Anti-Climacus

LINK: I'd point out that, in addition to all the caveats Mystery Pollster makes about this most recent Newsweek poll (so breathlessly sent to me by Our Girl in Washington in a moment of schadenfreude), I'd add the following: 49-46 doesn't actually tell you very much about where things are going, because there's still some softness in the numbers. Given GOP gains in GOTV over the last two election cycles, Kerry probably has to be in the 51-52 range just prior to the election to have a reasonable chance of pulling it out. Couple of other points:

*Exactly how much higher can Kerry go? I've not looked at the post-convention numbers for Kerry, but he seems to have rarely, if ever, topped 50% in the head-to-heads (I could well be wrong about this). But if you believe, quite plausibly, that the general orientation of the voting public is around 50-50, you should expect the numbers to be converging to 50-50 as the election draws closer. Without a look at the methodology, I'd suspect what happened is that Kerry-voters who are loosely affiliated probably came back into the fold for the sample taken by Newsweek, and the numbers they get reflect that.

*This still doesn't say much of anything about Bush's numbers. I assume he's going to underperform relative to his actual support, because some of his loose affiliators have been temporarily scared off. It's also not entirely clear to me that we're seeing an actual reflection of the support of his base; you could probably make a decent argument that it doesn't really matter at all what he did in the last or any other debate, because his strategy does not seem to be based around them at all.