Today’s other Iowa poll

posted at 8:26 pm on December 13, 2011 by Karl

Commenters to my prior post about the PPP poll showing Ron Paul within a point of Newt Gingrich in Iowa pointed to the new Insider Advantage poll also released today:

Former House Speaker Newt Gingrich continues to hold a strong lead in Iowa, according to a new poll released Tuesday. Meanwhile, Texas Gov. Rick Perry has moved into third place ahead of former Massachusetts Gov. Mitt Romney.

Gingrich leads the field with 27 percent support while Texas Rep. Ron Paul comes in second with 17 percent (up from 13 percent in an Insider Advantage poll released in late November). Perry has 13 percent, up from 7 percent in the November poll, and Romney has 12 percent.

What to make of two polls concurrently showing Gingrich +10 and Gingrich +1? And Romney in third or Romney in fourth? Not as much as you might think. Both polls are of a little over 500 likely GOP caucus voters, with corresponding margins of error of 4% or more. Looking at differences between the two polls, Insider has Gingrich +5 from the PPP number, Romney -4, Paul -4 and Perry +4 — so the two polls are broadly consistent.

At the NYT, Nate Silver has posted his first polling-based projection for Iowa, including these polls — and has the order as Gingrich, Paul, Romney, Perry. FWIW, he finds Newt and Mitt losing momentum, while Paul has some, though not as much as his rivals have lost. Even so, unless paul overtakes Newt, it’s bad news for Mitt. Regarding that point, Silver notes that he stuck purely to polling; on Twitter, he noted that “If looking at subjective factors I might consider Paul almost a co-favorite.”

I don’t know if I buy Silver’s methodology just yet, but I am intrigued with this comment (at the NYT):

What’s a little different about these forecasts is that one of these factors — how recently the poll was conducted — really dominates everything else. We’ve analyzed literally thousands of primary and caucus polls dating to 1972, and what we’ve found is that you optimize forecast accuracy by being extremely aggressive about trying to identify the current trend. In the late stages of a primary or caucus race, a week is an eternity and even a couple of days can be meaningful.

Moreover, he has become convinced that momentum matters in primaries (but not in general election polling).

Blowback