Repayment Design and College Major Choice: Evidence from a Student Loan Reform

• In Event: What Shapes Student Trajectories? Evidence on Policies, Institutions, and Incentives

Abstract

Student loans are widely used across countries as a way to help students access higher education. However, a common problem with this type of loan is the affordability to repay once they enter the labor market. To solve this problem, many countries have adopted Income-Driven Repayments, which are loans that tie the amount paid by the student to their income.

These loan programs have been studied for their effects on borrowers’ decisions. Several papers have documented their effects on delinquency (Herbst 2023, Monarrez and Turner 2025), intermittence (Chava, Tookes, and Zhang 2023, Dinerstein, Yannelis, and Chen 2024, Boutros, Clara, and Gomes 2024), and mortgage defaults (Mueller and Yannelis 2019). More recently, several papers have studied their effects on labor supply, finding evidence of moral hazard (Matsuda and Mazur 2022,
Silva 2025, Fu, Li, and Zafar 2025). However, there is little evidence in the literature on their effects on human capital investment decisions. Income-Driven repayments can have unintended consequences on student decisions when they are entering college. On the one hand, students are less constrained by their payments, allowing them to choose more freely the major that they like the most or more expensive. On the other hand, they might try to game the system by choosing a less profitable major so that they can repay less in the future.

In this paper, I study the effects of repayment redesigns of student loans on human capital investment decisions in higher education. Using administrative data on applications, admissions, and enrollments in Chile, I examine how a major reform that reduced the interest rate of student loans and incorporated an optional income-driven repayment plan affected applications and field-of-study choices across selective universities and vocational institutions. To evaluate the impact of this reform on student outcomes, I use a differences-in-discontinuity research design, exploiting the eligibility cutoffs for the student loans. I find no effects in immediate enrollment across institutions,
but positive and significant effects on STEM enrollment for selective universities, and negative and significant effects for non-STEM enrollment in vocational institutions. These results suggest that changing the repayment redesign of student loans affects major program compositions, even with no overall increase in enrollment.

Author

• Eduardo Alfonso Barrueto Silva, University of Texas at Austin
  Presenting Author