As the traditional media scrambles desperately to make a horse race out of the systematic annihilation of the Republican era, both cable news and print media hacks have come to increasingly rely on Bradley Effect to explain why all the hard data "may be wrong" "according to some [unnamed] analysts." Thanks to the internet, however, there has been a significant, informed push back against this ratings-driven fear mongering, from humble bloggers pointing out obvious logical holes in the argument to statisticians taking apart the numbers. Recently, the pollster of Bradley's opponent even wrote an article to point out that his campaign's internals were, in fact, on the money.
In the article, Tarrance writes:
The hype surrounding the Bradley Effect has evolved to where some political pundits believe in 2008 that Obama must win in the national pre-election polls by 6-9 points before he can be assured a victory. That’s absurd. There won’t be a 6-9 point Bradley Effect –- there can’t be, since few national polls show a large enough amount of undecided voters and it's in the undecided column where racism supposedly hides.
The other reason I reject the Bradley Effect in 2008 is because there was not a Bradley Effect in the 1982 California Governor’s race, either. Even though Tom Bradley had been slightly ahead in the polls in 1982, due to sampling error, it was statistically too close to call.
Deukmejian goes on to point out that undecideds broke towards Bradley's opponent in the final week of the campaign (a week in which, not coincidentally, Bradley made several prominent gaffes). Certainly, it's possible that this could occur between Obama and McCain, though as I pointed out in my diary yesterday, the shrinking pool of undecideds are becoming increasingly irrelevant as Obama's coalition crosses the 50% threshold.
Today, the Bradley Myth took another hit. After ABCNews published an error-laden front page article that falsely asserted that "people lied to pollsters saying they would vote for Bradley when they did not" (in fact, the theory concerns undecideds, all the polls accurately predicted Bradley's share of the vote), their own pollster has now written a scathing article that debunks the statistical bogeyman as "a theory in search of data."
Gary Langer, showing more integrity than the traditional media has shown in months, points out the obvious historical facts:
Two problems: There are, in fact, beaucoups reasons beyond lying that those polls could have been wrong. And there have been plenty of accurate polls in such contests since.
Consider 2006: Good-quality polls were accurate in five U.S. Senate or gubernatorial elections in which white and African-American candidates faced off. (There was no good, publicly released polling in a sixth). One, in Massachusetts, understated the white candidate by 5 points (not that it mattered in a 55-35 percent blowout); another, in Pennsylvania, understated the black candidate by 5 points (in a 60-40 rout). But as for a consistent Bradley or Wilder effect: It didn’t happen.
Furthermore, Harold Ford won 47% of the vote in Tennessee, suggesting that race is not necessarily an overwhelming impediment, even in deep Appalachia (add to this the fact that Wilder, who did suffer from an alleged " Bradley Effect," still went on to win Virginia when that state had far less Dem-friendly demographics than it does today). Langer then goes on to point out a recent study that, while finding some variance between poll projections and election results, concluded that such differences have vanished and, more importantly, that the differences were always rather small (i.e., not the 5 to 10 point margin hyped by pundits and internet panic-stokers).
But what about the New York Times' suggestion that the angry racists may simply be hanging up on pollsters? Langer says it's unfounded:
Another concern focuses not on directly lying, but simply ducking the question by declining to state a vote preference at all. In our polls this year, however, the number of respondents who've either declined to answer the horse-race question, or say they have no opinion, is no higher than in past elections, nor do they look different demographically in any way that would suggest that their absence creates bias.
Langer also goes on to show that recent studies show that the race of the interviewer does not appear to have an effect on the respondent's answers, though as I've pointed out frequently, concern about interviewers has been made almost entirely irrelevant by the prevalent use of computerized polling.
So what are we to conclude from all this?
First of all, there is absolutely zero basis for the widespread suggestion that people are lying about supporting Barack Obama. The Bradley Effect has never been a matter of people lying about supporting the African-American candidate. This is nothing more than a further distortion of what may have always been a myth. "The Bradley Effect" has always been a matter of undecideds breaking heavily for the white candidate on the final day of the election.
However, studies have found that even the 'undecided' question is overstated. It has been 16 years since the last example of an alleged "Effect," and even then, it was only worth a point or two for the white candidate.
With that said, it may be reasonable to assume that John McCain will win the majority of undecideds on the basis of other factors, but with Obama earning more than 50% of the vote in nearly every battleground state, this should not be a significant concern. This simply means that Obama just needs to get out the vote, and there are many reasons to assume that he will have a much easier job in doing that. First of all, his coalition represents the majority of registered voters. Second, the fact that many McCain-leaning voters identify as 'undecided' means that many of these people may not even bother to head to the voting booth.
So again, those worried about a Bradley Effect should relax. There is no credible basis for assuming that the growing consensus of public polls is wrong. Obama has the coalition. He has the support. He just needs to get his 51% into the voting booth.