Two op-eds appear today — one in the New York Times and one in the Wall Street Journal — each debunking the original Bradley Effect. That is, each of them (the Times version is by a former Bradley aide, and the Journal version by a former aide to George Deukmejian, who won) makes the point that the best, most up-to-date polling data at the time showed that Tom Bradley was going to lose, and that it had nothing to do with race, but with his liberal policies and heavily Republican absentee ballots.
BTW, is it just me, or is it always (improperly) assumed that, when whites say they will vote for a black candidate and they don’t, that the lie is because they don’t want to vote for a black candidate — rather than that they just don’t want to admit that they aren’t planning to vote for the black candidate? That is, maybe the overstatement of white support for black candidates (assuming this does happen) isn’t because of white racism, but because of white reluctance to give politically incorrect answers to pollsters, which is quite different. Note that the same people who assert the existence of the Bradley Effect also discount surveys that show a decline in white racism, since they don’t trust whites to tell the (politically incorrect) truth to pollsters in THAT context.