Formulating efficient climate policies requires estimates of the impacts of climate change. An important category of impacts are climate amenities--the value people attach to temperature and precipitation. There is a large literature that attempts to value climate amenities using the fact that climate amenities will be capitalized into wages and property values. Many of these estimates assume that people are perfectly mobile and are based on estimates of national hedonic wage and property value functions. These functions will yield biased estimates of consumers' willingness-to-pay if consumers are not in locational equilibrium due to search or migration costs or if markets take time to adjust.

I provide estimates of the value of climate amenities in the US using a discrete model of residential location choice. I model the location choices of over 400,000 households who changed metropolitan statistical areas between 1995 and 2000 using the 5% PUMS data from the Census. To avoid making equilibrium assumptions, I face the migrants with the market conditions in each MSA.

The empirical model is motivated by a Random Utility Model framework, which posits that the utility that a household derives from living in an MSA depends on climate amenities along with earnings potential, housing costs and location-specific amenities. Households choose the MSA where they derive the maximum utility. The model is estimated using a two step procedure (Bayer, Keohane and Timmins, 2006). In the first stage, location-specific constants are estimated together with other parameters of the utility function. In the second stage, these location-specific intercepts are regressed on location-specific amenities to estimate the average utility attached to these amenities.

The dissertation estimates the marginal rate of substitution between climate variables and income. The results show that households facing an average winter temperature of 37 degrees Fahrenheit are willing to pay approximately about 3% of their income for an increase in average winter temperature by one degree. Willingness to pay to raise summer precipitation by an inch from a level of about 11 inches is roughly 3% of their income. The study also provides estimates of the quality of life in 297 Metropolitan Areas.