Perilous Flood Risk Assessments
Presented by:
Joakim Weill
University of California, Davis
Friday, January 14, 2022
12:00 pm-1:30 pm
Taylor-Hibbard Seminar Room (Rm103)
Online - https://go.wisc.edu/jmydkq
This paper investigates the impacts of providing new risk information and lowering information access costs on demand for disaster insurance. I compile new spatial data on a 15-year roll-out of digital flood maps provided by the National Flood Insurance Program and which depict the extent of zones subject to flooding across the United States. A priori, the impact of updated maps on household demand for flood insurance is ambiguous given competing effects of changing risk information, its cost of access, and insurance prices. Despite increasing flood risk and federal government interest in expanding insurance coverage, I find new maps caused demand to fall by 70,000 insurance policies overall. Using heterogeneity-robust difference-in-differences models, I estimate that households’ demand for insurance is highly sensitive to changes in the floodplain boundaries depicted in the maps, but that the mode of information provision (digital versus paper-based) does not impact demand. I also find that households respond to changes in nearby flood risk designations, suggesting spatial spillovers in beliefs. Leveraging a novel local synthetic control approach to quantify the distributional impacts of risk mapping, I estimate that updated maps caused demand to fall more in neighborhoods with a higher share of African Americans. This pattern is largely driven by rezoning these households outside of the high-risk floodplains. Finally, I compare federal maps with state-of-the-art estimates of flood risk and specify a model of insurance demand to quantify the welfare impacts of improved risk assessments.
Download File(s)