research

The Causes and Consequences of Urban Heat Islands (Job Market Paper)

with Jonathan Colmer and John Voorheis

This paper studies the causes and consequences of urban heat islands. Combining new administrative data with a novel proxy for experienced temperature at the neighborhood scale, we show that a hot day increases mortality by six additional deaths per 100,000 for the elderly population living in neighborhoods with a high concentration of impervious surfaces, relative to the median. These patterns hold even within counties and cannot be explained by selection. Moreover, the increase in mortality among elderly Black Americans following a hot day is three times that of elderly White Americans, and half of this disparity can be attributed to Black individuals living in more impervious neighborhoods. We then present suggestive evidence that imperviousness is driven by density zoning policies, and document that the racial incidence of density is reflected in a long historical process since the Great Migration.

Where Does Air Quality Matter? New Evidence from the Housing Market

with Eleanor Krause

The hedonic valuation approach often estimates demand for amenities from housing prices. We show that when housing supply is elastic, increased demand is met through quantity expansions, attenuating price capitalization and biasing hedonic estimates downward. Consistent with this, we find that declines in PM2.5 concentrations yield larger price effects in markets with inelastic housing supply and larger quantity effects in elastic markets. A spatial equilibrium model demonstrates that the traditional hedonic price coefficient reflects demand for an amenity attenuated by the supply elasticity. Incorporating elasticities into the hedonic framework increases the estimated benefits of PM2.5 reductions by over 12 percent.

Individual-Level Heat Disparities in the United States

with Jonathan Colmer and John Voorheis (draft available upon request)

Temperatures can vary substantially over short distances due to differences in land cover—a phenomenon known as the urban heat island effect. Combining 20 years of high-resolution satellite-derived land surface temperature data, measured over 9 billion cells, with new individual-level data containing detailed demographic, residential, and economic information for every citizen and permanent resident of the contiguous United States between 2000 and 2019, we provide the most comprehensive and systematic evaluation of surface temperature disparities to date. We document that within the same commuting-zone, Non-Hispanic Black individuals are exposed to higher surface temperatures than Non-Hispanic White individuals at every percentile of the income distribution. We show that individual economic circumstances can account for approximately 30 percent of the Black-White temperature gap, providing suggestive evidence that race rather than class is more important in determining heat disparities in the United States.