Are Educated Candidates Less Corrupt Bureaucrats? Evidence from Randomized Audits in Brazil

In this paper, we test whether more educated candidates make less corrupt public managers. Leveraging electoral regression discontinuity design and a randomized inspection program, we show that more educated candidates commit 32% fewer moderate infringements (associated with public management), but we do not find a differential in severe irregularities (corruption related). Exploiting data on judicial records, we find this effect does not stem from differences in corrupt behavior and might be explained by differences in managerial skills. In addition, more educated mayors have the same chance of running again and being reelected as their less educated peers. Taking advantage of administrative labor records, our examination of possible mechanisms shows that educated candidates have more previous experience in the public sector and high-skill positions. Finally, we find no difference in the provision of public goods between these two groups. In summary, we find that more educated candidates are more effective bureaucrats rather than better politicians.