The Determinants and Welfare Implications of US Workers’ Diverging Location Choices by Skill: 1980-2000 JOB MARKET PAPER

From 1980 to 2000, the substantial rise in the U.S. college-high school graduate wage gap coincided with an increase in geographic sorting as college graduates increasingly concentrated in high wage, high rent metropolitan areas, relative to lower skill workers. The increase in wage inequality may not reflect a similar increase in well-being inequality because high and low skill workers increasingly paid different housing costs and consumed different local amenities. This paper examines the determinants and welfare implications of the increased geographic skill sorting. I estimate a structural spatial equilibrium model of local labor demand, housing supply, labor supply, and amenity levels. The model allows local amenity and productivity levels to endogenously respond to a city’s skill-mix. I identify the model parameters using local labor demand changes driven by variation in cities’ industry mixes and interactions of these labor demand shocks with determinants of housing supply (land use regulations and land availability). The GMM estimates indicate that cross-city changes in firms’demands for high and low skill labor were the underlying forces of the increase in geographic skill sorting. An increase in labor demand for college relative to non-college workers increases a city’s college employment share, which then endogenously raises the local productivity of all workers and improves local amenities. Local wage and amenity growth generates in-migration, driving up rents. My estimates show that low skill workers are less willing to pay high housing costs to live in high-amenity cities, leading them to elect more affordable, low-amenity cities. I find that the combined effects of changes in cities’wages, rents, and endogenous amenities increased well-being inequality between high school and college graduates by a significantly larger amount than would be suggested by the increase in the college wage gap alone. ∗I am very grateful to my advisors Edward Glaeser, Lawrence Katz, and Ariel Pakes for their guidance and support. I also thank Nikhil Agarwal, Adam Guren, and participants at the Harvard Labor and Industrial Organization Workshops. I acknowledge support from a National Science Foundation Graduate Research Fellowship. The computations in this paper were run on the Odyssey cluster supported by the FAS Science Division Research Computing Group at Harvard University.