VITAL
FOREIGN PRESS
GOOD BLOGS
REFERENCE
« Previous Post | Main | Next Post »
Is everything the survey-takers know wrong this time?

It did not seem unusual when the post-convention polls yielded only a tiny bounce for Kerry--if you watched the Democrats in Boston, you knew why. But there are other clues that the art of the opinion survey may have fallen on hard times this year in view of the furious and contradictory public energies conjured by the Bush White House.
Zogby, the outfit that has performed best in the past few election cycles, had Kerry up on Bush 50-43 last week. The old lion Gallup, contrarily, gave W a 50-47 lead. A 10-point difference between major polling houses at this point is remarkable in itself. That's not the only thing. Even in disagreeing so widely from one to the other, few polls have budged much since the early gains of Kerry's post-Iowa surge were lost in the spring.
The media playbook answer is that voters are exceptionally polarized this time, and their minds have been made up from the start. No doubt this is true to a greater extent than usual, but the pollsters themselves are on more uncertain ground as well. Election surveys are based on statistical composites of the likely voter, and I don't think anyone has a clue what "the likely voter" looks like this time. Everyone who pays any attention at all to electoral politics is inflamed, or claims to be--perhaps most consequentially the young, who do not usually vote in great numbers but may this time.
I can only attest to what's going on in my circle of acquaintances. None of the Democrats I know plan to vote for Bush. A number of the Republicans I know plan to vote for Kerry. The polls say both parties have a defection rate around 5 percent. I think they're wrong. There will be more Republican turncoats and no-shows than the opinion-gatherers are catching. I suspect they are also missing a lot of traditional nonvoters who will show up on November 2 to register their distaste for Bush. How great is the margin of error involved? My hunch is that they're underestimating the Anybody-But-Bush groundswell by 5 points or more.
On the other hand, Kerry is such a poor vessel--for anti-Bush sentiment, or anything else--that there always remains a chance he will actually drive large numbers of those ABB voters back into despair and abstention by November.
Posted by Steve Perry at August 19, 2004 10:07 AM
« 10 Out of 10 Terrorists Want--Bush | Main | The Great American Lockdown »