The 'Big Sort': The New Segregation?

People prefer to be with people like themselves. For all the celebration of "diversity," it's sameness that dominates. Most people favor friendships with those who have similar backgrounds, interests and values. It makes for more shared experiences, easier conversations and more comfortable silences. Despite many exceptions, the urge is nearly universal. It's human nature.

Perhaps America's greatest glory is to rise above this self-absorption. People with many different heritages and beliefs have blended into a cohesive society. At some point, most people subordinate their own firmly held convictions and loyalties to the larger nation. This is more than patriotism; it's the identity of "being an American." But it is in constant tension with the differences that divide Americans.

The latest manifestation of this is what Bill Bishop calls "the Big Sort." By that, he means that Americans have increasingly "clustered in communities of sameness, among people with similar ways of life, beliefs, and, in the end, politics." Republican fundamentalists congregate with other Republican fundamentalists. Liberal Democrats herd with other liberal Democrats. Environmentalists decamp to Portland, Ore. Child-centered Republican families move to the exurbs of Dallas and Minneapolis.

The increasing segregation of America by social and cultural values -- not just by income -- helps explain America's growing political polarization, Bishop argues in his new book (naturally: "The Big Sort"). Because prosperity enables more Americans to live where they please, they gravitate to lifestyle ghettos -- and that has significant political implications. Citing studies of social psychology, Bishop says that group consciousness actually amplifies likes and dislikes. Views become more extreme. People become more self-righteous and more suspicious of outsiders.

It's not red and blue states so much as red and blue counties. Bishop -- a recovering newspaper columnist -- collaborated with Robert Cushing, a retired professor of sociology from the University of Texas, to examine voting patterns in presidential elections. They classified counties as politically lopsided if one candidate won by 20 percentage points or more. Their findings are stunning. In the 1960 Kennedy-Nixon election, a virtual dead heat, 33 percent of counties qualified. By 2000, also a dead heat, the share was 45 percent. In 2004, it was 48 percent. In 1976, it had been as low as 27 percent.

Not surprisingly, many neighborhoods today have mostly Obama or McCain yard signs, not a competitive mixture. Though he dislikes this sorting, Bishop, for good reason, is not contemptuous of it: He discovered it through personal experience. When he and his wife moved to Austin, they instinctively selected a neighborhood called Travis Heights. It had people like them, who turned out to be highly liberal. How liberal? Well, in 2000, almost 60 percent of Texans voted for George W. Bush. In Travis Heights, Bush finished third, after Al Gore and Ralph Nader. (Despite his views, Bishop's analysis is evenhanded.)

Although Bishop is on to something, I think his argument is slightly overdrawn. Today's residential segregation of like-minded people has ample precedent. For much of the 20th century, urban neighborhoods subdivided by ethnic group. The Irish had their blocks, the Italians theirs. But these enclaves were sufficiently compressed that they often coexisted within a single county (Bishop's measuring standard). More important, Bishop, like many others, has exaggerated the extent of the polarization. Evidence of growing differences of opinion among the general public-- as opposed to tinier political elites -- is slim.

Consider two decades of polls from the Pew Research Center. On many questions, there was little change. One question asked whether "government should care for those who can't care for themselves." In 1987, 71 percent agreed; in 2007, 69 percent did. Or take immigration. In 1992, when the question was first asked, 76 percent of respondents favored tougher restrictions; in 2007, 75 percent did. On some cultural issues, opinions converged. In 2007, only 28 percent thought that school boards should be able to "fire teachers who are known homosexuals," down from 51 percent in 1987. In 1987, only 48 percent thought it was "all right for blacks and whites to date each other"; by 2007, 83 percent did.

It's not that everyone agrees on everything (divisions remain strong on the Iraq war, abortion, gay marriage). But growing polarization predominates among political elites of both left and right. The "Big Sort" of residential segregation is still reshaping the political landscape, though more indirectly. With fewer competitive congressional districts, the real political struggles now often take place in primaries, where activists' views count the most. Candidates appeal to them and are driven toward the extremes.

What Arthur Schlesinger Jr. called "the vital center" is being slowly disenfranchised. Party "bases" become more important than their numbers justify. Passionate partisans dislike compromise and consensus. They want to demolish the other side. Whether from left or right, the danger is a tyranny of true believers.