Are Anti-corruption Audits Effective? Evidence from Brazilian Municipalities

• In Event: Frontiers of Research on Bureaucratic Politics and Management Mini-Conference: Internal Decision-making and Management

Abstract

Why do anti-corruption audits reduce corruption? How do audits affect governments and do these changes persist over time? This paper considers a program shown to be effective -- randomized audits in Brazilian municipalities -- and examines a set of causal mechanisms underlying it: restructuring in bureaucratic personnel and management practices. Considering the effects of 10 years of randomized anti-corruption audits on the careers of over eighty thousand high-ranking municipal officials, we find that anti-corruption audits have no impact on bureaucratic personnel. Employee dismissal and departure rates, as well as hiring patterns do not change post-audit, in both low and high-corruption municipalities. Audits reduce corruption primarily through improvements in public management. Managerial practices are more transparent, accountable and formalized after audits, with long-lasting effects. Our results suggest that anti-corruption campaigns fail to remove bureaucratic accomplices, highlighting the potential for recidivism. Managerial reforms help reduce the margin for corruption if properly institutionalized, with long-lasting consequences.

Authors

• Romain Reda Ferrali, Aix-Marseille School of Economics

• Galileu Kim, Princeton University