Sorting, Displacement, and Welfare Gains from Non-Local Firms

How much do consumers in a city benefit from the presence of in-person service firms originating in other cities? Although establishments of non-local firms are typically larger and likely more productive than their local counterparts, their welfare benefits may be smaller than these differences suggest. Because consumption in these industries is highly localized, non-local firms primarily benefit nearby consumers, and their larger size partly reflects their sorting into high-demand locations. Their entry also displaces local firms toward locations with lower demand, reducing the number of consumers served by those firms. After documenting these patterns across U.S. cities, I develop a quantitative urban model in which in-person service firms sort across heterogeneous locations within a city and compete for a limited supply of retail floor space. Quantifying the model for the 50 largest U.S. commuting zones, I find that the welfare gains from non-local firms are less than 1% on average. A decomposition shows that displacement offsets over 50% of the potential gains from improvements in the productivity distribution of active firms on average.