Association for Public Policy Analysis & Management

According to the Federal Reserve Bank of New York, outstanding student loan debt exceeded $1.1 trillion in December 2014. Concern over growing debt, delinquency, and default rates has stimulated various policy proposals, including income-based repayment (IBR) plans that link loan payments to borrowers’ earnings and "human capital contracts" (HCC) that would allow students to pledge a set fraction of post-college earnings for a specified period in exchange for tuition. In this project, we investigate the factors that affect students’ loan repayment choices, with an emphasis on exploring how behavioral biases in students’ decision processes affect both selection into income-driven schemes and effort exerted by borrowers in these plans after they enter repayment.

First, we test how the framing of the two key features that distinguish income-driven schemes – protection from unaffordable loan payments in periods of low earnings versus potentially higher costs over the life of the loan – affects empirical measures of adverse selection. In a survey University of Maryland  undergraduates, we elicit students’ preferences for income-driven plans relative to the standard plan by presenting students with scenarios in which both the percentage of income that payments represent and the framing of the alternative plan are randomly assigned. We examine the relationship between the level and variance of expected earnings and students’ preferences for IBR and HCC repayment plans. Furthermore, we test whether the relationship between earnings and preferences varies with the size of debt, the percent of income paid under the alternate plan, or the framing of the income-driven scheme.

Second, we use data generated from students’ choices in a lab experiment to investigate how effort decisions are affected by the available set of loan repayment options and whether borrowers are allowed to choose versus being forced into participation in an income-driven plan. Absent behavioral biases, agents who would have voluntarily selected such a plan should behave no differently in the event they are involuntarily assigned to the same plan. However, borrowers who voluntarily select into income-driven repayment may be more likely to focus on its insurance aspect and work hard to honor their contract while also earning more money for themselves. Conversely, borrowers who feel they have been forced into the scheme may resent their higher prospective payments and be discouraged from persisting in school or pursuing high paying but difficult jobs. Thus, the lab experiment allows us to test how voluntarily choosing versus being assigned to income-driven repayment affects costly effort.