scheme, "congressional aides had been bribed to write individual budget amendments financing the purchase of these ambulances with Health Ministry Funds" (Power and Taylor, 2011, 3; Carson and Prado, 2014). The scheme was not caught in TCU's auditing process, but was unveiled by another federal auditing institution, the CGU.

CGU is part of the executive branch. Despite being responsible for internal accountability within the executive branch, CGU can be considered an instrument of horizontal accountability for the purpose of this paper (Olivieri, 2006). As part of a program of random municipal audits (Ferraz and Finan, 2008; CGU, 2014), in 2004 the CGU came across irregularities in the public procurement processes used in some municipalities to purchase ambulances for the public health care system. Upon finding these irregularities, the CGU sent a document to the Minister of Health indicating that there was a group manipulating procurement processes at the local level across the country and embezzling public funds through the sale of overpriced ambulances to municipalities. The CGU asked the Minister to take the necessary measures to address the failures in procurement processes, while at the same time alerting the Federal Police about the case. In 2006, the federal police unveiled that a total of 1,000 ambulances had been purchased in this scheme, involving a total of US$55 million. In the same year, Congress started its own internal investigation through a CPI (Comissao Parlamentar de Inquerito). The CPI recommended that 72 congressional representatives be removed from office for being involved in the scheme, but none were expelled or faced any other penalty (Taylor, 2011; Rocha, 2012). The lack of sanctioning is a problem that we will address below.

The CGU's ability to catch an irregularity that had not been detected by the TCU is partially due to the fact that CGU analyzes the effectiveness of government programs, not only the formalities associated with expenditures (Loureiro et al. 2012). In the case of the bloodsuckers scheme, the CGU found that many municipalities had purchased non-operational or used vehicles despite the fact that the government had paid for new ones, and that many cities that had initiated procurement processes to acquire more vehicles had recently bought ambulances that were not in use (Folha de Sao Paulo, 2006a). These findings triggered a thorough investigation by the Federal Police (Folha de Sao Paulo, 2006b).

The bloodsucker scandal may be an example of compensation or complementarity. Some may argue that the TCU's failure to detect the scheme reflects deficiencies in its auditing process, while the CGU's success in identifying the irregularities indicates that their auditing methods are more effective. If so, this would be a case of compensation. On the other hand, one may claim that this is simply a result of distinct auditing methods. The monitoring techniques of the CGU were specifically designed with different parameters than those used by the TCU in order to increase the likelihood each institution could catch things undetected by the other (Interviews at TCU, April 2014). This is a case of complementarity. More specifically, in the detailed analyses required to assess program effectiveness, auditors came across details that revealed irregularities that are not captured by the analysis conducted by TCU. Regardless of which interpretation of the fact one adopts (compensation or complementarity), the point is that, in an attempt to analyze the effectiveness of a federal program at the local level, the CGU has uncovered a series of irregularities in the use of federal funds. Thus, this illustrates how overlapping oversight functions may increase the chances of spotting wrongdoing.

Institutional multiplicity has also been effective in overcoming the second problem of TCU, which Melo (2013, p. 21) describes as "a lack of connection between the professional work produced by [the TCU's] cadre of auditors and the political logic that underlies decision-making at its top decisionmaking body". There is some dispute as to how much this affects the functioning of TCU. (1) A political scientist claims that such impact is high: "the recommendations contained in the reports prepared by the TCU's technical personnel are usually not followed by its board of ministers for political reasons" (Figueiredo, 2003, p. 185). Based on that, another political scientist concludes that "[the TCU's] effectiveness depends primarily on the extent to which other actors--such as the media or opposition legislators--can publicize its audits" (Melo, 2013, p. 23). In this case, there is compensation, as other institutions are stepping in to compensate for the failures of the original institution.

A good example of the importance of institutional multiplicity in providing an alternative channel for the technical cadre of auditors to seek further investigation of suspicious activity is the case of corruption scandal involving the labor court in the state of Sao Paulo (also known as the TRT case). As Power and Taylor (2011, p. 2) describe "the cost of building the Regional Labor Court in Sao Paulo during the 1990s was inflated nearly fourfold, with proceeds on the order of US$100 million allegedly appropriated by Judge Nicolau dos Santos Neto (commonly referred to as "Lalau"), with the participation of a senator, Luiz Estevao, and the president and vice president of the construction company that won the building contract." The scandal came to the fore in 1998, but the auditors of the TCU had called attention to numerous irregularities back in 1992, when the building was simply a project.

An analysis of the paperwork revealed so many irregularities, that the 1992 auditing report suggested cancelling the contract with the construction company, withholding any future transfers from Congress to the project, the return of all money already transferred to the federal government and a thorough investigation of the project. In 1993, the report reached the highest echelon of the TCU, but the minister initially assigned to the case kept the case moving at glacial place, allowing for the construction (and the corruption scheme) to move forward. In 1995, the minister initially assigned to the case retired, and the new minister in charge took an entire year to evaluate it. In 1996, the TCU issued a decision: there were indeed irregularities, but since the construction had already started, and a lot of money had already been spent, the minister argued that interrupting project at this point would be costlier (due to the waste) than allowing for it to be concluded. It is estimated that those involved in the scheme embezzled around US$35,000/day during this period, which included kickbacks to senators who approved the transfer of federal funds to the construction project (Veja, 2000).

The TCU's failure in monitoring this project was remedied by a member of Congress, the Public Prosecutor's office and the media. (2) In 1996, a senator from an opposition party (Mr. Giovanni Queiroz) had blocked a special transfer to the project (amounting to US$5.5 million) during a regular congressional session, only to find out that the same transfer was eventually approved in December, when the house was in recess and only a few representatives remained at work to deal with urgent matters. Mr. Queiroz approached the Public Prosecutors' Office suggesting an investigation of the project. After assessing that the lifestyle of Lalau, which included a US$1million apartment in Miami, luxury cars and lavish parties in expensive restaurants, was incompatible with his means, the Public Prosecutors' Office started an investigation and in 1997, uncovered the connections between all the parties involved and the money transfers that siphoned public resources to private accounts (Veja, 2000).

In 1998, the Public Prosecutors' Office had collected enough evidence to request the judiciary to sequester all the assets of Lalau and all others involved in the scheme. After that, the scheme became public news and was covered heavily by the media. Public pressure led the judiciary to remove Lalau from his position as chief justice of the labor tribunal in September of that year, and forced the TCU to conduct a new audit of the project. This time, the TCU was able to determine quickly that there were irregularities and to decide that the construction should be suspended. It was only then that Congress suspended transfers of federal funds to the project, as recommended back in 1992. In 1999, in the aftermath of the scandal, in a move guided by electoral motives, Congress set up an investigation committee (Comissao Parlamentar de Inquerito) to assess corruption in the judiciary in general, but gave special attention to the labor law court. It was then revealed that Lalau alone had embezzled more than US$50 million in the scheme (Santiso, 2007, p. 26).

In sum, this example illustrates that the existence of alternative mechanisms of oversight and monitoring, especially if the institutions with overlapping jurisdiction have different governance structures and therefore different systems of incentives, may address the shortcomings of the existing auditing institutions. In this case, the TCU failed to act, but the error was "fixed" by a member of Congress, the Public Prosecutors' office and the media. Moreover, it is noticeable that after the creation of the CGU, the TCU has increased its activity. The MESICIC Committee (2012, p. 20) notes that the total number of processes (including audits, inspections, consultations, complaints, etc.) conducted by the TCU rose from 6,135 in 2006 to 8,019 in 2010. While no causal connection between these two events can be empirically tested or proven, they suggest that institutional multiplicity and the mechanisms through which it may impact on the operation of accountability institutions is something that deserves further investigation.

4.2 Institutional Multiplicity in Investigation

The central institutions performing investigative functions in Brazil are the Federal Public Prosecutors' Office or MPF (Ministerio Publico Federal) and the Federal Police or DFP (Departamento da Policia Federal). Oftentimes the MPF will conduct criminal investigations in collaboration with the DFP, especially in criminal cases. Indeed, there has been an increase in the number of investigations in the last 5 to 10 years (Carson and Prado, 2014), which appears to be the result not only of an increase in resources for DFP but also increased cooperation between DFP, MPF and other investigative bodies, such as state MPs, Revenue Service Inspectors and ministries (Arantes, 2011, p. 200 and 205). In many cases, joint task forces have been formed in order to better coordinate investigations. The results appear positive (Arantes, 2000), providing clear evidence of a situation in which some level of institutional multiplicity has led to collaboration.

While there seems to be a significant amount of collaboration between MPF and DFP in high profile cases (also known as "operations"),3 the same is not true for day-to-day investigations. In these cases, even quick consultations between the DPF and the MPF are rare. If the case involves a "serious" violation, the MPF consults with the DPF once a month, and all communication is on paper (memorandums). This frequency of communication is fairly low compared to other countries such as the United States and can hardly be classified as cooperation. While the MPF may request additional information, such petitions must be done via publicly-available memorandums, reducing the speed and efficacy of the investigative process. If the DPF does not comply with the MPF's request, the MPF may step in and conduct that investigation on its own, but such inquiries are rare. However, the constitutionality of such investigations remains unresolved. A case involving the issue has been pending before the Brazilian Supreme Court (STF) since 2008 (STF, RE 593727/2008), and, in the absence of a firm resolution regarding the legal status of such independent investigations, the STF has made clear those cases raise suspicions and may be invalidated. In addition to this constitutional ambiguity, the exercise of such powers faces political opposition; a bill proposing to eliminate MPF's investigative authority was introduced but voted down by the Brazilian Congress in 2013 (O Globo 2013). Given these legal and legislative developments, public prosecutors are very careful and selective about the cases in which they proceed alone (Authors Interviews, April 2014).

Despite being the exception rather than the rule, in a few cases the investigative powers of MPF seem to have compensated for the lack of police action. The cases in which MPF played a prominent role include the recent mensalao case. Indeed, according to the Attorney General (Procurador Geral da Republica) it would not have been possible to prosecute and convict those involved in the mensalao case without the investigation conducted by the MPF (Terra, 2013). However, given the current legal and political climate, it remains uncertain whether the MPF will retain these powers moving forward. If they are curtailed or, alternatively, confirmed or enhanced at some point, it will be an opportunity to test the hypothesis advanced in this paper. Such a change could allow assessment of the impact that institutional multiplicity may have on the overall level and success of criminal investigations in Brazil.

While there is multiplicity in criminal investigations, the DPF has no jurisdiction over civil investigations. In actions concerning administrative improbity, the MPF is the only institution that can officially conduct a civil investigation (called inquerito civil). Despite being only conducted by the MPF, this type of investigation can benefit from investigations in the criminal and administrative spheres. Since they relate to the same facts or actions, often the investigative efforts of MPF are similar to investigative efforts for administrative processes conducted by CGR (Corregedoria Geral da Uniao, the arm of CGU in charge of investigations). Similarly, civil and administrative investigations can feed into criminal investigations and vice-versa. Indeed, CGU has been working closely with the DPF in most of its operations (Interviews with members of CGR, April 2014). In such cases, while each institution may be focusing on a different aspect of the investigation, they can complement each other's work.

Two recent examples of this type of complementarity are the operations Two-Way Road (Mao Dupla), regarding a company hired by the government to build roads (Delta), and 13 of May (13 de maio), concerning the embezzlement of education funds by municipal officials. In the first case, in 2010, cooperation between the DPF and the CGU uncovered a scheme in which civil servants from the Department of Roads and Transportation of the State of Ceara allowed the company contracted by the Federal Department of Infrastructure and Transportation (DNIT) to charge more for the services and use lower quality materials than the ones specified in the contract in exchange for bribes. In the second operation, 13 of May, officials from the CGU and the DPF uncovered a scheme in which 26 municipalities in the state of Bahia were misappropriating funds from Fundeb (the national program to finance primary education). In most cases, civil servants and mayors had created shadow companies that pretended to sell services to the municipality (e.g., school buses, event organization services, etc.), but were simply redirecting federal funds to bank accounts owned by the people involved in the scheme.

Without discounting the potential of multiplicity to foster unproductive competition, we hypothesize that multiplicity has enhanced the investigatory capacity of Brazilian institutions, because it has increased the likelihood that an investigation may take place. Without a counterfactual scenario, it is hard to provide evidence that the increase in investigatory capacity was due to institutional multiplicity rather than initiatives started in 1998 and enhanced after 2003 to strengthen the DPF (Arantes, 2011). With increased budget, more personnel, a newly defined focus on corruption and the use of catchy names for operations to gain easy publicity, the DPF has become an increasingly potent force in fighting corruption in Brazil. Nevertheless, it is noticeable that the increase in criminal investigations and operations at the DFP between 2005 and 2009 (MESICIC, 2012, p. 27) has been followed by an increase in civil investigations initiated by the MPF on corruption and administrative impropriety between 2007 and 2011 (MESICIC, 2012, p. 36). Although we cannot prove a causal connection, our hypothesis is that the strengthening of the DPF may have had a positive impact on the MPF's performance and vice-versa.

While multiplicity may result from having independent institutions that can conduct investigations which will lead to criminal corruption charges (inter-institutional multiplicity), there is also some level of multiplicity in the fact that individual actors are independent to act within each of these institutions (intra-institutional multiplicity). For example, at the MPF, if one prosecutor receives information indicative of corruption and decides not to investigate the case, another prosecutor is free to proceed with the investigation. As argued by Sadek and Cavalcanti (2003), this lack of centralization which we consider an example of intra-institutional multiplicity--has largely strengthened MPF's power to act as an anti-corruption body. In contrast to the organizational structure of the TCU, described above, which grants politically-appointed ministers the discretion to halt inquiries into detected accounting or budget irregularities, the MPF's intra-institutional competition model may facilitate effective investigation and thus accountability for corruption-related offenses.

4.3. Institutional Multiplicity in Punishment

The level of institutional multiplicity in punishment in Brazil is considerably lower than oversight and investigation. Independently of the civil and criminal sanctions imposed by the judiciary, the CGU, TCU and internal accountability bodies can impose administrative sanctions on actors found to have engaged in corrupt activities. However, the ultimate sanctioning authority in Brazil remains the judiciary, which possesses the power to review and overturn punishments imposed by other entities (Federal Constitution, art 5, XXXV). As a result, Brazilian courts hold a monopoly power over the sanctioning of corruption in the country. Thus, unlike the other functions described here, the institutional multiplicity that exists in punishment can potentially be nullified by the fact that one single institution has the power to nullify all other sanctions. While institutional multiplicity still exists in the form of non-legal sanctions (e.g. reputational and electoral), at the level of legal sanctions, it is possible to say that Brazil is very close to having no institutional multiplicity in punishment.

This lack of institutional multiplicity may reduce the chances of punishment, because if the one institution in charge fails to perform its function, there is no other institution to fill the gap and correct the error. And this is exactly what is happening in Brazil. The Brazilian judiciary is characterized by very low rates of corruption convictions (Carson and Prado, 2014). Such underperformance is attributable not to sporadic errors in execution but rather to fundamental structural problems such as excessively formalistic processes, burdensome procedural rules and corruption. (Prillaman, 2000; Taylor, 2009; Taylor, 2011) These problems not only affect the enforcement of civil and criminal sanctions, which depend on a judicial decision, but they may also interfere with the enforcement of administrative sanctions. Administrative sanctions, in theory, do not depend on the judiciary to be enforced. However, all these decisions can be appealed to the judiciary. Once this happens, the judicial review suffers from the same ills that affect the civil and criminal procedures undermining the effectiveness of the administrative sanctions. For instance, the effectiveness of the sanctions imposed by TCU has been largely reduced by the availability of judicial appeals (Beck 2011).

There are a series of reforms in Brazil that can be interpreted as attempts to create institutional multiplicity in punishment for corruption. The heavy burden of proof in criminal cases has served as a motivation for an important reform that took place in 1992. The Law of Administrative Improbity of 1992 was enacted to expedite corruption cases and to empower the MP's office as a body of horizontal accountability. Prosecutors could choose between bringing civil or criminal charges depending on the amount and quality of evidence collected. As Arantes (2011) explains, this law provided an alternative to rather burdensome criminal trials as civil trials were expected to have a lower burden of proof. Available civil penalties included the removal of a public official from office, temporary suspension of political rights, and reimbursement to the public coffers.

However, the 1992 reform did not produce the expected results, as the strategy continued to rely on the Brazilian judiciary still plagued by the problems described earlier; very few civil corruption cases brought since 1992 have reached conclusion and resulted in sanction (Arantes 2011, MESICIC, 2012). "[T]aking stock of fifteen years' worth of experience with administrative improbity cases suggests that they are not entirely effective in the courts, whether as a result of the slowness of the proceedings, numerous dilatory appeals, or, frequently, judges' concerns about MP's authority to act in this arena, as when they fail to recognize the legal legitimacy of suits or the legality of procedures adopted during the investigation. Of 572 suits brought by the prosecutors in Sao Paulo since 1992, for example, fewer than 10 have reached a definitive conclusion to date." (Arantes, 2011, pp. 198-9) Indeed, the average time for trying cases concerning administrative improbity was approximately 5 years in the regional federal courts, which is quite lengthy (MESICIC, 2012, p. 47).

To remedy these procedural problems, many administrative bodies have been searching for administrative sanctions that have immediate effect. While these remain subject to judicial appeal, in the absence of an injunction, the judiciary's glacial pace guarantees that the sanction will be in place for a few years even if ultimately revoked. Indeed, the TCU has been concentrating more on imposing sanctions that can be made effective immediately and generate financial and reputational costs. Removal from office and blacklisting of companies that are blocked from public procurement processes are two examples (Speck, 2011). The CGU has been doing the same, while also ensuring that the administrative procedures closely follow the requirements imposed by the judiciary, in order to reduce the chances of the decision being reversed (Interviews with CGU officials, April 2014). Considering the low level of effectiveness of judicial sanctions, there are reasons to believe that currently the administrative sanctions are more effective in curtailing corruption in Brazil (Alencar, 2010; Alencar and Gico, 2011). Based on this assumption, the new Anti-Corruption Law that came into effect in January 2014 imposes mostly administrative sanctions on companies engaged in corruption. The bill was drafted in a way that avoided the judiciary as much as possible, as any judicially imposed sanction would drastically reduce its effectiveness (CGU, MJ, AGU, 2009, par. 10; Authors Interviews, May 2014).

