How do Bureaucrats Tackle Political Corruption in a Highly
Corrupt Setting?
The Case of Car Wash Operation in Brazil
Juan Pablo Briceno Student Number: s1879901
Master Thesis
MSc. Public Administration — Public Management and Leadership Faculty of Governance and Global Affairs
Leiden University The Hague, Netherlands
1st June 2020 Supervisor: Dr. K. Suzuki
Abstract
While existing research has found a robust relationship between meritocratic bureaucracies and lowered corruption, it has not come to terms with the mechanism underlying it. The present study fills that gap by asking: how do bureaucrats lower corruption? The objective of this study is to both test and develop hypotheses about the process in which bureaucrats tackle corruption. It does so by examining one of the largest anti-corruption operations ever launched in Brazil: the “Lava Jato” (Car Wash Operation). The research suggests that the relationship between meritocratic bureaucracies and lowered corruption is linked by the interaction between the broad institutional structure in which the bureaucrats are embedded and the voice of the individual bureaucrat. Particularly, for the operation to occur, a high degree of accountability between all relevant actors was necessary, as well as a high degree of legitimacy in the anti-corruption process, and a high degree of confidence in the bureaucrats advancing the operation. The research pinpoints a broad institutional framework (comprised not only of meritocratic recruitment, but also of a free press, popular expressions, bureaucrats keeping bureaucrats accountable, politicians keeping politicians accountable, transparent bureaucratic communication, and coercive judicial instruments) as key for these three conditions to appear, and therefore for the Car Wash operation to take place.
Keywords
Table of Contents
List of Acronyms 3
1. Introduction 4
2. Theoretical Framework 8
2.1 What Has Already Been Studied 8
2.2 What Remains to Be Studied 16
3. Data and Method 18
3.1 Process-tracing 18 3.2 Case Selection 18 3.3 Operationalization 20 3.4 Data 22 4. Findings 25 4.1 Research context 25
4.2 Examining the Mechanisms Proposed in the Literature 28 4.3 Examining Alternative Context-specific Mechanisms 36
5. Analysis 41
5.1 Mechanisms Proposed in the Literature 41
5.2 Mechanisms Not Accounted in the Literature But Present in the Case 44
6. Conclusion 47
6.1 Limitations and Future Directions 50
List of Acronyms
CCJ Justice and Citizenship Commission from the Brazilian Federal Senate (Portuguese:
Comissão de Constituição, Justiça e Cidadania)
CGU Brazilian General Comptroller’s Office (Portuguese: Controladoria-Geral da União) MBL Free Brazil Movement (Portuguese: Movimento Brasil Livre)
MDB Brazilian Democratic Movement (Portuguese: Movimento Democrático Brasileiro) MPF Brazilian Public Prosecutor’s Office (Portuguese: Ministério Público Federal) PF Brazilian Federal Police (Portuguese: Policia Federal)
PMDB Brazilian Democratic Movement Party PP Progressistas Party
PSDB Brazilian Social Democratic Party
PT Workers Party (Portuguese: Partido dos Trabalhadores) PTB Brazilian Labour Party
SD Solidariedade Party
1. Introduction
Ever since the late nineteenth century, scholars have maintained the view that bureaucratic recruitment based on merit—and not on the discretionary will of politicians—is a necessary component for avoiding the perils of patronage systems (Wilson, 1887; Goodnow, 1967; Weber, 1978, cited in Cooper 2018, 723) . This school has maintained that the bureaucratic insulation from politics is critical for creating an environment free from corruption. Such a view continues in modern scholarship, as a strong statistical relationship between meritocratic recruitment and bureaucrat’s proclivity to stand up against political corruption has been identified (Dahlström and Lapuente 2017, 8; Rauch and Evans 200, 5; Meyer-Sahling et al. 2016, 1105). Accordingly, the academy has suggested the existence of the relationship between meritocratic recruitment and lowered corruption.
Yet, the mechanism explaining this relationship remains to be settled (Meyer-Sahling et al. 2018, 279). The studies examining the relationship between the two variables have relied on different units of analysis, and as such, different mechanisms have been proposed explaining the relationship. Altogether, the proposed mechanisms have taken two overarching points of view. The first and most prevalent one embraces an institutional perspective. This view explains the negative relationship between meritocratic recruitment and corruption by highlighting the separation of careers between bureaucrats and politicians, and the accountability each one of them demands on the other (Dahlström and Lapuente, 2017; Neshkova and Kostadinova, 2012; Oliveros and Schuster, 2018; Meyer-Sahling and Mikkelsen, 2016; Boräng et al. 2018; Bersch et al., 2017b). This school argues that meritocratically recruited bureaucrats are likely to resist the pressure of collusion coming from politicians, given that their professional positions do not depend on the discretionary will of politicians but rather on their own capabilities. This independence is marked by the extent to which the interests of meritocratically recruited bureaucrats differ from those of politicians. The logic here is that with the separation of interests, both civil servants and politicians have the reason to hold each other on check, and can make the other accountable for corrupt demeanor. On the other hand, the second overarching view explains the relationship between meritocratic bureaucracies and corruption by focusing on bureaucratic behavior (Rauch and Evans, 2000; Cooper, 2018). The general idea underlying this mechanism is that bureaucratic structures following strict guidelines of meritocratic
recruitment tend to shape the behavior of civil servants towards a more neutral and objective one. Either for forming an esprit de corps among bureaucrats (Rauch and Evans, 2000), or for shaping the voice of the bureaucrat (Cooper, 2018), the psychological effect of the meritocratic structure on the bureaucrat herself is one that induces the bureaucrat to voice corrupt behavior in government.
Thus, while we know that meritocratic bureaucracies are statistically correlated to lower corruption, we do not know exactly in which way they are. Surely, context might play an important role in determining the mechanisms at play. Yet as of now, no scholarship has unraveled empirically which of the different theoretical mechanisms explaining this relation applies to which contextual factors (Meyer-Sahling et al. 2018, 279).
This study contributes to the theoretical discussion by filling such a gap. The research attempts to answer the research question of how do bureaucrats lower corruption, by delving into the Brazilian context, specifically into the “Lava Jato” (Car Wash Operation) case. The objective of examining this case is to test and develop hypotheses about the process in which anti-corruption efforts are advanced by bureaucrats in highly corrupt settings. To achieve this objective, the research applies a process-tracing method that both examines the existing institutional and behavioral theoretical propositions, and reconstructs a narrative that finds a sufficient explanation for the lowered corruption in Brazil. First, the research hypothesizes that the relationship between meritocratic bureaucracies and lowered corruption is linked by both institutional and behavioral mechanisms operating simultaneously. Then, and given the low likelihood that the mechanisms proposed in the theory provide a sufficient explanation for the specific case of Brazil, the research proposes—by means of an inductive method—a mechanism able to sufficiently account for the case outcome.
