I'm not a fan of SUSA's vice presidential matchups since all they've proven so far is that veeps with higher name ID (i.e. Edwards and Romney) do better than the folks no one recognize. That's not exactly a big revelation.
But given that Kansas Gov. Kathleen Sebelius is most likely on Obama's Veep shortlist, it is a good way to test the theory (which I firmly believe) that veep candidates don't deliver geography. So while Obama's path to victory doesn't run through Kansas, would she be able to deliver the state for Obama?
SurveyUSA. 5/16-18. Likely voters. MoE 4.1% (No trend lines)
McCain (R) 49
Obama (D) 39
That's actually not terrible considering that Obama wins Democrats by just 61-24. Yet another state in which the primary battle has depressed Obama's Democratic base. Probably related, he loses the woman vote 46-42 (though Kerry handily lost the woman vote in 2004 62-38, so perhaps it's not related). He wins Independents 45-38.
So what does Sebelius do to the ticket? Not much, actually. Obama/Sebelius win under none of the tested matchups. Obama/Sebelius versus McCain/Pawlenty loses by only four points, but swap out Edwards for Sebelius, and they lose by just two. It also just happens that Pawlenty is the only tested McCain veep with zero name recognition. The others are Huckabee, Romney and Lieberman.
Against McCain/Lieberman, Obama/Sebelius lose by seven, and it goes downhill from there. So at best, she may be worth three points.
Then again, that may be worth something. If you take Obama's Democratic performance and match it to McCain's (81-16), then my back-of-the-napkin calculations come up with the following number:
McCain (R) 47.2
Obama (D) 47.0
Undecided: 5.8
For reference, Kerry won Kansas Democrats 84-16 in 2004, while Bush won Republicans 94-5.
However, the 2004 exit poll sample was 50 R, 27 D, 23 I. In this SUSA poll, it's 42 R, 40 D, and 15 I. That's a 21-point swing in the sample in favor of Democrats.
Is it warranted given the leftward drift of the electorate the last four years? I don't know. It's the sort of educated guesses that pollsters are paid the big bucks to make.
