Population Matters When Modeling Hurricane Fatalities

Abstract

Jung et al. (2014) find gender bias leads to misperception of hurricane risk in both experimental and historical evidence, concluding that this bias has led to increased fatalities in the historical record. We affirm the rich literature on gender bias and its impacts. However, we argue that the empirical analysis leading to the Jung et al.’s stylized fact suffers from endogeneity. Once this issue is addressed, we find the previous results to be of questionable robustness. While there may be a gender bias in initial perception, especially in a limited-information experimental setting, historical evidence does not indicate gender bias in the number of deaths caused by hurricanes.