John Maginnis of the New Orleans Times-Picayune takes exception to the results of my recent co-authored study on the Duke vote in last year's Louisiana gubernatorial election.

Yet comparative vote totals don't tell the whole story. The study discounts ideology as a factor, especially in the Duke parishes, because Jindal and Terrell were both conservative Republicans. But that view overlooks the other side of the equation. It is reasonable to suggest that more rural conservatives voted for Blanco and not Landrieu because Blanco is more conservative than Landrieu. Terrell and the Republicans beat up on Landrieu for her tax votes and her abortion-rights stand. By contrast, Jindal did not lay a glove on Blanco, who protected her right flank on abortion, gun control and taxes while blasting Jindal on health care. Also, Blanco ran a better organized, more aggressive campaign in rural areas than did Jindal, Terrell or Landrieu.

Had a more liberal Democrat made the runoff, say, Buddy Leach, more conservatives, even in north Louisiana, could have gotten past Jindal's hue and voted for him, or just stayed home.

These criticisms seem misplaced. First, we did control for ideology. Our study looked at income levels to see if the Duke vote might have, in fact, been a economic populist vote. But even when you control for income, the Duke vote still went for Blanco over Jindal. And second, of course it helped that Blanco was a conservative Dem. Not all Duke voters went for Blanco and fewer would have if she were more liberal, but many of them did. But what's more interesting is why they did when other conservative and Republican voters didn't. Jindal did very well among Republicans in the New Orleans suburbs, but not in the northern parishes were Duke ran best. What possible reason, other than race, might explain this difference in voting patterns?

Maginnis goes on to add:

What makes people sensitive to the conclusions of the New York professors is the implication that the 671,000 people who voted for David Duke were racists. Surely, many were, but many others were adamantly anti-Edwards, and antipathy toward him ran deepest in rural north Louisiana. The message on the famous bumper sticker notwithstanding, it was more important for them to vote against "the crook."

Louisianans surely faced a Hobson's choice between Duke and Edwards, but just how many voters said, "I know Duke is a racist and a Nazi, and I find that repellent, but he's still the better of the two candidates?" Does this mean that every last one of Duke's 671,000 voters was a racist? I don't know, but if a vote for David Duke isn't an expression of racism, then what is?