Resilient Corrupted Coalitions: How Anti-Corruption Operations Reshaped Coalition-Building in Brazil’s Congress ()
1. Why Is Understanding the Impact of Anti-Corruption
Operations on Coalitional Presidentialism Important?
Coalitional presidentialism emerges when presidents fail to form a majority in Congress and seek political support through durable cross-party alliances (Abranches, 1988). Constitutions transfer certain legislative powers to the Executive in hopes of creating mutual dependence to prevent minority governments from failing due to legislative gridlock (Chaisty, Cheeseman, & Power, 2018). Raile, Pereira, and Power (2011) call these legislative powers the “presidential toolbox”, according to which the president stretches its persuasiveness by using agenda-setting power, budgetary allocation, cabinet management, direct influence over one or more parties, and country-specific informal institutions.
When coalitional presidentialism forms in a context of weak democratic institutions and ongoing widespread corruption, the mentioned tools may take the form of an exchange of favors, involving practices such as bribery, illegal electoral assistance, lucrative office appointments, and illegal benefit for businesses (Chaisty et al., 2018; Mello & Spektor, 2018). In these scenarios, to expand corrupt rents while keeping risks of exposure low, the Legislative and Executive branches may collude to shield illicit practices from judicial oversight and weaken control institutions, namely all bodies responsible for carrying out activities related to the defense of public property, laws, and increasing government transparency (Governo do Brasil, 2021; Mello & Spektor, 2018; Ministério Público Federal, n.d.a).
Sometimes, however, anti-corruption operations breach this “blockade,” confronting long-institutionalized favoritisms and shaking the bonds that hold “corrupt” coalitions together. Previous literature on coalitional presidentialism explores Executive-Legislative institutional dynamics, and how it relates to corruption studies, but glosses over the ways whereby presidential coalition-building powers interact with the challenges imposed by judicial and control institutions (e.g., heightened individual exposure and arising personal and political costs to perpetrators). Interestingly, anti-corruption operations may also feed into non-electoral mechanisms such as protests that inflate reputational costs, changing politicians’ incentives for certain types of misconduct (Balán, 2014).
This gap in literature is problematic for policy-makers, who may miss critical insights into how anti-corruption efforts can reshape power dynamics, reduce corruption, and strengthen governance. This may lead to ineffective or misguided policies that fail to account for the complex dynamics of institutionalized corruption and political stability.
Aiming to start bridging such gap, this article explores (i) how large anti-corruption operations impact informal methods of coalition formation, and, more specifically, (ii) how the Mensalão and Lava Jato impacted informal methods of coalition formation in Brazil. The article analyzes three Brazilian presidencies: Lula (2003-2010), Dilma (2011-2016), and Temer (2016-2018). This period is particularly interesting because, in less than 10 years, Brazil witnessed two large-scale corruption scandals: the Mensalão (2005) and Lava Jato (2014). The former entailed the Presidency making monthly payments to legislators in exchange for their votes, while the latter involved a complex scheme with different types of corrupt practices across social, professional, and political networks (Pereira, Power, & Raile, 2008).
Better comprehending the Brazilian case provides important lessons for other countries that present democratic competitiveness and require presidents to be coalitional formateurs (i.e., coalition builders). These include former Soviet states like Armenia, sub-Saharan countries like Benin and Kenya, South-Asian states such as Indonesia (which has relevant anti-corruption initiatives such as the Corruption Eradication Commission—KPK), and, crucially, Latin American countries such as Chile, Ecuador, and Argentina (Chaisty et al., 2018). Notably, the Lava Jato scandal had significant repercussions in Ecuador, where investigations revealed extensive bribery and corruption by high-ranking officials linked to the Brazilian constructor Odebrecht.
2. Minority Presidents and Coalition-Building under
Neo-Institutionalist Lenses
In 1990, Linz boldly declared that presidentialism was doomed to fail. According to the author, the president holds considerable powers and is directly elected by the people, as are members of Congress, creating a legitimacy dilemma of who’s “will” should prevail. This is worsened by the zero-sum nature of presidential elections (i.e. if one side wins, the other loses), whereby presidents become less attentive to smaller parties’ needs and embody a sense of power and mission. Linz (1990) further argued that the presidents’ fixed terms split the political process into discontinuous periods, and crises erupt in case of midterm termination of power.
Shugart and Carey (1992) indicated that fixed terms mean predictability, midterm terminations are the exception, and double legitimacy allows for checks and balances. Mainwaring (1993) suggested that presidentialism is problematic to democratic longevity only if mixed with multipartyism, for the latter increases the likelihood of political polarization and legislative deadlocks. According to the author, bipartidism is better because radical actors are barred by high-entry thresholds, and the president’s party usually holds a majority.
Chasquetti (2001) empirically demonstrated that multiparty presidentialism results in effective governance if coalitions are made. As legislators need presidents to advance programmatic concerns and presidents need legislative support to govern, the president is stimulated to assume a coordination role and act as a prime minister when forming a stable majority (Chaisty et al., 2018; Junior, Pereira, & Biderman, 2015; Power, 2010). Chasquetti’s (2001) findings on the sustainability of coalitional presidentialism lend theoretical support for institutions to determine and frame political interactions while admitting strategic behavior by political actors when deciding whether to seek coalitions and how to manage them.
The article borrows from Chasquetti’s (2001) findings and adopts a neo-institutionalist perspective, advancing that formal political institutions are composed of a coherent and stable set of norms that constraint political action (March & Olsen, 2008). Indeed, it acknowledges that individuals’ learning capacity and cost-benefit calculations based on personal and structural incentives can change a norms’ system under specific conditions.
This approach better aligns with rational choice neo-institutionalism (RCI), which predicts that individuals act in a utility-maximizing way to yield the highest personal gains when considering available information (Shepsle, 2010). After attributing value to each action and stipulating the involved risks, actors only change their action plan if the costs of a preferred action rise without a correspondent increase in benefits. For instance, new legislation and reframed social norms might change incentives for people to engage in corruption, halt their ability to conceal transgressions, or in cases of endemic corruption, raise rational preference for alternative forms of fraudulency (Carson, 2014).
Still, one should note that the analytical level offered by RCI is not self-contained, and complementary lenses such as historical neo-institutionalism are handy in gauging how the rules of the game that structure social interactions form over time at the macro-level.
Building on the idea that current outcomes are shaped by their historical trajectories (path-dependency) and that events form symbols that persist through time, significant and rapid changes require disruptions to the status quo through events that create uncertainty within the existing institutional framework (critical junctures) (Capoccia, 2015; Collier & Collier, 1991; Liu, Sibley, & Huang, 2014; March & Olsen, 2008; Williams & Baláž, 2008). Critical junctures trigger multi-level processes, enabling actors within a changing society and its institutions to introduce new practices and reshape normative structures, all while adapting to the evolving socio-political landscape.
Brazil has an institutional history characterized by democratic disruptions and political clientelism that ease the understanding of its political situation. The First Brazilian Republic (1889-1930) was installed after a military coup against the Brazilian monarchy (Fausto, 2019). This period was marked by systemic electoral fraud and patron-client relations remnants of monarchical patrimonialism—where the state is a continuation of royal affairs (Fausto, 2019; Weber, 2013). A weak central government and strong states supported by landowners laid ground to a multiparty system tinted by regional interests that, despite authoritarian interruptions (1930-1945 and 1964-1985), influence politics to date (Abranches, 1988).
Along with high levels of institutionalized corruption, the Brazilian experience seems to suggest that individual strategies and political defections play a role in determining the informal methods used by the Presidency. In fact, this article advances that, despite the endurance of coalitional presidentialism vis-à-vis crises, presidents’ ability to maintain a cohesive coalition and protect the political class from court convictions may have determined each Presidency’s fate.
To draw a parallel, Indonesia also offers a compelling case for applying neo-institutionalist approaches. The 1997-1998 financial crisis marked a critical juncture, sparking societal upheaval and protests against corruption, mismanagement, and political repression during President Suharto’s 32-year rule (Reid, 2015). His resignation ushered in the Reformasi period, characterized by constitutional amendments that curtailed Executive power, decentralized governance, expanded civil rights, and established institutions like the KPK to promote good governance (Nardi, 2018).
However, Indonesia’s transition to a multiparty democracy under Reformasi was tightly managed by elite consensus, resulting in a highly oligarchic system (Reid, 2015). Oversized coalitions, money politics, and entrenched corruption continued dominating the national legislature, with traditional elites in the Executive and Legislative branches often colluding to weaken reforms and good governance institutions (Nardi, 2018).
Despite institutional constraints, actors often pursued individual strategies. For instance, Constitutional Court Chief Justice Jimly Asshiddiqie implemented reforms enabling progressive NGOs to circumvent elite-dominated elected branches, and file petitions that shaped government policies (Nardi, 2018). NGOs succeeded by framing debates in legalistic terms and leveraging institutions like the KPK, Constitutional Court, and Human Rights Commission. Similar to Brazil, Indonesia’s coalitional presidentialism is deeply influenced by institutionalized informal coalition-building practices, and judicial and oversight bodies occasionally succeed in disrupting Executive-Legislative collusion. Finally, much like in Lava Jato, civil society has been instrumental in bolstering and legitimizing these institutions.
3. Hypothesis and the Informal Presidential Toolbox
By considering the sudden advent of the Mensalão and Lava Jato operations and the imprisonment of important politicians and businesspeople, breaking with a former condition of impunity, this article hypothesizes that anti-corruption operations affect coalition-building by disrupting favoritism networks and significantly raising the costs of informal methods of coalition-building. The alternative hypothesis is that anti-corruption operations have no impact on informal methods of coalition formation. This can happen because although anti-corruption operations increase the perception that corruption is being fought, no real shifts take place as the political elite blocks intrusive regulations and concedes shallow reforms while protecting the distribution of private goods (Mesquita & Smith, 2017; Rupnik, 2007).
This article operationalizes large anti-corruption operations as nationwide criminal investigations that seek to oppose or inhibit corruption by identifying and apprehending perpetrators, collecting evidence to support their conviction in court, and determining if a crime has been committed (Reilly, 2019). Investigations are usually conducted by controlling bodies, which in Brazil amount to the Polícia Federal-PF (Federal Police), Ministério Público Federal-MPF (Federal Public Prosecutor’s Office), Receita Federal (Federal Revenue Service), Controladoria-Geral da União-CGU (Federal Comptroller General), and the Tribunal de Contas da União-TCU (Brazilian Audit Court).
Additionally, the 1988 Brazilian Constitution provides for the creation of inquiry commissions by the Federal Senate and the Chamber of Deputies (hereinafter: the Senate and the Chamber, respectively) (Siqueira, n.d.). Such commissions are granted with some judicial powers, such as breaching the bank, tax, and telephone confidentiality of those investigated and should forward their final reports to the MPF.
Informal methods of coalition formation are operationalized following the categories delineated by Chaisty et al. (2018): (i) electoral assistance through financial and administrative resources, (ii) favorable business treatment, (iii) direct financial inducements, and (iv) lucrative or high-status appointments (pp. 211, 212). To grasp these methods’ illicit facets and adequately address the questions raised by this article, they are further specified: (i) electoral favors whereby funds are corruptly sourced, (ii) business perks, fraudulent contracts, and permits, (iii) personal bribes, and (iv) cabinet nomination accompanied by the raising of illicit funds.
Notably, the Presidency englobes the direct advisory bodies of the President, such as the Cabinet Office and the General Secretariat (Lima-Silva & Loureiro, 2018; Palermo, 2000). This implies that no normative claims regarding the presidents themselves are made. Similarly, by citing accusations, investigations, and indictments, this article reproduces facts of public knowledge without attributing guilt: every person accused of any crime is considered innocent until proven guilty by the competent judicial body.
4. What Can Be Considered Corruption?
The most popular definition of corruption for decades has been Rose-Ackerman’s (1999) “misuse of public power for private gain” (p. 91) for its operationality and generality, suiting governmental institutions well. Nevertheless, misuse is generally contrasted to law abidance, neglecting legislative capture by particularistic interests (Kurer, 2005; Olander, 2021). Moreover, the clear-cut distinction between “public” and “private” might blur when local-level interactions with officers who constantly transition between formality and informality unfold as well as because understandings of corruption are constructed through narratives (Gupta, 1995).
Klitgaard (1988) describes corruption as a process in which a patron entrusts responsibility to an agent, who betrays the patron for its own advantage in a scenario where the agent has monopoly over decision-making, discretion over implementation, and low accountability to the patron. This relational approach disregards the corruptibility of the principal, but broadens the conceptualization beyond the public sector, and enables Transparency International’s (TI) “misuse of entrusted power for private benefit” definition (Pozsgai-Alvarez, 2020: p. 435). Still, the definition refers to “private,” which inadvertently infers the public-private dichotomy, and might exclude morally dubious behavior.
A significant body of literature adds that “impartiality” norms should be emphasized for they exclude any particularisms, sustain the process by which dyadic actors cooperate and confer power to a TDR (Kurer, 2005; Rothstein, 2011; Sweet, 2000). Philp (2008) suggests that, while corruption comprises the office holder, the original beneficiary, and the actual beneficiary, corrupt behavior subverts office norms and regulations, jeopardizes the original beneficiaries, and favors sectional interests (pp. 311, 315). By replacing Philp’s (2008) “public official” with Friedrich’s (2002) “power-holder,” one might derive that corruption occurs whenever a power-holder subverts non-discrimination norms in favor of sectional interests and to the disadvantage of the original beneficiaries of that position.
The latter conceptualization allows for the contextualization of corruption according to locally defined norms, but presupposes a consensual agreement on society’s moral expectations. Rather than examining corruption through an emic lens in which social exchanges are studied apart from morality as suggested by Torsello and Venard (2015), this proposal advances that behavioral deviations from core societal values require both moral reflection and, to a certain extent, defection, making an appreciation of normative expectations more valuable for the current debate.
More generally, a definition that reaches the civil society denotes corruption as the perversion of something in a primitive state of purity, implying institutional or individual decay (Walton, 2015). This approach risks conceptual overstretching by containing virtually all ethically dubious acts, and ignores autonomy issues (e.g., physical addiction or dependency on others). Yet, it underscores a crucial element of decay when compared to the misuse of public office: Decay is often seen as unacceptable and harmful, followed by abuse of entrusted power, and, only then, abuse of public office. This links back to the notion that in post-colonial contexts, corruption and discourses of complaint enable the channeling of grievances concerning sundry forms of mismanagement, yielding deep moral and identity-related connotations (Orjuela, 2014). Anchoring corruption in terms of communal norms enables a nuanced understanding of collective action issues and effective grounds for moral disengagement of qualified defectors.
Although a conceptual agreement is unlikely to emerge anytime soon, this article adjusts TI’s definition and defines corruption as the systematic subversion of non-discrimination office norms in favor of sectional interests. This definition excludes persons in positions of power, yet it suggests the misuse of trust because it is the commitment to impartiality that guarantees reciprocity and retaliation in social contract theory. In a similar vein, “non-discrimination norms of office” contemplate the principles of equality and reciprocity, following macro-level normative structures that underpin institutional organization (Kurer, 2005; Sweet, 2000). Moreover, it is held that corruption is systematic to differentiate corruption from theft and other self-serving misdemeanors (Kurer, 2005).
Beware that when such benefits refer to the assignment of public offices by politicians or parties to loyal supporters, it is called patronage, while it is political clientelism when there is an iterated and contingent principal-agent distribution of any excludable benefits for political support (Berenschot & Aspinall, 2018; Hicken, 2011; Piattoni, 2001; Remmer, 2007; Scott, 1972, 1977; Stokes, 2007, 2011).
5. What Drives Corruption? The Million Dollar Question!
Debates on the fundamental drivers of corruption are often as spirited as those about its conceptualization since causes and effects frequently become cyclical. The purpose of this section is simply to outline key individual, social, and structural motivations that might encourage or perpetrate corrupt behavior.
At the micro-level, Bauhr and Nasiritousi (2011) present greed and need as key driving forces. Need refers to people being coerced or extorted into paying bribes to access critical services, whereas greed refers to collaboration to gain unfair advantages over others. By assuming that people make hedonistic calculations, a rational choice approach similarly argues that rational individuals have self-serving motivations to engage in corruption. Indeed, when the benefits of an action outweigh the costs (say, prison or the likelihood of being caught) and there is information asymmetry, people will engage in morally hazardous behavior (Eskeland & Thiele, 1999). While the rational choice approach glosses over moral reflection and cognitive biases, both approaches do not elaborate on instances of engagement due to action-seeking, belonging-seeking, socially frustrated, or even morally oblivious actors (Bjørgo, 2016).
Furthermore, actors do not operate in a vacuum, and their broader organizational context interacts with individual motives through prevalent forms of moral reasoning. Gorsira et al. (2018) suggest that organizational climates can be egoistic (decisions are mostly based on self-interest), benevolent (decisions mostly consider others’ well-being), and principled (decisions are primarily based on external codes), with the latter two encouraging ethical behavior. Likewise, Maesschalck (2005) examines how feelings of group belonging (i.e., the greater the incorporation into the group, the more subject to group decisions), and exposure to constraints (i.e., the more externally imposed prescription, the less room for individual negotiation) affect individual inclinations and incentives to behave (un)ethically.
At the structural level, Huntington (2002) contends that fast modernization and the expansion of government authority led to new processes of capital accumulation and dispossession, resulting in the creation of new “gray” zones that are ungoverned by or conflict with accepted norms. For example, where communal solidarity values are high and public service values are ambiguous or relatively low, people might feel more compelled to clientelist practices. This arises because of uncertainty rather than an intrinsic cultural predisposition for corruption.
According to Heidenheimer (2002), “black corruption” is widely condemned, and legally penalized, “white corruption” is minimal and ordinary, and “gray corruption” involves behaviors that are ethically dubious but not explicitly prohibited. Similarly, Pozsgai-Alvarez (2020) categorizes corruption as primary (breach of basic social norms of cooperation and the moral frameworks of society), secondary (breach of anti-corruption norms formed by evolving political and professional standards), and tertiary (individuals compelled to violate social norms).
Moreover, Rothstein (2011) argues that corruption occurs when, due to historical processes, people start fearing lack of reciprocity (thence being guided by (lack of) trust and not selfish and utility-maximizing principles), generating a process that resembles economists’ corrupt equilibrium. Societies with weak institutions and significant inequality may fall into a low trust-corruption-inequality trap, in which standard operating mechanisms (SOMs) are maintained to access key services, yet these services cannot be adequately provided owing to the fear of defection promoted by SOMs (Rothstein, 2011). For example, people do not pay taxes because they do not trust the government to manage resources properly, governments cannot deliver universal programs due to a lack of resources, and some government officials become rent-seekers to gain resources for themselves or their constituents.
Underlying grievance, in turn, are another driver of corruption, referring to historical cleavages or urgent needs. Orjuela (2014) indicates that corruption can both be used as a resistance or domination tool. Minorities who are excluded from the political system and fear losing social capital if they choose to express themselves in other ways may resort to corruption to obtain protection or have their grievances handled. Yet, petty corruption further reinforces the state’s partiality and continue to fragilize minorities by depleting their resources and sustaining their marginalization.
As revisiting the sociology of corruption requires a deeper reflection and this article’s purview is to offer a general institutional account of political corruption in contexts of coalitional presidentialism, the rational account approach suffices to the analysis. Still, one may classify the Brazilian Congress as pertaining to an egoistic organizational climate, while also recognizing Huntington’s gray zones in the formation of Brazilian political institutions.
6. Methodology
Well-defined theories of coalitional presidentialism and strategic behavior are taken to explore supposed causal mechanisms of large anti-corruption operations in shifting informal methods of coalition building. This requires in-depth knowledge that is best achieved with single-unit studies, which offer empirical observations of how variables interrelate, inductively contributing to general wisdom (Gerring, 2004, 2008). A typical single case is preferred as it yields wide representativeness and is well explained by existing theoretical models (Seawright & Gerring, 2008). Within the typical case, this article judges the evidence according to a specific theoretic pathway, exemplified here as the change of incentives for corruption, following a pattern-matching investigation.
The screening of Brazil as a typical case is theory-based. First, Brazil is a pure presidential State where the president is the main coalition-builder, also called formateur (Chaisty et al., 2018). Second, Brazil has constantly been ranked by the Freedom House (2020) as a democracy, meaning that Executive-Legislative dialogue is crucial for governance (Chaisty et al., 2018). Third, the institutionalization of a strong executive and crescent parliamentary fragmentation followed a period of democratic weakness (a negotiated transition from a broadly corrupt military regime to democracy), perpetuating informal practices in its public administration (Assis, 1983; Mello & Spektor, 2018).
Fourth, since 1990—the year symbolically ending Huntington’s (1991) third wave of democratization—the country had two nationwide politically relevant corruption scandals, the Mensalão (2005) and Lava Jato (2014) (Lagunes & Svejnar, 2020). Finally, as the Brazilian coalitional presidentialism remains stable after political and economic crises, it is a solid case of a growing form of presidential democracy, representing particularly well the Latin American context (Pereira & Melo, 2012). Therefore, it is important to study Brazil for its regional representativeness and to expand insights into the potential role of the judiciary and control institutions in rendering highly corrupt regimes more accountable in other coalitional presidentialism regimes.
This article employs theory-testing process-tracing to examine sequential events, unpack their causal mechanisms, and arrive at a midrange theory bound within the specific context of corrupt coalitional presidentialism (Beach & Pedersen, 2013; Falleti & Lynch, 2009). By applying process-tracing, one is expected to disaggregate a temporal sequence of key events, analyze consecutive changes and possible intervening causal processes, and relate such changes to an explanatory narrative (Beach, 2016; Collier, 2011; Gerring, 2004). In this manner, a theoretical framework and a hypothesized causal mechanism are deduced from literature, and then empirical evidence is used to determine whether the hypothesized causal mechanism was present and functioned as theorized (Beach & Pedersen, 2013).
Here, process-tracing might explain the impact of large anti-corruption operations (X—theorized cause) on informal methods of coalition formation (Y—predicted effect) by dissecting and analyzing the set of phenomena and supposed causal mechanisms that link X to Y (Beach, 2016).
The analyzed period spans from 2003 to 2021. To ensure data credibility, source triangulation was conducted using official government and law enforcement documents, reports from leading Brazilian newspapers, and academic articles from reputable databases (Polit & Beck, 2012). Media bias was mitigated by selecting sources from both neutral to center-right (e.g., Estadão and Veja) and neutral to center-left (e.g., Folha de São Paulo and Carta Capital) outlets. Meanwhile, the Globo (or G1) network usually presents mixed editorial stance, and IstoÉ is known for its critical reporting on corruption scandals regardless of political affiliation.
The material was found through purposive sampling and exponential discriminative snowballing. The former allows for the identification of available and information-rich sources from experts (Patton, 2015). The latter is the selection of citations used in other articles, after scrutiny of the researcher to find relevant information (Oregon State University, 2010).
The study’s main limitation revolves around the nature of informal mechanisms: they are not registered, their ethics are subjective, and when perceived as corrupt they are secretive. To alleviate these issues, findings are inferred and tested indirectly while contemplating standards of plausibility. Indirect evidence includes corruption allegations in tandem with shifts in cabinet representation, ideological alignment with the governing party, concessions on policy preferences, and transfers of particularistic resources (Pereira et al., 2008).
7. Lula’s First Presidential Term (January 2003-December
2006)
In 2003, President Lula’s Working Party (PT) was ideologically split upon assuming power. The President and his Chief of Staff, José Dirceu, were part of the dominant political right within the party, predominantly vote-seeking (Pereira et al., 2008). Nevertheless, about one-third of the PT’s deputies represented the radical left factions, predominantly policy-seeking, meaning that social welfare programs would be on the agenda regardless of President Lula’s electoral promises of sustaining austerity policies (Pereira et al., 2008; Sant’Ana, 2020). To free up resources for welfare, the Presidency advanced the tax and pension system reforms, contradicting the interests of the PT’s leftist faction, and reinforcing the need for a large coalition (Pereira et al., 2008).
Although presidents prefer making winning coalitions with a minimal number of parties to retain a tighter agenda control, formateurs expand their coalitions when a substantial number of votes is necessary to approve an unpopular political agenda (Carrubba & Volden, 2000). Moreover, as a traditionally combative party, the PT comes from a historically hegemonistic and suspicious political culture (Sant’Ana, 2020). Taken together, these factors influenced President Lula’s decision to co-opt a set of smaller political parties rather than costly traditional ones, avoiding altogether the strong influence of the Brazilian Democratic Movement Party (PMDB) in the government (Sant’Ana, 2020). Notably, the PMDB was the third-largest party at the Chamber (14.42%) and the largest at the Senate (23.46%), while the PT was the largest party at the Chamber (17.74%) and the third largest at the Senate (17.28%) (Tribunal Superior Eleitoral, 2003).
This highly fragmented scenario and President Lula’s coalitional choices led to a large and ideologically heterogeneous coalition, making it difficult to conceal divergent interests (Martin & Vanberg, 2014). Seeking to promote coalitional discipline, the Presidency turned to cabinet management and expanded the number of ministries and secretaries with ministerial status (Barbosa & Pompeu, 2017). This distribution yielded 24 out of the 35 ministries and special secretaries to the PT, which also placed partisans in lower offices to guarantee the coalition’s discipline through a process called politicization (Lima-Silva & Loureiro, 2018; Silva & Barbosa, 2019).
The overrepresentation of the PT is explained by Baron and Ferejohn (1989), who see it as a benefit of the formateur’s party. More unusual, however, is that the other parties were not represented proportionally to their legislative weight, something called low coalescence. Coalescence is measured from 0 to 1, being that a higher number equals more representation and satisfaction among the coalition (Bertholini & Pereira, 2017). President Lula started the government with a 0.64 coalescence, lowering it to 0.50 by early 2004 (Pereira et al., 2008).
Allies such as the center-right Progressive Party (PP) and the Brazilian Worker’s Party (PTB) demanded compensatory representation in key state-owned companies, e.g., the Banco do Brasil, and the National Development Bank (BNDES) (Congresso Nacional, 2016; Neto, 2007). In the first semester of 2004, these parties immobilized the Chamber until their appointee was nominated the director of Petrobras, namely the Brazilian state-owned oil multinational (Congresso Nacional, 2016).
Despite the Presidency’s efforts to make political transfers through cabinet allocation, the coalition secured only 213 of the 308 votes needed to pass strategic reforms (Pereira et al., 2008). Favoring autonomy over political decisions, the government decided to distribute budget amendments to the opposition, which is politically cheaper and helps political beneficiaries in their reelections. The opposition received 40.63% of the total money transfers destined to political parties, 75.70% of the total amount destined to individual legislators, and 89.43% of the total amount destined to state governors (Pereira et al., 2008). Budget distribution to legislators’ districts in exchange for votes is called pork barrel (Kunicová & Rose-Ackerman, 2005). The use of pork barrel certainly worked to approve legislation but caused discomfort within the coalition as allied politicians would compete against the same opposition legislators in the future.
Bertholini and Pereira (2017) indicate that the coalition’s size, coalescence, and ideological heterogeneity impact the costs of governance, operationalized by the Governing Costs Index (GCI) on a 0 - 100 scale. Such costs involve the cabinet portfolios that the Executive allocates, and the legislative amendments it includes in the federal budget. According to the index, the governing costs skyrocketed from a 37.2 GCI in former President Cardoso’s second presidential term to 63.3 in President Lula’s first term (2003-2006). If neither nominations nor budget allocations gave President Lula a solid majority, the question of how by the end of 2003, 11 out of 15 parties in the Chamber, and 73% of the deputies supported the government suggests additional coalition-building methods (Congresso Nacional, 2016). In May 2005, the Mensalão corruption scandal seemed to unveil at least some of these methods.
As per Figure 1, key directors and managers of the Correios, that is, the Brazilian postal service, were nominated by politicians (Congresso Nacional, 2016). The Correios hired certain companies which offered from 3% to 10% of the overpriced transaction’s values in bribes to officeholders (Junior, 2005). Two of these companies owned by Mr. Marcos Valério de Souza also had fraudulent contracts with the Banco do Brasil, Eletronorte, Ministry of Labor, and Ministry of Sports (Congresso Nacional, 2016). A congressional inquiry commission regarding denounces of vote-buying found that Mr. Dirceu (Chief of Staff of the Presidency) authorized the monthly financial transfers made by the PT’s treasurer to legislators (Congresso Nacional, 2016).
Figure 1. Simplified scheme of the Mensalão and Lava Jato corruption mechanisms.
The congressional inquiry commission of the Correios revealed that Mr. Souza raised funds in public and private banks and redistributed them to politicians under the guidance of the treasurer of the PT, to be used to cover electoral campaign costs and to buy votes in Congress already in January 2003 (Congresso Nacional, 2016). Due to Mr. Souza’s proximity to the Presidency, banks also offered him better terms so that he could intervene with the Presidency in their favor (Congresso Nacional, 2016).
When considering this “system of bribes”, one can easily note the institutionalization of corruption in the Brazilian public administration (Ribeiro et al., 2018). The Mensalão exposed the use of lucrative appointment-making envisioning the illicit collection of public funds, direct financial inducements (bribes), electoral favors (electoral donations of illicit money), and favorable business treatment (e.g., Mr. Souza’s companies) as informal methods of coalition-building (Força Tarefa Lava Jato, 2016). The support garnered by the Presidency through the patronizing distribution of influential positions and vote-buying served for two coalition-building purposes: increasing the coalition size, which due to party migration went from 254 to 325, and approving the government’s agenda, which suffered only eight defeats in 134 polls during Lula’s first term (Força Tarefa Lava Jato, 2016; Sant’Ana, 2020).
The Mensalão operation caused the PT’s leadership to step aside from the party and led to 24 convictions out of 40 defendants by 2014 (G1, 2012; Michener & Pereira, 2016). Moreover, although it seems puzzling that congressional inquiry commissions, sprouting from the legislature itself, investigated high-level politicians for crimes in a scenario of widespread corruption, the constant flow of evidence brought by Mensalão allowed for numerous agencies to assist and oversee legislators in their investigations (Michener & Pereira, 2016). Such agencies included the Federal Police, the Federal Public Prosecutor’s Office, the Federal Revenue Service, and the Brazilian Audit Court.
The scandal’s revelations further attracted intensive media coverage and popular demand for transparency, conveniently meeting the needs of a fragmented and opportunistic legislature willing to use concrete evidence to destabilize the government (Michener & Pereira, 2016; Pereira et al., 2008). Several coalition leaders lost their terms or political capital in Congress because of the accusations of corruption, thus weakening the Presidency, which had to increase the use of resources to sustain governability.
In this way, the development of Mensalão raised the costs of coalition formation. The Presidency resorted to methods that were beyond the perimeter of investigations and provided a higher amount of money (later discovered by the Lava Jato operation). Furthermore, to overcome the political crisis, the Presidency reduced its hegemonic character of cabinet allocation and redistributed important offices to the PMDB (Força Tarefa Lava Jato, 2016). Ultimately, the Presidency recovered its political capital and was able to approve important legislation, freeing funds to social policies and harvesting the benefits of a booming economy (Balán, 2014).
By joining judicial and controlling bodies under its umbrella, the Mensalão improved institutional coordination and enhanced information flows, which combined with intense media coverage, promoted procedural transparency in the investigations and consolidated the idea that high-level officials are accountable for their actions (Michener & Pereira, 2016). In response to the progress of the operation and mounting public pressure, the Brazilian Congress passed a series of anti-corruption laws: most notably, the 2010 Clean State Law, the 2011 Freedom of Information Law, and the 2013 Criminal Organizations Law. The latter, together with harsh sentences and consideration of spontaneous revelations in reducing the defendants’ sentences, strengthened the institution of plea bargain, which was widely used in the Lava Jato operation (Assessoria de Comunicação Social, 2014; Filho, 2017). Nevertheless, accompanying political stabilization, Congress retaliated against the judiciary by approving amendments limiting the investigative powers of the Federal Public Prosecutor’s Office and subjecting constitutional decisions by the Supreme Court to the Congress’ conditional approval (Michener & Pereira, 2016).
On one hand, the approval of substantial legislation that proved crucial to the upcoming Lava Jato operation shows the inability of Congress to contain the effects of the Mensalão and popular demand for increased transparency. On the other hand, the intense political backlash against judicial and control bodies once political stabilization was achieved, make it clear that Mensalão impacted the dynamics of Congress, and, in a way, made political alliances costlier due to the shutdown of some sources of illicit funds, and the political erosion of being associated with figures investigated for corruption. This reaction by the political elite contrasts with those predicted by Rupnik (2007) because in Brazil legislators passed criminally and judicially consequential laws (as opposed to shallow legislation) and actively sought to mitigate the political consequences of the operation, confirming the operation’s relevance.
8. Lula’s Second Presidential Term (January 2007-December
2010)
Reelected for his second term (2007-2010), Lula’s GCI reached 69.1 points (Bertholini & Pereira, 2017). Accordingly, retroactive proof of the rising costs of the government coalition surfaced with the advent of the 2014 Lava Jato operation. The corruption scheme later uncovered by the Lava Jato operation discovered the following: At least since 1990, a cartel of contractors worked together with the federal government in bypassing public bidding in exchange for articulating the payment of bribes (Força Tarefa Lava Jato, 2016). This cartel involved the nine biggest Brazilian contractors and operated fully and consistently at least between 2004 and 2013. In 2006, seven other contractors were admitted to the cartel, increasing the availability of illicit resources and the number of actors committed to that end.
Political parties nominated Petrobras’ board members, who then approved overpriced contracts offered by the cartel (Ministério Público Federal, n.d.b). The bribes equaled from 1% to 5% of the contract and were sent to money launderers who transformed them into licit money. In a first moment, contractors transferred cash to money launderers through overseas accounts or shell companies, while in a second moment the money went from the financial operator to the beneficiary in cash, by transfer abroad, or through the payment of goods.
The beneficiaries included Petrobras’ directors and their political patrons, political parties, and government officials. The main objectives were illicit enrichment, buying political support for the government’s agenda, and illicit campaign funding for perpetuation in power (Força Tarefa Lava Jato, 2016). In this context, the distribution, by the Presidency, of senior positions in the federal public administration to politicians was, in several situations, linked to a scheme of diversion of public money before, during, and after Mensalão.
In addition to lucrative cabinet allocation, documents discovered by the Federal Public Prosecutor’s Office in 2015 indicate that a decree issued by Lula’s Presidency in 2009 was “purchased” by lobbying in Congress and bribes to favor automobile assemblers (Matais & Fabrini, 2015). In 2017, the Presidency and other actors were accused of receiving bribes from the currently called J&F group (a Brazilian private investment holding company involved in the Lava Jato) to facilitate loans from the BNDES from as early as 2009 (Serapião, 2017).
The increase of PMDB’s cabinet share in the government, the persistence of monthly payments to legislators, and the expansion of the contractors’ cartel in Petrobras following the Mensalão show an apparent change of strategies in cabinet allocation and a shift in the preference order of informal methods: from direct financial inducements to a more complex system of mixed strategies indicated in Figure 1. Furthermore, innovative methods such as the selling of presidential decrees and facilitation of loans to interest groups appear to have emerged.
9. Dilma’s First Presidential Term (January 2011-December
2014)
After serving as Minister of Mines and Energy (2003-2005), whereby she presided over the Petrobras’ board of directors, and as Chief of Staff under Lula’s Presidency (2005-2010), Dilma became the first Brazilian female president in 2011. She gathered a big coalition that went from right-wing parties (PP) to left-wing parties (Communist Party of Brazil-PCdoB), occupying 338 out of the 513 Chamber seats and 48 out of 81 Senate seats (de Santi, 2010; G1, 2011).
As previously explained, large coalitions ease the approval of government-backed legislation at the cost of higher complexity of policy persuasion. By default, the coalition’s ideological heterogeneity implicates a sort of pragmatic governability, where ideologically distant parties trade vote alignment with the government’s preferences for benefits through participation in the government (Uribe, 2011). Cabinet allocation is used to accommodate divergent interests, gaining even more attractiveness in Brazil where such positions often represent opportunities to collect illicit funds.
Dilma’s Presidency maintained the PT’s hegemonic tradition, which upheld 17 out of 37 cabinets as opposed to the 6 allocated to the PMDB (Ribeiro, 2010). The lack of representativeness of other parties in executive cabinets is corroborated by a low coalescence of 0.59 in 1, and a high GCI of 71 points in 100 (Bertholini & Pereira, 2017). News reports from the period indicated that PMDB’s Congressmen complained about the party’s low representation, and that the replacement of four key directors of Petrobras in 2012 displeased the PP, PMDB, and PT, causing important defeats for the Presidency in Congress (Madueño, Lopes, & Samarco, 2010; Oswald, Jungblut, & Ordoñez, 2012). Amid political struggles, Dilma created two more secretaries with ministerial status, reaching 39 executive departments by 2015.
Before completing a full year in government, President Dilma saw six ministers fall, all of them denounced for corruption (Estadão, 2011). The crimes were committed under the previous administration but clearly revealed lucrative cabinet appointments as a key informal coalition method. Accordingly, the Republican Party (PR) left the government once its leader was removed from the Ministry of Transport, proving the pragmatic character of the alliance (Júnior, 2019). Nevertheless, as the bargaining utility of cabinet management dropped with the occupation of vacancies, the Presidency explored the use of executive decrees (agenda-setting power) and budget amendments (budgetary powers).
The annual average of decrees increased from 36 in Dilma’s first term in office (January 2011-December 2014) to 43 in Dilma’s second term (January 2015-August 2016), while the conversion rate into ordinary laws dropped from approximately 83% to 79%, the lowest being of 52% in 2014 (Matsui, 2017). As it is a manner of temporarily bypassing Congress, the overuse of vetoes or decrees might indicate less Executive-Legislative cooperation (Bedritichuk, 2021). In Dilma’s Presidency, a higher issuing of decrees coincided with her loss of political capital over the years due to a stagnating economy and the outbreak of the Lava Jato operation. Corroborating this trend, President Dilma was often accused of “governing by decree” at the time and criticized for not dialoguing enough with Congress (BBC, 2016).
Interestingly, Dilma’s Presidency seemed to have used her budgetary powers to safeguard party loyalty. In June 2013, huge protests against corruption and the costs inferred by the 2014 World Cup spread across the country. In one month, President Dilma’s popularity fell from 57% to 30%, leading her to propose an anti-corruption package of laws in Congress and a political reform informed by a public plebiscite (G1, 2013; Veja, 2013). Opposing a deep political reform, the PMDB, PP, and other allies formed a block to achieve higher bargaining power (Júnior, 2019). Under pressure, the Presidency liberated a record number of budget amendments for legislators (Patu & Costa, 2013).
In March 2014, the Lava Jato operation started at the Federal Court in Curitiba. Its task force, guided by the Federal Public Prosecutor’s Office, revealed that Petrobras’ director positions were distributed among the PP, PT, and PMDB (G1, 2015a). Moreover, as exemplified by Figure 2, Petrobras was divided into four cores. Whereas the political core involved the politicians responsible for nominating Petrobras’ directors, the administrative core referred to the board of directors (G1, 2015a). The board of directors, in turn, accepted bribes from the cartel—referred to as the economic core—which funneled these payments through money-launderers, known as the financial core, before ultimately reaching politicians (Lima, 2017).
Note. Adapted from Força Tarefa Lava Jato (2016).
Figure 2. Petrobras’ 4 cores.
In April, the opposition in Congress proposed creating a Senate parliamentary inquiry commission (CPI) and a joint commission with deputies (CPMI) to investigate allegations involving Petrobras (Jungmann, 2014). The government countered with identical proposals, accusing the opposition of leveraging Lava Jato to gauge electoral support. Dubbed the “CPI wars,” the dispute reached the Supreme Federal Court, which ruled all requests valid but prioritized the opposition’s due to chronology. Both the CPI and CPMI were established in May, and the CPMI’s final report recommended the indictment of 52 individuals.
President Dilma openly endorsed the Lava Jato operation from the outset, which the presidents of the Chamber and Senate—both members of the PMDB and under investigation—perceived as a direct attack. Despite divisions within the PMDB and the dissatisfaction of some coalition parties that the PT had declined to support in local elections, President Dilma secured re-election, albeit politically weakened compared to the start of her first term (Globo, 2014).
The rise of Lava Jato, combined with President Dilma’s somewhat permissive behavior towards it, hampered the Presidency’s political articulation capacity. Furthermore, in his plea bargain, Petrobras’ former director of supply denounced 25 deputies, 6 senators, and three governors, adding an element of tension to the Executive-Legislative relations (G1, 2015b). When drawing a parallel with Lula’s Presidency’s accommodation of the PMDB after the Mensalão, it appears that Lava Jato weakened President Dilma’s support base, leading her to grant ministries to other parties to retain her coalition’s cohesion, as well as increase her reliance on decrees.
10. Dilma’s Second Presidential Term (January 2015-August
2016)
The distribution of the ministerial team (39 positions) for President Dilma’s second term reflected the political weariness of the government. The PT had to cede three portfolios relative to her former mandate, and the PMDB gained one (Salomão, 2014). Despite the Presidency’s effort, the more conservative wing of the PT complained about having lost influence in Dilma’s government, while the PMDB assessed that it had no visible ministries.
At the end of 2014, Petrobras’ former supply director revealed a substantial part of the Lava Jato scheme, incriminating the PT, PP, and PMDB (Otoboni & Freire, 2020). As Lava Jato kept expanding in 2015, over 2 million people protested nationwide on March 15, both in support of the operation and against President Dilma’s government. Three days later, Dilma enacted a decree to enforce the anti-corruption law (12.846/13), imposing fines of 0.1% to 20% of a company’s gross revenue, with a minimum penalty equal to the illicit gains (Almeida & Yoshida, 2015). This sequence of events suggests that the decree was a direct response to public and political pressure.
By mid-2015, the presidents of all Brazilian largest contractors, Petrobras’ directors, and prominent figures of the PT were arrested (G1, 2015b). During the second half of 2015, the government’s leader in the Senate was arrested accused of attempted obstruction of justice, and denounces against the Chamber of Deputies’ president gained more consistency (G1, 2015b).
Still in 2015, Lava Jato unveiled that private interest groups that financed congressional election campaigns bribed high-ranking figures in Congress and the Executive for the preparation of favorable decrees (Bomfim, 2015). Dilma’s Presidency’s former Chief of Staff claimed that there was an exchange of favors involving bribes for amendments in at least nine hundred of the thousand decrees enacted in the four PT terms in office (Campos, 2018).
Faced with mounting political, public, and economic pressures, the Presidency eliminated eight ministries to cut costs, leaving the PT with just two more portfolios (9) than the PMDB (7) (Matoso, Alegretti, & Passarinho, 2015). This unusual move in any PT government shows President Dilma’s necessity to please the government’s coalition. Whereas the Presidency saw its formateur powers significantly decrease due to the arrest of key political leaders and the dismantling of the contractors’ cartel, a set of then-uncovered corruption schemes bolstered the position of the president of the Chamber, Mr. Eduardo Cunha (PMDB), relative to the Presidency.
Mr. Cunha’s office endowed him with agenda-setting and regimental powers. He decided the Chamber’s daily agenda and appointed the chairs of special and standing committees. The Antonov, Sépsis, and Lava Jato anti-corruption operations; J&F’s plea bargain; and President Dilma’s own testimonial to the magazine Carta Capital, all indicate that Mr. Cunha received millionaire bribes from companies to promote favorable legislation (Correio Braziliense, 2021; Estadão, 2017a; Estadão, 2021; G1, 2017a; Kaniak, 2020).
The money was used to buy votes for his election and support for his agenda. During the period in which Mr. Cunha presided over the Chamber, the votes taken to the plenary by the Legislative reached their highest historical number, 416 (De Carvalho, Viana, & Carlomagno, 2019). The shift from a more reactive Brazilian Legislative, often busy with proposals brought by the Executive, to a more active role reflects the changes in the Executive-Legislative relations (see Figure 3).
Note. Adapted from De Carvalho et al. (2019).
Figure 3. Roll call votes taken to the plenary by the Chamber’s Presidencies (2003-2017).
In addition to the resources brought by interest groups, President Dilma’s impeachment became a bargaining chip. The revelations of the J&F plea bargain and later by Mr. Cunha himself show that votes in favor of President Dilma’s impeachment had a monetary value, which would turn into political gains and protection against investigations in the eventuality of her impeachment (G1, 2017a). Corroborating this information, when vice-President Temer (PMDB), an ally of Mr. Cunha, assumed the Presidency in 2016, he appointed ministers who had been implicated in the investigations conducted by the Curitiba Federal Court (Costa, 2016; Jornal Nacional, 2017). In fact, ministers and legislators can only be prosecuted by the Supreme Court, and Curitiba would lose its jurisdiction over the appointed ministers. This characterizes a temporary method of coalition formation, namely “impeachment for political protection and resources.”
In December 2015, faced with compelling evidence from Lava Jato against Mr. Cunha, the PT voted to proceed with his impeachment process before the Chamber’s Ethics Council (Bragon & Uribe, 2015). In retaliation, Mr. Cunha authorized the advancement of an impeachment request based on financial crimes against President Dilma for a vote in the Chamber (Júnior, 2019).
As evidence against President Dilma’s mentor and former President Lula surfaced, she appointed him as her new Chief of Staff in March 2016 to shield him from the jurisdiction of the Curitiba Federal Court. Curitiba Federal Court Judge Moro leaked a conversation between President Dilma and former President Lula confirming attempted obstruction of justice and causing widespread public and political outcry (G1, 2016). In response, the Brazilian Republican Party (PRB) and the PMDB left the government. Sequentially, the Supreme Court suspended former President Lula’s nomination (Falcão, 2016).
Still in March, the Federal Public Prosecutor’s Office introduced an anti-corruption legislative package in Congress titled “10 Measures Against Corruption” (Ministério Público Federal, n.d.c). Supported by a popular campaign that began in July 2015, the initiative garnered over 2 million signatures, and showed how popular pressure and control bodies were actively affecting the legislative process.
Without being able to credibly nominate ministers, President Dilma’s impeachment had its way paved towards the final decision of the Senate in August 2016. To that point, around seventeen ministers of Dilma’s Presidency had been accused, charged, or condemned for embezzlement, favoring companies, and obstruction of justice (Praça, 2015). This number reached 18 by October 2016 (Estadão Conteúdo Web., 2016).
These events demonstrate that Lava Jato considerably impacted President Dilma’s coalition-building tools. First, by attracting enormous popular and media attention to the government, Lava Jato spurred changes in public tolerance for corruption, and it became politically costly to be aligned with the Presidency once scandals affected the winning coalition, lessening the President’s political influence. Second, the demise of the contractors’ cartel reduced the value of lucrative cabinet distribution to legislators while increasing its risks, diminishing the incentives to join the cabinet coalition. Third, the downfall of the Presidency’s formateur tools incited a partisan realignment within Congress to overthrow the government and return to the status quo of impunity existent before Lava Jato.
11. Temer’s Presidency (August 2016-December 2018)
Upon officially assuming the Presidency on August 31, 2016, President Temer faced the lowest popularity rates of any presidency (reaching 3% in September 2017) and a polarized Congress (Veja, 2017). Aware of having to pass difficult economic reforms, he focused on building a strong coalition. In the beginning, the government had strategic victories in Congress, where ally legislators of the Democrats (DEM) and PMDB won the House and Senate presidencies respectively (De Carvalho et al., 2019). This normalized the Executive-Legislative relations, which is explained by some of the Legislative’s disadvantages relative to the Executive. For instance, the Legislative cannot operate as a coordinating actor between Powers under normal conditions because legislative commissions are unstable and have little governance endowments.
In a context where left-wing parties were isolated after being in the center of corruption scandals, twenty parties transitioned to the Presidency’s sphere of influence (Ribeiro, 2018). President Temer reduced the number of cabinets to 23 and opted for a homogeneous center-right coalition, whose hardcore members were the PMDB, PP, DEM, Brazilian Social Democracy Party (PSDB), and Social Democratic Party (PSD). The PMDB abandoned the PT’s hegemonistic tradition and allocated visible and resource-rich offices to allied parties, such as the ministries of Health (to the PP), Economy (to the PSD), and Education (to the PSDB). This distribution yielded Temer the highest coalescence rate since democratization (0.80 out of 1), a 70% majority in Congress, and a GCI of approximately 15 (on a 0 - 100 scale) (De Carvalho et al., 2019; Ribeiro, 2018).
Alston and Mueller (2006) believe that presidents are more concerned with national policies as they are held directly accountable for them, while legislators try to collect resources for their constituencies. Before taking office, President Temer presided three times over the Chamber and knew of the “policy for pork” practice to pass important reforms (Ribeiro, 2018). From 2015 to 2016 the value in budget amendments released by the Presidency increased around 43.2%, while from 2016 to 2017, the increase was of 48%, reaching 10.7 billion BRL (Brazilian reais) (Estadão, 2018).
The winning coalition received 72.8% of the amendments, and relying on high loyalty rates from the allied base, the government “won” 76% of the times in Congress, which is above the historical average (Menegat, 2020; Ribeiro, 2018).
Nonetheless, the allegations stemming from Lava Jato ultimately extended to President Temer’s administration. As presented in Table 1, the values of budget amendments were substantially higher every time corruption allegations by the J&F or the Prosecutor’s Office placed President Temer in an unfavorable condition, matching the vote-buying allegations made by the opposition (Passarinho & Shalders, 2017).
Table 1. Budget amendments according to important events in 2017.
Date |
Event |
Description |
Monthly values in budget amendments (in BRL) |
Congress |
Senate |
22.03.2017 |
Outsourcing Reform. |
Allows the
outsourcing of any
activities of a
company. |
5.2 million in March. |
Approved. |
N/A |
27.04.2017 |
Labor
reform. |
Makes labor laws more flexible. |
5.5 million in April. |
Approved. |
Approved. |
06.05.2017 |
Privatization reform. |
Privatization of
infrastructure sectors. |
89.5 million in May. |
Approved. |
Approved. |
18.05.2017 (Effects felt in June and July) |
J&F’s plea bargain. |
Corruption accusation against Presidents Lula, Dilma, Temer, and a series of
high-rank
personalities. |
4266 billion in June/July. |
- |
- |
02/08/2017 |
Denounce against
President Temer. |
Passive corruption. |
199.2 million in August. |
Denied. |
Denied. |
04/10/2017 |
Political
reform. |
New campaign
financing rules. |
2273.9 billion in September/
October. |
Approved. |
N/A |
05/10/2017 |
Political
reform. |
Prohibition of
electoral coalitions. |
N/A |
Approved Approved. |
25/10/2017 |
Denounce |
Obstruction of justice. |
N/A |
Denied. Denied. |
Note. Modified from Faria and Perón (2018) and Ribeiro (2018).
J&F’s plea bargain revealed that from 2007 to 2010, the JBS (the founding company of the J&F) received 8 billion BRL from the Brazilian National Development Bank (BNDES), from which part was dedicated to the purchase of foreign companies and part to bribe politicians (Estadão, 2017b). The Brazilian champion donated 65 million BRL to several parties in the 2010 elections, and 366 million BRL in the 2014 elections through undeclared funds or bribes disguised as donations (G1, 2017a). In 2016, the J&F allegedly bought votes against the impeachment of President Dilma, whereas in 2017 the company allegedly bribed President Temer to approve decree No. 9.048/2017, favoring the company Rodrimar S/A (G1, 2017b; Serapião & Fabrini, 2017). Besides financing the campaign of 35% of Congressmembers, the J&F bribed Mr. Cunha and the former Minister of Economy, through whom the group had privileged access to the BNDES (G1, 2017c).
The dynamics brought by the J&F indicate that after Lava Jato started, the group influenced politics in favor of its partisan interests through direct bribes and loans facilitated by the BNDES for political favors. In this sense, private companies from sectors other than civil construction influenced portfolio distribution by the Executive, the issuing of decrees, and the purchase of votes in Congress. This strongly suggests that the use of lucrative cabinet appointments persisted in areas other than oil and gas and, together with direct bribes (in a process resembling Mensalão), strengthened the Presidency’s bargaining powers while meeting some Congressmembers’ “funding needs”.
In brief, Temer’s Presidency started the government with strategic coalitional and congressional victories, being able to shield key allies from justice and pass important reforms. The unfolding of Lava Jato contributed to rendering the Congress more expensive as, in 2017, the Presidency spent 4.5 billion BRL more than the previous administration in budget amendments. Following the restoration of stability within the coalition, actions were taken to curtail oversight/controlling bodies and centralize legal proceedings in judicial institutions based in Brasília, where politicians wield greater influence. These measures emerged in the aftermath of both the Mensalão and Lava Jato scandals.
Testifying the importance of Lava Jato, a political backlash followed political stabilization. The “10 Measures Against Corruption” package, for instance, was approved by the Chamber of Deputies in 2016 and by the Senate in 2019, following substantial revisions that significantly weakened its original strength. In 2019, the Senate approved a bill on abuse of power, which expanded the possibility of accusing judges and prosecutors (International Monetary Fund, 2020; Lagunes, 2020). Moreover, in the same year, the Supreme Court prohibited the imprisonment of people convicted of crimes by the appealing court, allowing defendants to appeal once more and freeing already convicted politicians, among which, current President Lula (January 2022-December 2025).
12. Lessons from Brazil: Navigating Corruption and
Coalitional Dynamics
Considering the effects of critical events as proposed by Capoccia (2015), it remains clear that the Mensalão and Lava Jato operations influenced the adoption of innovative methods of coalition formation in addition to the forms described by Chaisty et al. (2018). Both Lula’s and Dilma’s Presidencies allegedly deployed a scheme of sale of decrees and fraudulent loans favoring interest groups. Moreover, Dilma’s and Temer’s Presidencies used cabinet management with an innovative justification: protecting political allies. This means that anti-corruption operations forced Presidencies to resort to alternative coalition-building methods by lowering the likelihood of impunity and changing the incentives of being aligned with the government.
The more politically vulnerable the president, or the more exposed the informal coalition-building mechanisms, the more strategies are needed to secure the coalition’s cohesion. Here, institutionalized corruption determined the corrupt character of the new informal methods, but individual strategies reacting to changing incentives seemed to be ultimately responsible for their forms and intensity.
Additionally, the Mensalão and Lava Jato operations impacted informal methods of coalition formation by arguably changing the preference order between coalition methods. Lula’s Presidency increased its reliance on cabinet management and complex systems of mixed informal strategies (e.g., the scheme discovered by Lava Jato consisted of several informal methods). Dilma’s Presidency used cabinet management, budget amendments, and decrees vis-à-vis the shrinkage of resources previously retrieved from state-owned companies under investigation. Temer’s Presidency dealt with corruption accusations by intensely using budget allocation in exchange of favorable votes in Congress.
These results demonstrate that the increased cost of being associated with corrupt activities promoted by the government, caused in part by anti-corruption operations, led the presidents’ political agendas to be exchanged for activities related to patronage, clientelism, and pork barrel to expand their coalitions and sustain them in power (Carrubba & Volden, 2000).
The instability of Dilma’s presidency in maintaining coalition cohesion revealed an additional, unexpected outcome: a political realignment in Congress aimed at restoring the status quo of impunity that existed prior to the Lava Jato operation. During this period, when Congress wielded influence over certain informal coalition-building methods, two key behaviors emerged. First, interest groups financed members of Congress in exchange for favorable legislation. Second, votes for President Dilma’s impeachment were used as leverage by legislators to secure bribes, political backing, and positions within the subsequent administration. Therefore, President Dilma’s impeachment, much related to her loss of control of coalition-building methods and low support in Congress, recalls Chasquetti’s (2001) parallel between coalitional presidentialism and parliamentarism (more specifically, the vote of confidence).
The alternative hypothesis does not find ground as, contradicting Rupnik (2007) and Mesquita and Smith (2017), instead of passing shallow anti-corruption laws, Congress approved substantial legislation that reinforced judicial and controlling institutions, and de facto enabled Lava Jato’s breadth. Yet, legislative efforts were made to curtail the operations’ legacies once political stabilization was reached. Following the cost-benefit reasoning proposed by March & Olsen (2008), such findings suggest that, by constraining the Presidency’s coalitional powers, large anti-corruption operations alter incentives for the investigated corrupt practices. The operations represent a minor turning point within the prevailing political system, as they did not dramatically alter the course of institutional development, but still had moderate impact on subsequent political events. The operations forced the Presidency and legislators to reinvent themselves, while not precluding the alternative methods’ illegality.
As for the future of anti-corruption policy tools, this reflection underscores that institutional reforms are preventive and aim at altering the choice structure of individuals by managing the costs and gains of engaging in corruption. Reforms typically involve strengthening transparency mechanisms to promote accountability, expanding institutional oversight, enacting and enforcing laws or compliance rules, bolstering whistleblower protection, increasing penalties, conducting audits, implementing asset recovery mechanisms, and introducing competitive tendering, among other measures.
Similarly, Klitgaard (1988) emphasizes the significance of shifting people’s attitudes and fostering individual morality. His approach focuses on three key strategies: (i) penalizing corrupt behavior while restructuring rewards, (ii) enhancing information systems to improve detection, and (iii) eliminating positions where agents have monopoly over decision-making, discretion over implementation, and low accountability to the patron. Such approach is welcome and effective in curbing corruption-inducing opportunities, but its rigid understanding of corruption as cost calculations neglects behavioral and grievance drivers that render corruption fluid and adaptable to institutional challenges. This rationale behind anti-corruption policies may explain why institutional reforms often become obsolete in countries with systemic corruption.
Indeed, Rothstein (2011) notes that by employing an account of a utility-maximizer and risk-averse homo economicus, trust-based social interactions (such as corruption) could not fare. The reason being is that participants in the social contract would only cooperate as others cooperate, defecting when they anticipate being outsmarted or due to self-interest. Therefore, the mere awareness of opportunistic agents and their expected defection from agreements should lock parties into suboptimal outcomes of inaction or cyclical retaliation. In this scenario collaboration is ensured by reciprocity norms governing generally agreed procedures, rather than utility-maximization.
At their crudest form, corrupt relations are dyadic, contingent, relatively hierarchical, and iterative (Hicken, 2011). When dependency relations aimed at extracting rent are created along webs, corruption networks emerge, with agents (i.e., network nodes) offering transfers (i.e., vectors) to their “neighbours,” who then accept it and become accomplices, or refuse it and become witnesses (Ferrali, 2019; Yu, Kang, & Rhodes, 2020). These interactions are conscribed by organizational structure limits, and their sustainability depends on the successful activation of agents’ social capital, as cooperation requires reciprocity, solidarity, and trust (Junghagen & Aurvandil, 2020; Yu et al., 2020).
With this in mind, a purely cost-benefit approach to analyzing corruption—coupled with an analytical adherence to universalistic principles rooted in utility-maximizing and risk-averse human reasoning (isolated from broader social and contextual factors)—risks overlooking corruption’s long-term externalities and intangible effects. Moreover, this perspective may inadvertently legitimize and promote corrupt practices while fostering the replication of Western institutional models that were built on such assumptions. For example, where corruption is pervasive, greater transparency might instead reinforce the rationality of routinely prioritizing individual strategies over collective commitments, and when abuses are not redressed, transparency might generate frustration and feelings of powerlessness (Bauhr & Grimes, 2014). Under these political contingencies, increased transparency may lead citizens to resign from civic and political life, rather than voice their concerns.
Based on these considerations, future research should explore the main drivers of corruption networks as partial organizations as well as how they emerge and organize in political fora (Ahrne & Brunsson, 2011). As for the sustainability of governance, future research could explore how coalition stability interacts with street protests. Despite significant drops in popularity and corruption allegations against both administrations, stable coalitions under Lula’s and Temer’s Presidencies coincide with the lack of huge street demonstrations, which does not hold for President Dilma’s experience.
13. Conclusion
This article aimed to identify the mechanisms through which anti-corruption operations may impact informal methods of coalition formation in developing democracies. In Brazil, lucrative appointments, favorable business treatment, direct financial inducements, and electoral favors were used as coalition-building tools by all Presidencies. Interestingly, once President Dilma exceptionally “lost” the presidential coalition-making toolbox to Congress due to the gradually omnipresent Lava Jato operation and public pressure, her impeachment became a temporary bargaining chip for Congress-led coalition formation.
While anti-corruption operations were found to alter the order or form of informal coalition-building methods, and even encourage the adoption of meaningful anti-corruption legislation, politicians reinvented themselves within informality. This not only confirms path dependency and political backlash after stabilization, but indicates that purely institutional remedies and or alterations in incentives within political agents’ choice framework are not enough to curb corruption, for corruption’s fluid and networked nature as a standard operating mechanism adapts to such challenges. Further studies are needed to better understand the sociology of corruption and develop appropriate policy instruments.
Through analytic generalization, the article’s implications extend beyond Brazil to countries with political systems marked by coalition presidentialism, high political fragmentation, and self-serving organizational cultures. In such environments, informal coalition-building practices often play a crucial role in sustaining governance, especially when judicial and oversight bodies lack independence or the capacity to act effectively. As a result, countries like Argentina, Ecuador, and Indonesia, with similar coalition presidentialism and histories of systemic corruption, may display comparable dynamics. These findings also apply to some countries in Sub-Saharan Africa and Former Soviet states where minority presidents and entrenched informal coalition-building mechanisms prevail.
Ultimately, understanding the interplay between anti-corruption operations and informal coalition-building not only sheds light on the resilience of corrupt practices but also underscores the importance of rethinking policies and institution design to ensure that political stability is achieved without compromising accountability and governance integrity.