Prof. Susan Estrich writes about the Bradley effect:
If
you paid attention to the gushers in the press and punditry in the days
leading up to Super Tuesday, Hillary was on her way to the morgue,
murdered by her crazy husband’s loose talk, abandoned by young voters
and women and anti-war Democrats, and anyone else they could think of.
. . . That was,
according to the pollsters, the problem: about 10 percent of the
electorate claimed that they were going to vote for him, and in many
cases even told pollsters that they did, but they lied. . . . Racism in America. . . . If this is happening even among us good Democrats, what does that say about Obama’s strength in a general election?
Nothing. This has nothing to do with Senator Obama. Stop listening to the gushers in the press and punditry. Outlier polls get more publicity. Who wants to hype the poll that shows not much has changed? If you look at all of the data, instead of cherry-picking, you see a much clearer picture. The Center for Congressional and Presidential Studies says:
The polls appear to have been off by just one delegate in AZ, MO, NJ,
TN and UT, by two delegates in AL, CT and DE, and by 3 delegates in MA,
NY, and OK. [T]he poll-based estimates were off by 6 delegates [in California].
In other words, +/- 3%. Just like the pollsters said.