In the Name of Integrity? Anticorruption Discourse of Brazilian Operation Lava Jato

Abstract

This article identifies four frames of corruption in the discourse of three leaders of Operation Lava Jato, also known in English as Operation Car Wash, a large-scale Brazilian anticorruption operation (2014–2021). These frames are inequality, hidden pact, backwardness, and chronic disease. The frames were identified by analyzing a wide set of press interviews, opinion articles, and books by two prosecutors and one judge whose work has revealed scandals involving the state oil company Petrobras. The operation had a major impact on politics and the economy and left a controversial legacy. We noticed a contradiction between one frame invoking judicial activism (inequality) and three frames focusing on specific techniques that appeal to a more conventional view on the judiciary’s role (hidden pact, backwardness, and chronic disease). Furthermore, even when scholars were still largely positive about the operation, the discourse showed signs of judicial activism. This analysis contributes to the debate on Lava Jato and judicial activism by focusing on discourse rather than action.

Resumo

Este artigo identifica quatro enquadramentos da corrupção na retórica de três líderes da operação anticorrupção Lava Jato (Brasil 2014–2021): desigualdade, pacto oculto, atraso e doença crônica. Esses quadros foram identificados a partir de uma análise de enquadramentos de um amplo conjunto de entrevistas à imprensa, artigos de opinião e livros de dois procuradores e um juiz cujos trabalhos revelaram escândalos envolvendo a companhia estatal petrolífera Petrobras. A Lava Jato teve um grande impacto na política e na economia e deixou um legado controverso. Notamos uma contradição entre uma imagem invocando ativismo judicial (desigualdade) e três imagens focando em técnicas específicas, que apelam a uma ideia mais convencional do papel do judiciário (pacto oculto, atraso e doença crônica). Mesmo quando a maioria dos estudiosos ainda avaliava a operação de forma positiva, o discurso já mostrava sinais de ativismo judicial. Esta análise contribui para o debate sobre a Lava Jato por meio de um foco no discurso em vez da ação.

One of the world’s largest anticorruption probes, the Lava Jato (Car Wash) operation deeply affected Brazilian politics. From 2014 to 2021, businesspeople and politicians from several parties were implicated and convicted. It was also a decisive factor—together with deep economic crisis and conflicts with her main coalition partner, the PMDB party—in the impeachment of President Dilma Rousseff (Amorim Neto 2016). It also led to the jailing of ex-president Luiz Inácio Lula da Silva (Worker’s Party, Partido dos Trabalhadores, PT), which in turn led to the revoking of his political rights in the 2018 elections, which his far-right opponent Jair Bolsonaro won. Lava Jato (LJ) became a symbol of Brazil’s fall from grace in the 2010s, from the poster child of economic success and social inclusion to a crisis-ridden country balancing dangerously on the edge of collapse (Taylor 2020). Its repercussions spilled over into other countries in Latin America, with Peru and Mexico being the most strongly affected (Mattos Pimenta and Greene 2020). The scandal revealed kickback transactions involving state oil company Petrobras and its contractors, also causing huge political turmoil. In 2021, the Supreme Federal Court (STF in Portuguese) restored Lula’s political rights, enabling him to run for president. The court also agreed with Lula’s defense that Lula’s sentence had been handed down by a biased judge (Sergio Moro)—a case based on leaked messages among judge and prosecutors, which sparked the Vaza Jato media scandal (vazar means “to leak”).

While the literature on the operation has drawn attention to transgressions in its later stages (Avritzer 2018; Kerche and Feres Júnior 2018; Rodrigues 2020; Kerche and Marona 2022), little has been written on manifestations of politicization and judicial activism in its early years, when most scholars still held (relatively) positive views on LJ (Melo 2016; Praça 2017). Furthermore, little consideration has been given to how the operation’s leaders construed (anti)corruption—Sá e Silva (2020) being an important exception. Media coverage for the first five years of the operation could be described as sycophantic, and LJ appeared in Brazilian and foreign media as a kind of what McDonnell (2020) called a “pocket of bureaucratic effectiveness”—a unit able to build a system of excellence within an ineffective institutional environment (2), which helped give even more traction to its leaders’ anticorruption rhetoric.

This article analyses the discourse of Lava Jato’s three main protagonists—judge Sergio Moro (later minister of justice, 2019–2020), federal prosecutor Deltan Dallagnol (first coordinator of the LJ task force, 2014–2020), and prosecutor-general Rodrigo Janot (2013–2017) to examine how they framed corruption. After LJ, they got involved in political parties (Moro and Dallagnol were elected senator and federal deputy in 2022). Their exit from nonrepresentative institutions and onto electoral ballots makes studying their rhetoric even more salient. We use a methodological tool underused in corruption studies: frame analysis. Our frame analysis is based in cultural sociology and uses Entman’s (2006) influential concept of the frame. Specifically, we use a methodology by Lepianka (2015), who analyses frames in terms of four dimensions: problem definition, diagnosis and solution, the role of the corrupt (in Lepianka’s case, the poor), and framing devices (metaphors, visual imagery, catchphrases). This approach allowed us to identify and analyze four frames of corruption: inequality, hidden pact, backwardness, and chronic disease.

We detected a contradiction between a frame that shows signs of judicial activism (inequality) and three frames that align with more conventional understandings of the judiciary’s role in a democracy (hidden pact, backwardness, and chronic disease). Furthermore, we found that the inequality frame was present in the discourse throughout the operation, including in the beginning, when few scholars writing about it noted politicization or judicial activism.

Our contribution is twofold. First, we use a constructivist method that is rarely used in corruption studies, in order to shed light on how the protagonists of an influential anticorruption operation in one of the world’s largest democracies frame the problem of corruption and the fight against it. Second, we show that intentions of judicial activism, or more precisely, result-oriented judging (Kmiec 2004), were present in the discourse from the very beginning, before the more egregious transgressions and deviations from the judiciary’s proper role, according to classical democratic theory, manifested themselves, an aspect that the literature has overlooked.

