Higher education has changed dramatically in the last 40 years. What was the privilege of rich nations or poor countries' elites is now an integral part of international competition and development strategies. However, the rapid expansion in enrollment has posed significant challenges in terms of providing adequate financing, access and securing quality in higher education. This dissertation explores three aspects of these relevant issues.

The first chapter presents new estimates of the returns to higher education in two Latin American countries, Chile and Perú. Combining administrative records with a simple economic framework, I document large heterogeneity in the average returns to higher education, and I find negative net benefits of pursuing a number of degrees.

The second chapter studies a potential unintended consequence of student loans. Overall, student loans have proven to be effective in increasing college enrollment, especially among low-income students. Yet, loans' ability to improve student welfare depends on the pricing response of schools. If schools exert some degree of market power and set tuition strategically, they may react to loans by raising tuition in order to capture some portion of the aid. This generates a negative externality on ineligible students, who have to pay higher tuition fees than if loans did not exist. I develop an econometric model of supply and demand for higher education and study this phenomenon in the context of a student loan program that was implemented in Chile in 2006. I find that, on average, schools raise tuition by 6% in response to loans, which generate an average externality of US$178 per student per year.

Finally, the third chapter studies the endogenous formation of centralized admissions to college and its potential benefits. Policymakers around the world have adopted market-design-inspired centralized matching systems for assigning students to public schools. However, the question of whether policy intervention is necessary for such adoptions has been little studied. Examining a setting with application costs and heterogeneity in college quality, I show that sizeable application costs and small heterogeneity in college quality may lead to a voluntary transition to a centralized matching system. Using a 2012 system change in Chile, I demonstrate the plausibility of our theoretical setting and show that the enlarged pool of colleges in the centralized admission is welfare-improving, particularly for those students facing high application costs.