The Big Sort: Where you live, how you vote.

Still Undecided? Ask Your Neighbor.

Undecideds don't vote on issues. If they were interested in issues—or even in the personalities of candidates—they wouldn't be undecideds. "They are not radical, not liberal, not conservative, not reactionary," C. Wright Mills wrote in 1953, "they are inactionary; they are out of it." Mills was writing about the torpor of Eisenhower-era Americans, but this is also a good description of undecideds a day before the 2008 election.

How do you get "inactionaries" to vote? Mostly, undecideds split evenly. Or, Pew Research Center's Andy Kohut tells us, half of the undecideds don't vote at all.

In the era of the big sort, the strategy for turning out so-called base voters and undecideds is the same. Politics is no longer about convincing individual voters with long lists of proposals and policy papers. Campaigns move communities of people—neighborhoods, congregations, home-schoolers, hunters—to the polls. And as these "peoples" steam to the vote, they pull along the undecideds in their wake.

"Voting research trends have ... shown that independents are very susceptible to social influence, often turning to others in their local environment for advice on political matters," University of Maryland political scientist James Gimpel wrote at the Christian Science Monitor Web site Patchwork Nation. That makes sense. After all, if you don't know much about HD television, you turn to the neighborhood expert who can guide you through the "dynamic contrast ratios" of a purchase.

"As a result, independents' decision making can be predicted from knowing the viewpoints of those who live close to them," Gimpel continued. "Independents living in neighborhoods and households full of Democrats will be inclined to vote Democratic, and those interacting with Republicans will most often support the Republican. There are, of course, exceptions, but this pattern holds up in highly competitive elections."

A candidate doesn't win independents by moving to the ideological center. A campaign energizes partisans, and they will tow the undecided voters living nearby.

"In this sense, ‘base' and ‘independent' strategies complement one another and are not mutually exclusive," Gimpel wrote. "When independents begin to pay attention to the campaign, they'll typically turn to the partisans they live among for guidance. By playing to the base, a campaign is also likely to reach independent voters situated in the midst of that base."

Excited people increase turnout. Close races increase turnout. In this sense, both campaigns benefit from the sense that the polls are tightening.

Leaving a Red State Behind Might Turn It Blue

Over the past four years, more than 23 million Americans have moved from one state to another. Statistician Bob Cushing (co-author of The Big Sort) has been busy at the computer the last few weeks trying to understand how all those Mayflower trucks and U-Haul trailers might affect state results Tuesday.

We don't know who moved. We don't know their opinions or their party preferences. We do know, however, whether these movers came from a county that voted Democratic or Republican in 2004.* Most counties are overwhelmingly Republican or Democratic. So Bob counted the people moving into and out of the Republican and Democratic counties of swing states. What he found are some interesting changes in state populations over the last four years.  

For instance, Republican counties in Ohio lost 101,000 people while Democratic counties gained just over 6,000—a net change of 107,000 people in the Democrats' favor. (Bush's total margin in '04 was only 118,000 votes.)

Nevada was a red state in 2004. But according to Cushing's calculations, the net gain in Democratic counties topped 251,000 people. Meanwhile, Republican counties in Nevada lost 86,511 people.

New Hampshire, blue in '04, is still a tossup, perhaps because Republican counties there grew by 56,000 since 2004 while Democratic counties showed a net outflow of 40,000. Maybe that helps explain why John McCain zoomed to Peterborough on Sunday.

Virginia is trending Democratic partly because the Democratic counties in Virginia are growing faster than the Republican counties—a net increase of 32,000 people since the last election.

Some states showed little change at all. Indiana, for instance. Florida and Arizona showed huge increases in the population of Republican counties.

People move. When they do, they bring their politics with them. Over the last four years, there's been enough internal migration to change the vote in a half-dozen states. Those are the places both campaigns are visiting during these last few hours.

(*The Internal Revenue Service releases a file every year showing the number or returns and dependents who moved from one county to another during the previous years. This is a good count of how many families and people left one county and moved to another.)

This Campaign Has Gone Positively 19th Century

This election is getting out of hand. For one rally, I read, the locals rigged up six horses to a wagon big enough to carry a pipe organ and a glee club with 40 singers.

Was this a prop for another Barack Obama mega-rally? Or maybe a Sarah Palin revival? Nope. The six white horses were put to the service of Republican William McKinley in 1896. A two-score political singing group was nothing special back then. Campaigns were exercises in organizing large groups of marching partisans. In Sullivan County, Ind., that year, both parties organized glee clubs for their candidates—although only the Republicans thought of the wagon.

The parties discarded the mass rally as a political device before the turn of the last century, replacing the energy of the throng with a more isolated, individualistic kind of campaign advertising. The grand political rallies went missing for more than a century, but now they're back, and the raw expressions of political anger and feeling produced at these massive gatherings have shocked both candidates and the press. We're simply not accustomed to having people play such an active part in presidential political campaigns.

Just after the Civil War, campaigns were based on these mass movements. The war shaped both the language and the tactics of political campaigns. "Elections were treated like battles in which the two main armies (parties) concentrated on fielding the maximum number of troops (voters) on the battlefield (the polls) on election day," wrote historian Richard Jensen. Party organizers had been in the military and so that organized their parties like Civil War armies. The language we still use in politics came straight from the fields of Gettysburg and Antietam: the "opening gun," party "standard bearers," "last ditch stands," "war horses," "precinct captains," "rank and file" voters and "spoils of victory." Voters would chant battle cries, wave signs and flags, and they would march.

