Ongoing Research
Ongoing Research
Sorting Over Wildfire Risk Information - with Michael Fogarty
We study how information about climate risk changes household beliefs, housing demand, and residential sorting in equilibrium. We use the introduction of a wildfire risk disclosure policy in California as a shock to beliefs about wildfire risk that leaves the underlying physical hazard unchanged, allowing us to identify how changes in beliefs affect location choices and housing prices. We embed a Bayesian model of belief updating in a structural model of neighborhood choice in which households have heterogeneous preferences over housing prices and wildfire risk. Our estimates indicate that before the policy, households held diffuse beliefs about neighborhood wildfire risk, and disclosure widened the gap in believed annual burn probability between disclosed and non-disclosed neighborhoods by 0.47 percentage points. This change in beliefs re-sorts households by their sensitivity to wildfire risk: mortgage-financed buyers of primary homes, the most risk-sensitive group, shift away from disclosed neighborhoods, while buyers of non-primary homes, especially cash buyers, move in, and prices in disclosed neighborhoods fall relative to non-disclosed neighborhoods. Relative to a no-disclosure counterfactual simulation, the gap in average annual fire exposure between the neighborhoods chosen by cash buyers of non-primary homes and mortgage-financed buyers of primary homes increases by 29%. Alternative counterfactual exercises show that a finer-grained risk score achieve a similar or larger reallocation of risk with substantially less household re-sorting. Our findings show that information can reshape beliefs about wildfire risk, housing market outcomes, and the incidence of wildfire exposure, even when the underlying physical hazard is unchanged.
Home Financing Responses to Wildfire Risk Information and Insurance Availability (JMP) - with Michael Fogarty
Wildfire risk has become a key force shaping California’s housing market. We study home buyer responses to a 2021 reform that requires home sellers in California to disclose parcel-level wildfire risk and potential homeowners insurance constraints in designated fire-hazard zones. We show that following implementation of the risk disclosures, the share of all-cash purchases in fire hazard zones increased by 14 percentage points relative to mortgage-financed purchases. Concurrent with the rise in cash purchases, owner-occupancy fell by 13 percentage points, suggesting a reallocation of high-risk homes from primary residences toward second homes and vacation properties. The financing shifts are largest in state-managed wildfire areas, in older homes subject to tighter underwriting, and in ZIP codes with greater reliance on California’s homeowners insurer of last resort. While mortgage lenders require insurance, all-cash buyers may delay purchasing insurance, underinsure, or forgo insurance altogether. Thus, the shift from mortgage-financed to all-cash transactions suggests a reallocation of climate risk away from insurers and towards individual households.
Following Conceptions, Not Births: Wildfire Smoke, Selective Fetal Loss, and Birth Outcomes - with Jenna Nobles and Marcos A. Rangel
Wildfires have reduced air quality across the U.S., with implications for multiple dimensions of health. We examine the intergenerational effects of wildfire smoke by estimating its impact on cohorts of infants exposed during gestation. We advance research on early-life insults by recognizing that environmental factors shape cohort health through two competing channels: by determining which pregnancies are achieved and survive to live birth (selection), and by affecting the health of those that survive (scarring). We build georeferenced data on pregnancy outcomes by combining vital statistics records, menstrual- and pregnancy-tracking app data, and satellite and environmental data. We use weekly shifts in fire locations and wind direction to generate within-location variation in smoke exposure, and confirm that upwind fires reduce air quality relative to fires in other directions. We show that upwind fires reduce the probability of pregnancy implantation, increase pregnancy loss, and, among infants exposed in the first trimester, increase preterm birth. Simulations conservatively calibrated to documented parameters show that accounting for prenatal selection can make pollution effects on birth outcomes up to 35% larger. The findings reveal the substantial intergenerational reach of fire pollution. They also highlight pregnancy success as a relevant economic outcome: conception failure and pregnancy loss have welfare consequences for families and societies, and are sensitive to environmental hazards. By integrating data on early pregnancy, the study provides some of the first empirical evidence on the joint processes of selection and scarring that shape cohort health, even before birth.
Covid Exposure, Conception, and Pregnancy - with Jenna Nobles and Marcos A. Rangel
The COVID-19 pandemic has taken a large toll on population health and well-being. We examine the consequences of maternal exposure on conception probabilities and pregnancy loss. Using CDC county-level infection rates, we identify county-specific spikes in COVID-19 activity and leverage the fact that, although infections peaked nationally in fall 2021, the timing of local peaks varied across places. Linking these data to menstrual and pregnancy tracking app records, we are in the process of estimating event-study models centered on each county’s peak to assess potential effects on two margins of reproductive health: time to conception and early pregnancy loss. While prior research has emphasized COVID-19’s impacts on birth outcomes, we advance this literature by studying how the pandemic shaped the ability to conceive and which pregnancies survived to become live births. Understanding determinants of pregnancy survival has important intergenerational implications for population health. Our findings underscore the need for ongoing attention to maternal and fetal health during pandemics and public health crises more broadly.
with Carolina Barbosa, Andrew Breck, Grant King, Yoojin Kook, Amanda Honeycutt, and Dominick Esposito
2022, Journal of Comparative Effectiveness Research, 11 (2)
Impact analysis of expanding anti-TNF therapy for Crohn’s disease
with Amanda Honeycutt, Andrew Breck, and Dominick Esposito
2022, Journal of Comparative Effectiveness Research, 11 (2)
with Laurel Bates, Amanda Honeycutt, Timothy A Green, and Paul G Farnham
2021, JAIDS Journal of Acquired Immune Deficiency Syndromes, 88 (4), e28–e30
with Heidi S Harvie, Vivian W Sung, Simon J Neuwahl, Amanda A Honeycutt, Isuzu Meyer, Christopher J Chermansky, Shawn Menefee, Whitney K Hendrickson, Gena C Dunivan, Donna Mazloomdoost, and Marie G Gantz
2021, American Journal of Obstetrics and Gynecology, 225 (6), 651–e1
Validation of the Prevention Impacts Simulation Model (PRISM)
with Benjamin Yarnoff, Amanda Honeycutt, Christina Bradley, Olga Khavjou, Laurel Bates, Rachel Kaufmann, Lawrence Barker, and Peter Briss
2021, Preventing Chronic Disease, 18