Public costs vs private gain : assessing the effect of alternative information treatments on electoral accountability

Abstract

Do voters punish candidates with records of misuse of public funds? I address this question in the context of the 2008 and 2012 municipal elections for mayor and City Council legislator in Brazil. I use a unique dataset listing all candidates with accounts rejection records by any of the 34 Brazilian Audit Courts between 2004 and 2012 and electoral data. I test the effect of having accounts rejected on subsequent electoral returns. To test this effect I use matching and difference-in-difference. Results show evidence that voters punish both candidates for mayor and City Council legislator in most elections.