The Economic Consequences of Hurricanes

This dissertation focuses on the economic effects of hurricanes in the United States. Chapter 1 investigates the long-run consequences of prenatal hurricane exposure. Environmental conditions during gestation meaningfully affect human capital at birth, resulting in long-run economic consequences. We study the later-life effects of prenatal hurricane exposure for more than 1.4 million individuals born in the U.S. between 1980 and 1994, using administrative and survey data from the IRS and U.S. Census Bureau combined with satellite-derived hurricane data. We estimate that plausibly exogenous exposure to dangerous hurricane winds during gestation has no effects on later-life economic outcomes when hurricanes are forecasted, despite evidence from past work that hurricanes adversely affect birth outcomes. We also estimate that prenatal exposure to hurricanes outside of the forecasted area is associated with substantial losses: a $4,500 reduction in yearly adult earnings and a 5.4 percentile reduction in economic mobility. Back-of-the-envelope calculations suggest that in the absence of forecasts, total earnings lost from prenatal exposure would be $33 billion over the sample period. This highlights the value of advanced information in mitigating the translation of environmental shocks into economic damages.

The second chapter, joint with Jonathan Colmer and John Voorheis, studies the distributional effects of hurricanes. We document that there has been no systematic movement towards or away from riskier areas over the last two decades and that non-Hispanic Black and less-educated individuals are significantly more likely to be exposed to hurricanes than other Americans. Second, we estimate the dynamic causal effects of hurricane exposure, finding that individuals exposed to hurricane winds experience modest reductions in total income and earnings in the immediate aftermath of a hurricane, with a return to trend within five years. Earnings losses are largely offset through increases in transfer payments. However, our results suggest that Black and Hispanic Americans are disproportionately affected by hurricane exposure, particularly low and high income individuals. We estimate that these individuals experience reductions in total income and earnings that are six times larger than the average effects.

The third chapter studies the role of past experience with hurricanes. The relationship between hurricane exposure and economic damages will depend on decisions made, and these decisions cannot be perfectly optimized without perfect information. I study the role of hurricane experience by combining 37 years of satellite-derived hurricane exposure data with county-level income and defensive investments. I estimate that hurricane exposure in inexperienced counties reduces total income by about 1%, but past experience with hurricanes reduces this effect by about 30% per event. In addition, marginal defensive investments increase by about 18% per event experienced. These results may suggest that hurricane experience prompts learning and that people initially underinvest in self-protection against hurricanes.