The Car Wash operation is the largest anti-corruption investigation in Brazilian history. It began in 2014, and it has since revealed and recovered billions of dollars in illegal payments. The Brazilian Federal Police (PF) and the Public Prosecutor’s Office (MPF) initiated the investigation as part of their routine activities, without the influence of any political interest. Yet, this routine investigation later became a large-scale national anti-corruption investigation. The Car Wash operation was responsible, among others, for jailing Brazilian former President Luiz Inácio Lula da Silva (Lula), and bringing a sitting president, Dilma Roussef, into court based on corruption charges.
This operation was selected as a case study because it provides insights into the necessary conditions for the establishment of effective anti-corruption operations in contexts of high political corruption. The fact that the effective operation was advanced without political involvement, that is, purely from a bureaucratic point of view, and that it took place in a highly corrupt setting, makes this a rich and puzzling case where several insights can be drawn for lowering corruption. Besides, considering that the institution of meritocratic recruitment has been the method for selecting bureaucrats in the Brazilian State apparatus since the midst of the 20th century (Castelar et al. 2010, 81), this investigation sheds light on the nature of the relationship between meritocratically recruited bureaucrats and corrupt politicians in contexts where anti-corruption measures are advanced. The case shows the obstacles and maneuvers Brazilian bureaucrats faced and devised to make the anti-corruption endeavor possible.
The Car Wash operation case has been well documented. Primary data such as government reports, legal archival collections, and speeches from the key actors are available via government websites and international online databases. Moreover, the Brazilian General Comptroller’s Office (CGU) presents administrative data on more than 326,000 individual bureaucrats. This data comprises 95 of the most important federal agencies in Brazil, including the agencies relevant to the present research. Secondary data such as media articles; Dahlström et al. (2015) QoG expert survey data; and Bersch et al. (2017b) categorization of Brazilian agencies’ autonomy, are also consulted. Through the triangulation of data from different sources, the study attempts to control for the bias and selectivity intrinsic of qualitative studies (Beach and Pedersen 2013, 128).
The research is structured as follows. The first section reviews the existing literature and establishes the central role of bureaucrats in the prevention of corruption. This section also develops the theoretical argument indicating the mechanisms through which meritocratic recruitment lowers corruption. Then, the research specifies its design, data collection, and methods. Here, a broader justification for the selection of the Car Wash operation case is presented, as well as a specification of the limitations of the case, the operationalization of the empirical evidence, and the location of the data utilized. Afterward, a brief background to the case is shown. This section highlights the workings of the Brazilian public recruitment system, and describes the key events and actors revolving around the investigation. Next, the research investigates the theoretical claims empirically and collects facts about the case for
the construction of an empirical narrative. The findings are later analyzed, and a conclusion is given based on the analysis. This (single) case study lends credence to the theorized mechanisms that link meritocratic recruitment with lower corruption, and expands the theory by suggesting a broader institutional accountability framework as a necessary condition for explaining the relationship.
2. Theoretical Framework
2.1 What Has Already Been Studied
Corruption. Keeping corruption at bay is a key function of any government pursuing quality governance outcomes. Several studies indicate that corruption presents itself as a significant obstacle to economic growth, societal development, and state capacity. Accordingly, a 1 growing body of research investigates the causes of corruption and the ways to curtail its forces. 2
The literature pinpoints several determinants of corruption. The typology advanced by Brunetti and Weder classifies the causes of corruption in three different categories (Brunetti and Weder 2003, 1803). These are internal controls, external controls, and indirect determinants. Internal controls refer to those characteristics intrinsic in the bureaucratic structure that influence corruption, such as the methods for recruiting and promoting personnel, and or the salary of civil servants (Rauch and Evans 2000; Van Rijckeghem and Weder, 2001). Differently, external controls indicate variables outside of the boundaries of the bureaucratic structure that result in affecting corruption, such as the age of the regime, the degree of press freedom, and the share of legislators or government ministers who are female (Sung, 2004; Brunetti and Weder, 2003; Treisman 2007). The last category, indirect determinants, involves cultural factors and degrees of distortions in the economy that have repercussions on national levels corruption, such as the degree of ethnic diversity, and the degree of economic development, which seem to have an impact on corruption levels (Ades and Di Tella, 1999; Alesina et al. 1999).
1 For examples highlighting the negative effects of corruption on economic growth, see (Klitgaard, 1988; Mauro, 1995; Knack and Keefer, 1995; Brunetti et al., 1998; Mauro, 1996; Johnson et al., 1998; Rose-Ackerman, 1999; Tanzi, 1998; Jain, 2001; Aidt, 2003; Lambsdorff, 2006; Seldadyo and de Haan, 2006; Treisman 2007). For examples highlighting the negative effects of corruption on societal development, see (Kaufmann, 2004; Dimant et al., 2013; Azfar and Gurgur, 2008; Gupta et al., 2002). For examples highlighting the negative effects of corruption on state capacity, see (Kaufman and Wei, 1999; Rothstein, 2008; and Anderson and Tverdova, 2003).
2 For some examples, see (Tanzi, 1998; Bhattacharyya and Hodler, 2015; Saha et al., 2009; Dincer, 2008; Rivas, 2012; Badinger and Nindl, 2014; Goel and Nelson, 2010; Dell’Anno and Teobaldelli, 2015; Pellegrini and Gerlagh, 2008; Azfar and Nelson, 2007; Lindstedt and Naurin, 2010; Gokcekus and Knorich, 2006; Dong and Tongler, 2011; Justesen and Bjornskov, 2014; Campante, 2009; Angeles and Neanidis, 2015; Pellegrini and Gerlagh, 2008).
Table 1. Determinants of corruption. The table is updated based on Brunetti and Weder’s conceptualization (2003). 3
Internal Controls External Controls Indirect Determinants
Salary of the bureaucracy (Van Rijckeghem and
Weder, 2001)
Age of the regime (Sung, 2004)
Degree of economic development (Ades and Di Tella, 1999) Recruitment/promotion of
the bureaucracy (Dahlström and Lapuente, 2017)
Degree of press freedom
(Brunetti and Weder, 2003) Degree of ethnic diversity (Alesina et al. 1999) Tenure of the bureaucracy
(Oliveros and Schuster, 2018)
The psychological effect of the bureaucratic structure on the individual bureaucrat
(Rauch and Evans, 2000; Cooper, 2018)
Share of legislators or government ministers who are female (Treisman 2007)
Degree of government interventions in free markets
(Tanzi, 1994)
This research focuses, however, on the internal controls of corruption for two reasons. First, most of the corruption literature concentrates on the external and indirect variables affecting corruption, and little scholarly attention is left to the internal controls (Meyer-Sahling et al. 2018, 277). Second, focusing on the management of civil service implies focusing on the range of actionable policies that the state can maneuver. Arguably, governments might find more difficulties managing external and indirect controls affecting corruption, such as regime age and ethnic diversity, than managing internal controls concerning the structure of the bureaucracy. Not considering how the management of bureaucrats molds corruption might imply losing a decisive chance to tackle corrupt demeanor.
