OK, this post is about the presidential election, so if you find that kind of thing boring, just move along. I myself am fascinated by such things, and have been pondering a few questions:
First, does Mitt Romney’s erratic performance in the primary season, and especially his wipeout in MN, MO, and CO Tuesday, tell us anything about his chances in the fall? There doesn’t seem to be much evidence historically for a correlation between primary and general election performance, but I think we can learn something from the epic battle between Obama and Clinton last time around. Some trends in the primaries predictably disappeared in the general election; for example, most Latinos preferred Clinton over Obama, but they had no trouble preferring Obama over McCain.
But some patterns in the primary voting did persist, most notably Obama’s Appalachian problem. The demographics in the belt from West Virginia to Arkansas, and including parts of Pennsylvania and Ohio, was never going to be extra Barack-friendly, being largely devoid of liberal yuppies, black people, and universities, and his campaign invested few resrouces in that part of the country. Even so, it’s impressive to see how badly Clinton kicked his ass there: 67%-26% in WV, 65-30 in KY, even worse in AR (admittedly a state where Clintons are popular). And looking at the November results, you have to think that Appalachians just didn’t like him, whether it was Bittergate, or his somewhat swarthy complexion, or his taste in music, I don’t know. Obama actually lost Arkansas and Tennessee by wider margins than John Kerry had in ’04, and about as badly in WV, despite his overall much better national performance and despite the fact that Kerry was never mistaken for a coal miner’s daughter either.
Getting back to the present, I think Romney’s problem in South Carolina is comparable to Obama’s problem with Latinos: Southern white men are not going to vote for Obama even if the Republican nominee is Abraham Lincoln. But what to make of the more recent disasters? It’s not just that he lost Colorado, which he was suppsed to win easily, but that he was obliterated in the Midwestern races…I mean, in Minnesota I don’t think he won a single county, not even the suburban ring counties that are supposed to be his base. He lost by a mile to Santorum, but he also lost to Ron Paul, and it wasn’t close. I mean seriously, Ron Paul, the Andy Rooney of Presidential politics.
This looks to me like genuine distaste, something like what made so many residents of the border South unwilling to vote for someone smarter and darker than themselves in 2008. But while Democrats can afford to write off Appalachia with only a small pang of regret that this once-bluish region has gone off the rails, a Republican who writes off the Midwest is making his life very hard. The people in MN and MO who apparently loathe Mitt are pretty similar to people in Iowa and Michigan and Wisconsin.
To be sure, these were low-turnout affairs, and it could be that the sort of Bachmann-style irredentists who turned out Tuesday will all hold their noses and vote Mitt in November. But I wonder if it isn’t also a class thing—Midwesterners, like the rest of American, worhip wealth, but as an old girlfriend of mine from Massachusetts once commented, in Minnesota the rich seem to be just middle-class people with lots of money—they own six personal watercraft instead of one. Mitt may be sending out whatever the opposite of a dog-whistle is (my brother’s dog used to be deeply disturbed by the theme song for a soap opera, “Young and the Restless” maybe—I guess that would be the opposite of a dog-whistle. I gather that Newt has even accused him of speaking French. ‘Course, none of this will matter if he doesn’t win the nomination.
Hmm, this is getting a bit long, so I’ll just tackly one more thing quickly: What are the most likely battleground states?
The best place to start is the last election, but keep in mind that the battleground states won’t be the states that were close last time, but (this is obvious, right?) the states that would have been close in an overall close election. For example, Missouri and Montana were close in ’08, because Obama was romping to an easy victory, and if they’re close this year, the state-by-state details won’t matter much because the election will be another wipeout. So what we want are states where the margin was similar to Obama’s national margin of about 7%.
In the narrow 4%-10% window we have Virginia and Ohio (leaning slightly Republican), Iowa, Colorado, and New Hampshire (leaning slightly Democratic), with Florida, Minnesota, and Pennsylvania at the next level. (2-12%). One might also want to throw in Michigan, which Obama won by 16 points. That margin is a bit deceptive because the McCain campaign, in one of its many erratic and even petulant moves, publicly abandoned the state, pulling out its resources and presumably creating ill-will as well as puzzlement. It’s also possible that Arizona could be close, though it hasn’t been much of a battleground in recent years (of course 2008’s 8-point McCain victory was largely due to his native-son status).
Note: I get my stats from the lovely US Election Atlas:
And of course the guy who does this kind of stuff best is Nate Silver: http://fivethirtyeight.blogs.nytimes.com/