"A Storm Brewing: Costs of Climate-Induced Pollution in Drinking Water" [Working Paper]
This study assesses the health risks of water pollution caused by tropical cyclones which have been increasing in intensity with climate change. Using individual birth records and water quality measurements from Florida between 2013 and 2021, I estimate the effect of cyclones impacting drinking water sources on water pollution and infant health. Identification of the causal effect relies on the random timing and path of storms, which allows me to compare exposed and unexposed births for the same mother. I then differentiate between residential and source exposure to isolate water pollution as a mechanism by which cyclone exposure affects health. My results imply that source exposure to cyclones increased detection of coliform bacteria by 17% and preterm births by 5%. Results are robust to various threats to identification such as the comparability of exposed and unexposed mothers or selection based upon exposure. My estimates imply a health cost of exposure of between $296 and $322 million a year from 2013 - 2021. These findings illustrate the extent to which climate change can threaten health through its effect on drinking water quality.
"Pre-Existing Air Quality Conditions: The Case of Ozone and the NOx Budget Program." (with Lala Ma, Lucija Muehlenbachs, and Anicet Sagbo).
The air quality benefits of a policy depend not only on the direct impact of the targeted pollutant, but also on indirect reactions with other pollutants in the atmosphere. The reaction to create new pollutants, so-called “secondary” pollutants, are less often studied in the economics literature, even though these secondary pollutants and their health consequences may significantly alter the estimated monetary impact of the policy. We revisit the NOx Budget Program, to examine how pre-existing levels of air pollutants resulted in heterogeneous impacts from the policy. Using data on ambient air quality, we find the impact of the program on reducing ground-level ozone varied by five fold, depending on pre-existing levels of nitrogen oxides (NOx) and volatile organic compounds (VOCs). One complication is that the monitoring of VOCs historically has been lacking. We provide insights on using proxies for VOCs from data more readily available on traffic density, population density, and forest cover. Then by examining records from the CDC on mortality, we investigate the heterogeneous relationship between pollutant reduction and health outcomes.
"Housing Supply Responses to Floodplain Regulation." (with Sarah Brown)
We study the effect of regulation intended to promote resilience to flooding on the decisions of developers in Houston Texas using a dynamic model of housing supply. This regulation acts on both the intensive and extensive margin in increasing development costs in floodplains by raising the required elevation of houses in the 100-year floodplain and instituting a minimum elevation for houses in the 500-year floodplain. Our model will include forward-looking developers who use expectations of future prices and costs to make decisions on the type and timing of construction to initiate. To estimate this model, we combine CoreLogic individual housing transaction data with Houston county accessor data on the development status of lots in Houston over time. The results from this analysis will inform our understanding of how developers respond to regulation that promotes resiliency to climate threats. Understanding developer incentives is critical to guide construction in a way that improves communities resilience' to climate change.