Causes and consequences of the geographic partisan sort in the United States

What explains the geographic sorting of the population along partisan lines in the United States? We know that the number of "landslide" counties (i.e., counties in which the vast majority of the population regularly votes for candidates from one political party) and other geographic units has steadily increased since the 1970s. However, little empirical research has examined which factors turned some closely-contested counties into Republican enclaves and others into Democratic strongholds. This dissertation examines the increasing political homogenization of small geographic units and seeks to explain the causal mechanisms driving this phenomenon as well as its consequences for individual political attitudes and behavior among residents residing in these geographic units. This project helps explain the mechanisms that lead some counties to become magnets for Democrats, and others to draw increasing numbers of Republicans. In order to demonstrate that this process has important normative consequences, this dissertation also how examines this phenomenon influences individual views on politics. This project contributes to the field's understanding of political geography and provides another explanation for increasing levels of political polarization.;This dissertation posits that migration plays a key role in this sorting process. Different economic, demographic, and cultural community attributes tend to draw residents with different characteristics that are highly correlated with party identification. The political science literature has established that young Americans exhibit different voting patterns from older Americans; non-Hispanic whites differ from racial and ethnic minorities; married Americans differ from unmarried Americans; and social class and occupation are often a useful predictors of party identification. This dissertation demonstrates that these groups also exhibit different migratory patterns. Over time, the clustering of Americans by age, marital status, race/ethnicity, income, and occupation type leads to the domination of some communities by one political party or the other.;This geographic partisan sort is not simply a by-product of different migration patterns of different social groups. In contrast to previous literature, this dissertation demonstrates that politics itself determines the destinations many American migrants. Specifically, this dissertation provides strong evidence that the migratory patterns of non-Hispanic white Republicans are directly influenced by the local political context; as a county increasingly leans toward the Democratic Party in national elections, white Republicans living in that county become progressively less satisfied with their community and become increasingly likely to leave. This is demonstrated with both individual and aggregate data.;This dissertation also demonstrates that this phenomenon has important political consequences. Compared to residents in competitive counties, residents of landslide counties who identify with the party in the local majority are more likely to exhibit signs of polarization, such as strong ideological convictions, strong party identifications, and polarized attitudes toward major political figures. They are also more likely to be actively involved in politics. Thus, the geographic partisan sort is leading to a more polarized electorate.