Replication Data for: Anti-Political Class Bias in Corruption Sentencing

Are corruption trials that involve the highest ranks in the public sphere and large private companies biased against some groups? Existing research predominantly focuses on corruption prosecutions of politicians, leaving unresolved the extent to which judges apply differential treatment when convicting and sentencing the political class compared to other defendants, including those in the private sector. To address this gap, we investigate judicial bias within Brazil’s famous “Operação Lava Jato”, the largest corruption investigation carried out in history. Leveraging an original database that traces the trajectory of the universe of the 3,154 cases of Lava Jato, we show that judges’ sentencing decisions were not governed by a partisan logic. Instead, judges were more inclined to impose longer prison times and higher fines to elected politicians when compared to all other defendants, particularly those from the private sector. We interpret these findings as evidence of anti-political class bias. Qualitative analysis suggests that this bias arises from judges’ ethos centered on safeguarding society against State corruption and the pressure exerted by public opinion.