There's an unpleasant truth the Obama campaign may be avoiding: An important slice of voters who are still undecided about their presidential choice have some negative racial attitudes that could well be triggered before Election Day.
This shows up in private polls that probe more deeply into racial angles than the public polls have. This is potentially devastating for Barack Obama and the Democrats, but isn't an especially surprising finding when you think about it.
Most undecided voters simply haven't focused on the elections yet, something people who read columns like this find hard to believe. Those who have managed to ignore the loud and endless campaign tend to be less educated and well-off than the average voter, and more prone to negative racial feelings. If the white swing voters break in proportion to their racial attitudes, Obama could be sunk.
This polling indicates something else astonishing to the politically plugged in: Many undecideds haven't really connected their negative feelings about race to Obama yet. Their view of Obama is unformed, and their negative feelings toward African-Americans could be easily triggered when they finally tune in.
The Obama campaign has been accused of being hunkered down on "the race card," wishfully thinking 2008 is a normal campaign where the candidate's race is not a major new variable. They're said to believe that since all the meta-conditions (terrible economy, pessimistic attitudes about the direction of the nation, country at war, unpopular president) point to a Democratic win, running a tactically sound campaign with a traditional platform and a charismatic candidate is enough to win.
It may be.
But some who have been doing recent research on race believe there is a current of racism that has not been triggered and that is likely to be — perhaps triggered intentionally by Republicans, but also as a natural consequence of the undecided voters finally focusing. And plenty of pundits and advice-givers think Obama is not doing enough to minimize or counter the racial impulses of undecided voters. (I am not convinced there is any way to spin this: What is, is.)
Does this mean John McCain, to capture the undecided vote, needs to actively trigger subterranean prejudices? Hard to say. But it is clear that many legitimate issues in the campaign also have racial angles or, to be fair, will be perceived through a prejudiced lens by some voters.
The "inexperience" issue is one of them. Obama has, in fact, less big league experience in government and politics than McCain or any other party nominee since JFK in 1960. If a voter is so inclined, this can play out as Obama coming off as an "affirmative action candidate."
For example, in a recent column, I said, "If Obama were not black, if he were the same man (man, not woman) in white skin, he would most certainly be far ahead in the polls." I think this is an uncontroversial assertion: All the models political scientists use to predict presidential elections have the Democrat winning in a rout. But now the polls are tied in a knot. What is the variable that none of the models take into account? Race.
After the column ran, my inbox was besieged by angry missives declaring that I was a half-wit for not understanding that Obama was given the nomination, not because he earned it by getting more votes, but because he was black and got what he didn't deserve. He was an affirmative action candidate. These were not kind-hearted e-mails.
So as legitimate as the experience issue is, for some it triggers a notion that Obama is undeserving, presumptuous or "uppity," to use an old and ugly word. If Democratic pollsters see that, so do Republicans.
Similarly, many people felt McCain's famous ad that called Obama a celebrity like Paris Hilton was subliminally racist, subtly playing on racist impulses that fear black men with white women, or that preyed on the idea that black men succeed only in celebrity arenas like sports and music.
Since there has never been a viable black candidate for the White House, prediction is hazardous. But there is some evidence besides gut feelings, a sense of American history and this recent polling that leads to an uncharitable view of how race will affect the vote.
In the primaries, Obama did poorly with the kinds of people who tend to be undecided now — and they were Democrats.
In Obama's home state, Illinois, Obama creamed Clinton by 32 points, but Clinton carried 55 percent of the least educated white voters. In Ohio, Clinton beat Obama by a 57-point margin among less educated white voters. In Missouri, 17 percent of the white voters said race mattered to them and 65 percent of those voters went to Clinton.
The famous "Bradley effect" also played out in the Democratic primaries. In 1982, Tom Bradley, the African-American mayor of Los Angeles, lost the gubernatorial race despite leading in pre-election polls and election polls. In 1989 in Virginia, Doug Wilder was leading by 15 points in the final polls, but won by just 6,700 votes to become the country's first elected black governor.
Yes, that's old news. But an analysis by political scientist Anthony Greenwald for the Pew Center for People and the Press shows that in the early primaries and on Super Tuesday, pre-election polls did exaggerate Obama's performance.
The theory of the Bradley effect is that people don't like to admit prejudice to pollsters, or they don't recognize it in themselves. But working with the Gallup organization in 2005, political scientist David C. Wilson found that 17 percent of white people polled don't trust black people (39 percent of black people said they didn't trust white people).
There's another kind of polling worth noting. Since, as noted, few people are eager to reveal racism or bias, it is important to look at how African-Americans perceive racial attitudes. A CBS News/New York Times poll conducted this summer found that 70 percent of the black respondents said they encountered specific instances of racial discrimination. The figure is higher than it was in 2000.
If McCain's own campaign doesn't try to trigger racial antipathies, the Swift Boat types are likely to try. There surely is a current of resentment they can prey on, just as there is strong tide that carried Obama to the nomination. I doubt there are rhetorical devices and campaign strategies that can blunt something as elemental as racial prejudice. But if there are, I hope the Obama campaign uses them well.