Bill Bishop

May 2008By Comments

In The Big Sort, the Austin political blogger and Pulitzer finalist addresses America’s tendency to segment itself into tiny, like-minded groups (a phenomenon he calls clustering).

How did the “big sort” notion come to be, and what does it signify?

[Sociologist] Robert Cushing and I began exploring why some places produced technology and patents while others seemed to stagnate. We found that the country was sorting: The places where educated people moved to got richer; the places where young people moved produced more patents; basic beliefs varied place to place. The sort was cultural, economic, and political. What surprised us was that we live in a time of incredible choice in where and how to live, and yet we were constructing increasingly isolated lives. I think it signifies a retreat from our country’s early democratic promise, that diversity is an asset if we can find a way to talk to each other and listen.

So instead of bonding together over their similarities, Americans are dividing into sub-sub-sub-groups based on their differences?

Definitely, opposites do not attract. We talk about assortative mating when it comes to marriage. People tend to mate with those like themselves. (There is, by the way, a strong correlation between the ideological orientations of spouses—such that dating services now use political beliefs to help match couples.) What we’re talking about here is a society-wide case of assortative mating when it comes to how and where we live. We are finding comfort around those who share our interests—religion, sports, Internet sites, lifestyles, politics. Like does attract like. But, as you say, there is a power in people trying to avoid those who are different. Certainly, God help those who find themselves living in places where they are in the minority. Political scientists have known for half a century that political minorities vote less—political minorities are even less likely to participate in civic activities.

In my deeply Democratic neighborhood (Travis Heights in Austin), a Republican once ventured on to the Internet news group and was not so politely told to find someplace else to live. In one deeply Republican Hill Country town, Democrats built a July 4th float, but when it came time for the parade, nobody would ride. Nobody wanted to be publicly identified as a Democrat. Luckily, it rained and so the Ds were spared the ignominy of having a riderless float.

How does a national political campaign address such a socially schizophrenic populace?

The old way of running a political campaign was to think of the electorate as a pie cut into three large, and largely equal, chunks—Democrats, Republicans and Independents. The goal of a campaign was to convince enough Independents and those from the opposing party to pick your side. Campaigns were aimed at persuading voters. The key insight of the Bush campaign in 2003 was that there were just very few voters left out there to persuade. Actually, the number of true independents had been dropping since the 1970s, as were the number of people who split their ticket between the two parties. Most political journalism was—and still is—based on the outdated notion that one-third of the electorate was truly undecided.

What [former President Bush adviser] Matthew Dowd and others in the Bush campaign realized was the number of undecideds was actually under ten percent. So the Bush campaign concentrated on turnout rather than persuasion. (For fun, ask a group of people how many don’t know today which party they will support in November. There won’t be many. I did this last summer in a crowd of about 400. Maybe a dozen people raised their hands.) The second key insight of the Bush campaign was to understand that these like-minded social structures created over the past generation—neighborhoods, clubs, churches—couldn’t be organized from the outside. The campaign had to be run by people from the community. There was good research on this, too. The Bush people knew that a door-to-door campaign run by people who came from the neighborhood (or looked like they came from the neighborhood) was more effective than a door-to-door campaign of volunteers trucked in from out of state—the kind of campaign operated by the Democrats and MoveOn.org. Finally, the Bush campaign knew that they had to tap into the social networks being created. So they organized through home schoolers and through church networks and sports leagues. They didn’t use preachers to call their flocks into the voting booths. They found key leaders within the church who would support their campaign. (It turns out that preachers aren’t even very good at recruiting members for their own churches, much less for a political campaign.)

The Bush campaign gained members the way evangelical churches gain members—through friends recruiting friends. It’s called friendship evangelism, or relational evangelism in the church. So the same techniques that built megachurches in suburban and exurban counties also created the Bush majority in 2004. A shorter answer to your question is that you don’t run a national campaign. You run thousands of local campaigns that address the needs, desires, and ideologies of individuals. It’s target marketing or one-to-one marketing in the commercial world.

Are these homogeneous clusters really anything new? Economic and ethnic groups often tended to settle the same towns or neighborhoods.

