Publications:
The Impact of Dobbs v. Jackson on Medical School Applications. Economics Letters, 241 (2024). https://doi.org/10.1016/j.econlet.2024.111815.
I examine the relative impact of the Dobbs decision – which allowed states to restrict abortion access – and find schools in states with abortion bans saw a 0.65% increase in the 2022 share of women’s applications and a 1.17% increase in the 2023 share, but no measurable changes in the out-of-state share in either year. This is plausibly driven by applicants interested in an OB/GYN specialty.
Higher education: The impact of recreational marijuana on college applications. Contemporary Economic Policy, 1–19. https://doi.org/10.1111/coep.12633 (with Chris Blake and Danna Thomas)
Using a two-way fixed effects difference-in-differences model, we investigate the effects of local recreational marijuana (RMJ) policy changes on college applications and find that the three largest state public schools reaped, on average, an almost 54% increase in applications. This increase does not appear to come solely from low-ability students as both first and third quartiles of admitted student composite SAT scores to the largest three public schools do not decrease. Rather, they both increase by almost 3.8% though these estimates are not statistically significant. Robust difference-in-difference and event study models support the signs and magnitudes of these gains and show they diminish over time.
Uncertainty, learning, and local opposition to hydraulic fracturing. Resource and Energy Economics 55 (2019): 102-123, (with Dale Manning, Terry Iverson, and Harvey Cutler)
New extraction technologies, including hydraulic fracturing (fracking), have increased fossil fuel reserves in the United States. Despite local economic benefits, many jurisdictions have adopted bans on fracking. We develop a dynamic learning model parameterized with a computable general equilibrium (CGE) model to quantify the associated quasi-option value (QOV), and to explore if uncertainty about environmental damages with the potential to learn can rationalize such bans. The model is calibrated to a representative municipality in Colorado, the site of several fracking bans. With plausible damages, we find that the QOV increases the incentive to delay drilling within a range of energy prices. The results suggest that improving the ability to learn about fracking impacts could increase the prevalence of bans in the short run and lead to better policymaking over time. Incorporating CGE output into a dynamic learning framework permits calculation of the locality-wide QOV associated with a range of policies.
Working Papers:
Risky business: the effect of liberalized cannabis policy on risky teen sexual behavior (with Roy Choudhury) [under review]
We investigate the impact of recreational marijuana sales (RMJ) on teen cannabis use, sexual activity, risky activity, and risk sexual activity using data from the Youth Risk Behavior Surveillance System (YRBSS hereafter) and an event study approach. We find that RMJ are linked to higher rates of marijuana experimentation, with increased use among females and earlier initiation ages for males. Among females, marijuana use correlates with increased frequency of sexual activity and higher levels of risky sexual behavior, including younger ages of initiation, compared to non-users who report lower-risk profiles. For males, marijuana use shows no significant impact on sexual behavior, although non-users report more frequent risky sexual activities. RMJ are also associated with higher teen pregnancy rates and increased incidences of chlamydia and gonorrhea in affected states.
Budding enrollment: the impact of cannabis policy on US colleges and universities (with Danna Thomas) [under review]
We examine the impact of liberalized cannabis policy on US college and university enrollments. We find no statistically significant impact across all schools but early gains to the largest private school and latent gains to the largest public schools. This appears to be driven by out-of-state enrollment with the largest surge coming from border states. On average, the largest three public schools in a state saw a 4.2% gain, with 14.3% coming from out-of-state and 3.3% from border states. The largest public school had a 2.8% increase with 9.3% from out-of-state, and 17% from border states. The largest private school in the state saw a 5.2% increase in enrollment, with 5.7% coming from out-of-state, and 10% coming from border states.
Hold My Beer! The Effects of Alcohol on Risk Preferences in a College Bar (with Bentley Coffey and Elizabeth Watson) [under review]
We performed a field experiment to assess the impact of alcohol on the risky decisions of young adults in a ‘party’ environment – a college bar. Risk preferences were elicited using colorized Holt-Laury lottery choices to minimize cognitive load and duration. Because the Constant Relative Risk Aversion (CRRA) utility function is not formally identified with the standard Random Utility Model, we use a Random Parameter Model (RPM). To fully harness the limited information contained in the small samples typical of field experiment, we develop an Empirical Bayes (EB) approach for estimating such RPMs where the hyperparameters' prior distribution is intuitively set by pooled data moments in a Generalized Method of Moments (GMM) framework. This novel approach produces the precise individual-specific estimates of risk aversion that we need to analyze the covariation of risk aversion with Blood Alcohol Content (BAC). Results indicate the data favors a quadratic, rather than linear, relationship where risk aversion actually increases over low levels of blood alcohol content but then dramatically decreases at higher levels. On average, risk aversion reaches its maximum at BAC = 0.05 and preferences actually become risk seeking when BAC exceeds 0.16. A back-of-the-envelope parameterization of drunk driving deterrence policy implies that 48% of our participants would be willing to take the risk of driving (using our risk estimates), had measured BACs > 0.08, but self-assess their BAC as below the legal limit.
Green runs: Colorado cannabis sales during ski season (with Isabelle Bassock) [under review]
This study examines the impact of ski tourism on Colorado’s retail marijuana (RMJ) sales, leveraging county-level sales data and ski resort characteristics. Using a two-way fixed effects model, we analyze seasonal variations in RMJ sales across counties with differing ski resort acreages. Our findings indicate that ski resort counties experience significant increases in RMJ sales during peak ski months (December–March), with the most prominent effects observed in counties with the largest ski resorts. The largest ski resort county – Summit County – sees an estimated seasonal increase of $4.35 million in RMJ sales. We estimate the isolated extra-state ski tourism effects and find that while some sales may shift from non-ski counties, the net increase in RMJ sales for the top ski resort counties remains economically significant. Our set of results suggest ski tourism increases Colorado’s retail marijuana sales between $4.5M and $9.4M per year.
Estimating the Income Elasticity of Demand for Recreational Marijuana: Evidence from the Child Tax Credit. (with Jason DeBacker) [in progress]
Using high frequency data on consumers and a quasi-natural experiment, we estimate the income elasticity of demand for recreational marijuana. In 2021, many American taxpayers found advance payments of the child tax credit direct deposited into their bank accounts around the 15th of each month. We employ fixed effects regressions for direct estimation and a differences-in-differences design, comparing sales just before and after the 15th of the month across Public Use Microdata Areas in Washington state to estimate how these income shocks affected sales of recreational marijuana. Across all income groups we find that short-run income elasticity of demand (IED) to be 1.19. We estimate IED by income quintile and find the lowest quintile has the highest IED. This appears to decrease, but stays positive, as income increases. Our results have implications for policies that aim to affect, or are affected by, demand for marijuana such as excise tax revenue.