Essays on schooling and economic development: A micro-econometric approach for rural Pakistan and El Salvador

This dissertation is composed of four essays on schooling and economic development. Chapter 2 investigates the role of permanent and transitory income shocks in school attendance using household panel data from Pakistan. The results of school entrant and dropout regressions indicate that the transitory income movements affect children's schooling behavior significantly, implying that credit and insurance market imperfections exist. Our analysis also points out gender gap in education and resource competition among siblings. Human capital investment decision and intrahousehold schooling allocation seem to be affected by a need for self-insurance devices.;Chapter 3 investigates the sequential educational investment process by integrating field observations, economic theory, and econometric analysis. The field surveys were conducted twice in rural Pakistan, in order to collect retrospective data of 2365 children of 367 households. The most striking feature discovered is the high educational retention rate, conditional on school entry. We also find that at the higher education levels the schooling progression rates become comparable between male and female students. These results imply that under borrowing constraints, parents pick the 'winners' for educational specialization and allocate more resources to them.;Chapter 4, co-authored with Emmanuel Jimenez, measures the effects of decentralizing educational responsibility to communities and schools on student outcomes by using the example of El Salvador's Community-Managed Schools Program (EDUCO). This chapter finds that the rapid expansion of rural primary schools through EDUCO has positively affected student achievement in language skills through enhanced community involvement; and has diminished student absences, which may have longer-term effects on achievement.;Based on a principal-agent model, Chapter 5 investigates the organizational structure that made the EDUCO program successful by estimating the teacher compensation and effort functions, and input demand functions. While the EDUCO teachers receive a piece rate, depending on their performance, wage payment is relatively fixed in the traditional schools. Community participation seems to enhance the teacher effort level and thus increases students' academic performance indirectly. Moreover, parental associations can affect school-level inputs by decentralized school management. Our empirical results support the view that decentralization of education system should involve delegation of school administration and teacher management.