Voters (only men had the franchise) were extraordinarily loyal to their parties then. "Men spoke of political attachments in the same breath as loyalty to religion," Jensen wrote, "for as one Presbyterian historian explained, 'Every man ... is expected to stand up for the creed of his church as he does for the platform of his party.'" Nine out of 10 Americans were firmly committed to one party or the other. Local elections became proxies for the national battle between the parties.

All this roughly describes politics today, a time of intense party loyalty, increased straight-ticket voting, and stark political divisions directed, in part, by theology.

And, of course, the most visible sign that we have developed a 19th-century attitude toward politics is the return of the mass rally.

We should keep following this parallel between the 19th century and today because, as I said, the parties found reason to abandon the mass rally as a political device. The change began after Benjamin Harrison won with a military-style campaign in 1888. Harrison appointed John Wanamaker as his postmaster general, and Wanamaker set out to change the way Americans ran their elections.

John Wanamaker opened in Philadelphia what was considered the nation's first department store, and he brought his sales skills to Harrison's White House when he introduced the merchandizing style to politics. Wanamaker, by the way, is the guy who first said, "The customer is always right" and that half the money he spent on advertising was wasted but that he didn't know which half.

The military approach to political campaigns was unpredictable, Wanamaker realized. The problem with mass marches and parades led by horse-drawn wagons, according to one Wisconsin Democrat, was that the events "stir up the other side almost as much as their own. The trumpet that sounds the note of battle not only inspirits its friends but awakes its enemies."

The alternative was a campaign of "education" and isolation—that is advertising. Instead of relying on torch-lit parades, the merchandizing style printed pamphlets and newspaper ads. War cries were replaced with reasoned arguments. Instead of inspiring dependable supporters, the merchandizing campaign concentrated on identifying and then appealing to independents and waverers.

The goal wasn't to increase turnout. It was to control the vote. And it was effective. The merchandizing style replaced the mass marches and rallies within a few election cycles.

Voters were unimpressed. "The voters' reaction to the new style educational campaigns was lethargic," Jensen reported. Partisan loyalty declined, apathy increased, and voter turnout plummeted. Advertising was more expensive than torch lights, so fundraising became a political preoccupation. The parties had more control over the electorate, but voters were dispirited. Isolated by the merchandizing style, they stopped coming to the polls.

Now we have two candidates who began this campaign as leaders of a post-partisan future running in an election that has become partisan in the extreme. Nine out of 10 Republicans say they'll vote for McCain, and nine out of 10 Democrats say they'll vote for Obama—party loyalty that matches that of the military campaigns of the late 19th century. Local elections have become nationalized, and people show willingness to march and demonstrate in ways that have been missing for the past 100 years. Funny how things work out.

If This Is a "Change" Election, Then What's Changed?

Let's consider what's not new in this election.

There's a lot. The last five or six elections have been pushed along by trends that have been in place since the mid-1970s. Despite the extraordinary circumstances this year, the basic political contours of the country haven't changed (or haven't changed yet!).

If anything, 2008 appears to be more an extension of the 2006 midterms, an election that changed little in the country's basic political makeup from 2004—except, of course, for the name of the winning party. More on that tomorrow. Today, let's consider how static our politics have been.

Churchgoers Are Still Republicans
Thirty years ago, how often you went to church didn't mark you as a Democrat or a Republican. Evangelicals didn't have a party.

As the parties sorted according to lifestyle instead of class, weekly churchgoers and evangelicals became reliably Republican voters in presidential races. There's no evidence this is changing. Oh, there have been plenty of stories about the breakup of the evangelical vote. I'd read the stories, but the more hardheaded pollsters and religion scholars would find, as John Green did last month, that "Barack Obama's attempt to reach out to Christian voters ... is failing."

In the fall of 2004, George Bush had a 60.4 percent to 19.6 percent edge over John Kerry among evangelicals. This year, Green found, McCain leads Obama 57.2 percent to 19.9 percent. Maybe that will change, but it hasn't yet, according to Gallup. Evangelicals and churchgoers may be "lukewarm" about McCain, but they are still supporting him in numbers just a smidgen below 2004 levels.

Women Voting Democratic
In the 1970s, more women voted Republican than men. Over the past 30 years, they have increasingly voted Democratic. Again, there was a spate of stories about a reversal in this arrangement, but by late September Gallup had women supporting Obama 52 percent to 39 percent.

Fewer Genuine Independents
Political reporters love the story about the rise of the independent voter and the "decline of parties." But over the past 30 years, the number of true independents has declined, and allegiance to party has grown stronger. (Princeton's Larry Bartels wrote the most important paper on this phenomenon.)

Yes, there are more people who register as independent or tell pollsters they are independent. But almost all these people vote reliably for one party or the other. People tell pollsters that they are independents, but when pressed, they admit that they almost always vote for the same party.

Split-Ticket Voters Disappearing
Split-ticket voting has been declining for the past 30 years, too. We are less inclined to pick and choose between parties. People are picking sides and voting that way up and down the ballot. That was especially true in 2006, when Democrats, especially, cast large numbers of straight-ticket votes in New Hampshire, Iowa, Wisconsin, and Pennsylvania.

Rural Is Still Republican
Rural voters have been moving toward the Republican Party since the '70s. That trend continues, too. 

Replay 2006
Democrats thought the 2006 midterms were a turning point. They weren't. All these trends stayed in place. Just the results changed. Tomorrow we'll see why 2006 is a good model for 2008.