Bureaucracies and Corruption. The literature that considers the relationship between bureaucrats and corruption has mainly focused on two areas of civil service management: pay levels, and selection and recruitment of the bureaucracy (Meyer-Sahling et al. 2018, 278).
With regards to pay levels, the literature shows contradictory findings for its relationship with corruption. As for meritocratic recruitment, the literature presents a large consensus suggesting its negative relationship with corruption.
There is also a smaller literature on the effects of different features of the bureaucratic structure—such as promotions, employment protection, and job stability—on corruption. Similarly to the study of bureaucratic remuneration, this literature counts with mixed results. This scholarship has hypothesized that bureaucrats with tenure protection and job stability are free from being dismissed by politicians, who might otherwise pressure them to engage in collusion. The idea here is that bureaucrats with tenure protection and job stability are discouraged from sacrificing their long-term career prospects for short-term corruption gains (Herron et al., 2017; Rubin and Whitford 2008; Neshkova and Kostadinova 2012; Rauch and Evans 2000; Dahlström et al. 2012; Bersch et al. 2017b). However, none of these studies have found any significant statistical relationship supporting this logical supposition. In all, what scholars have found is that the determinant feature of the bureaucratic structure affecting corruption is the actual benchmark used to promote, protect, and recruit bureaucrats. As put by Meyer-Sahling et al. (2018), no evidence has been found proving that bureaucratic promotion reduces corruption in itself; rather, what appears to affect corruption is the criteria upon which promotion decisions are established—either by merit or by political discretion (280).
Concerning salary levels, scholars seem to disagree on how bureaucratic pay levels affect bureaucrats’ take on corruption. The temptation argument states that bureaucrats with lower pay levels are more tempted to engage in corrupt behavior when compared with those having higher remuneration (Becker and Stigler 1974; Besley and McLaren 1993; Van Rijckeghem and Weder 2001). However, more recent cross-country studies have placed doubts against the temptation argument, as they have not found any robust association supporting such perspective (Dahlström, Lapuente, and Teorell, 2012; Rauch and Evans, 2000; Treisman, 2000). Scholars have argued that, rather than the actual salary level, lowered levels of corruption are more likely to depend on how the pay system at large is structured (Meyer-Sahling et al. 2018, 282). Similar to the case of promotions systems, Meyer-Sahling et al. (2018) have found that asking how salary levels are determined—e.g. either through political/personal relations, qualifications or performance—provides a more appropriate approach for understanding the relationship between bureaucratic pay-levels and corruption
(Meyer-Sahling et al. 2018, 282). It is indeed this logic what has stirred the research focus to the criteria determining the selection, promotion, and salary levels of bureaucrats.
Unlike the literature that studies promotion systems, job stability, tenure protections, and pay-levels, the school that studies the criteria through which bureaucrats are recruited counts with a high degree of consensus. This literature differentiates between political and meritocratic criteria and argues that bureaucrats recruited and promoted based on merit are more likely to tackle corruption than those who were appointed based on political ties.
Meritocratic Bureaucracies. According to the International Encyclopedia of Political Science, all functionaries among the civil service of a polity are to follow a kind of “moral code” (Del Carmen Pardo 2011, 256). This ethical conduct has a basis on three central values reflected in the logic of the bureaucratic organization proposed originally by Max Weber. These three values are equality, meritocracy, and impartiality. Particularly, the reasons behind these three central codes are the following:
“1. Recruitment is based on the merits of the civil servants and not on patronage or inherited posts.
2. The idea of equality allows anyone to participate in the competition to occupy public posts, provided that they meet the prerequisites established for the position in question; and, once they are in the civil service, the system of evaluation and any sanctions should be comparable for all the functionaries.
3. Impartiality guarantees that the members of the civil service should maintain a neutral position toward political parties [...] to grant all applicants the same opportunities and, when dealing with the public, provide the same attention and quality in the services offered to all citizens regardless of their gender, religion, or partisan preference” (Del Carmen Pardo 2011, 256).
Several studies have presented robust evidence supporting the virtues of meritocratic recruitment. They have identified a theoretical relationship between the meritocratic features of bureaucracies and lower corruption, each one of which indicating that meritocratic bureaucracies constrain corruption (Rauch and Evans, 2000; Dahlström et al., 2012; Neshkova and Kostadinova, 2012; Meyer-Sahling et al., 2016; Oliveros and Schuster, 2018; and Bersch et al., 2017b; Boräng et al. 2017). This literature has linked meritocratic
recruitment to lower corruption through at least two overarching mechanisms: the institutional and the behavioral mechanisms.
First, the institutional mechanism has as its focal point the relationship between bureaucrats and politicians, and it bases itself on the idea that the bureaucratic institution operates independently from the political one (Neshkova and Kostadinova, 2012; Dahlström et al., 2012; Oliveros and Schuster, 2018; Meyer-Sahling and Mikkelsen, 2016; Boräng et al. 2018; Bersch et al., 2017b). This mechanism suggests that the institutional arrangement of separation of careers shields bureaucrats from political pressures, and creates the ideal environment for bureaucrats to adhere to the norms of impartiality stipulated in the Weberian bureaucratic structure. In other words, when political criteria are absent during bureaucratic recruitment, the decision-making process of civil servants is sheltered from the rent-seeking behavior of politicians. This is given that in such an institutional arrangement bureaucrats are not liable to the discretionary will of politicians, and therefore have an improved ability to cope with political pressures (Neshkova and Kostadinova 2012, 326-7). To the separation of careers explanation, the institutional mechanism adds the further theoretical claim of reciprocal accountability. Dahlström et al. (2012) argue that recruiting bureaucrats based on merit considerations allows for the emergence of a bureaucratic structure with interests different from those of the political structure. This difference of interests, in turn, incentivizes both politicians and bureaucrats to keep each other on the check and creates an institutional environment with high levels of accountability (Dahlström et al. 2012, 656). By highlighting the importance of the balance and mutual control inherent in countries with low corruption, these authors emphasize that it is not only politicians who can engage in corrupt practices, but also bureaucrats, and as such, that the institutional accountability should be reciprocal.
It is worth emphasizing that for the separation of careers and reciprocal accountability arrangements to take place, the constitutional stipulation of impartiality in bureaucracies is not enough. Several scholars have found that civil service laws contemplating meritocratic bureaucracies are negatively related to corruption only when they are correctly executed (Meyer-Sahling et al., 2016; Oliveros and Schuster, 2018; Neshkova and Kostadinova 2012). This is because institutions and rules alone do not prevent corruption, but rather induce in bureaucrats certain behaviors, and affect their psychology and motivations. Thus the importance of focusing on the behavior of the bureaucrat: to find a sufficient explanation for the relation. Although having many overlapping attributes, the behavioral focus zooms in at
the effect that the institutional structure poses on the psychology of the individual, and studies how the aggregate institutional level induces and shapes certain behaviors on the bureaucrat.