The main implication of these findings is that the discourse of judicial actors must be taken seriously in research on the politicization of the judiciary and judicial activism, which should not limit itself to observable action as it pertains to the relationships between branches of government. Discourse reveals intentions, and intentions are crucial to the concept of judicial activism.

Perspectives on corruption

A large body of positivist work has treated corruption as a variable interacting with other variables. These variables include development and growth (Huntington 1968; Nye 1967; Méon and Sekkat 2005), inequality (Rothstein and Uslaner 2005; Uslaner 2008; Jong-sung and Khagram 2016), attitudes toward or trust in government (Seligson 2002; Anderson and Tverdova 2003), political and economic competition (Acemoglu and Verdier 2000; Sandholtz and Koetzle 2000; Montinola and Jackman 2002; Treisman 2007), and many others.

However, corruption inherently has a normative aspect, which the minoritarian group of constructivist scholars of corruption argue that their positivist peers have overlooked. In Granovetter’s (2004, 10) words, “Because defining behavior as ‘corrupt’ inevitably entails a judgment about what behavior is legitimate and what is not, there is an irreducible sociological component that has been given surprisingly little attention.” Second, corruption is subject to historical and cultural relativity: the meanings of corruption and its antonym, integrity, vary by time and place (Granovetter 2004; Tänzler et al. 2012). For instance, while the hegemonic definition of corruption is the abuse of public office for private gain (Nye 1967; Girling 1997; Hough 2013; Johnston 2014), the separation between a private and a public sphere is itself historically situated (Tänzler et al. 2012). Third, for all the econometric literature’s infatuation with cause and effect, the failure to understand social and cultural dynamics surrounding potentially corrupt behavior leaves out “important causal determinants” of such behavior (Granovetter 2004, 7).

The conceptual subjectivity of corruption warrants us to reflect on different concepts and perspectives emerging from the literature, some of which LJ discourse unconsciously engages in dialogue with. There are two historical traditions of thought on corruption: a republican one (e.g., Plato, Aristotle, Machiavelli) that views it as a systemic and moral issue and a liberal one, more influential today (e.g., Hobbes, Madison, the Federalists), for which corruption is an “individual malfeasance” (Euben 1989). Closely related to the republican tradition was the medieval understanding of corruption, “a term used to trace the ‘health’ of the community to its moral qualities, and especially to the emulation of virtue among its members” (Buchan 2012, 73). Specifically, corruption was associated with disease of the body politic.

Within the rational-choice framework that emerged in the 1960s and 1970s, and specifically within the strand of new institutionalism, corruption has been defined in terms of principal-agent theory, distinguishing between a principal or superior, who “expresses a set of preferences which specify desired outcomes,” and an agent, “whom the superior has directed to achieve these outcomes” (Rose-Ackerman 1978, 6). Corruption, then, is a situation in which “some third person, who can benefit by the agent’s action, seeks to influence the agent’s decision by offering him a monetary payment which is not passed on to the principal” (Rose-Ackerman 1978, 6). Similarly, some constructivists have proposed treating corruption as a type of illegitimate or immoral exchange between two individuals or situated within social networks, whereby said legitimacy and morality are again culturally and historically situated (Granovetter 2004; Tänzler et al. 2012).

More contemporary conceptualizations of corruption include those related to inequality, exclusion, and democracy. For Rothstein and Uslaner (2005, 53–54), “corruption transfers resources from the mass public to the elites—and generally from the poor to the rich. It is functionally an extra tax on citizens, leaving less money for public expenditures.” Similarly, for Warren (2004, 333), who conceives of corruption as a violation of the democratic norm of inclusion, “the very logic of corruption involves exclusion: the corrupt use their control over resources to achieve gains at the expense of those excluded in collective decision making or organization of collective actions.” Warren also includes duplicity as a condition: “The corrupt hide their actions because they know, or at least intuit at some level, that they are violating collective norms. The very covertness of corruption pays tribute to the violated norms” (333).

Tänzler and colleagues (2012) situate the hegemonic understanding of corruption as the abuse of public office for private gain within modernization theory. They argue that this kind of corruption is often associated with prebureaucratic societies. As such, “the discursive function of corruption therefore consists in providing a symbolic representation of modernization deficiencies and cultural lags that can only be overcome by consistently observing the binding rules and procedures of administrative-technical professionalization” (Tänzler et al. 2012, para. 25).

These different conceptualizations map onto the structure-agency debate that is central to sociological theory (Walsh 1998, 8). In this debate, one position states that individuals make up society and shape it using their agency, whereas the other position states that social structures—defined as “systems of relationships”—shape individual behavior. Clearly, the liberal and rational choice traditions, as well as those definitions stressing illegitimate exchange between individuals, can be situated within the first position, whereas the republican tradition, the definitions emphasizing inequality and exclusion, and those linking corruption to modernization can be situated within the second.

The literature we engage with most directly is on discourse about corruption. Discourse analysis is an empirical methodology grounded in social constructivism. Scholars have applied it to media discourse about corruption scandals (Berti 2018; Breit 2011; Negro 2015), as well as to the (anti)corruption discourse of specific actors, such as heads of state and government (Kajsiu 2018; Hodge, Andrade, Zarza 2018) and members of civil society (Gephart 2014, 2016). The literature has a broad geographical scope, with cases ranging from Ghana to China (Hasty 2005; Wang 2017).

In Latin America, Malte Gephart (2014, 2016) has explored how international anticorruption discourse is locally embedded in, respectively, Paraguay and Chile. Hodge, Andrade, and Zarza (2019) have analyzed the contradictions in Mexican president Peña Nieto’s announcement of the launch of an anticorruption initiative simultaneously with his apology for past corrupt acts committed by himself.

Berti (2018) found that a frame of corruption as individual malfeasance dominated in New Zealand, whereas in Italy a systemic corruption frame prevailed. Another contrast was found by Kajsiu (2018) in comparing two Latin American presidents and one European one: Rafael Correa (Ecuador) mainly framed corruption as a problem affecting the private sector, while Edi Rama (Albania) framed it as a problem confined to the public sector. Juan Manuel Santos (Colombia) occupied an intermediate position: corruption as a problem derived from the interaction between the private and public sectors.

On Brazil, Sá e Silva (2020) has linked the discourse of what he calls LJ legal officers (Moro, Dallagnol, and others) to an ideology of illiberalism. Concretely, the discourse of the so-called LJ legal officers resembles illiberal ideology in the sense that, as authoritarian leaders, they identified an existential threat to the nation and presented themselves as protectors. The officials argued that the legal institutions are impediments to the fight against the threat, and the laws must be bent to counter it effectively. Any resistance from the institutions would be met with LJ’s attempts to mobilize the masses against them.

Lava Jato and judicial activism

Corruption has long been considered systemic and institutionalized in Brazil (Geddes and Neto 1992; Silva 1999; Carson and Prado 2014). Major corruption scandals erupted during Fernando Collor’s presidential term (1990–1992) and Lula da Silva’s first term (2003–2006), the vote-buying scandal known as the Mensalão being especially prominent during the latter. These scandals were overshadowed after 2014, when LJ began to reveal a massive kickback scheme between Petrobras and its contractors (Odebrecht, OAS, and other construction tycoons). The Anti-Corruption Act and other laws approved in 2013, after massive protests with roots in a crisis of political representation, empowered the anticorruption fight.

Judge Moro soon became a kind of cult figure, attracting popular demonstrations in his support. In 2017, he convicted Lula to nine years in prison for a beachfront penthouse in Guarujá that OAS allegedly refurbished by OAS for R$3,7 million as a bribe. Lula was jailed in April 2018 after a regional federal court rejected his appeal and increased his sentence to twelve years. Elected president in 2018, Bolsonaro appointed Moro to be his minister of justice. But in 2019, LJ’s reputation was dealt a huge blow when The Intercept published leaked Telegram messages between the operation’s officials. This Vaza Jato scandal suggested transgressions such as illegal and far-reaching cooperation between Moro and Dallagnol. Five months later, Lula was released from prison after STF ruled the imprisonment to have been unconstitutional, as it came after a court of appeals decision but before all other appeals had been exhausted.

After stepping down from his ministerial post in 2020, Moro experienced a setback when, in March 2021, a Supreme Court minister annulled Lula’s conviction, considering that the first-degree federal court in Paraná, where Moro worked, had no legal standing to judge the cases, and ordered a federal district judge to review the sentences. In that year, in the Prosecutor’s Office, the LJ task force was replaced by a special anti–organized crime group.

The ties among Petrobras officials, politicians, and executives of construction conglomerates gave LJ investigations and cases an unprecedented scale. Prosecutors and judges used a fourfold strategy (Silva 2020): division of the operation into a number of phases, each of which accompanied by a new set of search, seizure, and arrest orders; extensive publicity for case findings (e.g., press conferences); extensive use of plea deals to uncover new facts, even unrelated to the original Petrobras corruption scheme; and broad cooperation between domestic and foreign institutions, including agreements to locate assets and bank accounts funded by the Petrobras scheme. Kerche and Marona (2022, 133) stated that “corruption has become the most urgent and serious problem in the country, putting health, education, and security to the background, and served to criminalize politics, if not definitively, enough to fuel the crisis in which Brazil has been involved ever since” (our translation).

The operation has attracted praise and criticism alike in the academic literature. Melo (2016, 60) emphasized the number of arrests, convictions, indictments, search warrants, plea deals, and frozen assets secured by the operation. He also praised the “impressive efficiency and effectiveness” of Brazil’s judicial and law-enforcement institutions and celebrated the fact that levels of trust in these institutions had increased, highlighting that LJ’s leaders had become “national heroes” (62–63). “The independence and effectiveness of the Brazilian web of accountability institutions,” he continued, “are unparalleled in new democracies and probably even in a host of older democracies” (63). More moderate than Melo, Praça (2017) also asserted that LJ played an important role in reducing corruption.

More critical literature emerged in the later years of the operation. While Avritzer (2018) conceded that, in its early stages (2014), LJ made anticorruption advances within the doctrine of checks and balances, he argues that the operation began to manifest relative politicization after arresting contractors), and “complete politicization” after March 2016, with decisions relating to politics rather than the Petrobras system. Kerche, Marona, and Feres (2018) link LJ to Hirschl’s (2008) concept of the judicialization of megapolitics, focusing on the illegal wiretapping by Moro of a 2016 phone call between Dilma and Lula during the buildup to the former’s impeachment, in which the former offered the latter a minister’s post to shield him from investigation, as well as Lula’s arrest by LJ mere months before the 2018 presidential elections, preventing him from running. ¹ Likewise, Rodrigues (2020) noticed a strong sense of voluntarism in LJ, as the judiciary promoted a selective expansion of the prosecution capacity in Curitiba and, at a lower level, in Rio de Janeiro. ² For Rodrigues, the task force acted strategically to contribute to the series of events there and, more importantly, to their timing, again especially pertaining to Lula’s conviction just in time for the 2018 election. Based on this thesis, Rodrigues (2020) highlights two controversial issues: the partiality of the judge, which became especially clear after Moro’s adhesion to Bolsonaro’s cabinet, and the interaction between the selectiveness (targets of the probes) and the judicialization of politics through criminal persecution. Similarly, Kerche and Marona (2022) point to two events in 2019 that undermined the LJ’s reputation: Moro’s entry into the Bolsonaro administration and the aforementioned Vaza Jato scandal. Noteworthy is that all three works focus their criticism primarily on the later stages of the operation. Corruption has been linked to anti-PT sentiments but is not the only factor explaining the latter, which have been detected especially among Pentecostals (Araújo 2022).

Much of the critical literature on Lava Jato engages with debates on the relationship between the judiciary and politics and has invoked such concepts as the judicialization of politics, the politicization of justice, and judicial activism. These concepts are related but certainly not interchangeable. Ferejohn (2002, 41) defines the judicialization of politics as a post–World War II trend whereby courts have imposed limits on the power of legislative institutions, courts have become loci of policymaking, and judges have demonstrated increasing willingness to regulate political activity. This contrasts with the role of the judiciary according to “classical democratic theory,” which is the application of law, a “mostly technical matter of finding the right principles under which the dispute [about the application of the law] can be settled” (Ferejohn 2002, 49–50). In the Brazilian context, the judicialization of politics has been reflected in the criminalization of a long-standing form of creative accounting by the executive branch (pedaladas fiscais), then considered a sufficient basis to impeach president Dilma Rousseff in an effort of political elites to stop investigations of LJ against politicians (Limongi 2023, 167). Another example is the criminalization of caixa dois (illicit campaign financing) by the Ten Measures against Corruption (Dez Medidas contra a Corrupção), a legislative proposal created and popularized by Dallagnol, also in 2016 (Ramalho 2016). However, aside from legal officials, political actors played a crucial role in both cases, and so these are cases of “judicialization from within” (Vallinder 1994, 93).

The inevitable result of the judicialization of politics, according to Ferejohn (2002, 63–64), is the politicization of the judiciary itself, because “when courts can make politically consequential decisions … those interested in judicial decisions have reason to seek influence and … control appointments to the courts and legal institutions…. As a result, judicial decision-making tends to become politics carried on by other means.”

A concept displaying substantial overlap with the politicization of the judiciary is judicial activism. The difference seems to lie in the fact that, while the politicization of the judiciary is a macro trend, judicial activism is specific action carried out by courts and judges. As such, judicial activism can be seen as a symptom of the politicization of the judiciary. Kmiec (2004, 1444) identifies five meanings of judicial activism that have been used in legal and academic circles: “(1) invalidation of the arguably constitutional actions of other branches, (2) failure to adhere to precedent, (3) judicial ‘legislation,’ (4) departures from accepted interpretive methodology, and (5) result-oriented judging.” Kmiec sees the fifth meaning, result-oriented judging, as “particularly useful … because the scienter element [meaning that the act has been done knowingly] limits the universe of “activist” decisions more than any of the preceding definitions” (1476). Specifically, he defines result-oriented judging in terms of two elements: “(a) the judge has an ulterior motive for making the ruling; and (b) the decision departs from some “baseline” of correctness” (1476). The idea of result-oriented judging meshes particularly well with Rodrigues’s (2020) argument about Lava Jato’s timing of Lula’s arrest to preclude him from running in the 2018 election.

Data and methods

While frame analysis was employed facing many discourses, it has not been applied to LJ. For frame analysts (e.g., Gamson and Modigliani 1989; Benford and Snow 2000; Ferree et al. 2002), frames guide interpretations and worldviews, as they encode expectations about social relations and the effects of actions. Frames are a way of linking culture and behavior: rather than causing the behavior, which would imply a cause-effect relation, frames condition behavior (Small, Harding, and Lamont 2010)—for example, corruption frames are supposed to guide anticorruption strategies.

Entman (2006) saw framing as a useful research paradigm to study mass communication. Frames define problems, diagnose causes, make moral judgments about causal agents and their effects, and suggest remedies. Following Entman’s paradigm, we adopt the specific framework of Lepianka (2015), who adds framing devices such as metaphors, visual imagery, and catchphrases (Gamson and Lasch 1980) to Entman’s four dimensions.

We selected as our corpus all digital written press (newspapers, magazines) appearances (interviews or articles) of the three most important and visible LJ officials: Moro (until becoming minister in 2019), Dallagnol, and Janot. Aside from press appearances, totaling thirty texts, we incidentally cite an essay (Moro 2018) and their books (Dallagnol 2017; Janot 2019; Moro 2021), even if Janot and Moro wrote theirs after having left their positions—those quotations are mentioned cautiously as they do not constitute official rhetoric. ³ As our goal was to systematize corruption frames that emerge from the discourse of the LJ leaders, we limited our corpus to opinion texts and individual interviews. ⁴

From our readings, we inductively built a coding family with eighteen categories (see Appendix A) and 164 codes in Atlas.ti. After completing the coding family, we went back and recoded the entire corpus. We had become intimately familiar with the corpus, and with the help of the Cloud View function, we identified four frames (inequality, hidden pact, backwardness, and chronic disease), each with a problem definition, a diagnosis and solution, a role assigned to corrupt agents, and related framing devices. ⁵

In the following section, we discuss the four frames—inequality, hidden pact, backwardness, and chronic disease—that we identified using a methodological approach inspired by Entman (2006), Lepianka (2015), and Grangeia (2016).

Corruption frames in LJ discourse

This article focuses on the discourse of the three main LJ protagonists to identify and discuss corruption frames (Table 1). Starting with the inequality frame, the problem definition highlights the inequality in this crime committed by the powerful and victimizing ordinary citizens. The hidden pact frame sees corruption as the result of two or more actors from different economic spheres (public and private sectors) colluding illegally for mutual benefit. Both share individual or agent-based explanations, while the others deal with structural explanations. The backwardness frame conceives corruption as a stubborn remnant from the past; the chronic disease frame treats corruption as an endemic and damaging condition affecting society.

Table 1. Corruption frames

Source: Moro’s, Dallagnol’s, and Janot’s discourse (2014–2021).

Corruption as inequality

Problem definition

The inequality frame focuses on the inequality inherent in corruption, the sense that it is perpetrated by the powerful against the powerless. Corruption is described as a crime committed by and benefiting the “powerful,” “rich,” “elite,” or “ruling class.” It is said to disadvantage “citizens,” “society,” “people,” and “workers,” in other words, ordinary citizens. Two ways corruption harms the interests of society are highlighted. First, corruption is described as involving the siphoning of money destined for public services (e.g., health care, education) into criminals’ pockets, thereby compromising the well-being of society. Second, corruption is said to distort political representation as corrupt politicians serve their interests or the interests of their donors rather than those of the people who elected them. The actors frequently emphasize that corruption compromises democracy itself.

This perception of corruption as a crime committed by the privileged is illustrated particularly well by Dallagnol’s (2020) opinion article comparing the perceived negative reactions to anticorruption efforts with the reactions to the US civil rights movement. The then prosecutor quoted Dr. Martin Luther King Jr. to approach the perception of corruption as inequality: “Privileged classes never give up their privileges without strong resistance.” ⁶ Different frames may occur together, and by comparing the battle against corruption with the civil rights movement, Dallagnol also emphasized anticorruption as a progressive cause, an idea that also fits in the backwardness frame.

Diagnosis and solution

The main cause of corruption identified here is the inequality inherent in the justice system. The actors often describe Brazil’s justice system as generous in its dealings with the powerful perpetrators of corruption, which stands in stark contrast to its harsh treatment of ordinary people. With its manifold provisions protecting the defendant, such as extensive nondisclosure, the statute of limitations, and foro privilegiado (special jurisdiction in which superior courts have original jurisdiction over criminal charges against high federal official authorities), its inefficiency resulting from its ‘labyrinth’ of courts, slow functioning,’ ‘never ending’ processes and lenient punishments—instead of prison sentences, fines are applied, which come down to ‘mere taxation’—it is described as an ‘impunity machine’ that lets the powerful get away with corruption, rendering corruption worthwhile and even lucrative. ⁷ A frame of inequality among criminal acts is found in Moro’s (2021, 284) book: “The great truth is that the ‘Brazilian’s presumption of innocence’ is just an interpretative construction that aims to guarantee impunity for crimes committed by the ruling class, another reflection of our brutal social inequality…. The impunity of great corruption generated by this system is at the root of our patrimonial tradition and extractive culture” (our translation).

The main solution related to this frame is reforming the justice system to make it more effective and more egalitarian because it applies the same standards to powerful people and white-collar crime as it does to ordinary citizens and crime. An anticorruption initiative often mentioned by the actors, especially in 2015–2016, is the Ten Measures against Corruption, a set of bills precisely addressing the mentioned flaws in the justice system.

The actors emphasize the importance of spreading awareness among and mobilizing society in support of the anticorruption fight. Precisely the victims of corruption, the actors suggest, possess the power to bring an end to it, as they are numerous and, most importantly, can vote. As they argue, through petitions and street demonstrations, the people do and should express support for LJ and other anticorruption initiatives (e.g., Ten Measures). Aside from exerting pressure through these channels, people are expected to vote for honest politicians. In the long run, actors envision a solution involving “participatory democracy,” in which citizens create “popular initiative” legislative proposals. Hence, a solution suggested for the diminished quality of democracy caused by corruption is an expansion of the actors’ concept of democracy.

The role of the corrupt

The corrupt, as mentioned earlier, tend to be equated collectively with the powerful, the rich, and the elite. It is suggested that the corrupt practice corruption with a certain degree of carelessness because they know the justice system’s loopholes and know that they get away with it. They make cold cost-benefit calculations about the revenue of corruption and its possible costs or consequences. They, in a sense, abuse their power to enrich themselves at the expense of society. Furthermore, the corrupt are perceived as capable of anything within their powers to obstruct and sabotage anticorruption efforts.

Dallagnol forcefully mentions this idea after private messages between LJ officials were leaked in 2019. In two pieces around that time—one article with two colleagues and one interview—he stresses his suspicion that some of LJ’s defendants may be behind the hacking. ⁸ These powerful people, he argues, are “inconvenienced” by LJ, therefore using their far-reaching influence to “annul verdicts,” “block investigations,” “create conflicts between institutions,” “destroy reputations,” and “weaken LJ.” ⁹ Also, he emphasizes several times that he and other officials pay a “high personal price for confronting the rich and powerful.”

Framing devices

A typical device is the metaphor of corruption such as theft of national riches from state coffers or the robbery of the people or Brazil. This family of metaphors serves to emphasize that the victim of corruption is the society at large, robbed or assaulted by the powerful. Another striking omnipresent framing device is the use of the first-person plural in the anticorruption fight. One example: “We need to understand that, together as a society, we have the power to change our country.” ¹⁰ This device expresses a key idea of the inequality frame, that, while the justice system may favor the powerful, the common people collectively have the power to bring about change. Furthermore, by using the first-person plural, the official positions himself on the side of society and signals that he is part of it.

A closely related device is a personification of society as a monolithic being (one voice, one will). The third person singular is current. Among the three actors, an avid user of this device is Rodrigo Janot (prosecutor-general in 2013–2017), as illustrated by the following three statements. First, “the enthusiasm of the entire Brazilian citizenry with the perspective of the improvement of our legislation.” ¹¹ Second, “the country is tired of the temptation, the hypocrisy, the economic chicken flights sustained by special favor.” ¹² Third, “Lava Jato is not a matter of the Federal Prosecutor’s Office, it is a matter of Brazilian society.” ¹³ And finally, “everyone wants an efficient and modern abuse of authority law” ¹⁴ . Finally, Dallagnol framed anticorruption as a fight about who owns Brazil itself: “Brazil is not the private property of the corrupt, this land is ours” or “Brazil is not a no man’s land.” ¹⁵ So, within this frame, the anticorruption fight is portrayed as a battle of the people, represented as a force of the good, against the elite, represented as corrupt and rapacious.

Corruption as a hidden pact

Problem definition

In this frame, corruption is defined as a convergence of interests of at least two people or institutions to expand political or economic power. In brief, any act of corruption involves two or more agents bound by an obscure agreement assuring abuse in the public sector in exchange for a private benefit. In his memoir, Janot (2019, 233) provocatively illustrates this idea:

Throughout our history, the two sides [politicians and businesspeople] have negotiated on equal terms. The result has always been that both sides have gained advantages to the detriment of collective interests. In other words, they systematically broke the elementary, albeit undeclared, pact of every society: to live in a cohesive and solidary way. Now, if I was asked where this addictive way of business and doing politics comes from, I would say one possible explanation is the existence of ‘closed systems.’ Those are the rules for preserving the perpetuation of the same groups in charge of public administration and markets. Whoever is inside does not come out. Whoever is outside does not get in. (Our translation)

Diagnosis and solution

As corruption is attributed to an agreement of silence, it can be uncovered only if one breaks the agreement and reveals the scheme legitimized as a transaction of mutual interest. Then, the plea deal stands as an agreement in which the suspect agrees with the Prosecution Service to confess to crimes and identify collaborators in exchange for a reduced sentence. The Curitiba LJ nucleus promoted 209 plea deals, while there were 190 promoted in Rio de Janeiro and São Paulo (Kerche and Marona 2022).

In a 2015 interview, Dallagnol stressed that the pact of corruption can be uncovered only by plea deals. ¹⁶ As such, these were crucial for the success of LJ: “Unfortunately, it is very difficult to discover and prove corruption in Brazil—hence the importance of the plea deal as the kick-off of an investigation…. From [the deals], other pieces of evidence are gathered, which, if they are good enough, will allow for criminal prosecution.” Dallagnol even said that “the engine of LJ is the plea deals, which are only made by the defendant when he believes he will be punished” ¹⁷ . He illustrated the importance of the plea deal by arguing that it allowed for the recovery of R$6 billion from over a hundred defendants, instead of the mere R$26 million that would have been recovered from the first two defendants without a plea deal. All three actors shared the pro-plea rhetoric. Janot considered it hardly possible to uncover the corruption of criminal organizations without plea deals. ¹⁸ According to Moro (2018, 164), “crimes like corruption are committed to secret and usually only the criminals themselves are witness to their wrongdoing. Therefore, it is sometimes necessary to make a deal with a criminal to get evidence to build a case around more central players.”

At the root of those arguments is the logic of “the ends justify the means”: the plea bargain, a means that could be perceived as ethically dubious, is justified by its results in terms of the amount of money recovered and the relevance of the accused suspect (central players). Criticisms against plea deals were attributed by Moro (2021, 95) to the “breach of the great pact of silence that has always prevailed under these crimes in Brazil.”

Another LJ tool against corruption pacts was a judicial authorization of preventive arrests, justified on the grounds of its helping to prevent obstruction of justice on the part of suspects (e.g., by destroying evidence). However, some lawyers argued that it was used to persuade participants of a corruption scheme to denounce their accomplices. ¹⁹ When asked if there was an excess of prison orders without trial, Moro answered there had been only ten such cases, which he did not consider too many. ²⁰ Without reflecting on the ethics of preventive arrests, Moro dodged the issue by arguing that there had been few instances of the practice. There were 132 preventive arrest orders in Curitiba and 259 in Rio de Janeiro (Kerche and Marona 2022).

The role of the corrupt

Agents were portrayed as expecting to obtain mutual benefit from the accumulation of wealth and/or power, which is in turn construed as the root of their collusion in criminal practices, such as money laundering. For Janot, corruption “is not the fault of the public prosecutor, but of bad administrators who associate themselves with bad businessmen in odious performances assembled to continually plunder the national riches.” ²¹

Framing devices

We noticed a metaphor of corrupt pacts as a kind of symbiosis (interaction between two species in close physical association to the advantage of both), and this image is reinforced by its qualification as a “deleterious” relationship. More specifically, Janot described corruption as a “degenerate” way of doing politics and a “deleterious symbiosis that has been established between certain sectors of business and public agents for spurious gains and perpetuation in power.” ²² This frame emphasizes on the plea bargain as the preferred instrument for solving corruption cases, as already mentioned. For Moro, it is “undeniable that, with the collaboration of some of these individuals, there was a certain leap in the magnitude of the investigation.” ²³ Recurrent emphasis on the strategic use of the plea deal attests to how LJ authorities portray corruption as a hidden pact that deserves to be revealed for the benefit of the nation.

Corruption as backwardness

Problem definition

Corruption is described as a long-standing practice associated with the past and, more implicitly, with a backward society. An illustration of the temporal element of this frame was read in the following fragment from an essay by Moro (2018, 166–167) in Daedalus: “Hopefully, it will be possible to look back some years from now and say that LJ made the national economy, the rule of law, and democracy stronger in Brazil. Maybe it will be possible to say systemic corruption was overcome and that it became a sad memory of Brazil’s past.” Another example came from Dallagnol: “I also recognize the existence of forwarding and backward forces, which are not necessarily at the genesis of these laws and decisions but can influence the environment in which they arise and the course of history.” ²⁴

Diagnosis and solution

The backwardness frame does not specify a cause for corruption but associates it with a tradition or culture of impunity. This tradition or culture is associated with the past, and society can escape it only by evolving and reaching modernity, which is associated with a lack of corruption. This reasoning is exemplified by Moro when explaining that the United States was a “very corrupt country” around the beginning of the twentieth century but managed to drastically cut down on corruption in part through reforms implemented by Theodore Roosevelt. The moral of the story is that “no country is condemned to systemic corruption,” and that Brazil, like the US, can “evolve” and leave systemic corruption behind. ²⁵

Three solutions are referred to here. First is modernization of obsolete legislation, such as the abuse of the authority law of 1965. Janot called the law “very old” and “bad” and claimed that “anyone in their right mind wants a new law,” which must be “up to date” and “effective.” ²⁶ In this 2016 interview, he added that the new law must be “efficient and modern,” as the current law is “old” and “useless.” Strikingly, at no point, the interviewee explained why the law was insufficient. The very fact of its implementation decades ago was considered to speak for itself.

Second, they suggested that LJ had “ended” the tradition of impunity and frequently defined the operation as “unprecedented” in its scope and methods and linked with “drastic measures.” In a 2017 interview with the New York Times, Moro even said that “LJ represents the end of impunity as a rule in Brazil.” ²⁷ Hence, they suggested that only a rupture as unparalleled and radical as the LJ could break with the perceived tradition of impunity.

Third, when legitimizing anticorruption initiatives such as the Ten Measures and LJ, actors often invoke recommendations of global organizations—the United Nations or Transparency International (TI)—and experiences of developed countries. Dallagnol said TI was founded in Germany, a country that Brazilians associate with a high level of development. ²⁸ In an interview, he said that all Ten Measures were in force in “countries with a consolidated democracy.” Janot once made a similar point: “These instruments [included in the Ten Measures] exist in countries that are models and that we look up to, developed and democratic countries that respect human rights. They are based on international treaties. We are not inventing anything.” ²⁹ This last sentence is especially revealing, as it implies that a Brazilian tool in this field would be bad. Later, he reinforced that when saying that “there is no jabuticaba.” ³⁰ Finally, the actors invoke parallels between LJ and the Italian anticorruption operation of the 1990s Mani Pulite (Clean Hands), especially about both operations’ reliance on plea bargains.

In sum, the solutions to corruption according to them should be unprecedented in Brazil while solidly rooted in more developed countries. This is the basic idea underlying the frame of corruption as backwardness: all it takes to leave the old tradition of corruption behind is to develop as a country, following the example of developed countries.

The role of the corrupt

The corrupt in this frame tend to be described as hopelessly backward, being stuck in an obsolete livelihood or business model. A key quote from Janot illustrating this idea was when he emphasized the necessity to punish those “few who still insist on making their livelihood from the assault on the state coffers.” ³¹ In this frame, the activities of the corrupt are construed not so much as immoral but, above all, as outdated. They form an obstacle to the country’s progress and even “steal the future of the country.” ³²

Framing devices

A typical framing device here is the representation of the anticorruption fight as a historical opportunity, a window of opportunity related to an interplay of factors. These kinds of statements are often followed by assertions of the type that this opportunity can and must be seized on and that it is possible to bring about change if there is a will for it. There is considerable overlap between this sentiment of voluntarism and collectivism of the inequality frame, which often features together in these statements. The same statements using the first-person plural (“together, we …”) often also feature this idea of voluntarism (“can change the country”). These statements bridge the gap between both frames: while their collectivist aspect fits into the inequality frame, their voluntarist aspect fits into the backwardness frame. The metaphors of the window of opportunity and voluntarism reinforce that corruption was a firmly rooted tradition, but a singular opportunity has arisen that allows society to act and assume agency over its destiny. A key quote capturing this idea came from Janot:

We need wisdom in order not to miss this unique opportunity to change things and elevate the ethical standard of the politics developed in our country. This enterprise will only succeed if it can count on a true involvement of active citizenship: the power is of the people, for the people, and by the people. We need, therefore, to assume command of our destinies and promote the changes of tack that are indispensable for us to leave impunity in the oblivion of the past and attain the possible victory against the terrible ulcer of corruption that still blots the honor of our country. ³³

Aside from the major interrelated devices of opportunity and (collective) voluntarism, there was a relatively minor device: a frequently repeated mantra that the LJ is a “point outside the curve,” an outlier of sorts. This metaphor is especially popular among prosecutors (Dallagnol and Janot) and reinforces the perceived singularity of the operation in contrast to the banal and commonplace impunity perceived to characterize the Brazilian reality.

Corruption as a chronic disease

Problem definition

Regarding the disease element, corruption is often quite literally described as an infection or a wound, or more generally as a natural or biological phenomenon. When it comes to the chronic, corruption is portrayed as a systemic or endemic problem. Note that the word endemic is a medical metaphor that normally describes infectious disease. This frame sometimes specifies what corruption causes damage to (democracy, the rule of law, society), but never what that damage consists of. Like a disease, corruption’s maleficence is assumed to speak for itself.

Diagnosis and solution

One of the elements of a frame according to our methodology is the diagnosis of and solution to a given problem. It is natural that medical metaphors were often used concerning corruption. ³⁴ In our corpus, words such as diagnosis and treatment were at times used when the actors referred to solutions to corruption, but a broader, underlying idea can also be discerned: corruption requires, above all, a technical solution. The actors frequently emphasized that their work was “technical,” based on facts and established procedures (just like medical ones). If corruption is indeed a phenomenon ruled by the laws of nature or biology, then the solution must be scientific or technical, based on those laws. Hence, as in the frame of backwardness, corruption is depoliticized, and linear, technocratic solutions are put forward.

As doctors and surgeons, anticorruption fighters use “instruments,” which often refers to as plea deals or other methods and measures. Furthermore, disease-related metaphors are often followed by statements emphasizing the need to change the conditions leading to corruption. As diseases are caused by impersonal conditions such as genetic and environmental factors, so is corruption. Any solution must address these conditions. In the words of Dallagnol (2017, 387):

The way out to Brazil is not LJ. This investigation brings us a sigh of hope and belief in the functioning of our institutions, but this will all pass if we do not take the time to consolidate changes that guarantee justice in this and other cases. The operation acts as a drug on a focus of infection, but it is necessary to treat the source of this and other diseases. Until we act on the conditions that favor corruption in the country, we will continue to suffer from successive scandals. (Our translation)

The role of the corrupt

In this frame, corruption is not usually associated with actors. It is portrayed as simply existing, passively waiting to be cured by specialists. Sometimes, however, corruption was personified as an acting subject itself, such as a “cancer that sucks energy and resources that could lessen other social problems,” a “terrible ulcer that still stains the honor of our country,” “ailments … that defile the rule of law [and] bastardize society.” ³⁵ Corruption as a disease was associated with specific actors only once: “The systemic corruption of political agents and parties makes democracy ill.” ³⁶ In this case, corruption was not exactly construed as a disease but as something that causes illness.

Framing devices

The frame itself is a metaphor. Medical terminology (disease, ailment, infection, wound, cancer, diagnosis, treatment, cure) is used, but so are more general biological terms such as symbiosis and words related to impurity and decay, such as conspurcar (defile, pollute) and abastardar (bastardize). When it comes to the anticorruption fight, words such as instrument and technical frequently appear.

The metaphor of disease is so ingrained in corruption discourse that it is sometimes used contradictorily, as Moro illustrates. When asked if a plea deal or criminal investigation could bring the country to a halt, Moro denied it, stating that “what brings instability is corruption and not the confrontation of corruption. The problem is not in the cure but the disease.” ³⁷ In his 2018 essay, however, he contradicts this equation of corruption with the disease when emphasizing that “systemic corruption is not a natural phenomenon, and no country is destined to live with it” (Moro 2018, 143). Yet he invoked the metaphor of the cure: “Even if discovering and exposing corruption generates new challenges and painful resistance in the short run, these effects are part of the cure” (143). Hence, in explaining that corruption is not a natural phenomenon because it can be overcome, Moro confounded naturality with immutability, forgetting that the word cure construes corruption as a disease, a natural phenomenon but not necessarily an insurmountable one.

Discussion

All four of the identified frames can be linked to at least some of the literature previously discussed. The inequality frame echoes the idea of corruption as a resource transfer from the mass public to the elites, an extra cost to most of the citizens, on which we cited Rothstein and Uslaner (2005) and Warren (2004). The hidden pact frame can be related to Rose-Ackerman’s (1978) principal-agent concept of corruption. The ultimate principal, according to the LJ actors, is clearly “the people,” while politicians and public servants are the agents supposed to act on behalf of the people. Businesspeople, then, are the third actor, attempting to influence agents by offering payments that are not passed on to the principal. These secret, illegal agreements are precisely what we mean by the term hidden pact.

In line with Tänzler et al. (2012), The LJ protagonists also associate corruption with backwardness, a lack of modernization, and a lag on Brazil’s development, and the main proposed solution is to import administrative-technical procedures from countries that have allegedly reached full modernity or that are at a more advanced stage of modernization. Furthermore, and significantly, the actors rarely explain how the specific technical procedures they cite would contribute to reducing corruption. The discourse is rather crude here: when justifying their preference for this or that measure or this or that law, they consider it sufficient to refer to its age or the country or countries where it was developed or where it is in use. Finally, the frame of chronic disease goes back at least to medieval times.

Table 2 summarizes the identified frames, named after their problem definitions, in terms of their three main dimensions. Within each dimension, the frames can be subdivided into two or three categories per dimension. Starting with the causes, a contrast can be discerned between one frame that ascribes the causes of corruption to the deliberate actions of individuals (hidden pact) and the remaining frames, which others ascribe corruption to structural causes (inequality, backwardness, and chronic disease). Following the structure-agency dichotomy discussed earlier, we refer to the former category as “agent-based” and the latter one as “structure-based.”

Table 2. Classification of Frames

Source: Moro’s, Dallagnol’s, and Janot’s discourse (2014–2021).

Moving to the solutions, we see another dichotomy, which we will call “conventional,” relating to the role of the judiciary according to classical democratic theory, versus “activist,” in reference to the concept of judicial activism. Recall that the conventional role of the judiciary is to apply the law in a technical manner. ³⁸ In this sense, the hidden pact, backwardness, and chronic disease frames all propose attacking corruption using techniques that facilitate the application of the law, such as plea bargains and other legal instruments. The inequality frame expresses an ulterior motive beyond the desire to apply the law: to engender direct democracy. This is in line with Kmiec’s definition of judicial activism in terms of result-oriented judging.

Regarding the role of the corrupt, the frames of inequality and hidden pact construe the corrupt (the powerful, public, and private agents) as possessing agency, construing them as deliberately practicing corruption to achieve certain goals (accumulating wealth and power). In contrast, the backwardness and chronic disease frames hardly mention actors. When they do, no agency is ascribed to them. Corruption simply exists as a tradition or some undefined condition, and actors are said to be simply caught up in persistent habits or realities. Put differently, the corrupt are constrained by structure. The inequality frame occupies a special place here: while it identifies a structural cause (the inequality built into the justice system), it does reserve a role for actors: the rich and powerful who exploit the biased justice system for their own benefit.

Conclusion

We have identified four frames of corruption in an exhaustive sample of written texts from the three protagonists of operation Lava Jato. Three of the four frames identify structural causes of corruption, and one identifies individuals with agency as producing corruption. While two frames ascribe agency to the corrupt, two others see the corrupt as constrained by structure. The most relevant contrast, however, is that present in the solutions associated with the frames, between one frame emphasizing judicial activism, more specifically result-oriented judging, and another three proposing technical solutions that are more in line with the conventional ideas about the role of the judiciary in a democracy.

Our contribution is twofold. First, we employ a constructivist method, frame analysis, which is relatively little used in corruption studies, to shed light on how anticorruption actors themselves construe corruption, the struggle against it, and their role in that struggle. Second, we show that the inequality frame, associated with judicial activism, is present throughout the discourse, including during the operation’s initial stages, when it was still relatively uncontroversial in the literature and in society. Even the critical literature focuses its criticisms on transgressions LJ committed in its later stages, which raises the suggestion that the operation was founded on good intentions but got corrupted along the way. Our findings cast doubt on this narrative and demonstrate instead that, at least in the discourse, the intention of judicial activism was present from the beginning.

However, the inequality frame did not occur in a vacuum. The judicial activist discourse associated with that frame was present alongside the frames associated with conventional solutions. The role of the latter seemed to be to mask the former. Concretely, we found that, especially in the early years of the sample, bombastic declarations are made about how “we the people” must act on an opportunity to bring the corrupt elite to justice and eradicate the crime. However, every time LJ was criticized for committing transgressions or taking liberties with the law, the actors would adopt a rhetoric reflecting a commitment to technical application of the law, albeit mixed with a little bit of “the ends justify the means” when it came to the use of plea bargains and preventive arrests. But when more serious transgressions came to light—for instance, after the Vaza Jato—the discourse again adopted a belligerent tone, and rather than reflecting on the misconduct revealed by The Intercept, the actors, and especially Dallagnol, ascribe the hacking of their devices and the leaking of the messages to a conspiracy by powerful defendants in LJ cases. They are after the innocent underdogs again, who, at great personal cost and no personal interest, serve the interests of the nation, or so the narrative goes.

The main implication of this study is that discourse is a crucial aspect of phenomena such as the politicization of the judiciary and judicial activism, which tend to be studied mostly in terms of observable actions pertaining to interinstitutional relations. It would not have been possible to obtain the insight that there was an intention of judicial activism from the beginning of operation Lava Jato if we had focused purely on action, like most existing literature on LJ and on the broader phenomena that the operation is an instance of. Future research should apply discourse analysis in its different forms to other cases of (purported) judicial activism.