QoG lunch seminar with Guillermo Toral

Abstract:

Prosecutors are central figures in the fight against corruption and the rule of law more
broadly. Yet we lack systematic evidence about whether they are effective at reducing corruption
and, if so, why. I argue that prosecutors’ use of autonomy and discretion in anti-corruption
work benefits from physical proximity to the communities they monitor. I test this theory
through a causal event study of state prosecutors in Brazil, leveraging administrative data on
their deployment and behavior across municipalities. I find that prosecutor presence causes
increased anti-corruption action targeted at the local government. In response to prosecutor
presence, local politicians hire more bureaucrats on the civil service, rather than on temporary
contracts – a common vehicle for corruption in this setting. Consistent with prosecutor
presence constraining malfeasance, I find that municipal accounts executed right after the
arrival of a prosecutor have lower levels of corruption (as measured by federal auditors) than
those executed right before. I combine these quasi-experimental findings with insights from a
survey of politicians and in-depth interviews with prosecutors. Together, the results suggest
that physical presence can make prosecutors more effective at fighting corruption, and provide
rare causal evidence of the impact of autonomous prosecutors on local governance.