The Bradley Effect

Whenever someone runs for office who isn't white, male, Christian, and straight, there are always questions of whether people lie to pollsters about whether they will actually vote for the candidate. Specifically questions have been raised this election cycle that people are lying to pollsters that they will vote for Obama in November, but when they get in the ballot booth their prejudice will come out and Obama will underperform by two to three percentage points. This is known as the Bradley Effect. Baseball whiz and political blogger Nate Silver discusses the Bradley Effect at fivethirtyeight.com, and finds evidence that it may actually work in reverse - people may understate their propensity to vote for a black candidate. The last example was Harold Ford Jr. vs. Bob Corker in 2006, and Harold Ford Jr. (a Penn grad, and African American) outperformed the latest polling data in the final election count.