Cuz We Said So

Now, mind you, we've had more than a month where every poll under the sun showed the president between four and twelve points ahead. Two weeks ago, Michael Moore was telling everyone that the polls were all wrong because they were all over the map. ("Like diarrhea", he added, charmingly.) Three polls taken immediately after the first debate show no change to the race. Neither yesterday's nor today's Rassmussen poll shows a change. But here comes one poll from Newsweek, and:

Andrew Sullivan: Kerry pulls even
Josh Marhsall: Kerry pulls ahead

(Amusingly, pro-Bush pundits have been saying for a week that no matter how the debate came out, Kerry supporters and the mainstream press would trumpet the start of a Kerry comeback--which has happened. Of course, if you're pro-Kerry, you would probably assert that this was simply a technique to innoculate Bush in case the polls turned against him. Isn't game theory fun?)

Look, I'm not saying that my fellow Bush supporters haven't been doing much the same thing, confidently asserting that Kerry is toast for this or that reason. I recognize the desire on both sides to spin, to shore up the team. But the variance we're seeing in results from poll to poll is too wide for me to trust any of them as accurate. That's why I've had little to say about the polls since Bush appeared to take the lead following the convention. Consider, for example, two likely voter polls taken over similar time frames: the CNN/USAT/Gallup poll taken from September 24 to 26 had Bush at 52% and Kerry at 44%. The IBD/TIPP poll taken from September 22 to 27 had Kerry at 46% and Bush at 45%. That's a 9 point variance between two polls taken at roughly the same time. Consider the margins of error, and the actual difference could be wider still.

Kerry could indeed have closed the gap or pulled ahead in the polls--but one poll doesn't tell you that, especially when contradicted by others. Let's at least wait and see whether a trend is indicated before we start to wallow in our own credulity, shall we?

UPDATE: Joe Katzman at Winds of Change.NET has more on the Newsweek poll, complete with several good links to information on polling in general. I thought this point from Power Line was interesting: see how many of the people you see reporting on the poll tell you this part:

Newsweek's most recent poll included 345 Republicans, 364 Democrats and 278 independents. This compares to Newsweek's published data for their most recent prior poll, which showed President Bush with a comfortable lead: 391 Republicans, 300 Democrats and 270 independents. Yes, if you drop 46 Republicans and add 64 Democrats, you will get considerably better results for the Democratic nominee.