Sunday, February 3, 2008

Democrat Tsunami Tuesday Delegate Estimate: 2-3-2008 FINAL

OK, this is the last time I'm doing this prior to Tsunami Tuesday. Things might change between now and then, but I won't have time to look over the poll results Monday... and I doubt any polls will be released on the day itself, to avoid biasing the voters. (Ha ha ha. Yes, because all these polls BEFORE the election have no effect at all...)

Today's polling news is Zogby. I like Zogby's polls; they have a longstanding reputation for accuracy, and they've come closer than any other national polling service to predicting the outcome of all the primaries so far (except New Hampshire, which everybody blew). They also tend to have the largest sample sizes and lowest margins of error of any pollster... which is what makes these results so shocking.

* GEORGIA (primary, 87 delegates): (Zogby 1/30 - 2/2) C (28%) 32, O (48%) 55 (Swing: 7 to Obama.)

* MISSOURI (primary, 72 delegates): (Zogby 1/30 - 2/2) C (44%) 36, O (43%) 36 (Swing: 2 to Obama.)

And the really big shockers, the one that already has people shouting, "Waaaait a minute..."

* NEW JERSEY (primary, 107 delegates): (Zogby 1/30 - 2/2) C (43%) 54, O (42%) 53 (Swing: 6 to Obama.)

* CALIFORNIA (primary, 370 delegates): (Zogby 1/30 - 2/2) C (41%) 176, O (45%) 194, (Swing: 18 to Obama.)

That's right. Zogby calls New Jersey in a dead heat... and California with Obama leading outside the margin of error. (Of course, factor in the "Keith Number"- I watch Countdown with Keith Olbermann almost religiously- and there's a fudge factor of 18%, but still.)

In California, the Field Poll shows Clinton still ahead, but a truly massive undecided- 36-34. Mason-Dixon (polling for McClatchy-MSNBC) shows Clinton blowing Obama out of the water in a definite statistical outlier, 45-36. Zogby is, at present, the only poll showing Obama leading in California at all.

Mason-Dixon also represents the most optimistic poll for Clinton in Missouri, at 47-41 Clinton. Zogby isn't the best-case poll for Obama, though; American Research Group, which has pretty consistently weighed in Clinton's direction, shows Obama in a marginal lead, 44-42- the only poll to do so.

Lest we totally discount Mason-Dixon, though, in New Jersey their poll shows Clinton winning handily 46-39; the third New Jersey poll out today, Gannet, shows an even larger 50-36 Clinton win.

What has me leaning against Mason-Dixon is simple: Zogby has a sample size of more than double Mason-Dixon's. That, plus their past track record, adds up to me picking Zogby numbers over anyone else's.

National polling also tends to back up Zogby's claims. Although Rasmussen shows Clinton opening up to a double-digit lead in their three-day average, Gallup over the same period shows Clinton and Obama as close as they have ever been- only a two percentile difference, 44-42. Washington Post/ABC leans much more towards Gallup than Rasmussen, showing Clinton up only 47-43.

Other polls released today:

* NEW! OKLAHOMA (primary, 38 delegates): (Sooner Poll 1/27-30) C (41%) 27, O (17%) 11, E (24%) 0 (No swing.)

* NEW! ARIZONA (primary, 56 delegates): (Rass 1-31) C (46%) 30, O (41%) 26 (Swing: 2 to Obama)

Oh- and remember what I said about American Research Group leaning in Clinton's favor?

* NEW! DELAWARE (primary, 15 delegates): (ARG 1/31 - 2/1) C (44%) 8, O (42%) 7 (Swing: 3 to Obama.)

UNPOLLED

* ALASKA (caucus, 13 delegates): (not polled, WAG) C 7, O 8, E 0
* AMERICAN SAMOA (primary, 3 delegates): (not polled, WAG) C 2, O 1, E 0
* Dems ABROAD (primary, 7 delegates): (nat'l polls used) (Gallup 1-29-31) C (44%) 4, O (41%) 3
* KANSAS (caucus, 32 delegates): (not polled, WAG) C 13, O 19, E 0
* NORTH DAKOTA (caucus, 13 delegates): (not polled, WAG) C 6, O 7, E 0

OLD POLLS

* IDAHO (caucus, 18 delegates): (Greg Smith 7-12-2007) C (31%) 25, 9 (33%) 9, E (15%) 0
* NEW MEXICO (caucus, 26 delegates): (??? 9/5/2007) C (17%) 17, O (8%) 9, E (8%) 0
* UTAH (primary, 23 delegates): (??? 2-11-2007) C (31%) 15, O (18%) 8, E (9%) 0

