Unlikely Voter

I mentioned recently that broader polling pools favor Democrats, so when a big new poll of adults comes out from Gallup that shows Barack Obama to be in trouble, I take notice.

Earlier this week we caught an Iowa poll showing Rick Perry as a new leader in that state’s Republican Presidential race. Yesterday Public Policy Polling came out with a new Iowa poll as well.

Judge for yourself, but I’d say the broad strokes of the We Ask America poll are confirmed, at least when it comes to the big three candidates, though maybe not with Sarah Palin.

This week Gallup polled four Republican candidates against Barack Obama. For the three leading Republicans the results are typical, and do more to show the difference between polls of adults and polls of registered voters (Gallup polled both).

But oddly enough, Ron Paul was different.

By request, we have a somewhat unusual poll: We Ask America asked Iowa Republicans which candidate they don’t want as the Republican Presidential nominee in 2012.

As is usual this time of year, the poll is skewed by the inclusion of non-candidates. This time, overwhelmingly so.

By request I’m taking a look at a new poll of California Republicans by Probolsky Research. It shares problems I’m seeing in many early Presidential polls, but I am surprised at one finding that may be bad news for Mitt Romney.

Some polls at least deserve some serious scrutiny before they get dismissed as flawed. Others are just so laughable on their faces that I hesitate even to put them in the same posting category as serious works. One of these laughers is This Texas poll from the apparently new firm Azimuth Research Group.

According to Azimuth, Ron Paul runs better for President than Rick Perry statewide among Texas Republicans.

So Fox News put out a new poll of the race for the Republican Presidential nomination. It’s a typical poll in many ways, but Fox’s bit of analysis got me to thinking: Polls like this favor frontrunners and likely skew the race.

By request, I took a look at this poll by PPP for Daily Kos and SEIU. Markos Moulitsas himself is hyping the poll as showing an enthusiasm gap, which of course was one big indicator of the electoral wipeout we saw in 2010.

I think that he’s right, to a degree. However I read the figures as having two conclusions: First, the TEA party effect is still there, and Republicans are slightly more engaged than Democrats at this early point in the cycle. Second, the Union activism of this year is not having the same engagement effect with Democrats, that the TEA party, the ARRA, and the PPACA had with Republicans.

We all knew it was coming, and finally it’s here. The Swingometers have been updated. The House Swingometer includes the 2010 results as a new baseline, and the Electoral College Swingometer includes the 2010 Census-based reapportionment.

Some researchers led by Noah Smith at Carnegie Mellon tried an experiment: Could they predict the results of traditional polling, which has as its core feature a genuine random sample of people, with careful monitoring of Twitter?