Employing the existing judicial system to their advantage, a number of institutions have been trying to create a form of institutional multiplicity in punishment by resorting to non-legal sanctions. Indeed, the MPF has used investigations with strong media repercussions to impose reputational sanctions (Arantes, 2011, p. 199). Similarly, the CGU has maintained an open channel of communication with the public, operating sometimes as a mechanism of vertical accountability (Loureiro et al, 2012). Publicizing the list of companies blacklisted by the government seems to have generated reputational consequences, as other companies may also refuse to contract with or hire blacklisted companies (Authors Interviews, April 2014). Publicizing the reports of municipal audits, in turn, has generated electoral consequences (Ferraz and Finan, 2008). Also, CGU has occasionally leaked information about ongoing investigations in the media in an attempt to deter future wrongdoing. CGU has collected data that suggesting that irregularities declined after media reports of similar abnormalities (Authors Interviews, May 2014). This, of course, may mean that the practice has ceased, but it may also mean that it was simply structured in a different way. In any event, Brazilian governmental and non-governmental institutions are trying to create, to the extent they can, some level of institutional multiplicity by strengthening the impact of extra-legal punishment, so as to allow sanction to become effective without depending on the judiciary.

Perhaps the most relevant attempt to create some level of institutional multiplicity in legal punishment in Brazil is the Clean Record Law (Lei da Ficha Limpa) enacted in 2010. The Law prohibits candidates that have been sentenced for crimes or violations of electoral statutes from running for office. One of the most important features of the statute is that it authorizes another branch of the Brazilian judiciary--the Electoral Courts--to impose an electoral sanction (bar someone's from running in the next elections) based on a decision imposed by a regular court for a criminal offense. The electoral sanction is effective even if the criminal decision is subject to appeal. In other words, even if the criminal sanction is ultimately nullified, the Clean Record Law allows another sanction, imposed by another institution, to be derived from the first condemnation. The system seems to have been effective, barring 330 candidates in the year following its enactment (Panth, 2011). In this way, the sanction is not undermined by the cumbersome system of appeals that largely undermine the effectiveness of other legal sanctions in Brazil.

4.4. Risks and Challenges of Institutional Multiplicity

While the above examples and analysis highlight the advantages institutional multiplicity can offer in efforts to combat corruption, such a strategy clearly involves tradeoffs. Specifically, institutional multiplicity may exacerbate two of the problems that have motivated the creation of dedicated anti-corruption agencies (ACAs) in many countries: coordination problems and lack of expertise. Theoretically, the centralization of all anti-corruption activities into a single ACA should facilitate the sharing of information and intelligence, thus greatly reducing coordination problems, as well as allow managers and staff to gain specialized experience with and knowledge of the particular issues surrounding corruption (Meagher, 2005). While, as noted earlier, ACAs have often failed to meet such expectations or to contribute meaningfully to corruption reduction, the challenges presented by coordination problems and lack of specialization loom over the institutional multiplicity strategy and are examined in this section.

There are three types of legal punishment that can be imposed on someone engaged in corruption in Brazil: administrative, civil and criminal. Each is determined by separate judicial or administrative processes that run independently from one another (Law 8,112/90, art. 125). Our institutional multiplicity argument suggests that the independence among the processes might be positive: if one process is flawed, another may be able to compensate. However, the independence among the processes can also lead to inefficiencies and unnecessary duplication of efforts. A possible example of this is the simultaneous but non-integrated civil and criminal lawsuits in the Brazilian judiciary. As illustration, an in-depth analysis of the TRT case mentioned earlier shows that there were more than ten different proceedings related to the scandal pending before Brazilian courts as of May 2014 (Machado and Ferreira, forthcoming). This is a result of what Machado (2013) calls the "compartmentalization of legal knowledge," which demands that different spheres of the law (administrative, civil and criminal) deal with the same set of events independently of each other, despite clear interconnections. In the specific case of the TRT, this "compartmentalization" has been taken to the extreme, as subdivisions within each of the three spheres have led to an unnecessary repetition of proceedings, evidence and production of documents in cases that have included criminal trials for the accused individuals, appeals of the administrative sanctions, civil restitution proceedings to recover the embezzled funds, and bankruptcy proceedings involving the company at the center of the scheme (Machado and Ferreira, forthcoming).

While it may be advisable to keep all these cases separate under certain circumstances, such as when dealing with different evidentiary standards, in other cases there may be advantages in coordinating prosecutorial efforts. In the particular case of the TRT, enhanced communication and collaboration between those conducting the civil and the criminal trials could have saved a significant amount of resources for the legal system (Machado and Ferreira, forthcoming).

The lack of communication and coordination in the TRT case contrasts with the more recent examples of collaboration between the CGU and DPF. Indeed, such cooperation seems to be building on the fact that the Brazilian courts have authorized sharing of evidence provided in the criminal and the administrative spheres. Thus, evidence collected by the DPF in the criminal investigation can be used by CGU in the administrative process, and vice-versa (Alencar, 2010, p. 31). Bolstered by institutional rules and structures that allow and facilitate such cooperation, the CGU and DPF have increased efforts to share evidence and collaborate on cases. This type of coordination may not only reduce system inefficiencies, but can also enhance each entity's individual capacity to fight corruption in its respective sphere, thus strengthening each institution's ability to perform its own role effectively.

The lack of coordination is not only a problem in the judiciary. As discussed earlier, the individual independence of Brazilian public prosecutors, can be perceived as a form of intra-institutional multiplicity that is advantageous for the accountability system as a whole. However, it may also lead to duplication of efforts and inefficient use of resources. Individual prosecutors may not dialogue with other prosecutors working on diferent aspects of the same case, and they may also not share information about best practices that could potentially enhance efficiency (Authors interviews, April 2014). Thus, prosecutorial independence may also foster coordination problems that can ultimately impair anti-corruption efforts.

An example of an initiative that tries to curtail such problems is the State Strategy to Fight Coorruption (Programa Estrategia Estadual de Combate a Corrupcao, ECCO). ECCO won the Innovare Institute Prize for Best Strategy to Fight Corruption created by the Public Prosecutors office in 2013 (Innovare, 2013). ECCO provides individual prosecutors with a "toolkit of best practices" to adopt preventive measures to fight corruption. The use of the toolkit is discretionary, so the program does not impair the individual independence of prosecutors. Instead, the idea is that adoption of the tools by MP offices throughout the state will enhance efficiency by reducing the time each individual prosecutor needs to invest in order to determine the best course of action. Moreover, by providing prosecutors with a set of best practices, the toolkits foster consistency in prosecutorial criteria for adopting preventive measures, reducing both the risk and perception that prosecutions may be politically motivated. The initiative was well received: 75% of the offices of the MP in state of Rondonia have agreed to participate in the initiative. Unfortunately, there is no data on the effectiveness of the measures adopted by the individual offices.

From an institutional design perspective, these examples suggest that the supposed tradeoff between independence and coordination can be addressed effectively through institutional malleability. Creating institutional structures that allow--but do not require--otherwise independent entities to coordinate when feasible and beneficial can encourage efficient and effective inter- and intra-institutional collaboration while protecting organizational autonomy. In the context of the TRT cases, if the legal system had maintained its existing institutional multiplicity with separate spheres for civil, criminal, and administrative cases but introduced an element of malleability to permit prosecutors from those spheres to share evidence to the extent practicable and helpful, resources could have been saved while the integrity of the system was upheld. The ECCO initiative offers another example of the benefits of combining institutional multiplicity with malleability: it provides the tools for coordinated action but leaves room for independent action as well.

As the examples in the previous sections illustrate, the effects of institutional multiplicity in oversight and investigation in Brazil have not been uniform, sometimes resulting in compensation, sometimes complementarity, and, in other cases, collaboration. This diversity reflects the existence of malleable institutional structures that permit but do not force coordination. The design of structures that encourage and nourish coordination when advantageous for the system as a whole is clearly a challenging task. Praca and Taylor (2014) provide a map of the evolution to the web of accountability institutions in Brazil suggesting that it has been characterized by self-reinforcing reforms in multiple institutions, which ended up generating what they call an "autocatalytic process of reform." This may explain why there have been so many coordination efforts in Brazil in the last decade. While acknowledging their importance, exploring the mechanisms that may effectively promote productive coordination in a context of institutional multiplicity is beyond the scope of this paper.

Nevertheless, it is worth mentioning that the Brazilian anti-corruption system has created a mechanism to enhance institutional coordination, the National Strategy to Combat Corruption and Money Laundering (Estrategia Nacional de Combate a Corrupcao e Lavagem de Dinheiro, ENCCLA). Created in 2003 via Presidential Decree and located at the Ministry of Justice, ENCCLA was initially an attempt to coordinate the efforts of three branches of government, the public prosecutors office and civil society to fight money laundering. After making significant progress in this area, its mandate was expanded in 2006 to include corruption, changing the name from ENCLA to ENCCLA. ENCCLA promotes the exchange of information, fosters strategic cooperation among different institutions, and pushes for legislative and institutional reforms that can enhance collaboration (Authors Interviews, April and May 2014).

