« Rudy Didn't Fail: Guiliani Speech Sways More Voters than McCain or Arnold |
Main
|
Gallup: 7 Point Lead »
September 06, 2004
That Newsweek Poll: Too Biased in Favor of Kerry?
I have trouble believing this, but DJ Drummond at PoliPundit makes an interesting case.
As CrushKerry noted, the Newsweek poll sample had about 37% Republicans and 31% Democrats. (Most people believe actual party identification is something like 31-33% for both.) So, CrushKerry argued (and I agreed) that poll probably overstated Bush's true lead, at least by 20-25%. (And that doesn't even consider other problems the poll might have, like just hitting an unrepresentative sample.)
Well, looking deeper at the raw numbers, DJ finds, first, that independents broke this time for Bush 45-40, a big reversal from Kerry's previous lead with this group, and second, that the poll actually found even more people self-identifying as Republican, but the pollsters gave heavier weight to Democratic respondents in order to correct for a too-high sampling of Republicans.
Without that weighting, Bush's lead was even bigger. Huge, in fact. The technical mathematical term for it is "freakin' ginormous."
The dramatic shift in independents' sentiments alone should be plenty worrisome for Kerry.
I don't know what to make of that second point-- I don't think we can make much of it, since this country is obviously not 42% Republican, except to say that Newsweek did at least some counterweighting to reduce the Republican advantage in sampling.
Now, Fred Barnes has of course predicted eleven of the last zero poltical realignments in favor of Republicans. But if Time magazine similarly found they were getting a lot more self-identifying Republicans on the phone than they expected, Barnes might be tempted to write his twelfth column predicting that "the country is realigning to become plurality-Republican," and who knows? This time, he might even be right.
Susan Lucci eventually got her Emmy, after all.
(Take that, those who said I was unduly pessimistic last time out. I balance my undue pessimism with wild-eyed lunatic optimism.)
Update: I should have finished reading the analysis.
Actually, it's not as bad as all that for the Sinking Senator.
It may be worse.
But... Rasmussen shows just a single point lead. A lead of 1.2%. One point freakin' two.