The Big Sort

by Bill Bishop

  Nationally, Democrats increased their votes by 8 million over 2000. Republicans, however, bumped up their votes by 11.5 million, and that was the election. The Republican increase came from turning out Republicans, not by persuading independent voters.9 Kerry increased the Democratic Party's take of the independent vote over 2000, but his party still lost by more than 3 million votes. Micro-targeting probably helped increase Republican turnout, but Bush won because his campaign understood how communities and politics have changed over the past thirty years. His campaign knew that the way people organized their lives was more important than demographics. And the campaign realized that how people conceived of their communities was more important than issues. The Republicans won because they understood the Big Sort and the Democrats did not.

  A "Well-Slathered" Campaign

  Matthew Dowd sat in a marshmallowy chair in an Austin coffee shop and explained to me in 2005 what had guided the Bush campaign in 2004. Dowd had been Bush's pollster in 2000 and the campaign's chief strategist in 2004. The 2000 campaign had been standard-issue, late-twentieth-century politics, Dowd said. The presumption then was that the key to winning rested largely with independents, the ticket splitters who floated in the mushy middle between the parties waiting to be persuaded. This had also been the 1996 model for campaigns, Matt Dowd said, and the 1992 model, and the model on down through the years, as candidates aimed 80 percent of their money and time at the confused clump of voters that could be swayed to one side or the other.

  After the 2000 race, Dowd began to review records from the campaign and from recent political history. He found that the number of voters in the middle had shrunk to a nubbin. There were very few people who would listen to both sides of a political debate. In the 1980s, perhaps a fourth of the nation's voters were truly in the middle and available to be pulled or persuaded. But Dowd realized that the number of swing voters had since dropped by more than half, to less than 10 percent. Not only were there fewer people in the middle, but the 90 percent of Americans who had chosen sides were growing more adamant in their beliefs. Dowd studied polling results back to 1980 and saw that voters had become vastly more partisan. Democrats were more rigidly Democratic and Republicans more staunchly Republican. The key measure for Dowd was the gap in how Democrats and Republicans viewed the president. People were always more approving of presidents who came from their own party. But the gap between how Democrats and Republicans viewed the president had been growing: it was 36 percentage points for Richard Nixon in 1972, 42 percentage points for Jimmy Carter in 1980, 52 percentage points for Ronald Reagan in 1988, 55 percentage points for Bill Clinton in 1996, and 71 percentage points for George W. Bush in 2004. By the time of the 2004 campaign, 90 percent of Republicans approved of the way Bush was doing his job; 81 percent of Democrats disapproved.*10

  Dowd's insight was a genuine discovery. Political science textbooks were teaching students that voters "no longer strongly identify with one of the major parties as they once did." Journalists had written that the "most important phenomenon of American politics in the past quarter century has been the rise of independent voters."11 Political scientists could see that members of Congress were becoming more partisan in the 1980s and 1990s, but they still contended that for voters, "the two major parties are no longer as central as they once were in tying people's everyday concerns to their choices in the political system."† In early 2000, Princeton University's Larry Bartels wrote the first academic paper challenging the notion that voters had lost interest in political parties, a view he called "both exaggerated and outdated." Americans' attachment to political parties had declined until the mid-1970s, Bartels wrote, but then the trend had reversed.12 From the mid-1970s on, Americans had sorted themselves into political parties in the same way they had sorted themselves into like-minded communities, organizations, and churches. The middle of American politics dwindled, just like the middle of American religious life dwindled. When the Bush campaign strategists realized that the political middle consisted of a shallow pool of voters, they changed the way they planned for the 2004 election. The campaign wouldn't focus on persuading the undecided; instead, according to Dowd, Bush would spend 80 percent of his time and money motivating partisan Republicans to vote.

  The Republican strategists seized on another insight, too—inspired by a mid-priced restaurant chain. Dowd saw that American communities were "becoming very homogeneous." He believed that to a large degree, this clustering was defensive, the general reaction to a society, a country, and a world that were largely beyond an individual's control or understanding. For generations, people had used their clubs, their trust in a national government, and long-established religious denominations to make sense of the world. But those old institutions no longer provided a safe harbor. "What I think has happened," Dowd told me in the early summer of 2005, "is the general anxiety the country feels is building. We're no longer anchored."

  No kidding. Unsurpassed prosperity had set people free—free to think, speak, move, and drift. Unsurpassed prosperity had enriched Americans—and it had loosened long-established social moorings. Americans were scrambling to find a secure place, to make a secure place. In the post-materialist culture that emerged after the 1950s, people weren't finding safety or a sense of self by obeying or abiding or even joining. There wasn't the same satisfaction to be found in learning a trade and attending noon lunches with the Lions. But if the post-materialist person couldn't find meaning in being a carpenter or a Mason, it became incumbent on him or her to create an identity. Most Americans have done that by seeking out (or perhaps just gravitating toward) those who share their lifeworlds—made up of old, fundamental differences such as race, class, gender, and age, but also, now more than ever, personal tastes, beliefs, styles, opinions, and values.

  Americans joined communities, churches, and political parties in a manner that was almost tribal. "People were having a tendency to pick a team, and that team was having an effect up and down the ballot," Dowd said. People were siding with a party and then voting a straight ticket, from city council to president.* Political party affiliation had more to do with social identity than ideology. Choosing to be a Republican or a Democrat reflected a way of life.† Party membership shaped people's sense of reality. To Republicans, for example, the 2000 election had been a clear victory. To Democrats, it had been pure theft. People saw the world differently depending on their party affiliation, much as sports fans on opposite sides either cheer or boo a close call by a referee. Republicans understood that the 2004 contest would be driven by these tribal allegiances.

  Dowd described for me people who had lost attachment to their employers. "People no longer work in jobs for twenty years," he said. "We no longer have our pensions, and that has created a great anxiety." Americans worry about health care and a shifting morality, and they see a world that appears particularly dangerous and fragmented. In urban America, the commercial response to this sense of apprehension and isolation is Starbucks. "Starbucks has nothing to do with coffee," Dowd said. "It has to do with people who want to come to a place where they can be around people and not be isolated." The suburban equivalent to Starbucks is Applebee's, the low-slung eatery found in every exurban settlement. Applebee's, Dowd said, provided the social organizing strategy for the 2004 campaign.

  Dowd wasn't attracted by the food or the restaurant's design. Every Applebee's is like every other Applebee's in menu and architecture. (The food was characterized in one review as being "well-slathered. ")13 But each Applebee's is uniquely decorated in one way: it has a "neighborhood wall" just inside the front door. The wall is adorned with local photos, often from the region's agrarian past. There are bleached-out pictures of bonneted founding mothers and pioneer farmers wearing tall boots and droopy mustaches. Those links to the soil (the soil now covered by a Circuit City, the high school football stadium, and Applebee's own parking lot) are paired with present-day emblems of community: a yellow rubber jacket autographed by all the members of the volunt
eer fire department, photos of local sports heroes, a banner from the nearby high school. This is target nostalgia or micro-mythmaking, and it works because it reflects the real identities of the people hunkered down in the nearby booths. The business concept is to deliver a local experience along with Thousand Island dressing, to create a place where "our kind of people" take their dinner. Applebee's is "selling community" to the people enjoying the slathered entrées and the sepia photos.14

  "It's not just nostalgia," Dowd protested. And then he said something that makes sense only in a country where the same set-piece psychology can evoke community in both a glassy stack of "riblets" and a Republican presidential candidate. What people are looking for, he said, is "a present-tense desire to recapture something that was lost."

