Recently, I wrote here that the fundamentals of the 2008 race decisively favor Barack Obama. As is often the case, however, my words were met with a familiar riposte: What about the race factor? Are white Americans really ready to elect a black man as president? It's a recurrent refrain among Democrats and even some hopeful Republicans. As Andrew Kohut wrote here, "56 percent of Democrats believe that many people will not vote for Mr. Obama because he is black."

With the first ever African-American presidential candidate, race is certainly the great unknown of the 2008 campaign, but there is significant empirical evidence to suggest that Mr. Obama's skin color may be far less consequential than some believe -- and may even benefit him. At the very least, it is more complicated than many realize.

Arguments about race and the 2008 election play out on two levels: one, the notion that many white voters are "closet racists" and will not vote for a black man and two, that public opinion polling cannot be trusted because white voters are afraid to reveal their prejudices.

To the first point, it is incontrovertible that some whites will not vote for Barack Obama. We've come far as a nation; but we haven't come all the way. According to a recent Associated Press-Yahoo News poll, one-third of all white Democrats and independents have used a negative word to describe African-Americans, and racial antagonism may be costing Mr. Obama as much as six points in the polls.

However, these numbers are a bit suspect. The poll tested all Americans, not just voters and simply because someone has a partially negative view of African-Americans, it doesn't necessarily mean they won't vote for Barack Obama. The negative stereotypes that some white voters hold toward African-Americans may not necessarily have an impact on the way they think of Mr. Obama, particularly if he doesn't seem to fit those preconceived notions. And of course, many whites who would not vote for Barack Obama because he is African-American are unlikely to vote for any Democratic presidential candidate.

The likely voting patterns of whites are more complex than often assumed. First of all, there is a general belief that white working-class Americans vote solidly Republican. But outside of the South, Al Gore won this voting bloc in 2000 and John Kerry lost it barely in 2004. In the Deep South, however, white voters support Republicans by 3 to 1, and even higher margins (in Mississippi in 2004, 85 percent of whites voted for George Bush). As the political scientist Tom Schaller points out in his book "Whistling Past Dixie," Democrats don't have a white working-class problem; they have a Southern problem.

Indeed, for all the discussions about Mr. Obama's supposed exoticness to white Americans, a recent New York Times poll indicates that 66 percent of voters think Mr. Obama "shares the values most Americans live by" -- a higher number than that received by Mr. McCain.

But the real fly in the ointment of the white racist argument is that it ignores the other side of the equation: the behavior of non-white voters, which is approximately a third of the population.

For example, we take for granted the fact that Mr. Obama will likely reel in close to 95 percent of the African-American vote, but this could be a deciding factor on Election Day.

By some estimates an 8 percent jump in black turnout in Nevada, over 2004, would win him the state; in Florida, a 23 percent improvement could make the difference. In Ohio in 2004, George Bush won 16 percent of African-American voters; this year if Mr. Obama wins 95 percent of the black vote, he will not need a single additional voter over what John Kerry received in 2004.

In states like Michigan, Pennsylvania and Ohio the key for Democratic victory has generally been strong African-American turnout. So even in these three states if Mr. Obama loses some white voters because of his skin color he may cancel that out with support from non-white voters. And lest we forget, John Kerry and Al Gore both won Michigan and Pennsylvania while barely losing in Ohio.

But Mr. Obama's advantage stretches beyond African-Americans. Four years ago John Kerry won 53 percent of the Hispanic vote versus 44 percent for George Bush. According to the most recent poll of Hispanic voters, Mr. Obama leads in this crucial voting bloc by a whopping 66 percent to 23 percent margin. This advantage could be crucial in states like New Mexico, Colorado and Nevada and to a lesser extent, Iowa and Virginia.

Most telling, by 3-to-1, respondents said Mr. Obama's race would "help him as opposed to hurt him," putting to rest the notion of simmering racial tension between Hispanic and black voters. Indeed, Hillary Clinton's strong support among Hispanic voters in the Democratic primaries appears not to have been a case of "black-brown" rivalry, but simply that these voters preferred Mrs. Clinton then, and are returning to the Democratic fold now.