JANUARY POLLS

* ALABAMA (primary, 52 delegates): (Ras 2-1) C (46%) 27, O (41%) 25, E (9%) 0
* ARKANSAS (primary, 35 delegates): (??? 12-14-2007) C (57%) 27, O (17%) 8, E (14%) 0
* COLORADO (caucus, 55 delegates): (MD 1-21-3) C (32%) 27, O (34%) 28, E (17%) 0
* CONNECTICUT (primary, 48 delegates): (USA 1-30) C (44%) 22, O (48%) 26
* ILLINOIS (primary, 153 delegates): (Rass 1-30) C (24%) 44, O (60%) 109, E (10%) 0
* MASSACHUSETTS (primary, 93 delegates): (Rass 1-28) C (43%) 50, O (37%) 43, E (11%) 0
* MINNESOTA (caucus, 72 delegates): (Humphrey 1-30) C (40%) 39, O (33%) 33, E (12%) 0
* NEW YORK (primary, 232 delegates): (Rass 2-1) C (54%) 140, O (38%) 92
(Rass- Rasmussen, USA- Survey USA, MD- Mason-Dixon, PPP- Public Policy Polling, WAG- Wild Ass Guess)
* TENNESSEE (primary, 68 delegates): (Rass 2-1) C (49%) 40, O (35%) 28

New count:

Clinton 852, Obama 838 on February 5. (Swing: 38 to Obama.)

That's a delegate margin of only fourteen delegates. That's essentially a dead tie.

Now, leaving aside these numbers, what do I really expect on Tuesday? Bear in mind all my predictions on this election so far on the Democratic side have been pretty damn wrong. These are just guesses, with only an outsider's viewpoint to go by.

CLINTON WINS: Alabama, American Samoa, Arizona, Arkansas, California (narrowly, based on early votes), Massachusetts, Minnesota, Missouri, New Jersey, New Mexico, New York, Oklahoma, Tennessee and Utah.

OBAMA WINS: Alaska, Colorado, Connecticut, Delaware, Democrats Abroad, Georgia, Idaho, Illinois, Kansas, and North Dakota.

That's 14 Clinton states to 10 Obama states.

More importantly, I'm going to predict an actual delegate difference of just barely over 100 delegates in Clinton's favor when all the results are in. Added to the current pledged delegates (not counting superdelegates, which can always change), Obama will remain within 100 delegates of Clinton.

What does this mean? Well, to duplicate what I said on my LJ: an advantage of 200 or more delegates for Clinton on Tsunami Tuesday is a mortal wound for Obama. With more than half the pledged delegates in, Obama would be highly unlikely to overcome such a margin.

An advantage of 100 to 200 for Clinton amounts to a de facto tie, with Clinton's home states of Arkansas and New York behind her, along with Obama's home state of Illinois. The rest of February's contests consist of either caucus states (where Obama's money and organization serve him well) or primaries with large African-American electorates. The momentum from those states may well shift the remaining big-delegate states- Pennsylvania, Ohio and Texas- more into Obama's favor. (All three currently favor Clinton pretty heavily.)

A Clinton lead of less than 100 delegates favors Obama pretty strongly, and may well lead to endorsements from Edwards, Richardson, or both. It'd definitely outperform expectations.

If Obama gets more delegates out of Tuesday than Clinton, even one more delegate... or if he wins California, where both candidates have thrown everything they could into the contest... then Clinton's campaign is maimed, possibly fatally. If Obama wins Tsunami Tuesday even by the slimmest of margins, Clinton's support within the party will fade. All reports are that her supporters consist of either grudgingly-given endorsements based on arm-twisting or opportunistic politicians bought off early in the campaign with promises of inevitability. A shift of superdelegates to Obama and away from Clinton would spell doom for her campaign, although she's almost certain to keep fighting right on down to the Denver convention.

As for me, I'm done. I'll only be glancing at national polls tomorrow and maybe Tuesday. I think the delegate count I just painted is quite a bit too rosy- it doesn't take into account by-district distribution of delegates, and it uses polls which tend towards the extreme of the range of polling. I'd go so far as to call my estimate an Obama best-case scenario. I'm looking towards a narrow but substantial Clinton win, with Obama picking up enough states to not only remain viable but gain ground down the road.

But maybe we'll all get lucky, and Bill Clinton will make headlines today or tomorrow. I hope, I hope, I hope...

1 comment:

Anonymous said...

I think you're wrong with Utah (new poll landslide for Obama), Idaho and New Mexico.

I'd say:

Utah: C 10, O 13
Idaho: C 8, O 10
New Mexico: C 14, O 12