While initiatives such as ENCCLA hold significant promise, there are no guarantees that they will foster effective collaboration. So far there has been a great deal of goodwill among the various institutions participating in ENCCLA, but the level of commitment and involvement varies significantly from institution to institution, depending on leadership (Authors' Interview, May 2014). In addition, while ENCCLA has been able to promote a great deal of cooperation at the strategic level, it has encountered barriers to do the same at the operational level (OECD, 2011). In sum, how to effective foster inter-institutional collaboration in an environment of institutional multiplicity is a topic that certainly deserves more attention in the literature.

If it is not possible to adopt robust efforts to improve coordination, there are reasons to support the idea that inter-institutional communication should at least avoid destructive or uncooperative behavior. For example, while the relationship between the MPF and DPF and other investigative bodies can be collaborative and complementary, as described earlier, it can also be uncooperative. Indeed, there are cases in which fierce and unproductive competition between the two institutions has undermined overall investigative efforts. An example of constant tension is the relationship between DPF and MPF (Azevedo and Vasconcellos, 2011, Melo 2013, p.29). This shows that institutional multiplicity may also create coordination problems, and it may increase the risk of some actors undermining the functions performed by others. Thus, if malleable coordination such as the one promoted by ECCO and ENCCLA is not possible, trying to avoid uncooperative and destructive behavior may at least increase the chances of institutional multiplicity generating positive outcomes.

It is important to note that a great deal of the advances achieved in Brazil in general and by ENCCLA in particular, are due to the fact that specialized departments to combat money laundering and fight corruption have been created within the DFP, MPF, the Ministry of Justice and other institutions. This begs the question about the advantages of specialization. Perhaps the clearest example of its advantages comes from the creation of special "judicial bancs" (varas especializadas) to evaluate money-laundering cases. The proposal was initially formulated by ENCCLA (Recommendation 02/2001) and implemented in 2003 as an option for Federal Tribunals (CJF Resolutions 314/03, and 517/06). After showing significant results, the existence of these specialized "judicial bancs" became mandatory for federal appeals tribunals in 2013 (CJF Resolution 273/13). Specialized judges seem to be able to process the cases much faster than generalist judges, who are not often familiar with the complicated money laundering legislation. Thus, this initiative is often reported as one an example of success (Authors Interviews, April and May 2014). However, it is hard to determine to what extent the specialization in the judiciary alone is the sole factor that can possibly explained expedited decisions. Confounding variables include the specialization in the DPF and MPF (which may have generated claims based on more robust evidence) and a legislative change in 1998 that has defined money laundering as a stand-alone crime (Statute n[degrees] 9.613/98).

Based on the perception that the money laundering initiative was successful, ENCCLA is currently pushing for the creation of specialized "judicial bancs" for corruption cases (Authors Interviews, April and May 2014). While this may be a promising development, it is important to note that such initiatives may be more likely to generate positive outcomes if they are structured as an option to the existing system, not a replacement of it. An example is the solution adopted in Indonesia, where a separate judiciary was created to analyze corruption cases (Butt, 2009). The prosecutors actually have a choice between bringing the case to the regular courts or the special anti-corruption court (Schutte 2011, 2012). Despite the fact that the Indonesian Supreme Court has recently declared the anti-corruption judiciary unconstitutional (Butt, 2011), this may be an interesting experiment to be explored in the Brazilian context. More specifically, it may suggest that there are advantages in specialization, but these are likely to be particularly promising if they do not exclude attempts to promote institutional multiplicity (i.e. there is a choice to bring the case to a specialized court or not). In other words, there needs to be malleability in specialization as well.

As stated earlier in the paper, institutional multiplicity may generate alternative avenues to perform each one of the accountability functions (oversight, investigation and punishment), but it may also create problems. This section has indicated that both coordination and specialization may be undermined by institutional multiplicity, but this does not need to be necessarily the case. Indeed, Brazil provides a few examples of initiatives in which malleability has been used, and institutional multiplicity, coordination and specialization were successfully reconciled to produce promising results.

5. Implications for Other Countries

The track record of anti-corruption reforms in developing countries has not been extremely successful, raising questions about the reasons for these failures. While some point to obstacles to implementation that undermine the effectiveness of reforms in the long term, others argue that the core flaws lie in the design of reforms (Personn et al., 2012). One example of the latter is the argument that anti-corruption reforms have largely failed in Africa because they are based on a principal-agent model (Rose-Ackerman 1978, Klitgaard 1988, Groenendijk, 1997), when systemic corruption may be more accurately described as a collective action problem. It does not matter how careful and diligent the implementation, if the model itself is flawed or inapplicable to the circumstances (Personn et al., 2012).

Institutional multiplicity may be an interesting arrangement for reformers confronted with this type of academic debate, as it eliminates the need to understand why the existing institutions are not working (problems with implementation or design). Instead, multiplicity allows for different designs to be operating simultaneously, providing reformers with information about what may be working or not.

One area in which this may be particularly useful is the design of accountability institutions in Africa. There is much dispute as to whether institutions to fight corruption in Africa should be designed according to the principal-agent or the collective action model. As Bardhan shows, corruption may be framed as a collective action problem that can lead to either a high corruption or a low corruption equilibrium, depending on the starting point (Bardhan, 1997). Under this frequency-dependent equilibrium model of corruption, individuals' decisions to engage in corruption are influenced significantly by the number of other people in their environment that they expect to be corrupt; as more decision makers follow the example of those in their community, that majoritarian position gathers increasing fortitude and becomes more entrenched, thus pushing societies to relatively stable levels of corruption--whether low or high.

From a policy perspective, this model explain why relatively small and discrete interventions, such as the ones suggested by the principal-agent model, can tip the balance of an economy that is lying in the middle towards a low corruption equilibrium. In contrast, it also explains why the same small and discrete interventions are ineffective in countries with high corruption equilibrium, such as many African nations (Personn, et al., 2012). The collective action model shows that as the number of people involved in corruption increase, the costs of engaging in corruption decrease as the chances of an act of corruption being identified by a corrupt agent and not punished increases. The fact that there is more corrupt people in the system may decrease the individual benefit (bribes will go down as there is competition for them), but Bardhan shows that the net benefit of engaging in corruption is still positive in these cases (1997). In sum, this high-corruption equilibrium would make any institutional reforms designed to empower principals largely ineffective.

The focus on institutional multiplicity could be potentially translated into policies that address the respective concerns of these conflicting explanations for reform failures. For those who believe that accountability institutions should be designed according to the principal-agent model, institutional multiplicity allows for alternative implementation channels that may help overcome the obstacles associated with one particular institution by offering an alternative on which reformers can focus their efforts. Alternatively, if corruption is a collective action problem, institutional multiplicity increases the uncertainty associated with corrupt actions as no single entity retains a monopoly over each of the accountability functions necessary to enforce anti-corruption rules.

Conclusion

This paper hypothesizes that institutional multiplicity may be an important feature in explaining the successes and failures of Brazilian institutions performing each of the three accountability functions. More specifically, we argue that the country has had significant progress in oversight and investigation while making notably less progress in the punishment of corrupt actors and behaviors.

Our analysis concedes that institutional underperformance is more easily noticed at the punishment stage, after illicit activities have already been detected and formally charged. Moreover, there are intrinsic difficulties in determining the performance and efficacy of oversight mechanisms, as we cannot know the full scope of the corruption they fail to detect. The metrics improve at the investigatory stage, allowing for assessment of the proportion of problems identified through oversight or some other means (e.g., whistleblowers) that are then scrutinized by the police or prosecutors, as well as opening the potential for non-state actors to get involved, especially the press. However, we still face the difficulty of defining the counterfactual i.e. how much oversight or investigation would have been conducted without institutional multiplicity. This is an important limitation of the paper, but we believe that the significant increases in oversight and investigation at the traditional institutions (TCU and MPF) after the creation of a new oversight institution (CGU) or the strengthening of an investigative institution (DPF) suggests that institutional multiplicity may be producing positive effects in the Brazilian system. It is not possible to discard, however, an alternative hypothesis: the role of non-state actors including the press, the public, and civil society in either performing some of these functions or exerting pressure on institutions to do their jobs effectively. The two hypotheses are not mutually exclusive, but democratization may well be the most relevant factor in many of the changes that are currently taking place in Brazil.

If accurate, the hypothesis of institutional multiplicity developed in this paper may have important implications that go beyond the Brazilian case. From a policy perspective, under certain circumstances investing in multiple institutions may be a more promising avenue than creating a single behemoth of an anti-corruption agency. Indeed, a country may be better off having a multitude of different institutions with overlapping oversight, investigation and punishment functions to enhance the efforts to combat corruption. However, uncovering the specific circumstances under which institutional multiplicity may be particularly conducive to positive outcomes will require comparative analyses of the cases of successful and unsuccessful anticorruption efforts in low- and middle-income countries.

