What “bad apples” removed from barrels can tell us about corruption

Resumo

This study identified the distribution of administrative penalties imposed between 2003 and 2014 in the federal executive in Brazil, and analysed which civil servants are most likely to be sanctioned for corruption and which conditions differentiate them from those punished for other serious offences. Statistical analyses were performed on the subsample of sanctioned civil servants using the variables of gender, job tenure, status (career civil servant or occupant of a position of trust), and gross earnings. Analyses also included qualitative data gathered through 24 semi-structured interviews with civil servants directly involved with disciplinary sanctions. As this paper used real measures, and not survey answers, experiments or aggregate indexes, it explored methodological challenges of convictions databases, by questioning what can be learned regarding both bureaucratic corruption and official responses to it. Findings suggest that those punished most often for corruption were men and those who were not in the first five years of their careers. Apart from providing accurate portrayals of those caught and punished that are still scarce in the literature on corruption studies, the analysis on sanctions can help agencies better prevent and respond to misconduct. This can be done by identifying in advance employees who can either change their conduct or be subject to removal from the workforce or by identifying disparities that may occur in the context of broader inequality of sanction enforcement.