This analysis overstates the role of Iowa on later states and on candidates' likelihood of withdrawing from the race early.  I think the 3 leading candidates have too much money and organization to be "done" after one state, unless they come in fourth place, which is extremely unlikely.

Noam:

Good points, definitely.

Edwards cannot survive without a huge shot of momentum. Without that, he's done.

Stgla:

I dunno. This is the way this race is playing out. The 3 candidates are all retail campaigning in this states. If someone is crushed, then I don't think they deserve to continue...

Noam,

I definitely agree with you on Obama having more paths; for all that's happened over the past year, the race still seems to be between Hillary and whomever gets to be the anti-Hillary. If Edwards finishes third or a distant second I imagine a lot of his NH and SC supporters will break for Obama simply because he would seem to be the stronger liberal option to win the nomination.

What I'm wondering is if a Clinton win of three points means the race is over. I mean, how many votes difference are you looking at? From what I can gather from Wikipedia, that would be about 3,000 votes if the same number of people show as in 2004. (I'm not going to get out the data modeling software to try to account for the ridiculous weighting rules. BS'ing a number will need to do at this hour.) It seems like Obama could spin that as pretty damned good considering Hillary's institutional (Vilsack, her husband, the star f'ers at the DMR, etc.) and financial advantage. I mean, she was the prohibitive front-runner just a couple of months ago, right? I think a strong underdog plea like that could significantly reduce any great slippage in NH, and he might gain some of the anyone but Hillary crowd who were supporting Edwards.

An unrelated question: is there any recent, presumably solid data on the second preferences for non-big three supporters?  If the race is really this tight I imagine those votes might make a huge difference, and it's a wonderfully perverse pleasure to think of the (admittedly so remote as disbelief needs to be suspended) possibility of Kucinich being a king-maker. If Hillary wins by a tenth of a percentage point and Bill adds a research center for paranormal phenomena to his presidential library we'll know what deal went down.

I agree with stgla that people are overestimating the effect of the Iowa results on the New Hampshire results.  People are overgeneralizing from the 2004 experience, which may just stand for the more limited proposition that, if Iowa suddenly elevates a second-tier candidate, New Hampshire voters will give that candidate a second look.  But if Iowa splits between the three top-tier candidates, it's not clear that New Hampshire voters will have to rethink anything -- they've already been thinking a lot about those three.

The main prize in Iowa is the title of "The One Viable Alternative to Clinton."  And you're right that Obama has the advantage there -- Edwards needs a clear win to take that title, given his fundraising disadvantage, and his currently low numbers in New Hampshire.  But a close finish between Obama and Edwards could hurt Obama, too, by delaying the moment when he can credibly claim that title.

I think that you are over-emphasizing the importance of the early states, and forgetting that that is just the beginning and NOT the end! Remember that only a few delegates would be awarded until super Tuesday, which I predict would be a Hillary slaughter house, regardless of what happens in the early states (unless Obama sweeps!). Poll after poll shows Hillary to enjoy a huge advantage among the rank and file Democrats so that outside of the early states (where everybody - Republican and Democratic candidates - has been taking aim at her), I am not sure how large the "anti-Clinton" vote would be. The apt comparison would be Hart v. Mondale in 1984. Hart came on strongly in the early states, but Mondale's advantage was just too much elsewhere for Hart to hang on. Clinton would carry nearly every large delegate-rich state such as NY, NJ, CA, and TX. She has now already secured about 60% of the "super delegates." Also, I am not so sure that Obama would do that well in Dixie (if he does, it would be a measure of his overall appeal as the first truly viable black candidate). The only candidate that can falter in the early states and still win the nomination is Clinton. Obama must win IA convincingly or this won't be a fight at all. He must, at all cost, avoid coming out third in IA. He'd be finished because wavering black in SC would decide that he won't be viable in the long run, and flock to Hillary (they can't stand Edwards).

Your whole premise is flawed. Obama has a very narrow path to the the nomination, and it involves doing well in IA and sweeping the early states, which might fatally wound  Clinton. Short of that, Super Tuesday would put this thing out of reach,,,

But, dcshungu, your premise also seems flawed, since it's based on the idea that Iowa and New Hampshire are inexplicable outliers compared to the rest of the country, where Clinton polls well ahead.  But Clinton polled well ahead in Iowa and New Hampshire until the caucus was almost upon us and the voters actually started paying attention.  Why shouldn't we think that a similar effect will not occur elsewhere, as well?

