Author Timothy Heleniak is the director of the American Geographical Society.
PEOPLE who voted for Mitt Romney in the presidential election are amazed that he lost because they don't personally know many people who didn't vote for him.
Likewise for President Obama's supporters. Because they have few close acquaintances who voted for the other party, they are puzzled that the margin of victory wasn't higher.
As in many recent presidential elections, this year's was close in the popular vote. The reason for that closeness is that we have segregated ourselves geographically into such like-minded clusters.
The clearest reflection of that segregation is the Electoral College, which shows that there are a number of reliably red or reliably blue states and a small number of swing states.
But things are more obvious at the finer geographic scale of 3,000 U.S. counties, which are even more reliably red or reliably blue; often more than three-quarters of the electorate in many counties cast ballots for one party. Reliably Republican are rural, small-population counties in the Great Plains, the South, Utah and neighboring states and the South. Reliably Democratic are Washington, D.C., Indian reservations, black-majority counties in the South, major universities, high-amenity counties on the West Coast, large metropolitan counties and Vermont.
At an even finer geographic level of neighborhoods, political party preference is even more skewed toward one party or another. A major explanation for this is the high rate of mobility of the American population — 1 in 8 of us move each year — and how we end up residing near other like-minded persons.
Although there are many issues that cause people to vote for candidates — the proper role of government in people's lives, the place of the United States in the world, abortion and gay marriage — people don't usually select a place to live based on the political affiliation of neighbors, but rather on lifestyle and other characteristics and amenities that then translate into political preferences. But knowing where a person stands on one issue is a fairly reliable predictor of that person's political party affiliation and stand on other major issues.
A number of recent studies show that neighborhoods or clusters across the country explain how we have become so polarized. The most prominent describes "The Big Sort," or segregation by lifestyle. It showed that The Big Sort was also an arrangement by political party affiliation, which has become extreme since the 1970s. We also segregate in the clubs we belong to, in the TV and radio stations we listen to and in the periodicals we read.
We have become increasingly less exposed to viewpoints that differ from our own. We rarely know or associate with people who think differently from the way we do.
Think of the last time you were at a party that was divided roughly equally between liberals and conservatives. Typically, any gathering of friends or acquaintances will result in 90 percent or more consisting of like-minded persons with similar views on a range of social and political issues. For the one or two oddballs who hold views that are opposite yours, you will not only disagree and not understand them, you will likely question their mental stability.
This same segregation into opposite ends of the political spectrum is taking place in Congress as well, certainly in the House of Representatives, where redistricting creates safe and extreme districts. These are the people we voted for, so we have the Congress we deserve.
The high levels of racial segregation in the United States seemed to have peaked and are subsiding, but political segregation is increasing. One possible way to shift the political debate and expose the country to different solutions would be to eliminate the Electoral College. This is unlikely to happen because of the high hurdle of needing a constitutional amendment to do so and vested interests in retaining it. However, eliminating the Electoral College and bringing more voters into the discussion could produce a different set of issues, different candidates and different solutions. It's something worth trying, since the current system doesn't seem to be working.