After the Flood: Migration, Housing and Insurance Adjustment

Abstract: Do floods drive residential exit, and how do housing markets and flood insurance protection adjust in their aftermath? Combining satellite-observed inundation footprints for 24 U.S. floods with annual block-group migration flows, property-level transactions, mortgage records and NFIP policy and claims microdata for 2011–2019, I compare inundated and non-inundated block groups within the same affected county and flood event and inundated and non-inundated houses within the same block group. I find little evidence of broad residential exit following inundation. Instead, flooding localizes migration. Arrivals come from closer, non-inundated origins and departures from mapped high-risk areas do not shift toward lower-risk destinations. Housing markets capitalize the flood at the property. The same house sells for about 1.5% less with a smaller mortgage loan, the discount is highly spatial, and only inside the FEMA flood zone, where flood insurance is required, does the loan fall one-for-one with the price. These adjustments occur with little change in leverage or observable borrower characteristics. Local NFIP protection does not deepen where the losses occur. The post-flood rise in participation accrues to the non-inundated neighborhoods of the same counties, while the depth of protection falls where inundated homes lie inside the FEMA mapped flood zone. My results indicate that flooding reshuffles people, reprices housing, and spreads insurance participation without reducing exposure in the neighborhoods or counting as adaptation where it was realized.

Climate Shocks, Debt Defaults, and Investment in Climate Adaptation - Squaring an Impossible Trilemma. (with Constance de Soyres and Joanne Tan)

Abstract: Climate disasters tend to be associated with increased sovereign default risk. Countries face an “impossible trilemma”: scale up adaptation investment, keep debt sustainable amidst high borrowing costs and avoid the higher risk of default from delayed adaptation. Using a global panel of disaster event-level shocks, we find that a 1 percentage point increase in disaster related losses as a share of GDP raises the odds of sovereign default by approximately 2–3%. An additional US$1 billion in cumulative Official Development Assistance (ODA) is associated with a 0.13 point gain in a country’s adaptive capacity. Using average marginal effects and our predicted margins we then map concessional ODA finance to default probability and translate these relationships into a practical budgeting yardstick for calibrating needed ODA to sovereign default risk reduction targets. In a context of declining ODA, our findings highlight the crucial role of well-designed support in climate adaptation policies.