In this way, the behavioral mechanism is interested in the cues that the institutional arrangements suggest to the individual bureaucrat, and it is based on the idea that meritocratic recruitment shapes the behavior of civil servants towards a more honest, neutral and objective one (Rauch and Evans, 2000; Cooper, 2018). First, Rauch and Evans’ (2000) cross-national study argues that bureaucrats recruited based on their educational achievements or through competitive examinations are often concerned with their reputation among colleagues (5). This concern, in turn, creates an esprit de corps among officials, which leads them to greater adherence to norms of behavior such as integrity and honesty (Rauch and Evans 2000, 5). Second, Cooper (2018) argues that civil servants who believe to be working under a bureaucratic structure with high levels of meritocratic recruitment are more likely to express dissenting opinions upwards than those who believe not to be under a meritocratic bureaucracy (2018, 721). This is because the meritocratic structure induces in those civil servants confidence in their capabilities, and makes them less fearful of being reprimanded for raising their voices and providing candid suggestions. In other words, the perception of high levels of meritocratic recruitment signals civil servants that they are politically impartial, and thus that they are assumed to always provide objective and forthright advice (Cooper 2018, 724).
The concept of “credible commitment” advanced by Miller (2000), might help illustrate why a bureaucratic structure based on merit induces the individual bureaucrat to act according to the moral code of civil servants. Bureaucrats that are hired based on merit have nothing to gain from violating the trust of their managers and colleagues, for the simple reason that their initial entrance at the bureaucratic department is closely linked with them being honest and trustworthy (Miller 2000, 301). As a result, bureaucrats selected based on their capabilities, focus on strengthening those skills intrinsic to the moral code of civil servants, and have the “knowledge that they would be ultimately rewarded and not squeezed”, if they maintain those attributes through which they were valued in the first place (Miller 2000, 301). Thus the commitment to be an honest and trustworthy bureaucrat becomes credible.
Table 2. Overview of representative empirical studies on the topic of meritocratic recruitment Name of the author Relationship studied Mechanism explaining the relationship Method and Design Unit of analysis Rauch and Evans (2000) Meritocratic recruitment and Bureaucratic performance Characteristics of Weberian bureaucracy: formal and informal
rules Cross-country comparison (semi- and non-industrialized countries) Bureaucratic structure Oliveros and Schuster (2018) Meritocratic recruitment / Job stability and Bureaucratic behavior Separation of careers Conjoint experiment Bureaucratic structure Neshkova and Kostadinova (2012)
Civil service reform and
Government transparency / Foreign
direct investment
Separation of
careers Cross-country comparison (post-communist countries) National civil service reform Meyer-Sahlin g and Mikkelsen (2016) Bureaucratic perception of meritocratic recruitment and Degree of corruption rumors heard by bureaucrats Effective civil service laws that aim at meritocratic recruitment and separation of careers Cross-country comparison (post-communist countries) Individual public officials Boräng et al. (2018) Bureaucratic politicization and Production of policy knowledge Integration of
careers Within-case analysis and large-n time-series cross-country analysis. Argentinian statistical agency
Cooper (2018) Merit recruitment and Fear to voice Separation of careers and Structural suggestion of neutrality and objectivity Within-case analysis (Survey design) Federal public service employees in Canada Bersch et al. (2017a) Agency politicization and Corruption Partisan dominance
of federal agencies Comparative analysis (quantitative data sets) Brazil’s federal agencies Dahlström and Lapuente (2017) Professional bureaucracies and Corruption / governance Separations of careers and Reciprocal accountability Cross-country comparison (52 countries from the developed and developing world) Expert survey National bureaucracies
Altogether, the body of literature on the internal controls of corruption shows that the meritocratic criteria in the recruitment of bureaucrats is the variable to be most robustly correlated to lowered corruption. However, these studies have only mentioned tentative mechanisms for explaining their statistically significant correlations; they have not empirically examined the proposed and possibly alternative mechanisms explaining this relationship (Meyer-Sahling, Mikkelsen and Schuster 2018, 279). The next section develops a novel, multi-level argument to account for this omission.
Table 3. Theoretical mechanisms proposed in the literature linking meritocratic recruitment in bureaucracies with lowered corruption
Mechanism Explanation
Institutional Separation of careers and reciprocal accountability
2.2 What Remains to Be Studied
Although seemingly analogous, this research understands the behavioral and institutional explanations as different, given the different levels of analysis at which they operate. To increase the understanding of how meritocratic recruitment impacts corruption, the theoretical discussion might benefit significantly from approaching the issue in a multi-level framework, by taking a closer look at the interplay between the institutional rules and the individual behavior of bureaucrats in context specific instances.
Integrating the literature. Why is corruption more likely to be tackled when civil servants are recruited based on merit rather than on personal or political discretion? A reason for this might be given the emergence of 1) a bureaucratic shield against political pressure (separation of careers); 2) divergent interests between bureaucrats and politicians (reciprocal accountability); 3) an organizational ethic based on reputation (esprit de corps); or 4) confidence in the individual bureaucrat to voice a dissenting upward (employee voice).
Intuitively, one can argue that in contexts where anti-corruption measures are being effectively applied, such as the context presented in the operation Car Wash Case, each of the aforementioned mechanisms provides linkage for the relationship between meritocratic recruitment and lowered corruption. This is because each one of the theories examines reality from a different unit of analysis: while the separation of careers and reciprocal accountability mechanisms focus on the interplay between both the political and bureaucratic structures as the unit of analysis, the esprit de corps and employee voice mechanisms focuses on the individual bureaucrat as the unit of analysis.
From this follows that each suggested mechanism need not be excluded from the other in the cases where anti-corruption measures are advanced. It is rational to expect the emergence of all the suggested mechanisms in the Car Wash operation. The present study attempts to provide an answer to how anti-corruption measures were established in Brazil by hypothesizing that both mechanisms were present operating simultaneously at both the institutional and individual levels. The following section elaborates further on these suppositions as well as on the general method of the research. Eventually, the idea behind this deductive approach is to investigate whether it is possible to integrate seemingly reconcilable
theoretical frameworks, that to the date remain loose and not empirically related to each other.
