Bart, if the polls don't mean anything, how did they get it right in '04?

As to the Republicans, I am admittedly less informed--but attributing Huckabee to the neocons is absurd. He's a foreign policy neophyte (no-phyte?) and almost exclusively a phenomenon of the evangelicals.

Jeff, I'd pick apart your power rankings, except that they're exactly the order I filled in on my survey earlier today ;)

That said, Bart has good points, and these are things I wonder about too. I think that both Paul and Kucinich have huge grassroots support, and it will be interesting to see how successful media and establishment marginalization can be in the new tech world. Web 2.0, talk radio, etc. have come a long way since 2004. So it will be very interesting to see how both of them fare in the most deliberative system used in the primaries. Especially Paul.

Oh yeah, and I know this is going to be unpopular, but I want Paul to win the nomination. People seem to forget that a presidential election, in addition to choosing a new president, is a rare opportunity to have a large national discussion about where we want our country to go. Having anyone but Paul nominated on the Republican side will result in a general election that focuses on "whether or not" we want a continued occupation of Iraq -- a question the general public has pretty much gotten past.

With Paul as the nominee, we will have the chance to discuss HOW is the best way to get out if Iraq, and also domestic issues etc. We need and deserve the opportunity for such a discussion.

Would he be tough to beat? Hell yes. But anyone in the Democratic field is capable of it. They would just have to put their best foot forward to do so; there would be no cruising to victory while all the corporate lobbyists get a head start on steering their attention.

Oh, almost forgot:

GO HAWKS!!

Well, maybe next year.

From the Obama camp not long ago

I wanted to be the first to share the news, whether this email finds you tonight or you're reading this in the morning. A few minutes ago, the Des Moines Register released its final poll of Iowa Democrats ahead of Thursday's caucuses. The headline: 'Obama widens lead over Clinton.' Here are the numbers: Barack Obama 32% Hillary Clinton 25% John Edwards 24% The details get even better. Here are a couple of quotes from the accompanying article in tomorrow morning's paper: "The findings mark the largest lead of any of the Democratic candidates in the Register's poll all year, underscoring what has been a hard-fought battle among the three well-organized Iowa frontrunners." "Obama's rise is the result in part of a dramatic influx of first-time caucusgoers, including a sizable bloc of political independents. Both groups prefer the Illinois senator...."

Obama wins by 4.2% over Clinton. Thats my prediction. Like you said only the Register got it right in 04 and I'm banking on them now.

Matthew, from the Des Moines Register poll done 12/27-30 with 800 "likely caucus goers":

Democratic presidential candidate Barack Obama has widened his lead in Iowa over Hillary Clinton and John Edwards heading into Thursday's nominating caucuses [...] Obama's rise is the result in part of a dramatic influx of first-time caucusgoers, including a sizable bloc of political independents. Both groups prefer the Illinois senator in what has been a very competitive campaign. Obama was the choice of 32 percent of likely Democratic caucusgoers, up from 28 percent in the Register's last poll in late November, while Clinton, a New York senator, held steady at 25 percent and Edwards, a former North Carolina senator, was virtually unchanged at 24 percent.

Two things worth mentioning: I have been an ardent Obama fan since the Spring and my predictions have historically been terrible. More: the DM Register poll is widely admired as the most accurate, based mainly on their ability to identify who will actually turn up at the caucuses. And yet more: it's a little unclear which way causality runs here: in '04, when the Register got the order right, was that because they had read the voters right or because voters read the poll? If the latter, this will be a boost to Obama even if the poll was wrong.

A couple of thoughts:

1. Voters' second choices -- while obviously important -- are probably being overestimated a little. If you're an Edwards voter, but your second choice is Obama, or if you're an Obama voter who also likes Edwards, it won't really matter much because the viability threshold won't kick in (depending on precinct). Seems like there's some double counting happening.

2. I think the phenomenon of Obama underpolling because of cell phone use is real, assuming these folks show up. Also, polling in general seems very difficult right now, as pollsters and campaigns bombard every household in Iowa. If you're not picking up the phone, and plenty aren't, you're not being registered.

