-
sponsorship
One week after a historic presidential campaign that ended with the election of a man who billed himself as a post-partisan candidate of a unified America, this country is more divided than it was four or eight years ago. Or, less abstractly, while some danced in the streets over Barack Obama's victory, others bought guns. Here are some preliminary results from Tuesday's election:
• Communities are just as partisan now as they were in 2004.
Most counties in the United States have grown either more Republican or more Democratic since the 1970s. In 1976, 26.8 percent of the nation's voters lived in a county where either Gerald Ford or Jimmy Carter won by more than 20 percentage points. The number of people living in these "landslide counties" increased to 38 percent in 1992, to 45.3 percent in 2000, to 48.3 percent in 2004.
On Tuesday, 48.1 percent of the vote came from counties where either Obama or McCain won by 20 percentage points or more.
Partisanship is now more Democratic. Four years ago, 83 million people lived in Republican-landslide counties. Now it's closer to 53 million. In 2004, 64 million people lived in Democratic-landslide counties. In 2008, 94 million people lived in Obama-landslide counties.
That means that blue counties got bluer while red counties paled a tinge. The number of people living in counties that are overwhelmingly Republican or Democratic is unchanged from 2004.
• States are more divided.
Emory University political scientist Alan Abramowitz tracks the growing differences among states. He finds the same divisions in the country as four years ago, "only deeper."
In 1976, the average winning margin in the 50 states and Washington, D.C., was 10 percentage points. In 2000, it was 15 points. In 2004 it was 16 points. In 2008 the average winning margin for the 50 states and D.C. was 17.4 percentage points.
There are more landslide states (where a candidate won by 10 points or more). In 1976, there were 19 blowout states plus the District of Columbia. In 2004, there were 29 blowout states plus D.C.
In 2008, there were 36 landslide states plus D.C.
And there were fewer states where the election was relatively competitive, less than five percentage points. In 1976, 20 states were competitive. By 2004, there were only 11 competitive states. This time out, only seven states had margins of less than five points. (They were Missouri, North Carolina, Indiana, Florida, Montana, Ohio, and Virginia.)
"There may be only one United States of America, as Sen. Obama says," Abramowitz said, "but the divide between the red states and the blue states is deeper than at any time in the past 60 years."
• The division between rural America and urban America grew wider.
In 1976, there was little difference in the voting patterns of Americans living in rural and urban counties. By 2004, George Bush won 59 percent of the vote in the nation's rural counties while John Kerry took 40 percent.
Obama did better than Kerry, winning 42.8 percent of the rural vote overall, and he did much better than that in the battleground states. In Indiana, for example, Kerry won only 31.9 percent of the rural vote. Obama won 43 percent of the rural vote and, with it, the state.
Meanwhile, Obama's vote in the cities jumped dramatically over what the Democrat polled in '04. John Kerry took 51.6 percent of the urban vote. Barack Obama won 57 percent of the vote in the core metropolitan counties.
In a sense, 2008 was the inverse of 2004. Four years ago, George W. Bush tuned up the vote in rural and exurban counties and battled to stay close in urban America. In this election, Obama pumped out votes in urban precincts and then did a bit better than Kerry or Al Gore did in rural and exurban counties.
As a result, the gap between the urban and rural vote increased from '04 to ‘08.
Republican and Democratic counties were entirely different kinds of places. The average population of an Obama landslide county was 278,601. The average McCain landslide county had 37,475 people.
• Modern political campaigns continue to be designed to increase political divisions.
Obama's rural campaign consisted of finding the "urban" vote in small towns. For instance, Obama held a rally in Harrisonburg, Va., in the last week of the campaign. It was an unlikely place to find a Democrat. Harrisonburg had voted for every Republican presidential candidate since before the end of World War II. The last Democrat to visit that neck of the Shenandoah Valley was Stephen Douglas in 1860.