3. Data and Method
3.1 Process-tracing
This research examines whether the correlation between the independent variable of meritocratic bureaucracy and the outcome of lower corruption is causal by tracing the mechanisms that produced the outcome—in this case, the effective tackling of corruption. To do so, the research utilizes the explaining outcome case-centric process-tracing approach (Beach and Pedersen 2013, 283-5). This approach aims to find a convincing explanation for a puzzling outcome in a specific contextual case (Beach and Pedersen 2013, 285). In doing so, it pursues both deductive and inductive methods sequentially. First, through a hypothetical deductive manner (theory-testing), the research: 1) conceptualizes possible mechanisms that can explain the outcome; 2) formulates hypotheses based on the literature; and 2) collects evidence from the case to test such hypotheses. Then, through an inductive approach (theory-building), the research: 1) collects facts about the case to develop an empirical narrative; 2) infers the existence of particular phenomena; and 3) infers the existence of a causal mechanism. The research takes both deductive and inductive approaches because the theoretical literature rarely provides a sufficient explanation of the specific case. The empirical data collected through the inductive approach allows for identifying case-specific mechanisms and, consequently, a possible reconceptualization of the theory. Ultimately, this explaining outcome case-centric approach represents a research process in which initial mechanisms proposed by the theory are reconceptualized until a sufficient explanation is found. 4
3.2 Case Selection
In 2014, the PF and MPF initiated the Car Wash operation in Brazil, an anti-corruption investigation that revealed a money-laundering scheme of unprecedented scale in the country. The investigation involved, among other actors, the careful inspection of several political 5
4Note that this research understands mechanisms as systems of interacting parts that work together or on each other to produce an outcome. If one of the part is absent or weak, it can be expected that the outcome will also be deviated.
5The MPF is responsible for defending the social and individual rights of citizens before the STF, the Superior Court of Justice, federal regional courts, federal judges and electoral judges. The MPF has autonomy in the structure of the State, so much so that it has been regarded by scholars as a fourth governmental branch—apart from the judiciary, legislative, and executive) (Bersch et al., 2017a, 167).
actors coming from the whole spectrum of political parties, among which were the Brazilian Social Democratic Party (PSDB), Workers Party (PT), Progressistas Party (PP), Brazilian Democratic Movement Party (PMDB), Brazilian Labour Party (PTB), and Solidariedade Party (SD) (PF, 2019). This list of political parties included high ranked politicians such as former President of the Chamber of Deputies (Eduardo Cunha), and the former President of the Senate (Renan Calheiros), among others . The investigation has developed several new chapters and is currently ongoing.
The Car Wash operation was selected as the case study for the present research because it provides detailed information regarding the effective implementation of anti-corruption measures. The case presents itself as an opportunity to determine the necessary mechanisms allowing the establishment of anti-corruption measures in highly corrupted contexts. Given the far-reaching political consequences of the operation, it is puzzling that the anti-corruption efforts were initiated solely by bureaucrats without political influence whatsoever. Besides, it is an extreme case as, when compared to previous ones, it has been regarded as the largest anti-corruption investigation of Brazil (Castro and Ansari 2017, 351-2). Given that the theory emphasizes the relationship between meritocratically recruited bureaucrats and politicians as key for explaining the relationship between meritocratic recruitment and lower corruption, the Car Wash operation case presents itself as well fit for examining the mechanisms at play between the bureaucrats-politicians interaction. Last, reconstructing the chain of causal mechanisms of the Brazilian anti-corruption case could reveal theoretical aspects that might have been overlooked in past research. This is because pinpointing common relationships across cases in cross-case analyses—as, among others, Rauch and Evans (2000) and Dahlström and Lapuente (2017) did—may mitigate part of the conclusions intrinsic of individual cases.
The year 2014 was chosen as a starting point because it was when the Car Wash operation began. The study concludes its analysis in December 2019, with the rejection of Lula’s appeal by the judiciary, and his increased prison sentence (MPF, 2019). Given its extensive reach, the population of cases comparable to the Car Wash operation is relatively small. 6
6For an example, see the Mani Pulitecase, an Italian judicial investigation into political corruption that resulted in hundreds of arrests of congressional politicians.
The present design is not free from the typical qualitative limitations. Its main weaknesses are its low external validity. Given that the focus is on a single case study, and that the findings are likely to be highly context-dependent, the insights drawn from here are not to apply with certitude to other cases. The strength of it, however, is its high internal validity, namely the capacity to develop strong within-case inferences (Beach and Pedersen 2016, 24). For more theoretical generalizability of the findings of this study, new analyses on7 different anti-corruption cases are required.
3.3 Operationalization
As noted, the research assumes that both overarching mechanisms operate simultaneously at different levels of analysis. This section specifies three research hypotheses informed by the theory and the expected evidence that would prove or disprove each one of them.
The predicted evidence is evaluated based on the probabilities of its existence. As such, stronger inferences can be advanced when a highly unlikely piece of evidence is found (Beach and Pedersen 2013, 122). To do this, a specification of the likelihood of the predicted evidence is provided a priori. Eventually, testing if a theoretical mechanism is present in a given case entails examining if the theoretically-steered predictions of the evidence that is to be observed in that case match reality (Beach and Pedersen 2013, 124). Indeed, the research is not confined to search only for evidence that confirms the hypotheses. A thorough check on evidence that would disprove the hypothesis is also to be advanced (Thies 2002, 364).
Institutional mechanism. The research looks at the recruitment nature of the agencies that advanced the anti-corruption measures in Brazil. Namely, whether the bureaucrats conforming the Car Wash task-force were appointed by politicians or recruited based through formal civil service examinations.
Hypothesis 1 (H1): Bureaucrats that advanced the operation Car Wash were
meritocratically recruited and not politically appointed.
Given that the high number of scholars who suggest a negative correlation between meritocratic recruitment and lower corruption, it is highly likely that the study finds evidence confirming that all bureaucrats who advanced the operation Car Wash were recruited based on merit. The probabilities of finding pieces of evidence that would disprove or weaken H1 are therefore low. Because of the low probabilities, a single piece of evidence showing that one or more of the bureaucrats who advanced the operation Car Wash were politically appointed would be enough to challenge the solid theoretical base.
Also concerning the institutional focus, the research looks at the extent to which politicians were accountable to bureaucrats, as well as the extent to which bureaucrats were accountable to politicians. This is to determine whether in the case of Brazil, the accountability mechanism is reciprocal, as argued by Dahlström and Lapuente (2017), or one-sided—e.g. bureaucrats keeping politicians accountable but not vice-versa— as suspected by Meyer-Sahling et al. (2018, 279).
Hypothesis 2 (H2): The Car Wash operation showed the appearance of a reciprocal accountability mechanism.
Here again, the theoretically informed prediction gives high probabilities of finding evidence of the reciprocal accountability mechanism. Pieces of evidence confirming H2 would include bureaucrats taking politicians to court to testify for claims of corruption, and politicians taking bureaucrats to court to testify for claims of corruption; bureaucrats leaking information to the press that would reveal schemes of political corruption, and politicians leaking information to the press that would reveal schemes of bureaucratic corruption. On the other hand, there is a low likelihood of finding pieces of evidence that would disprove or weaken H2. Such evidence would include only one of the careers keeping the other one accountable: bureaucrats taking politicians to court to testify their doings, but politicians not doing the same with bureaucrats; politicians taking bureaucrats to court to testify their doings, but bureaucrats not doing the same with politicians.