Liebig:

There is nothing that says that the winner of the early states would move on to do much better elsewhere. Each state prides itself on its independence. As for the early states, it is not true that Hillary had "polled well ahead" in IA. Edwards had enjoyed an early lead Iowa until it became deadlocked and has remained that way for some time. The tightening of the race in the early states is hardly surprising. It usually happens and often the favored candidates are not the ones who wind up winning there. There reason the race is usually more competitive there can be explained by two factors: (1) all the candidates pour large amounts of resources there and (2) a firing squad is usually formed around the candidate who is perceived to be the front-runner, and this case that has been Hillary who has been attached by both the Dem and Repub candidates. It is, in fact, remarkable that she is still standing there and even appears to be rebounding and poised for a strong finish. Unless she is fatally wounded in the early states (i.e., unless Obama sweeps) Clinton will quickly be in more friendly territory on Feb 5, where the other candidates, having staked their campaigns on doing well in the early states, would not have enough time to mount an effective defense against Clinton, and would get slaughtered. If Clinton wins Iowa, this thing would be over before it starts. Obama has very little margin for error: He has to win Iowa AND sweep the early states in order to fatally wound Clinton, who has a significantly wider margin of error and more paths to take to the white house than Scheiber has suggested.

I can't agree, dschungu.  First, at least according to the averages on pollster.com, Clinton (not Edwards) was in the lead in Iowa from about July until just recently.  Secondly, it's just silly to say that other states will act differently because "each state prides itself on its independence."  How do you square that with the fact that Kerry won virtually everywhere after coming in first -- at the eleventh hour -- in Iowa?  Finally, as I understand it, Obama has just as much money to work with on Super Tuesday as Clinton does, so there's reason to think the voters there will be bombarded equally by both candidates, just as the Iowa voters have been.  And Obama can only benefit if, in the meantime, Edwards has faded as the perceived alternative to Clinton.

Liebig sez:

"How do you square that with the fact that Kerry won virtually everywhere after coming in first -- at the eleventh hour -- in Iowa?"

Bill Clinton lost badly in Iowa against Harkin, a native son, and came out second in NH, famously called himself "the comeback kid." He then went on to win afterwards because of his strength in the south on Super Tuesday. He finally put it away by dispatching Gov Brown after winning NY.  The Kerry model is not as apt (Hart v. Mondale would be), because he benefited from the perception that he would be more electable than Edwards, who had very little experience at the time, just like Obama this time around. In that last respect, Clinton would still be favored over Obama because most polls show that people (~60%) think that she would be more electable. The early states have generally not been as decisive as they were in 2004. There have always been surprises there. Obama simply must sweep the early states if he is to be viable.Hillary is really that strong and she is now again looking like the candidate to beat, which does not bode well for the competition, especially in Iowa which is do or die for the others...

As of late, Edwards has been directing his fire at Obama precisely because he has determined that Clinton would be Clinton and would go the distance but that he, Edwards, would be done unless he is perceived as the viable alternative to Clinton. This means that Obama must crash and burn in Iowa. That seems to be like a worthy move for Edwards but he would be weaker than Obama because he is broke. Finally, the compressed primary schedule just does not give the challengers enough time to counter Clinton's overwhelming advantage outside of the early states (see how she has recovered nationally at pollster.com)

Interesting debate in comments about Noam's item on post-Iowa scenarios for Clinton, Obama and Edwards

Dcshungu, a lot of what you're saying is plausible, but the problem is that same set of facts is consistent -- maybe more consistent -- with an alternative plausible hypothesis: namely, that the race in any state will tighten in a similar fashion as that state's election day nears.  Just as Iowa tightened near the end, New Hampshire is now tightening -- more so between Obama and Clinton, since Edwards is not spending as heavily there.  There is every reason to think the same thing might happen elsewhere.  The Bill Clinton example just contradicts your premise: he came in second in New Hampshire, but came back because he had the money to keep battling in later states -- just like Obama does.  

Your attempt to distinguish 2004 is pure post-hoc reasoning.  Kerry was far behind in the nationwide polls until he won in Iowa.  Because of his Iowa win, and because he was in a financial position to capitalize on it, he erased those deficits practically overnight.  

The Kerry win just demonstrates how soft those nationwide poll numbers are at this early stage.  Moreover, the polls that show that people think of Hillary Clinton as more electable are the same polls that show her leading nationwide -- and are just as likely to change as each state approaches its election day.  A poll of Oregon voters today is about as reliable as the polls matching various Democrats against various Republicans for November 2008.  It's just a measure of name recognition until the voters have actually been subjected to the media blitz -- both free and paid for media -- that will inevitably occur as they approach election day.  Any forecast that relies that much on nationwide poll figures at this early stage is bound to go wrong.  

That's especially true given the general antipathy toward Clinton among large portions of the electorate, which isn't going to go away even if she wins by a few points in Iowa.  There's an anti-Clinton vote out there just waiting to attach itself to one candidate or another -- in a way that was not at all true in 2004 (toward Kerry) or 1992 (toward Clinton).

