There is one Iowa poll apparently held in the highest esteem by professional pollsters and that is the Des Moines Register poll conducted by Ann Selzer. Well, Selzer is out with the final DMR numbers and they look like this:
Candidate ALL DEM IND Obama 32 27 39 Clinton 25 33 15 Edwards 24 25 24
According to the survey, which was in the field Thursday through Sunday and included 800 likely Democratic caucus goers for a margin of error of plus or minus 3.5 percentage points, 6 percent of respondents were undecided or uncommitted. No other candidate other than the three listed above received more than 6 percent support, suggesting that the second choices of supporters of these candidates could play a significant role Thursday night.
As you can see above, Obama's lead is entirely predicated on a likely voter model that predicts heavy independent turnout in the Democratic caucuses. Specifically, the poll predicts 40 percent of Democratic caucus-goers to be independent, and another 5 percent to be Republican. So while Clinton leads slightly among just Democratic voters, Obama's lead comes entirely from independent voters, who back him by a wide margin. So if indies turn out in a few nights, it looks like it's going to be a good night for the Obama camp. However, if indies are fickle and stay at home Thursday night, it might be a relatively long night in Chicago.
Update [2007-12-31 22:54:37 by Jonathan Singer]: Just to add...
An analysis of likely caucusgoers' second choices showed that the results would change little if the votes for the lower-rated candidates were redistributed among the front-runners.
Anyway, much more data here. Hope everyone has a happy and safe new years!
Update [2008-1-1 0:11:23 by Jerome Armstrong]: Regarding it's turnout projection, I'll just add a comment on this from the poll:
Ann Selzer may well have polled an outlier this year and turn into a toad, or Obama might prove her to be a poll god by turning Iowa's traditional caucus on its head by ramping up Independent & Republican participation in the Democratic event to an unprecedented record level-- we'll know in 72 hours.
Also, the Edwards camp has a memorandum out on the poll. I wrote the above prior to getting the note, but it seems that they see the same thing in their second point. I'll put the Edwards memo in the extended entry:
In response to questions, here are some observations about the Register poll released this evening.
Is the poll accurate? There are good reasons to think it is NOT.
The poll was conducted during the holiday season AND over the weekend. (One column says interviewing ended on Sunday while the article says interviewing was conducted Thursday, Friday, and Saturday.) Numerous professional studies indicate that either of these would make it more difficult to obtain a representative sample. The combination makes the problem of obtaining a valid sample geometrically worse.
The poll is at odds with history.
The poll says 60% of Democratic caucus participants will be first timers who have never before participated in a caucus. This matters in the horserace because the poll finds that about three-quarters of Obama's supporters are first-timers. The poll is at odds with known tenets of partisan caucus participation. The poll says 45% of those at Democratic caucuses will be Independents or Republicans. This matters because the entirety of Obama's "lead" is due to his advantage among non-Democrats.
The poll is at odds with other polls.
Several other polls show a closer race with other candidates leading. Others seem suspicious of the horserace numbers. David Yepsen's column highlights the fluidity of the results rather than the horserace.
What does the poll say? The race is close.
The race is yet to be decided. 34% say they could change their minds, and 6% do not express a preference. This means 40% have not made a final decision. As Yepsen points out,
21% of those in the 2004 entrance poll said they had made their decisions in the last 3 days -- something this poll cannot capture.
The trend is moving toward Edwards. The nightly results show Obama's support flat over the last two nights, Clinton's declining each night, and Edwards' support increasing each night.
|
|
|
Permalink :: 170 Comments :: Post a Comment
|