Obama targeted Harrisonburg because it was home to two universities, James Madison and Eastern Mennonite, and Mary Baldwin College was tucked into nearby Staunton (which also hadn't voted Democratic in more than 60 years). These were likely Obama voters, and so the campaign went after them. The James Madison precinct recorded the most votes ever-and Obama won Harrisonburg and Staunton.
The counties surrounding these two college towns, however, voted for McCain by a 7-to-3 margin.
Obama plucked these college towns from all over rural America. The only county Obama won in southeast Ohio was Athens, home of Ohio University. In western (Appalachian) North Carolina, Obama won Watauga County, home of Appalachian State University. There was one speck of blue in the Idaho panhandle: Latah County, where students attend the University of Idaho.
Political campaigns these days aren't designed to change minds. Candidates target voters who are already likely to support them-and then the campaigns feed these voters more of what they want to hear. That's a strategy that is all about polarization, and it works.
• People kept sorting.
From 2003 to 2007, people leaving counties that voted Democratic in 2004 likely moved to other Democratic counties, according to statistician Robert Cushing's parsing of federal migration data. The trend tended to increase the number of Democrats in counties that already voted for Democratic presidential candidates.
Republicans didn't cluster as strongly over that four-year stretch. In fact, Republican counties in Democratic states got a majority of their new residents from people who were moving from counties that voted for John Kerry. That could be a reason the Republican vote softened in some areas in '08.
This is the Big Sort, the self-selection of people into increasingly like-minded communities at microscopic levels of society.
• Young people clustered, and where they gathered, the vote was Democratic.
People of first-time voting age (18 to 21) increased their numbers in Obama-landslide counties four times faster than in McCain-landslide counties. According to Cushing's calculations, there were 10.5 million 18-to-21-year-old voters in counties Obama won. In McCain counties, there were 6.4 million of these young voters.
The Sort Continues
The vote last week was transformative in a sense. In many ways, however, the election produced no change at all. The country is split in much the same way it was divided four and eight years ago. People continue to sort by age and by way of life. As a result, our communities (and states) are growing more like-minded.
Oh, and there is the continuing and stark racial division in both the geography and how Americans live. In Republican-landslide counties, blacks and Hispanics are distinct minorities. Where McCain won by 20 percentage points or more, there were five Anglos of voting age for every black or Hispanic, Cushing found. In Obama-landslide counties, there are 1.3 whites for every black or Hispanic. Obama counties and McCain counties are very different places.
Liberals and Democrats seem to think the country's divisions have disappeared just because their man won. And it is easy to ignore people on the other side when they aren't your neighbors. But that doesn't mean the country is less polarized-because it isn't.
(Editor's note: This is the final post for Slate's Big Sort blog. You can read more from Bill Bishop at his Web site covering rural America, the Daily Yonder, and you can read more about the Big Sort itself in his book of the same name.)
-
sponsorship
The model for the modern political campaign is the evangelical megachurch.
This isn't a partisan observation. Both George Bush in 2004 and Barack Obama adopted the basic organizing techniques that many ministers have been using since the 1970s to grow their churches to stupendous size. And why not? They work.
The megachurch was built on an idea born in India by an American missionary. Donald McGavran spent half a century overseas, and he used much of that time to discover the way churches could convert large numbers of people to Christianity. McGavran observed that converts didn't come to the church one by one. They came in groups. And those groups were socially coherent—castes, villages, or families. The key to church growth wasn't in bringing individuals to Christianity but in converting groups, peoples. And these groups would come if they were appealed to as a "homogenous unit."
"The individual does not think of himself as a self-sufficient unit, but as a part of the group," McGavran wrote in this 1955 book, The Bridges of God. People don't want to come to a church where they hear a different language or eat strange foods. "Men like to become Christians without crossing racial, linguistic, or class barriers," McGavran wrote. McGavran said ministers needed to understand the culture of their constituents and recommended that they use the insights of anthropology to tailor their appeals to homogenous groups.