Crucial question. You are absolutely right. From the first day we’re alive, we learn that there is safety among those who are like ourselves—and danger in disagreeing with others. Birds of a feather flock because that’s the way birds survive. This has always been true, and America has at times been extraordinarily polarized geographically. (There was the Civil War, after all; and historian Robert Wiebe described America in the 19th century as a “nation of loosely connected islands.”) What was remarkable to us was that the country is growing more politically and culturally polarized now. We live in a time when day-to-day survival for most Americans is assured; when social safety nets reduce the need to depend on family; when Americans have unprecedented choice about where and how to live—but given all this freedom and opportunity to live where and how we like, the rates of political segmentation are increasing. The 19th century island communities were enforced by a rough kind of geography, by the demands of distance, by social convention, and by legal segregation. People’s survival was uncertain. It’s easy to see why people lived in what were certainly homogenous clusters. But why have our communities grown more segregated over the last 30 years? Why were we creating island communities now? That’s what The Big Sort is about.

Is this a uniquely American phenomenon?

As we looked back at the origins of this sorting, we could see that the country fractured in the 1960s—around 1965, to be precise. In that one year, trust in major institutions began to decline; membership in mainline churches started to drop; divorce and crime rates began to climb; allegiance to political party dissolved as people lost faith in traditional party labels; membership in long-standing civic organizations (the Elks, bowling leagues) started to drop, as did the percentage of daily newspaper readers. Society seemed to unravel all at once. Certainly there were plenty of events in 1965 to justify social revolution: “Bloody Sunday” in Selma; a rapid increase in troop deployments to Vietnam; the first bombing raids over North Vietnam; the Watts riots; the Selma to Montgomery civil rights march; a blackout that cut electricity to much of the Northeast; the first large-scale anti-war demonstrations and the first draft card burning. 1965 was both hopeful and terrifying—in the movies it was both The Sound of Music and Polanski’s Repulsion.

But here’s the thing—many of these same large-scale social changes took place around the same time in every industrialized country. Trust in major social institutions dropped in every economically advanced country. The Germans lost trust in their government and allegiance to their political parties roughly at the same time we did in the States. We like to think that “the ‘60s” destroyed American faith in government. Well, sure, the Watts riots had a powerful effect on the nation’s psyche. But New Zealanders lost their trust in government at roughly the same time. Americans like to explain things with nouns: Watergate, the ‘60s, Vietnam, assassination, Reagan, Clinton, the religious right. We could see that a lot of what was creating the big sort in the United States was happening at the same time in most other industrialized countries.

We get into this phenomenon in the book. The short explanation for why there is this uniform culture shift is that people who grow up in times when their survival is assured—when there’s plenty of food and people feel safe—develop different attitudes about politics and culture. They are less supportive of centralized authority and more concerned with individual freedom. They are less formally religious but increasingly concerned with individual spirituality. Economic security declines as an issue and environmental protection rises. This is true in the U.S., and it’s true in Japan, France and Sweden. Now the U.S. is an outlier. We remain more religious than most other industrialized countries. But the political pattern is the same. Ron Inglehart at the University of Michigan and Pippa Norris at Harvard report that in every advanced country, the majority of those who attend church at least once a week vote for the party on the right. The voting patterns of the faithful in the U.S. are, in fact, typical of those in other industrialized countries. The only unusual thing about America is that we have such a large number of churchgoers.

What cluster does Bill Bishop belong to?

When I went to the polls in 2006, I took pictures for a Web site that was collecting photos of American precincts. I snapped some shots of a brown Lab with voters in the background, then asked the owner the dog’s name. “Che,” she said. When a dog chosen at random at a polling place is named for a Marxist revolutionary, odds are that precinct skew blue.

Do you anticipate this trend changing course down the road?

We have lots of trends. Let’s think about two. The first trend is that of people seeking more likeminded company. All I see is for that to be accelerating. It’s now taken as a given that all Web sites will be tightly focused on homogenous groups. The New York Times recently wrote about problems with MySpace and Facebook. These sites looked dated to new software developers. The trouble with these massive social networks, according to the newspaper, was that “big Web sites attract masses of people who have dissimilar interests and, ultimately, little in common.” That, by the way, is considered a bad thing. Getting people together who have “dissimilar interests” is seen as a sure way to sink a new Web venture. Great!!