Behavioral mechanism. To test the behavioral proposition, the research looks at the extent to which bureaucrats that were involved in the investigation had the confidence to voice
dissenting opinions upwards. This is to determine the role that such confidence played in the making of an effective anti-corruption investigation.
Hypothesis 3 (H3): Bureaucrats that advanced the operation Car Wash had
high levels of confidence to voice dissenting opinions upwards.
Pieces of evidence that would prove H3 are most likely to be found. An example of a piece of evidence confirming H3 would include the continuation of the case by a bureaucrat after having received threats to his or her integrity for voicing the corruption plot. Differently, there is a low likelihood of finding pieces of evidence that would disprove H3. Such evidence would include bureaucrats who initiated the operation Car Wash being discouraged amid the process for risks posed against their integrity.
It is worth noting, however, that the present research understands the difficulties of empirically observing the general attitudes, mentality and confidence of the bureaucrats advancing the anti-corruption efforts without a survey, and recognizes this as part of the methodological limitations.
3.4 Data
Primary sources. All the primary data is available via government websites or international online databases. The MPF, for example, has documented all legal proceedings of the case on its website. Databases in the legislative and judiciary branches of the Brazilian government are also consulted; Brazilian legislation regarding the anti-corruption laws and investigative methods of anti-corruption organizations throughout the years of the operation Car Wash are taken into account for the construction of the “overarching narrative” (Bazeley 2013, 115). The research examines additional data on websites and communiqués of government investigative agencies, particularly the central agencies of MPF, PF, and CGU. Finally, the study looks into audiovisual interviews of the relevant actors: bureaucrats and politicians key in the Car Wash investigation. For this, the research accesses the youtube online database and examines only the interventions that count with no editions by third parties. This is to hinder the possibility of a third party transgressing the meaning intended by the studied actor.
Secondary sources. Secondary data are also consulted. These include the QoG expert survey data provided by Dahlström et al. (2015), which measures the national levels of the meritocracy of several countries, among which Brazil; and the sub-national categorization of agencies’ autonomy advanced by Bersch et al. (2017b). Also, the research analyses news coverage and published bulletins of the operation Car Wash anti-corruption efforts in major Brazilian newspapers covering a period of 7 years. It uses the Manchetômetro database from the Federal University of Rio de Janeiro, a website for monitoring major media coverage on different topics of national relevance, and with no political affiliation, to pinpoint the main Brazilian outlets focusing on the Car Wash operation (Manchetômetro, 2020). The study triangulates among a variety of primary and secondary sources to produce a more accurate representation of history (Thies 2002, 359). Table 3 presents an overview of the data sources used for the present research.
Table 4. Data sources
Governmental branch Institutional online platforms Parliamentary Commission of Inquiry on Car Wash
operation Pinpoint institutional reactions Pinpoint accountability of bureaucrats on politicians 3 institutional webpages (MPF, PF, and CGU) consulted, and approximately
1,000 transcribed pages read. Sergio Moro and Deltan Dallagnol (key anti-corruption officials) questioned Academia EBSCOhost The Quality of Government Institute (QoG)
Measures the level of meritocracy per individual agency in
Brazil Measures the national levels of the meritocracy of Brazil
Bersch et al. (2017b)
Dahlström et al. (2015)
Media Manchetômetro Factiva
Youtube Pinpointing alternative mechanisms 2,007 Brazilian and International newspaper headlines, 2014-2020 70 hr of interviews
held by key actors identified below in
4. Findings
4.1 Research context
Bureaucratic recruitment in Brazil. The pioneering QoG country-level data advanced by Dahlström et al., (2015), which was filled by 520 public administration experts from 97 Western, post-communist, and non-Western countries, presents, among others, Brazilian national levels of professionalism (meritocratic recruitment). The data shows that the level of professionalism in Brazil is slightly above average (45th). In the data, Brazil ranks higher than Western countries such as Italy (54tt) and Greece (65th), and all other Latin American counterparts but Uruguay (21st).
Yet, only focusing at the national level of bureaucratic meritocracy might overlook the significant differences existing within the country itself. As it is the case with Brazil, some agencies possess worldwide renown for their high levels of meritocracy and professionalism, and others are fully politicized (Bersch et al. 2017a, 158). This observation highlights the utility of looking at data not only on the expert’s perceptions and at a country-level, but also on the factual way the bureaucrat is recruited at an individual-level.
The Federal Transparency Website, online since November 2004, is provided by the CGU, and collects data listing all individual civil servants employed at the federal level, including their recruitment criteria. This data is updated monthly and the data used in this8 research is from April 2020. The data counts with four principal categories of bureaucrats: those recruited through RJU (Portuguese: Regime Jurídico Único) rules, those recruited through CLT (Portuguese: Consolidação das Leis do Trabalho) rules, and those promoted/appointed by DAS (Portuguese: Direção e Assessoramento Superior) offices. Bureaucrats recruited under an RJU and CLT contracts have passed civil service exams. Their contracts differ mainly in the provision of tenure protections. Bureaucrats holding a DAS contract, however, have been either promoted for technical reasons (administrative expertise) or appointed due to political criteria. The difference between the DAS promotion and the DAS appointment is that the promoted civil servants have passed the civil service examination, and the appointed ones have not. These appointed bureaucrats are outsiders who
hold their positions in government for purely political reasons, for example, due to political agreements/relationships, and or academic ties (Bersch et al. 2017a, 167-8).
Table 5. Different categories of bureaucratic recruitment in Brazil (Bersch et al. 2017a, 167-8).
RJU
contract contract CLT promotion DAS DAS appointment
Recruitment criteria Civil service examination Civil service examination Administrative expertise Political ties
The institution of the public tender (Portuguese: Concurso público) has been a strong method of bureaucratic recruitment in Brazil since the middle of the last century. This public tender refers to a selection process that aims to evaluate candidates competing for a permanent position in the Brazilian State apparatus (Castelar et al. 2010, 81). Competitions are generally required to assess the competence of candidates for the position for which they are running (Castelar et al. 2010, 81-2). The State is responsible for identifying and selecting the most appropriate bureaucrat according to objective criteria. The idea of public tender follows Article 37, item II of the Brazilian Federal Constitution, which reads that “Investiture in a public office [...] depends on prior approval of a public exam, [...] according to the nature and complexity of the job, as provided by law, except for nominations for office in a declared commission” (Presidência da República, 2016).
Three types of public tenders are offered based on the complexity of the position. These are: 1) the fundamental level (Portuguese: nivél fundamental), which offers positions that do not require formal education (e.g. firefighter, cook assistant, and bricklayer); 2) the middle-level (Portuguese: nivél médio), which offers positions that do require more elaborate educational preparation (e.g. electrician, receptionist, and dentist assistant); and 3) the superior level (Portuguese: nivél superior), which offers positions requiring the highest level of educational attainment (e.g. surgeon, lawyer, architect). The examinations for individuals aspiring for positions in the fundamental level involve the topics of Portuguese, mathematics, history, and “basic knowledge”. The examination for those aspiring for middle level positions involves maths, Portuguese, informatics, geography, and basic knowledge. And the
examination for those aspiring for the superior level involves informatics, geography, and history (Concursos no Brasil, 2020).