  In a sense, Applebee's applied the church growth techniques of Donald McGavran to restaurants. McGavran had learned as a missionary in India during the 1930s that trying to convince individuals to come to church one by one was infinitely slow. It was quicker to recruit whole villages or castes—"peoples," McGavran said—to enter a church as a "homogeneous unit." McGavran taught several generations of preachers that people wouldn't go to a church where services were conducted in languages they didn't understand or where members wore better clothes, were "obviously of a different sort," or ate strange foods. The trick to building larger congregations was to make the passage from community to church seamless and untroubled. The preachers who took McGavran's advice built suburban megachurches precisely to comfort the new communities Americans were moving to outside the cities. They removed the anxiety of coming to church by making church a piece of the community.

  That was exactly Applebee's strategy. To be successful, you don't entice individuals to a restaurant. You bring in entire communities by breaking down barriers between your establishment and the "village." You don't serve strange foods; you serve foods that are comfortable—well slathered. It worked in the same way the megachurches worked. Indeed, said Applebee's CEO Lloyd Hill, "there's something evangelistic about what we do."15 By the time of the 2004 presidential campaign, Applebee's was the largest sit-down food chain in the country.

  The Republicans took the Applebee's marketing concept—and, by extension, the tactics of the megachurch—and applied both to a political campaign. It was a marketing question, really. How do you build a following of the faithful in an age of increasing partisanship, widespread distrust, and homogeneous communities? How can a campaign for president use community as a political asset? The answer, the Bush campaign decided, was to create a tribe of Bush voters. The decision was to take Applebee's technique of adopting a trusted community and, Dowd said, "re-create that politically."*

  Savage Republicans

  "The [Bush] strategy, which we took, we actually implemented it exactly the way they said that we should," said Robert Thibodeaux. Thibodeaux sat in a booth at one of a jillion Caribou Coffee shops in overwhelmingly Republican Scott County, Minnesota. He always wears outrageous Hawaiian shirts. Thibodeaux had spoken several times at a state Republican Party convention, and people remembered his voluble nature and garish wear. So even though he had moved to Scott County only six years earlier, the party picked him to chair the local Bush campaign in 2004.

  Thibodeaux runs with a group of Republican activists from the little town of Savage. They are both geographically and ideologically Savage Republicans. A confidential briefing book put together by the Bush campaign in 2004 listed several "potential weaknesses" in Scott County. For one, the boys over in Savage were a little bit too Republican. "There is a significant number of ultra conservatives in the county, some of whom are disenchanted with some fiscal policies," the report said. For one thing, Thibodeaux and his group were hounding the party about Bush's ever-inflating budget, and it wasn't long before the state Republican Party had dubbed them the "Savage Mafia." But if being "ultra conservative" was a disadvantage in Scott County, you couldn't tell it from the vote in 2004. The Savage Mafia put together a campaign that won Bush more than 60 percent of the county vote. And, as Thibodeaux said, they did it all according to Dowd's plan.

  After Matt Dowd realized that the number of undecided voters was both small and dwindling, he and the rest of the Bush campaign adopted a strategy that bypassed independents and concentrated on turnout. They collected methods thought to boost the number of voters (those noxious computer-generated telephone calls, for example) and tested them in the 2002 off-year elections.

  In 2000, Donald Green and Alan Gerber at Yale University had published research measuring the effects of different get-out-the-vote techniques. The political scientists had conducted a test during a 1998 election in New Haven, Connecticut. Some residents had received a piece of mail asking them to vote. Paid solicitors had made the same pitch over the phone to another group of people. And a third group of residents had received a visit from a paid canvasser. When Gerber and Green checked who actually voted, they discovered that direct mail and, especially, phone calls had had little effect. But face-to-face canvassing had raised turnout rates from 44 percent of registered voters to 53 percent. In a subsequent experiment, the researchers found that for every twelve front-door contacts, one additional person voted.16 In the electoral world confronting both parties, these were huge increases. The four presidential elections from 1988 to 2000 were, on average, closer than in any comparable period over the past century. An increase of a fraction of a percent could win an election.

