The Spatial Variation of Geographic Polarization in the United States

Abstract

Introduction: In 1977, sociologists Knoke and Henry predicted that "future trends suggest a diminishing political difference between rural and urban populations" (51). Given recent events, their prognostication could not be further from the truth. The so-called "rural-urban divide" has received increased attention in American politics in recent years. Politicians, pundits, and political scientists often remark on the stark differences in political opinion and behavior of rural and urban citizens. These differences can be seen in any recent electoral map; small clusters of dark blue Democratic strongholds are surrounded by vast swaths of bright red Republican territory. Recent work in political geography has documented the extent to which rural and urban Americans differ in their political views and behavior, often examining the question at the national level. But it is possible that this political polarization differs across the country. For example, rural and urban Texans may be politically similar while rural and urban Wisconsinites are politically distinct. Uncovering this spatial heterogeneity is important because national-level measures of geographic polarization may understate or overstate the degree of polarization in particular areas. For example, if rural and urban Texans are politically similar, examining geographic polarization at the national level will overstate the degree to which Texans are polarized. Because of this gap in the literature, this project aims to answer the question: how does geographic political polarization vary across the country? The project provides greater understanding of the actual polarization on the ground by recognizing and measuring the spatial heterogeneity inherent in American politics. Both elites and ordinary citizens are concerned about polarization, but discuss it as if it is spatially uniform, equally strong throughout the country. In order to address political polarization, we ought to know where it is relatively high and where it is low. My project uses precinct-level presidential election results and voter registration data to estimate the degree of political polarization in each state. In so doing, I can identify spatial patterns of geographic polarization, determining which states are relatively highly polarized and which are relatively non-polarized. While this is a descriptive project and I do not have any strong predictions about the results, there are a few plausible outcomes. For example, states occupying a wider range along the rural-urban continuum may be more geographically polarized, as the state parties are better able to monopolize rural or urban places. In states that are more geographically homogenous, political cleavages cannot rest along geographic lines. In addition, southern states may be less geographically polarized, as they still contain a substantial population of rural blacks, who (unlike their white counterparts) overwhelmingly support Democrats. Finally, states with the largest media markets, such as New York and California, may be more geographically polarized than other states. For example, the media industry may attract political liberals to these states’ large cities, increasing geographic polarization within the state via residential sorting. If these states are highly polarized, it would explain political commentators’ fixation with geographic polarization and highlight the importance of describing the country’s spatial heterogeneity. Rather than affect everyone everywhere, geographic polarization may only (or especially) affect those with the loudest voices and most prominent platforms.