We have a historic nominee this year, and with that comes hours and hours of drivel from pundits in the traditional media. Much of that time has been spent on the "Bradley effect." From what I gather, this manifests itself in two ways: 1) people in pre-election polls saying they will vote for the black candidate or that they remain undecided, but in the end will vote for the white candidate, and 2) people in exit polls saying they voted for the black candidate when they have in-fact voted for the white candidate.
Personally, I've seen enough evidence to consider the matter debunked. The pollster for Bradley's opponent thinks that the pre-election polls were inaccurate. Wilder, last night on the Rachel Maddow Show, said that he thinks that polling methodology has vastly improved in the last 20 years. Additionally, a more rigorous analysis shows that the effect has disappeared.
However, the 2nd manifestation of the Bradley effect worries me. As we have witnessed repeatedly over this election, facts haven't stopped the traditional media from continuing any narrative they find interesting. So, this brings me to a thought I just had: will the perception of a Bradley effect change the way the networks call the election? Since the debacle in FL in 2000, the networks have already become very conservative in how they call races. If they are worried about the Bradley effect, will they delay calling states that they would normally call if the race were b/w two white candidates? In the end, I don't think that the network delays will have an impact on the outcome, but it could mean that Obama won't be declared the unofficial President-elect until late into the night.