Iowa Remains Unpredictable, With Polling Results Contradictory

The volatility of the Iowa caucus race can be seen in how little the polls agree with each other. While I noted yesterday that John Edwards had taken the lead in the Insider Advantage poll, today we have a new poll from the Washington Post-ABC News showing Edwards back in third place. As we saw in 2004, polls are poorly predictive of the outcome of the Iowa caucus and at this point it is possible that Obama, Clinton, or Edwards could wind up winning.

There are many problems in using polls to predict the Iowa caucus. There are many undecided voters, and even many of those who chose a candidate admit that they are likely to change their minds. Another factor is that it is difficult to predict who will turn out to participate in the Iowa caucus, which is far more difficult than just quickly showing up to vote. Edwards might have the edge here because more of his supporters have attended the caucus in the past, but other polls suggest a greater intensity and determination to attend among Obama’s supporters.

Until he fell behind in most polls it was assumed that Edwards would win in Iowa, and he very well still might do much better than is currently indicated in the polls. Edwards has many advantages having campaigned there in the previous election and having started so much earlier in this campaign cycle. He particularly benefits from having spent more time in rural areas as a candidate can pick up more delegates by winning more sparsely attended rural caucuses than even if his opponent wins by larger margins in the cities.

Being the second choice of many voters can move someone into the lead as supporters of candidates with the support of less than 15% attending must choose a different candidate. Some polls have showed that Edwards is the second choice of more supporters of the second tier candidates, but Obama has been eating into this lead. In 2004 Dennis Kucinich threw Edwards his support, helping him move into second, but Kucinich finally realized this year that Edwards is not worthy of such backing.

Independents tend to be more likely to support Obama, but they are less likely to vote in the caucus. This independent support might still help Obama as Iowa voters look towards the general election, especially as Edwards only does respectable in national polls because many voters in other states still see him as a moderate. People in Iowa, as well as The Des Moines Register which decided against repeating their endorsement, are more likely to realize that Edwards is quite different, and too divisive, this year.

Obama receives more support from students, but turn out by students is poor in the best of conditions. I wonder if it will be even worse this year due to the caucus falling while some students might not even be back after the holiday break. This might potentially have an impact on Ron Paul as well in the GOP race.

The 2004 caucus was notable for John Kerry moving from forth to first in the final days. While I would not totally exclude the possibility of a second tier candidate moving up, this looks far less likely this year. By early in 2004 there were stories about Kerry’s impressive ground game and some pundits were predicting he might sneak into second place. I had the opportunity to spend quite a bit of time talking with Teresa Heinz Kerry after a campaign stop in Michigan about a week before the Iowa caucus and she seemed quite optimistic about not only beating expectations but actually winning. I wonder if this was the expected optimism of a candidate’s spouse or if she had inside information which accurately predicted the outcome. Regardless, so far I don’t see anyone in the second tier showing signs of repeating Kerry’s surge, as much as I would like to see this happen.

With all the attention being placed on the Iowa caucus it is far from certain as to how much it will matter. If Obama or Clinton win big in Iowa they will be difficult to beat, but it is questionable if Edwards could use a win in Iowa as Kerry did to take the nomination. Kerry was already moving up in the New Hampshire polls even before the Iowa caucus, and Kerry’s views are far more in line with those of New Hampshire voters than Edwards’. As Edwards has devoted so much more time to Iowa anything less than a decisive victory would not be seen as very impressive. The Edwards campaign is also hampered by a poor organization beyond Iowa, in addition to a weak candidate, and even a bounce out of Iowa might not be enough to save his campaign.


  1. 1
    Ryan says:

    How do you think the media (and public opinion) would react to a Edwards victory in Iowa, if it happens? Meaning, which of these is the most likely: 1) Eh. It doesn’t mean much since he’s been there for three years and has all the advantages in Iowa that you mentioned. Anything less would have been devastating or 2) Whoa! Many of the polls had him in third place. What an upset! He will surely gain momentum with this stunning victory. (or something else entirely?) Just curious…

  2. 2
    Ron Chusid says:

    It is hard to say for sure as it depends partially on what expectations are set by the media before the vote and how much he wins by if it happens. Obviously a big win is more likely to give him a bounce than a narrow victory.

    Most likely a win by Edwards will help but not as much as it would help Obama or Clinton. While a win by Obama or Clinton would make them the overwhelming front runner (more so for Obama since Clinton’s campaign was based more on inevitability) a win by Edwards would make the difference between whether he is totally knocked out or if he becomes a contender in a to be decided race.

    New Hampshire will probably have a major impact if Edwards wins in Iowa. If Edwards can get a bounce out of Iowa and win in New Hampshire then he would be the probable winner. If someone else comes in respectably in Iowa and then wins in New Hampshire they might be able to use that momentum to win it all.

  3. 3
    Wayne says:

    Personally, my view on Polls is taken from the late Mike Royko (columnist for the Chicago Daily News, the Chicago Sun Times, and finally the Chicago Tribune), who advised people to lie to exit pollers, in order to make the media (of which Royko was an active participant) look bad if their early predictions were proven wrong.

  4. 4
    Ron Chusid says:


    I’ve seen speculation in the past that people lie more in exit polls as opposed to regular polls.

    My guess, which I have absolutely no evidence for, is that this might be true because those answering regular polls might want the poll to be accurate because either they want to influence others or they are interested in the results. On the other hand, once they voted they see the election as done and the poll serves no useful purpose to them.

Leave a comment