Saturday, September 18, 2004

The polls really are inaccurate 

I have had little patience for the constant discounting by lefty bloggers of polls showing Bush establishing a big lead over Kerry.

One reason for my tendency to write off the bloggers as self-deluded is that I have been hearing too much about how the samples of voters include more self-identified Republicans than Democrats. This argument is meaningless, because party-affiliation is not formally checked in these polls--people just say, "I'm a Republican" or "I'm a Democrat," and they say this because it reflects their decision to support a particular candidate. So a random sample that includes a majority of Rethugs just means that a lot of people at the moment self-identify as Repugs because they support Bush/Pilate, master of necessary blood-sacrifice.

The second reason for my impatience with the blogger poll criticism is that it seems fundamentally uninteresting. Over at Donkey Rising, Ruy Texeira and his followers analyze the polls in the most minute detail, always trying to show that the numbers actually indicate that things are going OK for Kerry. I don't discount Texeira's informative analyses, but the whole enterprise often strikes me as a way of denying the sickening horror of this campaign--the constant calls of Republicans to vote Republican as a way of honoring the dead, the implicit rebuke of anyone who would try to imagine a world with less death, the whole hermetic complex of 9/11-heroism sealed and preserved only by death and more death, the fact that an infantile fantasy image becomes even MORE impervious to analysis when the viewer knows with unconscious feelings of titillation and guilt that it covers a corpse...

And, most painful, the impotence of Kerry's replies. That's the kicker for me. I'd rather talk strategy than poll numbers, and Kerry's strategy has been suicidal. It's as if he's wanted to be the next victim of the RNC cult of ritualized human sacrifice. Sorry. I'm sick of observing the taboo (observed with an annoying combination of pathos and self-righteousness over at Kos and Atrios) against criticizing Kerry.

Now, he HAS been doing much better recently. But the fact that he has only NOW begun to mention, um, say, the issue of CREDIBILITY, is mind-boggling. This has always been known to be the linch-pin of the Democratic victory. I understand the impulse he indulged over the summer to let Bush self-destruct (Fallujah, Richard Clarke, etc.). But he STILL could have been sharpening the attacks for use when necessary, and getting the fucking inevitable and inescapable BUSH IS A LIAR message out there, if only through surrogates.

And the fact that he is only NOW talking about Iraq is equally mind-boggling. The fact that he could, on the stump, have come repeatedly to the Iraq question and said nothing more than, "I know what to do in Iraq; I defended this country as a young man and will defend it again"--the fact that he could have said this and NOTHING MORE is absolutely mind-boggling, indeed soul-destroying. Just thinking about it is killing me. I saw the man say this in person. It ruined about three weeks of my life. And this was essentially all he said at the DNC! The idea that the story of his Vietnam service could SUBSTITUTE for ADDRESSING THE IRAQ QUESTION IN ANY WAY--what kind of idea is that? It's demented. Vietnam-talk is fine, but not when you are cowering like a weakling behind it. THAT's why the Swift-Boat poison was so toxic--not because people really BELIEVED it, because it took Kerry even further away from the issue at hand--Iraq--and underscored his inability to assert himself regarding that issue.

Sorry. Just getting this off my chest. If I posted this on Atrios, I would be called a troll, or worse. Perhaps I am.

In any case, none of this is to the point. I started this post merely to note that in fact, while I dislike the "Kerry's actually doing great" theme, the polls ARE innacurate. The ones showing Bush taking a big lead are almost certainly wrong. They may even help Kerry by creating a feeling of unearned ease among the death-cultists.

Two things to note: it is very likely that in likely-voter surveys, Democratic turnout is being underestimated, as it was, grossly, in 2000. Gore was losing in the polls two days before the election. The unpredicted Democratic turnout should be larger this time. Second, and more importantly, today's Times/CBS poll showing Bush with a big lead really IS (despite what I said above) reflective of an over-sampling of Republicans. As Atrios points out, the kicker is the question about who the interviewee voted for in 2000. Among those polled by the Times, it was 36% Bush, 28% Gore--even though in the ACTUAL population, which is suppposedly being represented in the sample, Gore got more votes. That error alone accounts for the difference between Kerry and Bush in today's poll. And it is likely symptomatic of the inability of the Times (and Gallup) to find large numbers of Democrats who voted in 2000 and who will, in larger numbers, vote in 2004. These "marginal" Democrats (many minorities, many low-income) are the key, and they will be there, despite the inability of pollsters to acknowledge their existence.


This page is powered by Blogger. Isn't yours?