Walters seems to be in denial that the Bradley Effect even exists:
The basis for the theory is that Bradley was leading in the polls right up to Election Day, yet lost the election. What Bradley Effect theorists miss is that the polls were actually quite accurate – as far as they went. Bradley won among voters who cast ballots on Election Day, as "exit polling" of voters confirmed. Based on those polls, in fact, many news outlets immediately declared Bradley the winner.
He goes on to blame absentee ballots for Bradley's ultimate defeat, but what Walter's not saying is that their exit polls showed Bradley winning "then" with a huge margin, which wasn't reflected in the actual ballot tallies that day, so either the exit pollsters (and all the others) were wrong or the Bradley Effect kicked in.
His other point:
Finally, the story of the 1982 election is not that Bradley was the victim of a hidden anti-black vote, but that he did so well during an era in which Republicans had the upper hand. Then-Gov. Jerry Brown lost his 1982 bid for the U.S. Senate (to Republican Pete Wilson) by five times the margin Bradley lost the governorship and Republicans dominated top-of-the-ticket elections throughout the 1980s in California.
still misses the mark - he's talking about ideological shifts that don't explain away the Bradley Effect - the polls didn't match the results. If the "Republicans had the upper hand", those polled would be less likely to lie about it in an attempt to appear "color-blind".
I always thought there was a twist on the Bradley Effect during the Democrat primaries this year - all the "undecideds" would about break to Hillary Clinton at the polling site. The Bradley Effect is usually even more devastating in a general election, Barack Hussein Obama playing the "race card" will just make it stronger.