Research
I'm an applied microeconomist who studies human capital/education in diverse contexts. My research has addressed the selection biases into merit-aid, creating new insights on the responsiveness of students to different renewal targets, examined the role of peers in the formation of soft skills, and found that reducing piracy has deadly consequences for Somalians. My future research seeks to further develop these areas and deepen our understanding of human capital formation.
Publications
Where is the Value? The Impacts of Sow Gestation Crate Laws on Pork Supply and Consumer Value Perceptions [Link]
(With Clinton Neill and Peilu Zhang)
Food Policy, 2023
Submitted
2020-2021 JMP
Best Graduate Student Paper North American Regional Science Council 2020 (Page: http://www.narsc.org/newsite/awards-prizes/narsc-student-competitions/)
Financial aid renewal requirements are ubiquitous and give students a concrete goal which is a promising strategy for students transitioning beyond high school. Yet, our understanding of merit-based aid's renewal requirements could be deepened with alternative identification strategies that address selection into programs with publicly known academic eligibility cutoffs. This paper estimates how receiving West Virginia's merit-based Promise scholarship affects students' likelihood of meeting academic renewal requirements using an unmanipulated birthdate cutoff that produces an internally valid fuzzy regression discontinuity. In contrast to known entry test requirements, density tests do not find significant enrollment differences at the birth cutoff. Scholarship-receiving students are 24 percentage points (pp) more likely to exceed the credits requirement and 31 pp more likely to exceed the GPA requirement, on average. There are no significant effects of crossing this compulsory grade school entry birth cutoff, when crossing it does not also result in receiving the scholarship which means the effects are driven by students' responses to the requirements and not relative age effects. These estimates are externally valid for complier students away from the cutoff, but not to GPA targets of non-compliers due to never-takers out-performing untreated compliers. These estimates advance knowledge on how responsive students are to renewal goals, particularly their relative responses to credits or GPA targets and how externally valid those estimates are, informing design of financial aid renewal incentives.
(with Scott Cunningham and Matthew Pearson)
Non-cognitive skills play important roles in education and careers, but less is known about peers' influences. National Outdoor Leadership School conditionally randomly assigns students to wilderness classes which allows us to estimate causal effects of peers on self-reported non-cognitive skills. Being assigned to peers with higher self-rated non-cognitive ability reduces self-reported ratings in communication and leadership, particularly for females. This work suggests relying on strong peers may not be an effective way to bolster non-cognitive ability for college-aged individuals and may in fact contribute to fewer women in leadership positions.
Violence Displacement from Sea to Land: Evidence from Wind-Induced Somalian Piracy Reductions
Maritime piracy is a prominent form of violence in countries that face struggling institutions (e.g. Somalia), often attracting international wartime resources to combat violence on international waterways. An unintended consequence of piracy deterrence are reduced opportunity costs of violence on land. This paper examines how deterring pirates impacts land conflict. To overcome endogeneity, we exploit variation from ocean wind speeds using two stage least squares. We find one fewer pirate attack increases conflict by 18-30\% and causes 9.11 more land conflict deaths.
Corporate Acquisitions and Ownership of Veterinary Clinics: Local Market Power and Differing Targets
(with Clinton Neill)
Understanding how private equity acquisitions affect market concentration and subsequent consequences is critical. Yet our understanding of these in human healthcare is limited by markets that are highly concentrated. We ask how private equity acquisitions affect veterinarian markets, including concentration, revenue, and firm dynamics, and use 219 Banfield acquisitions from 1997-2020 to quantify the effects of acquisitions using two-way fixed effects. The results show that market concentration increases by 13.6\% of a standard deviation and the effect gets stronger over time as the market adjusts. This is due to an increase in number of firms that is offset by the revenue being concentrated within fewer firms.
Road Maintenance Over the Local Election Cycle
(with Margaret Bock)
Accepted, Public Choice
Professional Sports Facility Subsidies and Drug Asset Forfeiture
(with Brad Humphreys)
Revisions requested, Public Finance Review
Working Papers
Running Away to College? Domestic Terrorism and Student Location Preferences
Students exposed to violent events score lower on tests and miss school to protest, while in developing countries, exposure to armed conflict reduces height, health, test scores, and income. Exposure to violent events may also affect college choice which has implications for human capital accumulation, college success, college costs/debts, and lifetime earnings. Allowing transitory events to affect a decision with life-long consequences could be considered a sunk cost, but students may also not be forward-looking enough such that they over-value short-term feelings of safety and security. This paper leverages the randomness of foiled terrorist attacks and a large set of predictor variables to match students treated by a successful terrorist bombing during their senior year of high school to those who are in the county of a foiled terrorist bombing. Students exposed to a successful bombing are 3.28 pp less likely to attend college within 10 miles of home and 9.53 pp less likely to attend college within 50 miles of home. The most likely mechanism driving this is treated students are 8.5 pp more likely to cite getting away from home being a very important aspect of their choice, but they are also 1.1 pp less likely to apply to any other colleges, limiting their own possibilities. This violence also has consequences for regional educational attainment and debt, since students are also 4.29 pp less likely to attend in state and 4.34 pp less likely to attend a 4 year public college. This demonstrates the importance of short-term, violent transitory events for deciding where to attend college, a decision with long-term consequences.
(with Margaret Bock and Alexandre Scarcioffolo)
Although businesses bring benefits, they can also bring negative externalities. One such negative externality that has received substantial attention is heavy truck traffic, which is associated with hydraulic fracturing and agriculture. However, estimates of slaughterhouses on truck traffic and truck traffic on road damage have not been based on a causal framework. This paper estimates the aforementioned relationships using variation from a historic reason for agglomeration combined with the intuition that the pork industry prefers lower transportation costs. Estimates suggest 1 pork slaughterhouse increases truck traffic by 28.7 percent and a 10 percent increase in truck traffic leads to a 2.5 percent increase in road roughness. The IV estimates are 10X larger than OLS, suggesting that higher taxes on slaughterhouse trucks may be needed to ensure socially optimal truck traffic in equilibrium.
Healthy Guardians and Service Utilization: Evidence from Covid-19 Vaccination and Veterinary Services Visits
(with Clinton Neill)
Health is a durable capital stock which is heavily influenced by parents in adolescence. We identify the effect of parent health behavior on child service utilization in a context where fetal origins and genetics contribute nothing since they are impossible and there are no insurance spillovers. We use county-level data on Covid-19 vaccination hesitancy, takeup, and visits to veterinary services stores and multiple identification strategies. We find that a 1 percentage point (pp) increase in vaccine takeup is associated with 4.01-4.13 additional veterinary services visits, holding income, race-specific population, poverty, and unemployment constant. Selection on unobservables would have to be at least .808 that of selection on the aforementioned variables to nullify this result. Furthermore, the results are inconsistent with reopening being the main driver and two-stage least squares (in which hesitancy is used to instrument for takeup) finds similar results.