Monday, November 3, 2008

'Bradley Effect' Fiction

While conventional wisdom is that no Republican could win in this environment, none of the pre-convention polling data supports that conclusion. True, generic Democrats were beating generic Republicans by a large margin, but when the names were attached (Barack Obama and John McCain) the polls tightened to within a few points. The proof of this is the numerous almost countless stories during the summer asking why Obama's lead wasn't greater. The media then answered their own question and blamed it on "white blue-collar working class" voters convincing themselves Obama couldn't "close the deal" Unfortunately for John McCain, it appears he believed them even though there wasn't much evidence to support the claims that whites weren't going to vote for a black candidate. This may prove to be the first of many arguments that supports banning TV sets from campaign offices.

There has been a lot of hype about an anticipated "Bradley effect" also known as the "Wilder effect". The premise of these theories is that whites will tell pollsters they will vote for the black candidate, but in the voting booth, don't. Analysis of previous elections finds that this phenomenon was as high as 3.1 percent prior to 1996 but has been reduced to nearly insignificant now.

An analysis of the polls prior to the primary elections of 2008 showed no evidence whatsoever of a "Bradley effect" when compared to actual results. It should be readily apparent to any campaign analyst that a "Bradley effect" if existent, was not going to be a major factor with Barack Obama. The first sign of this was Iowa. Barack Obama won the Iowa Caucuses by significant margins. Iowa has a white population of nearly 95% and an African American population of nearly 2.5%. This media driven hype plays a big part in why McCain's campaign believes they still have a prayer and yet in State after State during the primary campaign, with the one exception of New Hampshire, Obama consistently ended up on election night +/- 2 points of where the polls had him. Even in New Hampshire, Obama received his predicted total but the large undecided number broke for Hillary Clinton. McCain enjoys no such undecided

So what does the summer polling tell us? Simply put, the polling, when analyzed by any metric, demonstrates that while a generic Republican should be in trouble, this Republican was different. This Republican could have won. In other words John McCain's narrative, war hero, statesman, straight-talking maverick, brought him 98% of the way back from the hole the generic Republican was in. What is killing John McCain's campaign is, frankly, John McCain's campaign.