Publications
Cardiovascular Mortality and Leaded Aviation Fuel: Evidence from Piston-Engine Air Traffic in North Carolina (with Heather Klemick, Dennis Guignet, Ron Shadbegian and Linda Bui).
[ Abstract | Published Version ]
Leaded fuel used by piston-engine aircraft is the largest source of airborne lead emissions in the United States. Previous studies have found higher blood lead levels in children living near airports where leaded aviation fuel is used. However, little is known about the health effects on adults. This study is the first to examine the association between exposure to aircraft operations that use leaded aviation fuel and adult cardiovascular mortality. We estimated the association between annual piston-engine air traffic and cardiovascular mortality among adults age 65 and older near 40 North Carolina airports during 2000 to 2017. We used several strategies to minimize the potential for bias due to omitted variables and confounding from other health hazards at airports, including coarsened exact matching, location-specific intercepts, and adjustment for jet-engine and other air traffic that does not use leaded fuel. Our findings are mixed but suggestive of adverse effects. We found higher rates of cardiovascular mortality within a few kilometers downwind of single- and multi-runway airports, though these results are not always statistically significant. We also found significantly higher cardiovascular mortality rates within a few kilometers and downwind of singlerunway airports in years with more piston-engine air traffic. We did not consistently find a statistically significant association between cardiovascular mortality rates and piston-engine air traffic near multi-runway airports, where there was greater uncertainty in our measure of the distance between populations and aviation exposures. These results suggest that (i) reducing lead emissions from aviation could yield health benefits for adults, and (ii) more refined data are needed to obtain more precise estimates of these benefits.
Working Papers
The Politics of (Cause of) Death (with Rebecca Thornton)
[ Abstract ]
In 20 U.S. states, coroners—often without formal medical training—are elected along partisan lines, potentially introducing bias into vital statistics. Using a regression discontinuity design and a novel dataset of coroner elections (2006–2023) linked to county-level death records, we estimate the causal effect of a Democrat versus Republican coroner on politically sensitive causes of death, including opioids, COVID-19, and gunshot fatalities. Our sample includes only contested, partisan coroner elections. Preliminary results show no strong discontinuities in cause-specific mortality around the electoral threshold, though opioid and unknown causes exhibit greater variation under Democratic coroners.
The Effect of Privatization on Prison Populations (with Ryan Quandt and Christopher Sharp)
[ Abstract ]
Using a panel dataset of all public and private state prisons, we examine whether the introduction of private prisons affects sentencing decisions. We utilize a doubly robust difference-in-differences design as well as an event study using to find that private prison openings lead to a modest but statistically insignificant increase in sentence lengths, concentrated entirely in the first two months. Public prison openings show no such effect.