The interesting thing about MS and Alabama, is that Mittens did quite well, where he usually blows (maybe ironically "the folks" respect more what "their leaders" are saying, mostly for Romney, because that is the culture, rather than the more sharp elbowed culture that most of us enjoy, where we tend to enjoy telling "the establishment" of our individual cohorts to just take a hike - just who the F do they think they are?), and blew where he usually rules. Demographics didn't matter much, where they have been key everywhere else. If Nate did not pick that up, he's missing something. The issue is why Mittens blew among the upper middle class, and whether that portends anything. Probably not, and the LDS thing needs to be touched upon, but it deserves more discussion than it is getting.
Nate, to is credit, actually pointed out that Romney does much more poorly in affluent suburban areas in the South and Midwest than the rest of the country. Sean Trende really nailed this with his demographic analysis, though.
OK, but why did Mittens do much better with the upper middle class in Georgia and SC? I don't think the 3 cornered (for some reason Nate didn't get out west with this) regionalism factor in Nate's model explains it all. Unless the upper middle class is unusually Evangelical in AL and MS. Did Nate touch on that?
I haven't seen anyone write on that. I don't think any people expected Romney's lackluster performance in places like DeSoto County, MS. It's a fascinating phenomenon that I wish I knew more about (the religiosity thing I mentioned is only my pet hypothesis.)
DeSoto isn't where you find the fancy suburbs of Memphis. Those are all in Shelby County. It's much more of a Wal-Mart kind of place.