Shifrah's job market paper: Rebuild or Relocate? Recovery after Natural Disasters

Abstract

This paper studies the distributional effects of natural disasters and the impact of post-disaster policies. Using flight data and original survey evidence from Puerto Rico after Hurricane Maria, I document that 7% of the population migrated to the mainland U.S. in the aftermath, though many later returned. High migration costs prevented an even larger number from leaving. Those who stayed faced widespread damages to housing, infrastructure, and the local economy. I find that age, wealth, and housing tenure shaped post-disaster choices: many homeowners with severe property damage defaulted on their mortgages, wealthier homeowners rebuilt with government assistance, and younger renters were most likely to leave the island permanently. These empirical findings inform a dynamic equilibrium model of migration, housing, and infrastructure with heterogeneous households. Homeowners with property damage experience significant welfare losses from direct reductions in home equity and housing consumption, while renters and undamaged homeowners also face welfare declines from infrastructure destruction and through general equilibrium price movements. Local infrastructure investment is a cost-effective policy, due to complementarities with both housing consumption and production. In contrast, rebuilding subsidies for homeowners, though effective in preventing mortgage defaults and mitigating housing shortages, are not cost-effective. Homeowners with property damage value these subsidies below cost because they are not guaranteed and do not provide any payout if the home is foreclosed on or sold. More flexible policies could yield large welfare improvements for similar costs.