Finally, aren't you contradicting yourself?  First you say that "the early states have generally not been as decisive as they were in 2004," but then you say that "Obama simply must sweep the early states if he is to be viable."  I'm agreeing with your first part, and disagreeing with your second part.  Obama's viability will hinge on (1) how much money he has, (and he's doing well there), (2) whether he can quickly establish himself as the single alternative to Clinton, and (3) the extent of anti-Clinton sentiment.

Norm is right on the money, isn't he. Obama is in the driver's seat. All he has to do is get "ties" or "wins" and he wins. SC blacks are looking to see if the "white folks" in Iowa and NH give the thumbs up to Obama. I remember when Obama had a blowout fundraising 2nd quarter this year, shooting past Big MaMa, the SC blacks started then to rally to Obama's side when he showed such firepower against the revered (and hated) Clintons. Ditto Iowa and NH. If Obama is even or ahead in those places, and delivers, then SC will go Obama.

Those national polls are like quicksand. Hilary could be up 45-20% in them and a few early state wins by Obama and those national numbers are suddenly Obama ahead 33-28%. It is that "fluid" and that much a game of momentum.

I think Obama is probably our next president, unless the Republicans nominate someone electable like Ron Paul. At least Obama will not break your kneecap to get to the White House; Hillary will.

This is for DC who mentioned that some folks think that Hillary is more "electable." What part of Anybody But Hillary do you Democrats not understand. Republicans and independent voters understand this extremely well. Mark Penn says that Hillary will get 25% of Republican women; this is clear proof that Hillary's pollster is on crack. I am very active in grassroots Republican politics. I don't know ANY Republican women who will vote for Hillary. She is having a hard enough tying down her female base among the Democrats.

     Hillary's base among females are women who are 1) too old to have a computer 2) to poor, undereducated, uniformed to have a computer and access to the internet. Because it they did, they would find a LOT of things about Hillary - her sewer mouth, abusive attitude and lesbianism for starters - that they would not like. If all those church going blue collar women who support Hillary knew what you and I know, Hillary would not get their votes, much less swing voters or Republican women.

"What part of Anybody But Hillary do you Democrats not understand."

This is stupid... really. Anybody but Clinton (ABC) is a slogan. Take a look at the number of who identify themselves at Dems, Repubs, and Indies...and then use your mind,

This is the way it is going to happen: The die-hard GOP folks would vote for their nominee. This means that 30-40% of the vote has already been preempted. The Dem nominee would get about the same number (automatically), which would leave about 20% in the middle: the so-called swing voters. Those people in the middle like someone like Hillary, who is a centrist!!!!! Her husband was too, and that is why he is the only Dem who got elected and re-elected in 30 years.

America is in the center of the divide. Kucinich will never win the presidency because he is too ideologically "pure" (he is a liberal), but nor would someone from the GOP  who is too far too the right...  

Liebig:

I do not think that we will resolve this one. I submit to you that Clinton is the stronger candidate for the Dems. Obama is an ambitious poll,  who should have waited his turn to run (No one knew anything about him other than that he was "articulate" at the 2004 Dem Convention). He wants to go from Junior High to CEO of America with very little experience. Why is he in such a rush (very few people write their. autobio by the tine they are 45. Obama has written two best-selling autobios by age 45). Meanwhile, he has been short-changing the people of Illinois by skippinbg votes and getting truning around and slamming people who bothered to show up for votes.

Please vote for Obama  since you have been taken in. I will support him if he is the nomimee but I think that he would be our worst nominee. We'll an election that we could not possible lose if we nominate Obama: He is too "green" and he will not win in Dixie (not his fault that America is not yet fully color-blind).

I am done, so cheers!.

DC -- seems like we went from talking about who was *likely* to win to talking about who we'd *like* to win.  Of course I disagree with your assertion that undecided voters will like Clinton best -- voting is not as rational or one-dimensional as you make it out to be.  That is, "independent" voters aren't necessarily "centrists," and voters don't necessarily pick the candidate nearest to them on the so-called spectrum.  The widespread antipathy toward Clinton affects even people whose views are generally in line with hers.  If it;s hurting her among Democratic primary voters, who generally remember the Bill Clinton presidency fondly, one can just imagine how it will hurt her among people who aren't active Democrats.  Meanwhile, many voters who do not consider themselves "liberal" remain interested in Obama.

None of that bears on what Obama needs to do to remain a viable candidate in later primaries, though, as you seem to recognize.  Maybe we're both guilty of some wishful thinkling there, but I stand by the points in my last post on that issue (which I notice you haven't tried to refute).