Corruption and the Voter's Decision : Experimental Evidence from Brazil

This Element explores the conditions under which corrupt politicians are reelected, focusing on court rulings on candidate malfeasance. It shows that low-income voters are responsive to such rulings but are only slightly more responsive to candidate attributes, party labels protect a candidate from corruption, and attributes have little effect.

Despite voters'' distaste for corruption, corrupt politicians frequently get reelected. This Element provides a framework for understanding when corrupt politicians are reelected. One unexplored source of electoral accountability is court rulings on candidate malfeasance, which are increasingly determining politicians'' electoral prospects. The findings suggest that (1) low-income voters – in contrast to higher-income voters – are responsive to such rulings. Unlike earlier studies, we explore multiple trade-offs voters weigh when confronting corrupt candidates, including the candidate''s party, policy positions, and personal attributes. The results also surprisingly show (2) low-income voters, like higher-income voters, weigh corruption allegations and policy positions similarly, and are slightly more responsive to candidate attributes. Moreover, irrespective of voter income, (3) party labels insulate candidates from corruption, and (4) candidate attributes like gender have little effect. The results have implications for when voters punish corrupt politicians, the success of anti-corruption campaigns, and the design and legitimacy of electoral institutions.