The Des Moines Register released their final pre caucus poll last night with some interesting results.
As the pundits have noted, this poll is a MAJOR outlier as far as recent polling trends. Most of the polls this week show a close three way tie with Clinton maintaining her support, Edwards surging and Obama slowly but steadily fading.
In fact, two polls released today show:
Clinton leading with 33%, followed by Obama at 31% and Edwards at 22%.
Clinton at 30% Obama at 26%, and John Edwards at 25%.
It’s interesting to note that Clinton wins among democrats. Looking closer at the DMR poll a lot of odd methodology is used. Most noticeably, the DMR poll is projecting a turnout between 180,000 and 200,000. By comparison, the turnout in 2004 was 124,000. This would be an astronomical turnout increase in a voting event which historically depresses turnout. Though given the increasing attention to this election I wouldn’t be suprised if turnout increased. But a 75,000 increase? Seems unlikely to me. But like I said, I’d give them leeway on turnout except for their projected breakdown of voter demographics.
2008 DMR poll
But in 2004 the actual demographic breakdown was
That would be a 360% increase in independent voters resulting in them comprising nearly half of the caucus goers. Thats crazy absurd. It definitely leads a vast chasm of doubt in the supposedly gold standard of Iowa caucus polls. I also think there’s something to be said about non democrats having such a large say in the selection of the democratic nominee. I’m normally for open primaries but considering that Iowa is a caucus and not a primary, and that its the first state to vote, which gives its SO MUCH value, this seems rather shady in my eyes if Obama does win. Particularly since he’s losing among democrats.