Friday, February 1, 2008

Predicting Super Tuesday

Despite all the hype, I think it's unlikely that Super Tuesday will result in a clear winner on the Dems side, particularly given the way in which the Democrats distribute delegates. Obviously, I expect Super Tuesday to put an end to Romney's hopes on the Republican side.

Unless Hillary manages to get about 60% of the overall vote on Tuesday or Obama wins a majority of the total vote plus a significant majority of the states, I think both sides will be able to spin just about any result as a victory for them and still pass the laugh test. I do expect that Hillary will come out clearly on top in both delegates and vote percentage, though by nowhere near enough to knock Obama out of the race or to plausibly argue that she is the presumptive nominee.

With that in mind, below is my best guess of the results in each of the Super Tuesday states based on the recent polling data I could find from Pollster, and intuition where recent polling is unavailable.

Solid Obama: Illinois, Georgia, Democrats Abroad
Lean Obama: Alabama, Idaho, Colorado, Delaware, Kansas, North Dakota
Toss-Up: Connecticut, New Mexico, Arizona, Minnesota, Missouri
Lean Clinton: New Jersey, Massachussetts, Utah, California, New Mexico, Tennessee
Solid Clinton: New York, Arkansas, Oklahoma

Looking at this, I noticed that my independently arrived at conclusions are pretty close to Marc Ambinder's predictions the other day. To be honest, I am quite surprised that I found Obama as the favorite in just as many states as Hillary, although Hillary does better in the most delegate-rich states, which is why I think she's likely to come out of the day with a higher delegate count and higher vote total. Still, if things shake out as I have it above, with the toss-up states being split evenly, it will be impossible for Hillary to breathe any kind of sigh of relief. It's just about impossible to see how this race gets decided without the Super Delegates- unless Obama takes all of the toss up states and steals at least two of the "Lean Clinton" states. Were that to happen, I could see him gaining enough momentum to take sizable majorities in the remaining primaries.

If you had asked me a few weeks ago, I would have said that a convention that comes down to Super Delegates would be bad news for Obama; but with the recent spate of endorsements he's received, I'm not so sure now. Hillary does have a 2:1 advantage currently in pledged Super Delegates, but the vast majority of Super Delegates have stayed out of the fray so far. Obama's endorsements will help him considerably with the hundreds of remaining Super Delegates.

One important caveat to all of this which has been under-covered: now that McCain's nomination seems almost inevitable, independent voters may turn out for Obama in somewhat heavier numbers in states with open primaries. I don't think it will make that much of a difference, but it could put him over the top in some of the toss-up states.

**UPDATE** A few explanations of some of the less intuitively obvious predictions:

California- Obama has narrowed the gap substantially here, and according to one recent poll has closed to within three points of Hillary, which would suggest the state is a toss-up. But that poll looks like it might be an outlier, as most polls have it scored as about a 10-15 point lead for Hillary. I'm guessing Hillary will win the state by about 5 points after Obama's momentum is factored in.

Delaware- there is no polling data here, but demographically the state is extremely favorable to Obama.

Kansas- again, no recent polling, but Obama has ties to the state, and it is the type of state in which Obama seems to be polling very strong (compare with Idaho, for example, where Obama had a lead in a July poll - well before he had any momentum, or with Iowa, which is where Obama started his surge).

Connecticut- one recent poll has Obama with a slight lead here and another has it tied, which would suggest that Obama should be the favorite. However, Connecticut is smack in the middle of Hillary country, and I would expect her to have a big advantage in the ground game here, which puts the state back in the toss-up category.

Massachussetts- despite the Kennedy endorsements, this is machine Democrat central, and Obama's initial deficit is far too much for him to overcome. The Kennedy endorsement will, however, make it much closer than it would have been.

New Jersey - recent polling has Obama closing fast here, and the state's high level of education allows Obama to be competitive. New Jersey's more urban black population is not going to be as pro-Obama as the rural black population in South Carolina, as New Jersey's cities are completely dominated by political machines that are Hillary's bread-and-butter. Obama will be helped in those areas by the support of highly-regarded Newark mayor Corey Booker, but New Jersey is just too much of a machine state for Hillary to lose.

Minnesota- polling still has Hillary with a lead of several points in the state, but demographically it is almost an ideal state for Obama. The history of Minnesota's politics lends itself even more heavily to Obama's favor.

North Dakota- this state is just an educated guess based more on Obama's win in Iowa than anything else.

New Mexico- there is no reliable polling in New Mexico right now, but the heavy Latino population in the state strongly favors Clinton, as do other demographic features like income level.

UPDATE #2: Today's Gallup Poll places Obama within the margin of error nationally against Hillary. This doesn't change the above analysis in any way, though.

More at memeorandum.