I JUST DON'T BELIEVE THE PALIN NUMBERS
McClatchy-Marist has a new poll out with the usual (mostly) grim news for President Obama:
A new McClatchy-Marist poll finds that Obama looks increasingly vulnerable in next year's election, with a majority of voters believing he'll lose to any Republican, a solid plurality saying they'll definitely vote against him and most potential Republican challengers gaining on him....
While most of the Republicans have gained on Obama, he still leads all of the announced candidates....
But this result is bizarre -- it completely contradicts every other comparable poll result I've seen this year:
The biggest gain came for Palin, the former Alaska governor who hasn't yet announced whether she'll jump into the fast-changing race for the 2012 Republican presidential nomination.
After trailing Obama by more than 20 percentage points in polls all year, the new national survey, taken Sept. 13-14, found Palin trailing the president by just 5 points, 49-44 percent. The key reason: She now leads Obama among independents, a sharp turnaround.
Real Clear Politics lists 34 polls taken this year of a head-to-head matchup between Palin and Obama. In every single one, Obama led by double digits. In 24 of them, he led by at least 15 points.
The last McClatchy-Marist poll of this matchup was taken in early August. Obama led Palin by 21. Now it's 5? She's closed 16 points of the gap in six weeks? Really?
By comparison, Obama had a 5-point lead over Romney in August and has a 2-point lead now. That makes sense. Obama had a 19-point lead over the not-quite-declared Perry in August and has a 9-point lead now. That makes sense. Obama had a 17-point lead over Bachmann in August and has a 13-point lead now. That makes sense.
The Palin numbers don't make sense. I don't believe them -- not until some other pollster replicates them. I think it's just a bizarre and unrepresentative sampling of "independents" (we're told that Palin "now leads Obama among independents, a sharp turnaround").*
****
However, I will note that this poll was taken in the two days after the last presidential debate -- the one in which Rick Perry supposedly took a lot of serious hits -- and Perry's leading Romney among Republicans and GOP leaners, 30%-22%. In the August poll, Romney led Perry 21%-18%. Still waiting for that Dean-scream decline from Perry....
****
*TO CLARIFY: If I had to guess, I'd assume that the pollsters ask two or three questions trying to determine whether a respondent is affiliated with a party, then whether the respondent is a leaner. I'm guessing this poll picked up way too many "independents" who insist they have no leaning even though they're down-the-line litmus-test Fox/talk radio/teabag ideologues.