  Both parties were familiar with Gerber and Green's research. The left-leaning ACT hired thousands of young people to canvass neighborhoods based on the face-to-face trial conducted in New Haven.17 But the Republicans learned from their 2002 experiments that door-to-door canvassing came with an asterisk. The more closely tied the person making the appeal was to the neighborhood, the more likely it was that the appeal would result in an additional vote. A friend worked best, but if the person coming to the door was clearly from the community, that also worked well. "There was like a four times higher likelihood that voters would respond if they were contacted by somebody they knew, or if it was somebody from the neighborhood," Dowd told me in reference to his party's 2002 tests. This was an old principle of marketing. Insurance salesmen had known since the 1960s that the more they resembled a prospective customer—in age, income, and especially political outlook—the better chance they had of making a sale. Similarly, evangelists would make sure that the newly converted were first met at the altar by people of the same sex, age, and race. In 2004, Republicans rediscovered the recruiting technique that had been refined by insurance salesmen and revivalists.*

  Republican strategists had also come to understand why it was important in these times to have somebody from the neighborhood make the appeal for Bush. Religious historian Robert Wuthnow had found a "search for connections" in America. In earlier times, Wuthnow wrote in 1998, people identified their faith by membership in a larger church. Now they were seeking kinship within smaller groups and communities.18 Similarly, the marketing director of McDonald's had said that the country had transformed from the "me decade" to the "age of I ... And by that [I] mean: I am an individual, but I don't want to feel alone. To what group do I belong?"19 Americans were seeking the company of others—albeit, others who were very much like themselves. J. Walker Smith of Yankelovich Partners described Americans as "hiving." Their reaction to the anxiety of the world—the uncertainty of work and health and the volatility of global politics—was to seek "the embrace of others in a safe setting abuzz with engagement and activity." They were devising personalized environments. "People want to live where they can enjoy the comforts of family, friends and neighbors," Smith told his clients. Americans were "finding comfort through connection," and the strongest connections were with people like themselves.20 "This self-selection is incredibly important," Republican political consultant Bill Greener told me in 2005. Greener lives in a heavily Democratic suburb of Washington, D.C., and has been struck by the way Americans are clustering in like-minded nei
ghborhoods. We live in an "age of invisibility," Greener said, and people have a "constant sense that everything is a hassle, where stress surrounds everything, whether it's driving to work or trying to get something accomplished in public policy. When it comes time to go home or it comes time to go to church or comes time to enjoy other social endeavors at some level, there is the thought that 'I don't want to have to listen to all that shit' [from those with different opinions]. And that translates into what we're observing."

  One person's social comfort, of course, is another's social prison, and as people seek the best place to live, they create a jagged cultural landscape. Places that are economically vibrant—that produce more technology and discover more marketable ideas—generally have looser social connections. People there are less likely to join clubs, volunteer, or attend church. These places, on average, vote Democratic.* Other people seem to prefer places with tighter social ties. Residents of these communities volunteer, join, and stay closer to their families. They largely vote Republican. This cultural difference had economic advantages for Democratic areas in 2004, but when it came time to build a grassroots political organization around community networks—the Apple-bee's strategy—the closer ties in Republican areas perfectly matched the Bush campaign strategy.†

  Put simply, Republicans had more social networks to tap. In Republican landslide counties, 20 percent of the residents went to church more than once a week. In Democratic landslide counties, 8 percent did. In addition, 45 percent of the people in Republican landslide counties belonged to a Bible study or prayer group.21 Those connections naturally transferred to politics. I visited with the Savage Republicans at a pizza party they held at the Tin Shed restaurant in town. I met Bob Stapleton, a commercial pilot. Stapleton had a deep interest in education and had become involved with the conservative group EdWatch, which opposes things such as "school-to-work" programs and other "European" innovations. He had also joined a Bible study group that included Randy Penrod, the chair of the Scott County Republican Party. Penrod had recruited Stapleton into the party, and soon Stapleton had become the education policy person on the party's county executive committee. "There's a synergism," Penrod explained. "Because Bob is involved with EdWatch, I asked him to sit on the executive committee. And I met him at Bible study."