From a descriptive perspective, institutional multiplicity offers a hypothesis that can be used to test the success or failure of certain reform efforts. For instance, institutional multiplicity may help to explain the success of Hong Kong, where the anti-corruption agency did not replace or displace an existing institution--the police--but instead performed the same function. The hypothesis may also offer an explanation for unsuccessful cases. Indeed, as we have suggested here, while Brazil's federal system shows much institutional multiplicity in the areas in which the country has achieved some degree of success in battling corruption--oversight and investigation--the stage at which accountability institutions appear most dysfunctional and in need of reform--enforcement--is charged to a single entity--the judiciary. If this institutional overlap hypothesis is correct, a potential policy implication would be to create institutional multiplicity at the punishment level, perhaps through the establishment of a special anti-corruption judiciary. The only country that seems to have experimented with the idea is Indonesia, and while the experience had been touted as a success, the country's Supreme Court struck it down as unconstitutional. Still, this may be a fruitful option for countries contemplating reforms, and it may be a particularly interesting idea in the Brazilian context.

While we argue that the institutional multiplicity hypothesis may provide insights that could be extended to contexts beyond Brazil, we acknowledge the strong aversion in much of the development literature against the idea of legal transplants or blueprints. Institutions are unlikely to operate in the same way when transported to a different political, social, economic or institutional environment (Trebilcock and Prado, 2011). However, this recognition of the limits of the institutional agenda does not mean abandoning it altogether, but instead redefining the meaning of "good governance" to account for the particular conditions and capabilities of the countries in which it is being promoted. Discussing the institutions of property rights and the rule of law in the African context, for instance, Khan (2012, pp. 6-7) articulates the idea very clearly: "a 'good governance' strategy of protecting and enforcing all existing property rights and a rule of law is unlikely to make much progress in the typical developing country. But a strategy to identify the most important property right transformations that are required for the development of most likely sectors, supported by strategies of learning by doing and the development of competitiveness in those sectors can pay dividends. (...) the strategies would have to be compatible with existing technocratic and political enforcement capabilities."

The institutional multiplicity hypothesis does not purport to offer a roadmap or standard reform package to be imported whole-cloth in other countries. Rather, it identifies a process through which countries may discover effective ways of fighting corruption, without prescribing specific institutional arrangements to be implemented. Multiplicity is a meta-concept (Davis and Prado, 2014), and its translation from theory to concrete policy recommendations for a given environment will require deep, context-aware examination and understanding of the existing institutional landscape. Thus, the institutional multiplicity hypothesis provides a new lens through which anti-corruption initiatives may be considered and evaluated in and across countries, not only in middle-income countries such as Brazil but across the developing world.

A deeper criticism of the analysis presented in this paper is the idea that not all kinds of corruption are detrimental to economic growth. This assumption has been particularly relevant to discuss the growth of Asian countries, many of which have achieved significant levels of economic development despite high rates of corruption (Khan 1998). This criticism may suggest that perhaps the fight against corruption should be focused on particular aspects of the phenomenon. We do not reject this possibility, despite the fact that we have not engaged with it in this paper. What we have offered here is an analysis of institutional strategies that may be effective in helping curtail corruption once a country has determined that it wants to engage into this fight. If there is no political will to do so (either because of the lack of empirical evidence that fighting particular types of corruption is conducive to growth, or because of self-interested resistance to maintain existing privileges), the analysis presented becomes largely irrelevant. However, in those countries in which such political and popular will does exist, we argue that the experience of Brazil may be able to offer lessons on how institutional structures and functions impact the efficacy of anti-corruption efforts.

NOTES

(1.) In an interview, a high level official at TCU argued that the percentage of cases in which there is this level of disagreement is low. The official emphasized that the decisions of the political cadre need to be justified when they are not in line with the technical report. This creates an accountability system, in which those trying to politically manipulate the process can be identified. The issue is whether there is any sanction associated with this. Except for possible reputational sanctions, there do not seem to be any consequences (especially if we consider that most of those occupying positions in the political body are politicians at the end of their careers). An official of an accounting tribunal at the state level (TCE) challenged this statement by suggesting that while the percentage of cases in which there is disagreement between the technical and the political body may be small, these may actually be the most significant cases. Only a comprehensive assessment of all cases would be able to provide insight as to who is right.

(2.) Brazil has witnessed a dramatic increase in investigative journalism since the return to democracy (Lima, 2006), and the press has made valuable contributions to the discovery and investigation of incidents of political corruption. However, political and ideological interests continue to exercise influence over the country's media (Abramo, 2007; Porto, 2011, p. 107), and journalists face harassment, censorship, and violence (Freedom House, 2013). According to the Committee to Protect Journalists, three Brazilian journalists--"all of them provincial journalists murdered after reporting on local crime and corruption"--were killed for their work in 2013, compared to four in 2012 and three in 2011 (Beiser, 2013).

(3.) For a detailed description and analysis of DPF operations, see Arantes 2011.

REFERENCES

Abbink, K. (2004), "Staff Rotation as an Anti-corruption Policy: An Experimental Study," European Journal of Political Economy 20: 887-906.

Abbink, K., and Serra, D. (2012), "Anticorruption Policies: Lessons from the Lab," in D. Serra and L. Wantchekon (eds.), New Advances in Experimental Research on Corruption (Research in Experimental Economics, Volume 15). Bingley: Emerald Group, 77-115.

Abramo, C. W. (2007), "Brazil: A Portrait of Disparities," Brazilian Journalism Research 3(1): 93-107.

Acemoglu, D., and Verdier, T. (2000), "The Choice between Market Failures and Corruption," American Economic Review 76(3): 495-520.

Ades, A., and Di Tella, R. (1999), "Rents, Competition, and Corruption," American Economic Review 89(4): 982-993.

Aidt, T. S. (2009), "Corruption, Institutions, and Economic Development," Oxford Review of Economic Policy 25(2): 271-291.

Alencar, C. H. R. (2010), "Prevencao e combate a corrupcao e eficacia judicial no Brasil. 5[degrees] Concurso de Monografias da Controladoria-Geral da Uniao--Controladoria-Geral da Uniao, Brasilia, DF. Available at http://www.cgu.gov.br/ concursos/Arquivos/5_ConcursoMonografias/2-lugar-profissionais-carloshigino-ribeiro-de-alencar. .pdf

Alencar, C. H. R., and Gico, I. T., Jr. (2011), "When Crime Pays: Measuring Judicial Performance against Corruption in Brazil," Law and Business Review of the Americas 17(3): 415-434.

Alston, L., Mello, M., Mueller, B., and Pereira, C. (2005), "Who Decides on Public Expenditures? A Political Economy Analysis of the Budget Process: The Case of Brazil" (Washington, DC: IDB, unpublished mimeo).

Amarribo Brasil (2012), UN Convention against Corruption Civil Society Review: Brazil 2012. [online]. Available at http://www.transparency.org/files/content/ publication/206-brazil-full-report.pdf [accessed Sep. 22, 2013].

Anderson, C. J., and Tverdora, Y. V. (2003), "Corruption, Political Allegiances, and Attitudes towards Government in Contemporary Democracies," American Journal of Political Science 47(1): 91-109.

Arantes, R. (2011), "The Federal Police and the Ministerio Publico," in T. J. Power and M. M. Taylor (eds.), Corruption and Democracy in Brazil: The Struggle for Accountability. Notre Dame, IN: University of Notre Dame Press, Ch. 8.

Artigo 19 (2013), Balanco de 1 ano de Lei de Acesso a Informacao Publica, available at http://artigo19.org/wp-content/uploads/2013/05/balancoLai.pdf.

Avritzer, L. (2008), "Indices de percepcao da corrupcao," in L. Avritzer, N. Bignotto, J. Guimaraes, and H. M. M. Starling (eds.), Corrupcao: ensaios e criticas. Belo Horizonte: Editora UFMG, 505-510.

Avritzer, L. (2011), "O real combate a corrupcao," Carta Capital, September 4. Available at http://www.cartacapital.com.br/sociedade/combate-a-corrupcaoe-programa-de-governo

Azevedo, R. G., and Vasconcellos, F. B. (2011), "O inquerito policial em questao: situacao atual e a percepcao dos delegados de policia sobre as fragilidades do modelo Brasileiro de investigacao criminal," Sociedade e Estado 26(1): 59-75.

Baldacci, E., Hillman, A. L., and Kojo, N. C. (2004), "Growth, Governance, and Fiscal Transmission Policies in Low-income Countries," European Journal of Political Economy 20: 517-549.

Banerjee, A., Chattopadhyay, R., Duflo, E., Keniston, D., and Singh, N. (2012), "Can Institutions Be Reformed from Within? Evidence from a Randomized Experiment with the Rajasthan Police," National Bureau of Economic Research Paper No. 17912. [online]. Available at http://economics.mit.edu/files/7581.

Bardhan, P. (1997), "Corruption and Development: A Review of Issues," Journal of Economic Literature 35(3): 1320-1346.

Beiser, E. (2013), "Syria, Iraq, Egypt Most Deadly Nations for Journalists," Committee to Protect Journalists. [online]. 30 December. Available at http://www.cpj. org/reports/2013/12/syria-iraq-egypt-most-deadly-nations-for-journalis.php.

Bertelsmann Stiftung (2014), BTI 2014--Brazil Country Report. Gutersloh: Bertelsmann Stiftung. [online]. Available at http://www.bti-project.de/uploads/tx_jp downloads/BTI_2014_Brazil.pdf.

Boas, T. C., Hidalgo, F. D., and Richardson, N. P. (2013), "The Spoils of Victory: Campaign Donations and Government Contracts in Brazil" [online]. Available at http://people.bu.edu/tboas/political_investment.pdf.

