After Barack Obama lost the New Hampshire primary, the punditocracy (including some pollsters) started speculating about the return of the "Wilder Effect," a.k.a. the "Bradley Effect." Named, respectively, for black gubernatorial candidates in Virginia and California, the Effects occur when white voters tell a pollster or exit pollster that they're voting for a black candidate, too embarrassed to admit that they voted for the white. In 1989, Douglas Wilder nearly lost his bid for governor of Virginia even though exit polls showed him winning by close to 10 points.
This year now-Richmond Mayor Wilder endorsed Barack Obama, and Obama won a historic landslide in the commonwealth. Final polls estimated that Obama would win by about 18 points. He won by 28 points. Exit polls estimated that 48 percent of white voters (a majority of men, a minority of women) went for Obama. The results basically prove that case. Look at the county-by-county map of Virginia, which shows a wave of Obama blueish-purple washing over the state and breaking in the southwest, at Bedford County. And look at some of the counties Obama won. Highland County, which borders West Virginia, is 99 percent white, and it went 54-45 for Obama. Floyd County, deep in Appalachia, is about 97 percent white, and it went 49-48 for Obama. Virginia has 11 congressional districts, all but two of them with white majorities, and Obama only lost one of them, the ultra-conservative 9th, which looks more like eastern Tennessee than the rest of Virginia.
I tried to tramp down dirt on the "Wilder Effect" last year, when it was pretty obvious that polls were accurate in six white-versus-black statewide races, and I was a little worried when the theory resurfaced this year. Could it surface again in a heated general election? Well, maybe. But not in Democratic primaries anymore. Barack Obama has read the rites and exorcised that ghost.