This paper studies the effect of spatial sorting on inequality through two channels: spatial differences in technology and the endogenous organization of production. First, I document a new fact on the spatial differences in the organization of production. The number of workers per manager is decreasing in city size, overall and within industries. I develop and quantify a model of a system of cities where workers of different skills organize in production teams. The model yields continuous wage distributions in cities of different sizes that resemble the data. I find that technology differs across cities in its productivity but also in its complexity, so there are no incentives for it to diffuse across cities. I then use the model to evaluate two local policies that are designed to address income inequality: a minimum wage and a housing subsidy. I find that a revenue neutral housing subsidy is more effective than a minimum wage at reducing inequality, measured by the variance of log wages.