Bolongaita, E. P. (2010), "An Exception to the Rule? Why Indonesia's AntiCorruption Commission Succeeds where Others Don't--A Comparison with the Philippines' Ombudsman," U4 Issue 4 (Aug. 2010), available at http://issuu.com/ cmi-norway/docs/3765-an-exception-to-the-rule?e=1246952/3174147#search.

Brinks, D. (2005), "Judicial Reform and Independence in Brazil and Argentina: The Beginning of a New Millennium?," Texas International Law Journal 40: 595.

Brollo, F. et al. (2010), "The Political Resource Curse," National Bureau of Economic Research Working Paper 15705.

Brollo, F. (2009), "Who Is Punishing Corrupt Politicians--Voters or the Central Government? Evidence from the Brazilian Anti-corruption Program," [online] Available at http://www.bu.edu/econ/files/2012/11/dp168.pdf.

Bulla, B. (2010), http://ultimainstancia.uol.com.br/conteudo/noticias/49686/cnj+age +contra+corrupcao+no+judiciario+e+gera+reacao+corporativa+de+juizes.shtml

Butt, S. (2009), "Indonesia's Anti-Corruption Drive and the Constitutional Court," Comparative Law Journal 4(2): 186-204.

Butt, S. (2011), "Anti-Corruption Reform in Indonesia: An Obituary? Bulletin of Indonesian Economic Studies, 47(3): 381-439

Carson, L., and Prado, M. M. (2014), "Diagnosing the Disease: Mapping Corruption & Its Institutional Determinants in Brazil," IRIBA Working Paper.

CGU (2014), "Programa de Fiscalizacao a partir de Sorteios Publicos" Ramdom Audits Program, [online]. Available at http://www.cgu.gov.br/english/Area AuditoriaFiscalizacao/ExecucaoProgramasGoverno/Sorteios/index.asp (English), http://www.cgu.gov.br/controleinterno/AvaliacaoExecucaoProgramasGoverno/S orteios/index.asp (Portuguese).

CGU, MJ, AGU (2009), Exposicao de Motivos da Proposta de Regulamentacao da Responsabilidade Administrativa e Civil de Pessoas Juridicas por Atos Lesivos a Administracao Publica Nacional e Estrangeira, EMI n. 00011/2009. Available at http://www.camara.gov.br/proposicoesWeb/prop_mostrarintegra;jsessionid=822 76AC6F4965256C98004B4D4817737.proposicoesWeb1?codteor=735505&filen ame=Tramitacao-PL+6826/2010.

Clausen, B., Kraay, A., and Nyiri, Z. (2009), "Corruption and Confidence in Public Institutions: Results from a Global Survey," World Bank Policy Research Working Paper No. 5157. Washington, DC: World Bank.

CNJ (2012), Meta 18. Available at http://www.cnj.jus.br/gestao-e-planejamento/ metas/metas-2013/meta-18

CNJ (2014), Report on Meta 18 Outcomes. Available at http://www.cnj.jus.br/ relatorio-meta-18.

della Porta, D., and Vannucci, A. (1999), Corrupt Exchanges: Actors, Resources, and Mechanisms of Political Corruption. New York: Aldine de Gruyter.

Djankov, S., La Porta, R., Lopez-de-Silanes, F., and Shleifer, A. (2010), "Disclosure by Politicians," American Economic Journal: Applied Economics 2(2): 179-209.

Doig, A., Watt, D., and Williams, R. (2005), Measuring "Success" in Five African Anti-Corruption Commissions--the Cases of Ghana, Malawi, Tanzania, Uganda & Zambia. U4 Report. Bergen: U4 Anti-Corruption Centre.

Epoca (2008), "Por que o Brasil pode vencer a corrupcao," Special "Debate" supplement. [online]. 22 July. Available at http://revistaepoca.globo.com/Revista/ Epoca/1,,EMI5188-15273,00.html

Estadao (2009), "Supremo mantem afastamento de desembargador do Amazonas," [online] 13 July. Available at http://www.estadao.com.br/noticias/nacional, supremo-mantem-afastamento-de-desembargador-do-amazonas,402126,0.htm

Estadao (2014), "Numero de processos contra juizes no CNJ mais que dobra em 2013," [online] 2 Jan. Available at http://blogs.estadao.com.br/fausto-macedo/ numero-de-processos-contra-juizes-no-cnj-mais-que-dobra-em-2013/.

Excelencias Database (2013), Parlamentares com Ocorrencias na Justica ou Tribunal de Contas, Available at http://www.excelencias.org.br/@casa.php?tribs.

EY (2013), Growing Beyond: A Place for Integrity. 12th Global Fraud Report. [online]. Available at http://www.ey.com/Publication/vwLUAssets/Global-Fraud -Survey-a-place-for-integrity-12th-Global-Fraud-Survey/$FILE/EY-12thGLOBAL-FRAUD-SURVEY.pdf.

Federacao das Industrias do Estado de Sao Paulo (FIESP) (2010), "Relatorio Corrupcao: custos economicos e propostas de combate," [online]. Available at http://www.fiesp.com.br/indices-pesquisas-e-publicacoes/relatorio-corrupcaocustos-economicos-e-propostas-de- combate/.

Feld, L., and Voigt, S. (2003), "Economic Growth and Judicial Independence: Crosscountry Evidence Using a New Set of Indicators," European Journal of Political Economy 19: 497-527.

Ferraz, C., and F. Finan (2008), "Exposing Corrupt Politicians: The Effects of Brazil's Publicly Released Audits on Electoral Outcomes," Quarterly Journal of Economics 123(2): 703-745.

Ferraz, C., and Finan, F. (2011), "Electoral Accountability and Corruption: Evidence from the Audits of Local Governments," American Economic Review 101(4): 1274-1311.

Figueiredo, A. C. (2003), "Routine Oversight Mechanisms under the 1988 Constitution: Legal and Organizational Empowerment and Actual Capacity," in S. Mainwaring and C. Welna (eds.), Democratic Accountability in Latin America. Oxford: Oxford University Press, 183-196.

Filgueiras, F. (2008), "Marcos Teoricos da Corrupcao," in L. Avritzer, N. Bignotto, J. Guimaraes, and H.M.M. Starling (eds.), Corrupcao: ensaios e criticas. Belo Horizonte: Editora UFMG, 299-306.

Filgueiras, F. (2011), "Transparencia e controle da corrupcao no Brasil," in L. Avritzer and F. Filgueiras (eds.), Corrupcao e sistema politico no Brasil. Rio de Janeiro: Civilizacao Brasileira, 133-162.

Filgueiras, F., and A. L. M. Aranha (2011), "Controle da Corrupcao e Burocracia da Linha de Frente: Regras, Discricionariedade e Reformas no Brasil," DADOS Revista de Ciencias Sociais 54(2): 349-387.

Fleischer, D. (2000), "Alem do Collorgate: perspectivas de consolidacao democratica no Brasil via reformas politicas," in K. Rosenn and R. Downes (eds.), Corrupcao e reforma politica no Brasil: p impacto do impeachment de Collor. Rio de Janeiro, Editora FGV, 81-110.

Folha de Sao Paulo (2006), "Quadrilha vendia ambulancias com defeitos a prefeitos," 23/05.

Folha de Sao Paulo (2010). "Supremo planeja mudar tramitacao de recursos," 29/12. Available at http://www1.folha.uol.com.br/fsp/poder/po2912201014.htm.

Freedom House (2013), "Freedom of the Press: Brazil," [online]. Available at http:// www. freedomhouse.org/report/freedom-press/2013/brazil#.Uu9Hd_k7um4.

Friedman, E., Johnson, S., Kaufmann, D., and Zoido-Lobaton, P. (2000), "Dodging the Grabbing Hand: The Determinants of Unofficial Activity in 69 Countries," Journal of Public Economics 76: 459-493.

Global Integrity (2009), Global Integrity scorecard: Brazil 2009. Washington, DC: Global Integrity. Available at http://report.globalintegrity.org/reportPDFS/2009/ Brazil.pdf [accessed Sep. 22, 2013].

Golden, M. A., and Picci, L. (2005), "Proposal for a New Measure of Corruption, Illustrated with Italian Data," Economics and Politics 17(1): 37-75.

Gray, C. W, and Kaufmann, D. (1998), "Corruption and Development," Finance and Development 35(1): 7-10.

Gupta, S., Davoodi, H., and Alonso-Terme, R. (2002), "Does Corruption Affect Income Inequality and Poverty?," Economics of Governance 3: 23-45.

Gupta, S., de Mello, L., and Sharan, R. (2001), "Corruption and the Provision of Health Care and Education Services," in A. K. Jain (ed.), The Political Economy of Corruption. London: Routledge, 111-141.

Habib, M., and Zurawicki, L. (2002), "Corruption and Foreign Direct Investment," Journal of International Business Studies 33(2): 291-307.

Hammergreen, L. (2007), "Fighting Judicial Corruption: A Comparative Perspective from Latin American," in Transparency International, Global Corruption Report --Corruption in Judicial Systems, 138-145.

Heller, T. C. (2003), "An Immodest Postscript," in E. G. Jensen and T. C. Heller (eds.), Beyond Common Knowledge: Empirical Approaches to the Rule of Law. Stanford, CA: Stanford University Press, 406-409.

