From 1980 to 2000, the rise in the U.S. college/high-school wage gap coincided with both an increase in the correlation of educational levels between husbands and wives, and an increase in geographic sorting as college graduates concentrated in high wage, high rent cities. In this paper, we argue that the two phenomena -- assortative marriage and geographical sorting -- and linked in that geographical sorting makes marriages more assortative, while marriage prospect reinforces geographical sorting. We build and estimate a spatial equilibrium model of marriage and location choice that simultaneously accounts for changes in wages, rents, marriage patterns, and local populations. Using the estimated model, we quantity the degree to which marriage and location choice influences each other. Welfare analyses show that once marriage is accounted for, the educational welfare gap between college and non-college workers increased more than their (real) wage gap, while the gender welfare gap decreased more than (real) gender wage gap.
