Thursday, November 15, 2012

Election 2012: Polarization Grows and The Big Sort Continues

In an earlier post, Politics in the USA: The Big Sort—1976-2008, we discussed Bill Bishop’s book The Big Sort: Why the Clustering of Like-Minded America Is Tearing Us Apart. Bishop presented rather compelling data from presidential elections in 1976, 2004, and 2008 that supported his contention that our nation was dividing itself into "tribes" of like-minded individuals. This like-mindedness, of course, extended to voting patterns. Bishop produced county level voting results that indicated those counties where one candidate won the county in a "landslide"—meaning by more than 20 percentage points. There was a clear progression over the years as the numbers of such landslide counties grew and spread over significant portions of the country map. The Republicans were accumulating large majorities of the votes in rural areas and the Democrats were doing the same in the more urban areas.

There seemed to be two major reasons for this tribalism. When people did have occasion to change location, they often had a choice of nearby regions and picked the one in which they felt most comfortable—the one in which the people were most like themselves. If a new job was to be filled in Denver, for example, a Republican-leaner might tend to live outside of town and commute in; a Democrat-leaner would more likely head for the heart of town. There is also a group dynamic taking place. Those who vote Democratic or Republican don’t just represent different political philosophies; they represent cultures with different lifestyles and different worldviews. People who migrate into a culture tend to assimilate into that culture. There will be subtle but definite pressures in urban areas, for example, to think and vote like the majority

I was quite interested in learning how this process has proceeded since the last election in 2008. I haven’t found data accumulated in exactly the form presented by Bishop, but Mark Newman of the University of Michigan has posted data that is perhaps even more informative. Newman provides this chart of the winning party in each county. Red is Romney/Republican and blue is Obama/Democrat.

The predominance of one party in specific geographic areas is clear, but the degree of dominance is not. It was startling to watch the election results being reported as they came in and see counties in which one party was winning not just by Bishop’s definition of a landslide, but by much larger margins.

Newman also provides a chart in which the color of the county varies between red and blue according to a scale in which a Republican vote of 70% or more is pure red, and a Democratic vote of 70% or more is pure blue.

Clearly, in the regions where one party dominates, there are significant areas in which the vote is reaching the 70% saturation point. That is rather incredible.

Where does this end? There was a report that there were a number of precincts in an Ohio city in which Romney received not a single vote.  Undiluted partisanship will breed more intense partisanship.  Who would want to be the last Democrat in a Republican precinct?

 Perhaps demographics will come to our rescue and define one party/culture to be dominant and temper this polarization, or realign us so that we fight over different issues.  Something better come to our rescue—and quickly.


  1. Very interesting. Do you have similar maps that show the changes from 2008 to 2010?

  2. I haven't seen anything equivalent for 2010. Off-year elections are strange. Only two-thirds of the voters show up and they seem to be biased towards Republicans. That is one reason they did so well in 2010.

  3. If we deem it healthy for the country to be less polarized - and I do - we should intervene at the ultimate cause of polarization: our voting technology. As an alternative, Ranked-Choice Voting (or Instant Runoff Voting) would fit with the political blueprint of the United States - geographically based districts which elect single members to political office - but open our elections to multiple candidacies, none of which have the power to "spoil" a presidential election in the manner that Nader spoiled 2000. Voters would rank the candidates rather than choose between Red and Blue. In fact, my guess is that elderly voters would likely rank Red and Blue 1-2 or vice versa. The depolarizing would come gradually as US voters move away from pledging to a my-one-and-only choice in favor of voting a rank-ordered Top 5 or 6 list. I am laboring on a boook about the subject, but interested parties should read the Wikipedia entries or visit the Fairvote USA website for more information at this time. TKuhlman
    USAT. Kuhlman


Lets Talk Books And Politics - Blogged