« Academics Weigh in on Brooks Article | Main | More On National Security Options For Democrats »


What's Real in the Latest IA/NH Poll?

So, faithful readers, if you did your homework about the extraordinary perils of polling Iowa, you're ready to digest today's big public opinion news, a new poll of Iowa and New Hampshire by CBS and the New York Times.

There's another bit of threshold info necessary here: this is the first CBS/NYT poll of these two states, so given the methodological variations between polling outfits, it's not very useful in terms of establishing trend lines with any precision. Thus, the finding that the Big Three Dems are essentially tied in Iowa, while some polls have shown a big HRC lead, or a relatively poor third-place Edwards position, isn't particularly illuminating, other than by reinforcing what pretty much everyone actually believes. Similarly, the poll shows John Edwards at 9% in NH, while other polls have shown him in the low-to-mid teens. This won't much matter until we see another poll from the same source.

But even in a "first poll" like this, some big trends may have real value. They are all basically on the GOP side. Mike Huckabee's 21% standing in Iowa represents a large enough shift to reflect something happening on the ground, particularly since the poll's internals show that his support levels are "harder" than those of front-runner Mitt Romney. And Fred Thompson's dismal sixth-place position in NH, well below Ron Paul, along with his single-digit showing in Iowa, isn't a good sign for the Big Fred Machine.

Poll internals from IA on the Democratic side are pretty interesting, since many published polls don't go into the deeper dynamics of candidate preference that particularly matter in that state's Caucuses. For example, the poll tests secondary preferences among voters supporting the lower-tier candidates (who presumably will struggle to meet the 15% threshold for winning delegates in any particular precinct), and both Edwards (30%) and Obama (27%) are winning more than double the percentage of such voters as compared to HRC (14%). Another set of interesting internals have to do with Caucus composition. Obama and Edwards are basically mirror images, with Obama's strength among the categories of voters (e.g., younger voters and self-identified independents) likely to matter the most if overall turnout is high, and Edwards doing especially well (though not much better than HRC) with the likeliest participants in a lower-turnout scenario: seniors, self-identified Democrats, and prior Caucus-attenders.

None of these IA Dem findings directly challenge the CW, though HRC's apparent second-choice weakness could pose a challenge to her Caucus-night tactics, where candidate-driven or spontaneous alliances with "nonviable" candidates are always a big deal in a large field. But such speculation depends on the accuracy of the poll in the first place, and to come full circle, we won't really know whether this or other IA polls are on mark until the results are in.