Daily Kos-SEIU polling banner

Public Policy Polling for Daily Kos & SEIU. 8/23-26. Registered voters. MoE ±3.1% (8/16-19 results):

Q: If the candidates for President this fall were Democrat Barack Obama and Republican Mitt Romney, who would you vote for?

Barack Obama: 50 (49)
Mitt Romney: 44 (45)
Undecided: 6 (7)

Back on July 22, President Obama hit his lowest mark in our head-to-head polling, tying Mitt Romney at 46 apiece. Since then, his standing has continually improved, going to +1, then +2, +2, +4, and now +6. Interestingly, PPP's zigging while most other pollsters are zagging: If you look at polling aggregates on sites like TPM and Pollster, you'll see the race generally tightening. So what gives?

You might look to the partisan composition of this survey, which is unusually blue: A very high 43 percent of voters describe themselves as Democrats, while 36 percent call themselves Republicans. That's actually a pretty high tally for the GOP, too, but this is a good example of how fluid party ID is: That late July poll which had the race tied also featured a seven-point edge for Democrats. It's also a good example of why you generally shouldn't put a whole lot of stock in part self-identification. But it's not irrelevant, either. (More on that in a moment.)

The more important issue, though, is that many firms have started switched from interviewing the broader universe of registered voters (RVs) to a narrower pool of so-called "likely" voters (LVs). Every pollster uses a different method of sussing out LVs, but generally speaking, LV polls will tend to tilt somewhat redder than RV polls, more often than not—it's simply a fact of life that Dem-leaning voters, for a variety of reasons, are typically less likely to go to the polls than Republicans.

So what I think we have here is, as I've alluded, a sample that's more left-leaning than usual—22 percent identify as "liberal" and 37 percent as "conservative," which is actually very liberal (typically the gap is 20 points—as it was in that July poll—or more). I'd therefore guess that Obama "really" has a lead that's more like two to three points, rather than six. And if this six-point edge is in fact just a blip, that would also tend to explain our outlier-ish trend. But after Labor Day, we expect to switch over to a likely voter model ourselves, so stay tuned to see what, if any, changes that brings to our horserace polling.

P.S. As always, you can find our complete poll archives on our Weekly Trends page.

Your Email has been sent.