Table 6. Different types of meritocratic paths for entering the public sector in Brazil (Concursos no Brasil, 2020).
Bureaucratic
complexity Example of positions Required knowledge for examination
Fundamental Firefighter, cook assistant, and bricklayer
Basic: Portuguese, mathematics, history, and
“basic knowledge” Middle Electrician, receptionist, and
dentist assistant
Basic: informatics, and geography
Advanced: Portuguese, maths, and basic knowledge Superior Surgeon, lawyer, architect Advanced: informatics,
geography and history
Car Wash operation. On 17 March 2014, the MPF made public an investigation concerning a corruption network that involved several Brazilian politicians and the CEOs from the largest state-owned and private companies in the country. The operation had started as a simple routine investigation, as the PF identified a series of bribes and money-laundering indicators stemming from an exchange office located in a Car Wash station (Lava Jato) in Brasilia. Little did the PF know that such a case was to become part of the largest corruption scheme in Brazil’s history, with far-reaching political and economic implications, not only in Brazil but in several Latin American countries. The investigations that led to the first arrests in 2014 date back to July 2013, when the PF in Curitiba, southern Brazil, discovered a money-laundering network that operated from Car Wash stations. In this opportunity, the firsts accused were Carlos Chater, the owner of the service station, and one of his partners, Alberto Yousseff (MPF, 2014). With these first captures, the MPF and PF began to unveil a network of bribes that involved top executives, such as the former director of the state-owned company Petrobras, Paulo Roberto Costa (MPF, 2014). These executives, in turn, provided testimonies that served to reveal more individuals responsible for the scheme. The
ramifications of the first accusations reached high levels in the political circle. In 2016, the name of Dilma Rousseff—Lula’s political heir—began to appear among those involved in corrupt activities, as she was accused of fiscal pedaling. With this scandal, Dilma started losing support among politicians and citizens (The New York Times, 2016). So much so that in August of that same year, Dilma was impeached and removed from her office (Senado Federal, 2016).
Yet the most symbolic case of the operation was that of Brazil’s former President Lula. In April 2018, Lula was sentenced to 25 years in prison for corruption and money-laundering (MPF, 2018). Specifically, the man who served 8 years as President was convicted of receiving an apartment as an exchange for profiting certain construction companies through the issuing of contracts with Petrobras. This sentence removed Lula from the 2018 presidential campaign, in which he was to compete against current President Jair Bolsonaro. Knowing about his arrest, Lula declared himself innocent and accused former Judge Sérgio Moro of bringing a political trial against him, and of advancing a coup d’etat (The Guardian, 2019; Brasil247, 2019) . 9
Furthermore, Senior executives of major construction companies have also been convicted in the process. Marcelo Odebrecht, president of the Odebrecht Group between 2008 and 2015, is serving a sentence of 19 years and 4 months at his residence, after having spent a couple of years in prison (MPF, 2015). Otávio Márques, a former CEO of the construction company Andrade Gutiérrez, was sentenced in 2016 to 18 years in prison and is also serving his sentence in house arrest (MPF, 2016). And Léo Pinheiro, the former president of the construction company OAS and one of the informers against Lula, was sentenced to 35 years in prison (MPF, 2016).
4.2 Examining the Mechanisms Proposed in the Literature
Institutional: Separation of careers. Despite the extensive reach of the operation Car Wash and the many civil servants involved in the investigation, the present research has identified key bureaucrats who played a critical role in the operation. Table 5 provides an
9 Sérgio Moro, the emblematic judge of the Car Wash operation, was designated Minister of Justice in 2019 by President Jair Bolsonaro. On April 23rd of 2020, Sérgio Moro resigned from his position as Minister. Moro claimed Bolsonaro was overusing his executive power, and that he was not to support such collusion (BBC News, 2020).
overview of who these actors are, their functions during the operation, and their selection criteria.
Table 7. Overview of key bureaucrats in the Car Wash operation.
Public agency Selection
criteria Role in the operation
Sérgio Moro Department of Justice. Regional Federal Court
of the 4th Region
Civil service
examination Judge in the Car Wash Former head Federal operation
Márcio Anselmo PF Civil service
examination Delegate of the PF for the Car Wash operation. Specialized in organized
crime Igor Romário
de Paula PF examination Civil service Delegate of the PF for the Car Wash operation. Specialized in financial crime Deltan Dallagnol MPF Civil service examination
Lead prosecutor of Car Wash operation
Erika Marena PF Civil service
examination Chief of the PF in the Car Wash Operation
Newton Ishii PF Civil service
examination participated in the arrest of Police Agent that key actors
Carlos Fernando dos
Santos Lima
MPF Civil service
examination investigated the Car Wash General Attorney that operation
Marco Aurelio
Canal The Special Department of Federal Revenue of Brazil (Receita Federal)
Civil service
examination Department of Federal Federal Auditor of the Revenue Paulo Roberto
Costa Petrobras Political ties Former Petrobras Supply Director Gabriela Hardt
Department of Justice. Regional Federal Court
of the 4th Region
Civil service examination
Substitute of Moro - head Federal Judge in the Car
Concerning the separation of careers mechanism, the research derives three important findings. First, all bureaucrats leading the task force of Car Wash operation had to pass a civil service examination exam. Second, not all these meritocratically recruited bureaucrats were exempt from corruption themselves. And third, the politically appointed bureaucrat identified at Petrobras was a key figure in the corruption scheme.
With regards to the first consideration, the research found that the Car Wash task-force was indeed recruited in compliance with the public tender article stipulated in the Brazilian Constitution. That is, documentary evidence was found proving that the bureaucrats advancing the operation had passed through formal civil service examination. This finding lends credence to H1.
