Do partisans sort? In particular, does residential access to public goods explain the contours of political behavior? I offer a test of the classic Tiebout model of residential sorting by studying revealed political preferences around the edges of all cities in the United States. Although prior research shows that voters do not tend to sort into neighborhoods based on partisan concerns alone, I argue that partisan differences in the demand for public goods can shape political outcomes at the borders of the cities that provide those services. My analysis is twofold; first, I take a regression-discontinuity approach to measuring geospatial variation in election outcomes at city limits; subsequently, I use voter-file-level data on U.S. voters that changed addresses to estimate the differential propensity between Democrats and Republicans to choose urban or rural residential environments. I show that support for Democratic candidates in recent presidential elections was roughly 3 percentage points lower in areas just outside municipal boundaries. This discontinuity holds in small cities as well as larger metropolitan areas. Preliminary evidence suggests that meaningful partisan differences exist in the decision to move to an incorporated city. This study is the first to examine the universe of U.S. cities and uncover sharp gaps around them in political outcomes, shedding light on the relationship between local political institutions, partisan geography, and the urban-rural divide in general.