Storms and Jobs: The Effect of Hurricanes on Individuals' Employment and Earnings over the Long Term

Abstract

We study the responsiveness of individuals' employment and earnings to the damages and disruption caused by Hurricanes Katrina and Rita, which struck the U.S. Gulf Coast in 2005. Our analysis is based on individual-level survey and administrative data that tracks workers over time, both in the immediate aftermath of the storm and over a seven-year period. For individuals who were employed at the time of the storm, we estimate models that compare the evolution of earnings for individuals who resided in storm-affected areas and individuals who resided in a set of control counties with pre-storm characteristics similar to those of the storm-affected areas prior to the storm. We find that, on average, the storms reduced the earnings of affected individuals during the first year after the storm. These losses reflect various aspects of the shortrun disruption caused by the hurricanes, including job separations, migration to other areas, and business contractions. Starting in the third year after the storms, however, we estimate that the storms increased the quarterly earnings of affected individuals. We provide evidence that the long-term earnings gains experienced by affected individuals were the result of differences in wage growth between the affected areas and the control areas, due to reduced labor supply and increased labor demand, especially in sectors related to rebuilding. Despite short-term earnings losses due to an increased rate of non-employment, we find a net increase in average quarterly earnings among affected individuals over the entire post-storm period. However, subgroups with large and persistent earnings losses after the storms had a net decrease in average quarterly earnings over the seven-year period due to the storms.

Long-Run Consequences of Exposure to Natural Disasters

“…Our findings suggest that the consequences of hurricanes, and potentially other natural disasters, can include costs such as reduced individual-level productivity, that do not manifest themselves until long after the immediate aftermath of the disaster. Our results also contrast some of the estimates of the short-and medium-run e¯ects of hurricane exposure on the adult population, such as Deryugina et al (2014), Groen et al (2016) and Ho et al (2016), who document positive e¯ects of exposure and associate the results with positive mortality selection, increased labor demand during the reconstruction, and the inflow of relief to the a¯ected locations. In the context of this study, where the subject of the "experiment" is a relatively fragile population, storms were unanticipated, and relief e¯orts were minimal, we find unambiguously negative long-term consequences.…”

“…Our analysis also provides a counterpoint to studies that find short-run benefits for survivors of natural disasters and a¯ected communities when examining the contemporaneous implications of positive mortality selection (Ho et al 2016) and labor market responses to natural disasters (Deryugina et al 2014;Groen et al 2016). Similarly, potential medium-run gains can arise due to induced migration (Nakamura et al 2016), and the transfer of wealth that occurs when disaster relief flows into a¯ected regions (Deryugina 2016).…”

Disasters in the United States: Frequency, Costs, and Compensation

“…A recent analysis by Groen, Kutzbach, and Polivka (2015) examines the effects of Hurricanes Katrina and Rita on individual labor market participants in Alabama, Louisiana, Mississippi, and Texas. The focus of their analysis is the changes in employment and earnings of individuals affected by hurricane damage.…”

Understanding and Mitigating the Impacts of Massive Relocations Due to Disasters

“…Moreover, Groen et al (2016) indicate that those who relocated suffered net earning losses, whereas those who remained in the affected area benefited from increased wages Bdue to reduced labor supply and increased labor demand^in hurricane-affected areas.…”