Research
Research
The Effects of Heat on Teachers and Their Students (Job Market Paper) [link]
Abstract: This paper provides the first evidence that extreme heat directly affects teachers, with consequences for student learning. Using matched student–teacher administrative data from North Carolina linked to high-resolution weather records, the analysis shows that older teachers, who are more sensitive to heat stress, experience greater productivity losses and higher absence rates during hot years. Students of older teachers see test score declines with each additional extreme heat day, effectively erasing the benefits of teaching experience. The cumulative impact of climate change–driven increases in heat is potentially large: equivalent to a shock on test scores comparable to Hurricane Katrina every ten years.
Electricity Disconnections, Hot Weather, and the Social Safety Net [link]
Abstract: Climate change leading to more volatile weather may have unanticipated spillover effects on take-up of government assistance programs. I use utility company-level monthly counts of involuntary electricity disconnections in Michigan from 2015 to 2021 and document how hot weather interacts with state shutoff protections and the social safety net. Using a two-way fixed effects panel regression, I find that an incremental hot weather shock (defined as a 1°F increase above 65°F) leads to a 9% increase in involuntary shutoffs. This effect is fully mitigated when state shutoff protections are in place. I then use these temperature shocks as an instrument to estimate the causal effect of involuntary shutoffs on Michigan Medicaid take-up. I find that an additional 1,000 shutoffs lead to a 17% increase over the unconditional mean in Michigan state-sponsored healthcare take-up. These findings provide evidence on how energy policy interacts with take-up of government assistance programs.
Social Safety Net Take-up and Energy Costs
Abstract: In this paper, I examine how changes in the costs of essential goods, particularly energy, affect households’ use of government support, in particular, the Supplemental Nutrition Assistance Program (SNAP). Using publicly available administrative data on SNAP participation linked with state-level prices for groceries and gasoline, I estimate how fluctuations in household expenses shape enrollment decisions. Preliminary results suggest that a ten-cent increase in gasoline prices is associated with a 2.7% rise in SNAP participation in the following month, controlling for grocery costs and local labor market conditions. These findings indicate that households respond quickly and directly to changes in volatile, unavoidable expenses. This work expands our understanding of program dynamics by showing that demand for assistance is not only determined by income and employment, but also by the shocks households face in meeting basic needs.