Holdheim, R. (2014), "Trust and the BRICs: Going Nowhere Slowly?," Edeleman. [blog]. 23 January. Available at http://www.edelman.com/post/trust-and-thebrics-going-nowhere-slowly/.

Huntington, S. (1968), Political Order in Changing Societies. New Haven, CT: Yale University Press.

Huther, J., and Shah, A. (2000), "Anticorruption Policies and Programs: A Framework for Evaluation," World Bank, Policy Research Working Paper 2501. Washington, DC: World Bank.

Innovare, I. (2013), "Praticas: Estrategia Estadual de Combate a Corrupcao," available at http://www.premioinnovare.com.br/praticas/estrategia-estadual-de-combate -a-corrupcao-ecco/

Iquiapaza, R., and Amaral, H. (2007), "Reflections on the Impact of Corruption on Economic Development: A Literature Review in the Brazilian Economy," MPRA Paper No. 1818. [online]. Available at http://mpra.ub.uni-muenchen.de/1818/.

Johnson, S., Kaufmann, D., McMillan, J., and Woodruff, C. (2000), "Why Do Firms Hide? Bribes and Unofficial Activity after Communism," Journal of Public Economics 76(3): 495-520.

Johnson, J., Hechler, H., De Sousa, L., and Mathison, H. (2011), "How to Monitor and Evaluate Anti-corruption Agencies: Guidelines for Agencies, Donors, and Evaluators," U4 Issue Paper No. 8. Bergen: U4 Anti-Corruption Centre.

Johnson, J., and Mason, P. (2013), "The Proxy Challenge: Why Bespoke Proxy Indicators Can Help Solve the Anti-corruption Measurement Problem," U4 Brief. Bergen: U4 Anti-Corruption Centre. [online]. Available at http://issuu.com/cminorway/docs/130913082222- 8ebe8df9a50149f9b68993699e70e626/1?e=0.

Kaufmann, D., Kraay, A., and Mastruzzi, M. (2009), "Governance Matters VIII: Aggregate and Individual Governance Indicators 1996-2008," World Bank Policy Research Working Paper No. 4978.

Kaufmann, D., Kraay, A., and Mastruzzi, M. (2006), "Measuring Governance Using Cross-country Perceptions Data," in S. Rose-Ackerman (ed.), The International Handbook on the Economics of Corruption. Camberley: Edward Elgar, 52-104.

Kenny, C., Klein, M., and Monika Sztajerowska, M. (2011), "A Trio of Perspectives on Corruption: Bias, Speed Money, & "Grand Theft Infrastructure," World Bank, Policy Research Working Paper No. 5889.

Knack, S. (2007), "Measuring Corruption: A Critique of Indicators in Eastern Europe and Central Asia," Journal of Public Policy 27(3): 255-291.

Lamsdorff, J. G. (2007), The Institutional Economics of Corruption and Reform: Theory, Evidence, and Policy. Cambridge: Cambridge University Press.

Lambsdorff, J. G. (2000), "The Precision and Regional Comparison of Perceived Levels of Corruption--Interpreting the Results," Background Paper to the 2000 Corruption Perceptions Index, Transparency International & Gottingen University. [online]. Available at http://www.icgg.org/downloads/2000_CPI_Precision.pdf

La Porta, R., et al. (2004), Judicial Checks and Balances," Journal of Political Economy 112(2): 445-470.

Li, H., Xu, L. C., and Zou, H.-F. (2000), "Corruption, Income Distribution, and Growth," Economics and Politics 12(2): 155-182.

Lima, V. (2006), Midia: crise politica e poder no Brasil. Sao Paulo: Editora de Fundacao Perseu Abramo.

Loureiro, M. R., Abrucio, F. L., Olivieri, C., and Teixeira, M. A. (2012), "Do controle interno ao controle social: a multipla atuacao da CGU na Democracia Brasileira" in Cadernos Gestao Publica e Cidadania, Sao Paulo 17(60): 54-67. Available at http://bibliotecadigital.fgv.br/ojs/index.php/cgpc/article/viewFile/ 3980/2851.

Lui, F. (1985), "An Equilibrium Queuing Model of Bribery," Journal of Political Economy 93(4): 760-781.

Machado, M. R. (2013), "Contra a departamentalizacao do saber juridico: a contribuicao dos estudos de caso para o campo direito e desenvolvimento," in Direito e Desenvolvimento no Seculo XXI. Brasilia: IPEA/Conpedi.

Machado, M. R. and Ferreira, L. M. A. (2015). Estudos sobre o Caso TRT. Sao Paulo, SP: Editora Saraiva. Forthcoming

Mauro, P. (1998), "Corruption and the Composition of Public Expenditure," Journal of Public Economics 69: 263-279.

McMillan, J., and Zoido, P. (2004), "How to Subvert Democracy: Montesinos in Peru," Journal of Economic Perspectives 18(4): 69-92.

Meagher, P. (2004), "Anti-corruption Agencies: A Review of Experience," Center for Institutional Reform and the Informal Sector (IRIS Center), Discussion Papers on Institutions and Development No. 02. College Park, MD: University of Maryland.

Meagher, P. (2005), "Anti-corruption Agencies: Rhetoric versus Reality," Journal of Policy Reform 8(1): 69-103.

Mechanism for Follow-Up on the Implementation of the Inter-American Convention against Corruption (MESICIC) (2012). Federative Republic of Brazil Final Report.

Melo, M. A. (2013), "Brazil: Democracy and Corruption," in Legatum Institute, Democracy Consensus Project Seminar. Rio de Janeiro, 15 May.

Melo, M. A., C. Pereira, and C. M. Figueiredo (2009), "Political and Institutional Checks on Corruption: Explaining the Performance of Brazilian Audit Institutions," Comparative Political Studies 42: 1217-1244.

Mo, P. H. (2001), "Corruption and Economic Growth," Journal of Comparative Economics 29: 66-79.

Morris, S. D., and Klesner, J. L. (2010), "Corruption and Trust: Theoretical Considerations and Evidence from Mexico," Comparative Political Studies 43(10): 1258-1285.

Musacchio, A., and Lazzarini, S. G. (2015), Reinventing State Capitalism: Leviathan in Business, Brazil and beyond. "Introduction" available at http://people.hbs.edu/ amusacchio/downloads/reinventing_state_capitalism.pdf. Forthcoming

Noonan, J. T., Jr. (1984), Bribes: An Intellectual History of a Moral Idea. New York: Macmillan.

Nye, J. S. (1967), "Corruption and Political Development: A Cost-benefit Analysis," American Review of Political Science 61(2): 417-427.

O'Donnell, G. A. (1998), "Horizontal Accountability in New Democracies," Journal of Democracy 9(3): 112-126.

O Globo (2013), "PEC 37 e votada e rejeitada na Camara dos Deputados," 25/06. Available at http://oglobo.globo.com/pais/pec-37-votada-rejeitada-na-camarados-deputados-8806597.

Olivieri, C. (2006), "O controle interno como instrumento de accountability horizontal: uma analise exploratoria da Controladoria Geral da Uniao," paper presented at the Meeting of Public Administration and Governance, available at http://admsp 20061.wikispaces.com/file/view/O+Controle+Interno+como+Instrumento+de+A ccountability+Horizontal+-+Cec%C3%ADlia+Olivieri.pdf.

Open Budget Index Report (2012). Available at http://internationalbudget.org/wpcontent/uploads/OBI2012-Report- English.pdf

Organisation for Economic Co-operation and Development, Convention on Combating Bribery of Foreign Public Officials in International Business Transactions (1997).

Organisation for Economic Co-operation and Development (2011), Report on the Assessment of Brazil's Integrity System of the Federal Public Administration. Available at http://cgu.gov.br/Publicacoes/AvaliacaoIntegridadeBrasileiraOCDE /AvaliacaoIntegridadeBrasileiraOCDE-integra-ingles.PDF

Panth, S. (2011), "The Ficha Limpa (Clean Record) Campaign," the World Bank Blog, May 9. Available at http://blogs.worldbank.org/publicsphere/ficha-limpaclean-record-campaign.

Persson, A., Rothstein, B., and Teorell, J. (2013), "Why Anticorruption Reforms Fail--Systemic Corruption as a Collective Action Problem," Governance 26(3): 449-471.

Pope, J. (2000), Transparency International Source Book 2000: Confronting Corruption: The Elements of a National Integrity System. Berlin: Transparency International. [online]. Available at http://archive.transparency.org/publications/ sourcebook [accessed Sep. 22, 2013].

Porto, M. P. (2011), "The Media and Political Accountability," in T. J. Power and M. M. Taylor (eds.), Corruption and Democracy in Brazil: The Struggle for Accountability. Notre Dame, IN: University of Notre Dame Press. Ch. 5.

Power, T., and Cyr, J. M. (2009), "Mapping Political Legitimacy in Latin America: A Multidimensional Approach," International Social Science Journal 60(196): 253-272.

Power T. J., and Taylor, M. M. (eds.) (2011), Corruption and Democracy in Brazil: The Struggle for Accountability. Notre Dame, IN: University of Notre Dame Press.

