Essays on Economics of Education

Abstract

In this dissertation, I use data from Chile to study the determinants of schooling trajectories. Chapters 1 and 2 focus on higher education, Chapter 3 on secondary education, and Chapter 4 on teachers. Chapter 1 explores the role of a socio-emotional ability, self-efficacy, in understanding why students with comparable qualifications transit different college paths. This study is the first of two papers dedicated to studying the role of self-efficacy in the path students follow after high school. This first paper discusses the psychology literature on self-efficacy and how it can be measured. Then, through exploratory factor analysis and rich administrative data from Chile, I show that self-efficacy is a construct different from pure cognitive ability. Finally, I estimate a discrete choice model for the decisions of taking the college admissions test, applying, enrolling, and graduating from college. I find that conditional on cognitive ability, a higher self-efficacy increases the probability of taking the college admissions test, applying, and graduating from college within eight years. Analyzing heterogeneous effects, I find a bigger effect among students from low-SES families, which are precisely the ones with lower base levels of these outcomes. In Chapter 2, I take a step further and explicitly model the role of self-efficacy on the trajectories students follow after high school using a structural approach. I estimate a multi-stage discrete choice model with unobserved heterogeneity to study the role of self-efficacy on college applications, enrollment, and graduation decisions. The results indicate that higher self-efficacy significantly increases the likelihood of taking the college admissions exam and submitting a college application, conditional on cognitive ability. For students who apply, increasing self-efficacy also increases their probability of enrolling in and graduating from college, even more than a comparable increase in cognitive ability. From the analysis of socioeconomic groups, I document that improving students' self-efficacy could reduce the socioeconomic gaps in the percentage of students who take the college admissions test, apply, and enroll in college. These findings suggest that policies oriented to boost students' self-efficacy could alleviate income-related inequalities in access to higher education. Chapter 3 is co-authored with Dr. Daniel Kraynak and Dr. Cristina Riquelme. It investigates how local economic conditions impact human capital accumulation in Chile's copper-producing zones using high-frequency data on copper prices, school attendance, and academic performance. To measure the exposure to copper price volatility, we created an index by determining the proportion of workers in the area associated with the metal mining industry. We performed a difference-in-differences analysis by comparing students in areas with low and high copper exposure during periods of varying prices. The results indicate that increasing copper prices in more exposed areas decreases the quarterly attendance of high school students in the same period. We also find that students compensate for this lower attendance by increasing their attendance in the next quarter. Analyzing test score performance, we find evidence of a positive effect of local economic conditions on students' math performance in the same period. However, this effect is completely offset in the following year. Chapter 4 is co-authored with Dr. Macarena Kutscher, Dr. Cristina Riquelme, and Dr. Sergio Urzúa. It explores the contribution of teachers to student performance in Chile's college admission test (PSU). Our analysis is based on a unique teacher-student matched dataset and decomposition methods. The findings suggest that teachers' performance on the PSU and the characteristics of their educational degrees are significant predictors of students' success. When controlling for students' and predetermined school characteristics, the gap between vouchers and public schools is reduced. Productivity differences emerge as key factors driving the disparities across school types. The analysis underscores the crucial role of teacher-student interactions in shaping student outcomes.