Concerning the second consideration, the PF agent Newton Ishii and the Federal Auditor Marco Aurelio Canal, both integral part of the Car Wash task-force, were found by the Brazilian Justice to be involved in different illegal activities. On one hand, Newton Ishii was sentenced to 4 years and 2 months in prison for facilitating smuggling activities into the country (Folha de S. Paulo, 2016a). Although the contraband Ishii was accused of was irrespective of the Car Wash operation, it is still considered an act of corruption inflicted by a bureaucrat recruited through a civil service examination exam. On the other hand, the illegal activities advanced by Marco Aurelio Canal were related to the Car Wash operation. According to investigations advanced by the MPF in October 2019, Canal participated in a bribery scheme formed by revenue inspectors who charged bribes to taxpayers targets of scrutiny. The scheme induced those investigated in the Car Wash operation to pay bribes to revenue inspectors in order not to receive any fine. The position of Federal Auditor held by Marco Aurélio Canal was part of an organism created to restore tax evasions to public coffers. According to the MPF, Canal had access to confidential information from the Car Wash operation by occupying this position. Upon learning novel information, Canal and his group, instead of collecting the fines determined by the superiors in the Treasury, negotiated bribes with those investigated. The investigation against Canal began with the declarations of Lelis Teixeira, former President of the transport company Fetranspor, who was also accused in the Car Wash operation. Teixeira reported that he negotiated the payment of US $ 750,000 with Canal to avoid a fine against his own company. The investigation led the PF to arrest Canal and twelve prosecutors from the Federal Tax Administration suspected of collecting bribes from people investigated in the Car Wash operation (MPF, 2019). Canal spent 2
months in prison, as in December 2019 an STF Minister affirmed that Canal freedom did not represent a risk to the Car Wash investigations (Estadão, 2019; G1, 2019b; O Antagonista, 2019). 10
Last, and concerning the third consideration, Paulo Roberto Costa was not recruited to the public position through civil service examinations but rather through a political appointment. Lula’s coalition had appointed Costa to his former position of supply director of Petrobras in 2004. His involvement in the corruption scheme was key, as he acted as the node connecting construction companies with politicians (MPF, 2014). Indeed, in March 2014 Costa was found guilty of corruption by the meritocratically recruited bureaucrats of the Lava Jato task-force.
Institutional: Reciprocal accountability. The findings concerning the reciprocal accountability mechanism in the Car Wash operation prove the theoretically informed hypothesis. The interplay of bureaucrats regulating politicians and politicians regulating bureaucrats is found throughout the 6 years of the operation.
Many instances were found in which bureaucrats kept politicians accountable. Indeed, what motivated this research was the significant number of politicians who were kept accountable by the Brazilian bureaucratic structure. To name one: the imprisonment of the President of the Chamber of Deputies of Brazil, Eduardo Cunha. On 19 October 2016, Cunha was arrested by the PF on the grounds of having hidden approximately US $ 40 million stemming from bribes in secret foreign bank accounts (MPF, 2016).
On the other hand, the research also found instances in which politicians kept bureaucrats accountable. One of these was when former Judge Sérgio Moro had to appear before the Justice and Citizen Commission (CCJ) of the Brazilian Senate, due to the publication of reports advanced by The Intercept Brasil revealing irregularities in the Car Wash investigation. The CCJ is the most important permanent commission of the Federal Senate of Brazil. It has the competence to appreciate through public hearings all the national issues passing through the Legislative Assembly (Senado Federal, 2020). In that instance testifying before politicians, Moro denounced having been the victim of a hacker attack. Four
10 The STF is the highest instance of the Judiciary. Its main task is to judge whether the laws and acts of the government are in accordance with the rules and principles provided for in the Federal Constitution. The court is composed of 11 ministers, all appointed by the President of the Republic. Those nominated by the president must have their name approved in the Senate by the vote of at least 41 of the 81 senators.
days after Moro’s defense, a hacker named Walter Delgatti Neto indeed confessed to the PF that he was responsible for breaking into the cell phones of Moro and the attorney Dallagnol when they both led the Car Wash operation. After a thorough investigation of the PF, the hacker, who already had a criminal record for fraud, forgery of documents and theft, was arrested (CCJ Sérgio Moro, 2019).
Another instance in which politicians were found to keep bureaucrats advancing anti-corruption efforts accountable, appeared when the National Council of the MPF (Conselho Nacional do Ministério Público) opened in December 2019 a new disciplinary administrative proceeding against Dallagnol, based on the accusations of a politician. The 11 complaint that resulted in the proceeding was presented by Senator Renan Calheiros (MDB). Calheiros had requested Dallagnol’s removal from his public service position for exercising partisan political activity. Yet, the majority of the council denied the request, and authorized only the opening of disciplinary administrative proceedings for breach of decorum—a proceeding that does not consider full expulsion, but rather public censorship (Conjur, 2019; Veja, 2019; Exame, 2019).
Behavioral: Employee voice. The case presents evidence of a series of instances in which bureaucrats advancing the Car Wash operation showed high levels of confidence to raise their voices. First, the research found evidence proving threats to the life of the former Judge Sérgio Moro. The Federal Judge in charge of the Car Wash operation in Rio de Janeiro, Marcelo Bretas, collected eight of these threats and published them on his social network account. All posts were published when the decision to release Lula was opposed by the Regional Federal Court of Curitiba, namely by Judge Moro. All of the messages compiled by Bretas refer to “killing Moro” (O Globo, 2017b; R7 Noticias, 2018; Veja, 2018).
Then, in the context of the impeachment of Dilma Rousseff in 2016, the same Judge Moro showed high levels of confidence when he leaked to the national press—through an illegal proceeding—a recording of a phone call between Dilma and Lula. In the dialogue, Dilma told Lula that she was going to send the former President a paper for him to use “in case of need”. The paper was to appoint Lula as Minister of State Chief of Staff of the
11 In Brazil, the National Council of the MPF is an external body responsible for controlling and inspecting the administrative and financial performance of the bodies comprising the MPF, as well as supervising the fulfillment of the functional duties of its members.
Presidency of the Republic (Ministro de Estado Chefe da Casa Civil da Presidência da República). To the notice, Lula replied only that he would wait for Dilma’s aide.
Conversation between Lula and Dilma “Dilma: Hello
Lula: Hello
Dilma: Lula, let me tell you something. Lula: Speak, dear.
Dilma: It is this, I’m sending Bessias along with the paper so we can have it, and only use it in case of need, which is the term of possession, okay?
Lula: Uhum. Okay, okay.
Dilma: Just that, you wait there that he is going there. Lula: Okay, I’m here, I’m waiting.
Dilma: Okay? Lula: Okay. Dilma: Bye.
Lula: Bye, dear”. (G1, 2016)
The recording of the conversation had taken place outside of the time that the court had ruled for it. According to a document of the PF, the dialogue between the two was intercepted at 1:32 pm on March 16, 2016. However, Moro had determined the end of Lula’s telephone interception at 11:20 am on that day. The investigators from the PF did not suspend the telephone recording when they were supposed to, and thus could identify the call. The PF agents informed Moro about the recording, and Moro authorized the release of this audio to the press.
Despite knowing the illegality of his action, Moro had the confidence to voice his judgment upward, and leaked the audio under the “public interest” claim. Moro claimed that12 “democracy in a free society requires the governed to know what the governors do, even when they seek to act protected by the shadows”. The recording was meant to incriminate both former Presidents, given that the appointment of Lula as Minister would signify that he was going to be tried by the STF through “Privileged Forum” (Foro Privilegiado), rather than
12 Moro’s action can be interpreted as a direct attack towards two of the most powerful individuals of the country.