The church establishment largely ignored McGavran. The missionary arrived in the United States at a time when the pews were crammed with parishioners. The mainline churches needed architects and builders in the 1950s, not some bearded missionary who had spent the last several decades in India. But church membership began to decline quite suddenly in the mid-1960s, and by the ‘70s a new generation of young ministers was looking for a way to stimulate church membership, to fulfill the Great Commission's call to "make disciples of all nations" (Matthew 28:16-20). They discovered Donald McGavran and began to use his ideas to create churches in the "homogenous units" (i.e., suburbs) being created on the edges of cities.
One of those young ministers was Rick Warren, who stumbled across an article on McGavran in an old Christian magazine. "As I sat there and read the article on Donald McGavran," Warren wrote in The Purpose-Driven Church, "I had no idea that it would dramatically impact the direction of my ministry."
Rick Warren's story is now megachurch legend. He founded his Saddleback Church in Orange County, Calif., and he used McGavran's techniques to attract members. Warren did his market research, even creating an anthropological composite of the target Saddleback member. He's "Saddleback Sam," a Docker-clad, cell-phone-carrying white guy who doesn't like organized religion and doesn't like neckties. That was the prototype for Warren's "homogenous unit," and so he built a church where Sam wouldn't have to cross many "barriers" to join. Sam wouldn't have to dress up for church, and he could look into the sanctuary and see more Sams, a "homogenous unit" of churchgoers. Warren stocked up on flowered shirts.
Ministers building these churches realized friends and neighbors were the best recruiters of new members. Like attracted like, which was an organizing tradition among evangelists. When supplicants answering the Rev. Billy Graham's altar call streamed to the foot of the stage, each would be met by one of the evangelist's helpers. The pairings weren't random. Graham insisted that young women meet young women. Older men greeted older men. Graham understood that the best way to cement the conversion was to show new believers a reflection of themselves within the church.
And the most effective recruiting tool was for friends to "witness" to friends their personal stories of salvation (Acts 1:8).
The marketing techniques all turned inward—friends talking to friends about their experiences in a church built for "people like us," which was the title of a popular book on church growth. The church wasn't designed to transform but to mirror a way of life. And the techniques worked.
It took until 2004 for what was common knowledge among megachurch chaplains to become the latest gimmick for selling a president. The Bush campaign in 2002 had experimented with different techniques to increase voter turnout, from door-to-door canvassers to the noxious (and utterly ineffective) robo-calls. Their tests found that personal contact with a voter was good but that an appeal coming from a friend or neighbor worked best. If it was clear to voters that the canvasser came from the same social hive, Bush campaign strategist Matthew Dowd told me, turnout jumped.
The Republican campaign in ‘04 recruited neighbors to contact neighbors, and it enlisted respected community members to serve as Bush "navigators," local surrogates for the president. The Bush organizers I talked with in Oregon and Minnesota instructed their all-volunteer canvassers to "witness" their support for the president. "We weren't there to convince anybody," one Oregon organizer told me. "We were there to give testimony of why we were for George Bush. And that's very religious."
The strategy was to reflect voters' beliefs and ways of life back on to themselves, so that the 2004 campaign wasn't as much about the re-election of a president and his policies as it was an affirmation of a local way of life.
The Democrats learned their lesson—they used paid workers who obviously were "not from around here" to do their canvassing in ‘04—and so this year the Obama campaign recruited an "army of persuasion" based on the Bush neighbor-to-neighbor model. In Wisconsin, the Obama campaign has hunters talking to hunters, women talking to women. At training sessions, "Obama Organizing Fellows" were taught to develop short, personal narratives that will explain to their neighbors how they came to support the Democrat—to witness.
Neighbors witnessing to neighbors is a marketing technique suited to Americans, who are increasingly sequestering themselves in communities, churches, and clubs with those who share similar ways of life and politics. The churches created over the last three decades have become some of the most politically segregated institutions in the country, a result of an organizing strategy built on the intentional molding of a "homogenous unit."
