Chapter 1 is The Effect of Honors College Participation on Student Outcomes. Honors education refers to programs for high achieving students at U.S. post-secondary institutions. These programs provide high achieving students benefits such as the ability to enroll in exclusive courses with small class sizes, to live in special dorms, and to enroll in classes earlier than non-honors students. These changes to a student’s college experience may change their academic outcomes in ways that concern students and policymakers. Results in most prior research on the effect of honors program participation on academic outcomes may be biased by unobserved differences between students in and not in an honors program. This paper addresses these unobserved differences by studying an honors college that uses GPA admissions cutoffs. The Michigan State University Honors College admits all students in the top 10% of the freshmen fall semester GPA distribution of each non-honors college. I use a regression discontinuity research design to compare outcomes of students above and below the cutoffs, and attribute differences in outcomes to differences in honors college participation. I find that participation in the honors college reduces the time for students to get their first degree and increases the probability that first-generation college students will graduate from MSU. However, the honors college has an insignificant effect on most outcomes for most groups I check, so the few significant findings may be due to random chance and doing many significance tests.Chapter 2 is How Low-Income Expectations Affect Student Loan Repayment Plan Choice: Survey Evidence from College Seniors. Income-driven repayment plans lower required payments for student loan borrowers when their income decreases. This helps to reduce student loan defaults. Despite universal availability, only a minority of student loan borrowers in the U.S. are in an income-driven repayment plan. In this study, I test whether a student’s choice of repayment plan is related to their expectations of earning a low income. Using an information experiment in a web survey, I create two groups of college seniors with an exogenous difference in low-income expectations. I find that respondents who see the major specific income information believe they, on average, have a higher probability of earning a low income. However, those respondents are not any more likely to choose the income-driven repayment plan. I conclude that students’ repayment plan preferences are not strongly related to their expectations of earning a low income. This may be due to students caring about things other than minimizing monthly payments when choosing a repayment plan. Chapter 3 is The Effect of Test Score Performance Labels on Postsecondary Educational Outcomes: Evidence from Michigan. Standardized test scores and the labels associated with those scores provide students and their parents with highly credible information about a student’s academic achievement. This information could cause students and their parents to change their beliefs regarding a student’s academic ability. This may then consequently change the student’s future educational choices and thus their future educational outcomes. In this chapter I use administrative data on Michigan students to look at the impact of receiving different labels summarizing a student’s performance on standardized tests on a student’s post-secondary educational outcomes. I use a regression discontinuity research design to compare students who have similar test scores but who receive different summary labels. While some of my estimates are significant, almost all lack robustness to using another bandwidth and I am likely to find some spurious effects given the large number of estimates in this chapter. I conclude that I do not find evidence of a large effect of performance labels on postsecondary outcomes.
Read
