| Chapter 1 examines the effects of college quality on students' learning, employment and earnings in Colombia. Scores on a national college entry test solely determine admission to many selective Colombian universities, creating exogenous peer and resource quality variation near admission cutoffs. In one regression discontinuity (RD) approach using applicant lists for Los Andes University, a top-ranked selective college, I find that relative to applicants below the cutoff, those just above score significantly higher on a national college exit test, are more likely to be employed after college and earn more, with the strongest effects for low-income applicants. In the second RD approach I generate a range of exit test-score effects from 25 nationwide selective universities that I relate to input differences. I find that peer quality and the fraction of full-time faculty relate positively and significantly to college exit scores.;Chapter 2 investigates the extent to which secondary and higher education supply constraints affected aggregate educational attainment in Colombia for cohorts born from 1909 to 1981. Until the 1950s, secondary and tertiary Colombian education was highly restrictive. After the 1950s the number of secondary schools and colleges increased considerably, and enrollment and attainment at both levels improved. Using variation in cohort size to proxy for changes in education demand I estimate that for cohorts born before 1945 a 10% increase in cohort size reduced the cohort high school completion rate by 6% and the college completion rate by 9%. For cohorts born after 1945, I find a significantly smaller elasticity, such that a 10% increase in cohort size reduced the high school completion rate by 3% and the college completion rate by 4%.;Chapter 3, with Eric Bettinger and Michael Kremer, takes advantage of an educational voucher program in Colombia for which spots were assigned by lottery to investigate whether vouchers increase educational productivity or are merely redistributive, benefiting recipients only through more desirable peers at the expense of other students. Among voucher applicants to vocational schools, lottery winners attended schools with less desirable peer characteristics. Still, lottery winners had better educational outcomes including higher graduation rates and reading test-scores. In this population vouchers improved student outcomes through two channels: economic incentives and the tighter fit of private vocational schools with the labor market. |