These tactics aren't designed to "sell" people something new or different but to show that the product (a church, a new concoction of PowerBar, a candidate) embodies the community's beliefs and lifestyle. "The message you've got to send, more than any other message, is that Barack Obama is just like us," Missouri Sen. Claire McCaskill told the Obama fellows, according to the Washington Post. Exactly.
Politicians have been packaging image from the beginning: McKinley sitting on the front porch, Truman speaking from the back of a train, Madison Avenue selling a new Nixon. In the end, however, the message was the same: "Vote for me." Campaigns today are doing something different. They attempt to manage behavior by creating a social environment that encourages people to vote for themselves. The most important message a campaign has to convey is one of flattery, that the candidate is "just like us."
Self-government, however, is the opposite of self-love. Democracy is about meeting and coming to terms with people who look, talk, believe, and think differently from us. Government might work better if that democratic exercise began for voters during the campaign rather than the day after inauguration.
-
sponsorship
Let's get something straight. There is no "women's vote." Women vote, of course. But reporters write about the "voting bloc" of white women as if it has meaning. It doesn't. Elections aren't about demography. They are about ways of life.
Marketing people have known that large demographic categories like "white women" are meaningless since at least 1973, when a New York adman asked in the Journal of Advertising, "Are Grace Slick and Tricia Nixon Cox the same person?" Both were white, young, urban, rich, former Finch College alums and women. Conventional market research in the early 1970s would have plugged in all this demographic data and concluded that, yes, the blond-haired daughter of the president and the lead singer for the psychedelic rock band Jefferson Airplane and author of the acid anthem "White Rabbit" were the same.
All of which made conventional market analysis meaningless, wrote John E. O'Toole, president of Foote, Cone & Belding Communications. It was so obvious the two had nothing in common that you didn't even need to go ask Alice. Tricia Nixon married a Republican White House aide on the White House lawn. The one time Slick was invited to the White House (a reception for Tricia's Finch classmates), she brought along her "bodyguard," Yippie founder and Chicago Seven defendant Abbie Hoffman. The two said they intended to spike the iced tea with LSD. The Secret Service didn't let them past the front door.
O'Toole wrote in '73 that statistics on income, age, and education—all the demographic "facts" we still use to divine presidential elections in 2008—had lost relevance because there had been a "Revolution of the Individual." People weren't living according to class or education or age. They were "forming liberation groups: black, feminist, gay, consumer, anything." Marketers had snoozed through the revolution and insulted people by discounting "their intelligence in favor of some vast common denominator."
People didn't define themselves by demographic markers, O'Toole wrote. They lived in groups "united by common attitudes or lifestyles or perceptions of themselves."
Marketing people long ago abandoned most of the demographic data that we still use to talk about politics. I talked with Chris Riley, a Portland, Ore., marketing guy who has worked for both Nike and Apple. Riley said people were forming communities of interest that had nothing to do with categories such as single, white, college-educated women. "I'm not allowed to use market research information, by dictate of (Apple founder) Steve Jobs," Riley said. "They don't trust it."
They don't trust it because demography—classifications such as a favorite from the Democratic primary, the "white working class"—doesn't get at how people live. "There is no (demographic) category for somebody who shapes his entire life around his concern for the environment," Riley explained.
After all, how many white, single women describe themselves that way. A San Diego woman told me recently that she was an "ocean oriented person." That's a more accurate political description in 2008.
Young ministers in the 1970s began designing churches for what one marketer called "image tribes." They created the modern American megachurch by catering to ways of life, not demographic types. Rick Warren of the Saddleback Church wears flowered shirts in part because his "target" audience, "Saddleback Sam," "prefers the casual and informal over the formal." Megachurches tailor their services to lifestyles. One California megachurch advertises different lifestyle venues for its Sunday morning service—a "Country Gospel" hoedown, a gathering known as "The Edge" (with Starbucks coffee and Mountain Dew), and a "Traditions" hall with music from a baby grand piano.