3. Edwards looked great on C-Span yesterday: very sharp, very crisp message. However, I think part of media's reporting on his surge is a result of collective guilt for ignoring him for months. I'm a strong Obama supporter, but I've long thought media were premature in essentially calling this as a two person race; I wonder how much of the reporting now is overcompensating for past unfair coverage.

4. The Clinton campaign refused to answer a question from a nine year-old reporter from the Scholastic News yesterday. I understand not doing press avails in these closing days, but I'd like to think there's some karmic principle at play here that will affect her chances Thursday. I'm not holding my breath though.

I can come up with very good reasons why any of the three top candidates should win Thursday night. We'll find out this week. I'm also intrigued by the New York Times piece speculating about the diminished importance of an inconclusive Iowa outcome. Expect to see that article widely circulated by the second and third place candidates -- no matter who they are -- immediately after the vote.

And Jeff, no one expected a sober column from you on New Year's Eve! Happy New Year's all.

Jeff, am I right in thinking that the time axis on your combined graph (thanks for the graphics work!) is since about September?

It looks like the most recent changes in these graphs are considerably "undecideds" deciding. Edwards has much recovered from an earlier decline, in which Obama may have been taking support from him, comparing the trend patterns. Yet Clinton is also up, as are Richardson and Biden on a chart on the polls.com site, and Obama is only slightly down.

We should keep in mind that the colored line endpoints are very close for all three. Without the scale this graphic creates an illusory perception of Obama being ahead in this comparison (i.e. I'm not talking about the single poll you & bdunn discuss). He's actually on a taller scale if I'm not mistaken, because he has been more volatile, almost wholly upwards. Even if the scales are the same & the effect is purely optical due to steepness of his climb, the best way to read these is from right to left.

There is a lot of buzz among Edwards' supporters too, it is hardly unique to Obama (probably everyone is pretty jazzed at this point). In the pro-Edwards discussions there is a lot of rhetoric about him as "unity" candidate, which at times however seems to mean either "likely to benefit dispropotionately in the second round" and perhaps less often as "point of unity for those with anybody-but-Clinton attitudes."

At the polls.com site Richardson is shown with 5-7% and Biden with 4-5%. In the case of Richardson, if his support has geographical concentrations, these numbers at least raise the possibility for me that he might get past cutoff thresholds in a number of places, which would mean fewer redistributed votes in the second round. That would be bad for Edwards if Edwards' supporters are right in thinking he tends to gain relatively in the second round.

Media reports make it seem that 15% is the most common cutoff for getting into the second round, so the possibility I'm raising about Richardson would be stronger if his statewide average were 10% or above.

But I don't really have a sense of how much geographical variation there is, or whether there may be a lower threshold in some bigger (urban?) precincts. I think there must be one or the other or both, because otherwise results in previous years, in which trailing candidates nonetheless get some % of actual votes cast & are under 15% don't make sense.

Perusal of the link provided by Matthew Sutton shows it to be a strongly pro-Obama blog. It gives some information about how Obama supporters are thinking & talking, though I can't know how typical it is. But if there's "horserace" info of a wider sort, it's buried a bit & I didn't find it on a quick look. This could be a late-in-the-campaign artifact.

Also, I'd like to point out, for fans of the "Kari Chisholm as duplicitous puppeteer mastermind" theory of BlueOregon, that Kari, an Edwards guy, clearly is not shaping this thread. Doubtless he's making the right choice about being involved in his son's movement from newborn to infant. (Off topic -- I'm not sure I've ever seen anything as amazing to me as the changes in those very early days, so basic yet so huge -- completely changed my perspective on the orders of magnitude in human complexity.)

The other two pro-Obama editors have shaped the thread, Jeff as author and Charlie as commenter.

And they have done so in an honest and forthright manner. They don't hid their preferences but consider information in a generally fair-minded light, exactly what one wants for interesting discussion informed by good will & critical thought. Thanks for setting a good tone, Jeff.

Charlie's telephone argument is interesting, but I am wondering if Obama is the only candidate implicated in its consequences. Are there reasons to think Obama supporters more likely to have cell phones or exclusively cell phones than supporters of other candidates?

Also, is there any reason to think that the DM Register poll has a superior technique for getting around that problem? Or for getting a better response rate on their phone calls, if as Charlie points out people are tired of being polled & not answering phones?

