| The three chapters of this dissertation address the economics of basic education in the state of Sao Paulo, Brazil. They use a unique data set that combines detailed information on schools with information on the socio-economic and political characteristics of municipalities. The topics addressed include the political economy of the provision of education quality, effects of decentralization of public education on public school resources and performance, and the resulting consequences on the quality of private schooling.; The first chapter empirically investigates the validity of the classic median ideal policy theorem, which states that political competition induces elected officials to satisfy the preferences of median voters. It analyzes the relation between different percentiles of the local income distribution and the public education spending that locally elected mayors choose. Our findings are consistent with the theorem when citizen-voters are heterogeneous in two dimensions; income and preference for education, and there is a private schooling alternative. However, the theorem is rejected when the only source heterogeneity across voters is income, as assumed by most of the existing literature.; The second chapter uses longitudinal data on primary schools to evaluate the effects of a decentralization reform implemented in Sao Paulo on several measures of public school resources and student performance. We find conflicting results for different school quality measures: decentralization increased dropout rates and failure rates across all primary school grades, but improved several school resources. Further investigation at the school grade level suggests that the combination of two effects were responsible for the worsening in school performance: democratization of school access which increased enrollments, and out-migration of above-average ability students from the decentralized schools.; The third chapter uses the same data-set to investigate the effects of the decentralization on private schools. Our findings are consistent with the results obtained for public schools: out-migrating students were of high ability relative to the public school, and low ability relative to the private school, resulting in a worsening of observed student performance in both private and public schools. |
