Research

(Job Market Paper)

Economic theory shows that public goods, like food pantries, will be under-provided due to the free rider problem. In this paper, I ask if donor preferences for giving to local charities lead to a mismatch between where well-funded food pantries exist and where food pantry customers live, indicating that the under-provision of charitable public goods is more extreme where charities are in highest demand. I find that food pantry customers living in high-income neighborhoods live near over 40 more food pantries than customers in low-income neighborhoods and customers in low-income neighborhoods face more restrictions, such as limits on the number of visits they are allowed, when visiting the food pantry. Using a random utility model, I identify the pantry characteristics that customers value the most, including the days-of-week and time-of day when customers prefer to visit and types of food that customers like. With the preferences derived from the model, I show that customers in high-income neighborhoods receive a higher per-visit benefit than customers in low-income neighborhoods. Finally, in a counterfactual estimation, I show that overall social welfare would increase if donors shifted their support to food pantries in lower-income areas.   [Working paper coming soon]

This paper explores the relationship between adult healthcare utilization and hydraulic fracturing on the Eagle Ford Shale formation in Texas. Using a staggered difference-in-differences estimation strategy, I find that inpatient claims for conditions plausibly related to hydraulic fracturing increase by 15 percent in shale zip codes after the entry of hydraulic fracturing, relative to zip codes outside of the shale formation. However, this result could be explained by pollution exposure from hydraulic fracturing or compositional changes related to migration. I therefore construct a novel dataset that combines public water source data with healthcare claims and oil and natural gas well permit data to compare healthcare utilization in counties with unconventional drilling within one kilometer of a public groundwater source to those with drilling farther away from their public groundwater source. I find that the increases in healthcare utilization are 13 percent larger in counties with drilling near a water source when compared to counties with drilling farther away. This result not only suggests that the increases in utilization are not likely driven by compositional changes, but also that groundwater contamination may serve as a primary pathway for unconventional drilling pollution exposure.  [Working Paper]