The Bush campaign in 2004 was the first to catch up with marketing techniques that had been refined over the past three decades. Bush identified individuals by how they lived—the cars they drove, magazines they read, television shows they watched, ring tones they downloaded.
John O'Toole in 1973 told marketers they misunderstood society by continuing to "shout at a crowd rather than talk to persons." Bush's campaign was the first to run a campaign aimed at individuals rather than crowds.
Barack Obama has essentially copied the Bush approach, identifying the "image tribes" we travel in rather than the bleak and only occasionally meaningful demographic categories that appear in exit polls and are coughed up in stories about politics.
So why do women seem to like Obama more than men? (Pew had women favoring Obama 54 percent to 37 percent at the end of September, while men backed McCain 47 percent to 43 percent.) We'll take that up tomorrow, and also answer the eternal happy hour question "Where are all the good men?"
-
sponsorship
"Palin and Biden Stake Their Claims on Change," rang the lead headline in the Times after the vice-presidential debate. Who could blame them? With several wars, Great Depression II, a blistering civil conflict in Pakistan, and European economic collapse, change would be welcome.
But there's something else going on with this "change" business, the reason why "change" is such an appealing message to Americans. It's not that an overwhelming number of voters want a change in particular policy. What Americans want universally is a change in the way government works.
Political writers say voters don't give a rip about governmental process. But these two campaigns are mostly about process, about how government works rather than what government does. Maybe that's because we care more about process than results, more about the way government operates rather than the policies government enacts.
That's the argument of two political scientists from the University of Nebraska, John Hibbing and Elizabeth Theiss-Morse. Policy doesn't drive people's vote, according to Hibbing and Theiss-Morse. People want good schools, peace, and a strong economy, certainly. But they don't have overwhelming concern about how these ends are achieved.
"What do people care about if they don't care about policy?" the two ask, and then answer. "We argue that people care deeply about process." (Hibbing and Theiss-Morse lay out their argument in their 2002 book, Stealth Democracy: Americans' Beliefs about How Government Should Work.)
People "despise pointless political conflict and they believe pointless political conflict is rampant in American politics today." And they overwhelmingly see a "political system in which decision makers—for no reason other than the fact that they are in a position to make decisions—accrue benefits at the expense of non-decision makers. Just as children are often less concerned with acquiring a privilege than with preventing their siblings from acquiring a privilege, citizens are usually less concerned with obtaining a policy outcome than with preventing others from using the process to feather their own nests."
OK, I'm like many of you. This sounds wrong, belittling. (The "children" reference is over the top.) Hibbing and Theiss-Morse aren't exactly thrilled about what they've found as they've polled and focus-grouped voters. But the Nebraska professors' thesis does explain why outsiders (governors from Georgia, California, Arkansas, and Texas) have won the presidency based largely on their outsiderness, that they didn't live "inside the Beltway."
Now we have the "maverick" (plus his Alaskan sidekick) and the guy who will deliver the "change we need/can believe in." Both candidates have promoted themselves as outsiders who promise to sweep out a government rotten with partisanship and corruption.
Hibbing and Theiss-Morse don't think people want to get rid of "insiders" so "the people" can take more direct control of government. They argue Americans have little interest in most policy proposals and even less interest in the day-to-day work of government. (Largely that's because people shy away from conflict, Hibbing and Theiss-Morse write, and politics is all about conflict.) What they want is for government to be open, accountable, and representative. Their perception is that government is none of those things, and so dissatisfaction with government "usually stems from perceptions of how government goes about its business, not what the government does."
Therefore, people's "main political goal is often limited to nothing more than achieving a process that will prevent decision makers from benefiting themselves."
The Nebraskans contend that the great power vested in government inevitably leads to this sense that politicians are "overpaid lackeys of special interests."
