| Progressive income taxes provide a disincentive for workers to live in high productivity local labor markets, potentially leading to a spatial misallocation of labor. Relative to previous work, Chapter 1 relaxes two key assumptions; 1) that workers are perfectly mobile and 2) that workers are homogeneous. These generalizations allow us to better quantify the impact of federal income taxes, as well as analyze the associated equity-efficiency trade-off, which has not previously been studied in a spatial context. To quantify these effects, we augment an empirical spatial equilibrium model (Diamond, 2015) to incorporate taxes and estimate it using Census data. We find that the optimal federal income tax code is substantially more progressive than the current tax code, i.e. that redistribution concerns outweigh the efficiency costs of income taxes in a spatial equilibrium.;High school graduates are substantially less likely to move between states than college graduates. If moving costs increase with distance, then a stronger spatial correlation in the value of nearby locations will decrease migration rates. In Chapter 2, I document that the spatial correlation in average (log) wages, by MSA, is stronger for high school graduates. I estimate a location choice model in the spirit of McFadden(1978) and Berry, Levinsohn and Pakes(2004) to assess the quantitative importance of this empirical relationship. Counterfactual experiments examine migration rates for high school graduates as if they faced the same spatial correlation in wages as college graduates. |