Meanwhile, the returns to living around those with similar interests are so great. Kids in Portland, Oregon, told me that one of the reasons they live there is that all the stuff they want—movies, books, bands, speakers—comes to their town. There is an advantage for consumers to live around those who are like themselves. It’s a way to insure that you can easily get your stuff—so that the music you want is played on the radio or at bars, the movies you’ve been wanting to see are screened, the authors who write the books you want to read come to the local bookstore. Moreover, the world isn’t getting any friendlier or easier to comprehend. “When people find themselves unable to control the world, they simply shrink the world to the size of their community,” observed sociologist Manuel Castells. That’s what we’re doing—and that’s what we’ll continue to do.

The second trend is that lifestyle is aligning so closely with political party and political beliefs—that the way a person thinks about, say, an automobile purchase, is a good predictor of how that person feels about the war in Iraq. And what that person feels about the war in Iraq is a good predictor for what he or she thinks about tax cuts or mass transport or foreign aid. History tells us those alignments won’t last. What breaks up these rigid connections are issues that cut across existing political divisions. There will be a new issue that doesn’t easily fit within the boundaries of what it means to be a Republican or a Democrat. I thought for a time that illegal immigration would be one of those crosscutting issues—where some Republicans found they had something in common with some Democrats. Or health care. Here you have the Service Employees union working with Wal-Mart and GM on producing a national health insurance plan. That’s a crosscutting relationship, which is something our country is missing right now. Democracy works best when people are friends one day on one issue and then opponents the next on another problem.

Today, our friends are our friends and our enemies are our enemies every day of the week. That’s why Congress is stymied and our political conversations are so unproductive. We don’t ever experience what it feels like to find agreement with our enemies. So we await the next crosscutting issue, knowing that one will arrive some time even if we don’t know when.

Which social groups are most hurt or disenfranchised by this clustering effect?

Lots of ways to think about this. First, economically, there are large portions of the country that are falling behind because they are losing their educated people. There’s a trade, in fact. Educated people move to a select group of cities (Dallas, Austin, and Houston in Texas) and those areas thrive while other communities are drained of their best talent. Meanwhile, as housing prices rise in these cities, poorer (and less educated) people move out. They move to where housing is cheaper. So some communities lose twice: they give up the people they educated with their tax dollars and then they become the homes for people who don’t have the skills to afford big city rents. In this sorting, lots of communities are losers.

In other ways, the nation is better off, because we have thousands of experiments going on locally in how to live and govern. This kind of diversity is healthy. But the nation comes up short because we don’t talk to one another. There’s no common ground between these vastly different communities, and so the country doesn’t gain the benefit of seeing those differences rub up against one another in a way that creates compromise and, perhaps, a better solution to our problems. I guess if I were to pick one group that is most disadvantaged, it would be those with graduate degrees. Diana Mutz, a University of Pennsylvania professor who has written a great book, Hearing the Other Side, tells us that Americans with the least education have the most diverse group of people with whom they discuss politics. Meanwhile, educated people like to think their lives are filled with diversity and worldly sophistication. But it turns out that those with the most homogenous political lives are Americans who have suffered through graduate school. The best-educated citizens are the least likely to have a political discussion with someone with a different opinion.

Have the political parties expressed any interest in your work?

I’ve had some back and forth with the [Hillary] Clinton campaign. They clearly see these differences in their contests with Senator [Barack] Obama. The maps of the primary results in Missouri, Ohio, and Texas look exactly like the Bush/Kerry election in ’04—only in this case, Clinton took all the “red” counties and Obama carried the “blue” communities. (I think this is something Bill Clinton knew instinctively. He campaigned in the Republican areas of Ohio and Texas—and these were the communities where Sen. Clinton won the two primaries.) The Clinton people seemed particularly interested in our finding that place overwhelmed other standard ways of measuring ideology. Self-described liberals who live in lopsidedly Republican counties are considerably more conservative than self-described liberals who live in, say, Austin.

Read an excerpt from The Big Sort.

Related Content