Essays in economics of education.

Abstract

In the first chapter we test the effects of competition between public and private schools on education outcomes. Theoretically we show that competition should generate a new equilibrium characterized by stratification (sorting) of students by school type and an ambiguous effect on school quality due to peer and efficiency changes. To sign this ambiguous effect, we study the Chilean case, where a generalized voucher system has been applied for more than 20 years. Exploiting the across market variation, we analyze the effect of more competition on test score and sorting. To solve the selection bias problem due to the nonrandom assignment of students to schools we aggregate all the schools at market level, and then analyze the effect of competition on the overall system. To control for the endogeneity of competition we use an instrumental variable approach. Even though that our results suggest that a higher competition generates a large effect on sorting, its net effect on aggregate test scores is positive though small. The second chapter evaluates the partial impact of a large extension in the length of schooling day using the Chilean experience. Since 1997, Chile has been implementing a national policy that increases by 30% the hours that each student stays at school. Using alternative specifications, I find a robust positive effect for the Full Day Schooling Program applied during last years. However, the positive effects of the program have been significantly larger among voucher schools than among public schools and larger for Language than Mathematics tests. Additionally, the small magnitude of the effects raises doubts about its cost-effectiveness, at least in its current generalized application. In the third chapter I develop an alternative methodology for developing countries to test causes which explain the large increment in relative wage for high skilled workers during last two decades. With a nested production function, and considering different skill groups, is possible to detangle the relative change in premium wages into non-markets components, relative supply and relative efficiency changes. The results for the Brazilian labor market indicate that the large increment in the supply of workers with secondary education played a relevant role in explaining the changes in relative wage between unskilled and middle skilled workers, but did not in explaining the larger premium for high skilled workers. For the 1980s, the higher relative wage premium for the latter group is mainly explained by efficiency factors, which could be associated with skilled biased technical changes. However, during the 1990s the additional wage premium for these workers was explained by efficiency changes as well as by larger rents obtained by workers hired in the Public Sector.