Share this:

Wendy-Tam-Cho-80x108Almost any electoral map of the United States clearly shows an association between political beliefs and geography; states in the South and Midwest tend to vote Republican, while the West Coast and Northeast reliably lean Democratic. Following this, Wendy K. Tam Cho looks for evidence of political self-segregation—that is, individuals moving to areas that contain clusters of other like-minded people—by tracking movement using voter records from 2004, 2006, and 2008. Her results indicate that, while it may not be the only factor, partisan sorting is a significant component of destination decisions.

News outlets in the United States seem to have adopted a uniform color scheme in maps for depicting Republican and Democratic support. When we view presidential election returns, we expect to see red and blue states, with red states signaling Republican support and blue states highlighting areas of Democratic strength. These maps succinctly convey a lot of information, and even those who are not particularly in tune with politics are struck by the plainly non-random displays of red and blue in U.S. elections. In particular, red states cover most of the landmass in the United States, while blue areas dominate the most densely populated regions. A view of these maps across time would further show that these patterns appear to be intensifying over time. Election after election, barring startling scandals, Democratic and Republican strongholds bear true to our pre-election expectations that their voters adhere to past partisan loyalties.


Credit: by and by (Creative Commons BY-ND 2.0)

In 2008, despite Barack Obama’s lopsided presidential election victory in many parts of the country, the geographic expression of rival partisan preferences appears to have heightened over previous elections.
The counties that previously favored one party awarded even greater victory margins to their preferred party, and the distribution of party registrants became increasingly askew across more locations. In the United States, Republicans prefer to live in close proximity to other Republicans, just as Democrats prefer their own kind.

This pattern of geographic clustering is intriguing, to be sure, but some have also claimed that it is alarming. Bill Bishop has lamented that geographic clustering of like-minded individuals breeds narrowness of viewpoint. He argues that heterogeneous communities teach their members to compromise by providing a neighborly forum for opposing opinions, whereas homogeneous communities promote extremism and ideological intensity because differing viewpoints are regularly dismissed without discussion or consideration. If we continue on this present course, Bishop admonishes, this cultural evolution will foster an increasingly intense intolerance that will tear the country apart at its seams.

The consequences of geographic clustering may also manifest at the institutional level where this type of geographic patterning may foster less responsive representatives. After all, why would politicians spend time courting constituents in noncompetitive homogeneous communities? Ironically, constituents reward, rather than penalize like-minded representatives because they place a higher premium on ideology than on responsiveness. The result, one might bewail, is a polity wherein citizens are deeply divided, parties are polarized, and political discourse is stifled.

We can speculate but we cannot know the precise future implications of geographic sorting. Although some might consider the Bishop account to be sensationalized, it is also believable enough to cause one to take pause. Is our preference to live among and associate with like-minded individuals likely to result in a country that none of us prefer—a country characterized by intolerance and dismissiveness?

In our research, we do not take a position on the political consequences of the geographic sorting of the electorate, but we acknowledge that consequences exist and that they have implications for democratic practices and traditions. The aggregate patterns of partisanship are clear when it comes to presidential preference. We are diffident about the consequences because we are unsure of the roots of the phenomenon. What creates, defines, and sustains these geographic patters of partisanship that shape the behavior of individuals as well as politicians? Bishop offers evidence that the American electorate has been sorting itself through decades of internal migration. His primary evidence comes from presidential voting data at the county level. Over three decades, he showed that counties that voted for the Democratic candidate have produced even more solid Democratic support, and Republican counties have become more reliably Republican. The maps, although remarkable in this regard, fall short of definitive as he presents aggregate patterning that is incapable of reliably shedding light on individual behavior. County-level relationships may be suggestive, but we have long known that their relationship with individual-level tendencies might not be in the same direction or of comparable magnitude. To obtain information about individual behavior, we must examine individual-level data.

We empirically examine individual voter migration on a large scale. Surely, many factors contribute to geographic variation in the balance of partisanship. Population migration is one obvious force, but other factors, such as the polarization of the national parties and the evolution of individual attitudes, would have a similar effect. There is much research that remains to understand the phenomena, and our contribution is in an empirical examination of voter migration.

To understand the effects of migration on the political landscape, we identify and track migration flows through voter files from 2004, 2006, and 2008 across seven states. For evidence that geographic sorting is occurring, we should see indications that Republicans move to areas that are more favorable to Republicans than their original locale. Similarly, Democrats should migrate toward areas that are more favorable to Democrats.

Milwaukee Area Labor Council Political Yard Signs

Credit: wisaflcio (Creative Commons BY-NC-SA 2.0)

It was evident in our analysis that not all partisans exhibit the same tendency to sort geographically. Partisan preference is regularly trumped by economic concerns. As one might expect, destination decisions are strongly related to destination characteristics related to racial composition, income, and population density. At the same time, our research indicates that partisan sorting is statistically significant for both Republicans and Democrats even after a whole host of neighborhood characteristics have been taken into account.

Certainly, migration is never only about politics or even principally about it. We hypothesized and our analysis contained evidence that jobs and family concerns remain the most important factors in migration decisions. Nonetheless, once the major decisions about relocation have been made—that a move is going to occur, that it will be to this state, to that metro area, and to this county—then the more micro decisions are made about specific neighborhoods, streets, and dwellings. At this level, the remaining political locales often still sport a significant amount of partisan variation. It is at this point that considerations might be more proximate to political values and where partisanship might play a more prominent role in the decision process. Whether the role of partisanship is central or ancillary, if it is any part of the decision process, it has the potential to recast the political landscape of the United States.

Please read our comments policy before commenting. 

Note:  This article gives the views of the authors, and not the position of USApp– American Politics and Policy, nor of the London School of Economics.

Shortened URL for this post:


About the author

Wendy-Tam-Cho-80x108Wendy K. Tam Cho – University of Illinois at Urbana-Champaign
Wendy K. Tam Cho is Professor in the Departments of Political Science and Statistics, Senior Research Scientist at the National Center for Supercomputing Applications, Faculty in the Illinois Informatics Institute, and Affiliate of the Cline Center for Democracy, the CyberGIS Center for Advanced Digital and Spatial Studies, and the Computational Science and Engineering Program at the University of Illinois at Urbana-Champaign.

Print Friendly