"[P]eople are amazingly sensitive to being played for suckers," they write. "Politicans are often in a position to do this, and as a result the people love to hate them." We are psychologically wired this way. It's not the policy positions politicians take that lead to our distrust, Hibbing and Theiss-Morse write; it is the power they hold.
Most people don't care about politics or policy. They just don't want to be ripped off. And so they vote, time and again, for the outsider, the maverick, the man from Hope, the peanut farmer, the Perot phenomenon in '92, the uniter not a divider, the guy who will deliver change from the bottom up.
This isn't the way I like to think about politics. And given the remarkable failures of the last eight years, it's hard to believe policy isn't a primary concern right now. But eight out of the last nine elections (including this one) have been shaped by the sense among voters that Washington is corrupt, politicians are crooks, and it will take a newcomer to get rid of the corruption and the partisanship.
I can't think of a single policy proposal that has had a similar impact.
(Top) Photo by SAUL LOEB/AFP/Getty Images; (Bottom) Photo by Peter Tompson/Getty Images
-
sponsorship
Rural voters in battleground states support John McCain by about the same percentage that they backed George Bush at this point in the 2004 election. In September 2004, Bush held a 13-point lead over John Kerry among rural voters in battleground states. This September, McCain holds a 10-point lead over Sen. Barack Obama.
Despite wars in two countries and a worldwide financial collapse, the 2008 election is static, locked in the divisions, rhetoric, and tactics of 2000 and 2004. The divide between rural and urban voters was among the most dramatic signs of geographic partisanship in 2004. George Bush came out of the nation's cities running 3.7 million votes behind John Kerry. He won rural counties by 4.1 million and then padded his margin in exurbia.
The poll of rural voters is just a bit more evidence that the talk about change and mavericks and a new kind of post-partisan politics is more than overheated. Americans are settling quite naturally into the voting patterns of the past two presidential elections.
The poll surveyed likely voters last week in 13 closely contested states. (They are Colorado, Florida, Iowa, Michigan, Minnesota, Missouri, Nevada, New Hampshire, New Mexico, Ohio, Pennsylvania, Virginia, and Wisconsin.) Greenberg Quinlan Rosner interviewed 742 likely voters living in rural communities. The poll was commissioned by the Center for Rural Strategies, a nonpartisan rural-advocacy group in Whitesburg, Ky. (As editor of the Daily Yonder, I am an employee of Rural Strategies.)
Greenberg Quinlan Rosner works exclusively for Democratic candidates. Bill Greener, a Republican strategist, helped shape the poll. The full poll can be found here.
Rural voters were clearly enamored with the selection of Gov. Sarah Palin for the Republican ticket, even if they were less certain she was ready to take over the presidency. More than half said the choice of the Alaskan governor made them more likely to vote for McCain. Thirty-one percent said they were less likely to vote for McCain because of Palin.
Rural voters liked Gov. Palin personally more than they were impressed by her qualifications. Some 65 percent of those polled said the Alaskan "represents the values of rural communities." Fewer, 54 percent, said she was "ready to be vice president and assume the presidency if need be."
Rural voters have warmed considerably to McCain since the spring. When asked who would do a "better job" on a range of issues, rural voters were increasingly likely to name McCain.
For example, in May rural voters thought Obama would do a better job than McCain on the nation's economy by a 44 percent to 36 percent margin. In this poll, however, rural voters now say McCain would do better with the economy by a 46 percent to 43 percent margin.
In September, Obama held a 10-point edge over McCain on the question of who would do a better job of "bringing the right kind of change." Now the two candidates are tied.
McCain moved up on every question—who is "on your side," who shares your values, who would do better in Iraq—while Obama lost ground or stayed the same since the May survey.
Fifty-three percent of those polled said their neighbors and their communities were "ready for a black president." Twenty-four percent said their neighbors and the people of their communities were not ready for a black president. A quite large number, 23 percent, answered this question by saying they didn't know, refused to answer, or that neither option was appropriate.
