Iowa Caucuses

We are just shy of three weeks out on the Iowa Caucuses.

Reports have Barak Obama pulling out front of Hillary Clinton, but the polls are not unanimous on that judgement. Real Clear Politics (RCP) publishes the latest polls. In all RCP publishes four polls and Barak Obama leads in two of those, but RCP combines the polls and dubs Barak Obama as the Democratic leader in Iowa.

Running the numbers shows the polls have a number of undecideds. The undecideds run from 5% to 13% with RCP's numbers yielding 9.6%. What I am trying to ferret out is the number of undecideds now compared to those in 2004.

Two weeks ago, polling had Edwards far behind then-front-runner Dean as well as Kerry and Gephardt. In pre-caucus polls from the campaign's final days, Kerry, Edwards, Dean and Gephardt were running so close, that the candidates were statistically tied. It was the closest Iowa contest since 1988.

Source: Fox News Online – Kerry Wins Iowa, Edwards Takes Second We are one week away from the same point as the excerpt and had Howard Dean out in front, I wonder what the number of undecideds were back then? My recollection is the number of undecideds were quite significant then, much more undecideds than now.

Iowa back in 2004 was a surprise. The MSM kept building up Howard Dean and the polls gave some support to the idea. However, in the end it was the person who had the aura of inevitability. The polls we see have enough margin of error and undecideds to swing Iowa to Edwards, Clinton, or Obama. I do not think this races is decided yet.

On the Republican side the undecideds (for the purpose of this blog I call the difference between the sum of the reported percentages & 100 as undecided) are greater in number, but unanimous for Mike Huckabee. In all cases but one the number of undecideds outnumber the difference between Huckabee and number two Mitt Romney.

Now, one last item I wonder about. From what populations do the polls RCP cite sample? The general Iowan or from the partisans who will decide the contests? Participation in the Iowa caucuses are limited to registered voters of the respective parties. I don't know, does the Iowa voter have to declare which side they belong to when they vote? Is it reasonable to generalize the Iowa population as tied to one party or the other? Even among those registered is it reasonable to expect the average Iowan party registered voter to show up at their caucus?

Obama and Huckabee lead the polls now, but polls do not count.