Source: Los Angeles Times

Race is the most enduring cleavage in American politics. It divided the authors of the Constitution and fueled the Civil War; afterward, Southern Democrats invoked white supremacy to maintain a one-party system. The success of the civil rights movement undermined that system; but after 1968, Republicans used a Southern strategy, employing both subtle and not-so-subtle racial appeals to lure white voters away from the Democratic Party.

As social scientists have demonstrated, Americans no longer subscribe to the explicit racism that denied blacks the vote and refused them entry into the same public facilities as whites. But many white Americans still harbor degrees of conscious or unconscious resentment against blacks. In an analysis of the extensive surveys conducted by the American National Election Studies, polling analyst Scott Winship and I found that this kind of resentment remains, most commonly among men rather than women, the less well-to-do rather than the wealthy, those who lack a college degree, those who work at blue-collar rather than white-collar jobs and those who live in small towns in the Midwest and South.

Thus, as the 2008 presidential election got underway in earnest last year, I certainly believed that these voters -- not just Republicans but Democrats as well -- would take exception to an African American presidential candidate and perhaps even deny Barack Obama the presidency. And the McCain campaign appeared to have been aware of these voters. While eschewing explicit racial appeals, it did play on the contrast between John McCain as an "American president Americans have been waiting for" and Obama's African roots.

McCain's campaign also portrayed Obama's tax cuts as "welfare to those who pay none" -- which is the kind of formulation Republicans have used repeatedly through the years to suggest that Democrats want to transfer money from hardworking whites to indolent blacks. And in the last weeks of the campaign, an independent Republican group ran incendiary ads trying to link Obama to the Rev. Jeremiah A. Wright Jr., his black former pastor.

Given all this, there was good reason to expect some kind of white backlash at the polls. And it did come -- among exactly those groups that Winship and I had identified. But in the end, it didn't matter enough to decide the election.

How do I know this? The exit polls ask only one question about race, and voters aren't always forthcoming on whether race figured in their vote, but one can draw some inferences from this question and from the results of the election. The polls asked whether race was "the single most important factor," "one of several important factors" or "a minor factor" in the voter's choice. Those who said it was the most important or one of several important factors -- 9% of overall voters -- backed Obama by 53% to 46%. That suggests Obama's race was a net plus for him. It brought blacks to the polls, and some whites as well who liked the idea of having an African American president of the United States.

But did Obama lose votes because of his race? There is evidence that he did. It comes from comparing how Obama did this year with how John Kerry did in 2004.

Leaving race aside, one can argue that Obama should have done better this year than Kerry did in 2004. He was a better campaigner; he had more money and a more extensive campaign; his opponent was not as popular among voters as George W. Bush had been in 2004; and he could place the blame for the growing economic crisis on his opponent's party. But despite that, there were states and counties where Obama did much worse than Kerry among white voters. In Alabama, for instance, Kerry won 19% of white voters in 2004; in 2008, Obama got only 10%. In Mississippi, Kerry won 18% of white males; Obama won a measly 9%.

Obama did noticeably worse than Kerry in counties in Texas, Louisiana, Arkansas, Tennessee, Oklahoma, Kentucky, Mississippi, Alabama, West Virginia and western Pennsylvania. These counties are predominantly white, rural or small town, and downscale. In Tennessee's Benton County, which is west of Nashville, Kerry got 54% of the vote in 2004, but Obama netted only 41% this year. Benton is more than 95% white, and almost 20% of the citizens fall below the poverty line. In western Pennsylvania, Obama lost two white, small-town counties -- Fayette and Beaver -- that Kerry won in 2004.

Obama won Pennsylvania because college-educated whites, blacks and Latinos in the cities and suburbs around Philadelphia backed him. But he lost all the other states where his margin among whites compared unfavorably with Kerry's. One can argue, of course, that Obama might not have won any of these states even if he had been white -- and that may be the case with Alabama, Mississippi, Oklahoma and Texas. But Bill Clinton won Arkansas, Louisiana, Kentucky, Tennessee and West Virginia twice. In a toxic year for Republican candidates, a white Democrat should have been able to win two or more of these states. The main reason Obama could not was because he was African American.

On the other hand, race didn't seem to have any effect in some states. Obama lost Nebraska, for instance, to McCain by 57% to 41%, but in the U.S. Senate race, white Democrat Scott Kleeb lost to Republican Mike Johanns by virtually the same percentage, 58% to 40%.

In predominantly white New Hampshire, Obama did better than Democratic Senate candidate Jeanne Shaheen. While she defeated incumbent John Sununu by 52% to 45%, Obama carried the state by 55% to 45%. It's probably fair to conclude that Obama's race was irrelevant to many of these voters.

Then there were states where Obama had previously encountered resistance on account of his race but that he carried in the election. These include some of the crucial battleground states. When Hillary Rodham Clinton defeated Obama in Ohio's Democratic primary last March, 20% of Ohio's voters said race was an "important factor" in their decision, according to exit polls. Of these, 59% voted for Clinton. That suggested about 12% of Ohio's white Democrats and Democratic-leaning independents opposed Obama partly or primarily because of his race. Primary voters, of course, are a small percentage of those who vote in a general election and, on average, are less likely to take race into account. So it was probably fair to assume that Obama could lose as much as 20% of the Democratic vote in November because of his race.

But in the general election, Obama carried Ohio Democrats by 9 to 1, and improved on Kerry's totals from 2004 among white voters. There was evidence that race was a factor -- 19% said it was "important" in their vote -- but Obama got these voters by 52% to 47%.

Why? In interviewing people who did canvassing for Obama, I heard the same refrain: "Economic issues trumped race." That doesn't mean white voters in Ohio harbored no racial prejudice; it means they put their feelings aside, or subordinated them, in deciding to vote for Obama over McCain. That's a psychological process that is common in many of the decisions we make. We do or decide something in spite of feelings to the contrary.

And that's what happened in many of these battleground states during this election.

Obama's success in the election is testimony to the "post-racial" campaign he ran and to many voters' preoccupation with a sinking economy. But it also shows the extent to which race is no longer the great dividing line in American politics.

Click here to see this article as it originally appeared in The Los Angeles Times.