The bounce

Put me down as being skeptical of that much-ballyhooed Newsweek poll that showed Sen. Barack Obama now enjoying a 15-point lead over Sen. John McCain. It's not so much that I think the result is impossible, it's that it's clearly due to their methodology for determining party ID, and that makes it harder to compare to other polls.

Obama's current lead also reflects the large party-identification advantage the Democrats now enjoy--55 percent of all voters call themselves Democrats or say they lean toward the party while just 36 percent call themselves Republicans or lean that way. Even as McCain seeks to gain voters by distancing himself from the unpopular Bush and emphasizing his maverick image, he is suffering from the GOP's poor reputation among many voters. Still, history provides hope for the GOP. [Pollster Larry] Hugick points out that in May 1988 when the primaries ended, Democrat Michael Dukakis enjoyed a 54 percent to 38 percent lead over George H.W. Bush. But Bush wound up winning handily. "Those results should give people pause," Hugick says, saying that a substantial number of voters, about 5 percent, have also moved into the undecided column. A significant improvement in the economy, or continued advances in Iraq--an issue McCain has identified with strongly as the senator who championed the "surge" first--could alter the Republican's fortunes.

Boy, how many times do you think that Dukakis comparison is going to get trotted out? Five thirty eight dispenses with that, and also explains why state level polling, for which we've also seen some wildly different results based on different models for determining party ID, suggests a more modest bounce for Obama. MyDD, on the other hand, is more bullish. I'd like to see some similar results before I feel too comfortable with it.

The main question I have in looking at this is do we have any partisan ID models besides Mike Baselice's that are being used in Texas polls? He says Republicans maintain an eight-point advantage statewide; I think that's too high. We had a couple of polls in May that suggested a much narrower gap, one each by Research 2000 and Rasmussen, then Rasmussen's subsequent poll swung sharply away from that, and that's about it. There's just not enough data to make a judgment, and unfortunately it looks like we're not going to get much more, at least not on a regular basis. Like it or not, it's what we've got to work with.

Posted by Charles Kuffner on June 23, 2008 to The making of the President