Friday, March 14, 2008

Can Obama win Pennsylvania?

My good buddy Kyle Moore at Comments from Left Field asked me that question this morning via e-mail. I live in Pennsylvania and I think I have a decent feel for Democratic politics around here, and I think it depends on several definitions of winning...

Can Barack Obama accumulate 50%+1 votes --- extremely unlikely, but plausible if Clinton withdraws or meltdowns

Can Barack Obama accumulate 50%+1 pledged delegates --- unlikely but much more plausible than winning the popular vote due to delegate allocation rules.

Can Barack Obama do what he needs to do to maintain his overall edge over Clinton in Pennsylvania --- highly likely.

Can Barack Obama accumulate 50%+1 votes (5% chance)
6 weeks out of the election, Barack Obama is trailing Hillary Clinton by approximately 14 points accordnig to Pollster.com's combined polling measure. He has been closing against Clinton since approximately September/October 2007 where his gradual uptrend inflected to a much greater slope. So can he close the remaining fourteen points in the next five and a half weeks? I don't think so.

We can count on Obama to continue to consolidate his African American base, as Rasmussan has him up 79-13 in this demographic. SurveyUSA has Obama winning the African American vote 76:22. If we project that he continues to win the black vote by roughly 90-10 margins, and that African American voters make up 15% of the Democratic Primary universe, then he can expect to close the 14 point gap by four net points.

He then needs to pick up a net of ten points from white voters. There is significant space for him to grow among self-identified liberal Democrats and voters who consider Iraq and foreign policy the areas of greatest importance. However there is most likely significant overlap within these subgroups. Clinton will roll-up the conservative Democrats in the central part of the state (excluding State College) and do well with the white working class voters in Northeastern and Southwestern Pennsylvania. If she continues to hammer on NAFTA and has the press throwing up its hands in frustration over how confusing the Canadian side of the story is, she'll do very well here, and Obama can not count on making significant in-roads.

I think the most likely outcome is that the Clinton campaign (which is doing a good job of organizing its likely voters in SW PA at least) will break even with Obama's ground game, and grind out a healthy six to eight point win in the state, so my call is 52: 46 with 2 points going to 'other'.

Can Barack Obama accumulate 50%+1 pledged delegates

Unlikely, but plausible (10%-15% chance)--- The distribution of support for Obama and Clinton need to be arrayed in a very precise, heterogenous pattern that is minimally impacted by both campaign's strategic interactions. RuralVotes is projecting a Clinton blow-out, but I don't think this is likely given the make-up of the districts. Chris Bowers is also projecting a Clinton delegate win but by much smaller and to my local eyes, more reasonable numbers. For Obama to win the delegate count, he needs to win Philadelphia by massive margins (PA-1 and PA-2 by 80:20 margins (unlikely, but plausible) to rack up large SE PA delegate counts, win PA-14 by 19 points for a three delegate pick-up, and then run no more than 10 to 12 points behind in all of the even delegate districts and keep it within 15 to 18 in the remaining odd-delegate districts so that the splits are either 2:1, or 3:2 for Clinton. He would need a lot of things to go right, but if we assume a 90/10 African American vote split for Obama, as I did earlier in the post, the base for urban blow-outs is present.

Can Barack Obama do what he needs to do to maintain his overall edge over Clinton in Pennsylvania (90%)

As I see it, he just needs to avoid being blown out in the pledged delegate count, and the popular vote. The schedule becomes more favorable to him after Pennsylvania, and the number of outstanding delegates left to be won is shrinking rapidly. If Michigan and Florida are resolved without a do-over, he'll be leaving May with a string of victories for local momentum and a very solid pledged delegate lead, and at current trends, breaking even with Hillary Clinton in the superdelegate race. He will not have hit the magic number, but he'll be much closer. In Pennsylvania, he needs to do decent and not stick his hand or other appendage into a buzz-saw in order to do what he needs. And given the past history of his campaign, I am confident that a minimized loss operation is highly likely.

No comments: