Abstract EN:

This thesis is composed of four independent evaluations of public policies. The first chapter assesses the impact of providing loans to students willing to enroll in university in the context of South Africa. It takes advantage of the fact that the loans are granted according to a credit score threshold to estimate the causal impact of loan obtainment based on a regression discontinuity approach. The results indicate that credit constraint is substantial, as it reduces the enrollment rate into higher education by more than 20 percentage points in a population of student loan applicants. The second chapter investigates the impact of mandated political reservation at village councils level for low caste and female candidates on the provision of school inputs. A feature of the selection procedure of reserved constituencies allows to implement a variance of the regression discontinuity design. The results indicate that neither female nor low castes representatives significantly favor the children from their category in terms of provision of school resources in the context of Bihar. The third chapter examines two channels through which the political affiliation of local representatives could affect the allocation of public goods. The allocation could differ because the representatives have different political views. Alternatively, it could be that elected representatives from the ruling party are favored. Using "close elections" comparisons provide quasi-experimental (RD) estimates of both effects on middle school construction. Large positive impacts of political preferences and political favoritism can be statistically rejected. Finally, the last chapter provides quasi-experimental estimates of the impact of a tax credit program targeted at less densely populated areas. . .