The lead isn't what it seems to be...

When you look at the polls on television, the internet, or (gasp!) in the newspaper- you see the headline that Obama is winning. Nearly every poll taken these days has Obama ahead (they don't poll places that aren't in question- and Obama is winning the swing states and the nationwide polls). But, is it really over, like the media would have you believe?

McCain has been saying that he is going to come from behind and win the election. People laugh and assume he is just trying to show confidence- or possibly believe he is getting senile. His staff has had some infighting and there have been sharp barbs launched at Governor Palin. The forecast looks bleak, no doubt about it.

But, when you take a closer look at the situation, McCain isn't doing so badly. Zogby's most recent poll gives Obama an edge by 49-45 with 6% still undecided- with a 2.9% margin of error. The race is close, and McCain is gaining ground. While it would take some massive movement to assume that all six percent of the undecided would break for McCain, there is reason to believe that the poll numbers may be skewed towards Mr. Obama. If this is true, the election may be much closer than even Mr. McCain realizes.

In some circles, the Bradley (or Wilder) Effect is a hot topic of conversation. To save time and space, I'll simply link to the wikipedia article and let you read up on the basics for yourself. The left, and often the media, dismiss the phenomenon out of hand without ever really investigating the reasons that we have seen this happen in the past, and without concern over whether it could be a possibility now. However, it is important enough for scholars at Harvard and Yale to examine, so I believe it is worthy of a mention in this humble space.

The crux of the issue is simple. Black candidates have, in the past, had much more success in pre-election surveys than they have had at the ballot box. The most striking example is that of Tom Bradley, who had a wide lead in polls leading up to the 1982 California governor's election, but lost on election day. However, that is far from the only example. In 1989 Douglas Wilder had a 15 point lead in the polls leading up to election day, but won by only 6,700 votes in a race where 1.7 million ballots were cast. On the same day, David Dinkins won the race for Mayor of New York by 2 points, when polls had him leading by as much as 18%. Simply put, there is a pattern of African-American candidates polling better in pre-election surveys than they do on election day. At times, the difference is quite large.

Now, is there reason to believe that this will take place in the 2008 Presidential election? I believe there is. In many of the "swing states" there is a history of racial problems. Virginia and North Carolina are both areas where racial tensions have run high in the past. It isn't all unreasonable to expect that white voters in those areas may feel the need to deflect accusations and claim they're voting for a black candidate when they have no plan to actually do so. But, what about Ohio and Pennsylvania? Is this a possibility in areas that have not had the same racial history as the others? The answer will not be known until the final election results are in, but I believe it is foolish to dismiss the possibility that Obama's support will be markedly low when compared to the surveys being conducted now.