Essays on financial aid and loan aversion

Chapter 1: In this paper, I use subtle differences in the federal aid application process to study how small costs affect student borrowing behavior by taking advantage of a policy change in which schools transitioned from the Federal Family Education Loan (FFEL) Program into the Direct Program. Using variation in when schools switch from FFEL into Direct Program in a difference-in-differences framework, I find a five percentage point increase in federal borrowing when students make one fewer decision under the more passive Direct Program. Additionally, average loan amounts among Direct loan borrowers are smaller. Finally, using Direct loan program status as an instrument for loan take-up, I find suggestive evidence that federal loan use increases retention.;Chapter 2: In response to tightened budgets and concerns about higher education costs, governments have introduced rules to limit the use of publicly funded aid, particularly at for-profit colleges. In this paper, we use an event study framework to estimate how tuition, enrollments, and loan use changed at for-profit colleges that became in-eligible for California's state grant aid program. Typical schools reduce tuition by 6 percent relative to unaffected colleges, indicating that the majority of the aid subsidy was captured by schools. In contrast, a subset of schools close to a federal eligibility threshold maintain or increase tuition. These results indicate that the Bennett hypothesis holds for targeted aid decreases in the for-profit sector.;Chapter 3: Among policymakers, there is growing concern some students are underinvesting in education because of their distaste for loans. To better measure this aversion and its relation to other attitudes, we conduct a lab experiment in which we elicit subjects' valuations of a money-earning task. Subjects must pay to earn money from the task, but in some treatments, we change all references of "pay" to "borrow" on the pricing list. We find people pay significantly less when offered an otherwise identical contract labeled as a loan across different levels of risk and loss aversion. Moreover, we find these effects are mitigated in a similar loan frame with all risk removed, but a subset of people continue to not borrow.