Tag Archives: Bradley effect

Elections have consequences

Here's something odd I've noticed.  Kathleen Parker's column used to appear regularly in the Chicago Tribune, but it almost never appeared in the Washington Post, despite her being syndicated by the Washington Post Writer's Group.  Now it appears regularly in the Post (whose op-ed page I read every day (though I am not really sure why–perhaps someone can suggest some other papers for me to read).  The difference between now and then of course is her arguing that Sarah Palin isn't qualified to be VP.  (No argument here on that score).  Perhaps she figured that if she continued to insist on what she has long been insisting on in the face of mountains of evidence to the contrary, she would continue to appear in the Tribune and on Fox, but not in the Washington Post and on CNN.  Whatever her personal motivation, it doesn't really matter.  Despite dumping McCain/Palin, she still reasons badly.  

Today she writes about a possible "reverse Bradley effect" in favor of Obama.  For those of you who don't know:

Among the hidden factors is the so-called Bradley Effect, meaning that whites lie to pollsters about their support for a black candidate. It is cited as the reason Los Angeles Mayor Tom Bradley lost to George Deukmejian in the 1982 California governor's race, despite polls showing him up to seven points ahead.

And what is the evidence for the soothing belief in an even bigger margin than the one Obama currently enjoys?

I've received too many e-mails and had too many conversations that began, "Just between you and me," and ended with, "I wouldn't want anyone at work to know," to believe that this is an insignificant trend.

Right.  And no one I know voted for Richard Nixon.  Among Zogby, Gallup, and so on, one does not see Kathleen Parker's email inbox.  Without any data, she continues to fantasize:

Sitting quietly at their desks are an unknown number of discreet conservatives who surprise themselves as they mull their options. Appalled by McCain's erratic behavior, both in dealing with the financial crisis and his selection of an unsuitable running mate, they will quietly (and with considerable trepidation) vote for Obama.

Are they are worried about higher taxes, a premature withdrawal from Iraq, and Obama's inexperience in matters executive? You betcha. But they do not want to vote for a divisive, anti-intellectual ticket headed by a man who, though they admire him, lately has made them embarrassed to be Republicans.

Should Obama win, it will be in part because some number of quiet, mostly white-collar men and women who speak Republican in public voted Democratic in private.

Notice that she has moved from the rather weak claim that there may be some of these reverse Bradley voters out there (something which may be true in some small way), to the rather more significant claim that they would be significantly responsible for an Obama victory, despite the fact that Obama is leading all over the place by significant margins.  This would mean that a vast number of people have consistently misrepresented their preference in the upcoming election, and that, get this, an even greater number of people are lying the other way.  So more people are lying that they won't vote for Obama than people are lying that they will.  That's some messed up reasoning. 

But this gets even more twisted.  She concludes,

Whatever the final tally, Obama should not interpret his victory as a mandate. Many of the Reverse-Bradley ballots won't have been votes cast for Obama, but against a campaign turned ugly. They also will have been delivered with solemn prayers that Obama will govern as the centrist, pragmatic leader he is capable of being.

Let me get this straight.  Because there could be a better opponent than McCain/Palin for Obama, people are voting for Obama because of that, and so any Obama victory is rather a defeat for McCain/Palin–but by no means an endorsement of Obama.  All this because of Parker's email poll.

Kathleen Parker, against McCain/Palin, but still loopy.