Charlie's warning to us about problems with late polls & polls & new technologies generally exemplify his fair-mindedness in my view.

FWIW the author of the blog Matthew cites, in a pro-Obama post which nonetheless stipulates that everything is highly fluid & up in the air, says that the DM Register pollster is the best in Iowa, in his opinion -- despite my earlier comment I don't get the sense that this is a one-off opinion voiced because this particular poll is favorable to Obama.

Chris, I think you're right about the segment shown--Pollster doesn't break it down by month, but eyeballing it, September seems about right.

Matthew, your testimony and that of others will make it easy for me to be an enthusiastic Obama backer if that's how things play out. The Kucinich semi-endorsement is most interesting.

On the other side, though, the Pollster.com site Jeff provided has links to a post from Taegan Goddard's Political Wire showing more than half a dozen polls with a lot of statistical dead heats, at least first round, and some showing each of the top three leading. They say that the widely disparate results reflect different polling assumptions about who are likely caucusers. So Charlie wins again: "It's all about turnout"™

The one I've seen Edwards supporters touting in some other places, our equivalent of the DM Register poll I suppose, is from Insider Advantage/Majority Opinion Research. Conducted Dec 28-29 (i.e. after the DM Register poll) and published Dec 31, it shows this first round spread: Clinton has 30%, Edwards 29%, Obama 22%, with 14% committed to other candidates and 5% undecided.

More significantly, though, they also look at second-round voting in a way restricted to the second preferences of those other than the "big three" -- addressing a point Charlie makes above. Their final conclusion:

Using the reallocation methodology InsiderAdvantage used in 2004 – which correctly indicated a fairly comfortable win for John Kerry – our new poll reveals that, if the caucuses were held today, the reallocated final outcome would be:

Edwards: 41% Clinton: 34% Obama: 25%

Now I know nothing about this outfit, their track record or neutrality, beyond the claim about Kerry above. But McClatchey gets somewhat similar though less extreme "second round" results: Edwards 36%, Obama 26%, Clinton 26%.

Apropos of widely varying turnout assumptions, Clinton's campaign in questioning the Des Moines Register poll thus (per Boston Globe):

The Clinton camp argues that the Register's model of who will turn out for the caucuses is different than in 2004 by including far more independents. If the same model was used in 2004, Clinton would be in the lead. Other recent polls show a virtual dead heat among the three Democrats. "The Des Moines Register poll adopts an unprecedented new turnout model for the caucuses, and its new poll is out of sync with the other polling done in the race," Mark Penn, Clinton's chief strategist, said in a memo posted on a campaign website late Monday.

Which might be right, or might be whistling past the graveyard of Obama's unique appeal (see Matthew's comment above).

Much depends I suppose on whether the pollster has evidence beyond the anecdotal suggesting increased historically unusual participation, and has correctly interpreted it, in making the underlying sampling choice.

Yep, we'll find out. My impression is that the different polls do include Is & Rs planning to caucus D to some degree, but in smaller numbers or calculated with different weighting.

The Clinton people are focusing on the DMR wider inclusion as a bad thing because misleading. Presumably the DMR chose what it did because its pollster thinks more such participation is likely this year. Apparently other pollsters disagree. As you say, we'll see.

Actually Barack Obama is such a good speaker that I'd be surprised if some people didn't just go to hear him. We've been starved for decent oratory for a while. If he were primary stumping here I'd go & take my kid even if I wasn't planning a primary vote for him.

Obama has a number of good qualities & if he wins the nomination I'll be able to campaign for him positively & enthusiastically; if it's Clinton I'm going to be campaigning more against whomever the Rs put up. But Edwards speaks more directly for me in a number of ways, that's all.

Bart Springfield,

I'm with LT on this one. Huckabee is a total heretic on the Grover Norquist no taxes throttle gummint scale.

He is the worst nightmare of the crowd that David Brock described in Blinded by the Right even though he got himself in trouble early on as governor of Arkansas doing their bidding in the Unending War on the Clintons.

That said, I only want to see him do well in the primaries to expose the hypocricy of his enemies inside his own party. His own tax policy is pretty much insane.

SO far it's been great fun watching them try to spin their inherent greed against Biblical New Testament princples.