The poll was taken during the economic turmoil of last week, and rural voters named the "economy and jobs" as the most pressing issues facing them. When asked to compare the importance of the economy and values, 61 percent of those polled said the economy was the "most important thing" in the election. Only 36 percent said it was most important that the next president "reflects my values."
Pollster Anna Greenberg said she was surprised McCain's lead wasn't larger. She said the sample in this poll was slightly more Republican than the rural poll taken in May. And although rural voters were seeing McCain more favorably, that was not translating into decisions to vote for the Republican.
Republican Greener interpreted the poll as more evidence that Democrats have been unable to cross the cultural division that has been defined by the geography of rural and urban America. "When Sen. Obama says that people living in small towns cling to their guns and religion due to bitterness, or his supporters attack Gov. Palin for not being qualified to serve by making light of her background as the mayor of a small city, this all contributes to separating the Democrats from voters in rural areas," Greener said.
Democrats have told themselves that they were able to win U.S. Senate seats in Montana, Missouri, and Virginia in 2006 because they put up candidates who could attract votes in rural areas. (Missouri's Claire McCaskill, Virginia's Jim Webb, and Jon Tester from Montana were nicknamed the "redneck caucus.") When you look at the county-by-county results, however, these races were won because Democrats increased their turnout in the cities.
Obama has been hustling in rural places, but this poll shows that over the past four months, rural voters in swing states have not moved his way. They have, however, found more to like in John McCain.
George Bush won in 2004 because he was able to turn out voters in rural and exurban communities. The tactical question this year may be whether Obama can pull the same trick in the cities.
-
sponsorship
Americans aren't moving to be around others who agree about single-payer health plans or the proper response to a nuclear Iran. We seek out comfort among people who live like we do, think like we do, act like we do. On Election Day, we tend to vote like our neighbors, and so it looks like we have sorted ourselves intentionally into Republican and Democratic enclaves, but those divisions are more about lifestyle than policy.
That's why the primary race between Barack Obama and Hillary Clinton was just as geographically polarized as the contest between John Kerry and George Bush. In 2004, half the voters in that very close election lived in a county where either Kerry or Bush won by more than 20 percentage points. In the dead-even 2008 primary, exactly half the voters lived in counties where Barack Obama or Hillary Clinton won by a landslide.
Democrats in '08 were evenly split along the same geographic lines as those that evenly split Republicans and Democrats in 2004. The better George Bush did in a county, the more votes Hillary Clinton won. The map of voting results by county in Missouri from the 2004 general election looks exactly like the map from the 2008 Democratic primary. Obama won the Kerry counties; Clinton and Bush won the rest.
The race between ideological opposites in 2004 had the same geographic divisions as the race between Obama and Clinton, who are ideological twins.
When this phenomenon was first discovered early in the primary season, there was a brief behind-the-scenes debate. The Clinton people thought this was proof that their candidate would be able to pull Republican votes in the fall. The Obama camp dismissed the comparison. You can't extrapolate primary results to the general election, they said. Two different kettles of fish.
The Clinton people looked at these results and began to change the campaign's itineraries. For the next several months, Bill Clinton spent most of his afternoons speaking from the back of a pick-up parked in the courthouse square of some red county. Obama, meanwhile, rolled up big leads in the cities. (He won the southern metro areas by the same margins that Hillary Clinton took Appalachia, though we didn't hear much complaining or surprise about those lopsided victories.)
The primary ended with the Democratic Party just as divided as the nation was in '04 and in exactly the same way.
Sen. Obama is stuck. He didn't find a way across the boundaries of lifestyle and culture that split Democrats. And now it's the middle of September, and he still is trying to find a way to bring these very different Americas together. (It might be that there isn't one.)
John McCain may have come to another conclusion: that he doesn't need to widen his net and can win by turning out the same voters who elected Bush. More on that shortly.