Eleições para senador no Brasil e gastos em campanha

Abstract

In most countries, financing of political campaigns is performed through the large business groups, which suggests that there is some set of interests embedded in their donations. In this sense, the effects of campaign spending on election outcomes have been widely investigated for various political systems all over the world with the aim to find empirical evidence to justify policy decisions regarding the regulation of campaign spending. The difficulties presented by the existing literature on the subject are to find appropriate solutions to the problems of reverse causality (higher probability of victory attracts more resources and higher spending increases the probability of victory) and omission of variables (voter’s characteristics, for example). In this work, the impact of campaign spending in the 2006 Senator’s election is estimated from a voter utility-maximizing framework by means of discrete choice models. The endogeneity of the variable expenses is treated with the inclusion of variables that can identify the not observed candidate’s “quality”. The results demonstrate th at expenditures are indeed important in determining election outcomes, but there is a difference on their returns whether the candidate is incumbent or not.