Abstract: | We test implications of economic geography by exploring spatial interactions among U.S. cities. We use a data set consisting of 1900–1990 metro area populations, and spatial measures including distance from the nearest larger city in a higher-tier, adjacency, and location within U.S. regions. We also date cities from their time of settlement. We find that among cities which enter the system, larger cities are more likely to locate near other cities. Moreover, older cities are more likely to have neighbors. Distance from the nearest higher-tier city is not always a significant determinant of size and growth. We find no evidence of persistent non-linear effects on urban growth of either size or distance, although distance is important for city size for some years. |