Essays in political economy

• Autor(a) principal: Oliveira, Marcel Ferreira de
• Data de Publicação: 2021
• Tipo de documento: Tese
• Idioma: eng
• Título da fonte: Biblioteca Digital de Teses e Dissertações da USP
• Texto Completo: https://www.teses.usp.br/teses/disponiveis/12/12138/tde-08122022-213753/
• topic: Close elections; Economia política; Eleições acirradas; Gender and local legislative elections; Gênero e eleições para o legislativo local; Political economy; Regressão descontínua; Regression discontinuity

This thesis is divided into two essays divided by chapters. In the first chapter, we investigate whether the size of legislative houses have any influence in the gender composition of the legislative house. To do so, we employ a regression discontinuity design to exploit an exogenous variation in the size of local legislatures in Brazil. This variation resulted from a rule that determined precisely the size of local legislatures according to population thresholds for the elections of 2004 and 2008. This quasi-experiment was investigated previously by Correa e Madeira (2014), which had found that the number of seats in the legislature has a significant and positive impact on the share of women in the legislature. The authors claim that this is result from a cannibalization effect among male candidates. As the number of seats increase, more candidates participate in the elections. However, most of these candidates are men and they compete with each other. This competition does not affect women as much, which results in an improvement on the chance that women will be elected. We develop a theoretical framework to formalize this argument. We develop a model of imperfect coordination in which voters have distinct preferences for male and female candidates. Under the assumption that it is costlier for women to participate in the election and that the voters are sufficiently biased against women, we are able to model and replicate this cannibalization effect. Moreover, the authors also find that this increase in female representation was followed by improvements in early education, antenatal and infant health care and social and community assistance. We re-examine their findings using non-parametric methods. We find very similar results, serving as additional exercises for the robustness of these results. In the second chapter, we take a different approach to examine the entry of women in politics and its effects on the provision of public goods. We investigate the impacts of electing a woman to a local legislative house in the context of close elections in Brazil. In contrast to the previous study and to other studies, we find no evidence that there is any effect of the gender of the legislator on policy outcomes related to early education, prenatal and child health care and government spending. Our regressions did find significant decrease in underweight births; however these findings are not robust when we employ the Romano-Wolf procedure to control for the Familywise Error Rate (FWER). That means we cannot rule out the possibility that this result is just a false positive. This lack of concrete results is in contrast with previous literature. We argue that simply electing women might not be enough to observe significant impacts in policy-making, leading us to believe other political factors might serve as prerequisite for their effective participation.