Even among smaller segments of the population, Mr. Obama is seeing an advantage. For example, among Arab-Americans, a small but important voting bloc in states like Michigan, Virginia and Ohio, the Democrat has a 21-point advantage.

Finally, there are surely some white voters for which Mr. Obama's race is a positive rather than a negative. This seems particularly true among younger voters, who are generally more racially tolerant and are supporting Mr. Obama in record numbers. One recent poll gives Mr. Obama a 27-point advantage among these voters; a stark contrast to 2004 when John Kerry won this bloc by a mere nine points. It's hard not to imagine that Mr. Obama's message of change, which is personified in large measure by his race, is influencing these voters.

Of course, we don't know for sure what the turnout levels will be for minority and young voters, but in any examination of race on the campaign trail they must be factored into consideration. However, when Mr. Obama's race is discussed, the positive political aspects of his background never seem to receive the same level of attention as the supposedly negative aspects.

The common retort to this avalanche of empirical evidence is to suggest that public opinion polling cannot be trusted. Skeptics evoke the so-called Bradley effect, which refers to the 1982 governor's race in California when Tom Bradley, who is African-American, lost the race to his white opponent George Deukmejian, even though he consistently led in pre-election polls. The idea is that white voters will lie to pollsters about their true intentions on Election Day for fear of appearing to hold racist views. However, the continued existence of a Bradley effect has been largely discredited.

Indeed, during the 2008 Democratic primaries there was no discernible Bradley effect, if anything there was a reverse Bradley effect, with Mr. Obama frequently outperforming pre-election polling results. The 2006 Senate race in Tennessee between Harold Ford and Bob Corker, one of the most racially charged in recent memory, saw the same phenomenon. Mr. Ford who is African-American overperformed pre-election polls. In fact, according to a recent survey by the Princeton Election Consortium, since 1996, black candidates have actually polled 0.3 percent lower than the final result.

Of course, some will argue that a vote for president is different than voting for a governor or senator. Perhaps. But in a fascinating breakdown of the Bradley effect, Adam Berinsky, a professor of political science at MIT, argues that in racial voting, political context matters. For example, Mr. Berinsky cites the 1989 New York mayoral race in which the African-American candidate, David Dinkins, badly underperformed in pre-election polling. But this may have been a case of white voters in the heavily Democratic city being uncomfortable telling pollsters they were going to vote against a black candidate who shared their party affiliation.

Four year later, after Mr. Dinkins' rather unsatisfactory performance as mayor, public opinion polls were spot on. After all, there was little reason to lie; one could simply explain their vote by the fact that Mr. Dinkins had done a poor job.

The same dynamic may dictate the 2008 race. Those who don't want to vote for Barack Obama because he is black can simply say they are voting for John McCain because he is the better candidate or more in tune with their political values. There is little reason to lie. Indeed, there would have been more reason to mislead in the Democratic primary where the two candidates held relatively similar policy positions and the vote was more about personal characteristics. But again, no Bradley effect was seen.

If there are large blocs of anti-Obama voters in current polling they would most likely be lurking in the "don't know" category of public opinion polling. But in the Gallup daily tracking poll, the number of don't knows is generally around 7 percent.

This group is a tiny segment of the voting public, and we don't know where these voters are situated (they could be a white union member in Massachusetts or a white businessman in Alabama, and neither of these states is up for grabs) or whether these individuals are lying in the first place. In short, those who argue that there will be a Bradley effect in the 2008 race have little more than supposition backing up their argument. The evidence simply isn't there.

In the end, we don't know how the race factor will play out on Election Day. No pollster can look into the soul of a voter; and the verdict cast in the voting booth is a highly personal decision where a whole series of political and social factors have an impact on the final choice. But examining what we do know about voting patterns suggests that fears of racial animus determining the presidential election are wildly overstated. Race may play a factor on Election Day; but then again it may not. And even if it does, it may provide more, not less benefit to Barack Obama.