The impact of conditional loan forgiveness: Evidence from the Texas B-On-Time program

• In Event: Unpacking the Black Box of Institutional Quality in Higher Education

Abstract

College completion rates, particularly on-time completion, remain relatively low. Access to student loans expand access to college and typically yield high individual and societal returns, yet the effects of debt on student decisions may depend on program design. This paper evaluates a program designed to tackle both of these problems simultaneously: the Texas B-On-Time (BOT) loan program. This program offered eligible undergraduates interest-free loans that could be fully forgiven upon on-time degree completion. This program provides additional access to capital to credit-constrained students, which may allow students to reduce outside work and/or increase their enrollment intensity. It also incentivizes effort throughout college through minimum eligibility requirements that use a students’ cumulative GPA. At the same time, this is an opt-in program which might lead only students who expect to graduate on-time to enroll, suggesting a null effect of the program overall. Finally, due to the time constraints and GPA requirements, it may induce students to select or switch into “easier” majors.  We exploit two features of the program's eligibility rules using a regression discontinuity design: a GPA threshold for continued loan receipt and a credit hour limit for loan forgiveness. Comparing students just above and below these cutoffs, we estimate the causal effects of receiving a conditional loan on educational choices—including course loads, major selection, and time-to-degree—and of receiving loan forgiveness on post-graduation outcomes such as employment, earnings, and graduate school enrollment. We complement these designs with difference-in-differences designs to examine the effect of the program introduction. We use detailed administrative records from the Texas Education Research Center, including K-12 records, postsecondary enrollment and financial aid data, and labor market earnings. This work is preliminary and on-going. Our findings will inform whether conditional forgiveness effectively promotes timely completion and improves labor market outcomes, or whether simpler aid structures may better serve students.

Author

• Meredith S Welch, American University
  Presenting Author