The Effects of Flooding on Migration Patterns in the U.S. States with High Storm Surge Risk

Abstract

Based on a wide range of existing literature (Tol 2004; Nordhaus 2007; Friedlingstein, Knutti, Plattner, and Soloman 2009; Haigh and Zhong 2013; and the IPCC Fifth Assessment Report 2014), this analysis accepts climate change as a result of human activity and studies one of its many ramifications. Because environmentally-induced migration triggers both economic and human costs, understanding any factors that can affect migration patterns becomes advantageous. Specifically, by using migration, flood, and environmental data at the county level, my study examines the effects of flooding on migration patterns in the 19 U.S. states with high storm surge risk. I implement difference-in-differences, fixed effects, and multilevel regressions in order to provide a reliable relationship between net migration and flooding during the years 2001 to 2011. Each regression technique captures a slightly nuanced aspect of the relationship. The difference-in-differences regressions reveal flooding likely only affects net migration after a county reaches an extreme level of flood events. The fixed effects regressions demonstrate that a slight majority of states encounter out-migration in counties when the number of flood events in the previous year increases. Finally, the multilevel analysis indicates that in general, states experience a positive level of baseline net migration in the absence of flooding, and see a negative relationship between net migration and flood events. While all three regression techniques provide intuitive results for many of the 19 states, only a select number of states yield significant results. Furthermore, many of the states still witness a positive relationship between net migration and flooding. This study presents logical explanations for any complications in the results but largely acknowledges that flooding influences out-migration in a county after it reaches a sufficiently dangerous and costly number of flood events.