Josh Marshall, having been mocked for not bothering to do any homework on racial voting patterns in the Democratic presidential campaign, coincidentally discovered Brendan Nyhan’s analyses of the subject.  Indeed, Nyhan’s most recent analysis deconstructs recent nuttiness on the topic from David Sirota, concluding that “the claim that Obama does uniquely poorly with whites in states with moderate black populations is not well supported.”

To save face, Marshall claims that he is “not sure this apparent disagreement is anything other than the same point expressed a different way.”  And Nyhan writes that he tends to agree with that.  The problem is that Nyhan’s original analysis and his March 5 update showed a statistically insignificant decrease in white support for Obama as a state’s black population increases, but a statistically significant quadratic relationship showing an increase in white support for Obama as a state’s black population increases.

Another problem is that there are several variables explaining voting patterns in the Democratic campaign beyond mere black and white.  Indeed, Nyhan, like Jay Cost, identifies “southernness” as one of them.  As the Daily Howler observes of Sirota’s analysis:

Long ago, we observed an unfortunate fact; some men go into stand-up comedy to ridicule women from a bright stage, with the help of a microphone. Similarly, some people seem to be become “progressives” so they can accuse average schmoes of racism, thus displaying their own moral grandeur.

Then, after noting that Obama won the Washington state primary by a much smaller margin than he won the Washington state caucus:

So which is it? Is Washington a 68-31 state? Or is its “real” margin 51-46? And what would have happened if those other states had conducted primaries instead of caucuses? Once again, there’s no way to know—and the force of Sirota’s lusty charge stems from the big margins achieved in low-turnout caucus events. But in a great deal of modern “progressive” politics, the real purpose of the exercise is fairly clear—the real goal is the desire to brand “low information voters” as slobbering racists. In Sirota’s hands, this led to some very slapdash analysis. (But then again, just look at his link to our piece!) By the way: However good this sort of name-calling feels, we’ll also assume that it’s a very bad way to get Dems elected to office.

True.  Then again, picking your standard-bearer on the basis of margins achieved in low-turnout caucus events also may be risky.  Presidential elections are 50 state winner-take-all elections with no caucuses.  If the Democrats had used that system for nominating their candidate, Hillary Clinton would likely be their nominee.  Instead, their likely nominee is Barack Obama, who won in states like Idaho that he has little chance of winning in a general election.