In TNR, Clay Risen revisited the question of voting patterns in the South in this year’s presidential elections, and responded to a point of criticism I raised here on his previous take. Being an incurable nitpicker, I’m still not altogether convinced.

In his previous take, Risen justifiably sounded critical notes about how the South was presented in some of the electoral analysis, which was all about how its “backward ways are increasingly irrelevant to the American scene”. He pointed out that hey, in much of the South Obama actually did better than Kerry had done, thank you very much. The “red splotches” on the electoral map that showed a shift to McCain only covered a specific band of counties stretching from Kentucky to Oklahoma, while “across the “Deep South” [..] the map is almost entirely blue.” Ergo, what the map showed was “not a waning South, but a fissured and rapidly changing one”, and “what is really surprising is not how stalwart the South is in its ways” but “that broad swaths of the region look just like the rest of the country.”

The former is certainly a good point. My own stab at electoral analysis here highlighted how the Gulf states and Atlantic states seem to be heading down different paths altogether. But the latter point has a problem or two.

For one, the South was actually further out of sync with the national vote this year than it was in either of the two previous election cycles. When it comes to the margin between the Democratic and Republican presidential candidates, the South was off from the national standard by 12.5% in 2000; 13% in 2004; and 15% in 2008. Small differences, for sure, but still – if anything, then, the South looks less like the rest of the country this year than in either 2004 or 2000.

Moreover, was the point I raised last time, in as far as Obama made gains compared to Kerry in the Deep South, it was almost entirely thanks to his mobilisation of black voters. White voters in Louisiana, Mississippi and Alabama actually moved toward McCain, and more strongly so than those in the “red splotches” a tad further north. And although of course white and black Southerners are all Southerners, “an increasingly racially polarised political landscape like that of Alabama and Mississippi, where whites rallied even more behind the Republican banner while the Democrat gained ground by even better mobilising the black vote, is not exactly what you found in the rest of the country.”

This is the point Risen now revisits. “Biracial coalitions supposedly cohered in the rest of the country,” he summarises the criticism, “while racial division defined the South”. But, he writes, an analysis by two MIT political scientists in the Boston Review now debunks this notion. After all, its authors write, “Racial polarization in American voting patterns (the difference between black support for Democrats and white support for Democrats) was the highest it has been since the 1984 election.” In short, Risen concludes, “much of the rest of the country” saw the same “significant increase in racially polarized voting” as “those Alabama and Mississippi counties”.

Electoral map reprise: electoral shifts among white voters, 2004-2008

How has the white vote shifted between 2004 and 2008? In this map, McCain getting 25% more of the white vote in a state than Bush did in '04 would colour the state a fiery red; McCain getting 25% less would make it the coolest blue. The map shows that whites in much of the Deep South swung to McCain, while whites in the Mountain and Pacific West, the Midwest and the Atlantic South swung strongly to Obama.

You can’t argue with the data, of course. The exit polls do show that the difference between black and white support for Democrats this year was higher than in many previous years. To quote the Boston Review article: “The percentage of blacks voting for the Democratic presidential candidate rose from 88 percent in 2004 to 95 percent in 2008,” but “white voters [..] increased their support of the Democratic candidate by just 2 percentage points, from 41 percent for Kerry to 43 percent for Obama.” The difference between the two was thus larger this year (52%) than in 2004 (47%), or, for that matter, 2000 (48%), 1996 (41%), 1992 (44%) – et cetera. Racial polarisation increased!

What a paradox: Obama’s “victory, built upon the highest degree of racial polarization seen in many years, has ushered in a period of racial good feelings.” But is it really ironic, as Risen says? I don’t know. The increased racial polarisation on a national level, and even more so in the Northeast, Midwest and West, consists of whites moving toward the Democrat … but just not as fast as blacks and hispanics. That’s the bottom line: in the first time ever that the Democrats ran a black presidential candidate, he actually got a higher share of the white vote than his white predecessors had gotten. The only other time the Democratic candidate got as much as 43% of the white vote in the last twenty years was when Clinton ran for reelection in 1996. Who would not have “racial good feelings” about that?

And this is exactly where the South, with the striking exception of the Atlantic states, is different from the rest of the country. Yes, racial polarisation increased throughout the country to the extent that the white shift to Obama did not keep pace with the black and hispanic shift in his direction. But outside the Deep South and the “hillbilly belt” ranging from Kentucky to Oklahoma, there was a significant white shift to Obama. In no less than 22 states, covering the Pacific coast and much of the Western mountains and plains, the Great Lakes area and New England, Obama got at least 5% more of the white vote than Kerry, on an above-average swing. It was only in parts of the South that whites and blacks actually moved in opposite directions. In Louisiana, Alabama, Arkansas and Mississippi, whites moved further right and blacks moved left. The same holds for most of rural Kentucky, Tennessee and Oklahoma. And that’s a pattern you did not see barely anywhere else.