Praca, S., and Taylor, M. M. (2014), "Inching Toward Accountability: The Evolution of Brazil's Anticorruption Institutions, 1985-2010," Latin American Politics and Society 56(2): 27-48.

Prillaman, W. (2000), The Judiciary and Democratic Decay in Latin America: Declining Confidence in the Rule of Law. Westport, CT: Greenwood Press.

Rauch, J. F., and Evans, P. B. (2000), "Bureaucratic Structure and Bureaucratic Performance in Less Developed Countries," Journal of Public Economics 75: 49-71.

Rios-Figueroa, J. (2006), Judicial Independence: Definition, Measurement, and Its Effects on Corruption. PhD Dissertation, New York University, Department of Politics.

Rocha, A. (2003), "Especializacao e autonomia funcional no ambito do Tribunal de Contas da Uniao," Revista de Informacao Legislativa 40(157): 223-251.

Rocha, M. (2012), "Como se Desvia Dinheiro no Brasil," Epoca [online] 22 July. Available at http://revistaepoca.globo.com/ideias/noticia/2012/01/como-sedesvia-dinheiro-no-brasil.html.

Rose-Ackerman, S. (1999), Corruption and Government: Causes, Consequences, and Reform. Cambridge: Cambridge University Press.

Rose-Ackerman, S., and Truex, R. (2012), "Corruption and Policy Reform," Working Paper, Copenhagen Consensus Project. [online]. Available at http://www.copenhagenconsensus.com/sites/default/files/Working%2Bpaper_Corruption.pdf.

Sadek, M. T. (2008), "Ministerio Publico," in L. Avritzer, N. Bignotto, J. Guimaraes, and H. M. M. Starling (eds.), Corrupcao: ensaios e criticas. Belo Horizonte: Editora UFMG, 454-460.

Sadek, M. T., and Cavalcanti, R. B. (2003), "The New Brazilian Public Prosecution: an Agent of Accountability," in S. Mainwaring and C. Welna (eds.), Democratic Accountability in Latin America. Oxford: Oxford University Press, 201-227.

Santiso, C. (2007), "Eyes Wide Shut? The Politics of Autonomous Audit Agencies in Emerging Economies," Centro de Implementation de Politicas Publicas para la Equidad y el Crecimiento (CIPPEC), Document de Trabajo. Available at http://www.cippec.org/documents/10179/51827/5+DT+Transparencia%2C%20E yes+wide+shut%2C%20Santiso%2C%202007.pdf/08aaea45-a6de-4317-9e0018d53a718020.

Sarsfield, R. (2012), "The Bribe Game: Microfoundations of Corruption in Mexico," Justice System Journal 33(2): 215-234.

Schutte, S. A. (2011), "Appointing Top Officials in a Democratic Indonesia: The Corruption Eradication Commission," Bulletin of Indonesian Economic Studies 47(3): 355-379.

Schutte, S. A. (2012), "Against the Odds: Anti-Corruption Reform in Indonesia," Public Administration & Development 32: 38-48.

Seligson, M. A. (2002), "The Impact of Corruption on Regime Legitimacy: A Comparative Study of Four Latin American Countries," Journal of Politics 64(2): 408-433.

Seligson, M. A., Smith, A. E., and Zechmeister, E. J. (eds.) (2012), "The Political Culture of Democracy in the Americas, 2012: Towards Equality of Opportunity," [online]. Available at http://www.vanderbilt.edu/lapop/ab2012/AB2012-comparative-Report-V7 -Final-Cover-01.25.13.pdf.

Shah, A. (2007), "Tailoring the Fight against Corruption to Country Circumstances," in A. Shah (ed.), Performance Accountability and Combating Corruption. Washington, DC: World Bank, 233-254.

Shleifer, A., and Vishny, R. W. (1993), "Corruption," Quarterly Journal of Economics 108: 599-617.

Silva, M. F. G., Garcia, F., and Bandeira, A. C. (2001), "How Does Corruption Hurt Growth? Evidences about the Effects of Corruption on Factors Productivity and Per Capita Income," [online]. Available at http://bibliotecadigital.fgv.br/dspace /bitstream/10438/1953/1/TD103.pdf.

Speck, B. W. (2000), "Mensurando a corrupcao: uma revisao de dados provenientes de pesquisas empiricas," in Cadernos Adenauer 10: Os custos da corrupcao. Sao Paulo: Fundcao Konrad Adenauer.

Speck, B. W. (2011), "Auditing Institutions," in T. J. Power and M. M. Taylor (eds.), Corruption and Democracy in Brazil: The Struggle for Accountability. Notre Dame, IN: University of Notre Dame Press. Ch. 6.

Stocker, F. T. (2012), Anti-corruption Developments in the BRIC Countries: A MAPI Series. Arlington, VA: Manufacturers Alliance for Productivity and Innovation. [online]. Available at http://www.mapi.net/system/files/PA-107_2.pdf [accessed Sep. 22, 2013].

Swaleheen, M. (2011), "Economic Growth with Endogenous Corruption: An Empirical Study," Public Choice 146: 23-41.

Tanzi, V. (1998), "Corruption around the World: Causes, Consequences, Scope, and Cures," IMF Staff Papers 45(4): 559-594.

Tanzi, V., and Davoodi, H. (1997), "Corruption, Public Investment, and Growth," Working Paper of the International Monetary Fund No. 139.

Taylor, M. M. (2009), "Corruption, Accountability Reforms and Democracy in Brazil," in C. H Blake and S. Morris (eds.), Corruption and Democracy in Latin America. Pittsburgh, PA: University of Pittsburgh Press, 150-168.

Taylor, M. M., and Vinicius C. Buranelli (2007), "Ending Up in Pizza: Accountability as a Problem of Institutional Arrangement in Brazil," Latin American Politics and Society 49(1): 59-87.

Terra (2013), "Gurgel: sem investigacao do MP, acao do mensalao nao teria existido," 11/03. Available at http://noticias.terra.com.br/brasil/politica/julgamento-domensalao/gurgel-sem-investigacao-do- mp-acao-do-mensalao-nao-teria-existido, d3b5f2532da5d310VgnVCM5000009ccceb0aRCRD.html

Transparency International (2013a), Corruption Perceptions Index.

Transparency International (2013b), Global Corruption Barometer.

Transparency International (2012), Putting Corruption Out of Business. Data available at http://www.transparency.org/research/bps2011.

Transparency International (2011a), Bribe Payers ' Index.

Transparency International (2011b), Global Corruption Barometer.

Transparency International (2007), Global Corruption Report--Corruption in Judicial Systems.

Treisman, D. (2000), "The Causes of Corruption: A Cross-national Study," Journal of Public Economics 76: 399-457.

Treisman, D. (2007), "What Have We Learned about the Causes of Corruption from Ten Years of Cross-national Empirical Research?," Annual Review of Political Science 10: 211-44.

Van Rijickeghem, C., and Weder, B. (2001), "Bureaucratic Corruption and the Rate of Temptation: Do Wages in the Civil Service Affect Corruption, and by How Much?," Journal of Development Economics 65: 307-331.

Veja (2000), A Anatomia do Crime, Edicao 1660, August 2. veja.abril.com.br/ 020800/p_038.html

Warren, M. E. (2004), "What Does Corruption Mean in a Democracy?," American Journal of Political Science 48(2): 328-343.

Wei, S.-J. (2000), "How Taxing Is Corruption on International Investors?," Review of Economics and Statistics 1(11): 1-11.

World Bank (2000), "Anti-corruption in Transition," [online]. Available at http:// siteresources.worldbank.org/INTWBIGOVANTCOR/Resources/contribution.pdf.

World Bank (2009), "Enterprise Survey: Country Report on Brazil," [online]. Available at http://www.enterprisesurveys.org/~/media/FPDKM/EnterpriseSurveys/ Documents/Profiles/English/brazil-2009.pdf.

World Bank (2013), "Development Indicators," [online]. Available at http://databank.worldbank.org/data/home.aspx.

World Economic Forum (2013), The Global Competitiveness Report 2013-2014: Full Data Edition. [online]. Available at http://www3.weforum.org/docs/WEF_ GlobalCompetitivenessReport_2013-14.pdf.

Zamboni Filho, Y., and S. Litschwig (2006), "Law Enforcement and Local Governance in Brazil: Evidence from Random Audit Reports," paper presented at the Northeast Universities Development Consortium Conference.

Zurawicki, L., and Habib, M. (2010), "Corruption and Foreign Direct Investment: What Have We Learned?," International Business and Economics Research Journal 9(7): 1-10.

Received 20 April 2015 * Received in revised form 23 June 2015

Accepted 24 June 2015 * Available online 15 September 2015

MARIANA MOTA PRADO

mariana.prado@utoronto.ca

Faculty of Law, University of Toronto

LINDSEY CARSON

Johns Hopkins School of Advanced International Studies

(SAIS), Washington, DC
COPYRIGHT 2016 Addleton Academic Publishers
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Prado, Mariana Mota; Carson, Lindsey
Publication:Journal of Self-Governance and Management Economics
Article Type:Report
Geographic Code:3BRAZ
Date:Jan 1, 2016
Words:15735
Previous Article:Strategies for reducing moral hazard in construction procurement: a conceptual framework.
Next Article:The post-2015 global development agenda: a critical analysis.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters