Civil Society and Presidential Accountability: Evidence from Brazil

Why does popular mobilization by civil society lead to the impeachment of some corrupt presidents but not others? Around two decades ago, O'Donnell (2003) argued that civil society activism is inevitable to bridge the weakness, and sometimes reluctance, of horizontal accountability institutions to hold presidents accountable. According to Transparency International’s Global Corruption Barometer 2013-2017, elected representatives are perceived as the second most corrupt actor, just after the police. Carothers (2020) counts 48 cases where presidents or prime ministers were ousted before the end of their term over corruption-related charges. In many of these cases, the ouster of politicians was triggered by popular mobilization by civil society groups and legal action. Brazil is a case in point. Since its transition to democracy thirty years ago, Brazil has witnessed two successful and two failed presidential impeachment attempts because of political corruption scandals that were accompanied by popular mobilization by civil society. Despite the long tradition in acknowledging the importance of the role of civil society in accountability, the literature has two gaps. First, most of the literature focuses on the relationship between the role of civil society in inducing accountability (Diamond 1994; Kim 2014; Perez-Linan 2007; Peruzzotti and Smulovitz 2006; Schedler 1999) without necessarily identifying the causal mechanisms linking popular mobilization by civil society to presidential accountability. In particular, when does popular mobilization by civil society lead to activating and inducing horizontal accountability institutions to sanction presidents and when does it fail? Second, the literature so far has focused on positive cases, that is cases when presidents were actually removed from office after corruption scandals (Damgaard 2018; Kada 2003; Ramalho 2007), rather than cases when they went unpunished. This paper bridges both gaps by examining the following question: why did popular mobilization by civil society in Brazil lead to the impeachment of Presidents Collor de Mello in 1992 and Dilma Rousseff in 2016, but not Lula Da Silva in 2005 and Michel Temer in 2018? The paper uses process tracing (Beach and Pedersen 2013) to uncover the causal mechanisms at work in the four cases of presidential accountability and non-accountability. Brazil is a perfect case because no other case gives this variety in the outcome and non-outcome at the presidential level. Doing this, the paper expands the scope of the presidential impeachment literature (Kim 2014; Perez-Linan 2007) by elucidating when and how popular mobilization by civil society induces horizontal accountability institutions such as parliaments or courts and updates the scholarly understanding on how the accountability of corrupt politicians is made possible.