[Posted by Karl]
The bloggy thing to do this morning would be to link the results of the new Quinnipiac University Swing State Poll of FL, OH and PA, toss in today’s media focus on VA from ABC News and others, then do some analysis of the strategies the campigns might be to pursue some given set of swing states. Indeed, I have done posts in that vein before (albeit with some nuance I won’t get into here). But today I feel more contrarian and nitpicky.
First, these polls and media stories merely confirm what we would have surmised a year ago: FL and OH are going to be close, PA remains a tough get for the GOP, and VA has been trending Democratic but not a sure thing for Obama in light of the 2009 and 2010 elections there.
Second, as Nate Silver notes, state polling is still noisy at theis point in the campaign.
Third, as political scientist Andrew Gelman notes, the past several decades have seen a steady decline in the variation of statewide vote swings. Come November, the swing in the swing states will likely mirror the swing nationally. Electionate makes a similar point, although I have some disagreement with the underlying reasoning:
There’s a growing chorus arguing that Obama has an electoral college advantage. The underlying assumption is that the race is close nationally and yet Obama seems poised to secure well over 300 electoral votes. In my view, that argument is misguided for a simple reason: the race isn’t close nationally, and the electoral college consequently reflects an Obama advantage.
Electionate’s claim that the race isn’t close nationally is based in large part on the argument that Rasmussen and Gallup are skewing perceptions of the race. I will not rehash the claims against Rasmussen; some of them are quite reasonable, others less so. Gallup defends its polling here and here. Rather, I will note that Electionate’s plot excluding Gallup and Rasmussen tends to show a slowly tightening race, which is what you see with Gallup and Ras in the mix. Eyeballing the plot suggests Obama currently has an edge of a bit over 4% — but today’s RCP average gives Obama an edge of 3.6%. That’s not much a difference, particularly when considering that head-to-head polls at this point in the election cycle explain less than 50% of eventual results. [Note: Electionate does not name RCP as an offender on this score.]
However, this is another reason to focus more on Obama’s job approval number than any Electoral College map at the moment. The current RCP averages are 48.3% approve, 47.4% disapprove. If you exclude Gallup and Ras, 47.8% approve and 47.3% disapprove. Again, judiciously including Gallup and Ras has no significant effect on the numbers; if anything, they boost Obama’s approval number. For the zombies focused on the 2004 campaign analogy, note that while Bush had declining job approval eight years ago, he went into the election with a 49.8% job approval by the RCP average.
None of this will keep we political junkies from obsessing over polls and maps. It’s fun to do that. Just keep in mind that at this point in the campaign, they probably do not tell you what you really want to know.