Research
Dissertation Papers
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 then 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 considers for admission 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 may reduce the time for students to get their first degree and increases the probability that first-generation college students will graduate from MSU.
How Low Income Expectations Affect Student Loan Repayment Plan Choice: Survey Evidence from College Seniors
Revision Based on Feedback from Journal Referees: (pdf)(SSRN page)(data, do files, and survey screenshots)
This material is based upon work supported by the National Science Foundation under Grant No. 2049358. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.
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.
Standardized test scores and the labels associated with those scores provide students and their parents 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 change a student’s future educational choices and thus their future educational outcomes. In this paper 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 of robustness to using another bandwidth and I am likely to find some spurious effects given the large number of estimates in this paper. I conclude that I do not find evidence of a large effect of performance labels on postsecondary outcomes.
Correction: An earlier version of this paper contained the following sentence: "Unlike Papay et al. (2016) this study looks at all students who attended a public school in 11th grade rather than only looking at low-income urban students." That sentence was incorrect. On page 361 the authors of Papay et al. (2016) write: "We confirmed that there are no effects of labeling on the ELA examination or for students who were higher income or from suburban schools on the low-stakes cutoffs we examine here." In a footnote the authors write that detailed results are available upon request to the first author. See Papay, John P., Richard J. Murnane and John B. Willett. 2016. “The Impact of Test Score Labels on Human-Capital Investment Decisions.” The Journal of Human Resources. Vol. 52(2) pp 357 – 388
Pre-Dissertation Research
This paper was last edited in 2021.
Income-based student loan repayment provides borrowers with insurance against high repayment obligations relative to their income. Borrowers on income-based repayment rarely default on their loans and there is research which suggests that this is casual. Despite this benefit most students choose to stay on the default 10-year repayment plan. This study investigates whether the specific numerical parameters of the U.S. Federal Government’s income-based repayment plans contribute to the low take up of income-based repayment. I do this by piloting a survey experiment to undergraduate students at Michigan State University. The survey provides evidence that students are more or less likely to choose an income-based repayment plan over the standard repayment plan depending on the specific value of the parameters of the income-based repayment plan.