Jared C. Carbone, Sul-Ki Lee and Yuzhou Shen
We estimate how climate amenities influence where households decide to reside in the United States with two main objectives in mind: (i) to produce estimates with sufficient demographic detail to inform demographic population projections for use in climate impact analysis; (ii) to study the robustness of estimates from the existing literature. With respect to the former goal, we find important differences in job-related migration motives by age group and in the overall propensity to migrate among households with children. With respect to the latter aim, our framework shares a common methodological approach with other, recent attempts to recover climate preferences, allowing us to explore the consequences of a number of key assumptions in a systematic manner. Consistent with the existing literature, we find relatively robust estimates of the impact of the frequency of extreme heat days on household location decisions. The impacts of other common measures of climate, including the frequency of extreme cold days, average summer and winter temperatures, annual precipitation, humidity and frequency of sunshine, do not show a strong enough signal in the data to be estimated with precision.