Research

Job Market Paper

Gambling with Graduation: Mobile Sports Betting Legalization and Student Persistence [draft]

This paper examines whether the legalization of mobile sports betting affects college student persistence. I construct an original dataset identifying the precise date each state first offered legal mobile sports betting, distinct from the date of legislative authorization, and combine it with an annual panel of 1,488 four-year colleges from the Integrated Postsecondary Education Data System (IPEDS), 2014–2024. Using a difference-in-differences design that accounts for staggered treatment timing (Callaway and Sant'Anna, 2021), I find that legalization is associated with a statistically significant 1.21 percentage point decline in four-year graduation rates, with effects increasing in cohorts' cumulative exposure to legal betting access. The effect is absent from both six-year graduation rates and first-year retention, the latter serving as a falsification check since most entering freshmen are below the legal betting age of 21. Checks on applicant, admissions, and enrollment composition show no evidence that the results are driven by selection into treated institutions. The findings suggest mobile sports betting access delays, rather than prevents, college completion.

Working Papers

If You Build It, They Will Learn: The Effect of Construction Spending on Construction Trades Graduates with Joshua Hess & John Killingsworth [under review]

We examine the impact of construction activity on construction trade degrees and find lagged declines in trade degree completions associated with increased activity. Using county-level construction spending data matched with institution-level degree conferral data from the Integrated Postsecondary Education Data System (IPEDS) for non-metropolitan U.S. counties from 2010 to 2019, we estimate fixed-effects regression models that account for persistent differences across institutions, counties, and years, and allows for lagged effects. We split the sample by spending quartiles. We find a 1% increase in spending leads to an immediate 6% decline in degree completions and a 10% decline two years later for the bottom quartile. The top quartile sees 16% decline one year after a 1% increase in spending somewhat mitigated by a 10% increase two years after.

Workforce by Design? The Effect of Targeted Scholarships on College Major Choice

This paper studies whether adding a supplemental scholarship for STEM majors to an existing merit-based aid program increases enrollment and graduation in targeted fields. I examine South Carolina's Math and Science Scholarship Enhancement, which layers an additional $2,500 per year onto the state's LIFE and Palmetto Fellows merit scholarships for students majoring in an approved STEM field, using student-level enrollment and financial aid data from the University of South Carolina and treating ineligible out-of-state students as a control group. I find that South Carolina resident students were 2.47 percentage points more likely to choose an enhancement-eligible major after the policy took effect, an effect concentrated among male students, with no evidence that this reflects students sorting into majors with relatively higher pre-enhancement GPAs. I also find a modest decline in the likelihood of pursuing a double major, consistent with time constraints or added caution around meeting the scholarships' 3.0 GPA requirement. Despite the shift in major choice, I find no statistically significant increase in the rate of students graduating with a STEM degree, suggesting that targeted scholarship funding alone may not be sufficient to generate new STEM graduates.

Hail to the Chief: The Impact of New Presidents on College Employment

This paper examines how changes in college presidents affect institutional employment. Using annual IPEDS data from 2013–2023 and an event study design that exploits the timing of leadership transitions at 720 primarily baccalaureate institutions, we track staffing adjustments across twelve occupational categories. We find that presidential transitions are followed by gradual but persistent declines in total employment, driven by reductions in instructional staff, student and academic affairs personnel, office and administrative support staff, and facilities and maintenance workers. These effects typically emerge two to three years after the transition and grow over time, consistent with restructuring through attrition, hiring freezes, and operational consolidation rather than immediate layoffs. In contrast, eight occupational categories— including IT staff, graduate assistants, management, healthcare personnel, and service workers—show no systematic response to new leadership. The results suggest that presidents do matter for how institutions allocate labor, but their influence is uneven, targeted, and realized slowly through routine budget cycles rather than dramatic short-term shifts.